diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.5252605122701985,
+  "epoch": 0.7704762173712499,
   "eval_steps": 500,
-  "global_step": 223200,
+  "global_step": 327400,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -312487,6 +312487,145886 @@
       "learning_rate": 0.00016785002554023598,
       "loss": 2.0624,
       "step": 223200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016784866777342625,
+      "loss": 2.0614,
+      "step": 223205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016784730998343816,
+      "loss": 2.115,
+      "step": 223210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00016784595217027222,
+      "loss": 2.0661,
+      "step": 223215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016784459433392886,
+      "loss": 1.9062,
+      "step": 223220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016784323647440856,
+      "loss": 2.2478,
+      "step": 223225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016784187859171176,
+      "loss": 2.152,
+      "step": 223230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016784052068583897,
+      "loss": 2.1974,
+      "step": 223235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001678391627567906,
+      "loss": 2.0411,
+      "step": 223240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001678378048045672,
+      "loss": 2.0948,
+      "step": 223245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001678364468291691,
+      "loss": 2.1167,
+      "step": 223250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001678350888305969,
+      "loss": 2.1689,
+      "step": 223255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016783373080885098,
+      "loss": 2.2333,
+      "step": 223260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016783237276393184,
+      "loss": 2.3798,
+      "step": 223265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001678310146958399,
+      "loss": 1.9692,
+      "step": 223270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016782965660457565,
+      "loss": 2.1187,
+      "step": 223275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016782829849013958,
+      "loss": 1.9444,
+      "step": 223280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016782694035253216,
+      "loss": 2.1884,
+      "step": 223285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016782558219175382,
+      "loss": 2.014,
+      "step": 223290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000167824224007805,
+      "loss": 2.092,
+      "step": 223295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016782286580068622,
+      "loss": 2.2491,
+      "step": 223300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001678215075703979,
+      "loss": 2.1327,
+      "step": 223305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016782014931694054,
+      "loss": 1.889,
+      "step": 223310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016781879104031458,
+      "loss": 2.0293,
+      "step": 223315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016781743274052048,
+      "loss": 2.1208,
+      "step": 223320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016781607441755872,
+      "loss": 2.1021,
+      "step": 223325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016781471607142976,
+      "loss": 1.8657,
+      "step": 223330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016781335770213408,
+      "loss": 2.1163,
+      "step": 223335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016781199930967212,
+      "loss": 2.1405,
+      "step": 223340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016781064089404436,
+      "loss": 2.2499,
+      "step": 223345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001678092824552512,
+      "loss": 2.0042,
+      "step": 223350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016780792399329322,
+      "loss": 2.0092,
+      "step": 223355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016780656550817078,
+      "loss": 2.2444,
+      "step": 223360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016780520699988442,
+      "loss": 2.0866,
+      "step": 223365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016780384846843454,
+      "loss": 2.298,
+      "step": 223370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001678024899138217,
+      "loss": 2.0394,
+      "step": 223375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016780113133604626,
+      "loss": 2.1933,
+      "step": 223380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001677997727351087,
+      "loss": 2.0903,
+      "step": 223385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016779841411100954,
+      "loss": 1.9954,
+      "step": 223390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001677970554637492,
+      "loss": 2.2318,
+      "step": 223395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016779569679332816,
+      "loss": 2.034,
+      "step": 223400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016779433809974688,
+      "loss": 1.9693,
+      "step": 223405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001677929793830058,
+      "loss": 1.8557,
+      "step": 223410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016779162064310544,
+      "loss": 2.1375,
+      "step": 223415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016779026188004626,
+      "loss": 2.1387,
+      "step": 223420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016778890309382863,
+      "loss": 2.176,
+      "step": 223425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016778754428445312,
+      "loss": 2.0683,
+      "step": 223430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016778618545192016,
+      "loss": 2.1118,
+      "step": 223435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016778482659623016,
+      "loss": 2.1417,
+      "step": 223440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001677834677173837,
+      "loss": 2.0838,
+      "step": 223445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016778210881538116,
+      "loss": 2.0649,
+      "step": 223450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016778074989022302,
+      "loss": 1.9166,
+      "step": 223455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00016777939094190974,
+      "loss": 2.1928,
+      "step": 223460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016777803197044175,
+      "loss": 2.0387,
+      "step": 223465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016777667297581962,
+      "loss": 2.1306,
+      "step": 223470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016777531395804373,
+      "loss": 2.2465,
+      "step": 223475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016777395491711453,
+      "loss": 2.2141,
+      "step": 223480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016777259585303256,
+      "loss": 1.9562,
+      "step": 223485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016777123676579823,
+      "loss": 2.2729,
+      "step": 223490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000167769877655412,
+      "loss": 2.1015,
+      "step": 223495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001677685185218744,
+      "loss": 2.1217,
+      "step": 223500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001677671593651858,
+      "loss": 2.2734,
+      "step": 223505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001677658001853467,
+      "loss": 2.2404,
+      "step": 223510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001677644409823576,
+      "loss": 2.1713,
+      "step": 223515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016776308175621893,
+      "loss": 2.2013,
+      "step": 223520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016776172250693115,
+      "loss": 1.9616,
+      "step": 223525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016776036323449476,
+      "loss": 2.1091,
+      "step": 223530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001677590039389102,
+      "loss": 2.0422,
+      "step": 223535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016775764462017794,
+      "loss": 2.2032,
+      "step": 223540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016775628527829838,
+      "loss": 1.9506,
+      "step": 223545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001677549259132721,
+      "loss": 2.1739,
+      "step": 223550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016775356652509952,
+      "loss": 2.1874,
+      "step": 223555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016775220711378106,
+      "loss": 2.0794,
+      "step": 223560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016775084767931724,
+      "loss": 2.2806,
+      "step": 223565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016774948822170847,
+      "loss": 2.1333,
+      "step": 223570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001677481287409553,
+      "loss": 2.1682,
+      "step": 223575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001677467692370581,
+      "loss": 2.029,
+      "step": 223580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.00016774540971001736,
+      "loss": 2.0161,
+      "step": 223585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001677440501598336,
+      "loss": 2.0973,
+      "step": 223590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016774269058650722,
+      "loss": 2.03,
+      "step": 223595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016774133099003871,
+      "loss": 2.074,
+      "step": 223600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016773997137042854,
+      "loss": 2.1499,
+      "step": 223605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016773861172767718,
+      "loss": 2.0921,
+      "step": 223610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016773725206178504,
+      "loss": 2.1176,
+      "step": 223615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016773589237275267,
+      "loss": 2.0839,
+      "step": 223620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016773453266058047,
+      "loss": 2.106,
+      "step": 223625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001677331729252689,
+      "loss": 2.0781,
+      "step": 223630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001677318131668185,
+      "loss": 2.1352,
+      "step": 223635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016773045338522966,
+      "loss": 1.8356,
+      "step": 223640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001677290935805029,
+      "loss": 1.7767,
+      "step": 223645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016772773375263857,
+      "loss": 2.1617,
+      "step": 223650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001677263739016373,
+      "loss": 2.0495,
+      "step": 223655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016772501402749943,
+      "loss": 2.1114,
+      "step": 223660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016772365413022548,
+      "loss": 2.0496,
+      "step": 223665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001677222942098159,
+      "loss": 1.9153,
+      "step": 223670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016772093426627117,
+      "loss": 2.2178,
+      "step": 223675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016771957429959173,
+      "loss": 2.1511,
+      "step": 223680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016771821430977806,
+      "loss": 2.0338,
+      "step": 223685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001677168542968306,
+      "loss": 2.1014,
+      "step": 223690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016771549426074987,
+      "loss": 2.0127,
+      "step": 223695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016771413420153626,
+      "loss": 2.1937,
+      "step": 223700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016771277411919028,
+      "loss": 2.1109,
+      "step": 223705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001677114140137124,
+      "loss": 2.1753,
+      "step": 223710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001677100538851031,
+      "loss": 2.1652,
+      "step": 223715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016770869373336278,
+      "loss": 2.1016,
+      "step": 223720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016770733355849196,
+      "loss": 2.0954,
+      "step": 223725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016770597336049107,
+      "loss": 2.2484,
+      "step": 223730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001677046131393606,
+      "loss": 2.0109,
+      "step": 223735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000167703252895101,
+      "loss": 2.0007,
+      "step": 223740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016770189262771274,
+      "loss": 2.0675,
+      "step": 223745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001677005323371963,
+      "loss": 2.3203,
+      "step": 223750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016769917202355214,
+      "loss": 2.0814,
+      "step": 223755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016769781168678069,
+      "loss": 2.0983,
+      "step": 223760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016769645132688244,
+      "loss": 2.1787,
+      "step": 223765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001676950909438579,
+      "loss": 2.1107,
+      "step": 223770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016769373053770746,
+      "loss": 2.0695,
+      "step": 223775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001676923701084316,
+      "loss": 2.189,
+      "step": 223780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016769100965603078,
+      "loss": 2.1511,
+      "step": 223785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016768964918050552,
+      "loss": 2.0867,
+      "step": 223790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001676882886818562,
+      "loss": 2.0117,
+      "step": 223795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001676869281600834,
+      "loss": 2.1335,
+      "step": 223800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016768556761518748,
+      "loss": 2.0662,
+      "step": 223805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016768420704716896,
+      "loss": 2.0357,
+      "step": 223810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001676828464560283,
+      "loss": 2.0148,
+      "step": 223815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00016768148584176592,
+      "loss": 2.0808,
+      "step": 223820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001676801252043823,
+      "loss": 2.2122,
+      "step": 223825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016767876454387796,
+      "loss": 2.006,
+      "step": 223830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016767740386025332,
+      "loss": 2.1231,
+      "step": 223835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016767604315350885,
+      "loss": 2.1647,
+      "step": 223840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000167674682423645,
+      "loss": 2.0659,
+      "step": 223845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016767332167066227,
+      "loss": 2.2394,
+      "step": 223850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016767196089456112,
+      "loss": 2.1918,
+      "step": 223855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016767060009534196,
+      "loss": 2.1973,
+      "step": 223860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001676692392730053,
+      "loss": 2.255,
+      "step": 223865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016766787842755162,
+      "loss": 2.1685,
+      "step": 223870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016766651755898137,
+      "loss": 2.1433,
+      "step": 223875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000167665156667295,
+      "loss": 2.1434,
+      "step": 223880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000167663795752493,
+      "loss": 2.227,
+      "step": 223885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016766243481457582,
+      "loss": 2.0986,
+      "step": 223890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001676610738535439,
+      "loss": 2.1875,
+      "step": 223895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016765971286939775,
+      "loss": 2.3529,
+      "step": 223900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016765835186213781,
+      "loss": 2.2211,
+      "step": 223905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016765699083176453,
+      "loss": 2.105,
+      "step": 223910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016765562977827844,
+      "loss": 1.9976,
+      "step": 223915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001676542687016799,
+      "loss": 2.1406,
+      "step": 223920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001676529076019695,
+      "loss": 2.0645,
+      "step": 223925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001676515464791476,
+      "loss": 1.8824,
+      "step": 223930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001676501853332147,
+      "loss": 1.7713,
+      "step": 223935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001676488241641713,
+      "loss": 1.9488,
+      "step": 223940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001676474629720178,
+      "loss": 2.1877,
+      "step": 223945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016764610175675472,
+      "loss": 2.1536,
+      "step": 223950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00016764474051838253,
+      "loss": 2.213,
+      "step": 223955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016764337925690163,
+      "loss": 2.0799,
+      "step": 223960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016764201797231254,
+      "loss": 1.962,
+      "step": 223965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016764065666461574,
+      "loss": 2.1876,
+      "step": 223970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001676392953338116,
+      "loss": 2.2039,
+      "step": 223975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016763793397990066,
+      "loss": 2.0506,
+      "step": 223980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0001676365726028834,
+      "loss": 1.9909,
+      "step": 223985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001676352112027603,
+      "loss": 2.0048,
+      "step": 223990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016763384977953172,
+      "loss": 1.9602,
+      "step": 223995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001676324883331982,
+      "loss": 2.2895,
+      "step": 224000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001676311268637602,
+      "loss": 2.0123,
+      "step": 224005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016762976537121823,
+      "loss": 2.1189,
+      "step": 224010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016762840385557267,
+      "loss": 2.0907,
+      "step": 224015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016762704231682398,
+      "loss": 1.9527,
+      "step": 224020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001676256807549727,
+      "loss": 2.0739,
+      "step": 224025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016762431917001925,
+      "loss": 2.3182,
+      "step": 224030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016762295756196413,
+      "loss": 2.1654,
+      "step": 224035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016762159593080778,
+      "loss": 2.0564,
+      "step": 224040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016762023427655066,
+      "loss": 2.0787,
+      "step": 224045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016761887259919323,
+      "loss": 2.0527,
+      "step": 224050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016761751089873597,
+      "loss": 2.2896,
+      "step": 224055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016761614917517936,
+      "loss": 2.1998,
+      "step": 224060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016761478742852385,
+      "loss": 1.9585,
+      "step": 224065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016761342565876986,
+      "loss": 1.9621,
+      "step": 224070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016761206386591793,
+      "loss": 2.1449,
+      "step": 224075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016761070204996848,
+      "loss": 2.1979,
+      "step": 224080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016760934021092198,
+      "loss": 2.0694,
+      "step": 224085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016760797834877893,
+      "loss": 2.1454,
+      "step": 224090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016760661646353974,
+      "loss": 1.9941,
+      "step": 224095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016760525455520494,
+      "loss": 2.1685,
+      "step": 224100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016760389262377495,
+      "loss": 2.1148,
+      "step": 224105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001676025306692502,
+      "loss": 2.0929,
+      "step": 224110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016760116869163124,
+      "loss": 2.0931,
+      "step": 224115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016759980669091848,
+      "loss": 2.0977,
+      "step": 224120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016759844466711243,
+      "loss": 2.0319,
+      "step": 224125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001675970826202135,
+      "loss": 2.1289,
+      "step": 224130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016759572055022217,
+      "loss": 2.1874,
+      "step": 224135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016759435845713894,
+      "loss": 2.2475,
+      "step": 224140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016759299634096423,
+      "loss": 2.1755,
+      "step": 224145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016759163420169854,
+      "loss": 2.0827,
+      "step": 224150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016759027203934232,
+      "loss": 1.9803,
+      "step": 224155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016758890985389604,
+      "loss": 2.1666,
+      "step": 224160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016758754764536015,
+      "loss": 2.0215,
+      "step": 224165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016758618541373515,
+      "loss": 1.9848,
+      "step": 224170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016758482315902147,
+      "loss": 2.0406,
+      "step": 224175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001675834608812196,
+      "loss": 2.2479,
+      "step": 224180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016758209858032998,
+      "loss": 2.0684,
+      "step": 224185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016758073625635308,
+      "loss": 2.281,
+      "step": 224190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016757937390928938,
+      "loss": 2.0221,
+      "step": 224195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016757801153913934,
+      "loss": 2.0275,
+      "step": 224200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016757664914590344,
+      "loss": 2.143,
+      "step": 224205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001675752867295821,
+      "loss": 1.945,
+      "step": 224210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016757392429017585,
+      "loss": 2.0877,
+      "step": 224215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016757256182768512,
+      "loss": 1.9527,
+      "step": 224220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016757119934211038,
+      "loss": 1.9235,
+      "step": 224225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001675698368334521,
+      "loss": 2.2098,
+      "step": 224230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001675684743017107,
+      "loss": 1.9457,
+      "step": 224235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001675671117468867,
+      "loss": 2.0998,
+      "step": 224240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016756574916898057,
+      "loss": 2.1693,
+      "step": 224245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016756438656799274,
+      "loss": 2.0632,
+      "step": 224250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001675630239439237,
+      "loss": 2.2877,
+      "step": 224255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016756166129677387,
+      "loss": 2.2667,
+      "step": 224260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001675602986265438,
+      "loss": 2.2282,
+      "step": 224265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016755893593323385,
+      "loss": 2.189,
+      "step": 224270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016755757321684458,
+      "loss": 1.7126,
+      "step": 224275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001675562104773764,
+      "loss": 2.1515,
+      "step": 224280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016755484771482983,
+      "loss": 2.3555,
+      "step": 224285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016755348492920527,
+      "loss": 2.0678,
+      "step": 224290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001675521221205032,
+      "loss": 2.1572,
+      "step": 224295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016755075928872413,
+      "loss": 2.1364,
+      "step": 224300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001675493964338685,
+      "loss": 2.0422,
+      "step": 224305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016754803355593673,
+      "loss": 2.1243,
+      "step": 224310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016754667065492935,
+      "loss": 2.1393,
+      "step": 224315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016754530773084684,
+      "loss": 2.1499,
+      "step": 224320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016754394478368955,
+      "loss": 1.9112,
+      "step": 224325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016754258181345808,
+      "loss": 1.8714,
+      "step": 224330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016754121882015283,
+      "loss": 2.0473,
+      "step": 224335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016753985580377428,
+      "loss": 2.0317,
+      "step": 224340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016753849276432287,
+      "loss": 2.0135,
+      "step": 224345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001675371297017991,
+      "loss": 2.2548,
+      "step": 224350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001675357666162034,
+      "loss": 2.0043,
+      "step": 224355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016753440350753628,
+      "loss": 2.1243,
+      "step": 224360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001675330403757982,
+      "loss": 1.865,
+      "step": 224365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016753167722098958,
+      "loss": 2.1193,
+      "step": 224370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001675303140431109,
+      "loss": 2.0943,
+      "step": 224375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016752895084216267,
+      "loss": 2.1673,
+      "step": 224380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016752758761814533,
+      "loss": 2.1234,
+      "step": 224385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001675262243710593,
+      "loss": 2.345,
+      "step": 224390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016752486110090513,
+      "loss": 2.0154,
+      "step": 224395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016752349780768325,
+      "loss": 2.0537,
+      "step": 224400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016752213449139407,
+      "loss": 1.9641,
+      "step": 224405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016752077115203814,
+      "loss": 2.0838,
+      "step": 224410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016751940778961588,
+      "loss": 2.1554,
+      "step": 224415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016751804440412777,
+      "loss": 2.2202,
+      "step": 224420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016751668099557426,
+      "loss": 2.0321,
+      "step": 224425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016751531756395584,
+      "loss": 2.1781,
+      "step": 224430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016751395410927295,
+      "loss": 2.047,
+      "step": 224435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016751259063152606,
+      "loss": 2.1339,
+      "step": 224440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016751122713071567,
+      "loss": 1.9913,
+      "step": 224445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016750986360684222,
+      "loss": 1.8366,
+      "step": 224450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016750850005990616,
+      "loss": 2.1112,
+      "step": 224455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00016750713648990798,
+      "loss": 2.1041,
+      "step": 224460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016750577289684814,
+      "loss": 1.8296,
+      "step": 224465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016750440928072714,
+      "loss": 2.1356,
+      "step": 224470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016750304564154536,
+      "loss": 2.0426,
+      "step": 224475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016750168197930335,
+      "loss": 2.1694,
+      "step": 224480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001675003182940015,
+      "loss": 1.8403,
+      "step": 224485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016749895458564033,
+      "loss": 2.0597,
+      "step": 224490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016749759085422027,
+      "loss": 2.0299,
+      "step": 224495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016749622709974187,
+      "loss": 2.1757,
+      "step": 224500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001674948633222055,
+      "loss": 2.3122,
+      "step": 224505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016749349952161166,
+      "loss": 2.1141,
+      "step": 224510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001674921356979608,
+      "loss": 2.157,
+      "step": 224515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016749077185125341,
+      "loss": 2.2004,
+      "step": 224520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016748940798148998,
+      "loss": 2.1459,
+      "step": 224525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016748804408867093,
+      "loss": 2.0933,
+      "step": 224530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016748668017279675,
+      "loss": 2.0403,
+      "step": 224535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016748531623386785,
+      "loss": 2.2275,
+      "step": 224540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016748395227188478,
+      "loss": 2.0041,
+      "step": 224545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016748258828684794,
+      "loss": 2.1924,
+      "step": 224550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016748122427875787,
+      "loss": 2.1074,
+      "step": 224555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016747986024761498,
+      "loss": 2.1561,
+      "step": 224560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001674784961934197,
+      "loss": 2.13,
+      "step": 224565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001674771321161726,
+      "loss": 2.0236,
+      "step": 224570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016747576801587403,
+      "loss": 2.0618,
+      "step": 224575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016747440389252457,
+      "loss": 2.0475,
+      "step": 224580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001674730397461246,
+      "loss": 2.236,
+      "step": 224585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001674716755766746,
+      "loss": 1.9974,
+      "step": 224590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016747031138417508,
+      "loss": 2.1044,
+      "step": 224595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016746894716862648,
+      "loss": 2.1788,
+      "step": 224600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016746758293002925,
+      "loss": 2.2005,
+      "step": 224605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016746621866838387,
+      "loss": 1.8215,
+      "step": 224610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016746485438369078,
+      "loss": 2.0811,
+      "step": 224615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016746349007595052,
+      "loss": 2.0062,
+      "step": 224620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001674621257451635,
+      "loss": 2.0652,
+      "step": 224625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001674607613913302,
+      "loss": 2.1361,
+      "step": 224630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016745939701445106,
+      "loss": 2.1326,
+      "step": 224635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016745803261452658,
+      "loss": 2.1093,
+      "step": 224640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001674566681915572,
+      "loss": 2.1228,
+      "step": 224645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001674553037455434,
+      "loss": 2.1815,
+      "step": 224650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016745393927648566,
+      "loss": 1.9667,
+      "step": 224655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016745257478438438,
+      "loss": 2.2277,
+      "step": 224660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016745121026924014,
+      "loss": 2.239,
+      "step": 224665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016744984573105332,
+      "loss": 1.9545,
+      "step": 224670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001674484811698244,
+      "loss": 2.1444,
+      "step": 224675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016744711658555389,
+      "loss": 2.204,
+      "step": 224680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016744575197824216,
+      "loss": 1.8763,
+      "step": 224685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016744438734788978,
+      "loss": 2.0609,
+      "step": 224690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016744302269449716,
+      "loss": 1.9844,
+      "step": 224695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016744165801806479,
+      "loss": 1.9922,
+      "step": 224700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016744029331859314,
+      "loss": 2.0278,
+      "step": 224705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016743892859608263,
+      "loss": 2.2616,
+      "step": 224710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001674375638505338,
+      "loss": 1.9632,
+      "step": 224715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016743619908194705,
+      "loss": 2.0238,
+      "step": 224720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016743483429032289,
+      "loss": 2.0376,
+      "step": 224725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016743346947566172,
+      "loss": 1.9867,
+      "step": 224730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001674321046379641,
+      "loss": 2.1762,
+      "step": 224735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016743073977723043,
+      "loss": 1.9296,
+      "step": 224740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001674293748934612,
+      "loss": 1.8124,
+      "step": 224745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016742800998665686,
+      "loss": 1.9338,
+      "step": 224750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016742664505681792,
+      "loss": 1.9998,
+      "step": 224755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016742528010394478,
+      "loss": 2.1223,
+      "step": 224760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016742391512803798,
+      "loss": 2.1353,
+      "step": 224765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016742255012909793,
+      "loss": 1.9697,
+      "step": 224770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001674211851071251,
+      "loss": 1.9201,
+      "step": 224775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016741982006212,
+      "loss": 2.1079,
+      "step": 224780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000167418454994083,
+      "loss": 2.0327,
+      "step": 224785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001674170899030147,
+      "loss": 2.0631,
+      "step": 224790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016741572478891549,
+      "loss": 2.0843,
+      "step": 224795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016741435965178584,
+      "loss": 1.9199,
+      "step": 224800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00016741299449162623,
+      "loss": 2.0417,
+      "step": 224805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016741162930843708,
+      "loss": 2.1065,
+      "step": 224810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016741026410221893,
+      "loss": 1.9607,
+      "step": 224815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016740889887297222,
+      "loss": 2.0227,
+      "step": 224820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016740753362069738,
+      "loss": 2.061,
+      "step": 224825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001674061683453949,
+      "loss": 2.2245,
+      "step": 224830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016740480304706527,
+      "loss": 2.0349,
+      "step": 224835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016740343772570895,
+      "loss": 2.096,
+      "step": 224840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016740207238132636,
+      "loss": 1.968,
+      "step": 224845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016740070701391804,
+      "loss": 1.8633,
+      "step": 224850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016739934162348437,
+      "loss": 2.1696,
+      "step": 224855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016739797621002588,
+      "loss": 2.1427,
+      "step": 224860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016739661077354304,
+      "loss": 2.023,
+      "step": 224865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001673952453140363,
+      "loss": 2.1294,
+      "step": 224870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001673938798315061,
+      "loss": 2.1735,
+      "step": 224875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001673925143259529,
+      "loss": 1.8106,
+      "step": 224880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001673911487973772,
+      "loss": 2.0765,
+      "step": 224885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001673897832457795,
+      "loss": 2.1711,
+      "step": 224890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016738841767116022,
+      "loss": 2.1361,
+      "step": 224895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016738705207351983,
+      "loss": 2.0692,
+      "step": 224900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016738568645285881,
+      "loss": 2.2706,
+      "step": 224905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016738432080917759,
+      "loss": 2.2092,
+      "step": 224910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016738295514247668,
+      "loss": 2.1034,
+      "step": 224915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016738158945275652,
+      "loss": 2.0061,
+      "step": 224920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001673802237400176,
+      "loss": 2.1954,
+      "step": 224925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001673788580042604,
+      "loss": 2.1443,
+      "step": 224930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001673774922454853,
+      "loss": 2.163,
+      "step": 224935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016737612646369288,
+      "loss": 2.1422,
+      "step": 224940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001673747606588835,
+      "loss": 2.2119,
+      "step": 224945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016737339483105773,
+      "loss": 2.153,
+      "step": 224950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016737202898021596,
+      "loss": 2.0396,
+      "step": 224955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001673706631063587,
+      "loss": 2.0548,
+      "step": 224960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016736929720948638,
+      "loss": 2.2874,
+      "step": 224965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001673679312895995,
+      "loss": 1.9928,
+      "step": 224970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016736656534669853,
+      "loss": 2.0494,
+      "step": 224975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016736519938078388,
+      "loss": 2.0744,
+      "step": 224980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016736383339185608,
+      "loss": 2.2727,
+      "step": 224985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016736246737991553,
+      "loss": 1.9735,
+      "step": 224990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001673611013449628,
+      "loss": 2.1968,
+      "step": 224995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016735973528699826,
+      "loss": 2.1038,
+      "step": 225000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016735836920602246,
+      "loss": 2.2032,
+      "step": 225005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016735700310203578,
+      "loss": 2.0357,
+      "step": 225010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016735563697503874,
+      "loss": 2.0758,
+      "step": 225015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016735427082503174,
+      "loss": 2.1736,
+      "step": 225020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016735290465201536,
+      "loss": 2.1155,
+      "step": 225025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016735153845599002,
+      "loss": 2.278,
+      "step": 225030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016735017223695612,
+      "loss": 1.95,
+      "step": 225035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001673488059949142,
+      "loss": 2.0959,
+      "step": 225040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001673474397298647,
+      "loss": 2.1132,
+      "step": 225045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001673460734418081,
+      "loss": 2.0646,
+      "step": 225050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016734470713074487,
+      "loss": 2.1058,
+      "step": 225055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016734334079667547,
+      "loss": 2.1618,
+      "step": 225060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016734197443960035,
+      "loss": 1.9385,
+      "step": 225065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016734060805951998,
+      "loss": 2.0609,
+      "step": 225070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016733924165643487,
+      "loss": 2.1957,
+      "step": 225075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016733787523034543,
+      "loss": 2.0521,
+      "step": 225080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016733650878125213,
+      "loss": 2.0092,
+      "step": 225085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001673351423091555,
+      "loss": 2.2573,
+      "step": 225090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016733377581405593,
+      "loss": 2.1364,
+      "step": 225095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016733240929595392,
+      "loss": 2.1083,
+      "step": 225100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016733104275484995,
+      "loss": 1.735,
+      "step": 225105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016732967619074448,
+      "loss": 2.0144,
+      "step": 225110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016732830960363798,
+      "loss": 1.9336,
+      "step": 225115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016732694299353092,
+      "loss": 2.1483,
+      "step": 225120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016732557636042373,
+      "loss": 2.1075,
+      "step": 225125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001673242097043169,
+      "loss": 2.0138,
+      "step": 225130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001673228430252109,
+      "loss": 1.9901,
+      "step": 225135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001673214763231062,
+      "loss": 2.1665,
+      "step": 225140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016732010959800327,
+      "loss": 2.155,
+      "step": 225145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016731874284990256,
+      "loss": 2.1294,
+      "step": 225150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016731737607880454,
+      "loss": 2.2497,
+      "step": 225155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001673160092847097,
+      "loss": 2.1812,
+      "step": 225160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001673146424676185,
+      "loss": 2.03,
+      "step": 225165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001673132756275314,
+      "loss": 2.0007,
+      "step": 225170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016731190876444885,
+      "loss": 2.0307,
+      "step": 225175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001673105418783713,
+      "loss": 2.0284,
+      "step": 225180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016730917496929928,
+      "loss": 2.1311,
+      "step": 225185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016730780803723324,
+      "loss": 2.1737,
+      "step": 225190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001673064410821736,
+      "loss": 2.1671,
+      "step": 225195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016730507410412087,
+      "loss": 2.2306,
+      "step": 225200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001673037071030755,
+      "loss": 2.1227,
+      "step": 225205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016730234007903796,
+      "loss": 2.1294,
+      "step": 225210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016730097303200873,
+      "loss": 2.1021,
+      "step": 225215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016729960596198828,
+      "loss": 2.0048,
+      "step": 225220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016729823886897703,
+      "loss": 1.9097,
+      "step": 225225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016729687175297552,
+      "loss": 1.8844,
+      "step": 225230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016729550461398415,
+      "loss": 2.1507,
+      "step": 225235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016729413745200343,
+      "loss": 2.2852,
+      "step": 225240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016729277026703381,
+      "loss": 2.0234,
+      "step": 225245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016729140305907575,
+      "loss": 2.1127,
+      "step": 225250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016729003582812975,
+      "loss": 2.0935,
+      "step": 225255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001672886685741962,
+      "loss": 2.1234,
+      "step": 225260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001672873012972757,
+      "loss": 2.2147,
+      "step": 225265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016728593399736858,
+      "loss": 2.1102,
+      "step": 225270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016728456667447538,
+      "loss": 2.1515,
+      "step": 225275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016728319932859655,
+      "loss": 2.0761,
+      "step": 225280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016728183195973255,
+      "loss": 1.9824,
+      "step": 225285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016728046456788387,
+      "loss": 1.993,
+      "step": 225290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016727909715305095,
+      "loss": 2.0976,
+      "step": 225295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016727772971523427,
+      "loss": 2.2588,
+      "step": 225300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016727636225443435,
+      "loss": 2.1587,
+      "step": 225305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016727499477065157,
+      "loss": 2.0479,
+      "step": 225310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001672736272638864,
+      "loss": 2.136,
+      "step": 225315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016727225973413935,
+      "loss": 2.0887,
+      "step": 225320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00016727089218141088,
+      "loss": 2.241,
+      "step": 225325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001672695246057015,
+      "loss": 2.3281,
+      "step": 225330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016726815700701154,
+      "loss": 2.1933,
+      "step": 225335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016726678938534163,
+      "loss": 2.1865,
+      "step": 225340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016726542174069214,
+      "loss": 2.0344,
+      "step": 225345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016726405407306358,
+      "loss": 2.2205,
+      "step": 225350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016726268638245638,
+      "loss": 2.063,
+      "step": 225355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016726131866887103,
+      "loss": 2.2039,
+      "step": 225360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016725995093230797,
+      "loss": 2.1378,
+      "step": 225365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016725858317276772,
+      "loss": 1.9045,
+      "step": 225370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016725721539025075,
+      "loss": 1.8199,
+      "step": 225375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016725584758475742,
+      "loss": 2.1212,
+      "step": 225380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016725447975628833,
+      "loss": 2.2127,
+      "step": 225385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001672531119048439,
+      "loss": 2.258,
+      "step": 225390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016725174403042455,
+      "loss": 2.1973,
+      "step": 225395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00016725037613303075,
+      "loss": 2.1435,
+      "step": 225400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016724900821266306,
+      "loss": 2.1792,
+      "step": 225405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001672476402693219,
+      "loss": 2.0665,
+      "step": 225410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016724627230300772,
+      "loss": 2.2792,
+      "step": 225415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016724490431372093,
+      "loss": 2.218,
+      "step": 225420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001672435363014621,
+      "loss": 2.0782,
+      "step": 225425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001672421682662317,
+      "loss": 2.1866,
+      "step": 225430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016724080020803013,
+      "loss": 1.9457,
+      "step": 225435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016723943212685786,
+      "loss": 1.8663,
+      "step": 225440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016723806402271538,
+      "loss": 1.9637,
+      "step": 225445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016723669589560318,
+      "loss": 2.0389,
+      "step": 225450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016723532774552167,
+      "loss": 2.1586,
+      "step": 225455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001672339595724714,
+      "loss": 2.0467,
+      "step": 225460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016723259137645279,
+      "loss": 2.1627,
+      "step": 225465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001672312231574663,
+      "loss": 2.1164,
+      "step": 225470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016722985491551237,
+      "loss": 2.206,
+      "step": 225475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016722848665059152,
+      "loss": 2.0797,
+      "step": 225480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001672271183627042,
+      "loss": 2.1059,
+      "step": 225485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001672257500518509,
+      "loss": 2.1175,
+      "step": 225490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016722438171803202,
+      "loss": 2.0158,
+      "step": 225495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001672230133612481,
+      "loss": 2.0372,
+      "step": 225500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016722164498149958,
+      "loss": 1.8849,
+      "step": 225505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016722027657878696,
+      "loss": 2.1389,
+      "step": 225510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016721890815311064,
+      "loss": 2.2261,
+      "step": 225515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001672175397044711,
+      "loss": 1.9868,
+      "step": 225520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016721617123286885,
+      "loss": 2.2461,
+      "step": 225525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016721480273830435,
+      "loss": 1.889,
+      "step": 225530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016721343422077803,
+      "loss": 2.1215,
+      "step": 225535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001672120656802904,
+      "loss": 2.1114,
+      "step": 225540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001672106971168419,
+      "loss": 2.2292,
+      "step": 225545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016720932853043303,
+      "loss": 2.2395,
+      "step": 225550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016720795992106419,
+      "loss": 1.8407,
+      "step": 225555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016720659128873592,
+      "loss": 2.0796,
+      "step": 225560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016720522263344867,
+      "loss": 2.0821,
+      "step": 225565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016720385395520285,
+      "loss": 2.0669,
+      "step": 225570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016720248525399901,
+      "loss": 2.1795,
+      "step": 225575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016720111652983758,
+      "loss": 2.1113,
+      "step": 225580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016719974778271906,
+      "loss": 2.1214,
+      "step": 225585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016719837901264385,
+      "loss": 2.1333,
+      "step": 225590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016719701021961246,
+      "loss": 2.0297,
+      "step": 225595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016719564140362536,
+      "loss": 2.0344,
+      "step": 225600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000167194272564683,
+      "loss": 1.9645,
+      "step": 225605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016719290370278587,
+      "loss": 2.1981,
+      "step": 225610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016719153481793442,
+      "loss": 2.1313,
+      "step": 225615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016719016591012912,
+      "loss": 1.9975,
+      "step": 225620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016718879697937044,
+      "loss": 2.1146,
+      "step": 225625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016718742802565884,
+      "loss": 2.2846,
+      "step": 225630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016718605904899483,
+      "loss": 2.0413,
+      "step": 225635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001671846900493788,
+      "loss": 1.9648,
+      "step": 225640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016718332102681128,
+      "loss": 2.0874,
+      "step": 225645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016718195198129273,
+      "loss": 1.9966,
+      "step": 225650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001671805829128236,
+      "loss": 2.0166,
+      "step": 225655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016717921382140437,
+      "loss": 2.1337,
+      "step": 225660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016717784470703548,
+      "loss": 2.1524,
+      "step": 225665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016717647556971745,
+      "loss": 2.1875,
+      "step": 225670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001671751064094507,
+      "loss": 2.0088,
+      "step": 225675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016717373722623574,
+      "loss": 2.0043,
+      "step": 225680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016717236802007296,
+      "loss": 2.0339,
+      "step": 225685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016717099879096293,
+      "loss": 2.2919,
+      "step": 225690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016716962953890603,
+      "loss": 2.1854,
+      "step": 225695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016716826026390278,
+      "loss": 1.9591,
+      "step": 225700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016716689096595366,
+      "loss": 2.0476,
+      "step": 225705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001671655216450591,
+      "loss": 2.2371,
+      "step": 225710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016716415230121957,
+      "loss": 2.1439,
+      "step": 225715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016716278293443554,
+      "loss": 2.1391,
+      "step": 225720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001671614135447075,
+      "loss": 2.0213,
+      "step": 225725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001671600441320359,
+      "loss": 1.9183,
+      "step": 225730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001671586746964212,
+      "loss": 2.2286,
+      "step": 225735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016715730523786387,
+      "loss": 2.0003,
+      "step": 225740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016715593575636443,
+      "loss": 1.9477,
+      "step": 225745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016715456625192326,
+      "loss": 2.0199,
+      "step": 225750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016715319672454092,
+      "loss": 2.0403,
+      "step": 225755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001671518271742178,
+      "loss": 1.9793,
+      "step": 225760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016715045760095439,
+      "loss": 2.0004,
+      "step": 225765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016714908800475117,
+      "loss": 2.0948,
+      "step": 225770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001671477183856086,
+      "loss": 2.224,
+      "step": 225775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016714634874352718,
+      "loss": 1.9854,
+      "step": 225780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016714497907850733,
+      "loss": 1.993,
+      "step": 225785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016714360939054955,
+      "loss": 1.9187,
+      "step": 225790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016714223967965425,
+      "loss": 2.224,
+      "step": 225795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016714086994582198,
+      "loss": 2.1901,
+      "step": 225800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016713950018905317,
+      "loss": 2.0844,
+      "step": 225805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001671381304093483,
+      "loss": 2.1445,
+      "step": 225810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001671367606067078,
+      "loss": 1.863,
+      "step": 225815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016713539078113215,
+      "loss": 1.9108,
+      "step": 225820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001671340209326219,
+      "loss": 2.1791,
+      "step": 225825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001671326510611774,
+      "loss": 2.0383,
+      "step": 225830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001671312811667992,
+      "loss": 2.0632,
+      "step": 225835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016712991124948772,
+      "loss": 2.1125,
+      "step": 225840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016712854130924342,
+      "loss": 2.1094,
+      "step": 225845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016712717134606683,
+      "loss": 2.0276,
+      "step": 225850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016712580135995835,
+      "loss": 2.2109,
+      "step": 225855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016712443135091852,
+      "loss": 1.9244,
+      "step": 225860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016712306131894775,
+      "loss": 2.1575,
+      "step": 225865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001671216912640465,
+      "loss": 1.9265,
+      "step": 225870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016712032118621525,
+      "loss": 1.9335,
+      "step": 225875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016711895108545454,
+      "loss": 2.0357,
+      "step": 225880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016711758096176475,
+      "loss": 2.1281,
+      "step": 225885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016711621081514638,
+      "loss": 2.2174,
+      "step": 225890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016711484064559987,
+      "loss": 2.126,
+      "step": 225895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016711347045312577,
+      "loss": 1.9401,
+      "step": 225900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016711210023772445,
+      "loss": 2.3341,
+      "step": 225905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001671107299993964,
+      "loss": 2.1614,
+      "step": 225910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016710935973814214,
+      "loss": 2.2193,
+      "step": 225915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016710798945396212,
+      "loss": 2.0211,
+      "step": 225920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016710661914685676,
+      "loss": 2.2786,
+      "step": 225925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001671052488168266,
+      "loss": 2.005,
+      "step": 225930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.000167103878463872,
+      "loss": 2.0759,
+      "step": 225935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016710250808799358,
+      "loss": 2.0978,
+      "step": 225940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016710113768919168,
+      "loss": 2.1152,
+      "step": 225945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016709976726746683,
+      "loss": 2.0392,
+      "step": 225950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016709839682281947,
+      "loss": 2.1068,
+      "step": 225955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001670970263552501,
+      "loss": 1.9014,
+      "step": 225960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016709565586475912,
+      "loss": 1.9234,
+      "step": 225965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001670942853513471,
+      "loss": 1.982,
+      "step": 225970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016709291481501444,
+      "loss": 2.1244,
+      "step": 225975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016709154425576162,
+      "loss": 2.1797,
+      "step": 225980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016709017367358913,
+      "loss": 1.85,
+      "step": 225985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016708880306849739,
+      "loss": 2.1188,
+      "step": 225990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001670874324404869,
+      "loss": 2.218,
+      "step": 225995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016708606178955816,
+      "loss": 2.1935,
+      "step": 226000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016708469111571158,
+      "loss": 2.2047,
+      "step": 226005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016708332041894764,
+      "loss": 1.995,
+      "step": 226010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016708194969926681,
+      "loss": 2.1052,
+      "step": 226015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001670805789566696,
+      "loss": 1.9338,
+      "step": 226020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016707920819115647,
+      "loss": 2.2151,
+      "step": 226025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016707783740272785,
+      "loss": 2.1414,
+      "step": 226030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001670764665913842,
+      "loss": 2.2273,
+      "step": 226035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016707509575712603,
+      "loss": 2.156,
+      "step": 226040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001670737248999538,
+      "loss": 2.0739,
+      "step": 226045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016707235401986793,
+      "loss": 2.1358,
+      "step": 226050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016707098311686898,
+      "loss": 2.2234,
+      "step": 226055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016706961219095735,
+      "loss": 1.9028,
+      "step": 226060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016706824124213348,
+      "loss": 2.2123,
+      "step": 226065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016706687027039793,
+      "loss": 1.9919,
+      "step": 226070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016706549927575113,
+      "loss": 2.088,
+      "step": 226075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001670641282581935,
+      "loss": 2.1412,
+      "step": 226080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016706275721772555,
+      "loss": 1.9737,
+      "step": 226085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016706138615434776,
+      "loss": 2.1445,
+      "step": 226090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016706001506806059,
+      "loss": 1.9769,
+      "step": 226095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001670586439588645,
+      "loss": 2.0458,
+      "step": 226100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016705727282675996,
+      "loss": 2.133,
+      "step": 226105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016705590167174746,
+      "loss": 2.086,
+      "step": 226110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001670545304938274,
+      "loss": 2.0251,
+      "step": 226115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016705315929300033,
+      "loss": 1.9418,
+      "step": 226120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016705178806926667,
+      "loss": 2.0869,
+      "step": 226125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016705041682262692,
+      "loss": 2.0858,
+      "step": 226130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016704904555308152,
+      "loss": 2.1162,
+      "step": 226135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00016704767426063095,
+      "loss": 2.0443,
+      "step": 226140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016704630294527565,
+      "loss": 2.1018,
+      "step": 226145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016704493160701618,
+      "loss": 2.0169,
+      "step": 226150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001670435602458529,
+      "loss": 2.1431,
+      "step": 226155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016704218886178632,
+      "loss": 2.1997,
+      "step": 226160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016704081745481694,
+      "loss": 2.1376,
+      "step": 226165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016703944602494518,
+      "loss": 2.1503,
+      "step": 226170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016703807457217154,
+      "loss": 2.0526,
+      "step": 226175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001670367030964965,
+      "loss": 2.0479,
+      "step": 226180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016703533159792046,
+      "loss": 2.2134,
+      "step": 226185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000167033960076444,
+      "loss": 1.9799,
+      "step": 226190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016703258853206744,
+      "loss": 2.1898,
+      "step": 226195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016703121696479137,
+      "loss": 2.1148,
+      "step": 226200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016702984537461625,
+      "loss": 2.2088,
+      "step": 226205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016702847376154248,
+      "loss": 2.2661,
+      "step": 226210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016702710212557057,
+      "loss": 2.083,
+      "step": 226215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000167025730466701,
+      "loss": 2.1214,
+      "step": 226220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016702435878493424,
+      "loss": 2.1803,
+      "step": 226225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001670229870802707,
+      "loss": 2.0806,
+      "step": 226230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016702161535271092,
+      "loss": 2.1935,
+      "step": 226235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.578125,
+      "learning_rate": 0.00016702024360225535,
+      "loss": 2.2326,
+      "step": 226240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016701887182890443,
+      "loss": 2.088,
+      "step": 226245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016701750003265863,
+      "loss": 2.1104,
+      "step": 226250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001670161282135185,
+      "loss": 2.1038,
+      "step": 226255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016701475637148438,
+      "loss": 2.245,
+      "step": 226260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016701338450655687,
+      "loss": 2.0126,
+      "step": 226265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016701201261873632,
+      "loss": 2.033,
+      "step": 226270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016701064070802325,
+      "loss": 2.0731,
+      "step": 226275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016700926877441817,
+      "loss": 1.9742,
+      "step": 226280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016700789681792146,
+      "loss": 2.2874,
+      "step": 226285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016700652483853364,
+      "loss": 1.9707,
+      "step": 226290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016700515283625523,
+      "loss": 2.0702,
+      "step": 226295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016700378081108662,
+      "loss": 2.1186,
+      "step": 226300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016700240876302828,
+      "loss": 2.1816,
+      "step": 226305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016700103669208073,
+      "loss": 2.0872,
+      "step": 226310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001669996645982444,
+      "loss": 2.0689,
+      "step": 226315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016699829248151976,
+      "loss": 2.1493,
+      "step": 226320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016699692034190727,
+      "loss": 2.0269,
+      "step": 226325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016699554817940743,
+      "loss": 1.9816,
+      "step": 226330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016699417599402074,
+      "loss": 2.1549,
+      "step": 226335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016699280378574757,
+      "loss": 1.9599,
+      "step": 226340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016699143155458847,
+      "loss": 2.0924,
+      "step": 226345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016699005930054388,
+      "loss": 2.1541,
+      "step": 226350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001669886870236143,
+      "loss": 2.0595,
+      "step": 226355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001669873147238001,
+      "loss": 2.0781,
+      "step": 226360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016698594240110185,
+      "loss": 2.161,
+      "step": 226365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016698457005552003,
+      "loss": 2.2664,
+      "step": 226370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000166983197687055,
+      "loss": 2.1661,
+      "step": 226375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016698182529570734,
+      "loss": 2.0139,
+      "step": 226380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016698045288147744,
+      "loss": 2.0861,
+      "step": 226385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016697908044436584,
+      "loss": 2.1505,
+      "step": 226390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016697770798437293,
+      "loss": 2.3362,
+      "step": 226395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016697633550149924,
+      "loss": 2.1799,
+      "step": 226400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016697496299574523,
+      "loss": 2.0832,
+      "step": 226405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016697359046711135,
+      "loss": 2.0252,
+      "step": 226410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016697221791559807,
+      "loss": 2.0737,
+      "step": 226415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016697084534120587,
+      "loss": 2.1916,
+      "step": 226420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016696947274393522,
+      "loss": 1.9643,
+      "step": 226425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001669681001237866,
+      "loss": 2.0293,
+      "step": 226430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016696672748076042,
+      "loss": 2.2264,
+      "step": 226435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016696535481485722,
+      "loss": 2.2,
+      "step": 226440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001669639821260774,
+      "loss": 1.9523,
+      "step": 226445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016696260941442153,
+      "loss": 2.2412,
+      "step": 226450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016696123667989002,
+      "loss": 2.1204,
+      "step": 226455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016695986392248327,
+      "loss": 2.1559,
+      "step": 226460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016695849114220187,
+      "loss": 1.9523,
+      "step": 226465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016695711833904618,
+      "loss": 2.1192,
+      "step": 226470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016695574551301682,
+      "loss": 2.1241,
+      "step": 226475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016695437266411406,
+      "loss": 2.2478,
+      "step": 226480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016695299979233857,
+      "loss": 2.2378,
+      "step": 226485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016695162689769064,
+      "loss": 2.0636,
+      "step": 226490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016695025398017084,
+      "loss": 2.0933,
+      "step": 226495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016694888103977964,
+      "loss": 2.0811,
+      "step": 226500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001669475080765175,
+      "loss": 2.1952,
+      "step": 226505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016694613509038483,
+      "loss": 2.1136,
+      "step": 226510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016694476208138216,
+      "loss": 1.9957,
+      "step": 226515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016694338904950996,
+      "loss": 1.9777,
+      "step": 226520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001669420159947687,
+      "loss": 2.0851,
+      "step": 226525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001669406429171588,
+      "loss": 2.3595,
+      "step": 226530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016693926981668078,
+      "loss": 2.0292,
+      "step": 226535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001669378966933351,
+      "loss": 1.938,
+      "step": 226540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016693652354712218,
+      "loss": 2.0014,
+      "step": 226545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016693515037804256,
+      "loss": 2.1282,
+      "step": 226550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001669337771860967,
+      "loss": 2.1429,
+      "step": 226555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000166932403971285,
+      "loss": 2.1206,
+      "step": 226560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000166931030733608,
+      "loss": 2.1515,
+      "step": 226565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016692965747306616,
+      "loss": 2.0156,
+      "step": 226570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016692828418965992,
+      "loss": 2.1701,
+      "step": 226575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016692691088338973,
+      "loss": 2.1524,
+      "step": 226580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016692553755425616,
+      "loss": 1.9636,
+      "step": 226585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016692416420225958,
+      "loss": 2.0361,
+      "step": 226590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016692279082740049,
+      "loss": 2.1252,
+      "step": 226595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016692141742967936,
+      "loss": 2.1368,
+      "step": 226600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016692004400909664,
+      "loss": 2.0423,
+      "step": 226605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016691867056565287,
+      "loss": 2.0441,
+      "step": 226610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016691729709934845,
+      "loss": 2.0176,
+      "step": 226615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016691592361018387,
+      "loss": 2.2291,
+      "step": 226620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016691455009815956,
+      "loss": 2.2132,
+      "step": 226625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016691317656327605,
+      "loss": 2.0603,
+      "step": 226630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016691180300553379,
+      "loss": 2.0693,
+      "step": 226635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016691042942493323,
+      "loss": 1.9251,
+      "step": 226640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016690905582147488,
+      "loss": 2.0382,
+      "step": 226645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001669076821951592,
+      "loss": 2.1923,
+      "step": 226650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001669063085459866,
+      "loss": 2.133,
+      "step": 226655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016690493487395757,
+      "loss": 2.1227,
+      "step": 226660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016690356117907268,
+      "loss": 2.1879,
+      "step": 226665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016690218746133226,
+      "loss": 2.0603,
+      "step": 226670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016690081372073686,
+      "loss": 2.127,
+      "step": 226675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001668994399572869,
+      "loss": 2.0479,
+      "step": 226680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001668980661709829,
+      "loss": 1.9758,
+      "step": 226685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016689669236182532,
+      "loss": 2.1045,
+      "step": 226690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001668953185298146,
+      "loss": 1.9469,
+      "step": 226695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016689394467495124,
+      "loss": 2.1006,
+      "step": 226700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016689257079723565,
+      "loss": 2.1683,
+      "step": 226705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016689119689666842,
+      "loss": 2.1942,
+      "step": 226710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016688982297324989,
+      "loss": 2.0801,
+      "step": 226715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001668884490269806,
+      "loss": 1.8688,
+      "step": 226720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000166887075057861,
+      "loss": 2.1138,
+      "step": 226725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016688570106589156,
+      "loss": 1.9386,
+      "step": 226730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016688432705107276,
+      "loss": 2.2483,
+      "step": 226735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016688295301340507,
+      "loss": 2.0032,
+      "step": 226740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016688157895288893,
+      "loss": 2.0813,
+      "step": 226745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016688020486952483,
+      "loss": 2.1364,
+      "step": 226750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016687883076331326,
+      "loss": 2.0994,
+      "step": 226755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016687745663425461,
+      "loss": 2.0584,
+      "step": 226760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016687608248234948,
+      "loss": 1.9834,
+      "step": 226765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016687470830759824,
+      "loss": 2.0564,
+      "step": 226770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016687333411000136,
+      "loss": 2.1603,
+      "step": 226775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016687195988955938,
+      "loss": 2.0419,
+      "step": 226780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001668705856462727,
+      "loss": 2.0848,
+      "step": 226785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016686921138014182,
+      "loss": 2.0189,
+      "step": 226790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001668678370911672,
+      "loss": 2.142,
+      "step": 226795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016686646277934934,
+      "loss": 2.2631,
+      "step": 226800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016686508844468868,
+      "loss": 2.1383,
+      "step": 226805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016686371408718565,
+      "loss": 2.0895,
+      "step": 226810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016686233970684077,
+      "loss": 2.2143,
+      "step": 226815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016686096530365454,
+      "loss": 1.9039,
+      "step": 226820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016685959087762737,
+      "loss": 2.0736,
+      "step": 226825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016685821642875977,
+      "loss": 2.0676,
+      "step": 226830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016685684195705218,
+      "loss": 2.0491,
+      "step": 226835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016685546746250505,
+      "loss": 2.0902,
+      "step": 226840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001668540929451189,
+      "loss": 2.0073,
+      "step": 226845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00016685271840489422,
+      "loss": 2.1626,
+      "step": 226850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001668513438418314,
+      "loss": 2.1901,
+      "step": 226855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016684996925593097,
+      "loss": 1.93,
+      "step": 226860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016684859464719334,
+      "loss": 2.205,
+      "step": 226865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016684722001561903,
+      "loss": 2.202,
+      "step": 226870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016684584536120852,
+      "loss": 2.046,
+      "step": 226875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016684447068396224,
+      "loss": 1.9277,
+      "step": 226880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001668430959838807,
+      "loss": 2.119,
+      "step": 226885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016684172126096432,
+      "loss": 2.079,
+      "step": 226890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016684034651521357,
+      "loss": 2.1282,
+      "step": 226895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016683897174662898,
+      "loss": 2.1136,
+      "step": 226900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000166837596955211,
+      "loss": 2.1413,
+      "step": 226905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016683622214096007,
+      "loss": 2.1099,
+      "step": 226910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016683484730387667,
+      "loss": 2.1065,
+      "step": 226915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016683347244396125,
+      "loss": 1.987,
+      "step": 226920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016683209756121435,
+      "loss": 1.9823,
+      "step": 226925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016683072265563634,
+      "loss": 2.0962,
+      "step": 226930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001668293477272278,
+      "loss": 2.1155,
+      "step": 226935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001668279727759891,
+      "loss": 2.2511,
+      "step": 226940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016682659780192078,
+      "loss": 2.0853,
+      "step": 226945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001668252228050233,
+      "loss": 2.0691,
+      "step": 226950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016682384778529707,
+      "loss": 1.9802,
+      "step": 226955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016682247274274264,
+      "loss": 1.9225,
+      "step": 226960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016682109767736042,
+      "loss": 1.9457,
+      "step": 226965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016681972258915088,
+      "loss": 2.122,
+      "step": 226970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001668183474781145,
+      "loss": 2.2601,
+      "step": 226975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001668169723442518,
+      "loss": 2.2112,
+      "step": 226980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016681559718756322,
+      "loss": 2.2145,
+      "step": 226985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001668142220080492,
+      "loss": 2.094,
+      "step": 226990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016681284680571023,
+      "loss": 2.0966,
+      "step": 226995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016681147158054675,
+      "loss": 2.1887,
+      "step": 227000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016681009633255932,
+      "loss": 2.0379,
+      "step": 227005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016680872106174832,
+      "loss": 2.1464,
+      "step": 227010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016680734576811424,
+      "loss": 2.0195,
+      "step": 227015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016680597045165758,
+      "loss": 2.2449,
+      "step": 227020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016680459511237878,
+      "loss": 2.1593,
+      "step": 227025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001668032197502783,
+      "loss": 1.9733,
+      "step": 227030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016680184436535664,
+      "loss": 1.936,
+      "step": 227035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016680046895761427,
+      "loss": 1.9377,
+      "step": 227040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016679909352705166,
+      "loss": 2.1622,
+      "step": 227045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016679771807366923,
+      "loss": 2.2319,
+      "step": 227050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001667963425974675,
+      "loss": 1.9923,
+      "step": 227055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016679496709844695,
+      "loss": 2.1608,
+      "step": 227060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.000166793591576608,
+      "loss": 2.0086,
+      "step": 227065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001667922160319512,
+      "loss": 1.7817,
+      "step": 227070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001667908404644769,
+      "loss": 1.7801,
+      "step": 227075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016678946487418567,
+      "loss": 1.953,
+      "step": 227080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016678808926107793,
+      "loss": 2.1241,
+      "step": 227085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016678671362515418,
+      "loss": 2.1178,
+      "step": 227090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001667853379664149,
+      "loss": 2.0946,
+      "step": 227095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001667839622848605,
+      "loss": 2.177,
+      "step": 227100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001667825865804915,
+      "loss": 2.1575,
+      "step": 227105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016678121085330835,
+      "loss": 2.1636,
+      "step": 227110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016677983510331155,
+      "loss": 2.0585,
+      "step": 227115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016677845933050155,
+      "loss": 2.1594,
+      "step": 227120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001667770835348788,
+      "loss": 2.0439,
+      "step": 227125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001667757077164438,
+      "loss": 2.2852,
+      "step": 227130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000166774331875197,
+      "loss": 2.1423,
+      "step": 227135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016677295601113886,
+      "loss": 1.9414,
+      "step": 227140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016677158012426988,
+      "loss": 1.9752,
+      "step": 227145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016677020421459053,
+      "loss": 2.1755,
+      "step": 227150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016676882828210124,
+      "loss": 2.1111,
+      "step": 227155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016676745232680254,
+      "loss": 2.2083,
+      "step": 227160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016676607634869483,
+      "loss": 2.1951,
+      "step": 227165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016676470034777865,
+      "loss": 2.1697,
+      "step": 227170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001667633243240544,
+      "loss": 2.1274,
+      "step": 227175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016676194827752264,
+      "loss": 2.1706,
+      "step": 227180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016676057220818376,
+      "loss": 2.0595,
+      "step": 227185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016675919611603825,
+      "loss": 2.1628,
+      "step": 227190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001667578200010866,
+      "loss": 2.1899,
+      "step": 227195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016675644386332926,
+      "loss": 2.074,
+      "step": 227200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016675506770276672,
+      "loss": 2.1432,
+      "step": 227205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016675369151939942,
+      "loss": 2.1337,
+      "step": 227210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016675231531322786,
+      "loss": 2.1212,
+      "step": 227215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016675093908425249,
+      "loss": 2.2207,
+      "step": 227220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016674956283247382,
+      "loss": 2.1427,
+      "step": 227225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016674818655789225,
+      "loss": 2.0407,
+      "step": 227230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016674681026050833,
+      "loss": 2.0663,
+      "step": 227235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016674543394032244,
+      "loss": 2.1355,
+      "step": 227240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016674405759733513,
+      "loss": 1.9899,
+      "step": 227245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001667426812315468,
+      "loss": 2.0177,
+      "step": 227250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016674130484295801,
+      "loss": 2.0063,
+      "step": 227255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016673992843156915,
+      "loss": 1.936,
+      "step": 227260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016673855199738076,
+      "loss": 2.0558,
+      "step": 227265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001667371755403932,
+      "loss": 2.0761,
+      "step": 227270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016673579906060707,
+      "loss": 2.0173,
+      "step": 227275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016673442255802279,
+      "loss": 2.1777,
+      "step": 227280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001667330460326408,
+      "loss": 2.0502,
+      "step": 227285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016673166948446156,
+      "loss": 2.2279,
+      "step": 227290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016673029291348562,
+      "loss": 2.0584,
+      "step": 227295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016672891631971335,
+      "loss": 2.0024,
+      "step": 227300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016672753970314533,
+      "loss": 1.9443,
+      "step": 227305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016672616306378192,
+      "loss": 2.1164,
+      "step": 227310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016672478640162366,
+      "loss": 2.1258,
+      "step": 227315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000166723409716671,
+      "loss": 1.9154,
+      "step": 227320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016672203300892443,
+      "loss": 1.9741,
+      "step": 227325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016672065627838437,
+      "loss": 2.2499,
+      "step": 227330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016671927952505137,
+      "loss": 2.0222,
+      "step": 227335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016671790274892583,
+      "loss": 2.0583,
+      "step": 227340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016671652595000826,
+      "loss": 2.0529,
+      "step": 227345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001667151491282991,
+      "loss": 1.964,
+      "step": 227350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016671377228379884,
+      "loss": 2.1319,
+      "step": 227355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016671239541650793,
+      "loss": 1.9464,
+      "step": 227360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016671101852642686,
+      "loss": 2.1487,
+      "step": 227365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016670964161355611,
+      "loss": 2.038,
+      "step": 227370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016670826467789614,
+      "loss": 2.0674,
+      "step": 227375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001667068877194474,
+      "loss": 2.1769,
+      "step": 227380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001667055107382104,
+      "loss": 2.1633,
+      "step": 227385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016670413373418556,
+      "loss": 2.0524,
+      "step": 227390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016670275670737337,
+      "loss": 2.213,
+      "step": 227395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016670137965777438,
+      "loss": 2.0188,
+      "step": 227400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00016670000258538892,
+      "loss": 2.0074,
+      "step": 227405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016669862549021752,
+      "loss": 2.1091,
+      "step": 227410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001666972483722607,
+      "loss": 2.1829,
+      "step": 227415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016669587123151888,
+      "loss": 2.0781,
+      "step": 227420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016669449406799255,
+      "loss": 2.2238,
+      "step": 227425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016669311688168213,
+      "loss": 2.0045,
+      "step": 227430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001666917396725882,
+      "loss": 2.1803,
+      "step": 227435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001666903624407111,
+      "loss": 2.0761,
+      "step": 227440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001666889851860514,
+      "loss": 2.1402,
+      "step": 227445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016668760790860954,
+      "loss": 2.0707,
+      "step": 227450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016668623060838594,
+      "loss": 1.9634,
+      "step": 227455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016668485328538113,
+      "loss": 2.0619,
+      "step": 227460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016668347593959556,
+      "loss": 2.0075,
+      "step": 227465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016668209857102973,
+      "loss": 2.1519,
+      "step": 227470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016668072117968408,
+      "loss": 2.0424,
+      "step": 227475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016667934376555908,
+      "loss": 2.0645,
+      "step": 227480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001666779663286552,
+      "loss": 2.0424,
+      "step": 227485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016667658886897294,
+      "loss": 1.9815,
+      "step": 227490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016667521138651273,
+      "loss": 2.1869,
+      "step": 227495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016667383388127503,
+      "loss": 2.2342,
+      "step": 227500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016667245635326042,
+      "loss": 2.0567,
+      "step": 227505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001666710788024692,
+      "loss": 2.0429,
+      "step": 227510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016666970122890197,
+      "loss": 2.2557,
+      "step": 227515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016666832363255917,
+      "loss": 1.9959,
+      "step": 227520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016666694601344127,
+      "loss": 2.0976,
+      "step": 227525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001666655683715487,
+      "loss": 2.1221,
+      "step": 227530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016666419070688197,
+      "loss": 2.1762,
+      "step": 227535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016666281301944155,
+      "loss": 2.1346,
+      "step": 227540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001666614353092279,
+      "loss": 1.9221,
+      "step": 227545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00016666005757624152,
+      "loss": 2.2234,
+      "step": 227550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016665867982048281,
+      "loss": 2.154,
+      "step": 227555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016665730204195234,
+      "loss": 2.2444,
+      "step": 227560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001666559242406505,
+      "loss": 2.0649,
+      "step": 227565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016665454641657775,
+      "loss": 2.0758,
+      "step": 227570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016665316856973465,
+      "loss": 2.21,
+      "step": 227575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016665179070012162,
+      "loss": 2.1177,
+      "step": 227580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001666504128077391,
+      "loss": 2.0402,
+      "step": 227585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001666490348925876,
+      "loss": 2.0689,
+      "step": 227590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001666476569546676,
+      "loss": 2.1132,
+      "step": 227595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001666462789939795,
+      "loss": 2.1207,
+      "step": 227600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016664490101052387,
+      "loss": 2.0266,
+      "step": 227605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016664352300430112,
+      "loss": 2.144,
+      "step": 227610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016664214497531173,
+      "loss": 2.1462,
+      "step": 227615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001666407669235562,
+      "loss": 2.0952,
+      "step": 227620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016663938884903495,
+      "loss": 2.1096,
+      "step": 227625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016663801075174846,
+      "loss": 2.0823,
+      "step": 227630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016663663263169725,
+      "loss": 1.9739,
+      "step": 227635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016663525448888176,
+      "loss": 2.0601,
+      "step": 227640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016663387632330243,
+      "loss": 2.0495,
+      "step": 227645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016663249813495977,
+      "loss": 2.0749,
+      "step": 227650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016663111992385425,
+      "loss": 2.2324,
+      "step": 227655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016662974168998632,
+      "loss": 2.0547,
+      "step": 227660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016662836343335648,
+      "loss": 2.0075,
+      "step": 227665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016662698515396515,
+      "loss": 2.2086,
+      "step": 227670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016662560685181286,
+      "loss": 2.2882,
+      "step": 227675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016662422852690005,
+      "loss": 1.964,
+      "step": 227680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016662285017922718,
+      "loss": 2.0905,
+      "step": 227685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016662147180879474,
+      "loss": 1.98,
+      "step": 227690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016662009341560322,
+      "loss": 2.057,
+      "step": 227695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016661871499965304,
+      "loss": 1.992,
+      "step": 227700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001666173365609447,
+      "loss": 1.9822,
+      "step": 227705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001666159580994787,
+      "loss": 2.1015,
+      "step": 227710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001666145796152555,
+      "loss": 2.0941,
+      "step": 227715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016661320110827548,
+      "loss": 1.9997,
+      "step": 227720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001666118225785392,
+      "loss": 1.9329,
+      "step": 227725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016661044402604716,
+      "loss": 2.09,
+      "step": 227730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016660906545079972,
+      "loss": 2.1058,
+      "step": 227735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016660768685279747,
+      "loss": 2.0707,
+      "step": 227740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016660630823204083,
+      "loss": 1.9808,
+      "step": 227745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001666049295885302,
+      "loss": 2.0797,
+      "step": 227750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016660355092226617,
+      "loss": 2.0289,
+      "step": 227755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016660217223324914,
+      "loss": 2.1408,
+      "step": 227760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001666007935214796,
+      "loss": 2.1035,
+      "step": 227765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016659941478695806,
+      "loss": 2.102,
+      "step": 227770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001665980360296849,
+      "loss": 2.2273,
+      "step": 227775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016659665724966068,
+      "loss": 2.1798,
+      "step": 227780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016659527844688581,
+      "loss": 2.288,
+      "step": 227785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001665938996213608,
+      "loss": 1.9418,
+      "step": 227790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001665925207730861,
+      "loss": 2.0359,
+      "step": 227795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001665911419020622,
+      "loss": 2.1148,
+      "step": 227800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016658976300828955,
+      "loss": 2.0962,
+      "step": 227805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016658838409176863,
+      "loss": 2.0381,
+      "step": 227810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001665870051524999,
+      "loss": 2.0692,
+      "step": 227815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016658562619048384,
+      "loss": 2.0239,
+      "step": 227820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016658424720572096,
+      "loss": 2.2928,
+      "step": 227825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016658286819821162,
+      "loss": 2.284,
+      "step": 227830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016658148916795642,
+      "loss": 2.0192,
+      "step": 227835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016658011011495575,
+      "loss": 2.1866,
+      "step": 227840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016657873103921014,
+      "loss": 2.126,
+      "step": 227845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016657735194072,
+      "loss": 2.085,
+      "step": 227850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016657597281948583,
+      "loss": 2.1276,
+      "step": 227855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001665745936755081,
+      "loss": 2.1104,
+      "step": 227860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001665732145087873,
+      "loss": 2.089,
+      "step": 227865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016657183531932387,
+      "loss": 1.9451,
+      "step": 227870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001665704561071183,
+      "loss": 2.1231,
+      "step": 227875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016656907687217105,
+      "loss": 2.0101,
+      "step": 227880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001665676976144826,
+      "loss": 1.9738,
+      "step": 227885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016656631833405342,
+      "loss": 2.0802,
+      "step": 227890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016656493903088396,
+      "loss": 2.0845,
+      "step": 227895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016656355970497472,
+      "loss": 2.1788,
+      "step": 227900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016656218035632616,
+      "loss": 1.8256,
+      "step": 227905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016656080098493873,
+      "loss": 1.9537,
+      "step": 227910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016655942159081298,
+      "loss": 1.9783,
+      "step": 227915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001665580421739493,
+      "loss": 1.9067,
+      "step": 227920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001665566627343482,
+      "loss": 2.1108,
+      "step": 227925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016655528327201008,
+      "loss": 1.8994,
+      "step": 227930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001665539037869355,
+      "loss": 2.1909,
+      "step": 227935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001665525242791249,
+      "loss": 2.1336,
+      "step": 227940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016655114474857876,
+      "loss": 2.2849,
+      "step": 227945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016654976519529755,
+      "loss": 2.0084,
+      "step": 227950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001665483856192817,
+      "loss": 2.1381,
+      "step": 227955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016654700602053175,
+      "loss": 2.1662,
+      "step": 227960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001665456263990481,
+      "loss": 2.2582,
+      "step": 227965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001665442467548313,
+      "loss": 2.2118,
+      "step": 227970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016654286708788174,
+      "loss": 1.9994,
+      "step": 227975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016654148739819996,
+      "loss": 2.2805,
+      "step": 227980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016654010768578638,
+      "loss": 2.1217,
+      "step": 227985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016653872795064148,
+      "loss": 2.2538,
+      "step": 227990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016653734819276577,
+      "loss": 2.0518,
+      "step": 227995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016653596841215967,
+      "loss": 1.9322,
+      "step": 228000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001665345886088237,
+      "loss": 2.1247,
+      "step": 228005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016653320878275832,
+      "loss": 2.1693,
+      "step": 228010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00016653182893396396,
+      "loss": 2.0958,
+      "step": 228015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016653044906244114,
+      "loss": 2.1548,
+      "step": 228020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016652906916819028,
+      "loss": 2.1922,
+      "step": 228025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016652768925121193,
+      "loss": 2.2441,
+      "step": 228030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001665263093115065,
+      "loss": 2.2523,
+      "step": 228035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016652492934907446,
+      "loss": 2.0326,
+      "step": 228040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001665235493639163,
+      "loss": 2.2575,
+      "step": 228045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016652216935603248,
+      "loss": 1.9908,
+      "step": 228050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001665207893254235,
+      "loss": 2.1259,
+      "step": 228055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.00016651940927208983,
+      "loss": 2.1646,
+      "step": 228060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001665180291960319,
+      "loss": 2.2283,
+      "step": 228065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001665166490972502,
+      "loss": 2.0997,
+      "step": 228070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016651526897574521,
+      "loss": 2.0167,
+      "step": 228075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016651388883151736,
+      "loss": 2.1016,
+      "step": 228080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016651250866456723,
+      "loss": 2.0014,
+      "step": 228085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016651112847489518,
+      "loss": 1.9704,
+      "step": 228090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016650974826250175,
+      "loss": 2.2007,
+      "step": 228095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016650836802738735,
+      "loss": 2.224,
+      "step": 228100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001665069877695525,
+      "loss": 1.9737,
+      "step": 228105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016650560748899767,
+      "loss": 2.2061,
+      "step": 228110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001665042271857233,
+      "loss": 2.1192,
+      "step": 228115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016650284685972987,
+      "loss": 2.2966,
+      "step": 228120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016650146651101788,
+      "loss": 2.2105,
+      "step": 228125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001665000861395878,
+      "loss": 1.9974,
+      "step": 228130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016649870574544004,
+      "loss": 2.2312,
+      "step": 228135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016649732532857513,
+      "loss": 1.9468,
+      "step": 228140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016649594488899356,
+      "loss": 2.1854,
+      "step": 228145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016649456442669575,
+      "loss": 2.2386,
+      "step": 228150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016649318394168218,
+      "loss": 2.1919,
+      "step": 228155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016649180343395335,
+      "loss": 2.0821,
+      "step": 228160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016649042290350972,
+      "loss": 2.0096,
+      "step": 228165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001664890423503517,
+      "loss": 2.1628,
+      "step": 228170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001664876617744799,
+      "loss": 1.9984,
+      "step": 228175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016648628117589464,
+      "loss": 2.0917,
+      "step": 228180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001664849005545965,
+      "loss": 2.1929,
+      "step": 228185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001664835199105859,
+      "loss": 1.9217,
+      "step": 228190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016648213924386332,
+      "loss": 2.034,
+      "step": 228195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016648075855442925,
+      "loss": 2.1661,
+      "step": 228200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001664793778422841,
+      "loss": 2.0274,
+      "step": 228205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016647799710742844,
+      "loss": 2.1258,
+      "step": 228210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016647661634986268,
+      "loss": 1.8329,
+      "step": 228215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001664752355695873,
+      "loss": 2.0554,
+      "step": 228220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016647385476660278,
+      "loss": 2.1641,
+      "step": 228225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016647247394090958,
+      "loss": 1.7879,
+      "step": 228230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016647109309250813,
+      "loss": 2.1709,
+      "step": 228235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016646971222139903,
+      "loss": 2.1197,
+      "step": 228240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016646833132758263,
+      "loss": 2.1874,
+      "step": 228245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016646695041105947,
+      "loss": 2.0846,
+      "step": 228250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016646556947182996,
+      "loss": 2.2033,
+      "step": 228255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016646418850989458,
+      "loss": 1.889,
+      "step": 228260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001664628075252539,
+      "loss": 2.0373,
+      "step": 228265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016646142651790825,
+      "loss": 2.2566,
+      "step": 228270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016646004548785822,
+      "loss": 2.2838,
+      "step": 228275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016645866443510424,
+      "loss": 2.127,
+      "step": 228280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016645728335964674,
+      "loss": 2.0681,
+      "step": 228285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016645590226148623,
+      "loss": 2.1174,
+      "step": 228290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001664545211406232,
+      "loss": 2.0658,
+      "step": 228295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016645313999705808,
+      "loss": 2.1833,
+      "step": 228300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001664517588307914,
+      "loss": 2.1213,
+      "step": 228305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016645037764182353,
+      "loss": 2.0624,
+      "step": 228310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016644899643015505,
+      "loss": 2.1605,
+      "step": 228315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001664476151957864,
+      "loss": 2.0021,
+      "step": 228320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000166446233938718,
+      "loss": 2.0339,
+      "step": 228325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016644485265895038,
+      "loss": 2.1268,
+      "step": 228330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016644347135648401,
+      "loss": 2.0063,
+      "step": 228335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016644209003131933,
+      "loss": 1.9588,
+      "step": 228340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00016644070868345682,
+      "loss": 2.0213,
+      "step": 228345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016643932731289697,
+      "loss": 1.9878,
+      "step": 228350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016643794591964022,
+      "loss": 2.2331,
+      "step": 228355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001664365645036871,
+      "loss": 2.0962,
+      "step": 228360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.000166435183065038,
+      "loss": 2.2843,
+      "step": 228365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016643380160369348,
+      "loss": 2.1842,
+      "step": 228370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016643242011965393,
+      "loss": 2.022,
+      "step": 228375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001664310386129199,
+      "loss": 1.9995,
+      "step": 228380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016642965708349179,
+      "loss": 2.1455,
+      "step": 228385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001664282755313701,
+      "loss": 2.2716,
+      "step": 228390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016642689395655533,
+      "loss": 2.0366,
+      "step": 228395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016642551235904792,
+      "loss": 2.0516,
+      "step": 228400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016642413073884838,
+      "loss": 2.0226,
+      "step": 228405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001664227490959571,
+      "loss": 2.2638,
+      "step": 228410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016642136743037463,
+      "loss": 2.1715,
+      "step": 228415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001664199857421014,
+      "loss": 2.0386,
+      "step": 228420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016641860403113792,
+      "loss": 2.0321,
+      "step": 228425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016641722229748465,
+      "loss": 2.1283,
+      "step": 228430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000166415840541142,
+      "loss": 2.2379,
+      "step": 228435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016641445876211056,
+      "loss": 2.1502,
+      "step": 228440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001664130769603907,
+      "loss": 1.9696,
+      "step": 228445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016641169513598292,
+      "loss": 2.0434,
+      "step": 228450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016641031328888773,
+      "loss": 1.9799,
+      "step": 228455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016640893141910555,
+      "loss": 2.0192,
+      "step": 228460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016640754952663688,
+      "loss": 2.2068,
+      "step": 228465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016640616761148217,
+      "loss": 2.0303,
+      "step": 228470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016640478567364194,
+      "loss": 2.1259,
+      "step": 228475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016640340371311662,
+      "loss": 2.0382,
+      "step": 228480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001664020217299067,
+      "loss": 2.0533,
+      "step": 228485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016640063972401262,
+      "loss": 2.247,
+      "step": 228490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016639925769543491,
+      "loss": 2.1207,
+      "step": 228495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016639787564417396,
+      "loss": 2.1029,
+      "step": 228500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016639649357023034,
+      "loss": 2.0737,
+      "step": 228505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016639511147360447,
+      "loss": 2.1207,
+      "step": 228510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001663937293542968,
+      "loss": 2.1242,
+      "step": 228515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016639234721230786,
+      "loss": 2.0541,
+      "step": 228520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016639096504763806,
+      "loss": 2.1884,
+      "step": 228525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001663895828602879,
+      "loss": 2.07,
+      "step": 228530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016638820065025786,
+      "loss": 2.3075,
+      "step": 228535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00016638681841754844,
+      "loss": 2.2288,
+      "step": 228540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016638543616216006,
+      "loss": 2.309,
+      "step": 228545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001663840538840932,
+      "loss": 2.1331,
+      "step": 228550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016638267158334836,
+      "loss": 2.1137,
+      "step": 228555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016638128925992597,
+      "loss": 2.0404,
+      "step": 228560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016637990691382654,
+      "loss": 2.1399,
+      "step": 228565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016637852454505053,
+      "loss": 2.1085,
+      "step": 228570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001663771421535984,
+      "loss": 2.1108,
+      "step": 228575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016637575973947065,
+      "loss": 2.1734,
+      "step": 228580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016637437730266774,
+      "loss": 2.1075,
+      "step": 228585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016637299484319014,
+      "loss": 2.1133,
+      "step": 228590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016637161236103832,
+      "loss": 2.0075,
+      "step": 228595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016637022985621273,
+      "loss": 2.207,
+      "step": 228600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016636884732871388,
+      "loss": 2.1151,
+      "step": 228605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00016636746477854223,
+      "loss": 2.0367,
+      "step": 228610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016636608220569826,
+      "loss": 2.1034,
+      "step": 228615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016636469961018243,
+      "loss": 2.047,
+      "step": 228620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001663633169919952,
+      "loss": 2.2037,
+      "step": 228625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016636193435113708,
+      "loss": 2.2726,
+      "step": 228630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016636055168760852,
+      "loss": 2.0631,
+      "step": 228635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016635916900140996,
+      "loss": 2.1586,
+      "step": 228640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016635778629254193,
+      "loss": 2.2307,
+      "step": 228645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001663564035610049,
+      "loss": 2.2214,
+      "step": 228650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016635502080679927,
+      "loss": 2.2318,
+      "step": 228655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016635363802992558,
+      "loss": 2.1233,
+      "step": 228660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016635225523038428,
+      "loss": 1.9587,
+      "step": 228665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016635087240817584,
+      "loss": 2.0058,
+      "step": 228670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016634948956330075,
+      "loss": 1.9764,
+      "step": 228675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016634810669575946,
+      "loss": 2.09,
+      "step": 228680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001663467238055525,
+      "loss": 2.0888,
+      "step": 228685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016634534089268025,
+      "loss": 2.0556,
+      "step": 228690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016634395795714322,
+      "loss": 2.1276,
+      "step": 228695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016634257499894192,
+      "loss": 2.0734,
+      "step": 228700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016634119201807676,
+      "loss": 2.0925,
+      "step": 228705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001663398090145483,
+      "loss": 2.1025,
+      "step": 228710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016633842598835693,
+      "loss": 2.0773,
+      "step": 228715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016633704293950313,
+      "loss": 1.9973,
+      "step": 228720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001663356598679874,
+      "loss": 2.1334,
+      "step": 228725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016633427677381022,
+      "loss": 2.2406,
+      "step": 228730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016633289365697204,
+      "loss": 2.118,
+      "step": 228735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016633151051747332,
+      "loss": 2.0451,
+      "step": 228740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016633012735531462,
+      "loss": 2.0376,
+      "step": 228745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016632874417049629,
+      "loss": 2.1821,
+      "step": 228750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016632736096301888,
+      "loss": 2.0872,
+      "step": 228755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016632597773288282,
+      "loss": 2.0738,
+      "step": 228760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001663245944800886,
+      "loss": 2.2399,
+      "step": 228765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016632321120463672,
+      "loss": 2.1218,
+      "step": 228770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001663218279065276,
+      "loss": 2.1609,
+      "step": 228775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016632044458576178,
+      "loss": 2.0005,
+      "step": 228780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016631906124233968,
+      "loss": 2.1893,
+      "step": 228785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016631767787626177,
+      "loss": 2.0053,
+      "step": 228790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016631629448752856,
+      "loss": 2.0342,
+      "step": 228795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00016631491107614048,
+      "loss": 2.1282,
+      "step": 228800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016631352764209803,
+      "loss": 2.1159,
+      "step": 228805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016631214418540169,
+      "loss": 2.2761,
+      "step": 228810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016631076070605189,
+      "loss": 2.0934,
+      "step": 228815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016630937720404915,
+      "loss": 2.1516,
+      "step": 228820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016630799367939393,
+      "loss": 2.1696,
+      "step": 228825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016630661013208666,
+      "loss": 2.1244,
+      "step": 228830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001663052265621279,
+      "loss": 2.1565,
+      "step": 228835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016630384296951805,
+      "loss": 2.1469,
+      "step": 228840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001663024593542576,
+      "loss": 1.9315,
+      "step": 228845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000166301075716347,
+      "loss": 2.012,
+      "step": 228850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016629969205578682,
+      "loss": 2.02,
+      "step": 228855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016629830837257742,
+      "loss": 2.1341,
+      "step": 228860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016629692466671933,
+      "loss": 2.0585,
+      "step": 228865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016629554093821298,
+      "loss": 2.1756,
+      "step": 228870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001662941571870589,
+      "loss": 1.9368,
+      "step": 228875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001662927734132575,
+      "loss": 2.0659,
+      "step": 228880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016629138961680933,
+      "loss": 2.0198,
+      "step": 228885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001662900057977148,
+      "loss": 2.0158,
+      "step": 228890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001662886219559744,
+      "loss": 2.3462,
+      "step": 228895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016628723809158862,
+      "loss": 2.0204,
+      "step": 228900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016628585420455787,
+      "loss": 1.9568,
+      "step": 228905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001662844702948827,
+      "loss": 2.0133,
+      "step": 228910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016628308636256356,
+      "loss": 2.2844,
+      "step": 228915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016628170240760092,
+      "loss": 2.0885,
+      "step": 228920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016628031842999525,
+      "loss": 2.1312,
+      "step": 228925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.000166278934429747,
+      "loss": 2.1316,
+      "step": 228930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001662775504068567,
+      "loss": 2.2054,
+      "step": 228935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016627616636132473,
+      "loss": 2.43,
+      "step": 228940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016627478229315167,
+      "loss": 2.0909,
+      "step": 228945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016627339820233795,
+      "loss": 2.1736,
+      "step": 228950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016627201408888397,
+      "loss": 2.0042,
+      "step": 228955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001662706299527903,
+      "loss": 2.2507,
+      "step": 228960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001662692457940574,
+      "loss": 2.2052,
+      "step": 228965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016626786161268574,
+      "loss": 2.1869,
+      "step": 228970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016626647740867573,
+      "loss": 2.086,
+      "step": 228975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001662650931820279,
+      "loss": 2.1538,
+      "step": 228980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016626370893274271,
+      "loss": 2.0471,
+      "step": 228985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016626232466082066,
+      "loss": 2.2609,
+      "step": 228990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016626094036626218,
+      "loss": 2.0602,
+      "step": 228995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016625955604906775,
+      "loss": 1.9928,
+      "step": 229000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016625817170923789,
+      "loss": 1.9682,
+      "step": 229005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000166256787346773,
+      "loss": 2.0656,
+      "step": 229010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016625540296167358,
+      "loss": 1.9512,
+      "step": 229015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016625401855394016,
+      "loss": 1.9228,
+      "step": 229020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016625263412357316,
+      "loss": 2.0691,
+      "step": 229025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016625124967057305,
+      "loss": 2.0181,
+      "step": 229030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016624986519494028,
+      "loss": 2.1336,
+      "step": 229035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001662484806966754,
+      "loss": 2.2738,
+      "step": 229040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001662470961757788,
+      "loss": 2.1458,
+      "step": 229045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016624571163225102,
+      "loss": 2.2232,
+      "step": 229050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001662443270660925,
+      "loss": 2.0892,
+      "step": 229055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016624294247730368,
+      "loss": 2.1004,
+      "step": 229060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001662415578658851,
+      "loss": 2.0832,
+      "step": 229065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001662401732318372,
+      "loss": 2.1235,
+      "step": 229070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016623878857516044,
+      "loss": 2.0747,
+      "step": 229075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001662374038958553,
+      "loss": 2.3826,
+      "step": 229080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016623601919392233,
+      "loss": 2.1438,
+      "step": 229085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016623463446936186,
+      "loss": 2.073,
+      "step": 229090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016623324972217448,
+      "loss": 2.1488,
+      "step": 229095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001662318649523606,
+      "loss": 2.0668,
+      "step": 229100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016623048015992073,
+      "loss": 2.1234,
+      "step": 229105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001662290953448553,
+      "loss": 2.2155,
+      "step": 229110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016622771050716484,
+      "loss": 2.0841,
+      "step": 229115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016622632564684977,
+      "loss": 2.0214,
+      "step": 229120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001662249407639106,
+      "loss": 2.0038,
+      "step": 229125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016622355585834778,
+      "loss": 2.2442,
+      "step": 229130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001662221709301618,
+      "loss": 2.1914,
+      "step": 229135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016622078597935313,
+      "loss": 2.1831,
+      "step": 229140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016621940100592224,
+      "loss": 2.2108,
+      "step": 229145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001662180160098696,
+      "loss": 2.1574,
+      "step": 229150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016621663099119566,
+      "loss": 2.2834,
+      "step": 229155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016621524594990096,
+      "loss": 2.0692,
+      "step": 229160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001662138608859859,
+      "loss": 2.1087,
+      "step": 229165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016621247579945097,
+      "loss": 2.1942,
+      "step": 229170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016621109069029667,
+      "loss": 1.9995,
+      "step": 229175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016620970555852348,
+      "loss": 2.2304,
+      "step": 229180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016620832040413185,
+      "loss": 2.0937,
+      "step": 229185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016620693522712224,
+      "loss": 2.1459,
+      "step": 229190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016620555002749517,
+      "loss": 2.0982,
+      "step": 229195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016620416480525103,
+      "loss": 2.0578,
+      "step": 229200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001662027795603904,
+      "loss": 2.0333,
+      "step": 229205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016620139429291368,
+      "loss": 2.1665,
+      "step": 229210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016620000900282138,
+      "loss": 2.108,
+      "step": 229215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016619862369011393,
+      "loss": 2.1176,
+      "step": 229220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016619723835479183,
+      "loss": 2.0469,
+      "step": 229225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016619585299685555,
+      "loss": 2.1017,
+      "step": 229230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016619446761630558,
+      "loss": 2.2598,
+      "step": 229235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016619308221314238,
+      "loss": 2.1513,
+      "step": 229240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016619169678736643,
+      "loss": 2.1661,
+      "step": 229245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016619031133897818,
+      "loss": 2.0472,
+      "step": 229250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016618892586797813,
+      "loss": 2.2956,
+      "step": 229255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001661875403743667,
+      "loss": 2.0257,
+      "step": 229260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016618615485814448,
+      "loss": 2.0886,
+      "step": 229265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016618476931931183,
+      "loss": 2.1481,
+      "step": 229270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001661833837578693,
+      "loss": 2.114,
+      "step": 229275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016618199817381726,
+      "loss": 2.1768,
+      "step": 229280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001661806125671563,
+      "loss": 1.9972,
+      "step": 229285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001661792269378868,
+      "loss": 2.1019,
+      "step": 229290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001661778412860093,
+      "loss": 2.1102,
+      "step": 229295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016617645561152426,
+      "loss": 2.377,
+      "step": 229300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016617506991443211,
+      "loss": 2.1544,
+      "step": 229305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016617368419473338,
+      "loss": 2.033,
+      "step": 229310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001661722984524285,
+      "loss": 2.2133,
+      "step": 229315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000166170912687518,
+      "loss": 2.1178,
+      "step": 229320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001661695269000023,
+      "loss": 2.1604,
+      "step": 229325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016616814108988187,
+      "loss": 2.1545,
+      "step": 229330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016616675525715722,
+      "loss": 2.2623,
+      "step": 229335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001661653694018288,
+      "loss": 1.9676,
+      "step": 229340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001661639835238971,
+      "loss": 2.1546,
+      "step": 229345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016616259762336258,
+      "loss": 2.2564,
+      "step": 229350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016616121170022572,
+      "loss": 2.2763,
+      "step": 229355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016615982575448696,
+      "loss": 2.1642,
+      "step": 229360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016615843978614683,
+      "loss": 2.0927,
+      "step": 229365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016615705379520579,
+      "loss": 2.0498,
+      "step": 229370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001661556677816643,
+      "loss": 2.1697,
+      "step": 229375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001661542817455228,
+      "loss": 2.1024,
+      "step": 229380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001661528956867818,
+      "loss": 2.0309,
+      "step": 229385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001661515096054418,
+      "loss": 2.2364,
+      "step": 229390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016615012350150324,
+      "loss": 2.0681,
+      "step": 229395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016614873737496658,
+      "loss": 2.2783,
+      "step": 229400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016614735122583233,
+      "loss": 1.9235,
+      "step": 229405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016614596505410092,
+      "loss": 2.0502,
+      "step": 229410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016614457885977287,
+      "loss": 2.0633,
+      "step": 229415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016614319264284864,
+      "loss": 2.1153,
+      "step": 229420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001661418064033287,
+      "loss": 2.0505,
+      "step": 229425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001661404201412135,
+      "loss": 2.1517,
+      "step": 229430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001661390338565035,
+      "loss": 2.2472,
+      "step": 229435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016613764754919928,
+      "loss": 2.0214,
+      "step": 229440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001661362612193012,
+      "loss": 2.054,
+      "step": 229445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016613487486680977,
+      "loss": 1.9278,
+      "step": 229450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016613348849172548,
+      "loss": 2.0713,
+      "step": 229455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016613210209404878,
+      "loss": 2.1613,
+      "step": 229460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016613071567378012,
+      "loss": 2.0405,
+      "step": 229465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016612932923092008,
+      "loss": 2.1934,
+      "step": 229470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016612794276546903,
+      "loss": 2.0422,
+      "step": 229475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016612655627742747,
+      "loss": 1.9081,
+      "step": 229480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001661251697667959,
+      "loss": 2.2105,
+      "step": 229485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016612378323357474,
+      "loss": 2.0251,
+      "step": 229490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016612239667776452,
+      "loss": 2.2126,
+      "step": 229495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016612101009936568,
+      "loss": 2.0819,
+      "step": 229500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001661196234983787,
+      "loss": 2.0941,
+      "step": 229505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016611823687480407,
+      "loss": 2.1217,
+      "step": 229510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016611685022864227,
+      "loss": 2.1744,
+      "step": 229515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016611546355989372,
+      "loss": 2.1207,
+      "step": 229520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00016611407686855894,
+      "loss": 2.1432,
+      "step": 229525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016611269015463838,
+      "loss": 2.1947,
+      "step": 229530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016611130341813253,
+      "loss": 1.9591,
+      "step": 229535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001661099166590419,
+      "loss": 1.9048,
+      "step": 229540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016610852987736688,
+      "loss": 2.0873,
+      "step": 229545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000166107143073108,
+      "loss": 2.1391,
+      "step": 229550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001661057562462657,
+      "loss": 2.0755,
+      "step": 229555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016610436939684052,
+      "loss": 2.0058,
+      "step": 229560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016610298252483285,
+      "loss": 2.1128,
+      "step": 229565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016610159563024322,
+      "loss": 2.01,
+      "step": 229570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016610020871307208,
+      "loss": 2.0347,
+      "step": 229575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001660988217733199,
+      "loss": 1.9985,
+      "step": 229580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001660974348109872,
+      "loss": 1.913,
+      "step": 229585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016609604782607436,
+      "loss": 2.1345,
+      "step": 229590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016609466081858197,
+      "loss": 2.0307,
+      "step": 229595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016609327378851041,
+      "loss": 1.9968,
+      "step": 229600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016609188673586021,
+      "loss": 2.1021,
+      "step": 229605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001660904996606318,
+      "loss": 2.1712,
+      "step": 229610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016608911256282567,
+      "loss": 1.8236,
+      "step": 229615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016608772544244232,
+      "loss": 2.1729,
+      "step": 229620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016608633829948223,
+      "loss": 2.1796,
+      "step": 229625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001660849511339458,
+      "loss": 2.0584,
+      "step": 229630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016608356394583358,
+      "loss": 2.2177,
+      "step": 229635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000166082176735146,
+      "loss": 2.0576,
+      "step": 229640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016608078950188358,
+      "loss": 2.1569,
+      "step": 229645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001660794022460467,
+      "loss": 2.2879,
+      "step": 229650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016607801496763594,
+      "loss": 2.2147,
+      "step": 229655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016607662766665175,
+      "loss": 2.0574,
+      "step": 229660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016607524034309456,
+      "loss": 1.9301,
+      "step": 229665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001660738529969649,
+      "loss": 2.1244,
+      "step": 229670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001660724656282632,
+      "loss": 2.1414,
+      "step": 229675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016607107823698991,
+      "loss": 2.0004,
+      "step": 229680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016606969082314557,
+      "loss": 2.3029,
+      "step": 229685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016606830338673062,
+      "loss": 2.0659,
+      "step": 229690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016606691592774557,
+      "loss": 2.0594,
+      "step": 229695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016606552844619082,
+      "loss": 2.051,
+      "step": 229700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001660641409420669,
+      "loss": 2.0306,
+      "step": 229705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001660627534153743,
+      "loss": 2.2099,
+      "step": 229710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016606136586611343,
+      "loss": 1.9669,
+      "step": 229715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001660599782942848,
+      "loss": 1.897,
+      "step": 229720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016605859069988892,
+      "loss": 2.2686,
+      "step": 229725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016605720308292618,
+      "loss": 2.0601,
+      "step": 229730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016605581544339713,
+      "loss": 2.1381,
+      "step": 229735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001660544277813022,
+      "loss": 2.1013,
+      "step": 229740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001660530400966419,
+      "loss": 2.1395,
+      "step": 229745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016605165238941667,
+      "loss": 2.0051,
+      "step": 229750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.000166050264659627,
+      "loss": 1.8915,
+      "step": 229755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00016604887690727335,
+      "loss": 2.0036,
+      "step": 229760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001660474891323562,
+      "loss": 2.0932,
+      "step": 229765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016604610133487605,
+      "loss": 2.0614,
+      "step": 229770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016604471351483338,
+      "loss": 2.0763,
+      "step": 229775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001660433256722286,
+      "loss": 2.0302,
+      "step": 229780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001660419378070622,
+      "loss": 2.0566,
+      "step": 229785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016604054991933474,
+      "loss": 2.133,
+      "step": 229790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001660391620090466,
+      "loss": 1.8798,
+      "step": 229795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016603777407619828,
+      "loss": 2.249,
+      "step": 229800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016603638612079026,
+      "loss": 2.0539,
+      "step": 229805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016603499814282302,
+      "loss": 2.0199,
+      "step": 229810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000166033610142297,
+      "loss": 2.1305,
+      "step": 229815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001660322221192127,
+      "loss": 2.2188,
+      "step": 229820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016603083407357066,
+      "loss": 2.0451,
+      "step": 229825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016602944600537124,
+      "loss": 2.055,
+      "step": 229830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016602805791461497,
+      "loss": 2.1545,
+      "step": 229835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001660266698013023,
+      "loss": 2.2067,
+      "step": 229840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016602528166543374,
+      "loss": 2.2159,
+      "step": 229845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016602389350700976,
+      "loss": 2.066,
+      "step": 229850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016602250532603082,
+      "loss": 2.2165,
+      "step": 229855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016602111712249738,
+      "loss": 2.0252,
+      "step": 229860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016601972889640992,
+      "loss": 2.1399,
+      "step": 229865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016601834064776893,
+      "loss": 2.2492,
+      "step": 229870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016601695237657488,
+      "loss": 2.0073,
+      "step": 229875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016601556408282826,
+      "loss": 1.9373,
+      "step": 229880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016601417576652948,
+      "loss": 2.0205,
+      "step": 229885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016601278742767913,
+      "loss": 2.0456,
+      "step": 229890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016601139906627753,
+      "loss": 2.0304,
+      "step": 229895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016601001068232532,
+      "loss": 2.0577,
+      "step": 229900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016600862227582284,
+      "loss": 1.8778,
+      "step": 229905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016600723384677064,
+      "loss": 2.2288,
+      "step": 229910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016600584539516917,
+      "loss": 2.1279,
+      "step": 229915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.15625,
+      "learning_rate": 0.0001660044569210189,
+      "loss": 2.0154,
+      "step": 229920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00016600306842432032,
+      "loss": 2.1789,
+      "step": 229925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016600167990507388,
+      "loss": 2.2788,
+      "step": 229930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016600029136328007,
+      "loss": 1.9928,
+      "step": 229935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016599890279893938,
+      "loss": 2.3711,
+      "step": 229940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016599751421205228,
+      "loss": 2.1047,
+      "step": 229945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001659961256026192,
+      "loss": 2.114,
+      "step": 229950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016599473697064067,
+      "loss": 2.1598,
+      "step": 229955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001659933483161171,
+      "loss": 2.1801,
+      "step": 229960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016599195963904904,
+      "loss": 2.2613,
+      "step": 229965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016599057093943692,
+      "loss": 2.0658,
+      "step": 229970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001659891822172812,
+      "loss": 1.9758,
+      "step": 229975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016598779347258243,
+      "loss": 1.9962,
+      "step": 229980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000165986404705341,
+      "loss": 1.9437,
+      "step": 229985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016598501591555745,
+      "loss": 2.3095,
+      "step": 229990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001659836271032322,
+      "loss": 1.9837,
+      "step": 229995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016598223826836574,
+      "loss": 2.2264,
+      "step": 230000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016598084941095854,
+      "loss": 2.1608,
+      "step": 230005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001659794605310111,
+      "loss": 2.1123,
+      "step": 230010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001659780716285239,
+      "loss": 2.3223,
+      "step": 230015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016597668270349738,
+      "loss": 2.2672,
+      "step": 230020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016597529375593202,
+      "loss": 2.0875,
+      "step": 230025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001659739047858283,
+      "loss": 2.0925,
+      "step": 230030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016597251579318672,
+      "loss": 2.1179,
+      "step": 230035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001659711267780077,
+      "loss": 2.0268,
+      "step": 230040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016596973774029173,
+      "loss": 1.9052,
+      "step": 230045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016596834868003936,
+      "loss": 2.2169,
+      "step": 230050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016596695959725097,
+      "loss": 2.2387,
+      "step": 230055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001659655704919271,
+      "loss": 2.0676,
+      "step": 230060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016596418136406819,
+      "loss": 2.1814,
+      "step": 230065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016596279221367467,
+      "loss": 2.1939,
+      "step": 230070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016596140304074713,
+      "loss": 2.0707,
+      "step": 230075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001659600138452859,
+      "loss": 2.0096,
+      "step": 230080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016595862462729158,
+      "loss": 2.0423,
+      "step": 230085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016595723538676463,
+      "loss": 2.1695,
+      "step": 230090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016595584612370546,
+      "loss": 2.1451,
+      "step": 230095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016595445683811457,
+      "loss": 1.9112,
+      "step": 230100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016595306752999244,
+      "loss": 2.0639,
+      "step": 230105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016595167819933955,
+      "loss": 2.2092,
+      "step": 230110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016595028884615638,
+      "loss": 2.1001,
+      "step": 230115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016594889947044338,
+      "loss": 2.2222,
+      "step": 230120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016594751007220103,
+      "loss": 2.1081,
+      "step": 230125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016594612065142986,
+      "loss": 2.0366,
+      "step": 230130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016594473120813027,
+      "loss": 2.1872,
+      "step": 230135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016594334174230277,
+      "loss": 2.0259,
+      "step": 230140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016594195225394783,
+      "loss": 1.852,
+      "step": 230145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001659405627430659,
+      "loss": 2.0953,
+      "step": 230150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001659391732096575,
+      "loss": 2.0893,
+      "step": 230155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016593778365372308,
+      "loss": 2.2511,
+      "step": 230160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001659363940752631,
+      "loss": 2.0769,
+      "step": 230165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016593500447427807,
+      "loss": 2.2704,
+      "step": 230170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016593361485076847,
+      "loss": 2.2805,
+      "step": 230175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001659322252047347,
+      "loss": 2.1656,
+      "step": 230180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016593083553617733,
+      "loss": 2.1656,
+      "step": 230185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016592944584509675,
+      "loss": 2.044,
+      "step": 230190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016592805613149352,
+      "loss": 2.1355,
+      "step": 230195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016592666639536806,
+      "loss": 2.0305,
+      "step": 230200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016592527663672084,
+      "loss": 2.0905,
+      "step": 230205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016592388685555232,
+      "loss": 2.2179,
+      "step": 230210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016592249705186304,
+      "loss": 2.1932,
+      "step": 230215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016592110722565346,
+      "loss": 1.9358,
+      "step": 230220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016591971737692403,
+      "loss": 1.8901,
+      "step": 230225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001659183275056752,
+      "loss": 2.0705,
+      "step": 230230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001659169376119075,
+      "loss": 2.0239,
+      "step": 230235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016591554769562135,
+      "loss": 2.2552,
+      "step": 230240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016591415775681724,
+      "loss": 2.1492,
+      "step": 230245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016591276779549569,
+      "loss": 2.1452,
+      "step": 230250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016591137781165715,
+      "loss": 2.0282,
+      "step": 230255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001659099878053021,
+      "loss": 2.0929,
+      "step": 230260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000165908597776431,
+      "loss": 1.9355,
+      "step": 230265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016590720772504427,
+      "loss": 2.1482,
+      "step": 230270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016590581765114245,
+      "loss": 2.0788,
+      "step": 230275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016590442755472606,
+      "loss": 2.2292,
+      "step": 230280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001659030374357955,
+      "loss": 2.1249,
+      "step": 230285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016590164729435127,
+      "loss": 2.0617,
+      "step": 230290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016590025713039382,
+      "loss": 2.2094,
+      "step": 230295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001658988669439237,
+      "loss": 2.0756,
+      "step": 230300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016589747673494127,
+      "loss": 1.9482,
+      "step": 230305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.640625,
+      "learning_rate": 0.0001658960865034471,
+      "loss": 1.9923,
+      "step": 230310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016589469624944163,
+      "loss": 2.1314,
+      "step": 230315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016589330597292535,
+      "loss": 2.0328,
+      "step": 230320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016589191567389867,
+      "loss": 2.0447,
+      "step": 230325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016589052535236217,
+      "loss": 2.051,
+      "step": 230330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016588913500831628,
+      "loss": 2.2616,
+      "step": 230335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001658877446417614,
+      "loss": 2.1574,
+      "step": 230340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016588635425269813,
+      "loss": 2.0481,
+      "step": 230345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016588496384112685,
+      "loss": 2.2079,
+      "step": 230350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016588357340704808,
+      "loss": 2.0992,
+      "step": 230355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001658821829504623,
+      "loss": 2.0341,
+      "step": 230360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016588079247136997,
+      "loss": 1.9772,
+      "step": 230365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016587940196977158,
+      "loss": 2.1177,
+      "step": 230370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016587801144566755,
+      "loss": 2.1356,
+      "step": 230375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016587662089905845,
+      "loss": 2.224,
+      "step": 230380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016587523032994466,
+      "loss": 2.1265,
+      "step": 230385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001658738397383267,
+      "loss": 2.0495,
+      "step": 230390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016587244912420504,
+      "loss": 2.2481,
+      "step": 230395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016587105848758017,
+      "loss": 1.9715,
+      "step": 230400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00016586966782845255,
+      "loss": 2.1342,
+      "step": 230405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016586827714682267,
+      "loss": 2.026,
+      "step": 230410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016586688644269098,
+      "loss": 2.0808,
+      "step": 230415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016586549571605795,
+      "loss": 2.058,
+      "step": 230420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001658641049669241,
+      "loss": 2.1068,
+      "step": 230425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016586271419528984,
+      "loss": 2.2217,
+      "step": 230430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001658613234011557,
+      "loss": 1.9623,
+      "step": 230435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016585993258452216,
+      "loss": 2.1107,
+      "step": 230440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001658585417453897,
+      "loss": 2.0561,
+      "step": 230445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016585715088375872,
+      "loss": 2.2171,
+      "step": 230450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016585575999962973,
+      "loss": 2.0931,
+      "step": 230455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016585436909300323,
+      "loss": 2.0571,
+      "step": 230460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016585297816387967,
+      "loss": 2.0363,
+      "step": 230465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016585158721225957,
+      "loss": 2.1662,
+      "step": 230470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016585019623814335,
+      "loss": 2.0479,
+      "step": 230475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016584880524153154,
+      "loss": 2.0536,
+      "step": 230480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016584741422242454,
+      "loss": 2.0283,
+      "step": 230485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016584602318082289,
+      "loss": 2.2546,
+      "step": 230490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016584463211672708,
+      "loss": 1.7948,
+      "step": 230495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016584324103013752,
+      "loss": 2.0559,
+      "step": 230500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001658418499210547,
+      "loss": 2.033,
+      "step": 230505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016584045878947912,
+      "loss": 2.1807,
+      "step": 230510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016583906763541125,
+      "loss": 2.0889,
+      "step": 230515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016583767645885153,
+      "loss": 2.1184,
+      "step": 230520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001658362852598005,
+      "loss": 2.1264,
+      "step": 230525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001658348940382586,
+      "loss": 2.1837,
+      "step": 230530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016583350279422628,
+      "loss": 2.2068,
+      "step": 230535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016583211152770407,
+      "loss": 2.1397,
+      "step": 230540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016583072023869242,
+      "loss": 2.0977,
+      "step": 230545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016582932892719179,
+      "loss": 2.0801,
+      "step": 230550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016582793759320269,
+      "loss": 2.2342,
+      "step": 230555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016582654623672552,
+      "loss": 2.3001,
+      "step": 230560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016582515485776082,
+      "loss": 2.0512,
+      "step": 230565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016582376345630908,
+      "loss": 2.068,
+      "step": 230570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016582237203237074,
+      "loss": 2.0333,
+      "step": 230575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001658209805859463,
+      "loss": 1.9796,
+      "step": 230580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001658195891170362,
+      "loss": 2.1391,
+      "step": 230585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016581819762564095,
+      "loss": 2.0415,
+      "step": 230590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016581680611176097,
+      "loss": 2.1863,
+      "step": 230595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001658154145753968,
+      "loss": 2.0945,
+      "step": 230600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016581402301654892,
+      "loss": 2.0808,
+      "step": 230605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016581263143521774,
+      "loss": 2.0685,
+      "step": 230610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001658112398314038,
+      "loss": 2.2292,
+      "step": 230615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001658098482051075,
+      "loss": 2.0437,
+      "step": 230620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016580845655632943,
+      "loss": 2.2257,
+      "step": 230625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016580706488506994,
+      "loss": 2.0289,
+      "step": 230630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016580567319132955,
+      "loss": 2.0648,
+      "step": 230635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001658042814751088,
+      "loss": 2.1781,
+      "step": 230640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016580288973640808,
+      "loss": 2.103,
+      "step": 230645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016580149797522793,
+      "loss": 2.1077,
+      "step": 230650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001658001061915688,
+      "loss": 2.0936,
+      "step": 230655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016579871438543111,
+      "loss": 2.1761,
+      "step": 230660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001657973225568154,
+      "loss": 2.3544,
+      "step": 230665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016579593070572215,
+      "loss": 1.9453,
+      "step": 230670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016579453883215184,
+      "loss": 2.1549,
+      "step": 230675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016579314693610487,
+      "loss": 2.1118,
+      "step": 230680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001657917550175818,
+      "loss": 2.0874,
+      "step": 230685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001657903630765831,
+      "loss": 2.1437,
+      "step": 230690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001657889711131092,
+      "loss": 2.045,
+      "step": 230695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016578757912716054,
+      "loss": 2.2446,
+      "step": 230700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016578618711873767,
+      "loss": 2.1385,
+      "step": 230705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001657847950878411,
+      "loss": 2.1467,
+      "step": 230710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001657834030344712,
+      "loss": 1.8784,
+      "step": 230715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016578201095862851,
+      "loss": 2.2402,
+      "step": 230720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016578061886031347,
+      "loss": 2.0561,
+      "step": 230725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001657792267395266,
+      "loss": 2.0811,
+      "step": 230730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016577783459626838,
+      "loss": 2.142,
+      "step": 230735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016577644243053925,
+      "loss": 2.2327,
+      "step": 230740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016577505024233967,
+      "loss": 2.2351,
+      "step": 230745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016577365803167014,
+      "loss": 2.0976,
+      "step": 230750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016577226579853115,
+      "loss": 2.2778,
+      "step": 230755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016577087354292314,
+      "loss": 1.9572,
+      "step": 230760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016576948126484662,
+      "loss": 2.2512,
+      "step": 230765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016576808896430207,
+      "loss": 2.2079,
+      "step": 230770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016576669664128995,
+      "loss": 2.0384,
+      "step": 230775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016576530429581067,
+      "loss": 1.9425,
+      "step": 230780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016576391192786482,
+      "loss": 2.1375,
+      "step": 230785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016576251953745284,
+      "loss": 1.9237,
+      "step": 230790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016576112712457514,
+      "loss": 2.1332,
+      "step": 230795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001657597346892323,
+      "loss": 2.1126,
+      "step": 230800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001657583422314247,
+      "loss": 2.1069,
+      "step": 230805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001657569497511529,
+      "loss": 2.0326,
+      "step": 230810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001657555572484173,
+      "loss": 2.1458,
+      "step": 230815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001657541647232184,
+      "loss": 2.1446,
+      "step": 230820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016575277217555673,
+      "loss": 2.0306,
+      "step": 230825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016575137960543268,
+      "loss": 2.1383,
+      "step": 230830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016574998701284678,
+      "loss": 2.0099,
+      "step": 230835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001657485943977995,
+      "loss": 2.1802,
+      "step": 230840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001657472017602913,
+      "loss": 2.2346,
+      "step": 230845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016574580910032268,
+      "loss": 2.1417,
+      "step": 230850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016574441641789407,
+      "loss": 2.1827,
+      "step": 230855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.75,
+      "learning_rate": 0.000165743023713006,
+      "loss": 2.1257,
+      "step": 230860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001657416309856589,
+      "loss": 2.0827,
+      "step": 230865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016574023823585327,
+      "loss": 2.2552,
+      "step": 230870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016573884546358956,
+      "loss": 2.1205,
+      "step": 230875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016573745266886832,
+      "loss": 2.1254,
+      "step": 230880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016573605985168994,
+      "loss": 2.0176,
+      "step": 230885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00016573466701205495,
+      "loss": 2.1461,
+      "step": 230890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001657332741499638,
+      "loss": 2.1344,
+      "step": 230895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016573188126541697,
+      "loss": 2.1812,
+      "step": 230900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016573048835841493,
+      "loss": 2.0479,
+      "step": 230905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016572909542895815,
+      "loss": 1.9725,
+      "step": 230910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016572770247704713,
+      "loss": 2.0626,
+      "step": 230915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016572630950268237,
+      "loss": 2.2152,
+      "step": 230920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016572491650586427,
+      "loss": 2.1217,
+      "step": 230925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016572352348659335,
+      "loss": 2.1176,
+      "step": 230930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001657221304448701,
+      "loss": 2.1997,
+      "step": 230935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016572073738069495,
+      "loss": 2.0223,
+      "step": 230940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016571934429406842,
+      "loss": 2.2474,
+      "step": 230945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016571795118499097,
+      "loss": 2.2025,
+      "step": 230950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001657165580534631,
+      "loss": 2.1187,
+      "step": 230955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016571516489948524,
+      "loss": 2.2168,
+      "step": 230960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016571377172305788,
+      "loss": 2.0361,
+      "step": 230965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016571237852418152,
+      "loss": 2.0586,
+      "step": 230970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016571098530285662,
+      "loss": 2.1501,
+      "step": 230975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016570959205908364,
+      "loss": 2.1014,
+      "step": 230980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001657081987928631,
+      "loss": 1.9841,
+      "step": 230985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016570680550419538,
+      "loss": 1.8608,
+      "step": 230990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016570541219308109,
+      "loss": 2.0511,
+      "step": 230995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00016570401885952066,
+      "loss": 1.9398,
+      "step": 231000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016570262550351448,
+      "loss": 2.0843,
+      "step": 231005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001657012321250631,
+      "loss": 2.2819,
+      "step": 231010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016569983872416702,
+      "loss": 2.0784,
+      "step": 231015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016569844530082666,
+      "loss": 2.1486,
+      "step": 231020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016569705185504253,
+      "loss": 2.0991,
+      "step": 231025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016569565838681508,
+      "loss": 1.9866,
+      "step": 231030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016569426489614484,
+      "loss": 2.0092,
+      "step": 231035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016569287138303222,
+      "loss": 1.9818,
+      "step": 231040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016569147784747772,
+      "loss": 1.9559,
+      "step": 231045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016569008428948183,
+      "loss": 2.0861,
+      "step": 231050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000165688690709045,
+      "loss": 2.0312,
+      "step": 231055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016568729710616774,
+      "loss": 2.0323,
+      "step": 231060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016568590348085051,
+      "loss": 2.1555,
+      "step": 231065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016568450983309377,
+      "loss": 1.9592,
+      "step": 231070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016568311616289803,
+      "loss": 2.0643,
+      "step": 231075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016568172247026372,
+      "loss": 1.9047,
+      "step": 231080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016568032875519134,
+      "loss": 2.0786,
+      "step": 231085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001656789350176814,
+      "loss": 2.0274,
+      "step": 231090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016567754125773433,
+      "loss": 2.0239,
+      "step": 231095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016567614747535063,
+      "loss": 2.0214,
+      "step": 231100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016567475367053076,
+      "loss": 2.1831,
+      "step": 231105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016567335984327518,
+      "loss": 2.1461,
+      "step": 231110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001656719659935844,
+      "loss": 2.0025,
+      "step": 231115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001656705721214589,
+      "loss": 2.0171,
+      "step": 231120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016566917822689913,
+      "loss": 2.0439,
+      "step": 231125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0001656677843099056,
+      "loss": 1.9454,
+      "step": 231130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016566639037047875,
+      "loss": 2.1231,
+      "step": 231135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016566499640861907,
+      "loss": 2.0438,
+      "step": 231140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016566360242432703,
+      "loss": 2.0703,
+      "step": 231145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016566220841760312,
+      "loss": 2.0836,
+      "step": 231150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001656608143884478,
+      "loss": 2.1607,
+      "step": 231155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016565942033686156,
+      "loss": 2.0987,
+      "step": 231160
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016565802626284488,
+      "loss": 2.091,
+      "step": 231165
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001656566321663982,
+      "loss": 2.0031,
+      "step": 231170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016565523804752205,
+      "loss": 2.0973,
+      "step": 231175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001656538439062169,
+      "loss": 2.1926,
+      "step": 231180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016565244974248316,
+      "loss": 2.1798,
+      "step": 231185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016565105555632137,
+      "loss": 2.2001,
+      "step": 231190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016564966134773198,
+      "loss": 2.1532,
+      "step": 231195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016564826711671548,
+      "loss": 2.0333,
+      "step": 231200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016564687286327234,
+      "loss": 2.0459,
+      "step": 231205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016564547858740304,
+      "loss": 1.7872,
+      "step": 231210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016564408428910808,
+      "loss": 2.1454,
+      "step": 231215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016564268996838787,
+      "loss": 1.9639,
+      "step": 231220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016564129562524296,
+      "loss": 1.9454,
+      "step": 231225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001656399012596738,
+      "loss": 2.0726,
+      "step": 231230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016563850687168082,
+      "loss": 2.1394,
+      "step": 231235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016563711246126453,
+      "loss": 2.0982,
+      "step": 231240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016563571802842545,
+      "loss": 2.0501,
+      "step": 231245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000165634323573164,
+      "loss": 2.1405,
+      "step": 231250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001656329290954807,
+      "loss": 2.1968,
+      "step": 231255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016563153459537598,
+      "loss": 2.1593,
+      "step": 231260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016563014007285032,
+      "loss": 1.9989,
+      "step": 231265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016562874552790422,
+      "loss": 2.2339,
+      "step": 231270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016562735096053817,
+      "loss": 2.0511,
+      "step": 231275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001656259563707526,
+      "loss": 2.152,
+      "step": 231280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016562456175854803,
+      "loss": 2.0326,
+      "step": 231285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001656231671239249,
+      "loss": 2.1139,
+      "step": 231290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016562177246688377,
+      "loss": 2.0318,
+      "step": 231295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016562037778742502,
+      "loss": 2.0711,
+      "step": 231300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001656189830855491,
+      "loss": 2.2211,
+      "step": 231305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016561758836125662,
+      "loss": 2.0179,
+      "step": 231310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016561619361454795,
+      "loss": 2.0247,
+      "step": 231315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001656147988454236,
+      "loss": 2.0623,
+      "step": 231320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016561340405388405,
+      "loss": 2.1639,
+      "step": 231325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016561200923992977,
+      "loss": 1.7734,
+      "step": 231330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00016561061440356124,
+      "loss": 2.2224,
+      "step": 231335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016560921954477892,
+      "loss": 1.9217,
+      "step": 231340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001656078246635833,
+      "loss": 2.2967,
+      "step": 231345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016560642975997487,
+      "loss": 2.1397,
+      "step": 231350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016560503483395412,
+      "loss": 2.1807,
+      "step": 231355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016560363988552145,
+      "loss": 2.2132,
+      "step": 231360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016560224491467742,
+      "loss": 1.9456,
+      "step": 231365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016560084992142245,
+      "loss": 2.2423,
+      "step": 231370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016559945490575705,
+      "loss": 1.9918,
+      "step": 231375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001655980598676817,
+      "loss": 2.0556,
+      "step": 231380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016559666480719685,
+      "loss": 2.1585,
+      "step": 231385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000165595269724303,
+      "loss": 2.1861,
+      "step": 231390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001655938746190006,
+      "loss": 2.1242,
+      "step": 231395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016559247949129015,
+      "loss": 2.2612,
+      "step": 231400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016559108434117213,
+      "loss": 2.0656,
+      "step": 231405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016558968916864699,
+      "loss": 2.2023,
+      "step": 231410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016558829397371523,
+      "loss": 2.1264,
+      "step": 231415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001655868987563773,
+      "loss": 2.0369,
+      "step": 231420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016558550351663374,
+      "loss": 2.0756,
+      "step": 231425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016558410825448494,
+      "loss": 2.067,
+      "step": 231430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016558271296993144,
+      "loss": 1.7041,
+      "step": 231435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016558131766297368,
+      "loss": 2.1518,
+      "step": 231440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016557992233361216,
+      "loss": 2.0177,
+      "step": 231445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016557852698184737,
+      "loss": 1.8963,
+      "step": 231450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016557713160767973,
+      "loss": 2.2765,
+      "step": 231455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001655757362111098,
+      "loss": 2.1112,
+      "step": 231460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016557434079213796,
+      "loss": 2.1047,
+      "step": 231465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016557294535076477,
+      "loss": 1.9901,
+      "step": 231470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016557154988699067,
+      "loss": 2.07,
+      "step": 231475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001655701544008161,
+      "loss": 2.1513,
+      "step": 231480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001655687588922416,
+      "loss": 1.925,
+      "step": 231485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016556736336126762,
+      "loss": 2.0251,
+      "step": 231490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016556596780789468,
+      "loss": 1.9056,
+      "step": 231495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016556457223212317,
+      "loss": 2.124,
+      "step": 231500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001655631766339536,
+      "loss": 2.2156,
+      "step": 231505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016556178101338653,
+      "loss": 1.8693,
+      "step": 231510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016556038537042232,
+      "loss": 2.2195,
+      "step": 231515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016555898970506148,
+      "loss": 2.1219,
+      "step": 231520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016555759401730452,
+      "loss": 1.9436,
+      "step": 231525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001655561983071519,
+      "loss": 2.0657,
+      "step": 231530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001655548025746041,
+      "loss": 1.9584,
+      "step": 231535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016555340681966157,
+      "loss": 2.1742,
+      "step": 231540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001655520110423248,
+      "loss": 2.1539,
+      "step": 231545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001655506152425943,
+      "loss": 2.0053,
+      "step": 231550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001655492194204705,
+      "loss": 2.0094,
+      "step": 231555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001655478235759539,
+      "loss": 2.0493,
+      "step": 231560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000165546427709045,
+      "loss": 2.0742,
+      "step": 231565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016554503181974423,
+      "loss": 2.2378,
+      "step": 231570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016554363590805208,
+      "loss": 2.0973,
+      "step": 231575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016554223997396905,
+      "loss": 1.9912,
+      "step": 231580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016554084401749561,
+      "loss": 2.1372,
+      "step": 231585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001655394480386322,
+      "loss": 2.0426,
+      "step": 231590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016553805203737936,
+      "loss": 2.0472,
+      "step": 231595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001655366560137375,
+      "loss": 2.2184,
+      "step": 231600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016553525996770719,
+      "loss": 2.1399,
+      "step": 231605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016553386389928876,
+      "loss": 2.0991,
+      "step": 231610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016553246780848284,
+      "loss": 2.0929,
+      "step": 231615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001655310716952898,
+      "loss": 2.076,
+      "step": 231620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001655296755597102,
+      "loss": 2.0791,
+      "step": 231625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016552827940174443,
+      "loss": 2.2058,
+      "step": 231630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016552688322139304,
+      "loss": 2.1219,
+      "step": 231635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016552548701865647,
+      "loss": 2.1986,
+      "step": 231640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001655240907935352,
+      "loss": 2.1006,
+      "step": 231645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016552269454602974,
+      "loss": 2.1103,
+      "step": 231650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016552129827614048,
+      "loss": 2.0396,
+      "step": 231655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016551990198386803,
+      "loss": 2.0743,
+      "step": 231660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016551850566921274,
+      "loss": 2.0783,
+      "step": 231665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016551710933217516,
+      "loss": 2.0427,
+      "step": 231670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016551571297275573,
+      "loss": 2.2275,
+      "step": 231675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016551431659095496,
+      "loss": 2.06,
+      "step": 231680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016551292018677333,
+      "loss": 2.202,
+      "step": 231685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016551152376021125,
+      "loss": 1.9766,
+      "step": 231690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016551012731126928,
+      "loss": 2.0314,
+      "step": 231695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016550873083994783,
+      "loss": 2.1667,
+      "step": 231700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00016550733434624748,
+      "loss": 2.1149,
+      "step": 231705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016550593783016857,
+      "loss": 2.0668,
+      "step": 231710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016550454129171167,
+      "loss": 1.9384,
+      "step": 231715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016550314473087722,
+      "loss": 2.0404,
+      "step": 231720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016550174814766573,
+      "loss": 1.983,
+      "step": 231725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001655003515420776,
+      "loss": 2.1151,
+      "step": 231730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016549895491411342,
+      "loss": 2.2115,
+      "step": 231735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016549755826377358,
+      "loss": 2.2424,
+      "step": 231740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001654961615910586,
+      "loss": 2.1554,
+      "step": 231745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016549476489596896,
+      "loss": 1.9161,
+      "step": 231750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016549336817850508,
+      "loss": 2.2215,
+      "step": 231755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001654919714386675,
+      "loss": 2.0766,
+      "step": 231760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001654905746764567,
+      "loss": 1.9679,
+      "step": 231765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016548917789187308,
+      "loss": 2.1633,
+      "step": 231770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016548778108491717,
+      "loss": 2.2363,
+      "step": 231775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016548638425558949,
+      "loss": 2.0852,
+      "step": 231780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016548498740389045,
+      "loss": 2.1458,
+      "step": 231785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016548359052982056,
+      "loss": 2.0355,
+      "step": 231790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016548219363338027,
+      "loss": 2.0516,
+      "step": 231795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001654807967145701,
+      "loss": 2.1348,
+      "step": 231800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016547939977339048,
+      "loss": 2.0502,
+      "step": 231805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016547800280984195,
+      "loss": 2.2258,
+      "step": 231810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001654766058239249,
+      "loss": 2.0153,
+      "step": 231815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016547520881563988,
+      "loss": 2.2169,
+      "step": 231820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016547381178498731,
+      "loss": 2.136,
+      "step": 231825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016547241473196774,
+      "loss": 1.9558,
+      "step": 231830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016547101765658156,
+      "loss": 2.0232,
+      "step": 231835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001654696205588293,
+      "loss": 2.1319,
+      "step": 231840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016546822343871146,
+      "loss": 2.0436,
+      "step": 231845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016546682629622848,
+      "loss": 2.1472,
+      "step": 231850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016546542913138082,
+      "loss": 2.1092,
+      "step": 231855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016546403194416903,
+      "loss": 2.001,
+      "step": 231860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001654626347345935,
+      "loss": 2.1604,
+      "step": 231865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016546123750265477,
+      "loss": 1.9533,
+      "step": 231870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016545984024835325,
+      "loss": 2.0826,
+      "step": 231875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016545844297168952,
+      "loss": 2.1829,
+      "step": 231880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016545704567266399,
+      "loss": 1.9949,
+      "step": 231885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001654556483512771,
+      "loss": 2.0716,
+      "step": 231890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001654542510075294,
+      "loss": 2.1396,
+      "step": 231895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016545285364142134,
+      "loss": 1.869,
+      "step": 231900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016545145625295339,
+      "loss": 2.2132,
+      "step": 231905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001654500588421261,
+      "loss": 2.1023,
+      "step": 231910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.46875,
+      "learning_rate": 0.0001654486614089398,
+      "loss": 2.2409,
+      "step": 231915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001654472639533951,
+      "loss": 2.2479,
+      "step": 231920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001654458664754924,
+      "loss": 2.2633,
+      "step": 231925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016544446897523218,
+      "loss": 1.9934,
+      "step": 231930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.000165443071452615,
+      "loss": 2.1328,
+      "step": 231935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016544167390764126,
+      "loss": 2.1446,
+      "step": 231940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016544027634031143,
+      "loss": 2.2955,
+      "step": 231945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016543887875062601,
+      "loss": 2.0341,
+      "step": 231950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016543748113858554,
+      "loss": 1.985,
+      "step": 231955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016543608350419039,
+      "loss": 2.2793,
+      "step": 231960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001654346858474411,
+      "loss": 2.0976,
+      "step": 231965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016543328816833816,
+      "loss": 2.2435,
+      "step": 231970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.000165431890466882,
+      "loss": 2.0337,
+      "step": 231975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016543049274307312,
+      "loss": 2.1871,
+      "step": 231980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016542909499691198,
+      "loss": 2.2252,
+      "step": 231985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001654276972283991,
+      "loss": 2.2484,
+      "step": 231990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016542629943753495,
+      "loss": 2.048,
+      "step": 231995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016542490162431997,
+      "loss": 2.1458,
+      "step": 232000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016542350378875463,
+      "loss": 1.9235,
+      "step": 232005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016542210593083946,
+      "loss": 2.0683,
+      "step": 232010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016542070805057493,
+      "loss": 1.9768,
+      "step": 232015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001654193101479615,
+      "loss": 2.2824,
+      "step": 232020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016541791222299962,
+      "loss": 2.0524,
+      "step": 232025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016541651427568979,
+      "loss": 2.1142,
+      "step": 232030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001654151163060325,
+      "loss": 2.1193,
+      "step": 232035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001654137183140282,
+      "loss": 2.0449,
+      "step": 232040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016541232029967741,
+      "loss": 1.9248,
+      "step": 232045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001654109222629806,
+      "loss": 2.0319,
+      "step": 232050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001654095242039382,
+      "loss": 2.3285,
+      "step": 232055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016540812612255077,
+      "loss": 2.0882,
+      "step": 232060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001654067280188187,
+      "loss": 2.0596,
+      "step": 232065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001654053298927425,
+      "loss": 2.232,
+      "step": 232070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016540393174432265,
+      "loss": 1.9571,
+      "step": 232075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016540253357355967,
+      "loss": 2.0577,
+      "step": 232080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016540113538045397,
+      "loss": 2.0401,
+      "step": 232085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016539973716500606,
+      "loss": 2.0786,
+      "step": 232090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001653983389272164,
+      "loss": 2.0491,
+      "step": 232095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016539694066708548,
+      "loss": 2.1468,
+      "step": 232100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016539554238461382,
+      "loss": 2.0132,
+      "step": 232105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016539414407980182,
+      "loss": 2.232,
+      "step": 232110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016539274575265,
+      "loss": 2.0263,
+      "step": 232115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016539134740315884,
+      "loss": 2.0882,
+      "step": 232120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001653899490313288,
+      "loss": 2.239,
+      "step": 232125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016538855063716036,
+      "loss": 1.9238,
+      "step": 232130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016538715222065404,
+      "loss": 2.1495,
+      "step": 232135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016538575378181026,
+      "loss": 2.1886,
+      "step": 232140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016538435532062953,
+      "loss": 2.0751,
+      "step": 232145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.0001653829568371123,
+      "loss": 2.0239,
+      "step": 232150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016538155833125908,
+      "loss": 2.2341,
+      "step": 232155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016538015980307035,
+      "loss": 2.0801,
+      "step": 232160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016537876125254656,
+      "loss": 2.0953,
+      "step": 232165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016537736267968817,
+      "loss": 2.1348,
+      "step": 232170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016537596408449572,
+      "loss": 1.9867,
+      "step": 232175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016537456546696967,
+      "loss": 2.1703,
+      "step": 232180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016537316682711044,
+      "loss": 2.198,
+      "step": 232185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016537176816491854,
+      "loss": 2.1201,
+      "step": 232190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016537036948039452,
+      "loss": 1.9501,
+      "step": 232195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016536897077353875,
+      "loss": 2.1865,
+      "step": 232200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016536757204435175,
+      "loss": 2.0922,
+      "step": 232205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016536617329283404,
+      "loss": 2.0269,
+      "step": 232210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016536477451898604,
+      "loss": 2.1552,
+      "step": 232215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016536337572280825,
+      "loss": 2.1698,
+      "step": 232220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016536197690430112,
+      "loss": 2.0373,
+      "step": 232225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016536057806346518,
+      "loss": 1.9526,
+      "step": 232230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016535917920030085,
+      "loss": 2.233,
+      "step": 232235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001653577803148087,
+      "loss": 2.0929,
+      "step": 232240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016535638140698913,
+      "loss": 2.0084,
+      "step": 232245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001653549824768426,
+      "loss": 2.1359,
+      "step": 232250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016535358352436964,
+      "loss": 2.0151,
+      "step": 232255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001653521845495707,
+      "loss": 2.061,
+      "step": 232260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001653507855524463,
+      "loss": 2.0571,
+      "step": 232265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016534938653299685,
+      "loss": 2.1856,
+      "step": 232270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001653479874912229,
+      "loss": 1.8412,
+      "step": 232275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016534658842712484,
+      "loss": 2.2326,
+      "step": 232280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016534518934070326,
+      "loss": 2.1477,
+      "step": 232285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016534379023195852,
+      "loss": 2.1571,
+      "step": 232290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001653423911008912,
+      "loss": 2.187,
+      "step": 232295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016534099194750172,
+      "loss": 2.3593,
+      "step": 232300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00016533959277179058,
+      "loss": 2.2122,
+      "step": 232305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016533819357375824,
+      "loss": 2.1948,
+      "step": 232310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001653367943534052,
+      "loss": 2.0074,
+      "step": 232315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001653353951107319,
+      "loss": 2.0132,
+      "step": 232320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016533399584573886,
+      "loss": 2.0223,
+      "step": 232325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016533259655842654,
+      "loss": 1.9376,
+      "step": 232330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016533119724879543,
+      "loss": 2.0486,
+      "step": 232335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000165329797916846,
+      "loss": 2.1291,
+      "step": 232340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001653283985625787,
+      "loss": 2.0671,
+      "step": 232345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016532699918599407,
+      "loss": 2.1151,
+      "step": 232350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016532559978709252,
+      "loss": 2.2415,
+      "step": 232355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016532420036587459,
+      "loss": 2.0811,
+      "step": 232360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016532280092234068,
+      "loss": 2.0335,
+      "step": 232365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016532140145649138,
+      "loss": 1.8834,
+      "step": 232370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016532000196832705,
+      "loss": 2.2635,
+      "step": 232375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00016531860245784828,
+      "loss": 1.9189,
+      "step": 232380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016531720292505545,
+      "loss": 2.1357,
+      "step": 232385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016531580336994907,
+      "loss": 2.2638,
+      "step": 232390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00016531440379252964,
+      "loss": 2.0607,
+      "step": 232395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016531300419279766,
+      "loss": 2.1292,
+      "step": 232400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016531160457075353,
+      "loss": 2.0327,
+      "step": 232405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001653102049263978,
+      "loss": 2.0262,
+      "step": 232410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016530880525973088,
+      "loss": 2.0712,
+      "step": 232415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001653074055707533,
+      "loss": 2.0773,
+      "step": 232420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016530600585946555,
+      "loss": 2.1794,
+      "step": 232425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001653046061258681,
+      "loss": 2.2909,
+      "step": 232430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016530320636996138,
+      "loss": 2.0971,
+      "step": 232435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001653018065917459,
+      "loss": 2.0102,
+      "step": 232440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001653004067912221,
+      "loss": 2.2825,
+      "step": 232445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016529900696839056,
+      "loss": 1.9564,
+      "step": 232450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001652976071232517,
+      "loss": 2.1608,
+      "step": 232455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016529620725580594,
+      "loss": 1.9923,
+      "step": 232460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016529480736605386,
+      "loss": 2.1307,
+      "step": 232465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016529340745399584,
+      "loss": 2.1731,
+      "step": 232470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016529200751963242,
+      "loss": 2.0467,
+      "step": 232475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016529060756296412,
+      "loss": 2.2221,
+      "step": 232480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001652892075839913,
+      "loss": 2.1802,
+      "step": 232485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016528780758271453,
+      "loss": 2.1723,
+      "step": 232490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016528640755913426,
+      "loss": 2.039,
+      "step": 232495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016528500751325097,
+      "loss": 2.239,
+      "step": 232500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016528360744506515,
+      "loss": 2.1492,
+      "step": 232505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016528220735457727,
+      "loss": 2.033,
+      "step": 232510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016528080724178778,
+      "loss": 1.9669,
+      "step": 232515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016527940710669715,
+      "loss": 2.0751,
+      "step": 232520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016527800694930594,
+      "loss": 2.2327,
+      "step": 232525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001652766067696146,
+      "loss": 2.0472,
+      "step": 232530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016527520656762354,
+      "loss": 2.0317,
+      "step": 232535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001652738063433333,
+      "loss": 2.0916,
+      "step": 232540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016527240609674434,
+      "loss": 2.157,
+      "step": 232545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016527100582785711,
+      "loss": 2.1879,
+      "step": 232550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001652696055366722,
+      "loss": 2.2779,
+      "step": 232555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016526820522318997,
+      "loss": 2.0138,
+      "step": 232560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001652668048874109,
+      "loss": 2.0207,
+      "step": 232565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016526540452933556,
+      "loss": 2.0764,
+      "step": 232570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016526400414896436,
+      "loss": 2.0453,
+      "step": 232575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016526260374629778,
+      "loss": 2.0638,
+      "step": 232580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016526120332133632,
+      "loss": 2.0392,
+      "step": 232585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016525980287408046,
+      "loss": 2.1259,
+      "step": 232590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016525840240453065,
+      "loss": 2.1302,
+      "step": 232595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016525700191268738,
+      "loss": 2.0239,
+      "step": 232600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016525560139855113,
+      "loss": 1.9505,
+      "step": 232605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001652542008621224,
+      "loss": 2.1406,
+      "step": 232610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016525280030340166,
+      "loss": 2.0767,
+      "step": 232615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016525139972238938,
+      "loss": 2.0487,
+      "step": 232620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016524999911908604,
+      "loss": 2.0604,
+      "step": 232625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016524859849349207,
+      "loss": 2.047,
+      "step": 232630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016524719784560805,
+      "loss": 1.9716,
+      "step": 232635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016524579717543437,
+      "loss": 2.055,
+      "step": 232640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016524439648297155,
+      "loss": 2.2031,
+      "step": 232645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001652429957682201,
+      "loss": 2.1028,
+      "step": 232650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016524159503118041,
+      "loss": 2.2209,
+      "step": 232655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000165240194271853,
+      "loss": 2.08,
+      "step": 232660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016523879349023842,
+      "loss": 2.0371,
+      "step": 232665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016523739268633705,
+      "loss": 2.1089,
+      "step": 232670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001652359918601494,
+      "loss": 2.109,
+      "step": 232675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016523459101167593,
+      "loss": 2.0806,
+      "step": 232680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016523319014091715,
+      "loss": 2.0011,
+      "step": 232685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016523178924787358,
+      "loss": 1.9548,
+      "step": 232690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001652303883325456,
+      "loss": 2.1016,
+      "step": 232695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016522898739493376,
+      "loss": 2.0826,
+      "step": 232700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016522758643503846,
+      "loss": 2.146,
+      "step": 232705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016522618545286027,
+      "loss": 2.0576,
+      "step": 232710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016522478444839966,
+      "loss": 2.0112,
+      "step": 232715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016522338342165707,
+      "loss": 2.2248,
+      "step": 232720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016522198237263298,
+      "loss": 2.1405,
+      "step": 232725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016522058130132787,
+      "loss": 1.993,
+      "step": 232730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016521918020774223,
+      "loss": 2.1973,
+      "step": 232735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016521777909187652,
+      "loss": 2.0923,
+      "step": 232740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016521637795373126,
+      "loss": 2.1143,
+      "step": 232745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001652149767933069,
+      "loss": 2.0514,
+      "step": 232750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001652135756106039,
+      "loss": 2.1225,
+      "step": 232755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001652121744056228,
+      "loss": 1.9602,
+      "step": 232760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000165210773178364,
+      "loss": 2.262,
+      "step": 232765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016520937192882805,
+      "loss": 2.0431,
+      "step": 232770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016520797065701536,
+      "loss": 2.0259,
+      "step": 232775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016520656936292648,
+      "loss": 2.1066,
+      "step": 232780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016520516804656177,
+      "loss": 1.9432,
+      "step": 232785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001652037667079219,
+      "loss": 1.9836,
+      "step": 232790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016520236534700719,
+      "loss": 1.9907,
+      "step": 232795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016520096396381817,
+      "loss": 2.0784,
+      "step": 232800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016519956255835533,
+      "loss": 2.0577,
+      "step": 232805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016519816113061913,
+      "loss": 2.1327,
+      "step": 232810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016519675968061005,
+      "loss": 2.3037,
+      "step": 232815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001651953582083286,
+      "loss": 2.1993,
+      "step": 232820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001651939567137752,
+      "loss": 2.0868,
+      "step": 232825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016519255519695038,
+      "loss": 2.1073,
+      "step": 232830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001651911536578546,
+      "loss": 2.1266,
+      "step": 232835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016518975209648833,
+      "loss": 2.1755,
+      "step": 232840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016518835051285206,
+      "loss": 2.1221,
+      "step": 232845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001651869489069463,
+      "loss": 2.2773,
+      "step": 232850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016518554727877144,
+      "loss": 2.1878,
+      "step": 232855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016518414562832808,
+      "loss": 2.111,
+      "step": 232860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016518274395561656,
+      "loss": 2.0702,
+      "step": 232865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016518134226063748,
+      "loss": 1.8593,
+      "step": 232870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016517994054339127,
+      "loss": 2.0467,
+      "step": 232875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001651785388038784,
+      "loss": 1.9863,
+      "step": 232880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016517713704209937,
+      "loss": 2.1592,
+      "step": 232885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016517573525805463,
+      "loss": 1.897,
+      "step": 232890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016517433345174468,
+      "loss": 2.0617,
+      "step": 232895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016517293162317,
+      "loss": 2.0007,
+      "step": 232900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016517152977233107,
+      "loss": 2.202,
+      "step": 232905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016517012789922836,
+      "loss": 2.1832,
+      "step": 232910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016516872600386235,
+      "loss": 1.9518,
+      "step": 232915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016516732408623352,
+      "loss": 2.2086,
+      "step": 232920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016516592214634235,
+      "loss": 2.019,
+      "step": 232925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016516452018418932,
+      "loss": 2.1563,
+      "step": 232930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001651631181997749,
+      "loss": 1.9789,
+      "step": 232935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001651617161930996,
+      "loss": 2.0028,
+      "step": 232940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016516031416416387,
+      "loss": 2.0151,
+      "step": 232945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016515891211296818,
+      "loss": 2.024,
+      "step": 232950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016515751003951304,
+      "loss": 2.0662,
+      "step": 232955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016515610794379888,
+      "loss": 1.9431,
+      "step": 232960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016515470582582626,
+      "loss": 2.1703,
+      "step": 232965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016515330368559557,
+      "loss": 2.0313,
+      "step": 232970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016515190152310735,
+      "loss": 2.11,
+      "step": 232975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016515049933836203,
+      "loss": 1.9752,
+      "step": 232980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016514909713136018,
+      "loss": 2.1361,
+      "step": 232985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016514769490210218,
+      "loss": 1.8895,
+      "step": 232990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016514629265058854,
+      "loss": 1.8554,
+      "step": 232995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016514489037681976,
+      "loss": 2.1442,
+      "step": 233000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016514348808079627,
+      "loss": 2.0958,
+      "step": 233005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016514208576251863,
+      "loss": 2.1756,
+      "step": 233010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016514068342198725,
+      "loss": 2.0949,
+      "step": 233015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016513928105920265,
+      "loss": 2.0382,
+      "step": 233020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016513787867416526,
+      "loss": 2.1112,
+      "step": 233025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016513647626687557,
+      "loss": 2.1345,
+      "step": 233030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016513507383733413,
+      "loss": 2.174,
+      "step": 233035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016513367138554135,
+      "loss": 2.041,
+      "step": 233040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016513226891149774,
+      "loss": 2.1603,
+      "step": 233045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016513086641520376,
+      "loss": 2.237,
+      "step": 233050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016512946389665985,
+      "loss": 2.0248,
+      "step": 233055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016512806135586655,
+      "loss": 2.1823,
+      "step": 233060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00016512665879282436,
+      "loss": 2.196,
+      "step": 233065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00016512525620753372,
+      "loss": 2.0137,
+      "step": 233070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001651238535999951,
+      "loss": 2.1008,
+      "step": 233075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016512245097020895,
+      "loss": 1.9883,
+      "step": 233080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00016512104831817586,
+      "loss": 1.8592,
+      "step": 233085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001651196456438962,
+      "loss": 1.9847,
+      "step": 233090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001651182429473705,
+      "loss": 2.4037,
+      "step": 233095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016511684022859922,
+      "loss": 2.0581,
+      "step": 233100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016511543748758283,
+      "loss": 1.9971,
+      "step": 233105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016511403472432187,
+      "loss": 2.0328,
+      "step": 233110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016511263193881673,
+      "loss": 2.0173,
+      "step": 233115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016511122913106795,
+      "loss": 2.1629,
+      "step": 233120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000165109826301076,
+      "loss": 2.28,
+      "step": 233125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016510842344884133,
+      "loss": 2.0908,
+      "step": 233130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016510702057436445,
+      "loss": 2.0508,
+      "step": 233135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016510561767764585,
+      "loss": 2.1005,
+      "step": 233140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016510421475868597,
+      "loss": 2.0586,
+      "step": 233145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016510281181748534,
+      "loss": 2.1613,
+      "step": 233150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016510140885404436,
+      "loss": 2.1285,
+      "step": 233155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016510000586836358,
+      "loss": 1.907,
+      "step": 233160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016509860286044348,
+      "loss": 2.1145,
+      "step": 233165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016509719983028452,
+      "loss": 2.0377,
+      "step": 233170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016509579677788714,
+      "loss": 2.1653,
+      "step": 233175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016509439370325187,
+      "loss": 2.1218,
+      "step": 233180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016509299060637916,
+      "loss": 2.1073,
+      "step": 233185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016509158748726952,
+      "loss": 2.0932,
+      "step": 233190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016509018434592343,
+      "loss": 2.171,
+      "step": 233195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001650887811823413,
+      "loss": 2.1132,
+      "step": 233200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016508737799652372,
+      "loss": 2.209,
+      "step": 233205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001650859747884711,
+      "loss": 2.1452,
+      "step": 233210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001650845715581839,
+      "loss": 2.1188,
+      "step": 233215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016508316830566265,
+      "loss": 2.0617,
+      "step": 233220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016508176503090782,
+      "loss": 2.129,
+      "step": 233225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016508036173391988,
+      "loss": 2.0359,
+      "step": 233230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016507895841469928,
+      "loss": 2.2095,
+      "step": 233235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016507755507324653,
+      "loss": 2.2402,
+      "step": 233240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016507615170956213,
+      "loss": 2.2108,
+      "step": 233245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001650747483236465,
+      "loss": 2.1602,
+      "step": 233250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001650733449155002,
+      "loss": 1.9558,
+      "step": 233255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016507194148512367,
+      "loss": 2.03,
+      "step": 233260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016507053803251735,
+      "loss": 2.1465,
+      "step": 233265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016506913455768176,
+      "loss": 2.1071,
+      "step": 233270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016506773106061737,
+      "loss": 2.0231,
+      "step": 233275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00016506632754132466,
+      "loss": 2.0077,
+      "step": 233280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016506492399980416,
+      "loss": 2.1344,
+      "step": 233285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016506352043605627,
+      "loss": 2.1512,
+      "step": 233290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016506211685008148,
+      "loss": 2.2609,
+      "step": 233295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001650607132418803,
+      "loss": 2.013,
+      "step": 233300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001650593096114532,
+      "loss": 2.0685,
+      "step": 233305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016505790595880068,
+      "loss": 2.1622,
+      "step": 233310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001650565022839232,
+      "loss": 2.0418,
+      "step": 233315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016505509858682124,
+      "loss": 2.0989,
+      "step": 233320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016505369486749524,
+      "loss": 2.114,
+      "step": 233325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016505229112594573,
+      "loss": 1.9204,
+      "step": 233330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016505088736217322,
+      "loss": 2.281,
+      "step": 233335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001650494835761781,
+      "loss": 1.9901,
+      "step": 233340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016504807976796092,
+      "loss": 2.0953,
+      "step": 233345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016504667593752212,
+      "loss": 2.241,
+      "step": 233350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001650452720848622,
+      "loss": 2.1136,
+      "step": 233355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016504386820998166,
+      "loss": 1.8998,
+      "step": 233360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016504246431288096,
+      "loss": 2.4034,
+      "step": 233365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016504106039356056,
+      "loss": 2.1821,
+      "step": 233370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001650396564520209,
+      "loss": 2.2526,
+      "step": 233375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016503825248826258,
+      "loss": 2.0471,
+      "step": 233380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.000165036848502286,
+      "loss": 2.1426,
+      "step": 233385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016503544449409163,
+      "loss": 2.2541,
+      "step": 233390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016503404046368,
+      "loss": 2.1365,
+      "step": 233395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016503263641105159,
+      "loss": 2.1806,
+      "step": 233400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016503123233620678,
+      "loss": 1.9799,
+      "step": 233405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016502982823914618,
+      "loss": 2.0885,
+      "step": 233410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016502842411987017,
+      "loss": 2.1115,
+      "step": 233415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001650270199783793,
+      "loss": 2.1237,
+      "step": 233420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.000165025615814674,
+      "loss": 2.1533,
+      "step": 233425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016502421162875477,
+      "loss": 2.0937,
+      "step": 233430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001650228074206221,
+      "loss": 2.0888,
+      "step": 233435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016502140319027644,
+      "loss": 2.1814,
+      "step": 233440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016501999893771832,
+      "loss": 2.1741,
+      "step": 233445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016501859466294818,
+      "loss": 2.1742,
+      "step": 233450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001650171903659665,
+      "loss": 1.8904,
+      "step": 233455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016501578604677378,
+      "loss": 1.9874,
+      "step": 233460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016501438170537047,
+      "loss": 2.1102,
+      "step": 233465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016501297734175708,
+      "loss": 2.2044,
+      "step": 233470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001650115729559341,
+      "loss": 1.9763,
+      "step": 233475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00016501016854790197,
+      "loss": 2.1018,
+      "step": 233480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016500876411766114,
+      "loss": 2.0304,
+      "step": 233485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016500735966521222,
+      "loss": 2.1682,
+      "step": 233490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016500595519055558,
+      "loss": 2.0684,
+      "step": 233495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001650045506936917,
+      "loss": 2.0575,
+      "step": 233500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001650031461746211,
+      "loss": 2.0366,
+      "step": 233505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016500174163334424,
+      "loss": 1.8696,
+      "step": 233510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016500033706986163,
+      "loss": 2.2472,
+      "step": 233515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001649989324841737,
+      "loss": 2.2168,
+      "step": 233520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016499752787628097,
+      "loss": 1.992,
+      "step": 233525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001649961232461839,
+      "loss": 2.1523,
+      "step": 233530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016499471859388298,
+      "loss": 2.1304,
+      "step": 233535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016499331391937866,
+      "loss": 2.1488,
+      "step": 233540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00016499190922267148,
+      "loss": 2.1881,
+      "step": 233545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016499050450376188,
+      "loss": 2.0661,
+      "step": 233550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016498909976265033,
+      "loss": 2.1829,
+      "step": 233555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016498769499933733,
+      "loss": 1.9386,
+      "step": 233560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016498629021382337,
+      "loss": 2.0296,
+      "step": 233565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001649848854061089,
+      "loss": 2.1206,
+      "step": 233570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016498348057619443,
+      "loss": 2.0688,
+      "step": 233575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001649820757240804,
+      "loss": 1.8445,
+      "step": 233580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016498067084976732,
+      "loss": 2.2454,
+      "step": 233585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016497926595325567,
+      "loss": 2.0528,
+      "step": 233590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001649778610345459,
+      "loss": 2.1995,
+      "step": 233595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016497645609363855,
+      "loss": 1.8481,
+      "step": 233600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016497505113053406,
+      "loss": 2.0463,
+      "step": 233605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016497364614523292,
+      "loss": 2.0035,
+      "step": 233610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016497224113773559,
+      "loss": 2.1527,
+      "step": 233615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016497083610804257,
+      "loss": 2.0843,
+      "step": 233620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001649694310561543,
+      "loss": 1.9801,
+      "step": 233625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016496802598207135,
+      "loss": 2.3273,
+      "step": 233630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016496662088579412,
+      "loss": 2.1256,
+      "step": 233635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001649652157673231,
+      "loss": 2.0212,
+      "step": 233640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001649638106266588,
+      "loss": 2.1791,
+      "step": 233645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016496240546380166,
+      "loss": 2.1996,
+      "step": 233650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016496100027875224,
+      "loss": 2.1564,
+      "step": 233655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001649595950715109,
+      "loss": 2.0632,
+      "step": 233660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016495818984207822,
+      "loss": 2.1771,
+      "step": 233665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016495678459045463,
+      "loss": 2.1149,
+      "step": 233670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016495537931664062,
+      "loss": 2.0362,
+      "step": 233675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016495397402063668,
+      "loss": 2.0523,
+      "step": 233680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016495256870244332,
+      "loss": 2.0504,
+      "step": 233685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016495116336206093,
+      "loss": 2.0266,
+      "step": 233690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001649497579994901,
+      "loss": 2.0957,
+      "step": 233695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016494835261473122,
+      "loss": 1.9599,
+      "step": 233700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001649469472077848,
+      "loss": 2.236,
+      "step": 233705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016494554177865132,
+      "loss": 2.1724,
+      "step": 233710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016494413632733127,
+      "loss": 2.0724,
+      "step": 233715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016494273085382514,
+      "loss": 2.1014,
+      "step": 233720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016494132535813338,
+      "loss": 1.8836,
+      "step": 233725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001649399198402565,
+      "loss": 1.9889,
+      "step": 233730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016493851430019496,
+      "loss": 2.2088,
+      "step": 233735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016493710873794926,
+      "loss": 2.1077,
+      "step": 233740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016493570315351984,
+      "loss": 2.0231,
+      "step": 233745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016493429754690721,
+      "loss": 2.1419,
+      "step": 233750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016493289191811186,
+      "loss": 2.0265,
+      "step": 233755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016493148626713425,
+      "loss": 2.1689,
+      "step": 233760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001649300805939749,
+      "loss": 2.0388,
+      "step": 233765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001649286748986342,
+      "loss": 1.96,
+      "step": 233770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016492726918111272,
+      "loss": 2.208,
+      "step": 233775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001649258634414109,
+      "loss": 1.9881,
+      "step": 233780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001649244576795292,
+      "loss": 1.883,
+      "step": 233785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016492305189546818,
+      "loss": 2.1963,
+      "step": 233790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001649216460892282,
+      "loss": 2.0688,
+      "step": 233795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016492024026080988,
+      "loss": 2.1087,
+      "step": 233800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001649188344102136,
+      "loss": 1.9916,
+      "step": 233805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016491742853743987,
+      "loss": 2.1374,
+      "step": 233810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016491602264248915,
+      "loss": 2.1051,
+      "step": 233815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016491461672536197,
+      "loss": 2.0564,
+      "step": 233820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016491321078605876,
+      "loss": 2.3015,
+      "step": 233825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016491180482458003,
+      "loss": 2.1141,
+      "step": 233830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016491039884092626,
+      "loss": 1.9501,
+      "step": 233835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001649089928350979,
+      "loss": 2.1356,
+      "step": 233840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016490758680709545,
+      "loss": 1.9673,
+      "step": 233845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001649061807569194,
+      "loss": 1.9838,
+      "step": 233850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016490477468457021,
+      "loss": 2.2289,
+      "step": 233855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001649033685900484,
+      "loss": 2.0173,
+      "step": 233860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016490196247335442,
+      "loss": 2.0326,
+      "step": 233865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001649005563344887,
+      "loss": 2.1424,
+      "step": 233870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001648991501734518,
+      "loss": 2.133,
+      "step": 233875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001648977439902442,
+      "loss": 2.0656,
+      "step": 233880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016489633778486634,
+      "loss": 2.082,
+      "step": 233885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001648949315573187,
+      "loss": 1.9641,
+      "step": 233890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001648935253076018,
+      "loss": 2.0453,
+      "step": 233895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016489211903571608,
+      "loss": 2.1168,
+      "step": 233900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016489071274166202,
+      "loss": 2.0667,
+      "step": 233905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016488930642544013,
+      "loss": 2.1095,
+      "step": 233910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001648879000870509,
+      "loss": 2.1114,
+      "step": 233915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016488649372649477,
+      "loss": 2.1265,
+      "step": 233920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.6875,
+      "learning_rate": 0.00016488508734377222,
+      "loss": 2.2611,
+      "step": 233925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016488368093888376,
+      "loss": 2.0471,
+      "step": 233930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016488227451182984,
+      "loss": 1.9309,
+      "step": 233935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016488086806261098,
+      "loss": 1.9514,
+      "step": 233940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016487946159122763,
+      "loss": 2.0431,
+      "step": 233945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016487805509768027,
+      "loss": 2.1396,
+      "step": 233950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001648766485819694,
+      "loss": 2.0278,
+      "step": 233955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001648752420440955,
+      "loss": 2.0829,
+      "step": 233960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016487383548405903,
+      "loss": 2.1368,
+      "step": 233965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016487242890186047,
+      "loss": 2.1113,
+      "step": 233970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016487102229750035,
+      "loss": 2.2007,
+      "step": 233975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001648696156709791,
+      "loss": 2.1226,
+      "step": 233980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001648682090222972,
+      "loss": 2.2397,
+      "step": 233985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016486680235145516,
+      "loss": 2.0074,
+      "step": 233990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016486539565845342,
+      "loss": 2.1443,
+      "step": 233995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001648639889432925,
+      "loss": 1.9022,
+      "step": 234000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016486258220597283,
+      "loss": 2.1789,
+      "step": 234005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016486117544649496,
+      "loss": 2.2281,
+      "step": 234010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016485976866485933,
+      "loss": 2.1792,
+      "step": 234015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016485836186106643,
+      "loss": 2.2278,
+      "step": 234020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001648569550351167,
+      "loss": 2.103,
+      "step": 234025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016485554818701072,
+      "loss": 2.1959,
+      "step": 234030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001648541413167489,
+      "loss": 2.0364,
+      "step": 234035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016485273442433168,
+      "loss": 2.205,
+      "step": 234040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001648513275097596,
+      "loss": 2.1106,
+      "step": 234045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016484992057303317,
+      "loss": 2.0132,
+      "step": 234050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001648485136141528,
+      "loss": 1.9529,
+      "step": 234055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016484710663311903,
+      "loss": 2.1581,
+      "step": 234060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001648456996299323,
+      "loss": 2.1363,
+      "step": 234065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001648442926045931,
+      "loss": 2.0863,
+      "step": 234070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001648428855571019,
+      "loss": 2.0861,
+      "step": 234075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001648414784874592,
+      "loss": 2.0427,
+      "step": 234080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016484007139566548,
+      "loss": 2.0954,
+      "step": 234085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016483866428172123,
+      "loss": 2.2506,
+      "step": 234090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016483725714562688,
+      "loss": 2.1842,
+      "step": 234095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016483584998738297,
+      "loss": 2.1774,
+      "step": 234100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016483444280698996,
+      "loss": 2.1022,
+      "step": 234105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016483303560444834,
+      "loss": 1.9205,
+      "step": 234110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016483162837975857,
+      "loss": 2.2253,
+      "step": 234115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016483022113292112,
+      "loss": 2.0548,
+      "step": 234120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001648288138639365,
+      "loss": 2.1004,
+      "step": 234125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016482740657280517,
+      "loss": 2.1876,
+      "step": 234130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016482599925952767,
+      "loss": 2.2499,
+      "step": 234135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001648245919241044,
+      "loss": 1.9432,
+      "step": 234140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016482318456653586,
+      "loss": 2.1414,
+      "step": 234145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016482177718682257,
+      "loss": 1.9788,
+      "step": 234150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016482036978496498,
+      "loss": 2.0217,
+      "step": 234155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016481896236096355,
+      "loss": 2.173,
+      "step": 234160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016481755491481884,
+      "loss": 2.0516,
+      "step": 234165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016481614744653125,
+      "loss": 2.1219,
+      "step": 234170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016481473995610127,
+      "loss": 1.9318,
+      "step": 234175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001648133324435294,
+      "loss": 1.9758,
+      "step": 234180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016481192490881615,
+      "loss": 2.1473,
+      "step": 234185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016481051735196193,
+      "loss": 2.0009,
+      "step": 234190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001648091097729673,
+      "loss": 2.1123,
+      "step": 234195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001648077021718327,
+      "loss": 2.2495,
+      "step": 234200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001648062945485586,
+      "loss": 1.953,
+      "step": 234205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016480488690314552,
+      "loss": 1.9266,
+      "step": 234210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016480347923559388,
+      "loss": 2.053,
+      "step": 234215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001648020715459042,
+      "loss": 2.0727,
+      "step": 234220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016480066383407698,
+      "loss": 2.2944,
+      "step": 234225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016479925610011264,
+      "loss": 1.9688,
+      "step": 234230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016479784834401174,
+      "loss": 2.0289,
+      "step": 234235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001647964405657747,
+      "loss": 2.0692,
+      "step": 234240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016479503276540201,
+      "loss": 2.4282,
+      "step": 234245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001647936249428942,
+      "loss": 1.9792,
+      "step": 234250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016479221709825168,
+      "loss": 2.1483,
+      "step": 234255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016479080923147496,
+      "loss": 2.1967,
+      "step": 234260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016478940134256453,
+      "loss": 2.2012,
+      "step": 234265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001647879934315209,
+      "loss": 2.1816,
+      "step": 234270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016478658549834446,
+      "loss": 2.202,
+      "step": 234275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001647851775430358,
+      "loss": 2.0543,
+      "step": 234280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001647837695655953,
+      "loss": 2.2018,
+      "step": 234285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016478236156602352,
+      "loss": 2.1855,
+      "step": 234290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001647809535443209,
+      "loss": 2.0628,
+      "step": 234295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016477954550048795,
+      "loss": 2.103,
+      "step": 234300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001647781374345251,
+      "loss": 2.3183,
+      "step": 234305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016477672934643288,
+      "loss": 2.0242,
+      "step": 234310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016477532123621176,
+      "loss": 2.1962,
+      "step": 234315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016477391310386225,
+      "loss": 2.1699,
+      "step": 234320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016477250494938473,
+      "loss": 2.1048,
+      "step": 234325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016477109677277978,
+      "loss": 2.1186,
+      "step": 234330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016476968857404784,
+      "loss": 2.1534,
+      "step": 234335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.0001647682803531894,
+      "loss": 2.0685,
+      "step": 234340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016476687211020496,
+      "loss": 1.9332,
+      "step": 234345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016476546384509497,
+      "loss": 2.137,
+      "step": 234350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016476405555785993,
+      "loss": 2.1296,
+      "step": 234355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001647626472485003,
+      "loss": 2.0169,
+      "step": 234360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016476123891701655,
+      "loss": 2.156,
+      "step": 234365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016475983056340923,
+      "loss": 2.0376,
+      "step": 234370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016475842218767873,
+      "loss": 2.2379,
+      "step": 234375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016475701378982564,
+      "loss": 1.9314,
+      "step": 234380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016475560536985035,
+      "loss": 2.0573,
+      "step": 234385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016475419692775338,
+      "loss": 2.2047,
+      "step": 234390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016475278846353518,
+      "loss": 2.0021,
+      "step": 234395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00016475137997719626,
+      "loss": 1.973,
+      "step": 234400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016474997146873708,
+      "loss": 2.0131,
+      "step": 234405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016474856293815818,
+      "loss": 2.1007,
+      "step": 234410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016474715438545998,
+      "loss": 2.1248,
+      "step": 234415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016474574581064295,
+      "loss": 1.9907,
+      "step": 234420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016474433721370762,
+      "loss": 2.216,
+      "step": 234425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016474292859465446,
+      "loss": 1.9077,
+      "step": 234430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016474151995348396,
+      "loss": 2.0811,
+      "step": 234435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016474011129019657,
+      "loss": 2.1329,
+      "step": 234440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016473870260479273,
+      "loss": 1.9828,
+      "step": 234445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000164737293897273,
+      "loss": 2.1455,
+      "step": 234450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016473588516763784,
+      "loss": 2.0135,
+      "step": 234455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016473447641588776,
+      "loss": 2.2734,
+      "step": 234460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001647330676420232,
+      "loss": 2.1474,
+      "step": 234465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016473165884604463,
+      "loss": 2.0642,
+      "step": 234470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016473025002795253,
+      "loss": 2.2637,
+      "step": 234475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016472884118774744,
+      "loss": 2.1919,
+      "step": 234480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001647274323254298,
+      "loss": 2.1927,
+      "step": 234485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001647260234410001,
+      "loss": 2.172,
+      "step": 234490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001647246145344588,
+      "loss": 2.1494,
+      "step": 234495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001647232056058064,
+      "loss": 1.9919,
+      "step": 234500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001647217966550434,
+      "loss": 2.1271,
+      "step": 234505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016472038768217023,
+      "loss": 2.131,
+      "step": 234510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001647189786871874,
+      "loss": 1.9593,
+      "step": 234515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016471756967009543,
+      "loss": 1.9279,
+      "step": 234520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016471616063089473,
+      "loss": 2.1443,
+      "step": 234525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001647147515695858,
+      "loss": 2.0708,
+      "step": 234530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016471334248616918,
+      "loss": 2.2026,
+      "step": 234535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001647119333806453,
+      "loss": 2.1625,
+      "step": 234540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016471052425301466,
+      "loss": 1.9578,
+      "step": 234545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001647091151032777,
+      "loss": 2.1918,
+      "step": 234550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016470770593143494,
+      "loss": 2.1607,
+      "step": 234555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016470629673748688,
+      "loss": 2.1954,
+      "step": 234560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016470488752143395,
+      "loss": 2.1338,
+      "step": 234565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016470347828327666,
+      "loss": 2.1181,
+      "step": 234570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016470206902301552,
+      "loss": 2.2989,
+      "step": 234575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001647006597406509,
+      "loss": 2.2361,
+      "step": 234580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016469925043618343,
+      "loss": 2.1353,
+      "step": 234585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016469784110961352,
+      "loss": 2.0241,
+      "step": 234590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016469643176094163,
+      "loss": 2.0579,
+      "step": 234595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001646950223901683,
+      "loss": 2.1769,
+      "step": 234600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016469361299729395,
+      "loss": 2.1385,
+      "step": 234605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016469220358231908,
+      "loss": 2.1783,
+      "step": 234610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001646907941452442,
+      "loss": 2.1324,
+      "step": 234615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016468938468606976,
+      "loss": 2.257,
+      "step": 234620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016468797520479623,
+      "loss": 1.9573,
+      "step": 234625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016468656570142417,
+      "loss": 2.1908,
+      "step": 234630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016468515617595398,
+      "loss": 2.2722,
+      "step": 234635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016468374662838615,
+      "loss": 2.0145,
+      "step": 234640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001646823370587212,
+      "loss": 2.0866,
+      "step": 234645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016468092746695957,
+      "loss": 1.9876,
+      "step": 234650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016467951785310178,
+      "loss": 1.9964,
+      "step": 234655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001646781082171483,
+      "loss": 1.9901,
+      "step": 234660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016467669855909958,
+      "loss": 2.0842,
+      "step": 234665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016467528887895613,
+      "loss": 2.2358,
+      "step": 234670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016467387917671843,
+      "loss": 2.0546,
+      "step": 234675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.000164672469452387,
+      "loss": 2.0997,
+      "step": 234680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016467105970596223,
+      "loss": 2.0768,
+      "step": 234685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001646696499374447,
+      "loss": 2.0133,
+      "step": 234690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016466824014683482,
+      "loss": 2.0554,
+      "step": 234695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016466683033413307,
+      "loss": 2.0176,
+      "step": 234700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016466542049934,
+      "loss": 2.1334,
+      "step": 234705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016466401064245604,
+      "loss": 2.0273,
+      "step": 234710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016466260076348165,
+      "loss": 2.2297,
+      "step": 234715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.00016466119086241737,
+      "loss": 2.0155,
+      "step": 234720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016465978093926367,
+      "loss": 2.0541,
+      "step": 234725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000164658370994021,
+      "loss": 2.1749,
+      "step": 234730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016465696102668987,
+      "loss": 2.0645,
+      "step": 234735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016465555103727073,
+      "loss": 2.0793,
+      "step": 234740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001646541410257641,
+      "loss": 2.2086,
+      "step": 234745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001646527309921704,
+      "loss": 2.1345,
+      "step": 234750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001646513209364902,
+      "loss": 1.9394,
+      "step": 234755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016464991085872392,
+      "loss": 2.2335,
+      "step": 234760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016464850075887207,
+      "loss": 2.0767,
+      "step": 234765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016464709063693508,
+      "loss": 2.2057,
+      "step": 234770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016464568049291348,
+      "loss": 2.2131,
+      "step": 234775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001646442703268078,
+      "loss": 2.1718,
+      "step": 234780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016464286013861843,
+      "loss": 2.0591,
+      "step": 234785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016464144992834587,
+      "loss": 2.282,
+      "step": 234790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016464003969599063,
+      "loss": 2.0387,
+      "step": 234795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016463862944155318,
+      "loss": 2.0059,
+      "step": 234800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000164637219165034,
+      "loss": 2.1157,
+      "step": 234805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016463580886643357,
+      "loss": 2.1803,
+      "step": 234810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001646343985457524,
+      "loss": 1.9675,
+      "step": 234815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016463298820299093,
+      "loss": 2.1267,
+      "step": 234820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016463157783814966,
+      "loss": 2.2049,
+      "step": 234825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016463016745122904,
+      "loss": 2.1585,
+      "step": 234830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016462875704222966,
+      "loss": 2.22,
+      "step": 234835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016462734661115187,
+      "loss": 2.0757,
+      "step": 234840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001646259361579962,
+      "loss": 2.1368,
+      "step": 234845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016462452568276315,
+      "loss": 2.1878,
+      "step": 234850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001646231151854532,
+      "loss": 2.138,
+      "step": 234855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001646217046660668,
+      "loss": 2.1836,
+      "step": 234860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016462029412460448,
+      "loss": 2.1311,
+      "step": 234865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001646188835610667,
+      "loss": 2.1061,
+      "step": 234870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001646174729754539,
+      "loss": 2.0149,
+      "step": 234875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001646160623677666,
+      "loss": 2.1259,
+      "step": 234880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016461465173800533,
+      "loss": 2.124,
+      "step": 234885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016461324108617048,
+      "loss": 2.0821,
+      "step": 234890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001646118304122626,
+      "loss": 2.1419,
+      "step": 234895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016461041971628214,
+      "loss": 2.196,
+      "step": 234900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016460900899822957,
+      "loss": 2.0488,
+      "step": 234905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001646075982581054,
+      "loss": 2.1104,
+      "step": 234910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016460618749591013,
+      "loss": 2.1559,
+      "step": 234915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001646047767116442,
+      "loss": 1.872,
+      "step": 234920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016460336590530807,
+      "loss": 2.0531,
+      "step": 234925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016460195507690228,
+      "loss": 2.1812,
+      "step": 234930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016460054422642731,
+      "loss": 2.0548,
+      "step": 234935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016459913335388363,
+      "loss": 2.1702,
+      "step": 234940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016459772245927168,
+      "loss": 2.1701,
+      "step": 234945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000164596311542592,
+      "loss": 2.0262,
+      "step": 234950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.75,
+      "learning_rate": 0.000164594900603845,
+      "loss": 2.1125,
+      "step": 234955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001645934896430313,
+      "loss": 2.1882,
+      "step": 234960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016459207866015123,
+      "loss": 1.9909,
+      "step": 234965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016459066765520537,
+      "loss": 2.2299,
+      "step": 234970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00016458925662819412,
+      "loss": 2.0568,
+      "step": 234975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016458784557911807,
+      "loss": 1.9455,
+      "step": 234980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.00016458643450797757,
+      "loss": 2.1633,
+      "step": 234985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001645850234147732,
+      "loss": 2.1515,
+      "step": 234990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016458361229950544,
+      "loss": 2.1133,
+      "step": 234995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001645822011621747,
+      "loss": 2.1792,
+      "step": 235000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016458079000278156,
+      "loss": 2.0106,
+      "step": 235005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016457937882132642,
+      "loss": 2.0777,
+      "step": 235010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001645779676178098,
+      "loss": 2.2669,
+      "step": 235015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016457655639223216,
+      "loss": 2.0997,
+      "step": 235020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016457514514459405,
+      "loss": 2.0542,
+      "step": 235025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016457373387489586,
+      "loss": 2.0984,
+      "step": 235030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001645723225831381,
+      "loss": 2.07,
+      "step": 235035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016457091126932126,
+      "loss": 2.1218,
+      "step": 235040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016456949993344587,
+      "loss": 2.0241,
+      "step": 235045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016456808857551232,
+      "loss": 2.0647,
+      "step": 235050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016456667719552117,
+      "loss": 1.941,
+      "step": 235055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016456526579347288,
+      "loss": 1.9583,
+      "step": 235060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001645638543693679,
+      "loss": 2.3221,
+      "step": 235065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016456244292320673,
+      "loss": 2.1297,
+      "step": 235070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016456103145498987,
+      "loss": 2.2583,
+      "step": 235075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001645596199647178,
+      "loss": 2.015,
+      "step": 235080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000164558208452391,
+      "loss": 2.1515,
+      "step": 235085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001645567969180099,
+      "loss": 1.9993,
+      "step": 235090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016455538536157507,
+      "loss": 2.1613,
+      "step": 235095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016455397378308694,
+      "loss": 2.1013,
+      "step": 235100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000164552562182546,
+      "loss": 2.3083,
+      "step": 235105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016455115055995275,
+      "loss": 1.9571,
+      "step": 235110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016454973891530763,
+      "loss": 2.0825,
+      "step": 235115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016454832724861117,
+      "loss": 2.0529,
+      "step": 235120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001645469155598638,
+      "loss": 2.0925,
+      "step": 235125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016454550384906606,
+      "loss": 2.1387,
+      "step": 235130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.00016454409211621838,
+      "loss": 2.0417,
+      "step": 235135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001645426803613213,
+      "loss": 2.1477,
+      "step": 235140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016454126858437524,
+      "loss": 2.019,
+      "step": 235145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016453985678538074,
+      "loss": 2.2878,
+      "step": 235150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016453844496433826,
+      "loss": 2.0763,
+      "step": 235155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016453703312124823,
+      "loss": 1.9506,
+      "step": 235160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016453562125611125,
+      "loss": 2.061,
+      "step": 235165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016453420936892764,
+      "loss": 1.8463,
+      "step": 235170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016453279745969805,
+      "loss": 2.1002,
+      "step": 235175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016453138552842284,
+      "loss": 2.1383,
+      "step": 235180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016452997357510254,
+      "loss": 2.1641,
+      "step": 235185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016452856159973766,
+      "loss": 2.107,
+      "step": 235190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016452714960232866,
+      "loss": 2.1008,
+      "step": 235195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016452573758287596,
+      "loss": 1.9425,
+      "step": 235200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016452432554138016,
+      "loss": 1.9738,
+      "step": 235205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016452291347784166,
+      "loss": 2.0493,
+      "step": 235210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016452150139226098,
+      "loss": 2.1177,
+      "step": 235215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016452008928463852,
+      "loss": 2.1038,
+      "step": 235220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001645186771549749,
+      "loss": 2.307,
+      "step": 235225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016451726500327048,
+      "loss": 2.3316,
+      "step": 235230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001645158528295258,
+      "loss": 2.1601,
+      "step": 235235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016451444063374137,
+      "loss": 1.9189,
+      "step": 235240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016451302841591762,
+      "loss": 2.1478,
+      "step": 235245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00016451161617605503,
+      "loss": 2.0206,
+      "step": 235250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001645102039141541,
+      "loss": 2.201,
+      "step": 235255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016450879163021532,
+      "loss": 2.1432,
+      "step": 235260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001645073793242392,
+      "loss": 2.0089,
+      "step": 235265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016450596699622616,
+      "loss": 2.0359,
+      "step": 235270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016450455464617672,
+      "loss": 2.0843,
+      "step": 235275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016450314227409133,
+      "loss": 1.9677,
+      "step": 235280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001645017298799705,
+      "loss": 2.2053,
+      "step": 235285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016450031746381474,
+      "loss": 2.093,
+      "step": 235290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001644989050256245,
+      "loss": 2.1818,
+      "step": 235295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016449749256540025,
+      "loss": 2.0822,
+      "step": 235300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001644960800831425,
+      "loss": 1.9017,
+      "step": 235305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001644946675788517,
+      "loss": 2.156,
+      "step": 235310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016449325505252834,
+      "loss": 2.1037,
+      "step": 235315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016449184250417296,
+      "loss": 2.1176,
+      "step": 235320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.875,
+      "learning_rate": 0.000164490429933786,
+      "loss": 2.0273,
+      "step": 235325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016448901734136787,
+      "loss": 2.0831,
+      "step": 235330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016448760472691917,
+      "loss": 2.1124,
+      "step": 235335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016448619209044033,
+      "loss": 2.129,
+      "step": 235340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016448477943193184,
+      "loss": 2.005,
+      "step": 235345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016448336675139416,
+      "loss": 2.1641,
+      "step": 235350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016448195404882783,
+      "loss": 2.0128,
+      "step": 235355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016448054132423327,
+      "loss": 2.0822,
+      "step": 235360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000164479128577611,
+      "loss": 1.9507,
+      "step": 235365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016447771580896148,
+      "loss": 1.9741,
+      "step": 235370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001644763030182852,
+      "loss": 2.1036,
+      "step": 235375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016447489020558265,
+      "loss": 2.1186,
+      "step": 235380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00016447347737085427,
+      "loss": 2.2285,
+      "step": 235385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016447206451410064,
+      "loss": 2.1046,
+      "step": 235390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016447065163532217,
+      "loss": 2.1418,
+      "step": 235395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016446923873451934,
+      "loss": 2.0698,
+      "step": 235400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001644678258116927,
+      "loss": 2.0936,
+      "step": 235405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001644664128668426,
+      "loss": 2.0313,
+      "step": 235410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016446499989996964,
+      "loss": 2.2028,
+      "step": 235415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016446358691107427,
+      "loss": 2.0517,
+      "step": 235420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016446217390015698,
+      "loss": 2.1539,
+      "step": 235425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016446076086721822,
+      "loss": 2.1374,
+      "step": 235430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001644593478122585,
+      "loss": 2.1012,
+      "step": 235435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016445793473527832,
+      "loss": 2.0511,
+      "step": 235440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016445652163627812,
+      "loss": 2.1175,
+      "step": 235445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001644551085152584,
+      "loss": 1.9809,
+      "step": 235450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016445369537221966,
+      "loss": 2.085,
+      "step": 235455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001644522822071624,
+      "loss": 2.0481,
+      "step": 235460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000164450869020087,
+      "loss": 2.1123,
+      "step": 235465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016444945581099403,
+      "loss": 2.0918,
+      "step": 235470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.000164448042579884,
+      "loss": 2.1283,
+      "step": 235475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016444662932675732,
+      "loss": 2.1268,
+      "step": 235480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016444521605161452,
+      "loss": 2.0662,
+      "step": 235485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016444380275445605,
+      "loss": 2.0917,
+      "step": 235490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001644423894352824,
+      "loss": 2.101,
+      "step": 235495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016444097609409405,
+      "loss": 2.1484,
+      "step": 235500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016443956273089154,
+      "loss": 2.0278,
+      "step": 235505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001644381493456753,
+      "loss": 1.9831,
+      "step": 235510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001644367359384458,
+      "loss": 2.042,
+      "step": 235515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016443532250920353,
+      "loss": 1.9036,
+      "step": 235520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016443390905794902,
+      "loss": 1.8906,
+      "step": 235525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016443249558468267,
+      "loss": 2.1706,
+      "step": 235530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016443108208940506,
+      "loss": 2.1526,
+      "step": 235535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001644296685721166,
+      "loss": 2.1004,
+      "step": 235540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016442825503281777,
+      "loss": 2.0461,
+      "step": 235545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016442684147150913,
+      "loss": 2.0562,
+      "step": 235550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001644254278881911,
+      "loss": 2.1805,
+      "step": 235555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016442401428286415,
+      "loss": 2.0259,
+      "step": 235560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016442260065552882,
+      "loss": 2.1092,
+      "step": 235565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016442118700618554,
+      "loss": 2.1262,
+      "step": 235570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016441977333483482,
+      "loss": 2.0731,
+      "step": 235575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016441835964147713,
+      "loss": 1.848,
+      "step": 235580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016441694592611298,
+      "loss": 2.0529,
+      "step": 235585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016441553218874284,
+      "loss": 1.9884,
+      "step": 235590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00016441411842936716,
+      "loss": 1.9963,
+      "step": 235595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016441270464798646,
+      "loss": 2.1713,
+      "step": 235600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016441129084460122,
+      "loss": 2.1415,
+      "step": 235605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016440987701921188,
+      "loss": 1.957,
+      "step": 235610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016440846317181896,
+      "loss": 2.1154,
+      "step": 235615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000164407049302423,
+      "loss": 2.1054,
+      "step": 235620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016440563541102437,
+      "loss": 2.0844,
+      "step": 235625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016440422149762362,
+      "loss": 2.0928,
+      "step": 235630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001644028075622212,
+      "loss": 2.2372,
+      "step": 235635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016440139360481766,
+      "loss": 2.3017,
+      "step": 235640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016439997962541342,
+      "loss": 2.2462,
+      "step": 235645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016439856562400898,
+      "loss": 2.0462,
+      "step": 235650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016439715160060478,
+      "loss": 2.0272,
+      "step": 235655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001643957375552014,
+      "loss": 2.2825,
+      "step": 235660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016439432348779924,
+      "loss": 2.0079,
+      "step": 235665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001643929093983988,
+      "loss": 2.0544,
+      "step": 235670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016439149528700063,
+      "loss": 2.2545,
+      "step": 235675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001643900811536051,
+      "loss": 2.0489,
+      "step": 235680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016438866699821278,
+      "loss": 2.2081,
+      "step": 235685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001643872528208241,
+      "loss": 2.1695,
+      "step": 235690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001643858386214396,
+      "loss": 2.0594,
+      "step": 235695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001643844244000597,
+      "loss": 2.0175,
+      "step": 235700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016438301015668492,
+      "loss": 2.3268,
+      "step": 235705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016438159589131574,
+      "loss": 2.017,
+      "step": 235710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001643801816039526,
+      "loss": 2.0397,
+      "step": 235715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001643787672945961,
+      "loss": 2.1308,
+      "step": 235720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016437735296324658,
+      "loss": 2.1264,
+      "step": 235725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016437593860990464,
+      "loss": 2.1446,
+      "step": 235730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016437452423457067,
+      "loss": 2.2048,
+      "step": 235735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016437310983724522,
+      "loss": 2.0671,
+      "step": 235740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016437169541792874,
+      "loss": 2.156,
+      "step": 235745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016437028097662172,
+      "loss": 1.9422,
+      "step": 235750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016436886651332463,
+      "loss": 2.1054,
+      "step": 235755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016436745202803797,
+      "loss": 2.0844,
+      "step": 235760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016436603752076226,
+      "loss": 2.0947,
+      "step": 235765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001643646229914979,
+      "loss": 2.1185,
+      "step": 235770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016436320844024548,
+      "loss": 2.0169,
+      "step": 235775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016436179386700536,
+      "loss": 1.9408,
+      "step": 235780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001643603792717781,
+      "loss": 1.8744,
+      "step": 235785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016435896465456416,
+      "loss": 2.0907,
+      "step": 235790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016435755001536402,
+      "loss": 2.0927,
+      "step": 235795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016435613535417822,
+      "loss": 1.9512,
+      "step": 235800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016435472067100716,
+      "loss": 2.0404,
+      "step": 235805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016435330596585137,
+      "loss": 2.0755,
+      "step": 235810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001643518912387113,
+      "loss": 2.1292,
+      "step": 235815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001643504764895875,
+      "loss": 2.214,
+      "step": 235820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001643490617184804,
+      "loss": 2.1534,
+      "step": 235825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001643476469253905,
+      "loss": 1.9856,
+      "step": 235830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016434623211031824,
+      "loss": 2.0545,
+      "step": 235835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016434481727326414,
+      "loss": 2.0054,
+      "step": 235840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016434340241422868,
+      "loss": 2.136,
+      "step": 235845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016434198753321238,
+      "loss": 2.2279,
+      "step": 235850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016434057263021572,
+      "loss": 2.0424,
+      "step": 235855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001643391577052391,
+      "loss": 2.0737,
+      "step": 235860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016433774275828307,
+      "loss": 2.0225,
+      "step": 235865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001643363277893481,
+      "loss": 2.2369,
+      "step": 235870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016433491279843466,
+      "loss": 1.9267,
+      "step": 235875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016433349778554327,
+      "loss": 2.1226,
+      "step": 235880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001643320827506744,
+      "loss": 2.0757,
+      "step": 235885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001643306676938285,
+      "loss": 2.0502,
+      "step": 235890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016432925261500607,
+      "loss": 2.0446,
+      "step": 235895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016432783751420762,
+      "loss": 2.1477,
+      "step": 235900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001643264223914336,
+      "loss": 2.1569,
+      "step": 235905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001643250072466845,
+      "loss": 1.9987,
+      "step": 235910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016432359207996087,
+      "loss": 2.066,
+      "step": 235915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016432217689126306,
+      "loss": 2.1044,
+      "step": 235920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016432076168059167,
+      "loss": 2.1306,
+      "step": 235925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016431934644794714,
+      "loss": 2.1365,
+      "step": 235930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016431793119332994,
+      "loss": 2.102,
+      "step": 235935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016431651591674058,
+      "loss": 2.0966,
+      "step": 235940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001643151006181795,
+      "loss": 2.1245,
+      "step": 235945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016431368529764727,
+      "loss": 1.9661,
+      "step": 235950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016431226995514427,
+      "loss": 2.2344,
+      "step": 235955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016431085459067105,
+      "loss": 2.0533,
+      "step": 235960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001643094392042281,
+      "loss": 2.0856,
+      "step": 235965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016430802379581585,
+      "loss": 2.1337,
+      "step": 235970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016430660836543482,
+      "loss": 2.2114,
+      "step": 235975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016430519291308547,
+      "loss": 2.0203,
+      "step": 235980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001643037774387683,
+      "loss": 2.1763,
+      "step": 235985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016430236194248384,
+      "loss": 2.0854,
+      "step": 235990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016430094642423248,
+      "loss": 2.0257,
+      "step": 235995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016429953088401475,
+      "loss": 2.0294,
+      "step": 236000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016429811532183116,
+      "loss": 2.0038,
+      "step": 236005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016429669973768216,
+      "loss": 2.1309,
+      "step": 236010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016429528413156825,
+      "loss": 2.1608,
+      "step": 236015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001642938685034899,
+      "loss": 2.0041,
+      "step": 236020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016429245285344758,
+      "loss": 2.191,
+      "step": 236025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016429103718144182,
+      "loss": 1.9689,
+      "step": 236030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016428962148747306,
+      "loss": 1.9574,
+      "step": 236035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016428820577154177,
+      "loss": 1.9476,
+      "step": 236040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001642867900336485,
+      "loss": 1.9372,
+      "step": 236045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016428537427379372,
+      "loss": 2.0324,
+      "step": 236050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016428395849197783,
+      "loss": 2.1761,
+      "step": 236055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001642825426882014,
+      "loss": 1.9597,
+      "step": 236060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016428112686246488,
+      "loss": 2.1232,
+      "step": 236065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016427971101476877,
+      "loss": 2.1255,
+      "step": 236070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016427829514511358,
+      "loss": 2.1454,
+      "step": 236075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016427687925349972,
+      "loss": 1.9564,
+      "step": 236080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016427546333992773,
+      "loss": 2.1395,
+      "step": 236085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016427404740439804,
+      "loss": 2.1355,
+      "step": 236090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016427263144691117,
+      "loss": 1.9909,
+      "step": 236095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016427121546746767,
+      "loss": 1.9783,
+      "step": 236100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016426979946606787,
+      "loss": 2.206,
+      "step": 236105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016426838344271242,
+      "loss": 1.864,
+      "step": 236110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016426696739740168,
+      "loss": 2.0172,
+      "step": 236115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001642655513301362,
+      "loss": 2.0217,
+      "step": 236120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016426413524091643,
+      "loss": 2.2331,
+      "step": 236125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016426271912974288,
+      "loss": 1.9618,
+      "step": 236130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016426130299661596,
+      "loss": 2.0777,
+      "step": 236135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016425988684153628,
+      "loss": 2.1418,
+      "step": 236140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016425847066450425,
+      "loss": 2.1508,
+      "step": 236145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016425705446552032,
+      "loss": 2.1599,
+      "step": 236150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016425563824458506,
+      "loss": 2.1584,
+      "step": 236155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001642542220016989,
+      "loss": 2.4395,
+      "step": 236160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016425280573686233,
+      "loss": 2.169,
+      "step": 236165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001642513894500758,
+      "loss": 2.0983,
+      "step": 236170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016424997314133988,
+      "loss": 1.8838,
+      "step": 236175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000164248556810655,
+      "loss": 1.9038,
+      "step": 236180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001642471404580216,
+      "loss": 2.1168,
+      "step": 236185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016424572408344024,
+      "loss": 2.1408,
+      "step": 236190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001642443076869114,
+      "loss": 2.1103,
+      "step": 236195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001642428912684355,
+      "loss": 2.0649,
+      "step": 236200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001642414748280131,
+      "loss": 1.9478,
+      "step": 236205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016424005836564463,
+      "loss": 2.0669,
+      "step": 236210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016423864188133057,
+      "loss": 1.9787,
+      "step": 236215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00016423722537507146,
+      "loss": 2.0498,
+      "step": 236220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001642358088468677,
+      "loss": 2.2474,
+      "step": 236225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016423439229671988,
+      "loss": 2.3647,
+      "step": 236230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016423297572462842,
+      "loss": 2.1109,
+      "step": 236235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016423155913059377,
+      "loss": 2.0545,
+      "step": 236240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016423014251461647,
+      "loss": 2.1622,
+      "step": 236245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.000164228725876697,
+      "loss": 2.1172,
+      "step": 236250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016422730921683584,
+      "loss": 2.0913,
+      "step": 236255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016422589253503348,
+      "loss": 2.0228,
+      "step": 236260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016422447583129034,
+      "loss": 2.0353,
+      "step": 236265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016422305910560698,
+      "loss": 2.0059,
+      "step": 236270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016422164235798387,
+      "loss": 2.0913,
+      "step": 236275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016422022558842145,
+      "loss": 1.9484,
+      "step": 236280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016421880879692027,
+      "loss": 2.1114,
+      "step": 236285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016421739198348075,
+      "loss": 2.1865,
+      "step": 236290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016421597514810343,
+      "loss": 2.1389,
+      "step": 236295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016421455829078875,
+      "loss": 2.0691,
+      "step": 236300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016421314141153723,
+      "loss": 2.2729,
+      "step": 236305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016421172451034933,
+      "loss": 2.0366,
+      "step": 236310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016421030758722552,
+      "loss": 2.0985,
+      "step": 236315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016420889064216632,
+      "loss": 2.0442,
+      "step": 236320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016420747367517223,
+      "loss": 1.9534,
+      "step": 236325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016420605668624364,
+      "loss": 1.8465,
+      "step": 236330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016420463967538115,
+      "loss": 2.0331,
+      "step": 236335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001642032226425852,
+      "loss": 1.8148,
+      "step": 236340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016420180558785624,
+      "loss": 2.0074,
+      "step": 236345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001642003885111947,
+      "loss": 2.0157,
+      "step": 236350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016419897141260127,
+      "loss": 2.0882,
+      "step": 236355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016419755429207625,
+      "loss": 2.0628,
+      "step": 236360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001641961371496202,
+      "loss": 2.1415,
+      "step": 236365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016419471998523356,
+      "loss": 2.0216,
+      "step": 236370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016419330279891686,
+      "loss": 1.9983,
+      "step": 236375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016419188559067054,
+      "loss": 2.0643,
+      "step": 236380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016419046836049512,
+      "loss": 2.2954,
+      "step": 236385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001641890511083911,
+      "loss": 2.1008,
+      "step": 236390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016418763383435888,
+      "loss": 2.0702,
+      "step": 236395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016418621653839903,
+      "loss": 1.9124,
+      "step": 236400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.000164184799220512,
+      "loss": 1.9727,
+      "step": 236405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016418338188069827,
+      "loss": 2.1771,
+      "step": 236410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016418196451895836,
+      "loss": 2.1083,
+      "step": 236415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001641805471352927,
+      "loss": 2.0802,
+      "step": 236420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016417912972970184,
+      "loss": 2.1867,
+      "step": 236425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001641777123021862,
+      "loss": 2.2223,
+      "step": 236430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016417629485274629,
+      "loss": 2.0767,
+      "step": 236435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016417487738138258,
+      "loss": 2.1428,
+      "step": 236440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016417345988809556,
+      "loss": 2.1934,
+      "step": 236445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016417204237288577,
+      "loss": 2.0076,
+      "step": 236450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001641706248357536,
+      "loss": 2.1026,
+      "step": 236455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001641692072766996,
+      "loss": 2.078,
+      "step": 236460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016416778969572423,
+      "loss": 2.1749,
+      "step": 236465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016416637209282803,
+      "loss": 2.1476,
+      "step": 236470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016416495446801136,
+      "loss": 1.9448,
+      "step": 236475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001641635368212748,
+      "loss": 2.2771,
+      "step": 236480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016416211915261882,
+      "loss": 2.0394,
+      "step": 236485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001641607014620439,
+      "loss": 2.072,
+      "step": 236490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001641592837495505,
+      "loss": 2.0363,
+      "step": 236495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016415786601513914,
+      "loss": 2.1494,
+      "step": 236500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016415644825881032,
+      "loss": 2.13,
+      "step": 236505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016415503048056442,
+      "loss": 1.9922,
+      "step": 236510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016415361268040207,
+      "loss": 2.0497,
+      "step": 236515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016415219485832366,
+      "loss": 2.2476,
+      "step": 236520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016415077701432965,
+      "loss": 2.0742,
+      "step": 236525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016414935914842063,
+      "loss": 1.9839,
+      "step": 236530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000164147941260597,
+      "loss": 2.125,
+      "step": 236535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016414652335085928,
+      "loss": 1.8685,
+      "step": 236540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016414510541920793,
+      "loss": 1.8866,
+      "step": 236545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016414368746564347,
+      "loss": 2.0044,
+      "step": 236550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016414226949016636,
+      "loss": 2.0519,
+      "step": 236555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001641408514927771,
+      "loss": 2.2354,
+      "step": 236560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001641394334734761,
+      "loss": 2.0883,
+      "step": 236565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016413801543226398,
+      "loss": 2.245,
+      "step": 236570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001641365973691411,
+      "loss": 2.0522,
+      "step": 236575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016413517928410803,
+      "loss": 1.9713,
+      "step": 236580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016413376117716523,
+      "loss": 2.0621,
+      "step": 236585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016413234304831308,
+      "loss": 2.0477,
+      "step": 236590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016413092489755225,
+      "loss": 2.1019,
+      "step": 236595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001641295067248831,
+      "loss": 2.0686,
+      "step": 236600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016412808853030617,
+      "loss": 2.0989,
+      "step": 236605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016412667031382193,
+      "loss": 2.1239,
+      "step": 236610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001641252520754308,
+      "loss": 2.1964,
+      "step": 236615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.96875,
+      "learning_rate": 0.00016412383381513335,
+      "loss": 2.089,
+      "step": 236620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016412241553293003,
+      "loss": 2.0827,
+      "step": 236625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016412099722882138,
+      "loss": 2.1512,
+      "step": 236630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016411957890280778,
+      "loss": 2.2248,
+      "step": 236635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001641181605548898,
+      "loss": 2.0504,
+      "step": 236640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016411674218506786,
+      "loss": 2.1466,
+      "step": 236645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001641153237933425,
+      "loss": 1.8449,
+      "step": 236650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016411390537971417,
+      "loss": 2.0876,
+      "step": 236655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016411248694418338,
+      "loss": 2.1328,
+      "step": 236660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00016411106848675063,
+      "loss": 2.037,
+      "step": 236665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001641096500074163,
+      "loss": 2.0823,
+      "step": 236670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000164108231506181,
+      "loss": 2.1204,
+      "step": 236675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001641068129830452,
+      "loss": 2.1314,
+      "step": 236680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001641053944380093,
+      "loss": 2.1475,
+      "step": 236685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016410397587107383,
+      "loss": 2.1685,
+      "step": 236690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001641025572822393,
+      "loss": 2.189,
+      "step": 236695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016410113867150616,
+      "loss": 2.1452,
+      "step": 236700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00016409972003887492,
+      "loss": 2.2411,
+      "step": 236705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016409830138434606,
+      "loss": 2.0914,
+      "step": 236710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016409688270792002,
+      "loss": 1.9498,
+      "step": 236715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016409546400959738,
+      "loss": 2.1478,
+      "step": 236720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016409404528937852,
+      "loss": 2.2497,
+      "step": 236725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016409262654726398,
+      "loss": 2.0989,
+      "step": 236730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016409120778325425,
+      "loss": 2.1558,
+      "step": 236735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001640897889973498,
+      "loss": 2.1159,
+      "step": 236740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001640883701895511,
+      "loss": 2.0932,
+      "step": 236745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00016408695135985868,
+      "loss": 2.1853,
+      "step": 236750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016408553250827298,
+      "loss": 2.0237,
+      "step": 236755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.125,
+      "learning_rate": 0.0001640841136347945,
+      "loss": 2.1243,
+      "step": 236760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001640826947394237,
+      "loss": 2.1331,
+      "step": 236765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016408127582216112,
+      "loss": 2.1211,
+      "step": 236770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016407985688300723,
+      "loss": 2.2354,
+      "step": 236775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016407843792196246,
+      "loss": 1.9324,
+      "step": 236780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016407701893902737,
+      "loss": 2.1907,
+      "step": 236785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016407559993420235,
+      "loss": 2.1799,
+      "step": 236790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.000164074180907488,
+      "loss": 2.0473,
+      "step": 236795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016407276185888473,
+      "loss": 2.173,
+      "step": 236800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000164071342788393,
+      "loss": 2.0864,
+      "step": 236805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001640699236960134,
+      "loss": 2.2741,
+      "step": 236810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016406850458174633,
+      "loss": 1.8943,
+      "step": 236815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016406708544559228,
+      "loss": 2.128,
+      "step": 236820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001640656662875518,
+      "loss": 1.9898,
+      "step": 236825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016406424710762528,
+      "loss": 2.1129,
+      "step": 236830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016406282790581327,
+      "loss": 2.2576,
+      "step": 236835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001640614086821162,
+      "loss": 1.893,
+      "step": 236840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016405998943653464,
+      "loss": 2.2225,
+      "step": 236845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016405857016906902,
+      "loss": 2.0907,
+      "step": 236850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016405715087971983,
+      "loss": 2.2779,
+      "step": 236855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001640557315684875,
+      "loss": 2.1128,
+      "step": 236860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016405431223537263,
+      "loss": 2.0809,
+      "step": 236865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016405289288037565,
+      "loss": 2.3163,
+      "step": 236870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016405147350349698,
+      "loss": 2.0996,
+      "step": 236875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016405005410473724,
+      "loss": 1.9537,
+      "step": 236880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016404863468409678,
+      "loss": 1.9204,
+      "step": 236885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016404721524157617,
+      "loss": 2.0424,
+      "step": 236890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001640457957771759,
+      "loss": 1.9137,
+      "step": 236895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016404437629089638,
+      "loss": 2.1189,
+      "step": 236900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016404295678273815,
+      "loss": 1.9212,
+      "step": 236905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001640415372527017,
+      "loss": 2.0178,
+      "step": 236910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016404011770078747,
+      "loss": 1.9685,
+      "step": 236915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000164038698126996,
+      "loss": 2.1201,
+      "step": 236920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016403727853132774,
+      "loss": 2.0337,
+      "step": 236925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016403585891378318,
+      "loss": 1.973,
+      "step": 236930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001640344392743628,
+      "loss": 1.9857,
+      "step": 236935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001640330196130671,
+      "loss": 2.1035,
+      "step": 236940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001640315999298966,
+      "loss": 2.1408,
+      "step": 236945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001640301802248517,
+      "loss": 2.1289,
+      "step": 236950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016402876049793295,
+      "loss": 2.1766,
+      "step": 236955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016402734074914078,
+      "loss": 2.0699,
+      "step": 236960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016402592097847576,
+      "loss": 2.0787,
+      "step": 236965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001640245011859383,
+      "loss": 2.1146,
+      "step": 236970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016402308137152888,
+      "loss": 2.2816,
+      "step": 236975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016402166153524808,
+      "loss": 2.0271,
+      "step": 236980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016402024167709627,
+      "loss": 2.1751,
+      "step": 236985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000164018821797074,
+      "loss": 2.0793,
+      "step": 236990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001640174018951817,
+      "loss": 2.0822,
+      "step": 236995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016401598197141994,
+      "loss": 2.1021,
+      "step": 237000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016401456202578917,
+      "loss": 2.3105,
+      "step": 237005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016401314205828984,
+      "loss": 2.1879,
+      "step": 237010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001640117220689225,
+      "loss": 1.9693,
+      "step": 237015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001640103020576875,
+      "loss": 2.0179,
+      "step": 237020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001640088820245855,
+      "loss": 2.1548,
+      "step": 237025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001640074619696169,
+      "loss": 1.9547,
+      "step": 237030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016400604189278215,
+      "loss": 2.0953,
+      "step": 237035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001640046217940818,
+      "loss": 2.0644,
+      "step": 237040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016400320167351631,
+      "loss": 2.2322,
+      "step": 237045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016400178153108618,
+      "loss": 2.1325,
+      "step": 237050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016400036136679185,
+      "loss": 2.172,
+      "step": 237055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016399894118063386,
+      "loss": 2.1476,
+      "step": 237060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016399752097261266,
+      "loss": 1.9811,
+      "step": 237065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00016399610074272874,
+      "loss": 2.0617,
+      "step": 237070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016399468049098262,
+      "loss": 2.099,
+      "step": 237075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016399326021737472,
+      "loss": 1.9553,
+      "step": 237080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001639918399219056,
+      "loss": 1.9608,
+      "step": 237085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016399041960457567,
+      "loss": 2.1665,
+      "step": 237090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016398899926538547,
+      "loss": 2.1356,
+      "step": 237095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016398757890433545,
+      "loss": 2.21,
+      "step": 237100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016398615852142613,
+      "loss": 2.0518,
+      "step": 237105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016398473811665798,
+      "loss": 2.2251,
+      "step": 237110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016398331769003149,
+      "loss": 2.2654,
+      "step": 237115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016398189724154715,
+      "loss": 2.0406,
+      "step": 237120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016398047677120537,
+      "loss": 2.0543,
+      "step": 237125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016397905627900673,
+      "loss": 2.0972,
+      "step": 237130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016397763576495172,
+      "loss": 2.1031,
+      "step": 237135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016397621522904076,
+      "loss": 1.9062,
+      "step": 237140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016397479467127437,
+      "loss": 1.9339,
+      "step": 237145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.000163973374091653,
+      "loss": 2.1806,
+      "step": 237150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001639719534901772,
+      "loss": 2.1202,
+      "step": 237155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001639705328668474,
+      "loss": 2.2527,
+      "step": 237160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016396911222166415,
+      "loss": 2.0669,
+      "step": 237165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016396769155462785,
+      "loss": 2.0766,
+      "step": 237170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016396627086573903,
+      "loss": 2.2111,
+      "step": 237175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016396485015499816,
+      "loss": 2.1999,
+      "step": 237180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016396342942240577,
+      "loss": 2.1167,
+      "step": 237185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001639620086679623,
+      "loss": 2.0536,
+      "step": 237190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016396058789166826,
+      "loss": 1.9028,
+      "step": 237195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001639591670935241,
+      "loss": 2.1541,
+      "step": 237200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001639577462735303,
+      "loss": 2.0945,
+      "step": 237205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016395632543168742,
+      "loss": 2.1714,
+      "step": 237210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001639549045679959,
+      "loss": 2.021,
+      "step": 237215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016395348368245622,
+      "loss": 2.3085,
+      "step": 237220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016395206277506887,
+      "loss": 2.0996,
+      "step": 237225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016395064184583432,
+      "loss": 1.9663,
+      "step": 237230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016394922089475308,
+      "loss": 2.0892,
+      "step": 237235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016394779992182564,
+      "loss": 2.0043,
+      "step": 237240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016394637892705246,
+      "loss": 1.9404,
+      "step": 237245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016394495791043404,
+      "loss": 2.1917,
+      "step": 237250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016394353687197087,
+      "loss": 2.1435,
+      "step": 237255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001639421158116634,
+      "loss": 2.2607,
+      "step": 237260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016394069472951218,
+      "loss": 2.1128,
+      "step": 237265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016393927362551763,
+      "loss": 2.0495,
+      "step": 237270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001639378524996803,
+      "loss": 2.1459,
+      "step": 237275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016393643135200058,
+      "loss": 2.0434,
+      "step": 237280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016393501018247905,
+      "loss": 2.093,
+      "step": 237285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016393358899111614,
+      "loss": 2.1494,
+      "step": 237290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016393216777791237,
+      "loss": 2.1119,
+      "step": 237295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016393074654286825,
+      "loss": 2.2171,
+      "step": 237300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016392932528598418,
+      "loss": 2.2289,
+      "step": 237305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016392790400726072,
+      "loss": 2.2364,
+      "step": 237310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016392648270669832,
+      "loss": 1.8852,
+      "step": 237315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016392506138429748,
+      "loss": 2.2004,
+      "step": 237320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016392364004005867,
+      "loss": 1.9832,
+      "step": 237325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001639222186739824,
+      "loss": 2.0343,
+      "step": 237330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001639207972860691,
+      "loss": 2.0524,
+      "step": 237335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016391937587631933,
+      "loss": 2.1002,
+      "step": 237340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016391795444473356,
+      "loss": 2.1603,
+      "step": 237345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016391653299131223,
+      "loss": 2.0393,
+      "step": 237350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016391511151605585,
+      "loss": 2.0653,
+      "step": 237355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016391369001896494,
+      "loss": 2.1788,
+      "step": 237360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001639122685000399,
+      "loss": 2.1863,
+      "step": 237365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001639108469592813,
+      "loss": 2.021,
+      "step": 237370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001639094253966896,
+      "loss": 2.1294,
+      "step": 237375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016390800381226527,
+      "loss": 1.8821,
+      "step": 237380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016390658220600884,
+      "loss": 2.1881,
+      "step": 237385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016390516057792072,
+      "loss": 2.0315,
+      "step": 237390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016390373892800146,
+      "loss": 2.2772,
+      "step": 237395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016390231725625152,
+      "loss": 2.2707,
+      "step": 237400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016390089556267138,
+      "loss": 2.1211,
+      "step": 237405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016389947384726154,
+      "loss": 1.9369,
+      "step": 237410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016389805211002248,
+      "loss": 2.109,
+      "step": 237415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001638966303509547,
+      "loss": 2.1268,
+      "step": 237420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016389520857005866,
+      "loss": 1.9727,
+      "step": 237425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016389378676733487,
+      "loss": 2.2,
+      "step": 237430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001638923649427838,
+      "loss": 2.2315,
+      "step": 237435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001638909430964059,
+      "loss": 2.0162,
+      "step": 237440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016388952122820176,
+      "loss": 2.1868,
+      "step": 237445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016388809933817178,
+      "loss": 2.0798,
+      "step": 237450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016388667742631645,
+      "loss": 1.8249,
+      "step": 237455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016388525549263628,
+      "loss": 2.1156,
+      "step": 237460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016388383353713174,
+      "loss": 2.2162,
+      "step": 237465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016388241155980334,
+      "loss": 2.1751,
+      "step": 237470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016388098956065153,
+      "loss": 2.2631,
+      "step": 237475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016387956753967684,
+      "loss": 1.9981,
+      "step": 237480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001638781454968797,
+      "loss": 2.1622,
+      "step": 237485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016387672343226066,
+      "loss": 2.1624,
+      "step": 237490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016387530134582016,
+      "loss": 1.9918,
+      "step": 237495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001638738792375587,
+      "loss": 1.8492,
+      "step": 237500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016387245710747673,
+      "loss": 2.1108,
+      "step": 237505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016387103495557482,
+      "loss": 2.1573,
+      "step": 237510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001638696127818534,
+      "loss": 2.1954,
+      "step": 237515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016386819058631294,
+      "loss": 2.0844,
+      "step": 237520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016386676836895397,
+      "loss": 2.0919,
+      "step": 237525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001638653461297769,
+      "loss": 2.1161,
+      "step": 237530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016386392386878233,
+      "loss": 2.0566,
+      "step": 237535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016386250158597067,
+      "loss": 1.8916,
+      "step": 237540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001638610792813424,
+      "loss": 1.9306,
+      "step": 237545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016385965695489807,
+      "loss": 2.2412,
+      "step": 237550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016385823460663806,
+      "loss": 1.9636,
+      "step": 237555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016385681223656296,
+      "loss": 2.1039,
+      "step": 237560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016385538984467322,
+      "loss": 2.2294,
+      "step": 237565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016385396743096934,
+      "loss": 2.0897,
+      "step": 237570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016385254499545174,
+      "loss": 2.1124,
+      "step": 237575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016385112253812097,
+      "loss": 2.1214,
+      "step": 237580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016384970005897749,
+      "loss": 1.9401,
+      "step": 237585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016384827755802178,
+      "loss": 2.1185,
+      "step": 237590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001638468550352544,
+      "loss": 2.2767,
+      "step": 237595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016384543249067573,
+      "loss": 2.1875,
+      "step": 237600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001638440099242863,
+      "loss": 2.0676,
+      "step": 237605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016384258733608662,
+      "loss": 2.0355,
+      "step": 237610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016384116472607714,
+      "loss": 1.9912,
+      "step": 237615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016383974209425837,
+      "loss": 2.1206,
+      "step": 237620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016383831944063077,
+      "loss": 2.0197,
+      "step": 237625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016383689676519487,
+      "loss": 1.9341,
+      "step": 237630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001638354740679511,
+      "loss": 2.1138,
+      "step": 237635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016383405134889998,
+      "loss": 2.1398,
+      "step": 237640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016383262860804202,
+      "loss": 2.1885,
+      "step": 237645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016383120584537766,
+      "loss": 1.9628,
+      "step": 237650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016382978306090737,
+      "loss": 1.9635,
+      "step": 237655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001638283602546317,
+      "loss": 2.0729,
+      "step": 237660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001638269374265511,
+      "loss": 2.2237,
+      "step": 237665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016382551457666607,
+      "loss": 2.0079,
+      "step": 237670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016382409170497705,
+      "loss": 2.0208,
+      "step": 237675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001638226688114846,
+      "loss": 2.061,
+      "step": 237680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016382124589618916,
+      "loss": 2.0284,
+      "step": 237685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016381982295909123,
+      "loss": 2.2767,
+      "step": 237690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016381840000019126,
+      "loss": 2.1374,
+      "step": 237695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001638169770194898,
+      "loss": 2.1089,
+      "step": 237700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001638155540169873,
+      "loss": 2.1721,
+      "step": 237705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016381413099268426,
+      "loss": 2.1642,
+      "step": 237710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016381270794658113,
+      "loss": 1.8799,
+      "step": 237715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001638112848786784,
+      "loss": 2.0987,
+      "step": 237720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016380986178897664,
+      "loss": 2.0767,
+      "step": 237725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016380843867747625,
+      "loss": 2.2675,
+      "step": 237730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016380701554417773,
+      "loss": 2.1231,
+      "step": 237735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016380559238908155,
+      "loss": 2.0115,
+      "step": 237740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016380416921218827,
+      "loss": 2.2072,
+      "step": 237745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016380274601349832,
+      "loss": 2.0216,
+      "step": 237750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016380132279301217,
+      "loss": 2.121,
+      "step": 237755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016379989955073033,
+      "loss": 2.1664,
+      "step": 237760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001637984762866533,
+      "loss": 2.1848,
+      "step": 237765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016379705300078156,
+      "loss": 1.9854,
+      "step": 237770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001637956296931156,
+      "loss": 2.0641,
+      "step": 237775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001637942063636559,
+      "loss": 2.0315,
+      "step": 237780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016379278301240292,
+      "loss": 2.073,
+      "step": 237785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016379135963935716,
+      "loss": 1.9338,
+      "step": 237790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016378993624451913,
+      "loss": 1.88,
+      "step": 237795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016378851282788928,
+      "loss": 1.9963,
+      "step": 237800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016378708938946814,
+      "loss": 1.9173,
+      "step": 237805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016378566592925616,
+      "loss": 2.2182,
+      "step": 237810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016378424244725389,
+      "loss": 2.1218,
+      "step": 237815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001637828189434617,
+      "loss": 2.0945,
+      "step": 237820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016378139541788018,
+      "loss": 2.1395,
+      "step": 237825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016377997187050976,
+      "loss": 2.21,
+      "step": 237830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016377854830135094,
+      "loss": 2.1046,
+      "step": 237835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016377712471040424,
+      "loss": 2.0241,
+      "step": 237840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001637757010976701,
+      "loss": 2.0729,
+      "step": 237845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.000163774277463149,
+      "loss": 2.3081,
+      "step": 237850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001637728538068415,
+      "loss": 2.215,
+      "step": 237855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000163771430128748,
+      "loss": 2.0241,
+      "step": 237860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000163770006428869,
+      "loss": 2.0266,
+      "step": 237865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001637685827072051,
+      "loss": 2.0535,
+      "step": 237870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001637671589637566,
+      "loss": 2.0685,
+      "step": 237875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016376573519852412,
+      "loss": 2.1847,
+      "step": 237880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016376431141150812,
+      "loss": 2.1222,
+      "step": 237885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016376288760270905,
+      "loss": 2.0768,
+      "step": 237890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016376146377212745,
+      "loss": 2.0672,
+      "step": 237895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016376003991976377,
+      "loss": 2.0435,
+      "step": 237900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016375861604561848,
+      "loss": 2.0533,
+      "step": 237905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001637571921496921,
+      "loss": 2.2757,
+      "step": 237910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001637557682319851,
+      "loss": 1.9576,
+      "step": 237915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016375434429249797,
+      "loss": 2.0024,
+      "step": 237920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016375292033123126,
+      "loss": 2.24,
+      "step": 237925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016375149634818533,
+      "loss": 1.9039,
+      "step": 237930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016375007234336076,
+      "loss": 2.2137,
+      "step": 237935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.000163748648316758,
+      "loss": 2.0245,
+      "step": 237940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016374722426837755,
+      "loss": 1.9817,
+      "step": 237945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001637458001982199,
+      "loss": 2.0872,
+      "step": 237950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016374437610628553,
+      "loss": 1.8798,
+      "step": 237955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001637429519925749,
+      "loss": 2.0662,
+      "step": 237960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016374152785708854,
+      "loss": 2.1307,
+      "step": 237965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001637401036998269,
+      "loss": 1.838,
+      "step": 237970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016373867952079052,
+      "loss": 2.2036,
+      "step": 237975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016373725531997982,
+      "loss": 2.2268,
+      "step": 237980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001637358310973953,
+      "loss": 1.8581,
+      "step": 237985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016373440685303755,
+      "loss": 2.1113,
+      "step": 237990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016373298258690687,
+      "loss": 2.0987,
+      "step": 237995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016373155829900392,
+      "loss": 2.188,
+      "step": 238000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016373013398932908,
+      "loss": 2.2521,
+      "step": 238005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016372870965788286,
+      "loss": 2.0784,
+      "step": 238010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016372728530466576,
+      "loss": 1.9715,
+      "step": 238015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001637258609296783,
+      "loss": 2.1878,
+      "step": 238020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016372443653292092,
+      "loss": 2.1172,
+      "step": 238025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001637230121143941,
+      "loss": 1.9717,
+      "step": 238030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016372158767409832,
+      "loss": 2.121,
+      "step": 238035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.875,
+      "learning_rate": 0.00016372016321203412,
+      "loss": 1.9882,
+      "step": 238040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016371873872820194,
+      "loss": 2.0635,
+      "step": 238045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001637173142226023,
+      "loss": 2.2071,
+      "step": 238050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016371588969523568,
+      "loss": 1.9663,
+      "step": 238055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016371446514610254,
+      "loss": 1.9703,
+      "step": 238060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001637130405752034,
+      "loss": 2.2993,
+      "step": 238065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001637116159825387,
+      "loss": 2.0683,
+      "step": 238070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016371019136810896,
+      "loss": 1.9569,
+      "step": 238075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016370876673191468,
+      "loss": 1.9813,
+      "step": 238080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016370734207395632,
+      "loss": 2.1379,
+      "step": 238085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00016370591739423438,
+      "loss": 2.0088,
+      "step": 238090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016370449269274934,
+      "loss": 2.1458,
+      "step": 238095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001637030679695017,
+      "loss": 2.2256,
+      "step": 238100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016370164322449194,
+      "loss": 1.9775,
+      "step": 238105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016370021845772052,
+      "loss": 1.9988,
+      "step": 238110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016369879366918798,
+      "loss": 2.0121,
+      "step": 238115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016369736885889474,
+      "loss": 1.9983,
+      "step": 238120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016369594402684134,
+      "loss": 2.0842,
+      "step": 238125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016369451917302828,
+      "loss": 2.133,
+      "step": 238130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016369309429745599,
+      "loss": 2.0255,
+      "step": 238135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016369166940012497,
+      "loss": 2.076,
+      "step": 238140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016369024448103573,
+      "loss": 1.9766,
+      "step": 238145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016368881954018875,
+      "loss": 2.1351,
+      "step": 238150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016368739457758453,
+      "loss": 2.0694,
+      "step": 238155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016368596959322353,
+      "loss": 2.0706,
+      "step": 238160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016368454458710625,
+      "loss": 2.3435,
+      "step": 238165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016368311955923317,
+      "loss": 2.0786,
+      "step": 238170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016368169450960477,
+      "loss": 1.7615,
+      "step": 238175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016368026943822158,
+      "loss": 2.3003,
+      "step": 238180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016367884434508403,
+      "loss": 2.0912,
+      "step": 238185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016367741923019265,
+      "loss": 2.0723,
+      "step": 238190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001636759940935479,
+      "loss": 2.2271,
+      "step": 238195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016367456893515026,
+      "loss": 2.0273,
+      "step": 238200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016367314375500026,
+      "loss": 2.0898,
+      "step": 238205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016367171855309834,
+      "loss": 2.125,
+      "step": 238210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000163670293329445,
+      "loss": 1.9758,
+      "step": 238215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016366886808404077,
+      "loss": 2.0345,
+      "step": 238220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016366744281688604,
+      "loss": 2.1804,
+      "step": 238225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016366601752798137,
+      "loss": 2.1439,
+      "step": 238230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016366459221732726,
+      "loss": 2.0655,
+      "step": 238235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016366316688492418,
+      "loss": 2.2466,
+      "step": 238240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016366174153077258,
+      "loss": 1.9916,
+      "step": 238245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000163660316154873,
+      "loss": 2.0265,
+      "step": 238250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016365889075722588,
+      "loss": 2.0654,
+      "step": 238255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016365746533783174,
+      "loss": 2.0231,
+      "step": 238260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016365603989669105,
+      "loss": 1.9046,
+      "step": 238265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001636546144338043,
+      "loss": 2.2835,
+      "step": 238270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.000163653188949172,
+      "loss": 2.1424,
+      "step": 238275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001636517634427946,
+      "loss": 2.185,
+      "step": 238280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016365033791467258,
+      "loss": 2.0452,
+      "step": 238285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001636489123648065,
+      "loss": 2.2859,
+      "step": 238290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016364748679319674,
+      "loss": 2.0399,
+      "step": 238295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016364606119984392,
+      "loss": 2.2518,
+      "step": 238300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016364463558474839,
+      "loss": 2.258,
+      "step": 238305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001636432099479107,
+      "loss": 2.0028,
+      "step": 238310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016364178428933136,
+      "loss": 2.1904,
+      "step": 238315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001636403586090108,
+      "loss": 1.9416,
+      "step": 238320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016363893290694957,
+      "loss": 2.0533,
+      "step": 238325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016363750718314813,
+      "loss": 2.1244,
+      "step": 238330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016363608143760695,
+      "loss": 2.28,
+      "step": 238335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016363465567032653,
+      "loss": 2.0627,
+      "step": 238340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016363322988130735,
+      "loss": 2.0756,
+      "step": 238345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001636318040705499,
+      "loss": 2.1029,
+      "step": 238350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001636303782380547,
+      "loss": 2.2154,
+      "step": 238355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001636289523838222,
+      "loss": 2.0692,
+      "step": 238360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001636275265078529,
+      "loss": 2.3228,
+      "step": 238365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016362610061014724,
+      "loss": 2.0796,
+      "step": 238370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001636246746907058,
+      "loss": 2.0486,
+      "step": 238375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.000163623248749529,
+      "loss": 2.1616,
+      "step": 238380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016362182278661735,
+      "loss": 2.2491,
+      "step": 238385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001636203968019713,
+      "loss": 2.0726,
+      "step": 238390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016361897079559143,
+      "loss": 2.0875,
+      "step": 238395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016361754476747813,
+      "loss": 2.0064,
+      "step": 238400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016361611871763192,
+      "loss": 2.1781,
+      "step": 238405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001636146926460533,
+      "loss": 2.1864,
+      "step": 238410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016361326655274272,
+      "loss": 2.1904,
+      "step": 238415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016361184043770072,
+      "loss": 2.2261,
+      "step": 238420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016361041430092775,
+      "loss": 1.94,
+      "step": 238425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001636089881424243,
+      "loss": 2.2233,
+      "step": 238430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016360756196219091,
+      "loss": 2.0303,
+      "step": 238435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000163606135760228,
+      "loss": 2.152,
+      "step": 238440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016360470953653605,
+      "loss": 2.167,
+      "step": 238445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016360328329111562,
+      "loss": 1.9637,
+      "step": 238450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016360185702396716,
+      "loss": 2.0937,
+      "step": 238455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016360043073509111,
+      "loss": 2.132,
+      "step": 238460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016359900442448801,
+      "loss": 2.0825,
+      "step": 238465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016359757809215837,
+      "loss": 2.1178,
+      "step": 238470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016359615173810262,
+      "loss": 2.1837,
+      "step": 238475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016359472536232127,
+      "loss": 1.8291,
+      "step": 238480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016359329896481478,
+      "loss": 2.0198,
+      "step": 238485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016359187254558373,
+      "loss": 2.0515,
+      "step": 238490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016359044610462847,
+      "loss": 2.1529,
+      "step": 238495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016358901964194962,
+      "loss": 2.1043,
+      "step": 238500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016358759315754758,
+      "loss": 2.1454,
+      "step": 238505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016358616665142287,
+      "loss": 2.0564,
+      "step": 238510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000163584740123576,
+      "loss": 2.0383,
+      "step": 238515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016358331357400739,
+      "loss": 2.1467,
+      "step": 238520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016358188700271758,
+      "loss": 2.0941,
+      "step": 238525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016358046040970705,
+      "loss": 2.1416,
+      "step": 238530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016357903379497624,
+      "loss": 2.2068,
+      "step": 238535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016357760715852575,
+      "loss": 2.1271,
+      "step": 238540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016357618050035595,
+      "loss": 2.2206,
+      "step": 238545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016357475382046738,
+      "loss": 2.139,
+      "step": 238550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016357332711886056,
+      "loss": 2.2122,
+      "step": 238555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016357190039553588,
+      "loss": 2.0627,
+      "step": 238560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016357047365049392,
+      "loss": 2.0626,
+      "step": 238565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001635690468837351,
+      "loss": 2.1454,
+      "step": 238570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00016356762009525996,
+      "loss": 2.1829,
+      "step": 238575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000163566193285069,
+      "loss": 2.067,
+      "step": 238580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016356476645316262,
+      "loss": 2.0457,
+      "step": 238585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016356333959954142,
+      "loss": 2.0139,
+      "step": 238590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016356191272420581,
+      "loss": 2.0257,
+      "step": 238595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016356048582715627,
+      "loss": 1.9356,
+      "step": 238600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016355905890839332,
+      "loss": 2.1168,
+      "step": 238605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016355763196791747,
+      "loss": 1.9705,
+      "step": 238610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001635562050057292,
+      "loss": 2.049,
+      "step": 238615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016355477802182893,
+      "loss": 2.1427,
+      "step": 238620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001635533510162172,
+      "loss": 2.0321,
+      "step": 238625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001635519239888945,
+      "loss": 1.9431,
+      "step": 238630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001635504969398613,
+      "loss": 2.156,
+      "step": 238635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016354906986911814,
+      "loss": 1.8881,
+      "step": 238640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016354764277666544,
+      "loss": 2.034,
+      "step": 238645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001635462156625037,
+      "loss": 1.9955,
+      "step": 238650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016354478852663342,
+      "loss": 1.9286,
+      "step": 238655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001635433613690551,
+      "loss": 1.9269,
+      "step": 238660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016354193418976923,
+      "loss": 2.0769,
+      "step": 238665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016354050698877626,
+      "loss": 2.0954,
+      "step": 238670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001635390797660767,
+      "loss": 2.1032,
+      "step": 238675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016353765252167103,
+      "loss": 1.9193,
+      "step": 238680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016353622525555977,
+      "loss": 2.1083,
+      "step": 238685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016353479796774338,
+      "loss": 1.9477,
+      "step": 238690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016353337065822234,
+      "loss": 2.0656,
+      "step": 238695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016353194332699715,
+      "loss": 1.9707,
+      "step": 238700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016353051597406827,
+      "loss": 1.9857,
+      "step": 238705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016352908859943629,
+      "loss": 2.2545,
+      "step": 238710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016352766120310156,
+      "loss": 1.9626,
+      "step": 238715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016352623378506462,
+      "loss": 2.1021,
+      "step": 238720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016352480634532597,
+      "loss": 2.0535,
+      "step": 238725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016352337888388612,
+      "loss": 2.2038,
+      "step": 238730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016352195140074553,
+      "loss": 2.2423,
+      "step": 238735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001635205238959047,
+      "loss": 2.017,
+      "step": 238740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016351909636936409,
+      "loss": 1.8668,
+      "step": 238745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016351766882112423,
+      "loss": 2.1624,
+      "step": 238750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016351624125118553,
+      "loss": 2.1337,
+      "step": 238755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016351481365954856,
+      "loss": 2.2479,
+      "step": 238760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016351338604621377,
+      "loss": 2.1048,
+      "step": 238765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016351195841118166,
+      "loss": 2.245,
+      "step": 238770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016351053075445271,
+      "loss": 2.059,
+      "step": 238775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016350910307602742,
+      "loss": 2.1981,
+      "step": 238780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001635076753759063,
+      "loss": 1.9796,
+      "step": 238785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016350624765408973,
+      "loss": 1.9187,
+      "step": 238790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016350481991057834,
+      "loss": 2.1376,
+      "step": 238795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001635033921453725,
+      "loss": 2.2581,
+      "step": 238800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016350196435847278,
+      "loss": 2.2848,
+      "step": 238805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016350053654987963,
+      "loss": 2.0037,
+      "step": 238810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016349910871959352,
+      "loss": 2.1735,
+      "step": 238815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000163497680867615,
+      "loss": 2.0147,
+      "step": 238820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016349625299394451,
+      "loss": 2.0463,
+      "step": 238825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016349482509858256,
+      "loss": 2.0478,
+      "step": 238830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016349339718152962,
+      "loss": 2.1505,
+      "step": 238835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001634919692427862,
+      "loss": 1.9574,
+      "step": 238840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016349054128235273,
+      "loss": 2.0515,
+      "step": 238845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016348911330022975,
+      "loss": 2.1004,
+      "step": 238850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016348768529641776,
+      "loss": 2.0757,
+      "step": 238855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016348625727091723,
+      "loss": 2.138,
+      "step": 238860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016348482922372862,
+      "loss": 1.9693,
+      "step": 238865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016348340115485244,
+      "loss": 2.0505,
+      "step": 238870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016348197306428918,
+      "loss": 2.0635,
+      "step": 238875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016348054495203936,
+      "loss": 2.0754,
+      "step": 238880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001634791168181034,
+      "loss": 2.1979,
+      "step": 238885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001634776886624818,
+      "loss": 2.026,
+      "step": 238890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016347626048517511,
+      "loss": 2.1005,
+      "step": 238895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016347483228618378,
+      "loss": 2.1473,
+      "step": 238900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001634734040655083,
+      "loss": 1.9545,
+      "step": 238905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001634719758231491,
+      "loss": 2.0797,
+      "step": 238910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016347054755910679,
+      "loss": 2.1178,
+      "step": 238915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016346911927338174,
+      "loss": 2.1578,
+      "step": 238920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016346769096597453,
+      "loss": 1.9078,
+      "step": 238925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016346626263688557,
+      "loss": 2.1165,
+      "step": 238930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016346483428611542,
+      "loss": 1.9991,
+      "step": 238935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001634634059136645,
+      "loss": 2.1092,
+      "step": 238940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016346197751953334,
+      "loss": 2.2007,
+      "step": 238945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016346054910372238,
+      "loss": 2.0365,
+      "step": 238950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016345912066623218,
+      "loss": 2.3508,
+      "step": 238955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016345769220706322,
+      "loss": 2.037,
+      "step": 238960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016345626372621594,
+      "loss": 2.1589,
+      "step": 238965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016345483522369085,
+      "loss": 2.087,
+      "step": 238970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016345340669948844,
+      "loss": 2.1327,
+      "step": 238975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016345197815360917,
+      "loss": 2.18,
+      "step": 238980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016345054958605358,
+      "loss": 2.4007,
+      "step": 238985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016344912099682213,
+      "loss": 1.9364,
+      "step": 238990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001634476923859153,
+      "loss": 2.0408,
+      "step": 238995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016344626375333358,
+      "loss": 2.1486,
+      "step": 239000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016344483509907746,
+      "loss": 2.0172,
+      "step": 239005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016344340642314748,
+      "loss": 2.0362,
+      "step": 239010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016344197772554404,
+      "loss": 2.1235,
+      "step": 239015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001634405490062677,
+      "loss": 2.1086,
+      "step": 239020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001634391202653189,
+      "loss": 2.2,
+      "step": 239025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016343769150269815,
+      "loss": 1.9334,
+      "step": 239030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016343626271840593,
+      "loss": 1.879,
+      "step": 239035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016343483391244271,
+      "loss": 2.1661,
+      "step": 239040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016343340508480903,
+      "loss": 2.0944,
+      "step": 239045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016343197623550535,
+      "loss": 2.2248,
+      "step": 239050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016343054736453214,
+      "loss": 2.2037,
+      "step": 239055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016342911847188991,
+      "loss": 2.166,
+      "step": 239060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016342768955757915,
+      "loss": 2.164,
+      "step": 239065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016342626062160033,
+      "loss": 2.0886,
+      "step": 239070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016342483166395397,
+      "loss": 2.0798,
+      "step": 239075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001634234026846405,
+      "loss": 2.0294,
+      "step": 239080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016342197368366047,
+      "loss": 2.0206,
+      "step": 239085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016342054466101435,
+      "loss": 1.7766,
+      "step": 239090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016341911561670261,
+      "loss": 2.0776,
+      "step": 239095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016341768655072576,
+      "loss": 2.1722,
+      "step": 239100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001634162574630843,
+      "loss": 2.1524,
+      "step": 239105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016341482835377863,
+      "loss": 2.091,
+      "step": 239110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016341339922280933,
+      "loss": 2.0762,
+      "step": 239115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016341197007017688,
+      "loss": 2.3187,
+      "step": 239120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016341054089588178,
+      "loss": 2.0121,
+      "step": 239125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016340911169992442,
+      "loss": 2.0865,
+      "step": 239130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016340768248230541,
+      "loss": 1.9518,
+      "step": 239135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016340625324302516,
+      "loss": 2.1509,
+      "step": 239140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001634048239820842,
+      "loss": 2.0493,
+      "step": 239145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000163403394699483,
+      "loss": 2.0189,
+      "step": 239150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016340196539522205,
+      "loss": 1.9825,
+      "step": 239155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016340053606930184,
+      "loss": 2.1107,
+      "step": 239160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016339910672172282,
+      "loss": 2.0644,
+      "step": 239165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016339767735248554,
+      "loss": 2.1008,
+      "step": 239170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016339624796159048,
+      "loss": 2.1276,
+      "step": 239175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016339481854903812,
+      "loss": 2.143,
+      "step": 239180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016339338911482894,
+      "loss": 1.9707,
+      "step": 239185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001633919596589634,
+      "loss": 2.1675,
+      "step": 239190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000163390530181442,
+      "loss": 2.1525,
+      "step": 239195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001633891006822653,
+      "loss": 2.263,
+      "step": 239200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001633876711614337,
+      "loss": 2.2406,
+      "step": 239205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016338624161894773,
+      "loss": 2.0247,
+      "step": 239210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016338481205480787,
+      "loss": 2.1664,
+      "step": 239215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001633833824690146,
+      "loss": 2.0554,
+      "step": 239220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016338195286156842,
+      "loss": 2.0665,
+      "step": 239225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016338052323246983,
+      "loss": 2.2443,
+      "step": 239230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001633790935817193,
+      "loss": 2.1111,
+      "step": 239235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001633776639093173,
+      "loss": 2.1051,
+      "step": 239240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016337623421526436,
+      "loss": 2.1641,
+      "step": 239245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016337480449956092,
+      "loss": 2.0968,
+      "step": 239250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016337337476220754,
+      "loss": 2.0805,
+      "step": 239255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016337194500320464,
+      "loss": 2.0613,
+      "step": 239260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016337051522255275,
+      "loss": 2.017,
+      "step": 239265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016336908542025232,
+      "loss": 2.0423,
+      "step": 239270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001633676555963039,
+      "loss": 2.0425,
+      "step": 239275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001633662257507079,
+      "loss": 2.0032,
+      "step": 239280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001633647958834649,
+      "loss": 2.2365,
+      "step": 239285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016336336599457528,
+      "loss": 2.0099,
+      "step": 239290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016336193608403957,
+      "loss": 2.2175,
+      "step": 239295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00016336050615185834,
+      "loss": 2.0747,
+      "step": 239300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016335907619803198,
+      "loss": 2.1027,
+      "step": 239305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000163357646222561,
+      "loss": 2.1521,
+      "step": 239310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001633562162254459,
+      "loss": 2.1634,
+      "step": 239315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016335478620668723,
+      "loss": 2.0578,
+      "step": 239320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016335335616628533,
+      "loss": 1.9706,
+      "step": 239325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016335192610424082,
+      "loss": 2.2783,
+      "step": 239330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016335049602055413,
+      "loss": 2.1295,
+      "step": 239335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016334906591522578,
+      "loss": 2.1578,
+      "step": 239340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016334763578825624,
+      "loss": 2.2323,
+      "step": 239345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016334620563964594,
+      "loss": 2.1836,
+      "step": 239350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016334477546939548,
+      "loss": 2.0804,
+      "step": 239355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016334334527750527,
+      "loss": 2.2006,
+      "step": 239360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016334191506397585,
+      "loss": 2.2051,
+      "step": 239365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016334048482880768,
+      "loss": 2.0319,
+      "step": 239370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016333905457200122,
+      "loss": 2.1246,
+      "step": 239375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016333762429355703,
+      "loss": 2.108,
+      "step": 239380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016333619399347556,
+      "loss": 2.1462,
+      "step": 239385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016333476367175727,
+      "loss": 2.1408,
+      "step": 239390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001633333333284027,
+      "loss": 1.9505,
+      "step": 239395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001633319029634123,
+      "loss": 1.968,
+      "step": 239400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016333047257678656,
+      "loss": 2.1882,
+      "step": 239405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000163329042168526,
+      "loss": 2.1446,
+      "step": 239410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016332761173863108,
+      "loss": 2.0234,
+      "step": 239415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016332618128710234,
+      "loss": 1.9732,
+      "step": 239420
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001633247508139402,
+      "loss": 2.0175,
+      "step": 239425
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016332332031914515,
+      "loss": 2.0207,
+      "step": 239430
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016332188980271772,
+      "loss": 2.1324,
+      "step": 239435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001633204592646584,
+      "loss": 2.118,
+      "step": 239440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016331902870496764,
+      "loss": 2.0787,
+      "step": 239445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016331759812364597,
+      "loss": 2.1308,
+      "step": 239450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016331616752069384,
+      "loss": 2.1132,
+      "step": 239455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016331473689611177,
+      "loss": 2.1263,
+      "step": 239460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016331330624990028,
+      "loss": 2.0933,
+      "step": 239465
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001633118755820598,
+      "loss": 2.2494,
+      "step": 239470
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001633104448925908,
+      "loss": 2.0077,
+      "step": 239475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016330901418149383,
+      "loss": 1.9722,
+      "step": 239480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016330758344876932,
+      "loss": 2.1803,
+      "step": 239485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001633061526944178,
+      "loss": 2.1633,
+      "step": 239490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016330472191843976,
+      "loss": 2.0772,
+      "step": 239495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001633032911208357,
+      "loss": 2.133,
+      "step": 239500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016330186030160605,
+      "loss": 2.201,
+      "step": 239505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00016330042946075137,
+      "loss": 1.9784,
+      "step": 239510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00016329899859827206,
+      "loss": 2.1376,
+      "step": 239515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016329756771416872,
+      "loss": 2.1766,
+      "step": 239520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016329613680844178,
+      "loss": 2.0841,
+      "step": 239525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016329470588109172,
+      "loss": 2.1042,
+      "step": 239530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016329327493211905,
+      "loss": 2.1163,
+      "step": 239535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001632918439615242,
+      "loss": 1.9572,
+      "step": 239540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016329041296930775,
+      "loss": 2.0014,
+      "step": 239545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016328898195547016,
+      "loss": 2.0719,
+      "step": 239550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001632875509200119,
+      "loss": 1.9353,
+      "step": 239555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016328611986293344,
+      "loss": 2.0705,
+      "step": 239560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016328468878423533,
+      "loss": 2.2064,
+      "step": 239565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000163283257683918,
+      "loss": 1.994,
+      "step": 239570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016328182656198196,
+      "loss": 2.0983,
+      "step": 239575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001632803954184277,
+      "loss": 2.235,
+      "step": 239580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016327896425325574,
+      "loss": 2.1964,
+      "step": 239585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016327753306646646,
+      "loss": 2.0437,
+      "step": 239590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001632761018580605,
+      "loss": 1.9947,
+      "step": 239595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016327467062803825,
+      "loss": 2.0593,
+      "step": 239600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016327323937640023,
+      "loss": 2.0482,
+      "step": 239605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016327180810314691,
+      "loss": 1.9935,
+      "step": 239610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016327037680827883,
+      "loss": 2.0424,
+      "step": 239615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001632689454917964,
+      "loss": 2.1087,
+      "step": 239620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016326751415370017,
+      "loss": 2.19,
+      "step": 239625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001632660827939906,
+      "loss": 2.1051,
+      "step": 239630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001632646514126682,
+      "loss": 2.1092,
+      "step": 239635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016326322000973347,
+      "loss": 2.0704,
+      "step": 239640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001632617885851868,
+      "loss": 2.0719,
+      "step": 239645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016326035713902886,
+      "loss": 2.2074,
+      "step": 239650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016325892567125994,
+      "loss": 2.1641,
+      "step": 239655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001632574941818807,
+      "loss": 2.0342,
+      "step": 239660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001632560626708915,
+      "loss": 2.0717,
+      "step": 239665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016325463113829288,
+      "loss": 2.2294,
+      "step": 239670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016325319958408535,
+      "loss": 2.005,
+      "step": 239675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001632517680082694,
+      "loss": 2.0495,
+      "step": 239680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016325033641084548,
+      "loss": 2.0714,
+      "step": 239685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016324890479181412,
+      "loss": 2.0304,
+      "step": 239690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016324747315117576,
+      "loss": 1.9859,
+      "step": 239695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001632460414889309,
+      "loss": 1.9387,
+      "step": 239700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016324460980508007,
+      "loss": 2.1401,
+      "step": 239705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016324317809962374,
+      "loss": 2.0068,
+      "step": 239710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016324174637256236,
+      "loss": 2.2251,
+      "step": 239715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001632403146238965,
+      "loss": 2.0364,
+      "step": 239720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016323888285362656,
+      "loss": 2.234,
+      "step": 239725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001632374510617531,
+      "loss": 1.9669,
+      "step": 239730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016323601924827656,
+      "loss": 2.1464,
+      "step": 239735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016323458741319746,
+      "loss": 1.9105,
+      "step": 239740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016323315555651628,
+      "loss": 2.0199,
+      "step": 239745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001632317236782335,
+      "loss": 1.9066,
+      "step": 239750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001632302917783496,
+      "loss": 2.2293,
+      "step": 239755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016322885985686512,
+      "loss": 1.8936,
+      "step": 239760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001632274279137805,
+      "loss": 1.9097,
+      "step": 239765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016322599594909625,
+      "loss": 2.1727,
+      "step": 239770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016322456396281285,
+      "loss": 2.2617,
+      "step": 239775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016322313195493078,
+      "loss": 1.9894,
+      "step": 239780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016322169992545057,
+      "loss": 2.1105,
+      "step": 239785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016322026787437264,
+      "loss": 2.116,
+      "step": 239790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016321883580169754,
+      "loss": 2.1512,
+      "step": 239795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016321740370742576,
+      "loss": 2.0025,
+      "step": 239800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016321597159155773,
+      "loss": 2.275,
+      "step": 239805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016321453945409398,
+      "loss": 1.9502,
+      "step": 239810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016321310729503502,
+      "loss": 2.4019,
+      "step": 239815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001632116751143813,
+      "loss": 2.0561,
+      "step": 239820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016321024291213334,
+      "loss": 2.1417,
+      "step": 239825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001632088106882916,
+      "loss": 2.0695,
+      "step": 239830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016320737844285657,
+      "loss": 2.2442,
+      "step": 239835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001632059461758288,
+      "loss": 2.1403,
+      "step": 239840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001632045138872087,
+      "loss": 2.1162,
+      "step": 239845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016320308157699681,
+      "loss": 2.0564,
+      "step": 239850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001632016492451936,
+      "loss": 2.125,
+      "step": 239855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016320021689179956,
+      "loss": 2.1616,
+      "step": 239860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016319878451681515,
+      "loss": 2.2194,
+      "step": 239865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001631973521202409,
+      "loss": 2.2199,
+      "step": 239870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00016319591970207734,
+      "loss": 2.0338,
+      "step": 239875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016319448726232486,
+      "loss": 2.2206,
+      "step": 239880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000163193054800984,
+      "loss": 2.0828,
+      "step": 239885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016319162231805522,
+      "loss": 2.2058,
+      "step": 239890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001631901898135391,
+      "loss": 2.2438,
+      "step": 239895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016318875728743603,
+      "loss": 2.0632,
+      "step": 239900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016318732473974654,
+      "loss": 2.3137,
+      "step": 239905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016318589217047112,
+      "loss": 1.9481,
+      "step": 239910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016318445957961018,
+      "loss": 2.1655,
+      "step": 239915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016318302696716435,
+      "loss": 2.0155,
+      "step": 239920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016318159433313407,
+      "loss": 1.8922,
+      "step": 239925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016318016167751978,
+      "loss": 2.0108,
+      "step": 239930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016317872900032204,
+      "loss": 2.1395,
+      "step": 239935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016317729630154127,
+      "loss": 2.2637,
+      "step": 239940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016317586358117797,
+      "loss": 2.0879,
+      "step": 239945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016317443083923268,
+      "loss": 2.1634,
+      "step": 239950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016317299807570585,
+      "loss": 2.1402,
+      "step": 239955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016317156529059798,
+      "loss": 2.1211,
+      "step": 239960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016317013248390953,
+      "loss": 2.1819,
+      "step": 239965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016316869965564106,
+      "loss": 2.0208,
+      "step": 239970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016316726680579299,
+      "loss": 2.223,
+      "step": 239975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016316583393436584,
+      "loss": 1.9659,
+      "step": 239980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001631644010413601,
+      "loss": 2.0907,
+      "step": 239985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016316296812677624,
+      "loss": 2.0439,
+      "step": 239990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001631615351906148,
+      "loss": 1.9836,
+      "step": 239995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016316010223287622,
+      "loss": 2.0259,
+      "step": 240000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000163158669253561,
+      "loss": 2.0979,
+      "step": 240005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016315723625266962,
+      "loss": 2.2108,
+      "step": 240010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001631558032302026,
+      "loss": 2.0766,
+      "step": 240015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001631543701861604,
+      "loss": 2.3185,
+      "step": 240020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001631529371205435,
+      "loss": 1.902,
+      "step": 240025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016315150403335243,
+      "loss": 2.1473,
+      "step": 240030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001631500709245877,
+      "loss": 1.9616,
+      "step": 240035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001631486377942497,
+      "loss": 2.1448,
+      "step": 240040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.000163147204642339,
+      "loss": 1.8804,
+      "step": 240045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001631457714688561,
+      "loss": 2.1275,
+      "step": 240050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016314433827380143,
+      "loss": 2.0522,
+      "step": 240055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016314290505717554,
+      "loss": 2.124,
+      "step": 240060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016314147181897885,
+      "loss": 1.9543,
+      "step": 240065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001631400385592119,
+      "loss": 2.1887,
+      "step": 240070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016313860527787517,
+      "loss": 2.2244,
+      "step": 240075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016313717197496917,
+      "loss": 2.0636,
+      "step": 240080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016313573865049433,
+      "loss": 2.1254,
+      "step": 240085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001631343053044512,
+      "loss": 2.0815,
+      "step": 240090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016313287193684024,
+      "loss": 2.0405,
+      "step": 240095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016313143854766192,
+      "loss": 2.1361,
+      "step": 240100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016313000513691682,
+      "loss": 2.1539,
+      "step": 240105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001631285717046053,
+      "loss": 2.2306,
+      "step": 240110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016312713825072795,
+      "loss": 2.0958,
+      "step": 240115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016312570477528522,
+      "loss": 1.9962,
+      "step": 240120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001631242712782776,
+      "loss": 2.0915,
+      "step": 240125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016312283775970555,
+      "loss": 2.1307,
+      "step": 240130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016312140421956964,
+      "loss": 2.2107,
+      "step": 240135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001631199706578703,
+      "loss": 2.1942,
+      "step": 240140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016311853707460806,
+      "loss": 2.0838,
+      "step": 240145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016311710346978336,
+      "loss": 2.047,
+      "step": 240150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001631156698433967,
+      "loss": 1.9797,
+      "step": 240155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016311423619544857,
+      "loss": 2.1503,
+      "step": 240160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001631128025259395,
+      "loss": 2.1654,
+      "step": 240165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016311136883486994,
+      "loss": 2.1302,
+      "step": 240170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016310993512224042,
+      "loss": 2.1618,
+      "step": 240175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016310850138805138,
+      "loss": 2.0177,
+      "step": 240180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.0001631070676323033,
+      "loss": 2.1108,
+      "step": 240185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016310563385499676,
+      "loss": 2.0429,
+      "step": 240190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016310420005613216,
+      "loss": 2.0223,
+      "step": 240195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016310276623571004,
+      "loss": 1.9308,
+      "step": 240200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016310133239373084,
+      "loss": 2.0869,
+      "step": 240205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016309989853019506,
+      "loss": 1.9232,
+      "step": 240210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016309846464510327,
+      "loss": 2.1588,
+      "step": 240215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016309703073845588,
+      "loss": 2.254,
+      "step": 240220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016309559681025338,
+      "loss": 2.1745,
+      "step": 240225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016309416286049632,
+      "loss": 2.0807,
+      "step": 240230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001630927288891851,
+      "loss": 1.9542,
+      "step": 240235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001630912948963203,
+      "loss": 1.9314,
+      "step": 240240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016308986088190236,
+      "loss": 2.2037,
+      "step": 240245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016308842684593176,
+      "loss": 2.1533,
+      "step": 240250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016308699278840905,
+      "loss": 2.0753,
+      "step": 240255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016308555870933464,
+      "loss": 2.005,
+      "step": 240260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016308412460870907,
+      "loss": 2.0246,
+      "step": 240265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016308269048653283,
+      "loss": 2.0839,
+      "step": 240270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001630812563428064,
+      "loss": 2.1116,
+      "step": 240275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016307982217753024,
+      "loss": 2.1802,
+      "step": 240280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001630783879907049,
+      "loss": 2.1905,
+      "step": 240285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016307695378233082,
+      "loss": 2.2413,
+      "step": 240290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016307551955240852,
+      "loss": 2.1144,
+      "step": 240295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016307408530093844,
+      "loss": 2.0515,
+      "step": 240300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016307265102792116,
+      "loss": 2.0556,
+      "step": 240305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001630712167333571,
+      "loss": 2.162,
+      "step": 240310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016306978241724678,
+      "loss": 2.11,
+      "step": 240315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016306834807959067,
+      "loss": 2.1737,
+      "step": 240320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016306691372038924,
+      "loss": 1.9687,
+      "step": 240325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016306547933964304,
+      "loss": 2.1532,
+      "step": 240330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016306404493735251,
+      "loss": 2.146,
+      "step": 240335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001630626105135182,
+      "loss": 2.1039,
+      "step": 240340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001630611760681405,
+      "loss": 2.2102,
+      "step": 240345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016305974160122,
+      "loss": 2.1747,
+      "step": 240350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016305830711275712,
+      "loss": 1.9708,
+      "step": 240355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016305687260275242,
+      "loss": 1.9312,
+      "step": 240360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016305543807120635,
+      "loss": 2.286,
+      "step": 240365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016305400351811935,
+      "loss": 2.1764,
+      "step": 240370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016305256894349197,
+      "loss": 2.0657,
+      "step": 240375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001630511343473247,
+      "loss": 2.1541,
+      "step": 240380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016304969972961804,
+      "loss": 2.0898,
+      "step": 240385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001630482650903724,
+      "loss": 2.2495,
+      "step": 240390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016304683042958837,
+      "loss": 2.1631,
+      "step": 240395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001630453957472664,
+      "loss": 2.0967,
+      "step": 240400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.000163043961043407,
+      "loss": 1.9774,
+      "step": 240405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001630425263180106,
+      "loss": 2.2891,
+      "step": 240410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016304109157107773,
+      "loss": 2.1295,
+      "step": 240415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001630396568026089,
+      "loss": 1.9128,
+      "step": 240420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001630382220126046,
+      "loss": 2.1375,
+      "step": 240425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016303678720106527,
+      "loss": 2.1413,
+      "step": 240430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001630353523679914,
+      "loss": 2.2072,
+      "step": 240435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016303391751338358,
+      "loss": 2.0985,
+      "step": 240440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.0001630324826372422,
+      "loss": 2.1182,
+      "step": 240445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016303104773956776,
+      "loss": 2.0093,
+      "step": 240450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016302961282036078,
+      "loss": 2.1874,
+      "step": 240455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016302817787962174,
+      "loss": 2.0918,
+      "step": 240460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016302674291735118,
+      "loss": 2.1312,
+      "step": 240465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016302530793354947,
+      "loss": 2.0977,
+      "step": 240470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016302387292821722,
+      "loss": 2.2606,
+      "step": 240475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016302243790135485,
+      "loss": 2.0164,
+      "step": 240480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016302100285296288,
+      "loss": 2.0206,
+      "step": 240485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016301956778304178,
+      "loss": 2.1313,
+      "step": 240490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001630181326915921,
+      "loss": 2.2047,
+      "step": 240495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016301669757861422,
+      "loss": 2.0239,
+      "step": 240500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.0001630152624441087,
+      "loss": 2.1588,
+      "step": 240505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001630138272880761,
+      "loss": 2.1423,
+      "step": 240510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016301239211051675,
+      "loss": 1.906,
+      "step": 240515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016301095691143128,
+      "loss": 2.2104,
+      "step": 240520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016300952169082006,
+      "loss": 1.9507,
+      "step": 240525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016300808644868368,
+      "loss": 2.163,
+      "step": 240530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016300665118502262,
+      "loss": 1.9272,
+      "step": 240535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001630052158998373,
+      "loss": 2.1216,
+      "step": 240540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016300378059312827,
+      "loss": 2.0297,
+      "step": 240545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016300234526489603,
+      "loss": 2.1696,
+      "step": 240550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016300090991514101,
+      "loss": 2.3286,
+      "step": 240555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016299947454386375,
+      "loss": 1.9007,
+      "step": 240560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016299803915106474,
+      "loss": 2.1138,
+      "step": 240565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016299660373674443,
+      "loss": 2.1915,
+      "step": 240570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016299516830090339,
+      "loss": 2.1563,
+      "step": 240575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000162993732843542,
+      "loss": 2.2295,
+      "step": 240580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016299229736466085,
+      "loss": 2.1479,
+      "step": 240585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001629908618642604,
+      "loss": 1.8041,
+      "step": 240590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016298942634234106,
+      "loss": 2.2716,
+      "step": 240595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016298799079890347,
+      "loss": 2.159,
+      "step": 240600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016298655523394796,
+      "loss": 1.9839,
+      "step": 240605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016298511964747515,
+      "loss": 1.9177,
+      "step": 240610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016298368403948546,
+      "loss": 1.9632,
+      "step": 240615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016298224840997943,
+      "loss": 2.0903,
+      "step": 240620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001629808127589575,
+      "loss": 2.1219,
+      "step": 240625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001629793770864202,
+      "loss": 2.1374,
+      "step": 240630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016297794139236797,
+      "loss": 2.1772,
+      "step": 240635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016297650567680133,
+      "loss": 2.1317,
+      "step": 240640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016297506993972083,
+      "loss": 2.1561,
+      "step": 240645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016297363418112684,
+      "loss": 2.1339,
+      "step": 240650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016297219840101995,
+      "loss": 2.0196,
+      "step": 240655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016297076259940064,
+      "loss": 2.1477,
+      "step": 240660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016296932677626935,
+      "loss": 1.9701,
+      "step": 240665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016296789093162657,
+      "loss": 2.1825,
+      "step": 240670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016296645506547286,
+      "loss": 2.1227,
+      "step": 240675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016296501917780862,
+      "loss": 2.1859,
+      "step": 240680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016296358326863445,
+      "loss": 2.0263,
+      "step": 240685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001629621473379507,
+      "loss": 1.9677,
+      "step": 240690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000162960711385758,
+      "loss": 2.1708,
+      "step": 240695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016295927541205675,
+      "loss": 2.0436,
+      "step": 240700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016295783941684748,
+      "loss": 1.9738,
+      "step": 240705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016295640340013064,
+      "loss": 2.0958,
+      "step": 240710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016295496736190682,
+      "loss": 2.1187,
+      "step": 240715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016295353130217638,
+      "loss": 2.2101,
+      "step": 240720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001629520952209399,
+      "loss": 2.0852,
+      "step": 240725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016295065911819782,
+      "loss": 1.9301,
+      "step": 240730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016294922299395069,
+      "loss": 1.8519,
+      "step": 240735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016294778684819894,
+      "loss": 2.0118,
+      "step": 240740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016294635068094307,
+      "loss": 2.2016,
+      "step": 240745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001629449144921836,
+      "loss": 2.0126,
+      "step": 240750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016294347828192101,
+      "loss": 2.0438,
+      "step": 240755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016294204205015577,
+      "loss": 2.2381,
+      "step": 240760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001629406057968884,
+      "loss": 2.2937,
+      "step": 240765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016293916952211937,
+      "loss": 2.1748,
+      "step": 240770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001629377332258492,
+      "loss": 2.1741,
+      "step": 240775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016293629690807836,
+      "loss": 2.1286,
+      "step": 240780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016293486056880732,
+      "loss": 2.1866,
+      "step": 240785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001629334242080366,
+      "loss": 2.1209,
+      "step": 240790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016293198782576664,
+      "loss": 2.0907,
+      "step": 240795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016293055142199803,
+      "loss": 2.1485,
+      "step": 240800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016292911499673118,
+      "loss": 2.0602,
+      "step": 240805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001629276785499666,
+      "loss": 2.149,
+      "step": 240810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016292624208170478,
+      "loss": 2.2313,
+      "step": 240815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016292480559194623,
+      "loss": 2.0247,
+      "step": 240820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016292336908069142,
+      "loss": 2.069,
+      "step": 240825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016292193254794084,
+      "loss": 2.1907,
+      "step": 240830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016292049599369497,
+      "loss": 2.1331,
+      "step": 240835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016291905941795437,
+      "loss": 2.0912,
+      "step": 240840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016291762282071942,
+      "loss": 2.1197,
+      "step": 240845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016291618620199069,
+      "loss": 2.053,
+      "step": 240850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016291474956176867,
+      "loss": 1.9875,
+      "step": 240855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001629133129000538,
+      "loss": 2.1291,
+      "step": 240860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016291187621684662,
+      "loss": 2.0708,
+      "step": 240865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001629104395121476,
+      "loss": 1.9184,
+      "step": 240870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001629090027859572,
+      "loss": 1.9138,
+      "step": 240875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.000162907566038276,
+      "loss": 2.1355,
+      "step": 240880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016290612926910442,
+      "loss": 2.2322,
+      "step": 240885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016290469247844293,
+      "loss": 2.1623,
+      "step": 240890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001629032556662921,
+      "loss": 2.1757,
+      "step": 240895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016290181883265234,
+      "loss": 1.9605,
+      "step": 240900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016290038197752418,
+      "loss": 2.1228,
+      "step": 240905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016289894510090812,
+      "loss": 2.2147,
+      "step": 240910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016289750820280466,
+      "loss": 2.2452,
+      "step": 240915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016289607128321424,
+      "loss": 1.9044,
+      "step": 240920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001628946343421374,
+      "loss": 2.0122,
+      "step": 240925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001628931973795746,
+      "loss": 2.2382,
+      "step": 240930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016289176039552635,
+      "loss": 2.2092,
+      "step": 240935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016289032338999315,
+      "loss": 2.2012,
+      "step": 240940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016288888636297545,
+      "loss": 2.0318,
+      "step": 240945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016288744931447377,
+      "loss": 1.9581,
+      "step": 240950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001628860122444886,
+      "loss": 2.105,
+      "step": 240955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016288457515302045,
+      "loss": 2.0398,
+      "step": 240960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016288313804006975,
+      "loss": 2.1553,
+      "step": 240965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00016288170090563705,
+      "loss": 1.9389,
+      "step": 240970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016288026374972284,
+      "loss": 1.9875,
+      "step": 240975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016287882657232755,
+      "loss": 2.1486,
+      "step": 240980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016287738937345174,
+      "loss": 2.283,
+      "step": 240985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016287595215309586,
+      "loss": 1.9538,
+      "step": 240990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016287451491126044,
+      "loss": 2.0556,
+      "step": 240995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001628730776479459,
+      "loss": 2.1689,
+      "step": 241000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001628716403631528,
+      "loss": 2.223,
+      "step": 241005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001628702030568816,
+      "loss": 1.8622,
+      "step": 241010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016286876572913284,
+      "loss": 2.187,
+      "step": 241015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016286732837990692,
+      "loss": 2.0677,
+      "step": 241020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016286589100920443,
+      "loss": 2.0249,
+      "step": 241025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016286445361702576,
+      "loss": 2.2167,
+      "step": 241030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016286301620337148,
+      "loss": 2.0865,
+      "step": 241035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016286157876824203,
+      "loss": 2.0493,
+      "step": 241040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016286014131163792,
+      "loss": 2.0523,
+      "step": 241045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016285870383355968,
+      "loss": 2.1104,
+      "step": 241050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016285726633400778,
+      "loss": 2.2748,
+      "step": 241055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016285582881298265,
+      "loss": 2.081,
+      "step": 241060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016285439127048483,
+      "loss": 1.8653,
+      "step": 241065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016285295370651484,
+      "loss": 2.0063,
+      "step": 241070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016285151612107312,
+      "loss": 1.9783,
+      "step": 241075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001628500785141602,
+      "loss": 2.0367,
+      "step": 241080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00016284864088577655,
+      "loss": 2.2885,
+      "step": 241085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016284720323592265,
+      "loss": 2.0339,
+      "step": 241090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000162845765564599,
+      "loss": 2.308,
+      "step": 241095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016284432787180614,
+      "loss": 2.1815,
+      "step": 241100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001628428901575445,
+      "loss": 2.0413,
+      "step": 241105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016284145242181458,
+      "loss": 2.1536,
+      "step": 241110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016284001466461686,
+      "loss": 2.0941,
+      "step": 241115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016283857688595188,
+      "loss": 2.1361,
+      "step": 241120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001628371390858201,
+      "loss": 2.0521,
+      "step": 241125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000162835701264222,
+      "loss": 2.0266,
+      "step": 241130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016283426342115812,
+      "loss": 2.1408,
+      "step": 241135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016283282555662885,
+      "loss": 2.0972,
+      "step": 241140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001628313876706348,
+      "loss": 2.202,
+      "step": 241145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016282994976317639,
+      "loss": 2.1924,
+      "step": 241150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016282851183425413,
+      "loss": 1.9726,
+      "step": 241155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001628270738838685,
+      "loss": 2.1651,
+      "step": 241160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016282563591202002,
+      "loss": 1.9986,
+      "step": 241165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016282419791870914,
+      "loss": 2.1805,
+      "step": 241170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016282275990393643,
+      "loss": 2.1794,
+      "step": 241175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016282132186770228,
+      "loss": 2.0696,
+      "step": 241180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016281988381000722,
+      "loss": 2.0179,
+      "step": 241185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001628184457308518,
+      "loss": 2.2027,
+      "step": 241190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016281700763023645,
+      "loss": 2.1113,
+      "step": 241195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016281556950816162,
+      "loss": 2.1552,
+      "step": 241200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016281413136462786,
+      "loss": 2.1134,
+      "step": 241205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016281269319963569,
+      "loss": 1.9964,
+      "step": 241210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016281125501318555,
+      "loss": 2.1368,
+      "step": 241215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016280981680527793,
+      "loss": 2.2225,
+      "step": 241220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00016280837857591337,
+      "loss": 1.947,
+      "step": 241225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001628069403250923,
+      "loss": 2.1382,
+      "step": 241230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016280550205281525,
+      "loss": 2.0884,
+      "step": 241235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016280406375908268,
+      "loss": 2.3719,
+      "step": 241240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016280262544389514,
+      "loss": 2.123,
+      "step": 241245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016280118710725305,
+      "loss": 2.0799,
+      "step": 241250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016279974874915695,
+      "loss": 2.3546,
+      "step": 241255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016279831036960734,
+      "loss": 2.1315,
+      "step": 241260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016279687196860468,
+      "loss": 2.0638,
+      "step": 241265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016279543354614945,
+      "loss": 2.1174,
+      "step": 241270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016279399510224217,
+      "loss": 2.1517,
+      "step": 241275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001627925566368833,
+      "loss": 2.1414,
+      "step": 241280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016279111815007338,
+      "loss": 2.1266,
+      "step": 241285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001627896796418129,
+      "loss": 2.2835,
+      "step": 241290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016278824111210226,
+      "loss": 2.1052,
+      "step": 241295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016278680256094207,
+      "loss": 2.3478,
+      "step": 241300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016278536398833275,
+      "loss": 2.1774,
+      "step": 241305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001627839253942748,
+      "loss": 2.2355,
+      "step": 241310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016278248677876878,
+      "loss": 2.0726,
+      "step": 241315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016278104814181508,
+      "loss": 2.0544,
+      "step": 241320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016277960948341425,
+      "loss": 2.0653,
+      "step": 241325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016277817080356672,
+      "loss": 2.3017,
+      "step": 241330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016277673210227307,
+      "loss": 1.9667,
+      "step": 241335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016277529337953373,
+      "loss": 2.3555,
+      "step": 241340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016277385463534923,
+      "loss": 2.1666,
+      "step": 241345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016277241586972006,
+      "loss": 2.0503,
+      "step": 241350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016277097708264666,
+      "loss": 2.0628,
+      "step": 241355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001627695382741296,
+      "loss": 2.0161,
+      "step": 241360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016276809944416925,
+      "loss": 2.0833,
+      "step": 241365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016276666059276623,
+      "loss": 2.0144,
+      "step": 241370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016276522171992098,
+      "loss": 2.1488,
+      "step": 241375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016276378282563397,
+      "loss": 2.1852,
+      "step": 241380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016276234390990573,
+      "loss": 2.147,
+      "step": 241385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016276090497273672,
+      "loss": 2.0003,
+      "step": 241390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001627594660141275,
+      "loss": 2.0704,
+      "step": 241395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016275802703407844,
+      "loss": 2.2133,
+      "step": 241400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016275658803259016,
+      "loss": 1.9811,
+      "step": 241405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016275514900966306,
+      "loss": 2.1415,
+      "step": 241410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016275370996529764,
+      "loss": 2.0747,
+      "step": 241415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016275227089949446,
+      "loss": 2.0319,
+      "step": 241420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016275083181225392,
+      "loss": 2.0862,
+      "step": 241425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001627493927035766,
+      "loss": 1.8984,
+      "step": 241430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016274795357346293,
+      "loss": 2.2727,
+      "step": 241435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001627465144219134,
+      "loss": 2.1844,
+      "step": 241440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016274507524892856,
+      "loss": 2.032,
+      "step": 241445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016274363605450884,
+      "loss": 2.2855,
+      "step": 241450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016274219683865479,
+      "loss": 2.053,
+      "step": 241455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016274075760136682,
+      "loss": 2.0278,
+      "step": 241460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001627393183426455,
+      "loss": 2.174,
+      "step": 241465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016273787906249127,
+      "loss": 2.0122,
+      "step": 241470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016273643976090466,
+      "loss": 2.2298,
+      "step": 241475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016273500043788617,
+      "loss": 2.083,
+      "step": 241480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016273356109343623,
+      "loss": 2.1377,
+      "step": 241485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016273212172755538,
+      "loss": 2.2188,
+      "step": 241490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016273068234024408,
+      "loss": 2.126,
+      "step": 241495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016272924293150284,
+      "loss": 2.0478,
+      "step": 241500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016272780350133216,
+      "loss": 2.2236,
+      "step": 241505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016272636404973255,
+      "loss": 2.1515,
+      "step": 241510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016272492457670445,
+      "loss": 2.0877,
+      "step": 241515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016272348508224838,
+      "loss": 2.0889,
+      "step": 241520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016272204556636485,
+      "loss": 2.2513,
+      "step": 241525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016272060602905432,
+      "loss": 2.1555,
+      "step": 241530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001627191664703173,
+      "loss": 2.2146,
+      "step": 241535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016271772689015426,
+      "loss": 1.9563,
+      "step": 241540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001627162872885657,
+      "loss": 2.1215,
+      "step": 241545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016271484766555214,
+      "loss": 2.0255,
+      "step": 241550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016271340802111406,
+      "loss": 2.0136,
+      "step": 241555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016271196835525192,
+      "loss": 2.2132,
+      "step": 241560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016271052866796625,
+      "loss": 2.0631,
+      "step": 241565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016270908895925753,
+      "loss": 2.2026,
+      "step": 241570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016270764922912621,
+      "loss": 2.0804,
+      "step": 241575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016270620947757286,
+      "loss": 2.052,
+      "step": 241580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001627047697045979,
+      "loss": 2.1253,
+      "step": 241585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016270332991020188,
+      "loss": 2.0399,
+      "step": 241590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016270189009438525,
+      "loss": 2.2214,
+      "step": 241595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016270045025714854,
+      "loss": 2.3083,
+      "step": 241600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001626990103984922,
+      "loss": 2.1579,
+      "step": 241605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016269757051841675,
+      "loss": 2.1187,
+      "step": 241610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016269613061692265,
+      "loss": 2.003,
+      "step": 241615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016269469069401043,
+      "loss": 2.239,
+      "step": 241620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016269325074968056,
+      "loss": 1.9506,
+      "step": 241625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016269181078393355,
+      "loss": 2.1357,
+      "step": 241630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016269037079676986,
+      "loss": 2.2055,
+      "step": 241635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016268893078819004,
+      "loss": 1.9302,
+      "step": 241640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001626874907581945,
+      "loss": 2.0306,
+      "step": 241645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016268605070678382,
+      "loss": 2.1421,
+      "step": 241650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001626846106339584,
+      "loss": 1.9397,
+      "step": 241655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001626831705397188,
+      "loss": 1.9545,
+      "step": 241660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016268173042406555,
+      "loss": 1.9875,
+      "step": 241665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000162680290286999,
+      "loss": 2.1954,
+      "step": 241670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016267885012851978,
+      "loss": 2.0472,
+      "step": 241675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016267740994862828,
+      "loss": 1.9668,
+      "step": 241680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016267596974732507,
+      "loss": 1.968,
+      "step": 241685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016267452952461062,
+      "loss": 2.1129,
+      "step": 241690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016267308928048538,
+      "loss": 2.0685,
+      "step": 241695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016267164901494994,
+      "loss": 2.0694,
+      "step": 241700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016267020872800467,
+      "loss": 2.0771,
+      "step": 241705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016266876841965013,
+      "loss": 2.0944,
+      "step": 241710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001626673280898868,
+      "loss": 1.9881,
+      "step": 241715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016266588773871518,
+      "loss": 2.1764,
+      "step": 241720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016266444736613578,
+      "loss": 1.9756,
+      "step": 241725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016266300697214905,
+      "loss": 2.1182,
+      "step": 241730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00016266156655675548,
+      "loss": 1.8241,
+      "step": 241735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016266012611995562,
+      "loss": 2.0864,
+      "step": 241740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016265868566174988,
+      "loss": 2.1661,
+      "step": 241745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016265724518213883,
+      "loss": 2.064,
+      "step": 241750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016265580468112294,
+      "loss": 1.9569,
+      "step": 241755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016265436415870266,
+      "loss": 2.0789,
+      "step": 241760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001626529236148785,
+      "loss": 2.1066,
+      "step": 241765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016265148304965102,
+      "loss": 2.0817,
+      "step": 241770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016265004246302062,
+      "loss": 2.1285,
+      "step": 241775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016264860185498785,
+      "loss": 2.0725,
+      "step": 241780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016264716122555317,
+      "loss": 2.1089,
+      "step": 241785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016264572057471708,
+      "loss": 2.0697,
+      "step": 241790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016264427990248006,
+      "loss": 2.0488,
+      "step": 241795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016264283920884264,
+      "loss": 2.1826,
+      "step": 241800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001626413984938053,
+      "loss": 1.9772,
+      "step": 241805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016263995775736852,
+      "loss": 2.1082,
+      "step": 241810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016263851699953275,
+      "loss": 1.894,
+      "step": 241815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016263707622029855,
+      "loss": 2.1519,
+      "step": 241820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001626356354196664,
+      "loss": 2.1473,
+      "step": 241825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001626341945976368,
+      "loss": 2.0061,
+      "step": 241830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016263275375421018,
+      "loss": 2.184,
+      "step": 241835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016263131288938708,
+      "loss": 2.1166,
+      "step": 241840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016262987200316804,
+      "loss": 1.9282,
+      "step": 241845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016262843109555344,
+      "loss": 2.0547,
+      "step": 241850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016262699016654386,
+      "loss": 2.1323,
+      "step": 241855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016262554921613978,
+      "loss": 2.3125,
+      "step": 241860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016262410824434165,
+      "loss": 2.1359,
+      "step": 241865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016262266725115,
+      "loss": 2.1331,
+      "step": 241870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016262122623656532,
+      "loss": 2.0478,
+      "step": 241875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016261978520058806,
+      "loss": 2.1862,
+      "step": 241880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016261834414321876,
+      "loss": 2.0143,
+      "step": 241885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001626169030644579,
+      "loss": 2.2747,
+      "step": 241890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016261546196430596,
+      "loss": 1.9426,
+      "step": 241895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016261402084276347,
+      "loss": 2.158,
+      "step": 241900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016261257969983088,
+      "loss": 2.1513,
+      "step": 241905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001626111385355087,
+      "loss": 1.984,
+      "step": 241910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016260969734979743,
+      "loss": 2.0157,
+      "step": 241915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016260825614269753,
+      "loss": 2.1611,
+      "step": 241920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016260681491420953,
+      "loss": 2.1545,
+      "step": 241925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001626053736643339,
+      "loss": 2.0908,
+      "step": 241930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016260393239307113,
+      "loss": 2.0077,
+      "step": 241935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016260249110042175,
+      "loss": 2.0985,
+      "step": 241940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001626010497863862,
+      "loss": 2.2619,
+      "step": 241945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000162599608450965,
+      "loss": 2.1337,
+      "step": 241950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016259816709415862,
+      "loss": 2.1908,
+      "step": 241955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001625967257159676,
+      "loss": 2.1741,
+      "step": 241960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016259528431639242,
+      "loss": 2.0366,
+      "step": 241965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016259384289543354,
+      "loss": 1.9519,
+      "step": 241970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016259240145309142,
+      "loss": 2.208,
+      "step": 241975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016259095998936665,
+      "loss": 2.1157,
+      "step": 241980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016258951850425967,
+      "loss": 2.0659,
+      "step": 241985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016258807699777097,
+      "loss": 2.0572,
+      "step": 241990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016258663546990106,
+      "loss": 1.8807,
+      "step": 241995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016258519392065037,
+      "loss": 2.1837,
+      "step": 242000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016258375235001947,
+      "loss": 1.9683,
+      "step": 242005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016258231075800885,
+      "loss": 2.1674,
+      "step": 242010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016258086914461894,
+      "loss": 2.188,
+      "step": 242015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016257942750985032,
+      "loss": 2.1158,
+      "step": 242020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001625779858537034,
+      "loss": 2.1515,
+      "step": 242025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001625765441761787,
+      "loss": 2.2331,
+      "step": 242030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016257510247727673,
+      "loss": 2.1572,
+      "step": 242035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016257366075699795,
+      "loss": 2.1965,
+      "step": 242040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001625722190153429,
+      "loss": 2.0738,
+      "step": 242045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016257077725231206,
+      "loss": 2.0575,
+      "step": 242050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016256933546790584,
+      "loss": 2.0542,
+      "step": 242055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016256789366212484,
+      "loss": 2.0747,
+      "step": 242060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016256645183496953,
+      "loss": 2.0324,
+      "step": 242065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016256500998644037,
+      "loss": 2.0323,
+      "step": 242070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016256356811653787,
+      "loss": 2.122,
+      "step": 242075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016256212622526252,
+      "loss": 2.0975,
+      "step": 242080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001625606843126148,
+      "loss": 2.2552,
+      "step": 242085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016255924237859522,
+      "loss": 2.061,
+      "step": 242090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016255780042320427,
+      "loss": 2.0407,
+      "step": 242095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016255635844644246,
+      "loss": 2.1122,
+      "step": 242100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016255491644831028,
+      "loss": 1.8113,
+      "step": 242105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016255347442880816,
+      "loss": 2.1069,
+      "step": 242110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016255203238793665,
+      "loss": 2.1302,
+      "step": 242115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016255059032569622,
+      "loss": 2.1615,
+      "step": 242120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001625491482420874,
+      "loss": 2.0137,
+      "step": 242125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016254770613711064,
+      "loss": 2.0983,
+      "step": 242130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016254626401076644,
+      "loss": 2.1165,
+      "step": 242135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016254482186305532,
+      "loss": 2.128,
+      "step": 242140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016254337969397775,
+      "loss": 2.133,
+      "step": 242145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016254193750353424,
+      "loss": 2.0353,
+      "step": 242150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016254049529172528,
+      "loss": 1.9892,
+      "step": 242155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016253905305855132,
+      "loss": 2.0341,
+      "step": 242160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001625376108040129,
+      "loss": 2.2581,
+      "step": 242165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016253616852811046,
+      "loss": 1.8061,
+      "step": 242170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016253472623084457,
+      "loss": 2.0126,
+      "step": 242175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001625332839122157,
+      "loss": 2.0093,
+      "step": 242180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001625318415722243,
+      "loss": 2.1743,
+      "step": 242185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016253039921087088,
+      "loss": 2.1917,
+      "step": 242190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016252895682815596,
+      "loss": 2.2216,
+      "step": 242195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016252751442408,
+      "loss": 1.9263,
+      "step": 242200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016252607199864353,
+      "loss": 1.8795,
+      "step": 242205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000162524629551847,
+      "loss": 2.081,
+      "step": 242210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001625231870836909,
+      "loss": 2.1312,
+      "step": 242215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001625217445941758,
+      "loss": 2.2016,
+      "step": 242220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016252030208330212,
+      "loss": 2.0121,
+      "step": 242225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016251885955107035,
+      "loss": 2.196,
+      "step": 242230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016251741699748105,
+      "loss": 2.0703,
+      "step": 242235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016251597442253464,
+      "loss": 1.9404,
+      "step": 242240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001625145318262316,
+      "loss": 2.218,
+      "step": 242245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016251308920857248,
+      "loss": 2.1882,
+      "step": 242250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001625116465695578,
+      "loss": 2.1125,
+      "step": 242255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016251020390918797,
+      "loss": 2.1707,
+      "step": 242260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016250876122746353,
+      "loss": 2.1776,
+      "step": 242265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016250731852438498,
+      "loss": 2.102,
+      "step": 242270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00016250587579995276,
+      "loss": 2.1425,
+      "step": 242275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016250443305416742,
+      "loss": 1.8357,
+      "step": 242280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016250299028702943,
+      "loss": 1.9432,
+      "step": 242285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001625015474985393,
+      "loss": 1.9897,
+      "step": 242290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016250010468869746,
+      "loss": 1.9298,
+      "step": 242295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016249866185750447,
+      "loss": 2.1031,
+      "step": 242300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016249721900496084,
+      "loss": 2.2169,
+      "step": 242305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016249577613106702,
+      "loss": 2.2817,
+      "step": 242310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001624943332358235,
+      "loss": 2.2458,
+      "step": 242315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016249289031923075,
+      "loss": 2.1845,
+      "step": 242320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001624914473812893,
+      "loss": 2.052,
+      "step": 242325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016249000442199968,
+      "loss": 2.0592,
+      "step": 242330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016248856144136233,
+      "loss": 2.1323,
+      "step": 242335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016248711843937775,
+      "loss": 2.1365,
+      "step": 242340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016248567541604642,
+      "loss": 1.981,
+      "step": 242345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016248423237136885,
+      "loss": 2.2399,
+      "step": 242350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016248278930534556,
+      "loss": 2.0516,
+      "step": 242355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000162481346217977,
+      "loss": 2.0117,
+      "step": 242360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016247990310926368,
+      "loss": 2.2148,
+      "step": 242365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001624784599792061,
+      "loss": 2.0355,
+      "step": 242370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016247701682780473,
+      "loss": 2.0561,
+      "step": 242375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016247557365506006,
+      "loss": 2.1555,
+      "step": 242380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016247413046097264,
+      "loss": 2.1889,
+      "step": 242385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016247268724554293,
+      "loss": 1.8728,
+      "step": 242390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016247124400877138,
+      "loss": 2.1358,
+      "step": 242395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016246980075065854,
+      "loss": 2.0306,
+      "step": 242400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016246835747120488,
+      "loss": 1.928,
+      "step": 242405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001624669141704109,
+      "loss": 2.081,
+      "step": 242410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016246547084827707,
+      "loss": 2.1761,
+      "step": 242415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016246402750480393,
+      "loss": 2.1083,
+      "step": 242420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001624625841399919,
+      "loss": 1.8799,
+      "step": 242425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016246114075384158,
+      "loss": 2.0353,
+      "step": 242430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016245969734635338,
+      "loss": 1.941,
+      "step": 242435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001624582539175278,
+      "loss": 1.9624,
+      "step": 242440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016245681046736534,
+      "loss": 2.0283,
+      "step": 242445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016245536699586652,
+      "loss": 2.1343,
+      "step": 242450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016245392350303182,
+      "loss": 2.0604,
+      "step": 242455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001624524799888617,
+      "loss": 2.1539,
+      "step": 242460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001624510364533567,
+      "loss": 2.0564,
+      "step": 242465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016244959289651728,
+      "loss": 2.187,
+      "step": 242470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016244814931834397,
+      "loss": 2.2426,
+      "step": 242475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016244670571883725,
+      "loss": 2.2746,
+      "step": 242480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016244526209799754,
+      "loss": 2.0842,
+      "step": 242485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016244381845582546,
+      "loss": 1.9698,
+      "step": 242490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.0001624423747923214,
+      "loss": 2.0601,
+      "step": 242495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001624409311074859,
+      "loss": 2.139,
+      "step": 242500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016243948740131945,
+      "loss": 2.1183,
+      "step": 242505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016243804367382253,
+      "loss": 1.9259,
+      "step": 242510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016243659992499568,
+      "loss": 1.9696,
+      "step": 242515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001624351561548393,
+      "loss": 2.0051,
+      "step": 242520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016243371236335398,
+      "loss": 2.0817,
+      "step": 242525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016243226855054016,
+      "loss": 2.033,
+      "step": 242530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016243082471639837,
+      "loss": 2.2324,
+      "step": 242535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016242938086092902,
+      "loss": 2.115,
+      "step": 242540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001624279369841327,
+      "loss": 1.81,
+      "step": 242545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016242649308600987,
+      "loss": 2.0514,
+      "step": 242550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016242504916656102,
+      "loss": 2.2077,
+      "step": 242555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001624236052257866,
+      "loss": 2.319,
+      "step": 242560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016242216126368715,
+      "loss": 2.1112,
+      "step": 242565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001624207172802632,
+      "loss": 1.9317,
+      "step": 242570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016241927327551516,
+      "loss": 1.9253,
+      "step": 242575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001624178292494436,
+      "loss": 1.9451,
+      "step": 242580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016241638520204897,
+      "loss": 2.1469,
+      "step": 242585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016241494113333174,
+      "loss": 2.1879,
+      "step": 242590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016241349704329246,
+      "loss": 2.1158,
+      "step": 242595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001624120529319316,
+      "loss": 2.1432,
+      "step": 242600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016241060879924963,
+      "loss": 2.2005,
+      "step": 242605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016240916464524706,
+      "loss": 2.2013,
+      "step": 242610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016240772046992443,
+      "loss": 2.0787,
+      "step": 242615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016240627627328214,
+      "loss": 2.0622,
+      "step": 242620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016240483205532077,
+      "loss": 2.001,
+      "step": 242625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001624033878160408,
+      "loss": 2.1232,
+      "step": 242630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016240194355544266,
+      "loss": 2.0576,
+      "step": 242635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016240049927352686,
+      "loss": 2.2203,
+      "step": 242640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016239905497029394,
+      "loss": 2.157,
+      "step": 242645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001623976106457444,
+      "loss": 2.1065,
+      "step": 242650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016239616629987872,
+      "loss": 2.1397,
+      "step": 242655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001623947219326973,
+      "loss": 2.2181,
+      "step": 242660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016239327754420077,
+      "loss": 2.0974,
+      "step": 242665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016239183313438953,
+      "loss": 2.09,
+      "step": 242670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016239038870326415,
+      "loss": 2.0382,
+      "step": 242675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016238894425082506,
+      "loss": 2.0483,
+      "step": 242680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016238749977707278,
+      "loss": 2.0826,
+      "step": 242685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001623860552820078,
+      "loss": 2.0282,
+      "step": 242690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001623846107656306,
+      "loss": 2.1544,
+      "step": 242695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016238316622794167,
+      "loss": 2.2494,
+      "step": 242700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016238172166894155,
+      "loss": 2.1182,
+      "step": 242705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001623802770886307,
+      "loss": 2.1892,
+      "step": 242710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001623788324870096,
+      "loss": 2.1234,
+      "step": 242715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016237738786407878,
+      "loss": 2.1379,
+      "step": 242720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001623759432198387,
+      "loss": 2.0231,
+      "step": 242725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016237449855428988,
+      "loss": 2.1204,
+      "step": 242730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016237305386743276,
+      "loss": 2.0601,
+      "step": 242735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016237160915926792,
+      "loss": 2.0856,
+      "step": 242740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016237016442979577,
+      "loss": 2.0219,
+      "step": 242745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016236871967901685,
+      "loss": 1.9823,
+      "step": 242750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016236727490693168,
+      "loss": 1.9979,
+      "step": 242755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016236583011354069,
+      "loss": 1.9839,
+      "step": 242760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001623643852988444,
+      "loss": 2.1642,
+      "step": 242765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016236294046284333,
+      "loss": 2.01,
+      "step": 242770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016236149560553792,
+      "loss": 2.179,
+      "step": 242775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001623600507269287,
+      "loss": 2.1464,
+      "step": 242780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001623586058270161,
+      "loss": 2.154,
+      "step": 242785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016235716090580077,
+      "loss": 2.1262,
+      "step": 242790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016235571596328306,
+      "loss": 2.1153,
+      "step": 242795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001623542709994635,
+      "loss": 2.06,
+      "step": 242800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016235282601434258,
+      "loss": 2.0054,
+      "step": 242805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016235138100792083,
+      "loss": 2.0623,
+      "step": 242810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001623499359801987,
+      "loss": 2.1492,
+      "step": 242815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001623484909311767,
+      "loss": 2.1875,
+      "step": 242820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016234704586085534,
+      "loss": 2.0022,
+      "step": 242825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016234560076923508,
+      "loss": 2.1736,
+      "step": 242830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016234415565631645,
+      "loss": 2.0798,
+      "step": 242835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001623427105220999,
+      "loss": 2.0267,
+      "step": 242840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00016234126536658597,
+      "loss": 2.299,
+      "step": 242845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016233982018977512,
+      "loss": 2.0949,
+      "step": 242850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001623383749916679,
+      "loss": 1.9831,
+      "step": 242855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001623369297722647,
+      "loss": 2.0629,
+      "step": 242860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.0001623354845315661,
+      "loss": 2.0812,
+      "step": 242865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016233403926957257,
+      "loss": 2.0467,
+      "step": 242870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016233259398628461,
+      "loss": 1.9761,
+      "step": 242875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001623311486817027,
+      "loss": 2.3658,
+      "step": 242880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001623297033558273,
+      "loss": 1.9623,
+      "step": 242885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000162328258008659,
+      "loss": 2.1964,
+      "step": 242890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016232681264019823,
+      "loss": 2.2405,
+      "step": 242895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016232536725044547,
+      "loss": 1.9885,
+      "step": 242900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016232392183940125,
+      "loss": 2.0126,
+      "step": 242905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016232247640706601,
+      "loss": 1.9778,
+      "step": 242910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00016232103095344032,
+      "loss": 2.0736,
+      "step": 242915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016231958547852462,
+      "loss": 1.9959,
+      "step": 242920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016231813998231944,
+      "loss": 2.0487,
+      "step": 242925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016231669446482524,
+      "loss": 2.0028,
+      "step": 242930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016231524892604252,
+      "loss": 2.0135,
+      "step": 242935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016231380336597179,
+      "loss": 2.0432,
+      "step": 242940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001623123577846135,
+      "loss": 2.0405,
+      "step": 242945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016231091218196823,
+      "loss": 2.3028,
+      "step": 242950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001623094665580364,
+      "loss": 1.8949,
+      "step": 242955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016230802091281855,
+      "loss": 2.1375,
+      "step": 242960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001623065752463151,
+      "loss": 2.052,
+      "step": 242965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016230512955852663,
+      "loss": 2.1933,
+      "step": 242970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016230368384945358,
+      "loss": 2.0923,
+      "step": 242975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016230223811909648,
+      "loss": 2.0305,
+      "step": 242980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001623007923674558,
+      "loss": 1.9883,
+      "step": 242985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016229934659453202,
+      "loss": 2.0341,
+      "step": 242990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016229790080032568,
+      "loss": 1.9074,
+      "step": 242995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016229645498483722,
+      "loss": 2.1592,
+      "step": 243000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001622950091480672,
+      "loss": 2.22,
+      "step": 243005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016229356329001604,
+      "loss": 2.0933,
+      "step": 243010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016229211741068428,
+      "loss": 1.9407,
+      "step": 243015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001622906715100724,
+      "loss": 2.0646,
+      "step": 243020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001622892255881809,
+      "loss": 1.9113,
+      "step": 243025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016228777964501028,
+      "loss": 2.2016,
+      "step": 243030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000162286333680561,
+      "loss": 2.2557,
+      "step": 243035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001622848876948336,
+      "loss": 2.1419,
+      "step": 243040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016228344168782853,
+      "loss": 2.1504,
+      "step": 243045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016228199565954636,
+      "loss": 2.2832,
+      "step": 243050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016228054960998748,
+      "loss": 1.9306,
+      "step": 243055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016227910353915247,
+      "loss": 2.0384,
+      "step": 243060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016227765744704175,
+      "loss": 2.121,
+      "step": 243065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016227621133365586,
+      "loss": 2.0263,
+      "step": 243070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016227476519899533,
+      "loss": 1.9886,
+      "step": 243075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016227331904306055,
+      "loss": 2.2237,
+      "step": 243080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016227187286585213,
+      "loss": 1.9999,
+      "step": 243085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016227042666737046,
+      "loss": 2.1867,
+      "step": 243090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001622689804476161,
+      "loss": 2.1688,
+      "step": 243095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016226753420658954,
+      "loss": 1.8996,
+      "step": 243100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016226608794429125,
+      "loss": 2.1317,
+      "step": 243105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016226464166072178,
+      "loss": 2.2509,
+      "step": 243110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016226319535588154,
+      "loss": 2.1757,
+      "step": 243115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016226174902977104,
+      "loss": 2.1409,
+      "step": 243120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001622603026823908,
+      "loss": 2.2809,
+      "step": 243125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016225885631374135,
+      "loss": 2.0366,
+      "step": 243130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016225740992382312,
+      "loss": 2.1708,
+      "step": 243135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016225596351263666,
+      "loss": 1.923,
+      "step": 243140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001622545170801824,
+      "loss": 1.9399,
+      "step": 243145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016225307062646088,
+      "loss": 1.9361,
+      "step": 243150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016225162415147256,
+      "loss": 1.9725,
+      "step": 243155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016225017765521797,
+      "loss": 1.9891,
+      "step": 243160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001622487311376976,
+      "loss": 2.1264,
+      "step": 243165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016224728459891195,
+      "loss": 1.9731,
+      "step": 243170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016224583803886147,
+      "loss": 1.9731,
+      "step": 243175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016224439145754666,
+      "loss": 2.0929,
+      "step": 243180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001622429448549681,
+      "loss": 2.145,
+      "step": 243185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016224149823112618,
+      "loss": 2.0461,
+      "step": 243190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016224005158602144,
+      "loss": 2.1061,
+      "step": 243195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001622386049196544,
+      "loss": 1.9086,
+      "step": 243200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001622371582320255,
+      "loss": 2.2414,
+      "step": 243205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016223571152313525,
+      "loss": 2.1832,
+      "step": 243210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016223426479298415,
+      "loss": 1.9817,
+      "step": 243215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016223281804157271,
+      "loss": 2.0647,
+      "step": 243220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016223137126890143,
+      "loss": 2.0268,
+      "step": 243225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016222992447497073,
+      "loss": 2.1603,
+      "step": 243230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001622284776597812,
+      "loss": 2.0324,
+      "step": 243235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016222703082333326,
+      "loss": 1.877,
+      "step": 243240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016222558396562748,
+      "loss": 2.3329,
+      "step": 243245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001622241370866643,
+      "loss": 2.0018,
+      "step": 243250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016222269018644422,
+      "loss": 1.9975,
+      "step": 243255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016222124326496773,
+      "loss": 2.1787,
+      "step": 243260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016221979632223535,
+      "loss": 2.0101,
+      "step": 243265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016221834935824757,
+      "loss": 1.9894,
+      "step": 243270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016221690237300484,
+      "loss": 2.0704,
+      "step": 243275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016221545536650772,
+      "loss": 1.9397,
+      "step": 243280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.671875,
+      "learning_rate": 0.00016221400833875665,
+      "loss": 2.183,
+      "step": 243285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016221256128975216,
+      "loss": 1.9678,
+      "step": 243290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016221111421949473,
+      "loss": 2.0414,
+      "step": 243295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016220966712798484,
+      "loss": 2.0509,
+      "step": 243300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016220822001522303,
+      "loss": 1.9574,
+      "step": 243305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016220677288120974,
+      "loss": 2.1114,
+      "step": 243310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.25,
+      "learning_rate": 0.00016220532572594547,
+      "loss": 2.0657,
+      "step": 243315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016220387854943077,
+      "loss": 1.9917,
+      "step": 243320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016220243135166607,
+      "loss": 1.9491,
+      "step": 243325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001622009841326519,
+      "loss": 2.2193,
+      "step": 243330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016219953689238875,
+      "loss": 1.9876,
+      "step": 243335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001621980896308771,
+      "loss": 1.9722,
+      "step": 243340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016219664234811746,
+      "loss": 2.0596,
+      "step": 243345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016219519504411033,
+      "loss": 1.8539,
+      "step": 243350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016219374771885622,
+      "loss": 2.0212,
+      "step": 243355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016219230037235556,
+      "loss": 2.1319,
+      "step": 243360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001621908530046089,
+      "loss": 2.0707,
+      "step": 243365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016218940561561667,
+      "loss": 2.0203,
+      "step": 243370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016218795820537943,
+      "loss": 1.9577,
+      "step": 243375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001621865107738977,
+      "loss": 2.1335,
+      "step": 243380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001621850633211719,
+      "loss": 2.0427,
+      "step": 243385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016218361584720254,
+      "loss": 1.9967,
+      "step": 243390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016218216835199014,
+      "loss": 2.0986,
+      "step": 243395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016218072083553516,
+      "loss": 2.1421,
+      "step": 243400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016217927329783817,
+      "loss": 2.1897,
+      "step": 243405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001621778257388996,
+      "loss": 2.0803,
+      "step": 243410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00016217637815871993,
+      "loss": 2.0056,
+      "step": 243415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016217493055729966,
+      "loss": 2.0841,
+      "step": 243420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016217348293463938,
+      "loss": 1.8939,
+      "step": 243425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016217203529073947,
+      "loss": 2.2057,
+      "step": 243430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016217058762560047,
+      "loss": 2.0085,
+      "step": 243435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016216913993922287,
+      "loss": 2.1189,
+      "step": 243440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016216769223160714,
+      "loss": 2.2421,
+      "step": 243445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001621662445027538,
+      "loss": 1.9619,
+      "step": 243450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016216479675266337,
+      "loss": 2.1725,
+      "step": 243455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016216334898133634,
+      "loss": 2.2154,
+      "step": 243460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016216190118877318,
+      "loss": 2.111,
+      "step": 243465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016216045337497434,
+      "loss": 2.0166,
+      "step": 243470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016215900553994038,
+      "loss": 2.1515,
+      "step": 243475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016215755768367177,
+      "loss": 2.0156,
+      "step": 243480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000162156109806169,
+      "loss": 2.1528,
+      "step": 243485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016215466190743264,
+      "loss": 2.2043,
+      "step": 243490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001621532139874631,
+      "loss": 1.8991,
+      "step": 243495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00016215176604626082,
+      "loss": 2.1131,
+      "step": 243500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016215031808382643,
+      "loss": 2.3821,
+      "step": 243505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016214887010016036,
+      "loss": 2.3339,
+      "step": 243510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001621474220952631,
+      "loss": 1.9778,
+      "step": 243515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016214597406913515,
+      "loss": 1.9066,
+      "step": 243520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016214452602177702,
+      "loss": 2.1346,
+      "step": 243525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001621430779531892,
+      "loss": 2.0968,
+      "step": 243530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016214162986337216,
+      "loss": 2.1052,
+      "step": 243535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016214018175232645,
+      "loss": 1.916,
+      "step": 243540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016213873362005247,
+      "loss": 1.8045,
+      "step": 243545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001621372854665508,
+      "loss": 1.9866,
+      "step": 243550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016213583729182189,
+      "loss": 1.9016,
+      "step": 243555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001621343890958663,
+      "loss": 2.1208,
+      "step": 243560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016213294087868442,
+      "loss": 2.0597,
+      "step": 243565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016213149264027684,
+      "loss": 2.1534,
+      "step": 243570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016213004438064398,
+      "loss": 2.0574,
+      "step": 243575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016212859609978638,
+      "loss": 2.0418,
+      "step": 243580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016212714779770452,
+      "loss": 2.1283,
+      "step": 243585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016212569947439892,
+      "loss": 2.2303,
+      "step": 243590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016212425112987005,
+      "loss": 1.993,
+      "step": 243595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001621228027641184,
+      "loss": 2.0657,
+      "step": 243600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016212135437714448,
+      "loss": 1.9216,
+      "step": 243605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016211990596894877,
+      "loss": 2.0786,
+      "step": 243610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016211845753953178,
+      "loss": 1.9651,
+      "step": 243615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016211700908889403,
+      "loss": 2.0617,
+      "step": 243620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016211556061703592,
+      "loss": 1.9852,
+      "step": 243625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016211411212395805,
+      "loss": 2.0824,
+      "step": 243630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016211266360966086,
+      "loss": 2.2345,
+      "step": 243635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016211121507414487,
+      "loss": 2.1369,
+      "step": 243640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016210976651741054,
+      "loss": 2.0844,
+      "step": 243645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016210831793945843,
+      "loss": 2.184,
+      "step": 243650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016210686934028896,
+      "loss": 2.1063,
+      "step": 243655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016210542071990264,
+      "loss": 2.0166,
+      "step": 243660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001621039720783,
+      "loss": 1.9508,
+      "step": 243665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016210252341548152,
+      "loss": 2.0953,
+      "step": 243670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001621010747314477,
+      "loss": 2.0107,
+      "step": 243675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016209962602619903,
+      "loss": 2.1169,
+      "step": 243680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016209817729973596,
+      "loss": 2.2113,
+      "step": 243685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016209672855205907,
+      "loss": 1.9652,
+      "step": 243690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016209527978316879,
+      "loss": 2.0958,
+      "step": 243695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016209383099306565,
+      "loss": 2.0709,
+      "step": 243700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.0001620923821817501,
+      "loss": 1.9998,
+      "step": 243705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016209093334922269,
+      "loss": 1.936,
+      "step": 243710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016208948449548388,
+      "loss": 2.0817,
+      "step": 243715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001620880356205342,
+      "loss": 2.1067,
+      "step": 243720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016208658672437412,
+      "loss": 2.1491,
+      "step": 243725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001620851378070041,
+      "loss": 2.1145,
+      "step": 243730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016208368886842475,
+      "loss": 2.1538,
+      "step": 243735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001620822399086364,
+      "loss": 1.9922,
+      "step": 243740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016208079092763965,
+      "loss": 2.1032,
+      "step": 243745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000162079341925435,
+      "loss": 2.0743,
+      "step": 243750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016207789290202288,
+      "loss": 2.0773,
+      "step": 243755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00016207644385740389,
+      "loss": 2.0181,
+      "step": 243760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001620749947915784,
+      "loss": 2.0924,
+      "step": 243765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016207354570454697,
+      "loss": 2.1239,
+      "step": 243770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016207209659631016,
+      "loss": 2.2263,
+      "step": 243775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016207064746686834,
+      "loss": 2.0985,
+      "step": 243780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016206919831622204,
+      "loss": 2.1724,
+      "step": 243785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016206774914437184,
+      "loss": 2.2547,
+      "step": 243790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001620662999513181,
+      "loss": 2.0797,
+      "step": 243795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016206485073706146,
+      "loss": 2.1272,
+      "step": 243800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016206340150160228,
+      "loss": 2.0214,
+      "step": 243805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016206195224494115,
+      "loss": 1.9523,
+      "step": 243810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016206050296707853,
+      "loss": 2.0019,
+      "step": 243815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001620590536680149,
+      "loss": 2.121,
+      "step": 243820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016205760434775078,
+      "loss": 2.0177,
+      "step": 243825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016205615500628666,
+      "loss": 2.0049,
+      "step": 243830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016205470564362303,
+      "loss": 1.9818,
+      "step": 243835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001620532562597604,
+      "loss": 2.0863,
+      "step": 243840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016205180685469923,
+      "loss": 2.2781,
+      "step": 243845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016205035742844004,
+      "loss": 2.1818,
+      "step": 243850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016204890798098335,
+      "loss": 2.205,
+      "step": 243855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016204745851232957,
+      "loss": 2.0403,
+      "step": 243860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001620460090224793,
+      "loss": 2.1052,
+      "step": 243865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000162044559511433,
+      "loss": 2.0559,
+      "step": 243870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00016204310997919113,
+      "loss": 2.0854,
+      "step": 243875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001620416604257542,
+      "loss": 2.175,
+      "step": 243880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001620402108511227,
+      "loss": 2.1419,
+      "step": 243885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001620387612552972,
+      "loss": 1.9974,
+      "step": 243890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016203731163827808,
+      "loss": 1.8637,
+      "step": 243895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001620358620000659,
+      "loss": 2.0769,
+      "step": 243900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016203441234066116,
+      "loss": 2.0284,
+      "step": 243905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016203296266006434,
+      "loss": 2.0555,
+      "step": 243910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001620315129582759,
+      "loss": 2.2678,
+      "step": 243915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001620300632352964,
+      "loss": 2.1226,
+      "step": 243920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016202861349112632,
+      "loss": 2.0532,
+      "step": 243925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001620271637257661,
+      "loss": 2.0111,
+      "step": 243930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016202571393921628,
+      "loss": 2.0061,
+      "step": 243935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001620242641314774,
+      "loss": 2.1677,
+      "step": 243940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016202281430254987,
+      "loss": 2.1236,
+      "step": 243945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001620213644524342,
+      "loss": 2.1445,
+      "step": 243950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016201991458113097,
+      "loss": 2.2612,
+      "step": 243955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016201846468864056,
+      "loss": 2.0092,
+      "step": 243960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001620170147749635,
+      "loss": 2.0305,
+      "step": 243965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016201556484010036,
+      "loss": 2.1279,
+      "step": 243970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016201411488405156,
+      "loss": 1.9355,
+      "step": 243975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001620126649068176,
+      "loss": 2.2214,
+      "step": 243980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016201121490839898,
+      "loss": 2.1001,
+      "step": 243985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016200976488879625,
+      "loss": 2.0858,
+      "step": 243990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016200831484800983,
+      "loss": 2.1211,
+      "step": 243995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016200686478604027,
+      "loss": 2.1583,
+      "step": 244000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000162005414702888,
+      "loss": 2.2043,
+      "step": 244005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016200396459855357,
+      "loss": 2.1466,
+      "step": 244010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.00016200251447303745,
+      "loss": 2.1168,
+      "step": 244015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016200106432634018,
+      "loss": 1.9808,
+      "step": 244020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001619996141584622,
+      "loss": 2.0198,
+      "step": 244025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000161998163969404,
+      "loss": 1.9645,
+      "step": 244030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016199671375916616,
+      "loss": 2.0652,
+      "step": 244035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016199526352774908,
+      "loss": 2.0821,
+      "step": 244040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001619938132751533,
+      "loss": 2.0753,
+      "step": 244045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016199236300137933,
+      "loss": 2.0366,
+      "step": 244050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016199091270642762,
+      "loss": 2.1245,
+      "step": 244055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001619894623902987,
+      "loss": 2.1949,
+      "step": 244060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016198801205299302,
+      "loss": 1.9073,
+      "step": 244065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001619865616945112,
+      "loss": 2.2828,
+      "step": 244070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016198511131485356,
+      "loss": 1.9337,
+      "step": 244075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001619836609140207,
+      "loss": 2.0147,
+      "step": 244080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016198221049201314,
+      "loss": 2.2223,
+      "step": 244085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016198076004883126,
+      "loss": 2.0928,
+      "step": 244090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001619793095844757,
+      "loss": 2.0901,
+      "step": 244095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016197785909894685,
+      "loss": 2.2035,
+      "step": 244100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016197640859224525,
+      "loss": 2.1545,
+      "step": 244105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016197495806437139,
+      "loss": 2.0926,
+      "step": 244110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016197350751532573,
+      "loss": 2.0775,
+      "step": 244115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001619720569451088,
+      "loss": 2.1513,
+      "step": 244120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001619706063537211,
+      "loss": 2.0407,
+      "step": 244125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016196915574116316,
+      "loss": 2.2966,
+      "step": 244130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001619677051074354,
+      "loss": 2.3476,
+      "step": 244135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016196625445253835,
+      "loss": 2.1523,
+      "step": 244140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016196480377647247,
+      "loss": 2.1069,
+      "step": 244145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016196335307923834,
+      "loss": 2.1929,
+      "step": 244150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016196190236083638,
+      "loss": 2.1024,
+      "step": 244155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001619604516212671,
+      "loss": 2.2931,
+      "step": 244160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.000161959000860531,
+      "loss": 2.1808,
+      "step": 244165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001619575500786286,
+      "loss": 1.8822,
+      "step": 244170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001619560992755604,
+      "loss": 2.225,
+      "step": 244175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016195464845132683,
+      "loss": 2.186,
+      "step": 244180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016195319760592847,
+      "loss": 2.1865,
+      "step": 244185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016195174673936576,
+      "loss": 2.2539,
+      "step": 244190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001619502958516392,
+      "loss": 2.0785,
+      "step": 244195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016194884494274926,
+      "loss": 2.0313,
+      "step": 244200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001619473940126965,
+      "loss": 2.1133,
+      "step": 244205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016194594306148143,
+      "loss": 2.1939,
+      "step": 244210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016194449208910446,
+      "loss": 2.1998,
+      "step": 244215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016194304109556616,
+      "loss": 2.0206,
+      "step": 244220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016194159008086695,
+      "loss": 2.1475,
+      "step": 244225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001619401390450074,
+      "loss": 2.0961,
+      "step": 244230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016193868798798796,
+      "loss": 2.0551,
+      "step": 244235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016193723690980914,
+      "loss": 1.9822,
+      "step": 244240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016193578581047145,
+      "loss": 2.2031,
+      "step": 244245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016193433468997536,
+      "loss": 2.0945,
+      "step": 244250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016193288354832136,
+      "loss": 2.1204,
+      "step": 244255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016193143238551,
+      "loss": 2.0915,
+      "step": 244260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016192998120154173,
+      "loss": 2.0794,
+      "step": 244265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016192852999641705,
+      "loss": 2.273,
+      "step": 244270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016192707877013646,
+      "loss": 2.122,
+      "step": 244275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016192562752270043,
+      "loss": 1.9358,
+      "step": 244280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016192417625410954,
+      "loss": 1.8256,
+      "step": 244285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016192272496436417,
+      "loss": 2.1049,
+      "step": 244290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016192127365346494,
+      "loss": 2.2172,
+      "step": 244295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016191982232141222,
+      "loss": 2.0817,
+      "step": 244300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016191837096820658,
+      "loss": 2.0416,
+      "step": 244305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016191691959384854,
+      "loss": 2.142,
+      "step": 244310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016191546819833852,
+      "loss": 2.1264,
+      "step": 244315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016191401678167705,
+      "loss": 1.9854,
+      "step": 244320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016191256534386466,
+      "loss": 1.9003,
+      "step": 244325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001619111138849018,
+      "loss": 2.2064,
+      "step": 244330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016190966240478896,
+      "loss": 2.1195,
+      "step": 244335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016190821090352668,
+      "loss": 2.1748,
+      "step": 244340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001619067593811154,
+      "loss": 2.1547,
+      "step": 244345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001619053078375557,
+      "loss": 1.9497,
+      "step": 244350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016190385627284798,
+      "loss": 2.2321,
+      "step": 244355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001619024046869928,
+      "loss": 2.0406,
+      "step": 244360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016190095307999064,
+      "loss": 2.1328,
+      "step": 244365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016189950145184198,
+      "loss": 2.1921,
+      "step": 244370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016189804980254734,
+      "loss": 2.0696,
+      "step": 244375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016189659813210718,
+      "loss": 2.1017,
+      "step": 244380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016189514644052202,
+      "loss": 2.2345,
+      "step": 244385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001618936947277924,
+      "loss": 1.9192,
+      "step": 244390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016189224299391873,
+      "loss": 2.0996,
+      "step": 244395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016189079123890156,
+      "loss": 2.2466,
+      "step": 244400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016188933946274139,
+      "loss": 1.9729,
+      "step": 244405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016188788766543868,
+      "loss": 2.1324,
+      "step": 244410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016188643584699394,
+      "loss": 2.1552,
+      "step": 244415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016188498400740767,
+      "loss": 2.1038,
+      "step": 244420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016188353214668038,
+      "loss": 2.1589,
+      "step": 244425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016188208026481256,
+      "loss": 1.9728,
+      "step": 244430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001618806283618047,
+      "loss": 2.0987,
+      "step": 244435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016187917643765726,
+      "loss": 1.9886,
+      "step": 244440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001618777244923708,
+      "loss": 2.1912,
+      "step": 244445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016187627252594577,
+      "loss": 2.0024,
+      "step": 244450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001618748205383827,
+      "loss": 2.0312,
+      "step": 244455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016187336852968207,
+      "loss": 2.0619,
+      "step": 244460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016187191649984434,
+      "loss": 2.2478,
+      "step": 244465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001618704644488701,
+      "loss": 2.0945,
+      "step": 244470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016186901237675975,
+      "loss": 2.0433,
+      "step": 244475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016186756028351384,
+      "loss": 2.2363,
+      "step": 244480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016186610816913284,
+      "loss": 1.9911,
+      "step": 244485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016186465603361725,
+      "loss": 2.2077,
+      "step": 244490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016186320387696757,
+      "loss": 2.1612,
+      "step": 244495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001618617516991843,
+      "loss": 2.1976,
+      "step": 244500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016186029950026796,
+      "loss": 2.2004,
+      "step": 244505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000161858847280219,
+      "loss": 2.1969,
+      "step": 244510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016185739503903792,
+      "loss": 2.2674,
+      "step": 244515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016185594277672523,
+      "loss": 2.1393,
+      "step": 244520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016185449049328145,
+      "loss": 1.9235,
+      "step": 244525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016185303818870702,
+      "loss": 2.2119,
+      "step": 244530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016185158586300253,
+      "loss": 1.9484,
+      "step": 244535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001618501335161684,
+      "loss": 2.1013,
+      "step": 244540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001618486811482051,
+      "loss": 2.0704,
+      "step": 244545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001618472287591132,
+      "loss": 1.8286,
+      "step": 244550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016184577634889316,
+      "loss": 2.1974,
+      "step": 244555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001618443239175455,
+      "loss": 2.2088,
+      "step": 244560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016184287146507067,
+      "loss": 2.1875,
+      "step": 244565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001618414189914692,
+      "loss": 2.0851,
+      "step": 244570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016183996649674157,
+      "loss": 2.1193,
+      "step": 244575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016183851398088832,
+      "loss": 2.013,
+      "step": 244580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001618370614439099,
+      "loss": 2.0402,
+      "step": 244585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001618356088858068,
+      "loss": 2.0117,
+      "step": 244590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016183415630657955,
+      "loss": 1.9771,
+      "step": 244595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016183270370622862,
+      "loss": 2.0339,
+      "step": 244600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016183125108475453,
+      "loss": 1.9946,
+      "step": 244605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016182979844215774,
+      "loss": 1.9832,
+      "step": 244610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016182834577843877,
+      "loss": 2.2106,
+      "step": 244615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016182689309359813,
+      "loss": 2.184,
+      "step": 244620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001618254403876363,
+      "loss": 2.1178,
+      "step": 244625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001618239876605538,
+      "loss": 2.1153,
+      "step": 244630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001618225349123511,
+      "loss": 2.1642,
+      "step": 244635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016182108214302866,
+      "loss": 2.0679,
+      "step": 244640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016181962935258706,
+      "loss": 1.853,
+      "step": 244645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001618181765410267,
+      "loss": 2.0919,
+      "step": 244650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001618167237083482,
+      "loss": 2.0377,
+      "step": 244655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016181527085455194,
+      "loss": 1.9312,
+      "step": 244660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001618138179796385,
+      "loss": 1.9311,
+      "step": 244665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001618123650836083,
+      "loss": 1.9943,
+      "step": 244670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016181091216646188,
+      "loss": 1.9418,
+      "step": 244675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016180945922819975,
+      "loss": 2.0679,
+      "step": 244680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016180800626882234,
+      "loss": 2.0522,
+      "step": 244685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016180655328833026,
+      "loss": 2.1064,
+      "step": 244690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016180510028672394,
+      "loss": 1.9315,
+      "step": 244695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016180364726400384,
+      "loss": 2.1551,
+      "step": 244700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001618021942201705,
+      "loss": 2.0132,
+      "step": 244705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016180074115522438,
+      "loss": 2.0845,
+      "step": 244710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016179928806916606,
+      "loss": 2.0617,
+      "step": 244715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016179783496199596,
+      "loss": 2.0131,
+      "step": 244720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001617963818337146,
+      "loss": 1.8438,
+      "step": 244725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016179492868432247,
+      "loss": 2.1164,
+      "step": 244730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016179347551382008,
+      "loss": 2.0268,
+      "step": 244735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016179202232220792,
+      "loss": 2.221,
+      "step": 244740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016179056910948646,
+      "loss": 1.9749,
+      "step": 244745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016178911587565623,
+      "loss": 2.1116,
+      "step": 244750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016178766262071775,
+      "loss": 2.0811,
+      "step": 244755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016178620934467145,
+      "loss": 2.0612,
+      "step": 244760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016178475604751785,
+      "loss": 2.2229,
+      "step": 244765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001617833027292575,
+      "loss": 2.1681,
+      "step": 244770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00016178184938989084,
+      "loss": 2.2352,
+      "step": 244775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016178039602941837,
+      "loss": 2.2372,
+      "step": 244780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016177894264784063,
+      "loss": 2.1863,
+      "step": 244785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000161777489245158,
+      "loss": 2.0823,
+      "step": 244790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016177603582137113,
+      "loss": 2.0898,
+      "step": 244795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016177458237648043,
+      "loss": 2.1202,
+      "step": 244800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001617731289104864,
+      "loss": 2.1084,
+      "step": 244805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016177167542338956,
+      "loss": 2.1394,
+      "step": 244810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00016177022191519037,
+      "loss": 2.0114,
+      "step": 244815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016176876838588938,
+      "loss": 2.1974,
+      "step": 244820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016176731483548705,
+      "loss": 2.2128,
+      "step": 244825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001617658612639839,
+      "loss": 2.1429,
+      "step": 244830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001617644076713804,
+      "loss": 2.1899,
+      "step": 244835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016176295405767705,
+      "loss": 2.0824,
+      "step": 244840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016176150042287437,
+      "loss": 2.1317,
+      "step": 244845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016176004676697283,
+      "loss": 2.2175,
+      "step": 244850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016175859308997293,
+      "loss": 2.1298,
+      "step": 244855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001617571393918752,
+      "loss": 2.198,
+      "step": 244860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016175568567268012,
+      "loss": 1.9294,
+      "step": 244865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.90625,
+      "learning_rate": 0.00016175423193238813,
+      "loss": 2.1683,
+      "step": 244870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001617527781709998,
+      "loss": 2.1178,
+      "step": 244875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016175132438851558,
+      "loss": 2.1225,
+      "step": 244880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016174987058493602,
+      "loss": 2.0036,
+      "step": 244885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001617484167602616,
+      "loss": 2.0881,
+      "step": 244890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016174696291449274,
+      "loss": 2.1082,
+      "step": 244895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016174550904763003,
+      "loss": 2.0042,
+      "step": 244900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016174405515967394,
+      "loss": 2.2027,
+      "step": 244905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016174260125062495,
+      "loss": 2.0497,
+      "step": 244910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001617411473204836,
+      "loss": 2.1086,
+      "step": 244915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016173969336925033,
+      "loss": 2.0886,
+      "step": 244920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016173823939692565,
+      "loss": 2.0612,
+      "step": 244925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016173678540351008,
+      "loss": 2.0386,
+      "step": 244930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016173533138900408,
+      "loss": 2.0625,
+      "step": 244935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016173387735340822,
+      "loss": 2.0097,
+      "step": 244940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001617324232967229,
+      "loss": 2.0537,
+      "step": 244945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016173096921894868,
+      "loss": 2.2367,
+      "step": 244950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016172951512008607,
+      "loss": 2.3426,
+      "step": 244955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016172806100013548,
+      "loss": 2.1918,
+      "step": 244960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001617266068590975,
+      "loss": 2.1682,
+      "step": 244965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016172515269697262,
+      "loss": 2.0194,
+      "step": 244970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016172369851376125,
+      "loss": 2.1844,
+      "step": 244975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016172224430946398,
+      "loss": 1.9342,
+      "step": 244980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016172079008408127,
+      "loss": 2.0615,
+      "step": 244985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016171933583761363,
+      "loss": 2.2708,
+      "step": 244990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016171788157006155,
+      "loss": 2.1431,
+      "step": 244995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016171642728142547,
+      "loss": 2.1969,
+      "step": 245000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016171497297170597,
+      "loss": 2.0116,
+      "step": 245005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016171351864090351,
+      "loss": 2.2086,
+      "step": 245010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016171206428901863,
+      "loss": 2.0548,
+      "step": 245015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016171060991605177,
+      "loss": 2.1626,
+      "step": 245020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016170915552200343,
+      "loss": 1.9222,
+      "step": 245025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001617077011068741,
+      "loss": 1.9779,
+      "step": 245030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016170624667066435,
+      "loss": 2.1609,
+      "step": 245035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016170479221337464,
+      "loss": 2.0174,
+      "step": 245040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001617033377350054,
+      "loss": 2.0852,
+      "step": 245045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001617018832355572,
+      "loss": 2.0845,
+      "step": 245050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016170042871503052,
+      "loss": 2.2371,
+      "step": 245055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016169897417342584,
+      "loss": 1.9791,
+      "step": 245060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016169751961074368,
+      "loss": 2.1609,
+      "step": 245065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016169606502698455,
+      "loss": 2.2317,
+      "step": 245070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001616946104221489,
+      "loss": 1.937,
+      "step": 245075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016169315579623726,
+      "loss": 2.1228,
+      "step": 245080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016169170114925012,
+      "loss": 1.959,
+      "step": 245085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016169024648118798,
+      "loss": 2.0282,
+      "step": 245090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016168879179205133,
+      "loss": 2.1022,
+      "step": 245095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016168733708184067,
+      "loss": 2.1504,
+      "step": 245100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001616858823505565,
+      "loss": 2.2325,
+      "step": 245105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.0001616844275981993,
+      "loss": 2.0102,
+      "step": 245110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001616829728247696,
+      "loss": 2.0851,
+      "step": 245115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016168151803026788,
+      "loss": 1.9974,
+      "step": 245120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001616800632146946,
+      "loss": 2.0659,
+      "step": 245125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016167860837805034,
+      "loss": 2.2563,
+      "step": 245130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001616771535203355,
+      "loss": 2.158,
+      "step": 245135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016167569864155068,
+      "loss": 2.2661,
+      "step": 245140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016167424374169626,
+      "loss": 2.1785,
+      "step": 245145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016167278882077285,
+      "loss": 2.1581,
+      "step": 245150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016167133387878088,
+      "loss": 1.9351,
+      "step": 245155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016166987891572088,
+      "loss": 2.1496,
+      "step": 245160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001616684239315933,
+      "loss": 1.9553,
+      "step": 245165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016166696892639868,
+      "loss": 2.1932,
+      "step": 245170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016166551390013752,
+      "loss": 1.7865,
+      "step": 245175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001616640588528103,
+      "loss": 2.0491,
+      "step": 245180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001616626037844175,
+      "loss": 2.1157,
+      "step": 245185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016166114869495962,
+      "loss": 2.2046,
+      "step": 245190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016165969358443723,
+      "loss": 1.9869,
+      "step": 245195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016165823845285072,
+      "loss": 2.1504,
+      "step": 245200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016165678330020065,
+      "loss": 2.1088,
+      "step": 245205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001616553281264875,
+      "loss": 2.1813,
+      "step": 245210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001616538729317118,
+      "loss": 2.0558,
+      "step": 245215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000161652417715874,
+      "loss": 2.2341,
+      "step": 245220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001616509624789746,
+      "loss": 1.9652,
+      "step": 245225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016164950722101414,
+      "loss": 2.055,
+      "step": 245230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016164805194199307,
+      "loss": 2.0241,
+      "step": 245235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016164659664191192,
+      "loss": 2.0222,
+      "step": 245240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016164514132077117,
+      "loss": 2.0121,
+      "step": 245245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016164368597857136,
+      "loss": 2.104,
+      "step": 245250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001616422306153129,
+      "loss": 2.0366,
+      "step": 245255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016164077523099634,
+      "loss": 2.1436,
+      "step": 245260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016163931982562222,
+      "loss": 2.0681,
+      "step": 245265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016163786439919094,
+      "loss": 2.0829,
+      "step": 245270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00016163640895170304,
+      "loss": 2.0823,
+      "step": 245275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016163495348315904,
+      "loss": 1.9873,
+      "step": 245280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016163349799355945,
+      "loss": 2.0033,
+      "step": 245285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001616320424829047,
+      "loss": 1.9932,
+      "step": 245290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016163058695119534,
+      "loss": 2.324,
+      "step": 245295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016162913139843184,
+      "loss": 2.0862,
+      "step": 245300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016162767582461477,
+      "loss": 2.1456,
+      "step": 245305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016162622022974452,
+      "loss": 2.1234,
+      "step": 245310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016162476461382164,
+      "loss": 2.094,
+      "step": 245315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016162330897684664,
+      "loss": 2.1931,
+      "step": 245320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016162185331881996,
+      "loss": 2.0115,
+      "step": 245325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016162039763974217,
+      "loss": 2.3129,
+      "step": 245330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016161894193961372,
+      "loss": 2.2363,
+      "step": 245335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016161748621843514,
+      "loss": 2.0193,
+      "step": 245340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001616160304762069,
+      "loss": 2.1529,
+      "step": 245345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016161457471292948,
+      "loss": 1.9922,
+      "step": 245350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016161311892860344,
+      "loss": 2.0387,
+      "step": 245355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016161166312322921,
+      "loss": 2.0986,
+      "step": 245360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016161020729680737,
+      "loss": 2.2833,
+      "step": 245365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016160875144933834,
+      "loss": 2.1336,
+      "step": 245370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016160729558082262,
+      "loss": 2.3076,
+      "step": 245375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016160583969126075,
+      "loss": 2.1306,
+      "step": 245380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001616043837806532,
+      "loss": 1.8614,
+      "step": 245385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016160292784900051,
+      "loss": 2.0802,
+      "step": 245390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016160147189630308,
+      "loss": 2.1224,
+      "step": 245395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001616000159225615,
+      "loss": 2.1695,
+      "step": 245400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016159855992777623,
+      "loss": 1.8952,
+      "step": 245405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001615971039119478,
+      "loss": 2.1299,
+      "step": 245410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001615956478750767,
+      "loss": 1.8977,
+      "step": 245415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016159419181716335,
+      "loss": 2.1234,
+      "step": 245420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016159273573820835,
+      "loss": 2.0621,
+      "step": 245425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016159127963821213,
+      "loss": 2.0523,
+      "step": 245430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016158982351717523,
+      "loss": 1.9635,
+      "step": 245435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016158836737509811,
+      "loss": 2.1854,
+      "step": 245440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001615869112119813,
+      "loss": 2.2075,
+      "step": 245445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001615854550278253,
+      "loss": 2.2519,
+      "step": 245450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016158399882263056,
+      "loss": 2.079,
+      "step": 245455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016158254259639762,
+      "loss": 2.0821,
+      "step": 245460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016158108634912696,
+      "loss": 2.2457,
+      "step": 245465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016157963008081908,
+      "loss": 1.9545,
+      "step": 245470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016157817379147451,
+      "loss": 2.1714,
+      "step": 245475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016157671748109372,
+      "loss": 2.2045,
+      "step": 245480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016157526114967717,
+      "loss": 2.1908,
+      "step": 245485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016157380479722542,
+      "loss": 1.9578,
+      "step": 245490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016157234842373893,
+      "loss": 2.1834,
+      "step": 245495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016157089202921822,
+      "loss": 2.0547,
+      "step": 245500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016156943561366376,
+      "loss": 1.963,
+      "step": 245505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016156797917707606,
+      "loss": 1.948,
+      "step": 245510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016156652271945564,
+      "loss": 2.0811,
+      "step": 245515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016156506624080298,
+      "loss": 2.0465,
+      "step": 245520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016156360974111858,
+      "loss": 2.1993,
+      "step": 245525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016156215322040294,
+      "loss": 1.9701,
+      "step": 245530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016156069667865651,
+      "loss": 1.9402,
+      "step": 245535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016155924011587987,
+      "loss": 2.1364,
+      "step": 245540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001615577835320735,
+      "loss": 2.1232,
+      "step": 245545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016155632692723781,
+      "loss": 2.1437,
+      "step": 245550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001615548703013734,
+      "loss": 2.0292,
+      "step": 245555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016155341365448072,
+      "loss": 1.9886,
+      "step": 245560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016155195698656027,
+      "loss": 2.0481,
+      "step": 245565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016155050029761258,
+      "loss": 2.1079,
+      "step": 245570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.00016154904358763812,
+      "loss": 2.2481,
+      "step": 245575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016154758685663735,
+      "loss": 2.1309,
+      "step": 245580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016154613010461086,
+      "loss": 2.1996,
+      "step": 245585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016154467333155906,
+      "loss": 2.1353,
+      "step": 245590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001615432165374825,
+      "loss": 2.1509,
+      "step": 245595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016154175972238165,
+      "loss": 2.1992,
+      "step": 245600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016154030288625704,
+      "loss": 2.0396,
+      "step": 245605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016153884602910913,
+      "loss": 2.1624,
+      "step": 245610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016153738915093842,
+      "loss": 1.9738,
+      "step": 245615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016153593225174545,
+      "loss": 2.167,
+      "step": 245620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001615344753315307,
+      "loss": 2.1681,
+      "step": 245625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016153301839029463,
+      "loss": 2.1329,
+      "step": 245630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016153156142803776,
+      "loss": 2.1263,
+      "step": 245635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016153010444476063,
+      "loss": 2.1456,
+      "step": 245640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016152864744046363,
+      "loss": 2.1248,
+      "step": 245645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016152719041514737,
+      "loss": 2.0554,
+      "step": 245650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001615257333688123,
+      "loss": 2.0264,
+      "step": 245655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00016152427630145897,
+      "loss": 2.085,
+      "step": 245660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001615228192130878,
+      "loss": 1.9379,
+      "step": 245665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001615213621036993,
+      "loss": 2.1146,
+      "step": 245670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016151990497329397,
+      "loss": 2.0741,
+      "step": 245675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016151844782187237,
+      "loss": 2.1385,
+      "step": 245680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016151699064943494,
+      "loss": 1.9357,
+      "step": 245685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016151553345598217,
+      "loss": 2.044,
+      "step": 245690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001615140762415146,
+      "loss": 1.9619,
+      "step": 245695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001615126190060327,
+      "loss": 2.0911,
+      "step": 245700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016151116174953697,
+      "loss": 2.1128,
+      "step": 245705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016150970447202791,
+      "loss": 1.9678,
+      "step": 245710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016150824717350607,
+      "loss": 2.1697,
+      "step": 245715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016150678985397184,
+      "loss": 2.0203,
+      "step": 245720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016150533251342578,
+      "loss": 2.1254,
+      "step": 245725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001615038751518684,
+      "loss": 2.1905,
+      "step": 245730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016150241776930016,
+      "loss": 2.0769,
+      "step": 245735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001615009603657216,
+      "loss": 2.1355,
+      "step": 245740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016149950294113316,
+      "loss": 1.9965,
+      "step": 245745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001614980454955354,
+      "loss": 2.2588,
+      "step": 245750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016149658802892879,
+      "loss": 2.1492,
+      "step": 245755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016149513054131384,
+      "loss": 2.1526,
+      "step": 245760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016149367303269104,
+      "loss": 2.0189,
+      "step": 245765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016149221550306088,
+      "loss": 2.037,
+      "step": 245770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016149075795242386,
+      "loss": 2.14,
+      "step": 245775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016148930038078048,
+      "loss": 1.9081,
+      "step": 245780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016148784278813123,
+      "loss": 2.0503,
+      "step": 245785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016148638517447664,
+      "loss": 2.0687,
+      "step": 245790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016148492753981718,
+      "loss": 2.1589,
+      "step": 245795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016148346988415337,
+      "loss": 2.1751,
+      "step": 245800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016148201220748566,
+      "loss": 2.038,
+      "step": 245805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001614805545098146,
+      "loss": 2.1336,
+      "step": 245810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016147909679114063,
+      "loss": 2.1234,
+      "step": 245815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016147763905146435,
+      "loss": 1.9703,
+      "step": 245820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016147618129078615,
+      "loss": 2.1326,
+      "step": 245825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016147472350910657,
+      "loss": 2.2606,
+      "step": 245830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001614732657064261,
+      "loss": 2.0488,
+      "step": 245835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001614718078827453,
+      "loss": 2.0926,
+      "step": 245840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016147035003806457,
+      "loss": 2.1121,
+      "step": 245845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016146889217238445,
+      "loss": 2.219,
+      "step": 245850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016146743428570548,
+      "loss": 1.9972,
+      "step": 245855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016146597637802808,
+      "loss": 2.1783,
+      "step": 245860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001614645184493528,
+      "loss": 2.1919,
+      "step": 245865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016146306049968014,
+      "loss": 2.0654,
+      "step": 245870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001614616025290106,
+      "loss": 1.9791,
+      "step": 245875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016146014453734462,
+      "loss": 2.147,
+      "step": 245880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016145868652468274,
+      "loss": 2.0098,
+      "step": 245885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001614572284910255,
+      "loss": 2.0366,
+      "step": 245890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016145577043637333,
+      "loss": 2.0296,
+      "step": 245895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016145431236072675,
+      "loss": 2.2234,
+      "step": 245900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016145285426408625,
+      "loss": 2.1973,
+      "step": 245905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016145139614645238,
+      "loss": 2.2995,
+      "step": 245910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016144993800782555,
+      "loss": 2.1124,
+      "step": 245915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016144847984820635,
+      "loss": 2.1455,
+      "step": 245920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001614470216675952,
+      "loss": 2.187,
+      "step": 245925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016144556346599267,
+      "loss": 2.1805,
+      "step": 245930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001614441052433992,
+      "loss": 1.9228,
+      "step": 245935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001614426469998153,
+      "loss": 2.0185,
+      "step": 245940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001614411887352415,
+      "loss": 2.2561,
+      "step": 245945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016143973044967827,
+      "loss": 2.1862,
+      "step": 245950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001614382721431261,
+      "loss": 2.0879,
+      "step": 245955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001614368138155855,
+      "loss": 2.1666,
+      "step": 245960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000161435355467057,
+      "loss": 2.034,
+      "step": 245965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016143389709754105,
+      "loss": 2.1595,
+      "step": 245970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016143243870703815,
+      "loss": 2.115,
+      "step": 245975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016143098029554884,
+      "loss": 2.1472,
+      "step": 245980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016142952186307358,
+      "loss": 2.0569,
+      "step": 245985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001614280634096129,
+      "loss": 2.0236,
+      "step": 245990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016142660493516725,
+      "loss": 2.0646,
+      "step": 245995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016142514643973716,
+      "loss": 2.1793,
+      "step": 246000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016142368792332315,
+      "loss": 2.122,
+      "step": 246005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001614222293859257,
+      "loss": 2.0557,
+      "step": 246010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016142077082754528,
+      "loss": 2.1116,
+      "step": 246015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001614193122481824,
+      "loss": 2.2457,
+      "step": 246020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001614178536478376,
+      "loss": 2.0383,
+      "step": 246025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001614163950265113,
+      "loss": 2.3221,
+      "step": 246030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016141493638420409,
+      "loss": 1.9299,
+      "step": 246035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001614134777209164,
+      "loss": 2.1349,
+      "step": 246040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016141201903664877,
+      "loss": 2.2228,
+      "step": 246045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016141056033140168,
+      "loss": 2.1134,
+      "step": 246050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001614091016051756,
+      "loss": 2.1493,
+      "step": 246055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016140764285797111,
+      "loss": 2.0756,
+      "step": 246060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001614061840897886,
+      "loss": 2.1205,
+      "step": 246065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001614047253006287,
+      "loss": 2.1553,
+      "step": 246070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016140326649049174,
+      "loss": 2.3397,
+      "step": 246075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016140180765937835,
+      "loss": 2.0659,
+      "step": 246080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000161400348807289,
+      "loss": 2.2222,
+      "step": 246085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016139888993422417,
+      "loss": 2.1499,
+      "step": 246090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016139743104018437,
+      "loss": 2.062,
+      "step": 246095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016139597212517007,
+      "loss": 2.0593,
+      "step": 246100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016139451318918183,
+      "loss": 2.1024,
+      "step": 246105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016139305423222008,
+      "loss": 2.0931,
+      "step": 246110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016139159525428538,
+      "loss": 2.123,
+      "step": 246115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016139013625537818,
+      "loss": 2.1535,
+      "step": 246120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016138867723549898,
+      "loss": 2.211,
+      "step": 246125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001613872181946483,
+      "loss": 2.1362,
+      "step": 246130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016138575913282664,
+      "loss": 2.0803,
+      "step": 246135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001613843000500345,
+      "loss": 2.1387,
+      "step": 246140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016138284094627237,
+      "loss": 2.126,
+      "step": 246145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016138138182154074,
+      "loss": 2.0505,
+      "step": 246150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016137992267584012,
+      "loss": 2.0681,
+      "step": 246155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016137846350917099,
+      "loss": 2.199,
+      "step": 246160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001613770043215339,
+      "loss": 2.1982,
+      "step": 246165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001613755451129293,
+      "loss": 1.9706,
+      "step": 246170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016137408588335768,
+      "loss": 2.1779,
+      "step": 246175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016137262663281957,
+      "loss": 2.1145,
+      "step": 246180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016137116736131545,
+      "loss": 2.0188,
+      "step": 246185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016136970806884584,
+      "loss": 2.0496,
+      "step": 246190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001613682487554112,
+      "loss": 2.0513,
+      "step": 246195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016136678942101207,
+      "loss": 1.9593,
+      "step": 246200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016136533006564892,
+      "loss": 1.9083,
+      "step": 246205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016136387068932227,
+      "loss": 2.0939,
+      "step": 246210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016136241129203262,
+      "loss": 2.3025,
+      "step": 246215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016136095187378047,
+      "loss": 2.0323,
+      "step": 246220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016135949243456627,
+      "loss": 2.166,
+      "step": 246225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001613580329743906,
+      "loss": 2.0957,
+      "step": 246230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016135657349325387,
+      "loss": 2.1152,
+      "step": 246235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016135511399115662,
+      "loss": 2.2261,
+      "step": 246240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016135365446809936,
+      "loss": 2.1811,
+      "step": 246245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001613521949240826,
+      "loss": 2.2211,
+      "step": 246250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001613507353591068,
+      "loss": 2.1396,
+      "step": 246255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016134927577317248,
+      "loss": 2.1177,
+      "step": 246260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016134781616628012,
+      "loss": 2.1363,
+      "step": 246265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016134635653843025,
+      "loss": 2.2313,
+      "step": 246270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016134489688962337,
+      "loss": 2.1739,
+      "step": 246275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016134343721985994,
+      "loss": 1.9735,
+      "step": 246280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016134197752914046,
+      "loss": 2.0005,
+      "step": 246285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016134051781746544,
+      "loss": 2.329,
+      "step": 246290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016133905808483543,
+      "loss": 2.2063,
+      "step": 246295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001613375983312509,
+      "loss": 2.0296,
+      "step": 246300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001613361385567123,
+      "loss": 2.1675,
+      "step": 246305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016133467876122015,
+      "loss": 2.0395,
+      "step": 246310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016133321894477498,
+      "loss": 2.1656,
+      "step": 246315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016133175910737726,
+      "loss": 1.9797,
+      "step": 246320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001613302992490275,
+      "loss": 2.1696,
+      "step": 246325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001613288393697262,
+      "loss": 2.0304,
+      "step": 246330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016132737946947384,
+      "loss": 2.095,
+      "step": 246335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016132591954827097,
+      "loss": 2.0681,
+      "step": 246340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016132445960611805,
+      "loss": 2.0584,
+      "step": 246345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016132299964301557,
+      "loss": 2.0948,
+      "step": 246350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016132153965896405,
+      "loss": 1.9852,
+      "step": 246355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016132007965396396,
+      "loss": 2.0895,
+      "step": 246360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001613186196280158,
+      "loss": 1.9703,
+      "step": 246365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016131715958112012,
+      "loss": 1.9652,
+      "step": 246370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016131569951327742,
+      "loss": 2.1219,
+      "step": 246375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016131423942448812,
+      "loss": 2.0528,
+      "step": 246380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016131277931475278,
+      "loss": 2.0828,
+      "step": 246385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016131131918407186,
+      "loss": 2.1399,
+      "step": 246390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001613098590324459,
+      "loss": 2.0998,
+      "step": 246395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016130839885987537,
+      "loss": 2.1228,
+      "step": 246400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016130693866636076,
+      "loss": 1.9298,
+      "step": 246405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016130547845190262,
+      "loss": 2.0454,
+      "step": 246410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016130401821650144,
+      "loss": 2.0709,
+      "step": 246415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016130255796015765,
+      "loss": 2.0569,
+      "step": 246420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.0001613010976828718,
+      "loss": 2.2877,
+      "step": 246425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001612996373846444,
+      "loss": 2.0293,
+      "step": 246430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016129817706547593,
+      "loss": 2.0176,
+      "step": 246435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001612967167253669,
+      "loss": 2.0852,
+      "step": 246440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001612952563643178,
+      "loss": 2.1889,
+      "step": 246445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016129379598232908,
+      "loss": 1.9743,
+      "step": 246450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016129233557940134,
+      "loss": 2.1706,
+      "step": 246455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.000161290875155535,
+      "loss": 2.1751,
+      "step": 246460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001612894147107306,
+      "loss": 1.9603,
+      "step": 246465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001612879542449886,
+      "loss": 2.0959,
+      "step": 246470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016128649375830956,
+      "loss": 2.0088,
+      "step": 246475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016128503325069392,
+      "loss": 2.0685,
+      "step": 246480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001612835727221422,
+      "loss": 2.0779,
+      "step": 246485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016128211217265493,
+      "loss": 2.1524,
+      "step": 246490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016128065160223254,
+      "loss": 2.0307,
+      "step": 246495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016127919101087559,
+      "loss": 2.0501,
+      "step": 246500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016127773039858453,
+      "loss": 2.0376,
+      "step": 246505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016127626976535992,
+      "loss": 2.2103,
+      "step": 246510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.0001612748091112022,
+      "loss": 2.264,
+      "step": 246515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016127334843611194,
+      "loss": 2.0401,
+      "step": 246520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016127188774008955,
+      "loss": 2.0886,
+      "step": 246525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016127042702313555,
+      "loss": 2.1917,
+      "step": 246530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001612689662852505,
+      "loss": 2.1701,
+      "step": 246535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016126750552643485,
+      "loss": 2.1387,
+      "step": 246540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016126604474668913,
+      "loss": 2.1697,
+      "step": 246545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016126458394601377,
+      "loss": 2.182,
+      "step": 246550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00016126312312440935,
+      "loss": 2.3505,
+      "step": 246555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016126166228187634,
+      "loss": 2.2176,
+      "step": 246560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016126020141841522,
+      "loss": 2.0826,
+      "step": 246565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001612587405340265,
+      "loss": 2.1347,
+      "step": 246570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016125727962871072,
+      "loss": 2.2283,
+      "step": 246575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001612558187024683,
+      "loss": 2.108,
+      "step": 246580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001612543577552998,
+      "loss": 2.1674,
+      "step": 246585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001612528967872057,
+      "loss": 2.1882,
+      "step": 246590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001612514357981865,
+      "loss": 2.1483,
+      "step": 246595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001612499747882427,
+      "loss": 2.0994,
+      "step": 246600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001612485137573748,
+      "loss": 2.1155,
+      "step": 246605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016124705270558328,
+      "loss": 2.0118,
+      "step": 246610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016124559163286866,
+      "loss": 2.0018,
+      "step": 246615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016124413053923142,
+      "loss": 2.1728,
+      "step": 246620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016124266942467213,
+      "loss": 2.0487,
+      "step": 246625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016124120828919122,
+      "loss": 2.0021,
+      "step": 246630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016123974713278915,
+      "loss": 2.0051,
+      "step": 246635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016123828595546652,
+      "loss": 1.9951,
+      "step": 246640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016123682475722374,
+      "loss": 2.3749,
+      "step": 246645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016123536353806136,
+      "loss": 2.3174,
+      "step": 246650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001612339022979799,
+      "loss": 2.1442,
+      "step": 246655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016123244103697984,
+      "loss": 2.1237,
+      "step": 246660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001612309797550616,
+      "loss": 2.0505,
+      "step": 246665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001612295184522258,
+      "loss": 2.1114,
+      "step": 246670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016122805712847287,
+      "loss": 2.2151,
+      "step": 246675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016122659578380333,
+      "loss": 1.9324,
+      "step": 246680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016122513441821766,
+      "loss": 2.081,
+      "step": 246685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001612236730317164,
+      "loss": 1.9975,
+      "step": 246690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001612222116243,
+      "loss": 2.2418,
+      "step": 246695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016122075019596902,
+      "loss": 2.1506,
+      "step": 246700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016121928874672387,
+      "loss": 2.1867,
+      "step": 246705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001612178272765651,
+      "loss": 2.1026,
+      "step": 246710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016121636578549325,
+      "loss": 2.2934,
+      "step": 246715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016121490427350877,
+      "loss": 2.139,
+      "step": 246720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016121344274061216,
+      "loss": 2.0483,
+      "step": 246725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016121198118680395,
+      "loss": 2.1374,
+      "step": 246730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016121051961208457,
+      "loss": 2.0794,
+      "step": 246735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016120905801645462,
+      "loss": 2.2831,
+      "step": 246740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001612075963999145,
+      "loss": 2.0268,
+      "step": 246745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016120613476246475,
+      "loss": 1.8638,
+      "step": 246750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016120467310410592,
+      "loss": 2.1987,
+      "step": 246755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016120321142483845,
+      "loss": 1.9756,
+      "step": 246760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016120174972466284,
+      "loss": 2.2819,
+      "step": 246765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001612002880035796,
+      "loss": 1.9758,
+      "step": 246770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016119882626158924,
+      "loss": 2.0384,
+      "step": 246775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00016119736449869224,
+      "loss": 2.214,
+      "step": 246780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016119590271488913,
+      "loss": 2.1647,
+      "step": 246785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016119444091018038,
+      "loss": 2.0669,
+      "step": 246790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00016119297908456648,
+      "loss": 1.9754,
+      "step": 246795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016119151723804797,
+      "loss": 2.1502,
+      "step": 246800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016119005537062531,
+      "loss": 2.2444,
+      "step": 246805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016118859348229907,
+      "loss": 2.1128,
+      "step": 246810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016118713157306966,
+      "loss": 2.0786,
+      "step": 246815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016118566964293759,
+      "loss": 2.0698,
+      "step": 246820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016118420769190342,
+      "loss": 2.0863,
+      "step": 246825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001611827457199676,
+      "loss": 2.1351,
+      "step": 246830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016118128372713067,
+      "loss": 2.2094,
+      "step": 246835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016117982171339309,
+      "loss": 1.9099,
+      "step": 246840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016117835967875536,
+      "loss": 1.9199,
+      "step": 246845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016117689762321799,
+      "loss": 2.2062,
+      "step": 246850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001611754355467815,
+      "loss": 2.1393,
+      "step": 246855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016117397344944635,
+      "loss": 2.2367,
+      "step": 246860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016117251133121308,
+      "loss": 2.0782,
+      "step": 246865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001611710491920822,
+      "loss": 2.1023,
+      "step": 246870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016116958703205414,
+      "loss": 2.1894,
+      "step": 246875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016116812485112943,
+      "loss": 2.1254,
+      "step": 246880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001611666626493086,
+      "loss": 2.0242,
+      "step": 246885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016116520042659213,
+      "loss": 2.2314,
+      "step": 246890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016116373818298048,
+      "loss": 2.1352,
+      "step": 246895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016116227591847423,
+      "loss": 2.2931,
+      "step": 246900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016116081363307382,
+      "loss": 2.1227,
+      "step": 246905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016115935132677978,
+      "loss": 2.0837,
+      "step": 246910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001611578889995926,
+      "loss": 2.0367,
+      "step": 246915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016115642665151278,
+      "loss": 2.0638,
+      "step": 246920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001611549642825408,
+      "loss": 2.2167,
+      "step": 246925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016115350189267716,
+      "loss": 2.2555,
+      "step": 246930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016115203948192242,
+      "loss": 2.1266,
+      "step": 246935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000161150577050277,
+      "loss": 2.147,
+      "step": 246940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016114911459774142,
+      "loss": 2.1691,
+      "step": 246945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001611476521243162,
+      "loss": 2.3334,
+      "step": 246950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016114618963000186,
+      "loss": 2.2862,
+      "step": 246955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016114472711479887,
+      "loss": 2.0922,
+      "step": 246960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00016114326457870772,
+      "loss": 1.987,
+      "step": 246965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016114180202172893,
+      "loss": 2.0348,
+      "step": 246970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016114033944386298,
+      "loss": 2.0774,
+      "step": 246975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001611388768451104,
+      "loss": 2.2096,
+      "step": 246980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016113741422547163,
+      "loss": 2.0124,
+      "step": 246985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016113595158494723,
+      "loss": 2.1586,
+      "step": 246990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001611344889235377,
+      "loss": 2.0891,
+      "step": 246995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016113302624124352,
+      "loss": 2.1283,
+      "step": 247000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016113156353806518,
+      "loss": 2.165,
+      "step": 247005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016113010081400317,
+      "loss": 2.1124,
+      "step": 247010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00016112863806905803,
+      "loss": 2.0618,
+      "step": 247015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016112717530323022,
+      "loss": 2.0798,
+      "step": 247020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016112571251652029,
+      "loss": 2.045,
+      "step": 247025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001611242497089287,
+      "loss": 2.0983,
+      "step": 247030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016112278688045596,
+      "loss": 1.8506,
+      "step": 247035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001611213240311025,
+      "loss": 2.1618,
+      "step": 247040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000161119861160869,
+      "loss": 2.2042,
+      "step": 247045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016111839826975577,
+      "loss": 2.0715,
+      "step": 247050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016111693535776343,
+      "loss": 2.0455,
+      "step": 247055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001611154724248924,
+      "loss": 1.9512,
+      "step": 247060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001611140094711432,
+      "loss": 2.0474,
+      "step": 247065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001611125464965164,
+      "loss": 2.0633,
+      "step": 247070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016111108350101242,
+      "loss": 2.0703,
+      "step": 247075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001611096204846318,
+      "loss": 1.9434,
+      "step": 247080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000161108157447375,
+      "loss": 2.0264,
+      "step": 247085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016110669438924253,
+      "loss": 2.068,
+      "step": 247090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016110523131023495,
+      "loss": 2.0439,
+      "step": 247095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001611037682103527,
+      "loss": 2.0038,
+      "step": 247100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001611023050895963,
+      "loss": 1.8301,
+      "step": 247105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016110084194796622,
+      "loss": 2.1414,
+      "step": 247110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000161099378785463,
+      "loss": 2.0813,
+      "step": 247115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016109791560208711,
+      "loss": 2.0878,
+      "step": 247120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001610964523978391,
+      "loss": 2.4067,
+      "step": 247125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001610949891727194,
+      "loss": 1.9606,
+      "step": 247130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016109352592672858,
+      "loss": 2.2623,
+      "step": 247135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016109206265986706,
+      "loss": 2.0056,
+      "step": 247140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001610905993721354,
+      "loss": 2.1425,
+      "step": 247145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016108913606353409,
+      "loss": 2.0653,
+      "step": 247150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.25,
+      "learning_rate": 0.0001610876727340636,
+      "loss": 2.0839,
+      "step": 247155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001610862093837245,
+      "loss": 1.9244,
+      "step": 247160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001610847460125172,
+      "loss": 2.0724,
+      "step": 247165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016108328262044226,
+      "loss": 2.1587,
+      "step": 247170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016108181920750016,
+      "loss": 2.1854,
+      "step": 247175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016108035577369143,
+      "loss": 2.0331,
+      "step": 247180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001610788923190165,
+      "loss": 1.9785,
+      "step": 247185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016107742884347595,
+      "loss": 2.184,
+      "step": 247190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001610759653470702,
+      "loss": 2.1114,
+      "step": 247195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001610745018297998,
+      "loss": 2.1486,
+      "step": 247200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016107303829166527,
+      "loss": 2.0008,
+      "step": 247205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016107157473266707,
+      "loss": 1.8922,
+      "step": 247210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016107011115280574,
+      "loss": 2.2083,
+      "step": 247215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001610686475520817,
+      "loss": 1.9015,
+      "step": 247220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016106718393049554,
+      "loss": 2.0233,
+      "step": 247225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016106572028804766,
+      "loss": 2.0449,
+      "step": 247230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016106425662473873,
+      "loss": 2.173,
+      "step": 247235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016106279294056906,
+      "loss": 2.0917,
+      "step": 247240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016106132923553927,
+      "loss": 1.993,
+      "step": 247245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001610598655096498,
+      "loss": 2.1475,
+      "step": 247250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016105840176290116,
+      "loss": 1.7746,
+      "step": 247255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016105693799529387,
+      "loss": 2.1879,
+      "step": 247260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016105547420682846,
+      "loss": 2.2091,
+      "step": 247265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001610540103975054,
+      "loss": 2.1778,
+      "step": 247270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016105254656732513,
+      "loss": 2.035,
+      "step": 247275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016105108271628825,
+      "loss": 2.0866,
+      "step": 247280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016104961884439516,
+      "loss": 2.256,
+      "step": 247285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016104815495164644,
+      "loss": 2.0897,
+      "step": 247290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016104669103804257,
+      "loss": 2.173,
+      "step": 247295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016104522710358405,
+      "loss": 2.1617,
+      "step": 247300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016104376314827134,
+      "loss": 1.9677,
+      "step": 247305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000161042299172105,
+      "loss": 2.3106,
+      "step": 247310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001610408351750855,
+      "loss": 2.1628,
+      "step": 247315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016103937115721334,
+      "loss": 1.9631,
+      "step": 247320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016103790711848902,
+      "loss": 2.0536,
+      "step": 247325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016103644305891308,
+      "loss": 2.2338,
+      "step": 247330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016103497897848593,
+      "loss": 2.0575,
+      "step": 247335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016103351487720816,
+      "loss": 1.9867,
+      "step": 247340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001610320507550802,
+      "loss": 1.9875,
+      "step": 247345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016103058661210264,
+      "loss": 2.0641,
+      "step": 247350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001610291224482759,
+      "loss": 2.0937,
+      "step": 247355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016102765826360048,
+      "loss": 1.9886,
+      "step": 247360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001610261940580769,
+      "loss": 2.0352,
+      "step": 247365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016102472983170569,
+      "loss": 2.2285,
+      "step": 247370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00016102326558448734,
+      "loss": 1.9859,
+      "step": 247375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001610218013164223,
+      "loss": 2.1853,
+      "step": 247380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016102033702751112,
+      "loss": 1.9668,
+      "step": 247385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016101887271775431,
+      "loss": 2.0374,
+      "step": 247390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016101740838715234,
+      "loss": 2.2322,
+      "step": 247395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016101594403570568,
+      "loss": 2.2458,
+      "step": 247400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001610144796634149,
+      "loss": 2.0434,
+      "step": 247405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016101301527028045,
+      "loss": 2.112,
+      "step": 247410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016101155085630284,
+      "loss": 2.1233,
+      "step": 247415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001610100864214826,
+      "loss": 2.1014,
+      "step": 247420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001610086219658202,
+      "loss": 2.118,
+      "step": 247425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016100715748931617,
+      "loss": 2.2006,
+      "step": 247430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016100569299197094,
+      "loss": 2.1531,
+      "step": 247435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001610042284737851,
+      "loss": 2.1201,
+      "step": 247440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001610027639347591,
+      "loss": 2.1494,
+      "step": 247445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016100129937489345,
+      "loss": 2.0908,
+      "step": 247450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016099983479418866,
+      "loss": 2.2122,
+      "step": 247455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016099837019264518,
+      "loss": 2.1997,
+      "step": 247460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016099690557026358,
+      "loss": 2.1294,
+      "step": 247465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001609954409270443,
+      "loss": 2.2915,
+      "step": 247470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001609939762629879,
+      "loss": 2.1293,
+      "step": 247475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016099251157809487,
+      "loss": 2.0805,
+      "step": 247480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016099104687236567,
+      "loss": 2.043,
+      "step": 247485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016098958214580083,
+      "loss": 1.8556,
+      "step": 247490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016098811739840083,
+      "loss": 1.9687,
+      "step": 247495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001609866526301662,
+      "loss": 2.0977,
+      "step": 247500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016098518784109743,
+      "loss": 2.0405,
+      "step": 247505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000160983723031195,
+      "loss": 2.1969,
+      "step": 247510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016098225820045938,
+      "loss": 2.1512,
+      "step": 247515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016098079334889116,
+      "loss": 2.01,
+      "step": 247520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001609793284764908,
+      "loss": 1.9659,
+      "step": 247525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001609778635832588,
+      "loss": 2.132,
+      "step": 247530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016097639866919562,
+      "loss": 1.9157,
+      "step": 247535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016097493373430183,
+      "loss": 2.0721,
+      "step": 247540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016097346877857785,
+      "loss": 2.0258,
+      "step": 247545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001609720038020243,
+      "loss": 2.1161,
+      "step": 247550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016097053880464155,
+      "loss": 1.985,
+      "step": 247555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016096907378643018,
+      "loss": 1.9653,
+      "step": 247560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016096760874739066,
+      "loss": 2.102,
+      "step": 247565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001609661436875235,
+      "loss": 2.277,
+      "step": 247570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016096467860682922,
+      "loss": 2.1286,
+      "step": 247575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001609632135053083,
+      "loss": 2.0481,
+      "step": 247580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001609617483829612,
+      "loss": 2.106,
+      "step": 247585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001609602832397885,
+      "loss": 1.9815,
+      "step": 247590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016095881807579065,
+      "loss": 2.13,
+      "step": 247595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016095735289096817,
+      "loss": 2.2701,
+      "step": 247600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016095588768532152,
+      "loss": 1.9719,
+      "step": 247605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016095442245885129,
+      "loss": 2.1195,
+      "step": 247610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001609529572115579,
+      "loss": 2.1945,
+      "step": 247615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00016095149194344185,
+      "loss": 1.8866,
+      "step": 247620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001609500266545037,
+      "loss": 2.0452,
+      "step": 247625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016094856134474386,
+      "loss": 2.1501,
+      "step": 247630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016094709601416294,
+      "loss": 2.1752,
+      "step": 247635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016094563066276134,
+      "loss": 1.9725,
+      "step": 247640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016094416529053966,
+      "loss": 1.7923,
+      "step": 247645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016094269989749832,
+      "loss": 2.1688,
+      "step": 247650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016094123448363786,
+      "loss": 2.1611,
+      "step": 247655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016093976904895874,
+      "loss": 2.2568,
+      "step": 247660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016093830359346152,
+      "loss": 2.1338,
+      "step": 247665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016093683811714665,
+      "loss": 1.9022,
+      "step": 247670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016093537262001467,
+      "loss": 2.0523,
+      "step": 247675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016093390710206603,
+      "loss": 2.003,
+      "step": 247680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001609324415633013,
+      "loss": 2.0054,
+      "step": 247685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016093097600372092,
+      "loss": 2.2685,
+      "step": 247690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016092951042332544,
+      "loss": 2.0196,
+      "step": 247695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016092804482211533,
+      "loss": 2.1148,
+      "step": 247700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016092657920009108,
+      "loss": 2.075,
+      "step": 247705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001609251135572532,
+      "loss": 2.1215,
+      "step": 247710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001609236478936022,
+      "loss": 2.0938,
+      "step": 247715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001609221822091386,
+      "loss": 2.0825,
+      "step": 247720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.1875,
+      "learning_rate": 0.00016092071650386286,
+      "loss": 2.3089,
+      "step": 247725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016091925077777548,
+      "loss": 1.9791,
+      "step": 247730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000160917785030877,
+      "loss": 2.0771,
+      "step": 247735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001609163192631679,
+      "loss": 2.0777,
+      "step": 247740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016091485347464866,
+      "loss": 2.0776,
+      "step": 247745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016091338766531982,
+      "loss": 2.0682,
+      "step": 247750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016091192183518186,
+      "loss": 2.0908,
+      "step": 247755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001609104559842353,
+      "loss": 1.9719,
+      "step": 247760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016090899011248058,
+      "loss": 2.1213,
+      "step": 247765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016090752421991827,
+      "loss": 1.9131,
+      "step": 247770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016090605830654885,
+      "loss": 2.0582,
+      "step": 247775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001609045923723728,
+      "loss": 2.1974,
+      "step": 247780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016090312641739066,
+      "loss": 2.1217,
+      "step": 247785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00016090166044160288,
+      "loss": 2.0225,
+      "step": 247790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016090019444501001,
+      "loss": 2.0883,
+      "step": 247795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016089872842761253,
+      "loss": 2.174,
+      "step": 247800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016089726238941094,
+      "loss": 2.1074,
+      "step": 247805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016089579633040573,
+      "loss": 2.0583,
+      "step": 247810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016089433025059743,
+      "loss": 1.9963,
+      "step": 247815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016089286414998648,
+      "loss": 2.1518,
+      "step": 247820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016089139802857345,
+      "loss": 2.1605,
+      "step": 247825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016088993188635882,
+      "loss": 2.0261,
+      "step": 247830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001608884657233431,
+      "loss": 2.0979,
+      "step": 247835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016088699953952673,
+      "loss": 2.0194,
+      "step": 247840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016088553333491027,
+      "loss": 2.1453,
+      "step": 247845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016088406710949423,
+      "loss": 2.011,
+      "step": 247850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016088260086327908,
+      "loss": 2.1975,
+      "step": 247855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016088113459626532,
+      "loss": 1.9943,
+      "step": 247860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001608796683084535,
+      "loss": 2.1085,
+      "step": 247865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016087820199984403,
+      "loss": 2.2104,
+      "step": 247870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016087673567043747,
+      "loss": 1.9381,
+      "step": 247875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016087526932023435,
+      "loss": 1.9957,
+      "step": 247880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001608738029492351,
+      "loss": 2.2038,
+      "step": 247885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016087233655744028,
+      "loss": 1.8612,
+      "step": 247890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016087087014485032,
+      "loss": 1.9024,
+      "step": 247895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001608694037114658,
+      "loss": 1.9625,
+      "step": 247900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016086793725728717,
+      "loss": 1.9925,
+      "step": 247905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016086647078231497,
+      "loss": 2.3482,
+      "step": 247910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016086500428654968,
+      "loss": 1.9693,
+      "step": 247915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016086353776999178,
+      "loss": 2.0291,
+      "step": 247920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001608620712326418,
+      "loss": 2.2282,
+      "step": 247925
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016086060467450025,
+      "loss": 2.0764,
+      "step": 247930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001608591380955676,
+      "loss": 2.0693,
+      "step": 247935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016085767149584437,
+      "loss": 1.9975,
+      "step": 247940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016085620487533105,
+      "loss": 1.9895,
+      "step": 247945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016085473823402815,
+      "loss": 1.9593,
+      "step": 247950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016085327157193614,
+      "loss": 2.2375,
+      "step": 247955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016085180488905557,
+      "loss": 2.1231,
+      "step": 247960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00016085033818538695,
+      "loss": 1.9745,
+      "step": 247965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016084887146093074,
+      "loss": 2.1764,
+      "step": 247970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001608474047156874,
+      "loss": 2.2645,
+      "step": 247975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00016084593794965754,
+      "loss": 2.236,
+      "step": 247980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001608444711628416,
+      "loss": 1.93,
+      "step": 247985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016084300435524008,
+      "loss": 1.9467,
+      "step": 247990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016084153752685346,
+      "loss": 2.0753,
+      "step": 247995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016084007067768228,
+      "loss": 2.1193,
+      "step": 248000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016083860380772702,
+      "loss": 2.1874,
+      "step": 248005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001608371369169882,
+      "loss": 2.1992,
+      "step": 248010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001608356700054663,
+      "loss": 2.1647,
+      "step": 248015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016083420307316185,
+      "loss": 2.0095,
+      "step": 248020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016083273612007534,
+      "loss": 2.0374,
+      "step": 248025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016083126914620723,
+      "loss": 1.9626,
+      "step": 248030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001608298021515581,
+      "loss": 2.0285,
+      "step": 248035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001608283351361284,
+      "loss": 2.1554,
+      "step": 248040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001608268680999186,
+      "loss": 2.0565,
+      "step": 248045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016082540104292927,
+      "loss": 2.0254,
+      "step": 248050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016082393396516084,
+      "loss": 2.0874,
+      "step": 248055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001608224668666139,
+      "loss": 2.1613,
+      "step": 248060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016082099974728888,
+      "loss": 2.1105,
+      "step": 248065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001608195326071863,
+      "loss": 2.0724,
+      "step": 248070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016081806544630668,
+      "loss": 1.9995,
+      "step": 248075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016081659826465048,
+      "loss": 2.0055,
+      "step": 248080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016081513106221823,
+      "loss": 2.1348,
+      "step": 248085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016081366383901045,
+      "loss": 2.2742,
+      "step": 248090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001608121965950276,
+      "loss": 2.0474,
+      "step": 248095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016081072933027022,
+      "loss": 2.1263,
+      "step": 248100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016080926204473878,
+      "loss": 2.1559,
+      "step": 248105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001608077947384338,
+      "loss": 2.1588,
+      "step": 248110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016080632741135578,
+      "loss": 2.0285,
+      "step": 248115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016080486006350518,
+      "loss": 2.1306,
+      "step": 248120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016080339269488258,
+      "loss": 2.1592,
+      "step": 248125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016080192530548842,
+      "loss": 2.0407,
+      "step": 248130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016080045789532323,
+      "loss": 2.0934,
+      "step": 248135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016079899046438748,
+      "loss": 1.9939,
+      "step": 248140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001607975230126817,
+      "loss": 1.8946,
+      "step": 248145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001607960555402064,
+      "loss": 2.1501,
+      "step": 248150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016079458804696206,
+      "loss": 1.8761,
+      "step": 248155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016079312053294919,
+      "loss": 2.1116,
+      "step": 248160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016079165299816826,
+      "loss": 2.207,
+      "step": 248165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016079018544261982,
+      "loss": 2.213,
+      "step": 248170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016078871786630433,
+      "loss": 2.1401,
+      "step": 248175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016078725026922236,
+      "loss": 2.1103,
+      "step": 248180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001607857826513743,
+      "loss": 2.1652,
+      "step": 248185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016078431501276075,
+      "loss": 2.1869,
+      "step": 248190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016078284735338217,
+      "loss": 1.9704,
+      "step": 248195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001607813796732391,
+      "loss": 2.1672,
+      "step": 248200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016077991197233193,
+      "loss": 2.1337,
+      "step": 248205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016077844425066128,
+      "loss": 1.7952,
+      "step": 248210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016077697650822764,
+      "loss": 2.2578,
+      "step": 248215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016077550874503145,
+      "loss": 1.9921,
+      "step": 248220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016077404096107327,
+      "loss": 1.9765,
+      "step": 248225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016077257315635354,
+      "loss": 2.0089,
+      "step": 248230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016077110533087283,
+      "loss": 1.9699,
+      "step": 248235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016076963748463158,
+      "loss": 2.0287,
+      "step": 248240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016076816961763031,
+      "loss": 1.9863,
+      "step": 248245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016076670172986956,
+      "loss": 2.0949,
+      "step": 248250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001607652338213498,
+      "loss": 2.2847,
+      "step": 248255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016076376589207154,
+      "loss": 2.229,
+      "step": 248260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016076229794203528,
+      "loss": 2.177,
+      "step": 248265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001607608299712415,
+      "loss": 2.1497,
+      "step": 248270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016075936197969073,
+      "loss": 1.9829,
+      "step": 248275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016075789396738344,
+      "loss": 2.0721,
+      "step": 248280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016075642593432017,
+      "loss": 2.0636,
+      "step": 248285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016075495788050138,
+      "loss": 2.2095,
+      "step": 248290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001607534898059276,
+      "loss": 2.0251,
+      "step": 248295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00016075202171059936,
+      "loss": 1.9425,
+      "step": 248300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016075055359451713,
+      "loss": 2.0148,
+      "step": 248305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016074908545768137,
+      "loss": 2.0183,
+      "step": 248310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016074761730009262,
+      "loss": 2.1092,
+      "step": 248315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001607461491217514,
+      "loss": 1.9774,
+      "step": 248320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016074468092265817,
+      "loss": 2.1695,
+      "step": 248325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001607432127028135,
+      "loss": 2.2284,
+      "step": 248330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001607417444622178,
+      "loss": 2.2511,
+      "step": 248335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016074027620087167,
+      "loss": 2.03,
+      "step": 248340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016073880791877553,
+      "loss": 2.0723,
+      "step": 248345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001607373396159299,
+      "loss": 2.1333,
+      "step": 248350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001607358712923353,
+      "loss": 2.1428,
+      "step": 248355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016073440294799224,
+      "loss": 2.0563,
+      "step": 248360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016073293458290119,
+      "loss": 2.1174,
+      "step": 248365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001607314661970627,
+      "loss": 2.0692,
+      "step": 248370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016072999779047722,
+      "loss": 2.0053,
+      "step": 248375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016072852936314523,
+      "loss": 2.0848,
+      "step": 248380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00016072706091506734,
+      "loss": 2.1394,
+      "step": 248385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016072559244624394,
+      "loss": 1.9962,
+      "step": 248390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001607241239566756,
+      "loss": 1.9117,
+      "step": 248395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001607226554463628,
+      "loss": 2.2559,
+      "step": 248400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.000160721186915306,
+      "loss": 2.1177,
+      "step": 248405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016071971836350576,
+      "loss": 2.0991,
+      "step": 248410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001607182497909626,
+      "loss": 2.1076,
+      "step": 248415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016071678119767695,
+      "loss": 2.3141,
+      "step": 248420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016071531258364936,
+      "loss": 2.1604,
+      "step": 248425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016071384394888028,
+      "loss": 2.2029,
+      "step": 248430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001607123752933703,
+      "loss": 2.2167,
+      "step": 248435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016071090661711986,
+      "loss": 2.2181,
+      "step": 248440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016070943792012945,
+      "loss": 2.086,
+      "step": 248445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001607079692023996,
+      "loss": 2.1083,
+      "step": 248450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016070650046393087,
+      "loss": 2.0791,
+      "step": 248455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016070503170472362,
+      "loss": 2.1631,
+      "step": 248460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016070356292477843,
+      "loss": 2.1122,
+      "step": 248465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016070209412409584,
+      "loss": 2.0152,
+      "step": 248470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016070062530267634,
+      "loss": 2.086,
+      "step": 248475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016069915646052034,
+      "loss": 1.952,
+      "step": 248480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016069768759762845,
+      "loss": 2.0803,
+      "step": 248485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001606962187140011,
+      "loss": 2.1509,
+      "step": 248490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016069474980963885,
+      "loss": 1.9769,
+      "step": 248495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016069328088454213,
+      "loss": 1.9989,
+      "step": 248500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016069181193871152,
+      "loss": 2.0494,
+      "step": 248505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001606903429721475,
+      "loss": 2.1036,
+      "step": 248510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016068887398485053,
+      "loss": 2.0359,
+      "step": 248515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016068740497682114,
+      "loss": 2.0033,
+      "step": 248520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016068593594805984,
+      "loss": 2.1572,
+      "step": 248525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001606844668985671,
+      "loss": 2.2693,
+      "step": 248530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00016068299782834352,
+      "loss": 2.0517,
+      "step": 248535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016068152873738943,
+      "loss": 1.9254,
+      "step": 248540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016068005962570548,
+      "loss": 2.2647,
+      "step": 248545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016067859049329213,
+      "loss": 2.0522,
+      "step": 248550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016067712134014986,
+      "loss": 2.1424,
+      "step": 248555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016067565216627917,
+      "loss": 2.0142,
+      "step": 248560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016067418297168059,
+      "loss": 2.1119,
+      "step": 248565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001606727137563546,
+      "loss": 1.9854,
+      "step": 248570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016067124452030175,
+      "loss": 1.9834,
+      "step": 248575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016066977526352244,
+      "loss": 1.8993,
+      "step": 248580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016066830598601727,
+      "loss": 2.1552,
+      "step": 248585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016066683668778672,
+      "loss": 2.0058,
+      "step": 248590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016066536736883123,
+      "loss": 2.0888,
+      "step": 248595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001606638980291514,
+      "loss": 2.2339,
+      "step": 248600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001606624286687477,
+      "loss": 1.9342,
+      "step": 248605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016066095928762058,
+      "loss": 2.2163,
+      "step": 248610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016065948988577055,
+      "loss": 2.1218,
+      "step": 248615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016065802046319816,
+      "loss": 2.0826,
+      "step": 248620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001606565510199039,
+      "loss": 2.1696,
+      "step": 248625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016065508155588824,
+      "loss": 2.0538,
+      "step": 248630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016065361207115176,
+      "loss": 2.0802,
+      "step": 248635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016065214256569482,
+      "loss": 2.0705,
+      "step": 248640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016065067303951805,
+      "loss": 2.2548,
+      "step": 248645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001606492034926219,
+      "loss": 2.2293,
+      "step": 248650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016064773392500693,
+      "loss": 2.2218,
+      "step": 248655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016064626433667356,
+      "loss": 2.0443,
+      "step": 248660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016064479472762232,
+      "loss": 2.0672,
+      "step": 248665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016064332509785372,
+      "loss": 2.092,
+      "step": 248670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016064185544736827,
+      "loss": 2.1311,
+      "step": 248675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016064038577616647,
+      "loss": 1.9221,
+      "step": 248680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001606389160842488,
+      "loss": 2.0726,
+      "step": 248685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016063744637161576,
+      "loss": 2.1993,
+      "step": 248690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001606359766382679,
+      "loss": 1.9611,
+      "step": 248695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016063450688420566,
+      "loss": 2.1066,
+      "step": 248700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001606330371094296,
+      "loss": 2.0923,
+      "step": 248705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016063156731394018,
+      "loss": 2.1441,
+      "step": 248710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016063009749773795,
+      "loss": 2.0643,
+      "step": 248715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001606286276608233,
+      "loss": 2.1149,
+      "step": 248720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016062715780319688,
+      "loss": 2.2516,
+      "step": 248725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001606256879248591,
+      "loss": 2.021,
+      "step": 248730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016062421802581049,
+      "loss": 2.2328,
+      "step": 248735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016062274810605155,
+      "loss": 2.0717,
+      "step": 248740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016062127816558278,
+      "loss": 2.146,
+      "step": 248745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001606198082044047,
+      "loss": 1.9262,
+      "step": 248750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016061833822251777,
+      "loss": 2.1994,
+      "step": 248755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016061686821992253,
+      "loss": 2.0111,
+      "step": 248760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016061539819661947,
+      "loss": 1.9845,
+      "step": 248765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016061392815260907,
+      "loss": 2.0238,
+      "step": 248770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016061245808789186,
+      "loss": 1.9875,
+      "step": 248775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016061098800246833,
+      "loss": 2.1798,
+      "step": 248780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016060951789633897,
+      "loss": 2.0621,
+      "step": 248785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016060804776950435,
+      "loss": 2.2202,
+      "step": 248790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001606065776219649,
+      "loss": 2.0967,
+      "step": 248795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016060510745372116,
+      "loss": 2.0976,
+      "step": 248800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016060363726477358,
+      "loss": 2.0548,
+      "step": 248805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001606021670551227,
+      "loss": 2.1374,
+      "step": 248810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016060069682476903,
+      "loss": 2.1594,
+      "step": 248815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016059922657371306,
+      "loss": 2.0964,
+      "step": 248820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016059775630195531,
+      "loss": 2.0157,
+      "step": 248825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016059628600949627,
+      "loss": 2.1175,
+      "step": 248830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016059481569633642,
+      "loss": 2.0638,
+      "step": 248835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001605933453624763,
+      "loss": 2.0382,
+      "step": 248840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001605918750079164,
+      "loss": 2.1214,
+      "step": 248845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001605904046326572,
+      "loss": 1.9501,
+      "step": 248850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016058893423669923,
+      "loss": 2.152,
+      "step": 248855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016058746382004296,
+      "loss": 2.2877,
+      "step": 248860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016058599338268894,
+      "loss": 2.1184,
+      "step": 248865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001605845229246376,
+      "loss": 2.0839,
+      "step": 248870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016058305244588957,
+      "loss": 2.0389,
+      "step": 248875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016058158194644522,
+      "loss": 2.0057,
+      "step": 248880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016058011142630507,
+      "loss": 2.0567,
+      "step": 248885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001605786408854697,
+      "loss": 2.1571,
+      "step": 248890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016057717032393954,
+      "loss": 2.1189,
+      "step": 248895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016057569974171516,
+      "loss": 1.901,
+      "step": 248900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016057422913879702,
+      "loss": 2.1568,
+      "step": 248905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016057275851518558,
+      "loss": 2.2074,
+      "step": 248910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016057128787088143,
+      "loss": 2.005,
+      "step": 248915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000160569817205885,
+      "loss": 2.0098,
+      "step": 248920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001605683465201968,
+      "loss": 2.1705,
+      "step": 248925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016056687581381742,
+      "loss": 2.054,
+      "step": 248930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016056540508674726,
+      "loss": 2.1552,
+      "step": 248935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016056393433898686,
+      "loss": 2.0593,
+      "step": 248940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016056246357053672,
+      "loss": 2.1739,
+      "step": 248945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016056099278139737,
+      "loss": 2.0976,
+      "step": 248950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00016055952197156924,
+      "loss": 2.1171,
+      "step": 248955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016055805114105294,
+      "loss": 2.0115,
+      "step": 248960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016055658028984885,
+      "loss": 2.0234,
+      "step": 248965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016055510941795757,
+      "loss": 2.0711,
+      "step": 248970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00016055363852537954,
+      "loss": 2.0739,
+      "step": 248975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016055216761211533,
+      "loss": 2.1173,
+      "step": 248980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016055069667816538,
+      "loss": 2.1431,
+      "step": 248985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016054922572353018,
+      "loss": 2.2279,
+      "step": 248990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001605477547482103,
+      "loss": 2.1159,
+      "step": 248995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001605462837522062,
+      "loss": 2.0286,
+      "step": 249000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001605448127355184,
+      "loss": 2.2871,
+      "step": 249005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016054334169814742,
+      "loss": 1.9917,
+      "step": 249010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001605418706400937,
+      "loss": 2.0696,
+      "step": 249015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016054039956135777,
+      "loss": 2.2806,
+      "step": 249020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016053892846194017,
+      "loss": 2.2152,
+      "step": 249025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016053745734184136,
+      "loss": 2.0777,
+      "step": 249030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016053598620106188,
+      "loss": 2.1223,
+      "step": 249035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016053451503960219,
+      "loss": 1.9606,
+      "step": 249040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016053304385746278,
+      "loss": 2.0868,
+      "step": 249045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00016053157265464422,
+      "loss": 1.9548,
+      "step": 249050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016053010143114697,
+      "loss": 2.2131,
+      "step": 249055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016052863018697154,
+      "loss": 2.042,
+      "step": 249060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016052715892211846,
+      "loss": 2.1218,
+      "step": 249065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016052568763658817,
+      "loss": 2.032,
+      "step": 249070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001605242163303812,
+      "loss": 2.1738,
+      "step": 249075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001605227450034981,
+      "loss": 1.9398,
+      "step": 249080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001605212736559393,
+      "loss": 2.077,
+      "step": 249085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016051980228770535,
+      "loss": 2.2556,
+      "step": 249090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001605183308987967,
+      "loss": 1.9039,
+      "step": 249095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00016051685948921394,
+      "loss": 2.0746,
+      "step": 249100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016051538805895752,
+      "loss": 2.0183,
+      "step": 249105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016051391660802794,
+      "loss": 2.069,
+      "step": 249110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016051244513642568,
+      "loss": 2.044,
+      "step": 249115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001605109736441513,
+      "loss": 1.9984,
+      "step": 249120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016050950213120525,
+      "loss": 2.0901,
+      "step": 249125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016050803059758806,
+      "loss": 2.1297,
+      "step": 249130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016050655904330027,
+      "loss": 2.1109,
+      "step": 249135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001605050874683423,
+      "loss": 2.0318,
+      "step": 249140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016050361587271472,
+      "loss": 2.0749,
+      "step": 249145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016050214425641798,
+      "loss": 2.2674,
+      "step": 249150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016050067261945262,
+      "loss": 1.9969,
+      "step": 249155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016049920096181914,
+      "loss": 2.1555,
+      "step": 249160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016049772928351805,
+      "loss": 2.0551,
+      "step": 249165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016049625758454982,
+      "loss": 2.0911,
+      "step": 249170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016049478586491498,
+      "loss": 2.1649,
+      "step": 249175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000160493314124614,
+      "loss": 2.0401,
+      "step": 249180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016049184236364744,
+      "loss": 2.1519,
+      "step": 249185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016049037058201576,
+      "loss": 2.0636,
+      "step": 249190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016048889877971946,
+      "loss": 2.2381,
+      "step": 249195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016048742695675907,
+      "loss": 2.1087,
+      "step": 249200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016048595511313507,
+      "loss": 1.9338,
+      "step": 249205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016048448324884794,
+      "loss": 2.0491,
+      "step": 249210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016048301136389827,
+      "loss": 1.9614,
+      "step": 249215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016048153945828647,
+      "loss": 2.1174,
+      "step": 249220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001604800675320131,
+      "loss": 1.9734,
+      "step": 249225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016047859558507863,
+      "loss": 2.0769,
+      "step": 249230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016047712361748356,
+      "loss": 2.149,
+      "step": 249235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016047565162922845,
+      "loss": 2.1214,
+      "step": 249240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016047417962031374,
+      "loss": 2.2332,
+      "step": 249245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016047270759073995,
+      "loss": 2.1648,
+      "step": 249250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001604712355405076,
+      "loss": 1.9586,
+      "step": 249255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016046976346961714,
+      "loss": 2.0441,
+      "step": 249260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016046829137806917,
+      "loss": 2.2019,
+      "step": 249265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016046681926586408,
+      "loss": 2.1113,
+      "step": 249270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016046534713300243,
+      "loss": 2.1292,
+      "step": 249275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016046387497948477,
+      "loss": 2.1654,
+      "step": 249280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016046240280531154,
+      "loss": 2.0695,
+      "step": 249285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016046093061048322,
+      "loss": 2.0406,
+      "step": 249290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016045945839500037,
+      "loss": 2.197,
+      "step": 249295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016045798615886344,
+      "loss": 2.2299,
+      "step": 249300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000160456513902073,
+      "loss": 2.1044,
+      "step": 249305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016045504162462955,
+      "loss": 2.1436,
+      "step": 249310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016045356932653353,
+      "loss": 2.0971,
+      "step": 249315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016045209700778546,
+      "loss": 2.1791,
+      "step": 249320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016045062466838586,
+      "loss": 2.0852,
+      "step": 249325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016044915230833525,
+      "loss": 2.1677,
+      "step": 249330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001604476799276341,
+      "loss": 2.2167,
+      "step": 249335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016044620752628293,
+      "loss": 2.1572,
+      "step": 249340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016044473510428225,
+      "loss": 2.0579,
+      "step": 249345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001604432626616325,
+      "loss": 2.0304,
+      "step": 249350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001604417901983343,
+      "loss": 2.0036,
+      "step": 249355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001604403177143881,
+      "loss": 2.1266,
+      "step": 249360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016043884520979432,
+      "loss": 2.1585,
+      "step": 249365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016043737268455356,
+      "loss": 1.9957,
+      "step": 249370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016043590013866633,
+      "loss": 2.1728,
+      "step": 249375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001604344275721331,
+      "loss": 2.2423,
+      "step": 249380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016043295498495433,
+      "loss": 2.1849,
+      "step": 249385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016043148237713059,
+      "loss": 2.158,
+      "step": 249390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016043000974866236,
+      "loss": 2.2215,
+      "step": 249395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016042853709955014,
+      "loss": 2.15,
+      "step": 249400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016042706442979444,
+      "loss": 2.0304,
+      "step": 249405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016042559173939575,
+      "loss": 1.9935,
+      "step": 249410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001604241190283546,
+      "loss": 1.9113,
+      "step": 249415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016042264629667144,
+      "loss": 2.0372,
+      "step": 249420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016042117354434684,
+      "loss": 2.1402,
+      "step": 249425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016041970077138127,
+      "loss": 2.0624,
+      "step": 249430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016041822797777523,
+      "loss": 2.0918,
+      "step": 249435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001604167551635292,
+      "loss": 2.0405,
+      "step": 249440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00016041528232864377,
+      "loss": 2.1522,
+      "step": 249445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016041380947311932,
+      "loss": 2.2112,
+      "step": 249450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016041233659695644,
+      "loss": 2.1128,
+      "step": 249455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016041086370015563,
+      "loss": 2.0143,
+      "step": 249460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016040939078271739,
+      "loss": 1.9098,
+      "step": 249465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016040791784464217,
+      "loss": 2.0713,
+      "step": 249470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016040644488593049,
+      "loss": 2.0719,
+      "step": 249475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016040497190658291,
+      "loss": 2.0313,
+      "step": 249480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001604034989065999,
+      "loss": 2.2916,
+      "step": 249485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016040202588598192,
+      "loss": 2.2233,
+      "step": 249490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016040055284472955,
+      "loss": 2.068,
+      "step": 249495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016039907978284324,
+      "loss": 2.1139,
+      "step": 249500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001603976067003235,
+      "loss": 1.97,
+      "step": 249505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016039613359717086,
+      "loss": 2.1373,
+      "step": 249510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001603946604733858,
+      "loss": 2.2411,
+      "step": 249515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016039318732896885,
+      "loss": 2.1849,
+      "step": 249520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016039171416392045,
+      "loss": 2.134,
+      "step": 249525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016039024097824113,
+      "loss": 2.1018,
+      "step": 249530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016038876777193145,
+      "loss": 2.1119,
+      "step": 249535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001603872945449919,
+      "loss": 1.9932,
+      "step": 249540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001603858212974229,
+      "loss": 1.851,
+      "step": 249545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016038434802922503,
+      "loss": 2.1092,
+      "step": 249550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016038287474039878,
+      "loss": 2.1743,
+      "step": 249555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016038140143094463,
+      "loss": 2.1459,
+      "step": 249560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001603799281008631,
+      "loss": 2.0802,
+      "step": 249565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001603784547501547,
+      "loss": 1.9944,
+      "step": 249570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016037698137881992,
+      "loss": 2.1417,
+      "step": 249575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016037550798685926,
+      "loss": 2.0784,
+      "step": 249580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016037403457427323,
+      "loss": 1.9103,
+      "step": 249585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016037256114106234,
+      "loss": 2.1568,
+      "step": 249590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016037108768722708,
+      "loss": 2.1673,
+      "step": 249595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000160369614212768,
+      "loss": 1.9508,
+      "step": 249600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016036814071768553,
+      "loss": 2.1264,
+      "step": 249605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016036666720198022,
+      "loss": 1.8894,
+      "step": 249610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00016036519366565254,
+      "loss": 2.0404,
+      "step": 249615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016036372010870306,
+      "loss": 2.212,
+      "step": 249620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001603622465311322,
+      "loss": 2.0184,
+      "step": 249625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016036077293294051,
+      "loss": 2.0803,
+      "step": 249630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016035929931412847,
+      "loss": 2.1017,
+      "step": 249635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016035782567469663,
+      "loss": 2.0681,
+      "step": 249640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016035635201464547,
+      "loss": 2.1519,
+      "step": 249645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016035487833397543,
+      "loss": 1.9718,
+      "step": 249650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00016035340463268713,
+      "loss": 2.1328,
+      "step": 249655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.000160351930910781,
+      "loss": 2.1878,
+      "step": 249660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001603504571682575,
+      "loss": 2.0418,
+      "step": 249665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016034898340511726,
+      "loss": 2.1732,
+      "step": 249670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016034750962136068,
+      "loss": 2.1108,
+      "step": 249675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001603460358169883,
+      "loss": 1.9359,
+      "step": 249680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00016034456199200063,
+      "loss": 2.2812,
+      "step": 249685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016034308814639813,
+      "loss": 2.176,
+      "step": 249690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016034161428018137,
+      "loss": 2.0936,
+      "step": 249695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016034014039335083,
+      "loss": 2.1356,
+      "step": 249700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016033866648590696,
+      "loss": 2.0725,
+      "step": 249705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016033719255785034,
+      "loss": 2.2689,
+      "step": 249710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016033571860918144,
+      "loss": 2.0581,
+      "step": 249715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016033424463990077,
+      "loss": 2.159,
+      "step": 249720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001603327706500088,
+      "loss": 2.0794,
+      "step": 249725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00016033129663950608,
+      "loss": 2.0933,
+      "step": 249730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.0001603298226083931,
+      "loss": 2.0949,
+      "step": 249735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016032834855667034,
+      "loss": 2.2813,
+      "step": 249740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016032687448433832,
+      "loss": 2.0293,
+      "step": 249745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016032540039139756,
+      "loss": 1.9951,
+      "step": 249750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016032392627784856,
+      "loss": 1.978,
+      "step": 249755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001603224521436918,
+      "loss": 2.0945,
+      "step": 249760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016032097798892778,
+      "loss": 2.0237,
+      "step": 249765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016031950381355703,
+      "loss": 1.9892,
+      "step": 249770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016031802961758004,
+      "loss": 2.1538,
+      "step": 249775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016031655540099733,
+      "loss": 2.2499,
+      "step": 249780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001603150811638094,
+      "loss": 2.1395,
+      "step": 249785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00016031360690601674,
+      "loss": 1.9864,
+      "step": 249790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016031213262761984,
+      "loss": 2.2,
+      "step": 249795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016031065832861923,
+      "loss": 1.963,
+      "step": 249800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001603091840090154,
+      "loss": 2.0679,
+      "step": 249805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016030770966880885,
+      "loss": 2.0643,
+      "step": 249810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016030623530800014,
+      "loss": 1.9657,
+      "step": 249815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016030476092658967,
+      "loss": 1.9648,
+      "step": 249820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016030328652457803,
+      "loss": 1.9104,
+      "step": 249825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016030181210196568,
+      "loss": 2.1826,
+      "step": 249830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016030033765875316,
+      "loss": 2.0452,
+      "step": 249835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016029886319494092,
+      "loss": 2.0998,
+      "step": 249840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001602973887105295,
+      "loss": 1.9772,
+      "step": 249845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00016029591420551942,
+      "loss": 2.1459,
+      "step": 249850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016029443967991114,
+      "loss": 2.0353,
+      "step": 249855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016029296513370522,
+      "loss": 2.2153,
+      "step": 249860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00016029149056690208,
+      "loss": 2.1019,
+      "step": 249865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016029001597950233,
+      "loss": 2.1027,
+      "step": 249870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00016028854137150637,
+      "loss": 2.1016,
+      "step": 249875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016028706674291478,
+      "loss": 2.153,
+      "step": 249880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016028559209372802,
+      "loss": 2.2048,
+      "step": 249885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001602841174239466,
+      "loss": 2.0387,
+      "step": 249890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016028264273357107,
+      "loss": 1.9376,
+      "step": 249895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016028116802260187,
+      "loss": 2.1332,
+      "step": 249900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001602796932910395,
+      "loss": 2.076,
+      "step": 249905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016027821853888452,
+      "loss": 1.8904,
+      "step": 249910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016027674376613743,
+      "loss": 2.0167,
+      "step": 249915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001602752689727987,
+      "loss": 2.0937,
+      "step": 249920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016027379415886882,
+      "loss": 2.0648,
+      "step": 249925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001602723193243483,
+      "loss": 2.0933,
+      "step": 249930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016027084446923774,
+      "loss": 2.313,
+      "step": 249935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016026936959353752,
+      "loss": 2.0699,
+      "step": 249940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001602678946972482,
+      "loss": 2.0615,
+      "step": 249945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016026641978037027,
+      "loss": 2.0708,
+      "step": 249950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016026494484290424,
+      "loss": 2.0403,
+      "step": 249955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001602634698848506,
+      "loss": 1.981,
+      "step": 249960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016026199490620986,
+      "loss": 2.1843,
+      "step": 249965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016026051990698255,
+      "loss": 2.1733,
+      "step": 249970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016025904488716918,
+      "loss": 1.9295,
+      "step": 249975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001602575698467702,
+      "loss": 2.1136,
+      "step": 249980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001602560947857861,
+      "loss": 1.9829,
+      "step": 249985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016025461970421744,
+      "loss": 1.9359,
+      "step": 249990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016025314460206476,
+      "loss": 2.1187,
+      "step": 249995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00016025166947932848,
+      "loss": 2.1338,
+      "step": 250000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016025019433600914,
+      "loss": 2.2105,
+      "step": 250005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016024871917210724,
+      "loss": 1.9185,
+      "step": 250010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.0001602472439876233,
+      "loss": 2.2134,
+      "step": 250015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016024576878255776,
+      "loss": 2.0198,
+      "step": 250020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016024429355691123,
+      "loss": 2.0509,
+      "step": 250025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016024281831068416,
+      "loss": 1.8983,
+      "step": 250030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016024134304387698,
+      "loss": 1.9043,
+      "step": 250035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016023986775649034,
+      "loss": 2.1442,
+      "step": 250040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016023839244852463,
+      "loss": 2.2606,
+      "step": 250045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016023691711998043,
+      "loss": 2.1158,
+      "step": 250050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016023544177085815,
+      "loss": 2.1233,
+      "step": 250055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016023396640115837,
+      "loss": 2.0581,
+      "step": 250060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001602324910108816,
+      "loss": 1.9962,
+      "step": 250065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016023101560002826,
+      "loss": 2.2773,
+      "step": 250070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016022954016859896,
+      "loss": 2.1946,
+      "step": 250075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016022806471659417,
+      "loss": 1.9394,
+      "step": 250080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016022658924401436,
+      "loss": 2.0726,
+      "step": 250085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016022511375086005,
+      "loss": 1.8969,
+      "step": 250090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.15625,
+      "learning_rate": 0.0001602236382371318,
+      "loss": 2.2985,
+      "step": 250095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016022216270282998,
+      "loss": 2.0598,
+      "step": 250100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016022068714795522,
+      "loss": 2.0493,
+      "step": 250105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016021921157250799,
+      "loss": 2.1306,
+      "step": 250110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016021773597648875,
+      "loss": 2.2886,
+      "step": 250115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016021626035989806,
+      "loss": 2.1275,
+      "step": 250120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001602147847227364,
+      "loss": 2.0832,
+      "step": 250125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00016021330906500428,
+      "loss": 2.1481,
+      "step": 250130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016021183338670223,
+      "loss": 1.996,
+      "step": 250135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001602103576878307,
+      "loss": 1.945,
+      "step": 250140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001602088819683902,
+      "loss": 1.9862,
+      "step": 250145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00016020740622838128,
+      "loss": 2.021,
+      "step": 250150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001602059304678044,
+      "loss": 2.0781,
+      "step": 250155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016020445468666012,
+      "loss": 2.1532,
+      "step": 250160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001602029788849489,
+      "loss": 2.3043,
+      "step": 250165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001602015030626712,
+      "loss": 2.0232,
+      "step": 250170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001602000272198276,
+      "loss": 2.1717,
+      "step": 250175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001601985513564186,
+      "loss": 1.962,
+      "step": 250180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016019707547244467,
+      "loss": 2.0157,
+      "step": 250185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001601955995679063,
+      "loss": 2.0215,
+      "step": 250190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016019412364280406,
+      "loss": 2.0977,
+      "step": 250195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001601926476971384,
+      "loss": 2.0626,
+      "step": 250200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016019117173090983,
+      "loss": 2.1539,
+      "step": 250205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016018969574411887,
+      "loss": 2.1123,
+      "step": 250210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016018821973676602,
+      "loss": 2.2606,
+      "step": 250215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016018674370885181,
+      "loss": 1.9981,
+      "step": 250220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001601852676603767,
+      "loss": 2.3355,
+      "step": 250225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016018379159134122,
+      "loss": 2.1771,
+      "step": 250230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016018231550174583,
+      "loss": 2.1496,
+      "step": 250235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016018083939159107,
+      "loss": 2.1358,
+      "step": 250240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001601793632608775,
+      "loss": 2.2026,
+      "step": 250245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016017788710960554,
+      "loss": 2.0996,
+      "step": 250250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001601764109377757,
+      "loss": 1.9684,
+      "step": 250255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016017493474538853,
+      "loss": 2.092,
+      "step": 250260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00016017345853244446,
+      "loss": 2.1885,
+      "step": 250265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001601719822989441,
+      "loss": 2.0989,
+      "step": 250270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016017050604488788,
+      "loss": 2.0168,
+      "step": 250275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016016902977027635,
+      "loss": 2.0401,
+      "step": 250280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016016755347510997,
+      "loss": 2.0849,
+      "step": 250285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016016607715938925,
+      "loss": 2.1426,
+      "step": 250290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001601646008231147,
+      "loss": 2.0303,
+      "step": 250295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016016312446628686,
+      "loss": 2.2843,
+      "step": 250300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001601616480889062,
+      "loss": 2.0593,
+      "step": 250305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016016017169097322,
+      "loss": 1.8645,
+      "step": 250310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016015869527248846,
+      "loss": 2.0658,
+      "step": 250315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00016015721883345235,
+      "loss": 1.8915,
+      "step": 250320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001601557423738655,
+      "loss": 2.1027,
+      "step": 250325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016015426589372834,
+      "loss": 2.1438,
+      "step": 250330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00016015278939304135,
+      "loss": 2.1778,
+      "step": 250335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001601513128718051,
+      "loss": 2.1842,
+      "step": 250340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001601498363300201,
+      "loss": 2.0766,
+      "step": 250345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001601483597676868,
+      "loss": 1.9505,
+      "step": 250350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016014688318480573,
+      "loss": 1.8011,
+      "step": 250355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001601454065813774,
+      "loss": 2.0081,
+      "step": 250360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016014392995740225,
+      "loss": 2.0576,
+      "step": 250365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016014245331288094,
+      "loss": 2.2054,
+      "step": 250370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016014097664781383,
+      "loss": 2.2231,
+      "step": 250375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016013949996220147,
+      "loss": 2.2874,
+      "step": 250380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016013802325604438,
+      "loss": 2.13,
+      "step": 250385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016013654652934304,
+      "loss": 2.0168,
+      "step": 250390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016013506978209796,
+      "loss": 2.0548,
+      "step": 250395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016013359301430966,
+      "loss": 2.1646,
+      "step": 250400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001601321162259786,
+      "loss": 2.1881,
+      "step": 250405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00016013063941710536,
+      "loss": 2.2139,
+      "step": 250410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001601291625876904,
+      "loss": 2.0586,
+      "step": 250415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001601276857377342,
+      "loss": 1.9314,
+      "step": 250420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00016012620886723733,
+      "loss": 2.1044,
+      "step": 250425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00016012473197620026,
+      "loss": 2.0021,
+      "step": 250430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016012325506462345,
+      "loss": 2.2389,
+      "step": 250435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00016012177813250745,
+      "loss": 2.03,
+      "step": 250440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001601203011798528,
+      "loss": 2.064,
+      "step": 250445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016011882420665993,
+      "loss": 2.1517,
+      "step": 250450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001601173472129294,
+      "loss": 2.0504,
+      "step": 250455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016011587019866167,
+      "loss": 2.1684,
+      "step": 250460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016011439316385728,
+      "loss": 2.1439,
+      "step": 250465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016011291610851672,
+      "loss": 2.072,
+      "step": 250470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001601114390326405,
+      "loss": 2.3016,
+      "step": 250475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00016010996193622913,
+      "loss": 2.1856,
+      "step": 250480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016010848481928311,
+      "loss": 2.2848,
+      "step": 250485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016010700768180294,
+      "loss": 2.089,
+      "step": 250490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001601055305237891,
+      "loss": 2.0027,
+      "step": 250495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016010405334524215,
+      "loss": 2.1037,
+      "step": 250500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016010257614616256,
+      "loss": 1.9858,
+      "step": 250505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016010109892655082,
+      "loss": 2.0746,
+      "step": 250510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016009962168640749,
+      "loss": 1.9858,
+      "step": 250515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016009814442573302,
+      "loss": 2.1086,
+      "step": 250520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016009666714452792,
+      "loss": 2.116,
+      "step": 250525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016009518984279274,
+      "loss": 2.1338,
+      "step": 250530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016009371252052792,
+      "loss": 2.0435,
+      "step": 250535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016009223517773403,
+      "loss": 1.9589,
+      "step": 250540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016009075781441153,
+      "loss": 2.1084,
+      "step": 250545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001600892804305609,
+      "loss": 2.0904,
+      "step": 250550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016008780302618271,
+      "loss": 1.969,
+      "step": 250555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016008632560127745,
+      "loss": 2.0995,
+      "step": 250560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001600848481558456,
+      "loss": 2.1454,
+      "step": 250565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001600833706898877,
+      "loss": 2.1804,
+      "step": 250570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016008189320340419,
+      "loss": 2.118,
+      "step": 250575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016008041569639563,
+      "loss": 1.9901,
+      "step": 250580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016007893816886252,
+      "loss": 1.9929,
+      "step": 250585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016007746062080537,
+      "loss": 2.0063,
+      "step": 250590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016007598305222464,
+      "loss": 2.1477,
+      "step": 250595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016007450546312086,
+      "loss": 2.0159,
+      "step": 250600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00016007302785349457,
+      "loss": 2.0822,
+      "step": 250605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001600715502233462,
+      "loss": 2.017,
+      "step": 250610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016007007257267633,
+      "loss": 2.0676,
+      "step": 250615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016006859490148547,
+      "loss": 1.9321,
+      "step": 250620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016006711720977404,
+      "loss": 2.0931,
+      "step": 250625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001600656394975426,
+      "loss": 2.264,
+      "step": 250630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016006416176479163,
+      "loss": 2.328,
+      "step": 250635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016006268401152173,
+      "loss": 2.0508,
+      "step": 250640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00016006120623773324,
+      "loss": 1.9165,
+      "step": 250645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016005972844342677,
+      "loss": 2.0327,
+      "step": 250650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016005825062860284,
+      "loss": 2.3587,
+      "step": 250655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001600567727932619,
+      "loss": 2.1044,
+      "step": 250660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00016005529493740448,
+      "loss": 2.1216,
+      "step": 250665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016005381706103112,
+      "loss": 1.8123,
+      "step": 250670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00016005233916414224,
+      "loss": 2.0185,
+      "step": 250675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00016005086124673838,
+      "loss": 2.1875,
+      "step": 250680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001600493833088201,
+      "loss": 1.9561,
+      "step": 250685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016004790535038783,
+      "loss": 2.0948,
+      "step": 250690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016004642737144211,
+      "loss": 1.8907,
+      "step": 250695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016004494937198345,
+      "loss": 1.9109,
+      "step": 250700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016004347135201237,
+      "loss": 2.1522,
+      "step": 250705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001600419933115293,
+      "loss": 2.0585,
+      "step": 250710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016004051525053483,
+      "loss": 2.0651,
+      "step": 250715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016003903716902944,
+      "loss": 2.218,
+      "step": 250720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001600375590670136,
+      "loss": 1.7863,
+      "step": 250725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00016003608094448786,
+      "loss": 2.1892,
+      "step": 250730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001600346028014527,
+      "loss": 2.0713,
+      "step": 250735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00016003312463790863,
+      "loss": 2.076,
+      "step": 250740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016003164645385614,
+      "loss": 2.1764,
+      "step": 250745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001600301682492958,
+      "loss": 2.0751,
+      "step": 250750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000160028690024228,
+      "loss": 2.2925,
+      "step": 250755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00016002721177865335,
+      "loss": 2.0972,
+      "step": 250760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001600257335125723,
+      "loss": 2.2228,
+      "step": 250765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016002425522598538,
+      "loss": 2.1749,
+      "step": 250770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016002277691889308,
+      "loss": 2.0432,
+      "step": 250775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00016002129859129594,
+      "loss": 1.9521,
+      "step": 250780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001600198202431944,
+      "loss": 1.9846,
+      "step": 250785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.000160018341874589,
+      "loss": 2.3014,
+      "step": 250790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00016001686348548027,
+      "loss": 1.9832,
+      "step": 250795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016001538507586869,
+      "loss": 2.204,
+      "step": 250800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00016001390664575476,
+      "loss": 1.9668,
+      "step": 250805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016001242819513895,
+      "loss": 2.039,
+      "step": 250810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00016001094972402185,
+      "loss": 2.0433,
+      "step": 250815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00016000947123240392,
+      "loss": 2.149,
+      "step": 250820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.00016000799272028566,
+      "loss": 2.0236,
+      "step": 250825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016000651418766757,
+      "loss": 2.0125,
+      "step": 250830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001600050356345502,
+      "loss": 2.0323,
+      "step": 250835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016000355706093398,
+      "loss": 2.0678,
+      "step": 250840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016000207846681946,
+      "loss": 2.1788,
+      "step": 250845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00016000059985220717,
+      "loss": 2.0195,
+      "step": 250850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015999912121709758,
+      "loss": 2.0635,
+      "step": 250855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001599976425614912,
+      "loss": 2.0795,
+      "step": 250860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015999616388538853,
+      "loss": 2.1605,
+      "step": 250865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001599946851887901,
+      "loss": 1.8528,
+      "step": 250870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001599932064716964,
+      "loss": 1.9879,
+      "step": 250875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001599917277341079,
+      "loss": 2.1989,
+      "step": 250880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015999024897602512,
+      "loss": 2.1593,
+      "step": 250885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015998877019744864,
+      "loss": 2.1961,
+      "step": 250890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015998729139837887,
+      "loss": 2.2013,
+      "step": 250895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015998581257881637,
+      "loss": 2.0338,
+      "step": 250900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015998433373876165,
+      "loss": 2.0584,
+      "step": 250905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015998285487821515,
+      "loss": 2.227,
+      "step": 250910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015998137599717747,
+      "loss": 2.1877,
+      "step": 250915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015997989709564902,
+      "loss": 2.0329,
+      "step": 250920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015997841817363035,
+      "loss": 2.1278,
+      "step": 250925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015997693923112198,
+      "loss": 2.1937,
+      "step": 250930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001599754602681244,
+      "loss": 1.9799,
+      "step": 250935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015997398128463812,
+      "loss": 2.058,
+      "step": 250940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015997250228066363,
+      "loss": 2.1719,
+      "step": 250945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015997102325620143,
+      "loss": 2.0926,
+      "step": 250950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015996954421125206,
+      "loss": 2.1463,
+      "step": 250955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.000159968065145816,
+      "loss": 1.9402,
+      "step": 250960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015996658605989378,
+      "loss": 2.0698,
+      "step": 250965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001599651069534859,
+      "loss": 2.2511,
+      "step": 250970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001599636278265928,
+      "loss": 2.2597,
+      "step": 250975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015996214867921506,
+      "loss": 2.0918,
+      "step": 250980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015996066951135318,
+      "loss": 2.2762,
+      "step": 250985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015995919032300764,
+      "loss": 1.9907,
+      "step": 250990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015995771111417892,
+      "loss": 2.0677,
+      "step": 250995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015995623188486762,
+      "loss": 2.2815,
+      "step": 251000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001599547526350741,
+      "loss": 2.1566,
+      "step": 251005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000159953273364799,
+      "loss": 2.1967,
+      "step": 251010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001599517940740428,
+      "loss": 2.1588,
+      "step": 251015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015995031476280595,
+      "loss": 2.0009,
+      "step": 251020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015994883543108898,
+      "loss": 1.9773,
+      "step": 251025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015994735607889238,
+      "loss": 2.1914,
+      "step": 251030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015994587670621673,
+      "loss": 2.1263,
+      "step": 251035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015994439731306247,
+      "loss": 2.092,
+      "step": 251040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001599429178994301,
+      "loss": 2.205,
+      "step": 251045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015994143846532013,
+      "loss": 2.172,
+      "step": 251050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015993995901073308,
+      "loss": 2.1918,
+      "step": 251055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015993847953566946,
+      "loss": 2.0036,
+      "step": 251060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001599370000401298,
+      "loss": 1.9257,
+      "step": 251065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015993552052411453,
+      "loss": 2.2296,
+      "step": 251070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015993404098762421,
+      "loss": 2.0322,
+      "step": 251075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015993256143065937,
+      "loss": 1.9281,
+      "step": 251080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015993108185322042,
+      "loss": 2.1584,
+      "step": 251085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015992960225530797,
+      "loss": 2.2082,
+      "step": 251090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015992812263692247,
+      "loss": 2.1195,
+      "step": 251095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015992664299806445,
+      "loss": 2.2512,
+      "step": 251100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015992516333873438,
+      "loss": 2.0015,
+      "step": 251105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015992368365893276,
+      "loss": 2.1152,
+      "step": 251110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001599222039586602,
+      "loss": 2.1487,
+      "step": 251115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015992072423791703,
+      "loss": 2.317,
+      "step": 251120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015991924449670393,
+      "loss": 1.9772,
+      "step": 251125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001599177647350213,
+      "loss": 2.2225,
+      "step": 251130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015991628495286967,
+      "loss": 1.9591,
+      "step": 251135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015991480515024957,
+      "loss": 1.9652,
+      "step": 251140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001599133253271615,
+      "loss": 2.0559,
+      "step": 251145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015991184548360592,
+      "loss": 2.0872,
+      "step": 251150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015991036561958336,
+      "loss": 2.1731,
+      "step": 251155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015990888573509437,
+      "loss": 2.1721,
+      "step": 251160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001599074058301394,
+      "loss": 1.9429,
+      "step": 251165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015990592590471895,
+      "loss": 2.1037,
+      "step": 251170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001599044459588336,
+      "loss": 2.0053,
+      "step": 251175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015990296599248377,
+      "loss": 2.1393,
+      "step": 251180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015990148600567,
+      "loss": 2.0736,
+      "step": 251185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001599000059983928,
+      "loss": 2.1589,
+      "step": 251190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001598985259706527,
+      "loss": 2.0988,
+      "step": 251195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001598970459224501,
+      "loss": 1.9904,
+      "step": 251200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015989556585378563,
+      "loss": 2.0738,
+      "step": 251205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015989408576465978,
+      "loss": 2.1321,
+      "step": 251210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000159892605655073,
+      "loss": 2.126,
+      "step": 251215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015989112552502582,
+      "loss": 2.1969,
+      "step": 251220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015988964537451872,
+      "loss": 1.917,
+      "step": 251225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015988816520355227,
+      "loss": 2.0158,
+      "step": 251230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015988668501212692,
+      "loss": 1.9792,
+      "step": 251235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001598852048002432,
+      "loss": 2.0751,
+      "step": 251240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001598837245679016,
+      "loss": 1.9554,
+      "step": 251245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015988224431510263,
+      "loss": 2.0419,
+      "step": 251250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001598807640418468,
+      "loss": 2.034,
+      "step": 251255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001598792837481346,
+      "loss": 2.0715,
+      "step": 251260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015987780343396657,
+      "loss": 1.9942,
+      "step": 251265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001598763230993432,
+      "loss": 2.1517,
+      "step": 251270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015987484274426498,
+      "loss": 2.0768,
+      "step": 251275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015987336236873242,
+      "loss": 1.8083,
+      "step": 251280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015987188197274604,
+      "loss": 2.1357,
+      "step": 251285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015987040155630638,
+      "loss": 2.0794,
+      "step": 251290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015986892111941387,
+      "loss": 1.9357,
+      "step": 251295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015986744066206904,
+      "loss": 2.0238,
+      "step": 251300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015986596018427243,
+      "loss": 2.0771,
+      "step": 251305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001598644796860245,
+      "loss": 1.9867,
+      "step": 251310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015986299916732582,
+      "loss": 2.2,
+      "step": 251315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001598615186281768,
+      "loss": 2.0483,
+      "step": 251320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000159860038068578,
+      "loss": 2.2034,
+      "step": 251325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015985855748852995,
+      "loss": 2.071,
+      "step": 251330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015985707688803314,
+      "loss": 2.0951,
+      "step": 251335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015985559626708807,
+      "loss": 2.1206,
+      "step": 251340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015985411562569523,
+      "loss": 2.1137,
+      "step": 251345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001598526349638551,
+      "loss": 2.1325,
+      "step": 251350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015985115428156829,
+      "loss": 2.0657,
+      "step": 251355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001598496735788352,
+      "loss": 1.9893,
+      "step": 251360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001598481928556564,
+      "loss": 2.1188,
+      "step": 251365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015984671211203237,
+      "loss": 2.0944,
+      "step": 251370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001598452313479636,
+      "loss": 1.9673,
+      "step": 251375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015984375056345062,
+      "loss": 2.2881,
+      "step": 251380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015984226975849394,
+      "loss": 1.9345,
+      "step": 251385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015984078893309405,
+      "loss": 2.1798,
+      "step": 251390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015983930808725146,
+      "loss": 2.0492,
+      "step": 251395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015983782722096666,
+      "loss": 2.0017,
+      "step": 251400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015983634633424018,
+      "loss": 1.792,
+      "step": 251405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015983486542707254,
+      "loss": 2.2011,
+      "step": 251410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001598333844994642,
+      "loss": 2.1411,
+      "step": 251415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015983190355141572,
+      "loss": 2.1668,
+      "step": 251420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015983042258292755,
+      "loss": 2.0741,
+      "step": 251425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015982894159400024,
+      "loss": 2.1152,
+      "step": 251430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015982746058463426,
+      "loss": 2.1734,
+      "step": 251435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015982597955483018,
+      "loss": 2.1049,
+      "step": 251440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001598244985045884,
+      "loss": 1.93,
+      "step": 251445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001598230174339095,
+      "loss": 2.2901,
+      "step": 251450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000159821536342794,
+      "loss": 1.9958,
+      "step": 251455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015982005523124235,
+      "loss": 2.0649,
+      "step": 251460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00015981857409925508,
+      "loss": 2.2142,
+      "step": 251465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015981709294683274,
+      "loss": 1.9441,
+      "step": 251470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015981561177397577,
+      "loss": 2.0636,
+      "step": 251475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001598141305806847,
+      "loss": 1.8519,
+      "step": 251480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015981264936696003,
+      "loss": 1.9629,
+      "step": 251485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001598111681328023,
+      "loss": 2.1785,
+      "step": 251490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015980968687821196,
+      "loss": 1.8992,
+      "step": 251495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015980820560318958,
+      "loss": 2.118,
+      "step": 251500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001598067243077356,
+      "loss": 2.0815,
+      "step": 251505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015980524299185058,
+      "loss": 2.0673,
+      "step": 251510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000159803761655535,
+      "loss": 1.9509,
+      "step": 251515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015980228029878937,
+      "loss": 2.0835,
+      "step": 251520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.796875,
+      "learning_rate": 0.0001598007989216142,
+      "loss": 2.2168,
+      "step": 251525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015979931752401,
+      "loss": 1.9919,
+      "step": 251530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015979783610597725,
+      "loss": 2.0032,
+      "step": 251535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001597963546675165,
+      "loss": 2.0387,
+      "step": 251540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015979487320862818,
+      "loss": 2.251,
+      "step": 251545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001597933917293129,
+      "loss": 2.2321,
+      "step": 251550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015979191022957108,
+      "loss": 2.1001,
+      "step": 251555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015979042870940328,
+      "loss": 2.2615,
+      "step": 251560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015978894716880997,
+      "loss": 2.2584,
+      "step": 251565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015978746560779168,
+      "loss": 2.1185,
+      "step": 251570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001597859840263489,
+      "loss": 1.9933,
+      "step": 251575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015978450242448215,
+      "loss": 2.0986,
+      "step": 251580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015978302080219193,
+      "loss": 2.0861,
+      "step": 251585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015978153915947873,
+      "loss": 2.0833,
+      "step": 251590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001597800574963431,
+      "loss": 2.071,
+      "step": 251595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015977857581278551,
+      "loss": 2.0,
+      "step": 251600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015977709410880647,
+      "loss": 2.0191,
+      "step": 251605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015977561238440647,
+      "loss": 2.2386,
+      "step": 251610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001597741306395861,
+      "loss": 2.3438,
+      "step": 251615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015977264887434572,
+      "loss": 2.0363,
+      "step": 251620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00015977116708868597,
+      "loss": 1.9905,
+      "step": 251625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001597696852826073,
+      "loss": 2.1649,
+      "step": 251630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001597682034561102,
+      "loss": 2.1689,
+      "step": 251635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015976672160919524,
+      "loss": 2.0708,
+      "step": 251640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001597652397418628,
+      "loss": 2.1665,
+      "step": 251645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015976375785411355,
+      "loss": 2.096,
+      "step": 251650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015976227594594787,
+      "loss": 2.1197,
+      "step": 251655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015976079401736637,
+      "loss": 2.1138,
+      "step": 251660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015975931206836945,
+      "loss": 2.1659,
+      "step": 251665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015975783009895768,
+      "loss": 1.9413,
+      "step": 251670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015975634810913155,
+      "loss": 2.0923,
+      "step": 251675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001597548660988916,
+      "loss": 2.0469,
+      "step": 251680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015975338406823826,
+      "loss": 2.1641,
+      "step": 251685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001597519020171721,
+      "loss": 1.9993,
+      "step": 251690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015975041994569357,
+      "loss": 2.1409,
+      "step": 251695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015974893785380325,
+      "loss": 2.277,
+      "step": 251700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001597474557415016,
+      "loss": 2.3297,
+      "step": 251705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015974597360878914,
+      "loss": 2.0414,
+      "step": 251710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015974449145566636,
+      "loss": 1.9488,
+      "step": 251715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015974300928213382,
+      "loss": 1.9796,
+      "step": 251720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015974152708819194,
+      "loss": 2.0274,
+      "step": 251725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015974004487384125,
+      "loss": 2.0158,
+      "step": 251730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015973856263908233,
+      "loss": 2.1004,
+      "step": 251735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015973708038391563,
+      "loss": 1.9583,
+      "step": 251740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00015973559810834165,
+      "loss": 1.9657,
+      "step": 251745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001597341158123609,
+      "loss": 1.9195,
+      "step": 251750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015973263349597388,
+      "loss": 2.1277,
+      "step": 251755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015973115115918113,
+      "loss": 2.1766,
+      "step": 251760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015972966880198312,
+      "loss": 1.9001,
+      "step": 251765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001597281864243804,
+      "loss": 2.0187,
+      "step": 251770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015972670402637342,
+      "loss": 2.103,
+      "step": 251775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001597252216079627,
+      "loss": 2.101,
+      "step": 251780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001597237391691488,
+      "loss": 2.1551,
+      "step": 251785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015972225670993217,
+      "loss": 2.1075,
+      "step": 251790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015972077423031332,
+      "loss": 2.1685,
+      "step": 251795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015971929173029282,
+      "loss": 1.9976,
+      "step": 251800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015971780920987104,
+      "loss": 2.3116,
+      "step": 251805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015971632666904865,
+      "loss": 2.1115,
+      "step": 251810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015971484410782605,
+      "loss": 2.1678,
+      "step": 251815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001597133615262038,
+      "loss": 2.006,
+      "step": 251820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015971187892418237,
+      "loss": 2.1649,
+      "step": 251825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015971039630176227,
+      "loss": 2.0805,
+      "step": 251830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015970891365894402,
+      "loss": 2.194,
+      "step": 251835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001597074309957281,
+      "loss": 2.0472,
+      "step": 251840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001597059483121151,
+      "loss": 1.993,
+      "step": 251845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001597044656081054,
+      "loss": 2.155,
+      "step": 251850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001597029828836996,
+      "loss": 2.2526,
+      "step": 251855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015970150013889817,
+      "loss": 2.0008,
+      "step": 251860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015970001737370163,
+      "loss": 2.0977,
+      "step": 251865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001596985345881105,
+      "loss": 1.9942,
+      "step": 251870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015969705178212523,
+      "loss": 2.0752,
+      "step": 251875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001596955689557464,
+      "loss": 2.2065,
+      "step": 251880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015969408610897446,
+      "loss": 2.1156,
+      "step": 251885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015969260324180993,
+      "loss": 2.1147,
+      "step": 251890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015969112035425336,
+      "loss": 2.1721,
+      "step": 251895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015968963744630518,
+      "loss": 2.0941,
+      "step": 251900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015968815451796595,
+      "loss": 1.9414,
+      "step": 251905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015968667156923618,
+      "loss": 1.913,
+      "step": 251910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015968518860011633,
+      "loss": 2.1832,
+      "step": 251915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015968370561060698,
+      "loss": 2.238,
+      "step": 251920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015968222260070856,
+      "loss": 2.1188,
+      "step": 251925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015968073957042163,
+      "loss": 1.9834,
+      "step": 251930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001596792565197467,
+      "loss": 1.9456,
+      "step": 251935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015967777344868416,
+      "loss": 2.118,
+      "step": 251940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00015967629035723468,
+      "loss": 1.9601,
+      "step": 251945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001596748072453987,
+      "loss": 2.0772,
+      "step": 251950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015967332411317671,
+      "loss": 2.1303,
+      "step": 251955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015967184096056923,
+      "loss": 2.0719,
+      "step": 251960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015967035778757677,
+      "loss": 2.0221,
+      "step": 251965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015966887459419984,
+      "loss": 1.9819,
+      "step": 251970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015966739138043894,
+      "loss": 1.9628,
+      "step": 251975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015966590814629457,
+      "loss": 1.9712,
+      "step": 251980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015966442489176723,
+      "loss": 2.1492,
+      "step": 251985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015966294161685746,
+      "loss": 1.8165,
+      "step": 251990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015966145832156575,
+      "loss": 2.0329,
+      "step": 251995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001596599750058926,
+      "loss": 2.0888,
+      "step": 252000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015965849166983852,
+      "loss": 2.099,
+      "step": 252005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000159657008313404,
+      "loss": 2.139,
+      "step": 252010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001596555249365896,
+      "loss": 2.1556,
+      "step": 252015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015965404153939574,
+      "loss": 2.0652,
+      "step": 252020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000159652558121823,
+      "loss": 1.9825,
+      "step": 252025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015965107468387187,
+      "loss": 2.1332,
+      "step": 252030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015964959122554285,
+      "loss": 2.1126,
+      "step": 252035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015964810774683646,
+      "loss": 2.1855,
+      "step": 252040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00015964662424775318,
+      "loss": 2.2262,
+      "step": 252045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015964514072829353,
+      "loss": 2.1727,
+      "step": 252050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000159643657188458,
+      "loss": 2.2923,
+      "step": 252055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015964217362824718,
+      "loss": 2.0987,
+      "step": 252060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015964069004766146,
+      "loss": 2.2015,
+      "step": 252065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015963920644670142,
+      "loss": 2.2311,
+      "step": 252070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001596377228253675,
+      "loss": 1.9827,
+      "step": 252075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015963623918366032,
+      "loss": 2.0656,
+      "step": 252080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015963475552158027,
+      "loss": 2.2417,
+      "step": 252085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015963327183912793,
+      "loss": 1.9916,
+      "step": 252090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015963178813630378,
+      "loss": 2.0208,
+      "step": 252095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015963030441310832,
+      "loss": 2.2356,
+      "step": 252100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015962882066954205,
+      "loss": 2.3867,
+      "step": 252105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001596273369056055,
+      "loss": 1.9624,
+      "step": 252110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001596258531212992,
+      "loss": 2.1091,
+      "step": 252115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015962436931662364,
+      "loss": 2.0265,
+      "step": 252120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015962288549157926,
+      "loss": 2.0649,
+      "step": 252125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015962140164616667,
+      "loss": 2.0196,
+      "step": 252130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015961991778038628,
+      "loss": 2.1189,
+      "step": 252135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015961843389423867,
+      "loss": 1.9697,
+      "step": 252140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015961694998772432,
+      "loss": 2.0704,
+      "step": 252145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015961546606084375,
+      "loss": 2.1739,
+      "step": 252150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015961398211359744,
+      "loss": 2.1838,
+      "step": 252155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015961249814598592,
+      "loss": 2.0774,
+      "step": 252160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015961101415800969,
+      "loss": 2.0923,
+      "step": 252165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015960953014966924,
+      "loss": 2.013,
+      "step": 252170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015960804612096511,
+      "loss": 2.0533,
+      "step": 252175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001596065620718978,
+      "loss": 2.1012,
+      "step": 252180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015960507800246777,
+      "loss": 2.1684,
+      "step": 252185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001596035939126756,
+      "loss": 2.1392,
+      "step": 252190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015960210980252176,
+      "loss": 2.1521,
+      "step": 252195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015960062567200675,
+      "loss": 2.0447,
+      "step": 252200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001595991415211311,
+      "loss": 2.0847,
+      "step": 252205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015959765734989526,
+      "loss": 2.1361,
+      "step": 252210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001595961731582998,
+      "loss": 2.1502,
+      "step": 252215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015959468894634522,
+      "loss": 2.12,
+      "step": 252220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000159593204714032,
+      "loss": 2.2033,
+      "step": 252225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015959172046136068,
+      "loss": 2.1311,
+      "step": 252230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001595902361883317,
+      "loss": 2.1387,
+      "step": 252235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015958875189494567,
+      "loss": 2.0924,
+      "step": 252240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000159587267581203,
+      "loss": 2.0361,
+      "step": 252245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015958578324710428,
+      "loss": 2.0267,
+      "step": 252250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015958429889264996,
+      "loss": 2.0533,
+      "step": 252255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015958281451784054,
+      "loss": 2.0961,
+      "step": 252260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015958133012267657,
+      "loss": 2.0326,
+      "step": 252265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015957984570715854,
+      "loss": 2.0999,
+      "step": 252270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015957836127128696,
+      "loss": 2.2405,
+      "step": 252275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015957687681506232,
+      "loss": 2.0859,
+      "step": 252280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015957539233848512,
+      "loss": 2.2768,
+      "step": 252285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001595739078415559,
+      "loss": 2.1692,
+      "step": 252290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015957242332427515,
+      "loss": 1.9057,
+      "step": 252295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00015957093878664339,
+      "loss": 2.0784,
+      "step": 252300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015956945422866112,
+      "loss": 2.0512,
+      "step": 252305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015956796965032886,
+      "loss": 1.9347,
+      "step": 252310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015956648505164705,
+      "loss": 2.1072,
+      "step": 252315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001595650004326163,
+      "loss": 2.19,
+      "step": 252320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015956351579323702,
+      "loss": 2.1033,
+      "step": 252325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001595620311335098,
+      "loss": 2.0666,
+      "step": 252330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001595605464534351,
+      "loss": 2.0586,
+      "step": 252335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015955906175301344,
+      "loss": 2.1552,
+      "step": 252340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015955757703224529,
+      "loss": 2.1183,
+      "step": 252345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015955609229113121,
+      "loss": 2.1198,
+      "step": 252350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001595546075296717,
+      "loss": 2.1536,
+      "step": 252355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015955312274786726,
+      "loss": 2.08,
+      "step": 252360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015955163794571838,
+      "loss": 1.9176,
+      "step": 252365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001595501531232256,
+      "loss": 2.1442,
+      "step": 252370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001595486682803894,
+      "loss": 2.2035,
+      "step": 252375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015954718341721028,
+      "loss": 2.1096,
+      "step": 252380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015954569853368877,
+      "loss": 2.0678,
+      "step": 252385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015954421362982538,
+      "loss": 2.1821,
+      "step": 252390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001595427287056206,
+      "loss": 2.0127,
+      "step": 252395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015954124376107496,
+      "loss": 2.1446,
+      "step": 252400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015953975879618893,
+      "loss": 2.1725,
+      "step": 252405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015953827381096303,
+      "loss": 2.0704,
+      "step": 252410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001595367888053978,
+      "loss": 2.37,
+      "step": 252415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015953530377949368,
+      "loss": 1.9053,
+      "step": 252420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015953381873325126,
+      "loss": 2.0475,
+      "step": 252425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.000159532333666671,
+      "loss": 2.0109,
+      "step": 252430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015953084857975343,
+      "loss": 2.1404,
+      "step": 252435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015952936347249903,
+      "loss": 1.8649,
+      "step": 252440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001595278783449083,
+      "loss": 2.1713,
+      "step": 252445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015952639319698176,
+      "loss": 2.164,
+      "step": 252450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015952490802871995,
+      "loss": 2.0598,
+      "step": 252455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015952342284012336,
+      "loss": 2.0701,
+      "step": 252460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001595219376311925,
+      "loss": 2.1155,
+      "step": 252465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015952045240192787,
+      "loss": 2.004,
+      "step": 252470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001595189671523299,
+      "loss": 2.094,
+      "step": 252475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015951748188239924,
+      "loss": 2.1287,
+      "step": 252480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001595159965921363,
+      "loss": 2.0485,
+      "step": 252485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015951451128154163,
+      "loss": 2.2208,
+      "step": 252490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015951302595061572,
+      "loss": 2.0216,
+      "step": 252495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001595115405993591,
+      "loss": 2.1909,
+      "step": 252500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015951005522777222,
+      "loss": 2.0451,
+      "step": 252505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015950856983585566,
+      "loss": 2.1315,
+      "step": 252510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015950708442360987,
+      "loss": 1.9218,
+      "step": 252515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001595055989910354,
+      "loss": 1.9114,
+      "step": 252520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015950411353813274,
+      "loss": 2.2,
+      "step": 252525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015950262806490234,
+      "loss": 1.9687,
+      "step": 252530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001595011425713448,
+      "loss": 2.0158,
+      "step": 252535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015949965705746062,
+      "loss": 2.1369,
+      "step": 252540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015949817152325024,
+      "loss": 2.0184,
+      "step": 252545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015949668596871422,
+      "loss": 1.9434,
+      "step": 252550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015949520039385307,
+      "loss": 2.0965,
+      "step": 252555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015949371479866725,
+      "loss": 1.9684,
+      "step": 252560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015949222918315733,
+      "loss": 2.1162,
+      "step": 252565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015949074354732376,
+      "loss": 2.0943,
+      "step": 252570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001594892578911671,
+      "loss": 2.2006,
+      "step": 252575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015948777221468777,
+      "loss": 2.1777,
+      "step": 252580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001594862865178864,
+      "loss": 2.1881,
+      "step": 252585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001594848008007634,
+      "loss": 2.0449,
+      "step": 252590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00015948331506331935,
+      "loss": 2.2308,
+      "step": 252595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001594818293055547,
+      "loss": 2.1885,
+      "step": 252600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015948034352746996,
+      "loss": 2.0264,
+      "step": 252605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015947885772906568,
+      "loss": 2.0211,
+      "step": 252610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015947737191034233,
+      "loss": 2.1013,
+      "step": 252615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015947588607130044,
+      "loss": 2.1049,
+      "step": 252620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015947440021194052,
+      "loss": 1.9822,
+      "step": 252625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015947291433226305,
+      "loss": 2.2905,
+      "step": 252630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015947142843226854,
+      "loss": 2.0553,
+      "step": 252635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015946994251195756,
+      "loss": 1.936,
+      "step": 252640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001594684565713305,
+      "loss": 2.0142,
+      "step": 252645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000159466970610388,
+      "loss": 2.1442,
+      "step": 252650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015946548462913046,
+      "loss": 2.1392,
+      "step": 252655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015946399862755848,
+      "loss": 1.9438,
+      "step": 252660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015946251260567245,
+      "loss": 2.1439,
+      "step": 252665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000159461026563473,
+      "loss": 2.03,
+      "step": 252670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001594595405009606,
+      "loss": 1.9713,
+      "step": 252675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001594580544181357,
+      "loss": 1.9498,
+      "step": 252680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015945656831499886,
+      "loss": 2.0775,
+      "step": 252685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015945508219155057,
+      "loss": 1.9094,
+      "step": 252690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015945359604779138,
+      "loss": 2.0307,
+      "step": 252695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015945210988372175,
+      "loss": 2.2087,
+      "step": 252700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015945062369934217,
+      "loss": 1.8994,
+      "step": 252705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015944913749465323,
+      "loss": 2.1025,
+      "step": 252710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015944765126965532,
+      "loss": 2.1389,
+      "step": 252715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015944616502434905,
+      "loss": 2.2125,
+      "step": 252720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015944467875873491,
+      "loss": 2.0692,
+      "step": 252725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015944319247281337,
+      "loss": 2.1247,
+      "step": 252730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015944170616658497,
+      "loss": 2.0503,
+      "step": 252735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001594402198400502,
+      "loss": 2.0037,
+      "step": 252740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015943873349320956,
+      "loss": 2.1196,
+      "step": 252745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015943724712606358,
+      "loss": 2.3367,
+      "step": 252750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015943576073861276,
+      "loss": 2.1636,
+      "step": 252755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001594342743308576,
+      "loss": 1.9997,
+      "step": 252760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015943278790279863,
+      "loss": 2.2046,
+      "step": 252765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001594313014544363,
+      "loss": 1.9881,
+      "step": 252770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015942981498577122,
+      "loss": 1.9309,
+      "step": 252775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001594283284968038,
+      "loss": 2.1146,
+      "step": 252780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015942684198753463,
+      "loss": 2.0507,
+      "step": 252785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001594253554579641,
+      "loss": 1.9488,
+      "step": 252790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001594238689080928,
+      "loss": 1.8954,
+      "step": 252795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015942238233792127,
+      "loss": 2.2077,
+      "step": 252800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015942089574744997,
+      "loss": 1.9944,
+      "step": 252805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001594194091366794,
+      "loss": 2.315,
+      "step": 252810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001594179225056101,
+      "loss": 2.0331,
+      "step": 252815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015941643585424254,
+      "loss": 2.1678,
+      "step": 252820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015941494918257724,
+      "loss": 2.1828,
+      "step": 252825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015941346249061475,
+      "loss": 1.855,
+      "step": 252830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001594119757783555,
+      "loss": 2.174,
+      "step": 252835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001594104890458001,
+      "loss": 2.0387,
+      "step": 252840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015940900229294898,
+      "loss": 2.1073,
+      "step": 252845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015940751551980263,
+      "loss": 2.1289,
+      "step": 252850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015940602872636163,
+      "loss": 1.9286,
+      "step": 252855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015940454191262643,
+      "loss": 2.1298,
+      "step": 252860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015940305507859757,
+      "loss": 2.1769,
+      "step": 252865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015940156822427556,
+      "loss": 1.9893,
+      "step": 252870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015940008134966092,
+      "loss": 1.99,
+      "step": 252875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015939859445475408,
+      "loss": 2.1068,
+      "step": 252880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015939710753955564,
+      "loss": 2.2404,
+      "step": 252885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015939562060406605,
+      "loss": 1.977,
+      "step": 252890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015939413364828586,
+      "loss": 2.194,
+      "step": 252895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015939264667221554,
+      "loss": 2.0628,
+      "step": 252900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001593911596758556,
+      "loss": 1.928,
+      "step": 252905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001593896726592066,
+      "loss": 2.2035,
+      "step": 252910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000159388185622269,
+      "loss": 2.0418,
+      "step": 252915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001593866985650433,
+      "loss": 2.0935,
+      "step": 252920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015938521148753005,
+      "loss": 2.0702,
+      "step": 252925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015938372438972968,
+      "loss": 2.0256,
+      "step": 252930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001593822372716428,
+      "loss": 2.0374,
+      "step": 252935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015938075013326988,
+      "loss": 2.2381,
+      "step": 252940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001593792629746114,
+      "loss": 1.9277,
+      "step": 252945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001593777757956679,
+      "loss": 2.1623,
+      "step": 252950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015937628859643988,
+      "loss": 2.0046,
+      "step": 252955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015937480137692778,
+      "loss": 2.0531,
+      "step": 252960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015937331413713225,
+      "loss": 2.081,
+      "step": 252965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001593718268770537,
+      "loss": 2.1196,
+      "step": 252970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015937033959669266,
+      "loss": 1.9931,
+      "step": 252975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001593688522960496,
+      "loss": 2.0939,
+      "step": 252980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015936736497512507,
+      "loss": 2.1748,
+      "step": 252985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001593658776339196,
+      "loss": 2.0575,
+      "step": 252990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015936439027243367,
+      "loss": 1.9551,
+      "step": 252995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015936290289066777,
+      "loss": 2.0004,
+      "step": 253000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015936141548862244,
+      "loss": 2.2452,
+      "step": 253005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015935992806629816,
+      "loss": 2.1059,
+      "step": 253010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015935844062369542,
+      "loss": 2.1556,
+      "step": 253015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015935695316081483,
+      "loss": 2.3003,
+      "step": 253020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001593554656776568,
+      "loss": 1.9721,
+      "step": 253025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015935397817422184,
+      "loss": 2.1363,
+      "step": 253030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001593524906505105,
+      "loss": 2.1288,
+      "step": 253035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001593510031065233,
+      "loss": 2.0586,
+      "step": 253040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001593495155422607,
+      "loss": 2.1084,
+      "step": 253045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001593480279577232,
+      "loss": 2.1553,
+      "step": 253050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015934654035291138,
+      "loss": 2.1358,
+      "step": 253055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015934505272782569,
+      "loss": 2.0576,
+      "step": 253060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015934356508246665,
+      "loss": 2.0052,
+      "step": 253065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015934207741683478,
+      "loss": 2.0748,
+      "step": 253070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015934058973093057,
+      "loss": 2.1363,
+      "step": 253075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00015933910202475454,
+      "loss": 2.1329,
+      "step": 253080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015933761429830722,
+      "loss": 2.1316,
+      "step": 253085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015933612655158908,
+      "loss": 1.9727,
+      "step": 253090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015933463878460064,
+      "loss": 2.071,
+      "step": 253095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001593331509973424,
+      "loss": 2.232,
+      "step": 253100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001593316631898149,
+      "loss": 2.1291,
+      "step": 253105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015933017536201862,
+      "loss": 2.0094,
+      "step": 253110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015932868751395408,
+      "loss": 2.1302,
+      "step": 253115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015932719964562177,
+      "loss": 2.0145,
+      "step": 253120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015932571175702223,
+      "loss": 2.0578,
+      "step": 253125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015932422384815593,
+      "loss": 2.2259,
+      "step": 253130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015932273591902344,
+      "loss": 1.9961,
+      "step": 253135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015932124796962517,
+      "loss": 2.1295,
+      "step": 253140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015931975999996172,
+      "loss": 2.1677,
+      "step": 253145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015931827201003357,
+      "loss": 2.1821,
+      "step": 253150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001593167839998412,
+      "loss": 1.9363,
+      "step": 253155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015931529596938514,
+      "loss": 2.0836,
+      "step": 253160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015931380791866593,
+      "loss": 2.1414,
+      "step": 253165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015931231984768403,
+      "loss": 2.0924,
+      "step": 253170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015931083175643995,
+      "loss": 1.9795,
+      "step": 253175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001593093436449342,
+      "loss": 2.0491,
+      "step": 253180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015930785551316739,
+      "loss": 2.0745,
+      "step": 253185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015930636736113986,
+      "loss": 2.2482,
+      "step": 253190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015930487918885223,
+      "loss": 2.1397,
+      "step": 253195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015930339099630496,
+      "loss": 2.0561,
+      "step": 253200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001593019027834986,
+      "loss": 2.1013,
+      "step": 253205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001593004145504336,
+      "loss": 2.2289,
+      "step": 253210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001592989262971105,
+      "loss": 2.0299,
+      "step": 253215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015929743802352989,
+      "loss": 2.0393,
+      "step": 253220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001592959497296921,
+      "loss": 2.0277,
+      "step": 253225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001592944614155978,
+      "loss": 2.1147,
+      "step": 253230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015929297308124742,
+      "loss": 2.1113,
+      "step": 253235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001592914847266415,
+      "loss": 2.1338,
+      "step": 253240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001592899963517805,
+      "loss": 1.9428,
+      "step": 253245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015928850795666498,
+      "loss": 2.0581,
+      "step": 253250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015928701954129543,
+      "loss": 1.992,
+      "step": 253255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015928553110567236,
+      "loss": 2.036,
+      "step": 253260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015928404264979628,
+      "loss": 2.0118,
+      "step": 253265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015928255417366769,
+      "loss": 2.0734,
+      "step": 253270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015928106567728712,
+      "loss": 2.0615,
+      "step": 253275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015927957716065503,
+      "loss": 2.1898,
+      "step": 253280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015927808862377198,
+      "loss": 2.0725,
+      "step": 253285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015927660006663847,
+      "loss": 2.083,
+      "step": 253290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000159275111489255,
+      "loss": 1.8864,
+      "step": 253295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015927362289162208,
+      "loss": 2.0295,
+      "step": 253300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015927213427374017,
+      "loss": 1.9445,
+      "step": 253305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00015927064563560985,
+      "loss": 1.8842,
+      "step": 253310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001592691569772316,
+      "loss": 2.0881,
+      "step": 253315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015926766829860595,
+      "loss": 2.1906,
+      "step": 253320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001592661795997334,
+      "loss": 2.1834,
+      "step": 253325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001592646908806144,
+      "loss": 1.982,
+      "step": 253330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015926320214124955,
+      "loss": 1.863,
+      "step": 253335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015926171338163928,
+      "loss": 2.2061,
+      "step": 253340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015926022460178416,
+      "loss": 2.0964,
+      "step": 253345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015925873580168468,
+      "loss": 2.2603,
+      "step": 253350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015925724698134133,
+      "loss": 2.0179,
+      "step": 253355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.90625,
+      "learning_rate": 0.0001592557581407546,
+      "loss": 2.2249,
+      "step": 253360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015925426927992505,
+      "loss": 2.1211,
+      "step": 253365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015925278039885318,
+      "loss": 2.1135,
+      "step": 253370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015925129149753947,
+      "loss": 2.0939,
+      "step": 253375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015924980257598446,
+      "loss": 1.9488,
+      "step": 253380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015924831363418865,
+      "loss": 2.0562,
+      "step": 253385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015924682467215252,
+      "loss": 2.0832,
+      "step": 253390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001592453356898766,
+      "loss": 2.2159,
+      "step": 253395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001592438466873614,
+      "loss": 2.1327,
+      "step": 253400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015924235766460744,
+      "loss": 2.0861,
+      "step": 253405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015924086862161523,
+      "loss": 2.2233,
+      "step": 253410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015923937955838522,
+      "loss": 2.1779,
+      "step": 253415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015923789047491799,
+      "loss": 2.1172,
+      "step": 253420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00015923640137121403,
+      "loss": 2.0287,
+      "step": 253425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015923491224727383,
+      "loss": 1.9244,
+      "step": 253430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001592334231030979,
+      "loss": 2.1265,
+      "step": 253435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015923193393868677,
+      "loss": 2.0503,
+      "step": 253440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015923044475404092,
+      "loss": 2.2646,
+      "step": 253445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001592289555491609,
+      "loss": 1.9391,
+      "step": 253450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00015922746632404718,
+      "loss": 2.2208,
+      "step": 253455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015922597707870032,
+      "loss": 2.1469,
+      "step": 253460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015922448781312073,
+      "loss": 1.9687,
+      "step": 253465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015922299852730903,
+      "loss": 1.9482,
+      "step": 253470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015922150922126566,
+      "loss": 2.0407,
+      "step": 253475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015922001989499113,
+      "loss": 2.0471,
+      "step": 253480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.000159218530548486,
+      "loss": 2.1765,
+      "step": 253485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015921704118175076,
+      "loss": 2.014,
+      "step": 253490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015921555179478585,
+      "loss": 2.0323,
+      "step": 253495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015921406238759186,
+      "loss": 2.0929,
+      "step": 253500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015921257296016927,
+      "loss": 2.0595,
+      "step": 253505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001592110835125186,
+      "loss": 2.0858,
+      "step": 253510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015920959404464036,
+      "loss": 2.1385,
+      "step": 253515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015920810455653502,
+      "loss": 2.2183,
+      "step": 253520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015920661504820313,
+      "loss": 2.1529,
+      "step": 253525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001592051255196452,
+      "loss": 2.1769,
+      "step": 253530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001592036359708617,
+      "loss": 2.1464,
+      "step": 253535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001592021464018532,
+      "loss": 1.9733,
+      "step": 253540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015920065681262013,
+      "loss": 2.2605,
+      "step": 253545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015919916720316305,
+      "loss": 1.9672,
+      "step": 253550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001591976775734825,
+      "loss": 1.9887,
+      "step": 253555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015919618792357892,
+      "loss": 2.1515,
+      "step": 253560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015919469825345287,
+      "loss": 2.0415,
+      "step": 253565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001591932085631048,
+      "loss": 2.1757,
+      "step": 253570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015919171885253526,
+      "loss": 2.1522,
+      "step": 253575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001591902291217448,
+      "loss": 2.0821,
+      "step": 253580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015918873937073388,
+      "loss": 2.0076,
+      "step": 253585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015918724959950296,
+      "loss": 2.0031,
+      "step": 253590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015918575980805263,
+      "loss": 2.0401,
+      "step": 253595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001591842699963834,
+      "loss": 2.1917,
+      "step": 253600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001591827801644957,
+      "loss": 1.9466,
+      "step": 253605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015918129031239011,
+      "loss": 2.013,
+      "step": 253610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015917980044006714,
+      "loss": 1.9291,
+      "step": 253615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015917831054752723,
+      "loss": 1.989,
+      "step": 253620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015917682063477096,
+      "loss": 1.9949,
+      "step": 253625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015917533070179882,
+      "loss": 2.0702,
+      "step": 253630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015917384074861133,
+      "loss": 2.037,
+      "step": 253635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015917235077520894,
+      "loss": 2.1237,
+      "step": 253640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015917086078159223,
+      "loss": 2.1471,
+      "step": 253645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015916937076776167,
+      "loss": 2.0678,
+      "step": 253650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001591678807337178,
+      "loss": 2.0062,
+      "step": 253655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015916639067946109,
+      "loss": 2.1068,
+      "step": 253660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015916490060499208,
+      "loss": 2.0126,
+      "step": 253665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015916341051031124,
+      "loss": 2.0463,
+      "step": 253670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015916192039541914,
+      "loss": 1.9809,
+      "step": 253675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015916043026031624,
+      "loss": 1.9738,
+      "step": 253680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015915894010500305,
+      "loss": 2.0297,
+      "step": 253685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015915744992948011,
+      "loss": 2.186,
+      "step": 253690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015915595973374792,
+      "loss": 2.1608,
+      "step": 253695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015915446951780696,
+      "loss": 2.1519,
+      "step": 253700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001591529792816578,
+      "loss": 2.1065,
+      "step": 253705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015915148902530087,
+      "loss": 2.0811,
+      "step": 253710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015914999874873674,
+      "loss": 2.1342,
+      "step": 253715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015914850845196588,
+      "loss": 2.1032,
+      "step": 253720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015914701813498882,
+      "loss": 2.1589,
+      "step": 253725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001591455277978061,
+      "loss": 2.0263,
+      "step": 253730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015914403744041815,
+      "loss": 2.0173,
+      "step": 253735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015914254706282552,
+      "loss": 2.0822,
+      "step": 253740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015914105666502875,
+      "loss": 2.0067,
+      "step": 253745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001591395662470283,
+      "loss": 2.0305,
+      "step": 253750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00015913807580882473,
+      "loss": 1.957,
+      "step": 253755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001591365853504185,
+      "loss": 1.7239,
+      "step": 253760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015913509487181014,
+      "loss": 2.2693,
+      "step": 253765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015913360437300018,
+      "loss": 2.142,
+      "step": 253770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001591321138539891,
+      "loss": 2.0337,
+      "step": 253775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015913062331477743,
+      "loss": 1.9832,
+      "step": 253780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015912913275536563,
+      "loss": 2.1081,
+      "step": 253785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015912764217575423,
+      "loss": 2.1047,
+      "step": 253790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015912615157594382,
+      "loss": 2.0926,
+      "step": 253795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015912466095593483,
+      "loss": 2.1745,
+      "step": 253800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015912317031572775,
+      "loss": 2.1871,
+      "step": 253805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015912167965532317,
+      "loss": 2.1794,
+      "step": 253810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015912018897472149,
+      "loss": 2.1562,
+      "step": 253815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001591186982739233,
+      "loss": 2.2159,
+      "step": 253820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015911720755292912,
+      "loss": 2.1252,
+      "step": 253825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015911571681173942,
+      "loss": 2.2314,
+      "step": 253830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015911422605035473,
+      "loss": 1.9596,
+      "step": 253835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015911273526877552,
+      "loss": 1.9001,
+      "step": 253840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015911124446700233,
+      "loss": 2.1808,
+      "step": 253845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001591097536450357,
+      "loss": 2.1504,
+      "step": 253850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015910826280287607,
+      "loss": 1.9192,
+      "step": 253855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.000159106771940524,
+      "loss": 2.0305,
+      "step": 253860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015910528105798,
+      "loss": 2.2244,
+      "step": 253865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015910379015524452,
+      "loss": 2.1539,
+      "step": 253870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015910229923231815,
+      "loss": 2.1814,
+      "step": 253875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015910080828920135,
+      "loss": 2.1222,
+      "step": 253880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015909931732589465,
+      "loss": 2.1012,
+      "step": 253885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015909782634239856,
+      "loss": 2.0277,
+      "step": 253890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015909633533871357,
+      "loss": 2.1784,
+      "step": 253895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015909484431484016,
+      "loss": 2.2216,
+      "step": 253900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001590933532707789,
+      "loss": 1.9155,
+      "step": 253905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001590918622065303,
+      "loss": 1.9758,
+      "step": 253910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015909037112209488,
+      "loss": 2.1816,
+      "step": 253915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015908888001747306,
+      "loss": 2.0078,
+      "step": 253920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015908738889266544,
+      "loss": 2.0522,
+      "step": 253925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015908589774767248,
+      "loss": 1.9652,
+      "step": 253930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001590844065824947,
+      "loss": 1.9235,
+      "step": 253935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001590829153971326,
+      "loss": 2.2779,
+      "step": 253940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015908142419158677,
+      "loss": 1.9812,
+      "step": 253945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001590799329658576,
+      "loss": 2.1611,
+      "step": 253950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015907844171994567,
+      "loss": 2.0816,
+      "step": 253955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015907695045385147,
+      "loss": 2.143,
+      "step": 253960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015907545916757552,
+      "loss": 2.2584,
+      "step": 253965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015907396786111832,
+      "loss": 2.0841,
+      "step": 253970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015907247653448037,
+      "loss": 2.1516,
+      "step": 253975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001590709851876622,
+      "loss": 2.0684,
+      "step": 253980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015906949382066428,
+      "loss": 2.1331,
+      "step": 253985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001590680024334872,
+      "loss": 1.8799,
+      "step": 253990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001590665110261314,
+      "loss": 2.1603,
+      "step": 253995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001590650195985974,
+      "loss": 1.9982,
+      "step": 254000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001590635281508857,
+      "loss": 1.8934,
+      "step": 254005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015906203668299685,
+      "loss": 2.1134,
+      "step": 254010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015906054519493137,
+      "loss": 1.8827,
+      "step": 254015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001590590536866897,
+      "loss": 2.0409,
+      "step": 254020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001590575621582724,
+      "loss": 2.1226,
+      "step": 254025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015905607060967992,
+      "loss": 2.1232,
+      "step": 254030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015905457904091285,
+      "loss": 2.1547,
+      "step": 254035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015905308745197167,
+      "loss": 1.9748,
+      "step": 254040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015905159584285689,
+      "loss": 1.8612,
+      "step": 254045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015905010421356897,
+      "loss": 2.1143,
+      "step": 254050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001590486125641085,
+      "loss": 2.1884,
+      "step": 254055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015904712089447599,
+      "loss": 2.1162,
+      "step": 254060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015904562920467186,
+      "loss": 2.0745,
+      "step": 254065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015904413749469668,
+      "loss": 2.1372,
+      "step": 254070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015904264576455094,
+      "loss": 2.0424,
+      "step": 254075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015904115401423521,
+      "loss": 2.1914,
+      "step": 254080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001590396622437499,
+      "loss": 2.1725,
+      "step": 254085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015903817045309559,
+      "loss": 1.957,
+      "step": 254090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015903667864227276,
+      "loss": 1.9665,
+      "step": 254095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015903518681128196,
+      "loss": 1.902,
+      "step": 254100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015903369496012365,
+      "loss": 1.9686,
+      "step": 254105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015903220308879833,
+      "loss": 2.147,
+      "step": 254110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015903071119730656,
+      "loss": 2.1302,
+      "step": 254115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015902921928564884,
+      "loss": 2.0379,
+      "step": 254120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001590277273538257,
+      "loss": 2.1177,
+      "step": 254125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015902623540183755,
+      "loss": 1.9667,
+      "step": 254130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015902474342968501,
+      "loss": 2.2029,
+      "step": 254135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015902325143736854,
+      "loss": 2.011,
+      "step": 254140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015902175942488866,
+      "loss": 2.3041,
+      "step": 254145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015902026739224586,
+      "loss": 2.169,
+      "step": 254150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015901877533944068,
+      "loss": 2.0858,
+      "step": 254155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015901728326647363,
+      "loss": 2.1485,
+      "step": 254160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015901579117334516,
+      "loss": 2.0465,
+      "step": 254165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015901429906005587,
+      "loss": 2.2947,
+      "step": 254170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001590128069266062,
+      "loss": 1.8982,
+      "step": 254175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015901131477299668,
+      "loss": 2.043,
+      "step": 254180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015900982259922785,
+      "loss": 2.0654,
+      "step": 254185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015900833040530019,
+      "loss": 1.8409,
+      "step": 254190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001590068381912142,
+      "loss": 1.9545,
+      "step": 254195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001590053459569704,
+      "loss": 1.9744,
+      "step": 254200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001590038537025693,
+      "loss": 2.1115,
+      "step": 254205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015900236142801143,
+      "loss": 2.0583,
+      "step": 254210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015900086913329728,
+      "loss": 1.9782,
+      "step": 254215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015899937681842737,
+      "loss": 2.0191,
+      "step": 254220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.765625,
+      "learning_rate": 0.0001589978844834022,
+      "loss": 2.1652,
+      "step": 254225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001589963921282223,
+      "loss": 2.1229,
+      "step": 254230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015899489975288814,
+      "loss": 1.9703,
+      "step": 254235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015899340735740024,
+      "loss": 2.1778,
+      "step": 254240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015899191494175914,
+      "loss": 1.9862,
+      "step": 254245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015899042250596532,
+      "loss": 2.0869,
+      "step": 254250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001589889300500193,
+      "loss": 2.0757,
+      "step": 254255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015898743757392162,
+      "loss": 2.0843,
+      "step": 254260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015898594507767272,
+      "loss": 2.0423,
+      "step": 254265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001589844525612732,
+      "loss": 1.9589,
+      "step": 254270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015898296002472349,
+      "loss": 2.0346,
+      "step": 254275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001589814674680241,
+      "loss": 2.0729,
+      "step": 254280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015897997489117564,
+      "loss": 1.9831,
+      "step": 254285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015897848229417852,
+      "loss": 2.025,
+      "step": 254290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015897698967703327,
+      "loss": 2.0209,
+      "step": 254295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015897549703974045,
+      "loss": 2.1052,
+      "step": 254300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015897400438230047,
+      "loss": 2.1232,
+      "step": 254305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015897251170471396,
+      "loss": 2.0479,
+      "step": 254310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015897101900698134,
+      "loss": 1.9098,
+      "step": 254315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015896952628910316,
+      "loss": 2.1803,
+      "step": 254320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015896803355107988,
+      "loss": 2.1068,
+      "step": 254325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001589665407929121,
+      "loss": 2.0398,
+      "step": 254330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015896504801460025,
+      "loss": 1.9933,
+      "step": 254335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001589635552161449,
+      "loss": 1.9856,
+      "step": 254340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001589620623975465,
+      "loss": 2.1521,
+      "step": 254345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015896056955880562,
+      "loss": 2.228,
+      "step": 254350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015895907669992272,
+      "loss": 2.1257,
+      "step": 254355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015895758382089833,
+      "loss": 1.996,
+      "step": 254360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015895609092173295,
+      "loss": 2.2922,
+      "step": 254365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015895459800242714,
+      "loss": 2.1508,
+      "step": 254370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015895310506298133,
+      "loss": 2.0249,
+      "step": 254375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015895161210339608,
+      "loss": 2.1153,
+      "step": 254380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001589501191236719,
+      "loss": 2.017,
+      "step": 254385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001589486261238093,
+      "loss": 2.0203,
+      "step": 254390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015894713310380873,
+      "loss": 2.1348,
+      "step": 254395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001589456400636708,
+      "loss": 1.9783,
+      "step": 254400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015894414700339598,
+      "loss": 2.1854,
+      "step": 254405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015894265392298471,
+      "loss": 2.1988,
+      "step": 254410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001589411608224376,
+      "loss": 2.3014,
+      "step": 254415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015893966770175514,
+      "loss": 2.033,
+      "step": 254420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015893817456093777,
+      "loss": 2.1027,
+      "step": 254425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001589366813999861,
+      "loss": 2.1138,
+      "step": 254430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015893518821890056,
+      "loss": 2.2355,
+      "step": 254435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015893369501768172,
+      "loss": 1.9911,
+      "step": 254440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015893220179633003,
+      "loss": 1.9759,
+      "step": 254445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015893070855484603,
+      "loss": 2.2439,
+      "step": 254450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015892921529323025,
+      "loss": 1.9289,
+      "step": 254455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015892772201148318,
+      "loss": 2.1549,
+      "step": 254460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001589262287096053,
+      "loss": 2.2775,
+      "step": 254465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001589247353875972,
+      "loss": 2.1748,
+      "step": 254470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015892324204545933,
+      "loss": 2.1834,
+      "step": 254475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001589217486831922,
+      "loss": 2.094,
+      "step": 254480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015892025530079632,
+      "loss": 2.0036,
+      "step": 254485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015891876189827223,
+      "loss": 2.1334,
+      "step": 254490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001589172684756204,
+      "loss": 2.0622,
+      "step": 254495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015891577503284138,
+      "loss": 2.2691,
+      "step": 254500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015891428156993568,
+      "loss": 1.9581,
+      "step": 254505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015891278808690377,
+      "loss": 2.1923,
+      "step": 254510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015891129458374617,
+      "loss": 2.0947,
+      "step": 254515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015890980106046345,
+      "loss": 2.2488,
+      "step": 254520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015890830751705607,
+      "loss": 2.1117,
+      "step": 254525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015890681395352448,
+      "loss": 2.1605,
+      "step": 254530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015890532036986932,
+      "loss": 2.1719,
+      "step": 254535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015890382676609098,
+      "loss": 2.0895,
+      "step": 254540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015890233314219007,
+      "loss": 2.1037,
+      "step": 254545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015890083949816704,
+      "loss": 2.1992,
+      "step": 254550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015889934583402241,
+      "loss": 1.9198,
+      "step": 254555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001588978521497567,
+      "loss": 2.2719,
+      "step": 254560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001588963584453704,
+      "loss": 2.2121,
+      "step": 254565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015889486472086406,
+      "loss": 2.1303,
+      "step": 254570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015889337097623813,
+      "loss": 2.1546,
+      "step": 254575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001588918772114932,
+      "loss": 2.2089,
+      "step": 254580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001588903834266297,
+      "loss": 2.1937,
+      "step": 254585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.00015888888962164823,
+      "loss": 2.0862,
+      "step": 254590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015888739579654918,
+      "loss": 1.9055,
+      "step": 254595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015888590195133317,
+      "loss": 2.0589,
+      "step": 254600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015888440808600068,
+      "loss": 2.0647,
+      "step": 254605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015888291420055218,
+      "loss": 1.9721,
+      "step": 254610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001588814202949882,
+      "loss": 1.9945,
+      "step": 254615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001588799263693093,
+      "loss": 2.0488,
+      "step": 254620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015887843242351593,
+      "loss": 2.1296,
+      "step": 254625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001588769384576086,
+      "loss": 2.1659,
+      "step": 254630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015887544447158787,
+      "loss": 2.1839,
+      "step": 254635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015887395046545417,
+      "loss": 2.1228,
+      "step": 254640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001588724564392081,
+      "loss": 2.3018,
+      "step": 254645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015887096239285012,
+      "loss": 2.0152,
+      "step": 254650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015886946832638075,
+      "loss": 2.146,
+      "step": 254655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015886797423980052,
+      "loss": 1.9704,
+      "step": 254660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001588664801331099,
+      "loss": 1.9451,
+      "step": 254665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015886498600630947,
+      "loss": 2.0937,
+      "step": 254670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015886349185939965,
+      "loss": 2.1563,
+      "step": 254675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000158861997692381,
+      "loss": 2.1042,
+      "step": 254680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015886050350525402,
+      "loss": 1.9214,
+      "step": 254685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015885900929801921,
+      "loss": 2.1416,
+      "step": 254690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015885751507067712,
+      "loss": 1.9957,
+      "step": 254695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015885602082322823,
+      "loss": 2.2365,
+      "step": 254700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015885452655567305,
+      "loss": 1.9033,
+      "step": 254705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015885303226801208,
+      "loss": 1.9949,
+      "step": 254710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001588515379602459,
+      "loss": 2.132,
+      "step": 254715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015885004363237493,
+      "loss": 2.1109,
+      "step": 254720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001588485492843997,
+      "loss": 1.9387,
+      "step": 254725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001588470549163208,
+      "loss": 2.0435,
+      "step": 254730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015884556052813865,
+      "loss": 2.1598,
+      "step": 254735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015884406611985376,
+      "loss": 2.111,
+      "step": 254740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015884257169146668,
+      "loss": 2.048,
+      "step": 254745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015884107724297794,
+      "loss": 2.0423,
+      "step": 254750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000158839582774388,
+      "loss": 2.0257,
+      "step": 254755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015883808828569737,
+      "loss": 2.2127,
+      "step": 254760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001588365937769066,
+      "loss": 1.9915,
+      "step": 254765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001588350992480162,
+      "loss": 2.0294,
+      "step": 254770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015883360469902668,
+      "loss": 2.0446,
+      "step": 254775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015883211012993848,
+      "loss": 2.1225,
+      "step": 254780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015883061554075218,
+      "loss": 1.938,
+      "step": 254785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001588291209314683,
+      "loss": 2.1307,
+      "step": 254790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015882762630208732,
+      "loss": 2.0799,
+      "step": 254795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015882613165260973,
+      "loss": 1.971,
+      "step": 254800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015882463698303607,
+      "loss": 2.1593,
+      "step": 254805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015882314229336686,
+      "loss": 2.1831,
+      "step": 254810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015882164758360262,
+      "loss": 1.982,
+      "step": 254815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001588201528537438,
+      "loss": 1.9748,
+      "step": 254820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015881865810379097,
+      "loss": 2.0508,
+      "step": 254825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015881716333374462,
+      "loss": 2.1912,
+      "step": 254830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015881566854360522,
+      "loss": 2.0965,
+      "step": 254835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015881417373337338,
+      "loss": 1.8598,
+      "step": 254840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001588126789030495,
+      "loss": 2.0222,
+      "step": 254845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001588111840526342,
+      "loss": 2.0336,
+      "step": 254850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001588096891821279,
+      "loss": 1.9496,
+      "step": 254855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001588081942915311,
+      "loss": 2.1252,
+      "step": 254860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001588066993808444,
+      "loss": 1.9914,
+      "step": 254865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001588052044500683,
+      "loss": 1.9076,
+      "step": 254870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015880370949920323,
+      "loss": 2.0497,
+      "step": 254875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015880221452824978,
+      "loss": 2.1142,
+      "step": 254880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015880071953720836,
+      "loss": 2.3404,
+      "step": 254885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001587992245260796,
+      "loss": 2.0731,
+      "step": 254890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015879772949486395,
+      "loss": 2.2535,
+      "step": 254895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015879623444356192,
+      "loss": 2.0703,
+      "step": 254900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015879473937217405,
+      "loss": 2.0891,
+      "step": 254905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015879324428070085,
+      "loss": 2.0263,
+      "step": 254910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015879174916914277,
+      "loss": 2.1182,
+      "step": 254915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015879025403750036,
+      "loss": 1.8902,
+      "step": 254920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015878875888577414,
+      "loss": 2.061,
+      "step": 254925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015878726371396464,
+      "loss": 2.1624,
+      "step": 254930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001587857685220723,
+      "loss": 1.9508,
+      "step": 254935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001587842733100977,
+      "loss": 1.979,
+      "step": 254940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001587827780780413,
+      "loss": 2.0839,
+      "step": 254945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001587812828259037,
+      "loss": 2.2664,
+      "step": 254950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001587797875536853,
+      "loss": 2.0203,
+      "step": 254955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015877829226138669,
+      "loss": 2.0474,
+      "step": 254960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015877679694900834,
+      "loss": 2.0258,
+      "step": 254965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015877530161655072,
+      "loss": 2.156,
+      "step": 254970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015877380626401445,
+      "loss": 1.9905,
+      "step": 254975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015877231089139998,
+      "loss": 2.124,
+      "step": 254980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015877081549870776,
+      "loss": 2.1647,
+      "step": 254985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015876932008593842,
+      "loss": 2.0341,
+      "step": 254990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001587678246530924,
+      "loss": 2.0192,
+      "step": 254995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015876632920017023,
+      "loss": 2.0594,
+      "step": 255000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001587648337271724,
+      "loss": 2.0192,
+      "step": 255005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015876333823409947,
+      "loss": 2.032,
+      "step": 255010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015876184272095188,
+      "loss": 1.9432,
+      "step": 255015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001587603471877302,
+      "loss": 1.9081,
+      "step": 255020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015875885163443493,
+      "loss": 1.9347,
+      "step": 255025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8125,
+      "learning_rate": 0.00015875735606106657,
+      "loss": 1.9837,
+      "step": 255030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001587558604676256,
+      "loss": 2.1726,
+      "step": 255035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001587543648541126,
+      "loss": 2.1533,
+      "step": 255040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000158752869220528,
+      "loss": 1.9274,
+      "step": 255045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015875137356687237,
+      "loss": 2.0766,
+      "step": 255050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015874987789314625,
+      "loss": 2.0787,
+      "step": 255055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015874838219935005,
+      "loss": 2.0006,
+      "step": 255060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015874688648548436,
+      "loss": 2.1357,
+      "step": 255065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015874539075154967,
+      "loss": 2.1198,
+      "step": 255070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015874389499754646,
+      "loss": 1.966,
+      "step": 255075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015874239922347532,
+      "loss": 2.1759,
+      "step": 255080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001587409034293367,
+      "loss": 2.0736,
+      "step": 255085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015873940761513107,
+      "loss": 2.3104,
+      "step": 255090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015873791178085906,
+      "loss": 2.0345,
+      "step": 255095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001587364159265211,
+      "loss": 2.1513,
+      "step": 255100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015873492005211766,
+      "loss": 2.2374,
+      "step": 255105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001587334241576494,
+      "loss": 1.999,
+      "step": 255110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015873192824311666,
+      "loss": 2.1764,
+      "step": 255115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015873043230852006,
+      "loss": 2.2139,
+      "step": 255120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015872893635386008,
+      "loss": 2.0536,
+      "step": 255125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015872744037913718,
+      "loss": 2.0078,
+      "step": 255130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015872594438435197,
+      "loss": 2.1641,
+      "step": 255135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015872444836950492,
+      "loss": 2.0671,
+      "step": 255140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015872295233459652,
+      "loss": 1.891,
+      "step": 255145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015872145627962732,
+      "loss": 2.1572,
+      "step": 255150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015871996020459773,
+      "loss": 2.004,
+      "step": 255155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001587184641095084,
+      "loss": 1.9633,
+      "step": 255160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001587169679943598,
+      "loss": 2.1813,
+      "step": 255165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015871547185915237,
+      "loss": 2.0368,
+      "step": 255170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015871397570388666,
+      "loss": 1.9956,
+      "step": 255175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015871247952856321,
+      "loss": 1.8572,
+      "step": 255180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015871098333318252,
+      "loss": 2.2709,
+      "step": 255185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015870948711774508,
+      "loss": 2.0935,
+      "step": 255190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015870799088225142,
+      "loss": 1.954,
+      "step": 255195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015870649462670206,
+      "loss": 2.008,
+      "step": 255200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001587049983510975,
+      "loss": 2.0905,
+      "step": 255205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001587035020554382,
+      "loss": 1.9125,
+      "step": 255210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001587020057397248,
+      "loss": 2.2672,
+      "step": 255215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015870050940395764,
+      "loss": 2.053,
+      "step": 255220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015869901304813738,
+      "loss": 2.0982,
+      "step": 255225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015869751667226445,
+      "loss": 1.946,
+      "step": 255230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001586960202763394,
+      "loss": 2.0171,
+      "step": 255235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001586945238603627,
+      "loss": 2.0251,
+      "step": 255240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00015869302742433492,
+      "loss": 1.9498,
+      "step": 255245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001586915309682565,
+      "loss": 2.1705,
+      "step": 255250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000158690034492128,
+      "loss": 2.1988,
+      "step": 255255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001586885379959499,
+      "loss": 2.0463,
+      "step": 255260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001586870414797228,
+      "loss": 2.0786,
+      "step": 255265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001586855449434471,
+      "loss": 2.0109,
+      "step": 255270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015868404838712333,
+      "loss": 1.9729,
+      "step": 255275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015868255181075206,
+      "loss": 2.0442,
+      "step": 255280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015868105521433376,
+      "loss": 2.2218,
+      "step": 255285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015867955859786897,
+      "loss": 2.0335,
+      "step": 255290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015867806196135812,
+      "loss": 2.1174,
+      "step": 255295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001586765653048018,
+      "loss": 2.1782,
+      "step": 255300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015867506862820053,
+      "loss": 2.0657,
+      "step": 255305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015867357193155476,
+      "loss": 2.1518,
+      "step": 255310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015867207521486507,
+      "loss": 2.1528,
+      "step": 255315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001586705784781319,
+      "loss": 2.0446,
+      "step": 255320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015866908172135582,
+      "loss": 2.0899,
+      "step": 255325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015866758494453728,
+      "loss": 2.0936,
+      "step": 255330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015866608814767686,
+      "loss": 2.0876,
+      "step": 255335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015866459133077507,
+      "loss": 1.9865,
+      "step": 255340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015866309449383232,
+      "loss": 2.1239,
+      "step": 255345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015866159763684923,
+      "loss": 1.975,
+      "step": 255350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001586601007598263,
+      "loss": 1.9507,
+      "step": 255355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.000158658603862764,
+      "loss": 2.1759,
+      "step": 255360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015865710694566284,
+      "loss": 2.1681,
+      "step": 255365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015865561000852333,
+      "loss": 2.008,
+      "step": 255370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015865411305134605,
+      "loss": 1.6932,
+      "step": 255375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015865261607413142,
+      "loss": 2.1466,
+      "step": 255380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015865111907688003,
+      "loss": 2.0812,
+      "step": 255385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015864962205959233,
+      "loss": 2.1129,
+      "step": 255390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015864812502226885,
+      "loss": 1.8881,
+      "step": 255395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001586466279649101,
+      "loss": 2.1183,
+      "step": 255400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015864513088751663,
+      "loss": 2.0661,
+      "step": 255405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001586436337900889,
+      "loss": 2.1293,
+      "step": 255410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015864213667262747,
+      "loss": 2.0863,
+      "step": 255415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001586406395351328,
+      "loss": 2.273,
+      "step": 255420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001586391423776054,
+      "loss": 1.9126,
+      "step": 255425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015863764520004583,
+      "loss": 1.9706,
+      "step": 255430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015863614800245456,
+      "loss": 2.0902,
+      "step": 255435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015863465078483214,
+      "loss": 2.0641,
+      "step": 255440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015863315354717908,
+      "loss": 2.0486,
+      "step": 255445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015863165628949585,
+      "loss": 2.1399,
+      "step": 255450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015863015901178295,
+      "loss": 1.9993,
+      "step": 255455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015862866171404094,
+      "loss": 2.1671,
+      "step": 255460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015862716439627036,
+      "loss": 2.1858,
+      "step": 255465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015862566705847163,
+      "loss": 2.0274,
+      "step": 255470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015862416970064533,
+      "loss": 1.9735,
+      "step": 255475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015862267232279192,
+      "loss": 2.0803,
+      "step": 255480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015862117492491197,
+      "loss": 2.131,
+      "step": 255485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015861967750700596,
+      "loss": 2.1243,
+      "step": 255490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015861818006907442,
+      "loss": 2.089,
+      "step": 255495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015861668261111785,
+      "loss": 2.0015,
+      "step": 255500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015861518513313668,
+      "loss": 1.9826,
+      "step": 255505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015861368763513158,
+      "loss": 2.0588,
+      "step": 255510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015861219011710297,
+      "loss": 2.1443,
+      "step": 255515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015861069257905137,
+      "loss": 2.2595,
+      "step": 255520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001586091950209773,
+      "loss": 2.2112,
+      "step": 255525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015860769744288123,
+      "loss": 2.1076,
+      "step": 255530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015860619984476375,
+      "loss": 1.9314,
+      "step": 255535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001586047022266253,
+      "loss": 1.9143,
+      "step": 255540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015860320458846645,
+      "loss": 1.9712,
+      "step": 255545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015860170693028763,
+      "loss": 1.9029,
+      "step": 255550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015860020925208946,
+      "loss": 2.2476,
+      "step": 255555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015859871155387236,
+      "loss": 1.9362,
+      "step": 255560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001585972138356369,
+      "loss": 2.0565,
+      "step": 255565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015859571609738358,
+      "loss": 1.9948,
+      "step": 255570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015859421833911284,
+      "loss": 2.1619,
+      "step": 255575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015859272056082532,
+      "loss": 1.9717,
+      "step": 255580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015859122276252143,
+      "loss": 2.0148,
+      "step": 255585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015858972494420172,
+      "loss": 2.0931,
+      "step": 255590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001585882271058667,
+      "loss": 2.0109,
+      "step": 255595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001585867292475169,
+      "loss": 1.9726,
+      "step": 255600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015858523136915277,
+      "loss": 2.1728,
+      "step": 255605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015858373347077488,
+      "loss": 1.8003,
+      "step": 255610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015858223555238374,
+      "loss": 2.1411,
+      "step": 255615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001585807376139798,
+      "loss": 1.9698,
+      "step": 255620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001585792396555637,
+      "loss": 2.161,
+      "step": 255625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001585777416771358,
+      "loss": 2.0452,
+      "step": 255630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001585762436786967,
+      "loss": 2.0826,
+      "step": 255635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015857474566024688,
+      "loss": 1.9902,
+      "step": 255640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001585732476217869,
+      "loss": 2.0319,
+      "step": 255645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015857174956331721,
+      "loss": 2.0555,
+      "step": 255650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015857025148483835,
+      "loss": 2.151,
+      "step": 255655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015856875338635083,
+      "loss": 2.1477,
+      "step": 255660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015856725526785517,
+      "loss": 2.1518,
+      "step": 255665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015856575712935186,
+      "loss": 2.1455,
+      "step": 255670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015856425897084142,
+      "loss": 1.9489,
+      "step": 255675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001585627607923244,
+      "loss": 2.1832,
+      "step": 255680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015856126259380124,
+      "loss": 2.1389,
+      "step": 255685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001585597643752725,
+      "loss": 2.067,
+      "step": 255690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015855826613673873,
+      "loss": 2.0972,
+      "step": 255695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015855676787820033,
+      "loss": 1.9175,
+      "step": 255700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001585552695996579,
+      "loss": 2.0243,
+      "step": 255705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015855377130111195,
+      "loss": 2.1555,
+      "step": 255710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015855227298256294,
+      "loss": 2.2768,
+      "step": 255715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001585507746440114,
+      "loss": 2.0534,
+      "step": 255720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001585492762854579,
+      "loss": 1.9731,
+      "step": 255725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015854777790690286,
+      "loss": 2.2586,
+      "step": 255730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015854627950834686,
+      "loss": 2.0046,
+      "step": 255735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015854478108979041,
+      "loss": 2.1181,
+      "step": 255740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015854328265123394,
+      "loss": 2.0687,
+      "step": 255745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015854178419267808,
+      "loss": 2.0585,
+      "step": 255750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015854028571412327,
+      "loss": 2.2108,
+      "step": 255755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015853878721557005,
+      "loss": 2.1213,
+      "step": 255760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001585372886970189,
+      "loss": 2.2286,
+      "step": 255765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015853579015847032,
+      "loss": 2.0618,
+      "step": 255770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015853429159992487,
+      "loss": 1.9109,
+      "step": 255775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001585327930213831,
+      "loss": 2.1142,
+      "step": 255780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015853129442284543,
+      "loss": 2.1082,
+      "step": 255785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015852979580431237,
+      "loss": 2.0769,
+      "step": 255790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00015852829716578453,
+      "loss": 2.0132,
+      "step": 255795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001585267985072623,
+      "loss": 2.0493,
+      "step": 255800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015852529982874632,
+      "loss": 2.2128,
+      "step": 255805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000158523801130237,
+      "loss": 2.056,
+      "step": 255810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015852230241173492,
+      "loss": 1.9536,
+      "step": 255815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015852080367324053,
+      "loss": 2.0501,
+      "step": 255820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015851930491475434,
+      "loss": 1.9019,
+      "step": 255825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015851780613627693,
+      "loss": 2.1688,
+      "step": 255830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015851630733780878,
+      "loss": 2.1115,
+      "step": 255835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001585148085193504,
+      "loss": 2.1844,
+      "step": 255840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001585133096809023,
+      "loss": 2.2261,
+      "step": 255845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015851181082246498,
+      "loss": 2.0797,
+      "step": 255850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015851031194403898,
+      "loss": 2.2052,
+      "step": 255855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001585088130456248,
+      "loss": 2.2179,
+      "step": 255860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001585073141272229,
+      "loss": 2.1754,
+      "step": 255865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001585058151888339,
+      "loss": 2.1274,
+      "step": 255870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015850431623045823,
+      "loss": 1.9827,
+      "step": 255875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001585028172520964,
+      "loss": 2.1743,
+      "step": 255880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015850131825374898,
+      "loss": 2.1712,
+      "step": 255885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015849981923541644,
+      "loss": 1.946,
+      "step": 255890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001584983201970993,
+      "loss": 2.1084,
+      "step": 255895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001584968211387981,
+      "loss": 1.8885,
+      "step": 255900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015849532206051327,
+      "loss": 2.1904,
+      "step": 255905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001584938229622454,
+      "loss": 1.8624,
+      "step": 255910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015849232384399498,
+      "loss": 2.1419,
+      "step": 255915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015849082470576257,
+      "loss": 2.1196,
+      "step": 255920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015848932554754857,
+      "loss": 2.164,
+      "step": 255925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015848782636935355,
+      "loss": 2.147,
+      "step": 255930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015848632717117806,
+      "loss": 2.0994,
+      "step": 255935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015848482795302257,
+      "loss": 2.1292,
+      "step": 255940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001584833287148876,
+      "loss": 2.1095,
+      "step": 255945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015848182945677367,
+      "loss": 2.0079,
+      "step": 255950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015848033017868129,
+      "loss": 2.033,
+      "step": 255955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000158478830880611,
+      "loss": 2.2177,
+      "step": 255960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001584773315625632,
+      "loss": 2.1911,
+      "step": 255965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015847583222453854,
+      "loss": 2.1774,
+      "step": 255970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015847433286653745,
+      "loss": 2.035,
+      "step": 255975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015847283348856047,
+      "loss": 2.0753,
+      "step": 255980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015847133409060813,
+      "loss": 2.0372,
+      "step": 255985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001584698346726809,
+      "loss": 2.0978,
+      "step": 255990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015846833523477932,
+      "loss": 2.1094,
+      "step": 255995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001584668357769039,
+      "loss": 2.2921,
+      "step": 256000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015846533629905517,
+      "loss": 2.1474,
+      "step": 256005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015846383680123357,
+      "loss": 2.1624,
+      "step": 256010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015846233728343973,
+      "loss": 2.1935,
+      "step": 256015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015846083774567404,
+      "loss": 2.3534,
+      "step": 256020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001584593381879371,
+      "loss": 2.1537,
+      "step": 256025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015845783861022939,
+      "loss": 2.0127,
+      "step": 256030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001584563390125514,
+      "loss": 2.0432,
+      "step": 256035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015845483939490367,
+      "loss": 1.9862,
+      "step": 256040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001584533397572867,
+      "loss": 2.0178,
+      "step": 256045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015845184009970103,
+      "loss": 2.1259,
+      "step": 256050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015845034042214715,
+      "loss": 2.2191,
+      "step": 256055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015844884072462555,
+      "loss": 2.088,
+      "step": 256060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015844734100713676,
+      "loss": 1.9555,
+      "step": 256065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015844584126968134,
+      "loss": 2.137,
+      "step": 256070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015844434151225976,
+      "loss": 2.1613,
+      "step": 256075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001584428417348725,
+      "loss": 2.0678,
+      "step": 256080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001584413419375201,
+      "loss": 2.3281,
+      "step": 256085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001584398421202031,
+      "loss": 2.2709,
+      "step": 256090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015843834228292198,
+      "loss": 2.1292,
+      "step": 256095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015843684242567728,
+      "loss": 2.1677,
+      "step": 256100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015843534254846947,
+      "loss": 2.2742,
+      "step": 256105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001584338426512991,
+      "loss": 2.0415,
+      "step": 256110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001584323427341667,
+      "loss": 2.0084,
+      "step": 256115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001584308427970727,
+      "loss": 2.0029,
+      "step": 256120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015842934284001768,
+      "loss": 2.0384,
+      "step": 256125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015842784286300215,
+      "loss": 2.08,
+      "step": 256130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001584263428660266,
+      "loss": 2.1925,
+      "step": 256135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015842484284909155,
+      "loss": 2.2111,
+      "step": 256140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001584233428121975,
+      "loss": 1.9986,
+      "step": 256145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.000158421842755345,
+      "loss": 2.1368,
+      "step": 256150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015842034267853455,
+      "loss": 2.1652,
+      "step": 256155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015841884258176663,
+      "loss": 2.252,
+      "step": 256160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015841734246504178,
+      "loss": 2.1473,
+      "step": 256165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001584158423283605,
+      "loss": 2.1794,
+      "step": 256170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001584143421717233,
+      "loss": 1.8776,
+      "step": 256175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015841284199513074,
+      "loss": 2.0502,
+      "step": 256180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015841134179858326,
+      "loss": 2.2136,
+      "step": 256185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015840984158208143,
+      "loss": 2.0451,
+      "step": 256190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001584083413456257,
+      "loss": 1.9379,
+      "step": 256195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015840684108921663,
+      "loss": 2.1534,
+      "step": 256200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015840534081285474,
+      "loss": 2.0607,
+      "step": 256205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015840384051654053,
+      "loss": 2.0017,
+      "step": 256210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001584023402002745,
+      "loss": 2.2571,
+      "step": 256215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001584008398640572,
+      "loss": 2.0414,
+      "step": 256220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015839933950788907,
+      "loss": 2.0799,
+      "step": 256225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001583978391317707,
+      "loss": 1.9743,
+      "step": 256230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015839633873570254,
+      "loss": 2.0604,
+      "step": 256235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015839483831968517,
+      "loss": 1.9878,
+      "step": 256240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015839333788371902,
+      "loss": 2.2106,
+      "step": 256245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015839183742780467,
+      "loss": 1.8846,
+      "step": 256250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015839033695194262,
+      "loss": 2.0291,
+      "step": 256255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015838883645613336,
+      "loss": 2.1031,
+      "step": 256260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001583873359403774,
+      "loss": 1.9465,
+      "step": 256265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001583858354046753,
+      "loss": 2.158,
+      "step": 256270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015838433484902751,
+      "loss": 2.174,
+      "step": 256275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015838283427343456,
+      "loss": 2.0442,
+      "step": 256280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.000158381333677897,
+      "loss": 2.1007,
+      "step": 256285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015837983306241532,
+      "loss": 1.8808,
+      "step": 256290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015837833242699002,
+      "loss": 2.058,
+      "step": 256295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015837683177162163,
+      "loss": 2.0821,
+      "step": 256300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015837533109631067,
+      "loss": 2.3237,
+      "step": 256305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015837383040105763,
+      "loss": 1.9916,
+      "step": 256310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015837232968586302,
+      "loss": 2.1862,
+      "step": 256315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015837082895072736,
+      "loss": 2.1043,
+      "step": 256320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015836932819565119,
+      "loss": 2.0691,
+      "step": 256325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015836782742063496,
+      "loss": 2.1477,
+      "step": 256330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015836632662567923,
+      "loss": 2.1759,
+      "step": 256335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001583648258107845,
+      "loss": 2.1182,
+      "step": 256340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015836332497595136,
+      "loss": 2.1111,
+      "step": 256345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015836182412118016,
+      "loss": 1.8705,
+      "step": 256350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015836032324647155,
+      "loss": 2.1159,
+      "step": 256355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015835882235182598,
+      "loss": 2.1783,
+      "step": 256360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015835732143724398,
+      "loss": 2.1127,
+      "step": 256365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015835582050272606,
+      "loss": 1.9129,
+      "step": 256370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015835431954827273,
+      "loss": 2.2006,
+      "step": 256375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015835281857388452,
+      "loss": 2.1503,
+      "step": 256380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001583513175795619,
+      "loss": 2.0974,
+      "step": 256385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001583498165653054,
+      "loss": 2.0929,
+      "step": 256390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001583483155311156,
+      "loss": 2.0819,
+      "step": 256395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015834681447699292,
+      "loss": 1.954,
+      "step": 256400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015834531340293794,
+      "loss": 1.939,
+      "step": 256405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001583438123089511,
+      "loss": 2.2945,
+      "step": 256410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.000158342311195033,
+      "loss": 2.0157,
+      "step": 256415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001583408100611841,
+      "loss": 2.1832,
+      "step": 256420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001583393089074049,
+      "loss": 2.1333,
+      "step": 256425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015833780773369594,
+      "loss": 2.0353,
+      "step": 256430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015833630654005774,
+      "loss": 2.0555,
+      "step": 256435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001583348053264908,
+      "loss": 2.0885,
+      "step": 256440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001583333040929956,
+      "loss": 2.0324,
+      "step": 256445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001583318028395727,
+      "loss": 2.0691,
+      "step": 256450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001583303015662226,
+      "loss": 2.0615,
+      "step": 256455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015832880027294584,
+      "loss": 2.1695,
+      "step": 256460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001583272989597429,
+      "loss": 2.1655,
+      "step": 256465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001583257976266143,
+      "loss": 2.0988,
+      "step": 256470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015832429627356048,
+      "loss": 2.1314,
+      "step": 256475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001583227949005821,
+      "loss": 2.2716,
+      "step": 256480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015832129350767956,
+      "loss": 2.0632,
+      "step": 256485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001583197920948534,
+      "loss": 2.1263,
+      "step": 256490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015831829066210418,
+      "loss": 1.9778,
+      "step": 256495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015831678920943232,
+      "loss": 2.1158,
+      "step": 256500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015831528773683844,
+      "loss": 2.012,
+      "step": 256505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015831378624432297,
+      "loss": 2.0277,
+      "step": 256510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015831228473188645,
+      "loss": 2.0779,
+      "step": 256515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015831078319952942,
+      "loss": 2.0223,
+      "step": 256520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015830928164725235,
+      "loss": 2.1779,
+      "step": 256525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015830778007505577,
+      "loss": 2.1374,
+      "step": 256530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001583062784829402,
+      "loss": 1.8884,
+      "step": 256535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015830477687090615,
+      "loss": 2.0456,
+      "step": 256540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015830327523895415,
+      "loss": 2.2666,
+      "step": 256545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015830177358708468,
+      "loss": 2.0693,
+      "step": 256550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015830027191529825,
+      "loss": 2.1751,
+      "step": 256555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001582987702235954,
+      "loss": 2.1937,
+      "step": 256560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015829726851197663,
+      "loss": 1.9588,
+      "step": 256565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015829576678044247,
+      "loss": 2.2475,
+      "step": 256570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001582942650289934,
+      "loss": 2.2357,
+      "step": 256575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015829276325762996,
+      "loss": 2.048,
+      "step": 256580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015829126146635266,
+      "loss": 2.0459,
+      "step": 256585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000158289759655162,
+      "loss": 1.9784,
+      "step": 256590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001582882578240585,
+      "loss": 2.1848,
+      "step": 256595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015828675597304268,
+      "loss": 1.9392,
+      "step": 256600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015828525410211506,
+      "loss": 1.8932,
+      "step": 256605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015828375221127613,
+      "loss": 2.0721,
+      "step": 256610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015828225030052639,
+      "loss": 2.07,
+      "step": 256615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015828074836986638,
+      "loss": 2.1496,
+      "step": 256620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015827924641929664,
+      "loss": 2.2337,
+      "step": 256625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015827774444881763,
+      "loss": 2.0577,
+      "step": 256630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001582762424584299,
+      "loss": 2.0564,
+      "step": 256635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015827474044813396,
+      "loss": 2.1277,
+      "step": 256640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015827323841793026,
+      "loss": 2.1017,
+      "step": 256645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015827173636781943,
+      "loss": 1.9987,
+      "step": 256650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015827023429780188,
+      "loss": 2.002,
+      "step": 256655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015826873220787813,
+      "loss": 2.2385,
+      "step": 256660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001582672300980488,
+      "loss": 2.0005,
+      "step": 256665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015826572796831428,
+      "loss": 2.1346,
+      "step": 256670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015826422581867515,
+      "loss": 2.1457,
+      "step": 256675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015826272364913185,
+      "loss": 2.033,
+      "step": 256680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015826122145968502,
+      "loss": 2.3289,
+      "step": 256685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.00015825971925033508,
+      "loss": 1.9328,
+      "step": 256690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015825821702108254,
+      "loss": 1.8879,
+      "step": 256695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015825671477192796,
+      "loss": 2.2427,
+      "step": 256700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015825521250287182,
+      "loss": 2.156,
+      "step": 256705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015825371021391463,
+      "loss": 2.1083,
+      "step": 256710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015825220790505691,
+      "loss": 2.2251,
+      "step": 256715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001582507055762992,
+      "loss": 2.1039,
+      "step": 256720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000158249203227642,
+      "loss": 1.8801,
+      "step": 256725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015824770085908583,
+      "loss": 2.3187,
+      "step": 256730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015824619847063117,
+      "loss": 2.0328,
+      "step": 256735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015824469606227855,
+      "loss": 2.1334,
+      "step": 256740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015824319363402846,
+      "loss": 2.1018,
+      "step": 256745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015824169118588147,
+      "loss": 2.1653,
+      "step": 256750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015824018871783804,
+      "loss": 2.0577,
+      "step": 256755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015823868622989872,
+      "loss": 2.0249,
+      "step": 256760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015823718372206402,
+      "loss": 2.0198,
+      "step": 256765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015823568119433443,
+      "loss": 1.9854,
+      "step": 256770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015823417864671048,
+      "loss": 2.0159,
+      "step": 256775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015823267607919267,
+      "loss": 1.9968,
+      "step": 256780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015823117349178154,
+      "loss": 2.215,
+      "step": 256785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015822967088447755,
+      "loss": 1.9204,
+      "step": 256790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015822816825728127,
+      "loss": 1.856,
+      "step": 256795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015822666561019318,
+      "loss": 2.1096,
+      "step": 256800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015822516294321384,
+      "loss": 2.1158,
+      "step": 256805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001582236602563437,
+      "loss": 2.0091,
+      "step": 256810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001582221575495833,
+      "loss": 2.1051,
+      "step": 256815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015822065482293317,
+      "loss": 2.0815,
+      "step": 256820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015821915207639377,
+      "loss": 2.0524,
+      "step": 256825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001582176493099657,
+      "loss": 2.1566,
+      "step": 256830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001582161465236494,
+      "loss": 2.0663,
+      "step": 256835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015821464371744542,
+      "loss": 2.1038,
+      "step": 256840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015821314089135426,
+      "loss": 2.0975,
+      "step": 256845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015821163804537641,
+      "loss": 2.0649,
+      "step": 256850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015821013517951243,
+      "loss": 2.0042,
+      "step": 256855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015820863229376283,
+      "loss": 2.0324,
+      "step": 256860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015820712938812809,
+      "loss": 2.1143,
+      "step": 256865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015820562646260872,
+      "loss": 2.0539,
+      "step": 256870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015820412351720529,
+      "loss": 2.0722,
+      "step": 256875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015820262055191826,
+      "loss": 2.1271,
+      "step": 256880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015820111756674815,
+      "loss": 1.99,
+      "step": 256885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001581996145616955,
+      "loss": 2.1525,
+      "step": 256890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015819811153676075,
+      "loss": 2.008,
+      "step": 256895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015819660849194454,
+      "loss": 2.0106,
+      "step": 256900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015819510542724728,
+      "loss": 1.8544,
+      "step": 256905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001581936023426695,
+      "loss": 2.1565,
+      "step": 256910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015819209923821176,
+      "loss": 2.0482,
+      "step": 256915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015819059611387455,
+      "loss": 2.2377,
+      "step": 256920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015818909296965835,
+      "loss": 2.1978,
+      "step": 256925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015818758980556372,
+      "loss": 2.0835,
+      "step": 256930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00015818608662159116,
+      "loss": 2.003,
+      "step": 256935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015818458341774116,
+      "loss": 2.1139,
+      "step": 256940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015818308019401424,
+      "loss": 2.0343,
+      "step": 256945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015818157695041097,
+      "loss": 2.0337,
+      "step": 256950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015818007368693178,
+      "loss": 2.0671,
+      "step": 256955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015817857040357724,
+      "loss": 2.2022,
+      "step": 256960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015817706710034783,
+      "loss": 2.1158,
+      "step": 256965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015817556377724408,
+      "loss": 2.0099,
+      "step": 256970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015817406043426647,
+      "loss": 2.1291,
+      "step": 256975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001581725570714156,
+      "loss": 2.041,
+      "step": 256980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015817105368869192,
+      "loss": 2.1317,
+      "step": 256985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015816955028609596,
+      "loss": 2.065,
+      "step": 256990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015816804686362822,
+      "loss": 2.1188,
+      "step": 256995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001581665434212892,
+      "loss": 2.1298,
+      "step": 257000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015816503995907946,
+      "loss": 2.0266,
+      "step": 257005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015816353647699947,
+      "loss": 1.9546,
+      "step": 257010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015816203297504974,
+      "loss": 2.148,
+      "step": 257015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015816052945323087,
+      "loss": 2.2743,
+      "step": 257020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015815902591154326,
+      "loss": 1.9878,
+      "step": 257025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015815752234998745,
+      "loss": 2.1296,
+      "step": 257030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000158156018768564,
+      "loss": 2.0945,
+      "step": 257035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015815451516727343,
+      "loss": 2.1623,
+      "step": 257040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015815301154611618,
+      "loss": 1.8366,
+      "step": 257045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015815150790509283,
+      "loss": 2.0995,
+      "step": 257050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015815000424420386,
+      "loss": 2.0358,
+      "step": 257055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001581485005634498,
+      "loss": 2.1544,
+      "step": 257060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015814699686283117,
+      "loss": 2.0888,
+      "step": 257065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015814549314234846,
+      "loss": 2.033,
+      "step": 257070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001581439894020022,
+      "loss": 2.2934,
+      "step": 257075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015814248564179287,
+      "loss": 1.8342,
+      "step": 257080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015814098186172105,
+      "loss": 2.1329,
+      "step": 257085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015813947806178716,
+      "loss": 2.3082,
+      "step": 257090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015813797424199183,
+      "loss": 2.2243,
+      "step": 257095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001581364704023355,
+      "loss": 1.9885,
+      "step": 257100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001581349665428187,
+      "loss": 2.13,
+      "step": 257105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001581334626634419,
+      "loss": 2.0839,
+      "step": 257110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001581319587642057,
+      "loss": 2.2453,
+      "step": 257115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015813045484511058,
+      "loss": 2.2295,
+      "step": 257120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000158128950906157,
+      "loss": 2.0697,
+      "step": 257125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015812744694734555,
+      "loss": 2.0084,
+      "step": 257130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015812594296867668,
+      "loss": 2.2438,
+      "step": 257135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015812443897015094,
+      "loss": 1.9745,
+      "step": 257140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015812293495176885,
+      "loss": 1.9276,
+      "step": 257145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001581214309135309,
+      "loss": 2.1594,
+      "step": 257150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015811992685543764,
+      "loss": 2.1101,
+      "step": 257155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015811842277748954,
+      "loss": 2.0898,
+      "step": 257160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015811691867968712,
+      "loss": 2.1141,
+      "step": 257165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015811541456203093,
+      "loss": 1.9908,
+      "step": 257170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015811391042452145,
+      "loss": 2.2908,
+      "step": 257175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001581124062671592,
+      "loss": 2.0022,
+      "step": 257180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001581109020899447,
+      "loss": 2.2076,
+      "step": 257185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015810939789287845,
+      "loss": 2.0854,
+      "step": 257190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.000158107893675961,
+      "loss": 2.1253,
+      "step": 257195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015810638943919284,
+      "loss": 2.1631,
+      "step": 257200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015810488518257446,
+      "loss": 2.1047,
+      "step": 257205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00015810338090610642,
+      "loss": 2.031,
+      "step": 257210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015810187660978922,
+      "loss": 2.0501,
+      "step": 257215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001581003722936233,
+      "loss": 2.1208,
+      "step": 257220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001580988679576093,
+      "loss": 2.1616,
+      "step": 257225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015809736360174767,
+      "loss": 1.9059,
+      "step": 257230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015809585922603892,
+      "loss": 2.0028,
+      "step": 257235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015809435483048358,
+      "loss": 2.1541,
+      "step": 257240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015809285041508213,
+      "loss": 1.9809,
+      "step": 257245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015809134597983513,
+      "loss": 2.2273,
+      "step": 257250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015808984152474307,
+      "loss": 2.2535,
+      "step": 257255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015808833704980644,
+      "loss": 2.1615,
+      "step": 257260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001580868325550258,
+      "loss": 2.1534,
+      "step": 257265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015808532804040165,
+      "loss": 1.8284,
+      "step": 257270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001580838235059345,
+      "loss": 2.1518,
+      "step": 257275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015808231895162483,
+      "loss": 2.1378,
+      "step": 257280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015808081437747323,
+      "loss": 2.1341,
+      "step": 257285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015807930978348016,
+      "loss": 2.0997,
+      "step": 257290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015807780516964614,
+      "loss": 1.9478,
+      "step": 257295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015807630053597167,
+      "loss": 2.1685,
+      "step": 257300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001580747958824573,
+      "loss": 2.0089,
+      "step": 257305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015807329120910352,
+      "loss": 2.1607,
+      "step": 257310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015807178651591088,
+      "loss": 1.9142,
+      "step": 257315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015807028180287981,
+      "loss": 2.0893,
+      "step": 257320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001580687770700109,
+      "loss": 1.8619,
+      "step": 257325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015806727231730468,
+      "loss": 2.0749,
+      "step": 257330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001580657675447616,
+      "loss": 2.0618,
+      "step": 257335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015806426275238219,
+      "loss": 2.0522,
+      "step": 257340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015806275794016697,
+      "loss": 2.0951,
+      "step": 257345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015806125310811647,
+      "loss": 2.1291,
+      "step": 257350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001580597482562312,
+      "loss": 1.976,
+      "step": 257355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015805824338451167,
+      "loss": 2.1875,
+      "step": 257360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015805673849295836,
+      "loss": 2.0852,
+      "step": 257365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015805523358157188,
+      "loss": 2.0043,
+      "step": 257370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015805372865035263,
+      "loss": 2.3188,
+      "step": 257375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015805222369930116,
+      "loss": 2.0225,
+      "step": 257380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000158050718728418,
+      "loss": 2.0213,
+      "step": 257385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001580492137377037,
+      "loss": 1.9776,
+      "step": 257390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00015804770872715872,
+      "loss": 2.1666,
+      "step": 257395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001580462036967836,
+      "loss": 2.0996,
+      "step": 257400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001580446986465788,
+      "loss": 2.0187,
+      "step": 257405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015804319357654492,
+      "loss": 2.2312,
+      "step": 257410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015804168848668245,
+      "loss": 2.2553,
+      "step": 257415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015804018337699184,
+      "loss": 2.0428,
+      "step": 257420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015803867824747367,
+      "loss": 2.3554,
+      "step": 257425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015803717309812842,
+      "loss": 1.9247,
+      "step": 257430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015803566792895666,
+      "loss": 2.1258,
+      "step": 257435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001580341627399588,
+      "loss": 2.0129,
+      "step": 257440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015803265753113548,
+      "loss": 2.2053,
+      "step": 257445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0,
+      "learning_rate": 0.0001580311523024871,
+      "loss": 2.0658,
+      "step": 257450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015802964705401425,
+      "loss": 1.9619,
+      "step": 257455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015802814178571744,
+      "loss": 1.9815,
+      "step": 257460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015802663649759714,
+      "loss": 2.1131,
+      "step": 257465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001580251311896539,
+      "loss": 2.031,
+      "step": 257470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015802362586188822,
+      "loss": 2.167,
+      "step": 257475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015802212051430062,
+      "loss": 1.9607,
+      "step": 257480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015802061514689158,
+      "loss": 2.2268,
+      "step": 257485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015801910975966167,
+      "loss": 2.0438,
+      "step": 257490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001580176043526114,
+      "loss": 2.1178,
+      "step": 257495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015801609892574124,
+      "loss": 2.1259,
+      "step": 257500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015801459347905173,
+      "loss": 2.0354,
+      "step": 257505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015801308801254339,
+      "loss": 2.0213,
+      "step": 257510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015801158252621672,
+      "loss": 2.1773,
+      "step": 257515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015801007702007225,
+      "loss": 2.1763,
+      "step": 257520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00015800857149411045,
+      "loss": 2.01,
+      "step": 257525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001580070659483319,
+      "loss": 2.0351,
+      "step": 257530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001580055603827371,
+      "loss": 2.107,
+      "step": 257535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00015800405479732653,
+      "loss": 2.0008,
+      "step": 257540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015800254919210073,
+      "loss": 2.0772,
+      "step": 257545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001580010435670602,
+      "loss": 2.0611,
+      "step": 257550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015799953792220548,
+      "loss": 2.0847,
+      "step": 257555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015799803225753704,
+      "loss": 2.1061,
+      "step": 257560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001579965265730554,
+      "loss": 2.1582,
+      "step": 257565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015799502086876114,
+      "loss": 2.0505,
+      "step": 257570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015799351514465473,
+      "loss": 2.0563,
+      "step": 257575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015799200940073667,
+      "loss": 2.1176,
+      "step": 257580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001579905036370075,
+      "loss": 2.104,
+      "step": 257585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001579889978534677,
+      "loss": 1.9757,
+      "step": 257590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015798749205011783,
+      "loss": 2.1226,
+      "step": 257595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015798598622695836,
+      "loss": 2.0606,
+      "step": 257600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015798448038398984,
+      "loss": 1.9299,
+      "step": 257605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015798297452121273,
+      "loss": 2.0033,
+      "step": 257610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015798146863862762,
+      "loss": 2.0,
+      "step": 257615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015797996273623498,
+      "loss": 2.02,
+      "step": 257620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015797845681403533,
+      "loss": 1.9788,
+      "step": 257625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015797695087202922,
+      "loss": 2.0655,
+      "step": 257630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001579754449102171,
+      "loss": 1.9769,
+      "step": 257635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001579739389285995,
+      "loss": 1.918,
+      "step": 257640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015797243292717698,
+      "loss": 2.1126,
+      "step": 257645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015797092690595003,
+      "loss": 2.1126,
+      "step": 257650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015796942086491916,
+      "loss": 1.9934,
+      "step": 257655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015796791480408485,
+      "loss": 2.0759,
+      "step": 257660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001579664087234477,
+      "loss": 2.2315,
+      "step": 257665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015796490262300815,
+      "loss": 2.1766,
+      "step": 257670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015796339650276671,
+      "loss": 2.1588,
+      "step": 257675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015796189036272398,
+      "loss": 2.1396,
+      "step": 257680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015796038420288038,
+      "loss": 2.1282,
+      "step": 257685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015795887802323648,
+      "loss": 2.0753,
+      "step": 257690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015795737182379272,
+      "loss": 2.187,
+      "step": 257695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015795586560454972,
+      "loss": 1.9925,
+      "step": 257700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015795435936550795,
+      "loss": 1.967,
+      "step": 257705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001579528531066679,
+      "loss": 2.0444,
+      "step": 257710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015795134682803011,
+      "loss": 2.1113,
+      "step": 257715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015794984052959512,
+      "loss": 1.8714,
+      "step": 257720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015794833421136338,
+      "loss": 2.11,
+      "step": 257725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015794682787333543,
+      "loss": 2.025,
+      "step": 257730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015794532151551183,
+      "loss": 1.9164,
+      "step": 257735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000157943815137893,
+      "loss": 2.1165,
+      "step": 257740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015794230874047958,
+      "loss": 1.9858,
+      "step": 257745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015794080232327197,
+      "loss": 2.1046,
+      "step": 257750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015793929588627073,
+      "loss": 2.1622,
+      "step": 257755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015793778942947638,
+      "loss": 2.1068,
+      "step": 257760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015793628295288946,
+      "loss": 2.1077,
+      "step": 257765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015793477645651042,
+      "loss": 2.2246,
+      "step": 257770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015793326994033982,
+      "loss": 2.0997,
+      "step": 257775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015793176340437817,
+      "loss": 2.1077,
+      "step": 257780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015793025684862597,
+      "loss": 2.0471,
+      "step": 257785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015792875027308376,
+      "loss": 2.0418,
+      "step": 257790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015792724367775206,
+      "loss": 2.1743,
+      "step": 257795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001579257370626313,
+      "loss": 2.1045,
+      "step": 257800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015792423042772208,
+      "loss": 2.0647,
+      "step": 257805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015792272377302492,
+      "loss": 2.1634,
+      "step": 257810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015792121709854028,
+      "loss": 2.1343,
+      "step": 257815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001579197104042687,
+      "loss": 1.8192,
+      "step": 257820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001579182036902107,
+      "loss": 1.9161,
+      "step": 257825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015791669695636683,
+      "loss": 2.1447,
+      "step": 257830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001579151902027375,
+      "loss": 1.9782,
+      "step": 257835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015791368342932334,
+      "loss": 2.0159,
+      "step": 257840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001579121766361248,
+      "loss": 1.9223,
+      "step": 257845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001579106698231424,
+      "loss": 2.2456,
+      "step": 257850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001579091629903767,
+      "loss": 2.2108,
+      "step": 257855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015790765613782814,
+      "loss": 2.0329,
+      "step": 257860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001579061492654973,
+      "loss": 2.1342,
+      "step": 257865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015790464237338468,
+      "loss": 2.0906,
+      "step": 257870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015790313546149075,
+      "loss": 2.1254,
+      "step": 257875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015790162852981606,
+      "loss": 2.0052,
+      "step": 257880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015790012157836114,
+      "loss": 1.9921,
+      "step": 257885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015789861460712646,
+      "loss": 2.1991,
+      "step": 257890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001578971076161126,
+      "loss": 2.1532,
+      "step": 257895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015789560060532003,
+      "loss": 2.1209,
+      "step": 257900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015789409357474927,
+      "loss": 2.0714,
+      "step": 257905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015789258652440085,
+      "loss": 1.9869,
+      "step": 257910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015789107945427523,
+      "loss": 2.0465,
+      "step": 257915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75,
+      "learning_rate": 0.000157889572364373,
+      "loss": 2.0806,
+      "step": 257920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015788806525469465,
+      "loss": 2.0726,
+      "step": 257925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015788655812524067,
+      "loss": 2.0744,
+      "step": 257930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001578850509760116,
+      "loss": 1.9704,
+      "step": 257935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001578835438070079,
+      "loss": 2.0193,
+      "step": 257940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015788203661823023,
+      "loss": 2.0561,
+      "step": 257945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015788052940967894,
+      "loss": 2.0155,
+      "step": 257950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001578790221813546,
+      "loss": 2.1598,
+      "step": 257955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015787751493325777,
+      "loss": 2.0924,
+      "step": 257960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015787600766538896,
+      "loss": 2.1404,
+      "step": 257965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015787450037774858,
+      "loss": 2.1717,
+      "step": 257970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015787299307033726,
+      "loss": 1.9973,
+      "step": 257975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015787148574315546,
+      "loss": 1.9063,
+      "step": 257980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015786997839620375,
+      "loss": 2.154,
+      "step": 257985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015786847102948256,
+      "loss": 1.9311,
+      "step": 257990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015786696364299246,
+      "loss": 2.0598,
+      "step": 257995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015786545623673397,
+      "loss": 2.2695,
+      "step": 258000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001578639488107076,
+      "loss": 2.1492,
+      "step": 258005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015786244136491387,
+      "loss": 2.0999,
+      "step": 258010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015786093389935323,
+      "loss": 2.0971,
+      "step": 258015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015785942641402623,
+      "loss": 2.0055,
+      "step": 258020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001578579189089335,
+      "loss": 2.3035,
+      "step": 258025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00015785641138407538,
+      "loss": 2.0354,
+      "step": 258030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015785490383945247,
+      "loss": 2.2869,
+      "step": 258035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001578533962750653,
+      "loss": 2.2942,
+      "step": 258040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015785188869091435,
+      "loss": 2.0022,
+      "step": 258045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001578503810870001,
+      "loss": 2.1172,
+      "step": 258050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015784887346332316,
+      "loss": 2.0316,
+      "step": 258055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015784736581988398,
+      "loss": 2.3004,
+      "step": 258060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001578458581566831,
+      "loss": 2.1542,
+      "step": 258065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015784435047372104,
+      "loss": 1.9967,
+      "step": 258070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015784284277099825,
+      "loss": 2.0013,
+      "step": 258075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015784133504851534,
+      "loss": 2.0502,
+      "step": 258080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015783982730627275,
+      "loss": 2.008,
+      "step": 258085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015783831954427105,
+      "loss": 2.0562,
+      "step": 258090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015783681176251074,
+      "loss": 1.8641,
+      "step": 258095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015783530396099228,
+      "loss": 2.0587,
+      "step": 258100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001578337961397163,
+      "loss": 2.0849,
+      "step": 258105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015783228829868321,
+      "loss": 2.1134,
+      "step": 258110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015783078043789356,
+      "loss": 2.1332,
+      "step": 258115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015782927255734788,
+      "loss": 2.0904,
+      "step": 258120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015782776465704664,
+      "loss": 2.1847,
+      "step": 258125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001578262567369904,
+      "loss": 2.1111,
+      "step": 258130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015782474879717966,
+      "loss": 2.2565,
+      "step": 258135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015782324083761498,
+      "loss": 2.0564,
+      "step": 258140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015782173285829677,
+      "loss": 1.9674,
+      "step": 258145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015782022485922562,
+      "loss": 2.1116,
+      "step": 258150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015781871684040206,
+      "loss": 2.0373,
+      "step": 258155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015781720880182656,
+      "loss": 2.1718,
+      "step": 258160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015781570074349966,
+      "loss": 2.0354,
+      "step": 258165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015781419266542188,
+      "loss": 2.1068,
+      "step": 258170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001578126845675937,
+      "loss": 2.1033,
+      "step": 258175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015781117645001567,
+      "loss": 2.0303,
+      "step": 258180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001578096683126883,
+      "loss": 2.1012,
+      "step": 258185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015780816015561208,
+      "loss": 2.0199,
+      "step": 258190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015780665197878755,
+      "loss": 2.1202,
+      "step": 258195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015780514378221523,
+      "loss": 1.8594,
+      "step": 258200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001578036355658956,
+      "loss": 1.9663,
+      "step": 258205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015780212732982923,
+      "loss": 2.2284,
+      "step": 258210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001578006190740166,
+      "loss": 1.9998,
+      "step": 258215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015779911079845822,
+      "loss": 2.2498,
+      "step": 258220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015779760250315462,
+      "loss": 1.9099,
+      "step": 258225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001577960941881063,
+      "loss": 2.0471,
+      "step": 258230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001577945858533138,
+      "loss": 1.9787,
+      "step": 258235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015779307749877764,
+      "loss": 2.0555,
+      "step": 258240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001577915691244983,
+      "loss": 2.1375,
+      "step": 258245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015779006073047628,
+      "loss": 2.0823,
+      "step": 258250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015778855231671216,
+      "loss": 2.0556,
+      "step": 258255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001577870438832064,
+      "loss": 1.8974,
+      "step": 258260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015778553542995956,
+      "loss": 2.0616,
+      "step": 258265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.8125,
+      "learning_rate": 0.00015778402695697214,
+      "loss": 2.2491,
+      "step": 258270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001577825184642446,
+      "loss": 1.871,
+      "step": 258275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001577810099517776,
+      "loss": 2.0702,
+      "step": 258280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015777950141957145,
+      "loss": 2.1612,
+      "step": 258285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015777799286762682,
+      "loss": 2.1065,
+      "step": 258290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001577764842959442,
+      "loss": 1.9332,
+      "step": 258295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015777497570452408,
+      "loss": 2.1577,
+      "step": 258300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015777346709336696,
+      "loss": 2.1848,
+      "step": 258305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015777195846247336,
+      "loss": 1.9375,
+      "step": 258310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015777044981184383,
+      "loss": 2.0413,
+      "step": 258315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015776894114147888,
+      "loss": 2.2582,
+      "step": 258320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015776743245137902,
+      "loss": 2.0694,
+      "step": 258325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015776592374154474,
+      "loss": 2.0134,
+      "step": 258330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015776441501197657,
+      "loss": 2.0086,
+      "step": 258335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015776290626267502,
+      "loss": 2.2448,
+      "step": 258340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015776139749364062,
+      "loss": 2.086,
+      "step": 258345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015775988870487387,
+      "loss": 2.1022,
+      "step": 258350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015775837989637532,
+      "loss": 2.0582,
+      "step": 258355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015775687106814546,
+      "loss": 2.1687,
+      "step": 258360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001577553622201848,
+      "loss": 2.0661,
+      "step": 258365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015775385335249384,
+      "loss": 2.111,
+      "step": 258370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015775234446507313,
+      "loss": 2.0506,
+      "step": 258375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015775083555792318,
+      "loss": 2.0347,
+      "step": 258380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001577493266310445,
+      "loss": 1.9719,
+      "step": 258385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001577478176844376,
+      "loss": 2.1863,
+      "step": 258390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015774630871810295,
+      "loss": 1.9812,
+      "step": 258395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015774479973204117,
+      "loss": 2.161,
+      "step": 258400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015774329072625268,
+      "loss": 2.033,
+      "step": 258405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015774178170073808,
+      "loss": 2.2954,
+      "step": 258410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015774027265549783,
+      "loss": 2.2651,
+      "step": 258415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015773876359053242,
+      "loss": 1.8821,
+      "step": 258420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015773725450584242,
+      "loss": 1.8881,
+      "step": 258425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015773574540142835,
+      "loss": 2.0103,
+      "step": 258430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015773423627729068,
+      "loss": 2.0571,
+      "step": 258435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015773272713342994,
+      "loss": 2.3214,
+      "step": 258440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001577312179698467,
+      "loss": 2.1433,
+      "step": 258445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015772970878654137,
+      "loss": 2.0115,
+      "step": 258450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015772819958351456,
+      "loss": 2.0925,
+      "step": 258455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015772669036076672,
+      "loss": 1.9481,
+      "step": 258460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015772518111829844,
+      "loss": 2.1374,
+      "step": 258465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015772367185611016,
+      "loss": 2.1498,
+      "step": 258470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015772216257420242,
+      "loss": 2.0472,
+      "step": 258475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015772065327257577,
+      "loss": 2.1233,
+      "step": 258480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015771914395123067,
+      "loss": 2.1368,
+      "step": 258485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015771763461016767,
+      "loss": 2.1244,
+      "step": 258490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001577161252493873,
+      "loss": 2.0275,
+      "step": 258495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015771461586889005,
+      "loss": 2.1122,
+      "step": 258500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001577131064686764,
+      "loss": 2.3152,
+      "step": 258505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015771159704874697,
+      "loss": 2.0933,
+      "step": 258510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015771008760910216,
+      "loss": 2.137,
+      "step": 258515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015770857814974257,
+      "loss": 2.0155,
+      "step": 258520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.00015770706867066867,
+      "loss": 2.0627,
+      "step": 258525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015770555917188098,
+      "loss": 1.9383,
+      "step": 258530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015770404965338,
+      "loss": 2.0955,
+      "step": 258535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015770254011516634,
+      "loss": 1.8807,
+      "step": 258540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015770103055724039,
+      "loss": 2.1476,
+      "step": 258545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015769952097960277,
+      "loss": 2.119,
+      "step": 258550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001576980113822539,
+      "loss": 2.2225,
+      "step": 258555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015769650176519434,
+      "loss": 2.0408,
+      "step": 258560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00015769499212842463,
+      "loss": 1.7825,
+      "step": 258565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001576934824719453,
+      "loss": 2.0267,
+      "step": 258570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015769197279575677,
+      "loss": 2.1989,
+      "step": 258575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015769046309985964,
+      "loss": 2.1873,
+      "step": 258580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001576889533842544,
+      "loss": 1.8907,
+      "step": 258585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015768744364894158,
+      "loss": 2.0833,
+      "step": 258590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015768593389392166,
+      "loss": 2.0854,
+      "step": 258595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001576844241191952,
+      "loss": 2.1624,
+      "step": 258600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015768291432476268,
+      "loss": 2.1073,
+      "step": 258605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015768140451062463,
+      "loss": 2.1695,
+      "step": 258610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015767989467678159,
+      "loss": 2.014,
+      "step": 258615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015767838482323403,
+      "loss": 2.0014,
+      "step": 258620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001576768749499825,
+      "loss": 2.0247,
+      "step": 258625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015767536505702746,
+      "loss": 2.1652,
+      "step": 258630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001576738551443695,
+      "loss": 1.9825,
+      "step": 258635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001576723452120091,
+      "loss": 2.0831,
+      "step": 258640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015767083525994682,
+      "loss": 1.9748,
+      "step": 258645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001576693252881831,
+      "loss": 2.0027,
+      "step": 258650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001576678152967185,
+      "loss": 2.0462,
+      "step": 258655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015766630528555353,
+      "loss": 2.0887,
+      "step": 258660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001576647952546887,
+      "loss": 1.9841,
+      "step": 258665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015766328520412453,
+      "loss": 2.1131,
+      "step": 258670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015766177513386152,
+      "loss": 1.9782,
+      "step": 258675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015766026504390024,
+      "loss": 2.0679,
+      "step": 258680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015765875493424112,
+      "loss": 1.9906,
+      "step": 258685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015765724480488476,
+      "loss": 2.0422,
+      "step": 258690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015765573465583165,
+      "loss": 2.1836,
+      "step": 258695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015765422448708226,
+      "loss": 2.0054,
+      "step": 258700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015765271429863718,
+      "loss": 2.0918,
+      "step": 258705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015765120409049686,
+      "loss": 2.0843,
+      "step": 258710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015764969386266185,
+      "loss": 1.9461,
+      "step": 258715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015764818361513263,
+      "loss": 2.1174,
+      "step": 258720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001576466733479098,
+      "loss": 2.2601,
+      "step": 258725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015764516306099378,
+      "loss": 2.0981,
+      "step": 258730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015764365275438514,
+      "loss": 2.1217,
+      "step": 258735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015764214242808437,
+      "loss": 2.1462,
+      "step": 258740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000157640632082092,
+      "loss": 1.9766,
+      "step": 258745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015763912171640854,
+      "loss": 2.0771,
+      "step": 258750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015763761133103454,
+      "loss": 1.9638,
+      "step": 258755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001576361009259705,
+      "loss": 2.0369,
+      "step": 258760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015763459050121688,
+      "loss": 2.1531,
+      "step": 258765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015763308005677424,
+      "loss": 2.1424,
+      "step": 258770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015763156959264313,
+      "loss": 2.0091,
+      "step": 258775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000157630059108824,
+      "loss": 2.0517,
+      "step": 258780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001576285486053174,
+      "loss": 2.2477,
+      "step": 258785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015762703808212388,
+      "loss": 2.207,
+      "step": 258790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00015762552753924387,
+      "loss": 2.2305,
+      "step": 258795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015762401697667794,
+      "loss": 1.8575,
+      "step": 258800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001576225063944266,
+      "loss": 2.038,
+      "step": 258805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001576209957924904,
+      "loss": 2.1847,
+      "step": 258810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001576194851708698,
+      "loss": 2.1158,
+      "step": 258815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015761797452956534,
+      "loss": 2.0528,
+      "step": 258820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015761646386857755,
+      "loss": 2.002,
+      "step": 258825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001576149531879069,
+      "loss": 2.0897,
+      "step": 258830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015761344248755395,
+      "loss": 2.07,
+      "step": 258835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015761193176751923,
+      "loss": 2.1159,
+      "step": 258840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001576104210278032,
+      "loss": 2.3427,
+      "step": 258845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001576089102684064,
+      "loss": 2.1149,
+      "step": 258850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015760739948932938,
+      "loss": 2.0239,
+      "step": 258855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001576058886905726,
+      "loss": 2.0677,
+      "step": 258860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001576043778721366,
+      "loss": 2.0403,
+      "step": 258865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015760286703402193,
+      "loss": 2.2333,
+      "step": 258870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015760135617622905,
+      "loss": 2.1571,
+      "step": 258875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001575998452987585,
+      "loss": 2.1442,
+      "step": 258880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015759833440161082,
+      "loss": 2.1519,
+      "step": 258885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001575968234847865,
+      "loss": 2.2556,
+      "step": 258890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015759531254828606,
+      "loss": 1.9351,
+      "step": 258895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015759380159211002,
+      "loss": 1.9992,
+      "step": 258900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015759229061625886,
+      "loss": 1.9665,
+      "step": 258905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015759077962073317,
+      "loss": 2.1466,
+      "step": 258910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001575892686055334,
+      "loss": 2.0342,
+      "step": 258915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015758775757066011,
+      "loss": 2.0721,
+      "step": 258920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001575862465161138,
+      "loss": 1.9886,
+      "step": 258925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015758473544189494,
+      "loss": 2.0143,
+      "step": 258930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001575832243480041,
+      "loss": 2.0529,
+      "step": 258935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015758171323444184,
+      "loss": 2.2241,
+      "step": 258940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015758020210120862,
+      "loss": 2.0889,
+      "step": 258945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001575786909483049,
+      "loss": 2.0023,
+      "step": 258950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001575771797757313,
+      "loss": 2.2689,
+      "step": 258955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015757566858348824,
+      "loss": 2.1945,
+      "step": 258960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015757415737157633,
+      "loss": 2.0765,
+      "step": 258965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015757264613999606,
+      "loss": 2.0934,
+      "step": 258970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015757113488874792,
+      "loss": 1.9487,
+      "step": 258975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001575696236178324,
+      "loss": 2.1589,
+      "step": 258980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015756811232725006,
+      "loss": 2.0323,
+      "step": 258985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015756660101700146,
+      "loss": 2.1935,
+      "step": 258990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015756508968708702,
+      "loss": 2.1538,
+      "step": 258995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001575635783375073,
+      "loss": 2.0926,
+      "step": 259000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015756206696826284,
+      "loss": 2.116,
+      "step": 259005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015756055557935412,
+      "loss": 2.1991,
+      "step": 259010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015755904417078166,
+      "loss": 2.027,
+      "step": 259015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000157557532742546,
+      "loss": 2.0083,
+      "step": 259020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015755602129464763,
+      "loss": 2.1387,
+      "step": 259025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001575545098270871,
+      "loss": 2.0708,
+      "step": 259030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015755299833986492,
+      "loss": 1.8278,
+      "step": 259035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015755148683298153,
+      "loss": 2.13,
+      "step": 259040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00015754997530643755,
+      "loss": 2.024,
+      "step": 259045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015754846376023348,
+      "loss": 2.1883,
+      "step": 259050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015754695219436976,
+      "loss": 2.142,
+      "step": 259055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015754544060884703,
+      "loss": 2.131,
+      "step": 259060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00015754392900366564,
+      "loss": 2.2377,
+      "step": 259065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015754241737882623,
+      "loss": 2.1313,
+      "step": 259070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015754090573432935,
+      "loss": 2.0958,
+      "step": 259075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001575393940701754,
+      "loss": 2.0629,
+      "step": 259080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015753788238636495,
+      "loss": 1.9316,
+      "step": 259085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001575363706828985,
+      "loss": 1.9058,
+      "step": 259090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001575348589597766,
+      "loss": 2.2585,
+      "step": 259095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015753334721699974,
+      "loss": 2.0274,
+      "step": 259100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015753183545456846,
+      "loss": 1.9055,
+      "step": 259105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015753032367248326,
+      "loss": 1.9591,
+      "step": 259110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015752881187074466,
+      "loss": 2.2939,
+      "step": 259115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015752730004935317,
+      "loss": 2.2178,
+      "step": 259120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001575257882083093,
+      "loss": 1.9989,
+      "step": 259125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001575242763476136,
+      "loss": 2.0581,
+      "step": 259130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015752276446726654,
+      "loss": 1.9999,
+      "step": 259135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015752125256726868,
+      "loss": 1.9311,
+      "step": 259140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001575197406476205,
+      "loss": 2.0651,
+      "step": 259145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015751822870832256,
+      "loss": 2.1342,
+      "step": 259150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015751671674937533,
+      "loss": 2.1042,
+      "step": 259155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015751520477077935,
+      "loss": 2.0333,
+      "step": 259160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015751369277253513,
+      "loss": 1.8954,
+      "step": 259165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001575121807546432,
+      "loss": 2.2095,
+      "step": 259170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015751066871710404,
+      "loss": 1.9478,
+      "step": 259175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015750915665991823,
+      "loss": 2.0355,
+      "step": 259180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015750764458308622,
+      "loss": 1.8848,
+      "step": 259185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015750613248660856,
+      "loss": 2.0527,
+      "step": 259190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001575046203704858,
+      "loss": 2.0583,
+      "step": 259195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015750310823471833,
+      "loss": 2.0477,
+      "step": 259200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015750159607930684,
+      "loss": 2.0339,
+      "step": 259205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015750008390425174,
+      "loss": 2.1815,
+      "step": 259210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001574985717095536,
+      "loss": 2.0424,
+      "step": 259215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015749705949521286,
+      "loss": 1.865,
+      "step": 259220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015749554726123005,
+      "loss": 2.1749,
+      "step": 259225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015749403500760578,
+      "loss": 2.0945,
+      "step": 259230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015749252273434045,
+      "loss": 2.2829,
+      "step": 259235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015749101044143466,
+      "loss": 2.0296,
+      "step": 259240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015748949812888894,
+      "loss": 2.0674,
+      "step": 259245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015748798579670373,
+      "loss": 2.1796,
+      "step": 259250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015748647344487955,
+      "loss": 1.9804,
+      "step": 259255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000157484961073417,
+      "loss": 1.9742,
+      "step": 259260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015748344868231652,
+      "loss": 2.2029,
+      "step": 259265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015748193627157864,
+      "loss": 2.1683,
+      "step": 259270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001574804238412039,
+      "loss": 2.0776,
+      "step": 259275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001574789113911928,
+      "loss": 1.8306,
+      "step": 259280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015747739892154586,
+      "loss": 1.9821,
+      "step": 259285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001574758864322636,
+      "loss": 2.1434,
+      "step": 259290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015747437392334655,
+      "loss": 2.1969,
+      "step": 259295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001574728613947952,
+      "loss": 2.1517,
+      "step": 259300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015747134884661006,
+      "loss": 2.1906,
+      "step": 259305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001574698362787917,
+      "loss": 2.0653,
+      "step": 259310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001574683236913406,
+      "loss": 2.1237,
+      "step": 259315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015746681108425725,
+      "loss": 2.1737,
+      "step": 259320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001574652984575422,
+      "loss": 2.1609,
+      "step": 259325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015746378581119597,
+      "loss": 2.1339,
+      "step": 259330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001574622731452191,
+      "loss": 2.1355,
+      "step": 259335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015746076045961203,
+      "loss": 2.2688,
+      "step": 259340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015745924775437535,
+      "loss": 2.0772,
+      "step": 259345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015745773502950954,
+      "loss": 2.0109,
+      "step": 259350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015745622228501512,
+      "loss": 1.9627,
+      "step": 259355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001574547095208926,
+      "loss": 1.9484,
+      "step": 259360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015745319673714253,
+      "loss": 2.0518,
+      "step": 259365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015745168393376538,
+      "loss": 2.0794,
+      "step": 259370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015745017111076171,
+      "loss": 1.9942,
+      "step": 259375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015744865826813203,
+      "loss": 1.9388,
+      "step": 259380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015744714540587683,
+      "loss": 2.0687,
+      "step": 259385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015744563252399668,
+      "loss": 1.9182,
+      "step": 259390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015744411962249203,
+      "loss": 2.1207,
+      "step": 259395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015744260670136343,
+      "loss": 2.1584,
+      "step": 259400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001574410937606114,
+      "loss": 2.0584,
+      "step": 259405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015743958080023643,
+      "loss": 2.0339,
+      "step": 259410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015743806782023904,
+      "loss": 2.1332,
+      "step": 259415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015743655482061981,
+      "loss": 2.215,
+      "step": 259420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001574350418013792,
+      "loss": 2.1854,
+      "step": 259425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015743352876251775,
+      "loss": 2.1677,
+      "step": 259430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015743201570403596,
+      "loss": 1.9795,
+      "step": 259435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001574305026259343,
+      "loss": 1.9967,
+      "step": 259440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015742898952821342,
+      "loss": 2.2259,
+      "step": 259445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001574274764108737,
+      "loss": 2.0594,
+      "step": 259450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015742596327391573,
+      "loss": 1.9585,
+      "step": 259455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015742445011734,
+      "loss": 2.2038,
+      "step": 259460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015742293694114706,
+      "loss": 2.1545,
+      "step": 259465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015742142374533736,
+      "loss": 2.0009,
+      "step": 259470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015741991052991148,
+      "loss": 2.2519,
+      "step": 259475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015741839729486995,
+      "loss": 1.8598,
+      "step": 259480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00015741688404021322,
+      "loss": 2.1488,
+      "step": 259485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015741537076594187,
+      "loss": 2.0282,
+      "step": 259490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015741385747205635,
+      "loss": 1.8156,
+      "step": 259495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015741234415855723,
+      "loss": 1.9921,
+      "step": 259500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00015741083082544503,
+      "loss": 2.2845,
+      "step": 259505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015740931747272021,
+      "loss": 2.0583,
+      "step": 259510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015740780410038337,
+      "loss": 2.0314,
+      "step": 259515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015740629070843496,
+      "loss": 2.221,
+      "step": 259520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001574047772968755,
+      "loss": 2.0086,
+      "step": 259525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015740326386570557,
+      "loss": 2.1162,
+      "step": 259530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001574017504149256,
+      "loss": 2.1266,
+      "step": 259535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001574002369445362,
+      "loss": 2.1653,
+      "step": 259540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001573987234545378,
+      "loss": 2.1566,
+      "step": 259545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015739720994493097,
+      "loss": 1.9833,
+      "step": 259550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001573956964157162,
+      "loss": 2.0042,
+      "step": 259555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015739418286689406,
+      "loss": 2.1745,
+      "step": 259560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000157392669298465,
+      "loss": 2.0108,
+      "step": 259565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015739115571042955,
+      "loss": 1.978,
+      "step": 259570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015738964210278822,
+      "loss": 2.1679,
+      "step": 259575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001573881284755416,
+      "loss": 2.1749,
+      "step": 259580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015738661482869013,
+      "loss": 2.0844,
+      "step": 259585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015738510116223436,
+      "loss": 1.9769,
+      "step": 259590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015738358747617478,
+      "loss": 2.079,
+      "step": 259595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015738207377051193,
+      "loss": 2.1282,
+      "step": 259600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015738056004524636,
+      "loss": 2.0481,
+      "step": 259605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001573790463003785,
+      "loss": 2.1321,
+      "step": 259610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015737753253590895,
+      "loss": 1.8048,
+      "step": 259615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001573760187518382,
+      "loss": 2.1381,
+      "step": 259620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015737450494816672,
+      "loss": 2.0068,
+      "step": 259625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015737299112489507,
+      "loss": 2.0622,
+      "step": 259630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001573714772820238,
+      "loss": 2.2241,
+      "step": 259635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001573699634195534,
+      "loss": 2.0034,
+      "step": 259640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015736844953748435,
+      "loss": 2.0617,
+      "step": 259645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001573669356358172,
+      "loss": 2.1334,
+      "step": 259650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015736542171455246,
+      "loss": 2.1003,
+      "step": 259655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015736390777369068,
+      "loss": 2.215,
+      "step": 259660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015736239381323235,
+      "loss": 2.0295,
+      "step": 259665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015736087983317796,
+      "loss": 1.97,
+      "step": 259670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015735936583352806,
+      "loss": 2.0447,
+      "step": 259675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015735785181428315,
+      "loss": 1.9143,
+      "step": 259680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015735633777544375,
+      "loss": 2.1591,
+      "step": 259685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001573548237170104,
+      "loss": 2.1718,
+      "step": 259690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015735330963898364,
+      "loss": 2.1351,
+      "step": 259695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015735179554136392,
+      "loss": 2.1777,
+      "step": 259700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015735028142415178,
+      "loss": 2.051,
+      "step": 259705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015734876728734772,
+      "loss": 2.0611,
+      "step": 259710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001573472531309523,
+      "loss": 2.0517,
+      "step": 259715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015734573895496604,
+      "loss": 2.0121,
+      "step": 259720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015734422475938945,
+      "loss": 2.138,
+      "step": 259725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015734271054422297,
+      "loss": 2.1141,
+      "step": 259730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015734119630946721,
+      "loss": 2.1079,
+      "step": 259735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00015733968205512269,
+      "loss": 2.0785,
+      "step": 259740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015733816778118985,
+      "loss": 2.0038,
+      "step": 259745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015733665348766925,
+      "loss": 2.0107,
+      "step": 259750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015733513917456145,
+      "loss": 2.2688,
+      "step": 259755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015733362484186692,
+      "loss": 2.0696,
+      "step": 259760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015733211048958614,
+      "loss": 2.0844,
+      "step": 259765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001573305961177197,
+      "loss": 2.0469,
+      "step": 259770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001573290817262681,
+      "loss": 1.9226,
+      "step": 259775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001573275673152318,
+      "loss": 1.9658,
+      "step": 259780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015732605288461144,
+      "loss": 2.0737,
+      "step": 259785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015732453843440738,
+      "loss": 2.285,
+      "step": 259790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015732302396462026,
+      "loss": 1.9575,
+      "step": 259795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015732150947525057,
+      "loss": 2.0512,
+      "step": 259800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015731999496629878,
+      "loss": 2.3359,
+      "step": 259805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015731848043776546,
+      "loss": 2.0119,
+      "step": 259810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015731696588965107,
+      "loss": 2.2238,
+      "step": 259815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015731545132195623,
+      "loss": 2.1928,
+      "step": 259820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015731393673468135,
+      "loss": 2.0474,
+      "step": 259825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000157312422127827,
+      "loss": 1.9137,
+      "step": 259830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001573109075013937,
+      "loss": 2.1458,
+      "step": 259835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015730939285538194,
+      "loss": 1.9317,
+      "step": 259840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015730787818979224,
+      "loss": 1.9374,
+      "step": 259845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015730636350462515,
+      "loss": 2.1162,
+      "step": 259850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015730484879988117,
+      "loss": 2.1743,
+      "step": 259855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001573033340755608,
+      "loss": 2.2437,
+      "step": 259860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015730181933166457,
+      "loss": 2.1213,
+      "step": 259865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000157300304568193,
+      "loss": 2.0976,
+      "step": 259870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001572987897851466,
+      "loss": 2.314,
+      "step": 259875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001572972749825259,
+      "loss": 2.3204,
+      "step": 259880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015729576016033144,
+      "loss": 2.1244,
+      "step": 259885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015729424531856368,
+      "loss": 2.1679,
+      "step": 259890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015729273045722315,
+      "loss": 1.9699,
+      "step": 259895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015729121557631043,
+      "loss": 2.1508,
+      "step": 259900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015728970067582594,
+      "loss": 2.2471,
+      "step": 259905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001572881857557703,
+      "loss": 1.8595,
+      "step": 259910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015728667081614398,
+      "loss": 2.0562,
+      "step": 259915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015728515585694745,
+      "loss": 2.1046,
+      "step": 259920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015728364087818129,
+      "loss": 2.0301,
+      "step": 259925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015728212587984598,
+      "loss": 2.0363,
+      "step": 259930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001572806108619421,
+      "loss": 1.8667,
+      "step": 259935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001572790958244701,
+      "loss": 2.0962,
+      "step": 259940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015727758076743053,
+      "loss": 1.8739,
+      "step": 259945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015727606569082388,
+      "loss": 2.1614,
+      "step": 259950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001572745505946507,
+      "loss": 2.0199,
+      "step": 259955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001572730354789115,
+      "loss": 2.1077,
+      "step": 259960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015727152034360678,
+      "loss": 2.1047,
+      "step": 259965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001572700051887371,
+      "loss": 2.1383,
+      "step": 259970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001572684900143029,
+      "loss": 1.9994,
+      "step": 259975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015726697482030478,
+      "loss": 2.1005,
+      "step": 259980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015726545960674323,
+      "loss": 2.1081,
+      "step": 259985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015726394437361874,
+      "loss": 1.9627,
+      "step": 259990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015726242912093182,
+      "loss": 2.0975,
+      "step": 259995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015726091384868306,
+      "loss": 2.238,
+      "step": 260000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015725939855687292,
+      "loss": 2.0287,
+      "step": 260005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001572578832455019,
+      "loss": 1.9935,
+      "step": 260010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015725636791457062,
+      "loss": 2.1068,
+      "step": 260015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015725485256407948,
+      "loss": 1.9877,
+      "step": 260020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015725333719402907,
+      "loss": 2.1134,
+      "step": 260025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015725182180441983,
+      "loss": 2.0758,
+      "step": 260030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015725030639525238,
+      "loss": 2.2073,
+      "step": 260035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001572487909665272,
+      "loss": 2.0296,
+      "step": 260040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015724727551824473,
+      "loss": 2.1231,
+      "step": 260045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001572457600504056,
+      "loss": 2.0256,
+      "step": 260050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015724424456301027,
+      "loss": 2.093,
+      "step": 260055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015724272905605924,
+      "loss": 2.1813,
+      "step": 260060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015724121352955308,
+      "loss": 2.0381,
+      "step": 260065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015723969798349228,
+      "loss": 2.0009,
+      "step": 260070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015723818241787738,
+      "loss": 2.0945,
+      "step": 260075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015723666683270887,
+      "loss": 1.9847,
+      "step": 260080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015723515122798725,
+      "loss": 2.2085,
+      "step": 260085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015723363560371308,
+      "loss": 2.1531,
+      "step": 260090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015723211995988687,
+      "loss": 1.9834,
+      "step": 260095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015723060429650914,
+      "loss": 2.1039,
+      "step": 260100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001572290886135804,
+      "loss": 2.1752,
+      "step": 260105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015722757291110112,
+      "loss": 2.0422,
+      "step": 260110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001572260571890719,
+      "loss": 1.913,
+      "step": 260115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001572245414474932,
+      "loss": 2.058,
+      "step": 260120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001572230256863656,
+      "loss": 2.0071,
+      "step": 260125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015722150990568954,
+      "loss": 2.1108,
+      "step": 260130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001572199941054656,
+      "loss": 2.0084,
+      "step": 260135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015721847828569424,
+      "loss": 1.9903,
+      "step": 260140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000157216962446376,
+      "loss": 2.0111,
+      "step": 260145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015721544658751143,
+      "loss": 2.0779,
+      "step": 260150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015721393070910103,
+      "loss": 2.1675,
+      "step": 260155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015721241481114534,
+      "loss": 2.1956,
+      "step": 260160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001572108988936448,
+      "loss": 2.0829,
+      "step": 260165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001572093829566,
+      "loss": 1.971,
+      "step": 260170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015720786700001147,
+      "loss": 2.0432,
+      "step": 260175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015720635102387965,
+      "loss": 1.9724,
+      "step": 260180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015720483502820515,
+      "loss": 2.0262,
+      "step": 260185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015720331901298838,
+      "loss": 2.0258,
+      "step": 260190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015720180297822996,
+      "loss": 2.0746,
+      "step": 260195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015720028692393036,
+      "loss": 1.9427,
+      "step": 260200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015719877085009008,
+      "loss": 2.2385,
+      "step": 260205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001571972547567097,
+      "loss": 2.2167,
+      "step": 260210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015719573864378966,
+      "loss": 2.1203,
+      "step": 260215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015719422251133053,
+      "loss": 2.0663,
+      "step": 260220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015719270635933285,
+      "loss": 2.11,
+      "step": 260225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001571911901877971,
+      "loss": 2.0029,
+      "step": 260230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015718967399672376,
+      "loss": 2.2264,
+      "step": 260235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001571881577861134,
+      "loss": 2.0027,
+      "step": 260240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015718664155596655,
+      "loss": 2.2405,
+      "step": 260245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015718512530628372,
+      "loss": 1.9805,
+      "step": 260250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015718360903706538,
+      "loss": 2.1693,
+      "step": 260255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.0001571820927483121,
+      "loss": 2.2364,
+      "step": 260260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015718057644002439,
+      "loss": 1.923,
+      "step": 260265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015717906011220273,
+      "loss": 2.0757,
+      "step": 260270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015717754376484767,
+      "loss": 2.2703,
+      "step": 260275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015717602739795974,
+      "loss": 2.144,
+      "step": 260280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015717451101153945,
+      "loss": 2.0448,
+      "step": 260285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015717299460558728,
+      "loss": 2.1002,
+      "step": 260290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001571714781801038,
+      "loss": 2.2522,
+      "step": 260295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001571699617350895,
+      "loss": 2.0979,
+      "step": 260300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015716844527054492,
+      "loss": 2.0576,
+      "step": 260305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015716692878647054,
+      "loss": 1.9743,
+      "step": 260310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015716541228286688,
+      "loss": 2.0935,
+      "step": 260315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015716389575973452,
+      "loss": 2.012,
+      "step": 260320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015716237921707392,
+      "loss": 2.0728,
+      "step": 260325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015716086265488561,
+      "loss": 2.0155,
+      "step": 260330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015715934607317013,
+      "loss": 2.0737,
+      "step": 260335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015715782947192794,
+      "loss": 2.0859,
+      "step": 260340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00015715631285115965,
+      "loss": 2.1825,
+      "step": 260345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015715479621086572,
+      "loss": 1.9347,
+      "step": 260350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015715327955104664,
+      "loss": 1.853,
+      "step": 260355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015715176287170297,
+      "loss": 2.0694,
+      "step": 260360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015715024617283526,
+      "loss": 2.1339,
+      "step": 260365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015714872945444394,
+      "loss": 2.0618,
+      "step": 260370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015714721271652962,
+      "loss": 1.9685,
+      "step": 260375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015714569595909274,
+      "loss": 2.1848,
+      "step": 260380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015714417918213388,
+      "loss": 1.9958,
+      "step": 260385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015714266238565352,
+      "loss": 2.1531,
+      "step": 260390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015714114556965222,
+      "loss": 2.089,
+      "step": 260395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015713962873413042,
+      "loss": 2.0147,
+      "step": 260400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001571381118790887,
+      "loss": 2.0335,
+      "step": 260405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001571365950045276,
+      "loss": 2.2016,
+      "step": 260410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015713507811044756,
+      "loss": 2.1096,
+      "step": 260415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015713356119684914,
+      "loss": 2.079,
+      "step": 260420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.0001571320442637329,
+      "loss": 1.9923,
+      "step": 260425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015713052731109928,
+      "loss": 2.237,
+      "step": 260430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015712901033894885,
+      "loss": 1.9676,
+      "step": 260435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015712749334728212,
+      "loss": 1.8743,
+      "step": 260440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001571259763360996,
+      "loss": 2.0557,
+      "step": 260445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001571244593054018,
+      "loss": 2.2501,
+      "step": 260450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015712294225518926,
+      "loss": 2.1071,
+      "step": 260455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015712142518546247,
+      "loss": 2.0563,
+      "step": 260460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000157119908096222,
+      "loss": 2.2183,
+      "step": 260465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001571183909874683,
+      "loss": 1.9573,
+      "step": 260470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015711687385920194,
+      "loss": 2.2055,
+      "step": 260475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001571153567114234,
+      "loss": 1.9601,
+      "step": 260480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001571138395441332,
+      "loss": 1.9011,
+      "step": 260485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015711232235733193,
+      "loss": 1.9674,
+      "step": 260490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015711080515102004,
+      "loss": 2.0872,
+      "step": 260495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001571092879251981,
+      "loss": 2.0944,
+      "step": 260500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001571077706798665,
+      "loss": 2.2469,
+      "step": 260505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015710625341502593,
+      "loss": 2.1582,
+      "step": 260510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015710473613067677,
+      "loss": 1.8998,
+      "step": 260515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001571032188268196,
+      "loss": 2.0832,
+      "step": 260520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015710170150345497,
+      "loss": 2.0107,
+      "step": 260525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015710018416058335,
+      "loss": 2.1958,
+      "step": 260530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001570986667982053,
+      "loss": 2.1823,
+      "step": 260535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015709714941632122,
+      "loss": 2.175,
+      "step": 260540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015709563201493182,
+      "loss": 2.1282,
+      "step": 260545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015709411459403748,
+      "loss": 2.0146,
+      "step": 260550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015709259715363875,
+      "loss": 1.9134,
+      "step": 260555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00015709107969373617,
+      "loss": 2.1692,
+      "step": 260560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001570895622143302,
+      "loss": 1.9857,
+      "step": 260565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001570880447154214,
+      "loss": 2.0148,
+      "step": 260570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015708652719701032,
+      "loss": 1.9626,
+      "step": 260575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015708500965909745,
+      "loss": 2.1339,
+      "step": 260580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015708349210168332,
+      "loss": 2.025,
+      "step": 260585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015708197452476841,
+      "loss": 1.9732,
+      "step": 260590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015708045692835326,
+      "loss": 2.1032,
+      "step": 260595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015707893931243838,
+      "loss": 2.1118,
+      "step": 260600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015707742167702434,
+      "loss": 2.2663,
+      "step": 260605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015707590402211157,
+      "loss": 2.0754,
+      "step": 260610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015707438634770065,
+      "loss": 2.1687,
+      "step": 260615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015707286865379206,
+      "loss": 2.0245,
+      "step": 260620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001570713509403864,
+      "loss": 2.0373,
+      "step": 260625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015706983320748412,
+      "loss": 1.8071,
+      "step": 260630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015706831545508573,
+      "loss": 1.988,
+      "step": 260635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001570667976831918,
+      "loss": 2.0387,
+      "step": 260640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015706527989180278,
+      "loss": 2.0949,
+      "step": 260645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015706376208091921,
+      "loss": 2.0239,
+      "step": 260650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015706224425054166,
+      "loss": 2.077,
+      "step": 260655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015706072640067062,
+      "loss": 2.0587,
+      "step": 260660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015705920853130656,
+      "loss": 1.9752,
+      "step": 260665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015705769064245007,
+      "loss": 2.006,
+      "step": 260670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015705617273410163,
+      "loss": 2.2758,
+      "step": 260675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001570546548062618,
+      "loss": 1.8232,
+      "step": 260680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015705313685893102,
+      "loss": 1.9555,
+      "step": 260685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015705161889210985,
+      "loss": 2.1569,
+      "step": 260690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015705010090579885,
+      "loss": 2.0756,
+      "step": 260695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015704858289999844,
+      "loss": 2.008,
+      "step": 260700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015704706487470926,
+      "loss": 2.1922,
+      "step": 260705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015704554682993176,
+      "loss": 2.045,
+      "step": 260710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015704402876566645,
+      "loss": 2.0604,
+      "step": 260715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015704251068191386,
+      "loss": 2.0764,
+      "step": 260720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015704099257867453,
+      "loss": 2.0595,
+      "step": 260725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015703947445594896,
+      "loss": 1.995,
+      "step": 260730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015703795631373765,
+      "loss": 2.0802,
+      "step": 260735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015703643815204117,
+      "loss": 2.0124,
+      "step": 260740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015703491997086,
+      "loss": 1.9294,
+      "step": 260745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015703340177019467,
+      "loss": 2.096,
+      "step": 260750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015703188355004565,
+      "loss": 2.1893,
+      "step": 260755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015703036531041357,
+      "loss": 2.1209,
+      "step": 260760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015702884705129884,
+      "loss": 1.8559,
+      "step": 260765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015702732877270203,
+      "loss": 1.7734,
+      "step": 260770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015702581047462365,
+      "loss": 1.9918,
+      "step": 260775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015702429215706424,
+      "loss": 2.2052,
+      "step": 260780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.0001570227738200243,
+      "loss": 2.0951,
+      "step": 260785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001570212554635043,
+      "loss": 2.2135,
+      "step": 260790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015701973708750482,
+      "loss": 2.0751,
+      "step": 260795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001570182186920264,
+      "loss": 2.0525,
+      "step": 260800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001570167002770695,
+      "loss": 2.2845,
+      "step": 260805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015701518184263466,
+      "loss": 2.0449,
+      "step": 260810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015701366338872238,
+      "loss": 1.9974,
+      "step": 260815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00015701214491533325,
+      "loss": 2.0989,
+      "step": 260820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001570106264224677,
+      "loss": 1.9022,
+      "step": 260825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001570091079101263,
+      "loss": 2.1828,
+      "step": 260830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015700758937830953,
+      "loss": 2.0107,
+      "step": 260835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015700607082701798,
+      "loss": 2.0165,
+      "step": 260840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015700455225625207,
+      "loss": 1.9834,
+      "step": 260845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001570030336660124,
+      "loss": 2.0578,
+      "step": 260850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001570015150562995,
+      "loss": 2.1629,
+      "step": 260855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015699999642711373,
+      "loss": 2.1889,
+      "step": 260860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015699847777845584,
+      "loss": 1.9207,
+      "step": 260865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001569969591103262,
+      "loss": 2.0961,
+      "step": 260870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015699544042272536,
+      "loss": 1.8406,
+      "step": 260875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015699392171565383,
+      "loss": 2.0867,
+      "step": 260880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015699240298911216,
+      "loss": 2.1545,
+      "step": 260885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015699088424310083,
+      "loss": 2.1786,
+      "step": 260890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015698936547762045,
+      "loss": 2.0816,
+      "step": 260895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015698784669267138,
+      "loss": 2.0495,
+      "step": 260900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001569863278882543,
+      "loss": 2.1754,
+      "step": 260905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015698480906436963,
+      "loss": 1.9026,
+      "step": 260910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001569832902210179,
+      "loss": 2.1509,
+      "step": 260915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015698177135819964,
+      "loss": 2.0733,
+      "step": 260920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015698025247591542,
+      "loss": 2.109,
+      "step": 260925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00015697873357416566,
+      "loss": 1.9573,
+      "step": 260930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015697721465295097,
+      "loss": 1.993,
+      "step": 260935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001569756957122718,
+      "loss": 2.1943,
+      "step": 260940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001569741767521287,
+      "loss": 1.9053,
+      "step": 260945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001569726577725222,
+      "loss": 2.2602,
+      "step": 260950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015697113877345282,
+      "loss": 2.0673,
+      "step": 260955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015696961975492104,
+      "loss": 2.0312,
+      "step": 260960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015696810071692743,
+      "loss": 2.1169,
+      "step": 260965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015696658165947243,
+      "loss": 2.1634,
+      "step": 260970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001569650625825567,
+      "loss": 2.0364,
+      "step": 260975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001569635434861806,
+      "loss": 1.9395,
+      "step": 260980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015696202437034476,
+      "loss": 1.9801,
+      "step": 260985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015696050523504967,
+      "loss": 2.0036,
+      "step": 260990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015695898608029577,
+      "loss": 2.1055,
+      "step": 260995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001569574669060837,
+      "loss": 2.0412,
+      "step": 261000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015695594771241393,
+      "loss": 2.0912,
+      "step": 261005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015695442849928694,
+      "loss": 2.0946,
+      "step": 261010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015695290926670335,
+      "loss": 1.9201,
+      "step": 261015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015695139001466353,
+      "loss": 2.197,
+      "step": 261020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001569498707431681,
+      "loss": 2.0621,
+      "step": 261025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001569483514522176,
+      "loss": 2.047,
+      "step": 261030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015694683214181253,
+      "loss": 1.9658,
+      "step": 261035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015694531281195334,
+      "loss": 2.1841,
+      "step": 261040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001569437934626406,
+      "loss": 2.0129,
+      "step": 261045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015694227409387482,
+      "loss": 2.1401,
+      "step": 261050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015694075470565656,
+      "loss": 2.0521,
+      "step": 261055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001569392352979863,
+      "loss": 1.9213,
+      "step": 261060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015693771587086457,
+      "loss": 2.0445,
+      "step": 261065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015693619642429186,
+      "loss": 2.1727,
+      "step": 261070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001569346769582687,
+      "loss": 1.9824,
+      "step": 261075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015693315747279565,
+      "loss": 1.9692,
+      "step": 261080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001569316379678732,
+      "loss": 2.0745,
+      "step": 261085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015693011844350186,
+      "loss": 2.1197,
+      "step": 261090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015692859889968218,
+      "loss": 2.1342,
+      "step": 261095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015692707933641465,
+      "loss": 2.0432,
+      "step": 261100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015692555975369975,
+      "loss": 2.026,
+      "step": 261105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001569240401515381,
+      "loss": 2.3132,
+      "step": 261110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015692252052993015,
+      "loss": 1.9524,
+      "step": 261115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015692100088887645,
+      "loss": 2.0472,
+      "step": 261120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015691948122837748,
+      "loss": 2.0273,
+      "step": 261125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001569179615484338,
+      "loss": 2.0196,
+      "step": 261130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001569164418490459,
+      "loss": 1.9551,
+      "step": 261135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001569149221302143,
+      "loss": 1.8719,
+      "step": 261140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015691340239193955,
+      "loss": 2.1025,
+      "step": 261145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015691188263422214,
+      "loss": 1.9106,
+      "step": 261150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001569103628570626,
+      "loss": 2.0865,
+      "step": 261155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015690884306046146,
+      "loss": 1.9914,
+      "step": 261160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015690732324441924,
+      "loss": 2.0919,
+      "step": 261165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015690580340893643,
+      "loss": 2.0318,
+      "step": 261170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015690428355401355,
+      "loss": 1.8012,
+      "step": 261175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015690276367965114,
+      "loss": 2.056,
+      "step": 261180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001569012437858497,
+      "loss": 1.8003,
+      "step": 261185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001568997238726098,
+      "loss": 1.9647,
+      "step": 261190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001568982039399319,
+      "loss": 2.0809,
+      "step": 261195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015689668398781654,
+      "loss": 2.088,
+      "step": 261200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015689516401626427,
+      "loss": 2.0217,
+      "step": 261205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015689364402527554,
+      "loss": 1.8842,
+      "step": 261210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001568921240148509,
+      "loss": 2.0353,
+      "step": 261215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015689060398499092,
+      "loss": 2.0593,
+      "step": 261220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015688908393569608,
+      "loss": 2.089,
+      "step": 261225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001568875638669669,
+      "loss": 1.9886,
+      "step": 261230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015688604377880387,
+      "loss": 1.9353,
+      "step": 261235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015688452367120754,
+      "loss": 2.0387,
+      "step": 261240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015688300354417843,
+      "loss": 2.1723,
+      "step": 261245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015688148339771706,
+      "loss": 2.0268,
+      "step": 261250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015687996323182393,
+      "loss": 2.004,
+      "step": 261255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015687844304649958,
+      "loss": 1.9027,
+      "step": 261260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001568769228417445,
+      "loss": 1.8646,
+      "step": 261265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015687540261755925,
+      "loss": 1.9517,
+      "step": 261270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015687388237394436,
+      "loss": 2.2397,
+      "step": 261275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015687236211090028,
+      "loss": 1.8839,
+      "step": 261280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015687084182842759,
+      "loss": 2.0445,
+      "step": 261285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015686932152652675,
+      "loss": 2.0548,
+      "step": 261290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00015686780120519837,
+      "loss": 1.9655,
+      "step": 261295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001568662808644429,
+      "loss": 2.2051,
+      "step": 261300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015686476050426088,
+      "loss": 1.9344,
+      "step": 261305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015686324012465283,
+      "loss": 1.9717,
+      "step": 261310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015686171972561925,
+      "loss": 2.0243,
+      "step": 261315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015686019930716069,
+      "loss": 2.0565,
+      "step": 261320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015685867886927764,
+      "loss": 1.9238,
+      "step": 261325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015685715841197063,
+      "loss": 1.9821,
+      "step": 261330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001568556379352402,
+      "loss": 2.1154,
+      "step": 261335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015685411743908686,
+      "loss": 2.0464,
+      "step": 261340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001568525969235111,
+      "loss": 2.1393,
+      "step": 261345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001568510763885135,
+      "loss": 2.0376,
+      "step": 261350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015684955583409452,
+      "loss": 2.177,
+      "step": 261355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001568480352602547,
+      "loss": 2.1179,
+      "step": 261360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015684651466699452,
+      "loss": 2.2344,
+      "step": 261365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001568449940543146,
+      "loss": 2.0786,
+      "step": 261370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015684347342221537,
+      "loss": 2.0835,
+      "step": 261375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015684195277069736,
+      "loss": 2.2159,
+      "step": 261380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015684043209976116,
+      "loss": 2.0441,
+      "step": 261385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001568389114094072,
+      "loss": 2.1513,
+      "step": 261390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015683739069963603,
+      "loss": 2.2152,
+      "step": 261395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001568358699704482,
+      "loss": 2.0972,
+      "step": 261400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015683434922184421,
+      "loss": 1.9079,
+      "step": 261405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015683282845382456,
+      "loss": 2.086,
+      "step": 261410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015683130766638978,
+      "loss": 2.2592,
+      "step": 261415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001568297868595404,
+      "loss": 2.0255,
+      "step": 261420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015682826603327695,
+      "loss": 2.1184,
+      "step": 261425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001568267451875999,
+      "loss": 2.0827,
+      "step": 261430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015682522432250984,
+      "loss": 2.3114,
+      "step": 261435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015682370343800725,
+      "loss": 2.2212,
+      "step": 261440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015682218253409262,
+      "loss": 2.0748,
+      "step": 261445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001568206616107665,
+      "loss": 2.0165,
+      "step": 261450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015681914066802946,
+      "loss": 2.0244,
+      "step": 261455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015681761970588195,
+      "loss": 2.0682,
+      "step": 261460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001568160987243245,
+      "loss": 2.0374,
+      "step": 261465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015681457772335767,
+      "loss": 2.0682,
+      "step": 261470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001568130567029819,
+      "loss": 2.2666,
+      "step": 261475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015681153566319777,
+      "loss": 2.0143,
+      "step": 261480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015681001460400581,
+      "loss": 2.1061,
+      "step": 261485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015680849352540653,
+      "loss": 2.1335,
+      "step": 261490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001568069724274004,
+      "loss": 2.0194,
+      "step": 261495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000156805451309988,
+      "loss": 2.2749,
+      "step": 261500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001568039301731698,
+      "loss": 2.1034,
+      "step": 261505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015680240901694642,
+      "loss": 2.1085,
+      "step": 261510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015680088784131825,
+      "loss": 2.1734,
+      "step": 261515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015679936664628587,
+      "loss": 2.1874,
+      "step": 261520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015679784543184982,
+      "loss": 2.0225,
+      "step": 261525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015679632419801057,
+      "loss": 2.2112,
+      "step": 261530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015679480294476867,
+      "loss": 2.1078,
+      "step": 261535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015679328167212463,
+      "loss": 2.0656,
+      "step": 261540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015679176038007898,
+      "loss": 2.0379,
+      "step": 261545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015679023906863226,
+      "loss": 2.2091,
+      "step": 261550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001567887177377849,
+      "loss": 2.0398,
+      "step": 261555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015678719638753755,
+      "loss": 1.8276,
+      "step": 261560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015678567501789065,
+      "loss": 1.9303,
+      "step": 261565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015678415362884473,
+      "loss": 2.1204,
+      "step": 261570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015678263222040033,
+      "loss": 1.936,
+      "step": 261575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001567811107925579,
+      "loss": 2.1672,
+      "step": 261580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015677958934531806,
+      "loss": 2.1268,
+      "step": 261585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015677806787868127,
+      "loss": 2.0855,
+      "step": 261590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015677654639264805,
+      "loss": 2.0748,
+      "step": 261595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015677502488721894,
+      "loss": 1.9503,
+      "step": 261600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015677350336239443,
+      "loss": 1.9675,
+      "step": 261605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015677198181817512,
+      "loss": 2.0109,
+      "step": 261610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015677046025456144,
+      "loss": 2.1333,
+      "step": 261615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015676893867155392,
+      "loss": 1.9923,
+      "step": 261620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015676741706915312,
+      "loss": 2.0955,
+      "step": 261625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015676589544735955,
+      "loss": 2.004,
+      "step": 261630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001567643738061737,
+      "loss": 2.1532,
+      "step": 261635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015676285214559613,
+      "loss": 2.0752,
+      "step": 261640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015676133046562733,
+      "loss": 2.1776,
+      "step": 261645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015675980876626785,
+      "loss": 2.0506,
+      "step": 261650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015675828704751817,
+      "loss": 2.1382,
+      "step": 261655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001567567653093788,
+      "loss": 2.1097,
+      "step": 261660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001567552435518503,
+      "loss": 2.232,
+      "step": 261665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001567537217749332,
+      "loss": 2.1511,
+      "step": 261670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015675219997862802,
+      "loss": 2.0282,
+      "step": 261675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015675067816293525,
+      "loss": 1.908,
+      "step": 261680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015674915632785542,
+      "loss": 1.9922,
+      "step": 261685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000156747634473389,
+      "loss": 1.9259,
+      "step": 261690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001567461125995366,
+      "loss": 2.2491,
+      "step": 261695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001567445907062987,
+      "loss": 2.0582,
+      "step": 261700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015674306879367582,
+      "loss": 2.2127,
+      "step": 261705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015674154686166845,
+      "loss": 2.0246,
+      "step": 261710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015674002491027718,
+      "loss": 2.0739,
+      "step": 261715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015673850293950245,
+      "loss": 2.1409,
+      "step": 261720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015673698094934485,
+      "loss": 1.9664,
+      "step": 261725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015673545893980486,
+      "loss": 2.1006,
+      "step": 261730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000156733936910883,
+      "loss": 2.0409,
+      "step": 261735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001567324148625798,
+      "loss": 2.1069,
+      "step": 261740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015673089279489576,
+      "loss": 2.0769,
+      "step": 261745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015672937070783144,
+      "loss": 2.1925,
+      "step": 261750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015672784860138735,
+      "loss": 2.0417,
+      "step": 261755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015672632647556397,
+      "loss": 2.1555,
+      "step": 261760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00015672480433036188,
+      "loss": 1.938,
+      "step": 261765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015672328216578154,
+      "loss": 1.9077,
+      "step": 261770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001567217599818235,
+      "loss": 2.011,
+      "step": 261775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001567202377784883,
+      "loss": 2.1721,
+      "step": 261780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015671871555577641,
+      "loss": 2.1755,
+      "step": 261785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001567171933136884,
+      "loss": 2.0018,
+      "step": 261790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015671567105222478,
+      "loss": 2.0882,
+      "step": 261795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015671414877138602,
+      "loss": 2.0487,
+      "step": 261800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001567126264711727,
+      "loss": 2.2296,
+      "step": 261805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015671110415158533,
+      "loss": 2.0452,
+      "step": 261810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015670958181262438,
+      "loss": 2.1406,
+      "step": 261815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015670805945429045,
+      "loss": 2.1292,
+      "step": 261820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015670653707658397,
+      "loss": 2.1341,
+      "step": 261825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015670501467950556,
+      "loss": 1.9824,
+      "step": 261830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015670349226305566,
+      "loss": 2.2863,
+      "step": 261835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015670196982723485,
+      "loss": 2.0002,
+      "step": 261840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015670044737204356,
+      "loss": 2.2802,
+      "step": 261845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015669892489748242,
+      "loss": 2.2398,
+      "step": 261850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015669740240355185,
+      "loss": 2.1463,
+      "step": 261855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015669587989025247,
+      "loss": 1.9052,
+      "step": 261860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015669435735758474,
+      "loss": 2.0459,
+      "step": 261865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001566928348055492,
+      "loss": 2.156,
+      "step": 261870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015669131223414632,
+      "loss": 2.1327,
+      "step": 261875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015668978964337668,
+      "loss": 1.9966,
+      "step": 261880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001566882670332408,
+      "loss": 2.075,
+      "step": 261885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015668674440373913,
+      "loss": 2.0735,
+      "step": 261890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015668522175487227,
+      "loss": 1.9194,
+      "step": 261895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001566836990866407,
+      "loss": 1.9809,
+      "step": 261900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015668217639904498,
+      "loss": 2.2345,
+      "step": 261905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015668065369208557,
+      "loss": 2.0647,
+      "step": 261910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015667913096576304,
+      "loss": 2.1915,
+      "step": 261915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015667760822007788,
+      "loss": 1.8515,
+      "step": 261920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015667608545503063,
+      "loss": 1.9588,
+      "step": 261925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015667456267062178,
+      "loss": 2.0598,
+      "step": 261930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001566730398668519,
+      "loss": 2.3005,
+      "step": 261935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015667151704372144,
+      "loss": 2.0383,
+      "step": 261940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000156669994201231,
+      "loss": 2.0073,
+      "step": 261945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015666847133938107,
+      "loss": 2.1042,
+      "step": 261950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015666694845817216,
+      "loss": 2.2161,
+      "step": 261955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015666542555760473,
+      "loss": 1.9975,
+      "step": 261960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015666390263767942,
+      "loss": 2.1988,
+      "step": 261965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015666237969839668,
+      "loss": 2.0652,
+      "step": 261970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015666085673975708,
+      "loss": 2.0624,
+      "step": 261975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015665933376176106,
+      "loss": 2.1151,
+      "step": 261980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015665781076440918,
+      "loss": 2.1829,
+      "step": 261985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015665628774770197,
+      "loss": 2.2207,
+      "step": 261990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015665476471163995,
+      "loss": 1.9882,
+      "step": 261995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015665324165622366,
+      "loss": 2.2202,
+      "step": 262000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015665171858145357,
+      "loss": 2.1265,
+      "step": 262005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015665019548733025,
+      "loss": 2.0484,
+      "step": 262010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015664867237385414,
+      "loss": 2.0088,
+      "step": 262015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015664714924102586,
+      "loss": 2.1027,
+      "step": 262020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015664562608884588,
+      "loss": 2.149,
+      "step": 262025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015664410291731473,
+      "loss": 2.1007,
+      "step": 262030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015664257972643292,
+      "loss": 1.9672,
+      "step": 262035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015664105651620096,
+      "loss": 2.1489,
+      "step": 262040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015663953328661943,
+      "loss": 2.0877,
+      "step": 262045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015663801003768877,
+      "loss": 2.0176,
+      "step": 262050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015663648676940957,
+      "loss": 2.1486,
+      "step": 262055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015663496348178232,
+      "loss": 2.0543,
+      "step": 262060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015663344017480748,
+      "loss": 2.0489,
+      "step": 262065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015663191684848568,
+      "loss": 1.7001,
+      "step": 262070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015663039350281735,
+      "loss": 2.2213,
+      "step": 262075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001566288701378031,
+      "loss": 2.2749,
+      "step": 262080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001566273467534434,
+      "loss": 2.1288,
+      "step": 262085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015662582334973876,
+      "loss": 2.2776,
+      "step": 262090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015662429992668967,
+      "loss": 2.2276,
+      "step": 262095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015662277648429676,
+      "loss": 2.014,
+      "step": 262100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015662125302256047,
+      "loss": 1.8729,
+      "step": 262105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001566197295414813,
+      "loss": 1.9627,
+      "step": 262110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001566182060410598,
+      "loss": 2.255,
+      "step": 262115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001566166825212965,
+      "loss": 2.2049,
+      "step": 262120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015661515898219194,
+      "loss": 2.098,
+      "step": 262125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015661363542374658,
+      "loss": 1.8555,
+      "step": 262130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015661211184596102,
+      "loss": 1.902,
+      "step": 262135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015661058824883572,
+      "loss": 2.0935,
+      "step": 262140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0001566090646323712,
+      "loss": 2.0608,
+      "step": 262145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015660754099656803,
+      "loss": 2.075,
+      "step": 262150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015660601734142666,
+      "loss": 2.0528,
+      "step": 262155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015660449366694765,
+      "loss": 2.1061,
+      "step": 262160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015660296997313153,
+      "loss": 2.1627,
+      "step": 262165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001566014462599788,
+      "loss": 2.024,
+      "step": 262170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015659992252749002,
+      "loss": 1.8643,
+      "step": 262175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015659839877566565,
+      "loss": 1.9367,
+      "step": 262180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015659687500450627,
+      "loss": 2.0495,
+      "step": 262185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015659535121401231,
+      "loss": 2.121,
+      "step": 262190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001565938274041844,
+      "loss": 2.1461,
+      "step": 262195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015659230357502302,
+      "loss": 2.098,
+      "step": 262200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015659077972652864,
+      "loss": 1.9526,
+      "step": 262205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015658925585870188,
+      "loss": 1.9843,
+      "step": 262210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015658773197154316,
+      "loss": 2.0365,
+      "step": 262215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015658620806505304,
+      "loss": 1.9699,
+      "step": 262220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015658468413923206,
+      "loss": 1.9679,
+      "step": 262225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015658316019408075,
+      "loss": 2.1333,
+      "step": 262230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001565816362295996,
+      "loss": 1.979,
+      "step": 262235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001565801122457891,
+      "loss": 2.0601,
+      "step": 262240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001565785882426498,
+      "loss": 2.0807,
+      "step": 262245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001565770642201823,
+      "loss": 2.0328,
+      "step": 262250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000156575540178387,
+      "loss": 2.0739,
+      "step": 262255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015657401611726446,
+      "loss": 2.0393,
+      "step": 262260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015657249203681524,
+      "loss": 2.0085,
+      "step": 262265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001565709679370398,
+      "loss": 2.0409,
+      "step": 262270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001565694438179387,
+      "loss": 2.0733,
+      "step": 262275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015656791967951246,
+      "loss": 2.0947,
+      "step": 262280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001565663955217616,
+      "loss": 2.2144,
+      "step": 262285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015656487134468663,
+      "loss": 2.0453,
+      "step": 262290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015656334714828806,
+      "loss": 1.99,
+      "step": 262295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015656182293256646,
+      "loss": 2.2404,
+      "step": 262300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00015656029869752226,
+      "loss": 2.2104,
+      "step": 262305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015655877444315606,
+      "loss": 2.0336,
+      "step": 262310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015655725016946835,
+      "loss": 1.9955,
+      "step": 262315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001565557258764597,
+      "loss": 2.2461,
+      "step": 262320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015655420156413053,
+      "loss": 1.9963,
+      "step": 262325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015655267723248145,
+      "loss": 1.9368,
+      "step": 262330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015655115288151298,
+      "loss": 2.066,
+      "step": 262335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015654962851122556,
+      "loss": 2.0169,
+      "step": 262340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001565481041216198,
+      "loss": 2.1368,
+      "step": 262345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015654657971269614,
+      "loss": 2.1964,
+      "step": 262350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015654505528445517,
+      "loss": 2.1829,
+      "step": 262355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001565435308368974,
+      "loss": 2.1272,
+      "step": 262360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001565420063700233,
+      "loss": 2.1199,
+      "step": 262365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015654048188383344,
+      "loss": 2.0607,
+      "step": 262370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015653895737832835,
+      "loss": 2.0242,
+      "step": 262375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015653743285350846,
+      "loss": 2.0487,
+      "step": 262380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001565359083093744,
+      "loss": 2.2478,
+      "step": 262385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015653438374592664,
+      "loss": 1.9035,
+      "step": 262390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001565328591631657,
+      "loss": 2.1207,
+      "step": 262395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015653133456109214,
+      "loss": 1.8758,
+      "step": 262400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015652980993970643,
+      "loss": 2.0504,
+      "step": 262405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015652828529900912,
+      "loss": 2.2763,
+      "step": 262410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00015652676063900074,
+      "loss": 2.1485,
+      "step": 262415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015652523595968178,
+      "loss": 2.1797,
+      "step": 262420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015652371126105276,
+      "loss": 2.227,
+      "step": 262425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001565221865431142,
+      "loss": 1.9153,
+      "step": 262430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001565206618058667,
+      "loss": 2.1192,
+      "step": 262435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015651913704931066,
+      "loss": 2.1216,
+      "step": 262440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015651761227344668,
+      "loss": 2.2116,
+      "step": 262445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015651608747827524,
+      "loss": 2.0848,
+      "step": 262450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001565145626637969,
+      "loss": 2.0151,
+      "step": 262455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015651303783001215,
+      "loss": 2.0483,
+      "step": 262460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001565115129769215,
+      "loss": 2.0827,
+      "step": 262465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015650998810452554,
+      "loss": 2.1165,
+      "step": 262470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015650846321282473,
+      "loss": 2.1453,
+      "step": 262475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015650693830181957,
+      "loss": 1.9924,
+      "step": 262480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015650541337151064,
+      "loss": 2.0318,
+      "step": 262485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015650388842189842,
+      "loss": 2.1071,
+      "step": 262490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001565023634529835,
+      "loss": 2.0484,
+      "step": 262495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001565008384647663,
+      "loss": 2.0491,
+      "step": 262500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001564993134572474,
+      "loss": 2.0172,
+      "step": 262505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015649778843042727,
+      "loss": 2.2025,
+      "step": 262510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015649626338430653,
+      "loss": 2.0886,
+      "step": 262515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015649473831888563,
+      "loss": 2.1686,
+      "step": 262520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001564932132341651,
+      "loss": 1.9276,
+      "step": 262525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015649168813014545,
+      "loss": 2.1036,
+      "step": 262530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001564901630068272,
+      "loss": 2.1481,
+      "step": 262535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015648863786421092,
+      "loss": 2.0852,
+      "step": 262540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001564871127022971,
+      "loss": 2.1005,
+      "step": 262545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015648558752108624,
+      "loss": 2.0783,
+      "step": 262550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015648406232057887,
+      "loss": 2.1676,
+      "step": 262555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015648253710077555,
+      "loss": 2.0944,
+      "step": 262560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015648101186167672,
+      "loss": 2.0089,
+      "step": 262565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000156479486603283,
+      "loss": 2.0866,
+      "step": 262570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015647796132559484,
+      "loss": 2.043,
+      "step": 262575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001564764360286128,
+      "loss": 2.2419,
+      "step": 262580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015647491071233736,
+      "loss": 2.2013,
+      "step": 262585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001564733853767691,
+      "loss": 2.0211,
+      "step": 262590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015647186002190847,
+      "loss": 2.1013,
+      "step": 262595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015647033464775606,
+      "loss": 2.2545,
+      "step": 262600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015646880925431237,
+      "loss": 1.983,
+      "step": 262605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001564672838415779,
+      "loss": 2.1284,
+      "step": 262610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015646575840955315,
+      "loss": 2.1482,
+      "step": 262615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015646423295823868,
+      "loss": 1.9932,
+      "step": 262620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015646270748763503,
+      "loss": 1.9484,
+      "step": 262625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001564611819977427,
+      "loss": 1.8914,
+      "step": 262630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015645965648856218,
+      "loss": 2.0211,
+      "step": 262635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015645813096009406,
+      "loss": 2.0685,
+      "step": 262640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015645660541233874,
+      "loss": 1.953,
+      "step": 262645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015645507984529688,
+      "loss": 2.1128,
+      "step": 262650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001564535542589689,
+      "loss": 2.2062,
+      "step": 262655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015645202865335542,
+      "loss": 2.1011,
+      "step": 262660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015645050302845687,
+      "loss": 2.1675,
+      "step": 262665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001564489773842738,
+      "loss": 2.1769,
+      "step": 262670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015644745172080672,
+      "loss": 2.042,
+      "step": 262675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015644592603805622,
+      "loss": 2.0749,
+      "step": 262680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001564444003360227,
+      "loss": 1.9643,
+      "step": 262685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015644287461470682,
+      "loss": 2.0939,
+      "step": 262690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000156441348874109,
+      "loss": 2.1276,
+      "step": 262695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015643982311422977,
+      "loss": 2.0794,
+      "step": 262700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015643829733506966,
+      "loss": 2.224,
+      "step": 262705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015643677153662927,
+      "loss": 1.8973,
+      "step": 262710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015643524571890902,
+      "loss": 2.0545,
+      "step": 262715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015643371988190947,
+      "loss": 1.9702,
+      "step": 262720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015643219402563112,
+      "loss": 1.9606,
+      "step": 262725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001564306681500745,
+      "loss": 2.2177,
+      "step": 262730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001564291422552402,
+      "loss": 2.1373,
+      "step": 262735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015642761634112863,
+      "loss": 2.1327,
+      "step": 262740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015642609040774036,
+      "loss": 2.1257,
+      "step": 262745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015642456445507597,
+      "loss": 1.8373,
+      "step": 262750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015642303848313586,
+      "loss": 2.0896,
+      "step": 262755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015642151249192066,
+      "loss": 1.9643,
+      "step": 262760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001564199864814308,
+      "loss": 2.1154,
+      "step": 262765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001564184604516669,
+      "loss": 2.049,
+      "step": 262770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015641693440262942,
+      "loss": 2.0689,
+      "step": 262775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015641540833431886,
+      "loss": 1.9986,
+      "step": 262780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001564138822467358,
+      "loss": 2.1484,
+      "step": 262785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015641235613988075,
+      "loss": 2.2054,
+      "step": 262790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001564108300137542,
+      "loss": 1.8852,
+      "step": 262795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015640930386835668,
+      "loss": 2.2288,
+      "step": 262800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001564077777036887,
+      "loss": 2.0739,
+      "step": 262805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015640625151975085,
+      "loss": 2.0607,
+      "step": 262810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015640472531654358,
+      "loss": 2.1535,
+      "step": 262815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015640319909406743,
+      "loss": 2.1379,
+      "step": 262820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001564016728523229,
+      "loss": 2.0589,
+      "step": 262825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015640014659131058,
+      "loss": 1.8876,
+      "step": 262830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001563986203110309,
+      "loss": 2.1349,
+      "step": 262835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015639709401148446,
+      "loss": 2.1312,
+      "step": 262840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015639556769267174,
+      "loss": 2.1507,
+      "step": 262845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015639404135459327,
+      "loss": 2.1611,
+      "step": 262850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015639251499724961,
+      "loss": 1.8749,
+      "step": 262855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001563909886206412,
+      "loss": 2.0533,
+      "step": 262860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015638946222476863,
+      "loss": 1.9435,
+      "step": 262865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001563879358096324,
+      "loss": 2.182,
+      "step": 262870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000156386409375233,
+      "loss": 2.0724,
+      "step": 262875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015638488292157098,
+      "loss": 2.2223,
+      "step": 262880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015638335644864687,
+      "loss": 1.9372,
+      "step": 262885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015638182995646117,
+      "loss": 2.2496,
+      "step": 262890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015638030344501444,
+      "loss": 1.9714,
+      "step": 262895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015637877691430718,
+      "loss": 1.9935,
+      "step": 262900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001563772503643399,
+      "loss": 2.0974,
+      "step": 262905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001563757237951131,
+      "loss": 2.1388,
+      "step": 262910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015637419720662736,
+      "loss": 2.0228,
+      "step": 262915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001563726705988832,
+      "loss": 1.9686,
+      "step": 262920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015637114397188106,
+      "loss": 2.0529,
+      "step": 262925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015636961732562154,
+      "loss": 2.17,
+      "step": 262930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015636809066010515,
+      "loss": 2.0519,
+      "step": 262935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015636656397533235,
+      "loss": 2.1789,
+      "step": 262940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015636503727130374,
+      "loss": 2.0731,
+      "step": 262945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015636351054801982,
+      "loss": 2.0191,
+      "step": 262950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015636198380548108,
+      "loss": 2.2861,
+      "step": 262955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015636045704368808,
+      "loss": 2.0795,
+      "step": 262960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015635893026264133,
+      "loss": 2.0828,
+      "step": 262965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015635740346234132,
+      "loss": 2.1504,
+      "step": 262970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001563558766427886,
+      "loss": 2.0506,
+      "step": 262975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015635434980398373,
+      "loss": 2.0952,
+      "step": 262980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015635282294592718,
+      "loss": 2.2958,
+      "step": 262985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015635129606861946,
+      "loss": 2.1823,
+      "step": 262990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015634976917206113,
+      "loss": 2.2475,
+      "step": 262995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015634824225625268,
+      "loss": 2.0236,
+      "step": 263000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001563467153211947,
+      "loss": 2.0107,
+      "step": 263005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001563451883668876,
+      "loss": 2.0649,
+      "step": 263010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.000156343661393332,
+      "loss": 1.8578,
+      "step": 263015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015634213440052837,
+      "loss": 2.1964,
+      "step": 263020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015634060738847723,
+      "loss": 2.1773,
+      "step": 263025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015633908035717916,
+      "loss": 2.0537,
+      "step": 263030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001563375533066346,
+      "loss": 2.0025,
+      "step": 263035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001563360262368441,
+      "loss": 2.2742,
+      "step": 263040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015633449914780823,
+      "loss": 1.9731,
+      "step": 263045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015633297203952745,
+      "loss": 1.9431,
+      "step": 263050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015633144491200232,
+      "loss": 2.2144,
+      "step": 263055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001563299177652333,
+      "loss": 1.9888,
+      "step": 263060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015632839059922103,
+      "loss": 2.015,
+      "step": 263065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015632686341396593,
+      "loss": 2.063,
+      "step": 263070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015632533620946851,
+      "loss": 1.9244,
+      "step": 263075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015632380898572939,
+      "loss": 2.2226,
+      "step": 263080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015632228174274903,
+      "loss": 2.2802,
+      "step": 263085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015632075448052793,
+      "loss": 2.2118,
+      "step": 263090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015631922719906667,
+      "loss": 2.2267,
+      "step": 263095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015631769989836572,
+      "loss": 2.0387,
+      "step": 263100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001563161725784256,
+      "loss": 2.1892,
+      "step": 263105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015631464523924691,
+      "loss": 2.2459,
+      "step": 263110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015631311788083006,
+      "loss": 1.9627,
+      "step": 263115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015631159050317564,
+      "loss": 2.0625,
+      "step": 263120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001563100631062842,
+      "loss": 2.0434,
+      "step": 263125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015630853569015618,
+      "loss": 2.2879,
+      "step": 263130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015630700825479217,
+      "loss": 2.0547,
+      "step": 263135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015630548080019266,
+      "loss": 2.138,
+      "step": 263140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015630395332635817,
+      "loss": 2.0112,
+      "step": 263145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001563024258332892,
+      "loss": 2.063,
+      "step": 263150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015630089832098634,
+      "loss": 2.0332,
+      "step": 263155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015629937078945005,
+      "loss": 1.9857,
+      "step": 263160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015629784323868088,
+      "loss": 2.0436,
+      "step": 263165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015629631566867937,
+      "loss": 2.263,
+      "step": 263170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015629478807944599,
+      "loss": 2.1491,
+      "step": 263175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001562932604709813,
+      "loss": 2.0648,
+      "step": 263180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001562917328432858,
+      "loss": 2.0112,
+      "step": 263185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015629020519636005,
+      "loss": 2.0658,
+      "step": 263190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001562886775302045,
+      "loss": 1.8803,
+      "step": 263195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015628714984481973,
+      "loss": 2.1375,
+      "step": 263200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001562856221402063,
+      "loss": 2.2529,
+      "step": 263205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001562840944163646,
+      "loss": 1.9751,
+      "step": 263210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001562825666732953,
+      "loss": 2.1772,
+      "step": 263215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001562810389109988,
+      "loss": 2.1031,
+      "step": 263220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001562795111294757,
+      "loss": 1.9741,
+      "step": 263225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015627798332872654,
+      "loss": 1.9449,
+      "step": 263230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015627645550875174,
+      "loss": 2.1345,
+      "step": 263235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001562749276695519,
+      "loss": 1.9947,
+      "step": 263240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015627339981112755,
+      "loss": 2.2084,
+      "step": 263245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015627187193347918,
+      "loss": 2.067,
+      "step": 263250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001562703440366073,
+      "loss": 2.0293,
+      "step": 263255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015626881612051244,
+      "loss": 2.002,
+      "step": 263260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015626728818519514,
+      "loss": 2.2324,
+      "step": 263265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001562657602306559,
+      "loss": 2.121,
+      "step": 263270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001562642322568953,
+      "loss": 2.1819,
+      "step": 263275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015626270426391382,
+      "loss": 2.2224,
+      "step": 263280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015626117625171192,
+      "loss": 2.0836,
+      "step": 263285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015625964822029023,
+      "loss": 2.2042,
+      "step": 263290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001562581201696492,
+      "loss": 2.0675,
+      "step": 263295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015625659209978938,
+      "loss": 2.0716,
+      "step": 263300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001562550640107113,
+      "loss": 2.0739,
+      "step": 263305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015625353590241546,
+      "loss": 2.0167,
+      "step": 263310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001562520077749024,
+      "loss": 1.9642,
+      "step": 263315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001562504796281726,
+      "loss": 2.0496,
+      "step": 263320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015624895146222666,
+      "loss": 1.9792,
+      "step": 263325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015624742327706505,
+      "loss": 2.2084,
+      "step": 263330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001562458950726883,
+      "loss": 2.1089,
+      "step": 263335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001562443668490969,
+      "loss": 2.122,
+      "step": 263340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015624283860629144,
+      "loss": 2.1587,
+      "step": 263345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001562413103442724,
+      "loss": 2.0549,
+      "step": 263350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001562397820630403,
+      "loss": 1.9069,
+      "step": 263355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015623825376259566,
+      "loss": 1.9024,
+      "step": 263360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015623672544293906,
+      "loss": 1.9718,
+      "step": 263365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015623519710407092,
+      "loss": 2.0116,
+      "step": 263370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015623366874599186,
+      "loss": 1.8823,
+      "step": 263375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015623214036870233,
+      "loss": 2.0208,
+      "step": 263380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015623061197220288,
+      "loss": 2.1661,
+      "step": 263385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015622908355649405,
+      "loss": 2.0429,
+      "step": 263390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015622755512157633,
+      "loss": 2.1821,
+      "step": 263395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015622602666745026,
+      "loss": 2.0241,
+      "step": 263400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015622449819411639,
+      "loss": 2.0489,
+      "step": 263405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015622296970157516,
+      "loss": 2.098,
+      "step": 263410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015622144118982717,
+      "loss": 2.1157,
+      "step": 263415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001562199126588729,
+      "loss": 2.0182,
+      "step": 263420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001562183841087129,
+      "loss": 2.0498,
+      "step": 263425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001562168555393477,
+      "loss": 2.0626,
+      "step": 263430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015621532695077778,
+      "loss": 1.9509,
+      "step": 263435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001562137983430037,
+      "loss": 1.8464,
+      "step": 263440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015621226971602596,
+      "loss": 2.0825,
+      "step": 263445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015621074106984505,
+      "loss": 2.1936,
+      "step": 263450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001562092124044616,
+      "loss": 2.1138,
+      "step": 263455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015620768371987605,
+      "loss": 2.0245,
+      "step": 263460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001562061550160889,
+      "loss": 2.0372,
+      "step": 263465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015620462629310072,
+      "loss": 2.0465,
+      "step": 263470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015620309755091203,
+      "loss": 2.1529,
+      "step": 263475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015620156878952334,
+      "loss": 2.3526,
+      "step": 263480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015620004000893519,
+      "loss": 2.0047,
+      "step": 263485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015619851120914806,
+      "loss": 2.0413,
+      "step": 263490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001561969823901625,
+      "loss": 2.122,
+      "step": 263495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015619545355197907,
+      "loss": 2.1244,
+      "step": 263500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001561939246945982,
+      "loss": 2.1151,
+      "step": 263505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001561923958180205,
+      "loss": 1.7343,
+      "step": 263510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015619086692224645,
+      "loss": 2.117,
+      "step": 263515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015618933800727656,
+      "loss": 2.0847,
+      "step": 263520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015618780907311145,
+      "loss": 2.1577,
+      "step": 263525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015618628011975148,
+      "loss": 2.0625,
+      "step": 263530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015618475114719728,
+      "loss": 2.0924,
+      "step": 263535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015618322215544937,
+      "loss": 2.0081,
+      "step": 263540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015618169314450824,
+      "loss": 2.2305,
+      "step": 263545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015618016411437445,
+      "loss": 2.2128,
+      "step": 263550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015617863506504844,
+      "loss": 2.0711,
+      "step": 263555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015617710599653083,
+      "loss": 2.117,
+      "step": 263560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015617557690882208,
+      "loss": 2.1647,
+      "step": 263565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015617404780192274,
+      "loss": 2.0611,
+      "step": 263570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015617251867583334,
+      "loss": 2.0154,
+      "step": 263575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001561709895305544,
+      "loss": 2.0515,
+      "step": 263580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001561694603660864,
+      "loss": 2.0225,
+      "step": 263585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001561679311824299,
+      "loss": 1.9671,
+      "step": 263590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015616640197958542,
+      "loss": 2.2592,
+      "step": 263595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001561648727575535,
+      "loss": 2.0697,
+      "step": 263600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015616334351633464,
+      "loss": 1.8839,
+      "step": 263605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015616181425592932,
+      "loss": 2.166,
+      "step": 263610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015616028497633812,
+      "loss": 2.2266,
+      "step": 263615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001561587556775616,
+      "loss": 1.941,
+      "step": 263620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015615722635960017,
+      "loss": 2.1769,
+      "step": 263625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015615569702245443,
+      "loss": 2.1212,
+      "step": 263630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001561541676661249,
+      "loss": 2.0813,
+      "step": 263635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015615263829061207,
+      "loss": 2.1364,
+      "step": 263640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001561511088959165,
+      "loss": 1.9874,
+      "step": 263645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015614957948203867,
+      "loss": 2.1089,
+      "step": 263650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015614805004897914,
+      "loss": 2.2117,
+      "step": 263655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015614652059673843,
+      "loss": 2.1694,
+      "step": 263660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.000156144991125317,
+      "loss": 1.8949,
+      "step": 263665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.28125,
+      "learning_rate": 0.00015614346163471545,
+      "loss": 2.125,
+      "step": 263670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001561419321249343,
+      "loss": 2.0694,
+      "step": 263675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015614040259597403,
+      "loss": 2.0338,
+      "step": 263680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015613887304783517,
+      "loss": 2.1063,
+      "step": 263685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015613734348051826,
+      "loss": 2.1188,
+      "step": 263690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015613581389402383,
+      "loss": 2.0919,
+      "step": 263695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015613428428835237,
+      "loss": 1.9866,
+      "step": 263700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015613275466350442,
+      "loss": 2.0035,
+      "step": 263705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015613122501948053,
+      "loss": 2.1053,
+      "step": 263710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015612969535628115,
+      "loss": 2.0717,
+      "step": 263715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015612816567390686,
+      "loss": 2.1514,
+      "step": 263720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015612663597235822,
+      "loss": 2.0697,
+      "step": 263725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015612510625163565,
+      "loss": 1.9833,
+      "step": 263730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015612357651173975,
+      "loss": 2.112,
+      "step": 263735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000156122046752671,
+      "loss": 2.1166,
+      "step": 263740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015612051697442994,
+      "loss": 2.2014,
+      "step": 263745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015611898717701714,
+      "loss": 2.0887,
+      "step": 263750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00015611745736043302,
+      "loss": 2.0627,
+      "step": 263755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015611592752467818,
+      "loss": 2.041,
+      "step": 263760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015611439766975314,
+      "loss": 2.0651,
+      "step": 263765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015611286779565838,
+      "loss": 2.0742,
+      "step": 263770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015611133790239444,
+      "loss": 2.1035,
+      "step": 263775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015610980798996186,
+      "loss": 1.9506,
+      "step": 263780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015610827805836116,
+      "loss": 2.2247,
+      "step": 263785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015610674810759284,
+      "loss": 1.9769,
+      "step": 263790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015610521813765746,
+      "loss": 2.0442,
+      "step": 263795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001561036881485555,
+      "loss": 2.0772,
+      "step": 263800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015610215814028752,
+      "loss": 2.1813,
+      "step": 263805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000156100628112854,
+      "loss": 2.0279,
+      "step": 263810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001560990980662555,
+      "loss": 1.763,
+      "step": 263815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015609756800049257,
+      "loss": 2.0438,
+      "step": 263820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015609603791556564,
+      "loss": 2.1451,
+      "step": 263825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001560945078114753,
+      "loss": 2.1995,
+      "step": 263830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015609297768822206,
+      "loss": 2.0328,
+      "step": 263835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015609144754580645,
+      "loss": 2.0797,
+      "step": 263840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015608991738422896,
+      "loss": 1.9799,
+      "step": 263845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015608838720349016,
+      "loss": 2.2218,
+      "step": 263850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015608685700359054,
+      "loss": 2.0819,
+      "step": 263855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00015608532678453062,
+      "loss": 2.0718,
+      "step": 263860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015608379654631098,
+      "loss": 1.964,
+      "step": 263865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015608226628893208,
+      "loss": 2.2388,
+      "step": 263870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015608073601239446,
+      "loss": 2.0505,
+      "step": 263875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001560792057166986,
+      "loss": 1.9972,
+      "step": 263880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001560776754018451,
+      "loss": 1.8091,
+      "step": 263885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015607614506783446,
+      "loss": 2.0433,
+      "step": 263890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015607461471466719,
+      "loss": 2.1671,
+      "step": 263895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001560730843423438,
+      "loss": 2.006,
+      "step": 263900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001560715539508648,
+      "loss": 2.0723,
+      "step": 263905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001560700235402308,
+      "loss": 1.9627,
+      "step": 263910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015606849311044222,
+      "loss": 2.0252,
+      "step": 263915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015606696266149967,
+      "loss": 1.9231,
+      "step": 263920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001560654321934036,
+      "loss": 2.1488,
+      "step": 263925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015606390170615453,
+      "loss": 2.0855,
+      "step": 263930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015606237119975303,
+      "loss": 2.0571,
+      "step": 263935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015606084067419965,
+      "loss": 2.1831,
+      "step": 263940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00015605931012949484,
+      "loss": 1.8992,
+      "step": 263945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015605777956563917,
+      "loss": 2.1627,
+      "step": 263950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015605624898263313,
+      "loss": 2.134,
+      "step": 263955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015605471838047724,
+      "loss": 2.0035,
+      "step": 263960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015605318775917208,
+      "loss": 2.0738,
+      "step": 263965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015605165711871812,
+      "loss": 1.9898,
+      "step": 263970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001560501264591159,
+      "loss": 2.05,
+      "step": 263975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015604859578036595,
+      "loss": 1.9577,
+      "step": 263980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015604706508246877,
+      "loss": 2.0585,
+      "step": 263985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015604553436542488,
+      "loss": 2.1609,
+      "step": 263990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001560440036292348,
+      "loss": 2.1794,
+      "step": 263995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015604247287389912,
+      "loss": 1.8571,
+      "step": 264000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001560409420994183,
+      "loss": 2.0956,
+      "step": 264005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001560394113057929,
+      "loss": 2.0326,
+      "step": 264010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015603788049302336,
+      "loss": 2.1147,
+      "step": 264015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015603634966111034,
+      "loss": 1.9399,
+      "step": 264020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015603481881005423,
+      "loss": 2.1011,
+      "step": 264025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015603328793985566,
+      "loss": 2.1054,
+      "step": 264030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015603175705051507,
+      "loss": 2.2591,
+      "step": 264035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000156030226142033,
+      "loss": 1.8599,
+      "step": 264040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015602869521441002,
+      "loss": 1.8729,
+      "step": 264045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001560271642676466,
+      "loss": 2.1033,
+      "step": 264050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015602563330174328,
+      "loss": 2.1751,
+      "step": 264055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015602410231670063,
+      "loss": 2.1342,
+      "step": 264060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015602257131251907,
+      "loss": 2.1388,
+      "step": 264065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015602104028919923,
+      "loss": 2.2686,
+      "step": 264070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015601950924674156,
+      "loss": 1.9607,
+      "step": 264075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015601797818514662,
+      "loss": 1.9078,
+      "step": 264080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001560164471044149,
+      "loss": 2.0217,
+      "step": 264085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015601491600454698,
+      "loss": 2.1459,
+      "step": 264090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015601338488554334,
+      "loss": 1.9933,
+      "step": 264095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001560118537474045,
+      "loss": 2.1078,
+      "step": 264100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015601032259013103,
+      "loss": 2.1787,
+      "step": 264105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001560087914137234,
+      "loss": 2.1399,
+      "step": 264110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015600726021818212,
+      "loss": 2.3149,
+      "step": 264115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015600572900350778,
+      "loss": 2.1728,
+      "step": 264120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015600419776970082,
+      "loss": 2.0969,
+      "step": 264125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015600266651676187,
+      "loss": 2.0855,
+      "step": 264130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015600113524469138,
+      "loss": 2.0837,
+      "step": 264135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015599960395348988,
+      "loss": 1.9615,
+      "step": 264140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001559980726431579,
+      "loss": 2.1898,
+      "step": 264145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015599654131369595,
+      "loss": 2.1227,
+      "step": 264150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015599500996510453,
+      "loss": 2.1725,
+      "step": 264155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015599347859738427,
+      "loss": 2.0606,
+      "step": 264160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001559919472105356,
+      "loss": 2.1491,
+      "step": 264165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015599041580455905,
+      "loss": 1.9339,
+      "step": 264170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015598888437945516,
+      "loss": 2.1492,
+      "step": 264175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015598735293522449,
+      "loss": 2.1656,
+      "step": 264180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015598582147186748,
+      "loss": 2.1862,
+      "step": 264185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015598428998938474,
+      "loss": 2.0868,
+      "step": 264190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001559827584877767,
+      "loss": 2.0725,
+      "step": 264195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015598122696704395,
+      "loss": 2.136,
+      "step": 264200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000155979695427187,
+      "loss": 2.1973,
+      "step": 264205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001559781638682064,
+      "loss": 1.9327,
+      "step": 264210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015597663229010263,
+      "loss": 2.1047,
+      "step": 264215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001559751006928762,
+      "loss": 1.9255,
+      "step": 264220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001559735690765277,
+      "loss": 2.1157,
+      "step": 264225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015597203744105757,
+      "loss": 2.0021,
+      "step": 264230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015597050578646638,
+      "loss": 2.2917,
+      "step": 264235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015596897411275471,
+      "loss": 1.9087,
+      "step": 264240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015596744241992298,
+      "loss": 2.0493,
+      "step": 264245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015596591070797175,
+      "loss": 2.0641,
+      "step": 264250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015596437897690154,
+      "loss": 1.8188,
+      "step": 264255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001559628472267129,
+      "loss": 2.1606,
+      "step": 264260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015596131545740635,
+      "loss": 2.0713,
+      "step": 264265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015595978366898238,
+      "loss": 2.1308,
+      "step": 264270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015595825186144157,
+      "loss": 2.0576,
+      "step": 264275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015595672003478436,
+      "loss": 1.8956,
+      "step": 264280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015595518818901132,
+      "loss": 2.1327,
+      "step": 264285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015595365632412302,
+      "loss": 2.0788,
+      "step": 264290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001559521244401199,
+      "loss": 1.8934,
+      "step": 264295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015595059253700253,
+      "loss": 2.1101,
+      "step": 264300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001559490606147714,
+      "loss": 2.1357,
+      "step": 264305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015594752867342711,
+      "loss": 2.0702,
+      "step": 264310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015594599671297008,
+      "loss": 2.0524,
+      "step": 264315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001559444647334009,
+      "loss": 2.0868,
+      "step": 264320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015594293273472006,
+      "loss": 2.062,
+      "step": 264325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015594140071692814,
+      "loss": 2.1693,
+      "step": 264330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015593986868002558,
+      "loss": 2.0269,
+      "step": 264335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015593833662401298,
+      "loss": 1.9559,
+      "step": 264340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001559368045488908,
+      "loss": 2.1963,
+      "step": 264345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001559352724546596,
+      "loss": 1.9614,
+      "step": 264350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001559337403413199,
+      "loss": 1.9087,
+      "step": 264355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015593220820887222,
+      "loss": 2.1743,
+      "step": 264360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001559306760573171,
+      "loss": 2.1381,
+      "step": 264365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015592914388665502,
+      "loss": 1.9608,
+      "step": 264370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015592761169688654,
+      "loss": 1.9764,
+      "step": 264375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001559260794880122,
+      "loss": 2.0257,
+      "step": 264380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015592454726003245,
+      "loss": 2.1569,
+      "step": 264385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001559230150129479,
+      "loss": 2.085,
+      "step": 264390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.000155921482746759,
+      "loss": 2.2014,
+      "step": 264395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015591995046146634,
+      "loss": 1.9083,
+      "step": 264400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001559184181570704,
+      "loss": 1.9662,
+      "step": 264405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001559168858335717,
+      "loss": 1.935,
+      "step": 264410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001559153534909708,
+      "loss": 1.9545,
+      "step": 264415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001559138211292682,
+      "loss": 2.0378,
+      "step": 264420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001559122887484644,
+      "loss": 1.9834,
+      "step": 264425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015591075634855997,
+      "loss": 2.0433,
+      "step": 264430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015590922392955537,
+      "loss": 2.0586,
+      "step": 264435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001559076914914512,
+      "loss": 2.0681,
+      "step": 264440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015590615903424796,
+      "loss": 2.056,
+      "step": 264445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015590462655794615,
+      "loss": 2.0775,
+      "step": 264450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015590309406254631,
+      "loss": 2.1542,
+      "step": 264455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015590156154804893,
+      "loss": 2.0519,
+      "step": 264460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001559000290144546,
+      "loss": 2.1407,
+      "step": 264465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001558984964617638,
+      "loss": 1.9441,
+      "step": 264470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015589696388997704,
+      "loss": 2.0414,
+      "step": 264475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015589543129909488,
+      "loss": 2.0299,
+      "step": 264480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001558938986891178,
+      "loss": 2.0483,
+      "step": 264485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001558923660600464,
+      "loss": 1.9684,
+      "step": 264490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015589083341188112,
+      "loss": 2.1409,
+      "step": 264495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015588930074462252,
+      "loss": 2.2024,
+      "step": 264500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015588776805827115,
+      "loss": 2.14,
+      "step": 264505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015588623535282747,
+      "loss": 2.13,
+      "step": 264510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015588470262829204,
+      "loss": 1.9393,
+      "step": 264515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015588316988466542,
+      "loss": 2.1271,
+      "step": 264520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015588163712194802,
+      "loss": 2.0592,
+      "step": 264525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001558801043401405,
+      "loss": 2.0178,
+      "step": 264530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015587857153924331,
+      "loss": 1.9647,
+      "step": 264535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000155877038719257,
+      "loss": 1.9506,
+      "step": 264540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015587550588018204,
+      "loss": 2.2425,
+      "step": 264545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015587397302201903,
+      "loss": 2.1592,
+      "step": 264550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015587244014476847,
+      "loss": 1.9201,
+      "step": 264555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015587090724843082,
+      "loss": 2.0469,
+      "step": 264560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001558693743330067,
+      "loss": 2.1225,
+      "step": 264565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001558678413984966,
+      "loss": 2.1687,
+      "step": 264570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015586630844490097,
+      "loss": 2.1574,
+      "step": 264575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015586477547222045,
+      "loss": 2.0949,
+      "step": 264580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015586324248045548,
+      "loss": 2.029,
+      "step": 264585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015586170946960663,
+      "loss": 2.1313,
+      "step": 264590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001558601764396744,
+      "loss": 2.0946,
+      "step": 264595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001558586433906593,
+      "loss": 2.2181,
+      "step": 264600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001558571103225619,
+      "loss": 2.0914,
+      "step": 264605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015585557723538272,
+      "loss": 2.1195,
+      "step": 264610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001558540441291222,
+      "loss": 1.9334,
+      "step": 264615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015585251100378098,
+      "loss": 2.0992,
+      "step": 264620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001558509778593595,
+      "loss": 2.1893,
+      "step": 264625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015584944469585832,
+      "loss": 1.9804,
+      "step": 264630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015584791151327794,
+      "loss": 1.8628,
+      "step": 264635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015584637831161894,
+      "loss": 2.217,
+      "step": 264640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001558448450908818,
+      "loss": 1.9793,
+      "step": 264645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015584331185106702,
+      "loss": 1.9933,
+      "step": 264650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015584177859217518,
+      "loss": 2.149,
+      "step": 264655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015584024531420675,
+      "loss": 2.0461,
+      "step": 264660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001558387120171623,
+      "loss": 2.0869,
+      "step": 264665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015583717870104233,
+      "loss": 2.0953,
+      "step": 264670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015583564536584733,
+      "loss": 2.0105,
+      "step": 264675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001558341120115779,
+      "loss": 2.0785,
+      "step": 264680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001558325786382345,
+      "loss": 2.0731,
+      "step": 264685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015583104524581772,
+      "loss": 2.1009,
+      "step": 264690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000155829511834328,
+      "loss": 2.2113,
+      "step": 264695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015582797840376595,
+      "loss": 2.1659,
+      "step": 264700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015582644495413202,
+      "loss": 1.9477,
+      "step": 264705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015582491148542676,
+      "loss": 2.0943,
+      "step": 264710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001558233779976507,
+      "loss": 2.1479,
+      "step": 264715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015582184449080435,
+      "loss": 2.0295,
+      "step": 264720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015582031096488826,
+      "loss": 2.0165,
+      "step": 264725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015581877741990295,
+      "loss": 2.1149,
+      "step": 264730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015581724385584891,
+      "loss": 2.1845,
+      "step": 264735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001558157102727267,
+      "loss": 2.1835,
+      "step": 264740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015581417667053687,
+      "loss": 2.1971,
+      "step": 264745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015581264304927986,
+      "loss": 2.0638,
+      "step": 264750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00015581110940895624,
+      "loss": 2.1819,
+      "step": 264755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015580957574956654,
+      "loss": 2.0455,
+      "step": 264760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015580804207111125,
+      "loss": 1.9996,
+      "step": 264765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015580650837359098,
+      "loss": 1.8842,
+      "step": 264770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015580497465700615,
+      "loss": 1.8497,
+      "step": 264775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015580344092135735,
+      "loss": 2.0635,
+      "step": 264780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001558019071666451,
+      "loss": 2.1792,
+      "step": 264785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015580037339286987,
+      "loss": 2.0414,
+      "step": 264790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001557988396000322,
+      "loss": 1.9357,
+      "step": 264795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015579730578813266,
+      "loss": 2.0411,
+      "step": 264800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001557957719571718,
+      "loss": 1.9731,
+      "step": 264805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015579423810715002,
+      "loss": 1.9543,
+      "step": 264810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015579270423806793,
+      "loss": 2.0303,
+      "step": 264815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015579117034992605,
+      "loss": 1.98,
+      "step": 264820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015578963644272492,
+      "loss": 2.0511,
+      "step": 264825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000155788102516465,
+      "loss": 2.1549,
+      "step": 264830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015578656857114686,
+      "loss": 2.0622,
+      "step": 264835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015578503460677104,
+      "loss": 2.2361,
+      "step": 264840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000155783500623338,
+      "loss": 2.1671,
+      "step": 264845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015578196662084835,
+      "loss": 2.2,
+      "step": 264850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015578043259930252,
+      "loss": 2.1241,
+      "step": 264855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001557788985587011,
+      "loss": 2.0722,
+      "step": 264860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015577736449904462,
+      "loss": 2.2418,
+      "step": 264865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015577583042033358,
+      "loss": 2.0492,
+      "step": 264870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015577429632256847,
+      "loss": 2.0034,
+      "step": 264875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001557727622057499,
+      "loss": 2.1407,
+      "step": 264880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001557712280698783,
+      "loss": 1.9958,
+      "step": 264885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015576969391495427,
+      "loss": 2.0968,
+      "step": 264890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015576815974097826,
+      "loss": 2.096,
+      "step": 264895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015576662554795084,
+      "loss": 2.1662,
+      "step": 264900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015576509133587258,
+      "loss": 1.9178,
+      "step": 264905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001557635571047439,
+      "loss": 2.0429,
+      "step": 264910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001557620228545654,
+      "loss": 2.0648,
+      "step": 264915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00015576048858533757,
+      "loss": 1.999,
+      "step": 264920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015575895429706095,
+      "loss": 2.1049,
+      "step": 264925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015575741998973607,
+      "loss": 1.8919,
+      "step": 264930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015575588566336346,
+      "loss": 2.0501,
+      "step": 264935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001557543513179436,
+      "loss": 2.05,
+      "step": 264940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015575281695347706,
+      "loss": 1.9816,
+      "step": 264945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015575128256996431,
+      "loss": 1.9569,
+      "step": 264950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015574974816740596,
+      "loss": 2.163,
+      "step": 264955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015574821374580243,
+      "loss": 2.0403,
+      "step": 264960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015574667930515435,
+      "loss": 2.1348,
+      "step": 264965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001557451448454622,
+      "loss": 2.1756,
+      "step": 264970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015574361036672644,
+      "loss": 2.1802,
+      "step": 264975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015574207586894768,
+      "loss": 2.1133,
+      "step": 264980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015574054135212643,
+      "loss": 2.0204,
+      "step": 264985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001557390068162632,
+      "loss": 2.0524,
+      "step": 264990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001557374722613585,
+      "loss": 2.182,
+      "step": 264995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015573593768741288,
+      "loss": 2.1274,
+      "step": 265000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015573440309442683,
+      "loss": 2.0228,
+      "step": 265005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001557328684824009,
+      "loss": 1.9619,
+      "step": 265010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015573133385133562,
+      "loss": 2.0758,
+      "step": 265015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015572979920123152,
+      "loss": 2.1539,
+      "step": 265020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015572826453208913,
+      "loss": 1.935,
+      "step": 265025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001557267298439089,
+      "loss": 2.1184,
+      "step": 265030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015572519513669142,
+      "loss": 2.0571,
+      "step": 265035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015572366041043722,
+      "loss": 2.201,
+      "step": 265040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001557221256651468,
+      "loss": 2.078,
+      "step": 265045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001557205909008207,
+      "loss": 2.0988,
+      "step": 265050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015571905611745944,
+      "loss": 2.0826,
+      "step": 265055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001557175213150635,
+      "loss": 2.09,
+      "step": 265060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001557159864936335,
+      "loss": 2.186,
+      "step": 265065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015571445165316988,
+      "loss": 2.1448,
+      "step": 265070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001557129167936732,
+      "loss": 2.1028,
+      "step": 265075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000155711381915144,
+      "loss": 2.0363,
+      "step": 265080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015570984701758272,
+      "loss": 2.0098,
+      "step": 265085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015570831210098997,
+      "loss": 2.0014,
+      "step": 265090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015570677716536627,
+      "loss": 2.0967,
+      "step": 265095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015570524221071209,
+      "loss": 2.0632,
+      "step": 265100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015570370723702804,
+      "loss": 2.0994,
+      "step": 265105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015570217224431456,
+      "loss": 1.9753,
+      "step": 265110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001557006372325722,
+      "loss": 1.9549,
+      "step": 265115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001556991022018015,
+      "loss": 2.1442,
+      "step": 265120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015569756715200298,
+      "loss": 1.8038,
+      "step": 265125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.0001556960320831772,
+      "loss": 2.2634,
+      "step": 265130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001556944969953246,
+      "loss": 2.0907,
+      "step": 265135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015569296188844574,
+      "loss": 2.1612,
+      "step": 265140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015569142676254117,
+      "loss": 1.9637,
+      "step": 265145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001556898916176114,
+      "loss": 1.9704,
+      "step": 265150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015568835645365695,
+      "loss": 2.1281,
+      "step": 265155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015568682127067838,
+      "loss": 1.9517,
+      "step": 265160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015568528606867615,
+      "loss": 2.1085,
+      "step": 265165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001556837508476508,
+      "loss": 2.0307,
+      "step": 265170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001556822156076029,
+      "loss": 1.7668,
+      "step": 265175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015568068034853294,
+      "loss": 2.3329,
+      "step": 265180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015567914507044143,
+      "loss": 1.9439,
+      "step": 265185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015567760977332892,
+      "loss": 1.9021,
+      "step": 265190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015567607445719593,
+      "loss": 2.3384,
+      "step": 265195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015567453912204303,
+      "loss": 2.039,
+      "step": 265200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015567300376787067,
+      "loss": 2.0828,
+      "step": 265205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001556714683946794,
+      "loss": 2.2207,
+      "step": 265210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001556699330024697,
+      "loss": 1.9099,
+      "step": 265215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015566839759124218,
+      "loss": 1.9927,
+      "step": 265220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015566686216099735,
+      "loss": 2.162,
+      "step": 265225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001556653267117357,
+      "loss": 2.2625,
+      "step": 265230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015566379124345775,
+      "loss": 2.0514,
+      "step": 265235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015566225575616404,
+      "loss": 2.0695,
+      "step": 265240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015566072024985508,
+      "loss": 2.1716,
+      "step": 265245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001556591847245314,
+      "loss": 1.9639,
+      "step": 265250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015565764918019356,
+      "loss": 2.1668,
+      "step": 265255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015565611361684206,
+      "loss": 2.136,
+      "step": 265260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015565457803447743,
+      "loss": 1.9208,
+      "step": 265265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015565304243310015,
+      "loss": 1.9789,
+      "step": 265270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001556515068127108,
+      "loss": 2.0775,
+      "step": 265275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015564997117330993,
+      "loss": 2.097,
+      "step": 265280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015564843551489797,
+      "loss": 2.2264,
+      "step": 265285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001556468998374755,
+      "loss": 1.9557,
+      "step": 265290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015564536414104305,
+      "loss": 1.9064,
+      "step": 265295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015564382842560114,
+      "loss": 2.0747,
+      "step": 265300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015564229269115026,
+      "loss": 2.0722,
+      "step": 265305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000155640756937691,
+      "loss": 2.2764,
+      "step": 265310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015563922116522384,
+      "loss": 2.3305,
+      "step": 265315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001556376853737493,
+      "loss": 2.1887,
+      "step": 265320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001556361495632679,
+      "loss": 2.0658,
+      "step": 265325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00015563461373378022,
+      "loss": 1.8626,
+      "step": 265330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015563307788528673,
+      "loss": 2.0792,
+      "step": 265335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015563154201778798,
+      "loss": 2.1327,
+      "step": 265340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015563000613128444,
+      "loss": 2.0665,
+      "step": 265345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015562847022577674,
+      "loss": 2.1002,
+      "step": 265350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001556269343012653,
+      "loss": 2.1264,
+      "step": 265355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015562539835775073,
+      "loss": 2.0571,
+      "step": 265360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001556238623952335,
+      "loss": 2.2074,
+      "step": 265365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015562232641371418,
+      "loss": 1.8848,
+      "step": 265370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001556207904131932,
+      "loss": 2.0728,
+      "step": 265375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001556192543936712,
+      "loss": 2.0664,
+      "step": 265380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015561771835514865,
+      "loss": 2.038,
+      "step": 265385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015561618229762605,
+      "loss": 2.0699,
+      "step": 265390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000155614646221104,
+      "loss": 2.1134,
+      "step": 265395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015561311012558292,
+      "loss": 2.4508,
+      "step": 265400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015561157401106342,
+      "loss": 1.9886,
+      "step": 265405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000155610037877546,
+      "loss": 1.8899,
+      "step": 265410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015560850172503118,
+      "loss": 2.0901,
+      "step": 265415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001556069655535195,
+      "loss": 1.9111,
+      "step": 265420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00015560542936301147,
+      "loss": 1.9541,
+      "step": 265425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001556038931535076,
+      "loss": 2.0025,
+      "step": 265430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015560235692500842,
+      "loss": 2.0432,
+      "step": 265435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001556008206775145,
+      "loss": 2.1729,
+      "step": 265440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015559928441102633,
+      "loss": 2.0394,
+      "step": 265445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015559774812554442,
+      "loss": 2.1465,
+      "step": 265450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015559621182106934,
+      "loss": 2.1437,
+      "step": 265455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015559467549760154,
+      "loss": 2.0798,
+      "step": 265460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001555931391551416,
+      "loss": 2.1589,
+      "step": 265465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015559160279369005,
+      "loss": 1.9671,
+      "step": 265470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001555900664132474,
+      "loss": 2.0544,
+      "step": 265475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015558853001381418,
+      "loss": 2.164,
+      "step": 265480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001555869935953909,
+      "loss": 2.2867,
+      "step": 265485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001555854571579781,
+      "loss": 2.1359,
+      "step": 265490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001555839207015763,
+      "loss": 2.1249,
+      "step": 265495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000155582384226186,
+      "loss": 2.1105,
+      "step": 265500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001555808477318078,
+      "loss": 2.2342,
+      "step": 265505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015557931121844214,
+      "loss": 1.9963,
+      "step": 265510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015557777468608956,
+      "loss": 2.2407,
+      "step": 265515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015557623813475065,
+      "loss": 2.1123,
+      "step": 265520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015557470156442587,
+      "loss": 2.0987,
+      "step": 265525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015557316497511576,
+      "loss": 2.0387,
+      "step": 265530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015557162836682083,
+      "loss": 2.0567,
+      "step": 265535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015557009173954164,
+      "loss": 2.1424,
+      "step": 265540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015556855509327871,
+      "loss": 2.0384,
+      "step": 265545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015556701842803255,
+      "loss": 2.0286,
+      "step": 265550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001555654817438037,
+      "loss": 2.1088,
+      "step": 265555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015556394504059266,
+      "loss": 2.0878,
+      "step": 265560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015556240831839996,
+      "loss": 1.9062,
+      "step": 265565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00015556087157722614,
+      "loss": 2.0504,
+      "step": 265570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001555593348170717,
+      "loss": 2.0732,
+      "step": 265575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015555779803793723,
+      "loss": 2.0122,
+      "step": 265580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015555626123982318,
+      "loss": 2.1567,
+      "step": 265585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001555547244227301,
+      "loss": 1.9127,
+      "step": 265590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015555318758665853,
+      "loss": 1.973,
+      "step": 265595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000155551650731609,
+      "loss": 2.1141,
+      "step": 265600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015555011385758199,
+      "loss": 2.0236,
+      "step": 265605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015554857696457807,
+      "loss": 1.9788,
+      "step": 265610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015554704005259776,
+      "loss": 2.0945,
+      "step": 265615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001555455031216415,
+      "loss": 1.9495,
+      "step": 265620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015554396617170998,
+      "loss": 2.1902,
+      "step": 265625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015554242920280358,
+      "loss": 2.0252,
+      "step": 265630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001555408922149229,
+      "loss": 2.0786,
+      "step": 265635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015553935520806845,
+      "loss": 2.1914,
+      "step": 265640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001555378181822407,
+      "loss": 2.0962,
+      "step": 265645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015553628113744027,
+      "loss": 1.9242,
+      "step": 265650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015553474407366765,
+      "loss": 2.0303,
+      "step": 265655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015553320699092333,
+      "loss": 2.1178,
+      "step": 265660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015553166988920787,
+      "loss": 2.0247,
+      "step": 265665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015553013276852179,
+      "loss": 2.0584,
+      "step": 265670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015552859562886558,
+      "loss": 2.0294,
+      "step": 265675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015552705847023983,
+      "loss": 2.1609,
+      "step": 265680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015552552129264498,
+      "loss": 2.1416,
+      "step": 265685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015552398409608166,
+      "loss": 2.021,
+      "step": 265690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015552244688055034,
+      "loss": 1.9335,
+      "step": 265695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001555209096460515,
+      "loss": 1.8568,
+      "step": 265700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015551937239258572,
+      "loss": 2.2151,
+      "step": 265705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015551783512015354,
+      "loss": 2.1268,
+      "step": 265710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015551629782875547,
+      "loss": 2.0022,
+      "step": 265715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015551476051839198,
+      "loss": 2.0576,
+      "step": 265720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015551322318906367,
+      "loss": 2.1611,
+      "step": 265725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015551168584077103,
+      "loss": 1.9703,
+      "step": 265730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001555101484735146,
+      "loss": 1.9715,
+      "step": 265735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001555086110872949,
+      "loss": 2.1111,
+      "step": 265740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015550707368211243,
+      "loss": 2.0612,
+      "step": 265745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015550553625796778,
+      "loss": 1.9891,
+      "step": 265750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015550399881486136,
+      "loss": 2.1332,
+      "step": 265755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015550246135279382,
+      "loss": 2.1905,
+      "step": 265760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015550092387176561,
+      "loss": 2.0813,
+      "step": 265765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001554993863717773,
+      "loss": 1.9323,
+      "step": 265770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001554978488528294,
+      "loss": 2.0392,
+      "step": 265775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015549631131492238,
+      "loss": 2.0919,
+      "step": 265780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015549477375805687,
+      "loss": 2.0694,
+      "step": 265785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001554932361822333,
+      "loss": 1.9532,
+      "step": 265790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015549169858745224,
+      "loss": 2.1246,
+      "step": 265795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015549016097371422,
+      "loss": 2.0056,
+      "step": 265800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015548862334101976,
+      "loss": 2.254,
+      "step": 265805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015548708568936936,
+      "loss": 2.036,
+      "step": 265810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015548554801876358,
+      "loss": 2.0784,
+      "step": 265815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015548401032920294,
+      "loss": 2.2649,
+      "step": 265820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015548247262068795,
+      "loss": 2.0898,
+      "step": 265825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015548093489321912,
+      "loss": 2.1983,
+      "step": 265830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015547939714679704,
+      "loss": 2.2221,
+      "step": 265835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015547785938142215,
+      "loss": 2.0495,
+      "step": 265840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00015547632159709502,
+      "loss": 1.9844,
+      "step": 265845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015547478379381618,
+      "loss": 2.1751,
+      "step": 265850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015547324597158614,
+      "loss": 2.0219,
+      "step": 265855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015547170813040543,
+      "loss": 2.1154,
+      "step": 265860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001554701702702746,
+      "loss": 1.8924,
+      "step": 265865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015546863239119415,
+      "loss": 2.1128,
+      "step": 265870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001554670944931646,
+      "loss": 2.0886,
+      "step": 265875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015546555657618648,
+      "loss": 2.008,
+      "step": 265880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015546401864026036,
+      "loss": 1.8569,
+      "step": 265885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001554624806853867,
+      "loss": 2.0624,
+      "step": 265890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015546094271156604,
+      "loss": 2.0383,
+      "step": 265895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015545940471879888,
+      "loss": 2.2237,
+      "step": 265900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015545786670708586,
+      "loss": 1.9059,
+      "step": 265905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015545632867642738,
+      "loss": 2.0101,
+      "step": 265910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000155454790626824,
+      "loss": 2.2454,
+      "step": 265915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001554532525582763,
+      "loss": 2.0305,
+      "step": 265920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015545171447078475,
+      "loss": 2.0596,
+      "step": 265925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001554501763643499,
+      "loss": 1.9099,
+      "step": 265930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015544863823897223,
+      "loss": 2.0855,
+      "step": 265935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001554471000946523,
+      "loss": 1.9791,
+      "step": 265940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015544556193139065,
+      "loss": 2.0341,
+      "step": 265945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001554440237491878,
+      "loss": 2.2762,
+      "step": 265950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015544248554804426,
+      "loss": 2.024,
+      "step": 265955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015544094732796056,
+      "loss": 2.2012,
+      "step": 265960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001554394090889372,
+      "loss": 2.1325,
+      "step": 265965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00015543787083097476,
+      "loss": 2.018,
+      "step": 265970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015543633255407375,
+      "loss": 2.134,
+      "step": 265975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015543479425823468,
+      "loss": 1.93,
+      "step": 265980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015543325594345806,
+      "loss": 1.9809,
+      "step": 265985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015543171760974445,
+      "loss": 2.1443,
+      "step": 265990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015543017925709432,
+      "loss": 2.017,
+      "step": 265995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001554286408855083,
+      "loss": 1.9746,
+      "step": 266000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001554271024949868,
+      "loss": 2.1543,
+      "step": 266005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015542556408553045,
+      "loss": 2.2171,
+      "step": 266010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015542402565713967,
+      "loss": 2.2225,
+      "step": 266015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015542248720981503,
+      "loss": 2.0703,
+      "step": 266020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015542094874355708,
+      "loss": 1.9816,
+      "step": 266025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015541941025836638,
+      "loss": 2.2279,
+      "step": 266030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015541787175424334,
+      "loss": 1.9519,
+      "step": 266035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015541633323118856,
+      "loss": 2.0842,
+      "step": 266040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015541479468920256,
+      "loss": 2.2003,
+      "step": 266045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001554132561282859,
+      "loss": 2.1484,
+      "step": 266050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000155411717548439,
+      "loss": 2.284,
+      "step": 266055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015541017894966252,
+      "loss": 2.0894,
+      "step": 266060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001554086403319569,
+      "loss": 2.073,
+      "step": 266065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015540710169532264,
+      "loss": 1.9074,
+      "step": 266070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00015540556303976036,
+      "loss": 2.1652,
+      "step": 266075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015540402436527048,
+      "loss": 2.1546,
+      "step": 266080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015540248567185363,
+      "loss": 2.0468,
+      "step": 266085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015540094695951027,
+      "loss": 2.1181,
+      "step": 266090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015539940822824092,
+      "loss": 2.0476,
+      "step": 266095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015539786947804616,
+      "loss": 2.1932,
+      "step": 266100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015539633070892647,
+      "loss": 2.0952,
+      "step": 266105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001553947919208824,
+      "loss": 1.8547,
+      "step": 266110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015539325311391444,
+      "loss": 2.0417,
+      "step": 266115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015539171428802318,
+      "loss": 2.0452,
+      "step": 266120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015539017544320907,
+      "loss": 2.0986,
+      "step": 266125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015538863657947268,
+      "loss": 2.2103,
+      "step": 266130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015538709769681453,
+      "loss": 1.9946,
+      "step": 266135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015538555879523512,
+      "loss": 2.0151,
+      "step": 266140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015538401987473502,
+      "loss": 2.0674,
+      "step": 266145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015538248093531476,
+      "loss": 2.101,
+      "step": 266150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015538094197697482,
+      "loss": 2.0668,
+      "step": 266155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015537940299971572,
+      "loss": 2.0881,
+      "step": 266160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015537786400353804,
+      "loss": 2.1787,
+      "step": 266165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015537632498844228,
+      "loss": 2.1062,
+      "step": 266170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015537478595442896,
+      "loss": 2.0485,
+      "step": 266175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015537324690149856,
+      "loss": 2.0668,
+      "step": 266180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001553717078296517,
+      "loss": 2.1262,
+      "step": 266185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015537016873888888,
+      "loss": 2.0195,
+      "step": 266190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001553686296292106,
+      "loss": 2.0503,
+      "step": 266195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015536709050061737,
+      "loss": 2.0202,
+      "step": 266200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015536555135310973,
+      "loss": 2.1196,
+      "step": 266205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015536401218668822,
+      "loss": 2.0965,
+      "step": 266210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015536247300135337,
+      "loss": 2.1177,
+      "step": 266215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015536093379710567,
+      "loss": 2.106,
+      "step": 266220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015535939457394572,
+      "loss": 2.2068,
+      "step": 266225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015535785533187397,
+      "loss": 2.1099,
+      "step": 266230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015535631607089097,
+      "loss": 2.0373,
+      "step": 266235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015535477679099727,
+      "loss": 2.1818,
+      "step": 266240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015535323749219334,
+      "loss": 2.0751,
+      "step": 266245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015535169817447976,
+      "loss": 2.0757,
+      "step": 266250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015535015883785702,
+      "loss": 2.1544,
+      "step": 266255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015534861948232566,
+      "loss": 2.0359,
+      "step": 266260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015534708010788624,
+      "loss": 1.9575,
+      "step": 266265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001553455407145392,
+      "loss": 2.0773,
+      "step": 266270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015534400130228518,
+      "loss": 2.024,
+      "step": 266275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015534246187112462,
+      "loss": 2.1759,
+      "step": 266280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015534092242105806,
+      "loss": 2.0517,
+      "step": 266285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015533938295208606,
+      "loss": 2.1092,
+      "step": 266290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001553378434642091,
+      "loss": 2.1611,
+      "step": 266295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015533630395742773,
+      "loss": 2.2351,
+      "step": 266300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015533476443174252,
+      "loss": 1.9185,
+      "step": 266305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001553332248871539,
+      "loss": 2.094,
+      "step": 266310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015533168532366244,
+      "loss": 2.0045,
+      "step": 266315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001553301457412687,
+      "loss": 2.0756,
+      "step": 266320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001553286061399732,
+      "loss": 1.9931,
+      "step": 266325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001553270665197764,
+      "loss": 2.0062,
+      "step": 266330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.0001553255268806789,
+      "loss": 2.0837,
+      "step": 266335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015532398722268117,
+      "loss": 1.9706,
+      "step": 266340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015532244754578377,
+      "loss": 2.1231,
+      "step": 266345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015532090784998724,
+      "loss": 2.1974,
+      "step": 266350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015531936813529206,
+      "loss": 2.0444,
+      "step": 266355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001553178284016988,
+      "loss": 2.1774,
+      "step": 266360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015531628864920799,
+      "loss": 2.0587,
+      "step": 266365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015531474887782006,
+      "loss": 2.1042,
+      "step": 266370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015531320908753566,
+      "loss": 1.9306,
+      "step": 266375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015531166927835525,
+      "loss": 2.0854,
+      "step": 266380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001553101294502794,
+      "loss": 2.1405,
+      "step": 266385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015530858960330858,
+      "loss": 1.9991,
+      "step": 266390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001553070497374433,
+      "loss": 1.9564,
+      "step": 266395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001553055098526842,
+      "loss": 2.0775,
+      "step": 266400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001553039699490317,
+      "loss": 2.124,
+      "step": 266405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015530243002648638,
+      "loss": 2.0513,
+      "step": 266410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015530089008504872,
+      "loss": 2.0971,
+      "step": 266415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015529935012471933,
+      "loss": 2.0337,
+      "step": 266420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015529781014549863,
+      "loss": 2.0882,
+      "step": 266425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015529627014738718,
+      "loss": 2.153,
+      "step": 266430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015529473013038556,
+      "loss": 1.955,
+      "step": 266435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015529319009449424,
+      "loss": 2.1454,
+      "step": 266440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.0001552916500397138,
+      "loss": 2.2357,
+      "step": 266445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015529010996604464,
+      "loss": 2.0129,
+      "step": 266450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015528856987348744,
+      "loss": 2.2164,
+      "step": 266455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00015528702976204265,
+      "loss": 2.1114,
+      "step": 266460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015528548963171082,
+      "loss": 2.0279,
+      "step": 266465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015528394948249248,
+      "loss": 2.0146,
+      "step": 266470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001552824093143881,
+      "loss": 2.1289,
+      "step": 266475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015528086912739824,
+      "loss": 2.1142,
+      "step": 266480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015527932892152348,
+      "loss": 2.0794,
+      "step": 266485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015527778869676425,
+      "loss": 1.9861,
+      "step": 266490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015527624845312118,
+      "loss": 2.1371,
+      "step": 266495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001552747081905947,
+      "loss": 2.1231,
+      "step": 266500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015527316790918535,
+      "loss": 2.1431,
+      "step": 266505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015527162760889374,
+      "loss": 2.045,
+      "step": 266510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015527008728972032,
+      "loss": 2.1693,
+      "step": 266515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015526854695166564,
+      "loss": 2.1734,
+      "step": 266520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015526700659473022,
+      "loss": 2.2152,
+      "step": 266525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015526546621891457,
+      "loss": 2.2389,
+      "step": 266530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015526392582421926,
+      "loss": 2.0778,
+      "step": 266535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015526238541064478,
+      "loss": 2.0222,
+      "step": 266540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015526084497819167,
+      "loss": 2.0121,
+      "step": 266545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015525930452686045,
+      "loss": 1.9206,
+      "step": 266550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015525776405665161,
+      "loss": 2.0857,
+      "step": 266555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015525622356756577,
+      "loss": 1.7056,
+      "step": 266560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015525468305960341,
+      "loss": 2.1396,
+      "step": 266565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015525314253276502,
+      "loss": 2.1686,
+      "step": 266570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015525160198705119,
+      "loss": 2.0799,
+      "step": 266575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015525006142246234,
+      "loss": 2.0295,
+      "step": 266580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001552485208389991,
+      "loss": 2.3768,
+      "step": 266585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015524698023666197,
+      "loss": 2.0911,
+      "step": 266590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015524543961545147,
+      "loss": 2.1104,
+      "step": 266595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015524389897536814,
+      "loss": 2.1019,
+      "step": 266600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015524235831641247,
+      "loss": 2.0259,
+      "step": 266605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015524081763858502,
+      "loss": 2.2566,
+      "step": 266610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001552392769418863,
+      "loss": 2.2661,
+      "step": 266615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015523773622631683,
+      "loss": 2.1267,
+      "step": 266620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001552361954918772,
+      "loss": 2.1173,
+      "step": 266625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015523465473856783,
+      "loss": 1.9003,
+      "step": 266630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015523311396638931,
+      "loss": 1.9834,
+      "step": 266635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015523157317534215,
+      "loss": 2.0149,
+      "step": 266640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001552300323654269,
+      "loss": 2.0191,
+      "step": 266645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015522849153664408,
+      "loss": 2.0778,
+      "step": 266650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015522695068899417,
+      "loss": 2.1992,
+      "step": 266655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015522540982247776,
+      "loss": 2.038,
+      "step": 266660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015522386893709533,
+      "loss": 2.0672,
+      "step": 266665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015522232803284744,
+      "loss": 2.1651,
+      "step": 266670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001552207871097346,
+      "loss": 2.0446,
+      "step": 266675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015521924616775733,
+      "loss": 1.987,
+      "step": 266680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015521770520691617,
+      "loss": 1.9869,
+      "step": 266685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015521616422721162,
+      "loss": 2.0398,
+      "step": 266690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015521462322864424,
+      "loss": 2.1514,
+      "step": 266695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015521308221121455,
+      "loss": 2.0224,
+      "step": 266700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00015521154117492309,
+      "loss": 2.1976,
+      "step": 266705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015521000011977032,
+      "loss": 2.0453,
+      "step": 266710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015520845904575686,
+      "loss": 1.9601,
+      "step": 266715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015520691795288313,
+      "loss": 2.0665,
+      "step": 266720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015520537684114979,
+      "loss": 2.278,
+      "step": 266725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015520383571055723,
+      "loss": 1.9985,
+      "step": 266730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015520229456110607,
+      "loss": 2.0489,
+      "step": 266735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015520075339279678,
+      "loss": 2.0779,
+      "step": 266740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015519921220562993,
+      "loss": 2.1244,
+      "step": 266745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015519767099960604,
+      "loss": 2.1285,
+      "step": 266750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015519612977472563,
+      "loss": 2.0347,
+      "step": 266755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015519458853098918,
+      "loss": 2.2007,
+      "step": 266760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001551930472683973,
+      "loss": 2.0002,
+      "step": 266765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015519150598695042,
+      "loss": 1.9314,
+      "step": 266770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015518996468664916,
+      "loss": 2.1325,
+      "step": 266775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000155188423367494,
+      "loss": 2.1032,
+      "step": 266780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015518688202948547,
+      "loss": 2.1253,
+      "step": 266785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001551853406726241,
+      "loss": 2.2122,
+      "step": 266790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015518379929691046,
+      "loss": 2.0879,
+      "step": 266795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015518225790234498,
+      "loss": 2.261,
+      "step": 266800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015518071648892823,
+      "loss": 1.8752,
+      "step": 266805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015517917505666076,
+      "loss": 2.0862,
+      "step": 266810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015517763360554314,
+      "loss": 2.1564,
+      "step": 266815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001551760921355758,
+      "loss": 1.8731,
+      "step": 266820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015517455064675928,
+      "loss": 2.0482,
+      "step": 266825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015517300913909417,
+      "loss": 1.8223,
+      "step": 266830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015517146761258094,
+      "loss": 2.1667,
+      "step": 266835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015516992606722015,
+      "loss": 1.9599,
+      "step": 266840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015516838450301226,
+      "loss": 2.0264,
+      "step": 266845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015516684291995788,
+      "loss": 1.9462,
+      "step": 266850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015516530131805752,
+      "loss": 2.1646,
+      "step": 266855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015516375969731172,
+      "loss": 2.232,
+      "step": 266860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015516221805772094,
+      "loss": 2.0252,
+      "step": 266865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015516067639928576,
+      "loss": 1.9518,
+      "step": 266870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001551591347220067,
+      "loss": 2.0338,
+      "step": 266875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015515759302588426,
+      "loss": 2.2113,
+      "step": 266880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015515605131091896,
+      "loss": 2.0713,
+      "step": 266885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015515450957711142,
+      "loss": 2.2526,
+      "step": 266890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015515296782446205,
+      "loss": 2.0671,
+      "step": 266895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015515142605297143,
+      "loss": 2.0439,
+      "step": 266900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001551498842626401,
+      "loss": 2.0658,
+      "step": 266905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015514834245346858,
+      "loss": 1.849,
+      "step": 266910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015514680062545736,
+      "loss": 2.1006,
+      "step": 266915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000155145258778607,
+      "loss": 2.1893,
+      "step": 266920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000155143716912918,
+      "loss": 1.9499,
+      "step": 266925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015514217502839094,
+      "loss": 1.9826,
+      "step": 266930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001551406331250263,
+      "loss": 2.1246,
+      "step": 266935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001551390912028246,
+      "loss": 2.1224,
+      "step": 266940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015513754926178643,
+      "loss": 2.0854,
+      "step": 266945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015513600730191228,
+      "loss": 2.0964,
+      "step": 266950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00015513446532320263,
+      "loss": 2.1148,
+      "step": 266955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015513292332565803,
+      "loss": 2.0073,
+      "step": 266960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015513138130927905,
+      "loss": 2.136,
+      "step": 266965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001551298392740662,
+      "loss": 1.9084,
+      "step": 266970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015512829722001998,
+      "loss": 2.2485,
+      "step": 266975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015512675514714092,
+      "loss": 2.0286,
+      "step": 266980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015512521305542957,
+      "loss": 2.1278,
+      "step": 266985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001551236709448865,
+      "loss": 2.0943,
+      "step": 266990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001551221288155121,
+      "loss": 2.1808,
+      "step": 266995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015512058666730707,
+      "loss": 2.1518,
+      "step": 267000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001551190445002718,
+      "loss": 2.2071,
+      "step": 267005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015511750231440684,
+      "loss": 2.1662,
+      "step": 267010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015511596010971278,
+      "loss": 2.2514,
+      "step": 267015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015511441788619007,
+      "loss": 2.0044,
+      "step": 267020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001551128756438393,
+      "loss": 2.1576,
+      "step": 267025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015511133338266098,
+      "loss": 2.1793,
+      "step": 267030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015510979110265564,
+      "loss": 2.2136,
+      "step": 267035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015510824880382378,
+      "loss": 2.0353,
+      "step": 267040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015510670648616594,
+      "loss": 2.2049,
+      "step": 267045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015510516414968267,
+      "loss": 2.015,
+      "step": 267050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015510362179437442,
+      "loss": 1.9761,
+      "step": 267055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015510207942024185,
+      "loss": 1.9911,
+      "step": 267060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015510053702728533,
+      "loss": 1.9312,
+      "step": 267065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015509899461550551,
+      "loss": 2.0282,
+      "step": 267070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001550974521849029,
+      "loss": 2.0787,
+      "step": 267075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015509590973547795,
+      "loss": 2.1191,
+      "step": 267080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015509436726723126,
+      "loss": 1.9622,
+      "step": 267085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015509282478016334,
+      "loss": 2.0395,
+      "step": 267090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015509128227427472,
+      "loss": 2.0949,
+      "step": 267095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001550897397495659,
+      "loss": 2.11,
+      "step": 267100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015508819720603745,
+      "loss": 2.202,
+      "step": 267105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015508665464368984,
+      "loss": 2.0583,
+      "step": 267110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015508511206252369,
+      "loss": 2.0392,
+      "step": 267115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015508356946253939,
+      "loss": 2.0485,
+      "step": 267120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001550820268437376,
+      "loss": 1.9904,
+      "step": 267125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015508048420611874,
+      "loss": 1.9379,
+      "step": 267130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015507894154968344,
+      "loss": 2.1181,
+      "step": 267135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015507739887443215,
+      "loss": 2.0879,
+      "step": 267140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001550758561803654,
+      "loss": 2.1851,
+      "step": 267145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015507431346748377,
+      "loss": 1.8843,
+      "step": 267150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015507277073578774,
+      "loss": 2.0375,
+      "step": 267155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015507122798527788,
+      "loss": 1.9061,
+      "step": 267160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015506968521595464,
+      "loss": 1.9907,
+      "step": 267165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015506814242781863,
+      "loss": 2.1378,
+      "step": 267170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015506659962087033,
+      "loss": 2.0401,
+      "step": 267175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001550650567951103,
+      "loss": 2.0931,
+      "step": 267180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015506351395053902,
+      "loss": 2.144,
+      "step": 267185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015506197108715707,
+      "loss": 2.2323,
+      "step": 267190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015506042820496494,
+      "loss": 2.1772,
+      "step": 267195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001550588853039632,
+      "loss": 2.2371,
+      "step": 267200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001550573423841523,
+      "loss": 2.2032,
+      "step": 267205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015505579944553283,
+      "loss": 2.1207,
+      "step": 267210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015505425648810531,
+      "loss": 2.163,
+      "step": 267215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015505271351187025,
+      "loss": 2.0042,
+      "step": 267220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015505117051682819,
+      "loss": 2.0464,
+      "step": 267225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015504962750297966,
+      "loss": 2.0423,
+      "step": 267230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015504808447032515,
+      "loss": 2.0551,
+      "step": 267235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015504654141886525,
+      "loss": 2.0287,
+      "step": 267240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.00015504499834860042,
+      "loss": 1.9449,
+      "step": 267245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015504345525953124,
+      "loss": 2.1918,
+      "step": 267250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001550419121516582,
+      "loss": 2.015,
+      "step": 267255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015504036902498186,
+      "loss": 2.1407,
+      "step": 267260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015503882587950275,
+      "loss": 1.8961,
+      "step": 267265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015503728271522136,
+      "loss": 2.0707,
+      "step": 267270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015503573953213823,
+      "loss": 2.0171,
+      "step": 267275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015503419633025386,
+      "loss": 2.1898,
+      "step": 267280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015503265310956888,
+      "loss": 2.2417,
+      "step": 267285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001550311098700837,
+      "loss": 2.0116,
+      "step": 267290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015502956661179893,
+      "loss": 2.1848,
+      "step": 267295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015502802333471505,
+      "loss": 2.1752,
+      "step": 267300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015502648003883258,
+      "loss": 2.1347,
+      "step": 267305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015502493672415207,
+      "loss": 2.0626,
+      "step": 267310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015502339339067405,
+      "loss": 1.886,
+      "step": 267315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015502185003839905,
+      "loss": 2.0342,
+      "step": 267320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001550203066673276,
+      "loss": 1.9898,
+      "step": 267325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015501876327746018,
+      "loss": 2.1192,
+      "step": 267330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015501721986879735,
+      "loss": 1.9925,
+      "step": 267335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015501567644133967,
+      "loss": 2.0255,
+      "step": 267340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015501413299508766,
+      "loss": 1.9549,
+      "step": 267345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015501258953004175,
+      "loss": 2.0823,
+      "step": 267350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015501104604620261,
+      "loss": 2.0474,
+      "step": 267355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015500950254357065,
+      "loss": 2.2106,
+      "step": 267360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015500795902214647,
+      "loss": 2.0464,
+      "step": 267365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015500641548193058,
+      "loss": 2.0644,
+      "step": 267370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015500487192292347,
+      "loss": 2.1312,
+      "step": 267375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015500332834512574,
+      "loss": 1.8464,
+      "step": 267380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015500178474853784,
+      "loss": 2.0935,
+      "step": 267385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015500024113316037,
+      "loss": 2.0782,
+      "step": 267390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015499869749899378,
+      "loss": 1.9273,
+      "step": 267395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015499715384603867,
+      "loss": 1.8939,
+      "step": 267400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015499561017429553,
+      "loss": 2.132,
+      "step": 267405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015499406648376488,
+      "loss": 1.8827,
+      "step": 267410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015499252277444725,
+      "loss": 2.0876,
+      "step": 267415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001549909790463432,
+      "loss": 1.8617,
+      "step": 267420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001549894352994532,
+      "loss": 1.964,
+      "step": 267425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015498789153377786,
+      "loss": 1.9793,
+      "step": 267430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015498634774931765,
+      "loss": 2.0044,
+      "step": 267435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015498480394607306,
+      "loss": 2.074,
+      "step": 267440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001549832601240447,
+      "loss": 2.0429,
+      "step": 267445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015498171628323306,
+      "loss": 2.0344,
+      "step": 267450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015498017242363864,
+      "loss": 2.144,
+      "step": 267455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015497862854526204,
+      "loss": 2.0258,
+      "step": 267460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001549770846481037,
+      "loss": 2.1857,
+      "step": 267465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015497554073216424,
+      "loss": 2.0,
+      "step": 267470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001549739967974441,
+      "loss": 1.8831,
+      "step": 267475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001549724528439439,
+      "loss": 2.0867,
+      "step": 267480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015497090887166403,
+      "loss": 1.987,
+      "step": 267485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015496936488060517,
+      "loss": 2.1312,
+      "step": 267490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015496782087076773,
+      "loss": 2.0237,
+      "step": 267495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001549662768421523,
+      "loss": 2.0567,
+      "step": 267500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015496473279475942,
+      "loss": 1.9859,
+      "step": 267505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015496318872858957,
+      "loss": 2.125,
+      "step": 267510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001549616446436433,
+      "loss": 2.3095,
+      "step": 267515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015496010053992108,
+      "loss": 1.9154,
+      "step": 267520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015495855641742357,
+      "loss": 1.9998,
+      "step": 267525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001549570122761512,
+      "loss": 2.0633,
+      "step": 267530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001549554681161045,
+      "loss": 2.1786,
+      "step": 267535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015495392393728404,
+      "loss": 1.9707,
+      "step": 267540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015495237973969035,
+      "loss": 1.8994,
+      "step": 267545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015495083552332387,
+      "loss": 1.834,
+      "step": 267550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001549492912881852,
+      "loss": 2.1657,
+      "step": 267555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015494774703427489,
+      "loss": 1.9304,
+      "step": 267560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001549462027615934,
+      "loss": 2.0692,
+      "step": 267565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015494465847014127,
+      "loss": 2.0167,
+      "step": 267570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001549431141599191,
+      "loss": 2.0509,
+      "step": 267575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015494156983092734,
+      "loss": 2.4049,
+      "step": 267580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015494002548316655,
+      "loss": 1.9811,
+      "step": 267585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015493848111663725,
+      "loss": 2.1182,
+      "step": 267590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015493693673133998,
+      "loss": 2.0855,
+      "step": 267595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015493539232727522,
+      "loss": 1.9933,
+      "step": 267600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015493384790444357,
+      "loss": 2.1248,
+      "step": 267605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001549323034628455,
+      "loss": 1.9958,
+      "step": 267610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015493075900248156,
+      "loss": 2.15,
+      "step": 267615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015492921452335228,
+      "loss": 2.1674,
+      "step": 267620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001549276700254582,
+      "loss": 2.1261,
+      "step": 267625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001549261255087998,
+      "loss": 2.16,
+      "step": 267630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015492458097337767,
+      "loss": 1.7709,
+      "step": 267635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001549230364191923,
+      "loss": 2.0285,
+      "step": 267640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015492149184624424,
+      "loss": 2.1381,
+      "step": 267645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015491994725453395,
+      "loss": 1.9521,
+      "step": 267650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.00015491840264406204,
+      "loss": 2.0121,
+      "step": 267655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015491685801482903,
+      "loss": 2.0918,
+      "step": 267660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001549153133668354,
+      "loss": 2.0589,
+      "step": 267665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001549137687000817,
+      "loss": 2.0839,
+      "step": 267670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015491222401456847,
+      "loss": 2.1159,
+      "step": 267675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001549106793102962,
+      "loss": 1.9815,
+      "step": 267680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001549091345872655,
+      "loss": 1.9188,
+      "step": 267685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015490758984547682,
+      "loss": 1.9267,
+      "step": 267690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015490604508493073,
+      "loss": 2.0899,
+      "step": 267695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015490450030562772,
+      "loss": 2.1883,
+      "step": 267700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015490295550756833,
+      "loss": 1.9772,
+      "step": 267705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015490141069075307,
+      "loss": 2.0111,
+      "step": 267710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015489986585518254,
+      "loss": 2.0893,
+      "step": 267715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015489832100085718,
+      "loss": 1.9326,
+      "step": 267720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001548967761277776,
+      "loss": 2.0894,
+      "step": 267725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015489523123594427,
+      "loss": 1.9942,
+      "step": 267730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001548936863253577,
+      "loss": 2.0115,
+      "step": 267735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001548921413960185,
+      "loss": 2.1919,
+      "step": 267740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015489059644792713,
+      "loss": 2.1046,
+      "step": 267745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001548890514810841,
+      "loss": 2.1082,
+      "step": 267750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015488750649549004,
+      "loss": 2.0771,
+      "step": 267755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001548859614911454,
+      "loss": 2.0189,
+      "step": 267760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001548844164680507,
+      "loss": 2.1933,
+      "step": 267765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001548828714262065,
+      "loss": 2.102,
+      "step": 267770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001548813263656133,
+      "loss": 1.8438,
+      "step": 267775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015487978128627162,
+      "loss": 2.0867,
+      "step": 267780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015487823618818208,
+      "loss": 2.0427,
+      "step": 267785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001548766910713451,
+      "loss": 2.0251,
+      "step": 267790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015487514593576123,
+      "loss": 2.2725,
+      "step": 267795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015487360078143104,
+      "loss": 2.214,
+      "step": 267800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015487205560835502,
+      "loss": 2.1203,
+      "step": 267805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015487051041653373,
+      "loss": 2.0737,
+      "step": 267810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015486896520596765,
+      "loss": 2.1658,
+      "step": 267815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015486741997665734,
+      "loss": 1.9633,
+      "step": 267820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015486587472860334,
+      "loss": 2.1853,
+      "step": 267825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015486432946180617,
+      "loss": 2.2042,
+      "step": 267830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015486278417626634,
+      "loss": 2.0096,
+      "step": 267835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015486123887198437,
+      "loss": 2.0081,
+      "step": 267840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015485969354896083,
+      "loss": 2.0988,
+      "step": 267845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001548581482071962,
+      "loss": 2.1328,
+      "step": 267850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001548566028466911,
+      "loss": 2.1982,
+      "step": 267855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015485505746744592,
+      "loss": 2.1436,
+      "step": 267860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015485351206946126,
+      "loss": 2.0586,
+      "step": 267865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015485196665273766,
+      "loss": 1.8982,
+      "step": 267870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.40625,
+      "learning_rate": 0.00015485042121727563,
+      "loss": 2.1878,
+      "step": 267875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015484887576307568,
+      "loss": 2.1574,
+      "step": 267880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001548473302901384,
+      "loss": 2.1986,
+      "step": 267885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015484578479846425,
+      "loss": 2.1135,
+      "step": 267890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001548442392880538,
+      "loss": 2.0288,
+      "step": 267895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015484269375890756,
+      "loss": 2.1405,
+      "step": 267900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015484114821102607,
+      "loss": 2.0909,
+      "step": 267905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015483960264440984,
+      "loss": 2.2247,
+      "step": 267910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015483805705905943,
+      "loss": 2.1476,
+      "step": 267915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015483651145497532,
+      "loss": 1.9922,
+      "step": 267920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015483496583215805,
+      "loss": 2.0296,
+      "step": 267925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001548334201906082,
+      "loss": 1.9648,
+      "step": 267930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001548318745303262,
+      "loss": 2.0526,
+      "step": 267935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015483032885131267,
+      "loss": 1.9643,
+      "step": 267940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001548287831535681,
+      "loss": 2.1524,
+      "step": 267945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015482723743709303,
+      "loss": 2.0596,
+      "step": 267950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000154825691701888,
+      "loss": 2.195,
+      "step": 267955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001548241459479535,
+      "loss": 2.0509,
+      "step": 267960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015482260017529007,
+      "loss": 1.9685,
+      "step": 267965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001548210543838983,
+      "loss": 1.9934,
+      "step": 267970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015481950857377857,
+      "loss": 2.1525,
+      "step": 267975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015481796274493157,
+      "loss": 2.043,
+      "step": 267980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015481641689735773,
+      "loss": 2.2512,
+      "step": 267985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001548148710310576,
+      "loss": 2.0744,
+      "step": 267990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015481332514603175,
+      "loss": 1.9875,
+      "step": 267995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015481177924228068,
+      "loss": 1.9271,
+      "step": 268000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015481023331980487,
+      "loss": 2.0263,
+      "step": 268005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001548086873786049,
+      "loss": 2.1775,
+      "step": 268010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015480714141868131,
+      "loss": 2.1176,
+      "step": 268015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015480559544003462,
+      "loss": 1.9744,
+      "step": 268020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001548040494426653,
+      "loss": 1.9841,
+      "step": 268025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015480250342657392,
+      "loss": 1.8739,
+      "step": 268030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015480095739176105,
+      "loss": 2.1004,
+      "step": 268035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015479941133822719,
+      "loss": 1.9637,
+      "step": 268040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001547978652659728,
+      "loss": 2.0805,
+      "step": 268045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001547963191749985,
+      "loss": 2.1354,
+      "step": 268050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015479477306530476,
+      "loss": 2.1427,
+      "step": 268055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00015479322693689218,
+      "loss": 2.0233,
+      "step": 268060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001547916807897612,
+      "loss": 2.1215,
+      "step": 268065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015479013462391241,
+      "loss": 1.9216,
+      "step": 268070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001547885884393463,
+      "loss": 2.2097,
+      "step": 268075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015478704223606342,
+      "loss": 1.9563,
+      "step": 268080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015478549601406427,
+      "loss": 1.893,
+      "step": 268085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015478394977334943,
+      "loss": 2.1414,
+      "step": 268090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015478240351391942,
+      "loss": 2.0338,
+      "step": 268095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001547808572357747,
+      "loss": 1.9547,
+      "step": 268100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015477931093891585,
+      "loss": 2.1762,
+      "step": 268105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015477776462334344,
+      "loss": 1.9312,
+      "step": 268110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015477621828905791,
+      "loss": 2.1029,
+      "step": 268115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015477467193605985,
+      "loss": 1.9686,
+      "step": 268120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015477312556434975,
+      "loss": 2.1452,
+      "step": 268125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001547715791739282,
+      "loss": 1.8262,
+      "step": 268130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015477003276479563,
+      "loss": 2.0838,
+      "step": 268135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015476848633695265,
+      "loss": 2.0577,
+      "step": 268140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015476693989039975,
+      "loss": 2.0826,
+      "step": 268145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001547653934251375,
+      "loss": 2.0804,
+      "step": 268150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015476384694116635,
+      "loss": 2.065,
+      "step": 268155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001547623004384869,
+      "loss": 2.0008,
+      "step": 268160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015476075391709966,
+      "loss": 2.1557,
+      "step": 268165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015475920737700516,
+      "loss": 1.9599,
+      "step": 268170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015475766081820392,
+      "loss": 2.1135,
+      "step": 268175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015475611424069645,
+      "loss": 2.0455,
+      "step": 268180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015475456764448331,
+      "loss": 2.0123,
+      "step": 268185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000154753021029565,
+      "loss": 2.0527,
+      "step": 268190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001547514743959421,
+      "loss": 2.1852,
+      "step": 268195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001547499277436151,
+      "loss": 2.1882,
+      "step": 268200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001547483810725845,
+      "loss": 1.9393,
+      "step": 268205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001547468343828509,
+      "loss": 2.1001,
+      "step": 268210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015474528767441475,
+      "loss": 2.2847,
+      "step": 268215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00015474374094727664,
+      "loss": 2.1143,
+      "step": 268220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015474219420143705,
+      "loss": 2.2171,
+      "step": 268225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015474064743689657,
+      "loss": 2.1027,
+      "step": 268230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001547391006536557,
+      "loss": 2.0845,
+      "step": 268235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015473755385171493,
+      "loss": 2.0145,
+      "step": 268240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001547360070310748,
+      "loss": 2.0787,
+      "step": 268245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001547344601917359,
+      "loss": 1.985,
+      "step": 268250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001547329133336987,
+      "loss": 2.1404,
+      "step": 268255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015473136645696373,
+      "loss": 2.0876,
+      "step": 268260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015472981956153156,
+      "loss": 2.2062,
+      "step": 268265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015472827264740265,
+      "loss": 1.9277,
+      "step": 268270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001547267257145776,
+      "loss": 2.1596,
+      "step": 268275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001547251787630569,
+      "loss": 1.9918,
+      "step": 268280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001547236317928411,
+      "loss": 2.1054,
+      "step": 268285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001547220848039307,
+      "loss": 1.996,
+      "step": 268290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015472053779632625,
+      "loss": 2.0489,
+      "step": 268295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015471899077002825,
+      "loss": 2.1346,
+      "step": 268300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015471744372503728,
+      "loss": 2.1009,
+      "step": 268305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015471589666135382,
+      "loss": 1.9678,
+      "step": 268310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.00015471434957897845,
+      "loss": 2.1898,
+      "step": 268315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015471280247791164,
+      "loss": 2.0511,
+      "step": 268320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015471125535815392,
+      "loss": 2.0811,
+      "step": 268325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015470970821970586,
+      "loss": 2.0648,
+      "step": 268330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000154708161062568,
+      "loss": 2.2051,
+      "step": 268335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015470661388674084,
+      "loss": 1.9945,
+      "step": 268340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001547050666922249,
+      "loss": 2.198,
+      "step": 268345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015470351947902068,
+      "loss": 2.1815,
+      "step": 268350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015470197224712876,
+      "loss": 2.0325,
+      "step": 268355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015470042499654966,
+      "loss": 2.1627,
+      "step": 268360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015469887772728394,
+      "loss": 2.055,
+      "step": 268365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015469733043933207,
+      "loss": 1.945,
+      "step": 268370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015469578313269457,
+      "loss": 2.0753,
+      "step": 268375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000154694235807372,
+      "loss": 2.1184,
+      "step": 268380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001546926884633649,
+      "loss": 1.9838,
+      "step": 268385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015469114110067383,
+      "loss": 2.0232,
+      "step": 268390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015468959371929924,
+      "loss": 1.9829,
+      "step": 268395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015468804631924168,
+      "loss": 2.0836,
+      "step": 268400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015468649890050168,
+      "loss": 2.0424,
+      "step": 268405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001546849514630798,
+      "loss": 2.106,
+      "step": 268410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015468340400697655,
+      "loss": 2.0453,
+      "step": 268415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015468185653219247,
+      "loss": 1.9004,
+      "step": 268420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015468030903872804,
+      "loss": 2.0949,
+      "step": 268425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015467876152658384,
+      "loss": 1.9415,
+      "step": 268430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001546772139957604,
+      "loss": 1.9472,
+      "step": 268435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015467566644625824,
+      "loss": 2.1292,
+      "step": 268440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015467411887807786,
+      "loss": 2.1995,
+      "step": 268445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001546725712912198,
+      "loss": 2.0283,
+      "step": 268450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015467102368568462,
+      "loss": 2.2097,
+      "step": 268455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015466947606147281,
+      "loss": 2.0976,
+      "step": 268460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015466792841858492,
+      "loss": 2.0227,
+      "step": 268465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015466638075702148,
+      "loss": 2.1556,
+      "step": 268470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015466483307678304,
+      "loss": 2.0307,
+      "step": 268475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015466328537787005,
+      "loss": 2.0937,
+      "step": 268480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015466173766028312,
+      "loss": 1.8989,
+      "step": 268485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015466018992402276,
+      "loss": 1.8489,
+      "step": 268490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015465864216908945,
+      "loss": 2.257,
+      "step": 268495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001546570943954838,
+      "loss": 2.1289,
+      "step": 268500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015465554660320624,
+      "loss": 2.0987,
+      "step": 268505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015465399879225738,
+      "loss": 2.185,
+      "step": 268510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015465245096263775,
+      "loss": 2.1324,
+      "step": 268515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001546509031143478,
+      "loss": 2.0435,
+      "step": 268520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015464935524738818,
+      "loss": 2.1242,
+      "step": 268525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001546478073617593,
+      "loss": 2.1176,
+      "step": 268530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015464625945746176,
+      "loss": 1.9683,
+      "step": 268535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015464471153449604,
+      "loss": 2.0089,
+      "step": 268540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015464316359286272,
+      "loss": 2.0773,
+      "step": 268545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001546416156325623,
+      "loss": 1.9125,
+      "step": 268550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001546400676535953,
+      "loss": 2.1455,
+      "step": 268555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015463851965596227,
+      "loss": 2.1979,
+      "step": 268560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001546369716396637,
+      "loss": 2.1126,
+      "step": 268565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001546354236047002,
+      "loss": 2.1949,
+      "step": 268570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001546338755510722,
+      "loss": 2.1126,
+      "step": 268575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015463232747878035,
+      "loss": 2.0409,
+      "step": 268580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015463077938782503,
+      "loss": 1.9258,
+      "step": 268585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015462923127820689,
+      "loss": 2.0376,
+      "step": 268590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001546276831499264,
+      "loss": 1.858,
+      "step": 268595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001546261350029841,
+      "loss": 1.9771,
+      "step": 268600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015462458683738052,
+      "loss": 2.1359,
+      "step": 268605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015462303865311622,
+      "loss": 2.0342,
+      "step": 268610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015462149045019166,
+      "loss": 2.199,
+      "step": 268615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001546199422286074,
+      "loss": 2.2406,
+      "step": 268620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000154618393988364,
+      "loss": 2.0701,
+      "step": 268625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000154616845729462,
+      "loss": 2.076,
+      "step": 268630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015461529745190185,
+      "loss": 1.9063,
+      "step": 268635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015461374915568413,
+      "loss": 2.094,
+      "step": 268640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015461220084080933,
+      "loss": 2.1322,
+      "step": 268645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001546106525072781,
+      "loss": 2.2089,
+      "step": 268650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0,
+      "learning_rate": 0.0001546091041550908,
+      "loss": 2.0319,
+      "step": 268655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015460755578424807,
+      "loss": 1.8846,
+      "step": 268660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015460600739475042,
+      "loss": 2.0954,
+      "step": 268665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015460445898659835,
+      "loss": 2.1294,
+      "step": 268670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001546029105597924,
+      "loss": 2.0467,
+      "step": 268675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015460136211433313,
+      "loss": 2.102,
+      "step": 268680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015459981365022104,
+      "loss": 2.1041,
+      "step": 268685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015459826516745664,
+      "loss": 2.0556,
+      "step": 268690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001545967166660405,
+      "loss": 2.0602,
+      "step": 268695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015459516814597313,
+      "loss": 1.9904,
+      "step": 268700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015459361960725506,
+      "loss": 2.0981,
+      "step": 268705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001545920710498868,
+      "loss": 2.2472,
+      "step": 268710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015459052247386893,
+      "loss": 1.9164,
+      "step": 268715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015458897387920194,
+      "loss": 2.0809,
+      "step": 268720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00015458742526588636,
+      "loss": 2.2621,
+      "step": 268725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015458587663392273,
+      "loss": 2.2023,
+      "step": 268730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015458432798331158,
+      "loss": 2.257,
+      "step": 268735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015458277931405343,
+      "loss": 1.9722,
+      "step": 268740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0001545812306261488,
+      "loss": 2.4157,
+      "step": 268745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015457968191959825,
+      "loss": 2.0477,
+      "step": 268750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015457813319440225,
+      "loss": 1.9317,
+      "step": 268755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015457658445056143,
+      "loss": 2.0865,
+      "step": 268760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015457503568807622,
+      "loss": 2.1523,
+      "step": 268765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001545734869069472,
+      "loss": 2.019,
+      "step": 268770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015457193810717486,
+      "loss": 2.2015,
+      "step": 268775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001545703892887598,
+      "loss": 1.9654,
+      "step": 268780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015456884045170248,
+      "loss": 2.0342,
+      "step": 268785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015456729159600346,
+      "loss": 2.1046,
+      "step": 268790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015456574272166326,
+      "loss": 2.1132,
+      "step": 268795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015456419382868243,
+      "loss": 2.1333,
+      "step": 268800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015456264491706145,
+      "loss": 2.2331,
+      "step": 268805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015456109598680092,
+      "loss": 2.0243,
+      "step": 268810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001545595470379013,
+      "loss": 2.3623,
+      "step": 268815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015455799807036316,
+      "loss": 2.2,
+      "step": 268820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.000154556449084187,
+      "loss": 2.0664,
+      "step": 268825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001545549000793734,
+      "loss": 2.1096,
+      "step": 268830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015455335105592284,
+      "loss": 1.9129,
+      "step": 268835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015455180201383585,
+      "loss": 2.3464,
+      "step": 268840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000154550252953113,
+      "loss": 2.0862,
+      "step": 268845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015454870387375478,
+      "loss": 2.0365,
+      "step": 268850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015454715477576176,
+      "loss": 1.9319,
+      "step": 268855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001545456056591344,
+      "loss": 2.0242,
+      "step": 268860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001545440565238733,
+      "loss": 1.8453,
+      "step": 268865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015454250736997898,
+      "loss": 2.1376,
+      "step": 268870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001545409581974519,
+      "loss": 2.0435,
+      "step": 268875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015453940900629268,
+      "loss": 2.0564,
+      "step": 268880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001545378597965018,
+      "loss": 2.0429,
+      "step": 268885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015453631056807978,
+      "loss": 2.0455,
+      "step": 268890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00015453476132102717,
+      "loss": 1.9044,
+      "step": 268895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015453321205534452,
+      "loss": 2.115,
+      "step": 268900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015453166277103233,
+      "loss": 2.1011,
+      "step": 268905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015453011346809112,
+      "loss": 2.0671,
+      "step": 268910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015452856414652143,
+      "loss": 1.9574,
+      "step": 268915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015452701480632382,
+      "loss": 1.7429,
+      "step": 268920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015452546544749877,
+      "loss": 2.0927,
+      "step": 268925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015452391607004683,
+      "loss": 1.9469,
+      "step": 268930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015452236667396857,
+      "loss": 2.24,
+      "step": 268935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015452081725926444,
+      "loss": 1.9674,
+      "step": 268940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015451926782593503,
+      "loss": 2.1962,
+      "step": 268945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001545177183739808,
+      "loss": 2.1796,
+      "step": 268950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015451616890340238,
+      "loss": 2.1962,
+      "step": 268955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015451461941420023,
+      "loss": 2.2699,
+      "step": 268960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001545130699063749,
+      "loss": 2.2293,
+      "step": 268965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015451152037992695,
+      "loss": 2.0642,
+      "step": 268970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015450997083485684,
+      "loss": 2.1775,
+      "step": 268975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015450842127116515,
+      "loss": 1.9802,
+      "step": 268980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001545068716888524,
+      "loss": 1.8502,
+      "step": 268985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001545053220879191,
+      "loss": 2.0234,
+      "step": 268990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015450377246836578,
+      "loss": 2.0263,
+      "step": 268995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000154502222830193,
+      "loss": 2.0106,
+      "step": 269000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001545006731734013,
+      "loss": 2.1556,
+      "step": 269005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015449912349799113,
+      "loss": 2.1319,
+      "step": 269010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001544975738039631,
+      "loss": 2.0694,
+      "step": 269015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015449602409131768,
+      "loss": 1.9643,
+      "step": 269020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015449447436005547,
+      "loss": 2.1074,
+      "step": 269025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015449292461017694,
+      "loss": 2.0848,
+      "step": 269030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015449137484168263,
+      "loss": 2.2357,
+      "step": 269035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015448982505457311,
+      "loss": 2.0791,
+      "step": 269040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015448827524884885,
+      "loss": 2.279,
+      "step": 269045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001544867254245104,
+      "loss": 1.9558,
+      "step": 269050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001544851755815583,
+      "loss": 1.9822,
+      "step": 269055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015448362571999312,
+      "loss": 2.2896,
+      "step": 269060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015448207583981528,
+      "loss": 2.186,
+      "step": 269065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015448052594102543,
+      "loss": 2.0483,
+      "step": 269070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000154478976023624,
+      "loss": 2.045,
+      "step": 269075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001544774260876116,
+      "loss": 2.1504,
+      "step": 269080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001544758761329887,
+      "loss": 2.1009,
+      "step": 269085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015447432615975587,
+      "loss": 2.1112,
+      "step": 269090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001544727761679136,
+      "loss": 2.4221,
+      "step": 269095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015447122615746246,
+      "loss": 2.0489,
+      "step": 269100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015446967612840295,
+      "loss": 1.9847,
+      "step": 269105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001544681260807356,
+      "loss": 2.1068,
+      "step": 269110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000154466576014461,
+      "loss": 2.0141,
+      "step": 269115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001544650259295796,
+      "loss": 1.9917,
+      "step": 269120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015446347582609196,
+      "loss": 1.978,
+      "step": 269125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001544619257039986,
+      "loss": 2.2076,
+      "step": 269130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015446037556330003,
+      "loss": 2.0704,
+      "step": 269135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015445882540399686,
+      "loss": 2.1612,
+      "step": 269140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015445727522608955,
+      "loss": 2.0277,
+      "step": 269145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015445572502957865,
+      "loss": 1.9074,
+      "step": 269150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00015445417481446466,
+      "loss": 2.079,
+      "step": 269155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001544526245807482,
+      "loss": 1.9592,
+      "step": 269160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015445107432842967,
+      "loss": 2.111,
+      "step": 269165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001544495240575097,
+      "loss": 1.9142,
+      "step": 269170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001544479737679888,
+      "loss": 2.0346,
+      "step": 269175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015444642345986744,
+      "loss": 1.9352,
+      "step": 269180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015444487313314622,
+      "loss": 2.0266,
+      "step": 269185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015444332278782561,
+      "loss": 2.0591,
+      "step": 269190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015444177242390623,
+      "loss": 1.9461,
+      "step": 269195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015444022204138852,
+      "loss": 2.0756,
+      "step": 269200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015443867164027303,
+      "loss": 1.9936,
+      "step": 269205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015443712122056034,
+      "loss": 2.2106,
+      "step": 269210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001544355707822509,
+      "loss": 2.0653,
+      "step": 269215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015443402032534534,
+      "loss": 2.0711,
+      "step": 269220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015443246984984406,
+      "loss": 2.085,
+      "step": 269225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015443091935574772,
+      "loss": 2.0482,
+      "step": 269230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015442936884305675,
+      "loss": 1.9934,
+      "step": 269235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015442781831177173,
+      "loss": 2.0991,
+      "step": 269240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015442626776189319,
+      "loss": 2.178,
+      "step": 269245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015442471719342166,
+      "loss": 2.1212,
+      "step": 269250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015442316660635767,
+      "loss": 2.1161,
+      "step": 269255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015442161600070172,
+      "loss": 2.1677,
+      "step": 269260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015442006537645433,
+      "loss": 2.2306,
+      "step": 269265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015441851473361608,
+      "loss": 1.9981,
+      "step": 269270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001544169640721875,
+      "loss": 2.1386,
+      "step": 269275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015441541339216908,
+      "loss": 1.9905,
+      "step": 269280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015441386269356136,
+      "loss": 2.1362,
+      "step": 269285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015441231197636487,
+      "loss": 1.911,
+      "step": 269290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015441076124058018,
+      "loss": 2.0213,
+      "step": 269295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015440921048620777,
+      "loss": 2.1109,
+      "step": 269300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015440765971324818,
+      "loss": 1.8991,
+      "step": 269305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015440610892170198,
+      "loss": 2.3343,
+      "step": 269310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015440455811156962,
+      "loss": 2.1147,
+      "step": 269315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001544030072828517,
+      "loss": 2.1705,
+      "step": 269320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015440145643554873,
+      "loss": 2.1183,
+      "step": 269325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015439990556966122,
+      "loss": 2.0755,
+      "step": 269330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015439835468518976,
+      "loss": 2.263,
+      "step": 269335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015439680378213477,
+      "loss": 2.1524,
+      "step": 269340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015439525286049688,
+      "loss": 2.0364,
+      "step": 269345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015439370192027658,
+      "loss": 2.1384,
+      "step": 269350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001543921509614744,
+      "loss": 1.9784,
+      "step": 269355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015439059998409088,
+      "loss": 2.0433,
+      "step": 269360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015438904898812657,
+      "loss": 1.9851,
+      "step": 269365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015438749797358194,
+      "loss": 2.1065,
+      "step": 269370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015438594694045757,
+      "loss": 2.1638,
+      "step": 269375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015438439588875394,
+      "loss": 2.1114,
+      "step": 269380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015438284481847166,
+      "loss": 2.2034,
+      "step": 269385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001543812937296112,
+      "loss": 2.1377,
+      "step": 269390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001543797426221731,
+      "loss": 2.0554,
+      "step": 269395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001543781914961579,
+      "loss": 2.0169,
+      "step": 269400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001543766403515661,
+      "loss": 2.1384,
+      "step": 269405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015437508918839828,
+      "loss": 1.9575,
+      "step": 269410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015437353800665493,
+      "loss": 2.1056,
+      "step": 269415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001543719868063366,
+      "loss": 2.0351,
+      "step": 269420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001543704355874438,
+      "loss": 2.0883,
+      "step": 269425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001543688843499771,
+      "loss": 2.1695,
+      "step": 269430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.000154367333093937,
+      "loss": 2.0334,
+      "step": 269435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000154365781819324,
+      "loss": 2.2006,
+      "step": 269440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015436423052613872,
+      "loss": 2.0337,
+      "step": 269445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015436267921438158,
+      "loss": 1.9703,
+      "step": 269450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015436112788405317,
+      "loss": 2.2964,
+      "step": 269455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015435957653515403,
+      "loss": 2.1959,
+      "step": 269460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015435802516768464,
+      "loss": 2.0483,
+      "step": 269465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015435647378164559,
+      "loss": 2.1348,
+      "step": 269470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001543549223770374,
+      "loss": 2.307,
+      "step": 269475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015435337095386054,
+      "loss": 2.1515,
+      "step": 269480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001543518195121156,
+      "loss": 2.1702,
+      "step": 269485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015435026805180312,
+      "loss": 2.1793,
+      "step": 269490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001543487165729236,
+      "loss": 2.0871,
+      "step": 269495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015434716507547753,
+      "loss": 1.9805,
+      "step": 269500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001543456135594655,
+      "loss": 2.1817,
+      "step": 269505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015434406202488802,
+      "loss": 2.1581,
+      "step": 269510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015434251047174565,
+      "loss": 2.1152,
+      "step": 269515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015434095890003887,
+      "loss": 2.25,
+      "step": 269520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015433940730976827,
+      "loss": 1.9053,
+      "step": 269525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001543378557009343,
+      "loss": 2.2987,
+      "step": 269530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015433630407353754,
+      "loss": 1.8653,
+      "step": 269535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001543347524275785,
+      "loss": 1.9812,
+      "step": 269540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015433320076305775,
+      "loss": 1.9938,
+      "step": 269545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015433164907997576,
+      "loss": 1.9429,
+      "step": 269550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015433009737833312,
+      "loss": 1.9678,
+      "step": 269555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015432854565813035,
+      "loss": 2.1173,
+      "step": 269560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015432699391936794,
+      "loss": 2.1445,
+      "step": 269565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015432544216204645,
+      "loss": 2.1535,
+      "step": 269570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001543238903861664,
+      "loss": 2.0486,
+      "step": 269575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015432233859172834,
+      "loss": 2.1621,
+      "step": 269580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015432078677873274,
+      "loss": 1.8954,
+      "step": 269585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001543192349471802,
+      "loss": 2.2142,
+      "step": 269590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015431768309707124,
+      "loss": 1.9542,
+      "step": 269595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015431613122840636,
+      "loss": 2.1955,
+      "step": 269600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015431457934118612,
+      "loss": 2.1458,
+      "step": 269605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015431302743541098,
+      "loss": 2.1414,
+      "step": 269610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001543114755110816,
+      "loss": 1.9991,
+      "step": 269615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001543099235681984,
+      "loss": 2.1977,
+      "step": 269620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015430837160676194,
+      "loss": 1.938,
+      "step": 269625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015430681962677278,
+      "loss": 2.1897,
+      "step": 269630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015430526762823138,
+      "loss": 2.2454,
+      "step": 269635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015430371561113836,
+      "loss": 2.2574,
+      "step": 269640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015430216357549417,
+      "loss": 2.0214,
+      "step": 269645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001543006115212994,
+      "loss": 2.078,
+      "step": 269650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015429905944855453,
+      "loss": 2.0411,
+      "step": 269655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015429750735726017,
+      "loss": 2.3071,
+      "step": 269660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015429595524741677,
+      "loss": 2.2077,
+      "step": 269665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015429440311902487,
+      "loss": 2.1128,
+      "step": 269670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015429285097208503,
+      "loss": 2.0139,
+      "step": 269675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015429129880659777,
+      "loss": 2.1969,
+      "step": 269680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001542897466225636,
+      "loss": 1.9878,
+      "step": 269685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015428819441998308,
+      "loss": 1.8734,
+      "step": 269690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015428664219885673,
+      "loss": 2.2284,
+      "step": 269695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015428508995918506,
+      "loss": 2.2654,
+      "step": 269700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015428353770096867,
+      "loss": 1.9324,
+      "step": 269705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000154281985424208,
+      "loss": 2.1299,
+      "step": 269710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001542804331289036,
+      "loss": 2.2031,
+      "step": 269715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015427888081505605,
+      "loss": 2.0001,
+      "step": 269720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015427732848266585,
+      "loss": 1.9827,
+      "step": 269725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015427577613173353,
+      "loss": 2.1462,
+      "step": 269730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015427422376225962,
+      "loss": 2.0487,
+      "step": 269735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015427267137424464,
+      "loss": 2.033,
+      "step": 269740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00015427111896768912,
+      "loss": 2.251,
+      "step": 269745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015426956654259364,
+      "loss": 2.0446,
+      "step": 269750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015426801409895865,
+      "loss": 2.0739,
+      "step": 269755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015426646163678475,
+      "loss": 2.1973,
+      "step": 269760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015426490915607243,
+      "loss": 2.3043,
+      "step": 269765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015426335665682226,
+      "loss": 2.1242,
+      "step": 269770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001542618041390347,
+      "loss": 2.1524,
+      "step": 269775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015426025160271035,
+      "loss": 2.0408,
+      "step": 269780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015425869904784972,
+      "loss": 2.0996,
+      "step": 269785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015425714647445332,
+      "loss": 2.0352,
+      "step": 269790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001542555938825217,
+      "loss": 1.9682,
+      "step": 269795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015425404127205537,
+      "loss": 2.2162,
+      "step": 269800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015425248864305488,
+      "loss": 2.0912,
+      "step": 269805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001542509359955208,
+      "loss": 2.0814,
+      "step": 269810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015424938332945357,
+      "loss": 2.1863,
+      "step": 269815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015424783064485378,
+      "loss": 2.3277,
+      "step": 269820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015424627794172194,
+      "loss": 2.0933,
+      "step": 269825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001542447252200586,
+      "loss": 1.974,
+      "step": 269830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015424317247986427,
+      "loss": 2.1418,
+      "step": 269835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015424161972113945,
+      "loss": 2.0202,
+      "step": 269840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015424006694388478,
+      "loss": 2.23,
+      "step": 269845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015423851414810068,
+      "loss": 2.1732,
+      "step": 269850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015423696133378773,
+      "loss": 1.8395,
+      "step": 269855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015423540850094644,
+      "loss": 2.0405,
+      "step": 269860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015423385564957738,
+      "loss": 2.1237,
+      "step": 269865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015423230277968104,
+      "loss": 2.0334,
+      "step": 269870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015423074989125794,
+      "loss": 2.0174,
+      "step": 269875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.578125,
+      "learning_rate": 0.00015422919698430864,
+      "loss": 2.0512,
+      "step": 269880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015422764405883366,
+      "loss": 2.1753,
+      "step": 269885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015422609111483354,
+      "loss": 2.1168,
+      "step": 269890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015422453815230883,
+      "loss": 1.976,
+      "step": 269895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015422298517126,
+      "loss": 2.0676,
+      "step": 269900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015422143217168758,
+      "loss": 2.0388,
+      "step": 269905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015421987915359222,
+      "loss": 1.9634,
+      "step": 269910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015421832611697432,
+      "loss": 2.1608,
+      "step": 269915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015421677306183445,
+      "loss": 2.1431,
+      "step": 269920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015421521998817316,
+      "loss": 2.0493,
+      "step": 269925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015421366689599095,
+      "loss": 2.0187,
+      "step": 269930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001542121137852884,
+      "loss": 2.1889,
+      "step": 269935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015421056065606598,
+      "loss": 2.1401,
+      "step": 269940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015420900750832427,
+      "loss": 1.9941,
+      "step": 269945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001542074543420638,
+      "loss": 2.1145,
+      "step": 269950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015420590115728503,
+      "loss": 2.1051,
+      "step": 269955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015420434795398858,
+      "loss": 2.1271,
+      "step": 269960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001542027947321749,
+      "loss": 2.207,
+      "step": 269965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001542012414918446,
+      "loss": 1.9711,
+      "step": 269970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015419968823299817,
+      "loss": 2.1262,
+      "step": 269975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015419813495563613,
+      "loss": 1.9943,
+      "step": 269980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000154196581659759,
+      "loss": 2.0682,
+      "step": 269985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015419502834536738,
+      "loss": 1.9983,
+      "step": 269990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015419347501246174,
+      "loss": 2.1363,
+      "step": 269995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015419192166104263,
+      "loss": 2.0446,
+      "step": 270000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00015419036829111057,
+      "loss": 1.8469,
+      "step": 270005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001541888149026661,
+      "loss": 2.0582,
+      "step": 270010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015418726149570974,
+      "loss": 2.219,
+      "step": 270015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015418570807024205,
+      "loss": 2.1197,
+      "step": 270020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015418415462626353,
+      "loss": 2.1973,
+      "step": 270025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001541826011637747,
+      "loss": 2.0024,
+      "step": 270030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015418104768277614,
+      "loss": 2.1264,
+      "step": 270035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001541794941832683,
+      "loss": 2.1557,
+      "step": 270040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015417794066525185,
+      "loss": 2.1192,
+      "step": 270045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015417638712872714,
+      "loss": 2.0253,
+      "step": 270050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015417483357369486,
+      "loss": 1.9789,
+      "step": 270055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015417328000015546,
+      "loss": 2.0297,
+      "step": 270060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015417172640810946,
+      "loss": 2.1199,
+      "step": 270065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015417017279755743,
+      "loss": 2.192,
+      "step": 270070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015416861916849988,
+      "loss": 2.1928,
+      "step": 270075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001541670655209374,
+      "loss": 2.0419,
+      "step": 270080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001541655118548704,
+      "loss": 2.0727,
+      "step": 270085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015416395817029947,
+      "loss": 2.0828,
+      "step": 270090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015416240446722518,
+      "loss": 2.1265,
+      "step": 270095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015416085074564806,
+      "loss": 1.9823,
+      "step": 270100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015415929700556856,
+      "loss": 2.1123,
+      "step": 270105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015415774324698732,
+      "loss": 2.1392,
+      "step": 270110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015415618946990476,
+      "loss": 1.9947,
+      "step": 270115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015415463567432146,
+      "loss": 2.0633,
+      "step": 270120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015415308186023802,
+      "loss": 2.1411,
+      "step": 270125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015415152802765483,
+      "loss": 2.1712,
+      "step": 270130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00015414997417657254,
+      "loss": 1.9907,
+      "step": 270135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001541484203069916,
+      "loss": 2.2117,
+      "step": 270140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001541468664189126,
+      "loss": 2.1531,
+      "step": 270145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015414531251233606,
+      "loss": 1.9954,
+      "step": 270150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015414375858726248,
+      "loss": 1.9882,
+      "step": 270155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015414220464369244,
+      "loss": 2.0949,
+      "step": 270160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001541406506816264,
+      "loss": 2.1557,
+      "step": 270165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015413909670106496,
+      "loss": 2.1436,
+      "step": 270170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001541375427020086,
+      "loss": 2.173,
+      "step": 270175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001541359886844579,
+      "loss": 2.1032,
+      "step": 270180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015413443464841334,
+      "loss": 2.026,
+      "step": 270185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015413288059387548,
+      "loss": 2.0394,
+      "step": 270190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015413132652084487,
+      "loss": 2.1445,
+      "step": 270195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015412977242932198,
+      "loss": 2.2326,
+      "step": 270200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015412821831930744,
+      "loss": 1.9813,
+      "step": 270205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015412666419080165,
+      "loss": 2.0066,
+      "step": 270210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015412511004380525,
+      "loss": 2.1352,
+      "step": 270215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015412355587831873,
+      "loss": 1.974,
+      "step": 270220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00015412200169434258,
+      "loss": 2.1714,
+      "step": 270225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00015412044749187742,
+      "loss": 2.0544,
+      "step": 270230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001541188932709237,
+      "loss": 1.9841,
+      "step": 270235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015411733903148204,
+      "loss": 2.1563,
+      "step": 270240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015411578477355286,
+      "loss": 1.925,
+      "step": 270245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015411423049713677,
+      "loss": 2.04,
+      "step": 270250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015411267620223424,
+      "loss": 2.1004,
+      "step": 270255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001541111218888459,
+      "loss": 2.0809,
+      "step": 270260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001541095675569722,
+      "loss": 2.2631,
+      "step": 270265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015410801320661367,
+      "loss": 2.0463,
+      "step": 270270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015410645883777086,
+      "loss": 2.0818,
+      "step": 270275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001541049044504443,
+      "loss": 2.144,
+      "step": 270280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015410335004463455,
+      "loss": 2.0728,
+      "step": 270285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001541017956203421,
+      "loss": 2.1181,
+      "step": 270290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001541002411775675,
+      "loss": 2.0411,
+      "step": 270295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015409868671631126,
+      "loss": 2.1237,
+      "step": 270300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015409713223657394,
+      "loss": 2.0472,
+      "step": 270305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015409557773835605,
+      "loss": 2.0727,
+      "step": 270310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015409402322165815,
+      "loss": 2.1369,
+      "step": 270315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015409246868648073,
+      "loss": 2.0816,
+      "step": 270320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015409091413282435,
+      "loss": 2.1609,
+      "step": 270325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015408935956068954,
+      "loss": 2.3129,
+      "step": 270330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015408780497007683,
+      "loss": 2.0775,
+      "step": 270335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015408625036098672,
+      "loss": 2.1313,
+      "step": 270340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001540846957334198,
+      "loss": 1.9786,
+      "step": 270345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015408314108737652,
+      "loss": 2.0414,
+      "step": 270350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001540815864228575,
+      "loss": 2.246,
+      "step": 270355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001540800317398632,
+      "loss": 2.1597,
+      "step": 270360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001540784770383942,
+      "loss": 1.9781,
+      "step": 270365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015407692231845098,
+      "loss": 2.0681,
+      "step": 270370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015407536758003416,
+      "loss": 2.0266,
+      "step": 270375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00015407381282314417,
+      "loss": 2.2024,
+      "step": 270380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015407225804778157,
+      "loss": 2.0068,
+      "step": 270385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015407070325394693,
+      "loss": 2.1461,
+      "step": 270390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015406914844164077,
+      "loss": 2.1415,
+      "step": 270395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.0001540675936108636,
+      "loss": 2.1831,
+      "step": 270400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015406603876161597,
+      "loss": 2.0865,
+      "step": 270405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00015406448389389837,
+      "loss": 2.1723,
+      "step": 270410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015406292900771137,
+      "loss": 2.0652,
+      "step": 270415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015406137410305551,
+      "loss": 2.2236,
+      "step": 270420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001540598191799313,
+      "loss": 2.1827,
+      "step": 270425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015405826423833927,
+      "loss": 2.0833,
+      "step": 270430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00015405670927827995,
+      "loss": 2.1695,
+      "step": 270435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015405515429975388,
+      "loss": 2.2528,
+      "step": 270440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001540535993027616,
+      "loss": 2.0014,
+      "step": 270445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015405204428730362,
+      "loss": 2.1519,
+      "step": 270450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015405048925338052,
+      "loss": 1.9433,
+      "step": 270455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015404893420099274,
+      "loss": 1.8497,
+      "step": 270460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015404737913014087,
+      "loss": 1.9716,
+      "step": 270465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015404582404082544,
+      "loss": 1.8258,
+      "step": 270470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000154044268933047,
+      "loss": 2.0576,
+      "step": 270475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015404271380680605,
+      "loss": 2.02,
+      "step": 270480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015404115866210312,
+      "loss": 1.9844,
+      "step": 270485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015403960349893872,
+      "loss": 1.9966,
+      "step": 270490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00015403804831731343,
+      "loss": 2.0724,
+      "step": 270495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001540364931172278,
+      "loss": 2.1185,
+      "step": 270500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001540349378986823,
+      "loss": 2.046,
+      "step": 270505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015403338266167747,
+      "loss": 2.1215,
+      "step": 270510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015403182740621389,
+      "loss": 2.1336,
+      "step": 270515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015403027213229203,
+      "loss": 2.0662,
+      "step": 270520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015402871683991244,
+      "loss": 2.0596,
+      "step": 270525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001540271615290757,
+      "loss": 2.1674,
+      "step": 270530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015402560619978228,
+      "loss": 2.0862,
+      "step": 270535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015402405085203274,
+      "loss": 2.1365,
+      "step": 270540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015402249548582758,
+      "loss": 1.9614,
+      "step": 270545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015402094010116737,
+      "loss": 2.0439,
+      "step": 270550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015401938469805263,
+      "loss": 2.0884,
+      "step": 270555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015401782927648388,
+      "loss": 2.0914,
+      "step": 270560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015401627383646165,
+      "loss": 2.0997,
+      "step": 270565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015401471837798653,
+      "loss": 2.0365,
+      "step": 270570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015401316290105895,
+      "loss": 1.9911,
+      "step": 270575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015401160740567953,
+      "loss": 2.189,
+      "step": 270580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015401005189184875,
+      "loss": 2.0993,
+      "step": 270585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015400849635956716,
+      "loss": 2.3244,
+      "step": 270590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015400694080883529,
+      "loss": 1.8857,
+      "step": 270595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015400538523965365,
+      "loss": 1.9333,
+      "step": 270600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001540038296520228,
+      "loss": 1.9729,
+      "step": 270605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001540022740459433,
+      "loss": 2.1072,
+      "step": 270610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015400071842141562,
+      "loss": 2.0472,
+      "step": 270615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015399916277844032,
+      "loss": 2.096,
+      "step": 270620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015399760711701788,
+      "loss": 2.2921,
+      "step": 270625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001539960514371489,
+      "loss": 2.0265,
+      "step": 270630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001539944957388339,
+      "loss": 2.0611,
+      "step": 270635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015399294002207343,
+      "loss": 1.977,
+      "step": 270640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015399138428686798,
+      "loss": 2.1124,
+      "step": 270645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015398982853321807,
+      "loss": 2.0107,
+      "step": 270650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015398827276112424,
+      "loss": 2.1588,
+      "step": 270655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015398671697058707,
+      "loss": 2.0447,
+      "step": 270660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015398516116160707,
+      "loss": 2.1223,
+      "step": 270665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015398360533418472,
+      "loss": 1.8775,
+      "step": 270670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015398204948832063,
+      "loss": 1.991,
+      "step": 270675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015398049362401527,
+      "loss": 1.9498,
+      "step": 270680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001539789377412692,
+      "loss": 2.0597,
+      "step": 270685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015397738184008293,
+      "loss": 2.0514,
+      "step": 270690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015397582592045704,
+      "loss": 2.038,
+      "step": 270695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015397426998239202,
+      "loss": 2.0581,
+      "step": 270700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015397271402588838,
+      "loss": 1.8371,
+      "step": 270705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001539711580509467,
+      "loss": 2.0674,
+      "step": 270710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015396960205756749,
+      "loss": 2.1368,
+      "step": 270715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001539680460457513,
+      "loss": 2.1633,
+      "step": 270720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015396649001549862,
+      "loss": 2.013,
+      "step": 270725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015396493396681006,
+      "loss": 2.0289,
+      "step": 270730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015396337789968606,
+      "loss": 2.147,
+      "step": 270735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001539618218141272,
+      "loss": 1.9455,
+      "step": 270740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015396026571013399,
+      "loss": 2.1291,
+      "step": 270745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000153958709587707,
+      "loss": 1.8746,
+      "step": 270750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001539571534468467,
+      "loss": 2.0326,
+      "step": 270755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001539555972875537,
+      "loss": 2.1958,
+      "step": 270760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015395404110982844,
+      "loss": 2.0234,
+      "step": 270765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015395248491367153,
+      "loss": 2.1035,
+      "step": 270770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015395092869908348,
+      "loss": 2.0023,
+      "step": 270775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001539493724660648,
+      "loss": 1.995,
+      "step": 270780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015394781621461609,
+      "loss": 2.0994,
+      "step": 270785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015394625994473774,
+      "loss": 1.9968,
+      "step": 270790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015394470365643042,
+      "loss": 2.0338,
+      "step": 270795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015394314734969462,
+      "loss": 2.1405,
+      "step": 270800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00015394159102453085,
+      "loss": 1.7539,
+      "step": 270805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015394003468093966,
+      "loss": 2.0798,
+      "step": 270810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015393847831892156,
+      "loss": 1.9211,
+      "step": 270815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015393692193847706,
+      "loss": 2.1276,
+      "step": 270820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015393536553960679,
+      "loss": 1.9227,
+      "step": 270825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001539338091223112,
+      "loss": 2.0088,
+      "step": 270830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015393225268659085,
+      "loss": 2.1126,
+      "step": 270835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00015393069623244627,
+      "loss": 2.0085,
+      "step": 270840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015392913975987798,
+      "loss": 2.1484,
+      "step": 270845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001539275832688865,
+      "loss": 2.1627,
+      "step": 270850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015392602675947238,
+      "loss": 2.2501,
+      "step": 270855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015392447023163615,
+      "loss": 2.0268,
+      "step": 270860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015392291368537835,
+      "loss": 2.0776,
+      "step": 270865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015392135712069953,
+      "loss": 2.0379,
+      "step": 270870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001539198005376002,
+      "loss": 2.2015,
+      "step": 270875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015391824393608085,
+      "loss": 2.0978,
+      "step": 270880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015391668731614208,
+      "loss": 2.032,
+      "step": 270885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015391513067778437,
+      "loss": 2.1432,
+      "step": 270890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015391357402100829,
+      "loss": 1.9237,
+      "step": 270895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015391201734581432,
+      "loss": 2.012,
+      "step": 270900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015391046065220306,
+      "loss": 2.0108,
+      "step": 270905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000153908903940175,
+      "loss": 2.121,
+      "step": 270910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015390734720973072,
+      "loss": 2.0148,
+      "step": 270915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001539057904608707,
+      "loss": 1.9316,
+      "step": 270920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001539042336935954,
+      "loss": 2.1334,
+      "step": 270925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001539026769079055,
+      "loss": 2.0578,
+      "step": 270930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015390112010380146,
+      "loss": 2.1202,
+      "step": 270935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015389956328128385,
+      "loss": 2.0881,
+      "step": 270940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015389800644035316,
+      "loss": 2.0602,
+      "step": 270945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001538964495810099,
+      "loss": 2.0156,
+      "step": 270950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015389489270325462,
+      "loss": 2.0724,
+      "step": 270955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015389333580708795,
+      "loss": 2.1047,
+      "step": 270960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015389177889251027,
+      "loss": 2.0892,
+      "step": 270965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015389022195952222,
+      "loss": 2.1687,
+      "step": 270970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015388866500812425,
+      "loss": 2.2907,
+      "step": 270975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000153887108038317,
+      "loss": 2.0036,
+      "step": 270980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015388555105010088,
+      "loss": 2.1166,
+      "step": 270985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001538839940434765,
+      "loss": 2.0436,
+      "step": 270990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015388243701844437,
+      "loss": 2.1372,
+      "step": 270995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000153880879975005,
+      "loss": 2.0319,
+      "step": 271000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015387932291315893,
+      "loss": 2.1576,
+      "step": 271005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015387776583290677,
+      "loss": 2.0651,
+      "step": 271010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015387620873424896,
+      "loss": 2.0478,
+      "step": 271015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015387465161718607,
+      "loss": 2.1167,
+      "step": 271020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015387309448171857,
+      "loss": 2.1268,
+      "step": 271025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001538715373278471,
+      "loss": 2.0071,
+      "step": 271030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015386998015557213,
+      "loss": 2.0778,
+      "step": 271035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015386842296489416,
+      "loss": 2.2082,
+      "step": 271040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015386686575581378,
+      "loss": 2.2039,
+      "step": 271045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001538653085283315,
+      "loss": 2.0868,
+      "step": 271050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001538637512824479,
+      "loss": 2.2781,
+      "step": 271055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001538621940181634,
+      "loss": 2.0046,
+      "step": 271060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015386063673547863,
+      "loss": 2.0727,
+      "step": 271065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015385907943439407,
+      "loss": 2.1836,
+      "step": 271070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001538575221149103,
+      "loss": 2.0851,
+      "step": 271075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015385596477702782,
+      "loss": 2.2361,
+      "step": 271080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015385440742074713,
+      "loss": 2.0759,
+      "step": 271085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015385285004606883,
+      "loss": 2.0823,
+      "step": 271090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015385129265299343,
+      "loss": 2.1352,
+      "step": 271095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015384973524152143,
+      "loss": 2.1434,
+      "step": 271100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001538481778116534,
+      "loss": 1.9453,
+      "step": 271105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015384662036338983,
+      "loss": 1.9951,
+      "step": 271110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015384506289673125,
+      "loss": 2.1826,
+      "step": 271115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001538435054116783,
+      "loss": 1.9744,
+      "step": 271120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015384194790823142,
+      "loss": 2.0694,
+      "step": 271125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001538403903863911,
+      "loss": 2.2213,
+      "step": 271130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015383883284615792,
+      "loss": 1.9828,
+      "step": 271135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001538372752875325,
+      "loss": 2.0069,
+      "step": 271140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015383571771051524,
+      "loss": 2.2152,
+      "step": 271145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001538341601151067,
+      "loss": 2.0389,
+      "step": 271150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015383260250130749,
+      "loss": 2.1477,
+      "step": 271155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015383104486911802,
+      "loss": 2.1456,
+      "step": 271160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015382948721853894,
+      "loss": 1.9876,
+      "step": 271165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015382792954957071,
+      "loss": 2.0184,
+      "step": 271170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001538263718622139,
+      "loss": 2.0788,
+      "step": 271175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015382481415646902,
+      "loss": 2.0965,
+      "step": 271180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015382325643233658,
+      "loss": 2.1086,
+      "step": 271185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015382169868981716,
+      "loss": 2.0036,
+      "step": 271190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015382014092891127,
+      "loss": 1.9092,
+      "step": 271195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015381858314961945,
+      "loss": 2.097,
+      "step": 271200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015381702535194225,
+      "loss": 2.1431,
+      "step": 271205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015381546753588015,
+      "loss": 2.0329,
+      "step": 271210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015381390970143372,
+      "loss": 2.2347,
+      "step": 271215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015381235184860342,
+      "loss": 2.191,
+      "step": 271220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015381079397738993,
+      "loss": 2.145,
+      "step": 271225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015380923608779366,
+      "loss": 1.9249,
+      "step": 271230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015380767817981518,
+      "loss": 2.1741,
+      "step": 271235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015380612025345505,
+      "loss": 2.1847,
+      "step": 271240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001538045623087137,
+      "loss": 2.0667,
+      "step": 271245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001538030043455918,
+      "loss": 1.9737,
+      "step": 271250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015380144636408983,
+      "loss": 2.0031,
+      "step": 271255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015379988836420827,
+      "loss": 2.0637,
+      "step": 271260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001537983303459477,
+      "loss": 2.1624,
+      "step": 271265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015379677230930864,
+      "loss": 1.9853,
+      "step": 271270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015379521425429164,
+      "loss": 2.0194,
+      "step": 271275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015379365618089724,
+      "loss": 2.0873,
+      "step": 271280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001537920980891259,
+      "loss": 1.9406,
+      "step": 271285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015379053997897826,
+      "loss": 2.1838,
+      "step": 271290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015378898185045475,
+      "loss": 2.23,
+      "step": 271295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015378742370355597,
+      "loss": 2.0272,
+      "step": 271300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015378586553828244,
+      "loss": 2.0214,
+      "step": 271305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015378430735463465,
+      "loss": 1.9843,
+      "step": 271310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001537827491526132,
+      "loss": 2.0759,
+      "step": 271315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015378119093221857,
+      "loss": 2.0327,
+      "step": 271320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001537796326934513,
+      "loss": 2.0819,
+      "step": 271325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015377807443631194,
+      "loss": 2.122,
+      "step": 271330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000153776516160801,
+      "loss": 2.1867,
+      "step": 271335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015377495786691908,
+      "loss": 2.167,
+      "step": 271340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001537733995546666,
+      "loss": 1.9258,
+      "step": 271345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015377184122404417,
+      "loss": 2.1293,
+      "step": 271350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001537702828750523,
+      "loss": 1.9768,
+      "step": 271355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015376872450769153,
+      "loss": 2.0183,
+      "step": 271360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001537671661219624,
+      "loss": 1.9972,
+      "step": 271365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015376560771786547,
+      "loss": 2.2574,
+      "step": 271370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015376404929540115,
+      "loss": 2.0386,
+      "step": 271375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015376249085457012,
+      "loss": 2.1339,
+      "step": 271380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.875,
+      "learning_rate": 0.0001537609323953728,
+      "loss": 2.0234,
+      "step": 271385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001537593739178098,
+      "loss": 2.1668,
+      "step": 271390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015375781542188163,
+      "loss": 2.166,
+      "step": 271395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001537562569075888,
+      "loss": 2.0048,
+      "step": 271400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015375469837493184,
+      "loss": 1.9352,
+      "step": 271405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001537531398239113,
+      "loss": 2.0344,
+      "step": 271410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015375158125452776,
+      "loss": 2.0019,
+      "step": 271415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015375002266678167,
+      "loss": 2.06,
+      "step": 271420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015374846406067362,
+      "loss": 1.8287,
+      "step": 271425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015374690543620408,
+      "loss": 2.0749,
+      "step": 271430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015374534679337364,
+      "loss": 2.1871,
+      "step": 271435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015374378813218283,
+      "loss": 1.935,
+      "step": 271440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015374222945263214,
+      "loss": 2.1034,
+      "step": 271445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015374067075472217,
+      "loss": 2.0998,
+      "step": 271450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001537391120384534,
+      "loss": 2.1003,
+      "step": 271455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015373755330382632,
+      "loss": 1.8446,
+      "step": 271460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001537359945508416,
+      "loss": 2.0424,
+      "step": 271465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015373443577949963,
+      "loss": 2.0378,
+      "step": 271470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00015373287698980102,
+      "loss": 1.9598,
+      "step": 271475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001537313181817463,
+      "loss": 2.028,
+      "step": 271480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015372975935533595,
+      "loss": 2.0477,
+      "step": 271485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001537282005105706,
+      "loss": 1.9392,
+      "step": 271490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015372664164745067,
+      "loss": 1.9699,
+      "step": 271495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015372508276597678,
+      "loss": 2.1736,
+      "step": 271500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001537235238661494,
+      "loss": 1.9612,
+      "step": 271505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015372196494796908,
+      "loss": 2.1755,
+      "step": 271510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015372040601143637,
+      "loss": 2.1288,
+      "step": 271515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001537188470565518,
+      "loss": 1.9847,
+      "step": 271520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015371728808331593,
+      "loss": 2.1008,
+      "step": 271525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015371572909172928,
+      "loss": 2.1472,
+      "step": 271530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001537141700817923,
+      "loss": 2.0608,
+      "step": 271535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015371261105350556,
+      "loss": 2.0856,
+      "step": 271540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015371105200686968,
+      "loss": 2.0507,
+      "step": 271545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015370949294188512,
+      "loss": 1.9956,
+      "step": 271550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015370793385855242,
+      "loss": 2.2151,
+      "step": 271555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001537063747568721,
+      "loss": 2.1425,
+      "step": 271560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015370481563684472,
+      "loss": 1.8928,
+      "step": 271565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015370325649847078,
+      "loss": 1.9993,
+      "step": 271570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015370169734175087,
+      "loss": 2.2505,
+      "step": 271575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015370013816668547,
+      "loss": 2.0897,
+      "step": 271580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015369857897327513,
+      "loss": 2.1036,
+      "step": 271585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001536970197615204,
+      "loss": 2.2242,
+      "step": 271590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015369546053142175,
+      "loss": 1.9556,
+      "step": 271595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001536939012829798,
+      "loss": 2.0134,
+      "step": 271600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000153692342016195,
+      "loss": 1.8041,
+      "step": 271605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015369078273106795,
+      "loss": 2.0642,
+      "step": 271610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015368922342759916,
+      "loss": 1.989,
+      "step": 271615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001536876641057891,
+      "loss": 2.1974,
+      "step": 271620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015368610476563843,
+      "loss": 1.9757,
+      "step": 271625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001536845454071476,
+      "loss": 2.2546,
+      "step": 271630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015368298603031714,
+      "loss": 2.1979,
+      "step": 271635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001536814266351476,
+      "loss": 2.0615,
+      "step": 271640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015367986722163952,
+      "loss": 1.968,
+      "step": 271645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001536783077897934,
+      "loss": 2.0773,
+      "step": 271650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.90625,
+      "learning_rate": 0.0001536767483396098,
+      "loss": 1.9834,
+      "step": 271655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015367518887108927,
+      "loss": 2.0313,
+      "step": 271660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015367362938423234,
+      "loss": 2.0925,
+      "step": 271665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001536720698790395,
+      "loss": 2.0522,
+      "step": 271670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001536705103555113,
+      "loss": 1.998,
+      "step": 271675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015366895081364828,
+      "loss": 1.9459,
+      "step": 271680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000153667391253451,
+      "loss": 2.0547,
+      "step": 271685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015366583167491995,
+      "loss": 2.1673,
+      "step": 271690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015366427207805567,
+      "loss": 2.1043,
+      "step": 271695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001536627124628587,
+      "loss": 1.9913,
+      "step": 271700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015366115282932955,
+      "loss": 2.0317,
+      "step": 271705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015365959317746884,
+      "loss": 2.1126,
+      "step": 271710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015365803350727702,
+      "loss": 1.9175,
+      "step": 271715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001536564738187546,
+      "loss": 1.9411,
+      "step": 271720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015365491411190217,
+      "loss": 2.1411,
+      "step": 271725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015365335438672031,
+      "loss": 1.9896,
+      "step": 271730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015365179464320945,
+      "loss": 1.8543,
+      "step": 271735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015365023488137015,
+      "loss": 1.9157,
+      "step": 271740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015364867510120296,
+      "loss": 2.1583,
+      "step": 271745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001536471153027084,
+      "loss": 2.1327,
+      "step": 271750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015364555548588705,
+      "loss": 2.1239,
+      "step": 271755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001536439956507394,
+      "loss": 2.1411,
+      "step": 271760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015364243579726593,
+      "loss": 2.0262,
+      "step": 271765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015364087592546728,
+      "loss": 2.1293,
+      "step": 271770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015363931603534393,
+      "loss": 2.267,
+      "step": 271775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015363775612689643,
+      "loss": 1.9616,
+      "step": 271780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015363619620012528,
+      "loss": 2.1137,
+      "step": 271785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015363463625503102,
+      "loss": 2.2921,
+      "step": 271790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015363307629161422,
+      "loss": 1.9236,
+      "step": 271795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015363151630987538,
+      "loss": 2.0413,
+      "step": 271800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015362995630981504,
+      "loss": 1.8565,
+      "step": 271805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015362839629143372,
+      "loss": 1.9988,
+      "step": 271810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000153626836254732,
+      "loss": 1.9986,
+      "step": 271815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015362527619971035,
+      "loss": 1.8511,
+      "step": 271820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015362371612636936,
+      "loss": 2.2359,
+      "step": 271825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015362215603470953,
+      "loss": 2.2273,
+      "step": 271830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015362059592473136,
+      "loss": 1.9267,
+      "step": 271835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015361903579643547,
+      "loss": 1.8895,
+      "step": 271840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015361747564982233,
+      "loss": 1.9864,
+      "step": 271845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001536159154848925,
+      "loss": 2.0914,
+      "step": 271850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015361435530164647,
+      "loss": 1.862,
+      "step": 271855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001536127951000848,
+      "loss": 2.2046,
+      "step": 271860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001536112348802081,
+      "loss": 2.027,
+      "step": 271865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015360967464201676,
+      "loss": 1.8713,
+      "step": 271870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015360811438551138,
+      "loss": 2.0504,
+      "step": 271875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015360655411069253,
+      "loss": 2.0621,
+      "step": 271880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001536049938175607,
+      "loss": 2.0486,
+      "step": 271885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001536034335061164,
+      "loss": 2.2444,
+      "step": 271890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015360187317636025,
+      "loss": 2.1318,
+      "step": 271895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001536003128282927,
+      "loss": 1.9641,
+      "step": 271900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001535987524619143,
+      "loss": 2.1494,
+      "step": 271905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001535971920772256,
+      "loss": 2.1345,
+      "step": 271910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015359563167422714,
+      "loss": 2.1501,
+      "step": 271915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015359407125291942,
+      "loss": 1.8498,
+      "step": 271920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000153592510813303,
+      "loss": 2.0959,
+      "step": 271925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015359095035537843,
+      "loss": 2.0282,
+      "step": 271930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001535893898791462,
+      "loss": 2.0747,
+      "step": 271935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015358782938460684,
+      "loss": 2.1862,
+      "step": 271940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015358626887176095,
+      "loss": 2.1818,
+      "step": 271945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.000153584708340609,
+      "loss": 2.0081,
+      "step": 271950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015358314779115152,
+      "loss": 2.0881,
+      "step": 271955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001535815872233891,
+      "loss": 2.032,
+      "step": 271960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015358002663732223,
+      "loss": 2.2539,
+      "step": 271965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015357846603295143,
+      "loss": 2.0549,
+      "step": 271970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015357690541027726,
+      "loss": 2.0365,
+      "step": 271975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001535753447693003,
+      "loss": 2.0147,
+      "step": 271980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015357378411002097,
+      "loss": 1.9719,
+      "step": 271985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015357222343243988,
+      "loss": 2.0942,
+      "step": 271990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015357066273655754,
+      "loss": 2.1721,
+      "step": 271995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00015356910202237448,
+      "loss": 1.9698,
+      "step": 272000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001535675412898913,
+      "loss": 2.1252,
+      "step": 272005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015356598053910842,
+      "loss": 2.0726,
+      "step": 272010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015356441977002646,
+      "loss": 2.053,
+      "step": 272015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015356285898264588,
+      "loss": 2.0515,
+      "step": 272020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001535612981769673,
+      "loss": 1.9292,
+      "step": 272025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015355973735299122,
+      "loss": 2.1353,
+      "step": 272030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001535581765107181,
+      "loss": 2.0879,
+      "step": 272035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001535566156501486,
+      "loss": 2.1982,
+      "step": 272040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015355505477128315,
+      "loss": 2.1265,
+      "step": 272045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015355349387412233,
+      "loss": 1.9751,
+      "step": 272050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015355193295866666,
+      "loss": 2.0317,
+      "step": 272055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001535503720249167,
+      "loss": 2.3638,
+      "step": 272060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015354881107287294,
+      "loss": 2.0311,
+      "step": 272065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015354725010253596,
+      "loss": 1.9983,
+      "step": 272070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015354568911390623,
+      "loss": 2.0935,
+      "step": 272075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015354412810698436,
+      "loss": 2.1418,
+      "step": 272080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001535425670817708,
+      "loss": 1.8783,
+      "step": 272085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015354100603826617,
+      "loss": 1.9174,
+      "step": 272090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015353944497647096,
+      "loss": 2.0569,
+      "step": 272095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001535378838963857,
+      "loss": 2.2237,
+      "step": 272100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015353632279801088,
+      "loss": 2.0224,
+      "step": 272105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015353476168134713,
+      "loss": 2.1088,
+      "step": 272110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015353320054639492,
+      "loss": 1.9745,
+      "step": 272115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015353163939315481,
+      "loss": 2.0795,
+      "step": 272120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015353007822162734,
+      "loss": 2.157,
+      "step": 272125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015352851703181298,
+      "loss": 1.9446,
+      "step": 272130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015352695582371233,
+      "loss": 2.0713,
+      "step": 272135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001535253945973259,
+      "loss": 2.0237,
+      "step": 272140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015352383335265422,
+      "loss": 2.1942,
+      "step": 272145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015352227208969782,
+      "loss": 2.1421,
+      "step": 272150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015352071080845724,
+      "loss": 2.0606,
+      "step": 272155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015351914950893306,
+      "loss": 2.2371,
+      "step": 272160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015351758819112573,
+      "loss": 1.8852,
+      "step": 272165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015351602685503583,
+      "loss": 1.9728,
+      "step": 272170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00015351446550066386,
+      "loss": 1.9677,
+      "step": 272175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015351290412801038,
+      "loss": 2.1432,
+      "step": 272180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015351134273707597,
+      "loss": 1.9937,
+      "step": 272185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015350978132786108,
+      "loss": 2.0818,
+      "step": 272190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001535082199003663,
+      "loss": 2.016,
+      "step": 272195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015350665845459212,
+      "loss": 2.1909,
+      "step": 272200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015350509699053907,
+      "loss": 2.0388,
+      "step": 272205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015350353550820775,
+      "loss": 2.3006,
+      "step": 272210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015350197400759865,
+      "loss": 1.8512,
+      "step": 272215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015350041248871229,
+      "loss": 2.1707,
+      "step": 272220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001534988509515492,
+      "loss": 2.0044,
+      "step": 272225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015349728939610997,
+      "loss": 1.8538,
+      "step": 272230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001534957278223951,
+      "loss": 1.9993,
+      "step": 272235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015349416623040507,
+      "loss": 2.1678,
+      "step": 272240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001534926046201405,
+      "loss": 2.0453,
+      "step": 272245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015349104299160188,
+      "loss": 2.1334,
+      "step": 272250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015348948134478974,
+      "loss": 1.7482,
+      "step": 272255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015348791967970463,
+      "loss": 2.1939,
+      "step": 272260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015348635799634707,
+      "loss": 2.1145,
+      "step": 272265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015348479629471765,
+      "loss": 2.0622,
+      "step": 272270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001534832345748168,
+      "loss": 2.0904,
+      "step": 272275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015348167283664514,
+      "loss": 2.0705,
+      "step": 272280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015348011108020314,
+      "loss": 2.2281,
+      "step": 272285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015347854930549133,
+      "loss": 2.1454,
+      "step": 272290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015347698751251034,
+      "loss": 2.1752,
+      "step": 272295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015347542570126062,
+      "loss": 1.9426,
+      "step": 272300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015347386387174275,
+      "loss": 2.2019,
+      "step": 272305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001534723020239572,
+      "loss": 1.9757,
+      "step": 272310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015347074015790456,
+      "loss": 2.2331,
+      "step": 272315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015346917827358536,
+      "loss": 2.0528,
+      "step": 272320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001534676163710001,
+      "loss": 1.916,
+      "step": 272325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015346605445014936,
+      "loss": 2.1542,
+      "step": 272330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001534644925110336,
+      "loss": 2.1193,
+      "step": 272335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001534629305536534,
+      "loss": 1.904,
+      "step": 272340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015346136857800934,
+      "loss": 1.959,
+      "step": 272345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015345980658410189,
+      "loss": 2.0468,
+      "step": 272350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015345824457193156,
+      "loss": 2.075,
+      "step": 272355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015345668254149896,
+      "loss": 1.9146,
+      "step": 272360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001534551204928046,
+      "loss": 1.9544,
+      "step": 272365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015345355842584897,
+      "loss": 2.017,
+      "step": 272370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015345199634063264,
+      "loss": 2.137,
+      "step": 272375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015345043423715615,
+      "loss": 2.1435,
+      "step": 272380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015344887211542004,
+      "loss": 2.149,
+      "step": 272385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015344730997542483,
+      "loss": 2.0939,
+      "step": 272390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000153445747817171,
+      "loss": 2.1542,
+      "step": 272395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015344418564065915,
+      "loss": 2.2022,
+      "step": 272400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001534426234458898,
+      "loss": 2.1173,
+      "step": 272405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015344106123286352,
+      "loss": 2.0662,
+      "step": 272410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015343949900158076,
+      "loss": 2.1339,
+      "step": 272415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015343793675204212,
+      "loss": 2.135,
+      "step": 272420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015343637448424807,
+      "loss": 2.0223,
+      "step": 272425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015343481219819922,
+      "loss": 1.9025,
+      "step": 272430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001534332498938961,
+      "loss": 2.0986,
+      "step": 272435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015343168757133915,
+      "loss": 1.9905,
+      "step": 272440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015343012523052902,
+      "loss": 2.156,
+      "step": 272445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015342856287146615,
+      "loss": 1.9648,
+      "step": 272450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015342700049415114,
+      "loss": 2.1476,
+      "step": 272455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.00015342543809858448,
+      "loss": 2.0217,
+      "step": 272460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015342387568476672,
+      "loss": 1.9998,
+      "step": 272465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001534223132526984,
+      "loss": 2.3569,
+      "step": 272470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015342075080238004,
+      "loss": 1.9658,
+      "step": 272475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001534191883338122,
+      "loss": 2.1808,
+      "step": 272480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001534176258469954,
+      "loss": 2.1042,
+      "step": 272485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015341606334193017,
+      "loss": 2.2853,
+      "step": 272490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015341450081861704,
+      "loss": 1.9559,
+      "step": 272495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015341293827705654,
+      "loss": 2.0791,
+      "step": 272500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001534113757172492,
+      "loss": 1.9284,
+      "step": 272505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001534098131391956,
+      "loss": 2.2151,
+      "step": 272510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001534082505428962,
+      "loss": 2.1994,
+      "step": 272515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001534066879283516,
+      "loss": 2.0784,
+      "step": 272520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001534051252955623,
+      "loss": 2.1275,
+      "step": 272525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015340356264452885,
+      "loss": 1.9924,
+      "step": 272530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015340199997525175,
+      "loss": 2.0603,
+      "step": 272535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015340043728773157,
+      "loss": 2.1161,
+      "step": 272540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015339887458196884,
+      "loss": 2.2499,
+      "step": 272545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001533973118579641,
+      "loss": 2.1139,
+      "step": 272550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015339574911571783,
+      "loss": 2.027,
+      "step": 272555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015339418635523062,
+      "loss": 2.1014,
+      "step": 272560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015339262357650299,
+      "loss": 2.1446,
+      "step": 272565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015339106077953552,
+      "loss": 2.0395,
+      "step": 272570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015338949796432863,
+      "loss": 1.9609,
+      "step": 272575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015338793513088294,
+      "loss": 2.1069,
+      "step": 272580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001533863722791989,
+      "loss": 1.8513,
+      "step": 272585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001533848094092772,
+      "loss": 2.167,
+      "step": 272590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001533832465211183,
+      "loss": 2.0363,
+      "step": 272595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015338168361472265,
+      "loss": 2.0234,
+      "step": 272600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015338012069009084,
+      "loss": 2.2624,
+      "step": 272605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015337855774722343,
+      "loss": 2.133,
+      "step": 272610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015337699478612095,
+      "loss": 2.0091,
+      "step": 272615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015337543180678393,
+      "loss": 2.0664,
+      "step": 272620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015337386880921288,
+      "loss": 1.9083,
+      "step": 272625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015337230579340832,
+      "loss": 2.143,
+      "step": 272630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015337074275937086,
+      "loss": 2.0859,
+      "step": 272635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015336917970710094,
+      "loss": 2.0453,
+      "step": 272640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001533676166365992,
+      "loss": 2.1147,
+      "step": 272645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015336605354786603,
+      "loss": 2.1414,
+      "step": 272650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015336449044090213,
+      "loss": 2.1339,
+      "step": 272655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001533629273157079,
+      "loss": 2.0929,
+      "step": 272660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015336136417228396,
+      "loss": 2.1079,
+      "step": 272665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015335980101063078,
+      "loss": 2.121,
+      "step": 272670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015335823783074894,
+      "loss": 2.2212,
+      "step": 272675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015335667463263895,
+      "loss": 2.0827,
+      "step": 272680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015335511141630138,
+      "loss": 1.9945,
+      "step": 272685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015335354818173672,
+      "loss": 2.0439,
+      "step": 272690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015335198492894548,
+      "loss": 1.9545,
+      "step": 272695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001533504216579283,
+      "loss": 2.2506,
+      "step": 272700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015334885836868563,
+      "loss": 2.1215,
+      "step": 272705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015334729506121802,
+      "loss": 2.1981,
+      "step": 272710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000153345731735526,
+      "loss": 2.0584,
+      "step": 272715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001533441683916101,
+      "loss": 1.9352,
+      "step": 272720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015334260502947086,
+      "loss": 1.9093,
+      "step": 272725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015334104164910884,
+      "loss": 2.1715,
+      "step": 272730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015333947825052455,
+      "loss": 2.1606,
+      "step": 272735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015333791483371854,
+      "loss": 2.1092,
+      "step": 272740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015333635139869127,
+      "loss": 2.0575,
+      "step": 272745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001533347879454434,
+      "loss": 2.152,
+      "step": 272750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001533332244739754,
+      "loss": 1.9305,
+      "step": 272755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015333166098428777,
+      "loss": 2.0719,
+      "step": 272760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015333009747638108,
+      "loss": 2.3572,
+      "step": 272765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015332853395025583,
+      "loss": 2.0854,
+      "step": 272770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015332697040591265,
+      "loss": 2.01,
+      "step": 272775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000153325406843352,
+      "loss": 2.0267,
+      "step": 272780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001533238432625744,
+      "loss": 1.9693,
+      "step": 272785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015332227966358044,
+      "loss": 2.0079,
+      "step": 272790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015332071604637058,
+      "loss": 2.0706,
+      "step": 272795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015331915241094542,
+      "loss": 2.0675,
+      "step": 272800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015331758875730545,
+      "loss": 2.2974,
+      "step": 272805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015331602508545123,
+      "loss": 2.1753,
+      "step": 272810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015331446139538332,
+      "loss": 2.0592,
+      "step": 272815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001533128976871022,
+      "loss": 2.0955,
+      "step": 272820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015331133396060843,
+      "loss": 1.9793,
+      "step": 272825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015330977021590255,
+      "loss": 2.0668,
+      "step": 272830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015330820645298508,
+      "loss": 2.0357,
+      "step": 272835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015330664267185654,
+      "loss": 2.2015,
+      "step": 272840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015330507887251752,
+      "loss": 2.0074,
+      "step": 272845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001533035150549685,
+      "loss": 2.0564,
+      "step": 272850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015330195121921004,
+      "loss": 2.1819,
+      "step": 272855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015330038736524262,
+      "loss": 2.0254,
+      "step": 272860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001532988234930669,
+      "loss": 2.0462,
+      "step": 272865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015329725960268328,
+      "loss": 1.9215,
+      "step": 272870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015329569569409238,
+      "loss": 1.9649,
+      "step": 272875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015329413176729467,
+      "loss": 1.9705,
+      "step": 272880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015329256782229073,
+      "loss": 2.1825,
+      "step": 272885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015329100385908114,
+      "loss": 2.0782,
+      "step": 272890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001532894398776663,
+      "loss": 2.1568,
+      "step": 272895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015328787587804685,
+      "loss": 1.9258,
+      "step": 272900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015328631186022326,
+      "loss": 1.9747,
+      "step": 272905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015328474782419615,
+      "loss": 1.9071,
+      "step": 272910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015328318376996598,
+      "loss": 2.2019,
+      "step": 272915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015328161969753332,
+      "loss": 1.9405,
+      "step": 272920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001532800556068987,
+      "loss": 2.0032,
+      "step": 272925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015327849149806262,
+      "loss": 1.9593,
+      "step": 272930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015327692737102564,
+      "loss": 2.0927,
+      "step": 272935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001532753632257883,
+      "loss": 1.9162,
+      "step": 272940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015327379906235117,
+      "loss": 1.9818,
+      "step": 272945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015327223488071468,
+      "loss": 2.2628,
+      "step": 272950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001532706706808795,
+      "loss": 2.1458,
+      "step": 272955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000153269106462846,
+      "loss": 2.037,
+      "step": 272960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001532675422266149,
+      "loss": 2.2756,
+      "step": 272965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001532659779721866,
+      "loss": 1.9916,
+      "step": 272970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015326441369956166,
+      "loss": 2.1008,
+      "step": 272975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001532628494087407,
+      "loss": 2.0901,
+      "step": 272980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001532612850997241,
+      "loss": 2.2028,
+      "step": 272985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015325972077251252,
+      "loss": 1.9828,
+      "step": 272990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001532581564271064,
+      "loss": 1.9996,
+      "step": 272995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001532565920635064,
+      "loss": 2.1985,
+      "step": 273000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015325502768171297,
+      "loss": 1.9649,
+      "step": 273005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015325346328172663,
+      "loss": 2.0741,
+      "step": 273010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015325189886354794,
+      "loss": 1.9899,
+      "step": 273015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015325033442717745,
+      "loss": 2.0806,
+      "step": 273020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001532487699726157,
+      "loss": 2.0893,
+      "step": 273025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015324720549986318,
+      "loss": 2.083,
+      "step": 273030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015324564100892046,
+      "loss": 1.9232,
+      "step": 273035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015324407649978806,
+      "loss": 2.2101,
+      "step": 273040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001532425119724665,
+      "loss": 2.0212,
+      "step": 273045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015324094742695635,
+      "loss": 1.8805,
+      "step": 273050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001532393828632581,
+      "loss": 1.984,
+      "step": 273055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015323781828137235,
+      "loss": 1.8424,
+      "step": 273060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015323625368129958,
+      "loss": 2.0518,
+      "step": 273065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015323468906304035,
+      "loss": 2.2186,
+      "step": 273070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015323312442659515,
+      "loss": 2.0378,
+      "step": 273075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001532315597719646,
+      "loss": 2.1519,
+      "step": 273080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015322999509914914,
+      "loss": 2.2036,
+      "step": 273085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015322843040814938,
+      "loss": 1.9639,
+      "step": 273090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015322686569896581,
+      "loss": 2.1421,
+      "step": 273095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015322530097159899,
+      "loss": 2.0011,
+      "step": 273100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001532237362260494,
+      "loss": 2.0583,
+      "step": 273105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015322217146231766,
+      "loss": 2.1288,
+      "step": 273110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015322060668040424,
+      "loss": 1.9587,
+      "step": 273115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001532190418803097,
+      "loss": 2.0937,
+      "step": 273120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015321747706203458,
+      "loss": 2.1543,
+      "step": 273125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001532159122255794,
+      "loss": 2.2168,
+      "step": 273130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015321434737094469,
+      "loss": 2.1393,
+      "step": 273135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000153212782498131,
+      "loss": 2.2106,
+      "step": 273140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015321121760713885,
+      "loss": 1.9591,
+      "step": 273145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001532096526979688,
+      "loss": 2.1285,
+      "step": 273150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015320808777062134,
+      "loss": 2.0013,
+      "step": 273155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001532065228250971,
+      "loss": 1.9855,
+      "step": 273160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015320495786139645,
+      "loss": 2.0873,
+      "step": 273165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001532033928795201,
+      "loss": 2.2037,
+      "step": 273170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015320182787946846,
+      "loss": 2.1385,
+      "step": 273175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015320026286124211,
+      "loss": 1.9492,
+      "step": 273180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015319869782484162,
+      "loss": 1.9581,
+      "step": 273185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00015319713277026748,
+      "loss": 2.0032,
+      "step": 273190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015319556769752022,
+      "loss": 2.0147,
+      "step": 273195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015319400260660037,
+      "loss": 2.0351,
+      "step": 273200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015319243749750854,
+      "loss": 1.9984,
+      "step": 273205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015319087237024518,
+      "loss": 2.0502,
+      "step": 273210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015318930722481083,
+      "loss": 1.8222,
+      "step": 273215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015318774206120607,
+      "loss": 2.1495,
+      "step": 273220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001531861768794314,
+      "loss": 1.9167,
+      "step": 273225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015318461167948742,
+      "loss": 2.1641,
+      "step": 273230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015318304646137456,
+      "loss": 2.1552,
+      "step": 273235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001531814812250934,
+      "loss": 2.2225,
+      "step": 273240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001531799159706445,
+      "loss": 2.1461,
+      "step": 273245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001531783506980284,
+      "loss": 2.0667,
+      "step": 273250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015317678540724557,
+      "loss": 2.1927,
+      "step": 273255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001531752200982966,
+      "loss": 2.0091,
+      "step": 273260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000153173654771182,
+      "loss": 2.0501,
+      "step": 273265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015317208942590234,
+      "loss": 2.0555,
+      "step": 273270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015317052406245814,
+      "loss": 1.9991,
+      "step": 273275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001531689586808499,
+      "loss": 2.1565,
+      "step": 273280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015316739328107817,
+      "loss": 2.0525,
+      "step": 273285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015316582786314352,
+      "loss": 2.1674,
+      "step": 273290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001531642624270464,
+      "loss": 2.0914,
+      "step": 273295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001531626969727875,
+      "loss": 2.1347,
+      "step": 273300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015316113150036717,
+      "loss": 2.0168,
+      "step": 273305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015315956600978607,
+      "loss": 2.0375,
+      "step": 273310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001531580005010447,
+      "loss": 2.0672,
+      "step": 273315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001531564349741436,
+      "loss": 2.1063,
+      "step": 273320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001531548694290833,
+      "loss": 2.1759,
+      "step": 273325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015315330386586433,
+      "loss": 1.9206,
+      "step": 273330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001531517382844872,
+      "loss": 2.2278,
+      "step": 273335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001531501726849525,
+      "loss": 2.2017,
+      "step": 273340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015314860706726072,
+      "loss": 1.9975,
+      "step": 273345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015314704143141243,
+      "loss": 1.9735,
+      "step": 273350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015314547577740815,
+      "loss": 2.1451,
+      "step": 273355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015314391010524836,
+      "loss": 2.1461,
+      "step": 273360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001531423444149337,
+      "loss": 2.0379,
+      "step": 273365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015314077870646467,
+      "loss": 2.0412,
+      "step": 273370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015313921297984174,
+      "loss": 2.0121,
+      "step": 273375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001531376472350655,
+      "loss": 2.0818,
+      "step": 273380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001531360814721365,
+      "loss": 2.2183,
+      "step": 273385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001531345156910552,
+      "loss": 1.868,
+      "step": 273390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015313294989182223,
+      "loss": 1.8938,
+      "step": 273395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001531313840744381,
+      "loss": 2.029,
+      "step": 273400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015312981823890323,
+      "loss": 1.886,
+      "step": 273405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015312825238521834,
+      "loss": 2.1746,
+      "step": 273410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015312668651338383,
+      "loss": 2.026,
+      "step": 273415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015312512062340031,
+      "loss": 2.0524,
+      "step": 273420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015312355471526828,
+      "loss": 2.2201,
+      "step": 273425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015312198878898826,
+      "loss": 2.0806,
+      "step": 273430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015312042284456085,
+      "loss": 2.1977,
+      "step": 273435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015311885688198647,
+      "loss": 1.956,
+      "step": 273440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015311729090126578,
+      "loss": 2.1284,
+      "step": 273445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015311572490239926,
+      "loss": 2.1882,
+      "step": 273450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.28125,
+      "learning_rate": 0.00015311415888538744,
+      "loss": 2.2329,
+      "step": 273455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015311259285023088,
+      "loss": 2.1145,
+      "step": 273460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015311102679693006,
+      "loss": 2.1623,
+      "step": 273465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015310946072548553,
+      "loss": 2.1435,
+      "step": 273470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015310789463589787,
+      "loss": 2.0886,
+      "step": 273475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001531063285281676,
+      "loss": 2.1458,
+      "step": 273480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015310476240229526,
+      "loss": 2.1289,
+      "step": 273485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001531031962582813,
+      "loss": 2.1119,
+      "step": 273490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015310163009612637,
+      "loss": 2.0798,
+      "step": 273495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000153100063915831,
+      "loss": 2.0486,
+      "step": 273500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015309849771739565,
+      "loss": 1.8828,
+      "step": 273505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015309693150082087,
+      "loss": 1.9562,
+      "step": 273510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015309536526610725,
+      "loss": 2.1796,
+      "step": 273515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015309379901325528,
+      "loss": 2.2117,
+      "step": 273520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015309223274226547,
+      "loss": 1.9687,
+      "step": 273525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001530906664531384,
+      "loss": 2.2133,
+      "step": 273530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015308910014587464,
+      "loss": 2.1353,
+      "step": 273535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015308753382047466,
+      "loss": 1.8261,
+      "step": 273540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000153085967476939,
+      "loss": 1.9258,
+      "step": 273545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001530844011152682,
+      "loss": 2.1056,
+      "step": 273550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015308283473546282,
+      "loss": 2.1858,
+      "step": 273555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001530812683375234,
+      "loss": 1.9731,
+      "step": 273560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015307970192145044,
+      "loss": 2.0179,
+      "step": 273565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001530781354872445,
+      "loss": 2.0893,
+      "step": 273570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001530765690349061,
+      "loss": 2.124,
+      "step": 273575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001530750025644358,
+      "loss": 1.8349,
+      "step": 273580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001530734360758341,
+      "loss": 2.125,
+      "step": 273585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015307186956910152,
+      "loss": 2.0232,
+      "step": 273590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015307030304423868,
+      "loss": 2.0454,
+      "step": 273595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015306873650124602,
+      "loss": 2.1181,
+      "step": 273600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015306716994012415,
+      "loss": 2.2362,
+      "step": 273605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015306560336087355,
+      "loss": 2.0934,
+      "step": 273610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015306403676349482,
+      "loss": 2.131,
+      "step": 273615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001530624701479884,
+      "loss": 2.0576,
+      "step": 273620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015306090351435491,
+      "loss": 1.8169,
+      "step": 273625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015305933686259482,
+      "loss": 2.1407,
+      "step": 273630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015305777019270872,
+      "loss": 2.0021,
+      "step": 273635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015305620350469714,
+      "loss": 2.1715,
+      "step": 273640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015305463679856057,
+      "loss": 2.0408,
+      "step": 273645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015305307007429958,
+      "loss": 2.1576,
+      "step": 273650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001530515033319147,
+      "loss": 1.9108,
+      "step": 273655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015304993657140647,
+      "loss": 2.0454,
+      "step": 273660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015304836979277542,
+      "loss": 2.0914,
+      "step": 273665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001530468029960221,
+      "loss": 2.0329,
+      "step": 273670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015304523618114702,
+      "loss": 2.0208,
+      "step": 273675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001530436693481507,
+      "loss": 1.9309,
+      "step": 273680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001530421024970337,
+      "loss": 2.201,
+      "step": 273685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015304053562779658,
+      "loss": 2.0145,
+      "step": 273690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015303896874043987,
+      "loss": 2.1438,
+      "step": 273695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015303740183496406,
+      "loss": 2.0823,
+      "step": 273700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001530358349113697,
+      "loss": 1.9022,
+      "step": 273705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015303426796965734,
+      "loss": 1.8862,
+      "step": 273710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015303270100982755,
+      "loss": 2.0421,
+      "step": 273715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00015303113403188076,
+      "loss": 1.9849,
+      "step": 273720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015302956703581763,
+      "loss": 2.1173,
+      "step": 273725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015302800002163863,
+      "loss": 1.998,
+      "step": 273730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015302643298934426,
+      "loss": 2.0163,
+      "step": 273735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015302486593893514,
+      "loss": 1.8658,
+      "step": 273740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015302329887041175,
+      "loss": 2.0578,
+      "step": 273745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015302173178377467,
+      "loss": 2.106,
+      "step": 273750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001530201646790244,
+      "loss": 2.0808,
+      "step": 273755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015301859755616147,
+      "loss": 2.1098,
+      "step": 273760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015301703041518638,
+      "loss": 2.2353,
+      "step": 273765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015301546325609974,
+      "loss": 2.1728,
+      "step": 273770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001530138960789021,
+      "loss": 2.1318,
+      "step": 273775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015301232888359392,
+      "loss": 2.099,
+      "step": 273780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015301076167017577,
+      "loss": 1.9336,
+      "step": 273785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015300919443864814,
+      "loss": 2.1137,
+      "step": 273790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015300762718901164,
+      "loss": 2.1761,
+      "step": 273795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001530060599212668,
+      "loss": 2.2985,
+      "step": 273800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001530044926354141,
+      "loss": 2.0507,
+      "step": 273805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015300292533145413,
+      "loss": 2.1403,
+      "step": 273810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015300135800938738,
+      "loss": 2.0147,
+      "step": 273815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001529997906692144,
+      "loss": 1.9107,
+      "step": 273820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015299822331093572,
+      "loss": 2.0497,
+      "step": 273825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015299665593455187,
+      "loss": 2.0949,
+      "step": 273830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015299508854006345,
+      "loss": 2.0507,
+      "step": 273835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015299352112747093,
+      "loss": 1.9414,
+      "step": 273840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015299195369677487,
+      "loss": 1.9387,
+      "step": 273845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015299038624797578,
+      "loss": 2.1068,
+      "step": 273850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015298881878107422,
+      "loss": 2.016,
+      "step": 273855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001529872512960707,
+      "loss": 2.0802,
+      "step": 273860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015298568379296584,
+      "loss": 2.0825,
+      "step": 273865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015298411627176,
+      "loss": 1.9898,
+      "step": 273870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015298254873245392,
+      "loss": 2.0317,
+      "step": 273875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.000152980981175048,
+      "loss": 2.0266,
+      "step": 273880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015297941359954283,
+      "loss": 2.1928,
+      "step": 273885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015297784600593894,
+      "loss": 1.9808,
+      "step": 273890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015297627839423682,
+      "loss": 2.1247,
+      "step": 273895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015297471076443706,
+      "loss": 2.081,
+      "step": 273900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015297314311654017,
+      "loss": 1.9932,
+      "step": 273905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015297157545054673,
+      "loss": 2.0787,
+      "step": 273910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015297000776645723,
+      "loss": 1.9449,
+      "step": 273915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015296844006427218,
+      "loss": 1.9997,
+      "step": 273920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015296687234399217,
+      "loss": 1.9184,
+      "step": 273925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015296530460561774,
+      "loss": 2.0084,
+      "step": 273930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015296373684914934,
+      "loss": 1.7387,
+      "step": 273935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001529621690745876,
+      "loss": 2.1553,
+      "step": 273940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015296060128193304,
+      "loss": 2.0837,
+      "step": 273945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015295903347118617,
+      "loss": 1.9806,
+      "step": 273950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015295746564234755,
+      "loss": 2.2032,
+      "step": 273955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015295589779541766,
+      "loss": 1.9079,
+      "step": 273960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001529543299303971,
+      "loss": 2.2473,
+      "step": 273965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015295276204728637,
+      "loss": 1.9657,
+      "step": 273970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015295119414608603,
+      "loss": 2.1828,
+      "step": 273975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015294962622679656,
+      "loss": 2.1698,
+      "step": 273980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001529480582894186,
+      "loss": 2.0367,
+      "step": 273985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015294649033395258,
+      "loss": 2.0893,
+      "step": 273990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015294492236039907,
+      "loss": 1.9964,
+      "step": 273995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015294335436875865,
+      "loss": 2.1663,
+      "step": 274000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001529417863590318,
+      "loss": 2.151,
+      "step": 274005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015294021833121908,
+      "loss": 2.1425,
+      "step": 274010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000152938650285321,
+      "loss": 2.122,
+      "step": 274015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015293708222133815,
+      "loss": 2.1916,
+      "step": 274020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015293551413927105,
+      "loss": 2.1567,
+      "step": 274025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015293394603912018,
+      "loss": 1.9458,
+      "step": 274030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015293237792088612,
+      "loss": 2.2294,
+      "step": 274035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015293080978456938,
+      "loss": 1.9904,
+      "step": 274040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015292924163017057,
+      "loss": 1.9406,
+      "step": 274045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015292767345769014,
+      "loss": 2.1477,
+      "step": 274050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015292610526712865,
+      "loss": 2.1343,
+      "step": 274055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015292453705848663,
+      "loss": 2.0153,
+      "step": 274060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015292296883176465,
+      "loss": 2.2419,
+      "step": 274065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015292140058696324,
+      "loss": 2.0613,
+      "step": 274070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001529198323240829,
+      "loss": 2.0917,
+      "step": 274075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015291826404312419,
+      "loss": 2.1337,
+      "step": 274080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001529166957440876,
+      "loss": 2.0127,
+      "step": 274085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015291512742697378,
+      "loss": 2.2806,
+      "step": 274090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015291355909178314,
+      "loss": 2.1957,
+      "step": 274095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001529119907385163,
+      "loss": 2.0927,
+      "step": 274100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015291042236717373,
+      "loss": 2.2612,
+      "step": 274105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015290885397775602,
+      "loss": 2.292,
+      "step": 274110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001529072855702637,
+      "loss": 2.12,
+      "step": 274115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001529057171446973,
+      "loss": 2.041,
+      "step": 274120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001529041487010573,
+      "loss": 2.1627,
+      "step": 274125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015290258023934432,
+      "loss": 2.1651,
+      "step": 274130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015290101175955884,
+      "loss": 2.1893,
+      "step": 274135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015289944326170142,
+      "loss": 2.0825,
+      "step": 274140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001528978747457726,
+      "loss": 2.0747,
+      "step": 274145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015289630621177292,
+      "loss": 1.8688,
+      "step": 274150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015289473765970287,
+      "loss": 2.1347,
+      "step": 274155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015289316908956307,
+      "loss": 2.1394,
+      "step": 274160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001528916005013539,
+      "loss": 1.9857,
+      "step": 274165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015289003189507608,
+      "loss": 2.1032,
+      "step": 274170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015288846327073005,
+      "loss": 2.1221,
+      "step": 274175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001528868946283164,
+      "loss": 1.9563,
+      "step": 274180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001528853259678356,
+      "loss": 2.0618,
+      "step": 274185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015288375728928817,
+      "loss": 2.051,
+      "step": 274190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015288218859267473,
+      "loss": 2.1102,
+      "step": 274195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001528806198779958,
+      "loss": 2.2329,
+      "step": 274200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00015287905114525185,
+      "loss": 2.1473,
+      "step": 274205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015287748239444345,
+      "loss": 2.0863,
+      "step": 274210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015287591362557116,
+      "loss": 2.114,
+      "step": 274215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015287434483863548,
+      "loss": 2.0699,
+      "step": 274220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.000152872776033637,
+      "loss": 2.1126,
+      "step": 274225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001528712072105762,
+      "loss": 2.1015,
+      "step": 274230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015286963836945368,
+      "loss": 2.2643,
+      "step": 274235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015286806951026987,
+      "loss": 1.873,
+      "step": 274240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001528665006330254,
+      "loss": 1.9612,
+      "step": 274245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015286493173772076,
+      "loss": 2.2318,
+      "step": 274250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001528633628243565,
+      "loss": 2.098,
+      "step": 274255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015286179389293318,
+      "loss": 2.0774,
+      "step": 274260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001528602249434513,
+      "loss": 1.9582,
+      "step": 274265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001528586559759114,
+      "loss": 2.0096,
+      "step": 274270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015285708699031402,
+      "loss": 2.1645,
+      "step": 274275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015285551798665972,
+      "loss": 2.2376,
+      "step": 274280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015285394896494902,
+      "loss": 2.0143,
+      "step": 274285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001528523799251824,
+      "loss": 1.9623,
+      "step": 274290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015285081086736054,
+      "loss": 2.0683,
+      "step": 274295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015284924179148382,
+      "loss": 2.2344,
+      "step": 274300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015284767269755284,
+      "loss": 1.9528,
+      "step": 274305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015284610358556815,
+      "loss": 2.1377,
+      "step": 274310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015284453445553028,
+      "loss": 1.8866,
+      "step": 274315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001528429653074398,
+      "loss": 1.9994,
+      "step": 274320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015284139614129712,
+      "loss": 2.101,
+      "step": 274325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015283982695710288,
+      "loss": 2.0996,
+      "step": 274330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001528382577548576,
+      "loss": 2.13,
+      "step": 274335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015283668853456186,
+      "loss": 1.9684,
+      "step": 274340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015283511929621612,
+      "loss": 2.0106,
+      "step": 274345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001528335500398209,
+      "loss": 2.1627,
+      "step": 274350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015283198076537682,
+      "loss": 2.1145,
+      "step": 274355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001528304114728844,
+      "loss": 2.236,
+      "step": 274360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015282884216234416,
+      "loss": 2.0972,
+      "step": 274365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015282727283375658,
+      "loss": 2.0696,
+      "step": 274370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015282570348712228,
+      "loss": 2.0688,
+      "step": 274375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015282413412244172,
+      "loss": 2.0451,
+      "step": 274380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015282256473971552,
+      "loss": 2.1581,
+      "step": 274385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015282099533894415,
+      "loss": 2.2365,
+      "step": 274390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001528194259201282,
+      "loss": 2.1732,
+      "step": 274395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015281785648326818,
+      "loss": 2.12,
+      "step": 274400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015281628702836457,
+      "loss": 1.9528,
+      "step": 274405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015281471755541797,
+      "loss": 1.9743,
+      "step": 274410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015281314806442895,
+      "loss": 1.9971,
+      "step": 274415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015281157855539796,
+      "loss": 2.1367,
+      "step": 274420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015281000902832558,
+      "loss": 2.2029,
+      "step": 274425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015280843948321237,
+      "loss": 2.0107,
+      "step": 274430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015280686992005883,
+      "loss": 2.0429,
+      "step": 274435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015280530033886546,
+      "loss": 2.0387,
+      "step": 274440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001528037307396329,
+      "loss": 1.8888,
+      "step": 274445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001528021611223616,
+      "loss": 2.0421,
+      "step": 274450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015280059148705215,
+      "loss": 2.1903,
+      "step": 274455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015279902183370502,
+      "loss": 2.1538,
+      "step": 274460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015279745216232083,
+      "loss": 1.9381,
+      "step": 274465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015279588247290001,
+      "loss": 1.9874,
+      "step": 274470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015279431276544322,
+      "loss": 2.1526,
+      "step": 274475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015279274303995095,
+      "loss": 2.0715,
+      "step": 274480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015279117329642365,
+      "loss": 1.9691,
+      "step": 274485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015278960353486197,
+      "loss": 2.0017,
+      "step": 274490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001527880337552664,
+      "loss": 2.0951,
+      "step": 274495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001527864639576375,
+      "loss": 2.0838,
+      "step": 274500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015278489414197576,
+      "loss": 2.1219,
+      "step": 274505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015278332430828173,
+      "loss": 1.9804,
+      "step": 274510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015278175445655599,
+      "loss": 2.0984,
+      "step": 274515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015278018458679906,
+      "loss": 2.0635,
+      "step": 274520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015277861469901143,
+      "loss": 2.1013,
+      "step": 274525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015277704479319367,
+      "loss": 2.07,
+      "step": 274530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001527754748693463,
+      "loss": 2.1554,
+      "step": 274535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015277390492746988,
+      "loss": 1.9851,
+      "step": 274540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015277233496756496,
+      "loss": 2.059,
+      "step": 274545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00015277076498963204,
+      "loss": 2.2684,
+      "step": 274550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015276919499367165,
+      "loss": 2.0545,
+      "step": 274555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015276762497968437,
+      "loss": 1.943,
+      "step": 274560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015276605494767072,
+      "loss": 2.0884,
+      "step": 274565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001527644848976312,
+      "loss": 2.1537,
+      "step": 274570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001527629148295664,
+      "loss": 2.062,
+      "step": 274575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001527613447434768,
+      "loss": 1.8999,
+      "step": 274580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015275977463936302,
+      "loss": 2.0711,
+      "step": 274585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015275820451722552,
+      "loss": 2.0405,
+      "step": 274590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015275663437706483,
+      "loss": 1.9153,
+      "step": 274595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015275506421888154,
+      "loss": 2.0199,
+      "step": 274600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015275349404267615,
+      "loss": 1.8864,
+      "step": 274605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015275192384844925,
+      "loss": 2.0563,
+      "step": 274610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015275035363620132,
+      "loss": 2.0938,
+      "step": 274615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001527487834059329,
+      "loss": 1.9541,
+      "step": 274620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015274721315764454,
+      "loss": 2.161,
+      "step": 274625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015274564289133674,
+      "loss": 2.0657,
+      "step": 274630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015274407260701013,
+      "loss": 2.0521,
+      "step": 274635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015274250230466518,
+      "loss": 2.1839,
+      "step": 274640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015274093198430243,
+      "loss": 2.0797,
+      "step": 274645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015273936164592237,
+      "loss": 1.99,
+      "step": 274650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015273779128952564,
+      "loss": 1.9271,
+      "step": 274655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015273622091511272,
+      "loss": 2.0353,
+      "step": 274660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015273465052268416,
+      "loss": 1.9755,
+      "step": 274665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015273308011224047,
+      "loss": 2.1376,
+      "step": 274670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015273150968378218,
+      "loss": 2.0769,
+      "step": 274675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001527299392373099,
+      "loss": 1.9524,
+      "step": 274680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015272836877282408,
+      "loss": 1.9315,
+      "step": 274685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001527267982903253,
+      "loss": 2.0795,
+      "step": 274690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001527252277898141,
+      "loss": 2.1542,
+      "step": 274695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000152723657271291,
+      "loss": 2.0722,
+      "step": 274700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00015272208673475658,
+      "loss": 2.1595,
+      "step": 274705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015272051618021125,
+      "loss": 2.1792,
+      "step": 274710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001527189456076557,
+      "loss": 1.9804,
+      "step": 274715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001527173750170904,
+      "loss": 2.0574,
+      "step": 274720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015271580440851586,
+      "loss": 2.1066,
+      "step": 274725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015271423378193267,
+      "loss": 2.0664,
+      "step": 274730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015271266313734134,
+      "loss": 2.0814,
+      "step": 274735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001527110924747424,
+      "loss": 2.0292,
+      "step": 274740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001527095217941364,
+      "loss": 2.1647,
+      "step": 274745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015270795109552387,
+      "loss": 1.9651,
+      "step": 274750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015270638037890533,
+      "loss": 2.2002,
+      "step": 274755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015270480964428137,
+      "loss": 2.2234,
+      "step": 274760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015270323889165246,
+      "loss": 2.0046,
+      "step": 274765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001527016681210192,
+      "loss": 1.9242,
+      "step": 274770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001527000973323821,
+      "loss": 2.0836,
+      "step": 274775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015269852652574165,
+      "loss": 2.1326,
+      "step": 274780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015269695570109844,
+      "loss": 2.1303,
+      "step": 274785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015269538485845297,
+      "loss": 2.015,
+      "step": 274790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015269381399780584,
+      "loss": 2.0948,
+      "step": 274795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015269224311915754,
+      "loss": 1.8871,
+      "step": 274800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001526906722225086,
+      "loss": 2.2513,
+      "step": 274805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015268910130785958,
+      "loss": 2.1892,
+      "step": 274810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015268753037521102,
+      "loss": 2.0117,
+      "step": 274815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015268595942456346,
+      "loss": 2.2075,
+      "step": 274820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015268438845591738,
+      "loss": 2.0402,
+      "step": 274825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015268281746927338,
+      "loss": 2.0463,
+      "step": 274830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015268124646463193,
+      "loss": 1.7823,
+      "step": 274835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015267967544199367,
+      "loss": 2.1625,
+      "step": 274840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015267810440135904,
+      "loss": 2.1396,
+      "step": 274845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015267653334272862,
+      "loss": 2.171,
+      "step": 274850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015267496226610295,
+      "loss": 2.0757,
+      "step": 274855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015267339117148256,
+      "loss": 2.2283,
+      "step": 274860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015267182005886798,
+      "loss": 2.2226,
+      "step": 274865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015267024892825975,
+      "loss": 1.8105,
+      "step": 274870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015266867777965842,
+      "loss": 2.1871,
+      "step": 274875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015266710661306448,
+      "loss": 2.1349,
+      "step": 274880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015266553542847853,
+      "loss": 2.0729,
+      "step": 274885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015266396422590107,
+      "loss": 2.0151,
+      "step": 274890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015266239300533265,
+      "loss": 1.9333,
+      "step": 274895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015266082176677382,
+      "loss": 2.0331,
+      "step": 274900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015265925051022506,
+      "loss": 2.2205,
+      "step": 274905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015265767923568698,
+      "loss": 1.9182,
+      "step": 274910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015265610794316006,
+      "loss": 1.9351,
+      "step": 274915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015265453663264485,
+      "loss": 2.0866,
+      "step": 274920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001526529653041419,
+      "loss": 2.2224,
+      "step": 274925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015265139395765175,
+      "loss": 2.1758,
+      "step": 274930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015264982259317495,
+      "loss": 2.21,
+      "step": 274935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015264825121071195,
+      "loss": 2.0027,
+      "step": 274940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001526466798102634,
+      "loss": 2.1383,
+      "step": 274945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001526451083918298,
+      "loss": 1.9898,
+      "step": 274950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001526435369554117,
+      "loss": 2.3037,
+      "step": 274955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015264196550100956,
+      "loss": 2.2006,
+      "step": 274960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015264039402862397,
+      "loss": 2.0108,
+      "step": 274965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015263882253825547,
+      "loss": 2.0131,
+      "step": 274970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015263725102990464,
+      "loss": 2.1245,
+      "step": 274975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015263567950357193,
+      "loss": 1.961,
+      "step": 274980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015263410795925792,
+      "loss": 1.9601,
+      "step": 274985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015263253639696313,
+      "loss": 1.9002,
+      "step": 274990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001526309648166881,
+      "loss": 2.2387,
+      "step": 274995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015262939321843345,
+      "loss": 2.1436,
+      "step": 275000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001526278216021996,
+      "loss": 2.2425,
+      "step": 275005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015262624996798714,
+      "loss": 2.1674,
+      "step": 275010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015262467831579657,
+      "loss": 2.0087,
+      "step": 275015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015262310664562848,
+      "loss": 2.1152,
+      "step": 275020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015262153495748337,
+      "loss": 2.0849,
+      "step": 275025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001526199632513618,
+      "loss": 2.105,
+      "step": 275030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001526183915272643,
+      "loss": 2.0451,
+      "step": 275035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001526168197851914,
+      "loss": 2.1771,
+      "step": 275040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015261524802514365,
+      "loss": 2.2507,
+      "step": 275045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015261367624712154,
+      "loss": 2.0525,
+      "step": 275050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015261210445112567,
+      "loss": 2.0406,
+      "step": 275055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015261053263715657,
+      "loss": 2.1031,
+      "step": 275060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015260896080521473,
+      "loss": 2.018,
+      "step": 275065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015260738895530075,
+      "loss": 2.3191,
+      "step": 275070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015260581708741508,
+      "loss": 2.0669,
+      "step": 275075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001526042452015583,
+      "loss": 2.1154,
+      "step": 275080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015260267329773105,
+      "loss": 2.0507,
+      "step": 275085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015260110137593367,
+      "loss": 2.16,
+      "step": 275090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015259952943616687,
+      "loss": 1.9459,
+      "step": 275095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001525979574784311,
+      "loss": 2.0114,
+      "step": 275100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015259638550272687,
+      "loss": 2.0926,
+      "step": 275105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00015259481350905483,
+      "loss": 2.0677,
+      "step": 275110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015259324149741538,
+      "loss": 2.1801,
+      "step": 275115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015259166946780917,
+      "loss": 2.1294,
+      "step": 275120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001525900974202367,
+      "loss": 2.0925,
+      "step": 275125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015258852535469843,
+      "loss": 1.9802,
+      "step": 275130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015258695327119502,
+      "loss": 1.972,
+      "step": 275135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015258538116972697,
+      "loss": 2.1114,
+      "step": 275140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015258380905029478,
+      "loss": 1.9866,
+      "step": 275145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015258223691289898,
+      "loss": 2.0126,
+      "step": 275150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015258066475754017,
+      "loss": 2.116,
+      "step": 275155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015257909258421882,
+      "loss": 1.9068,
+      "step": 275160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00015257752039293554,
+      "loss": 2.095,
+      "step": 275165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001525759481836908,
+      "loss": 1.9561,
+      "step": 275170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015257437595648515,
+      "loss": 1.9519,
+      "step": 275175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015257280371131917,
+      "loss": 1.9605,
+      "step": 275180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015257123144819335,
+      "loss": 2.0132,
+      "step": 275185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015256965916710824,
+      "loss": 2.1103,
+      "step": 275190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001525680868680644,
+      "loss": 2.0115,
+      "step": 275195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015256651455106232,
+      "loss": 2.1666,
+      "step": 275200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001525649422161026,
+      "loss": 2.1477,
+      "step": 275205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001525633698631857,
+      "loss": 1.9488,
+      "step": 275210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015256179749231224,
+      "loss": 1.988,
+      "step": 275215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001525602251034827,
+      "loss": 2.1755,
+      "step": 275220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015255865269669765,
+      "loss": 2.143,
+      "step": 275225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001525570802719576,
+      "loss": 2.0835,
+      "step": 275230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015255550782926304,
+      "loss": 1.9868,
+      "step": 275235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015255393536861464,
+      "loss": 2.142,
+      "step": 275240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015255236289001282,
+      "loss": 2.0829,
+      "step": 275245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001525507903934582,
+      "loss": 1.8827,
+      "step": 275250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015254921787895124,
+      "loss": 2.0269,
+      "step": 275255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015254764534649256,
+      "loss": 2.1809,
+      "step": 275260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001525460727960826,
+      "loss": 2.2752,
+      "step": 275265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015254450022772196,
+      "loss": 2.0023,
+      "step": 275270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001525429276414112,
+      "loss": 2.1756,
+      "step": 275275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001525413550371508,
+      "loss": 2.0003,
+      "step": 275280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001525397824149413,
+      "loss": 1.9319,
+      "step": 275285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015253820977478326,
+      "loss": 1.9854,
+      "step": 275290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015253663711667725,
+      "loss": 2.1015,
+      "step": 275295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015253506444062377,
+      "loss": 2.1869,
+      "step": 275300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015253349174662332,
+      "loss": 2.1382,
+      "step": 275305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001525319190346765,
+      "loss": 2.1029,
+      "step": 275310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00015253034630478385,
+      "loss": 2.0516,
+      "step": 275315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015252877355694585,
+      "loss": 2.2496,
+      "step": 275320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015252720079116306,
+      "loss": 2.0506,
+      "step": 275325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015252562800743605,
+      "loss": 2.1809,
+      "step": 275330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015252405520576533,
+      "loss": 2.227,
+      "step": 275335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015252248238615144,
+      "loss": 2.0786,
+      "step": 275340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015252090954859487,
+      "loss": 1.9816,
+      "step": 275345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015251933669309625,
+      "loss": 2.0605,
+      "step": 275350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001525177638196561,
+      "loss": 2.1306,
+      "step": 275355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015251619092827485,
+      "loss": 2.0403,
+      "step": 275360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001525146180189532,
+      "loss": 2.0938,
+      "step": 275365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00015251304509169154,
+      "loss": 1.9358,
+      "step": 275370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001525114721464905,
+      "loss": 1.9412,
+      "step": 275375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015250989918335058,
+      "loss": 2.0466,
+      "step": 275380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015250832620227234,
+      "loss": 1.91,
+      "step": 275385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015250675320325628,
+      "loss": 2.0686,
+      "step": 275390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000152505180186303,
+      "loss": 2.0723,
+      "step": 275395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015250360715141295,
+      "loss": 1.9609,
+      "step": 275400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015250203409858677,
+      "loss": 2.1728,
+      "step": 275405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001525004610278249,
+      "loss": 2.0825,
+      "step": 275410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015249888793912795,
+      "loss": 1.949,
+      "step": 275415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015249731483249642,
+      "loss": 1.9329,
+      "step": 275420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015249574170793083,
+      "loss": 2.1603,
+      "step": 275425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015249416856543175,
+      "loss": 1.9189,
+      "step": 275430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015249259540499975,
+      "loss": 2.0581,
+      "step": 275435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015249102222663526,
+      "loss": 2.1699,
+      "step": 275440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015248944903033894,
+      "loss": 2.0946,
+      "step": 275445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015248787581611124,
+      "loss": 2.1411,
+      "step": 275450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015248630258395273,
+      "loss": 2.2716,
+      "step": 275455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015248472933386399,
+      "loss": 2.3103,
+      "step": 275460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015248315606584547,
+      "loss": 1.9397,
+      "step": 275465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015248158277989776,
+      "loss": 1.9327,
+      "step": 275470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001524800094760214,
+      "loss": 1.8614,
+      "step": 275475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001524784361542169,
+      "loss": 2.239,
+      "step": 275480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015247686281448487,
+      "loss": 2.0934,
+      "step": 275485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015247528945682573,
+      "loss": 2.1734,
+      "step": 275490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015247371608124007,
+      "loss": 2.0766,
+      "step": 275495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001524721426877285,
+      "loss": 2.0775,
+      "step": 275500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015247056927629145,
+      "loss": 1.9191,
+      "step": 275505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001524689958469295,
+      "loss": 2.02,
+      "step": 275510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015246742239964322,
+      "loss": 2.0638,
+      "step": 275515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001524658489344331,
+      "loss": 1.9102,
+      "step": 275520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001524642754512997,
+      "loss": 1.8946,
+      "step": 275525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015246270195024353,
+      "loss": 1.9195,
+      "step": 275530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015246112843126516,
+      "loss": 1.9084,
+      "step": 275535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015245955489436513,
+      "loss": 1.9694,
+      "step": 275540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015245798133954396,
+      "loss": 2.0469,
+      "step": 275545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001524564077668022,
+      "loss": 2.2055,
+      "step": 275550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015245483417614035,
+      "loss": 1.9086,
+      "step": 275555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000152453260567559,
+      "loss": 2.226,
+      "step": 275560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015245168694105868,
+      "loss": 2.0533,
+      "step": 275565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015245011329663988,
+      "loss": 2.0852,
+      "step": 275570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001524485396343032,
+      "loss": 1.9979,
+      "step": 275575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015244696595404913,
+      "loss": 1.9275,
+      "step": 275580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001524453922558782,
+      "loss": 2.0924,
+      "step": 275585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000152443818539791,
+      "loss": 2.0145,
+      "step": 275590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015244224480578806,
+      "loss": 2.0223,
+      "step": 275595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015244067105386989,
+      "loss": 1.8695,
+      "step": 275600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000152439097284037,
+      "loss": 2.1406,
+      "step": 275605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015243752349629,
+      "loss": 2.2742,
+      "step": 275610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015243594969062936,
+      "loss": 2.091,
+      "step": 275615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015243437586705567,
+      "loss": 1.9441,
+      "step": 275620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015243280202556945,
+      "loss": 1.9609,
+      "step": 275625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015243122816617122,
+      "loss": 2.0332,
+      "step": 275630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015242965428886155,
+      "loss": 1.9565,
+      "step": 275635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015242808039364093,
+      "loss": 2.068,
+      "step": 275640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015242650648050995,
+      "loss": 1.8462,
+      "step": 275645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001524249325494691,
+      "loss": 2.2745,
+      "step": 275650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000152423358600519,
+      "loss": 1.9405,
+      "step": 275655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001524217846336601,
+      "loss": 2.1737,
+      "step": 275660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015242021064889292,
+      "loss": 2.0667,
+      "step": 275665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001524186366462181,
+      "loss": 2.2359,
+      "step": 275670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015241706262563608,
+      "loss": 1.9873,
+      "step": 275675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001524154885871475,
+      "loss": 2.0376,
+      "step": 275680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001524139145307528,
+      "loss": 2.1565,
+      "step": 275685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015241234045645254,
+      "loss": 2.112,
+      "step": 275690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001524107663642473,
+      "loss": 2.0902,
+      "step": 275695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001524091922541376,
+      "loss": 2.1289,
+      "step": 275700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015240761812612398,
+      "loss": 2.0904,
+      "step": 275705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015240604398020696,
+      "loss": 1.828,
+      "step": 275710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015240446981638705,
+      "loss": 2.0861,
+      "step": 275715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00015240289563466482,
+      "loss": 1.8641,
+      "step": 275720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015240132143504086,
+      "loss": 2.0687,
+      "step": 275725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015239974721751562,
+      "loss": 2.0717,
+      "step": 275730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015239817298208968,
+      "loss": 2.0556,
+      "step": 275735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001523965987287636,
+      "loss": 2.0989,
+      "step": 275740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015239502445753785,
+      "loss": 2.142,
+      "step": 275745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015239345016841303,
+      "loss": 1.8225,
+      "step": 275750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015239187586138967,
+      "loss": 2.1206,
+      "step": 275755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001523903015364683,
+      "loss": 1.8945,
+      "step": 275760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.00015238872719364942,
+      "loss": 2.0598,
+      "step": 275765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015238715283293362,
+      "loss": 1.8512,
+      "step": 275770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015238557845432142,
+      "loss": 2.3142,
+      "step": 275775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015238400405781337,
+      "loss": 2.0327,
+      "step": 275780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015238242964340997,
+      "loss": 1.9087,
+      "step": 275785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015238085521111175,
+      "loss": 2.1427,
+      "step": 275790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015237928076091935,
+      "loss": 2.109,
+      "step": 275795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001523777062928332,
+      "loss": 1.862,
+      "step": 275800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001523761318068539,
+      "loss": 2.0248,
+      "step": 275805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001523745573029819,
+      "loss": 2.0106,
+      "step": 275810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015237298278121788,
+      "loss": 1.8509,
+      "step": 275815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001523714082415623,
+      "loss": 2.2638,
+      "step": 275820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015236983368401564,
+      "loss": 2.048,
+      "step": 275825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001523682591085785,
+      "loss": 1.9757,
+      "step": 275830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015236668451525142,
+      "loss": 2.0811,
+      "step": 275835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000152365109904035,
+      "loss": 1.97,
+      "step": 275840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001523635352749296,
+      "loss": 2.01,
+      "step": 275845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015236196062793592,
+      "loss": 2.1862,
+      "step": 275850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015236038596305444,
+      "loss": 2.0159,
+      "step": 275855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015235881128028574,
+      "loss": 2.025,
+      "step": 275860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015235723657963028,
+      "loss": 2.1668,
+      "step": 275865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015235566186108863,
+      "loss": 2.1726,
+      "step": 275870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015235408712466138,
+      "loss": 1.9961,
+      "step": 275875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015235251237034894,
+      "loss": 2.1903,
+      "step": 275880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000152350937598152,
+      "loss": 2.0606,
+      "step": 275885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015234936280807105,
+      "loss": 2.1176,
+      "step": 275890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015234778800010657,
+      "loss": 2.0915,
+      "step": 275895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015234621317425912,
+      "loss": 1.8293,
+      "step": 275900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015234463833052928,
+      "loss": 2.2003,
+      "step": 275905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015234306346891756,
+      "loss": 2.0659,
+      "step": 275910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001523414885894245,
+      "loss": 2.0712,
+      "step": 275915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015233991369205066,
+      "loss": 2.0935,
+      "step": 275920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001523383387767965,
+      "loss": 1.9878,
+      "step": 275925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015233676384366268,
+      "loss": 2.0205,
+      "step": 275930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015233518889264962,
+      "loss": 2.1542,
+      "step": 275935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015233361392375793,
+      "loss": 2.1579,
+      "step": 275940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015233203893698813,
+      "loss": 1.8501,
+      "step": 275945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015233046393234073,
+      "loss": 2.1447,
+      "step": 275950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015232888890981636,
+      "loss": 2.1059,
+      "step": 275955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001523273138694154,
+      "loss": 2.0937,
+      "step": 275960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015232573881113855,
+      "loss": 2.0567,
+      "step": 275965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015232416373498628,
+      "loss": 2.1833,
+      "step": 275970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015232258864095912,
+      "loss": 2.0676,
+      "step": 275975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001523210135290576,
+      "loss": 2.0726,
+      "step": 275980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015231943839928225,
+      "loss": 2.2121,
+      "step": 275985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015231786325163365,
+      "loss": 2.055,
+      "step": 275990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015231628808611234,
+      "loss": 2.1216,
+      "step": 275995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015231471290271884,
+      "loss": 2.1138,
+      "step": 276000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015231313770145365,
+      "loss": 2.2417,
+      "step": 276005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015231156248231734,
+      "loss": 2.109,
+      "step": 276010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015230998724531048,
+      "loss": 2.2974,
+      "step": 276015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015230841199043358,
+      "loss": 2.1724,
+      "step": 276020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015230683671768716,
+      "loss": 2.2058,
+      "step": 276025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001523052614270718,
+      "loss": 2.2404,
+      "step": 276030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000152303686118588,
+      "loss": 2.0561,
+      "step": 276035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001523021107922363,
+      "loss": 1.8928,
+      "step": 276040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015230053544801727,
+      "loss": 2.1178,
+      "step": 276045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00015229896008593142,
+      "loss": 1.9151,
+      "step": 276050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001522973847059793,
+      "loss": 2.1537,
+      "step": 276055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015229580930816144,
+      "loss": 1.9578,
+      "step": 276060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015229423389247839,
+      "loss": 2.0696,
+      "step": 276065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001522926584589307,
+      "loss": 2.0813,
+      "step": 276070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015229108300751884,
+      "loss": 1.9117,
+      "step": 276075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015228950753824343,
+      "loss": 2.1013,
+      "step": 276080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000152287932051105,
+      "loss": 2.0365,
+      "step": 276085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015228635654610404,
+      "loss": 1.9479,
+      "step": 276090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015228478102324107,
+      "loss": 2.1037,
+      "step": 276095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001522832054825167,
+      "loss": 2.2358,
+      "step": 276100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015228162992393144,
+      "loss": 2.1713,
+      "step": 276105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015228005434748587,
+      "loss": 2.2231,
+      "step": 276110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015227847875318043,
+      "loss": 2.0957,
+      "step": 276115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001522769031410157,
+      "loss": 2.0557,
+      "step": 276120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015227532751099229,
+      "loss": 2.0913,
+      "step": 276125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015227375186311067,
+      "loss": 2.0318,
+      "step": 276130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015227217619737137,
+      "loss": 2.0115,
+      "step": 276135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015227060051377494,
+      "loss": 1.9684,
+      "step": 276140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015226902481232192,
+      "loss": 2.0418,
+      "step": 276145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015226744909301287,
+      "loss": 2.0989,
+      "step": 276150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015226587335584831,
+      "loss": 2.1621,
+      "step": 276155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001522642976008288,
+      "loss": 2.082,
+      "step": 276160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015226272182795482,
+      "loss": 2.2616,
+      "step": 276165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015226114603722696,
+      "loss": 2.1832,
+      "step": 276170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015225957022864574,
+      "loss": 2.1212,
+      "step": 276175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015225799440221174,
+      "loss": 2.0102,
+      "step": 276180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015225641855792543,
+      "loss": 1.9295,
+      "step": 276185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015225484269578736,
+      "loss": 2.1213,
+      "step": 276190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015225326681579811,
+      "loss": 1.9651,
+      "step": 276195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001522516909179582,
+      "loss": 2.0371,
+      "step": 276200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015225011500226817,
+      "loss": 2.0295,
+      "step": 276205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015224853906872852,
+      "loss": 2.149,
+      "step": 276210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015224696311733985,
+      "loss": 2.0967,
+      "step": 276215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015224538714810268,
+      "loss": 1.9318,
+      "step": 276220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015224381116101752,
+      "loss": 1.9634,
+      "step": 276225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015224223515608492,
+      "loss": 1.8199,
+      "step": 276230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015224065913330546,
+      "loss": 2.0327,
+      "step": 276235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001522390830926796,
+      "loss": 1.995,
+      "step": 276240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015223750703420797,
+      "loss": 2.1992,
+      "step": 276245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015223593095789104,
+      "loss": 2.1302,
+      "step": 276250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001522343548637293,
+      "loss": 1.8678,
+      "step": 276255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015223277875172343,
+      "loss": 2.0494,
+      "step": 276260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015223120262187388,
+      "loss": 2.0512,
+      "step": 276265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015222962647418124,
+      "loss": 2.159,
+      "step": 276270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015222805030864597,
+      "loss": 1.9483,
+      "step": 276275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015222647412526863,
+      "loss": 1.8763,
+      "step": 276280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015222489792404982,
+      "loss": 1.9282,
+      "step": 276285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015222332170499004,
+      "loss": 2.0814,
+      "step": 276290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015222174546808984,
+      "loss": 2.0988,
+      "step": 276295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001522201692133497,
+      "loss": 2.0825,
+      "step": 276300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015221859294077019,
+      "loss": 2.0128,
+      "step": 276305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001522170166503519,
+      "loss": 1.9043,
+      "step": 276310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001522154403420953,
+      "loss": 2.1482,
+      "step": 276315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000152213864016001,
+      "loss": 2.1194,
+      "step": 276320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015221228767206947,
+      "loss": 1.9112,
+      "step": 276325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001522107113103013,
+      "loss": 2.1647,
+      "step": 276330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015220913493069698,
+      "loss": 2.1015,
+      "step": 276335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001522075585332571,
+      "loss": 2.1302,
+      "step": 276340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015220598211798214,
+      "loss": 1.8756,
+      "step": 276345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015220440568487268,
+      "loss": 2.1639,
+      "step": 276350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015220282923392926,
+      "loss": 2.2334,
+      "step": 276355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015220125276515237,
+      "loss": 1.9982,
+      "step": 276360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001521996762785426,
+      "loss": 2.1791,
+      "step": 276365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001521980997741005,
+      "loss": 2.0946,
+      "step": 276370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015219652325182656,
+      "loss": 2.0815,
+      "step": 276375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015219494671172136,
+      "loss": 2.2524,
+      "step": 276380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015219337015378543,
+      "loss": 2.1769,
+      "step": 276385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015219179357801925,
+      "loss": 2.0174,
+      "step": 276390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015219021698442342,
+      "loss": 2.0666,
+      "step": 276395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015218864037299848,
+      "loss": 2.1083,
+      "step": 276400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015218706374374495,
+      "loss": 1.9499,
+      "step": 276405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001521854870966634,
+      "loss": 2.1413,
+      "step": 276410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001521839104317543,
+      "loss": 2.1293,
+      "step": 276415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015218233374901825,
+      "loss": 2.0787,
+      "step": 276420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015218075704845577,
+      "loss": 2.0632,
+      "step": 276425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001521791803300674,
+      "loss": 2.0322,
+      "step": 276430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015217760359385367,
+      "loss": 2.1809,
+      "step": 276435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015217602683981508,
+      "loss": 2.1097,
+      "step": 276440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001521744500679523,
+      "loss": 2.1877,
+      "step": 276445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015217287327826573,
+      "loss": 2.0899,
+      "step": 276450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015217129647075598,
+      "loss": 2.0667,
+      "step": 276455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015216971964542353,
+      "loss": 1.9623,
+      "step": 276460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015216814280226898,
+      "loss": 2.0487,
+      "step": 276465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015216656594129285,
+      "loss": 2.0211,
+      "step": 276470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015216498906249566,
+      "loss": 2.0157,
+      "step": 276475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015216341216587798,
+      "loss": 1.9078,
+      "step": 276480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015216183525144033,
+      "loss": 2.0526,
+      "step": 276485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015216025831918327,
+      "loss": 1.9542,
+      "step": 276490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001521586813691073,
+      "loss": 2.0468,
+      "step": 276495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00015215710440121298,
+      "loss": 2.2013,
+      "step": 276500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015215552741550082,
+      "loss": 2.0978,
+      "step": 276505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015215395041197143,
+      "loss": 1.9726,
+      "step": 276510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001521523733906253,
+      "loss": 2.0745,
+      "step": 276515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015215079635146295,
+      "loss": 2.0499,
+      "step": 276520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015214921929448494,
+      "loss": 2.2263,
+      "step": 276525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015214764221969183,
+      "loss": 1.9577,
+      "step": 276530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015214606512708412,
+      "loss": 2.0217,
+      "step": 276535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001521444880166624,
+      "loss": 1.9794,
+      "step": 276540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015214291088842715,
+      "loss": 2.0316,
+      "step": 276545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015214133374237893,
+      "loss": 2.17,
+      "step": 276550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015213975657851832,
+      "loss": 2.0186,
+      "step": 276555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001521381793968458,
+      "loss": 2.0532,
+      "step": 276560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015213660219736194,
+      "loss": 2.0795,
+      "step": 276565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015213502498006727,
+      "loss": 1.8957,
+      "step": 276570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001521334477449623,
+      "loss": 2.1298,
+      "step": 276575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015213187049204762,
+      "loss": 1.9398,
+      "step": 276580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015213029322132376,
+      "loss": 2.0171,
+      "step": 276585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015212871593279129,
+      "loss": 2.2962,
+      "step": 276590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015212713862645063,
+      "loss": 1.9914,
+      "step": 276595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001521255613023024,
+      "loss": 2.0792,
+      "step": 276600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015212398396034714,
+      "loss": 2.1571,
+      "step": 276605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015212240660058543,
+      "loss": 2.1361,
+      "step": 276610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015212082922301773,
+      "loss": 2.0914,
+      "step": 276615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001521192518276446,
+      "loss": 2.0388,
+      "step": 276620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015211767441446655,
+      "loss": 2.1813,
+      "step": 276625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015211609698348422,
+      "loss": 2.056,
+      "step": 276630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015211451953469807,
+      "loss": 2.131,
+      "step": 276635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015211294206810865,
+      "loss": 2.0674,
+      "step": 276640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015211136458371648,
+      "loss": 2.0651,
+      "step": 276645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015210978708152215,
+      "loss": 1.92,
+      "step": 276650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015210820956152617,
+      "loss": 2.0798,
+      "step": 276655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015210663202372908,
+      "loss": 2.1046,
+      "step": 276660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001521050544681314,
+      "loss": 2.0872,
+      "step": 276665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015210347689473372,
+      "loss": 2.0225,
+      "step": 276670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015210189930353653,
+      "loss": 2.0486,
+      "step": 276675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001521003216945404,
+      "loss": 2.1709,
+      "step": 276680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015209874406774585,
+      "loss": 2.1641,
+      "step": 276685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001520971664231534,
+      "loss": 2.1148,
+      "step": 276690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015209558876076365,
+      "loss": 2.0169,
+      "step": 276695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015209401108057707,
+      "loss": 2.2597,
+      "step": 276700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001520924333825943,
+      "loss": 2.1414,
+      "step": 276705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001520908556668157,
+      "loss": 1.9987,
+      "step": 276710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000152089277933242,
+      "loss": 2.0397,
+      "step": 276715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015208770018187363,
+      "loss": 1.9229,
+      "step": 276720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015208612241271116,
+      "loss": 2.1137,
+      "step": 276725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015208454462575514,
+      "loss": 2.0341,
+      "step": 276730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015208296682100607,
+      "loss": 2.0354,
+      "step": 276735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015208138899846453,
+      "loss": 1.9819,
+      "step": 276740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015207981115813104,
+      "loss": 2.1236,
+      "step": 276745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015207823330000615,
+      "loss": 2.1055,
+      "step": 276750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015207665542409037,
+      "loss": 2.1915,
+      "step": 276755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015207507753038426,
+      "loss": 2.0989,
+      "step": 276760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015207349961888835,
+      "loss": 2.1631,
+      "step": 276765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015207192168960324,
+      "loss": 2.015,
+      "step": 276770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015207034374252937,
+      "loss": 2.1221,
+      "step": 276775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015206876577766735,
+      "loss": 2.115,
+      "step": 276780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015206718779501766,
+      "loss": 2.243,
+      "step": 276785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001520656097945809,
+      "loss": 1.9264,
+      "step": 276790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015206403177635763,
+      "loss": 1.9804,
+      "step": 276795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015206245374034828,
+      "loss": 2.2562,
+      "step": 276800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015206087568655344,
+      "loss": 2.1903,
+      "step": 276805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015205929761497372,
+      "loss": 2.0726,
+      "step": 276810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015205771952560957,
+      "loss": 2.2546,
+      "step": 276815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015205614141846155,
+      "loss": 2.1512,
+      "step": 276820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015205456329353022,
+      "loss": 2.2456,
+      "step": 276825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001520529851508161,
+      "loss": 2.0053,
+      "step": 276830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015205140699031975,
+      "loss": 1.9632,
+      "step": 276835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001520498288120417,
+      "loss": 2.0632,
+      "step": 276840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015204825061598243,
+      "loss": 1.986,
+      "step": 276845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001520466724021426,
+      "loss": 2.0762,
+      "step": 276850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015204509417052263,
+      "loss": 1.9217,
+      "step": 276855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015204351592112316,
+      "loss": 2.1611,
+      "step": 276860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015204193765394464,
+      "loss": 2.4422,
+      "step": 276865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015204035936898768,
+      "loss": 1.9299,
+      "step": 276870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015203878106625278,
+      "loss": 2.0499,
+      "step": 276875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.00015203720274574047,
+      "loss": 2.0528,
+      "step": 276880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015203562440745135,
+      "loss": 2.0867,
+      "step": 276885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015203404605138588,
+      "loss": 2.1223,
+      "step": 276890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015203246767754464,
+      "loss": 2.1591,
+      "step": 276895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015203088928592817,
+      "loss": 1.9631,
+      "step": 276900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015202931087653703,
+      "loss": 2.0283,
+      "step": 276905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001520277324493717,
+      "loss": 1.889,
+      "step": 276910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015202615400443275,
+      "loss": 2.1596,
+      "step": 276915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015202457554172076,
+      "loss": 2.125,
+      "step": 276920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015202299706123623,
+      "loss": 1.8072,
+      "step": 276925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015202141856297967,
+      "loss": 2.0803,
+      "step": 276930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015201984004695163,
+      "loss": 2.0046,
+      "step": 276935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001520182615131527,
+      "loss": 2.1132,
+      "step": 276940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001520166829615834,
+      "loss": 2.0502,
+      "step": 276945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015201510439224425,
+      "loss": 2.0297,
+      "step": 276950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015201352580513578,
+      "loss": 2.1004,
+      "step": 276955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015201194720025854,
+      "loss": 2.1336,
+      "step": 276960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001520103685776131,
+      "loss": 2.0761,
+      "step": 276965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001520087899372,
+      "loss": 2.1022,
+      "step": 276970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001520072112790197,
+      "loss": 2.0524,
+      "step": 276975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015200563260307282,
+      "loss": 2.1686,
+      "step": 276980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015200405390935985,
+      "loss": 2.0527,
+      "step": 276985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015200247519788138,
+      "loss": 2.1413,
+      "step": 276990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001520008964686379,
+      "loss": 1.8015,
+      "step": 276995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015199931772163,
+      "loss": 2.0816,
+      "step": 277000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015199773895685818,
+      "loss": 2.0319,
+      "step": 277005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015199616017432297,
+      "loss": 2.0086,
+      "step": 277010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015199458137402493,
+      "loss": 2.1276,
+      "step": 277015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015199300255596464,
+      "loss": 2.0412,
+      "step": 277020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015199142372014258,
+      "loss": 2.0809,
+      "step": 277025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00015198984486655925,
+      "loss": 2.0464,
+      "step": 277030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015198826599521528,
+      "loss": 2.2925,
+      "step": 277035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015198668710611122,
+      "loss": 1.7814,
+      "step": 277040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001519851081992475,
+      "loss": 2.1928,
+      "step": 277045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001519835292746248,
+      "loss": 2.0972,
+      "step": 277050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0001519819503322435,
+      "loss": 2.0047,
+      "step": 277055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015198037137210423,
+      "loss": 2.1054,
+      "step": 277060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015197879239420758,
+      "loss": 1.9934,
+      "step": 277065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000151977213398554,
+      "loss": 2.2228,
+      "step": 277070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015197563438514407,
+      "loss": 2.0399,
+      "step": 277075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015197405535397832,
+      "loss": 2.1001,
+      "step": 277080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001519724763050573,
+      "loss": 2.0485,
+      "step": 277085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015197089723838152,
+      "loss": 2.1086,
+      "step": 277090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015196931815395154,
+      "loss": 2.0457,
+      "step": 277095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015196773905176792,
+      "loss": 1.9847,
+      "step": 277100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015196615993183114,
+      "loss": 2.0691,
+      "step": 277105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001519645807941418,
+      "loss": 1.9294,
+      "step": 277110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001519630016387004,
+      "loss": 2.1909,
+      "step": 277115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015196142246550753,
+      "loss": 2.1462,
+      "step": 277120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015195984327456366,
+      "loss": 2.1099,
+      "step": 277125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001519582640658694,
+      "loss": 2.0569,
+      "step": 277130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015195668483942523,
+      "loss": 2.0965,
+      "step": 277135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015195510559523173,
+      "loss": 1.9472,
+      "step": 277140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001519535263332894,
+      "loss": 2.1036,
+      "step": 277145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015195194705359884,
+      "loss": 2.0399,
+      "step": 277150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015195036775616055,
+      "loss": 2.0035,
+      "step": 277155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015194878844097503,
+      "loss": 2.0751,
+      "step": 277160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015194720910804288,
+      "loss": 2.0097,
+      "step": 277165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015194562975736462,
+      "loss": 2.1117,
+      "step": 277170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001519440503889408,
+      "loss": 2.1548,
+      "step": 277175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015194247100277196,
+      "loss": 2.0573,
+      "step": 277180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015194089159885864,
+      "loss": 2.0946,
+      "step": 277185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001519393121772013,
+      "loss": 2.1274,
+      "step": 277190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001519377327378006,
+      "loss": 2.1917,
+      "step": 277195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015193615328065705,
+      "loss": 2.2505,
+      "step": 277200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015193457380577115,
+      "loss": 2.1817,
+      "step": 277205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015193299431314343,
+      "loss": 2.0288,
+      "step": 277210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001519314148027745,
+      "loss": 2.1741,
+      "step": 277215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001519298352746648,
+      "loss": 2.0045,
+      "step": 277220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015192825572881498,
+      "loss": 2.123,
+      "step": 277225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001519266761652255,
+      "loss": 2.0776,
+      "step": 277230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015192509658389693,
+      "loss": 2.0569,
+      "step": 277235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001519235169848298,
+      "loss": 1.9758,
+      "step": 277240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015192193736802466,
+      "loss": 2.2056,
+      "step": 277245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015192035773348205,
+      "loss": 1.9975,
+      "step": 277250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001519187780812025,
+      "loss": 2.0541,
+      "step": 277255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015191719841118654,
+      "loss": 1.9048,
+      "step": 277260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015191561872343474,
+      "loss": 2.0993,
+      "step": 277265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015191403901794763,
+      "loss": 2.078,
+      "step": 277270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015191245929472574,
+      "loss": 2.0453,
+      "step": 277275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015191087955376958,
+      "loss": 1.7257,
+      "step": 277280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015190929979507978,
+      "loss": 2.0989,
+      "step": 277285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001519077200186568,
+      "loss": 2.0484,
+      "step": 277290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001519061402245012,
+      "loss": 1.8643,
+      "step": 277295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015190456041261347,
+      "loss": 2.129,
+      "step": 277300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015190298058299426,
+      "loss": 1.8421,
+      "step": 277305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015190140073564402,
+      "loss": 2.0617,
+      "step": 277310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015189982087056335,
+      "loss": 2.1012,
+      "step": 277315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015189824098775276,
+      "loss": 2.1054,
+      "step": 277320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015189666108721275,
+      "loss": 2.1916,
+      "step": 277325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015189508116894392,
+      "loss": 1.9377,
+      "step": 277330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001518935012329468,
+      "loss": 1.9112,
+      "step": 277335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015189192127922194,
+      "loss": 2.1168,
+      "step": 277340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015189034130776982,
+      "loss": 2.0846,
+      "step": 277345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000151888761318591,
+      "loss": 2.1314,
+      "step": 277350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015188718131168606,
+      "loss": 2.1181,
+      "step": 277355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015188560128705556,
+      "loss": 2.2409,
+      "step": 277360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015188402124469995,
+      "loss": 2.1027,
+      "step": 277365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015188244118461986,
+      "loss": 2.1227,
+      "step": 277370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015188086110681575,
+      "loss": 2.1066,
+      "step": 277375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015187928101128816,
+      "loss": 2.1725,
+      "step": 277380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00015187770089803774,
+      "loss": 2.0517,
+      "step": 277385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015187612076706492,
+      "loss": 2.0795,
+      "step": 277390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015187454061837028,
+      "loss": 2.0828,
+      "step": 277395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015187296045195435,
+      "loss": 1.9607,
+      "step": 277400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015187138026781768,
+      "loss": 2.1704,
+      "step": 277405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001518698000659608,
+      "loss": 1.986,
+      "step": 277410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015186821984638428,
+      "loss": 2.1595,
+      "step": 277415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015186663960908863,
+      "loss": 2.1162,
+      "step": 277420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001518650593540744,
+      "loss": 2.0645,
+      "step": 277425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001518634790813421,
+      "loss": 2.0744,
+      "step": 277430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001518618987908923,
+      "loss": 2.0007,
+      "step": 277435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015186031848272554,
+      "loss": 2.1101,
+      "step": 277440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015185873815684237,
+      "loss": 2.1697,
+      "step": 277445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015185715781324333,
+      "loss": 2.1611,
+      "step": 277450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001518555774519289,
+      "loss": 2.1759,
+      "step": 277455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015185399707289964,
+      "loss": 2.094,
+      "step": 277460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015185241667615618,
+      "loss": 2.1604,
+      "step": 277465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015185083626169897,
+      "loss": 2.1859,
+      "step": 277470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015184925582952858,
+      "loss": 1.9737,
+      "step": 277475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015184767537964553,
+      "loss": 1.9271,
+      "step": 277480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015184609491205038,
+      "loss": 2.0175,
+      "step": 277485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015184451442674365,
+      "loss": 2.1405,
+      "step": 277490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001518429339237259,
+      "loss": 2.1009,
+      "step": 277495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001518413534029977,
+      "loss": 2.0,
+      "step": 277500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015183977286455954,
+      "loss": 2.1242,
+      "step": 277505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015183819230841193,
+      "loss": 2.1987,
+      "step": 277510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015183661173455548,
+      "loss": 2.1223,
+      "step": 277515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001518350311429907,
+      "loss": 1.9222,
+      "step": 277520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015183345053371814,
+      "loss": 2.0404,
+      "step": 277525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015183186990673834,
+      "loss": 2.2276,
+      "step": 277530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015183028926205183,
+      "loss": 2.3563,
+      "step": 277535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015182870859965913,
+      "loss": 2.1862,
+      "step": 277540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015182712791956082,
+      "loss": 2.1382,
+      "step": 277545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015182554722175743,
+      "loss": 2.0686,
+      "step": 277550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015182396650624948,
+      "loss": 2.2443,
+      "step": 277555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015182238577303754,
+      "loss": 2.2119,
+      "step": 277560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001518208050221221,
+      "loss": 1.8551,
+      "step": 277565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015181922425350376,
+      "loss": 2.1758,
+      "step": 277570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015181764346718305,
+      "loss": 2.0443,
+      "step": 277575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015181606266316045,
+      "loss": 2.1935,
+      "step": 277580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015181448184143655,
+      "loss": 2.0345,
+      "step": 277585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015181290100201196,
+      "loss": 1.9034,
+      "step": 277590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015181132014488707,
+      "loss": 2.0509,
+      "step": 277595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015180973927006247,
+      "loss": 1.9841,
+      "step": 277600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00015180815837753878,
+      "loss": 2.067,
+      "step": 277605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015180657746731647,
+      "loss": 2.0735,
+      "step": 277610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015180499653939607,
+      "loss": 1.9398,
+      "step": 277615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015180341559377817,
+      "loss": 2.081,
+      "step": 277620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015180183463046326,
+      "loss": 2.1254,
+      "step": 277625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015180025364945195,
+      "loss": 2.1257,
+      "step": 277630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015179867265074467,
+      "loss": 2.253,
+      "step": 277635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015179709163434208,
+      "loss": 2.0048,
+      "step": 277640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015179551060024463,
+      "loss": 1.9939,
+      "step": 277645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001517939295484529,
+      "loss": 2.0948,
+      "step": 277650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015179234847896743,
+      "loss": 2.1473,
+      "step": 277655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015179076739178874,
+      "loss": 2.1219,
+      "step": 277660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015178918628691742,
+      "loss": 1.9671,
+      "step": 277665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015178760516435395,
+      "loss": 2.2923,
+      "step": 277670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015178602402409887,
+      "loss": 2.1292,
+      "step": 277675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001517844428661528,
+      "loss": 2.1795,
+      "step": 277680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001517828616905162,
+      "loss": 2.134,
+      "step": 277685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001517812804971896,
+      "loss": 2.1909,
+      "step": 277690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015177969928617362,
+      "loss": 1.9901,
+      "step": 277695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015177811805746876,
+      "loss": 2.1805,
+      "step": 277700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015177653681107552,
+      "loss": 2.2892,
+      "step": 277705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001517749555469945,
+      "loss": 1.9704,
+      "step": 277710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015177337426522622,
+      "loss": 2.3436,
+      "step": 277715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001517717929657712,
+      "loss": 2.1504,
+      "step": 277720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015177021164863,
+      "loss": 1.9679,
+      "step": 277725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00015176863031380315,
+      "loss": 2.1515,
+      "step": 277730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001517670489612912,
+      "loss": 2.1029,
+      "step": 277735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015176546759109472,
+      "loss": 2.1225,
+      "step": 277740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001517638862032142,
+      "loss": 2.2523,
+      "step": 277745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001517623047976502,
+      "loss": 2.0655,
+      "step": 277750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001517607233744032,
+      "loss": 2.0509,
+      "step": 277755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015175914193347388,
+      "loss": 2.1004,
+      "step": 277760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015175756047486268,
+      "loss": 2.0871,
+      "step": 277765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00015175597899857012,
+      "loss": 2.079,
+      "step": 277770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015175439750459682,
+      "loss": 1.986,
+      "step": 277775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015175281599294325,
+      "loss": 2.1882,
+      "step": 277780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015175123446360998,
+      "loss": 2.0513,
+      "step": 277785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001517496529165976,
+      "loss": 2.1436,
+      "step": 277790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015174807135190653,
+      "loss": 2.1683,
+      "step": 277795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015174648976953742,
+      "loss": 2.0369,
+      "step": 277800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015174490816949074,
+      "loss": 2.0563,
+      "step": 277805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001517433265517671,
+      "loss": 1.9162,
+      "step": 277810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000151741744916367,
+      "loss": 2.0922,
+      "step": 277815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015174016326329097,
+      "loss": 1.9348,
+      "step": 277820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015173858159253954,
+      "loss": 2.1282,
+      "step": 277825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001517369999041133,
+      "loss": 2.0123,
+      "step": 277830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015173541819801277,
+      "loss": 2.0898,
+      "step": 277835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015173383647423843,
+      "loss": 2.0611,
+      "step": 277840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015173225473279092,
+      "loss": 2.0447,
+      "step": 277845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015173067297367073,
+      "loss": 1.7669,
+      "step": 277850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015172909119687836,
+      "loss": 2.1605,
+      "step": 277855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015172750940241445,
+      "loss": 1.9956,
+      "step": 277860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015172592759027945,
+      "loss": 2.3048,
+      "step": 277865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015172434576047396,
+      "loss": 2.0106,
+      "step": 277870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001517227639129985,
+      "loss": 1.9401,
+      "step": 277875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001517211820478536,
+      "loss": 1.933,
+      "step": 277880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001517196001650398,
+      "loss": 2.2263,
+      "step": 277885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015171801826455763,
+      "loss": 2.0162,
+      "step": 277890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015171643634640764,
+      "loss": 2.125,
+      "step": 277895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001517148544105904,
+      "loss": 2.1035,
+      "step": 277900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015171327245710645,
+      "loss": 2.1496,
+      "step": 277905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001517116904859563,
+      "loss": 2.1464,
+      "step": 277910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015171010849714044,
+      "loss": 2.2354,
+      "step": 277915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015170852649065952,
+      "loss": 2.2528,
+      "step": 277920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015170694446651404,
+      "loss": 2.0821,
+      "step": 277925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001517053624247045,
+      "loss": 1.9758,
+      "step": 277930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001517037803652315,
+      "loss": 2.104,
+      "step": 277935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001517021982880955,
+      "loss": 2.0189,
+      "step": 277940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015170061619329715,
+      "loss": 2.1742,
+      "step": 277945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001516990340808369,
+      "loss": 1.9241,
+      "step": 277950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015169745195071533,
+      "loss": 2.1176,
+      "step": 277955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015169586980293297,
+      "loss": 2.0853,
+      "step": 277960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015169428763749037,
+      "loss": 2.2531,
+      "step": 277965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015169270545438806,
+      "loss": 2.1022,
+      "step": 277970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015169112325362657,
+      "loss": 2.253,
+      "step": 277975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015168954103520648,
+      "loss": 2.1822,
+      "step": 277980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001516879587991283,
+      "loss": 2.0715,
+      "step": 277985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015168637654539255,
+      "loss": 1.9929,
+      "step": 277990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015168479427399984,
+      "loss": 2.004,
+      "step": 277995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015168321198495065,
+      "loss": 2.1874,
+      "step": 278000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001516816296782455,
+      "loss": 1.9252,
+      "step": 278005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015168004735388498,
+      "loss": 1.991,
+      "step": 278010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015167846501186967,
+      "loss": 1.9951,
+      "step": 278015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001516768826522,
+      "loss": 2.0144,
+      "step": 278020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015167530027487662,
+      "loss": 2.068,
+      "step": 278025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015167371787989998,
+      "loss": 2.0564,
+      "step": 278030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015167213546727067,
+      "loss": 2.1362,
+      "step": 278035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015167055303698924,
+      "loss": 2.0925,
+      "step": 278040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015166897058905621,
+      "loss": 2.0156,
+      "step": 278045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001516673881234721,
+      "loss": 1.9817,
+      "step": 278050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015166580564023748,
+      "loss": 2.2014,
+      "step": 278055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001516642231393529,
+      "loss": 2.1613,
+      "step": 278060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001516626406208189,
+      "loss": 2.2499,
+      "step": 278065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015166105808463596,
+      "loss": 2.1864,
+      "step": 278070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015165947553080468,
+      "loss": 2.0859,
+      "step": 278075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015165789295932557,
+      "loss": 1.9064,
+      "step": 278080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015165631037019925,
+      "loss": 2.1678,
+      "step": 278085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015165472776342616,
+      "loss": 2.0375,
+      "step": 278090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015165314513900687,
+      "loss": 1.9646,
+      "step": 278095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.875,
+      "learning_rate": 0.0001516515624969419,
+      "loss": 2.1269,
+      "step": 278100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015164997983723187,
+      "loss": 2.0644,
+      "step": 278105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015164839715987725,
+      "loss": 2.0215,
+      "step": 278110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015164681446487861,
+      "loss": 2.1823,
+      "step": 278115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015164523175223646,
+      "loss": 1.9441,
+      "step": 278120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015164364902195136,
+      "loss": 2.0717,
+      "step": 278125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015164206627402388,
+      "loss": 1.97,
+      "step": 278130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015164048350845452,
+      "loss": 2.0659,
+      "step": 278135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.00015163890072524385,
+      "loss": 2.0776,
+      "step": 278140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015163731792439236,
+      "loss": 1.9971,
+      "step": 278145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00015163573510590066,
+      "loss": 2.2495,
+      "step": 278150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015163415226976924,
+      "loss": 2.0721,
+      "step": 278155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015163256941599866,
+      "loss": 2.1252,
+      "step": 278160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015163098654458947,
+      "loss": 2.169,
+      "step": 278165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015162940365554217,
+      "loss": 2.1533,
+      "step": 278170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015162782074885738,
+      "loss": 2.0202,
+      "step": 278175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015162623782453554,
+      "loss": 2.0608,
+      "step": 278180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015162465488257725,
+      "loss": 2.218,
+      "step": 278185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015162307192298307,
+      "loss": 1.9274,
+      "step": 278190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015162148894575346,
+      "loss": 1.991,
+      "step": 278195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015161990595088907,
+      "loss": 2.1081,
+      "step": 278200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015161832293839037,
+      "loss": 2.1309,
+      "step": 278205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015161673990825788,
+      "loss": 2.2124,
+      "step": 278210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015161515686049222,
+      "loss": 2.0981,
+      "step": 278215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015161357379509386,
+      "loss": 2.0317,
+      "step": 278220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00015161199071206337,
+      "loss": 2.0353,
+      "step": 278225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015161040761140128,
+      "loss": 2.1486,
+      "step": 278230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015160882449310816,
+      "loss": 2.0169,
+      "step": 278235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001516072413571845,
+      "loss": 2.2864,
+      "step": 278240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015160565820363092,
+      "loss": 2.0467,
+      "step": 278245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001516040750324479,
+      "loss": 2.0181,
+      "step": 278250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015160249184363592,
+      "loss": 2.2078,
+      "step": 278255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015160090863719568,
+      "loss": 1.9722,
+      "step": 278260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001515993254131276,
+      "loss": 2.1311,
+      "step": 278265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015159774217143225,
+      "loss": 1.8855,
+      "step": 278270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015159615891211016,
+      "loss": 2.0777,
+      "step": 278275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015159457563516192,
+      "loss": 1.9872,
+      "step": 278280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015159299234058802,
+      "loss": 2.0361,
+      "step": 278285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015159140902838903,
+      "loss": 2.0511,
+      "step": 278290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00015158982569856544,
+      "loss": 1.9724,
+      "step": 278295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015158824235111788,
+      "loss": 1.9398,
+      "step": 278300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015158665898604682,
+      "loss": 2.0153,
+      "step": 278305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015158507560335282,
+      "loss": 2.0954,
+      "step": 278310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015158349220303644,
+      "loss": 1.96,
+      "step": 278315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015158190878509817,
+      "loss": 2.1304,
+      "step": 278320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001515803253495386,
+      "loss": 2.0528,
+      "step": 278325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015157874189635826,
+      "loss": 2.0532,
+      "step": 278330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015157715842555774,
+      "loss": 2.2158,
+      "step": 278335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015157557493713743,
+      "loss": 2.1898,
+      "step": 278340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000151573991431098,
+      "loss": 2.153,
+      "step": 278345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015157240790744,
+      "loss": 1.9486,
+      "step": 278350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001515708243661639,
+      "loss": 2.1575,
+      "step": 278355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015156924080727025,
+      "loss": 2.0667,
+      "step": 278360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015156765723075966,
+      "loss": 2.0492,
+      "step": 278365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015156607363663258,
+      "loss": 1.8874,
+      "step": 278370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001515644900248896,
+      "loss": 2.0442,
+      "step": 278375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015156290639553127,
+      "loss": 1.9767,
+      "step": 278380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001515613227485581,
+      "loss": 2.0712,
+      "step": 278385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015155973908397068,
+      "loss": 1.9865,
+      "step": 278390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015155815540176948,
+      "loss": 1.9637,
+      "step": 278395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001515565717019551,
+      "loss": 2.1947,
+      "step": 278400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015155498798452807,
+      "loss": 2.1287,
+      "step": 278405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001515534042494889,
+      "loss": 1.974,
+      "step": 278410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015155182049683816,
+      "loss": 2.0918,
+      "step": 278415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015155023672657638,
+      "loss": 2.1426,
+      "step": 278420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001515486529387041,
+      "loss": 2.0611,
+      "step": 278425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015154706913322188,
+      "loss": 2.047,
+      "step": 278430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015154548531013024,
+      "loss": 2.0525,
+      "step": 278435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015154390146942973,
+      "loss": 1.9371,
+      "step": 278440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015154231761112087,
+      "loss": 2.1613,
+      "step": 278445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015154073373520424,
+      "loss": 2.1343,
+      "step": 278450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015153914984168036,
+      "loss": 2.1025,
+      "step": 278455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015153756593054975,
+      "loss": 2.2083,
+      "step": 278460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015153598200181301,
+      "loss": 1.9944,
+      "step": 278465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001515343980554706,
+      "loss": 2.0165,
+      "step": 278470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015153281409152313,
+      "loss": 2.1744,
+      "step": 278475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001515312301099711,
+      "loss": 2.0568,
+      "step": 278480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001515296461108151,
+      "loss": 1.7919,
+      "step": 278485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015152806209405562,
+      "loss": 1.9578,
+      "step": 278490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015152647805969323,
+      "loss": 1.9154,
+      "step": 278495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015152489400772846,
+      "loss": 1.9994,
+      "step": 278500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001515233099381618,
+      "loss": 2.1138,
+      "step": 278505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015152172585099392,
+      "loss": 1.9912,
+      "step": 278510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015152014174622523,
+      "loss": 1.9527,
+      "step": 278515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015151855762385636,
+      "loss": 1.8812,
+      "step": 278520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001515169734838878,
+      "loss": 2.1094,
+      "step": 278525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015151538932632009,
+      "loss": 1.8166,
+      "step": 278530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015151380515115382,
+      "loss": 1.8836,
+      "step": 278535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001515122209583895,
+      "loss": 1.8861,
+      "step": 278540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015151063674802764,
+      "loss": 2.0534,
+      "step": 278545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001515090525200688,
+      "loss": 2.0276,
+      "step": 278550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015150746827451356,
+      "loss": 2.1488,
+      "step": 278555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015150588401136248,
+      "loss": 2.0651,
+      "step": 278560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015150429973061598,
+      "loss": 2.1913,
+      "step": 278565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015150271543227476,
+      "loss": 1.8607,
+      "step": 278570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015150113111633918,
+      "loss": 2.1047,
+      "step": 278575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015149954678280993,
+      "loss": 2.132,
+      "step": 278580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001514979624316875,
+      "loss": 2.0214,
+      "step": 278585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015149637806297244,
+      "loss": 2.1416,
+      "step": 278590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015149479367666528,
+      "loss": 1.9654,
+      "step": 278595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015149320927276656,
+      "loss": 2.1833,
+      "step": 278600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001514916248512768,
+      "loss": 2.0843,
+      "step": 278605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015149004041219662,
+      "loss": 1.8981,
+      "step": 278610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015148845595552648,
+      "loss": 2.0124,
+      "step": 278615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015148687148126694,
+      "loss": 2.2497,
+      "step": 278620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015148528698941856,
+      "loss": 2.0369,
+      "step": 278625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001514837024799819,
+      "loss": 2.1533,
+      "step": 278630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015148211795295744,
+      "loss": 2.2164,
+      "step": 278635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015148053340834574,
+      "loss": 2.0765,
+      "step": 278640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015147894884614738,
+      "loss": 2.0912,
+      "step": 278645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015147736426636288,
+      "loss": 1.9863,
+      "step": 278650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001514757796689928,
+      "loss": 2.311,
+      "step": 278655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001514741950540376,
+      "loss": 2.0787,
+      "step": 278660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015147261042149792,
+      "loss": 1.9267,
+      "step": 278665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015147102577137424,
+      "loss": 2.1601,
+      "step": 278670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015146944110366718,
+      "loss": 1.8748,
+      "step": 278675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015146785641837718,
+      "loss": 2.135,
+      "step": 278680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015146627171550484,
+      "loss": 2.117,
+      "step": 278685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015146468699505066,
+      "loss": 2.0628,
+      "step": 278690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015146310225701526,
+      "loss": 2.2179,
+      "step": 278695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001514615175013991,
+      "loss": 1.8264,
+      "step": 278700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015145993272820273,
+      "loss": 2.1473,
+      "step": 278705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015145834793742674,
+      "loss": 2.0548,
+      "step": 278710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015145676312907166,
+      "loss": 2.0588,
+      "step": 278715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000151455178303138,
+      "loss": 1.9602,
+      "step": 278720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015145359345962632,
+      "loss": 2.038,
+      "step": 278725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015145200859853714,
+      "loss": 2.1575,
+      "step": 278730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015145042371987106,
+      "loss": 2.1271,
+      "step": 278735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015144883882362855,
+      "loss": 2.1181,
+      "step": 278740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015144725390981017,
+      "loss": 2.0942,
+      "step": 278745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015144566897841652,
+      "loss": 2.1082,
+      "step": 278750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015144408402944807,
+      "loss": 2.1213,
+      "step": 278755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001514424990629054,
+      "loss": 2.1055,
+      "step": 278760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00015144091407878903,
+      "loss": 1.9072,
+      "step": 278765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001514393290770995,
+      "loss": 2.0698,
+      "step": 278770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001514377440578374,
+      "loss": 2.0331,
+      "step": 278775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001514361590210032,
+      "loss": 2.089,
+      "step": 278780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001514345739665975,
+      "loss": 2.0014,
+      "step": 278785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001514329888946208,
+      "loss": 2.1553,
+      "step": 278790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015143140380507366,
+      "loss": 2.1141,
+      "step": 278795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015142981869795664,
+      "loss": 2.1606,
+      "step": 278800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015142823357327024,
+      "loss": 1.9037,
+      "step": 278805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015142664843101501,
+      "loss": 2.1459,
+      "step": 278810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00015142506327119152,
+      "loss": 2.04,
+      "step": 278815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015142347809380028,
+      "loss": 1.9481,
+      "step": 278820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015142189289884186,
+      "loss": 2.0393,
+      "step": 278825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015142030768631678,
+      "loss": 1.9496,
+      "step": 278830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015141872245622563,
+      "loss": 2.1523,
+      "step": 278835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015141713720856888,
+      "loss": 2.1367,
+      "step": 278840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001514155519433471,
+      "loss": 2.0978,
+      "step": 278845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015141396666056082,
+      "loss": 2.043,
+      "step": 278850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015141238136021063,
+      "loss": 2.0978,
+      "step": 278855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000151410796042297,
+      "loss": 2.0879,
+      "step": 278860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015140921070682056,
+      "loss": 2.2489,
+      "step": 278865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001514076253537818,
+      "loss": 2.0905,
+      "step": 278870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015140603998318124,
+      "loss": 2.1103,
+      "step": 278875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015140445459501943,
+      "loss": 2.157,
+      "step": 278880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015140286918929692,
+      "loss": 2.1865,
+      "step": 278885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015140128376601428,
+      "loss": 2.2374,
+      "step": 278890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015139969832517203,
+      "loss": 2.1157,
+      "step": 278895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015139811286677071,
+      "loss": 2.0958,
+      "step": 278900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015139652739081086,
+      "loss": 2.0665,
+      "step": 278905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015139494189729305,
+      "loss": 2.2528,
+      "step": 278910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015139335638621776,
+      "loss": 2.1382,
+      "step": 278915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015139177085758557,
+      "loss": 2.0213,
+      "step": 278920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015139018531139703,
+      "loss": 2.2045,
+      "step": 278925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015138859974765266,
+      "loss": 1.8991,
+      "step": 278930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015138701416635302,
+      "loss": 2.1788,
+      "step": 278935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015138542856749863,
+      "loss": 1.9534,
+      "step": 278940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015138384295109008,
+      "loss": 2.0668,
+      "step": 278945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015138225731712787,
+      "loss": 2.0175,
+      "step": 278950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015138067166561254,
+      "loss": 2.1675,
+      "step": 278955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001513790859965446,
+      "loss": 1.9149,
+      "step": 278960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015137750030992466,
+      "loss": 2.1058,
+      "step": 278965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015137591460575325,
+      "loss": 2.0654,
+      "step": 278970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015137432888403092,
+      "loss": 2.0777,
+      "step": 278975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015137274314475815,
+      "loss": 2.1426,
+      "step": 278980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001513711573879355,
+      "loss": 2.1675,
+      "step": 278985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015136957161356358,
+      "loss": 2.1925,
+      "step": 278990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015136798582164284,
+      "loss": 1.9291,
+      "step": 278995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001513664000121739,
+      "loss": 2.2082,
+      "step": 279000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015136481418515725,
+      "loss": 2.301,
+      "step": 279005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015136322834059346,
+      "loss": 2.1112,
+      "step": 279010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015136164247848304,
+      "loss": 2.0164,
+      "step": 279015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015136005659882655,
+      "loss": 2.2421,
+      "step": 279020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015135847070162455,
+      "loss": 2.0244,
+      "step": 279025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015135688478687756,
+      "loss": 2.0056,
+      "step": 279030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015135529885458611,
+      "loss": 2.1171,
+      "step": 279035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001513537129047508,
+      "loss": 2.1717,
+      "step": 279040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015135212693737207,
+      "loss": 2.175,
+      "step": 279045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015135054095245052,
+      "loss": 1.9412,
+      "step": 279050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015134895494998675,
+      "loss": 2.0855,
+      "step": 279055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015134736892998124,
+      "loss": 2.0172,
+      "step": 279060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001513457828924345,
+      "loss": 2.1815,
+      "step": 279065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015134419683734716,
+      "loss": 1.7729,
+      "step": 279070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015134261076471964,
+      "loss": 2.1312,
+      "step": 279075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015134102467455262,
+      "loss": 2.2101,
+      "step": 279080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015133943856684656,
+      "loss": 2.1648,
+      "step": 279085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000151337852441602,
+      "loss": 2.1464,
+      "step": 279090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001513362662988195,
+      "loss": 1.975,
+      "step": 279095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001513346801384996,
+      "loss": 2.0141,
+      "step": 279100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015133309396064286,
+      "loss": 1.842,
+      "step": 279105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015133150776524981,
+      "loss": 2.0262,
+      "step": 279110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015132992155232096,
+      "loss": 2.2324,
+      "step": 279115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015132833532185687,
+      "loss": 2.2336,
+      "step": 279120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015132674907385808,
+      "loss": 2.1742,
+      "step": 279125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001513251628083252,
+      "loss": 1.9535,
+      "step": 279130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015132357652525871,
+      "loss": 2.0687,
+      "step": 279135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001513219902246591,
+      "loss": 2.1036,
+      "step": 279140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015132040390652697,
+      "loss": 2.0646,
+      "step": 279145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001513188175708629,
+      "loss": 1.7562,
+      "step": 279150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001513172312176674,
+      "loss": 1.9907,
+      "step": 279155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015131564484694097,
+      "loss": 1.9552,
+      "step": 279160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015131405845868418,
+      "loss": 2.0565,
+      "step": 279165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001513124720528976,
+      "loss": 2.2389,
+      "step": 279170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015131088562958173,
+      "loss": 1.9077,
+      "step": 279175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015130929918873712,
+      "loss": 2.1468,
+      "step": 279180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015130771273036436,
+      "loss": 2.1846,
+      "step": 279185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015130612625446395,
+      "loss": 2.2275,
+      "step": 279190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001513045397610364,
+      "loss": 2.1993,
+      "step": 279195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001513029532500823,
+      "loss": 2.019,
+      "step": 279200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001513013667216022,
+      "loss": 2.1201,
+      "step": 279205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015129978017559663,
+      "loss": 2.0326,
+      "step": 279210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001512981936120661,
+      "loss": 1.8774,
+      "step": 279215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015129660703101121,
+      "loss": 2.1024,
+      "step": 279220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015129502043243243,
+      "loss": 2.0716,
+      "step": 279225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015129343381633037,
+      "loss": 2.1036,
+      "step": 279230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015129184718270554,
+      "loss": 1.9206,
+      "step": 279235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015129026053155843,
+      "loss": 2.0993,
+      "step": 279240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015128867386288972,
+      "loss": 2.1855,
+      "step": 279245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001512870871766998,
+      "loss": 2.2613,
+      "step": 279250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001512855004729893,
+      "loss": 1.9908,
+      "step": 279255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015128391375175877,
+      "loss": 1.9227,
+      "step": 279260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015128232701300872,
+      "loss": 2.0496,
+      "step": 279265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001512807402567397,
+      "loss": 2.0613,
+      "step": 279270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015127915348295222,
+      "loss": 1.9688,
+      "step": 279275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015127756669164685,
+      "loss": 2.0107,
+      "step": 279280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001512759798828242,
+      "loss": 2.0665,
+      "step": 279285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015127439305648465,
+      "loss": 2.0867,
+      "step": 279290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001512728062126289,
+      "loss": 2.1343,
+      "step": 279295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015127121935125742,
+      "loss": 2.1532,
+      "step": 279300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015126963247237075,
+      "loss": 2.2071,
+      "step": 279305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015126804557596946,
+      "loss": 2.217,
+      "step": 279310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015126645866205403,
+      "loss": 1.9515,
+      "step": 279315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001512648717306251,
+      "loss": 2.0341,
+      "step": 279320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015126328478168316,
+      "loss": 2.1269,
+      "step": 279325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001512616978152287,
+      "loss": 1.9368,
+      "step": 279330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015126011083126234,
+      "loss": 2.0175,
+      "step": 279335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001512585238297846,
+      "loss": 1.9762,
+      "step": 279340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015125693681079601,
+      "loss": 2.0756,
+      "step": 279345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015125534977429716,
+      "loss": 2.1649,
+      "step": 279350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015125376272028852,
+      "loss": 2.0927,
+      "step": 279355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015125217564877067,
+      "loss": 2.1987,
+      "step": 279360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015125058855974417,
+      "loss": 2.1091,
+      "step": 279365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001512490014532095,
+      "loss": 2.1139,
+      "step": 279370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015124741432916725,
+      "loss": 2.2736,
+      "step": 279375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015124582718761798,
+      "loss": 2.2853,
+      "step": 279380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015124424002856222,
+      "loss": 1.9248,
+      "step": 279385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015124265285200043,
+      "loss": 2.0837,
+      "step": 279390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015124106565793326,
+      "loss": 2.2657,
+      "step": 279395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015123947844636123,
+      "loss": 2.0746,
+      "step": 279400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015123789121728484,
+      "loss": 1.9987,
+      "step": 279405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015123630397070469,
+      "loss": 2.0845,
+      "step": 279410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015123471670662123,
+      "loss": 2.1967,
+      "step": 279415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015123312942503508,
+      "loss": 2.2084,
+      "step": 279420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001512315421259468,
+      "loss": 2.0644,
+      "step": 279425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015122995480935686,
+      "loss": 2.0321,
+      "step": 279430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015122836747526584,
+      "loss": 1.8781,
+      "step": 279435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015122678012367428,
+      "loss": 2.087,
+      "step": 279440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015122519275458278,
+      "loss": 2.054,
+      "step": 279445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015122360536799176,
+      "loss": 2.1881,
+      "step": 279450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015122201796390186,
+      "loss": 2.1157,
+      "step": 279455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015122043054231358,
+      "loss": 2.0315,
+      "step": 279460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015121884310322745,
+      "loss": 1.9869,
+      "step": 279465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015121725564664406,
+      "loss": 2.0892,
+      "step": 279470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015121566817256392,
+      "loss": 2.0175,
+      "step": 279475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001512140806809876,
+      "loss": 2.0751,
+      "step": 279480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015121249317191557,
+      "loss": 1.9965,
+      "step": 279485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015121090564534847,
+      "loss": 2.1551,
+      "step": 279490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001512093181012868,
+      "loss": 1.922,
+      "step": 279495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015120773053973105,
+      "loss": 2.1125,
+      "step": 279500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015120614296068184,
+      "loss": 2.0739,
+      "step": 279505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015120455536413968,
+      "loss": 2.0949,
+      "step": 279510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015120296775010514,
+      "loss": 2.0411,
+      "step": 279515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001512013801185787,
+      "loss": 2.2261,
+      "step": 279520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015119979246956095,
+      "loss": 2.1653,
+      "step": 279525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015119820480305243,
+      "loss": 2.0855,
+      "step": 279530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015119661711905367,
+      "loss": 2.1138,
+      "step": 279535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015119502941756524,
+      "loss": 2.088,
+      "step": 279540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015119344169858763,
+      "loss": 2.0581,
+      "step": 279545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001511918539621214,
+      "loss": 2.059,
+      "step": 279550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015119026620816713,
+      "loss": 2.0104,
+      "step": 279555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015118867843672535,
+      "loss": 1.9884,
+      "step": 279560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015118709064779656,
+      "loss": 1.9918,
+      "step": 279565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015118550284138132,
+      "loss": 1.9807,
+      "step": 279570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001511839150174802,
+      "loss": 2.0596,
+      "step": 279575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015118232717609376,
+      "loss": 2.0067,
+      "step": 279580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015118073931722249,
+      "loss": 1.9328,
+      "step": 279585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015117915144086694,
+      "loss": 2.1713,
+      "step": 279590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015117756354702765,
+      "loss": 1.99,
+      "step": 279595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015117597563570517,
+      "loss": 1.9788,
+      "step": 279600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001511743877069001,
+      "loss": 2.0777,
+      "step": 279605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001511727997606129,
+      "loss": 2.1742,
+      "step": 279610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015117121179684413,
+      "loss": 2.0427,
+      "step": 279615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015116962381559436,
+      "loss": 2.145,
+      "step": 279620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015116803581686415,
+      "loss": 2.071,
+      "step": 279625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015116644780065398,
+      "loss": 2.004,
+      "step": 279630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015116485976696443,
+      "loss": 2.0482,
+      "step": 279635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000151163271715796,
+      "loss": 1.9753,
+      "step": 279640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001511616836471493,
+      "loss": 1.9553,
+      "step": 279645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015116009556102484,
+      "loss": 2.0743,
+      "step": 279650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015115850745742317,
+      "loss": 2.1255,
+      "step": 279655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015115691933634482,
+      "loss": 1.8672,
+      "step": 279660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015115533119779032,
+      "loss": 1.8849,
+      "step": 279665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00015115374304176026,
+      "loss": 1.8779,
+      "step": 279670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015115215486825515,
+      "loss": 2.0814,
+      "step": 279675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015115056667727554,
+      "loss": 1.9823,
+      "step": 279680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015114897846882195,
+      "loss": 2.1167,
+      "step": 279685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015114739024289496,
+      "loss": 2.0921,
+      "step": 279690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001511458019994951,
+      "loss": 2.0025,
+      "step": 279695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015114421373862286,
+      "loss": 2.0718,
+      "step": 279700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015114262546027886,
+      "loss": 2.2334,
+      "step": 279705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001511410371644636,
+      "loss": 2.0838,
+      "step": 279710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015113944885117764,
+      "loss": 1.9171,
+      "step": 279715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015113786052042155,
+      "loss": 2.2149,
+      "step": 279720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001511362721721958,
+      "loss": 1.8883,
+      "step": 279725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015113468380650099,
+      "loss": 2.1753,
+      "step": 279730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001511330954233376,
+      "loss": 2.0662,
+      "step": 279735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015113150702270628,
+      "loss": 1.8561,
+      "step": 279740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015112991860460748,
+      "loss": 1.9543,
+      "step": 279745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015112833016904179,
+      "loss": 1.9901,
+      "step": 279750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001511267417160097,
+      "loss": 2.1552,
+      "step": 279755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015112515324551182,
+      "loss": 2.2352,
+      "step": 279760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015112356475754864,
+      "loss": 2.1452,
+      "step": 279765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015112197625212072,
+      "loss": 1.9732,
+      "step": 279770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015112038772922866,
+      "loss": 2.1432,
+      "step": 279775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015111879918887287,
+      "loss": 2.1414,
+      "step": 279780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015111721063105403,
+      "loss": 2.0032,
+      "step": 279785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001511156220557726,
+      "loss": 2.0603,
+      "step": 279790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00015111403346302913,
+      "loss": 2.0597,
+      "step": 279795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015111244485282421,
+      "loss": 2.2274,
+      "step": 279800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015111085622515833,
+      "loss": 2.1682,
+      "step": 279805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015110926758003207,
+      "loss": 1.8763,
+      "step": 279810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015110767891744596,
+      "loss": 2.0655,
+      "step": 279815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015110609023740053,
+      "loss": 2.199,
+      "step": 279820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015110450153989633,
+      "loss": 2.053,
+      "step": 279825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001511029128249339,
+      "loss": 2.0228,
+      "step": 279830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015110132409251384,
+      "loss": 1.9898,
+      "step": 279835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015109973534263658,
+      "loss": 2.1194,
+      "step": 279840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001510981465753027,
+      "loss": 2.0479,
+      "step": 279845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015109655779051284,
+      "loss": 2.1333,
+      "step": 279850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015109496898826745,
+      "loss": 2.1561,
+      "step": 279855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015109338016856708,
+      "loss": 2.0786,
+      "step": 279860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015109179133141227,
+      "loss": 1.9783,
+      "step": 279865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001510902024768036,
+      "loss": 2.0628,
+      "step": 279870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015108861360474158,
+      "loss": 2.1823,
+      "step": 279875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015108702471522677,
+      "loss": 1.993,
+      "step": 279880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001510854358082597,
+      "loss": 2.2977,
+      "step": 279885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015108384688384093,
+      "loss": 1.9839,
+      "step": 279890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015108225794197095,
+      "loss": 1.9797,
+      "step": 279895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015108066898265041,
+      "loss": 2.0321,
+      "step": 279900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015107908000587972,
+      "loss": 2.1513,
+      "step": 279905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015107749101165953,
+      "loss": 2.0133,
+      "step": 279910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015107590199999032,
+      "loss": 1.9245,
+      "step": 279915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001510743129708727,
+      "loss": 2.0943,
+      "step": 279920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00015107272392430714,
+      "loss": 2.1469,
+      "step": 279925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015107113486029417,
+      "loss": 2.1876,
+      "step": 279930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015106954577883444,
+      "loss": 1.9908,
+      "step": 279935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001510679566799284,
+      "loss": 2.0294,
+      "step": 279940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015106636756357663,
+      "loss": 2.1468,
+      "step": 279945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015106477842977966,
+      "loss": 2.0676,
+      "step": 279950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015106318927853802,
+      "loss": 2.1481,
+      "step": 279955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001510616001098523,
+      "loss": 2.1247,
+      "step": 279960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015106001092372298,
+      "loss": 2.1088,
+      "step": 279965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015105842172015064,
+      "loss": 2.1384,
+      "step": 279970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015105683249913582,
+      "loss": 2.1676,
+      "step": 279975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015105524326067905,
+      "loss": 2.1594,
+      "step": 279980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001510536540047809,
+      "loss": 2.0673,
+      "step": 279985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001510520647314419,
+      "loss": 2.2745,
+      "step": 279990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015105047544066257,
+      "loss": 2.1319,
+      "step": 279995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001510488861324435,
+      "loss": 2.1052,
+      "step": 280000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015104729680678514,
+      "loss": 1.9988,
+      "step": 280005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015104570746368815,
+      "loss": 2.1499,
+      "step": 280010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015104411810315304,
+      "loss": 2.0863,
+      "step": 280015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001510425287251803,
+      "loss": 2.0665,
+      "step": 280020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015104093932977053,
+      "loss": 2.0485,
+      "step": 280025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015103934991692421,
+      "loss": 2.1913,
+      "step": 280030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015103776048664195,
+      "loss": 2.1339,
+      "step": 280035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015103617103892427,
+      "loss": 2.1047,
+      "step": 280040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001510345815737717,
+      "loss": 2.107,
+      "step": 280045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015103299209118478,
+      "loss": 2.1135,
+      "step": 280050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015103140259116406,
+      "loss": 2.0994,
+      "step": 280055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015102981307371012,
+      "loss": 1.9768,
+      "step": 280060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015102822353882346,
+      "loss": 2.112,
+      "step": 280065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001510266339865046,
+      "loss": 2.1066,
+      "step": 280070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015102504441675415,
+      "loss": 2.0017,
+      "step": 280075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015102345482957263,
+      "loss": 2.2347,
+      "step": 280080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015102186522496054,
+      "loss": 1.9701,
+      "step": 280085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015102027560291847,
+      "loss": 2.1143,
+      "step": 280090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015101868596344696,
+      "loss": 2.12,
+      "step": 280095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001510170963065465,
+      "loss": 1.9431,
+      "step": 280100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015101550663221774,
+      "loss": 2.028,
+      "step": 280105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015101391694046109,
+      "loss": 2.1246,
+      "step": 280110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015101232723127721,
+      "loss": 1.9259,
+      "step": 280115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015101073750466656,
+      "loss": 2.1089,
+      "step": 280120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015100914776062973,
+      "loss": 2.0315,
+      "step": 280125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015100755799916728,
+      "loss": 2.1158,
+      "step": 280130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001510059682202797,
+      "loss": 2.1149,
+      "step": 280135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015100437842396753,
+      "loss": 2.0997,
+      "step": 280140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015100278861023136,
+      "loss": 2.1572,
+      "step": 280145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015100119877907173,
+      "loss": 2.1977,
+      "step": 280150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015099960893048917,
+      "loss": 2.1168,
+      "step": 280155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015099801906448418,
+      "loss": 1.9788,
+      "step": 280160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015099642918105737,
+      "loss": 1.8175,
+      "step": 280165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015099483928020924,
+      "loss": 2.2267,
+      "step": 280170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015099324936194037,
+      "loss": 2.0846,
+      "step": 280175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015099165942625127,
+      "loss": 1.9378,
+      "step": 280180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015099006947314248,
+      "loss": 1.9812,
+      "step": 280185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015098847950261455,
+      "loss": 2.0296,
+      "step": 280190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015098688951466804,
+      "loss": 2.0057,
+      "step": 280195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001509852995093035,
+      "loss": 2.0586,
+      "step": 280200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015098370948652144,
+      "loss": 2.1135,
+      "step": 280205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015098211944632243,
+      "loss": 2.0429,
+      "step": 280210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.000150980529388707,
+      "loss": 1.975,
+      "step": 280215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001509789393136757,
+      "loss": 1.9172,
+      "step": 280220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015097734922122906,
+      "loss": 2.2361,
+      "step": 280225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015097575911136765,
+      "loss": 1.9149,
+      "step": 280230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.000150974168984092,
+      "loss": 2.0895,
+      "step": 280235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015097257883940263,
+      "loss": 2.0964,
+      "step": 280240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001509709886773001,
+      "loss": 1.9495,
+      "step": 280245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015096939849778497,
+      "loss": 2.2756,
+      "step": 280250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015096780830085776,
+      "loss": 1.9771,
+      "step": 280255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015096621808651902,
+      "loss": 2.0282,
+      "step": 280260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015096462785476936,
+      "loss": 1.927,
+      "step": 280265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015096303760560916,
+      "loss": 2.1296,
+      "step": 280270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001509614473390391,
+      "loss": 2.0056,
+      "step": 280275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001509598570550597,
+      "loss": 2.0621,
+      "step": 280280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015095826675367148,
+      "loss": 2.0747,
+      "step": 280285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000150956676434875,
+      "loss": 2.2144,
+      "step": 280290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015095508609867076,
+      "loss": 1.9731,
+      "step": 280295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00015095349574505935,
+      "loss": 1.9891,
+      "step": 280300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001509519053740413,
+      "loss": 2.1127,
+      "step": 280305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015095031498561722,
+      "loss": 1.9864,
+      "step": 280310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015094872457978753,
+      "loss": 2.2069,
+      "step": 280315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015094713415655283,
+      "loss": 2.213,
+      "step": 280320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015094554371591364,
+      "loss": 1.9479,
+      "step": 280325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001509439532578706,
+      "loss": 2.1665,
+      "step": 280330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015094236278242416,
+      "loss": 2.0,
+      "step": 280335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015094077228957486,
+      "loss": 2.0782,
+      "step": 280340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015093918177932326,
+      "loss": 2.2282,
+      "step": 280345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015093759125166996,
+      "loss": 2.3543,
+      "step": 280350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001509360007066154,
+      "loss": 2.0192,
+      "step": 280355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001509344101441602,
+      "loss": 2.0617,
+      "step": 280360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001509328195643049,
+      "loss": 2.051,
+      "step": 280365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015093122896705,
+      "loss": 2.085,
+      "step": 280370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001509296383523961,
+      "loss": 2.0486,
+      "step": 280375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015092804772034367,
+      "loss": 2.1733,
+      "step": 280380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001509264570708933,
+      "loss": 2.2519,
+      "step": 280385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015092486640404555,
+      "loss": 1.8923,
+      "step": 280390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015092327571980096,
+      "loss": 2.2902,
+      "step": 280395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015092168501816002,
+      "loss": 2.2061,
+      "step": 280400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001509200942991233,
+      "loss": 2.1187,
+      "step": 280405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001509185035626914,
+      "loss": 1.9346,
+      "step": 280410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015091691280886478,
+      "loss": 2.0256,
+      "step": 280415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015091532203764403,
+      "loss": 1.9466,
+      "step": 280420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015091373124902967,
+      "loss": 1.9766,
+      "step": 280425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001509121404430223,
+      "loss": 1.9732,
+      "step": 280430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015091054961962236,
+      "loss": 2.2366,
+      "step": 280435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015090895877883047,
+      "loss": 1.9482,
+      "step": 280440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001509073679206472,
+      "loss": 2.2519,
+      "step": 280445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000150905777045073,
+      "loss": 2.1356,
+      "step": 280450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015090418615210848,
+      "loss": 1.9514,
+      "step": 280455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015090259524175415,
+      "loss": 2.1443,
+      "step": 280460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001509010043140106,
+      "loss": 2.0992,
+      "step": 280465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015089941336887832,
+      "loss": 2.1656,
+      "step": 280470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015089782240635787,
+      "loss": 2.0396,
+      "step": 280475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015089623142644984,
+      "loss": 2.0296,
+      "step": 280480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001508946404291547,
+      "loss": 2.127,
+      "step": 280485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015089304941447304,
+      "loss": 2.0985,
+      "step": 280490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015089145838240539,
+      "loss": 2.1184,
+      "step": 280495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001508898673329523,
+      "loss": 1.9901,
+      "step": 280500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001508882762661143,
+      "loss": 2.1551,
+      "step": 280505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015088668518189197,
+      "loss": 2.0511,
+      "step": 280510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015088509408028576,
+      "loss": 2.0613,
+      "step": 280515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015088350296129638,
+      "loss": 2.2777,
+      "step": 280520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001508819118249242,
+      "loss": 2.0636,
+      "step": 280525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015088032067116984,
+      "loss": 1.9337,
+      "step": 280530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015087872950003383,
+      "loss": 2.1256,
+      "step": 280535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015087713831151675,
+      "loss": 2.1734,
+      "step": 280540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001508755471056191,
+      "loss": 2.0976,
+      "step": 280545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015087395588234145,
+      "loss": 2.0961,
+      "step": 280550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015087236464168436,
+      "loss": 2.0851,
+      "step": 280555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015087077338364833,
+      "loss": 2.0534,
+      "step": 280560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015086918210823388,
+      "loss": 2.1071,
+      "step": 280565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015086759081544163,
+      "loss": 2.1015,
+      "step": 280570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015086599950527209,
+      "loss": 2.1385,
+      "step": 280575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015086440817772582,
+      "loss": 2.2376,
+      "step": 280580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015086281683280333,
+      "loss": 1.9725,
+      "step": 280585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015086122547050516,
+      "loss": 2.0724,
+      "step": 280590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015085963409083186,
+      "loss": 2.057,
+      "step": 280595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.000150858042693784,
+      "loss": 2.103,
+      "step": 280600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015085645127936215,
+      "loss": 2.0559,
+      "step": 280605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001508548598475668,
+      "loss": 1.991,
+      "step": 280610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015085326839839848,
+      "loss": 2.1343,
+      "step": 280615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015085167693185776,
+      "loss": 2.0279,
+      "step": 280620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015085008544794522,
+      "loss": 2.2106,
+      "step": 280625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015084849394666133,
+      "loss": 2.1701,
+      "step": 280630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015084690242800672,
+      "loss": 1.8385,
+      "step": 280635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015084531089198185,
+      "loss": 2.1329,
+      "step": 280640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015084371933858726,
+      "loss": 2.1389,
+      "step": 280645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015084212776782363,
+      "loss": 2.0442,
+      "step": 280650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015084053617969137,
+      "loss": 1.9913,
+      "step": 280655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015083894457419103,
+      "loss": 2.1503,
+      "step": 280660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001508373529513232,
+      "loss": 2.1614,
+      "step": 280665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001508357613110884,
+      "loss": 1.9971,
+      "step": 280670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001508341696534872,
+      "loss": 2.0319,
+      "step": 280675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001508325779785201,
+      "loss": 2.1737,
+      "step": 280680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001508309862861877,
+      "loss": 2.0415,
+      "step": 280685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001508293945764905,
+      "loss": 2.1731,
+      "step": 280690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015082780284942908,
+      "loss": 2.2296,
+      "step": 280695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015082621110500393,
+      "loss": 2.1623,
+      "step": 280700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015082461934321563,
+      "loss": 2.0791,
+      "step": 280705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001508230275640647,
+      "loss": 1.9342,
+      "step": 280710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015082143576755175,
+      "loss": 2.151,
+      "step": 280715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015081984395367724,
+      "loss": 1.9352,
+      "step": 280720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015081825212244176,
+      "loss": 1.9909,
+      "step": 280725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015081666027384584,
+      "loss": 1.9739,
+      "step": 280730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015081506840789003,
+      "loss": 2.0141,
+      "step": 280735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015081347652457487,
+      "loss": 2.17,
+      "step": 280740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015081188462390092,
+      "loss": 1.994,
+      "step": 280745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015081029270586869,
+      "loss": 1.9177,
+      "step": 280750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00015080870077047872,
+      "loss": 1.944,
+      "step": 280755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015080710881773163,
+      "loss": 2.2111,
+      "step": 280760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001508055168476279,
+      "loss": 2.0841,
+      "step": 280765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015080392486016802,
+      "loss": 1.9479,
+      "step": 280770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015080233285535267,
+      "loss": 1.8282,
+      "step": 280775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015080074083318227,
+      "loss": 1.9559,
+      "step": 280780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015079914879365748,
+      "loss": 2.1023,
+      "step": 280785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015079755673677873,
+      "loss": 1.9695,
+      "step": 280790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015079596466254663,
+      "loss": 2.0561,
+      "step": 280795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001507943725709617,
+      "loss": 2.0795,
+      "step": 280800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001507927804620245,
+      "loss": 1.9127,
+      "step": 280805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015079118833573554,
+      "loss": 2.26,
+      "step": 280810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015078959619209542,
+      "loss": 1.9745,
+      "step": 280815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015078800403110464,
+      "loss": 1.9876,
+      "step": 280820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015078641185276376,
+      "loss": 2.0494,
+      "step": 280825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001507848196570733,
+      "loss": 2.0936,
+      "step": 280830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015078322744403386,
+      "loss": 2.1693,
+      "step": 280835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015078163521364593,
+      "loss": 1.9775,
+      "step": 280840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001507800429659101,
+      "loss": 2.1967,
+      "step": 280845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015077845070082682,
+      "loss": 2.0717,
+      "step": 280850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015077685841839678,
+      "loss": 2.1101,
+      "step": 280855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015077526611862035,
+      "loss": 2.0538,
+      "step": 280860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015077367380149825,
+      "loss": 1.991,
+      "step": 280865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001507720814670309,
+      "loss": 2.1092,
+      "step": 280870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015077048911521894,
+      "loss": 1.9701,
+      "step": 280875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001507688967460628,
+      "loss": 2.0693,
+      "step": 280880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015076730435956313,
+      "loss": 2.0389,
+      "step": 280885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015076571195572037,
+      "loss": 2.1125,
+      "step": 280890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015076411953453517,
+      "loss": 1.8861,
+      "step": 280895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000150762527096008,
+      "loss": 2.1972,
+      "step": 280900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015076093464013948,
+      "loss": 2.1661,
+      "step": 280905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015075934216693003,
+      "loss": 2.0685,
+      "step": 280910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001507577496763803,
+      "loss": 2.1006,
+      "step": 280915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015075615716849083,
+      "loss": 2.0974,
+      "step": 280920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00015075456464326212,
+      "loss": 2.0954,
+      "step": 280925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001507529721006947,
+      "loss": 2.0975,
+      "step": 280930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015075137954078917,
+      "loss": 1.885,
+      "step": 280935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015074978696354602,
+      "loss": 2.2031,
+      "step": 280940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015074819436896588,
+      "loss": 2.1341,
+      "step": 280945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001507466017570492,
+      "loss": 2.0353,
+      "step": 280950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015074500912779656,
+      "loss": 2.0577,
+      "step": 280955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015074341648120852,
+      "loss": 2.0409,
+      "step": 280960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001507418238172856,
+      "loss": 2.0336,
+      "step": 280965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015074023113602832,
+      "loss": 1.995,
+      "step": 280970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001507386384374373,
+      "loss": 2.0327,
+      "step": 280975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015073704572151305,
+      "loss": 1.8432,
+      "step": 280980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015073545298825608,
+      "loss": 2.0309,
+      "step": 280985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015073386023766697,
+      "loss": 2.0344,
+      "step": 280990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015073226746974621,
+      "loss": 1.965,
+      "step": 280995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015073067468449443,
+      "loss": 2.1389,
+      "step": 281000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015072908188191214,
+      "loss": 2.1483,
+      "step": 281005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015072748906199985,
+      "loss": 1.9117,
+      "step": 281010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015072589622475813,
+      "loss": 2.1697,
+      "step": 281015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015072430337018753,
+      "loss": 2.167,
+      "step": 281020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001507227104982886,
+      "loss": 1.999,
+      "step": 281025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015072111760906188,
+      "loss": 2.0139,
+      "step": 281030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001507195247025079,
+      "loss": 2.0337,
+      "step": 281035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015071793177862718,
+      "loss": 2.0934,
+      "step": 281040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015071633883742035,
+      "loss": 1.95,
+      "step": 281045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015071474587888782,
+      "loss": 2.0502,
+      "step": 281050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001507131529030303,
+      "loss": 2.1712,
+      "step": 281055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00015071155990984818,
+      "loss": 1.9204,
+      "step": 281060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001507099668993421,
+      "loss": 1.9803,
+      "step": 281065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015070837387151257,
+      "loss": 2.0891,
+      "step": 281070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015070678082636013,
+      "loss": 2.1909,
+      "step": 281075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015070518776388536,
+      "loss": 2.016,
+      "step": 281080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00015070359468408875,
+      "loss": 2.3092,
+      "step": 281085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001507020015869709,
+      "loss": 1.9802,
+      "step": 281090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001507004084725323,
+      "loss": 2.0433,
+      "step": 281095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015069881534077355,
+      "loss": 1.9137,
+      "step": 281100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015069722219169512,
+      "loss": 2.0053,
+      "step": 281105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015069562902529765,
+      "loss": 2.1689,
+      "step": 281110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015069403584158162,
+      "loss": 2.0873,
+      "step": 281115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015069244264054755,
+      "loss": 1.9273,
+      "step": 281120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015069084942219608,
+      "loss": 2.046,
+      "step": 281125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015068925618652766,
+      "loss": 2.0476,
+      "step": 281130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015068766293354288,
+      "loss": 1.9741,
+      "step": 281135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015068606966324227,
+      "loss": 1.9969,
+      "step": 281140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001506844763756264,
+      "loss": 2.0414,
+      "step": 281145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001506828830706958,
+      "loss": 2.0162,
+      "step": 281150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015068128974845096,
+      "loss": 2.2992,
+      "step": 281155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015067969640889247,
+      "loss": 2.072,
+      "step": 281160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015067810305202092,
+      "loss": 2.2015,
+      "step": 281165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001506765096778368,
+      "loss": 1.9867,
+      "step": 281170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001506749162863407,
+      "loss": 1.9318,
+      "step": 281175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015067332287753307,
+      "loss": 2.0486,
+      "step": 281180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015067172945141454,
+      "loss": 2.0511,
+      "step": 281185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001506701360079856,
+      "loss": 1.9652,
+      "step": 281190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015066854254724686,
+      "loss": 2.1673,
+      "step": 281195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001506669490691988,
+      "loss": 2.1183,
+      "step": 281200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015066535557384204,
+      "loss": 1.9187,
+      "step": 281205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000150663762061177,
+      "loss": 2.0745,
+      "step": 281210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015066216853120437,
+      "loss": 1.9705,
+      "step": 281215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001506605749839246,
+      "loss": 2.1761,
+      "step": 281220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015065898141933823,
+      "loss": 2.0641,
+      "step": 281225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015065738783744584,
+      "loss": 2.0497,
+      "step": 281230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015065579423824798,
+      "loss": 2.0331,
+      "step": 281235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001506542006217452,
+      "loss": 2.0064,
+      "step": 281240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015065260698793803,
+      "loss": 2.0987,
+      "step": 281245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000150651013336827,
+      "loss": 2.0199,
+      "step": 281250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015064941966841264,
+      "loss": 2.0933,
+      "step": 281255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015064782598269553,
+      "loss": 2.0345,
+      "step": 281260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015064623227967623,
+      "loss": 1.9712,
+      "step": 281265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015064463855935523,
+      "loss": 2.1037,
+      "step": 281270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001506430448217331,
+      "loss": 1.9872,
+      "step": 281275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001506414510668104,
+      "loss": 2.2342,
+      "step": 281280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001506398572945877,
+      "loss": 2.0468,
+      "step": 281285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015063826350506543,
+      "loss": 2.0494,
+      "step": 281290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015063666969824426,
+      "loss": 2.0953,
+      "step": 281295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015063507587412467,
+      "loss": 2.1474,
+      "step": 281300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015063348203270724,
+      "loss": 1.9296,
+      "step": 281305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001506318881739925,
+      "loss": 2.3524,
+      "step": 281310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015063029429798094,
+      "loss": 2.0854,
+      "step": 281315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015062870040467318,
+      "loss": 2.1716,
+      "step": 281320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001506271064940697,
+      "loss": 2.2238,
+      "step": 281325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015062551256617116,
+      "loss": 2.1375,
+      "step": 281330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000150623918620978,
+      "loss": 1.9742,
+      "step": 281335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001506223246584908,
+      "loss": 2.0194,
+      "step": 281340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015062073067871002,
+      "loss": 1.9578,
+      "step": 281345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015061913668163633,
+      "loss": 2.0269,
+      "step": 281350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015061754266727025,
+      "loss": 2.1332,
+      "step": 281355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015061594863561228,
+      "loss": 2.1693,
+      "step": 281360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000150614354586663,
+      "loss": 2.2332,
+      "step": 281365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015061276052042287,
+      "loss": 2.0042,
+      "step": 281370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001506111664368926,
+      "loss": 1.9944,
+      "step": 281375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015060957233607259,
+      "loss": 2.103,
+      "step": 281380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015060797821796342,
+      "loss": 2.0514,
+      "step": 281385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015060638408256568,
+      "loss": 2.1691,
+      "step": 281390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015060478992987985,
+      "loss": 2.0235,
+      "step": 281395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015060319575990652,
+      "loss": 2.1324,
+      "step": 281400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015060160157264623,
+      "loss": 1.761,
+      "step": 281405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015060000736809949,
+      "loss": 2.2398,
+      "step": 281410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001505984131462669,
+      "loss": 2.1367,
+      "step": 281415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015059681890714895,
+      "loss": 2.1085,
+      "step": 281420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015059522465074622,
+      "loss": 1.8977,
+      "step": 281425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015059363037705925,
+      "loss": 1.9386,
+      "step": 281430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015059203608608856,
+      "loss": 2.0119,
+      "step": 281435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015059044177783473,
+      "loss": 2.0176,
+      "step": 281440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015058884745229828,
+      "loss": 2.1649,
+      "step": 281445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015058725310947976,
+      "loss": 2.0268,
+      "step": 281450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001505856587493797,
+      "loss": 1.9633,
+      "step": 281455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001505840643719987,
+      "loss": 1.9397,
+      "step": 281460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015058246997733728,
+      "loss": 2.1091,
+      "step": 281465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015058087556539591,
+      "loss": 2.0924,
+      "step": 281470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001505792811361752,
+      "loss": 2.0805,
+      "step": 281475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001505776866896757,
+      "loss": 2.1114,
+      "step": 281480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015057609222589798,
+      "loss": 2.193,
+      "step": 281485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015057449774484255,
+      "loss": 2.2144,
+      "step": 281490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015057290324650992,
+      "loss": 1.9594,
+      "step": 281495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015057130873090069,
+      "loss": 1.9758,
+      "step": 281500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015056971419801535,
+      "loss": 1.9887,
+      "step": 281505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015056811964785454,
+      "loss": 2.1143,
+      "step": 281510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015056652508041871,
+      "loss": 2.2196,
+      "step": 281515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00015056493049570843,
+      "loss": 1.9996,
+      "step": 281520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015056333589372425,
+      "loss": 2.0603,
+      "step": 281525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001505617412744667,
+      "loss": 2.1191,
+      "step": 281530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001505601466379364,
+      "loss": 1.9475,
+      "step": 281535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001505585519841338,
+      "loss": 2.0374,
+      "step": 281540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001505569573130595,
+      "loss": 2.1045,
+      "step": 281545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000150555362624714,
+      "loss": 2.0231,
+      "step": 281550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001505537679190979,
+      "loss": 2.0027,
+      "step": 281555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001505521731962117,
+      "loss": 2.0418,
+      "step": 281560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015055057845605595,
+      "loss": 1.8496,
+      "step": 281565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015054898369863123,
+      "loss": 2.1291,
+      "step": 281570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015054738892393804,
+      "loss": 1.9476,
+      "step": 281575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015054579413197695,
+      "loss": 2.0028,
+      "step": 281580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001505441993227485,
+      "loss": 2.0401,
+      "step": 281585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015054260449625325,
+      "loss": 2.1156,
+      "step": 281590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.0001505410096524917,
+      "loss": 2.1383,
+      "step": 281595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015053941479146446,
+      "loss": 2.0436,
+      "step": 281600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000150537819913172,
+      "loss": 2.0245,
+      "step": 281605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015053622501761492,
+      "loss": 2.0791,
+      "step": 281610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015053463010479377,
+      "loss": 2.0327,
+      "step": 281615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015053303517470904,
+      "loss": 1.9347,
+      "step": 281620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015053144022736135,
+      "loss": 1.9448,
+      "step": 281625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015052984526275117,
+      "loss": 2.0687,
+      "step": 281630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001505282502808791,
+      "loss": 2.1464,
+      "step": 281635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00015052665528174564,
+      "loss": 2.0364,
+      "step": 281640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015052506026535137,
+      "loss": 2.0888,
+      "step": 281645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015052346523169682,
+      "loss": 2.0637,
+      "step": 281650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015052187018078255,
+      "loss": 2.2682,
+      "step": 281655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001505202751126091,
+      "loss": 2.0296,
+      "step": 281660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015051868002717694,
+      "loss": 2.1171,
+      "step": 281665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015051708492448676,
+      "loss": 2.1062,
+      "step": 281670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015051548980453901,
+      "loss": 1.9663,
+      "step": 281675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015051389466733426,
+      "loss": 2.1431,
+      "step": 281680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015051229951287302,
+      "loss": 2.1463,
+      "step": 281685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015051070434115587,
+      "loss": 1.9957,
+      "step": 281690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00015050910915218334,
+      "loss": 2.1155,
+      "step": 281695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015050751394595603,
+      "loss": 2.0093,
+      "step": 281700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001505059187224744,
+      "loss": 2.2871,
+      "step": 281705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015050432348173902,
+      "loss": 1.9559,
+      "step": 281710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015050272822375046,
+      "loss": 2.0486,
+      "step": 281715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015050113294850926,
+      "loss": 1.9092,
+      "step": 281720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001504995376560159,
+      "loss": 2.0627,
+      "step": 281725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001504979423462711,
+      "loss": 2.1536,
+      "step": 281730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001504963470192752,
+      "loss": 1.9627,
+      "step": 281735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015049475167502885,
+      "loss": 2.158,
+      "step": 281740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015049315631353256,
+      "loss": 1.9753,
+      "step": 281745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015049156093478692,
+      "loss": 1.8833,
+      "step": 281750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015048996553879243,
+      "loss": 2.0888,
+      "step": 281755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00015048837012554968,
+      "loss": 1.9075,
+      "step": 281760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015048677469505916,
+      "loss": 1.9358,
+      "step": 281765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015048517924732145,
+      "loss": 2.0388,
+      "step": 281770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015048358378233706,
+      "loss": 2.2378,
+      "step": 281775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001504819883001066,
+      "loss": 2.221,
+      "step": 281780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015048039280063057,
+      "loss": 2.1294,
+      "step": 281785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015047879728390955,
+      "loss": 2.0229,
+      "step": 281790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015047720174994402,
+      "loss": 1.9843,
+      "step": 281795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015047560619873453,
+      "loss": 2.1782,
+      "step": 281800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015047401063028173,
+      "loss": 2.0704,
+      "step": 281805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015047241504458604,
+      "loss": 2.0447,
+      "step": 281810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015047081944164807,
+      "loss": 2.1753,
+      "step": 281815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00015046922382146836,
+      "loss": 2.2313,
+      "step": 281820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015046762818404742,
+      "loss": 1.9811,
+      "step": 281825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001504660325293859,
+      "loss": 1.8547,
+      "step": 281830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001504644368574842,
+      "loss": 2.084,
+      "step": 281835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015046284116834296,
+      "loss": 2.1705,
+      "step": 281840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00015046124546196267,
+      "loss": 2.0263,
+      "step": 281845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015045964973834391,
+      "loss": 2.0224,
+      "step": 281850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015045805399748724,
+      "loss": 2.2329,
+      "step": 281855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015045645823939316,
+      "loss": 1.9047,
+      "step": 281860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015045486246406225,
+      "loss": 2.0395,
+      "step": 281865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015045326667149507,
+      "loss": 1.9806,
+      "step": 281870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015045167086169214,
+      "loss": 2.229,
+      "step": 281875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015045007503465394,
+      "loss": 2.0338,
+      "step": 281880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001504484791903811,
+      "loss": 2.11,
+      "step": 281885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00015044688332887417,
+      "loss": 1.9366,
+      "step": 281890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015044528745013368,
+      "loss": 2.0464,
+      "step": 281895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00015044369155416018,
+      "loss": 2.12,
+      "step": 281900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015044209564095415,
+      "loss": 2.0592,
+      "step": 281905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001504404997105162,
+      "loss": 2.0561,
+      "step": 281910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015043890376284687,
+      "loss": 2.1079,
+      "step": 281915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001504373077979467,
+      "loss": 2.1147,
+      "step": 281920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015043571181581623,
+      "loss": 2.0815,
+      "step": 281925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000150434115816456,
+      "loss": 2.208,
+      "step": 281930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015043251979986652,
+      "loss": 1.9852,
+      "step": 281935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015043092376604842,
+      "loss": 2.0403,
+      "step": 281940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001504293277150022,
+      "loss": 1.7871,
+      "step": 281945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001504277316467284,
+      "loss": 2.2782,
+      "step": 281950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015042613556122758,
+      "loss": 2.1437,
+      "step": 281955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015042453945850025,
+      "loss": 2.0393,
+      "step": 281960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015042294333854702,
+      "loss": 1.9463,
+      "step": 281965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001504213472013684,
+      "loss": 2.1368,
+      "step": 281970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015041975104696493,
+      "loss": 1.9031,
+      "step": 281975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015041815487533712,
+      "loss": 2.008,
+      "step": 281980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001504165586864856,
+      "loss": 2.1866,
+      "step": 281985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00015041496248041086,
+      "loss": 2.1119,
+      "step": 281990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015041336625711344,
+      "loss": 1.9706,
+      "step": 281995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001504117700165939,
+      "loss": 2.0391,
+      "step": 282000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001504101737588528,
+      "loss": 1.9842,
+      "step": 282005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015040857748389064,
+      "loss": 2.0332,
+      "step": 282010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015040698119170802,
+      "loss": 1.9035,
+      "step": 282015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015040538488230544,
+      "loss": 2.0202,
+      "step": 282020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001504037885556835,
+      "loss": 2.1448,
+      "step": 282025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001504021922118427,
+      "loss": 2.027,
+      "step": 282030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001504005958507836,
+      "loss": 1.9568,
+      "step": 282035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001503989994725067,
+      "loss": 2.0867,
+      "step": 282040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015039740307701263,
+      "loss": 2.141,
+      "step": 282045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001503958066643019,
+      "loss": 2.0438,
+      "step": 282050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015039421023437503,
+      "loss": 2.1556,
+      "step": 282055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001503926137872326,
+      "loss": 2.1521,
+      "step": 282060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001503910173228751,
+      "loss": 1.863,
+      "step": 282065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015038942084130317,
+      "loss": 2.0311,
+      "step": 282070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015038782434251724,
+      "loss": 2.0547,
+      "step": 282075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015038622782651798,
+      "loss": 2.1886,
+      "step": 282080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015038463129330584,
+      "loss": 1.9982,
+      "step": 282085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015038303474288137,
+      "loss": 1.9882,
+      "step": 282090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015038143817524516,
+      "loss": 2.1045,
+      "step": 282095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015037984159039779,
+      "loss": 2.1834,
+      "step": 282100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015037824498833968,
+      "loss": 2.1751,
+      "step": 282105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015037664836907146,
+      "loss": 1.9776,
+      "step": 282110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015037505173259367,
+      "loss": 2.0438,
+      "step": 282115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00015037345507890683,
+      "loss": 1.8402,
+      "step": 282120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015037185840801156,
+      "loss": 2.0695,
+      "step": 282125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015037026171990832,
+      "loss": 1.8788,
+      "step": 282130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015036866501459768,
+      "loss": 2.1142,
+      "step": 282135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00015036706829208018,
+      "loss": 2.1177,
+      "step": 282140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001503654715523564,
+      "loss": 2.1305,
+      "step": 282145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015036387479542684,
+      "loss": 2.0481,
+      "step": 282150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001503622780212921,
+      "loss": 2.0329,
+      "step": 282155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015036068122995265,
+      "loss": 2.0531,
+      "step": 282160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001503590844214091,
+      "loss": 1.9704,
+      "step": 282165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000150357487595662,
+      "loss": 2.1182,
+      "step": 282170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015035589075271184,
+      "loss": 2.0976,
+      "step": 282175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015035429389255916,
+      "loss": 2.1044,
+      "step": 282180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001503526970152046,
+      "loss": 1.9968,
+      "step": 282185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015035110012064863,
+      "loss": 2.0532,
+      "step": 282190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001503495032088918,
+      "loss": 2.1174,
+      "step": 282195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015034790627993465,
+      "loss": 2.181,
+      "step": 282200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015034630933377776,
+      "loss": 2.1013,
+      "step": 282205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015034471237042167,
+      "loss": 1.9908,
+      "step": 282210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001503431153898669,
+      "loss": 2.2246,
+      "step": 282215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000150341518392114,
+      "loss": 2.159,
+      "step": 282220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015033992137716351,
+      "loss": 1.9949,
+      "step": 282225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015033832434501604,
+      "loss": 2.0452,
+      "step": 282230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015033672729567206,
+      "loss": 2.1578,
+      "step": 282235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015033513022913217,
+      "loss": 2.1684,
+      "step": 282240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015033353314539683,
+      "loss": 2.1607,
+      "step": 282245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001503319360444667,
+      "loss": 2.057,
+      "step": 282250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015033033892634218,
+      "loss": 2.0383,
+      "step": 282255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.000150328741791024,
+      "loss": 2.0724,
+      "step": 282260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00015032714463851255,
+      "loss": 2.1594,
+      "step": 282265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015032554746880847,
+      "loss": 2.0362,
+      "step": 282270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015032395028191227,
+      "loss": 1.9325,
+      "step": 282275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015032235307782448,
+      "loss": 2.0092,
+      "step": 282280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015032075585654563,
+      "loss": 2.102,
+      "step": 282285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015031915861807635,
+      "loss": 2.2028,
+      "step": 282290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015031756136241713,
+      "loss": 2.0015,
+      "step": 282295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015031596408956847,
+      "loss": 1.9713,
+      "step": 282300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015031436679953099,
+      "loss": 2.0638,
+      "step": 282305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001503127694923052,
+      "loss": 2.1819,
+      "step": 282310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001503111721678917,
+      "loss": 1.8671,
+      "step": 282315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015030957482629095,
+      "loss": 1.9631,
+      "step": 282320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015030797746750355,
+      "loss": 2.0789,
+      "step": 282325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015030638009153006,
+      "loss": 2.1037,
+      "step": 282330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015030478269837096,
+      "loss": 2.0837,
+      "step": 282335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015030318528802683,
+      "loss": 1.9947,
+      "step": 282340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015030158786049825,
+      "loss": 1.9252,
+      "step": 282345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001502999904157857,
+      "loss": 1.9092,
+      "step": 282350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001502983929538898,
+      "loss": 2.0285,
+      "step": 282355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000150296795474811,
+      "loss": 2.047,
+      "step": 282360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015029519797854995,
+      "loss": 1.959,
+      "step": 282365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015029360046510715,
+      "loss": 2.2173,
+      "step": 282370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015029200293448315,
+      "loss": 2.1548,
+      "step": 282375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00015029040538667844,
+      "loss": 2.1157,
+      "step": 282380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015028880782169365,
+      "loss": 1.9291,
+      "step": 282385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015028721023952927,
+      "loss": 2.0599,
+      "step": 282390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015028561264018593,
+      "loss": 2.1243,
+      "step": 282395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015028401502366405,
+      "loss": 1.9465,
+      "step": 282400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00015028241738996423,
+      "loss": 2.0756,
+      "step": 282405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015028081973908708,
+      "loss": 2.1605,
+      "step": 282410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015027922207103305,
+      "loss": 1.9433,
+      "step": 282415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015027762438580276,
+      "loss": 2.1721,
+      "step": 282420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001502760266833967,
+      "loss": 1.9653,
+      "step": 282425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015027442896381541,
+      "loss": 1.8863,
+      "step": 282430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015027283122705947,
+      "loss": 2.2313,
+      "step": 282435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015027123347312945,
+      "loss": 2.0559,
+      "step": 282440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00015026963570202584,
+      "loss": 2.0609,
+      "step": 282445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015026803791374924,
+      "loss": 2.1571,
+      "step": 282450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015026644010830012,
+      "loss": 2.1227,
+      "step": 282455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001502648422856791,
+      "loss": 2.0119,
+      "step": 282460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015026324444588672,
+      "loss": 2.0847,
+      "step": 282465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001502616465889235,
+      "loss": 1.9806,
+      "step": 282470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015026004871478995,
+      "loss": 2.1275,
+      "step": 282475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015025845082348668,
+      "loss": 2.087,
+      "step": 282480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001502568529150142,
+      "loss": 2.1793,
+      "step": 282485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001502552549893731,
+      "loss": 2.1336,
+      "step": 282490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015025365704656387,
+      "loss": 1.6673,
+      "step": 282495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015025205908658707,
+      "loss": 2.0694,
+      "step": 282500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015025046110944326,
+      "loss": 2.0524,
+      "step": 282505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000150248863115133,
+      "loss": 2.2037,
+      "step": 282510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015024726510365683,
+      "loss": 2.0237,
+      "step": 282515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00015024566707501523,
+      "loss": 2.1862,
+      "step": 282520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015024406902920884,
+      "loss": 2.2127,
+      "step": 282525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015024247096623812,
+      "loss": 1.9989,
+      "step": 282530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015024087288610373,
+      "loss": 2.1396,
+      "step": 282535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015023927478880614,
+      "loss": 2.1179,
+      "step": 282540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015023767667434585,
+      "loss": 2.2315,
+      "step": 282545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015023607854272345,
+      "loss": 2.0714,
+      "step": 282550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015023448039393954,
+      "loss": 2.1516,
+      "step": 282555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015023288222799463,
+      "loss": 2.1989,
+      "step": 282560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015023128404488922,
+      "loss": 1.935,
+      "step": 282565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001502296858446239,
+      "loss": 1.9654,
+      "step": 282570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015022808762719922,
+      "loss": 1.9279,
+      "step": 282575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001502264893926157,
+      "loss": 2.0306,
+      "step": 282580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001502248911408739,
+      "loss": 1.946,
+      "step": 282585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015022329287197437,
+      "loss": 2.0327,
+      "step": 282590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015022169458591767,
+      "loss": 2.1597,
+      "step": 282595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001502200962827043,
+      "loss": 2.3629,
+      "step": 282600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00015021849796233484,
+      "loss": 2.1481,
+      "step": 282605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015021689962480984,
+      "loss": 2.0651,
+      "step": 282610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015021530127012987,
+      "loss": 1.8853,
+      "step": 282615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015021370289829537,
+      "loss": 2.1362,
+      "step": 282620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015021210450930702,
+      "loss": 2.1583,
+      "step": 282625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00015021050610316528,
+      "loss": 2.1588,
+      "step": 282630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015020890767987073,
+      "loss": 2.0294,
+      "step": 282635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015020730923942388,
+      "loss": 1.992,
+      "step": 282640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001502057107818253,
+      "loss": 1.9025,
+      "step": 282645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001502041123070756,
+      "loss": 2.0011,
+      "step": 282650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015020251381517522,
+      "loss": 2.11,
+      "step": 282655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015020091530612475,
+      "loss": 2.0483,
+      "step": 282660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015019931677992472,
+      "loss": 2.0881,
+      "step": 282665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015019771823657574,
+      "loss": 1.9694,
+      "step": 282670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015019611967607828,
+      "loss": 2.297,
+      "step": 282675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015019452109843294,
+      "loss": 2.0965,
+      "step": 282680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015019292250364016,
+      "loss": 2.1769,
+      "step": 282685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015019132389170066,
+      "loss": 2.0017,
+      "step": 282690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015018972526261487,
+      "loss": 2.0542,
+      "step": 282695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00015018812661638333,
+      "loss": 2.0384,
+      "step": 282700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015018652795300666,
+      "loss": 2.0548,
+      "step": 282705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001501849292724853,
+      "loss": 1.9457,
+      "step": 282710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00015018333057481993,
+      "loss": 2.1457,
+      "step": 282715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015018173186001098,
+      "loss": 2.1142,
+      "step": 282720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015018013312805906,
+      "loss": 2.1208,
+      "step": 282725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001501785343789647,
+      "loss": 1.9113,
+      "step": 282730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001501769356127284,
+      "loss": 2.2905,
+      "step": 282735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015017533682935077,
+      "loss": 2.058,
+      "step": 282740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015017373802883236,
+      "loss": 2.0544,
+      "step": 282745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001501721392111737,
+      "loss": 1.9467,
+      "step": 282750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015017054037637525,
+      "loss": 2.033,
+      "step": 282755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001501689415244377,
+      "loss": 2.3167,
+      "step": 282760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001501673426553615,
+      "loss": 1.9267,
+      "step": 282765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015016574376914727,
+      "loss": 1.8744,
+      "step": 282770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015016414486579546,
+      "loss": 2.0631,
+      "step": 282775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001501625459453067,
+      "loss": 2.041,
+      "step": 282780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015016094700768151,
+      "loss": 2.173,
+      "step": 282785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001501593480529204,
+      "loss": 2.024,
+      "step": 282790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00015015774908102395,
+      "loss": 2.1099,
+      "step": 282795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015015615009199274,
+      "loss": 1.829,
+      "step": 282800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015015455108582728,
+      "loss": 2.0364,
+      "step": 282805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001501529520625281,
+      "loss": 2.0347,
+      "step": 282810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00015015135302209572,
+      "loss": 2.208,
+      "step": 282815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015014975396453076,
+      "loss": 1.9431,
+      "step": 282820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015014815488983372,
+      "loss": 1.9515,
+      "step": 282825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001501465557980052,
+      "loss": 2.1838,
+      "step": 282830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001501449566890457,
+      "loss": 2.1702,
+      "step": 282835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015014335756295574,
+      "loss": 2.0736,
+      "step": 282840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001501417584197359,
+      "loss": 2.0132,
+      "step": 282845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015014015925938676,
+      "loss": 2.0914,
+      "step": 282850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015013856008190882,
+      "loss": 1.9765,
+      "step": 282855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00015013696088730263,
+      "loss": 2.1018,
+      "step": 282860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015013536167556875,
+      "loss": 2.0076,
+      "step": 282865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001501337624467077,
+      "loss": 1.9992,
+      "step": 282870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001501321632007201,
+      "loss": 2.0243,
+      "step": 282875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001501305639376064,
+      "loss": 2.2273,
+      "step": 282880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001501289646573672,
+      "loss": 1.9689,
+      "step": 282885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015012736536000306,
+      "loss": 1.9333,
+      "step": 282890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015012576604551446,
+      "loss": 2.0639,
+      "step": 282895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000150124166713902,
+      "loss": 2.0561,
+      "step": 282900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015012256736516623,
+      "loss": 1.9417,
+      "step": 282905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015012096799930767,
+      "loss": 2.0101,
+      "step": 282910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001501193686163269,
+      "loss": 2.0507,
+      "step": 282915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015011776921622445,
+      "loss": 2.0746,
+      "step": 282920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015011616979900084,
+      "loss": 1.9378,
+      "step": 282925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015011457036465662,
+      "loss": 2.1924,
+      "step": 282930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015011297091319236,
+      "loss": 2.1116,
+      "step": 282935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00015011137144460863,
+      "loss": 1.9912,
+      "step": 282940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00015010977195890597,
+      "loss": 2.0725,
+      "step": 282945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015010817245608485,
+      "loss": 1.8939,
+      "step": 282950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015010657293614585,
+      "loss": 1.9718,
+      "step": 282955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00015010497339908957,
+      "loss": 2.0274,
+      "step": 282960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015010337384491654,
+      "loss": 1.9619,
+      "step": 282965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00015010177427362726,
+      "loss": 2.1048,
+      "step": 282970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015010017468522233,
+      "loss": 2.0161,
+      "step": 282975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015009857507970224,
+      "loss": 1.9194,
+      "step": 282980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001500969754570676,
+      "loss": 2.044,
+      "step": 282985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001500953758173189,
+      "loss": 1.9214,
+      "step": 282990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015009377616045675,
+      "loss": 2.1057,
+      "step": 282995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00015009217648648162,
+      "loss": 2.1803,
+      "step": 283000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001500905767953941,
+      "loss": 2.0629,
+      "step": 283005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015008897708719473,
+      "loss": 2.0778,
+      "step": 283010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015008737736188404,
+      "loss": 2.0699,
+      "step": 283015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015008577761946263,
+      "loss": 2.2923,
+      "step": 283020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00015008417785993097,
+      "loss": 2.0917,
+      "step": 283025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015008257808328968,
+      "loss": 2.1562,
+      "step": 283030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00015008097828953928,
+      "loss": 2.2408,
+      "step": 283035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001500793784786803,
+      "loss": 2.0086,
+      "step": 283040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00015007777865071328,
+      "loss": 1.908,
+      "step": 283045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001500761788056388,
+      "loss": 2.1503,
+      "step": 283050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015007457894345737,
+      "loss": 2.1166,
+      "step": 283055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015007297906416956,
+      "loss": 2.1915,
+      "step": 283060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015007137916777592,
+      "loss": 2.1373,
+      "step": 283065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00015006977925427698,
+      "loss": 2.1239,
+      "step": 283070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015006817932367332,
+      "loss": 2.0731,
+      "step": 283075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00015006657937596547,
+      "loss": 2.0068,
+      "step": 283080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015006497941115392,
+      "loss": 2.1029,
+      "step": 283085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001500633794292393,
+      "loss": 2.1842,
+      "step": 283090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001500617794302221,
+      "loss": 2.0183,
+      "step": 283095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00015006017941410291,
+      "loss": 2.095,
+      "step": 283100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015005857938088222,
+      "loss": 2.1721,
+      "step": 283105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015005697933056063,
+      "loss": 2.1653,
+      "step": 283110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015005537926313868,
+      "loss": 2.0928,
+      "step": 283115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001500537791786169,
+      "loss": 1.9898,
+      "step": 283120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00015005217907699585,
+      "loss": 2.2207,
+      "step": 283125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015005057895827606,
+      "loss": 2.0031,
+      "step": 283130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00015004897882245807,
+      "loss": 2.1355,
+      "step": 283135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015004737866954244,
+      "loss": 2.1333,
+      "step": 283140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015004577849952978,
+      "loss": 2.0772,
+      "step": 283145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001500441783124205,
+      "loss": 1.8625,
+      "step": 283150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00015004257810821521,
+      "loss": 2.0791,
+      "step": 283155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00015004097788691453,
+      "loss": 1.9828,
+      "step": 283160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001500393776485189,
+      "loss": 2.1906,
+      "step": 283165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015003777739302897,
+      "loss": 2.1346,
+      "step": 283170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015003617712044517,
+      "loss": 1.9198,
+      "step": 283175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001500345768307681,
+      "loss": 1.9517,
+      "step": 283180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015003297652399836,
+      "loss": 1.9264,
+      "step": 283185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001500313762001364,
+      "loss": 2.0968,
+      "step": 283190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015002977585918282,
+      "loss": 1.9218,
+      "step": 283195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001500281755011382,
+      "loss": 2.1003,
+      "step": 283200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015002657512600302,
+      "loss": 2.2077,
+      "step": 283205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015002497473377786,
+      "loss": 1.9425,
+      "step": 283210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015002337432446327,
+      "loss": 2.0597,
+      "step": 283215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015002177389805975,
+      "loss": 2.057,
+      "step": 283220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001500201734545679,
+      "loss": 1.9236,
+      "step": 283225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001500185729939883,
+      "loss": 2.1076,
+      "step": 283230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001500169725163214,
+      "loss": 2.0507,
+      "step": 283235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00015001537202156783,
+      "loss": 2.011,
+      "step": 283240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015001377150972807,
+      "loss": 2.1903,
+      "step": 283245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001500121709808027,
+      "loss": 1.9467,
+      "step": 283250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00015001057043479227,
+      "loss": 2.0779,
+      "step": 283255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015000896987169734,
+      "loss": 2.2855,
+      "step": 283260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00015000736929151841,
+      "loss": 2.044,
+      "step": 283265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00015000576869425607,
+      "loss": 1.8615,
+      "step": 283270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015000416807991085,
+      "loss": 1.9316,
+      "step": 283275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015000256744848332,
+      "loss": 2.0519,
+      "step": 283280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000150000966799974,
+      "loss": 2.1039,
+      "step": 283285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014999936613438342,
+      "loss": 2.0245,
+      "step": 283290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014999776545171215,
+      "loss": 2.1619,
+      "step": 283295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014999616475196072,
+      "loss": 1.9897,
+      "step": 283300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014999456403512975,
+      "loss": 2.1671,
+      "step": 283305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001499929633012197,
+      "loss": 1.9049,
+      "step": 283310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014999136255023114,
+      "loss": 2.027,
+      "step": 283315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014998976178216464,
+      "loss": 1.9525,
+      "step": 283320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014998816099702072,
+      "loss": 2.0413,
+      "step": 283325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014998656019479993,
+      "loss": 1.8537,
+      "step": 283330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014998495937550285,
+      "loss": 1.9691,
+      "step": 283335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014998335853912996,
+      "loss": 2.1501,
+      "step": 283340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001499817576856819,
+      "loss": 2.0322,
+      "step": 283345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014998015681515914,
+      "loss": 2.1209,
+      "step": 283350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014997855592756225,
+      "loss": 1.9343,
+      "step": 283355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014997695502289178,
+      "loss": 1.8597,
+      "step": 283360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014997535410114828,
+      "loss": 2.132,
+      "step": 283365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014997375316233227,
+      "loss": 2.1974,
+      "step": 283370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.0625,
+      "learning_rate": 0.00014997215220644437,
+      "loss": 1.9936,
+      "step": 283375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014997055123348502,
+      "loss": 1.9493,
+      "step": 283380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014996895024345486,
+      "loss": 2.0803,
+      "step": 283385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001499673492363544,
+      "loss": 2.1553,
+      "step": 283390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014996574821218416,
+      "loss": 2.1593,
+      "step": 283395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014996414717094475,
+      "loss": 2.1169,
+      "step": 283400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014996254611263662,
+      "loss": 2.0219,
+      "step": 283405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014996094503726042,
+      "loss": 2.2482,
+      "step": 283410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014995934394481668,
+      "loss": 2.1566,
+      "step": 283415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001499577428353059,
+      "loss": 2.144,
+      "step": 283420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014995614170872865,
+      "loss": 2.0828,
+      "step": 283425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014995454056508547,
+      "loss": 2.036,
+      "step": 283430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001499529394043769,
+      "loss": 1.9529,
+      "step": 283435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014995133822660352,
+      "loss": 1.9679,
+      "step": 283440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014994973703176586,
+      "loss": 1.9731,
+      "step": 283445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014994813581986445,
+      "loss": 2.109,
+      "step": 283450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014994653459089987,
+      "loss": 2.0608,
+      "step": 283455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014994493334487263,
+      "loss": 2.1019,
+      "step": 283460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014994333208178328,
+      "loss": 1.8931,
+      "step": 283465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001499417308016324,
+      "loss": 1.9933,
+      "step": 283470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014994012950442052,
+      "loss": 2.0572,
+      "step": 283475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001499385281901482,
+      "loss": 1.9835,
+      "step": 283480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014993692685881594,
+      "loss": 2.0896,
+      "step": 283485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014993532551042435,
+      "loss": 1.9922,
+      "step": 283490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014993372414497395,
+      "loss": 2.0771,
+      "step": 283495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014993212276246526,
+      "loss": 2.1818,
+      "step": 283500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014993052136289887,
+      "loss": 2.0127,
+      "step": 283505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014992891994627532,
+      "loss": 2.049,
+      "step": 283510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001499273185125951,
+      "loss": 1.9394,
+      "step": 283515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014992571706185884,
+      "loss": 2.149,
+      "step": 283520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014992411559406702,
+      "loss": 1.9651,
+      "step": 283525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014992251410922023,
+      "loss": 1.9981,
+      "step": 283530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014992091260731903,
+      "loss": 1.929,
+      "step": 283535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001499193110883639,
+      "loss": 2.2989,
+      "step": 283540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014991770955235544,
+      "loss": 1.9478,
+      "step": 283545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001499161079992942,
+      "loss": 1.9709,
+      "step": 283550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014991450642918072,
+      "loss": 2.1096,
+      "step": 283555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014991290484201553,
+      "loss": 1.9763,
+      "step": 283560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014991130323779915,
+      "loss": 1.9771,
+      "step": 283565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014990970161653222,
+      "loss": 2.0463,
+      "step": 283570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001499080999782152,
+      "loss": 2.1998,
+      "step": 283575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001499064983228487,
+      "loss": 1.9798,
+      "step": 283580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001499048966504332,
+      "loss": 2.1203,
+      "step": 283585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014990329496096928,
+      "loss": 1.8935,
+      "step": 283590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014990169325445748,
+      "loss": 2.0797,
+      "step": 283595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014990009153089838,
+      "loss": 2.0573,
+      "step": 283600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001498984897902925,
+      "loss": 1.8536,
+      "step": 283605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014989688803264043,
+      "loss": 2.214,
+      "step": 283610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014989528625794263,
+      "loss": 1.9664,
+      "step": 283615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014989368446619968,
+      "loss": 2.0879,
+      "step": 283620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014989208265741216,
+      "loss": 1.9108,
+      "step": 283625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014989048083158062,
+      "loss": 2.0927,
+      "step": 283630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001498888789887056,
+      "loss": 2.1684,
+      "step": 283635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014988727712878758,
+      "loss": 2.1169,
+      "step": 283640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014988567525182718,
+      "loss": 2.1424,
+      "step": 283645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014988407335782495,
+      "loss": 2.0175,
+      "step": 283650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001498824714467814,
+      "loss": 2.0123,
+      "step": 283655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001498808695186971,
+      "loss": 2.1895,
+      "step": 283660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014987926757357257,
+      "loss": 2.1047,
+      "step": 283665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014987766561140844,
+      "loss": 2.2978,
+      "step": 283670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014987606363220512,
+      "loss": 1.9828,
+      "step": 283675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014987446163596327,
+      "loss": 2.1057,
+      "step": 283680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014987285962268337,
+      "loss": 2.0478,
+      "step": 283685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014987125759236605,
+      "loss": 2.1739,
+      "step": 283690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014986965554501179,
+      "loss": 2.0739,
+      "step": 283695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001498680534806211,
+      "loss": 2.1969,
+      "step": 283700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001498664513991946,
+      "loss": 1.8855,
+      "step": 283705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014986484930073284,
+      "loss": 2.1997,
+      "step": 283710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014986324718523633,
+      "loss": 2.0362,
+      "step": 283715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014986164505270563,
+      "loss": 1.985,
+      "step": 283720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014986004290314128,
+      "loss": 2.1277,
+      "step": 283725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014985844073654386,
+      "loss": 2.0667,
+      "step": 283730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014985683855291387,
+      "loss": 2.0583,
+      "step": 283735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001498552363522519,
+      "loss": 1.9627,
+      "step": 283740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014985363413455845,
+      "loss": 2.0692,
+      "step": 283745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014985203189983413,
+      "loss": 2.1495,
+      "step": 283750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001498504296480794,
+      "loss": 2.1569,
+      "step": 283755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014984882737929488,
+      "loss": 1.9405,
+      "step": 283760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014984722509348112,
+      "loss": 1.9094,
+      "step": 283765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014984562279063863,
+      "loss": 2.2717,
+      "step": 283770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014984402047076798,
+      "loss": 2.2722,
+      "step": 283775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014984241813386967,
+      "loss": 2.2878,
+      "step": 283780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014984081577994434,
+      "loss": 1.9615,
+      "step": 283785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014983921340899246,
+      "loss": 2.0803,
+      "step": 283790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001498376110210146,
+      "loss": 2.0707,
+      "step": 283795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014983600861601129,
+      "loss": 1.9351,
+      "step": 283800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014983440619398316,
+      "loss": 1.9858,
+      "step": 283805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001498328037549306,
+      "loss": 2.2445,
+      "step": 283810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001498312012988543,
+      "loss": 2.1613,
+      "step": 283815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014982959882575475,
+      "loss": 1.9371,
+      "step": 283820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014982799633563251,
+      "loss": 2.1939,
+      "step": 283825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014982639382848814,
+      "loss": 2.0852,
+      "step": 283830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014982479130432212,
+      "loss": 2.0347,
+      "step": 283835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014982318876313508,
+      "loss": 1.8946,
+      "step": 283840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014982158620492755,
+      "loss": 2.0157,
+      "step": 283845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014981998362970005,
+      "loss": 1.8138,
+      "step": 283850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014981838103745316,
+      "loss": 2.0937,
+      "step": 283855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014981677842818735,
+      "loss": 1.9147,
+      "step": 283860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014981517580190326,
+      "loss": 1.9871,
+      "step": 283865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014981357315860141,
+      "loss": 1.9195,
+      "step": 283870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014981197049828234,
+      "loss": 2.0863,
+      "step": 283875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014981036782094658,
+      "loss": 2.0632,
+      "step": 283880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014980876512659473,
+      "loss": 2.1605,
+      "step": 283885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014980716241522725,
+      "loss": 2.0905,
+      "step": 283890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001498055596868448,
+      "loss": 2.1671,
+      "step": 283895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014980395694144782,
+      "loss": 2.048,
+      "step": 283900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014980235417903692,
+      "loss": 1.9487,
+      "step": 283905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014980075139961262,
+      "loss": 1.9886,
+      "step": 283910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001497991486031755,
+      "loss": 2.1308,
+      "step": 283915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001497975457897261,
+      "loss": 2.0982,
+      "step": 283920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014979594295926494,
+      "loss": 1.9668,
+      "step": 283925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014979434011179257,
+      "loss": 2.0462,
+      "step": 283930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014979273724730955,
+      "loss": 2.0648,
+      "step": 283935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014979113436581644,
+      "loss": 2.0116,
+      "step": 283940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014978953146731376,
+      "loss": 2.0056,
+      "step": 283945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001497879285518021,
+      "loss": 2.05,
+      "step": 283950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014978632561928198,
+      "loss": 2.0319,
+      "step": 283955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014978472266975393,
+      "loss": 2.1434,
+      "step": 283960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014978311970321851,
+      "loss": 2.0019,
+      "step": 283965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014978151671967628,
+      "loss": 2.1525,
+      "step": 283970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001497799137191278,
+      "loss": 1.8941,
+      "step": 283975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014977831070157356,
+      "loss": 2.0652,
+      "step": 283980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001497767076670142,
+      "loss": 1.9895,
+      "step": 283985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014977510461545017,
+      "loss": 2.1039,
+      "step": 283990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014977350154688206,
+      "loss": 2.4192,
+      "step": 283995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014977189846131044,
+      "loss": 2.0782,
+      "step": 284000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014977029535873584,
+      "loss": 2.1571,
+      "step": 284005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014976869223915878,
+      "loss": 2.0706,
+      "step": 284010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001497670891025799,
+      "loss": 1.9985,
+      "step": 284015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014976548594899957,
+      "loss": 2.0878,
+      "step": 284020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001497638827784185,
+      "loss": 1.9845,
+      "step": 284025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014976227959083722,
+      "loss": 2.1317,
+      "step": 284030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014976067638625622,
+      "loss": 2.0546,
+      "step": 284035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014975907316467605,
+      "loss": 1.9706,
+      "step": 284040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014975746992609729,
+      "loss": 2.2042,
+      "step": 284045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014975586667052048,
+      "loss": 2.0434,
+      "step": 284050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014975426339794616,
+      "loss": 2.0026,
+      "step": 284055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014975266010837487,
+      "loss": 1.8852,
+      "step": 284060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014975105680180718,
+      "loss": 1.9901,
+      "step": 284065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014974945347824366,
+      "loss": 2.0505,
+      "step": 284070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014974785013768476,
+      "loss": 2.0742,
+      "step": 284075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014974624678013113,
+      "loss": 2.12,
+      "step": 284080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001497446434055833,
+      "loss": 2.2379,
+      "step": 284085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014974304001404176,
+      "loss": 2.0458,
+      "step": 284090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001497414366055071,
+      "loss": 1.9977,
+      "step": 284095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001497398331799799,
+      "loss": 2.0899,
+      "step": 284100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014973822973746062,
+      "loss": 2.1728,
+      "step": 284105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014973662627794986,
+      "loss": 2.1159,
+      "step": 284110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001497350228014482,
+      "loss": 1.8965,
+      "step": 284115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014973341930795615,
+      "loss": 2.2302,
+      "step": 284120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014973181579747426,
+      "loss": 2.1445,
+      "step": 284125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014973021227000305,
+      "loss": 2.0435,
+      "step": 284130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014972860872554313,
+      "loss": 2.0776,
+      "step": 284135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000149727005164095,
+      "loss": 1.9753,
+      "step": 284140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014972540158565925,
+      "loss": 2.2005,
+      "step": 284145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014972379799023638,
+      "loss": 2.2622,
+      "step": 284150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014972219437782696,
+      "loss": 2.0071,
+      "step": 284155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014972059074843153,
+      "loss": 2.1454,
+      "step": 284160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014971898710205064,
+      "loss": 1.9854,
+      "step": 284165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014971738343868486,
+      "loss": 2.1887,
+      "step": 284170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014971577975833472,
+      "loss": 2.1524,
+      "step": 284175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014971417606100075,
+      "loss": 2.0271,
+      "step": 284180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014971257234668354,
+      "loss": 2.0472,
+      "step": 284185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001497109686153836,
+      "loss": 2.0662,
+      "step": 284190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014970936486710147,
+      "loss": 2.0563,
+      "step": 284195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014970776110183776,
+      "loss": 2.168,
+      "step": 284200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014970615731959295,
+      "loss": 2.0103,
+      "step": 284205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014970455352036763,
+      "loss": 2.0595,
+      "step": 284210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014970294970416233,
+      "loss": 2.1787,
+      "step": 284215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001497013458709776,
+      "loss": 2.0363,
+      "step": 284220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014969974202081398,
+      "loss": 2.1624,
+      "step": 284225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014969813815367204,
+      "loss": 2.1189,
+      "step": 284230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001496965342695523,
+      "loss": 1.904,
+      "step": 284235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001496949303684553,
+      "loss": 2.1007,
+      "step": 284240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014969332645038168,
+      "loss": 1.9622,
+      "step": 284245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014969172251533185,
+      "loss": 1.9821,
+      "step": 284250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014969011856330648,
+      "loss": 2.1754,
+      "step": 284255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014968851459430606,
+      "loss": 2.0528,
+      "step": 284260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001496869106083311,
+      "loss": 1.8538,
+      "step": 284265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001496853066053822,
+      "loss": 2.0503,
+      "step": 284270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014968370258545993,
+      "loss": 1.8764,
+      "step": 284275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014968209854856478,
+      "loss": 2.1815,
+      "step": 284280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014968049449469735,
+      "loss": 1.9166,
+      "step": 284285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001496788904238581,
+      "loss": 2.0305,
+      "step": 284290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001496772863360477,
+      "loss": 2.2339,
+      "step": 284295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014967568223126662,
+      "loss": 2.2442,
+      "step": 284300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00014967407810951544,
+      "loss": 2.0581,
+      "step": 284305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014967247397079469,
+      "loss": 1.9846,
+      "step": 284310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001496708698151049,
+      "loss": 2.1452,
+      "step": 284315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014966926564244663,
+      "loss": 2.1053,
+      "step": 284320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001496676614528205,
+      "loss": 2.2353,
+      "step": 284325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014966605724622695,
+      "loss": 2.0818,
+      "step": 284330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001496644530226666,
+      "loss": 2.2582,
+      "step": 284335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014966284878213996,
+      "loss": 2.0249,
+      "step": 284340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014966124452464756,
+      "loss": 2.0355,
+      "step": 284345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014965964025019,
+      "loss": 2.0164,
+      "step": 284350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001496580359587678,
+      "loss": 2.1147,
+      "step": 284355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014965643165038153,
+      "loss": 1.9631,
+      "step": 284360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001496548273250317,
+      "loss": 2.1518,
+      "step": 284365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001496532229827189,
+      "loss": 2.1172,
+      "step": 284370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014965161862344364,
+      "loss": 1.9388,
+      "step": 284375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014965001424720653,
+      "loss": 1.9229,
+      "step": 284380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014964840985400802,
+      "loss": 1.9689,
+      "step": 284385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014964680544384877,
+      "loss": 1.9912,
+      "step": 284390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014964520101672923,
+      "loss": 2.1599,
+      "step": 284395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014964359657264998,
+      "loss": 2.0462,
+      "step": 284400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001496419921116116,
+      "loss": 1.9837,
+      "step": 284405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014964038763361462,
+      "loss": 2.0593,
+      "step": 284410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014963878313865959,
+      "loss": 2.1376,
+      "step": 284415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000149637178626747,
+      "loss": 2.2003,
+      "step": 284420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014963557409787749,
+      "loss": 1.9717,
+      "step": 284425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014963396955205158,
+      "loss": 1.8948,
+      "step": 284430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014963236498926978,
+      "loss": 2.1074,
+      "step": 284435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001496307604095327,
+      "loss": 1.9334,
+      "step": 284440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001496291558128408,
+      "loss": 2.0375,
+      "step": 284445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014962755119919468,
+      "loss": 2.0466,
+      "step": 284450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014962594656859493,
+      "loss": 1.9957,
+      "step": 284455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014962434192104207,
+      "loss": 2.1141,
+      "step": 284460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014962273725653658,
+      "loss": 2.1662,
+      "step": 284465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001496211325750791,
+      "loss": 2.1925,
+      "step": 284470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014961952787667014,
+      "loss": 2.1591,
+      "step": 284475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001496179231613102,
+      "loss": 2.2824,
+      "step": 284480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.00014961631842899995,
+      "loss": 2.1453,
+      "step": 284485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014961471367973983,
+      "loss": 2.2024,
+      "step": 284490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014961310891353042,
+      "loss": 2.1706,
+      "step": 284495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014961150413037228,
+      "loss": 2.0181,
+      "step": 284500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014960989933026595,
+      "loss": 2.1326,
+      "step": 284505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014960829451321196,
+      "loss": 2.0934,
+      "step": 284510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001496066896792109,
+      "loss": 2.0248,
+      "step": 284515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014960508482826328,
+      "loss": 2.0593,
+      "step": 284520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014960347996036968,
+      "loss": 1.9641,
+      "step": 284525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014960187507553064,
+      "loss": 1.9798,
+      "step": 284530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014960027017374666,
+      "loss": 2.0766,
+      "step": 284535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014959866525501836,
+      "loss": 2.0055,
+      "step": 284540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014959706031934624,
+      "loss": 1.949,
+      "step": 284545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014959545536673087,
+      "loss": 2.0622,
+      "step": 284550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014959385039717282,
+      "loss": 1.9621,
+      "step": 284555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001495922454106726,
+      "loss": 2.0607,
+      "step": 284560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014959064040723073,
+      "loss": 2.1879,
+      "step": 284565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014958903538684782,
+      "loss": 1.9887,
+      "step": 284570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014958743034952442,
+      "loss": 2.1257,
+      "step": 284575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014958582529526104,
+      "loss": 2.2942,
+      "step": 284580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014958422022405822,
+      "loss": 2.118,
+      "step": 284585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014958261513591653,
+      "loss": 1.9145,
+      "step": 284590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014958101003083654,
+      "loss": 2.1003,
+      "step": 284595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014957940490881882,
+      "loss": 2.1131,
+      "step": 284600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001495777997698638,
+      "loss": 2.1404,
+      "step": 284605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014957619461397213,
+      "loss": 1.9354,
+      "step": 284610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014957458944114436,
+      "loss": 2.0461,
+      "step": 284615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000149572984251381,
+      "loss": 2.1182,
+      "step": 284620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014957137904468259,
+      "loss": 2.131,
+      "step": 284625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001495697738210497,
+      "loss": 1.9367,
+      "step": 284630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001495681685804829,
+      "loss": 2.0652,
+      "step": 284635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001495665633229827,
+      "loss": 2.0246,
+      "step": 284640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014956495804854965,
+      "loss": 2.208,
+      "step": 284645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014956335275718435,
+      "loss": 2.0376,
+      "step": 284650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014956174744888728,
+      "loss": 1.9314,
+      "step": 284655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000149560142123659,
+      "loss": 2.1227,
+      "step": 284660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001495585367815001,
+      "loss": 2.0503,
+      "step": 284665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014955693142241108,
+      "loss": 1.8338,
+      "step": 284670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014955532604639255,
+      "loss": 2.0359,
+      "step": 284675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000149553720653445,
+      "loss": 2.2133,
+      "step": 284680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014955211524356904,
+      "loss": 2.1572,
+      "step": 284685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014955050981676511,
+      "loss": 2.1528,
+      "step": 284690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014954890437303387,
+      "loss": 2.0798,
+      "step": 284695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014954729891237585,
+      "loss": 1.9378,
+      "step": 284700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014954569343479155,
+      "loss": 2.0357,
+      "step": 284705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001495440879402815,
+      "loss": 2.035,
+      "step": 284710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014954248242884637,
+      "loss": 2.0432,
+      "step": 284715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014954087690048656,
+      "loss": 1.9679,
+      "step": 284720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001495392713552027,
+      "loss": 2.0366,
+      "step": 284725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014953766579299535,
+      "loss": 1.958,
+      "step": 284730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014953606021386506,
+      "loss": 2.1156,
+      "step": 284735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014953445461781233,
+      "loss": 2.2891,
+      "step": 284740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014953284900483767,
+      "loss": 2.0624,
+      "step": 284745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014953124337494176,
+      "loss": 2.1925,
+      "step": 284750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001495296377281251,
+      "loss": 2.1718,
+      "step": 284755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014952803206438818,
+      "loss": 2.161,
+      "step": 284760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014952642638373158,
+      "loss": 2.0018,
+      "step": 284765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014952482068615583,
+      "loss": 1.9566,
+      "step": 284770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014952321497166152,
+      "loss": 2.1449,
+      "step": 284775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014952160924024924,
+      "loss": 2.0833,
+      "step": 284780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014952000349191942,
+      "loss": 1.873,
+      "step": 284785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014951839772667268,
+      "loss": 2.0323,
+      "step": 284790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014951679194450957,
+      "loss": 1.9167,
+      "step": 284795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001495151861454306,
+      "loss": 1.9089,
+      "step": 284800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014951358032943635,
+      "loss": 2.0343,
+      "step": 284805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014951197449652738,
+      "loss": 2.0829,
+      "step": 284810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014951036864670422,
+      "loss": 2.0742,
+      "step": 284815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014950876277996744,
+      "loss": 2.0549,
+      "step": 284820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014950715689631753,
+      "loss": 2.2657,
+      "step": 284825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001495055509957551,
+      "loss": 2.0292,
+      "step": 284830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014950394507828066,
+      "loss": 2.0655,
+      "step": 284835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014950233914389478,
+      "loss": 1.9157,
+      "step": 284840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000149500733192598,
+      "loss": 1.8636,
+      "step": 284845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001494991272243909,
+      "loss": 2.1721,
+      "step": 284850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014949752123927397,
+      "loss": 2.1427,
+      "step": 284855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001494959152372478,
+      "loss": 1.9416,
+      "step": 284860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014949430921831292,
+      "loss": 2.193,
+      "step": 284865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00014949270318246992,
+      "loss": 1.9905,
+      "step": 284870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014949109712971928,
+      "loss": 2.2089,
+      "step": 284875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014948949106006156,
+      "loss": 1.8498,
+      "step": 284880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014948788497349736,
+      "loss": 2.2484,
+      "step": 284885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001494862788700272,
+      "loss": 2.168,
+      "step": 284890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014948467274965166,
+      "loss": 2.0159,
+      "step": 284895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014948306661237122,
+      "loss": 1.9494,
+      "step": 284900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014948146045818645,
+      "loss": 2.0689,
+      "step": 284905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014947985428709795,
+      "loss": 1.9392,
+      "step": 284910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014947824809910624,
+      "loss": 2.1445,
+      "step": 284915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014947664189421186,
+      "loss": 2.3029,
+      "step": 284920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014947503567241537,
+      "loss": 2.092,
+      "step": 284925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014947342943371727,
+      "loss": 2.1465,
+      "step": 284930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014947182317811815,
+      "loss": 2.2725,
+      "step": 284935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014947021690561862,
+      "loss": 2.2341,
+      "step": 284940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001494686106162191,
+      "loss": 1.9781,
+      "step": 284945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014946700430992026,
+      "loss": 2.1043,
+      "step": 284950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014946539798672256,
+      "loss": 2.1627,
+      "step": 284955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014946379164662658,
+      "loss": 1.9966,
+      "step": 284960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014946218528963287,
+      "loss": 1.9962,
+      "step": 284965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.000149460578915742,
+      "loss": 1.9825,
+      "step": 284970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001494589725249545,
+      "loss": 2.0495,
+      "step": 284975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014945736611727092,
+      "loss": 2.099,
+      "step": 284980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014945575969269178,
+      "loss": 2.0245,
+      "step": 284985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001494541532512177,
+      "loss": 2.1754,
+      "step": 284990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014945254679284912,
+      "loss": 2.1451,
+      "step": 284995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014945094031758668,
+      "loss": 2.1734,
+      "step": 285000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014944933382543093,
+      "loss": 2.2233,
+      "step": 285005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001494477273163824,
+      "loss": 2.0507,
+      "step": 285010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014944612079044158,
+      "loss": 2.0729,
+      "step": 285015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014944451424760908,
+      "loss": 1.9555,
+      "step": 285020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014944290768788545,
+      "loss": 2.0287,
+      "step": 285025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00014944130111127125,
+      "loss": 2.0694,
+      "step": 285030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014943969451776696,
+      "loss": 2.1913,
+      "step": 285035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014943808790737317,
+      "loss": 2.2566,
+      "step": 285040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014943648128009046,
+      "loss": 2.034,
+      "step": 285045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014943487463591937,
+      "loss": 2.207,
+      "step": 285050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001494332679748604,
+      "loss": 2.0535,
+      "step": 285055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014943166129691412,
+      "loss": 1.9971,
+      "step": 285060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014943005460208111,
+      "loss": 2.0744,
+      "step": 285065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014942844789036188,
+      "loss": 2.0194,
+      "step": 285070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014942684116175703,
+      "loss": 2.1271,
+      "step": 285075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014942523441626704,
+      "loss": 1.9973,
+      "step": 285080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014942362765389252,
+      "loss": 2.3208,
+      "step": 285085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014942202087463396,
+      "loss": 2.0323,
+      "step": 285090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014942041407849197,
+      "loss": 2.0312,
+      "step": 285095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014941880726546705,
+      "loss": 2.1805,
+      "step": 285100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014941720043555977,
+      "loss": 2.1901,
+      "step": 285105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014941559358877068,
+      "loss": 2.0915,
+      "step": 285110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014941398672510035,
+      "loss": 2.0581,
+      "step": 285115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014941237984454929,
+      "loss": 2.0184,
+      "step": 285120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014941077294711806,
+      "loss": 2.0523,
+      "step": 285125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014940916603280722,
+      "loss": 2.0019,
+      "step": 285130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001494075591016173,
+      "loss": 1.9233,
+      "step": 285135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014940595215354887,
+      "loss": 2.0573,
+      "step": 285140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014940434518860246,
+      "loss": 1.9516,
+      "step": 285145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014940273820677867,
+      "loss": 2.1229,
+      "step": 285150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014940113120807797,
+      "loss": 2.132,
+      "step": 285155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014939952419250097,
+      "loss": 2.0995,
+      "step": 285160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001493979171600482,
+      "loss": 2.0987,
+      "step": 285165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014939631011072016,
+      "loss": 2.002,
+      "step": 285170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014939470304451747,
+      "loss": 2.0334,
+      "step": 285175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014939309596144067,
+      "loss": 2.077,
+      "step": 285180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001493914888614903,
+      "loss": 2.0619,
+      "step": 285185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014938988174466688,
+      "loss": 2.0983,
+      "step": 285190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014938827461097098,
+      "loss": 2.2156,
+      "step": 285195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014938666746040313,
+      "loss": 2.0806,
+      "step": 285200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014938506029296393,
+      "loss": 2.0568,
+      "step": 285205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001493834531086539,
+      "loss": 2.118,
+      "step": 285210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014938184590747357,
+      "loss": 2.0075,
+      "step": 285215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014938023868942351,
+      "loss": 2.1031,
+      "step": 285220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014937863145450428,
+      "loss": 2.092,
+      "step": 285225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001493770242027164,
+      "loss": 1.877,
+      "step": 285230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014937541693406042,
+      "loss": 2.129,
+      "step": 285235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014937380964853694,
+      "loss": 2.1432,
+      "step": 285240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014937220234614645,
+      "loss": 2.0416,
+      "step": 285245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001493705950268895,
+      "loss": 2.014,
+      "step": 285250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001493689876907667,
+      "loss": 2.3325,
+      "step": 285255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014936738033777852,
+      "loss": 2.01,
+      "step": 285260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014936577296792555,
+      "loss": 2.0111,
+      "step": 285265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014936416558120836,
+      "loss": 2.2066,
+      "step": 285270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014936255817762746,
+      "loss": 1.9998,
+      "step": 285275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014936095075718342,
+      "loss": 2.1258,
+      "step": 285280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001493593433198768,
+      "loss": 2.0108,
+      "step": 285285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001493577358657081,
+      "loss": 1.9463,
+      "step": 285290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014935612839467793,
+      "loss": 2.2392,
+      "step": 285295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014935452090678681,
+      "loss": 2.2141,
+      "step": 285300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014935291340203527,
+      "loss": 1.9908,
+      "step": 285305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001493513058804239,
+      "loss": 1.8483,
+      "step": 285310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001493496983419532,
+      "loss": 2.3083,
+      "step": 285315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014934809078662376,
+      "loss": 1.9466,
+      "step": 285320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014934648321443617,
+      "loss": 1.9668,
+      "step": 285325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001493448756253909,
+      "loss": 2.064,
+      "step": 285330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014934326801948846,
+      "loss": 1.9551,
+      "step": 285335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001493416603967295,
+      "loss": 2.0287,
+      "step": 285340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014934005275711457,
+      "loss": 2.1311,
+      "step": 285345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001493384451006442,
+      "loss": 1.8585,
+      "step": 285350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014933683742731886,
+      "loss": 2.088,
+      "step": 285355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014933522973713915,
+      "loss": 1.9681,
+      "step": 285360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014933362203010567,
+      "loss": 2.2422,
+      "step": 285365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014933201430621896,
+      "loss": 1.9644,
+      "step": 285370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001493304065654795,
+      "loss": 1.9913,
+      "step": 285375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014932879880788787,
+      "loss": 2.03,
+      "step": 285380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014932719103344464,
+      "loss": 1.9798,
+      "step": 285385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014932558324215033,
+      "loss": 2.0829,
+      "step": 285390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014932397543400555,
+      "loss": 2.1369,
+      "step": 285395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014932236760901077,
+      "loss": 2.2261,
+      "step": 285400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014932075976716657,
+      "loss": 2.046,
+      "step": 285405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001493191519084735,
+      "loss": 2.1696,
+      "step": 285410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014931754403293217,
+      "loss": 1.9055,
+      "step": 285415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014931593614054303,
+      "loss": 2.1825,
+      "step": 285420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014931432823130666,
+      "loss": 1.9982,
+      "step": 285425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014931272030522364,
+      "loss": 2.0837,
+      "step": 285430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001493111123622945,
+      "loss": 2.0892,
+      "step": 285435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014930950440251978,
+      "loss": 1.9648,
+      "step": 285440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014930789642590003,
+      "loss": 2.0652,
+      "step": 285445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014930628843243585,
+      "loss": 2.049,
+      "step": 285450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014930468042212772,
+      "loss": 2.0631,
+      "step": 285455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014930307239497623,
+      "loss": 2.0404,
+      "step": 285460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014930146435098189,
+      "loss": 2.2594,
+      "step": 285465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014929985629014527,
+      "loss": 2.0785,
+      "step": 285470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014929824821246696,
+      "loss": 1.9794,
+      "step": 285475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014929664011794746,
+      "loss": 2.1382,
+      "step": 285480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014929503200658735,
+      "loss": 1.9623,
+      "step": 285485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014929342387838714,
+      "loss": 2.0935,
+      "step": 285490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001492918157333474,
+      "loss": 2.0266,
+      "step": 285495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014929020757146867,
+      "loss": 2.0317,
+      "step": 285500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014928859939275158,
+      "loss": 1.9559,
+      "step": 285505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014928699119719653,
+      "loss": 2.0004,
+      "step": 285510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001492853829848042,
+      "loss": 2.1329,
+      "step": 285515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014928377475557504,
+      "loss": 1.9545,
+      "step": 285520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001492821665095097,
+      "loss": 1.9827,
+      "step": 285525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014928055824660867,
+      "loss": 2.0144,
+      "step": 285530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014927894996687248,
+      "loss": 2.09,
+      "step": 285535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001492773416703017,
+      "loss": 1.7355,
+      "step": 285540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014927573335689695,
+      "loss": 1.8066,
+      "step": 285545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014927412502665867,
+      "loss": 2.1735,
+      "step": 285550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014927251667958745,
+      "loss": 2.0498,
+      "step": 285555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014927090831568386,
+      "loss": 1.9479,
+      "step": 285560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014926929993494843,
+      "loss": 2.0133,
+      "step": 285565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001492676915373817,
+      "loss": 2.1224,
+      "step": 285570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00014926608312298428,
+      "loss": 1.9344,
+      "step": 285575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001492644746917566,
+      "loss": 2.1833,
+      "step": 285580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00014926286624369935,
+      "loss": 1.9728,
+      "step": 285585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014926125777881298,
+      "loss": 2.0455,
+      "step": 285590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014925964929709805,
+      "loss": 2.1661,
+      "step": 285595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001492580407985552,
+      "loss": 2.051,
+      "step": 285600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014925643228318482,
+      "loss": 1.984,
+      "step": 285605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001492548237509876,
+      "loss": 2.0661,
+      "step": 285610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000149253215201964,
+      "loss": 2.1767,
+      "step": 285615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014925160663611467,
+      "loss": 1.9278,
+      "step": 285620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014924999805344008,
+      "loss": 2.0344,
+      "step": 285625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014924838945394074,
+      "loss": 2.006,
+      "step": 285630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014924678083761732,
+      "loss": 1.995,
+      "step": 285635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014924517220447033,
+      "loss": 2.1624,
+      "step": 285640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014924356355450023,
+      "loss": 1.8534,
+      "step": 285645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014924195488770766,
+      "loss": 2.1462,
+      "step": 285650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014924034620409312,
+      "loss": 2.0181,
+      "step": 285655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014923873750365723,
+      "loss": 2.0077,
+      "step": 285660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014923712878640047,
+      "loss": 1.975,
+      "step": 285665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014923552005232345,
+      "loss": 1.9959,
+      "step": 285670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014923391130142663,
+      "loss": 2.0228,
+      "step": 285675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014923230253371063,
+      "loss": 2.0569,
+      "step": 285680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.000149230693749176,
+      "loss": 2.0817,
+      "step": 285685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014922908494782326,
+      "loss": 2.0941,
+      "step": 285690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014922747612965294,
+      "loss": 2.1912,
+      "step": 285695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014922586729466567,
+      "loss": 2.0319,
+      "step": 285700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014922425844286194,
+      "loss": 1.9665,
+      "step": 285705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001492226495742423,
+      "loss": 2.1544,
+      "step": 285710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001492210406888073,
+      "loss": 2.015,
+      "step": 285715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014921943178655754,
+      "loss": 2.3014,
+      "step": 285720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001492178228674935,
+      "loss": 2.0575,
+      "step": 285725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014921621393161575,
+      "loss": 2.0531,
+      "step": 285730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014921460497892486,
+      "loss": 2.1194,
+      "step": 285735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001492129960094214,
+      "loss": 2.0892,
+      "step": 285740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014921138702310584,
+      "loss": 2.0918,
+      "step": 285745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001492097780199788,
+      "loss": 2.1344,
+      "step": 285750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001492081690000408,
+      "loss": 2.0247,
+      "step": 285755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001492065599632924,
+      "loss": 2.0623,
+      "step": 285760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014920495090973415,
+      "loss": 2.0308,
+      "step": 285765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014920334183936656,
+      "loss": 2.1371,
+      "step": 285770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00014920173275219026,
+      "loss": 2.0054,
+      "step": 285775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014920012364820576,
+      "loss": 2.0855,
+      "step": 285780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001491985145274136,
+      "loss": 2.2298,
+      "step": 285785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001491969053898143,
+      "loss": 2.1307,
+      "step": 285790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001491952962354085,
+      "loss": 2.1149,
+      "step": 285795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014919368706419665,
+      "loss": 2.0503,
+      "step": 285800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014919207787617936,
+      "loss": 2.1975,
+      "step": 285805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014919046867135718,
+      "loss": 2.1021,
+      "step": 285810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001491888594497306,
+      "loss": 1.8436,
+      "step": 285815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014918725021130025,
+      "loss": 2.1674,
+      "step": 285820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014918564095606665,
+      "loss": 1.9608,
+      "step": 285825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014918403168403033,
+      "loss": 2.0427,
+      "step": 285830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014918242239519186,
+      "loss": 2.0082,
+      "step": 285835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014918081308955177,
+      "loss": 2.163,
+      "step": 285840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001491792037671106,
+      "loss": 2.0499,
+      "step": 285845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014917759442786897,
+      "loss": 1.9098,
+      "step": 285850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014917598507182734,
+      "loss": 1.95,
+      "step": 285855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014917437569898634,
+      "loss": 2.1404,
+      "step": 285860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014917276630934645,
+      "loss": 1.8905,
+      "step": 285865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014917115690290825,
+      "loss": 2.0998,
+      "step": 285870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014916954747967232,
+      "loss": 1.8316,
+      "step": 285875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014916793803963915,
+      "loss": 2.1059,
+      "step": 285880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014916632858280935,
+      "loss": 2.253,
+      "step": 285885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001491647191091834,
+      "loss": 1.9479,
+      "step": 285890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001491631096187619,
+      "loss": 2.0529,
+      "step": 285895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014916150011154544,
+      "loss": 1.9682,
+      "step": 285900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001491598905875345,
+      "loss": 1.9603,
+      "step": 285905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014915828104672963,
+      "loss": 1.9347,
+      "step": 285910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014915667148913142,
+      "loss": 2.0571,
+      "step": 285915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001491550619147404,
+      "loss": 2.0906,
+      "step": 285920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014915345232355707,
+      "loss": 2.154,
+      "step": 285925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014915184271558206,
+      "loss": 2.1482,
+      "step": 285930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001491502330908159,
+      "loss": 2.1995,
+      "step": 285935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014914862344925914,
+      "loss": 2.096,
+      "step": 285940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001491470137909123,
+      "loss": 2.0986,
+      "step": 285945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001491454041157759,
+      "loss": 2.2783,
+      "step": 285950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014914379442385061,
+      "loss": 2.1464,
+      "step": 285955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001491421847151369,
+      "loss": 2.1225,
+      "step": 285960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014914057498963533,
+      "loss": 1.8193,
+      "step": 285965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014913896524734643,
+      "loss": 1.9793,
+      "step": 285970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014913735548827077,
+      "loss": 1.8763,
+      "step": 285975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001491357457124089,
+      "loss": 2.003,
+      "step": 285980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014913413591976138,
+      "loss": 2.0456,
+      "step": 285985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014913252611032873,
+      "loss": 1.9706,
+      "step": 285990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014913091628411152,
+      "loss": 2.105,
+      "step": 285995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001491293064411103,
+      "loss": 2.0566,
+      "step": 286000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014912769658132562,
+      "loss": 1.9846,
+      "step": 286005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014912608670475803,
+      "loss": 2.2364,
+      "step": 286010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014912447681140806,
+      "loss": 2.0184,
+      "step": 286015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001491228669012763,
+      "loss": 2.0162,
+      "step": 286020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014912125697436328,
+      "loss": 2.0216,
+      "step": 286025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014911964703066954,
+      "loss": 1.8627,
+      "step": 286030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014911803707019563,
+      "loss": 2.0433,
+      "step": 286035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014911642709294212,
+      "loss": 2.2288,
+      "step": 286040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014911481709890953,
+      "loss": 2.0113,
+      "step": 286045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014911320708809847,
+      "loss": 1.9796,
+      "step": 286050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001491115970605094,
+      "loss": 2.092,
+      "step": 286055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014910998701614294,
+      "loss": 1.9866,
+      "step": 286060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001491083769549996,
+      "loss": 2.042,
+      "step": 286065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014910676687707996,
+      "loss": 1.8939,
+      "step": 286070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001491051567823846,
+      "loss": 2.0911,
+      "step": 286075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014910354667091398,
+      "loss": 2.0269,
+      "step": 286080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014910193654266867,
+      "loss": 2.0927,
+      "step": 286085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014910032639764927,
+      "loss": 1.8961,
+      "step": 286090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014909871623585636,
+      "loss": 2.128,
+      "step": 286095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001490971060572904,
+      "loss": 1.916,
+      "step": 286100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014909549586195197,
+      "loss": 2.1098,
+      "step": 286105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014909388564984162,
+      "loss": 2.1127,
+      "step": 286110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014909227542095994,
+      "loss": 1.9936,
+      "step": 286115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014909066517530744,
+      "loss": 1.9629,
+      "step": 286120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014908905491288469,
+      "loss": 2.0824,
+      "step": 286125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014908744463369217,
+      "loss": 2.0301,
+      "step": 286130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014908583433773052,
+      "loss": 2.1512,
+      "step": 286135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014908422402500028,
+      "loss": 2.0153,
+      "step": 286140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014908261369550195,
+      "loss": 2.0511,
+      "step": 286145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014908100334923614,
+      "loss": 2.1,
+      "step": 286150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014907939298620338,
+      "loss": 1.862,
+      "step": 286155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014907778260640417,
+      "loss": 1.8825,
+      "step": 286160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001490761722098391,
+      "loss": 2.1039,
+      "step": 286165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014907456179650874,
+      "loss": 1.8696,
+      "step": 286170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001490729513664136,
+      "loss": 2.0303,
+      "step": 286175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014907134091955425,
+      "loss": 1.9936,
+      "step": 286180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014906973045593127,
+      "loss": 1.9465,
+      "step": 286185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014906811997554515,
+      "loss": 2.1803,
+      "step": 286190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014906650947839648,
+      "loss": 2.0062,
+      "step": 286195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001490648989644858,
+      "loss": 1.9475,
+      "step": 286200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014906328843381368,
+      "loss": 2.1886,
+      "step": 286205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014906167788638066,
+      "loss": 2.1699,
+      "step": 286210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014906006732218725,
+      "loss": 2.1163,
+      "step": 286215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.000149058456741234,
+      "loss": 2.2578,
+      "step": 286220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014905684614352155,
+      "loss": 2.2271,
+      "step": 286225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014905523552905035,
+      "loss": 1.8495,
+      "step": 286230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014905362489782103,
+      "loss": 2.0842,
+      "step": 286235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014905201424983408,
+      "loss": 1.9964,
+      "step": 286240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014905040358509006,
+      "loss": 2.0403,
+      "step": 286245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014904879290358956,
+      "loss": 2.1392,
+      "step": 286250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001490471822053331,
+      "loss": 1.8434,
+      "step": 286255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001490455714903212,
+      "loss": 1.9977,
+      "step": 286260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001490439607585545,
+      "loss": 2.0133,
+      "step": 286265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014904235001003344,
+      "loss": 2.0551,
+      "step": 286270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00014904073924475867,
+      "loss": 2.0038,
+      "step": 286275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014903912846273064,
+      "loss": 2.0177,
+      "step": 286280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014903751766395,
+      "loss": 2.0335,
+      "step": 286285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014903590684841724,
+      "loss": 2.1778,
+      "step": 286290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014903429601613292,
+      "loss": 2.0288,
+      "step": 286295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001490326851670976,
+      "loss": 2.1999,
+      "step": 286300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001490310743013118,
+      "loss": 1.9933,
+      "step": 286305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014902946341877613,
+      "loss": 2.2019,
+      "step": 286310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001490278525194911,
+      "loss": 2.0162,
+      "step": 286315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014902624160345726,
+      "loss": 2.0445,
+      "step": 286320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014902463067067515,
+      "loss": 2.2054,
+      "step": 286325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014902301972114537,
+      "loss": 2.1599,
+      "step": 286330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014902140875486844,
+      "loss": 2.0099,
+      "step": 286335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014901979777184486,
+      "loss": 1.9726,
+      "step": 286340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001490181867720753,
+      "loss": 2.1637,
+      "step": 286345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001490165757555602,
+      "loss": 2.066,
+      "step": 286350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001490149647223001,
+      "loss": 2.1277,
+      "step": 286355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014901335367229566,
+      "loss": 2.051,
+      "step": 286360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014901174260554737,
+      "loss": 2.1448,
+      "step": 286365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00014901013152205576,
+      "loss": 1.9897,
+      "step": 286370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001490085204218214,
+      "loss": 2.1118,
+      "step": 286375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014900690930484483,
+      "loss": 2.243,
+      "step": 286380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014900529817112664,
+      "loss": 2.0834,
+      "step": 286385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014900368702066734,
+      "loss": 2.0903,
+      "step": 286390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014900207585346752,
+      "loss": 2.1335,
+      "step": 286395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014900046466952764,
+      "loss": 2.0855,
+      "step": 286400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014899885346884831,
+      "loss": 2.0914,
+      "step": 286405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014899724225143015,
+      "loss": 2.0239,
+      "step": 286410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001489956310172736,
+      "loss": 2.1873,
+      "step": 286415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00014899401976637927,
+      "loss": 2.2065,
+      "step": 286420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014899240849874769,
+      "loss": 2.0292,
+      "step": 286425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001489907972143794,
+      "loss": 2.1157,
+      "step": 286430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0,
+      "learning_rate": 0.000148989185913275,
+      "loss": 2.0965,
+      "step": 286435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014898757459543498,
+      "loss": 2.0459,
+      "step": 286440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001489859632608599,
+      "loss": 2.076,
+      "step": 286445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014898435190955036,
+      "loss": 2.2639,
+      "step": 286450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014898274054150686,
+      "loss": 2.1635,
+      "step": 286455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014898112915672998,
+      "loss": 2.1584,
+      "step": 286460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014897951775522022,
+      "loss": 1.934,
+      "step": 286465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014897790633697824,
+      "loss": 2.1097,
+      "step": 286470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014897629490200447,
+      "loss": 2.0175,
+      "step": 286475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001489746834502995,
+      "loss": 2.0678,
+      "step": 286480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001489730719818639,
+      "loss": 2.0914,
+      "step": 286485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014897146049669822,
+      "loss": 2.2293,
+      "step": 286490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.000148969848994803,
+      "loss": 2.0633,
+      "step": 286495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014896823747617878,
+      "loss": 2.0015,
+      "step": 286500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014896662594082616,
+      "loss": 2.0683,
+      "step": 286505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014896501438874563,
+      "loss": 2.045,
+      "step": 286510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014896340281993773,
+      "loss": 2.1251,
+      "step": 286515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014896179123440306,
+      "loss": 2.1164,
+      "step": 286520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001489601796321422,
+      "loss": 2.0994,
+      "step": 286525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001489585680131556,
+      "loss": 2.0398,
+      "step": 286530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001489569563774439,
+      "loss": 1.9845,
+      "step": 286535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014895534472500756,
+      "loss": 2.1545,
+      "step": 286540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014895373305584723,
+      "loss": 1.9729,
+      "step": 286545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014895212136996344,
+      "loss": 1.8699,
+      "step": 286550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014895050966735667,
+      "loss": 2.0259,
+      "step": 286555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014894889794802755,
+      "loss": 2.0097,
+      "step": 286560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014894728621197659,
+      "loss": 2.0494,
+      "step": 286565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014894567445920435,
+      "loss": 1.9146,
+      "step": 286570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014894406268971136,
+      "loss": 2.06,
+      "step": 286575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014894245090349824,
+      "loss": 2.0192,
+      "step": 286580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014894083910056545,
+      "loss": 1.9157,
+      "step": 286585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014893922728091358,
+      "loss": 2.0819,
+      "step": 286590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001489376154445432,
+      "loss": 1.9799,
+      "step": 286595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014893600359145484,
+      "loss": 2.1209,
+      "step": 286600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014893439172164905,
+      "loss": 2.0662,
+      "step": 286605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00014893277983512635,
+      "loss": 1.9604,
+      "step": 286610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001489311679318874,
+      "loss": 2.0433,
+      "step": 286615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001489295560119326,
+      "loss": 2.0523,
+      "step": 286620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014892794407526263,
+      "loss": 2.0878,
+      "step": 286625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014892633212187798,
+      "loss": 2.0061,
+      "step": 286630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014892472015177918,
+      "loss": 1.9734,
+      "step": 286635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014892310816496683,
+      "loss": 1.929,
+      "step": 286640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014892149616144143,
+      "loss": 1.9946,
+      "step": 286645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001489198841412036,
+      "loss": 2.0981,
+      "step": 286650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014891827210425381,
+      "loss": 2.3805,
+      "step": 286655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014891666005059268,
+      "loss": 1.9942,
+      "step": 286660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014891504798022077,
+      "loss": 1.9769,
+      "step": 286665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014891343589313853,
+      "loss": 2.1354,
+      "step": 286670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014891182378934658,
+      "loss": 1.8241,
+      "step": 286675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00014891021166884546,
+      "loss": 2.1936,
+      "step": 286680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014890859953163577,
+      "loss": 1.9066,
+      "step": 286685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014890698737771797,
+      "loss": 2.2228,
+      "step": 286690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014890537520709268,
+      "loss": 1.9526,
+      "step": 286695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014890376301976042,
+      "loss": 2.1585,
+      "step": 286700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.328125,
+      "learning_rate": 0.00014890215081572172,
+      "loss": 1.9834,
+      "step": 286705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014890053859497723,
+      "loss": 2.1535,
+      "step": 286710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.00014889892635752734,
+      "loss": 2.134,
+      "step": 286715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014889731410337275,
+      "loss": 1.9773,
+      "step": 286720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014889570183251394,
+      "loss": 2.2333,
+      "step": 286725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014889408954495145,
+      "loss": 2.0746,
+      "step": 286730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014889247724068588,
+      "loss": 2.0421,
+      "step": 286735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014889086491971773,
+      "loss": 1.9059,
+      "step": 286740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001488892525820476,
+      "loss": 2.1248,
+      "step": 286745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000148887640227676,
+      "loss": 2.0575,
+      "step": 286750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014888602785660348,
+      "loss": 2.0982,
+      "step": 286755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014888441546883062,
+      "loss": 2.1581,
+      "step": 286760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014888280306435796,
+      "loss": 2.1631,
+      "step": 286765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014888119064318604,
+      "loss": 2.0094,
+      "step": 286770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014887957820531545,
+      "loss": 2.162,
+      "step": 286775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014887796575074666,
+      "loss": 1.9946,
+      "step": 286780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001488763532794803,
+      "loss": 1.8707,
+      "step": 286785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001488747407915169,
+      "loss": 1.9873,
+      "step": 286790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.000148873128286857,
+      "loss": 2.0101,
+      "step": 286795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014887151576550113,
+      "loss": 2.0559,
+      "step": 286800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014886990322744988,
+      "loss": 1.9813,
+      "step": 286805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014886829067270376,
+      "loss": 1.9829,
+      "step": 286810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014886667810126337,
+      "loss": 1.8183,
+      "step": 286815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014886506551312924,
+      "loss": 1.9452,
+      "step": 286820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001488634529083019,
+      "loss": 1.8934,
+      "step": 286825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014886184028678194,
+      "loss": 2.1226,
+      "step": 286830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014886022764856987,
+      "loss": 2.0742,
+      "step": 286835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014885861499366624,
+      "loss": 1.8283,
+      "step": 286840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014885700232207167,
+      "loss": 1.9794,
+      "step": 286845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014885538963378664,
+      "loss": 2.2563,
+      "step": 286850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014885377692881174,
+      "loss": 2.0329,
+      "step": 286855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014885216420714746,
+      "loss": 2.0846,
+      "step": 286860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014885055146879442,
+      "loss": 2.0164,
+      "step": 286865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014884893871375317,
+      "loss": 2.1014,
+      "step": 286870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014884732594202423,
+      "loss": 1.958,
+      "step": 286875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014884571315360812,
+      "loss": 2.0533,
+      "step": 286880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014884410034850544,
+      "loss": 2.0734,
+      "step": 286885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014884248752671675,
+      "loss": 2.0687,
+      "step": 286890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014884087468824256,
+      "loss": 1.9702,
+      "step": 286895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014883926183308346,
+      "loss": 1.9385,
+      "step": 286900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014883764896123997,
+      "loss": 2.0146,
+      "step": 286905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014883603607271267,
+      "loss": 2.1144,
+      "step": 286910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014883442316750208,
+      "loss": 1.9551,
+      "step": 286915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014883281024560875,
+      "loss": 1.856,
+      "step": 286920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014883119730703327,
+      "loss": 2.0571,
+      "step": 286925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014882958435177618,
+      "loss": 2.1542,
+      "step": 286930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000148827971379838,
+      "loss": 2.0585,
+      "step": 286935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014882635839121928,
+      "loss": 2.0596,
+      "step": 286940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014882474538592059,
+      "loss": 1.9263,
+      "step": 286945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014882313236394253,
+      "loss": 2.0044,
+      "step": 286950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014882151932528554,
+      "loss": 2.0176,
+      "step": 286955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001488199062699503,
+      "loss": 2.0286,
+      "step": 286960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014881829319793725,
+      "loss": 1.9409,
+      "step": 286965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000148816680109247,
+      "loss": 2.051,
+      "step": 286970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014881506700388007,
+      "loss": 2.0902,
+      "step": 286975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014881345388183706,
+      "loss": 2.0637,
+      "step": 286980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014881184074311846,
+      "loss": 2.064,
+      "step": 286985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014881022758772488,
+      "loss": 1.9321,
+      "step": 286990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014880861441565678,
+      "loss": 2.1388,
+      "step": 286995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014880700122691482,
+      "loss": 2.0306,
+      "step": 287000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014880538802149952,
+      "loss": 2.1273,
+      "step": 287005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001488037747994114,
+      "loss": 2.1329,
+      "step": 287010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014880216156065102,
+      "loss": 2.2063,
+      "step": 287015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001488005483052189,
+      "loss": 1.9213,
+      "step": 287020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001487989350331157,
+      "loss": 2.0655,
+      "step": 287025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014879732174434183,
+      "loss": 2.1328,
+      "step": 287030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014879570843889795,
+      "loss": 2.0111,
+      "step": 287035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014879409511678454,
+      "loss": 2.1546,
+      "step": 287040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001487924817780022,
+      "loss": 1.9865,
+      "step": 287045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014879086842255147,
+      "loss": 2.1589,
+      "step": 287050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001487892550504329,
+      "loss": 2.1453,
+      "step": 287055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014878764166164702,
+      "loss": 2.1156,
+      "step": 287060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001487860282561944,
+      "loss": 1.945,
+      "step": 287065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001487844148340756,
+      "loss": 1.957,
+      "step": 287070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014878280139529114,
+      "loss": 1.9931,
+      "step": 287075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014878118793984162,
+      "loss": 2.1414,
+      "step": 287080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014877957446772752,
+      "loss": 1.9917,
+      "step": 287085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014877796097894947,
+      "loss": 2.0293,
+      "step": 287090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014877634747350796,
+      "loss": 2.2156,
+      "step": 287095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014877473395140357,
+      "loss": 2.0253,
+      "step": 287100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014877312041263686,
+      "loss": 2.2171,
+      "step": 287105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014877150685720835,
+      "loss": 2.1698,
+      "step": 287110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014876989328511863,
+      "loss": 2.0515,
+      "step": 287115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014876827969636822,
+      "loss": 2.1362,
+      "step": 287120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001487666660909577,
+      "loss": 2.1321,
+      "step": 287125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014876505246888755,
+      "loss": 2.0309,
+      "step": 287130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014876343883015838,
+      "loss": 2.0425,
+      "step": 287135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014876182517477077,
+      "loss": 1.9568,
+      "step": 287140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014876021150272522,
+      "loss": 1.9866,
+      "step": 287145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014875859781402233,
+      "loss": 1.9999,
+      "step": 287150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014875698410866253,
+      "loss": 2.1394,
+      "step": 287155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014875537038664657,
+      "loss": 2.0899,
+      "step": 287160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014875375664797483,
+      "loss": 2.0617,
+      "step": 287165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014875214289264795,
+      "loss": 2.1056,
+      "step": 287170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014875052912066643,
+      "loss": 1.9404,
+      "step": 287175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014874891533203085,
+      "loss": 2.0795,
+      "step": 287180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014874730152674177,
+      "loss": 2.1358,
+      "step": 287185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001487456877047997,
+      "loss": 2.1057,
+      "step": 287190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014874407386620522,
+      "loss": 2.1643,
+      "step": 287195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001487424600109589,
+      "loss": 1.9791,
+      "step": 287200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014874084613906128,
+      "loss": 2.0807,
+      "step": 287205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001487392322505129,
+      "loss": 2.1393,
+      "step": 287210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001487376183453143,
+      "loss": 2.1449,
+      "step": 287215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014873600442346605,
+      "loss": 2.0983,
+      "step": 287220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001487343904849687,
+      "loss": 2.0572,
+      "step": 287225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001487327765298228,
+      "loss": 1.9213,
+      "step": 287230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014873116255802887,
+      "loss": 2.1142,
+      "step": 287235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014872954856958755,
+      "loss": 2.0757,
+      "step": 287240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014872793456449928,
+      "loss": 1.9368,
+      "step": 287245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001487263205427647,
+      "loss": 2.0938,
+      "step": 287250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014872470650438433,
+      "loss": 2.0719,
+      "step": 287255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014872309244935868,
+      "loss": 2.0855,
+      "step": 287260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014872147837768833,
+      "loss": 1.7879,
+      "step": 287265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014871986428937386,
+      "loss": 1.8989,
+      "step": 287270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014871825018441582,
+      "loss": 2.2007,
+      "step": 287275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014871663606281474,
+      "loss": 2.0556,
+      "step": 287280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00014871502192457116,
+      "loss": 2.0066,
+      "step": 287285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001487134077696856,
+      "loss": 1.996,
+      "step": 287290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014871179359815873,
+      "loss": 2.0585,
+      "step": 287295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000148710179409991,
+      "loss": 2.224,
+      "step": 287300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.000148708565205183,
+      "loss": 2.0057,
+      "step": 287305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014870695098373525,
+      "loss": 2.0852,
+      "step": 287310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001487053367456483,
+      "loss": 2.0377,
+      "step": 287315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001487037224909228,
+      "loss": 2.2839,
+      "step": 287320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014870210821955917,
+      "loss": 1.9052,
+      "step": 287325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014870049393155802,
+      "loss": 1.9834,
+      "step": 287330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014869887962691992,
+      "loss": 2.1106,
+      "step": 287335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014869726530564536,
+      "loss": 2.1391,
+      "step": 287340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014869565096773497,
+      "loss": 2.0177,
+      "step": 287345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014869403661318923,
+      "loss": 2.0803,
+      "step": 287350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014869242224200874,
+      "loss": 2.0139,
+      "step": 287355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014869080785419404,
+      "loss": 2.169,
+      "step": 287360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014868919344974567,
+      "loss": 2.1505,
+      "step": 287365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001486875790286642,
+      "loss": 2.0679,
+      "step": 287370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014868596459095015,
+      "loss": 1.9704,
+      "step": 287375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014868435013660412,
+      "loss": 2.0123,
+      "step": 287380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001486827356656266,
+      "loss": 2.066,
+      "step": 287385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014868112117801823,
+      "loss": 2.1625,
+      "step": 287390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014867950667377943,
+      "loss": 2.1811,
+      "step": 287395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014867789215291087,
+      "loss": 2.0655,
+      "step": 287400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014867627761541306,
+      "loss": 2.0962,
+      "step": 287405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014867466306128655,
+      "loss": 2.1553,
+      "step": 287410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001486730484905319,
+      "loss": 2.1064,
+      "step": 287415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014867143390314964,
+      "loss": 2.0984,
+      "step": 287420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001486698192991403,
+      "loss": 2.0501,
+      "step": 287425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014866820467850453,
+      "loss": 2.3108,
+      "step": 287430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001486665900412428,
+      "loss": 2.0886,
+      "step": 287435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014866497538735567,
+      "loss": 2.1072,
+      "step": 287440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001486633607168437,
+      "loss": 2.0162,
+      "step": 287445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014866174602970743,
+      "loss": 2.0529,
+      "step": 287450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014866013132594746,
+      "loss": 2.0902,
+      "step": 287455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001486585166055643,
+      "loss": 2.058,
+      "step": 287460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001486569018685585,
+      "loss": 2.1272,
+      "step": 287465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014865528711493063,
+      "loss": 2.0738,
+      "step": 287470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001486536723446812,
+      "loss": 2.2041,
+      "step": 287475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001486520575578108,
+      "loss": 2.0972,
+      "step": 287480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014865044275432,
+      "loss": 1.9929,
+      "step": 287485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001486488279342093,
+      "loss": 2.1555,
+      "step": 287490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014864721309747932,
+      "loss": 2.1084,
+      "step": 287495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014864559824413051,
+      "loss": 2.1228,
+      "step": 287500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014864398337416352,
+      "loss": 2.2229,
+      "step": 287505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014864236848757887,
+      "loss": 1.8981,
+      "step": 287510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014864075358437707,
+      "loss": 1.9858,
+      "step": 287515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014863913866455873,
+      "loss": 2.0638,
+      "step": 287520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001486375237281244,
+      "loss": 2.0762,
+      "step": 287525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014863590877507455,
+      "loss": 1.9999,
+      "step": 287530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014863429380540985,
+      "loss": 2.1136,
+      "step": 287535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014863267881913076,
+      "loss": 2.0794,
+      "step": 287540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001486310638162379,
+      "loss": 2.0326,
+      "step": 287545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014862944879673174,
+      "loss": 2.1329,
+      "step": 287550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001486278337606129,
+      "loss": 1.9843,
+      "step": 287555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014862621870788192,
+      "loss": 2.0741,
+      "step": 287560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001486246036385393,
+      "loss": 2.0701,
+      "step": 287565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014862298855258568,
+      "loss": 2.1295,
+      "step": 287570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014862137345002157,
+      "loss": 2.174,
+      "step": 287575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014861975833084748,
+      "loss": 1.9444,
+      "step": 287580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000148618143195064,
+      "loss": 1.9636,
+      "step": 287585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014861652804267172,
+      "loss": 2.202,
+      "step": 287590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014861491287367114,
+      "loss": 2.0489,
+      "step": 287595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001486132976880628,
+      "loss": 2.0919,
+      "step": 287600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001486116824858473,
+      "loss": 1.9271,
+      "step": 287605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014861006726702512,
+      "loss": 2.0198,
+      "step": 287610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014860845203159694,
+      "loss": 1.8816,
+      "step": 287615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014860683677956317,
+      "loss": 1.9384,
+      "step": 287620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014860522151092443,
+      "loss": 2.1074,
+      "step": 287625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001486036062256813,
+      "loss": 2.0486,
+      "step": 287630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014860199092383428,
+      "loss": 1.9882,
+      "step": 287635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001486003756053839,
+      "loss": 1.8785,
+      "step": 287640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001485987602703308,
+      "loss": 1.9851,
+      "step": 287645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014859714491867546,
+      "loss": 1.9891,
+      "step": 287650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014859552955041844,
+      "loss": 2.1448,
+      "step": 287655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014859391416556034,
+      "loss": 2.1005,
+      "step": 287660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014859229876410166,
+      "loss": 2.0462,
+      "step": 287665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014859068334604296,
+      "loss": 2.1134,
+      "step": 287670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014858906791138482,
+      "loss": 2.2317,
+      "step": 287675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014858745246012776,
+      "loss": 1.818,
+      "step": 287680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014858583699227235,
+      "loss": 2.0769,
+      "step": 287685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001485842215078191,
+      "loss": 1.9404,
+      "step": 287690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001485826060067687,
+      "loss": 2.1169,
+      "step": 287695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001485809904891215,
+      "loss": 2.087,
+      "step": 287700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014857937495487818,
+      "loss": 1.9912,
+      "step": 287705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001485777594040393,
+      "loss": 2.1028,
+      "step": 287710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014857614383660534,
+      "loss": 2.051,
+      "step": 287715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014857452825257689,
+      "loss": 1.9221,
+      "step": 287720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001485729126519545,
+      "loss": 1.8901,
+      "step": 287725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014857129703473875,
+      "loss": 2.1195,
+      "step": 287730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014856968140093015,
+      "loss": 2.1747,
+      "step": 287735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014856806575052925,
+      "loss": 1.8392,
+      "step": 287740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001485664500835366,
+      "loss": 2.1239,
+      "step": 287745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014856483439995284,
+      "loss": 2.1344,
+      "step": 287750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001485632186997784,
+      "loss": 2.0999,
+      "step": 287755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001485616029830139,
+      "loss": 1.9849,
+      "step": 287760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014855998724965986,
+      "loss": 2.1215,
+      "step": 287765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014855837149971686,
+      "loss": 2.0287,
+      "step": 287770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014855675573318541,
+      "loss": 2.1408,
+      "step": 287775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014855513995006613,
+      "loss": 1.9422,
+      "step": 287780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014855352415035953,
+      "loss": 2.0853,
+      "step": 287785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014855190833406616,
+      "loss": 2.0218,
+      "step": 287790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014855029250118658,
+      "loss": 1.876,
+      "step": 287795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001485486766517213,
+      "loss": 1.988,
+      "step": 287800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014854706078567094,
+      "loss": 2.1161,
+      "step": 287805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014854544490303605,
+      "loss": 2.1674,
+      "step": 287810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014854382900381715,
+      "loss": 2.0382,
+      "step": 287815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001485422130880148,
+      "loss": 2.0491,
+      "step": 287820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014854059715562948,
+      "loss": 1.925,
+      "step": 287825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014853898120666188,
+      "loss": 1.9961,
+      "step": 287830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001485373652411125,
+      "loss": 2.0367,
+      "step": 287835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001485357492589818,
+      "loss": 1.9309,
+      "step": 287840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001485341332602705,
+      "loss": 2.0246,
+      "step": 287845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000148532517244979,
+      "loss": 2.1224,
+      "step": 287850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001485309012131079,
+      "loss": 2.0802,
+      "step": 287855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014852928516465778,
+      "loss": 2.0643,
+      "step": 287860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014852766909962921,
+      "loss": 2.071,
+      "step": 287865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014852605301802268,
+      "loss": 1.9547,
+      "step": 287870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014852443691983877,
+      "loss": 1.9975,
+      "step": 287875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000148522820805078,
+      "loss": 2.1623,
+      "step": 287880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000148521204673741,
+      "loss": 2.1905,
+      "step": 287885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014851958852582828,
+      "loss": 2.1391,
+      "step": 287890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014851797236134037,
+      "loss": 2.0087,
+      "step": 287895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014851635618027783,
+      "loss": 2.0563,
+      "step": 287900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014851473998264122,
+      "loss": 2.1523,
+      "step": 287905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014851312376843115,
+      "loss": 2.0776,
+      "step": 287910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014851150753764807,
+      "loss": 2.1336,
+      "step": 287915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014850989129029259,
+      "loss": 1.9629,
+      "step": 287920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014850827502636523,
+      "loss": 2.2015,
+      "step": 287925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001485066587458666,
+      "loss": 2.176,
+      "step": 287930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001485050424487972,
+      "loss": 2.0393,
+      "step": 287935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014850342613515757,
+      "loss": 2.1495,
+      "step": 287940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014850180980494835,
+      "loss": 2.0136,
+      "step": 287945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014850019345816998,
+      "loss": 2.0472,
+      "step": 287950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001484985770948231,
+      "loss": 2.005,
+      "step": 287955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001484969607149082,
+      "loss": 1.8915,
+      "step": 287960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014849534431842586,
+      "loss": 1.8865,
+      "step": 287965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014849372790537663,
+      "loss": 2.3355,
+      "step": 287970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014849211147576106,
+      "loss": 2.0504,
+      "step": 287975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.6875,
+      "learning_rate": 0.00014849049502957976,
+      "loss": 2.1585,
+      "step": 287980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014848887856683317,
+      "loss": 2.0194,
+      "step": 287985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001484872620875219,
+      "loss": 2.0734,
+      "step": 287990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014848564559164653,
+      "loss": 2.0963,
+      "step": 287995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014848402907920754,
+      "loss": 1.9471,
+      "step": 288000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014848241255020558,
+      "loss": 2.0466,
+      "step": 288005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014848079600464112,
+      "loss": 2.1084,
+      "step": 288010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014847917944251475,
+      "loss": 2.1905,
+      "step": 288015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000148477562863827,
+      "loss": 1.9626,
+      "step": 288020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014847594626857846,
+      "loss": 2.1633,
+      "step": 288025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014847432965676963,
+      "loss": 2.1002,
+      "step": 288030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001484727130284011,
+      "loss": 2.0625,
+      "step": 288035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001484710963834734,
+      "loss": 2.0641,
+      "step": 288040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014846947972198713,
+      "loss": 2.1707,
+      "step": 288045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001484678630439428,
+      "loss": 1.8971,
+      "step": 288050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014846624634934095,
+      "loss": 2.0816,
+      "step": 288055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014846462963818215,
+      "loss": 1.8752,
+      "step": 288060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014846301291046693,
+      "loss": 2.045,
+      "step": 288065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014846139616619592,
+      "loss": 1.9078,
+      "step": 288070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001484597794053696,
+      "loss": 2.0303,
+      "step": 288075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014845816262798852,
+      "loss": 2.0354,
+      "step": 288080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014845654583405327,
+      "loss": 2.1084,
+      "step": 288085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001484549290235644,
+      "loss": 1.9247,
+      "step": 288090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001484533121965224,
+      "loss": 2.0936,
+      "step": 288095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001484516953529279,
+      "loss": 2.1746,
+      "step": 288100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014845007849278142,
+      "loss": 2.1391,
+      "step": 288105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014844846161608352,
+      "loss": 2.0215,
+      "step": 288110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014844684472283475,
+      "loss": 1.976,
+      "step": 288115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014844522781303564,
+      "loss": 1.9423,
+      "step": 288120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014844361088668678,
+      "loss": 2.0509,
+      "step": 288125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001484419939437887,
+      "loss": 2.2192,
+      "step": 288130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014844037698434195,
+      "loss": 1.7791,
+      "step": 288135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001484387600083471,
+      "loss": 2.0001,
+      "step": 288140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001484371430158047,
+      "loss": 1.9371,
+      "step": 288145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014843552600671527,
+      "loss": 1.8301,
+      "step": 288150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014843390898107938,
+      "loss": 2.1416,
+      "step": 288155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001484322919388976,
+      "loss": 2.1268,
+      "step": 288160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014843067488017048,
+      "loss": 2.0986,
+      "step": 288165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014842905780489856,
+      "loss": 1.9788,
+      "step": 288170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014842744071308237,
+      "loss": 2.2046,
+      "step": 288175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001484258236047225,
+      "loss": 2.1621,
+      "step": 288180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014842420647981952,
+      "loss": 2.1536,
+      "step": 288185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014842258933837394,
+      "loss": 1.9739,
+      "step": 288190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014842097218038632,
+      "loss": 2.2179,
+      "step": 288195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001484193550058572,
+      "loss": 2.1328,
+      "step": 288200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014841773781478716,
+      "loss": 2.178,
+      "step": 288205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014841612060717676,
+      "loss": 2.327,
+      "step": 288210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001484145033830265,
+      "loss": 2.2772,
+      "step": 288215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014841288614233703,
+      "loss": 2.2987,
+      "step": 288220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014841126888510877,
+      "loss": 2.1056,
+      "step": 288225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014840965161134237,
+      "loss": 1.93,
+      "step": 288230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014840803432103835,
+      "loss": 1.8221,
+      "step": 288235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014840641701419728,
+      "loss": 2.1287,
+      "step": 288240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001484047996908197,
+      "loss": 2.0096,
+      "step": 288245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014840318235090615,
+      "loss": 1.9797,
+      "step": 288250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014840156499445722,
+      "loss": 1.8473,
+      "step": 288255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001483999476214734,
+      "loss": 2.0515,
+      "step": 288260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001483983302319553,
+      "loss": 1.938,
+      "step": 288265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014839671282590347,
+      "loss": 2.2187,
+      "step": 288270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014839509540331844,
+      "loss": 2.1782,
+      "step": 288275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014839347796420074,
+      "loss": 2.1551,
+      "step": 288280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014839186050855097,
+      "loss": 2.1462,
+      "step": 288285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014839024303636966,
+      "loss": 1.9843,
+      "step": 288290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014838862554765735,
+      "loss": 2.0747,
+      "step": 288295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014838700804241465,
+      "loss": 2.1598,
+      "step": 288300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014838539052064204,
+      "loss": 2.0588,
+      "step": 288305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001483837729823401,
+      "loss": 1.9024,
+      "step": 288310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014838215542750937,
+      "loss": 1.935,
+      "step": 288315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014838053785615047,
+      "loss": 1.9711,
+      "step": 288320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014837892026826386,
+      "loss": 2.1685,
+      "step": 288325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00014837730266385014,
+      "loss": 1.966,
+      "step": 288330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014837568504290984,
+      "loss": 1.8928,
+      "step": 288335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014837406740544359,
+      "loss": 1.9809,
+      "step": 288340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014837244975145181,
+      "loss": 1.878,
+      "step": 288345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014837083208093515,
+      "loss": 1.9739,
+      "step": 288350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001483692143938942,
+      "loss": 1.9,
+      "step": 288355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014836759669032933,
+      "loss": 1.9503,
+      "step": 288360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001483659789702413,
+      "loss": 1.9805,
+      "step": 288365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014836436123363052,
+      "loss": 2.158,
+      "step": 288370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014836274348049763,
+      "loss": 2.2122,
+      "step": 288375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014836112571084314,
+      "loss": 2.212,
+      "step": 288380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001483595079246676,
+      "loss": 2.0373,
+      "step": 288385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001483578901219716,
+      "loss": 2.104,
+      "step": 288390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014835627230275566,
+      "loss": 2.2241,
+      "step": 288395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014835465446702036,
+      "loss": 2.1763,
+      "step": 288400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001483530366147662,
+      "loss": 2.0355,
+      "step": 288405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001483514187459938,
+      "loss": 2.1415,
+      "step": 288410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001483498008607036,
+      "loss": 2.0217,
+      "step": 288415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014834818295889632,
+      "loss": 2.2492,
+      "step": 288420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014834656504057237,
+      "loss": 2.1358,
+      "step": 288425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001483449471057324,
+      "loss": 2.0581,
+      "step": 288430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001483433291543769,
+      "loss": 2.1389,
+      "step": 288435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001483417111865064,
+      "loss": 1.779,
+      "step": 288440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014834009320212156,
+      "loss": 2.1022,
+      "step": 288445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014833847520122283,
+      "loss": 2.1192,
+      "step": 288450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014833685718381085,
+      "loss": 2.0312,
+      "step": 288455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014833523914988607,
+      "loss": 2.0928,
+      "step": 288460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014833362109944914,
+      "loss": 2.0261,
+      "step": 288465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001483320030325005,
+      "loss": 1.947,
+      "step": 288470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001483303849490408,
+      "loss": 1.9743,
+      "step": 288475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001483287668490706,
+      "loss": 2.2023,
+      "step": 288480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001483271487325904,
+      "loss": 2.0412,
+      "step": 288485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014832553059960076,
+      "loss": 2.195,
+      "step": 288490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001483239124501022,
+      "loss": 2.0253,
+      "step": 288495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014832229428409538,
+      "loss": 2.0558,
+      "step": 288500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014832067610158077,
+      "loss": 2.0698,
+      "step": 288505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014831905790255894,
+      "loss": 2.1234,
+      "step": 288510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014831743968703043,
+      "loss": 2.1892,
+      "step": 288515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014831582145499581,
+      "loss": 2.0648,
+      "step": 288520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014831420320645563,
+      "loss": 1.9325,
+      "step": 288525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014831258494141044,
+      "loss": 1.9965,
+      "step": 288530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00014831096665986082,
+      "loss": 2.133,
+      "step": 288535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001483093483618073,
+      "loss": 2.1575,
+      "step": 288540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014830773004725038,
+      "loss": 2.0926,
+      "step": 288545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001483061117161907,
+      "loss": 2.0247,
+      "step": 288550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014830449336862876,
+      "loss": 2.1212,
+      "step": 288555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014830287500456515,
+      "loss": 2.0288,
+      "step": 288560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014830125662400039,
+      "loss": 2.0397,
+      "step": 288565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014829963822693504,
+      "loss": 2.108,
+      "step": 288570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014829801981336966,
+      "loss": 2.1811,
+      "step": 288575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00014829640138330479,
+      "loss": 1.9335,
+      "step": 288580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000148294782936741,
+      "loss": 1.9293,
+      "step": 288585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014829316447367884,
+      "loss": 2.0183,
+      "step": 288590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001482915459941189,
+      "loss": 2.0858,
+      "step": 288595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014828992749806163,
+      "loss": 2.1189,
+      "step": 288600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014828830898550764,
+      "loss": 2.1182,
+      "step": 288605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014828669045645752,
+      "loss": 2.0014,
+      "step": 288610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001482850719109118,
+      "loss": 1.9891,
+      "step": 288615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014828345334887103,
+      "loss": 2.0641,
+      "step": 288620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014828183477033573,
+      "loss": 1.9306,
+      "step": 288625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014828021617530648,
+      "loss": 1.9892,
+      "step": 288630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014827859756378386,
+      "loss": 1.9248,
+      "step": 288635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014827697893576839,
+      "loss": 2.1287,
+      "step": 288640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001482753602912606,
+      "loss": 2.0652,
+      "step": 288645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001482737416302611,
+      "loss": 2.042,
+      "step": 288650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014827212295277035,
+      "loss": 1.9051,
+      "step": 288655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014827050425878904,
+      "loss": 2.1671,
+      "step": 288660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014826888554831763,
+      "loss": 2.1377,
+      "step": 288665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.0,
+      "learning_rate": 0.00014826726682135667,
+      "loss": 1.9481,
+      "step": 288670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014826564807790676,
+      "loss": 2.0913,
+      "step": 288675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014826402931796844,
+      "loss": 2.0825,
+      "step": 288680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014826241054154225,
+      "loss": 2.0824,
+      "step": 288685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014826079174862873,
+      "loss": 1.9085,
+      "step": 288690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014825917293922845,
+      "loss": 1.9803,
+      "step": 288695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014825755411334194,
+      "loss": 2.0204,
+      "step": 288700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014825593527096982,
+      "loss": 2.1092,
+      "step": 288705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014825431641211256,
+      "loss": 1.9115,
+      "step": 288710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014825269753677075,
+      "loss": 1.9509,
+      "step": 288715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014825107864494498,
+      "loss": 2.1244,
+      "step": 288720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014824945973663573,
+      "loss": 2.0387,
+      "step": 288725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014824784081184363,
+      "loss": 2.1236,
+      "step": 288730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00014824622187056916,
+      "loss": 2.1441,
+      "step": 288735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001482446029128129,
+      "loss": 2.165,
+      "step": 288740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014824298393857543,
+      "loss": 2.2282,
+      "step": 288745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001482413649478573,
+      "loss": 2.0906,
+      "step": 288750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.000148239745940659,
+      "loss": 2.1082,
+      "step": 288755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014823812691698116,
+      "loss": 1.9321,
+      "step": 288760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014823650787682425,
+      "loss": 2.114,
+      "step": 288765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014823488882018893,
+      "loss": 2.0703,
+      "step": 288770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014823326974707567,
+      "loss": 2.1485,
+      "step": 288775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014823165065748508,
+      "loss": 2.1779,
+      "step": 288780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014823003155141766,
+      "loss": 2.1752,
+      "step": 288785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014822841242887396,
+      "loss": 2.1409,
+      "step": 288790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001482267932898546,
+      "loss": 1.9634,
+      "step": 288795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014822517413436007,
+      "loss": 1.9039,
+      "step": 288800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014822355496239096,
+      "loss": 2.0709,
+      "step": 288805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014822193577394783,
+      "loss": 1.925,
+      "step": 288810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014822031656903118,
+      "loss": 1.8884,
+      "step": 288815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001482186973476416,
+      "loss": 1.9721,
+      "step": 288820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014821707810977962,
+      "loss": 1.8981,
+      "step": 288825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014821545885544584,
+      "loss": 2.0787,
+      "step": 288830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014821383958464078,
+      "loss": 2.0044,
+      "step": 288835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.000148212220297365,
+      "loss": 2.2192,
+      "step": 288840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014821060099361907,
+      "loss": 2.048,
+      "step": 288845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001482089816734035,
+      "loss": 2.1376,
+      "step": 288850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014820736233671887,
+      "loss": 2.1911,
+      "step": 288855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001482057429835657,
+      "loss": 2.1245,
+      "step": 288860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014820412361394463,
+      "loss": 2.1671,
+      "step": 288865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001482025042278561,
+      "loss": 2.0764,
+      "step": 288870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014820088482530078,
+      "loss": 2.0126,
+      "step": 288875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001481992654062791,
+      "loss": 2.0629,
+      "step": 288880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014819764597079173,
+      "loss": 2.0246,
+      "step": 288885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014819602651883917,
+      "loss": 2.1383,
+      "step": 288890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014819440705042194,
+      "loss": 2.0809,
+      "step": 288895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014819278756554062,
+      "loss": 2.2841,
+      "step": 288900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014819116806419582,
+      "loss": 1.8853,
+      "step": 288905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014818954854638801,
+      "loss": 1.9967,
+      "step": 288910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00014818792901211779,
+      "loss": 2.2039,
+      "step": 288915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001481863094613857,
+      "loss": 2.0609,
+      "step": 288920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014818468989419225,
+      "loss": 2.0408,
+      "step": 288925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014818307031053807,
+      "loss": 2.1632,
+      "step": 288930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014818145071042368,
+      "loss": 2.1738,
+      "step": 288935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014817983109384966,
+      "loss": 2.1186,
+      "step": 288940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001481782114608165,
+      "loss": 2.0095,
+      "step": 288945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014817659181132477,
+      "loss": 2.0565,
+      "step": 288950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014817497214537507,
+      "loss": 1.8928,
+      "step": 288955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014817335246296794,
+      "loss": 2.1012,
+      "step": 288960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001481717327641039,
+      "loss": 1.9898,
+      "step": 288965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014817011304878354,
+      "loss": 2.0683,
+      "step": 288970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014816849331700737,
+      "loss": 2.1807,
+      "step": 288975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000148166873568776,
+      "loss": 2.055,
+      "step": 288980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014816525380408993,
+      "loss": 2.0925,
+      "step": 288985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014816363402294974,
+      "loss": 2.1145,
+      "step": 288990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014816201422535597,
+      "loss": 1.8308,
+      "step": 288995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014816039441130919,
+      "loss": 2.1832,
+      "step": 289000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014815877458080993,
+      "loss": 2.1523,
+      "step": 289005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014815715473385878,
+      "loss": 2.3392,
+      "step": 289010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014815553487045627,
+      "loss": 2.0749,
+      "step": 289015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014815391499060297,
+      "loss": 2.1894,
+      "step": 289020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001481522950942994,
+      "loss": 2.265,
+      "step": 289025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014815067518154614,
+      "loss": 2.1327,
+      "step": 289030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001481490552523437,
+      "loss": 1.9493,
+      "step": 289035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001481474353066927,
+      "loss": 2.1597,
+      "step": 289040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014814581534459368,
+      "loss": 1.9404,
+      "step": 289045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001481441953660472,
+      "loss": 2.0281,
+      "step": 289050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014814257537105373,
+      "loss": 2.1323,
+      "step": 289055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014814095535961388,
+      "loss": 1.9008,
+      "step": 289060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014813933533172825,
+      "loss": 2.0264,
+      "step": 289065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014813771528739733,
+      "loss": 1.8395,
+      "step": 289070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014813609522662172,
+      "loss": 2.2078,
+      "step": 289075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001481344751494019,
+      "loss": 1.9756,
+      "step": 289080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014813285505573846,
+      "loss": 2.1752,
+      "step": 289085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000148131234945632,
+      "loss": 2.0376,
+      "step": 289090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014812961481908306,
+      "loss": 2.0496,
+      "step": 289095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014812799467609213,
+      "loss": 1.9962,
+      "step": 289100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014812637451665982,
+      "loss": 2.2423,
+      "step": 289105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014812475434078666,
+      "loss": 2.096,
+      "step": 289110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014812313414847323,
+      "loss": 2.0973,
+      "step": 289115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014812151393972002,
+      "loss": 2.0734,
+      "step": 289120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014811989371452767,
+      "loss": 2.0613,
+      "step": 289125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014811827347289668,
+      "loss": 2.0847,
+      "step": 289130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001481166532148276,
+      "loss": 1.988,
+      "step": 289135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014811503294032103,
+      "loss": 2.1379,
+      "step": 289140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001481134126493775,
+      "loss": 2.0613,
+      "step": 289145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001481117923419975,
+      "loss": 2.158,
+      "step": 289150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014811017201818167,
+      "loss": 2.085,
+      "step": 289155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014810855167793056,
+      "loss": 1.968,
+      "step": 289160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014810693132124466,
+      "loss": 2.1183,
+      "step": 289165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014810531094812455,
+      "loss": 2.0253,
+      "step": 289170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014810369055857082,
+      "loss": 1.915,
+      "step": 289175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000148102070152584,
+      "loss": 1.9652,
+      "step": 289180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014810044973016464,
+      "loss": 2.2082,
+      "step": 289185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014809882929131326,
+      "loss": 1.8267,
+      "step": 289190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014809720883603047,
+      "loss": 2.0796,
+      "step": 289195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014809558836431682,
+      "loss": 2.1846,
+      "step": 289200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014809396787617284,
+      "loss": 1.9883,
+      "step": 289205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014809234737159907,
+      "loss": 1.8953,
+      "step": 289210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001480907268505961,
+      "loss": 2.2189,
+      "step": 289215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014808910631316442,
+      "loss": 2.031,
+      "step": 289220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001480874857593047,
+      "loss": 1.8659,
+      "step": 289225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014808586518901738,
+      "loss": 2.1095,
+      "step": 289230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014808424460230305,
+      "loss": 2.0041,
+      "step": 289235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014808262399916227,
+      "loss": 2.1611,
+      "step": 289240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001480810033795956,
+      "loss": 2.2393,
+      "step": 289245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001480793827436036,
+      "loss": 2.1002,
+      "step": 289250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001480777620911868,
+      "loss": 2.1218,
+      "step": 289255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014807614142234578,
+      "loss": 1.7466,
+      "step": 289260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014807452073708105,
+      "loss": 1.9956,
+      "step": 289265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014807290003539323,
+      "loss": 1.9415,
+      "step": 289270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001480712793172828,
+      "loss": 2.1741,
+      "step": 289275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014806965858275037,
+      "loss": 2.1604,
+      "step": 289280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014806803783179646,
+      "loss": 1.9478,
+      "step": 289285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014806641706442163,
+      "loss": 2.3317,
+      "step": 289290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014806479628062645,
+      "loss": 2.1365,
+      "step": 289295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014806317548041146,
+      "loss": 2.0435,
+      "step": 289300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014806155466377723,
+      "loss": 2.0654,
+      "step": 289305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001480599338307243,
+      "loss": 1.9876,
+      "step": 289310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014805831298125323,
+      "loss": 2.3093,
+      "step": 289315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014805669211536457,
+      "loss": 2.2273,
+      "step": 289320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014805507123305883,
+      "loss": 2.2142,
+      "step": 289325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014805345033433668,
+      "loss": 1.9817,
+      "step": 289330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014805182941919853,
+      "loss": 2.076,
+      "step": 289335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014805020848764503,
+      "loss": 2.1142,
+      "step": 289340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014804858753967673,
+      "loss": 2.0818,
+      "step": 289345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014804696657529413,
+      "loss": 2.1564,
+      "step": 289350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001480453455944978,
+      "loss": 1.9898,
+      "step": 289355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014804372459728836,
+      "loss": 2.1346,
+      "step": 289360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014804210358366628,
+      "loss": 2.1986,
+      "step": 289365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014804048255363217,
+      "loss": 2.0453,
+      "step": 289370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014803886150718652,
+      "loss": 2.0335,
+      "step": 289375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014803724044432993,
+      "loss": 1.9132,
+      "step": 289380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014803561936506298,
+      "loss": 1.9926,
+      "step": 289385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014803399826938618,
+      "loss": 2.0849,
+      "step": 289390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014803237715730008,
+      "loss": 2.0981,
+      "step": 289395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014803075602880524,
+      "loss": 1.9478,
+      "step": 289400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014802913488390227,
+      "loss": 2.1635,
+      "step": 289405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014802751372259163,
+      "loss": 2.1715,
+      "step": 289410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014802589254487393,
+      "loss": 2.0613,
+      "step": 289415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014802427135074972,
+      "loss": 2.0784,
+      "step": 289420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014802265014021956,
+      "loss": 2.017,
+      "step": 289425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014802102891328398,
+      "loss": 2.0443,
+      "step": 289430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014801940766994353,
+      "loss": 2.1038,
+      "step": 289435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001480177864101988,
+      "loss": 2.021,
+      "step": 289440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014801616513405031,
+      "loss": 2.0931,
+      "step": 289445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014801454384149863,
+      "loss": 2.0265,
+      "step": 289450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014801292253254435,
+      "loss": 1.8311,
+      "step": 289455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014801130120718794,
+      "loss": 1.9507,
+      "step": 289460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014800967986543,
+      "loss": 2.1257,
+      "step": 289465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014800805850727111,
+      "loss": 2.163,
+      "step": 289470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014800643713271176,
+      "loss": 2.1355,
+      "step": 289475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001480048157417526,
+      "loss": 2.2366,
+      "step": 289480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014800319433439408,
+      "loss": 1.9038,
+      "step": 289485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001480015729106368,
+      "loss": 2.0795,
+      "step": 289490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.0001479999514704813,
+      "loss": 2.0242,
+      "step": 289495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014799833001392817,
+      "loss": 2.0679,
+      "step": 289500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014799670854097793,
+      "loss": 2.0669,
+      "step": 289505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014799508705163114,
+      "loss": 2.0433,
+      "step": 289510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014799346554588834,
+      "loss": 2.1028,
+      "step": 289515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014799184402375016,
+      "loss": 2.0257,
+      "step": 289520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014799022248521707,
+      "loss": 2.016,
+      "step": 289525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014798860093028963,
+      "loss": 2.0972,
+      "step": 289530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014798697935896842,
+      "loss": 1.9544,
+      "step": 289535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014798535777125397,
+      "loss": 2.1681,
+      "step": 289540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001479837361671469,
+      "loss": 2.1712,
+      "step": 289545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014798211454664767,
+      "loss": 2.114,
+      "step": 289550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001479804929097569,
+      "loss": 2.0499,
+      "step": 289555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014797887125647514,
+      "loss": 2.0455,
+      "step": 289560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001479772495868029,
+      "loss": 2.1412,
+      "step": 289565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014797562790074077,
+      "loss": 2.058,
+      "step": 289570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001479740061982893,
+      "loss": 2.2402,
+      "step": 289575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014797238447944902,
+      "loss": 2.0853,
+      "step": 289580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014797076274422052,
+      "loss": 2.1705,
+      "step": 289585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014796914099260433,
+      "loss": 2.0109,
+      "step": 289590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014796751922460102,
+      "loss": 2.2378,
+      "step": 289595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014796589744021113,
+      "loss": 1.9695,
+      "step": 289600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014796427563943524,
+      "loss": 2.0328,
+      "step": 289605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014796265382227385,
+      "loss": 2.0862,
+      "step": 289610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001479610319887276,
+      "loss": 1.9318,
+      "step": 289615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014795941013879692,
+      "loss": 2.1722,
+      "step": 289620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001479577882724825,
+      "loss": 2.0993,
+      "step": 289625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014795616638978477,
+      "loss": 1.9838,
+      "step": 289630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001479545444907044,
+      "loss": 2.0963,
+      "step": 289635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014795292257524184,
+      "loss": 1.9779,
+      "step": 289640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014795130064339772,
+      "loss": 2.0123,
+      "step": 289645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014794967869517252,
+      "loss": 2.1953,
+      "step": 289650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001479480567305669,
+      "loss": 1.9833,
+      "step": 289655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014794643474958137,
+      "loss": 2.1847,
+      "step": 289660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014794481275221642,
+      "loss": 1.9859,
+      "step": 289665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014794319073847267,
+      "loss": 2.0885,
+      "step": 289670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014794156870835063,
+      "loss": 2.0574,
+      "step": 289675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001479399466618509,
+      "loss": 2.0762,
+      "step": 289680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014793832459897402,
+      "loss": 2.072,
+      "step": 289685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014793670251972055,
+      "loss": 1.9777,
+      "step": 289690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000147935080424091,
+      "loss": 1.8646,
+      "step": 289695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.000147933458312086,
+      "loss": 2.0999,
+      "step": 289700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014793183618370603,
+      "loss": 1.9364,
+      "step": 289705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014793021403895168,
+      "loss": 2.01,
+      "step": 289710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001479285918778235,
+      "loss": 2.0784,
+      "step": 289715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014792696970032204,
+      "loss": 2.024,
+      "step": 289720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014792534750644788,
+      "loss": 1.9636,
+      "step": 289725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014792372529620154,
+      "loss": 2.1108,
+      "step": 289730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001479221030695836,
+      "loss": 2.2024,
+      "step": 289735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014792048082659456,
+      "loss": 2.0575,
+      "step": 289740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014791885856723505,
+      "loss": 2.0068,
+      "step": 289745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014791723629150558,
+      "loss": 2.103,
+      "step": 289750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001479156139994067,
+      "loss": 2.1814,
+      "step": 289755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014791399169093898,
+      "loss": 1.992,
+      "step": 289760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.000147912369366103,
+      "loss": 2.1038,
+      "step": 289765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014791074702489925,
+      "loss": 1.9707,
+      "step": 289770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014790912466732835,
+      "loss": 2.118,
+      "step": 289775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001479075022933908,
+      "loss": 2.1178,
+      "step": 289780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00014790587990308717,
+      "loss": 1.8205,
+      "step": 289785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014790425749641804,
+      "loss": 2.0708,
+      "step": 289790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014790263507338396,
+      "loss": 2.2374,
+      "step": 289795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014790101263398547,
+      "loss": 2.1479,
+      "step": 289800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014789939017822312,
+      "loss": 1.9809,
+      "step": 289805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014789776770609742,
+      "loss": 2.1853,
+      "step": 289810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014789614521760905,
+      "loss": 1.988,
+      "step": 289815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014789452271275844,
+      "loss": 2.0071,
+      "step": 289820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014789290019154622,
+      "loss": 2.1485,
+      "step": 289825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001478912776539729,
+      "loss": 2.2276,
+      "step": 289830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014788965510003903,
+      "loss": 2.0651,
+      "step": 289835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014788803252974525,
+      "loss": 2.1506,
+      "step": 289840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000147886409943092,
+      "loss": 2.032,
+      "step": 289845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001478847873400799,
+      "loss": 2.1896,
+      "step": 289850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014788316472070948,
+      "loss": 1.8782,
+      "step": 289855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014788154208498128,
+      "loss": 2.0291,
+      "step": 289860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014787991943289593,
+      "loss": 2.0867,
+      "step": 289865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014787829676445388,
+      "loss": 2.0127,
+      "step": 289870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014787667407965575,
+      "loss": 1.9675,
+      "step": 289875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001478750513785021,
+      "loss": 2.0402,
+      "step": 289880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014787342866099345,
+      "loss": 2.2714,
+      "step": 289885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014787180592713036,
+      "loss": 2.0531,
+      "step": 289890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014787018317691342,
+      "loss": 1.9605,
+      "step": 289895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014786856041034313,
+      "loss": 2.111,
+      "step": 289900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014786693762742007,
+      "loss": 1.9644,
+      "step": 289905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014786531482814482,
+      "loss": 1.94,
+      "step": 289910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001478636920125179,
+      "loss": 1.9961,
+      "step": 289915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014786206918053987,
+      "loss": 1.9675,
+      "step": 289920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001478604463322113,
+      "loss": 2.066,
+      "step": 289925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014785882346753275,
+      "loss": 1.9302,
+      "step": 289930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014785720058650473,
+      "loss": 2.1209,
+      "step": 289935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014785557768912778,
+      "loss": 2.1254,
+      "step": 289940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014785395477540255,
+      "loss": 2.1264,
+      "step": 289945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014785233184532953,
+      "loss": 2.1495,
+      "step": 289950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014785070889890932,
+      "loss": 1.8619,
+      "step": 289955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014784908593614243,
+      "loss": 1.8672,
+      "step": 289960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014784746295702937,
+      "loss": 2.139,
+      "step": 289965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014784583996157078,
+      "loss": 2.109,
+      "step": 289970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001478442169497672,
+      "loss": 2.2069,
+      "step": 289975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014784259392161916,
+      "loss": 1.8893,
+      "step": 289980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001478409708771272,
+      "loss": 2.1866,
+      "step": 289985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014783934781629191,
+      "loss": 2.1302,
+      "step": 289990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014783772473911383,
+      "loss": 1.9571,
+      "step": 289995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014783610164559354,
+      "loss": 2.0109,
+      "step": 290000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014783447853573156,
+      "loss": 2.1529,
+      "step": 290005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014783285540952842,
+      "loss": 1.9412,
+      "step": 290010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014783123226698474,
+      "loss": 2.1154,
+      "step": 290015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014782960910810104,
+      "loss": 2.0621,
+      "step": 290020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014782798593287788,
+      "loss": 1.9615,
+      "step": 290025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014782636274131582,
+      "loss": 2.1032,
+      "step": 290030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014782473953341537,
+      "loss": 2.0525,
+      "step": 290035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014782311630917716,
+      "loss": 2.017,
+      "step": 290040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001478214930686017,
+      "loss": 2.1274,
+      "step": 290045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014781986981168954,
+      "loss": 2.1064,
+      "step": 290050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014781824653844122,
+      "loss": 2.1063,
+      "step": 290055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014781662324885736,
+      "loss": 1.9687,
+      "step": 290060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014781499994293845,
+      "loss": 1.9815,
+      "step": 290065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001478133766206851,
+      "loss": 1.9637,
+      "step": 290070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014781175328209783,
+      "loss": 2.1889,
+      "step": 290075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014781012992717715,
+      "loss": 2.1351,
+      "step": 290080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001478085065559237,
+      "loss": 1.9935,
+      "step": 290085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014780688316833802,
+      "loss": 1.9575,
+      "step": 290090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014780525976442062,
+      "loss": 2.0797,
+      "step": 290095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014780363634417208,
+      "loss": 2.1107,
+      "step": 290100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001478020129075929,
+      "loss": 2.2973,
+      "step": 290105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014780038945468374,
+      "loss": 2.1227,
+      "step": 290110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014779876598544509,
+      "loss": 1.8211,
+      "step": 290115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014779714249987753,
+      "loss": 2.1783,
+      "step": 290120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014779551899798157,
+      "loss": 2.1042,
+      "step": 290125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001477938954797578,
+      "loss": 1.9236,
+      "step": 290130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014779227194520677,
+      "loss": 2.1181,
+      "step": 290135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014779064839432902,
+      "loss": 2.2312,
+      "step": 290140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014778902482712517,
+      "loss": 2.1128,
+      "step": 290145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014778740124359567,
+      "loss": 2.1782,
+      "step": 290150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014778577764374114,
+      "loss": 2.0381,
+      "step": 290155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014778415402756214,
+      "loss": 1.7969,
+      "step": 290160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.796875,
+      "learning_rate": 0.00014778253039505917,
+      "loss": 2.1679,
+      "step": 290165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014778090674623287,
+      "loss": 2.1413,
+      "step": 290170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001477792830810837,
+      "loss": 2.1161,
+      "step": 290175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014777765939961228,
+      "loss": 2.1326,
+      "step": 290180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014777603570181915,
+      "loss": 2.1471,
+      "step": 290185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014777441198770484,
+      "loss": 2.1466,
+      "step": 290190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014777278825726996,
+      "loss": 1.8255,
+      "step": 290195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000147771164510515,
+      "loss": 1.9496,
+      "step": 290200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014776954074744057,
+      "loss": 2.0726,
+      "step": 290205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014776791696804716,
+      "loss": 2.1367,
+      "step": 290210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014776629317233536,
+      "loss": 1.9671,
+      "step": 290215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014776466936030578,
+      "loss": 2.0181,
+      "step": 290220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001477630455319589,
+      "loss": 1.9437,
+      "step": 290225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001477614216872953,
+      "loss": 1.8935,
+      "step": 290230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014775979782631548,
+      "loss": 2.0806,
+      "step": 290235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001477581739490201,
+      "loss": 2.178,
+      "step": 290240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014775655005540964,
+      "loss": 2.1003,
+      "step": 290245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001477549261454847,
+      "loss": 2.0298,
+      "step": 290250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001477533022192458,
+      "loss": 1.9888,
+      "step": 290255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014775167827669352,
+      "loss": 1.9965,
+      "step": 290260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014775005431782838,
+      "loss": 2.0493,
+      "step": 290265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014774843034265098,
+      "loss": 2.1263,
+      "step": 290270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014774680635116182,
+      "loss": 2.2496,
+      "step": 290275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001477451823433615,
+      "loss": 1.9411,
+      "step": 290280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014774355831925054,
+      "loss": 2.0062,
+      "step": 290285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014774193427882955,
+      "loss": 2.2843,
+      "step": 290290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014774031022209903,
+      "loss": 2.1563,
+      "step": 290295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014773868614905954,
+      "loss": 2.2372,
+      "step": 290300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014773706205971167,
+      "loss": 2.1728,
+      "step": 290305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014773543795405596,
+      "loss": 2.1428,
+      "step": 290310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014773381383209293,
+      "loss": 2.0485,
+      "step": 290315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014773218969382318,
+      "loss": 2.0234,
+      "step": 290320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014773056553924724,
+      "loss": 2.0604,
+      "step": 290325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014772894136836568,
+      "loss": 1.9763,
+      "step": 290330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014772731718117906,
+      "loss": 1.9478,
+      "step": 290335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014772569297768793,
+      "loss": 2.0118,
+      "step": 290340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001477240687578928,
+      "loss": 2.0703,
+      "step": 290345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014772244452179427,
+      "loss": 2.0019,
+      "step": 290350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001477208202693929,
+      "loss": 2.1966,
+      "step": 290355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014771919600068924,
+      "loss": 2.0923,
+      "step": 290360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014771757171568386,
+      "loss": 2.0411,
+      "step": 290365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014771594741437723,
+      "loss": 2.0,
+      "step": 290370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014771432309677,
+      "loss": 2.024,
+      "step": 290375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001477126987628627,
+      "loss": 1.9914,
+      "step": 290380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014771107441265588,
+      "loss": 2.0143,
+      "step": 290385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014770945004615008,
+      "loss": 2.0819,
+      "step": 290390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014770782566334587,
+      "loss": 1.9761,
+      "step": 290395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014770620126424377,
+      "loss": 2.1221,
+      "step": 290400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001477045768488444,
+      "loss": 1.997,
+      "step": 290405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014770295241714826,
+      "loss": 2.1679,
+      "step": 290410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014770132796915596,
+      "loss": 1.9729,
+      "step": 290415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.000147699703504868,
+      "loss": 2.1617,
+      "step": 290420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014769807902428495,
+      "loss": 2.1339,
+      "step": 290425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001476964545274074,
+      "loss": 2.1113,
+      "step": 290430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014769483001423583,
+      "loss": 2.1535,
+      "step": 290435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001476932054847709,
+      "loss": 2.098,
+      "step": 290440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014769158093901305,
+      "loss": 1.9504,
+      "step": 290445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014768995637696292,
+      "loss": 1.9791,
+      "step": 290450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014768833179862103,
+      "loss": 2.0718,
+      "step": 290455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014768670720398795,
+      "loss": 2.0795,
+      "step": 290460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001476850825930642,
+      "loss": 2.107,
+      "step": 290465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014768345796585037,
+      "loss": 2.0554,
+      "step": 290470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000147681833322347,
+      "loss": 2.1674,
+      "step": 290475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001476802086625547,
+      "loss": 2.1337,
+      "step": 290480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014767858398647392,
+      "loss": 1.9264,
+      "step": 290485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001476769592941053,
+      "loss": 2.1091,
+      "step": 290490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014767533458544935,
+      "loss": 2.0194,
+      "step": 290495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001476737098605067,
+      "loss": 2.0567,
+      "step": 290500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014767208511927776,
+      "loss": 2.0105,
+      "step": 290505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001476704603617632,
+      "loss": 2.0617,
+      "step": 290510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014766883558796354,
+      "loss": 2.0237,
+      "step": 290515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014766721079787938,
+      "loss": 1.7857,
+      "step": 290520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001476655859915112,
+      "loss": 2.0672,
+      "step": 290525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001476639611688596,
+      "loss": 2.1913,
+      "step": 290530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014766233632992513,
+      "loss": 2.2452,
+      "step": 290535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014766071147470834,
+      "loss": 1.9204,
+      "step": 290540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014765908660320978,
+      "loss": 2.1318,
+      "step": 290545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014765746171543003,
+      "loss": 2.1899,
+      "step": 290550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001476558368113696,
+      "loss": 2.0291,
+      "step": 290555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014765421189102908,
+      "loss": 2.0983,
+      "step": 290560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014765258695440906,
+      "loss": 2.0438,
+      "step": 290565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014765096200150998,
+      "loss": 2.0543,
+      "step": 290570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001476493370323325,
+      "loss": 2.1205,
+      "step": 290575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014764771204687715,
+      "loss": 2.0361,
+      "step": 290580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014764608704514448,
+      "loss": 1.9713,
+      "step": 290585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014764446202713503,
+      "loss": 2.0454,
+      "step": 290590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014764283699284937,
+      "loss": 2.165,
+      "step": 290595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014764121194228805,
+      "loss": 2.1788,
+      "step": 290600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014763958687545167,
+      "loss": 2.1036,
+      "step": 290605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014763796179234067,
+      "loss": 1.8695,
+      "step": 290610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014763633669295572,
+      "loss": 2.044,
+      "step": 290615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014763471157729733,
+      "loss": 2.0265,
+      "step": 290620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014763308644536607,
+      "loss": 2.1693,
+      "step": 290625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014763146129716245,
+      "loss": 1.8268,
+      "step": 290630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.0001476298361326871,
+      "loss": 1.9716,
+      "step": 290635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001476282109519405,
+      "loss": 2.0728,
+      "step": 290640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014762658575492326,
+      "loss": 2.0394,
+      "step": 290645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001476249605416359,
+      "loss": 2.0894,
+      "step": 290650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014762333531207899,
+      "loss": 2.0031,
+      "step": 290655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014762171006625313,
+      "loss": 1.9392,
+      "step": 290660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014762008480415876,
+      "loss": 2.1078,
+      "step": 290665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014761845952579653,
+      "loss": 2.1169,
+      "step": 290670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014761683423116697,
+      "loss": 2.1176,
+      "step": 290675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014761520892027068,
+      "loss": 1.8721,
+      "step": 290680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014761358359310814,
+      "loss": 1.9661,
+      "step": 290685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001476119582496799,
+      "loss": 2.1672,
+      "step": 290690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014761033288998658,
+      "loss": 2.1266,
+      "step": 290695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014760870751402873,
+      "loss": 1.9789,
+      "step": 290700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014760708212180687,
+      "loss": 1.9439,
+      "step": 290705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014760545671332155,
+      "loss": 2.0445,
+      "step": 290710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014760383128857335,
+      "loss": 2.1246,
+      "step": 290715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014760220584756278,
+      "loss": 2.0576,
+      "step": 290720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014760058039029053,
+      "loss": 2.0773,
+      "step": 290725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000147598954916757,
+      "loss": 2.0268,
+      "step": 290730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001475973294269628,
+      "loss": 1.8506,
+      "step": 290735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001475957039209085,
+      "loss": 2.0985,
+      "step": 290740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014759407839859463,
+      "loss": 2.1325,
+      "step": 290745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014759245286002177,
+      "loss": 1.9736,
+      "step": 290750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014759082730519046,
+      "loss": 2.0876,
+      "step": 290755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014758920173410127,
+      "loss": 1.9171,
+      "step": 290760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001475875761467547,
+      "loss": 2.1952,
+      "step": 290765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001475859505431514,
+      "loss": 2.1153,
+      "step": 290770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001475843249232919,
+      "loss": 2.1364,
+      "step": 290775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014758269928717666,
+      "loss": 2.0794,
+      "step": 290780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014758107363480635,
+      "loss": 2.0686,
+      "step": 290785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001475794479661815,
+      "loss": 2.139,
+      "step": 290790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014757782228130265,
+      "loss": 2.1372,
+      "step": 290795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001475761965801703,
+      "loss": 1.9307,
+      "step": 290800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014757457086278508,
+      "loss": 1.9966,
+      "step": 290805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00014757294512914753,
+      "loss": 2.1019,
+      "step": 290810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014757131937925822,
+      "loss": 2.0595,
+      "step": 290815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014756969361311766,
+      "loss": 2.0393,
+      "step": 290820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001475680678307264,
+      "loss": 2.1918,
+      "step": 290825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014756644203208508,
+      "loss": 2.1909,
+      "step": 290830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001475648162171942,
+      "loss": 2.0191,
+      "step": 290835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014756319038605428,
+      "loss": 2.0719,
+      "step": 290840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014756156453866592,
+      "loss": 2.0012,
+      "step": 290845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014755993867502967,
+      "loss": 2.2247,
+      "step": 290850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001475583127951461,
+      "loss": 2.2957,
+      "step": 290855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014755668689901574,
+      "loss": 1.9757,
+      "step": 290860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014755506098663914,
+      "loss": 2.0426,
+      "step": 290865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014755343505801688,
+      "loss": 1.9824,
+      "step": 290870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001475518091131495,
+      "loss": 2.0069,
+      "step": 290875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014755018315203757,
+      "loss": 2.1486,
+      "step": 290880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014754855717468161,
+      "loss": 2.2583,
+      "step": 290885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014754693118108222,
+      "loss": 2.1297,
+      "step": 290890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014754530517123993,
+      "loss": 2.3206,
+      "step": 290895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001475436791451553,
+      "loss": 1.9638,
+      "step": 290900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001475420531028289,
+      "loss": 2.1254,
+      "step": 290905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014754042704426124,
+      "loss": 2.0826,
+      "step": 290910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014753880096945294,
+      "loss": 2.1241,
+      "step": 290915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014753717487840453,
+      "loss": 1.8282,
+      "step": 290920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014753554877111654,
+      "loss": 1.7986,
+      "step": 290925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014753392264758955,
+      "loss": 1.898,
+      "step": 290930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001475322965078241,
+      "loss": 1.8202,
+      "step": 290935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014753067035182076,
+      "loss": 2.2058,
+      "step": 290940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014752904417958008,
+      "loss": 2.1073,
+      "step": 290945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014752741799110264,
+      "loss": 2.1171,
+      "step": 290950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014752579178638897,
+      "loss": 1.9546,
+      "step": 290955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014752416556543957,
+      "loss": 2.0803,
+      "step": 290960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001475225393282551,
+      "loss": 1.9398,
+      "step": 290965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014752091307483606,
+      "loss": 2.1438,
+      "step": 290970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014751928680518305,
+      "loss": 2.0844,
+      "step": 290975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014751766051929654,
+      "loss": 2.0167,
+      "step": 290980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014751603421717715,
+      "loss": 2.1149,
+      "step": 290985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001475144078988254,
+      "loss": 1.8893,
+      "step": 290990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001475127815642419,
+      "loss": 2.0283,
+      "step": 290995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014751115521342717,
+      "loss": 2.0474,
+      "step": 291000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014750952884638177,
+      "loss": 2.19,
+      "step": 291005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014750790246310623,
+      "loss": 2.0297,
+      "step": 291010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00014750627606360112,
+      "loss": 2.1464,
+      "step": 291015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014750464964786707,
+      "loss": 1.9837,
+      "step": 291020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014750302321590451,
+      "loss": 1.9082,
+      "step": 291025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014750139676771408,
+      "loss": 2.1552,
+      "step": 291030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001474997703032963,
+      "loss": 1.9556,
+      "step": 291035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014749814382265173,
+      "loss": 2.0143,
+      "step": 291040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014749651732578095,
+      "loss": 1.9868,
+      "step": 291045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001474948908126845,
+      "loss": 1.9234,
+      "step": 291050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001474932642833629,
+      "loss": 2.0811,
+      "step": 291055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014749163773781677,
+      "loss": 2.1942,
+      "step": 291060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014749001117604665,
+      "loss": 2.0454,
+      "step": 291065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014748838459805307,
+      "loss": 2.0156,
+      "step": 291070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014748675800383657,
+      "loss": 2.1289,
+      "step": 291075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014748513139339775,
+      "loss": 2.0703,
+      "step": 291080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00014748350476673713,
+      "loss": 2.0508,
+      "step": 291085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014748187812385532,
+      "loss": 2.1226,
+      "step": 291090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014748025146475282,
+      "loss": 2.1159,
+      "step": 291095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001474786247894302,
+      "loss": 2.0498,
+      "step": 291100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000147476998097888,
+      "loss": 1.9467,
+      "step": 291105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014747537139012687,
+      "loss": 1.9193,
+      "step": 291110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014747374466614723,
+      "loss": 2.0895,
+      "step": 291115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001474721179259497,
+      "loss": 2.083,
+      "step": 291120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014747049116953484,
+      "loss": 2.0368,
+      "step": 291125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001474688643969032,
+      "loss": 1.9404,
+      "step": 291130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014746723760805536,
+      "loss": 2.0913,
+      "step": 291135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014746561080299184,
+      "loss": 1.8342,
+      "step": 291140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001474639839817132,
+      "loss": 2.0361,
+      "step": 291145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014746235714421997,
+      "loss": 1.9887,
+      "step": 291150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001474607302905128,
+      "loss": 2.1351,
+      "step": 291155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014745910342059214,
+      "loss": 1.9967,
+      "step": 291160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001474574765344586,
+      "loss": 2.2381,
+      "step": 291165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014745584963211274,
+      "loss": 2.0748,
+      "step": 291170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014745422271355508,
+      "loss": 2.0227,
+      "step": 291175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001474525957787862,
+      "loss": 2.2086,
+      "step": 291180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014745096882780664,
+      "loss": 2.254,
+      "step": 291185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000147449341860617,
+      "loss": 2.1756,
+      "step": 291190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001474477148772178,
+      "loss": 2.1692,
+      "step": 291195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001474460878776096,
+      "loss": 1.9198,
+      "step": 291200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014744446086179292,
+      "loss": 1.9515,
+      "step": 291205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014744283382976838,
+      "loss": 1.8736,
+      "step": 291210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001474412067815365,
+      "loss": 1.9426,
+      "step": 291215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014743957971709786,
+      "loss": 2.1883,
+      "step": 291220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000147437952636453,
+      "loss": 2.1136,
+      "step": 291225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014743632553960244,
+      "loss": 2.0747,
+      "step": 291230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001474346984265468,
+      "loss": 2.0396,
+      "step": 291235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014743307129728658,
+      "loss": 2.0595,
+      "step": 291240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001474314441518224,
+      "loss": 2.0465,
+      "step": 291245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014742981699015476,
+      "loss": 1.955,
+      "step": 291250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014742818981228423,
+      "loss": 2.128,
+      "step": 291255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014742656261821139,
+      "loss": 1.943,
+      "step": 291260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014742493540793677,
+      "loss": 1.9821,
+      "step": 291265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014742330818146092,
+      "loss": 2.0736,
+      "step": 291270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001474216809387844,
+      "loss": 1.9589,
+      "step": 291275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001474200536799078,
+      "loss": 2.1766,
+      "step": 291280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014741842640483163,
+      "loss": 2.1239,
+      "step": 291285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014741679911355648,
+      "loss": 2.163,
+      "step": 291290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001474151718060829,
+      "loss": 2.0595,
+      "step": 291295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001474135444824114,
+      "loss": 1.9926,
+      "step": 291300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001474119171425426,
+      "loss": 2.03,
+      "step": 291305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000147410289786477,
+      "loss": 2.2281,
+      "step": 291310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014740866241421522,
+      "loss": 2.0982,
+      "step": 291315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014740703502575777,
+      "loss": 2.0449,
+      "step": 291320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014740540762110522,
+      "loss": 1.9843,
+      "step": 291325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001474037802002581,
+      "loss": 1.8693,
+      "step": 291330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014740215276321702,
+      "loss": 2.2938,
+      "step": 291335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001474005253099825,
+      "loss": 1.8911,
+      "step": 291340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001473988978405551,
+      "loss": 1.9358,
+      "step": 291345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014739727035493536,
+      "loss": 2.1203,
+      "step": 291350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014739564285312385,
+      "loss": 2.0754,
+      "step": 291355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014739401533512116,
+      "loss": 2.1326,
+      "step": 291360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014739238780092778,
+      "loss": 2.0637,
+      "step": 291365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014739076025054433,
+      "loss": 2.1323,
+      "step": 291370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001473891326839713,
+      "loss": 2.2153,
+      "step": 291375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014738750510120933,
+      "loss": 1.8468,
+      "step": 291380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.578125,
+      "learning_rate": 0.0001473858775022589,
+      "loss": 2.0204,
+      "step": 291385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001473842498871206,
+      "loss": 2.0071,
+      "step": 291390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014738262225579498,
+      "loss": 2.1125,
+      "step": 291395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001473809946082826,
+      "loss": 1.8938,
+      "step": 291400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014737936694458404,
+      "loss": 1.9864,
+      "step": 291405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001473777392646998,
+      "loss": 2.2015,
+      "step": 291410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014737611156863043,
+      "loss": 1.9981,
+      "step": 291415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00014737448385637654,
+      "loss": 1.7972,
+      "step": 291420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001473728561279387,
+      "loss": 2.1641,
+      "step": 291425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014737122838331745,
+      "loss": 2.2098,
+      "step": 291430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001473696006225133,
+      "loss": 2.1111,
+      "step": 291435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001473679728455268,
+      "loss": 2.0767,
+      "step": 291440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014736634505235858,
+      "loss": 2.1855,
+      "step": 291445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001473647172430092,
+      "loss": 2.0139,
+      "step": 291450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014736308941747908,
+      "loss": 2.132,
+      "step": 291455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014736146157576894,
+      "loss": 2.2099,
+      "step": 291460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014735983371787922,
+      "loss": 1.9548,
+      "step": 291465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014735820584381055,
+      "loss": 2.1045,
+      "step": 291470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014735657795356344,
+      "loss": 1.9745,
+      "step": 291475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014735495004713847,
+      "loss": 2.1109,
+      "step": 291480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001473533221245362,
+      "loss": 2.1091,
+      "step": 291485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014735169418575716,
+      "loss": 2.2913,
+      "step": 291490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014735006623080193,
+      "loss": 2.0047,
+      "step": 291495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001473484382596711,
+      "loss": 2.136,
+      "step": 291500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014734681027236512,
+      "loss": 2.0114,
+      "step": 291505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014734518226888462,
+      "loss": 1.9485,
+      "step": 291510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001473435542492302,
+      "loss": 2.1374,
+      "step": 291515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014734192621340233,
+      "loss": 2.0593,
+      "step": 291520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014734029816140158,
+      "loss": 2.1205,
+      "step": 291525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014733867009322852,
+      "loss": 2.0648,
+      "step": 291530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014733704200888375,
+      "loss": 2.0227,
+      "step": 291535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014733541390836777,
+      "loss": 2.0113,
+      "step": 291540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001473337857916812,
+      "loss": 2.1411,
+      "step": 291545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014733215765882446,
+      "loss": 2.0993,
+      "step": 291550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014733052950979825,
+      "loss": 2.1166,
+      "step": 291555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014732890134460308,
+      "loss": 2.2276,
+      "step": 291560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001473272731632395,
+      "loss": 2.1236,
+      "step": 291565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014732564496570803,
+      "loss": 2.2373,
+      "step": 291570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014732401675200926,
+      "loss": 1.9864,
+      "step": 291575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014732238852214376,
+      "loss": 2.0033,
+      "step": 291580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014732076027611213,
+      "loss": 2.1013,
+      "step": 291585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001473191320139148,
+      "loss": 2.0247,
+      "step": 291590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014731750373555243,
+      "loss": 2.1009,
+      "step": 291595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014731587544102555,
+      "loss": 2.0769,
+      "step": 291600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00014731424713033465,
+      "loss": 2.0795,
+      "step": 291605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014731261880348038,
+      "loss": 1.9951,
+      "step": 291610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001473109904604633,
+      "loss": 2.0679,
+      "step": 291615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001473093621012839,
+      "loss": 1.9908,
+      "step": 291620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014730773372594275,
+      "loss": 2.0596,
+      "step": 291625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014730610533444043,
+      "loss": 2.1011,
+      "step": 291630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014730447692677746,
+      "loss": 1.9171,
+      "step": 291635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014730284850295446,
+      "loss": 1.9372,
+      "step": 291640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014730122006297193,
+      "loss": 1.9045,
+      "step": 291645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014729959160683046,
+      "loss": 2.199,
+      "step": 291650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014729796313453058,
+      "loss": 2.088,
+      "step": 291655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014729633464607286,
+      "loss": 2.1017,
+      "step": 291660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014729470614145785,
+      "loss": 2.1463,
+      "step": 291665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014729307762068614,
+      "loss": 2.0956,
+      "step": 291670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014729144908375821,
+      "loss": 1.873,
+      "step": 291675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014728982053067473,
+      "loss": 2.2226,
+      "step": 291680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014728819196143612,
+      "loss": 2.1163,
+      "step": 291685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014728656337604303,
+      "loss": 2.0468,
+      "step": 291690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000147284934774496,
+      "loss": 2.0848,
+      "step": 291695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014728330615679559,
+      "loss": 2.0496,
+      "step": 291700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014728167752294232,
+      "loss": 1.9736,
+      "step": 291705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014728004887293677,
+      "loss": 2.0643,
+      "step": 291710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001472784202067795,
+      "loss": 2.1175,
+      "step": 291715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014727679152447108,
+      "loss": 2.1751,
+      "step": 291720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014727516282601205,
+      "loss": 2.0728,
+      "step": 291725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014727353411140296,
+      "loss": 2.1559,
+      "step": 291730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014727190538064438,
+      "loss": 2.3552,
+      "step": 291735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014727027663373685,
+      "loss": 2.1101,
+      "step": 291740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014726864787068095,
+      "loss": 2.1794,
+      "step": 291745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001472670190914772,
+      "loss": 2.2333,
+      "step": 291750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001472653902961262,
+      "loss": 2.1094,
+      "step": 291755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001472637614846285,
+      "loss": 2.0824,
+      "step": 291760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001472621326569846,
+      "loss": 1.9375,
+      "step": 291765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014726050381319513,
+      "loss": 1.9747,
+      "step": 291770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014725887495326058,
+      "loss": 2.1584,
+      "step": 291775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014725724607718156,
+      "loss": 2.0716,
+      "step": 291780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014725561718495863,
+      "loss": 1.8858,
+      "step": 291785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014725398827659228,
+      "loss": 2.2853,
+      "step": 291790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014725235935208314,
+      "loss": 2.1884,
+      "step": 291795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014725073041143172,
+      "loss": 1.9967,
+      "step": 291800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014724910145463862,
+      "loss": 2.3649,
+      "step": 291805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014724747248170435,
+      "loss": 1.9294,
+      "step": 291810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001472458434926295,
+      "loss": 2.0284,
+      "step": 291815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001472442144874146,
+      "loss": 1.9786,
+      "step": 291820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014724258546606022,
+      "loss": 2.0111,
+      "step": 291825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014724095642856694,
+      "loss": 2.0194,
+      "step": 291830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014723932737493526,
+      "loss": 2.114,
+      "step": 291835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001472376983051658,
+      "loss": 2.2148,
+      "step": 291840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014723606921925907,
+      "loss": 2.1582,
+      "step": 291845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014723444011721564,
+      "loss": 1.9124,
+      "step": 291850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014723281099903606,
+      "loss": 2.1378,
+      "step": 291855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001472311818647209,
+      "loss": 2.019,
+      "step": 291860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014722955271427075,
+      "loss": 2.2297,
+      "step": 291865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014722792354768607,
+      "loss": 1.9924,
+      "step": 291870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001472262943649675,
+      "loss": 2.0195,
+      "step": 291875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001472246651661156,
+      "loss": 2.069,
+      "step": 291880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014722303595113086,
+      "loss": 2.1195,
+      "step": 291885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014722140672001387,
+      "loss": 1.9247,
+      "step": 291890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001472197774727652,
+      "loss": 2.0338,
+      "step": 291895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014721814820938537,
+      "loss": 2.0654,
+      "step": 291900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014721651892987502,
+      "loss": 2.1753,
+      "step": 291905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014721488963423464,
+      "loss": 2.2191,
+      "step": 291910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001472132603224648,
+      "loss": 2.2043,
+      "step": 291915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014721163099456602,
+      "loss": 2.1386,
+      "step": 291920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001472100016505389,
+      "loss": 1.9045,
+      "step": 291925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.000147208372290384,
+      "loss": 2.1457,
+      "step": 291930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014720674291410186,
+      "loss": 2.0055,
+      "step": 291935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014720511352169305,
+      "loss": 2.3818,
+      "step": 291940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001472034841131581,
+      "loss": 2.14,
+      "step": 291945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001472018546884976,
+      "loss": 1.9401,
+      "step": 291950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014720022524771206,
+      "loss": 2.1382,
+      "step": 291955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001471985957908021,
+      "loss": 2.0124,
+      "step": 291960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014719696631776823,
+      "loss": 2.0346,
+      "step": 291965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.000147195336828611,
+      "loss": 2.1541,
+      "step": 291970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014719370732333106,
+      "loss": 2.1333,
+      "step": 291975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001471920778019288,
+      "loss": 2.1391,
+      "step": 291980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014719044826440493,
+      "loss": 2.049,
+      "step": 291985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014718881871075994,
+      "loss": 2.0646,
+      "step": 291990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014718718914099438,
+      "loss": 2.0675,
+      "step": 291995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014718555955510885,
+      "loss": 2.1418,
+      "step": 292000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001471839299531038,
+      "loss": 1.8898,
+      "step": 292005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014718230033497994,
+      "loss": 1.9219,
+      "step": 292010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014718067070073772,
+      "loss": 2.0619,
+      "step": 292015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014717904105037777,
+      "loss": 2.011,
+      "step": 292020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014717741138390054,
+      "loss": 2.209,
+      "step": 292025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014717578170130667,
+      "loss": 2.1159,
+      "step": 292030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001471741520025967,
+      "loss": 2.0276,
+      "step": 292035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001471725222877712,
+      "loss": 2.1565,
+      "step": 292040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001471708925568307,
+      "loss": 1.8948,
+      "step": 292045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014716926280977579,
+      "loss": 2.151,
+      "step": 292050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014716763304660697,
+      "loss": 2.057,
+      "step": 292055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014716600326732484,
+      "loss": 1.9094,
+      "step": 292060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014716437347192995,
+      "loss": 2.0566,
+      "step": 292065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014716274366042287,
+      "loss": 2.0064,
+      "step": 292070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001471611138328041,
+      "loss": 2.2413,
+      "step": 292075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001471594839890743,
+      "loss": 2.0359,
+      "step": 292080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014715785412923392,
+      "loss": 1.9864,
+      "step": 292085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014715622425328357,
+      "loss": 2.1607,
+      "step": 292090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001471545943612238,
+      "loss": 2.063,
+      "step": 292095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014715296445305516,
+      "loss": 1.9295,
+      "step": 292100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014715133452877823,
+      "loss": 2.1282,
+      "step": 292105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014714970458839353,
+      "loss": 2.1758,
+      "step": 292110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014714807463190165,
+      "loss": 2.2134,
+      "step": 292115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014714644465930313,
+      "loss": 2.0425,
+      "step": 292120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014714481467059852,
+      "loss": 1.9881,
+      "step": 292125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014714318466578838,
+      "loss": 2.0976,
+      "step": 292130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001471415546448733,
+      "loss": 2.1505,
+      "step": 292135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014713992460785378,
+      "loss": 2.0746,
+      "step": 292140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014713829455473043,
+      "loss": 1.977,
+      "step": 292145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014713666448550375,
+      "loss": 1.8669,
+      "step": 292150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014713503440017437,
+      "loss": 1.9328,
+      "step": 292155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014713340429874282,
+      "loss": 2.2676,
+      "step": 292160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001471317741812096,
+      "loss": 2.2073,
+      "step": 292165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014713014404757534,
+      "loss": 2.0983,
+      "step": 292170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014712851389784055,
+      "loss": 2.0743,
+      "step": 292175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014712688373200581,
+      "loss": 1.9158,
+      "step": 292180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014712525355007168,
+      "loss": 2.0911,
+      "step": 292185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001471236233520387,
+      "loss": 2.1262,
+      "step": 292190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014712199313790742,
+      "loss": 2.1597,
+      "step": 292195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014712036290767845,
+      "loss": 1.8985,
+      "step": 292200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001471187326613523,
+      "loss": 1.9901,
+      "step": 292205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001471171023989295,
+      "loss": 2.1529,
+      "step": 292210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014711547212041068,
+      "loss": 2.1219,
+      "step": 292215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014711384182579634,
+      "loss": 1.8513,
+      "step": 292220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014711221151508707,
+      "loss": 2.1127,
+      "step": 292225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001471105811882834,
+      "loss": 1.938,
+      "step": 292230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014710895084538592,
+      "loss": 2.0194,
+      "step": 292235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014710732048639518,
+      "loss": 2.0095,
+      "step": 292240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.0001471056901113117,
+      "loss": 1.9108,
+      "step": 292245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014710405972013607,
+      "loss": 2.0631,
+      "step": 292250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001471024293128688,
+      "loss": 2.0053,
+      "step": 292255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00014710079888951053,
+      "loss": 2.0057,
+      "step": 292260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014709916845006177,
+      "loss": 2.043,
+      "step": 292265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001470975379945231,
+      "loss": 2.0327,
+      "step": 292270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014709590752289502,
+      "loss": 2.1187,
+      "step": 292275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014709427703517812,
+      "loss": 2.0958,
+      "step": 292280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.000147092646531373,
+      "loss": 2.2107,
+      "step": 292285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014709101601148016,
+      "loss": 2.3567,
+      "step": 292290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014708938547550017,
+      "loss": 2.0771,
+      "step": 292295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014708775492343356,
+      "loss": 1.9915,
+      "step": 292300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014708612435528095,
+      "loss": 2.0417,
+      "step": 292305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014708449377104287,
+      "loss": 2.004,
+      "step": 292310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001470828631707199,
+      "loss": 2.1805,
+      "step": 292315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014708123255431255,
+      "loss": 2.1357,
+      "step": 292320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014707960192182135,
+      "loss": 1.9854,
+      "step": 292325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014707797127324693,
+      "loss": 2.01,
+      "step": 292330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014707634060858986,
+      "loss": 2.1394,
+      "step": 292335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001470747099278506,
+      "loss": 2.1985,
+      "step": 292340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001470730792310298,
+      "loss": 2.0354,
+      "step": 292345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000147071448518128,
+      "loss": 2.0834,
+      "step": 292350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014706981778914567,
+      "loss": 1.8954,
+      "step": 292355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014706818704408355,
+      "loss": 2.0423,
+      "step": 292360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000147066556282942,
+      "loss": 2.151,
+      "step": 292365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014706492550572168,
+      "loss": 1.9024,
+      "step": 292370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014706329471242312,
+      "loss": 2.133,
+      "step": 292375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014706166390304688,
+      "loss": 2.3522,
+      "step": 292380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014706003307759353,
+      "loss": 2.1247,
+      "step": 292385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014705840223606362,
+      "loss": 1.9568,
+      "step": 292390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014705677137845773,
+      "loss": 2.3078,
+      "step": 292395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014705514050477635,
+      "loss": 1.9398,
+      "step": 292400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014705350961502012,
+      "loss": 1.9886,
+      "step": 292405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014705187870918952,
+      "loss": 2.0814,
+      "step": 292410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014705024778728518,
+      "loss": 2.1841,
+      "step": 292415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001470486168493076,
+      "loss": 2.2216,
+      "step": 292420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014704698589525735,
+      "loss": 2.1894,
+      "step": 292425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014704535492513504,
+      "loss": 1.9841,
+      "step": 292430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014704372393894115,
+      "loss": 1.9694,
+      "step": 292435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001470420929366763,
+      "loss": 1.9953,
+      "step": 292440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014704046191834098,
+      "loss": 2.014,
+      "step": 292445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001470388308839358,
+      "loss": 2.1242,
+      "step": 292450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001470371998334613,
+      "loss": 1.9385,
+      "step": 292455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014703556876691803,
+      "loss": 1.9116,
+      "step": 292460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014703393768430658,
+      "loss": 2.0591,
+      "step": 292465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014703230658562747,
+      "loss": 2.0715,
+      "step": 292470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014703067547088128,
+      "loss": 2.1623,
+      "step": 292475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014702904434006855,
+      "loss": 2.1252,
+      "step": 292480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014702741319318983,
+      "loss": 2.1026,
+      "step": 292485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014702578203024572,
+      "loss": 2.0184,
+      "step": 292490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014702415085123674,
+      "loss": 2.1951,
+      "step": 292495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014702251965616347,
+      "loss": 2.0868,
+      "step": 292500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014702088844502644,
+      "loss": 1.9911,
+      "step": 292505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014701925721782622,
+      "loss": 2.1182,
+      "step": 292510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014701762597456338,
+      "loss": 2.3499,
+      "step": 292515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014701599471523847,
+      "loss": 2.0094,
+      "step": 292520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014701436343985202,
+      "loss": 1.9931,
+      "step": 292525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014701273214840464,
+      "loss": 1.9847,
+      "step": 292530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001470111008408968,
+      "loss": 1.9792,
+      "step": 292535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014700946951732918,
+      "loss": 2.0501,
+      "step": 292540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014700783817770226,
+      "loss": 1.9533,
+      "step": 292545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014700620682201658,
+      "loss": 2.1355,
+      "step": 292550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014700457545027275,
+      "loss": 2.2191,
+      "step": 292555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001470029440624713,
+      "loss": 2.1768,
+      "step": 292560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014700131265861282,
+      "loss": 1.856,
+      "step": 292565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014699968123869778,
+      "loss": 1.9628,
+      "step": 292570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014699804980272685,
+      "loss": 1.9583,
+      "step": 292575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014699641835070048,
+      "loss": 1.945,
+      "step": 292580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014699478688261935,
+      "loss": 2.0333,
+      "step": 292585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001469931553984839,
+      "loss": 2.084,
+      "step": 292590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014699152389829474,
+      "loss": 1.9977,
+      "step": 292595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014698989238205243,
+      "loss": 1.9774,
+      "step": 292600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001469882608497575,
+      "loss": 2.1866,
+      "step": 292605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001469866293014106,
+      "loss": 2.122,
+      "step": 292610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014698499773701216,
+      "loss": 2.1892,
+      "step": 292615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014698336615656275,
+      "loss": 1.9326,
+      "step": 292620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014698173456006303,
+      "loss": 1.9094,
+      "step": 292625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014698010294751353,
+      "loss": 1.8793,
+      "step": 292630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014697847131891472,
+      "loss": 2.1072,
+      "step": 292635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001469768396742672,
+      "loss": 2.0889,
+      "step": 292640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014697520801357157,
+      "loss": 2.0222,
+      "step": 292645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014697357633682834,
+      "loss": 2.1163,
+      "step": 292650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001469719446440381,
+      "loss": 2.1646,
+      "step": 292655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001469703129352014,
+      "loss": 2.0314,
+      "step": 292660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014696868121031876,
+      "loss": 2.0592,
+      "step": 292665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014696704946939076,
+      "loss": 2.1034,
+      "step": 292670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000146965417712418,
+      "loss": 1.7618,
+      "step": 292675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000146963785939401,
+      "loss": 2.0782,
+      "step": 292680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001469621541503403,
+      "loss": 2.17,
+      "step": 292685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014696052234523647,
+      "loss": 2.0414,
+      "step": 292690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014695889052409008,
+      "loss": 2.0383,
+      "step": 292695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014695725868690168,
+      "loss": 2.2203,
+      "step": 292700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014695562683367185,
+      "loss": 2.0533,
+      "step": 292705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001469539949644011,
+      "loss": 2.0269,
+      "step": 292710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014695236307909004,
+      "loss": 2.0168,
+      "step": 292715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001469507311777392,
+      "loss": 2.1567,
+      "step": 292720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014694909926034913,
+      "loss": 1.9902,
+      "step": 292725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014694746732692038,
+      "loss": 2.1333,
+      "step": 292730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014694583537745351,
+      "loss": 1.9251,
+      "step": 292735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014694420341194913,
+      "loss": 2.1392,
+      "step": 292740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014694257143040777,
+      "loss": 2.0946,
+      "step": 292745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014694093943282995,
+      "loss": 1.9931,
+      "step": 292750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014693930741921623,
+      "loss": 2.1486,
+      "step": 292755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00014693767538956722,
+      "loss": 1.9611,
+      "step": 292760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014693604334388346,
+      "loss": 1.9706,
+      "step": 292765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014693441128216548,
+      "loss": 2.1405,
+      "step": 292770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014693277920441386,
+      "loss": 2.1331,
+      "step": 292775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014693114711062915,
+      "loss": 2.0749,
+      "step": 292780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001469295150008119,
+      "loss": 2.0934,
+      "step": 292785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014692788287496268,
+      "loss": 2.1104,
+      "step": 292790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014692625073308205,
+      "loss": 1.9999,
+      "step": 292795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014692461857517057,
+      "loss": 2.0156,
+      "step": 292800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014692298640122878,
+      "loss": 2.2179,
+      "step": 292805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014692135421125725,
+      "loss": 2.1675,
+      "step": 292810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014691972200525652,
+      "loss": 2.1551,
+      "step": 292815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014691808978322716,
+      "loss": 2.0947,
+      "step": 292820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014691645754516975,
+      "loss": 2.0532,
+      "step": 292825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014691482529108482,
+      "loss": 2.1545,
+      "step": 292830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014691319302097294,
+      "loss": 1.9934,
+      "step": 292835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014691156073483465,
+      "loss": 2.1104,
+      "step": 292840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014690992843267053,
+      "loss": 2.0948,
+      "step": 292845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014690829611448112,
+      "loss": 2.0288,
+      "step": 292850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014690666378026698,
+      "loss": 2.0962,
+      "step": 292855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014690503143002868,
+      "loss": 1.9361,
+      "step": 292860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014690339906376677,
+      "loss": 2.1479,
+      "step": 292865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014690176668148182,
+      "loss": 1.965,
+      "step": 292870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014690013428317436,
+      "loss": 2.0662,
+      "step": 292875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000146898501868845,
+      "loss": 2.0799,
+      "step": 292880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014689686943849422,
+      "loss": 2.1047,
+      "step": 292885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014689523699212262,
+      "loss": 2.1834,
+      "step": 292890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014689360452973076,
+      "loss": 2.0078,
+      "step": 292895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001468919720513192,
+      "loss": 2.0569,
+      "step": 292900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014689033955688852,
+      "loss": 2.1412,
+      "step": 292905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001468887070464392,
+      "loss": 1.7825,
+      "step": 292910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014688707451997188,
+      "loss": 1.9783,
+      "step": 292915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014688544197748705,
+      "loss": 2.0836,
+      "step": 292920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014688380941898537,
+      "loss": 1.9762,
+      "step": 292925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014688217684446727,
+      "loss": 2.0384,
+      "step": 292930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014688054425393337,
+      "loss": 1.9675,
+      "step": 292935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014687891164738426,
+      "loss": 2.1104,
+      "step": 292940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014687727902482043,
+      "loss": 1.9819,
+      "step": 292945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001468756463862425,
+      "loss": 2.0765,
+      "step": 292950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.875,
+      "learning_rate": 0.000146874013731651,
+      "loss": 2.1899,
+      "step": 292955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014687238106104647,
+      "loss": 2.0268,
+      "step": 292960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014687074837442947,
+      "loss": 1.9239,
+      "step": 292965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014686911567180058,
+      "loss": 1.9354,
+      "step": 292970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014686748295316036,
+      "loss": 1.8826,
+      "step": 292975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014686585021850935,
+      "loss": 2.1738,
+      "step": 292980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014686421746784814,
+      "loss": 2.1681,
+      "step": 292985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014686258470117728,
+      "loss": 2.1899,
+      "step": 292990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014686095191849725,
+      "loss": 2.1977,
+      "step": 292995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001468593191198087,
+      "loss": 1.8427,
+      "step": 293000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014685768630511215,
+      "loss": 2.1958,
+      "step": 293005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014685605347440817,
+      "loss": 1.9853,
+      "step": 293010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001468544206276973,
+      "loss": 2.0991,
+      "step": 293015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014685278776498014,
+      "loss": 2.0181,
+      "step": 293020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014685115488625717,
+      "loss": 2.0521,
+      "step": 293025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014684952199152902,
+      "loss": 2.176,
+      "step": 293030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014684788908079623,
+      "loss": 2.1405,
+      "step": 293035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014684625615405937,
+      "loss": 2.0763,
+      "step": 293040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014684462321131898,
+      "loss": 2.082,
+      "step": 293045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014684299025257555,
+      "loss": 2.1861,
+      "step": 293050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014684135727782978,
+      "loss": 2.2044,
+      "step": 293055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014683972428708211,
+      "loss": 2.1845,
+      "step": 293060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014683809128033317,
+      "loss": 1.9585,
+      "step": 293065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014683645825758348,
+      "loss": 2.1411,
+      "step": 293070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014683482521883357,
+      "loss": 2.0181,
+      "step": 293075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014683319216408407,
+      "loss": 2.1342,
+      "step": 293080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014683155909333553,
+      "loss": 1.9418,
+      "step": 293085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014682992600658843,
+      "loss": 2.0281,
+      "step": 293090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001468282929038434,
+      "loss": 2.0795,
+      "step": 293095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014682665978510095,
+      "loss": 2.119,
+      "step": 293100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001468250266503617,
+      "loss": 2.1258,
+      "step": 293105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014682339349962614,
+      "loss": 2.2476,
+      "step": 293110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014682176033289487,
+      "loss": 2.0079,
+      "step": 293115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014682012715016844,
+      "loss": 2.0578,
+      "step": 293120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014681849395144743,
+      "loss": 2.0251,
+      "step": 293125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014681686073673235,
+      "loss": 2.0141,
+      "step": 293130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014681522750602377,
+      "loss": 2.113,
+      "step": 293135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014681359425932228,
+      "loss": 2.1054,
+      "step": 293140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014681196099662839,
+      "loss": 2.2413,
+      "step": 293145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014681032771794272,
+      "loss": 2.0245,
+      "step": 293150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014680869442326578,
+      "loss": 1.9411,
+      "step": 293155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014680706111259813,
+      "loss": 2.2148,
+      "step": 293160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014680542778594034,
+      "loss": 2.0298,
+      "step": 293165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014680379444329298,
+      "loss": 2.0962,
+      "step": 293170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014680216108465662,
+      "loss": 2.2339,
+      "step": 293175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014680052771003175,
+      "loss": 2.1242,
+      "step": 293180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014679889431941896,
+      "loss": 2.1274,
+      "step": 293185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014679726091281885,
+      "loss": 1.9925,
+      "step": 293190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014679562749023191,
+      "loss": 2.1638,
+      "step": 293195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014679399405165882,
+      "loss": 2.0426,
+      "step": 293200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014679236059709998,
+      "loss": 2.148,
+      "step": 293205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000146790727126556,
+      "loss": 2.0991,
+      "step": 293210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014678909364002752,
+      "loss": 1.8932,
+      "step": 293215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014678746013751503,
+      "loss": 2.1791,
+      "step": 293220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014678582661901908,
+      "loss": 2.0208,
+      "step": 293225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014678419308454023,
+      "loss": 2.1249,
+      "step": 293230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014678255953407904,
+      "loss": 1.9752,
+      "step": 293235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014678092596763612,
+      "loss": 1.9958,
+      "step": 293240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014677929238521197,
+      "loss": 1.8546,
+      "step": 293245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014677765878680718,
+      "loss": 2.2039,
+      "step": 293250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014677602517242227,
+      "loss": 2.0745,
+      "step": 293255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014677439154205782,
+      "loss": 1.9693,
+      "step": 293260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001467727578957144,
+      "loss": 2.0961,
+      "step": 293265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014677112423339256,
+      "loss": 2.1754,
+      "step": 293270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014676949055509285,
+      "loss": 2.1341,
+      "step": 293275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014676785686081585,
+      "loss": 2.0705,
+      "step": 293280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014676622315056207,
+      "loss": 2.0066,
+      "step": 293285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014676458942433214,
+      "loss": 2.1265,
+      "step": 293290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014676295568212655,
+      "loss": 2.037,
+      "step": 293295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001467613219239459,
+      "loss": 2.0438,
+      "step": 293300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001467596881497907,
+      "loss": 2.0985,
+      "step": 293305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014675805435966158,
+      "loss": 2.0537,
+      "step": 293310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014675642055355907,
+      "loss": 2.1657,
+      "step": 293315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001467547867314837,
+      "loss": 1.9894,
+      "step": 293320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014675315289343604,
+      "loss": 2.1148,
+      "step": 293325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014675151903941668,
+      "loss": 1.947,
+      "step": 293330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014674988516942615,
+      "loss": 2.0847,
+      "step": 293335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000146748251283465,
+      "loss": 2.1465,
+      "step": 293340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014674661738153373,
+      "loss": 1.9506,
+      "step": 293345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00014674498346363304,
+      "loss": 1.9151,
+      "step": 293350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014674334952976344,
+      "loss": 2.1883,
+      "step": 293355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014674171557992543,
+      "loss": 2.3021,
+      "step": 293360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014674008161411962,
+      "loss": 1.9765,
+      "step": 293365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001467384476323465,
+      "loss": 2.025,
+      "step": 293370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014673681363460672,
+      "loss": 2.1649,
+      "step": 293375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001467351796209008,
+      "loss": 2.174,
+      "step": 293380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001467335455912293,
+      "loss": 1.9229,
+      "step": 293385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014673191154559277,
+      "loss": 1.9328,
+      "step": 293390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014673027748399176,
+      "loss": 2.0821,
+      "step": 293395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014672864340642682,
+      "loss": 1.9234,
+      "step": 293400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014672700931289855,
+      "loss": 2.1491,
+      "step": 293405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001467253752034075,
+      "loss": 2.0381,
+      "step": 293410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00014672374107795418,
+      "loss": 2.1967,
+      "step": 293415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001467221069365392,
+      "loss": 2.1248,
+      "step": 293420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014672047277916313,
+      "loss": 2.053,
+      "step": 293425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014671883860582644,
+      "loss": 2.1066,
+      "step": 293430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014671720441652978,
+      "loss": 2.0073,
+      "step": 293435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014671557021127369,
+      "loss": 2.0147,
+      "step": 293440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014671393599005868,
+      "loss": 2.1234,
+      "step": 293445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014671230175288536,
+      "loss": 2.0343,
+      "step": 293450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001467106674997543,
+      "loss": 1.9905,
+      "step": 293455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000146709033230666,
+      "loss": 2.2096,
+      "step": 293460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014670739894562105,
+      "loss": 2.051,
+      "step": 293465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014670576464462001,
+      "loss": 1.9251,
+      "step": 293470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014670413032766343,
+      "loss": 2.1033,
+      "step": 293475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014670249599475183,
+      "loss": 1.8704,
+      "step": 293480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014670086164588587,
+      "loss": 1.9529,
+      "step": 293485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000146699227281066,
+      "loss": 2.0671,
+      "step": 293490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001466975929002929,
+      "loss": 1.9971,
+      "step": 293495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000146695958503567,
+      "loss": 2.1562,
+      "step": 293500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001466943240908889,
+      "loss": 2.1294,
+      "step": 293505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001466926896622592,
+      "loss": 2.0116,
+      "step": 293510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014669105521767845,
+      "loss": 2.2365,
+      "step": 293515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014668942075714715,
+      "loss": 1.8543,
+      "step": 293520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014668778628066593,
+      "loss": 2.1305,
+      "step": 293525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014668615178823529,
+      "loss": 2.197,
+      "step": 293530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014668451727985578,
+      "loss": 2.2201,
+      "step": 293535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014668288275552807,
+      "loss": 2.183,
+      "step": 293540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014668124821525262,
+      "loss": 2.1202,
+      "step": 293545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014667961365902999,
+      "loss": 2.0081,
+      "step": 293550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014667797908686077,
+      "loss": 2.0046,
+      "step": 293555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014667634449874547,
+      "loss": 1.9418,
+      "step": 293560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014667470989468472,
+      "loss": 2.0383,
+      "step": 293565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014667307527467904,
+      "loss": 1.93,
+      "step": 293570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00014667144063872897,
+      "loss": 2.093,
+      "step": 293575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014666980598683512,
+      "loss": 2.0712,
+      "step": 293580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000146668171318998,
+      "loss": 2.0494,
+      "step": 293585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.0001466665366352182,
+      "loss": 2.1669,
+      "step": 293590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014666490193549625,
+      "loss": 2.0158,
+      "step": 293595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014666326721983272,
+      "loss": 2.023,
+      "step": 293600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014666163248822818,
+      "loss": 2.3136,
+      "step": 293605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014665999774068317,
+      "loss": 1.9693,
+      "step": 293610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014665836297719827,
+      "loss": 2.2001,
+      "step": 293615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014665672819777403,
+      "loss": 2.1328,
+      "step": 293620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014665509340241098,
+      "loss": 2.1356,
+      "step": 293625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014665345859110975,
+      "loss": 2.1182,
+      "step": 293630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00014665182376387084,
+      "loss": 2.0475,
+      "step": 293635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014665018892069478,
+      "loss": 2.0784,
+      "step": 293640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001466485540615822,
+      "loss": 2.0731,
+      "step": 293645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014664691918653361,
+      "loss": 2.1437,
+      "step": 293650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014664528429554963,
+      "loss": 1.9879,
+      "step": 293655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014664364938863074,
+      "loss": 1.9458,
+      "step": 293660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014664201446577753,
+      "loss": 2.1187,
+      "step": 293665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014664037952699056,
+      "loss": 2.1028,
+      "step": 293670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014663874457227042,
+      "loss": 1.9527,
+      "step": 293675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001466371096016176,
+      "loss": 1.9897,
+      "step": 293680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014663547461503274,
+      "loss": 1.8836,
+      "step": 293685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014663383961251633,
+      "loss": 1.9516,
+      "step": 293690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014663220459406896,
+      "loss": 1.9457,
+      "step": 293695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014663056955969116,
+      "loss": 2.04,
+      "step": 293700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014662893450938355,
+      "loss": 1.74,
+      "step": 293705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014662729944314663,
+      "loss": 2.0676,
+      "step": 293710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014662566436098097,
+      "loss": 2.1303,
+      "step": 293715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014662402926288715,
+      "loss": 2.1791,
+      "step": 293720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001466223941488657,
+      "loss": 1.8714,
+      "step": 293725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014662075901891722,
+      "loss": 2.0898,
+      "step": 293730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014661912387304224,
+      "loss": 2.0547,
+      "step": 293735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014661748871124133,
+      "loss": 2.1153,
+      "step": 293740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000146615853533515,
+      "loss": 2.1577,
+      "step": 293745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014661421833986392,
+      "loss": 1.9827,
+      "step": 293750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001466125831302885,
+      "loss": 2.1311,
+      "step": 293755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014661094790478942,
+      "loss": 1.9198,
+      "step": 293760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014660931266336717,
+      "loss": 1.9705,
+      "step": 293765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014660767740602238,
+      "loss": 2.134,
+      "step": 293770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001466060421327555,
+      "loss": 2.2236,
+      "step": 293775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014660440684356718,
+      "loss": 2.1618,
+      "step": 293780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014660277153845795,
+      "loss": 1.9043,
+      "step": 293785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001466011362174284,
+      "loss": 1.954,
+      "step": 293790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014659950088047902,
+      "loss": 2.0366,
+      "step": 293795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014659786552761038,
+      "loss": 1.9641,
+      "step": 293800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001465962301588231,
+      "loss": 2.149,
+      "step": 293805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014659459477411773,
+      "loss": 2.221,
+      "step": 293810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014659295937349476,
+      "loss": 2.1186,
+      "step": 293815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014659132395695477,
+      "loss": 1.9141,
+      "step": 293820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014658968852449838,
+      "loss": 1.9637,
+      "step": 293825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014658805307612605,
+      "loss": 1.9468,
+      "step": 293830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014658641761183848,
+      "loss": 2.1122,
+      "step": 293835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001465847821316361,
+      "loss": 1.9548,
+      "step": 293840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014658314663551953,
+      "loss": 2.0485,
+      "step": 293845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001465815111234893,
+      "loss": 2.1687,
+      "step": 293850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014657987559554597,
+      "loss": 1.942,
+      "step": 293855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014657824005169013,
+      "loss": 1.9962,
+      "step": 293860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001465766044919223,
+      "loss": 2.1195,
+      "step": 293865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014657496891624307,
+      "loss": 2.0047,
+      "step": 293870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.000146573333324653,
+      "loss": 2.0234,
+      "step": 293875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001465716977171526,
+      "loss": 2.0866,
+      "step": 293880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014657006209374248,
+      "loss": 2.1241,
+      "step": 293885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001465684264544232,
+      "loss": 2.0007,
+      "step": 293890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014656679079919527,
+      "loss": 2.2066,
+      "step": 293895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001465651551280593,
+      "loss": 2.1885,
+      "step": 293900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014656351944101583,
+      "loss": 1.9781,
+      "step": 293905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014656188373806541,
+      "loss": 2.0048,
+      "step": 293910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001465602480192086,
+      "loss": 2.1176,
+      "step": 293915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014655861228444597,
+      "loss": 2.0549,
+      "step": 293920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001465569765337781,
+      "loss": 1.9916,
+      "step": 293925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001465553407672055,
+      "loss": 2.1407,
+      "step": 293930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014655370498472872,
+      "loss": 2.1061,
+      "step": 293935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014655206918634837,
+      "loss": 2.0303,
+      "step": 293940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000146550433372065,
+      "loss": 1.9592,
+      "step": 293945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014654879754187917,
+      "loss": 2.288,
+      "step": 293950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001465471616957914,
+      "loss": 1.98,
+      "step": 293955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014654552583380227,
+      "loss": 2.0916,
+      "step": 293960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014654388995591236,
+      "loss": 2.2368,
+      "step": 293965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014654225406212225,
+      "loss": 1.9486,
+      "step": 293970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014654061815243244,
+      "loss": 1.9405,
+      "step": 293975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001465389822268435,
+      "loss": 2.0267,
+      "step": 293980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014653734628535596,
+      "loss": 2.052,
+      "step": 293985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014653571032797045,
+      "loss": 2.027,
+      "step": 293990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001465340743546875,
+      "loss": 2.0229,
+      "step": 293995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014653243836550768,
+      "loss": 2.0594,
+      "step": 294000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014653080236043154,
+      "loss": 2.1758,
+      "step": 294005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001465291663394596,
+      "loss": 2.3119,
+      "step": 294010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014652753030259248,
+      "loss": 1.8493,
+      "step": 294015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001465258942498307,
+      "loss": 2.0218,
+      "step": 294020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014652425818117483,
+      "loss": 2.3503,
+      "step": 294025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014652262209662544,
+      "loss": 2.1704,
+      "step": 294030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014652098599618306,
+      "loss": 1.9782,
+      "step": 294035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014651934987984828,
+      "loss": 2.1147,
+      "step": 294040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014651771374762164,
+      "loss": 2.0421,
+      "step": 294045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014651607759950372,
+      "loss": 2.024,
+      "step": 294050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014651444143549508,
+      "loss": 2.1233,
+      "step": 294055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014651280525559624,
+      "loss": 2.239,
+      "step": 294060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014651116905980779,
+      "loss": 1.9557,
+      "step": 294065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014650953284813025,
+      "loss": 2.0345,
+      "step": 294070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014650789662056424,
+      "loss": 2.1088,
+      "step": 294075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00014650626037711025,
+      "loss": 2.1816,
+      "step": 294080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014650462411776896,
+      "loss": 2.1371,
+      "step": 294085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014650298784254077,
+      "loss": 1.9498,
+      "step": 294090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014650135155142631,
+      "loss": 2.0981,
+      "step": 294095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001464997152444262,
+      "loss": 1.9344,
+      "step": 294100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014649807892154094,
+      "loss": 2.0352,
+      "step": 294105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014649644258277105,
+      "loss": 2.0764,
+      "step": 294110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014649480622811717,
+      "loss": 2.1361,
+      "step": 294115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001464931698575798,
+      "loss": 1.9639,
+      "step": 294120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001464915334711595,
+      "loss": 2.1715,
+      "step": 294125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001464898970688569,
+      "loss": 2.1763,
+      "step": 294130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001464882606506725,
+      "loss": 1.9713,
+      "step": 294135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014648662421660685,
+      "loss": 2.0337,
+      "step": 294140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014648498776666054,
+      "loss": 1.9942,
+      "step": 294145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014648335130083409,
+      "loss": 2.0347,
+      "step": 294150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014648171481912808,
+      "loss": 2.2812,
+      "step": 294155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001464800783215431,
+      "loss": 2.1866,
+      "step": 294160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014647844180807967,
+      "loss": 1.8475,
+      "step": 294165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014647680527873834,
+      "loss": 2.1903,
+      "step": 294170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014647516873351975,
+      "loss": 2.2304,
+      "step": 294175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014647353217242434,
+      "loss": 1.8911,
+      "step": 294180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014647189559545277,
+      "loss": 1.9637,
+      "step": 294185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014647025900260552,
+      "loss": 2.01,
+      "step": 294190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014646862239388323,
+      "loss": 1.9023,
+      "step": 294195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014646698576928637,
+      "loss": 2.1864,
+      "step": 294200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014646534912881557,
+      "loss": 2.0432,
+      "step": 294205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014646371247247134,
+      "loss": 1.9422,
+      "step": 294210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014646207580025432,
+      "loss": 2.0742,
+      "step": 294215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014646043911216498,
+      "loss": 2.1263,
+      "step": 294220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001464588024082039,
+      "loss": 2.1013,
+      "step": 294225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014645716568837164,
+      "loss": 2.1594,
+      "step": 294230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014645552895266878,
+      "loss": 2.0243,
+      "step": 294235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014645389220109588,
+      "loss": 1.9857,
+      "step": 294240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014645225543365352,
+      "loss": 2.0498,
+      "step": 294245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.65625,
+      "learning_rate": 0.00014645061865034217,
+      "loss": 2.1475,
+      "step": 294250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014644898185116244,
+      "loss": 2.0638,
+      "step": 294255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014644734503611493,
+      "loss": 2.163,
+      "step": 294260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014644570820520017,
+      "loss": 2.0631,
+      "step": 294265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001464440713584187,
+      "loss": 2.0729,
+      "step": 294270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014644243449577111,
+      "loss": 1.9356,
+      "step": 294275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001464407976172579,
+      "loss": 1.8899,
+      "step": 294280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014643916072287972,
+      "loss": 2.1545,
+      "step": 294285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014643752381263706,
+      "loss": 2.1275,
+      "step": 294290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001464358868865305,
+      "loss": 2.1328,
+      "step": 294295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014643424994456057,
+      "loss": 2.1102,
+      "step": 294300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001464326129867279,
+      "loss": 2.0375,
+      "step": 294305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000146430976013033,
+      "loss": 2.1929,
+      "step": 294310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014642933902347642,
+      "loss": 2.1936,
+      "step": 294315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00014642770201805876,
+      "loss": 1.9113,
+      "step": 294320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014642606499678054,
+      "loss": 1.951,
+      "step": 294325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014642442795964234,
+      "loss": 2.3737,
+      "step": 294330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014642279090664468,
+      "loss": 2.1603,
+      "step": 294335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00014642115383778817,
+      "loss": 1.9611,
+      "step": 294340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014641951675307336,
+      "loss": 1.946,
+      "step": 294345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014641787965250083,
+      "loss": 2.261,
+      "step": 294350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014641624253607108,
+      "loss": 2.1597,
+      "step": 294355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014641460540378468,
+      "loss": 2.1301,
+      "step": 294360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014641296825564223,
+      "loss": 2.2058,
+      "step": 294365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001464113310916443,
+      "loss": 2.118,
+      "step": 294370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014640969391179138,
+      "loss": 2.0802,
+      "step": 294375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014640805671608408,
+      "loss": 2.1912,
+      "step": 294380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014640641950452294,
+      "loss": 2.0107,
+      "step": 294385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014640478227710853,
+      "loss": 2.1063,
+      "step": 294390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014640314503384138,
+      "loss": 1.9344,
+      "step": 294395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014640150777472211,
+      "loss": 2.105,
+      "step": 294400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014639987049975124,
+      "loss": 2.0556,
+      "step": 294405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001463982332089293,
+      "loss": 1.9236,
+      "step": 294410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001463965959022569,
+      "loss": 1.9831,
+      "step": 294415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014639495857973457,
+      "loss": 1.9807,
+      "step": 294420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014639332124136294,
+      "loss": 1.9299,
+      "step": 294425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014639168388714247,
+      "loss": 2.0862,
+      "step": 294430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014639004651707373,
+      "loss": 2.0049,
+      "step": 294435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014638840913115735,
+      "loss": 2.0214,
+      "step": 294440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014638677172939383,
+      "loss": 2.0806,
+      "step": 294445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014638513431178374,
+      "loss": 2.0247,
+      "step": 294450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014638349687832766,
+      "loss": 1.9869,
+      "step": 294455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014638185942902613,
+      "loss": 1.9471,
+      "step": 294460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014638022196387972,
+      "loss": 1.9599,
+      "step": 294465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000146378584482889,
+      "loss": 2.1081,
+      "step": 294470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001463769469860545,
+      "loss": 1.9504,
+      "step": 294475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014637530947337677,
+      "loss": 2.1153,
+      "step": 294480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014637367194485643,
+      "loss": 2.0015,
+      "step": 294485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014637203440049396,
+      "loss": 2.0335,
+      "step": 294490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014637039684028998,
+      "loss": 1.9324,
+      "step": 294495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014636875926424508,
+      "loss": 2.0157,
+      "step": 294500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014636712167235972,
+      "loss": 1.9984,
+      "step": 294505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001463654840646345,
+      "loss": 2.1011,
+      "step": 294510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014636384644107006,
+      "loss": 1.946,
+      "step": 294515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014636220880166683,
+      "loss": 2.0832,
+      "step": 294520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001463605711464254,
+      "loss": 2.1882,
+      "step": 294525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014635893347534642,
+      "loss": 1.9025,
+      "step": 294530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014635729578843035,
+      "loss": 1.9689,
+      "step": 294535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014635565808567783,
+      "loss": 2.212,
+      "step": 294540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014635402036708935,
+      "loss": 2.0379,
+      "step": 294545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001463523826326655,
+      "loss": 2.2936,
+      "step": 294550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014635074488240682,
+      "loss": 2.0335,
+      "step": 294555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001463491071163139,
+      "loss": 1.997,
+      "step": 294560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00014634746933438729,
+      "loss": 2.1175,
+      "step": 294565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014634583153662754,
+      "loss": 1.991,
+      "step": 294570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014634419372303517,
+      "loss": 2.1226,
+      "step": 294575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014634255589361086,
+      "loss": 2.0898,
+      "step": 294580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014634091804835504,
+      "loss": 2.0788,
+      "step": 294585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014633928018726832,
+      "loss": 2.2044,
+      "step": 294590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014633764231035127,
+      "loss": 2.0352,
+      "step": 294595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014633600441760447,
+      "loss": 2.1464,
+      "step": 294600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014633436650902842,
+      "loss": 2.0369,
+      "step": 294605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001463327285846237,
+      "loss": 2.1375,
+      "step": 294610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014633109064439092,
+      "loss": 2.1598,
+      "step": 294615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014632945268833056,
+      "loss": 2.0214,
+      "step": 294620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014632781471644323,
+      "loss": 2.0738,
+      "step": 294625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014632617672872948,
+      "loss": 2.1661,
+      "step": 294630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014632453872518987,
+      "loss": 2.0947,
+      "step": 294635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014632290070582497,
+      "loss": 1.9727,
+      "step": 294640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001463212626706353,
+      "loss": 2.2444,
+      "step": 294645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014631962461962147,
+      "loss": 2.2327,
+      "step": 294650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014631798655278402,
+      "loss": 2.1357,
+      "step": 294655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001463163484701235,
+      "loss": 2.218,
+      "step": 294660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014631471037164045,
+      "loss": 2.1605,
+      "step": 294665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014631307225733547,
+      "loss": 2.1369,
+      "step": 294670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014631143412720916,
+      "loss": 2.0144,
+      "step": 294675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014630979598126195,
+      "loss": 2.0591,
+      "step": 294680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014630815781949447,
+      "loss": 1.942,
+      "step": 294685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014630651964190735,
+      "loss": 2.0279,
+      "step": 294690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014630488144850104,
+      "loss": 1.9907,
+      "step": 294695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014630324323927613,
+      "loss": 2.1086,
+      "step": 294700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014630160501423323,
+      "loss": 2.0931,
+      "step": 294705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001462999667733728,
+      "loss": 2.0286,
+      "step": 294710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001462983285166955,
+      "loss": 2.1243,
+      "step": 294715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001462966902442019,
+      "loss": 2.2631,
+      "step": 294720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014629505195589245,
+      "loss": 2.0368,
+      "step": 294725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001462934136517678,
+      "loss": 2.2673,
+      "step": 294730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014629177533182842,
+      "loss": 2.1506,
+      "step": 294735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000146290136996075,
+      "loss": 2.0219,
+      "step": 294740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014628849864450801,
+      "loss": 2.067,
+      "step": 294745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014628686027712805,
+      "loss": 2.2296,
+      "step": 294750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001462852218939356,
+      "loss": 2.0886,
+      "step": 294755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001462835834949313,
+      "loss": 2.0967,
+      "step": 294760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014628194508011573,
+      "loss": 1.9892,
+      "step": 294765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014628030664948937,
+      "loss": 1.8561,
+      "step": 294770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014627866820305282,
+      "loss": 2.0649,
+      "step": 294775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014627702974080666,
+      "loss": 2.2404,
+      "step": 294780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001462753912627514,
+      "loss": 2.0597,
+      "step": 294785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014627375276888763,
+      "loss": 1.9256,
+      "step": 294790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001462721142592159,
+      "loss": 1.9476,
+      "step": 294795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014627047573373682,
+      "loss": 2.0638,
+      "step": 294800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014626883719245088,
+      "loss": 2.0248,
+      "step": 294805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014626719863535868,
+      "loss": 2.0419,
+      "step": 294810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014626556006246076,
+      "loss": 1.8024,
+      "step": 294815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014626392147375764,
+      "loss": 2.0281,
+      "step": 294820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014626228286924997,
+      "loss": 2.1412,
+      "step": 294825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014626064424893827,
+      "loss": 2.0827,
+      "step": 294830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001462590056128231,
+      "loss": 2.2078,
+      "step": 294835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.000146257366960905,
+      "loss": 1.9817,
+      "step": 294840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001462557282931845,
+      "loss": 1.8911,
+      "step": 294845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014625408960966226,
+      "loss": 2.038,
+      "step": 294850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014625245091033878,
+      "loss": 2.2508,
+      "step": 294855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014625081219521462,
+      "loss": 2.0599,
+      "step": 294860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014624917346429034,
+      "loss": 2.0407,
+      "step": 294865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014624753471756646,
+      "loss": 1.936,
+      "step": 294870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014624589595504364,
+      "loss": 2.134,
+      "step": 294875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001462442571767224,
+      "loss": 2.0711,
+      "step": 294880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014624261838260325,
+      "loss": 2.0659,
+      "step": 294885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014624097957268676,
+      "loss": 1.9891,
+      "step": 294890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014623934074697355,
+      "loss": 2.078,
+      "step": 294895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001462377019054641,
+      "loss": 2.1341,
+      "step": 294900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014623606304815906,
+      "loss": 2.2032,
+      "step": 294905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014623442417505892,
+      "loss": 2.1354,
+      "step": 294910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014623278528616427,
+      "loss": 2.1309,
+      "step": 294915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014623114638147566,
+      "loss": 2.0614,
+      "step": 294920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014622950746099364,
+      "loss": 2.0148,
+      "step": 294925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001462278685247188,
+      "loss": 2.0138,
+      "step": 294930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014622622957265167,
+      "loss": 2.0642,
+      "step": 294935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014622459060479283,
+      "loss": 2.0793,
+      "step": 294940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001462229516211428,
+      "loss": 2.1056,
+      "step": 294945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014622131262170222,
+      "loss": 1.8472,
+      "step": 294950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014621967360647158,
+      "loss": 2.0083,
+      "step": 294955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014621803457545146,
+      "loss": 1.8632,
+      "step": 294960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001462163955286424,
+      "loss": 2.0457,
+      "step": 294965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014621475646604503,
+      "loss": 2.018,
+      "step": 294970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014621311738765983,
+      "loss": 2.0102,
+      "step": 294975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001462114782934874,
+      "loss": 1.9433,
+      "step": 294980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014620983918352829,
+      "loss": 2.177,
+      "step": 294985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014620820005778305,
+      "loss": 1.9279,
+      "step": 294990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014620656091625226,
+      "loss": 1.9938,
+      "step": 294995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014620492175893647,
+      "loss": 2.0342,
+      "step": 295000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014620328258583624,
+      "loss": 1.9657,
+      "step": 295005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014620164339695214,
+      "loss": 2.1787,
+      "step": 295010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014620000419228473,
+      "loss": 2.0203,
+      "step": 295015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014619836497183452,
+      "loss": 2.129,
+      "step": 295020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014619672573560211,
+      "loss": 2.1426,
+      "step": 295025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0001461950864835881,
+      "loss": 1.9499,
+      "step": 295030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000146193447215793,
+      "loss": 1.8405,
+      "step": 295035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014619180793221735,
+      "loss": 2.1169,
+      "step": 295040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014619016863286178,
+      "loss": 2.0878,
+      "step": 295045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001461885293177268,
+      "loss": 1.9847,
+      "step": 295050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014618688998681298,
+      "loss": 2.2943,
+      "step": 295055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014618525064012086,
+      "loss": 2.0599,
+      "step": 295060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014618361127765106,
+      "loss": 2.0795,
+      "step": 295065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014618197189940406,
+      "loss": 2.0041,
+      "step": 295070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014618033250538047,
+      "loss": 2.1644,
+      "step": 295075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014617869309558086,
+      "loss": 2.1522,
+      "step": 295080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014617705367000577,
+      "loss": 2.0522,
+      "step": 295085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014617541422865576,
+      "loss": 2.0605,
+      "step": 295090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014617377477153137,
+      "loss": 2.0009,
+      "step": 295095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014617213529863321,
+      "loss": 1.9361,
+      "step": 295100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014617049580996182,
+      "loss": 1.9666,
+      "step": 295105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001461688563055177,
+      "loss": 2.0356,
+      "step": 295110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001461672167853015,
+      "loss": 1.9689,
+      "step": 295115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001461655772493137,
+      "loss": 2.0973,
+      "step": 295120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014616393769755496,
+      "loss": 1.9105,
+      "step": 295125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014616229813002574,
+      "loss": 2.2012,
+      "step": 295130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014616065854672664,
+      "loss": 1.9342,
+      "step": 295135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014615901894765823,
+      "loss": 1.9189,
+      "step": 295140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014615737933282107,
+      "loss": 2.2029,
+      "step": 295145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014615573970221574,
+      "loss": 2.1697,
+      "step": 295150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014615410005584274,
+      "loss": 2.0951,
+      "step": 295155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014615246039370266,
+      "loss": 2.1038,
+      "step": 295160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014615082071579605,
+      "loss": 1.9351,
+      "step": 295165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014614918102212352,
+      "loss": 1.8971,
+      "step": 295170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014614754131268558,
+      "loss": 2.0205,
+      "step": 295175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001461459015874828,
+      "loss": 1.9504,
+      "step": 295180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014614426184651573,
+      "loss": 2.0637,
+      "step": 295185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014614262208978494,
+      "loss": 1.9554,
+      "step": 295190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014614098231729102,
+      "loss": 1.9834,
+      "step": 295195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014613934252903448,
+      "loss": 1.9065,
+      "step": 295200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014613770272501591,
+      "loss": 2.1183,
+      "step": 295205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014613606290523587,
+      "loss": 2.2045,
+      "step": 295210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001461344230696949,
+      "loss": 1.9943,
+      "step": 295215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014613278321839359,
+      "loss": 2.0215,
+      "step": 295220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001461311433513325,
+      "loss": 1.9447,
+      "step": 295225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014612950346851213,
+      "loss": 2.0307,
+      "step": 295230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001461278635699331,
+      "loss": 2.0791,
+      "step": 295235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014612622365559597,
+      "loss": 2.1108,
+      "step": 295240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014612458372550128,
+      "loss": 2.0905,
+      "step": 295245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014612294377964961,
+      "loss": 2.1133,
+      "step": 295250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001461213038180415,
+      "loss": 1.7809,
+      "step": 295255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001461196638406775,
+      "loss": 2.1204,
+      "step": 295260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014611802384755822,
+      "loss": 1.87,
+      "step": 295265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014611638383868415,
+      "loss": 1.8477,
+      "step": 295270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001461147438140559,
+      "loss": 2.1887,
+      "step": 295275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000146113103773674,
+      "loss": 2.2084,
+      "step": 295280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014611146371753904,
+      "loss": 1.9487,
+      "step": 295285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001461098236456516,
+      "loss": 1.881,
+      "step": 295290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014610818355801218,
+      "loss": 2.0552,
+      "step": 295295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014610654345462132,
+      "loss": 1.8144,
+      "step": 295300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001461049033354797,
+      "loss": 2.0449,
+      "step": 295305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001461032632005878,
+      "loss": 1.9471,
+      "step": 295310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014610162304994618,
+      "loss": 2.0001,
+      "step": 295315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001460999828835554,
+      "loss": 2.1622,
+      "step": 295320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014609834270141602,
+      "loss": 2.1826,
+      "step": 295325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014609670250352865,
+      "loss": 1.9573,
+      "step": 295330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014609506228989379,
+      "loss": 2.1025,
+      "step": 295335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000146093422060512,
+      "loss": 2.0974,
+      "step": 295340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014609178181538388,
+      "loss": 2.0447,
+      "step": 295345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014609014155450995,
+      "loss": 2.1272,
+      "step": 295350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001460885012778908,
+      "loss": 2.0973,
+      "step": 295355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014608686098552698,
+      "loss": 2.2806,
+      "step": 295360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014608522067741908,
+      "loss": 1.8242,
+      "step": 295365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014608358035356762,
+      "loss": 2.14,
+      "step": 295370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014608194001397318,
+      "loss": 2.1072,
+      "step": 295375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001460802996586363,
+      "loss": 2.1253,
+      "step": 295380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014607865928755755,
+      "loss": 1.977,
+      "step": 295385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001460770189007375,
+      "loss": 2.1594,
+      "step": 295390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.25,
+      "learning_rate": 0.00014607537849817672,
+      "loss": 2.3058,
+      "step": 295395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014607373807987574,
+      "loss": 2.0789,
+      "step": 295400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014607209764583512,
+      "loss": 2.0859,
+      "step": 295405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014607045719605544,
+      "loss": 2.0071,
+      "step": 295410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001460688167305373,
+      "loss": 2.0234,
+      "step": 295415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014606717624928117,
+      "loss": 1.9266,
+      "step": 295420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014606553575228769,
+      "loss": 2.0769,
+      "step": 295425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014606389523955736,
+      "loss": 2.141,
+      "step": 295430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014606225471109076,
+      "loss": 1.8745,
+      "step": 295435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014606061416688848,
+      "loss": 2.0109,
+      "step": 295440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001460589736069511,
+      "loss": 2.158,
+      "step": 295445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014605733303127907,
+      "loss": 1.9667,
+      "step": 295450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00014605569243987304,
+      "loss": 2.2008,
+      "step": 295455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014605405183273353,
+      "loss": 2.0115,
+      "step": 295460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014605241120986117,
+      "loss": 2.0261,
+      "step": 295465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014605077057125643,
+      "loss": 1.9564,
+      "step": 295470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014604912991691995,
+      "loss": 2.0964,
+      "step": 295475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014604748924685224,
+      "loss": 2.1318,
+      "step": 295480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00014604584856105384,
+      "loss": 1.9976,
+      "step": 295485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00014604420785952537,
+      "loss": 2.0669,
+      "step": 295490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014604256714226735,
+      "loss": 2.2223,
+      "step": 295495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014604092640928037,
+      "loss": 2.3097,
+      "step": 295500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014603928566056496,
+      "loss": 2.0428,
+      "step": 295505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001460376448961217,
+      "loss": 2.1708,
+      "step": 295510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014603600411595115,
+      "loss": 2.0694,
+      "step": 295515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014603436332005386,
+      "loss": 2.1306,
+      "step": 295520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001460327225084304,
+      "loss": 2.2084,
+      "step": 295525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014603108168108136,
+      "loss": 1.9919,
+      "step": 295530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014602944083800723,
+      "loss": 2.1738,
+      "step": 295535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001460277999792086,
+      "loss": 2.1446,
+      "step": 295540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014602615910468605,
+      "loss": 2.1398,
+      "step": 295545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014602451821444013,
+      "loss": 2.1362,
+      "step": 295550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014602287730847139,
+      "loss": 2.0641,
+      "step": 295555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014602123638678043,
+      "loss": 2.0692,
+      "step": 295560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001460195954493678,
+      "loss": 2.0878,
+      "step": 295565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014601795449623397,
+      "loss": 2.1627,
+      "step": 295570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001460163135273796,
+      "loss": 2.0833,
+      "step": 295575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014601467254280525,
+      "loss": 2.0062,
+      "step": 295580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014601303154251143,
+      "loss": 1.8978,
+      "step": 295585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014601139052649873,
+      "loss": 2.1559,
+      "step": 295590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014600974949476769,
+      "loss": 2.1073,
+      "step": 295595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001460081084473189,
+      "loss": 2.0012,
+      "step": 295600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001460064673841529,
+      "loss": 2.1144,
+      "step": 295605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014600482630527025,
+      "loss": 2.0623,
+      "step": 295610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014600318521067152,
+      "loss": 2.1243,
+      "step": 295615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014600154410035726,
+      "loss": 2.1376,
+      "step": 295620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014599990297432808,
+      "loss": 2.1643,
+      "step": 295625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014599826183258447,
+      "loss": 2.0473,
+      "step": 295630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014599662067512703,
+      "loss": 2.047,
+      "step": 295635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014599497950195628,
+      "loss": 2.13,
+      "step": 295640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014599333831307284,
+      "loss": 2.0592,
+      "step": 295645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014599169710847724,
+      "loss": 2.1739,
+      "step": 295650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014599005588817002,
+      "loss": 2.0227,
+      "step": 295655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014598841465215178,
+      "loss": 2.0194,
+      "step": 295660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014598677340042307,
+      "loss": 2.0916,
+      "step": 295665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014598513213298444,
+      "loss": 2.1347,
+      "step": 295670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014598349084983645,
+      "loss": 2.1066,
+      "step": 295675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014598184955097966,
+      "loss": 2.0741,
+      "step": 295680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014598020823641464,
+      "loss": 2.1416,
+      "step": 295685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014597856690614196,
+      "loss": 2.1072,
+      "step": 295690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014597692556016215,
+      "loss": 1.9601,
+      "step": 295695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001459752841984758,
+      "loss": 2.0568,
+      "step": 295700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014597364282108342,
+      "loss": 2.1031,
+      "step": 295705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014597200142798565,
+      "loss": 2.1406,
+      "step": 295710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014597036001918302,
+      "loss": 2.107,
+      "step": 295715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001459687185946761,
+      "loss": 1.9504,
+      "step": 295720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014596707715446537,
+      "loss": 2.0776,
+      "step": 295725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014596543569855147,
+      "loss": 2.0535,
+      "step": 295730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014596379422693496,
+      "loss": 2.1258,
+      "step": 295735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001459621527396164,
+      "loss": 2.0313,
+      "step": 295740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014596051123659632,
+      "loss": 2.0791,
+      "step": 295745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014595886971787526,
+      "loss": 2.0589,
+      "step": 295750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014595722818345385,
+      "loss": 2.0192,
+      "step": 295755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014595558663333264,
+      "loss": 2.0418,
+      "step": 295760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.0001459539450675121,
+      "loss": 2.0791,
+      "step": 295765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014595230348599293,
+      "loss": 2.0435,
+      "step": 295770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001459506618887756,
+      "loss": 2.212,
+      "step": 295775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014594902027586067,
+      "loss": 2.0358,
+      "step": 295780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014594737864724873,
+      "loss": 2.094,
+      "step": 295785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014594573700294034,
+      "loss": 2.1328,
+      "step": 295790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014594409534293606,
+      "loss": 2.0639,
+      "step": 295795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014594245366723643,
+      "loss": 1.9395,
+      "step": 295800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.000145940811975842,
+      "loss": 1.9681,
+      "step": 295805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001459391702687534,
+      "loss": 2.0673,
+      "step": 295810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001459375285459711,
+      "loss": 1.9904,
+      "step": 295815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014593588680749574,
+      "loss": 1.9589,
+      "step": 295820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014593424505332785,
+      "loss": 2.1849,
+      "step": 295825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014593260328346798,
+      "loss": 2.1305,
+      "step": 295830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001459309614979167,
+      "loss": 2.1338,
+      "step": 295835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001459293196966746,
+      "loss": 2.1037,
+      "step": 295840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014592767787974215,
+      "loss": 1.9331,
+      "step": 295845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014592603604712,
+      "loss": 2.0264,
+      "step": 295850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014592439419880873,
+      "loss": 1.9648,
+      "step": 295855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014592275233480878,
+      "loss": 2.1536,
+      "step": 295860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00014592111045512083,
+      "loss": 1.9497,
+      "step": 295865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014591946855974536,
+      "loss": 2.0075,
+      "step": 295870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000145917826648683,
+      "loss": 1.9956,
+      "step": 295875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014591618472193429,
+      "loss": 2.2832,
+      "step": 295880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014591454277949976,
+      "loss": 1.9618,
+      "step": 295885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014591290082137997,
+      "loss": 2.2115,
+      "step": 295890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014591125884757551,
+      "loss": 2.1035,
+      "step": 295895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014590961685808694,
+      "loss": 2.0993,
+      "step": 295900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014590797485291483,
+      "loss": 2.1477,
+      "step": 295905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001459063328320597,
+      "loss": 2.0592,
+      "step": 295910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001459046907955221,
+      "loss": 2.0419,
+      "step": 295915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001459030487433027,
+      "loss": 1.9668,
+      "step": 295920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014590140667540194,
+      "loss": 2.1848,
+      "step": 295925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014589976459182045,
+      "loss": 2.134,
+      "step": 295930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014589812249255873,
+      "loss": 2.0543,
+      "step": 295935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014589648037761742,
+      "loss": 1.8883,
+      "step": 295940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014589483824699703,
+      "loss": 2.1033,
+      "step": 295945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001458931961006981,
+      "loss": 2.0241,
+      "step": 295950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014589155393872124,
+      "loss": 2.2195,
+      "step": 295955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014588991176106702,
+      "loss": 1.7733,
+      "step": 295960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014588826956773594,
+      "loss": 2.2185,
+      "step": 295965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001458866273587286,
+      "loss": 2.1056,
+      "step": 295970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014588498513404556,
+      "loss": 2.0914,
+      "step": 295975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001458833428936874,
+      "loss": 2.033,
+      "step": 295980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014588170063765463,
+      "loss": 1.9549,
+      "step": 295985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014588005836594788,
+      "loss": 1.9232,
+      "step": 295990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001458784160785676,
+      "loss": 2.0805,
+      "step": 295995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014587677377551444,
+      "loss": 1.8755,
+      "step": 296000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014587513145678899,
+      "loss": 1.9354,
+      "step": 296005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001458734891223917,
+      "loss": 2.1416,
+      "step": 296010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014587184677232327,
+      "loss": 2.0616,
+      "step": 296015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014587020440658415,
+      "loss": 2.1385,
+      "step": 296020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001458685620251749,
+      "loss": 2.0017,
+      "step": 296025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014586691962809614,
+      "loss": 1.9339,
+      "step": 296030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014586527721534843,
+      "loss": 2.0235,
+      "step": 296035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001458636347869323,
+      "loss": 2.0439,
+      "step": 296040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001458619923428483,
+      "loss": 2.1259,
+      "step": 296045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.000145860349883097,
+      "loss": 2.0977,
+      "step": 296050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014585870740767902,
+      "loss": 2.1067,
+      "step": 296055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014585706491659486,
+      "loss": 1.9511,
+      "step": 296060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014585542240984508,
+      "loss": 1.9911,
+      "step": 296065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014585377988743028,
+      "loss": 1.8659,
+      "step": 296070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014585213734935094,
+      "loss": 2.025,
+      "step": 296075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014585049479560775,
+      "loss": 2.1924,
+      "step": 296080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.734375,
+      "learning_rate": 0.00014584885222620117,
+      "loss": 2.1188,
+      "step": 296085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001458472096411318,
+      "loss": 2.1117,
+      "step": 296090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00014584556704040015,
+      "loss": 2.1463,
+      "step": 296095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014584392442400687,
+      "loss": 2.0679,
+      "step": 296100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014584228179195246,
+      "loss": 2.224,
+      "step": 296105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014584063914423748,
+      "loss": 2.2083,
+      "step": 296110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014583899648086253,
+      "loss": 2.1581,
+      "step": 296115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014583735380182812,
+      "loss": 1.9882,
+      "step": 296120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014583571110713485,
+      "loss": 2.019,
+      "step": 296125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014583406839678329,
+      "loss": 1.9681,
+      "step": 296130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014583242567077395,
+      "loss": 1.9932,
+      "step": 296135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014583078292910746,
+      "loss": 1.8908,
+      "step": 296140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001458291401717843,
+      "loss": 2.0751,
+      "step": 296145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014582749739880513,
+      "loss": 2.0345,
+      "step": 296150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001458258546101704,
+      "loss": 2.0798,
+      "step": 296155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014582421180588074,
+      "loss": 2.077,
+      "step": 296160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014582256898593672,
+      "loss": 1.9816,
+      "step": 296165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014582092615033888,
+      "loss": 2.0776,
+      "step": 296170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014581928329908777,
+      "loss": 2.1541,
+      "step": 296175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014581764043218395,
+      "loss": 2.1665,
+      "step": 296180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014581599754962798,
+      "loss": 2.0531,
+      "step": 296185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014581435465142045,
+      "loss": 2.1888,
+      "step": 296190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014581271173756192,
+      "loss": 1.8995,
+      "step": 296195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014581106880805295,
+      "loss": 2.0393,
+      "step": 296200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014580942586289405,
+      "loss": 2.1975,
+      "step": 296205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014580778290208582,
+      "loss": 2.0546,
+      "step": 296210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014580613992562887,
+      "loss": 2.0975,
+      "step": 296215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014580449693352368,
+      "loss": 1.9794,
+      "step": 296220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014580285392577083,
+      "loss": 2.1517,
+      "step": 296225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001458012109023709,
+      "loss": 2.1757,
+      "step": 296230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014579956786332448,
+      "loss": 2.2011,
+      "step": 296235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014579792480863205,
+      "loss": 2.1134,
+      "step": 296240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014579628173829422,
+      "loss": 2.3213,
+      "step": 296245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014579463865231157,
+      "loss": 2.0395,
+      "step": 296250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014579299555068464,
+      "loss": 1.9562,
+      "step": 296255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000145791352433414,
+      "loss": 2.0121,
+      "step": 296260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001457897093005002,
+      "loss": 2.1117,
+      "step": 296265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014578806615194378,
+      "loss": 2.1087,
+      "step": 296270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014578642298774535,
+      "loss": 2.1857,
+      "step": 296275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014578477980790544,
+      "loss": 2.0252,
+      "step": 296280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014578313661242463,
+      "loss": 1.9996,
+      "step": 296285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014578149340130345,
+      "loss": 2.0756,
+      "step": 296290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001457798501745425,
+      "loss": 1.9639,
+      "step": 296295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001457782069321423,
+      "loss": 2.1845,
+      "step": 296300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014577656367410345,
+      "loss": 2.1855,
+      "step": 296305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014577492040042652,
+      "loss": 2.0076,
+      "step": 296310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014577327711111203,
+      "loss": 2.1103,
+      "step": 296315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014577163380616052,
+      "loss": 2.1145,
+      "step": 296320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014576999048557262,
+      "loss": 2.0087,
+      "step": 296325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014576834714934888,
+      "loss": 2.1636,
+      "step": 296330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014576670379748982,
+      "loss": 2.0928,
+      "step": 296335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014576506042999603,
+      "loss": 2.1204,
+      "step": 296340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014576341704686805,
+      "loss": 1.8899,
+      "step": 296345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001457617736481065,
+      "loss": 2.0201,
+      "step": 296350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014576013023371185,
+      "loss": 2.021,
+      "step": 296355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014575848680368475,
+      "loss": 2.0405,
+      "step": 296360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014575684335802567,
+      "loss": 1.9987,
+      "step": 296365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014575519989673525,
+      "loss": 2.1186,
+      "step": 296370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014575355641981402,
+      "loss": 2.1357,
+      "step": 296375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014575191292726255,
+      "loss": 2.1946,
+      "step": 296380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001457502694190814,
+      "loss": 2.2856,
+      "step": 296385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014574862589527115,
+      "loss": 1.9553,
+      "step": 296390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.0001457469823558323,
+      "loss": 1.9781,
+      "step": 296395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014574533880076546,
+      "loss": 2.0928,
+      "step": 296400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001457436952300712,
+      "loss": 1.9424,
+      "step": 296405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014574205164375004,
+      "loss": 2.1669,
+      "step": 296410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014574040804180257,
+      "loss": 2.0538,
+      "step": 296415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014573876442422937,
+      "loss": 2.144,
+      "step": 296420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014573712079103096,
+      "loss": 2.0155,
+      "step": 296425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014573547714220793,
+      "loss": 2.088,
+      "step": 296430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014573383347776084,
+      "loss": 2.0076,
+      "step": 296435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001457321897976902,
+      "loss": 2.1598,
+      "step": 296440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001457305461019967,
+      "loss": 2.1877,
+      "step": 296445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001457289023906807,
+      "loss": 2.038,
+      "step": 296450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014572725866374294,
+      "loss": 2.2334,
+      "step": 296455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014572561492118394,
+      "loss": 2.0104,
+      "step": 296460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014572397116300424,
+      "loss": 2.0945,
+      "step": 296465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014572232738920438,
+      "loss": 2.0891,
+      "step": 296470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001457206835997849,
+      "loss": 2.1128,
+      "step": 296475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014571903979474646,
+      "loss": 2.0557,
+      "step": 296480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014571739597408955,
+      "loss": 2.1442,
+      "step": 296485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001457157521378148,
+      "loss": 2.278,
+      "step": 296490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014571410828592265,
+      "loss": 2.0966,
+      "step": 296495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014571246441841376,
+      "loss": 2.0683,
+      "step": 296500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014571082053528863,
+      "loss": 1.9146,
+      "step": 296505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014570917663654791,
+      "loss": 2.0812,
+      "step": 296510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014570753272219206,
+      "loss": 2.0878,
+      "step": 296515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014570588879222172,
+      "loss": 2.0001,
+      "step": 296520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001457042448466374,
+      "loss": 2.2266,
+      "step": 296525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014570260088543968,
+      "loss": 2.1113,
+      "step": 296530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00014570095690862912,
+      "loss": 1.9704,
+      "step": 296535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014569931291620632,
+      "loss": 2.017,
+      "step": 296540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014569766890817177,
+      "loss": 2.1791,
+      "step": 296545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001456960248845261,
+      "loss": 2.0614,
+      "step": 296550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001456943808452698,
+      "loss": 2.0027,
+      "step": 296555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014569273679040348,
+      "loss": 1.8993,
+      "step": 296560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014569109271992772,
+      "loss": 2.0729,
+      "step": 296565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014568944863384303,
+      "loss": 1.728,
+      "step": 296570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014568780453215,
+      "loss": 2.2141,
+      "step": 296575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001456861604148492,
+      "loss": 2.1806,
+      "step": 296580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014568451628194115,
+      "loss": 2.1782,
+      "step": 296585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014568287213342646,
+      "loss": 2.0128,
+      "step": 296590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014568122796930567,
+      "loss": 2.0817,
+      "step": 296595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014567958378957933,
+      "loss": 1.9612,
+      "step": 296600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014567793959424804,
+      "loss": 2.1273,
+      "step": 296605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014567629538331234,
+      "loss": 2.0434,
+      "step": 296610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014567465115677278,
+      "loss": 1.8936,
+      "step": 296615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014567300691462992,
+      "loss": 1.8743,
+      "step": 296620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014567136265688434,
+      "loss": 2.1582,
+      "step": 296625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014566971838353657,
+      "loss": 2.0617,
+      "step": 296630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014566807409458724,
+      "loss": 1.9436,
+      "step": 296635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014566642979003684,
+      "loss": 2.0815,
+      "step": 296640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014566478546988599,
+      "loss": 2.0467,
+      "step": 296645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001456631411341352,
+      "loss": 2.0819,
+      "step": 296650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014566149678278505,
+      "loss": 1.9648,
+      "step": 296655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014565985241583607,
+      "loss": 2.1248,
+      "step": 296660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001456582080332889,
+      "loss": 2.1674,
+      "step": 296665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014565656363514407,
+      "loss": 2.0659,
+      "step": 296670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014565491922140212,
+      "loss": 1.9888,
+      "step": 296675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001456532747920636,
+      "loss": 2.087,
+      "step": 296680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001456516303471291,
+      "loss": 2.1308,
+      "step": 296685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014564998588659917,
+      "loss": 1.9874,
+      "step": 296690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001456483414104744,
+      "loss": 2.14,
+      "step": 296695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001456466969187553,
+      "loss": 2.2366,
+      "step": 296700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014564505241144247,
+      "loss": 2.0297,
+      "step": 296705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014564340788853647,
+      "loss": 2.1429,
+      "step": 296710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014564176335003784,
+      "loss": 2.0403,
+      "step": 296715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001456401187959472,
+      "loss": 2.0131,
+      "step": 296720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014563847422626502,
+      "loss": 2.0994,
+      "step": 296725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014563682964099193,
+      "loss": 2.1755,
+      "step": 296730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014563518504012845,
+      "loss": 2.0147,
+      "step": 296735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001456335404236752,
+      "loss": 2.0981,
+      "step": 296740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014563189579163267,
+      "loss": 2.0178,
+      "step": 296745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014563025114400148,
+      "loss": 1.9577,
+      "step": 296750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014562860648078215,
+      "loss": 2.143,
+      "step": 296755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001456269618019753,
+      "loss": 2.2327,
+      "step": 296760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001456253171075814,
+      "loss": 2.1347,
+      "step": 296765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001456236723976011,
+      "loss": 2.1135,
+      "step": 296770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001456220276720349,
+      "loss": 2.1451,
+      "step": 296775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001456203829308834,
+      "loss": 1.9833,
+      "step": 296780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014561873817414716,
+      "loss": 2.1749,
+      "step": 296785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014561709340182673,
+      "loss": 2.1615,
+      "step": 296790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014561544861392266,
+      "loss": 1.9422,
+      "step": 296795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001456138038104355,
+      "loss": 2.0258,
+      "step": 296800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014561215899136588,
+      "loss": 2.0311,
+      "step": 296805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014561051415671434,
+      "loss": 2.0369,
+      "step": 296810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014560886930648138,
+      "loss": 2.1256,
+      "step": 296815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014560722444066762,
+      "loss": 2.1245,
+      "step": 296820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014560557955927358,
+      "loss": 2.0654,
+      "step": 296825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014560393466229988,
+      "loss": 2.0461,
+      "step": 296830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000145602289749747,
+      "loss": 2.2668,
+      "step": 296835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001456006448216156,
+      "loss": 2.0072,
+      "step": 296840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001455989998779062,
+      "loss": 2.1399,
+      "step": 296845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001455973549186193,
+      "loss": 2.2191,
+      "step": 296850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014559570994375558,
+      "loss": 2.1066,
+      "step": 296855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001455940649533155,
+      "loss": 1.958,
+      "step": 296860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014559241994729967,
+      "loss": 1.9723,
+      "step": 296865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014559077492570864,
+      "loss": 1.9401,
+      "step": 296870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000145589129888543,
+      "loss": 1.9823,
+      "step": 296875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014558748483580323,
+      "loss": 2.1177,
+      "step": 296880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014558583976748997,
+      "loss": 2.1002,
+      "step": 296885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014558419468360375,
+      "loss": 2.2544,
+      "step": 296890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014558254958414518,
+      "loss": 2.1169,
+      "step": 296895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014558090446911478,
+      "loss": 2.0655,
+      "step": 296900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014557925933851308,
+      "loss": 2.0452,
+      "step": 296905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001455776141923407,
+      "loss": 2.212,
+      "step": 296910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014557596903059817,
+      "loss": 2.1596,
+      "step": 296915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001455743238532861,
+      "loss": 2.058,
+      "step": 296920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.000145572678660405,
+      "loss": 2.0416,
+      "step": 296925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014557103345195542,
+      "loss": 2.1068,
+      "step": 296930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014556938822793796,
+      "loss": 1.9763,
+      "step": 296935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014556774298835317,
+      "loss": 2.0935,
+      "step": 296940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014556609773320163,
+      "loss": 2.24,
+      "step": 296945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014556445246248387,
+      "loss": 2.0368,
+      "step": 296950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014556280717620049,
+      "loss": 2.1727,
+      "step": 296955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014556116187435195,
+      "loss": 2.0352,
+      "step": 296960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014555951655693898,
+      "loss": 2.1017,
+      "step": 296965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014555787122396202,
+      "loss": 2.0211,
+      "step": 296970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014555622587542167,
+      "loss": 1.9721,
+      "step": 296975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014555458051131848,
+      "loss": 1.959,
+      "step": 296980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000145552935131653,
+      "loss": 2.0779,
+      "step": 296985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014555128973642585,
+      "loss": 2.0733,
+      "step": 296990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014554964432563755,
+      "loss": 1.9458,
+      "step": 296995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014554799889928864,
+      "loss": 2.075,
+      "step": 297000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014554635345737972,
+      "loss": 1.814,
+      "step": 297005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014554470799991133,
+      "loss": 2.0921,
+      "step": 297010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014554306252688405,
+      "loss": 1.9911,
+      "step": 297015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014554141703829843,
+      "loss": 2.1012,
+      "step": 297020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014553977153415504,
+      "loss": 2.1295,
+      "step": 297025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014553812601445445,
+      "loss": 2.1306,
+      "step": 297030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001455364804791972,
+      "loss": 1.9968,
+      "step": 297035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014553483492838384,
+      "loss": 2.0565,
+      "step": 297040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000145533189362015,
+      "loss": 2.1325,
+      "step": 297045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014553154378009115,
+      "loss": 2.1441,
+      "step": 297050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014552989818261293,
+      "loss": 2.0104,
+      "step": 297055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014552825256958086,
+      "loss": 2.0524,
+      "step": 297060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014552660694099552,
+      "loss": 1.8391,
+      "step": 297065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014552496129685745,
+      "loss": 1.9407,
+      "step": 297070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014552331563716722,
+      "loss": 1.9923,
+      "step": 297075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014552166996192547,
+      "loss": 2.1553,
+      "step": 297080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014552002427113263,
+      "loss": 2.0872,
+      "step": 297085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014551837856478935,
+      "loss": 2.0989,
+      "step": 297090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014551673284289612,
+      "loss": 2.3049,
+      "step": 297095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014551508710545358,
+      "loss": 2.149,
+      "step": 297100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014551344135246228,
+      "loss": 2.1366,
+      "step": 297105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014551179558392274,
+      "loss": 2.0279,
+      "step": 297110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014551014979983552,
+      "loss": 1.7897,
+      "step": 297115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014550850400020125,
+      "loss": 1.9284,
+      "step": 297120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014550685818502044,
+      "loss": 2.0764,
+      "step": 297125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014550521235429367,
+      "loss": 2.022,
+      "step": 297130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014550356650802146,
+      "loss": 2.0919,
+      "step": 297135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014550192064620443,
+      "loss": 2.1956,
+      "step": 297140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001455002747688431,
+      "loss": 2.1424,
+      "step": 297145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014549862887593807,
+      "loss": 2.2304,
+      "step": 297150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014549698296748988,
+      "loss": 1.88,
+      "step": 297155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014549533704349908,
+      "loss": 2.1131,
+      "step": 297160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001454936911039663,
+      "loss": 2.0737,
+      "step": 297165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000145492045148892,
+      "loss": 2.0593,
+      "step": 297170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001454903991782768,
+      "loss": 2.0855,
+      "step": 297175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014548875319212126,
+      "loss": 1.9413,
+      "step": 297180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014548710719042594,
+      "loss": 2.1183,
+      "step": 297185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001454854611731914,
+      "loss": 2.143,
+      "step": 297190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014548381514041823,
+      "loss": 1.9071,
+      "step": 297195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001454821690921069,
+      "loss": 2.0472,
+      "step": 297200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014548052302825807,
+      "loss": 2.1333,
+      "step": 297205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001454788769488723,
+      "loss": 2.0493,
+      "step": 297210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014547723085395005,
+      "loss": 2.0004,
+      "step": 297215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014547558474349202,
+      "loss": 2.0019,
+      "step": 297220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014547393861749866,
+      "loss": 1.9532,
+      "step": 297225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014547229247597059,
+      "loss": 2.12,
+      "step": 297230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001454706463189084,
+      "loss": 2.081,
+      "step": 297235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014546900014631255,
+      "loss": 1.9501,
+      "step": 297240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014546735395818368,
+      "loss": 2.0462,
+      "step": 297245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014546570775452236,
+      "loss": 1.8475,
+      "step": 297250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001454640615353291,
+      "loss": 2.0791,
+      "step": 297255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014546241530060455,
+      "loss": 2.0866,
+      "step": 297260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014546076905034915,
+      "loss": 2.0779,
+      "step": 297265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014545912278456354,
+      "loss": 1.897,
+      "step": 297270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001454574765032483,
+      "loss": 2.0447,
+      "step": 297275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014545583020640393,
+      "loss": 2.0786,
+      "step": 297280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014545418389403105,
+      "loss": 1.8882,
+      "step": 297285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001454525375661302,
+      "loss": 1.9337,
+      "step": 297290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001454508912227019,
+      "loss": 2.1144,
+      "step": 297295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001454492448637468,
+      "loss": 2.1493,
+      "step": 297300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014544759848926538,
+      "loss": 2.2015,
+      "step": 297305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014544595209925823,
+      "loss": 1.8696,
+      "step": 297310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014544430569372595,
+      "loss": 2.1112,
+      "step": 297315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014544265927266903,
+      "loss": 2.0737,
+      "step": 297320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014544101283608807,
+      "loss": 2.0064,
+      "step": 297325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001454393663839837,
+      "loss": 1.928,
+      "step": 297330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014543771991635637,
+      "loss": 2.0366,
+      "step": 297335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001454360734332067,
+      "loss": 2.0933,
+      "step": 297340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014543442693453523,
+      "loss": 2.0525,
+      "step": 297345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014543278042034258,
+      "loss": 2.0185,
+      "step": 297350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014543113389062923,
+      "loss": 2.0145,
+      "step": 297355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014542948734539582,
+      "loss": 2.0634,
+      "step": 297360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001454278407846428,
+      "loss": 2.1114,
+      "step": 297365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014542619420837083,
+      "loss": 1.8602,
+      "step": 297370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001454245476165805,
+      "loss": 1.708,
+      "step": 297375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001454229010092723,
+      "loss": 2.1208,
+      "step": 297380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001454212543864468,
+      "loss": 2.0,
+      "step": 297385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014541960774810454,
+      "loss": 2.0592,
+      "step": 297390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001454179610942462,
+      "loss": 2.138,
+      "step": 297395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001454163144248722,
+      "loss": 2.0995,
+      "step": 297400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014541466773998317,
+      "loss": 2.1487,
+      "step": 297405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001454130210395797,
+      "loss": 2.1297,
+      "step": 297410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014541137432366228,
+      "loss": 1.9417,
+      "step": 297415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014540972759223152,
+      "loss": 2.1688,
+      "step": 297420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014540808084528796,
+      "loss": 2.2467,
+      "step": 297425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001454064340828322,
+      "loss": 2.1935,
+      "step": 297430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014540478730486477,
+      "loss": 2.0934,
+      "step": 297435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014540314051138625,
+      "loss": 1.9375,
+      "step": 297440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001454014937023972,
+      "loss": 2.0092,
+      "step": 297445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014539984687789813,
+      "loss": 1.8989,
+      "step": 297450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014539820003788971,
+      "loss": 2.1008,
+      "step": 297455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001453965531823724,
+      "loss": 2.0244,
+      "step": 297460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014539490631134685,
+      "loss": 1.8716,
+      "step": 297465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001453932594248135,
+      "loss": 1.9663,
+      "step": 297470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014539161252277304,
+      "loss": 2.1179,
+      "step": 297475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014538996560522599,
+      "loss": 2.0977,
+      "step": 297480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001453883186721729,
+      "loss": 1.9906,
+      "step": 297485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014538667172361432,
+      "loss": 2.1927,
+      "step": 297490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00014538502475955082,
+      "loss": 2.0448,
+      "step": 297495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014538337777998298,
+      "loss": 1.9577,
+      "step": 297500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014538173078491138,
+      "loss": 2.0173,
+      "step": 297505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014538008377433656,
+      "loss": 1.9548,
+      "step": 297510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014537843674825905,
+      "loss": 2.2366,
+      "step": 297515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014537678970667948,
+      "loss": 1.9846,
+      "step": 297520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014537514264959832,
+      "loss": 2.2309,
+      "step": 297525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014537349557701624,
+      "loss": 2.1135,
+      "step": 297530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014537184848893375,
+      "loss": 2.0057,
+      "step": 297535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.96875,
+      "learning_rate": 0.00014537020138535137,
+      "loss": 2.1721,
+      "step": 297540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014536855426626973,
+      "loss": 1.9848,
+      "step": 297545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014536690713168936,
+      "loss": 2.1488,
+      "step": 297550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014536525998161087,
+      "loss": 2.0696,
+      "step": 297555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014536361281603473,
+      "loss": 2.1289,
+      "step": 297560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001453619656349616,
+      "loss": 2.1702,
+      "step": 297565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.000145360318438392,
+      "loss": 2.2235,
+      "step": 297570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014535867122632645,
+      "loss": 2.1356,
+      "step": 297575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001453570239987656,
+      "loss": 2.139,
+      "step": 297580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014535537675570994,
+      "loss": 2.2983,
+      "step": 297585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014535372949716008,
+      "loss": 2.2345,
+      "step": 297590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014535208222311654,
+      "loss": 2.1794,
+      "step": 297595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001453504349335799,
+      "loss": 2.1574,
+      "step": 297600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014534878762855076,
+      "loss": 2.0627,
+      "step": 297605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014534714030802964,
+      "loss": 2.0695,
+      "step": 297610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014534549297201714,
+      "loss": 2.1055,
+      "step": 297615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014534384562051375,
+      "loss": 1.9986,
+      "step": 297620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014534219825352015,
+      "loss": 2.0309,
+      "step": 297625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014534055087103676,
+      "loss": 2.0732,
+      "step": 297630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014533890347306422,
+      "loss": 2.0554,
+      "step": 297635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014533725605960312,
+      "loss": 2.0152,
+      "step": 297640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014533560863065398,
+      "loss": 1.9792,
+      "step": 297645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014533396118621738,
+      "loss": 1.9572,
+      "step": 297650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014533231372629386,
+      "loss": 2.1941,
+      "step": 297655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014533066625088402,
+      "loss": 2.027,
+      "step": 297660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014532901875998836,
+      "loss": 2.0722,
+      "step": 297665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014532737125360755,
+      "loss": 2.015,
+      "step": 297670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014532572373174207,
+      "loss": 2.0486,
+      "step": 297675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001453240761943925,
+      "loss": 1.9447,
+      "step": 297680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014532242864155934,
+      "loss": 2.1624,
+      "step": 297685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001453207810732433,
+      "loss": 1.8601,
+      "step": 297690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014531913348944482,
+      "loss": 1.9567,
+      "step": 297695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001453174858901645,
+      "loss": 2.0875,
+      "step": 297700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014531583827540292,
+      "loss": 2.1132,
+      "step": 297705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001453141906451606,
+      "loss": 2.0683,
+      "step": 297710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014531254299943815,
+      "loss": 2.0685,
+      "step": 297715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014531089533823612,
+      "loss": 2.0115,
+      "step": 297720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014530924766155507,
+      "loss": 2.0733,
+      "step": 297725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014530759996939555,
+      "loss": 2.1705,
+      "step": 297730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014530595226175814,
+      "loss": 2.082,
+      "step": 297735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014530430453864337,
+      "loss": 1.988,
+      "step": 297740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014530265680005183,
+      "loss": 2.0552,
+      "step": 297745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014530100904598412,
+      "loss": 2.0289,
+      "step": 297750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014529936127644072,
+      "loss": 1.9839,
+      "step": 297755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001452977134914223,
+      "loss": 2.0812,
+      "step": 297760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014529606569092929,
+      "loss": 2.0462,
+      "step": 297765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014529441787496233,
+      "loss": 2.161,
+      "step": 297770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000145292770043522,
+      "loss": 2.0083,
+      "step": 297775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014529112219660882,
+      "loss": 2.0775,
+      "step": 297780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014528947433422343,
+      "loss": 2.0766,
+      "step": 297785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014528782645636625,
+      "loss": 2.0067,
+      "step": 297790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014528617856303798,
+      "loss": 1.8469,
+      "step": 297795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001452845306542391,
+      "loss": 2.0257,
+      "step": 297800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014528288272997022,
+      "loss": 1.9657,
+      "step": 297805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014528123479023191,
+      "loss": 1.9827,
+      "step": 297810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014527958683502467,
+      "loss": 1.9673,
+      "step": 297815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001452779388643491,
+      "loss": 1.9928,
+      "step": 297820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001452762908782058,
+      "loss": 1.9915,
+      "step": 297825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014527464287659528,
+      "loss": 2.0399,
+      "step": 297830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014527299485951812,
+      "loss": 2.0777,
+      "step": 297835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00014527134682697487,
+      "loss": 2.1694,
+      "step": 297840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001452696987789661,
+      "loss": 2.0508,
+      "step": 297845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014526805071549243,
+      "loss": 2.1956,
+      "step": 297850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014526640263655432,
+      "loss": 1.9472,
+      "step": 297855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014526475454215243,
+      "loss": 2.0123,
+      "step": 297860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014526310643228724,
+      "loss": 2.0315,
+      "step": 297865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014526145830695938,
+      "loss": 2.0375,
+      "step": 297870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014525981016616935,
+      "loss": 1.9418,
+      "step": 297875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001452581620099178,
+      "loss": 1.9762,
+      "step": 297880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001452565138382052,
+      "loss": 2.23,
+      "step": 297885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014525486565103218,
+      "loss": 2.1272,
+      "step": 297890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014525321744839927,
+      "loss": 2.1562,
+      "step": 297895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000145251569230307,
+      "loss": 2.1673,
+      "step": 297900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014524992099675604,
+      "loss": 2.3556,
+      "step": 297905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014524827274774684,
+      "loss": 2.0839,
+      "step": 297910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014524662448328003,
+      "loss": 2.2099,
+      "step": 297915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014524497620335615,
+      "loss": 2.054,
+      "step": 297920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014524332790797573,
+      "loss": 2.031,
+      "step": 297925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.00014524167959713942,
+      "loss": 2.1032,
+      "step": 297930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001452400312708477,
+      "loss": 1.9968,
+      "step": 297935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014523838292910122,
+      "loss": 1.9993,
+      "step": 297940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014523673457190043,
+      "loss": 2.3065,
+      "step": 297945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014523508619924592,
+      "loss": 2.027,
+      "step": 297950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014523343781113837,
+      "loss": 2.1574,
+      "step": 297955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001452317894075782,
+      "loss": 1.9224,
+      "step": 297960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014523014098856605,
+      "loss": 1.964,
+      "step": 297965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014522849255410247,
+      "loss": 2.0851,
+      "step": 297970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000145226844104188,
+      "loss": 2.0784,
+      "step": 297975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014522519563882322,
+      "loss": 2.1749,
+      "step": 297980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014522354715800872,
+      "loss": 1.9379,
+      "step": 297985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014522189866174502,
+      "loss": 1.9814,
+      "step": 297990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014522025015003268,
+      "loss": 1.974,
+      "step": 297995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001452186016228723,
+      "loss": 1.9096,
+      "step": 298000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001452169530802644,
+      "loss": 2.1345,
+      "step": 298005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001452153045222096,
+      "loss": 2.1092,
+      "step": 298010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014521365594870846,
+      "loss": 2.1036,
+      "step": 298015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014521200735976146,
+      "loss": 2.1038,
+      "step": 298020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014521035875536927,
+      "loss": 2.0833,
+      "step": 298025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014520871013553235,
+      "loss": 2.001,
+      "step": 298030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014520706150025135,
+      "loss": 1.9279,
+      "step": 298035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014520541284952676,
+      "loss": 2.0771,
+      "step": 298040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001452037641833592,
+      "loss": 1.9995,
+      "step": 298045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014520211550174923,
+      "loss": 2.0737,
+      "step": 298050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001452004668046974,
+      "loss": 2.271,
+      "step": 298055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014519881809220422,
+      "loss": 2.1436,
+      "step": 298060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014519716936427034,
+      "loss": 2.0504,
+      "step": 298065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001451955206208963,
+      "loss": 2.0545,
+      "step": 298070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014519387186208262,
+      "loss": 1.9303,
+      "step": 298075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014519222308782994,
+      "loss": 2.2334,
+      "step": 298080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014519057429813873,
+      "loss": 2.1213,
+      "step": 298085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014518892549300963,
+      "loss": 2.1447,
+      "step": 298090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014518727667244314,
+      "loss": 2.0387,
+      "step": 298095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014518562783643991,
+      "loss": 2.069,
+      "step": 298100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014518397898500042,
+      "loss": 1.9998,
+      "step": 298105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014518233011812527,
+      "loss": 2.0952,
+      "step": 298110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014518068123581498,
+      "loss": 1.9472,
+      "step": 298115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001451790323380702,
+      "loss": 1.954,
+      "step": 298120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014517738342489142,
+      "loss": 2.0335,
+      "step": 298125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014517573449627925,
+      "loss": 1.9098,
+      "step": 298130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001451740855522342,
+      "loss": 1.9916,
+      "step": 298135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014517243659275685,
+      "loss": 2.1127,
+      "step": 298140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001451707876178478,
+      "loss": 1.9996,
+      "step": 298145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001451691386275076,
+      "loss": 1.9985,
+      "step": 298150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001451674896217368,
+      "loss": 1.9309,
+      "step": 298155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014516584060053593,
+      "loss": 1.9276,
+      "step": 298160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014516419156390562,
+      "loss": 2.1501,
+      "step": 298165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001451625425118464,
+      "loss": 1.9115,
+      "step": 298170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014516089344435883,
+      "loss": 2.116,
+      "step": 298175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001451592443614435,
+      "loss": 2.0392,
+      "step": 298180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014515759526310094,
+      "loss": 2.2641,
+      "step": 298185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014515594614933174,
+      "loss": 2.0268,
+      "step": 298190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014515429702013642,
+      "loss": 2.067,
+      "step": 298195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001451526478755156,
+      "loss": 2.0023,
+      "step": 298200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014515099871546979,
+      "loss": 1.9599,
+      "step": 298205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014514934953999958,
+      "loss": 2.0792,
+      "step": 298210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001451477003491056,
+      "loss": 2.0655,
+      "step": 298215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014514605114278824,
+      "loss": 2.1133,
+      "step": 298220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014514440192104825,
+      "loss": 2.0932,
+      "step": 298225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001451427526838861,
+      "loss": 1.8561,
+      "step": 298230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014514110343130237,
+      "loss": 2.0509,
+      "step": 298235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001451394541632976,
+      "loss": 1.9725,
+      "step": 298240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014513780487987235,
+      "loss": 2.0164,
+      "step": 298245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014513615558102726,
+      "loss": 2.1809,
+      "step": 298250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001451345062667628,
+      "loss": 1.9367,
+      "step": 298255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001451328569370796,
+      "loss": 2.1591,
+      "step": 298260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001451312075919782,
+      "loss": 2.1099,
+      "step": 298265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014512955823145916,
+      "loss": 2.0872,
+      "step": 298270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.000145127908855523,
+      "loss": 2.0354,
+      "step": 298275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014512625946417037,
+      "loss": 2.0524,
+      "step": 298280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001451246100574018,
+      "loss": 2.0776,
+      "step": 298285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014512296063521785,
+      "loss": 1.8152,
+      "step": 298290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014512131119761903,
+      "loss": 2.0976,
+      "step": 298295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014511966174460595,
+      "loss": 2.1479,
+      "step": 298300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014511801227617922,
+      "loss": 2.3119,
+      "step": 298305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014511636279233937,
+      "loss": 2.0432,
+      "step": 298310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014511471329308693,
+      "loss": 2.1391,
+      "step": 298315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014511306377842248,
+      "loss": 2.2375,
+      "step": 298320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014511141424834658,
+      "loss": 1.9407,
+      "step": 298325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014510976470285983,
+      "loss": 2.1626,
+      "step": 298330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014510811514196274,
+      "loss": 2.0419,
+      "step": 298335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001451064655656559,
+      "loss": 2.0384,
+      "step": 298340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014510481597393988,
+      "loss": 2.1636,
+      "step": 298345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014510316636681526,
+      "loss": 2.0464,
+      "step": 298350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014510151674428253,
+      "loss": 2.1396,
+      "step": 298355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014509986710634234,
+      "loss": 2.0816,
+      "step": 298360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001450982174529952,
+      "loss": 1.9723,
+      "step": 298365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001450965677842417,
+      "loss": 1.9865,
+      "step": 298370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001450949181000824,
+      "loss": 2.3117,
+      "step": 298375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014509326840051788,
+      "loss": 2.2287,
+      "step": 298380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014509161868554867,
+      "loss": 2.0456,
+      "step": 298385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001450899689551753,
+      "loss": 2.0664,
+      "step": 298390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014508831920939845,
+      "loss": 2.1221,
+      "step": 298395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014508666944821857,
+      "loss": 2.046,
+      "step": 298400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014508501967163624,
+      "loss": 2.1345,
+      "step": 298405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001450833698796521,
+      "loss": 2.0835,
+      "step": 298410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014508172007226663,
+      "loss": 2.3323,
+      "step": 298415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014508007024948045,
+      "loss": 2.0294,
+      "step": 298420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014507842041129408,
+      "loss": 2.0496,
+      "step": 298425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001450767705577081,
+      "loss": 2.2684,
+      "step": 298430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014507512068872308,
+      "loss": 1.886,
+      "step": 298435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014507347080433962,
+      "loss": 2.1476,
+      "step": 298440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014507182090455823,
+      "loss": 2.0429,
+      "step": 298445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014507017098937947,
+      "loss": 2.0224,
+      "step": 298450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014506852105880392,
+      "loss": 2.1045,
+      "step": 298455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014506687111283216,
+      "loss": 2.0155,
+      "step": 298460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014506522115146473,
+      "loss": 2.0606,
+      "step": 298465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001450635711747022,
+      "loss": 2.1065,
+      "step": 298470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014506192118254516,
+      "loss": 2.0901,
+      "step": 298475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014506027117499413,
+      "loss": 2.1304,
+      "step": 298480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014505862115204968,
+      "loss": 1.956,
+      "step": 298485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001450569711137124,
+      "loss": 2.2155,
+      "step": 298490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014505532105998285,
+      "loss": 2.2412,
+      "step": 298495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001450536709908616,
+      "loss": 2.0703,
+      "step": 298500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014505202090634917,
+      "loss": 2.1965,
+      "step": 298505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014505037080644618,
+      "loss": 2.1495,
+      "step": 298510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014504872069115313,
+      "loss": 2.0461,
+      "step": 298515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014504707056047064,
+      "loss": 2.0151,
+      "step": 298520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014504542041439926,
+      "loss": 2.0071,
+      "step": 298525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014504377025293956,
+      "loss": 2.2541,
+      "step": 298530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001450421200760921,
+      "loss": 2.1205,
+      "step": 298535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014504046988385736,
+      "loss": 1.9033,
+      "step": 298540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014503881967623601,
+      "loss": 1.9185,
+      "step": 298545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014503716945322862,
+      "loss": 2.0857,
+      "step": 298550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014503551921483571,
+      "loss": 2.0636,
+      "step": 298555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014503386896105785,
+      "loss": 1.9598,
+      "step": 298560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014503221869189558,
+      "loss": 1.9984,
+      "step": 298565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014503056840734947,
+      "loss": 2.1624,
+      "step": 298570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014502891810742015,
+      "loss": 1.962,
+      "step": 298575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014502726779210812,
+      "loss": 2.0271,
+      "step": 298580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014502561746141394,
+      "loss": 2.1528,
+      "step": 298585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001450239671153382,
+      "loss": 2.1037,
+      "step": 298590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00014502231675388145,
+      "loss": 2.0148,
+      "step": 298595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014502066637704427,
+      "loss": 1.9646,
+      "step": 298600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001450190159848272,
+      "loss": 1.8253,
+      "step": 298605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014501736557723084,
+      "loss": 1.989,
+      "step": 298610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014501571515425574,
+      "loss": 2.0589,
+      "step": 298615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014501406471590245,
+      "loss": 2.1608,
+      "step": 298620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014501241426217152,
+      "loss": 2.1393,
+      "step": 298625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014501076379306354,
+      "loss": 2.1573,
+      "step": 298630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014500911330857905,
+      "loss": 2.0094,
+      "step": 298635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014500746280871865,
+      "loss": 1.9909,
+      "step": 298640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014500581229348288,
+      "loss": 2.1627,
+      "step": 298645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001450041617628723,
+      "loss": 1.9594,
+      "step": 298650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001450025112168875,
+      "loss": 2.2511,
+      "step": 298655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014500086065552902,
+      "loss": 2.1168,
+      "step": 298660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014499921007879741,
+      "loss": 2.0833,
+      "step": 298665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001449975594866933,
+      "loss": 2.0062,
+      "step": 298670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014499590887921717,
+      "loss": 2.0754,
+      "step": 298675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014499425825636962,
+      "loss": 2.1076,
+      "step": 298680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014499260761815123,
+      "loss": 2.0818,
+      "step": 298685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014499095696456258,
+      "loss": 1.9677,
+      "step": 298690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014498930629560416,
+      "loss": 2.0665,
+      "step": 298695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014498765561127657,
+      "loss": 2.1529,
+      "step": 298700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001449860049115804,
+      "loss": 2.0391,
+      "step": 298705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014498435419651616,
+      "loss": 2.0388,
+      "step": 298710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014498270346608453,
+      "loss": 2.0304,
+      "step": 298715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014498105272028593,
+      "loss": 2.1089,
+      "step": 298720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000144979401959121,
+      "loss": 2.1453,
+      "step": 298725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014497775118259027,
+      "loss": 2.1986,
+      "step": 298730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014497610039069435,
+      "loss": 2.1697,
+      "step": 298735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001449744495834338,
+      "loss": 2.021,
+      "step": 298740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014497279876080912,
+      "loss": 1.9615,
+      "step": 298745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014497114792282094,
+      "loss": 2.1615,
+      "step": 298750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001449694970694698,
+      "loss": 2.0744,
+      "step": 298755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014496784620075624,
+      "loss": 2.0175,
+      "step": 298760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014496619531668086,
+      "loss": 1.846,
+      "step": 298765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014496454441724423,
+      "loss": 2.0368,
+      "step": 298770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014496289350244686,
+      "loss": 2.2012,
+      "step": 298775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001449612425722894,
+      "loss": 2.0599,
+      "step": 298780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014495959162677232,
+      "loss": 2.0472,
+      "step": 298785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014495794066589625,
+      "loss": 2.0029,
+      "step": 298790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014495628968966173,
+      "loss": 1.8681,
+      "step": 298795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014495463869806932,
+      "loss": 1.8748,
+      "step": 298800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001449529876911196,
+      "loss": 2.0012,
+      "step": 298805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001449513366688131,
+      "loss": 2.1023,
+      "step": 298810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014494968563115044,
+      "loss": 1.996,
+      "step": 298815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014494803457813213,
+      "loss": 2.1008,
+      "step": 298820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014494638350975877,
+      "loss": 2.0643,
+      "step": 298825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014494473242603092,
+      "loss": 1.9117,
+      "step": 298830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001449430813269491,
+      "loss": 1.9948,
+      "step": 298835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001449414302125139,
+      "loss": 2.1229,
+      "step": 298840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014493977908272595,
+      "loss": 2.0799,
+      "step": 298845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014493812793758574,
+      "loss": 2.2356,
+      "step": 298850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014493647677709382,
+      "loss": 2.0868,
+      "step": 298855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014493482560125082,
+      "loss": 2.1923,
+      "step": 298860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014493317441005723,
+      "loss": 2.2843,
+      "step": 298865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001449315232035137,
+      "loss": 1.9697,
+      "step": 298870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001449298719816207,
+      "loss": 1.9665,
+      "step": 298875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014492822074437886,
+      "loss": 2.1288,
+      "step": 298880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014492656949178874,
+      "loss": 2.0615,
+      "step": 298885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014492491822385085,
+      "loss": 2.1146,
+      "step": 298890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014492326694056582,
+      "loss": 2.1673,
+      "step": 298895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001449216156419342,
+      "loss": 2.0483,
+      "step": 298900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014491996432795652,
+      "loss": 1.9971,
+      "step": 298905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014491831299863337,
+      "loss": 2.0837,
+      "step": 298910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014491666165396532,
+      "loss": 1.9118,
+      "step": 298915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014491501029395288,
+      "loss": 1.9908,
+      "step": 298920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001449133589185967,
+      "loss": 2.1858,
+      "step": 298925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0001449117075278973,
+      "loss": 1.9746,
+      "step": 298930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014491005612185523,
+      "loss": 1.9842,
+      "step": 298935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014490840470047112,
+      "loss": 2.0731,
+      "step": 298940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014490675326374542,
+      "loss": 2.0027,
+      "step": 298945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014490510181167877,
+      "loss": 2.0706,
+      "step": 298950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014490345034427174,
+      "loss": 2.0796,
+      "step": 298955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014490179886152488,
+      "loss": 1.926,
+      "step": 298960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014490014736343878,
+      "loss": 1.9193,
+      "step": 298965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014489849585001393,
+      "loss": 1.9301,
+      "step": 298970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014489684432125093,
+      "loss": 2.1613,
+      "step": 298975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001448951927771504,
+      "loss": 1.9359,
+      "step": 298980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014489354121771281,
+      "loss": 2.1231,
+      "step": 298985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014489188964293882,
+      "loss": 1.9272,
+      "step": 298990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001448902380528289,
+      "loss": 1.9043,
+      "step": 298995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014488858644738368,
+      "loss": 2.1137,
+      "step": 299000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014488693482660368,
+      "loss": 1.9206,
+      "step": 299005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014488528319048954,
+      "loss": 2.1432,
+      "step": 299010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014488363153904177,
+      "loss": 2.0609,
+      "step": 299015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001448819798722609,
+      "loss": 2.0871,
+      "step": 299020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014488032819014754,
+      "loss": 2.0023,
+      "step": 299025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014487867649270225,
+      "loss": 2.0028,
+      "step": 299030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001448770247799256,
+      "loss": 2.0857,
+      "step": 299035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014487537305181815,
+      "loss": 2.0831,
+      "step": 299040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014487372130838044,
+      "loss": 1.9797,
+      "step": 299045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014487206954961305,
+      "loss": 2.0281,
+      "step": 299050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014487041777551656,
+      "loss": 2.0931,
+      "step": 299055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014486876598609151,
+      "loss": 2.0299,
+      "step": 299060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001448671141813385,
+      "loss": 2.078,
+      "step": 299065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014486546236125805,
+      "loss": 2.1877,
+      "step": 299070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014486381052585074,
+      "loss": 1.9425,
+      "step": 299075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014486215867511715,
+      "loss": 2.0423,
+      "step": 299080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014486050680905782,
+      "loss": 2.079,
+      "step": 299085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014485885492767335,
+      "loss": 2.1085,
+      "step": 299090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014485720303096427,
+      "loss": 1.9846,
+      "step": 299095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014485555111893117,
+      "loss": 2.0253,
+      "step": 299100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014485389919157457,
+      "loss": 1.9491,
+      "step": 299105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014485224724889508,
+      "loss": 2.1036,
+      "step": 299110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014485059529089324,
+      "loss": 2.0262,
+      "step": 299115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014484894331756964,
+      "loss": 2.1226,
+      "step": 299120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014484729132892483,
+      "loss": 2.2178,
+      "step": 299125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014484563932495935,
+      "loss": 2.0665,
+      "step": 299130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001448439873056738,
+      "loss": 2.0613,
+      "step": 299135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014484233527106873,
+      "loss": 1.8885,
+      "step": 299140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014484068322114473,
+      "loss": 2.1955,
+      "step": 299145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014483903115590228,
+      "loss": 2.0487,
+      "step": 299150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014483737907534203,
+      "loss": 2.0187,
+      "step": 299155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014483572697946453,
+      "loss": 1.9842,
+      "step": 299160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014483407486827035,
+      "loss": 1.9986,
+      "step": 299165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014483242274176,
+      "loss": 2.0834,
+      "step": 299170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001448307705999341,
+      "loss": 2.1992,
+      "step": 299175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001448291184427932,
+      "loss": 2.0221,
+      "step": 299180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001448274662703378,
+      "loss": 2.1396,
+      "step": 299185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014482581408256862,
+      "loss": 2.0231,
+      "step": 299190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014482416187948607,
+      "loss": 2.1891,
+      "step": 299195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001448225096610908,
+      "loss": 2.0914,
+      "step": 299200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014482085742738334,
+      "loss": 2.2452,
+      "step": 299205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014481920517836426,
+      "loss": 1.9875,
+      "step": 299210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001448175529140341,
+      "loss": 2.1956,
+      "step": 299215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014481590063439347,
+      "loss": 1.9685,
+      "step": 299220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014481424833944292,
+      "loss": 1.9674,
+      "step": 299225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014481259602918304,
+      "loss": 2.036,
+      "step": 299230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014481094370361435,
+      "loss": 2.0804,
+      "step": 299235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001448092913627374,
+      "loss": 1.9574,
+      "step": 299240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014480763900655277,
+      "loss": 2.0445,
+      "step": 299245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014480598663506108,
+      "loss": 1.9908,
+      "step": 299250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014480433424826283,
+      "loss": 2.0168,
+      "step": 299255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014480268184615864,
+      "loss": 2.1347,
+      "step": 299260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.000144801029428749,
+      "loss": 2.0476,
+      "step": 299265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014479937699603455,
+      "loss": 1.8589,
+      "step": 299270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001447977245480158,
+      "loss": 2.0451,
+      "step": 299275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014479607208469334,
+      "loss": 2.0505,
+      "step": 299280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014479441960606773,
+      "loss": 2.145,
+      "step": 299285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014479276711213952,
+      "loss": 1.9151,
+      "step": 299290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001447911146029093,
+      "loss": 2.1794,
+      "step": 299295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014478946207837764,
+      "loss": 2.2047,
+      "step": 299300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014478780953854504,
+      "loss": 2.3086,
+      "step": 299305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014478615698341216,
+      "loss": 1.9557,
+      "step": 299310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001447845044129795,
+      "loss": 1.9644,
+      "step": 299315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001447828518272476,
+      "loss": 2.0081,
+      "step": 299320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014478119922621714,
+      "loss": 2.1125,
+      "step": 299325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014477954660988858,
+      "loss": 2.0288,
+      "step": 299330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014477789397826252,
+      "loss": 1.9972,
+      "step": 299335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014477624133133949,
+      "loss": 1.9067,
+      "step": 299340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014477458866912012,
+      "loss": 2.1423,
+      "step": 299345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001447729359916049,
+      "loss": 2.1295,
+      "step": 299350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014477128329879447,
+      "loss": 2.104,
+      "step": 299355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014476963059068933,
+      "loss": 1.9357,
+      "step": 299360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001447679778672901,
+      "loss": 2.0187,
+      "step": 299365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014476632512859728,
+      "loss": 1.896,
+      "step": 299370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001447646723746115,
+      "loss": 1.9656,
+      "step": 299375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001447630196053333,
+      "loss": 2.1551,
+      "step": 299380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014476136682076322,
+      "loss": 2.0437,
+      "step": 299385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014475971402090184,
+      "loss": 2.3061,
+      "step": 299390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014475806120574976,
+      "loss": 2.0719,
+      "step": 299395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001447564083753075,
+      "loss": 2.1862,
+      "step": 299400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014475475552957562,
+      "loss": 1.9867,
+      "step": 299405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014475310266855472,
+      "loss": 2.1259,
+      "step": 299410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014475144979224537,
+      "loss": 2.0781,
+      "step": 299415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014474979690064812,
+      "loss": 2.0649,
+      "step": 299420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014474814399376346,
+      "loss": 1.9425,
+      "step": 299425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001447464910715921,
+      "loss": 1.9254,
+      "step": 299430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014474483813413445,
+      "loss": 2.1248,
+      "step": 299435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014474318518139124,
+      "loss": 2.0268,
+      "step": 299440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001447415322133629,
+      "loss": 2.0786,
+      "step": 299445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014473987923005002,
+      "loss": 2.0847,
+      "step": 299450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014473822623145322,
+      "loss": 2.1307,
+      "step": 299455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000144736573217573,
+      "loss": 2.3213,
+      "step": 299460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014473492018841,
+      "loss": 2.2443,
+      "step": 299465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001447332671439647,
+      "loss": 2.0808,
+      "step": 299470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014473161408423772,
+      "loss": 1.9113,
+      "step": 299475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014472996100922957,
+      "loss": 2.1397,
+      "step": 299480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014472830791894092,
+      "loss": 1.88,
+      "step": 299485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014472665481337225,
+      "loss": 2.0879,
+      "step": 299490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001447250016925241,
+      "loss": 2.0611,
+      "step": 299495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014472334855639713,
+      "loss": 1.8597,
+      "step": 299500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014472169540499184,
+      "loss": 1.9637,
+      "step": 299505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001447200422383088,
+      "loss": 1.8324,
+      "step": 299510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014471838905634859,
+      "loss": 1.9766,
+      "step": 299515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014471673585911176,
+      "loss": 2.1551,
+      "step": 299520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001447150826465989,
+      "loss": 2.203,
+      "step": 299525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014471342941881052,
+      "loss": 2.0691,
+      "step": 299530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014471177617574724,
+      "loss": 2.0737,
+      "step": 299535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001447101229174096,
+      "loss": 2.1328,
+      "step": 299540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014470846964379817,
+      "loss": 2.1536,
+      "step": 299545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.34375,
+      "learning_rate": 0.00014470681635491355,
+      "loss": 1.9822,
+      "step": 299550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014470516305075626,
+      "loss": 2.0299,
+      "step": 299555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014470350973132685,
+      "loss": 2.1045,
+      "step": 299560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001447018563966259,
+      "loss": 1.9548,
+      "step": 299565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014470020304665403,
+      "loss": 2.1457,
+      "step": 299570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014469854968141173,
+      "loss": 2.1927,
+      "step": 299575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00014469689630089961,
+      "loss": 2.0548,
+      "step": 299580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001446952429051182,
+      "loss": 1.9149,
+      "step": 299585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001446935894940681,
+      "loss": 2.0819,
+      "step": 299590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014469193606774986,
+      "loss": 1.9106,
+      "step": 299595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014469028262616406,
+      "loss": 2.1119,
+      "step": 299600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001446886291693112,
+      "loss": 1.976,
+      "step": 299605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014468697569719195,
+      "loss": 2.03,
+      "step": 299610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014468532220980674,
+      "loss": 1.9508,
+      "step": 299615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001446836687071563,
+      "loss": 1.9776,
+      "step": 299620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014468201518924107,
+      "loss": 2.1618,
+      "step": 299625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014468036165606166,
+      "loss": 2.0153,
+      "step": 299630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014467870810761863,
+      "loss": 2.1351,
+      "step": 299635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014467705454391253,
+      "loss": 2.0981,
+      "step": 299640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014467540096494395,
+      "loss": 2.0544,
+      "step": 299645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014467374737071344,
+      "loss": 1.9488,
+      "step": 299650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014467209376122157,
+      "loss": 2.0518,
+      "step": 299655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001446704401364689,
+      "loss": 2.0799,
+      "step": 299660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000144668786496456,
+      "loss": 2.2202,
+      "step": 299665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001446671328411834,
+      "loss": 1.9648,
+      "step": 299670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014466547917065174,
+      "loss": 1.9482,
+      "step": 299675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014466382548486152,
+      "loss": 2.0776,
+      "step": 299680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014466217178381335,
+      "loss": 2.0627,
+      "step": 299685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014466051806750777,
+      "loss": 2.073,
+      "step": 299690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001446588643359453,
+      "loss": 2.1171,
+      "step": 299695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001446572105891266,
+      "loss": 1.9878,
+      "step": 299700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014465555682705218,
+      "loss": 2.1329,
+      "step": 299705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001446539030497226,
+      "loss": 2.0724,
+      "step": 299710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014465224925713844,
+      "loss": 2.1735,
+      "step": 299715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014465059544930027,
+      "loss": 2.1154,
+      "step": 299720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014464894162620863,
+      "loss": 2.0246,
+      "step": 299725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014464728778786412,
+      "loss": 2.1185,
+      "step": 299730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001446456339342673,
+      "loss": 2.0966,
+      "step": 299735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014464398006541868,
+      "loss": 1.9356,
+      "step": 299740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014464232618131888,
+      "loss": 2.0709,
+      "step": 299745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014464067228196848,
+      "loss": 2.1348,
+      "step": 299750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000144639018367368,
+      "loss": 2.0177,
+      "step": 299755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000144637364437518,
+      "loss": 2.1645,
+      "step": 299760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001446357104924191,
+      "loss": 2.2044,
+      "step": 299765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014463405653207181,
+      "loss": 2.1049,
+      "step": 299770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014463240255647669,
+      "loss": 2.121,
+      "step": 299775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014463074856563442,
+      "loss": 1.8944,
+      "step": 299780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001446290945595454,
+      "loss": 2.1173,
+      "step": 299785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014462744053821028,
+      "loss": 2.029,
+      "step": 299790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014462578650162964,
+      "loss": 2.0914,
+      "step": 299795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014462413244980402,
+      "loss": 1.8696,
+      "step": 299800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.000144622478382734,
+      "loss": 1.913,
+      "step": 299805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001446208243004201,
+      "loss": 1.9559,
+      "step": 299810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014461917020286292,
+      "loss": 2.0538,
+      "step": 299815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014461751609006303,
+      "loss": 2.1384,
+      "step": 299820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000144615861962021,
+      "loss": 2.1423,
+      "step": 299825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014461420781873737,
+      "loss": 2.0934,
+      "step": 299830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014461255366021272,
+      "loss": 2.0339,
+      "step": 299835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001446108994864476,
+      "loss": 2.1615,
+      "step": 299840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001446092452974426,
+      "loss": 2.0511,
+      "step": 299845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014460759109319828,
+      "loss": 1.8502,
+      "step": 299850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014460593687371518,
+      "loss": 2.0598,
+      "step": 299855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014460428263899391,
+      "loss": 2.0241,
+      "step": 299860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014460262838903498,
+      "loss": 2.0507,
+      "step": 299865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014460097412383898,
+      "loss": 1.9429,
+      "step": 299870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014459931984340654,
+      "loss": 2.0354,
+      "step": 299875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001445976655477381,
+      "loss": 2.1628,
+      "step": 299880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014459601123683428,
+      "loss": 1.9994,
+      "step": 299885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014459435691069566,
+      "loss": 1.9021,
+      "step": 299890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014459270256932284,
+      "loss": 2.0342,
+      "step": 299895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014459104821271633,
+      "loss": 2.1744,
+      "step": 299900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001445893938408767,
+      "loss": 2.1009,
+      "step": 299905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001445877394538045,
+      "loss": 2.0459,
+      "step": 299910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014458608505150036,
+      "loss": 2.0491,
+      "step": 299915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001445844306339648,
+      "loss": 2.0571,
+      "step": 299920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014458277620119837,
+      "loss": 2.0791,
+      "step": 299925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001445811217532017,
+      "loss": 2.0729,
+      "step": 299930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014457946728997525,
+      "loss": 2.0889,
+      "step": 299935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014457781281151968,
+      "loss": 2.0496,
+      "step": 299940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001445761583178355,
+      "loss": 2.0774,
+      "step": 299945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001445745038089233,
+      "loss": 2.0697,
+      "step": 299950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014457284928478365,
+      "loss": 1.9413,
+      "step": 299955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014457119474541708,
+      "loss": 2.0892,
+      "step": 299960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014456954019082422,
+      "loss": 2.1211,
+      "step": 299965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014456788562100558,
+      "loss": 2.1096,
+      "step": 299970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014456623103596176,
+      "loss": 2.0023,
+      "step": 299975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014456457643569328,
+      "loss": 2.0739,
+      "step": 299980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014456292182020078,
+      "loss": 2.1075,
+      "step": 299985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001445612671894847,
+      "loss": 2.0263,
+      "step": 299990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014455961254354572,
+      "loss": 1.8767,
+      "step": 299995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001445579578823844,
+      "loss": 2.103,
+      "step": 300000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014455630320600124,
+      "loss": 2.0449,
+      "step": 300005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014455464851439686,
+      "loss": 2.1119,
+      "step": 300010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014455299380757177,
+      "loss": 2.0571,
+      "step": 300015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001445513390855266,
+      "loss": 2.1286,
+      "step": 300020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014454968434826187,
+      "loss": 2.1421,
+      "step": 300025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014454802959577818,
+      "loss": 2.2596,
+      "step": 300030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014454637482807606,
+      "loss": 2.0836,
+      "step": 300035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014454472004515604,
+      "loss": 2.0405,
+      "step": 300040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001445430652470188,
+      "loss": 2.16,
+      "step": 300045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014454141043366485,
+      "loss": 2.1224,
+      "step": 300050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014453975560509472,
+      "loss": 1.9955,
+      "step": 300055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000144538100761309,
+      "loss": 2.1357,
+      "step": 300060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014453644590230825,
+      "loss": 2.1594,
+      "step": 300065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014453479102809305,
+      "loss": 2.133,
+      "step": 300070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014453313613866397,
+      "loss": 2.1819,
+      "step": 300075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014453148123402156,
+      "loss": 1.9979,
+      "step": 300080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014452982631416638,
+      "loss": 1.9718,
+      "step": 300085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.000144528171379099,
+      "loss": 2.099,
+      "step": 300090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014452651642882,
+      "loss": 2.0763,
+      "step": 300095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014452486146332994,
+      "loss": 2.0255,
+      "step": 300100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014452320648262936,
+      "loss": 2.0161,
+      "step": 300105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014452155148671883,
+      "loss": 2.1936,
+      "step": 300110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014451989647559899,
+      "loss": 2.0552,
+      "step": 300115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014451824144927027,
+      "loss": 2.0532,
+      "step": 300120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014451658640773338,
+      "loss": 1.9652,
+      "step": 300125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014451493135098878,
+      "loss": 2.1065,
+      "step": 300130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001445132762790371,
+      "loss": 1.9318,
+      "step": 300135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014451162119187883,
+      "loss": 2.2012,
+      "step": 300140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00014450996608951464,
+      "loss": 1.83,
+      "step": 300145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.000144508310971945,
+      "loss": 2.0622,
+      "step": 300150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014450665583917052,
+      "loss": 2.4116,
+      "step": 300155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014450500069119175,
+      "loss": 2.0308,
+      "step": 300160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014450334552800932,
+      "loss": 2.2322,
+      "step": 300165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001445016903496237,
+      "loss": 2.0447,
+      "step": 300170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014450003515603546,
+      "loss": 1.9991,
+      "step": 300175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014449837994724523,
+      "loss": 1.9798,
+      "step": 300180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014449672472325356,
+      "loss": 2.0868,
+      "step": 300185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.000144495069484061,
+      "loss": 2.0178,
+      "step": 300190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001444934142296681,
+      "loss": 1.8749,
+      "step": 300195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014449175896007543,
+      "loss": 1.9573,
+      "step": 300200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001444901036752836,
+      "loss": 2.0543,
+      "step": 300205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014448844837529316,
+      "loss": 2.0754,
+      "step": 300210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014448679306010463,
+      "loss": 2.0995,
+      "step": 300215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001444851377297186,
+      "loss": 2.1027,
+      "step": 300220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014448348238413564,
+      "loss": 2.174,
+      "step": 300225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014448182702335633,
+      "loss": 2.1119,
+      "step": 300230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014448017164738123,
+      "loss": 2.1033,
+      "step": 300235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014447851625621087,
+      "loss": 2.0106,
+      "step": 300240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014447686084984585,
+      "loss": 2.0457,
+      "step": 300245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014447520542828674,
+      "loss": 2.0358,
+      "step": 300250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014447354999153408,
+      "loss": 1.9872,
+      "step": 300255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014447189453958845,
+      "loss": 1.9258,
+      "step": 300260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014447023907245043,
+      "loss": 2.0827,
+      "step": 300265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014446858359012054,
+      "loss": 2.1595,
+      "step": 300270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001444669280925994,
+      "loss": 1.7923,
+      "step": 300275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014446527257988754,
+      "loss": 2.156,
+      "step": 300280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014446361705198555,
+      "loss": 2.2459,
+      "step": 300285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014446196150889396,
+      "loss": 2.0784,
+      "step": 300290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014446030595061338,
+      "loss": 2.0612,
+      "step": 300295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014445865037714434,
+      "loss": 2.0014,
+      "step": 300300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014445699478848745,
+      "loss": 1.7843,
+      "step": 300305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014445533918464317,
+      "loss": 2.2314,
+      "step": 300310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001444536835656122,
+      "loss": 1.9773,
+      "step": 300315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014445202793139504,
+      "loss": 2.1626,
+      "step": 300320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014445037228199228,
+      "loss": 1.9309,
+      "step": 300325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014444871661740442,
+      "loss": 2.0578,
+      "step": 300330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014444706093763208,
+      "loss": 2.283,
+      "step": 300335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014444540524267583,
+      "loss": 2.0611,
+      "step": 300340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014444374953253623,
+      "loss": 2.022,
+      "step": 300345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014444209380721384,
+      "loss": 2.0422,
+      "step": 300350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014444043806670923,
+      "loss": 2.1417,
+      "step": 300355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014443878231102292,
+      "loss": 1.9965,
+      "step": 300360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014443712654015552,
+      "loss": 2.0575,
+      "step": 300365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014443547075410765,
+      "loss": 2.183,
+      "step": 300370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001444338149528798,
+      "loss": 2.0179,
+      "step": 300375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014443215913647253,
+      "loss": 1.9702,
+      "step": 300380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001444305033048864,
+      "loss": 1.967,
+      "step": 300385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014442884745812206,
+      "loss": 2.056,
+      "step": 300390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014442719159617998,
+      "loss": 1.946,
+      "step": 300395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001444255357190608,
+      "loss": 2.0061,
+      "step": 300400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014442387982676504,
+      "loss": 2.0636,
+      "step": 300405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014442222391929327,
+      "loss": 2.0977,
+      "step": 300410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014442056799664609,
+      "loss": 2.1007,
+      "step": 300415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014441891205882397,
+      "loss": 1.9834,
+      "step": 300420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014441725610582758,
+      "loss": 1.9191,
+      "step": 300425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014441560013765748,
+      "loss": 2.1495,
+      "step": 300430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014441394415431417,
+      "loss": 2.0898,
+      "step": 300435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014441228815579828,
+      "loss": 1.9713,
+      "step": 300440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001444106321421103,
+      "loss": 1.9258,
+      "step": 300445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014440897611325091,
+      "loss": 2.0601,
+      "step": 300450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014440732006922055,
+      "loss": 2.1237,
+      "step": 300455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014440566401001989,
+      "loss": 2.1108,
+      "step": 300460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014440400793564942,
+      "loss": 2.1109,
+      "step": 300465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014440235184610972,
+      "loss": 2.1343,
+      "step": 300470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001444006957414014,
+      "loss": 2.2314,
+      "step": 300475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000144399039621525,
+      "loss": 2.2577,
+      "step": 300480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001443973834864811,
+      "loss": 2.2752,
+      "step": 300485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001443957273362702,
+      "loss": 2.0226,
+      "step": 300490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001443940711708929,
+      "loss": 2.0894,
+      "step": 300495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014439241499034983,
+      "loss": 2.1599,
+      "step": 300500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001443907587946415,
+      "loss": 2.0484,
+      "step": 300505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014438910258376847,
+      "loss": 1.85,
+      "step": 300510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014438744635773135,
+      "loss": 2.1406,
+      "step": 300515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014438579011653063,
+      "loss": 2.0739,
+      "step": 300520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014438413386016693,
+      "loss": 1.9739,
+      "step": 300525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001443824775886408,
+      "loss": 1.9861,
+      "step": 300530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014438082130195282,
+      "loss": 2.007,
+      "step": 300535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014437916500010357,
+      "loss": 1.9348,
+      "step": 300540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014437750868309354,
+      "loss": 1.8939,
+      "step": 300545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001443758523509234,
+      "loss": 2.0668,
+      "step": 300550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014437419600359363,
+      "loss": 2.3721,
+      "step": 300555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001443725396411048,
+      "loss": 1.8579,
+      "step": 300560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014437088326345757,
+      "loss": 2.2251,
+      "step": 300565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014436922687065243,
+      "loss": 2.1226,
+      "step": 300570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014436757046268993,
+      "loss": 1.8587,
+      "step": 300575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014436591403957068,
+      "loss": 2.0911,
+      "step": 300580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014436425760129524,
+      "loss": 2.1358,
+      "step": 300585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014436260114786417,
+      "loss": 2.0314,
+      "step": 300590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014436094467927798,
+      "loss": 2.0811,
+      "step": 300595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014435928819553736,
+      "loss": 2.042,
+      "step": 300600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001443576316966427,
+      "loss": 2.0023,
+      "step": 300605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014435597518259476,
+      "loss": 2.215,
+      "step": 300610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000144354318653394,
+      "loss": 2.0768,
+      "step": 300615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000144352662109041,
+      "loss": 2.266,
+      "step": 300620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014435100554953628,
+      "loss": 2.0912,
+      "step": 300625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00014434934897488048,
+      "loss": 2.3051,
+      "step": 300630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014434769238507415,
+      "loss": 2.2578,
+      "step": 300635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014434603578011786,
+      "loss": 1.9423,
+      "step": 300640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014434437916001214,
+      "loss": 2.0769,
+      "step": 300645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014434272252475755,
+      "loss": 2.0386,
+      "step": 300650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001443410658743547,
+      "loss": 1.8368,
+      "step": 300655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014433940920880413,
+      "loss": 2.0595,
+      "step": 300660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014433775252810645,
+      "loss": 2.0333,
+      "step": 300665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014433609583226216,
+      "loss": 2.0025,
+      "step": 300670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014433443912127185,
+      "loss": 2.0951,
+      "step": 300675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001443327823951361,
+      "loss": 1.9652,
+      "step": 300680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001443311256538555,
+      "loss": 2.2131,
+      "step": 300685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014432946889743052,
+      "loss": 2.0276,
+      "step": 300690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014432781212586184,
+      "loss": 1.9575,
+      "step": 300695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014432615533914995,
+      "loss": 2.005,
+      "step": 300700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014432449853729544,
+      "loss": 1.9209,
+      "step": 300705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001443228417202989,
+      "loss": 2.0186,
+      "step": 300710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014432118488816084,
+      "loss": 1.9122,
+      "step": 300715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001443195280408819,
+      "loss": 1.9964,
+      "step": 300720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014431787117846257,
+      "loss": 1.9687,
+      "step": 300725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014431621430090348,
+      "loss": 2.1291,
+      "step": 300730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014431455740820517,
+      "loss": 2.1433,
+      "step": 300735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014431290050036816,
+      "loss": 2.0103,
+      "step": 300740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014431124357739309,
+      "loss": 2.0354,
+      "step": 300745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001443095866392805,
+      "loss": 2.2623,
+      "step": 300750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014430792968603094,
+      "loss": 2.0442,
+      "step": 300755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000144306272717645,
+      "loss": 2.0482,
+      "step": 300760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001443046157341232,
+      "loss": 2.0319,
+      "step": 300765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014430295873546617,
+      "loss": 2.0152,
+      "step": 300770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014430130172167445,
+      "loss": 2.1434,
+      "step": 300775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014429964469274862,
+      "loss": 2.1734,
+      "step": 300780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014429798764868917,
+      "loss": 2.1768,
+      "step": 300785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014429633058949675,
+      "loss": 2.1774,
+      "step": 300790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001442946735151719,
+      "loss": 2.1078,
+      "step": 300795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014429301642571524,
+      "loss": 1.8448,
+      "step": 300800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014429135932112722,
+      "loss": 2.0236,
+      "step": 300805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014428970220140848,
+      "loss": 1.8759,
+      "step": 300810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014428804506655957,
+      "loss": 2.0707,
+      "step": 300815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014428638791658108,
+      "loss": 1.9986,
+      "step": 300820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014428473075147352,
+      "loss": 2.0626,
+      "step": 300825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014428307357123755,
+      "loss": 2.0861,
+      "step": 300830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014428141637587362,
+      "loss": 2.1786,
+      "step": 300835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001442797591653824,
+      "loss": 1.9346,
+      "step": 300840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014427810193976437,
+      "loss": 2.128,
+      "step": 300845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014427644469902016,
+      "loss": 2.0568,
+      "step": 300850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001442747874431503,
+      "loss": 2.1947,
+      "step": 300855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014427313017215538,
+      "loss": 2.1062,
+      "step": 300860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014427147288603598,
+      "loss": 2.2094,
+      "step": 300865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001442698155847926,
+      "loss": 2.2155,
+      "step": 300870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014426815826842587,
+      "loss": 2.2123,
+      "step": 300875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014426650093693635,
+      "loss": 2.0292,
+      "step": 300880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014426484359032456,
+      "loss": 1.9195,
+      "step": 300885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001442631862285911,
+      "loss": 2.0782,
+      "step": 300890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014426152885173657,
+      "loss": 2.1853,
+      "step": 300895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014425987145976147,
+      "loss": 2.0296,
+      "step": 300900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014425821405266638,
+      "loss": 2.026,
+      "step": 300905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001442565566304519,
+      "loss": 2.0104,
+      "step": 300910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014425489919311858,
+      "loss": 1.9648,
+      "step": 300915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014425324174066698,
+      "loss": 2.1798,
+      "step": 300920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014425158427309768,
+      "loss": 2.1827,
+      "step": 300925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001442499267904112,
+      "loss": 1.9903,
+      "step": 300930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014424826929260818,
+      "loss": 1.8987,
+      "step": 300935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014424661177968914,
+      "loss": 2.0555,
+      "step": 300940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014424495425165466,
+      "loss": 2.087,
+      "step": 300945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014424329670850527,
+      "loss": 1.9888,
+      "step": 300950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001442416391502416,
+      "loss": 1.9933,
+      "step": 300955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014423998157686417,
+      "loss": 2.0851,
+      "step": 300960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014423832398837357,
+      "loss": 2.1281,
+      "step": 300965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014423666638477037,
+      "loss": 2.1318,
+      "step": 300970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001442350087660551,
+      "loss": 2.1259,
+      "step": 300975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014423335113222834,
+      "loss": 2.1176,
+      "step": 300980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014423169348329069,
+      "loss": 2.2405,
+      "step": 300985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001442300358192427,
+      "loss": 2.0744,
+      "step": 300990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014422837814008488,
+      "loss": 2.0184,
+      "step": 300995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014422672044581787,
+      "loss": 2.1289,
+      "step": 301000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001442250627364422,
+      "loss": 2.0439,
+      "step": 301005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014422340501195845,
+      "loss": 2.151,
+      "step": 301010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014422174727236718,
+      "loss": 1.9269,
+      "step": 301015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000144220089517669,
+      "loss": 1.8726,
+      "step": 301020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014421843174786437,
+      "loss": 2.1043,
+      "step": 301025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000144216773962954,
+      "loss": 2.0358,
+      "step": 301030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001442151161629383,
+      "loss": 1.9509,
+      "step": 301035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014421345834781795,
+      "loss": 2.0321,
+      "step": 301040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014421180051759348,
+      "loss": 1.9753,
+      "step": 301045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014421014267226547,
+      "loss": 1.9949,
+      "step": 301050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014420848481183448,
+      "loss": 2.1222,
+      "step": 301055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000144206826936301,
+      "loss": 2.0445,
+      "step": 301060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001442051690456657,
+      "loss": 2.0871,
+      "step": 301065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014420351113992916,
+      "loss": 1.8495,
+      "step": 301070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014420185321909187,
+      "loss": 2.1453,
+      "step": 301075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014420019528315443,
+      "loss": 2.136,
+      "step": 301080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001441985373321174,
+      "loss": 1.7832,
+      "step": 301085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014419687936598132,
+      "loss": 2.0646,
+      "step": 301090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014419522138474684,
+      "loss": 2.2157,
+      "step": 301095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014419356338841443,
+      "loss": 2.1467,
+      "step": 301100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014419190537698468,
+      "loss": 2.0413,
+      "step": 301105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014419024735045822,
+      "loss": 1.9568,
+      "step": 301110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014418858930883554,
+      "loss": 1.9542,
+      "step": 301115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014418693125211725,
+      "loss": 2.107,
+      "step": 301120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014418527318030392,
+      "loss": 2.2047,
+      "step": 301125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014418361509339604,
+      "loss": 2.2423,
+      "step": 301130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014418195699139429,
+      "loss": 1.9168,
+      "step": 301135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014418029887429916,
+      "loss": 2.345,
+      "step": 301140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014417864074211124,
+      "loss": 2.0382,
+      "step": 301145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001441769825948311,
+      "loss": 2.0158,
+      "step": 301150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001441753244324593,
+      "loss": 2.1389,
+      "step": 301155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001441736662549964,
+      "loss": 2.155,
+      "step": 301160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.000144172008062443,
+      "loss": 2.0454,
+      "step": 301165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014417034985479961,
+      "loss": 2.1115,
+      "step": 301170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014416869163206682,
+      "loss": 2.2007,
+      "step": 301175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014416703339424525,
+      "loss": 2.0752,
+      "step": 301180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001441653751413354,
+      "loss": 2.3633,
+      "step": 301185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014416371687333784,
+      "loss": 1.9864,
+      "step": 301190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014416205859025315,
+      "loss": 2.2019,
+      "step": 301195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014416040029208193,
+      "loss": 1.9476,
+      "step": 301200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001441587419788247,
+      "loss": 2.1225,
+      "step": 301205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014415708365048206,
+      "loss": 2.1395,
+      "step": 301210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014415542530705455,
+      "loss": 1.9568,
+      "step": 301215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014415376694854273,
+      "loss": 2.1086,
+      "step": 301220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001441521085749472,
+      "loss": 2.0689,
+      "step": 301225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001441504501862685,
+      "loss": 2.0082,
+      "step": 301230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001441487917825072,
+      "loss": 2.2367,
+      "step": 301235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014414713336366388,
+      "loss": 1.9865,
+      "step": 301240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001441454749297391,
+      "loss": 1.9583,
+      "step": 301245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001441438164807334,
+      "loss": 2.1798,
+      "step": 301250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014414215801664742,
+      "loss": 2.0835,
+      "step": 301255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014414049953748165,
+      "loss": 1.8509,
+      "step": 301260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001441388410432367,
+      "loss": 2.1083,
+      "step": 301265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001441371825339131,
+      "loss": 2.0361,
+      "step": 301270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014413552400951144,
+      "loss": 2.1772,
+      "step": 301275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001441338654700323,
+      "loss": 2.2128,
+      "step": 301280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014413220691547623,
+      "loss": 1.9457,
+      "step": 301285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014413054834584377,
+      "loss": 2.0555,
+      "step": 301290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014412888976113555,
+      "loss": 1.9185,
+      "step": 301295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014412723116135207,
+      "loss": 2.0425,
+      "step": 301300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014412557254649396,
+      "loss": 2.1298,
+      "step": 301305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014412391391656172,
+      "loss": 2.3018,
+      "step": 301310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014412225527155596,
+      "loss": 1.9313,
+      "step": 301315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014412059661147726,
+      "loss": 2.0149,
+      "step": 301320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014411893793632615,
+      "loss": 1.8797,
+      "step": 301325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001441172792461032,
+      "loss": 1.899,
+      "step": 301330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000144115620540809,
+      "loss": 2.1946,
+      "step": 301335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001441139618204441,
+      "loss": 1.8966,
+      "step": 301340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014411230308500908,
+      "loss": 2.1031,
+      "step": 301345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001441106443345045,
+      "loss": 2.014,
+      "step": 301350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014410898556893088,
+      "loss": 2.0455,
+      "step": 301355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014410732678828887,
+      "loss": 2.0711,
+      "step": 301360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000144105667992579,
+      "loss": 1.912,
+      "step": 301365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014410400918180185,
+      "loss": 2.1989,
+      "step": 301370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014410235035595793,
+      "loss": 1.8502,
+      "step": 301375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014410069151504784,
+      "loss": 2.0739,
+      "step": 301380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001440990326590722,
+      "loss": 2.0024,
+      "step": 301385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014409737378803152,
+      "loss": 2.058,
+      "step": 301390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014409571490192635,
+      "loss": 1.9073,
+      "step": 301395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014409405600075733,
+      "loss": 2.0158,
+      "step": 301400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014409239708452494,
+      "loss": 1.9665,
+      "step": 301405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014409073815322977,
+      "loss": 2.0461,
+      "step": 301410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014408907920687245,
+      "loss": 1.8106,
+      "step": 301415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014408742024545347,
+      "loss": 2.0835,
+      "step": 301420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014408576126897344,
+      "loss": 2.061,
+      "step": 301425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001440841022774329,
+      "loss": 2.1628,
+      "step": 301430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014408244327083247,
+      "loss": 2.0484,
+      "step": 301435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014408078424917265,
+      "loss": 1.9129,
+      "step": 301440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014407912521245404,
+      "loss": 2.1182,
+      "step": 301445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001440774661606772,
+      "loss": 2.1229,
+      "step": 301450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001440758070938427,
+      "loss": 2.0544,
+      "step": 301455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001440741480119511,
+      "loss": 2.1041,
+      "step": 301460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014407248891500296,
+      "loss": 2.0991,
+      "step": 301465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014407082980299885,
+      "loss": 2.081,
+      "step": 301470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001440691706759394,
+      "loss": 2.0357,
+      "step": 301475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014406751153382507,
+      "loss": 2.0606,
+      "step": 301480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014406585237665653,
+      "loss": 2.155,
+      "step": 301485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014406419320443426,
+      "loss": 2.1374,
+      "step": 301490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014406253401715887,
+      "loss": 2.0649,
+      "step": 301495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001440608748148309,
+      "loss": 1.919,
+      "step": 301500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014405921559745098,
+      "loss": 1.9968,
+      "step": 301505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014405755636501957,
+      "loss": 2.1487,
+      "step": 301510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014405589711753732,
+      "loss": 1.8352,
+      "step": 301515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001440542378550048,
+      "loss": 1.9179,
+      "step": 301520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014405257857742258,
+      "loss": 2.0721,
+      "step": 301525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014405091928479114,
+      "loss": 2.0242,
+      "step": 301530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014404925997711114,
+      "loss": 2.1447,
+      "step": 301535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001440476006543831,
+      "loss": 2.2048,
+      "step": 301540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001440459413166076,
+      "loss": 2.2165,
+      "step": 301545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014404428196378524,
+      "loss": 1.7796,
+      "step": 301550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014404262259591651,
+      "loss": 1.9915,
+      "step": 301555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014404096321300204,
+      "loss": 1.7891,
+      "step": 301560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001440393038150424,
+      "loss": 1.9355,
+      "step": 301565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001440376444020381,
+      "loss": 2.0312,
+      "step": 301570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014403598497398975,
+      "loss": 2.264,
+      "step": 301575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014403432553089792,
+      "loss": 2.1406,
+      "step": 301580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014403266607276315,
+      "loss": 1.9937,
+      "step": 301585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014403100659958603,
+      "loss": 1.9978,
+      "step": 301590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014402934711136713,
+      "loss": 1.9976,
+      "step": 301595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000144027687608107,
+      "loss": 2.0678,
+      "step": 301600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001440260280898062,
+      "loss": 2.03,
+      "step": 301605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014402436855646532,
+      "loss": 2.0382,
+      "step": 301610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00014402270900808493,
+      "loss": 2.1582,
+      "step": 301615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001440210494446656,
+      "loss": 2.133,
+      "step": 301620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014401938986620784,
+      "loss": 2.0192,
+      "step": 301625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00014401773027271228,
+      "loss": 2.1008,
+      "step": 301630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014401607066417946,
+      "loss": 2.1108,
+      "step": 301635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014401441104060995,
+      "loss": 2.1706,
+      "step": 301640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014401275140200434,
+      "loss": 1.9741,
+      "step": 301645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014401109174836313,
+      "loss": 1.8733,
+      "step": 301650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014400943207968695,
+      "loss": 1.901,
+      "step": 301655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014400777239597637,
+      "loss": 2.08,
+      "step": 301660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014400611269723193,
+      "loss": 2.0839,
+      "step": 301665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001440044529834542,
+      "loss": 2.187,
+      "step": 301670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014400279325464373,
+      "loss": 2.0271,
+      "step": 301675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014400113351080112,
+      "loss": 2.1577,
+      "step": 301680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014399947375192697,
+      "loss": 2.0907,
+      "step": 301685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014399781397802177,
+      "loss": 2.0273,
+      "step": 301690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001439961541890861,
+      "loss": 2.0307,
+      "step": 301695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014399449438512053,
+      "loss": 2.1877,
+      "step": 301700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014399283456612572,
+      "loss": 1.9484,
+      "step": 301705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014399117473210208,
+      "loss": 2.0127,
+      "step": 301710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001439895148830503,
+      "loss": 2.1457,
+      "step": 301715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001439878550189709,
+      "loss": 2.2387,
+      "step": 301720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014398619513986445,
+      "loss": 1.9363,
+      "step": 301725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001439845352457315,
+      "loss": 2.2575,
+      "step": 301730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014398287533657264,
+      "loss": 2.1217,
+      "step": 301735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014398121541238842,
+      "loss": 1.9685,
+      "step": 301740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014397955547317944,
+      "loss": 2.0363,
+      "step": 301745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014397789551894624,
+      "loss": 2.1828,
+      "step": 301750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014397623554968943,
+      "loss": 2.0133,
+      "step": 301755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001439745755654095,
+      "loss": 2.2465,
+      "step": 301760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014397291556610706,
+      "loss": 2.1341,
+      "step": 301765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014397125555178268,
+      "loss": 2.054,
+      "step": 301770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014396959552243692,
+      "loss": 2.0408,
+      "step": 301775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014396793547807036,
+      "loss": 2.0411,
+      "step": 301780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014396627541868353,
+      "loss": 1.9034,
+      "step": 301785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014396461534427703,
+      "loss": 2.2003,
+      "step": 301790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014396295525485143,
+      "loss": 1.9581,
+      "step": 301795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001439612951504073,
+      "loss": 2.2419,
+      "step": 301800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014395963503094518,
+      "loss": 2.077,
+      "step": 301805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014395797489646564,
+      "loss": 2.0756,
+      "step": 301810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014395631474696928,
+      "loss": 1.9681,
+      "step": 301815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014395465458245664,
+      "loss": 1.9243,
+      "step": 301820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014395299440292828,
+      "loss": 2.0553,
+      "step": 301825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001439513342083848,
+      "loss": 2.144,
+      "step": 301830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014394967399882673,
+      "loss": 1.9469,
+      "step": 301835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014394801377425463,
+      "loss": 1.8691,
+      "step": 301840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014394635353466917,
+      "loss": 1.8837,
+      "step": 301845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001439446932800708,
+      "loss": 1.9405,
+      "step": 301850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001439430330104601,
+      "loss": 2.221,
+      "step": 301855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014394137272583768,
+      "loss": 2.0295,
+      "step": 301860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014393971242620407,
+      "loss": 2.2162,
+      "step": 301865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001439380521115599,
+      "loss": 2.0541,
+      "step": 301870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014393639178190566,
+      "loss": 2.0235,
+      "step": 301875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014393473143724196,
+      "loss": 2.2723,
+      "step": 301880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014393307107756935,
+      "loss": 1.9482,
+      "step": 301885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014393141070288842,
+      "loss": 2.1,
+      "step": 301890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014392975031319972,
+      "loss": 2.1082,
+      "step": 301895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014392808990850383,
+      "loss": 2.0883,
+      "step": 301900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001439264294888013,
+      "loss": 1.7879,
+      "step": 301905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001439247690540927,
+      "loss": 2.066,
+      "step": 301910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014392310860437862,
+      "loss": 1.9178,
+      "step": 301915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014392144813965958,
+      "loss": 2.0361,
+      "step": 301920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014391978765993622,
+      "loss": 1.9713,
+      "step": 301925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.875,
+      "learning_rate": 0.000143918127165209,
+      "loss": 2.1217,
+      "step": 301930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014391646665547862,
+      "loss": 2.1753,
+      "step": 301935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014391480613074555,
+      "loss": 2.0778,
+      "step": 301940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014391314559101037,
+      "loss": 2.022,
+      "step": 301945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014391148503627366,
+      "loss": 2.1964,
+      "step": 301950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000143909824466536,
+      "loss": 2.0532,
+      "step": 301955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014390816388179796,
+      "loss": 2.205,
+      "step": 301960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001439065032820601,
+      "loss": 1.9712,
+      "step": 301965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014390484266732295,
+      "loss": 2.0562,
+      "step": 301970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014390318203758715,
+      "loss": 2.052,
+      "step": 301975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014390152139285323,
+      "loss": 1.9835,
+      "step": 301980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014389986073312173,
+      "loss": 2.1629,
+      "step": 301985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014389820005839322,
+      "loss": 1.9969,
+      "step": 301990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001438965393686683,
+      "loss": 2.1605,
+      "step": 301995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014389487866394758,
+      "loss": 2.1004,
+      "step": 302000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001438932179442315,
+      "loss": 2.0142,
+      "step": 302005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014389155720952073,
+      "loss": 1.7588,
+      "step": 302010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014388989645981582,
+      "loss": 1.8952,
+      "step": 302015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014388823569511732,
+      "loss": 2.0815,
+      "step": 302020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014388657491542578,
+      "loss": 2.1646,
+      "step": 302025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014388491412074182,
+      "loss": 2.2651,
+      "step": 302030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014388325331106598,
+      "loss": 2.0262,
+      "step": 302035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001438815924863988,
+      "loss": 2.0862,
+      "step": 302040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014387993164674087,
+      "loss": 2.0698,
+      "step": 302045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014387827079209278,
+      "loss": 2.0148,
+      "step": 302050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014387660992245507,
+      "loss": 1.9991,
+      "step": 302055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001438749490378283,
+      "loss": 2.0727,
+      "step": 302060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001438732881382131,
+      "loss": 2.042,
+      "step": 302065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014387162722360994,
+      "loss": 2.1082,
+      "step": 302070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014386996629401944,
+      "loss": 1.8999,
+      "step": 302075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014386830534944217,
+      "loss": 2.0707,
+      "step": 302080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001438666443898787,
+      "loss": 2.0118,
+      "step": 302085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014386498341532956,
+      "loss": 2.1941,
+      "step": 302090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014386332242579541,
+      "loss": 2.0455,
+      "step": 302095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001438616614212767,
+      "loss": 2.1688,
+      "step": 302100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014386000040177404,
+      "loss": 2.0871,
+      "step": 302105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014385833936728806,
+      "loss": 2.1593,
+      "step": 302110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014385667831781927,
+      "loss": 2.1521,
+      "step": 302115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014385501725336822,
+      "loss": 2.0375,
+      "step": 302120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001438533561739355,
+      "loss": 2.0265,
+      "step": 302125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014385169507952165,
+      "loss": 1.9954,
+      "step": 302130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001438500339701273,
+      "loss": 2.1545,
+      "step": 302135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014384837284575302,
+      "loss": 2.175,
+      "step": 302140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014384671170639926,
+      "loss": 2.1286,
+      "step": 302145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001438450505520667,
+      "loss": 2.2242,
+      "step": 302150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001438433893827559,
+      "loss": 2.1277,
+      "step": 302155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001438417281984674,
+      "loss": 2.0232,
+      "step": 302160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014384006699920176,
+      "loss": 1.9535,
+      "step": 302165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014383840578495956,
+      "loss": 2.111,
+      "step": 302170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014383674455574136,
+      "loss": 2.2384,
+      "step": 302175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014383508331154775,
+      "loss": 1.9745,
+      "step": 302180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014383342205237924,
+      "loss": 1.9801,
+      "step": 302185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014383176077823647,
+      "loss": 1.9629,
+      "step": 302190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014383009948911997,
+      "loss": 1.9855,
+      "step": 302195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014382843818503029,
+      "loss": 2.054,
+      "step": 302200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014382677686596807,
+      "loss": 2.072,
+      "step": 302205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014382511553193378,
+      "loss": 2.0798,
+      "step": 302210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014382345418292806,
+      "loss": 2.0473,
+      "step": 302215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014382179281895144,
+      "loss": 2.0555,
+      "step": 302220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001438201314400045,
+      "loss": 2.1463,
+      "step": 302225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014381847004608787,
+      "loss": 2.2112,
+      "step": 302230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000143816808637202,
+      "loss": 1.9447,
+      "step": 302235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001438151472133475,
+      "loss": 1.9712,
+      "step": 302240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014381348577452495,
+      "loss": 2.1419,
+      "step": 302245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014381182432073494,
+      "loss": 2.0318,
+      "step": 302250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014381016285197804,
+      "loss": 2.1076,
+      "step": 302255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014380850136825476,
+      "loss": 2.0375,
+      "step": 302260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001438068398695657,
+      "loss": 1.9888,
+      "step": 302265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014380517835591142,
+      "loss": 2.1547,
+      "step": 302270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014380351682729257,
+      "loss": 1.9662,
+      "step": 302275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.28125,
+      "learning_rate": 0.00014380185528370957,
+      "loss": 2.1602,
+      "step": 302280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014380019372516308,
+      "loss": 2.2227,
+      "step": 302285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014379853215165363,
+      "loss": 2.0515,
+      "step": 302290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014379687056318186,
+      "loss": 2.1324,
+      "step": 302295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014379520895974824,
+      "loss": 2.1468,
+      "step": 302300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001437935473413534,
+      "loss": 2.0577,
+      "step": 302305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014379188570799788,
+      "loss": 1.9009,
+      "step": 302310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014379022405968224,
+      "loss": 2.0827,
+      "step": 302315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001437885623964071,
+      "loss": 2.0028,
+      "step": 302320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000143786900718173,
+      "loss": 2.282,
+      "step": 302325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014378523902498048,
+      "loss": 2.0882,
+      "step": 302330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001437835773168301,
+      "loss": 1.8877,
+      "step": 302335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014378191559372253,
+      "loss": 1.9561,
+      "step": 302340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014378025385565818,
+      "loss": 1.8603,
+      "step": 302345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014377859210263776,
+      "loss": 2.1111,
+      "step": 302350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014377693033466174,
+      "loss": 2.1495,
+      "step": 302355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014377526855173075,
+      "loss": 1.9495,
+      "step": 302360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014377360675384534,
+      "loss": 1.8594,
+      "step": 302365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014377194494100606,
+      "loss": 1.9872,
+      "step": 302370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001437702831132135,
+      "loss": 1.827,
+      "step": 302375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00014376862127046819,
+      "loss": 2.1503,
+      "step": 302380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014376695941277075,
+      "loss": 2.1664,
+      "step": 302385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014376529754012174,
+      "loss": 1.8226,
+      "step": 302390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014376363565252168,
+      "loss": 2.0194,
+      "step": 302395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014376197374997115,
+      "loss": 2.0306,
+      "step": 302400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014376031183247078,
+      "loss": 2.316,
+      "step": 302405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001437586499000211,
+      "loss": 2.0858,
+      "step": 302410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014375698795262263,
+      "loss": 1.9992,
+      "step": 302415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000143755325990276,
+      "loss": 2.1278,
+      "step": 302420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014375366401298174,
+      "loss": 2.0156,
+      "step": 302425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00014375200202074043,
+      "loss": 2.0484,
+      "step": 302430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001437503400135527,
+      "loss": 1.9304,
+      "step": 302435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014374867799141902,
+      "loss": 2.1502,
+      "step": 302440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014374701595434,
+      "loss": 2.1406,
+      "step": 302445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014374535390231617,
+      "loss": 2.0337,
+      "step": 302450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014374369183534823,
+      "loss": 2.0669,
+      "step": 302455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001437420297534366,
+      "loss": 2.1975,
+      "step": 302460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014374036765658189,
+      "loss": 2.0933,
+      "step": 302465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014373870554478465,
+      "loss": 2.0059,
+      "step": 302470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001437370434180455,
+      "loss": 2.1001,
+      "step": 302475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014373538127636498,
+      "loss": 2.0659,
+      "step": 302480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014373371911974368,
+      "loss": 2.2427,
+      "step": 302485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014373205694818213,
+      "loss": 2.2461,
+      "step": 302490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014373039476168093,
+      "loss": 2.1043,
+      "step": 302495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001437287325602406,
+      "loss": 1.8936,
+      "step": 302500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001437270703438618,
+      "loss": 2.1955,
+      "step": 302505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000143725408112545,
+      "loss": 1.9919,
+      "step": 302510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001437237458662908,
+      "loss": 2.1238,
+      "step": 302515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001437220836050998,
+      "loss": 1.8788,
+      "step": 302520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014372042132897254,
+      "loss": 2.0906,
+      "step": 302525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001437187590379096,
+      "loss": 2.1057,
+      "step": 302530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014371709673191148,
+      "loss": 2.2165,
+      "step": 302535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014371543441097888,
+      "loss": 2.1617,
+      "step": 302540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014371377207511224,
+      "loss": 2.1244,
+      "step": 302545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014371210972431223,
+      "loss": 2.0533,
+      "step": 302550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014371044735857934,
+      "loss": 1.9455,
+      "step": 302555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014370878497791417,
+      "loss": 1.96,
+      "step": 302560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001437071225823173,
+      "loss": 2.0425,
+      "step": 302565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014370546017178929,
+      "loss": 1.7775,
+      "step": 302570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014370379774633068,
+      "loss": 2.0507,
+      "step": 302575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014370213530594207,
+      "loss": 2.1747,
+      "step": 302580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000143700472850624,
+      "loss": 1.9506,
+      "step": 302585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014369881038037708,
+      "loss": 2.0291,
+      "step": 302590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014369714789520187,
+      "loss": 2.192,
+      "step": 302595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001436954853950989,
+      "loss": 2.0311,
+      "step": 302600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014369382288006877,
+      "loss": 2.041,
+      "step": 302605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014369216035011202,
+      "loss": 1.9548,
+      "step": 302610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014369049780522922,
+      "loss": 2.0882,
+      "step": 302615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014368883524542097,
+      "loss": 1.8838,
+      "step": 302620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014368717267068784,
+      "loss": 2.2452,
+      "step": 302625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00014368551008103037,
+      "loss": 1.9331,
+      "step": 302630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014368384747644915,
+      "loss": 1.9802,
+      "step": 302635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.0001436821848569447,
+      "loss": 1.9898,
+      "step": 302640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014368052222251765,
+      "loss": 2.1218,
+      "step": 302645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014367885957316853,
+      "loss": 2.0826,
+      "step": 302650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014367719690889792,
+      "loss": 2.0775,
+      "step": 302655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014367553422970638,
+      "loss": 2.043,
+      "step": 302660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014367387153559454,
+      "loss": 1.9526,
+      "step": 302665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014367220882656286,
+      "loss": 1.9605,
+      "step": 302670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.0,
+      "learning_rate": 0.00014367054610261196,
+      "loss": 2.121,
+      "step": 302675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014366888336374243,
+      "loss": 1.9296,
+      "step": 302680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001436672206099548,
+      "loss": 1.9865,
+      "step": 302685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001436655578412497,
+      "loss": 1.9863,
+      "step": 302690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014366389505762758,
+      "loss": 2.0307,
+      "step": 302695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014366223225908913,
+      "loss": 1.9064,
+      "step": 302700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014366056944563487,
+      "loss": 2.0478,
+      "step": 302705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014365890661726535,
+      "loss": 1.9755,
+      "step": 302710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014365724377398114,
+      "loss": 2.0112,
+      "step": 302715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014365558091578286,
+      "loss": 2.0028,
+      "step": 302720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000143653918042671,
+      "loss": 1.9888,
+      "step": 302725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014365225515464624,
+      "loss": 2.0216,
+      "step": 302730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014365059225170901,
+      "loss": 2.1933,
+      "step": 302735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014364892933386,
+      "loss": 2.0808,
+      "step": 302740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014364726640109967,
+      "loss": 1.9521,
+      "step": 302745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001436456034534287,
+      "loss": 2.1219,
+      "step": 302750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014364394049084755,
+      "loss": 1.9927,
+      "step": 302755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014364227751335686,
+      "loss": 1.7838,
+      "step": 302760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014364061452095718,
+      "loss": 1.9626,
+      "step": 302765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014363895151364906,
+      "loss": 1.9932,
+      "step": 302770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001436372884914331,
+      "loss": 2.0343,
+      "step": 302775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014363562545430982,
+      "loss": 2.0765,
+      "step": 302780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014363396240227983,
+      "loss": 1.9189,
+      "step": 302785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001436322993353437,
+      "loss": 2.0661,
+      "step": 302790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014363063625350198,
+      "loss": 2.1437,
+      "step": 302795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014362897315675527,
+      "loss": 1.9171,
+      "step": 302800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014362731004510408,
+      "loss": 2.1012,
+      "step": 302805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014362564691854904,
+      "loss": 1.993,
+      "step": 302810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014362398377709066,
+      "loss": 2.0525,
+      "step": 302815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014362232062072953,
+      "loss": 2.0836,
+      "step": 302820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014362065744946622,
+      "loss": 2.0775,
+      "step": 302825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001436189942633013,
+      "loss": 2.1824,
+      "step": 302830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001436173310622354,
+      "loss": 2.0118,
+      "step": 302835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014361566784626897,
+      "loss": 1.9457,
+      "step": 302840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014361400461540268,
+      "loss": 2.0286,
+      "step": 302845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000143612341369637,
+      "loss": 1.9052,
+      "step": 302850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001436106781089726,
+      "loss": 2.1098,
+      "step": 302855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014360901483340997,
+      "loss": 2.1136,
+      "step": 302860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014360735154294972,
+      "loss": 2.0461,
+      "step": 302865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014360568823759242,
+      "loss": 2.0413,
+      "step": 302870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014360402491733862,
+      "loss": 2.1111,
+      "step": 302875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014360236158218888,
+      "loss": 1.8854,
+      "step": 302880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001436006982321438,
+      "loss": 2.0745,
+      "step": 302885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014359903486720394,
+      "loss": 2.0851,
+      "step": 302890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014359737148736984,
+      "loss": 1.9438,
+      "step": 302895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014359570809264208,
+      "loss": 1.9517,
+      "step": 302900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014359404468302123,
+      "loss": 1.9254,
+      "step": 302905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014359238125850792,
+      "loss": 2.1105,
+      "step": 302910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001435907178191026,
+      "loss": 1.9546,
+      "step": 302915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001435890543648059,
+      "loss": 1.9142,
+      "step": 302920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014358739089561842,
+      "loss": 2.2005,
+      "step": 302925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001435857274115407,
+      "loss": 2.2002,
+      "step": 302930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001435840639125733,
+      "loss": 2.166,
+      "step": 302935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014358240039871677,
+      "loss": 2.0232,
+      "step": 302940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014358073686997173,
+      "loss": 2.0923,
+      "step": 302945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001435790733263387,
+      "loss": 2.0389,
+      "step": 302950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001435774097678183,
+      "loss": 2.1296,
+      "step": 302955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00014357574619441104,
+      "loss": 2.0698,
+      "step": 302960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014357408260611752,
+      "loss": 2.1724,
+      "step": 302965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014357241900293826,
+      "loss": 2.0685,
+      "step": 302970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001435707553848739,
+      "loss": 2.0319,
+      "step": 302975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014356909175192503,
+      "loss": 2.016,
+      "step": 302980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014356742810409215,
+      "loss": 2.0806,
+      "step": 302985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001435657644413758,
+      "loss": 2.0574,
+      "step": 302990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001435641007637766,
+      "loss": 1.9891,
+      "step": 302995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014356243707129516,
+      "loss": 2.0251,
+      "step": 303000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014356077336393196,
+      "loss": 2.1148,
+      "step": 303005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014355910964168763,
+      "loss": 2.2145,
+      "step": 303010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001435574459045627,
+      "loss": 2.0412,
+      "step": 303015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014355578215255777,
+      "loss": 2.1226,
+      "step": 303020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014355411838567342,
+      "loss": 2.1287,
+      "step": 303025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014355245460391015,
+      "loss": 2.0067,
+      "step": 303030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014355079080726862,
+      "loss": 2.0366,
+      "step": 303035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014354912699574926,
+      "loss": 2.1851,
+      "step": 303040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014354746316935283,
+      "loss": 2.0904,
+      "step": 303045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014354579932807977,
+      "loss": 2.2849,
+      "step": 303050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014354413547193064,
+      "loss": 2.1591,
+      "step": 303055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014354247160090606,
+      "loss": 2.1389,
+      "step": 303060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014354080771500658,
+      "loss": 1.918,
+      "step": 303065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001435391438142328,
+      "loss": 2.0214,
+      "step": 303070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001435374798985852,
+      "loss": 2.2916,
+      "step": 303075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014353581596806447,
+      "loss": 2.0588,
+      "step": 303080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014353415202267107,
+      "loss": 2.086,
+      "step": 303085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014353248806240566,
+      "loss": 2.1723,
+      "step": 303090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014353082408726872,
+      "loss": 2.225,
+      "step": 303095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014352916009726088,
+      "loss": 1.9879,
+      "step": 303100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014352749609238268,
+      "loss": 1.9254,
+      "step": 303105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001435258320726347,
+      "loss": 2.1503,
+      "step": 303110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014352416803801755,
+      "loss": 2.1737,
+      "step": 303115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014352250398853167,
+      "loss": 2.1057,
+      "step": 303120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014352083992417774,
+      "loss": 2.2102,
+      "step": 303125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014351917584495635,
+      "loss": 2.102,
+      "step": 303130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000143517511750868,
+      "loss": 1.9824,
+      "step": 303135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014351584764191327,
+      "loss": 2.2433,
+      "step": 303140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014351418351809273,
+      "loss": 2.2658,
+      "step": 303145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014351251937940693,
+      "loss": 1.8642,
+      "step": 303150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001435108552258565,
+      "loss": 1.9965,
+      "step": 303155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014350919105744197,
+      "loss": 2.2188,
+      "step": 303160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014350752687416392,
+      "loss": 1.9619,
+      "step": 303165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014350586267602292,
+      "loss": 2.0202,
+      "step": 303170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014350419846301948,
+      "loss": 2.0207,
+      "step": 303175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014350253423515424,
+      "loss": 2.1142,
+      "step": 303180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014350086999242775,
+      "loss": 2.2183,
+      "step": 303185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001434992057348406,
+      "loss": 2.1468,
+      "step": 303190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014349754146239329,
+      "loss": 2.0229,
+      "step": 303195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001434958771750864,
+      "loss": 1.875,
+      "step": 303200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014349421287292063,
+      "loss": 2.0456,
+      "step": 303205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014349254855589637,
+      "loss": 2.0092,
+      "step": 303210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014349088422401432,
+      "loss": 2.1809,
+      "step": 303215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014348921987727495,
+      "loss": 2.0563,
+      "step": 303220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001434875555156789,
+      "loss": 2.0745,
+      "step": 303225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014348589113922667,
+      "loss": 2.0272,
+      "step": 303230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014348422674791893,
+      "loss": 2.139,
+      "step": 303235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014348256234175617,
+      "loss": 2.0095,
+      "step": 303240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014348089792073896,
+      "loss": 2.2463,
+      "step": 303245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001434792334848679,
+      "loss": 2.0426,
+      "step": 303250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014347756903414354,
+      "loss": 2.1813,
+      "step": 303255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014347590456856646,
+      "loss": 2.1211,
+      "step": 303260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0,
+      "learning_rate": 0.0001434742400881372,
+      "loss": 2.0479,
+      "step": 303265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014347257559285638,
+      "loss": 2.0549,
+      "step": 303270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014347091108272454,
+      "loss": 2.0295,
+      "step": 303275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014346924655774226,
+      "loss": 2.0042,
+      "step": 303280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014346758201791004,
+      "loss": 2.0735,
+      "step": 303285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014346591746322855,
+      "loss": 2.262,
+      "step": 303290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014346425289369833,
+      "loss": 1.9951,
+      "step": 303295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014346258830931992,
+      "loss": 2.1666,
+      "step": 303300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014346092371009389,
+      "loss": 1.9428,
+      "step": 303305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014345925909602078,
+      "loss": 2.0147,
+      "step": 303310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014345759446710124,
+      "loss": 1.9458,
+      "step": 303315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014345592982333583,
+      "loss": 1.8953,
+      "step": 303320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014345426516472506,
+      "loss": 2.0758,
+      "step": 303325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001434526004912695,
+      "loss": 2.1369,
+      "step": 303330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014345093580296975,
+      "loss": 2.124,
+      "step": 303335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001434492710998264,
+      "loss": 1.9712,
+      "step": 303340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014344760638183997,
+      "loss": 2.1404,
+      "step": 303345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014344594164901104,
+      "loss": 1.9894,
+      "step": 303350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001434442769013402,
+      "loss": 1.9768,
+      "step": 303355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014344261213882802,
+      "loss": 2.1925,
+      "step": 303360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014344094736147504,
+      "loss": 1.9375,
+      "step": 303365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014343928256928186,
+      "loss": 2.203,
+      "step": 303370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014343761776224902,
+      "loss": 2.1978,
+      "step": 303375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001434359529403771,
+      "loss": 2.0539,
+      "step": 303380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001434342881036667,
+      "loss": 1.8214,
+      "step": 303385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014343262325211832,
+      "loss": 1.8581,
+      "step": 303390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014343095838573256,
+      "loss": 1.955,
+      "step": 303395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014342929350451004,
+      "loss": 1.8323,
+      "step": 303400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014342762860845127,
+      "loss": 2.1686,
+      "step": 303405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014342596369755684,
+      "loss": 2.0235,
+      "step": 303410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014342429877182728,
+      "loss": 2.0445,
+      "step": 303415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00014342263383126323,
+      "loss": 2.134,
+      "step": 303420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001434209688758652,
+      "loss": 2.1552,
+      "step": 303425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014341930390563378,
+      "loss": 1.9434,
+      "step": 303430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014341763892056957,
+      "loss": 1.7728,
+      "step": 303435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014341597392067307,
+      "loss": 1.9572,
+      "step": 303440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001434143089059449,
+      "loss": 2.1185,
+      "step": 303445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001434126438763856,
+      "loss": 2.2908,
+      "step": 303450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001434109788319958,
+      "loss": 1.9933,
+      "step": 303455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014340931377277599,
+      "loss": 2.2393,
+      "step": 303460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014340764869872675,
+      "loss": 1.9079,
+      "step": 303465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014340598360984867,
+      "loss": 2.1444,
+      "step": 303470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014340431850614236,
+      "loss": 2.0699,
+      "step": 303475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.71875,
+      "learning_rate": 0.00014340265338760833,
+      "loss": 2.118,
+      "step": 303480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014340098825424717,
+      "loss": 2.1737,
+      "step": 303485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014339932310605943,
+      "loss": 1.8828,
+      "step": 303490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001433976579430457,
+      "loss": 2.0485,
+      "step": 303495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014339599276520654,
+      "loss": 2.132,
+      "step": 303500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014339432757254255,
+      "loss": 2.0095,
+      "step": 303505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014339266236505424,
+      "loss": 1.952,
+      "step": 303510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001433909971427422,
+      "loss": 2.0598,
+      "step": 303515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014338933190560703,
+      "loss": 2.216,
+      "step": 303520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014338766665364928,
+      "loss": 1.9016,
+      "step": 303525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001433860013868695,
+      "loss": 2.1628,
+      "step": 303530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001433843361052683,
+      "loss": 2.0023,
+      "step": 303535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001433826708088462,
+      "loss": 2.0137,
+      "step": 303540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001433810054976038,
+      "loss": 1.9752,
+      "step": 303545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014337934017154166,
+      "loss": 2.261,
+      "step": 303550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014337767483066035,
+      "loss": 2.0735,
+      "step": 303555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014337600947496044,
+      "loss": 2.1116,
+      "step": 303560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014337434410444252,
+      "loss": 2.1812,
+      "step": 303565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001433726787191071,
+      "loss": 1.9873,
+      "step": 303570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014337101331895482,
+      "loss": 2.2001,
+      "step": 303575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001433693479039862,
+      "loss": 2.0976,
+      "step": 303580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001433676824742018,
+      "loss": 1.9449,
+      "step": 303585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014336601702960228,
+      "loss": 2.0853,
+      "step": 303590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014336435157018807,
+      "loss": 2.0873,
+      "step": 303595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014336268609595987,
+      "loss": 2.0017,
+      "step": 303600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014336102060691813,
+      "loss": 1.9999,
+      "step": 303605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001433593551030635,
+      "loss": 2.2478,
+      "step": 303610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014335768958439657,
+      "loss": 1.8364,
+      "step": 303615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014335602405091783,
+      "loss": 1.949,
+      "step": 303620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001433543585026279,
+      "loss": 1.8906,
+      "step": 303625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014335269293952728,
+      "loss": 1.9855,
+      "step": 303630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014335102736161667,
+      "loss": 2.0574,
+      "step": 303635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014334936176889652,
+      "loss": 2.1376,
+      "step": 303640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014334769616136748,
+      "loss": 1.9932,
+      "step": 303645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014334603053903005,
+      "loss": 1.9805,
+      "step": 303650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014334436490188483,
+      "loss": 1.9983,
+      "step": 303655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001433426992499324,
+      "loss": 2.0983,
+      "step": 303660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001433410335831733,
+      "loss": 2.1676,
+      "step": 303665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014333936790160814,
+      "loss": 2.0734,
+      "step": 303670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014333770220523745,
+      "loss": 1.9731,
+      "step": 303675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014333603649406183,
+      "loss": 2.2596,
+      "step": 303680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001433343707680818,
+      "loss": 2.0149,
+      "step": 303685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000143332705027298,
+      "loss": 2.0909,
+      "step": 303690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014333103927171095,
+      "loss": 2.2844,
+      "step": 303695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014332937350132122,
+      "loss": 1.9932,
+      "step": 303700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001433277077161294,
+      "loss": 2.0298,
+      "step": 303705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014332604191613605,
+      "loss": 2.0068,
+      "step": 303710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014332437610134173,
+      "loss": 2.172,
+      "step": 303715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014332271027174705,
+      "loss": 2.0211,
+      "step": 303720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001433210444273525,
+      "loss": 2.114,
+      "step": 303725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014331937856815874,
+      "loss": 2.1477,
+      "step": 303730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014331771269416628,
+      "loss": 2.0107,
+      "step": 303735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014331604680537568,
+      "loss": 2.0468,
+      "step": 303740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014331438090178753,
+      "loss": 2.0531,
+      "step": 303745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014331271498340244,
+      "loss": 1.9449,
+      "step": 303750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001433110490502209,
+      "loss": 2.1722,
+      "step": 303755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014330938310224358,
+      "loss": 2.0206,
+      "step": 303760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001433077171394709,
+      "loss": 2.2092,
+      "step": 303765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001433060511619036,
+      "loss": 1.9884,
+      "step": 303770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014330438516954214,
+      "loss": 2.014,
+      "step": 303775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001433027191623871,
+      "loss": 2.05,
+      "step": 303780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001433010531404391,
+      "loss": 1.9961,
+      "step": 303785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014329938710369862,
+      "loss": 2.1853,
+      "step": 303790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014329772105216635,
+      "loss": 2.1806,
+      "step": 303795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014329605498584276,
+      "loss": 2.141,
+      "step": 303800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014329438890472844,
+      "loss": 2.2535,
+      "step": 303805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000143292722808824,
+      "loss": 2.0436,
+      "step": 303810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014329105669812995,
+      "loss": 2.2039,
+      "step": 303815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001432893905726469,
+      "loss": 2.0445,
+      "step": 303820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014328772443237541,
+      "loss": 2.3044,
+      "step": 303825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014328605827731606,
+      "loss": 2.0296,
+      "step": 303830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001432843921074694,
+      "loss": 2.1582,
+      "step": 303835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014328272592283603,
+      "loss": 2.0215,
+      "step": 303840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014328105972341646,
+      "loss": 2.1022,
+      "step": 303845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001432793935092113,
+      "loss": 2.11,
+      "step": 303850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014327772728022113,
+      "loss": 2.0909,
+      "step": 303855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014327606103644647,
+      "loss": 1.9843,
+      "step": 303860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000143274394777888,
+      "loss": 1.9786,
+      "step": 303865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014327272850454614,
+      "loss": 2.0468,
+      "step": 303870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014327106221642154,
+      "loss": 1.898,
+      "step": 303875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014326939591351477,
+      "loss": 2.0834,
+      "step": 303880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014326772959582642,
+      "loss": 2.1002,
+      "step": 303885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014326606326335698,
+      "loss": 2.0994,
+      "step": 303890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014326439691610712,
+      "loss": 2.1374,
+      "step": 303895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001432627305540773,
+      "loss": 2.0181,
+      "step": 303900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014326106417726818,
+      "loss": 2.0875,
+      "step": 303905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00014325939778568028,
+      "loss": 2.0466,
+      "step": 303910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001432577313793142,
+      "loss": 2.3566,
+      "step": 303915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001432560649581705,
+      "loss": 1.9995,
+      "step": 303920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001432543985222497,
+      "loss": 2.1571,
+      "step": 303925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014325273207155248,
+      "loss": 2.0677,
+      "step": 303930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001432510656060793,
+      "loss": 2.1883,
+      "step": 303935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014324939912583077,
+      "loss": 2.2263,
+      "step": 303940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014324773263080747,
+      "loss": 2.1445,
+      "step": 303945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014324606612100996,
+      "loss": 1.8759,
+      "step": 303950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001432443995964388,
+      "loss": 1.925,
+      "step": 303955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001432427330570946,
+      "loss": 2.074,
+      "step": 303960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014324106650297785,
+      "loss": 2.1232,
+      "step": 303965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001432393999340892,
+      "loss": 1.9404,
+      "step": 303970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014323773335042918,
+      "loss": 2.0177,
+      "step": 303975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014323606675199836,
+      "loss": 2.1004,
+      "step": 303980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014323440013879734,
+      "loss": 1.9848,
+      "step": 303985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014323273351082665,
+      "loss": 2.2051,
+      "step": 303990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014323106686808687,
+      "loss": 2.0672,
+      "step": 303995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014322940021057858,
+      "loss": 2.2707,
+      "step": 304000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014322773353830231,
+      "loss": 2.0675,
+      "step": 304005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001432260668512587,
+      "loss": 2.1738,
+      "step": 304010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014322440014944828,
+      "loss": 1.9911,
+      "step": 304015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001432227334328716,
+      "loss": 2.0263,
+      "step": 304020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014322106670152927,
+      "loss": 1.8386,
+      "step": 304025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014321939995542185,
+      "loss": 1.9097,
+      "step": 304030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014321773319454984,
+      "loss": 2.1332,
+      "step": 304035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014321606641891393,
+      "loss": 2.1626,
+      "step": 304040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014321439962851462,
+      "loss": 1.9062,
+      "step": 304045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014321273282335247,
+      "loss": 2.1051,
+      "step": 304050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001432110660034281,
+      "loss": 2.0232,
+      "step": 304055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000143209399168742,
+      "loss": 2.1878,
+      "step": 304060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001432077323192948,
+      "loss": 2.0063,
+      "step": 304065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014320606545508707,
+      "loss": 1.9283,
+      "step": 304070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014320439857611934,
+      "loss": 1.9346,
+      "step": 304075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014320273168239223,
+      "loss": 1.9911,
+      "step": 304080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014320106477390624,
+      "loss": 2.1609,
+      "step": 304085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00014319939785066202,
+      "loss": 1.8027,
+      "step": 304090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001431977309126601,
+      "loss": 1.8936,
+      "step": 304095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014319606395990107,
+      "loss": 2.1335,
+      "step": 304100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014319439699238544,
+      "loss": 2.1104,
+      "step": 304105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014319273001011385,
+      "loss": 2.1344,
+      "step": 304110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014319106301308683,
+      "loss": 1.9648,
+      "step": 304115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014318939600130494,
+      "loss": 1.9399,
+      "step": 304120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001431877289747688,
+      "loss": 1.8413,
+      "step": 304125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014318606193347892,
+      "loss": 1.9813,
+      "step": 304130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014318439487743593,
+      "loss": 2.0324,
+      "step": 304135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00014318272780664034,
+      "loss": 2.0843,
+      "step": 304140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014318106072109277,
+      "loss": 2.2884,
+      "step": 304145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014317939362079376,
+      "loss": 2.0579,
+      "step": 304150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014317772650574389,
+      "loss": 1.9667,
+      "step": 304155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001431760593759437,
+      "loss": 2.113,
+      "step": 304160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014317439223139383,
+      "loss": 2.1486,
+      "step": 304165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014317272507209476,
+      "loss": 2.0995,
+      "step": 304170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001431710578980471,
+      "loss": 1.8435,
+      "step": 304175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014316939070925145,
+      "loss": 2.1553,
+      "step": 304180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014316772350570837,
+      "loss": 2.0621,
+      "step": 304185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001431660562874184,
+      "loss": 1.9746,
+      "step": 304190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014316438905438207,
+      "loss": 2.1889,
+      "step": 304195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014316272180660007,
+      "loss": 2.1378,
+      "step": 304200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014316105454407287,
+      "loss": 2.217,
+      "step": 304205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014315938726680108,
+      "loss": 2.1669,
+      "step": 304210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014315771997478526,
+      "loss": 2.2743,
+      "step": 304215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014315605266802597,
+      "loss": 2.1374,
+      "step": 304220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001431543853465238,
+      "loss": 1.9643,
+      "step": 304225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001431527180102793,
+      "loss": 2.1674,
+      "step": 304230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014315105065929306,
+      "loss": 1.9656,
+      "step": 304235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014314938329356564,
+      "loss": 2.1254,
+      "step": 304240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001431477159130976,
+      "loss": 2.07,
+      "step": 304245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014314604851788952,
+      "loss": 2.1308,
+      "step": 304250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014314438110794197,
+      "loss": 2.2004,
+      "step": 304255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014314271368325553,
+      "loss": 2.0972,
+      "step": 304260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014314104624383074,
+      "loss": 2.0829,
+      "step": 304265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014313937878966817,
+      "loss": 2.053,
+      "step": 304270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014313771132076842,
+      "loss": 2.0005,
+      "step": 304275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014313604383713206,
+      "loss": 2.1288,
+      "step": 304280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014313437633875964,
+      "loss": 1.8792,
+      "step": 304285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014313270882565174,
+      "loss": 2.0933,
+      "step": 304290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014313104129780893,
+      "loss": 2.1801,
+      "step": 304295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001431293737552317,
+      "loss": 2.2623,
+      "step": 304300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014312770619792078,
+      "loss": 2.0246,
+      "step": 304305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014312603862587662,
+      "loss": 2.0999,
+      "step": 304310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014312437103909982,
+      "loss": 2.1738,
+      "step": 304315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014312270343759098,
+      "loss": 1.9748,
+      "step": 304320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014312103582135062,
+      "loss": 1.9437,
+      "step": 304325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014311936819037932,
+      "loss": 2.0798,
+      "step": 304330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001431177005446777,
+      "loss": 1.8808,
+      "step": 304335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014311603288424626,
+      "loss": 2.1885,
+      "step": 304340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014311436520908561,
+      "loss": 2.0243,
+      "step": 304345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014311269751919633,
+      "loss": 1.9265,
+      "step": 304350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001431110298145789,
+      "loss": 2.0052,
+      "step": 304355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014310936209523402,
+      "loss": 2.0764,
+      "step": 304360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001431076943611622,
+      "loss": 1.9999,
+      "step": 304365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000143106026612364,
+      "loss": 2.1506,
+      "step": 304370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014310435884884,
+      "loss": 2.0238,
+      "step": 304375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014310269107059077,
+      "loss": 1.9471,
+      "step": 304380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014310102327761686,
+      "loss": 2.2502,
+      "step": 304385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001430993554699189,
+      "loss": 2.0177,
+      "step": 304390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001430976876474974,
+      "loss": 2.2974,
+      "step": 304395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014309601981035296,
+      "loss": 2.1106,
+      "step": 304400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014309435195848611,
+      "loss": 2.1441,
+      "step": 304405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014309268409189743,
+      "loss": 2.1681,
+      "step": 304410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014309101621058756,
+      "loss": 2.198,
+      "step": 304415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000143089348314557,
+      "loss": 1.9463,
+      "step": 304420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014308768040380634,
+      "loss": 2.1842,
+      "step": 304425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014308601247833613,
+      "loss": 2.119,
+      "step": 304430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014308434453814696,
+      "loss": 2.2007,
+      "step": 304435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014308267658323938,
+      "loss": 2.1789,
+      "step": 304440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000143081008613614,
+      "loss": 1.9052,
+      "step": 304445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014307934062927137,
+      "loss": 2.0705,
+      "step": 304450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014307767263021208,
+      "loss": 1.8782,
+      "step": 304455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014307600461643664,
+      "loss": 1.9987,
+      "step": 304460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014307433658794565,
+      "loss": 1.8904,
+      "step": 304465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001430726685447397,
+      "loss": 1.932,
+      "step": 304470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014307100048681937,
+      "loss": 2.0982,
+      "step": 304475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014306933241418516,
+      "loss": 2.0138,
+      "step": 304480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001430676643268377,
+      "loss": 2.0082,
+      "step": 304485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014306599622477753,
+      "loss": 2.0378,
+      "step": 304490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014306432810800524,
+      "loss": 2.0569,
+      "step": 304495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014306265997652144,
+      "loss": 2.0449,
+      "step": 304500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014306099183032664,
+      "loss": 2.0991,
+      "step": 304505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014305932366942138,
+      "loss": 2.0694,
+      "step": 304510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001430576554938063,
+      "loss": 1.9928,
+      "step": 304515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014305598730348195,
+      "loss": 2.0836,
+      "step": 304520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014305431909844888,
+      "loss": 2.0593,
+      "step": 304525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001430526508787077,
+      "loss": 2.0441,
+      "step": 304530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014305098264425894,
+      "loss": 1.8469,
+      "step": 304535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014304931439510315,
+      "loss": 1.9566,
+      "step": 304540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014304764613124097,
+      "loss": 1.9755,
+      "step": 304545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014304597785267292,
+      "loss": 2.1876,
+      "step": 304550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001430443095593996,
+      "loss": 2.1827,
+      "step": 304555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014304264125142155,
+      "loss": 1.9064,
+      "step": 304560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014304097292873935,
+      "loss": 2.1059,
+      "step": 304565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014303930459135357,
+      "loss": 1.9911,
+      "step": 304570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001430376362392648,
+      "loss": 1.9724,
+      "step": 304575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014303596787247356,
+      "loss": 1.9517,
+      "step": 304580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014303429949098052,
+      "loss": 2.2353,
+      "step": 304585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014303263109478613,
+      "loss": 2.084,
+      "step": 304590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000143030962683891,
+      "loss": 2.0283,
+      "step": 304595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014302929425829574,
+      "loss": 2.1915,
+      "step": 304600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014302762581800088,
+      "loss": 2.1898,
+      "step": 304605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.000143025957363007,
+      "loss": 2.2147,
+      "step": 304610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014302428889331474,
+      "loss": 2.1922,
+      "step": 304615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001430226204089245,
+      "loss": 2.0287,
+      "step": 304620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014302095190983703,
+      "loss": 2.0247,
+      "step": 304625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014301928339605278,
+      "loss": 2.3038,
+      "step": 304630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001430176148675724,
+      "loss": 2.063,
+      "step": 304635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001430159463243964,
+      "loss": 2.1162,
+      "step": 304640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00014301427776652534,
+      "loss": 1.8824,
+      "step": 304645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014301260919395985,
+      "loss": 1.8499,
+      "step": 304650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014301094060670047,
+      "loss": 1.8817,
+      "step": 304655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014300927200474779,
+      "loss": 2.0574,
+      "step": 304660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014300760338810235,
+      "loss": 1.8891,
+      "step": 304665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014300593475676473,
+      "loss": 1.9338,
+      "step": 304670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014300426611073549,
+      "loss": 2.0774,
+      "step": 304675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014300259745001523,
+      "loss": 2.0321,
+      "step": 304680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001430009287746045,
+      "loss": 2.098,
+      "step": 304685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001429992600845039,
+      "loss": 2.0476,
+      "step": 304690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00014299759137971393,
+      "loss": 2.0167,
+      "step": 304695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014299592266023522,
+      "loss": 2.3073,
+      "step": 304700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014299425392606832,
+      "loss": 2.061,
+      "step": 304705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001429925851772138,
+      "loss": 1.9975,
+      "step": 304710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014299091641367223,
+      "loss": 1.9167,
+      "step": 304715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014298924763544418,
+      "loss": 2.1343,
+      "step": 304720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014298757884253024,
+      "loss": 2.1778,
+      "step": 304725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014298591003493097,
+      "loss": 2.033,
+      "step": 304730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014298424121264694,
+      "loss": 2.025,
+      "step": 304735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014298257237567867,
+      "loss": 1.9926,
+      "step": 304740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014298090352402683,
+      "loss": 1.9044,
+      "step": 304745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001429792346576919,
+      "loss": 1.9574,
+      "step": 304750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014297756577667448,
+      "loss": 2.1493,
+      "step": 304755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014297589688097513,
+      "loss": 1.9625,
+      "step": 304760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014297422797059447,
+      "loss": 2.0777,
+      "step": 304765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000142972559045533,
+      "loss": 2.0199,
+      "step": 304770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001429708901057914,
+      "loss": 1.9284,
+      "step": 304775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001429692211513701,
+      "loss": 2.1889,
+      "step": 304780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014296755218226972,
+      "loss": 2.0549,
+      "step": 304785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014296588319849085,
+      "loss": 2.2451,
+      "step": 304790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001429642142000341,
+      "loss": 1.9775,
+      "step": 304795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014296254518689997,
+      "loss": 1.8866,
+      "step": 304800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014296087615908905,
+      "loss": 2.0408,
+      "step": 304805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001429592071166019,
+      "loss": 2.0228,
+      "step": 304810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014295753805943913,
+      "loss": 2.0833,
+      "step": 304815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014295586898760128,
+      "loss": 2.2544,
+      "step": 304820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014295419990108897,
+      "loss": 2.1203,
+      "step": 304825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014295253079990268,
+      "loss": 2.0007,
+      "step": 304830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000142950861684043,
+      "loss": 2.1935,
+      "step": 304835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014294919255351056,
+      "loss": 2.0185,
+      "step": 304840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001429475234083059,
+      "loss": 1.9652,
+      "step": 304845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001429458542484296,
+      "loss": 1.9285,
+      "step": 304850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014294418507388218,
+      "loss": 2.1764,
+      "step": 304855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014294251588466427,
+      "loss": 2.1113,
+      "step": 304860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014294084668077642,
+      "loss": 2.0657,
+      "step": 304865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014293917746221918,
+      "loss": 2.2615,
+      "step": 304870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014293750822899314,
+      "loss": 2.0575,
+      "step": 304875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001429358389810989,
+      "loss": 1.8993,
+      "step": 304880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.000142934169718537,
+      "loss": 2.0791,
+      "step": 304885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014293250044130796,
+      "loss": 2.0526,
+      "step": 304890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001429308311494124,
+      "loss": 2.0555,
+      "step": 304895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014292916184285092,
+      "loss": 2.0618,
+      "step": 304900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014292749252162404,
+      "loss": 1.8578,
+      "step": 304905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001429258231857324,
+      "loss": 2.2018,
+      "step": 304910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014292415383517648,
+      "loss": 2.0685,
+      "step": 304915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014292248446995688,
+      "loss": 2.0399,
+      "step": 304920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001429208150900742,
+      "loss": 2.1109,
+      "step": 304925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000142919145695529,
+      "loss": 2.0897,
+      "step": 304930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014291747628632185,
+      "loss": 2.1197,
+      "step": 304935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014291580686245328,
+      "loss": 2.0543,
+      "step": 304940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014291413742392386,
+      "loss": 2.1713,
+      "step": 304945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014291246797073424,
+      "loss": 2.075,
+      "step": 304950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.000142910798502885,
+      "loss": 2.0614,
+      "step": 304955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014290912902037656,
+      "loss": 1.9064,
+      "step": 304960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001429074595232096,
+      "loss": 1.9285,
+      "step": 304965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014290579001138468,
+      "loss": 2.0665,
+      "step": 304970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001429041204849024,
+      "loss": 2.0989,
+      "step": 304975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014290245094376327,
+      "loss": 1.885,
+      "step": 304980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014290078138796787,
+      "loss": 1.921,
+      "step": 304985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001428991118175168,
+      "loss": 2.0689,
+      "step": 304990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014289744223241061,
+      "loss": 1.9754,
+      "step": 304995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001428957726326499,
+      "loss": 1.7303,
+      "step": 305000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014289410301823518,
+      "loss": 2.0573,
+      "step": 305005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014289243338916707,
+      "loss": 2.0512,
+      "step": 305010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014289076374544613,
+      "loss": 2.1155,
+      "step": 305015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014288909408707292,
+      "loss": 2.08,
+      "step": 305020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000142887424414048,
+      "loss": 1.9861,
+      "step": 305025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014288575472637201,
+      "loss": 2.1788,
+      "step": 305030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014288408502404542,
+      "loss": 2.1183,
+      "step": 305035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014288241530706886,
+      "loss": 2.1404,
+      "step": 305040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014288074557544292,
+      "loss": 1.9056,
+      "step": 305045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014287907582916808,
+      "loss": 2.0771,
+      "step": 305050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014287740606824502,
+      "loss": 2.0497,
+      "step": 305055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014287573629267425,
+      "loss": 2.2853,
+      "step": 305060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014287406650245635,
+      "loss": 2.1943,
+      "step": 305065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001428723966975919,
+      "loss": 2.0439,
+      "step": 305070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014287072687808145,
+      "loss": 2.1368,
+      "step": 305075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014286905704392555,
+      "loss": 1.8581,
+      "step": 305080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014286738719512484,
+      "loss": 2.0301,
+      "step": 305085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014286571733167986,
+      "loss": 2.1314,
+      "step": 305090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014286404745359117,
+      "loss": 2.1,
+      "step": 305095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014286237756085934,
+      "loss": 2.0653,
+      "step": 305100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001428607076534849,
+      "loss": 2.143,
+      "step": 305105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014285903773146854,
+      "loss": 2.1858,
+      "step": 305110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014285736779481073,
+      "loss": 2.0581,
+      "step": 305115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014285569784351206,
+      "loss": 2.1275,
+      "step": 305120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001428540278775731,
+      "loss": 2.0709,
+      "step": 305125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001428523578969944,
+      "loss": 1.9986,
+      "step": 305130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014285068790177661,
+      "loss": 1.9354,
+      "step": 305135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014284901789192022,
+      "loss": 2.0215,
+      "step": 305140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014284734786742585,
+      "loss": 1.8991,
+      "step": 305145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014284567782829401,
+      "loss": 1.9935,
+      "step": 305150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014284400777452534,
+      "loss": 2.0718,
+      "step": 305155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014284233770612038,
+      "loss": 1.8293,
+      "step": 305160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014284066762307968,
+      "loss": 1.8396,
+      "step": 305165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014283899752540383,
+      "loss": 1.9917,
+      "step": 305170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014283732741309342,
+      "loss": 1.8675,
+      "step": 305175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014283565728614902,
+      "loss": 2.0443,
+      "step": 305180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014283398714457114,
+      "loss": 2.1443,
+      "step": 305185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001428323169883604,
+      "loss": 2.0559,
+      "step": 305190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001428306468175174,
+      "loss": 2.1528,
+      "step": 305195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014282897663204263,
+      "loss": 1.9936,
+      "step": 305200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014282730643193676,
+      "loss": 2.2637,
+      "step": 305205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014282563621720025,
+      "loss": 2.0981,
+      "step": 305210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014282396598783373,
+      "loss": 2.1883,
+      "step": 305215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014282229574383778,
+      "loss": 1.9366,
+      "step": 305220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014282062548521299,
+      "loss": 2.0208,
+      "step": 305225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014281895521195986,
+      "loss": 1.848,
+      "step": 305230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000142817284924079,
+      "loss": 2.1894,
+      "step": 305235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014281561462157097,
+      "loss": 2.034,
+      "step": 305240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014281394430443637,
+      "loss": 2.0276,
+      "step": 305245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014281227397267576,
+      "loss": 1.9916,
+      "step": 305250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014281060362628968,
+      "loss": 2.0603,
+      "step": 305255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014280893326527874,
+      "loss": 2.0501,
+      "step": 305260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014280726288964346,
+      "loss": 2.2883,
+      "step": 305265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001428055924993845,
+      "loss": 1.961,
+      "step": 305270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001428039220945023,
+      "loss": 2.1463,
+      "step": 305275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014280225167499755,
+      "loss": 2.0295,
+      "step": 305280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014280058124087076,
+      "loss": 2.1542,
+      "step": 305285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014279891079212252,
+      "loss": 2.0429,
+      "step": 305290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014279724032875338,
+      "loss": 2.0286,
+      "step": 305295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014279556985076394,
+      "loss": 2.2127,
+      "step": 305300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014279389935815476,
+      "loss": 2.0455,
+      "step": 305305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014279222885092642,
+      "loss": 2.1161,
+      "step": 305310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014279055832907946,
+      "loss": 2.0209,
+      "step": 305315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014278888779261446,
+      "loss": 2.0598,
+      "step": 305320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014278721724153201,
+      "loss": 1.7997,
+      "step": 305325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014278554667583268,
+      "loss": 2.0749,
+      "step": 305330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014278387609551705,
+      "loss": 1.9958,
+      "step": 305335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014278220550058564,
+      "loss": 1.8407,
+      "step": 305340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014278053489103905,
+      "loss": 2.0234,
+      "step": 305345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014277886426687788,
+      "loss": 1.9844,
+      "step": 305350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014277719362810264,
+      "loss": 2.2593,
+      "step": 305355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014277552297471395,
+      "loss": 2.1204,
+      "step": 305360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014277385230671239,
+      "loss": 2.0554,
+      "step": 305365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014277218162409846,
+      "loss": 2.0188,
+      "step": 305370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014277051092687282,
+      "loss": 2.0113,
+      "step": 305375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014276884021503596,
+      "loss": 1.9837,
+      "step": 305380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014276716948858853,
+      "loss": 2.0445,
+      "step": 305385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014276549874753103,
+      "loss": 2.0035,
+      "step": 305390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014276382799186408,
+      "loss": 1.9279,
+      "step": 305395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014276215722158818,
+      "loss": 2.1856,
+      "step": 305400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014276048643670404,
+      "loss": 2.0384,
+      "step": 305405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014275881563721209,
+      "loss": 1.9465,
+      "step": 305410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014275714482311292,
+      "loss": 1.8414,
+      "step": 305415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001427554739944072,
+      "loss": 2.1596,
+      "step": 305420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014275380315109537,
+      "loss": 1.9489,
+      "step": 305425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014275213229317813,
+      "loss": 2.0721,
+      "step": 305430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014275046142065594,
+      "loss": 2.0342,
+      "step": 305435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014274879053352942,
+      "loss": 2.0073,
+      "step": 305440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014274711963179916,
+      "loss": 2.1736,
+      "step": 305445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014274544871546572,
+      "loss": 1.9776,
+      "step": 305450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001427437777845296,
+      "loss": 1.8193,
+      "step": 305455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001427421068389915,
+      "loss": 2.0315,
+      "step": 305460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001427404358788519,
+      "loss": 1.9829,
+      "step": 305465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014273876490411135,
+      "loss": 2.0701,
+      "step": 305470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001427370939147705,
+      "loss": 1.9032,
+      "step": 305475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001427354229108299,
+      "loss": 2.0902,
+      "step": 305480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014273375189229005,
+      "loss": 2.0386,
+      "step": 305485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001427320808591516,
+      "loss": 1.9379,
+      "step": 305490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014273040981141512,
+      "loss": 2.0708,
+      "step": 305495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014272873874908117,
+      "loss": 1.9508,
+      "step": 305500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014272706767215024,
+      "loss": 2.2197,
+      "step": 305505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014272539658062304,
+      "loss": 2.0039,
+      "step": 305510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001427237254745,
+      "loss": 2.0224,
+      "step": 305515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014272205435378183,
+      "loss": 2.0896,
+      "step": 305520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014272038321846902,
+      "loss": 2.0003,
+      "step": 305525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001427187120685621,
+      "loss": 1.8718,
+      "step": 305530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014271704090406174,
+      "loss": 2.1247,
+      "step": 305535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014271536972496843,
+      "loss": 2.0178,
+      "step": 305540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014271369853128283,
+      "loss": 1.976,
+      "step": 305545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001427120273230054,
+      "loss": 2.0434,
+      "step": 305550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014271035610013679,
+      "loss": 2.0453,
+      "step": 305555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014270868486267752,
+      "loss": 1.9549,
+      "step": 305560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014270701361062825,
+      "loss": 2.0223,
+      "step": 305565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014270534234398944,
+      "loss": 2.0031,
+      "step": 305570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014270367106276172,
+      "loss": 2.2809,
+      "step": 305575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.328125,
+      "learning_rate": 0.00014270199976694565,
+      "loss": 2.0019,
+      "step": 305580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014270032845654183,
+      "loss": 2.1415,
+      "step": 305585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014269865713155077,
+      "loss": 1.9583,
+      "step": 305590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014269698579197308,
+      "loss": 2.1183,
+      "step": 305595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014269531443780932,
+      "loss": 2.1609,
+      "step": 305600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001426936430690601,
+      "loss": 2.1395,
+      "step": 305605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014269197168572593,
+      "loss": 1.9055,
+      "step": 305610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001426903002878074,
+      "loss": 1.8918,
+      "step": 305615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001426886288753051,
+      "loss": 2.1959,
+      "step": 305620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001426869574482196,
+      "loss": 2.026,
+      "step": 305625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014268528600655146,
+      "loss": 2.3067,
+      "step": 305630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00014268361455030125,
+      "loss": 2.2303,
+      "step": 305635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014268194307946952,
+      "loss": 2.0329,
+      "step": 305640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014268027159405688,
+      "loss": 2.0766,
+      "step": 305645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014267860009406388,
+      "loss": 1.9874,
+      "step": 305650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001426769285794911,
+      "loss": 1.8967,
+      "step": 305655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014267525705033914,
+      "loss": 2.1365,
+      "step": 305660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001426735855066085,
+      "loss": 2.1902,
+      "step": 305665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001426719139482998,
+      "loss": 1.9084,
+      "step": 305670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001426702423754136,
+      "loss": 1.9825,
+      "step": 305675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014266857078795047,
+      "loss": 1.8793,
+      "step": 305680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014266689918591098,
+      "loss": 2.0058,
+      "step": 305685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014266522756929573,
+      "loss": 2.1004,
+      "step": 305690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014266355593810522,
+      "loss": 1.9713,
+      "step": 305695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001426618842923401,
+      "loss": 2.1484,
+      "step": 305700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001426602126320009,
+      "loss": 1.8532,
+      "step": 305705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014265854095708817,
+      "loss": 2.205,
+      "step": 305710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014265686926760254,
+      "loss": 1.9625,
+      "step": 305715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001426551975635445,
+      "loss": 2.1501,
+      "step": 305720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014265352584491475,
+      "loss": 1.9473,
+      "step": 305725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014265185411171374,
+      "loss": 2.1919,
+      "step": 305730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001426501823639421,
+      "loss": 2.0877,
+      "step": 305735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014264851060160033,
+      "loss": 2.1246,
+      "step": 305740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001426468388246891,
+      "loss": 2.0364,
+      "step": 305745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014264516703320893,
+      "loss": 1.7748,
+      "step": 305750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001426434952271604,
+      "loss": 2.168,
+      "step": 305755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014264182340654408,
+      "loss": 2.009,
+      "step": 305760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014264015157136055,
+      "loss": 2.0931,
+      "step": 305765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014263847972161036,
+      "loss": 2.0359,
+      "step": 305770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014263680785729408,
+      "loss": 1.9563,
+      "step": 305775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001426351359784123,
+      "loss": 2.0761,
+      "step": 305780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00014263346408496558,
+      "loss": 2.09,
+      "step": 305785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001426317921769545,
+      "loss": 2.0959,
+      "step": 305790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014263012025437965,
+      "loss": 2.1594,
+      "step": 305795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014262844831724154,
+      "loss": 2.0904,
+      "step": 305800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001426267763655408,
+      "loss": 1.8305,
+      "step": 305805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000142625104399278,
+      "loss": 1.9154,
+      "step": 305810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014262343241845367,
+      "loss": 2.095,
+      "step": 305815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014262176042306837,
+      "loss": 2.0738,
+      "step": 305820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014262008841312276,
+      "loss": 2.1114,
+      "step": 305825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001426184163886173,
+      "loss": 2.0326,
+      "step": 305830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014261674434955265,
+      "loss": 1.9979,
+      "step": 305835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014261507229592938,
+      "loss": 1.9829,
+      "step": 305840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014261340022774795,
+      "loss": 1.8016,
+      "step": 305845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014261172814500908,
+      "loss": 2.1739,
+      "step": 305850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001426100560477132,
+      "loss": 2.0814,
+      "step": 305855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000142608383935861,
+      "loss": 2.0181,
+      "step": 305860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014260671180945302,
+      "loss": 2.0295,
+      "step": 305865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001426050396684898,
+      "loss": 2.1096,
+      "step": 305870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001426033675129719,
+      "loss": 2.0045,
+      "step": 305875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014260169534289994,
+      "loss": 2.0676,
+      "step": 305880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014260002315827445,
+      "loss": 2.0804,
+      "step": 305885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014259835095909603,
+      "loss": 2.1243,
+      "step": 305890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014259667874536524,
+      "loss": 2.1184,
+      "step": 305895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014259500651708266,
+      "loss": 2.0611,
+      "step": 305900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014259333427424885,
+      "loss": 2.0786,
+      "step": 305905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001425916620168644,
+      "loss": 2.1187,
+      "step": 305910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014258998974492983,
+      "loss": 2.1257,
+      "step": 305915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014258831745844576,
+      "loss": 1.8331,
+      "step": 305920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014258664515741276,
+      "loss": 2.0179,
+      "step": 305925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001425849728418314,
+      "loss": 1.9512,
+      "step": 305930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001425833005117022,
+      "loss": 2.1009,
+      "step": 305935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001425816281670258,
+      "loss": 2.057,
+      "step": 305940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014257995580780272,
+      "loss": 2.0407,
+      "step": 305945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001425782834340336,
+      "loss": 2.1009,
+      "step": 305950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014257661104571896,
+      "loss": 2.1147,
+      "step": 305955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014257493864285934,
+      "loss": 2.0533,
+      "step": 305960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014257326622545534,
+      "loss": 1.9176,
+      "step": 305965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014257159379350758,
+      "loss": 2.1908,
+      "step": 305970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014256992134701658,
+      "loss": 1.9391,
+      "step": 305975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014256824888598293,
+      "loss": 1.9269,
+      "step": 305980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014256657641040717,
+      "loss": 2.0502,
+      "step": 305985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001425649039202899,
+      "loss": 2.0766,
+      "step": 305990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014256323141563173,
+      "loss": 2.0318,
+      "step": 305995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014256155889643315,
+      "loss": 2.1487,
+      "step": 306000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014255988636269478,
+      "loss": 2.0505,
+      "step": 306005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014255821381441717,
+      "loss": 1.9849,
+      "step": 306010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001425565412516009,
+      "loss": 1.9156,
+      "step": 306015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014255486867424658,
+      "loss": 2.1229,
+      "step": 306020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001425531960823547,
+      "loss": 2.214,
+      "step": 306025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001425515234759259,
+      "loss": 1.7641,
+      "step": 306030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001425498508549607,
+      "loss": 1.9728,
+      "step": 306035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014254817821945974,
+      "loss": 1.984,
+      "step": 306040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014254650556942352,
+      "loss": 2.0935,
+      "step": 306045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014254483290485265,
+      "loss": 2.2011,
+      "step": 306050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001425431602257477,
+      "loss": 2.0504,
+      "step": 306055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014254148753210924,
+      "loss": 1.9915,
+      "step": 306060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014253981482393783,
+      "loss": 2.0682,
+      "step": 306065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014253814210123404,
+      "loss": 2.034,
+      "step": 306070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014253646936399848,
+      "loss": 2.1686,
+      "step": 306075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014253479661223162,
+      "loss": 2.0676,
+      "step": 306080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014253312384593416,
+      "loss": 2.032,
+      "step": 306085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014253145106510665,
+      "loss": 2.3853,
+      "step": 306090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014252977826974953,
+      "loss": 1.8601,
+      "step": 306095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014252810545986353,
+      "loss": 2.1498,
+      "step": 306100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014252643263544914,
+      "loss": 2.0231,
+      "step": 306105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014252475979650698,
+      "loss": 2.0369,
+      "step": 306110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014252308694303754,
+      "loss": 2.0741,
+      "step": 306115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014252141407504145,
+      "loss": 2.2675,
+      "step": 306120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001425197411925193,
+      "loss": 2.1687,
+      "step": 306125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001425180682954716,
+      "loss": 2.0113,
+      "step": 306130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014251639538389902,
+      "loss": 1.9988,
+      "step": 306135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014251472245780202,
+      "loss": 2.2544,
+      "step": 306140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014251304951718122,
+      "loss": 1.9769,
+      "step": 306145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014251137656203717,
+      "loss": 2.1715,
+      "step": 306150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014250970359237054,
+      "loss": 1.8032,
+      "step": 306155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00014250803060818177,
+      "loss": 1.9579,
+      "step": 306160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014250635760947148,
+      "loss": 2.2319,
+      "step": 306165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014250468459624024,
+      "loss": 2.0749,
+      "step": 306170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014250301156848867,
+      "loss": 2.1534,
+      "step": 306175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014250133852621727,
+      "loss": 1.969,
+      "step": 306180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014249966546942667,
+      "loss": 2.1639,
+      "step": 306185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014249799239811735,
+      "loss": 1.9573,
+      "step": 306190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014249631931229003,
+      "loss": 2.0881,
+      "step": 306195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014249464621194513,
+      "loss": 2.201,
+      "step": 306200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001424929730970833,
+      "loss": 2.1951,
+      "step": 306205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014249129996770515,
+      "loss": 2.0685,
+      "step": 306210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014248962682381116,
+      "loss": 2.1371,
+      "step": 306215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014248795366540192,
+      "loss": 2.0096,
+      "step": 306220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014248628049247808,
+      "loss": 1.92,
+      "step": 306225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001424846073050401,
+      "loss": 2.083,
+      "step": 306230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014248293410308864,
+      "loss": 1.8994,
+      "step": 306235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014248126088662423,
+      "loss": 2.2307,
+      "step": 306240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014247958765564747,
+      "loss": 2.1603,
+      "step": 306245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014247791441015893,
+      "loss": 2.094,
+      "step": 306250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014247624115015908,
+      "loss": 2.1525,
+      "step": 306255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014247456787564862,
+      "loss": 1.9755,
+      "step": 306260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001424728945866281,
+      "loss": 2.0953,
+      "step": 306265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001424712212830981,
+      "loss": 2.2256,
+      "step": 306270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001424695479650591,
+      "loss": 2.1783,
+      "step": 306275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014246787463251175,
+      "loss": 1.9831,
+      "step": 306280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014246620128545657,
+      "loss": 2.0407,
+      "step": 306285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001424645279238942,
+      "loss": 1.9447,
+      "step": 306290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014246285454782518,
+      "loss": 2.243,
+      "step": 306295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001424611811572501,
+      "loss": 2.0215,
+      "step": 306300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014245950775216946,
+      "loss": 2.2021,
+      "step": 306305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001424578343325839,
+      "loss": 2.0898,
+      "step": 306310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000142456160898494,
+      "loss": 2.0362,
+      "step": 306315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014245448744990027,
+      "loss": 2.246,
+      "step": 306320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014245281398680333,
+      "loss": 1.9747,
+      "step": 306325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014245114050920373,
+      "loss": 2.0885,
+      "step": 306330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014244946701710208,
+      "loss": 2.3433,
+      "step": 306335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014244779351049887,
+      "loss": 1.9501,
+      "step": 306340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014244611998939476,
+      "loss": 2.0771,
+      "step": 306345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014244444645379029,
+      "loss": 2.0419,
+      "step": 306350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014244277290368604,
+      "loss": 2.0653,
+      "step": 306355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014244109933908255,
+      "loss": 2.0489,
+      "step": 306360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001424394257599804,
+      "loss": 1.9689,
+      "step": 306365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014243775216638018,
+      "loss": 2.0661,
+      "step": 306370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014243607855828245,
+      "loss": 1.9177,
+      "step": 306375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.0625,
+      "learning_rate": 0.00014243440493568778,
+      "loss": 2.0416,
+      "step": 306380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00014243273129859677,
+      "loss": 2.114,
+      "step": 306385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014243105764700993,
+      "loss": 2.1021,
+      "step": 306390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014242938398092793,
+      "loss": 2.0092,
+      "step": 306395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014242771030035123,
+      "loss": 2.0144,
+      "step": 306400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014242603660528051,
+      "loss": 2.1447,
+      "step": 306405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014242436289571626,
+      "loss": 2.0395,
+      "step": 306410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014242268917165904,
+      "loss": 1.9939,
+      "step": 306415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001424210154331095,
+      "loss": 1.8705,
+      "step": 306420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014241934168006816,
+      "loss": 2.0268,
+      "step": 306425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014241766791253562,
+      "loss": 2.266,
+      "step": 306430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014241599413051243,
+      "loss": 2.1171,
+      "step": 306435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014241432033399918,
+      "loss": 2.0286,
+      "step": 306440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014241264652299637,
+      "loss": 2.1976,
+      "step": 306445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001424109726975047,
+      "loss": 2.027,
+      "step": 306450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014240929885752464,
+      "loss": 2.105,
+      "step": 306455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001424076250030568,
+      "loss": 2.1266,
+      "step": 306460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014240595113410173,
+      "loss": 2.0246,
+      "step": 306465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014240427725066004,
+      "loss": 2.0327,
+      "step": 306470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014240260335273227,
+      "loss": 2.0333,
+      "step": 306475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.000142400929440319,
+      "loss": 2.0971,
+      "step": 306480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001423992555134208,
+      "loss": 1.9831,
+      "step": 306485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014239758157203827,
+      "loss": 2.15,
+      "step": 306490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001423959076161719,
+      "loss": 1.9907,
+      "step": 306495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001423942336458224,
+      "loss": 2.1017,
+      "step": 306500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014239255966099021,
+      "loss": 2.1811,
+      "step": 306505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014239088566167597,
+      "loss": 2.1168,
+      "step": 306510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014238921164788022,
+      "loss": 2.0894,
+      "step": 306515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014238753761960354,
+      "loss": 1.9981,
+      "step": 306520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001423858635768465,
+      "loss": 2.1288,
+      "step": 306525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014238418951960973,
+      "loss": 2.2083,
+      "step": 306530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001423825154478937,
+      "loss": 2.0736,
+      "step": 306535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014238084136169906,
+      "loss": 2.0542,
+      "step": 306540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014237916726102639,
+      "loss": 2.0983,
+      "step": 306545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014237749314587617,
+      "loss": 1.9974,
+      "step": 306550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014237581901624906,
+      "loss": 1.9806,
+      "step": 306555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014237414487214557,
+      "loss": 1.8767,
+      "step": 306560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014237247071356636,
+      "loss": 2.1457,
+      "step": 306565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014237079654051191,
+      "loss": 2.1272,
+      "step": 306570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014236912235298283,
+      "loss": 2.0126,
+      "step": 306575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014236744815097967,
+      "loss": 2.2385,
+      "step": 306580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00014236577393450306,
+      "loss": 2.2125,
+      "step": 306585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001423640997035535,
+      "loss": 2.179,
+      "step": 306590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001423624254581316,
+      "loss": 2.0254,
+      "step": 306595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014236075119823796,
+      "loss": 2.2306,
+      "step": 306600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014235907692387306,
+      "loss": 2.0234,
+      "step": 306605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014235740263503758,
+      "loss": 2.168,
+      "step": 306610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014235572833173204,
+      "loss": 1.9963,
+      "step": 306615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.00014235405401395702,
+      "loss": 2.0895,
+      "step": 306620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014235237968171304,
+      "loss": 2.0755,
+      "step": 306625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014235070533500076,
+      "loss": 1.9801,
+      "step": 306630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001423490309738207,
+      "loss": 2.017,
+      "step": 306635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014234735659817343,
+      "loss": 2.0778,
+      "step": 306640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014234568220805957,
+      "loss": 2.1082,
+      "step": 306645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014234400780347963,
+      "loss": 2.098,
+      "step": 306650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001423423333844342,
+      "loss": 2.1073,
+      "step": 306655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001423406589509239,
+      "loss": 2.0155,
+      "step": 306660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014233898450294923,
+      "loss": 2.0212,
+      "step": 306665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014233731004051082,
+      "loss": 1.9977,
+      "step": 306670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014233563556360917,
+      "loss": 2.0275,
+      "step": 306675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014233396107224497,
+      "loss": 2.0744,
+      "step": 306680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014233228656641866,
+      "loss": 2.1551,
+      "step": 306685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001423306120461309,
+      "loss": 2.1631,
+      "step": 306690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014232893751138224,
+      "loss": 1.9037,
+      "step": 306695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014232726296217326,
+      "loss": 2.0947,
+      "step": 306700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.09375,
+      "learning_rate": 0.00014232558839850448,
+      "loss": 2.0989,
+      "step": 306705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014232391382037652,
+      "loss": 1.9526,
+      "step": 306710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014232223922778993,
+      "loss": 2.094,
+      "step": 306715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014232056462074532,
+      "loss": 2.1423,
+      "step": 306720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014231888999924325,
+      "loss": 1.8558,
+      "step": 306725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014231721536328427,
+      "loss": 2.2498,
+      "step": 306730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014231554071286894,
+      "loss": 2.2636,
+      "step": 306735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014231386604799786,
+      "loss": 2.2013,
+      "step": 306740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014231219136867163,
+      "loss": 1.9799,
+      "step": 306745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014231051667489076,
+      "loss": 2.0728,
+      "step": 306750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014230884196665585,
+      "loss": 2.0707,
+      "step": 306755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001423071672439675,
+      "loss": 2.1145,
+      "step": 306760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014230549250682622,
+      "loss": 2.2162,
+      "step": 306765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014230381775523263,
+      "loss": 2.1509,
+      "step": 306770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014230214298918727,
+      "loss": 2.2256,
+      "step": 306775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014230046820869077,
+      "loss": 1.8277,
+      "step": 306780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014229879341374363,
+      "loss": 1.838,
+      "step": 306785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014229711860434647,
+      "loss": 2.0238,
+      "step": 306790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014229544378049983,
+      "loss": 1.9516,
+      "step": 306795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014229376894220433,
+      "loss": 2.1837,
+      "step": 306800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00014229209408946048,
+      "loss": 1.9971,
+      "step": 306805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001422904192222689,
+      "loss": 2.0864,
+      "step": 306810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014228874434063017,
+      "loss": 2.0827,
+      "step": 306815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014228706944454476,
+      "loss": 1.9223,
+      "step": 306820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014228539453401337,
+      "loss": 2.1777,
+      "step": 306825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014228371960903652,
+      "loss": 2.112,
+      "step": 306830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001422820446696148,
+      "loss": 2.1983,
+      "step": 306835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014228036971574877,
+      "loss": 2.0763,
+      "step": 306840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014227869474743896,
+      "loss": 1.9672,
+      "step": 306845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014227701976468602,
+      "loss": 2.0481,
+      "step": 306850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014227534476749044,
+      "loss": 2.1471,
+      "step": 306855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001422736697558529,
+      "loss": 2.1801,
+      "step": 306860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014227199472977386,
+      "loss": 2.0111,
+      "step": 306865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014227031968925392,
+      "loss": 2.0648,
+      "step": 306870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001422686446342937,
+      "loss": 1.9948,
+      "step": 306875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014226696956489378,
+      "loss": 2.0115,
+      "step": 306880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014226529448105465,
+      "loss": 1.9614,
+      "step": 306885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014226361938277698,
+      "loss": 2.0787,
+      "step": 306890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014226194427006123,
+      "loss": 2.0385,
+      "step": 306895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014226026914290803,
+      "loss": 2.1556,
+      "step": 306900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.375,
+      "learning_rate": 0.000142258594001318,
+      "loss": 2.1414,
+      "step": 306905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014225691884529166,
+      "loss": 2.0507,
+      "step": 306910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001422552436748296,
+      "loss": 1.9775,
+      "step": 306915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014225356848993236,
+      "loss": 1.93,
+      "step": 306920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001422518932906005,
+      "loss": 2.0978,
+      "step": 306925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014225021807683468,
+      "loss": 1.9915,
+      "step": 306930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001422485428486354,
+      "loss": 1.9093,
+      "step": 306935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014224686760600324,
+      "loss": 1.9428,
+      "step": 306940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014224519234893882,
+      "loss": 2.1301,
+      "step": 306945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014224351707744264,
+      "loss": 2.0916,
+      "step": 306950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014224184179151533,
+      "loss": 2.1727,
+      "step": 306955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014224016649115744,
+      "loss": 2.1369,
+      "step": 306960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014223849117636953,
+      "loss": 2.2109,
+      "step": 306965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014223681584715222,
+      "loss": 2.0362,
+      "step": 306970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014223514050350602,
+      "loss": 1.9996,
+      "step": 306975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014223346514543152,
+      "loss": 2.1,
+      "step": 306980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001422317897729293,
+      "loss": 2.2105,
+      "step": 306985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014223011438599997,
+      "loss": 2.0259,
+      "step": 306990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014222843898464404,
+      "loss": 2.2608,
+      "step": 306995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001422267635688621,
+      "loss": 2.0546,
+      "step": 307000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00014222508813865474,
+      "loss": 2.0183,
+      "step": 307005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014222341269402256,
+      "loss": 1.7579,
+      "step": 307010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014222173723496605,
+      "loss": 1.9995,
+      "step": 307015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014222006176148585,
+      "loss": 1.7573,
+      "step": 307020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001422183862735825,
+      "loss": 2.0603,
+      "step": 307025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001422167107712566,
+      "loss": 2.0378,
+      "step": 307030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014221503525450868,
+      "loss": 2.1051,
+      "step": 307035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001422133597233394,
+      "loss": 1.8501,
+      "step": 307040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001422116841777492,
+      "loss": 2.045,
+      "step": 307045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014221000861773875,
+      "loss": 2.1019,
+      "step": 307050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001422083330433086,
+      "loss": 2.0943,
+      "step": 307055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014220665745445928,
+      "loss": 1.9273,
+      "step": 307060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014220498185119147,
+      "loss": 2.14,
+      "step": 307065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014220330623350563,
+      "loss": 2.0034,
+      "step": 307070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014220163060140237,
+      "loss": 2.074,
+      "step": 307075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014219995495488229,
+      "loss": 2.1715,
+      "step": 307080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014219827929394592,
+      "loss": 2.1297,
+      "step": 307085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014219660361859386,
+      "loss": 2.0782,
+      "step": 307090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014219492792882666,
+      "loss": 2.1358,
+      "step": 307095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014219325222464494,
+      "loss": 2.1588,
+      "step": 307100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001421915765060492,
+      "loss": 2.0829,
+      "step": 307105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014218990077304012,
+      "loss": 1.9811,
+      "step": 307110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014218822502561814,
+      "loss": 2.0916,
+      "step": 307115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001421865492637839,
+      "loss": 2.2201,
+      "step": 307120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000142184873487538,
+      "loss": 1.8683,
+      "step": 307125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014218319769688096,
+      "loss": 1.9774,
+      "step": 307130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001421815218918134,
+      "loss": 2.0426,
+      "step": 307135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014217984607233584,
+      "loss": 2.1482,
+      "step": 307140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014217817023844887,
+      "loss": 1.8868,
+      "step": 307145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001421764943901531,
+      "loss": 2.1178,
+      "step": 307150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001421748185274491,
+      "loss": 2.1465,
+      "step": 307155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014217314265033737,
+      "loss": 2.0034,
+      "step": 307160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014217146675881852,
+      "loss": 2.2005,
+      "step": 307165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014216979085289317,
+      "loss": 1.9612,
+      "step": 307170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014216811493256187,
+      "loss": 1.9639,
+      "step": 307175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014216643899782515,
+      "loss": 2.0272,
+      "step": 307180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014216476304868358,
+      "loss": 2.1207,
+      "step": 307185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001421630870851378,
+      "loss": 1.9147,
+      "step": 307190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001421614111071883,
+      "loss": 2.076,
+      "step": 307195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00014215973511483575,
+      "loss": 1.9494,
+      "step": 307200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014215805910808066,
+      "loss": 2.1219,
+      "step": 307205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014215638308692362,
+      "loss": 1.9933,
+      "step": 307210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00014215470705136515,
+      "loss": 2.1863,
+      "step": 307215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001421530310014059,
+      "loss": 1.8774,
+      "step": 307220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014215135493704642,
+      "loss": 2.1705,
+      "step": 307225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014214967885828726,
+      "loss": 1.9952,
+      "step": 307230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014214800276512898,
+      "loss": 2.0042,
+      "step": 307235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014214632665757223,
+      "loss": 2.1847,
+      "step": 307240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014214465053561748,
+      "loss": 2.079,
+      "step": 307245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014214297439926537,
+      "loss": 1.992,
+      "step": 307250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014214129824851647,
+      "loss": 2.0311,
+      "step": 307255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014213962208337134,
+      "loss": 1.9889,
+      "step": 307260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014213794590383055,
+      "loss": 2.1466,
+      "step": 307265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014213626970989467,
+      "loss": 1.8897,
+      "step": 307270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014213459350156425,
+      "loss": 2.1372,
+      "step": 307275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001421329172788399,
+      "loss": 2.2553,
+      "step": 307280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014213124104172218,
+      "loss": 2.0687,
+      "step": 307285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001421295647902117,
+      "loss": 2.0037,
+      "step": 307290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014212788852430898,
+      "loss": 2.0936,
+      "step": 307295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014212621224401456,
+      "loss": 2.0266,
+      "step": 307300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014212453594932912,
+      "loss": 2.0477,
+      "step": 307305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014212285964025317,
+      "loss": 1.9398,
+      "step": 307310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014212118331678728,
+      "loss": 2.1865,
+      "step": 307315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014211950697893202,
+      "loss": 1.9747,
+      "step": 307320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014211783062668796,
+      "loss": 2.1192,
+      "step": 307325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014211615426005568,
+      "loss": 2.0284,
+      "step": 307330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001421144778790358,
+      "loss": 2.0242,
+      "step": 307335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014211280148362885,
+      "loss": 2.0037,
+      "step": 307340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014211112507383537,
+      "loss": 1.9769,
+      "step": 307345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000142109448649656,
+      "loss": 2.0693,
+      "step": 307350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014210777221109123,
+      "loss": 2.0508,
+      "step": 307355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001421060957581417,
+      "loss": 2.158,
+      "step": 307360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000142104419290808,
+      "loss": 2.0911,
+      "step": 307365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014210274280909065,
+      "loss": 2.0367,
+      "step": 307370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001421010663129902,
+      "loss": 2.0421,
+      "step": 307375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001420993898025073,
+      "loss": 1.9567,
+      "step": 307380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014209771327764248,
+      "loss": 2.1819,
+      "step": 307385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014209603673839632,
+      "loss": 1.9078,
+      "step": 307390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014209436018476938,
+      "loss": 2.0479,
+      "step": 307395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00014209268361676223,
+      "loss": 2.1354,
+      "step": 307400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001420910070343755,
+      "loss": 2.0707,
+      "step": 307405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014208933043760968,
+      "loss": 2.1169,
+      "step": 307410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014208765382646538,
+      "loss": 2.0005,
+      "step": 307415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014208597720094318,
+      "loss": 2.0352,
+      "step": 307420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014208430056104367,
+      "loss": 1.8952,
+      "step": 307425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014208262390676742,
+      "loss": 2.1481,
+      "step": 307430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014208094723811493,
+      "loss": 2.0451,
+      "step": 307435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001420792705550868,
+      "loss": 1.9457,
+      "step": 307440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014207759385768368,
+      "loss": 1.9524,
+      "step": 307445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014207591714590612,
+      "loss": 1.9931,
+      "step": 307450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014207424041975462,
+      "loss": 2.197,
+      "step": 307455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014207256367922976,
+      "loss": 2.1122,
+      "step": 307460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001420708869243322,
+      "loss": 1.9753,
+      "step": 307465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014206921015506247,
+      "loss": 2.058,
+      "step": 307470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014206753337142112,
+      "loss": 2.0225,
+      "step": 307475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014206585657340875,
+      "loss": 2.1496,
+      "step": 307480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014206417976102588,
+      "loss": 2.076,
+      "step": 307485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014206250293427312,
+      "loss": 2.1654,
+      "step": 307490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001420608260931511,
+      "loss": 2.1877,
+      "step": 307495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014205914923766033,
+      "loss": 2.1117,
+      "step": 307500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014205747236780138,
+      "loss": 2.0772,
+      "step": 307505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014205579548357483,
+      "loss": 1.8994,
+      "step": 307510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014205411858498126,
+      "loss": 2.0576,
+      "step": 307515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001420524416720212,
+      "loss": 2.1224,
+      "step": 307520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014205076474469531,
+      "loss": 2.1177,
+      "step": 307525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014204908780300412,
+      "loss": 2.0071,
+      "step": 307530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001420474108469482,
+      "loss": 2.0652,
+      "step": 307535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014204573387652808,
+      "loss": 1.9956,
+      "step": 307540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001420440568917444,
+      "loss": 2.0481,
+      "step": 307545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014204237989259773,
+      "loss": 2.0868,
+      "step": 307550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001420407028790886,
+      "loss": 2.0504,
+      "step": 307555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001420390258512176,
+      "loss": 2.0278,
+      "step": 307560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014203734880898533,
+      "loss": 2.0046,
+      "step": 307565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001420356717523923,
+      "loss": 2.1482,
+      "step": 307570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014203399468143914,
+      "loss": 2.2381,
+      "step": 307575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001420323175961264,
+      "loss": 2.071,
+      "step": 307580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001420306404964547,
+      "loss": 2.0733,
+      "step": 307585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014202896338242452,
+      "loss": 2.0434,
+      "step": 307590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001420272862540365,
+      "loss": 2.1594,
+      "step": 307595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001420256091112912,
+      "loss": 1.8767,
+      "step": 307600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014202393195418916,
+      "loss": 2.0036,
+      "step": 307605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014202225478273102,
+      "loss": 2.0806,
+      "step": 307610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014202057759691732,
+      "loss": 2.0757,
+      "step": 307615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001420189003967486,
+      "loss": 1.9769,
+      "step": 307620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014201722318222547,
+      "loss": 2.1965,
+      "step": 307625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014201554595334853,
+      "loss": 1.9432,
+      "step": 307630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014201386871011827,
+      "loss": 2.0181,
+      "step": 307635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014201219145253534,
+      "loss": 2.0905,
+      "step": 307640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014201051418060028,
+      "loss": 2.0198,
+      "step": 307645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014200883689431364,
+      "loss": 2.1285,
+      "step": 307650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014200715959367604,
+      "loss": 2.0758,
+      "step": 307655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014200548227868805,
+      "loss": 1.8613,
+      "step": 307660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001420038049493502,
+      "loss": 2.0923,
+      "step": 307665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014200212760566308,
+      "loss": 2.0248,
+      "step": 307670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001420004502476273,
+      "loss": 2.0426,
+      "step": 307675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014199877287524336,
+      "loss": 2.034,
+      "step": 307680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014199709548851192,
+      "loss": 2.066,
+      "step": 307685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001419954180874335,
+      "loss": 2.2157,
+      "step": 307690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014199374067200868,
+      "loss": 2.0644,
+      "step": 307695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014199206324223807,
+      "loss": 1.9759,
+      "step": 307700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014199038579812214,
+      "loss": 2.0066,
+      "step": 307705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014198870833966157,
+      "loss": 2.1027,
+      "step": 307710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001419870308668569,
+      "loss": 2.0748,
+      "step": 307715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014198535337970868,
+      "loss": 2.0338,
+      "step": 307720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.328125,
+      "learning_rate": 0.00014198367587821756,
+      "loss": 2.2258,
+      "step": 307725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014198199836238398,
+      "loss": 2.1615,
+      "step": 307730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001419803208322086,
+      "loss": 2.229,
+      "step": 307735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000141978643287692,
+      "loss": 2.1686,
+      "step": 307740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014197696572883478,
+      "loss": 1.9168,
+      "step": 307745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014197528815563742,
+      "loss": 2.067,
+      "step": 307750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014197361056810052,
+      "loss": 2.1104,
+      "step": 307755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001419719329662247,
+      "loss": 2.2372,
+      "step": 307760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014197025535001053,
+      "loss": 2.0997,
+      "step": 307765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001419685777194585,
+      "loss": 2.0104,
+      "step": 307770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014196690007456928,
+      "loss": 1.9983,
+      "step": 307775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001419652224153434,
+      "loss": 2.2685,
+      "step": 307780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001419635447417814,
+      "loss": 2.0323,
+      "step": 307785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014196186705388398,
+      "loss": 2.2758,
+      "step": 307790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014196018935165158,
+      "loss": 2.1411,
+      "step": 307795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001419585116350848,
+      "loss": 2.0269,
+      "step": 307800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014195683390418425,
+      "loss": 2.077,
+      "step": 307805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014195515615895047,
+      "loss": 2.0756,
+      "step": 307810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014195347839938407,
+      "loss": 2.0107,
+      "step": 307815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014195180062548558,
+      "loss": 2.1103,
+      "step": 307820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014195012283725562,
+      "loss": 2.1362,
+      "step": 307825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001419484450346947,
+      "loss": 2.0683,
+      "step": 307830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014194676721780347,
+      "loss": 1.9119,
+      "step": 307835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014194508938658243,
+      "loss": 2.042,
+      "step": 307840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014194341154103222,
+      "loss": 2.123,
+      "step": 307845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014194173368115335,
+      "loss": 2.173,
+      "step": 307850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014194005580694643,
+      "loss": 1.9493,
+      "step": 307855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014193837791841204,
+      "loss": 2.0162,
+      "step": 307860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014193670001555074,
+      "loss": 2.1068,
+      "step": 307865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014193502209836306,
+      "loss": 2.07,
+      "step": 307870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014193334416684968,
+      "loss": 1.9413,
+      "step": 307875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001419316662210111,
+      "loss": 2.2146,
+      "step": 307880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014192998826084783,
+      "loss": 2.0751,
+      "step": 307885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014192831028636058,
+      "loss": 2.0957,
+      "step": 307890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014192663229754983,
+      "loss": 2.0452,
+      "step": 307895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001419249542944162,
+      "loss": 2.1321,
+      "step": 307900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014192327627696028,
+      "loss": 1.9427,
+      "step": 307905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014192159824518257,
+      "loss": 1.9236,
+      "step": 307910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014191992019908363,
+      "loss": 2.0897,
+      "step": 307915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014191824213866415,
+      "loss": 2.1245,
+      "step": 307920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014191656406392465,
+      "loss": 2.0426,
+      "step": 307925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014191488597486567,
+      "loss": 2.1568,
+      "step": 307930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001419132078714878,
+      "loss": 2.0573,
+      "step": 307935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014191152975379163,
+      "loss": 1.9785,
+      "step": 307940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001419098516217777,
+      "loss": 2.0116,
+      "step": 307945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001419081734754466,
+      "loss": 1.9184,
+      "step": 307950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014190649531479893,
+      "loss": 2.0956,
+      "step": 307955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014190481713983522,
+      "loss": 2.11,
+      "step": 307960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014190313895055608,
+      "loss": 2.0874,
+      "step": 307965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001419014607469621,
+      "loss": 1.8444,
+      "step": 307970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014189978252905378,
+      "loss": 2.1079,
+      "step": 307975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014189810429683175,
+      "loss": 1.9769,
+      "step": 307980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014189642605029655,
+      "loss": 2.1337,
+      "step": 307985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001418947477894488,
+      "loss": 2.1529,
+      "step": 307990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000141893069514289,
+      "loss": 1.9985,
+      "step": 307995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001418913912248178,
+      "loss": 1.9396,
+      "step": 308000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014188971292103573,
+      "loss": 1.9756,
+      "step": 308005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014188803460294338,
+      "loss": 2.1082,
+      "step": 308010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001418863562705413,
+      "loss": 1.9945,
+      "step": 308015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014188467792383014,
+      "loss": 2.071,
+      "step": 308020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014188299956281034,
+      "loss": 2.0529,
+      "step": 308025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001418813211874826,
+      "loss": 1.9986,
+      "step": 308030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001418796427978474,
+      "loss": 2.1705,
+      "step": 308035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001418779643939054,
+      "loss": 1.8616,
+      "step": 308040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001418762859756571,
+      "loss": 2.214,
+      "step": 308045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001418746075431031,
+      "loss": 2.045,
+      "step": 308050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014187292909624398,
+      "loss": 2.0789,
+      "step": 308055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001418712506350803,
+      "loss": 2.1661,
+      "step": 308060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014186957215961266,
+      "loss": 2.1239,
+      "step": 308065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001418678936698416,
+      "loss": 2.1894,
+      "step": 308070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001418662151657677,
+      "loss": 2.1172,
+      "step": 308075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014186453664739155,
+      "loss": 2.0888,
+      "step": 308080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014186285811471373,
+      "loss": 2.0477,
+      "step": 308085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001418611795677348,
+      "loss": 2.1378,
+      "step": 308090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001418595010064553,
+      "loss": 2.0086,
+      "step": 308095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014185782243087585,
+      "loss": 2.0966,
+      "step": 308100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014185614384099702,
+      "loss": 1.9019,
+      "step": 308105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014185446523681935,
+      "loss": 2.0056,
+      "step": 308110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014185278661834347,
+      "loss": 2.1476,
+      "step": 308115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001418511079855699,
+      "loss": 2.0993,
+      "step": 308120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014184942933849923,
+      "loss": 2.1496,
+      "step": 308125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014184775067713204,
+      "loss": 2.1136,
+      "step": 308130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014184607200146888,
+      "loss": 2.1856,
+      "step": 308135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014184439331151036,
+      "loss": 1.975,
+      "step": 308140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014184271460725704,
+      "loss": 2.1651,
+      "step": 308145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014184103588870948,
+      "loss": 1.8362,
+      "step": 308150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001418393571558683,
+      "loss": 2.208,
+      "step": 308155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000141837678408734,
+      "loss": 1.8671,
+      "step": 308160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001418359996473072,
+      "loss": 1.9932,
+      "step": 308165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014183432087158844,
+      "loss": 2.0188,
+      "step": 308170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014183264208157836,
+      "loss": 2.0217,
+      "step": 308175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014183096327727746,
+      "loss": 2.3162,
+      "step": 308180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014182928445868634,
+      "loss": 1.8573,
+      "step": 308185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001418276056258056,
+      "loss": 2.07,
+      "step": 308190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014182592677863576,
+      "loss": 2.1616,
+      "step": 308195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014182424791717749,
+      "loss": 2.0807,
+      "step": 308200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014182256904143124,
+      "loss": 2.009,
+      "step": 308205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014182089015139763,
+      "loss": 2.1523,
+      "step": 308210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014181921124707726,
+      "loss": 2.1681,
+      "step": 308215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014181753232847072,
+      "loss": 2.1335,
+      "step": 308220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014181585339557854,
+      "loss": 2.1431,
+      "step": 308225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014181417444840132,
+      "loss": 1.9933,
+      "step": 308230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014181249548693957,
+      "loss": 2.2013,
+      "step": 308235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014181081651119397,
+      "loss": 2.1446,
+      "step": 308240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000141809137521165,
+      "loss": 1.9853,
+      "step": 308245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001418074585168533,
+      "loss": 1.9958,
+      "step": 308250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014180577949825937,
+      "loss": 2.1646,
+      "step": 308255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014180410046538386,
+      "loss": 2.064,
+      "step": 308260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001418024214182273,
+      "loss": 1.9802,
+      "step": 308265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014180074235679027,
+      "loss": 2.113,
+      "step": 308270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014179906328107338,
+      "loss": 1.9463,
+      "step": 308275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014179738419107713,
+      "loss": 1.9172,
+      "step": 308280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014179570508680216,
+      "loss": 2.1901,
+      "step": 308285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000141794025968249,
+      "loss": 2.069,
+      "step": 308290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014179234683541826,
+      "loss": 2.1296,
+      "step": 308295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001417906676883105,
+      "loss": 2.2515,
+      "step": 308300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014178898852692625,
+      "loss": 1.9494,
+      "step": 308305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001417873093512662,
+      "loss": 2.045,
+      "step": 308310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001417856301613308,
+      "loss": 2.0793,
+      "step": 308315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014178395095712065,
+      "loss": 2.1179,
+      "step": 308320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014178227173863636,
+      "loss": 2.2616,
+      "step": 308325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014178059250587852,
+      "loss": 1.9005,
+      "step": 308330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014177891325884765,
+      "loss": 1.9414,
+      "step": 308335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014177723399754434,
+      "loss": 2.0322,
+      "step": 308340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014177555472196914,
+      "loss": 1.8751,
+      "step": 308345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001417738754321227,
+      "loss": 2.1964,
+      "step": 308350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014177219612800557,
+      "loss": 1.8828,
+      "step": 308355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014177051680961826,
+      "loss": 2.2443,
+      "step": 308360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014176883747696136,
+      "loss": 1.9601,
+      "step": 308365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001417671581300355,
+      "loss": 2.0808,
+      "step": 308370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014176547876884122,
+      "loss": 1.8701,
+      "step": 308375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001417637993933791,
+      "loss": 2.0831,
+      "step": 308380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014176212000364968,
+      "loss": 2.1975,
+      "step": 308385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014176044059965357,
+      "loss": 1.9129,
+      "step": 308390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014175876118139138,
+      "loss": 2.3301,
+      "step": 308395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001417570817488636,
+      "loss": 2.1304,
+      "step": 308400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014175540230207086,
+      "loss": 2.0708,
+      "step": 308405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014175372284101369,
+      "loss": 1.9443,
+      "step": 308410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001417520433656927,
+      "loss": 2.1282,
+      "step": 308415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014175036387610847,
+      "loss": 2.0165,
+      "step": 308420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014174868437226155,
+      "loss": 2.2265,
+      "step": 308425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014174700485415254,
+      "loss": 2.0626,
+      "step": 308430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014174532532178197,
+      "loss": 2.0242,
+      "step": 308435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014174364577515045,
+      "loss": 2.204,
+      "step": 308440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014174196621425856,
+      "loss": 1.9903,
+      "step": 308445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014174028663910683,
+      "loss": 2.0076,
+      "step": 308450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014173860704969586,
+      "loss": 2.1546,
+      "step": 308455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014173692744602625,
+      "loss": 2.0079,
+      "step": 308460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014173524782809852,
+      "loss": 1.8796,
+      "step": 308465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001417335681959133,
+      "loss": 2.1328,
+      "step": 308470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014173188854947114,
+      "loss": 1.8814,
+      "step": 308475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014173020888877259,
+      "loss": 2.0342,
+      "step": 308480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014172852921381823,
+      "loss": 1.957,
+      "step": 308485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014172684952460868,
+      "loss": 1.9217,
+      "step": 308490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014172516982114447,
+      "loss": 2.1634,
+      "step": 308495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001417234901034262,
+      "loss": 2.1081,
+      "step": 308500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014172181037145438,
+      "loss": 2.2182,
+      "step": 308505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014172013062522967,
+      "loss": 2.1414,
+      "step": 308510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001417184508647526,
+      "loss": 2.1312,
+      "step": 308515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014171677109002376,
+      "loss": 2.3455,
+      "step": 308520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014171509130104373,
+      "loss": 2.0831,
+      "step": 308525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014171341149781303,
+      "loss": 1.9672,
+      "step": 308530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001417117316803323,
+      "loss": 2.016,
+      "step": 308535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014171005184860207,
+      "loss": 1.9579,
+      "step": 308540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014170837200262293,
+      "loss": 2.0024,
+      "step": 308545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014170669214239544,
+      "loss": 2.0647,
+      "step": 308550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001417050122679202,
+      "loss": 2.1378,
+      "step": 308555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014170333237919778,
+      "loss": 2.0357,
+      "step": 308560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014170165247622874,
+      "loss": 2.085,
+      "step": 308565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014169997255901367,
+      "loss": 2.1297,
+      "step": 308570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014169829262755313,
+      "loss": 2.1373,
+      "step": 308575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014169661268184769,
+      "loss": 1.9733,
+      "step": 308580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014169493272189794,
+      "loss": 2.0616,
+      "step": 308585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001416932527477044,
+      "loss": 2.0784,
+      "step": 308590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014169157275926774,
+      "loss": 2.0703,
+      "step": 308595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014168989275658845,
+      "loss": 1.965,
+      "step": 308600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014168821273966717,
+      "loss": 2.0358,
+      "step": 308605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014168653270850445,
+      "loss": 2.1016,
+      "step": 308610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014168485266310082,
+      "loss": 1.9742,
+      "step": 308615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014168317260345687,
+      "loss": 1.8794,
+      "step": 308620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014168149252957323,
+      "loss": 2.0351,
+      "step": 308625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014167981244145046,
+      "loss": 2.2531,
+      "step": 308630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014167813233908908,
+      "loss": 2.0755,
+      "step": 308635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014167645222248963,
+      "loss": 1.982,
+      "step": 308640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014167477209165282,
+      "loss": 2.0269,
+      "step": 308645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014167309194657916,
+      "loss": 2.1454,
+      "step": 308650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001416714117872692,
+      "loss": 2.1731,
+      "step": 308655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014166973161372352,
+      "loss": 2.0906,
+      "step": 308660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001416680514259427,
+      "loss": 2.2052,
+      "step": 308665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001416663712239273,
+      "loss": 2.1899,
+      "step": 308670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014166469100767798,
+      "loss": 1.9303,
+      "step": 308675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001416630107771952,
+      "loss": 2.0164,
+      "step": 308680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001416613305324796,
+      "loss": 1.8905,
+      "step": 308685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014165965027353172,
+      "loss": 2.0525,
+      "step": 308690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014165797000035215,
+      "loss": 2.0695,
+      "step": 308695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014165628971294145,
+      "loss": 2.0872,
+      "step": 308700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014165460941130021,
+      "loss": 2.0565,
+      "step": 308705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000141652929095429,
+      "loss": 2.0491,
+      "step": 308710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001416512487653284,
+      "loss": 1.7744,
+      "step": 308715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4375,
+      "learning_rate": 0.000141649568420999,
+      "loss": 1.9187,
+      "step": 308720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014164788806244133,
+      "loss": 2.0944,
+      "step": 308725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014164620768965597,
+      "loss": 1.9678,
+      "step": 308730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001416445273026435,
+      "loss": 2.2234,
+      "step": 308735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014164284690140454,
+      "loss": 1.9943,
+      "step": 308740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001416411664859396,
+      "loss": 2.0335,
+      "step": 308745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014163948605624933,
+      "loss": 2.0915,
+      "step": 308750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001416378056123342,
+      "loss": 2.199,
+      "step": 308755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001416361251541949,
+      "loss": 1.9705,
+      "step": 308760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014163444468183193,
+      "loss": 2.2712,
+      "step": 308765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014163276419524584,
+      "loss": 2.1422,
+      "step": 308770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014163108369443726,
+      "loss": 2.181,
+      "step": 308775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014162940317940674,
+      "loss": 2.1475,
+      "step": 308780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014162772265015488,
+      "loss": 2.055,
+      "step": 308785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014162604210668224,
+      "loss": 2.0824,
+      "step": 308790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014162436154898937,
+      "loss": 2.0578,
+      "step": 308795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00014162268097707686,
+      "loss": 2.1348,
+      "step": 308800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014162100039094528,
+      "loss": 1.8607,
+      "step": 308805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00014161931979059524,
+      "loss": 2.0658,
+      "step": 308810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001416176391760273,
+      "loss": 1.8439,
+      "step": 308815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014161595854724198,
+      "loss": 2.0698,
+      "step": 308820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014161427790423988,
+      "loss": 2.0879,
+      "step": 308825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014161259724702164,
+      "loss": 2.1598,
+      "step": 308830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014161091657558775,
+      "loss": 2.152,
+      "step": 308835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014160923588993883,
+      "loss": 2.207,
+      "step": 308840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014160755519007542,
+      "loss": 2.336,
+      "step": 308845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014160587447599815,
+      "loss": 2.0893,
+      "step": 308850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001416041937477075,
+      "loss": 1.9451,
+      "step": 308855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014160251300520416,
+      "loss": 2.1173,
+      "step": 308860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001416008322484886,
+      "loss": 2.0965,
+      "step": 308865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001415991514775615,
+      "loss": 2.1907,
+      "step": 308870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014159747069242333,
+      "loss": 1.9255,
+      "step": 308875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014159578989307472,
+      "loss": 2.0638,
+      "step": 308880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001415941090795162,
+      "loss": 2.199,
+      "step": 308885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014159242825174844,
+      "loss": 2.0251,
+      "step": 308890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014159074740977194,
+      "loss": 2.1067,
+      "step": 308895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014158906655358728,
+      "loss": 2.0874,
+      "step": 308900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014158738568319503,
+      "loss": 2.123,
+      "step": 308905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014158570479859577,
+      "loss": 2.1167,
+      "step": 308910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001415840238997901,
+      "loss": 1.9579,
+      "step": 308915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014158234298677856,
+      "loss": 2.0642,
+      "step": 308920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014158066205956175,
+      "loss": 2.0618,
+      "step": 308925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014157898111814021,
+      "loss": 2.0968,
+      "step": 308930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014157730016251452,
+      "loss": 2.1109,
+      "step": 308935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014157561919268528,
+      "loss": 2.0617,
+      "step": 308940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001415739382086531,
+      "loss": 2.089,
+      "step": 308945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014157225721041848,
+      "loss": 2.2277,
+      "step": 308950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014157057619798203,
+      "loss": 1.9504,
+      "step": 308955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001415688951713443,
+      "loss": 1.967,
+      "step": 308960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014156721413050586,
+      "loss": 2.2124,
+      "step": 308965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014156553307546736,
+      "loss": 2.2691,
+      "step": 308970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001415638520062293,
+      "loss": 1.9947,
+      "step": 308975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001415621709227923,
+      "loss": 1.8837,
+      "step": 308980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014156048982515686,
+      "loss": 2.1334,
+      "step": 308985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014155880871332362,
+      "loss": 2.0356,
+      "step": 308990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014155712758729313,
+      "loss": 2.0894,
+      "step": 308995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000141555446447066,
+      "loss": 1.8553,
+      "step": 309000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014155376529264275,
+      "loss": 2.1614,
+      "step": 309005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000141552084124024,
+      "loss": 2.0444,
+      "step": 309010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014155040294121026,
+      "loss": 1.8577,
+      "step": 309015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014154872174420218,
+      "loss": 2.2832,
+      "step": 309020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014154704053300033,
+      "loss": 2.0673,
+      "step": 309025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001415453593076052,
+      "loss": 2.0492,
+      "step": 309030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014154367806801748,
+      "loss": 2.1357,
+      "step": 309035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014154199681423764,
+      "loss": 2.1556,
+      "step": 309040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001415403155462663,
+      "loss": 2.0067,
+      "step": 309045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014153863426410407,
+      "loss": 2.0231,
+      "step": 309050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001415369529677515,
+      "loss": 2.0981,
+      "step": 309055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014153527165720913,
+      "loss": 2.1488,
+      "step": 309060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014153359033247754,
+      "loss": 2.0406,
+      "step": 309065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014153190899355732,
+      "loss": 1.9992,
+      "step": 309070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001415302276404491,
+      "loss": 1.9487,
+      "step": 309075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00014152854627315334,
+      "loss": 1.9964,
+      "step": 309080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014152686489167073,
+      "loss": 2.0403,
+      "step": 309085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014152518349600174,
+      "loss": 2.1361,
+      "step": 309090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000141523502086147,
+      "loss": 1.8395,
+      "step": 309095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001415218206621071,
+      "loss": 1.8927,
+      "step": 309100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001415201392238826,
+      "loss": 1.8661,
+      "step": 309105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014151845777147406,
+      "loss": 2.0881,
+      "step": 309110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014151677630488204,
+      "loss": 2.1383,
+      "step": 309115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014151509482410713,
+      "loss": 2.0076,
+      "step": 309120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014151341332914994,
+      "loss": 2.0311,
+      "step": 309125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014151173182001104,
+      "loss": 2.2378,
+      "step": 309130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014151005029669093,
+      "loss": 1.9951,
+      "step": 309135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014150836875919027,
+      "loss": 2.2302,
+      "step": 309140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014150668720750958,
+      "loss": 1.9963,
+      "step": 309145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014150500564164944,
+      "loss": 2.1928,
+      "step": 309150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014150332406161044,
+      "loss": 1.9875,
+      "step": 309155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014150164246739318,
+      "loss": 2.1267,
+      "step": 309160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014149996085899817,
+      "loss": 1.9665,
+      "step": 309165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014149827923642605,
+      "loss": 2.0251,
+      "step": 309170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014149659759967736,
+      "loss": 2.0322,
+      "step": 309175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014149491594875265,
+      "loss": 2.0473,
+      "step": 309180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014149323428365256,
+      "loss": 1.8494,
+      "step": 309185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001414915526043776,
+      "loss": 1.9241,
+      "step": 309190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001414898709109284,
+      "loss": 1.9742,
+      "step": 309195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001414881892033055,
+      "loss": 1.9935,
+      "step": 309200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014148650748150947,
+      "loss": 1.9344,
+      "step": 309205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014148482574554092,
+      "loss": 2.0839,
+      "step": 309210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014148314399540037,
+      "loss": 2.1187,
+      "step": 309215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014148146223108847,
+      "loss": 2.0093,
+      "step": 309220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014147978045260572,
+      "loss": 1.9449,
+      "step": 309225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001414780986599527,
+      "loss": 2.0988,
+      "step": 309230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014147641685313006,
+      "loss": 2.0427,
+      "step": 309235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001414747350321383,
+      "loss": 2.1225,
+      "step": 309240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014147305319697802,
+      "loss": 1.9709,
+      "step": 309245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001414713713476498,
+      "loss": 1.955,
+      "step": 309250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014146968948415417,
+      "loss": 2.033,
+      "step": 309255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014146800760649178,
+      "loss": 2.0924,
+      "step": 309260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014146632571466317,
+      "loss": 1.9762,
+      "step": 309265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014146464380866892,
+      "loss": 2.1541,
+      "step": 309270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014146296188850954,
+      "loss": 1.9964,
+      "step": 309275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001414612799541857,
+      "loss": 1.9357,
+      "step": 309280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014145959800569794,
+      "loss": 2.0604,
+      "step": 309285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001414579160430468,
+      "loss": 2.0698,
+      "step": 309290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014145623406623293,
+      "loss": 2.129,
+      "step": 309295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014145455207525683,
+      "loss": 1.9774,
+      "step": 309300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001414528700701191,
+      "loss": 2.0647,
+      "step": 309305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014145118805082032,
+      "loss": 2.1046,
+      "step": 309310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014144950601736108,
+      "loss": 1.9494,
+      "step": 309315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014144782396974189,
+      "loss": 1.9881,
+      "step": 309320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014144614190796342,
+      "loss": 2.1433,
+      "step": 309325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014144445983202617,
+      "loss": 2.115,
+      "step": 309330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014144277774193076,
+      "loss": 2.0635,
+      "step": 309335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001414410956376777,
+      "loss": 2.0376,
+      "step": 309340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001414394135192677,
+      "loss": 2.0419,
+      "step": 309345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014143773138670116,
+      "loss": 2.0365,
+      "step": 309350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014143604923997879,
+      "loss": 2.0086,
+      "step": 309355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001414343670791011,
+      "loss": 2.143,
+      "step": 309360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014143268490406864,
+      "loss": 2.0886,
+      "step": 309365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014143100271488205,
+      "loss": 2.0092,
+      "step": 309370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014142932051154189,
+      "loss": 1.9458,
+      "step": 309375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014142763829404876,
+      "loss": 2.088,
+      "step": 309380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014142595606240314,
+      "loss": 2.0591,
+      "step": 309385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014142427381660567,
+      "loss": 1.7999,
+      "step": 309390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014142259155665692,
+      "loss": 2.1118,
+      "step": 309395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014142090928255746,
+      "loss": 2.0561,
+      "step": 309400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001414192269943079,
+      "loss": 1.9729,
+      "step": 309405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014141754469190873,
+      "loss": 1.9541,
+      "step": 309410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014141586237536056,
+      "loss": 2.1902,
+      "step": 309415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014141418004466407,
+      "loss": 1.9393,
+      "step": 309420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.0001414124976998197,
+      "loss": 1.9575,
+      "step": 309425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014141081534082805,
+      "loss": 2.0718,
+      "step": 309430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014140913296768973,
+      "loss": 2.1135,
+      "step": 309435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014140745058040529,
+      "loss": 2.1708,
+      "step": 309440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014140576817897534,
+      "loss": 1.9195,
+      "step": 309445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014140408576340037,
+      "loss": 2.1304,
+      "step": 309450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014140240333368107,
+      "loss": 2.1389,
+      "step": 309455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014140072088981794,
+      "loss": 2.0326,
+      "step": 309460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014139903843181156,
+      "loss": 2.0156,
+      "step": 309465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014139735595966254,
+      "loss": 2.1072,
+      "step": 309470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014139567347337144,
+      "loss": 2.0909,
+      "step": 309475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001413939909729388,
+      "loss": 2.1726,
+      "step": 309480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014139230845836525,
+      "loss": 2.0175,
+      "step": 309485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014139062592965132,
+      "loss": 2.0015,
+      "step": 309490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001413889433867976,
+      "loss": 2.0649,
+      "step": 309495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014138726082980465,
+      "loss": 2.0011,
+      "step": 309500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001413855782586731,
+      "loss": 1.9747,
+      "step": 309505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014138389567340344,
+      "loss": 2.1428,
+      "step": 309510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014138221307399636,
+      "loss": 2.1136,
+      "step": 309515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001413805304604523,
+      "loss": 2.1428,
+      "step": 309520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001413788478327719,
+      "loss": 2.1433,
+      "step": 309525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014137716519095576,
+      "loss": 1.9556,
+      "step": 309530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014137548253500442,
+      "loss": 1.9537,
+      "step": 309535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001413737998649185,
+      "loss": 1.9592,
+      "step": 309540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014137211718069848,
+      "loss": 1.9271,
+      "step": 309545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014137043448234505,
+      "loss": 2.0909,
+      "step": 309550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001413687517698587,
+      "loss": 2.1084,
+      "step": 309555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00014136706904324003,
+      "loss": 2.2522,
+      "step": 309560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014136538630248963,
+      "loss": 2.2584,
+      "step": 309565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014136370354760807,
+      "loss": 2.0604,
+      "step": 309570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001413620207785959,
+      "loss": 1.9303,
+      "step": 309575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014136033799545374,
+      "loss": 2.1378,
+      "step": 309580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014135865519818212,
+      "loss": 1.985,
+      "step": 309585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014135697238678162,
+      "loss": 2.1098,
+      "step": 309590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014135528956125287,
+      "loss": 2.078,
+      "step": 309595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00014135360672159637,
+      "loss": 2.1252,
+      "step": 309600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014135192386781272,
+      "loss": 2.1282,
+      "step": 309605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014135024099990254,
+      "loss": 2.0878,
+      "step": 309610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001413485581178663,
+      "loss": 2.2952,
+      "step": 309615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014134687522170472,
+      "loss": 2.084,
+      "step": 309620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014134519231141825,
+      "loss": 2.0577,
+      "step": 309625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014134350938700753,
+      "loss": 1.9772,
+      "step": 309630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014134182644847312,
+      "loss": 1.9254,
+      "step": 309635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001413401434958156,
+      "loss": 1.9763,
+      "step": 309640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001413384605290355,
+      "loss": 1.9914,
+      "step": 309645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014133677754813348,
+      "loss": 2.1289,
+      "step": 309650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014133509455311005,
+      "loss": 1.9459,
+      "step": 309655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014133341154396579,
+      "loss": 2.1704,
+      "step": 309660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014133172852070128,
+      "loss": 2.1741,
+      "step": 309665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014133004548331712,
+      "loss": 2.083,
+      "step": 309670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014132836243181386,
+      "loss": 2.1788,
+      "step": 309675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001413266793661921,
+      "loss": 1.8915,
+      "step": 309680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014132499628645236,
+      "loss": 2.2053,
+      "step": 309685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014132331319259526,
+      "loss": 2.3304,
+      "step": 309690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001413216300846214,
+      "loss": 2.1826,
+      "step": 309695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001413199469625313,
+      "loss": 1.9743,
+      "step": 309700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014131826382632556,
+      "loss": 2.0583,
+      "step": 309705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014131658067600473,
+      "loss": 2.0015,
+      "step": 309710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00014131489751156943,
+      "loss": 2.1044,
+      "step": 309715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014131321433302022,
+      "loss": 2.0903,
+      "step": 309720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014131153114035766,
+      "loss": 2.0721,
+      "step": 309725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014130984793358234,
+      "loss": 1.9553,
+      "step": 309730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001413081647126948,
+      "loss": 2.0734,
+      "step": 309735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014130648147769566,
+      "loss": 2.2237,
+      "step": 309740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014130479822858547,
+      "loss": 2.0596,
+      "step": 309745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014130311496536482,
+      "loss": 2.2711,
+      "step": 309750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014130143168803426,
+      "loss": 2.0312,
+      "step": 309755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001412997483965944,
+      "loss": 2.0623,
+      "step": 309760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014129806509104577,
+      "loss": 2.0215,
+      "step": 309765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000141296381771389,
+      "loss": 2.1743,
+      "step": 309770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001412946984376246,
+      "loss": 2.1605,
+      "step": 309775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014129301508975322,
+      "loss": 2.0299,
+      "step": 309780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001412913317277754,
+      "loss": 2.0316,
+      "step": 309785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001412896483516917,
+      "loss": 2.1311,
+      "step": 309790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001412879649615027,
+      "loss": 2.115,
+      "step": 309795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000141286281557209,
+      "loss": 1.9487,
+      "step": 309800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014128459813881114,
+      "loss": 2.0673,
+      "step": 309805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014128291470630975,
+      "loss": 2.0647,
+      "step": 309810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014128123125970533,
+      "loss": 2.0089,
+      "step": 309815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014127954779899848,
+      "loss": 1.9743,
+      "step": 309820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001412778643241898,
+      "loss": 2.2316,
+      "step": 309825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001412761808352799,
+      "loss": 2.3167,
+      "step": 309830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014127449733226924,
+      "loss": 1.9494,
+      "step": 309835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001412728138151585,
+      "loss": 2.0158,
+      "step": 309840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014127113028394818,
+      "loss": 2.1581,
+      "step": 309845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014126944673863894,
+      "loss": 1.9529,
+      "step": 309850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001412677631792313,
+      "loss": 2.0644,
+      "step": 309855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014126607960572585,
+      "loss": 2.0548,
+      "step": 309860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014126439601812313,
+      "loss": 2.0673,
+      "step": 309865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014126271241642378,
+      "loss": 2.0916,
+      "step": 309870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001412610288006283,
+      "loss": 2.1034,
+      "step": 309875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001412593451707373,
+      "loss": 2.1747,
+      "step": 309880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001412576615267514,
+      "loss": 2.0577,
+      "step": 309885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014125597786867112,
+      "loss": 1.9913,
+      "step": 309890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014125429419649705,
+      "loss": 2.0143,
+      "step": 309895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014125261051022977,
+      "loss": 1.8833,
+      "step": 309900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001412509268098698,
+      "loss": 2.0259,
+      "step": 309905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014124924309541782,
+      "loss": 1.8769,
+      "step": 309910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014124755936687435,
+      "loss": 2.0665,
+      "step": 309915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014124587562423994,
+      "loss": 1.8048,
+      "step": 309920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001412441918675152,
+      "loss": 2.0347,
+      "step": 309925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001412425080967007,
+      "loss": 2.0129,
+      "step": 309930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.000141240824311797,
+      "loss": 2.0882,
+      "step": 309935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014123914051280471,
+      "loss": 2.0515,
+      "step": 309940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001412374566997244,
+      "loss": 2.0215,
+      "step": 309945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001412357728725566,
+      "loss": 2.0512,
+      "step": 309950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014123408903130187,
+      "loss": 2.0495,
+      "step": 309955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014123240517596085,
+      "loss": 2.1959,
+      "step": 309960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014123072130653412,
+      "loss": 2.199,
+      "step": 309965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014122903742302224,
+      "loss": 1.8983,
+      "step": 309970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014122735352542574,
+      "loss": 1.9094,
+      "step": 309975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001412256696137452,
+      "loss": 2.0158,
+      "step": 309980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014122398568798127,
+      "loss": 2.018,
+      "step": 309985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001412223017481345,
+      "loss": 1.9216,
+      "step": 309990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014122061779420542,
+      "loss": 2.1633,
+      "step": 309995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014121893382619462,
+      "loss": 1.991,
+      "step": 310000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014121724984410264,
+      "loss": 2.0588,
+      "step": 310005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014121556584793015,
+      "loss": 2.1599,
+      "step": 310010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001412138818376777,
+      "loss": 1.926,
+      "step": 310015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014121219781334582,
+      "loss": 1.9187,
+      "step": 310020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001412105137749351,
+      "loss": 2.0023,
+      "step": 310025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001412088297224461,
+      "loss": 2.1391,
+      "step": 310030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014120714565587944,
+      "loss": 2.0153,
+      "step": 310035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014120546157523568,
+      "loss": 2.2174,
+      "step": 310040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014120377748051536,
+      "loss": 2.2453,
+      "step": 310045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001412020933717191,
+      "loss": 1.8181,
+      "step": 310050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014120040924884745,
+      "loss": 1.9952,
+      "step": 310055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.000141198725111901,
+      "loss": 2.1409,
+      "step": 310060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001411970409608803,
+      "loss": 1.9052,
+      "step": 310065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000141195356795786,
+      "loss": 2.0695,
+      "step": 310070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014119367261661856,
+      "loss": 2.08,
+      "step": 310075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014119198842337864,
+      "loss": 1.984,
+      "step": 310080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014119030421606677,
+      "loss": 2.0594,
+      "step": 310085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014118861999468357,
+      "loss": 1.9138,
+      "step": 310090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014118693575922956,
+      "loss": 2.0594,
+      "step": 310095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014118525150970537,
+      "loss": 2.0503,
+      "step": 310100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014118356724611157,
+      "loss": 1.9208,
+      "step": 310105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0001411818829684487,
+      "loss": 2.0861,
+      "step": 310110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001411801986767173,
+      "loss": 2.0737,
+      "step": 310115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014117851437091803,
+      "loss": 2.0656,
+      "step": 310120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014117683005105146,
+      "loss": 2.0797,
+      "step": 310125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014117514571711817,
+      "loss": 1.8852,
+      "step": 310130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014117346136911863,
+      "loss": 1.9806,
+      "step": 310135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014117177700705347,
+      "loss": 2.1293,
+      "step": 310140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00014117009263092336,
+      "loss": 2.0153,
+      "step": 310145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001411684082407288,
+      "loss": 1.9879,
+      "step": 310150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001411667238364703,
+      "loss": 2.013,
+      "step": 310155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014116503941814855,
+      "loss": 2.0602,
+      "step": 310160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014116335498576404,
+      "loss": 2.032,
+      "step": 310165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001411616705393174,
+      "loss": 1.9323,
+      "step": 310170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001411599860788092,
+      "loss": 2.0903,
+      "step": 310175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014115830160424,
+      "loss": 1.9449,
+      "step": 310180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014115661711561038,
+      "loss": 2.0122,
+      "step": 310185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001411549326129209,
+      "loss": 2.1732,
+      "step": 310190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014115324809617214,
+      "loss": 2.2343,
+      "step": 310195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014115156356536468,
+      "loss": 2.2054,
+      "step": 310200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014114987902049913,
+      "loss": 1.8409,
+      "step": 310205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000141148194461576,
+      "loss": 2.1267,
+      "step": 310210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014114650988859592,
+      "loss": 2.0997,
+      "step": 310215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014114482530155944,
+      "loss": 2.1108,
+      "step": 310220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014114314070046716,
+      "loss": 2.0856,
+      "step": 310225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001411414560853196,
+      "loss": 2.0321,
+      "step": 310230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014113977145611742,
+      "loss": 2.1631,
+      "step": 310235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014113808681286114,
+      "loss": 1.9568,
+      "step": 310240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014113640215555133,
+      "loss": 2.2242,
+      "step": 310245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014113471748418855,
+      "loss": 2.0336,
+      "step": 310250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014113303279877343,
+      "loss": 2.1207,
+      "step": 310255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014113134809930653,
+      "loss": 1.9371,
+      "step": 310260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014112966338578842,
+      "loss": 1.8675,
+      "step": 310265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014112797865821966,
+      "loss": 2.2609,
+      "step": 310270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001411262939166008,
+      "loss": 1.9676,
+      "step": 310275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014112460916093248,
+      "loss": 1.973,
+      "step": 310280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014112292439121527,
+      "loss": 2.1766,
+      "step": 310285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001411212396074497,
+      "loss": 1.8915,
+      "step": 310290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001411195548096364,
+      "loss": 1.8715,
+      "step": 310295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014111786999777584,
+      "loss": 2.187,
+      "step": 310300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014111618517186875,
+      "loss": 2.0308,
+      "step": 310305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001411145003319156,
+      "loss": 2.032,
+      "step": 310310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.000141112815477917,
+      "loss": 2.2433,
+      "step": 310315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001411111306098735,
+      "loss": 2.1285,
+      "step": 310320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014110944572778567,
+      "loss": 2.1333,
+      "step": 310325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014110776083165415,
+      "loss": 2.0992,
+      "step": 310330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014110607592147947,
+      "loss": 2.1908,
+      "step": 310335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014110439099726218,
+      "loss": 2.1117,
+      "step": 310340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014110270605900292,
+      "loss": 1.8922,
+      "step": 310345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001411010211067022,
+      "loss": 2.0072,
+      "step": 310350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001410993361403606,
+      "loss": 1.9878,
+      "step": 310355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014109765115997877,
+      "loss": 2.1247,
+      "step": 310360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014109596616555722,
+      "loss": 2.2126,
+      "step": 310365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014109428115709655,
+      "loss": 2.1285,
+      "step": 310370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014109259613459733,
+      "loss": 1.9566,
+      "step": 310375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014109091109806013,
+      "loss": 1.9893,
+      "step": 310380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014108922604748553,
+      "loss": 2.045,
+      "step": 310385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001410875409828741,
+      "loss": 1.9483,
+      "step": 310390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014108585590422642,
+      "loss": 1.9291,
+      "step": 310395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001410841708115431,
+      "loss": 2.2473,
+      "step": 310400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014108248570482464,
+      "loss": 2.0282,
+      "step": 310405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014108080058407166,
+      "loss": 2.1569,
+      "step": 310410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00014107911544928477,
+      "loss": 1.9637,
+      "step": 310415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014107743030046445,
+      "loss": 1.8795,
+      "step": 310420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014107574513761142,
+      "loss": 2.2491,
+      "step": 310425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001410740599607261,
+      "loss": 1.9842,
+      "step": 310430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014107237476980913,
+      "loss": 1.8908,
+      "step": 310435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001410706895648611,
+      "loss": 1.9453,
+      "step": 310440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014106900434588264,
+      "loss": 2.1839,
+      "step": 310445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001410673191128742,
+      "loss": 1.9266,
+      "step": 310450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014106563386583645,
+      "loss": 2.0553,
+      "step": 310455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001410639486047699,
+      "loss": 2.1159,
+      "step": 310460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014106226332967522,
+      "loss": 1.9159,
+      "step": 310465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014106057804055285,
+      "loss": 2.1065,
+      "step": 310470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001410588927374035,
+      "loss": 2.222,
+      "step": 310475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014105720742022766,
+      "loss": 2.1206,
+      "step": 310480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014105552208902594,
+      "loss": 1.9379,
+      "step": 310485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014105383674379893,
+      "loss": 2.1984,
+      "step": 310490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014105215138454714,
+      "loss": 2.0076,
+      "step": 310495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014105046601127122,
+      "loss": 1.9984,
+      "step": 310500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014104878062397172,
+      "loss": 2.1374,
+      "step": 310505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001410470952226492,
+      "loss": 2.0027,
+      "step": 310510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014104540980730426,
+      "loss": 2.0259,
+      "step": 310515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014104372437793745,
+      "loss": 1.9928,
+      "step": 310520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014104203893454937,
+      "loss": 2.3619,
+      "step": 310525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014104035347714056,
+      "loss": 1.7744,
+      "step": 310530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001410386680057117,
+      "loss": 1.8933,
+      "step": 310535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014103698252026318,
+      "loss": 2.1744,
+      "step": 310540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014103529702079573,
+      "loss": 2.0311,
+      "step": 310545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014103361150730987,
+      "loss": 2.0541,
+      "step": 310550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014103192597980622,
+      "loss": 2.1781,
+      "step": 310555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001410302404382853,
+      "loss": 2.1332,
+      "step": 310560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014102855488274768,
+      "loss": 2.1752,
+      "step": 310565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014102686931319398,
+      "loss": 2.1758,
+      "step": 310570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014102518372962475,
+      "loss": 1.8907,
+      "step": 310575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014102349813204058,
+      "loss": 1.9501,
+      "step": 310580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014102181252044206,
+      "loss": 1.9955,
+      "step": 310585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014102012689482974,
+      "loss": 2.1638,
+      "step": 310590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014101844125520414,
+      "loss": 2.0197,
+      "step": 310595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014101675560156593,
+      "loss": 2.1648,
+      "step": 310600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014101506993391568,
+      "loss": 1.9561,
+      "step": 310605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014101338425225396,
+      "loss": 2.0561,
+      "step": 310610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014101169855658126,
+      "loss": 2.1648,
+      "step": 310615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014101001284689825,
+      "loss": 1.9621,
+      "step": 310620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014100832712320548,
+      "loss": 2.1195,
+      "step": 310625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001410066413855035,
+      "loss": 2.0315,
+      "step": 310630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014100495563379293,
+      "loss": 1.9878,
+      "step": 310635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014100326986807432,
+      "loss": 2.0611,
+      "step": 310640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014100158408834825,
+      "loss": 2.1709,
+      "step": 310645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014099989829461528,
+      "loss": 1.9097,
+      "step": 310650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000140998212486876,
+      "loss": 2.0865,
+      "step": 310655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014099652666513102,
+      "loss": 1.8678,
+      "step": 310660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014099484082938084,
+      "loss": 2.0772,
+      "step": 310665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014099315497962612,
+      "loss": 2.0485,
+      "step": 310670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014099146911586736,
+      "loss": 1.9675,
+      "step": 310675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014098978323810518,
+      "loss": 2.2583,
+      "step": 310680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014098809734634015,
+      "loss": 2.0149,
+      "step": 310685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014098641144057286,
+      "loss": 2.1087,
+      "step": 310690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014098472552080387,
+      "loss": 2.0753,
+      "step": 310695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014098303958703373,
+      "loss": 2.1296,
+      "step": 310700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.000140981353639263,
+      "loss": 1.9939,
+      "step": 310705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014097966767749238,
+      "loss": 2.0096,
+      "step": 310710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014097798170172233,
+      "loss": 2.0288,
+      "step": 310715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014097629571195345,
+      "loss": 2.1277,
+      "step": 310720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014097460970818635,
+      "loss": 2.0787,
+      "step": 310725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014097292369042152,
+      "loss": 1.9619,
+      "step": 310730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014097123765865964,
+      "loss": 2.1489,
+      "step": 310735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014096955161290126,
+      "loss": 1.9336,
+      "step": 310740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001409678655531469,
+      "loss": 1.9862,
+      "step": 310745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001409661794793972,
+      "loss": 2.0355,
+      "step": 310750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001409644933916527,
+      "loss": 2.0412,
+      "step": 310755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000140962807289914,
+      "loss": 1.8925,
+      "step": 310760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014096112117418163,
+      "loss": 2.0775,
+      "step": 310765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001409594350444562,
+      "loss": 1.9486,
+      "step": 310770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001409577489007383,
+      "loss": 2.1687,
+      "step": 310775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014095606274302852,
+      "loss": 2.0429,
+      "step": 310780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014095437657132736,
+      "loss": 1.8943,
+      "step": 310785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014095269038563548,
+      "loss": 2.0122,
+      "step": 310790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014095100418595337,
+      "loss": 2.2209,
+      "step": 310795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001409493179722817,
+      "loss": 1.9174,
+      "step": 310800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000140947631744621,
+      "loss": 2.2101,
+      "step": 310805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001409459455029718,
+      "loss": 2.0286,
+      "step": 310810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014094425924733475,
+      "loss": 2.1121,
+      "step": 310815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001409425729777104,
+      "loss": 2.1502,
+      "step": 310820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014094088669409932,
+      "loss": 1.9417,
+      "step": 310825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001409392003965021,
+      "loss": 2.1415,
+      "step": 310830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001409375140849193,
+      "loss": 2.1452,
+      "step": 310835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001409358277593515,
+      "loss": 2.0734,
+      "step": 310840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014093414141979927,
+      "loss": 1.9049,
+      "step": 310845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014093245506626323,
+      "loss": 2.0427,
+      "step": 310850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.00014093076869874392,
+      "loss": 1.994,
+      "step": 310855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001409290823172419,
+      "loss": 1.9386,
+      "step": 310860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014092739592175773,
+      "loss": 2.1206,
+      "step": 310865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014092570951229204,
+      "loss": 2.0298,
+      "step": 310870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014092402308884543,
+      "loss": 2.2455,
+      "step": 310875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001409223366514184,
+      "loss": 2.2073,
+      "step": 310880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014092065020001156,
+      "loss": 2.0186,
+      "step": 310885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014091896373462544,
+      "loss": 2.2604,
+      "step": 310890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001409172772552607,
+      "loss": 2.0627,
+      "step": 310895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014091559076191788,
+      "loss": 1.9509,
+      "step": 310900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014091390425459755,
+      "loss": 2.0211,
+      "step": 310905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014091221773330028,
+      "loss": 2.0376,
+      "step": 310910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014091053119802665,
+      "loss": 2.1711,
+      "step": 310915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014090884464877724,
+      "loss": 1.992,
+      "step": 310920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00014090715808555264,
+      "loss": 1.953,
+      "step": 310925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001409054715083534,
+      "loss": 2.0523,
+      "step": 310930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014090378491718014,
+      "loss": 2.0938,
+      "step": 310935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014090209831203336,
+      "loss": 2.1237,
+      "step": 310940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001409004116929137,
+      "loss": 2.1056,
+      "step": 310945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014089872505982173,
+      "loss": 1.9321,
+      "step": 310950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014089703841275802,
+      "loss": 2.0175,
+      "step": 310955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014089535175172314,
+      "loss": 2.1325,
+      "step": 310960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001408936650767176,
+      "loss": 2.0983,
+      "step": 310965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014089197838774214,
+      "loss": 2.0361,
+      "step": 310970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014089029168479718,
+      "loss": 1.9938,
+      "step": 310975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014088860496788333,
+      "loss": 2.1632,
+      "step": 310980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014088691823700125,
+      "loss": 2.0497,
+      "step": 310985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014088523149215145,
+      "loss": 2.1439,
+      "step": 310990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001408835447333345,
+      "loss": 2.0631,
+      "step": 310995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014088185796055097,
+      "loss": 2.0334,
+      "step": 311000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014088017117380147,
+      "loss": 1.9015,
+      "step": 311005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014087848437308658,
+      "loss": 1.994,
+      "step": 311010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014087679755840686,
+      "loss": 1.9376,
+      "step": 311015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00014087511072976287,
+      "loss": 2.1685,
+      "step": 311020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014087342388715516,
+      "loss": 2.1558,
+      "step": 311025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014087173703058442,
+      "loss": 2.2732,
+      "step": 311030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001408700501600511,
+      "loss": 2.3617,
+      "step": 311035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014086836327555587,
+      "loss": 2.1857,
+      "step": 311040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014086667637709926,
+      "loss": 1.9589,
+      "step": 311045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001408649894646818,
+      "loss": 1.9715,
+      "step": 311050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014086330253830417,
+      "loss": 1.9203,
+      "step": 311055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014086161559796688,
+      "loss": 1.9966,
+      "step": 311060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014085992864367052,
+      "loss": 2.1574,
+      "step": 311065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00014085824167541568,
+      "loss": 1.9909,
+      "step": 311070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001408565546932029,
+      "loss": 2.2353,
+      "step": 311075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001408548676970328,
+      "loss": 2.0228,
+      "step": 311080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014085318068690595,
+      "loss": 2.0614,
+      "step": 311085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014085149366282286,
+      "loss": 2.116,
+      "step": 311090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014084980662478418,
+      "loss": 1.9909,
+      "step": 311095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001408481195727905,
+      "loss": 2.1973,
+      "step": 311100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014084643250684232,
+      "loss": 2.0732,
+      "step": 311105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014084474542694028,
+      "loss": 1.9664,
+      "step": 311110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014084305833308492,
+      "loss": 2.1132,
+      "step": 311115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014084137122527684,
+      "loss": 1.9879,
+      "step": 311120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014083968410351662,
+      "loss": 2.0762,
+      "step": 311125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014083799696780483,
+      "loss": 1.8844,
+      "step": 311130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014083630981814198,
+      "loss": 2.0325,
+      "step": 311135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014083462265452877,
+      "loss": 1.9021,
+      "step": 311140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001408329354769657,
+      "loss": 2.1062,
+      "step": 311145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014083124828545335,
+      "loss": 2.2416,
+      "step": 311150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001408295610799923,
+      "loss": 1.86,
+      "step": 311155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014082787386058312,
+      "loss": 1.7782,
+      "step": 311160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014082618662722643,
+      "loss": 2.1064,
+      "step": 311165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014082449937992278,
+      "loss": 2.2043,
+      "step": 311170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014082281211867269,
+      "loss": 1.9792,
+      "step": 311175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001408211248434768,
+      "loss": 1.9514,
+      "step": 311180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014081943755433568,
+      "loss": 2.1021,
+      "step": 311185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001408177502512499,
+      "loss": 2.0371,
+      "step": 311190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014081606293422008,
+      "loss": 2.0888,
+      "step": 311195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001408143756032467,
+      "loss": 2.1001,
+      "step": 311200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001408126882583304,
+      "loss": 2.0082,
+      "step": 311205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014081100089947173,
+      "loss": 2.1309,
+      "step": 311210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014080931352667133,
+      "loss": 2.2285,
+      "step": 311215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001408076261399297,
+      "loss": 2.0303,
+      "step": 311220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014080593873924745,
+      "loss": 2.0967,
+      "step": 311225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014080425132462514,
+      "loss": 2.0459,
+      "step": 311230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014080256389606337,
+      "loss": 1.9571,
+      "step": 311235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001408008764535627,
+      "loss": 1.984,
+      "step": 311240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001407991889971237,
+      "loss": 1.9546,
+      "step": 311245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014079750152674696,
+      "loss": 2.0141,
+      "step": 311250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014079581404243308,
+      "loss": 2.0943,
+      "step": 311255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014079412654418257,
+      "loss": 2.0708,
+      "step": 311260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014079243903199605,
+      "loss": 1.9387,
+      "step": 311265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014079075150587412,
+      "loss": 2.2126,
+      "step": 311270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001407890639658173,
+      "loss": 2.0537,
+      "step": 311275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014078737641182623,
+      "loss": 2.2569,
+      "step": 311280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014078568884390148,
+      "loss": 2.0687,
+      "step": 311285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00014078400126204353,
+      "loss": 2.13,
+      "step": 311290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014078231366625304,
+      "loss": 2.0663,
+      "step": 311295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014078062605653058,
+      "loss": 2.1355,
+      "step": 311300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014077893843287672,
+      "loss": 2.0596,
+      "step": 311305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014077725079529208,
+      "loss": 2.0044,
+      "step": 311310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014077556314377713,
+      "loss": 2.1727,
+      "step": 311315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001407738754783325,
+      "loss": 1.9928,
+      "step": 311320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001407721877989588,
+      "loss": 1.9738,
+      "step": 311325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001407705001056566,
+      "loss": 1.8597,
+      "step": 311330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.00014076881239842646,
+      "loss": 2.0824,
+      "step": 311335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014076712467726895,
+      "loss": 1.9909,
+      "step": 311340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001407654369421846,
+      "loss": 1.8706,
+      "step": 311345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001407637491931741,
+      "loss": 2.1056,
+      "step": 311350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014076206143023794,
+      "loss": 1.9862,
+      "step": 311355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001407603736533767,
+      "loss": 2.0361,
+      "step": 311360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000140758685862591,
+      "loss": 1.9183,
+      "step": 311365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001407569980578814,
+      "loss": 1.9763,
+      "step": 311370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014075531023924844,
+      "loss": 1.8788,
+      "step": 311375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014075362240669275,
+      "loss": 2.1484,
+      "step": 311380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014075193456021488,
+      "loss": 2.0132,
+      "step": 311385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001407502466998154,
+      "loss": 1.8556,
+      "step": 311390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001407485588254949,
+      "loss": 2.1365,
+      "step": 311395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014074687093725395,
+      "loss": 1.9183,
+      "step": 311400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014074518303509317,
+      "loss": 2.0257,
+      "step": 311405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014074349511901307,
+      "loss": 2.0245,
+      "step": 311410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014074180718901422,
+      "loss": 1.9688,
+      "step": 311415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014074011924509726,
+      "loss": 1.9689,
+      "step": 311420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014073843128726275,
+      "loss": 2.0354,
+      "step": 311425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014073674331551122,
+      "loss": 1.9555,
+      "step": 311430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014073505532984331,
+      "loss": 1.9233,
+      "step": 311435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014073336733025954,
+      "loss": 2.1147,
+      "step": 311440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014073167931676056,
+      "loss": 2.0311,
+      "step": 311445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014072999128934686,
+      "loss": 1.9867,
+      "step": 311450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014072830324801905,
+      "loss": 1.8945,
+      "step": 311455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014072661519277772,
+      "loss": 2.3013,
+      "step": 311460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014072492712362345,
+      "loss": 2.1123,
+      "step": 311465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001407232390405568,
+      "loss": 2.202,
+      "step": 311470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014072155094357837,
+      "loss": 2.1844,
+      "step": 311475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014071986283268869,
+      "loss": 2.0373,
+      "step": 311480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014071817470788836,
+      "loss": 2.2328,
+      "step": 311485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014071648656917803,
+      "loss": 2.1429,
+      "step": 311490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014071479841655816,
+      "loss": 1.9968,
+      "step": 311495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014071311025002937,
+      "loss": 1.9879,
+      "step": 311500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014071142206959224,
+      "loss": 2.0949,
+      "step": 311505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001407097338752474,
+      "loss": 1.9668,
+      "step": 311510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014070804566699533,
+      "loss": 1.9487,
+      "step": 311515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014070635744483666,
+      "loss": 2.0666,
+      "step": 311520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014070466920877193,
+      "loss": 1.7787,
+      "step": 311525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001407029809588018,
+      "loss": 2.0618,
+      "step": 311530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014070129269492677,
+      "loss": 1.9755,
+      "step": 311535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014069960441714745,
+      "loss": 2.0284,
+      "step": 311540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001406979161254644,
+      "loss": 1.8223,
+      "step": 311545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014069622781987822,
+      "loss": 2.0687,
+      "step": 311550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014069453950038945,
+      "loss": 2.1628,
+      "step": 311555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001406928511669987,
+      "loss": 2.0162,
+      "step": 311560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001406911628197065,
+      "loss": 2.0495,
+      "step": 311565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014068947445851351,
+      "loss": 1.9747,
+      "step": 311570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014068778608342022,
+      "loss": 2.1292,
+      "step": 311575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014068609769442728,
+      "loss": 2.1377,
+      "step": 311580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001406844092915352,
+      "loss": 2.0144,
+      "step": 311585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001406827208747446,
+      "loss": 2.072,
+      "step": 311590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014068103244405604,
+      "loss": 2.0005,
+      "step": 311595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001406793439994701,
+      "loss": 2.1253,
+      "step": 311600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014067765554098737,
+      "loss": 1.83,
+      "step": 311605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014067596706860842,
+      "loss": 1.9618,
+      "step": 311610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001406742785823338,
+      "loss": 2.073,
+      "step": 311615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014067259008216412,
+      "loss": 2.0592,
+      "step": 311620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014067090156809995,
+      "loss": 2.0906,
+      "step": 311625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00014066921304014185,
+      "loss": 2.0946,
+      "step": 311630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014066752449829042,
+      "loss": 2.1209,
+      "step": 311635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014066583594254622,
+      "loss": 2.021,
+      "step": 311640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014066414737290984,
+      "loss": 2.1149,
+      "step": 311645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014066245878938184,
+      "loss": 2.1295,
+      "step": 311650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001406607701919628,
+      "loss": 2.0626,
+      "step": 311655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014065908158065332,
+      "loss": 2.0601,
+      "step": 311660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014065739295545396,
+      "loss": 1.9598,
+      "step": 311665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001406557043163653,
+      "loss": 2.2565,
+      "step": 311670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001406540156633879,
+      "loss": 2.1406,
+      "step": 311675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014065232699652234,
+      "loss": 2.1709,
+      "step": 311680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014065063831576922,
+      "loss": 1.974,
+      "step": 311685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014064894962112913,
+      "loss": 2.16,
+      "step": 311690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001406472609126026,
+      "loss": 2.0177,
+      "step": 311695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014064557219019023,
+      "loss": 1.9297,
+      "step": 311700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001406438834538926,
+      "loss": 2.1236,
+      "step": 311705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014064219470371028,
+      "loss": 2.0604,
+      "step": 311710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014064050593964384,
+      "loss": 2.2391,
+      "step": 311715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014063881716169387,
+      "loss": 2.0703,
+      "step": 311720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014063712836986093,
+      "loss": 2.2794,
+      "step": 311725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014063543956414565,
+      "loss": 2.0923,
+      "step": 311730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014063375074454854,
+      "loss": 2.1983,
+      "step": 311735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001406320619110702,
+      "loss": 2.1389,
+      "step": 311740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014063037306371123,
+      "loss": 2.0968,
+      "step": 311745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014062868420247217,
+      "loss": 2.069,
+      "step": 311750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001406269953273536,
+      "loss": 2.099,
+      "step": 311755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014062530643835615,
+      "loss": 1.9047,
+      "step": 311760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014062361753548034,
+      "loss": 2.0767,
+      "step": 311765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001406219286187268,
+      "loss": 1.8477,
+      "step": 311770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000140620239688096,
+      "loss": 2.1138,
+      "step": 311775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001406185507435886,
+      "loss": 2.14,
+      "step": 311780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014061686178520525,
+      "loss": 2.004,
+      "step": 311785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00014061517281294637,
+      "loss": 1.9124,
+      "step": 311790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014061348382681264,
+      "loss": 2.0773,
+      "step": 311795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014061179482680457,
+      "loss": 2.0697,
+      "step": 311800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014061010581292282,
+      "loss": 2.1809,
+      "step": 311805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001406084167851679,
+      "loss": 2.0552,
+      "step": 311810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001406067277435404,
+      "loss": 2.043,
+      "step": 311815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014060503868804092,
+      "loss": 2.0821,
+      "step": 311820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014060334961867,
+      "loss": 2.0539,
+      "step": 311825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014060166053542827,
+      "loss": 2.0312,
+      "step": 311830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014059997143831625,
+      "loss": 1.8779,
+      "step": 311835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014059828232733455,
+      "loss": 2.1381,
+      "step": 311840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014059659320248376,
+      "loss": 2.0748,
+      "step": 311845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001405949040637644,
+      "loss": 2.0634,
+      "step": 311850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014059321491117715,
+      "loss": 1.9781,
+      "step": 311855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014059152574472244,
+      "loss": 2.0124,
+      "step": 311860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000140589836564401,
+      "loss": 2.0086,
+      "step": 311865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014058814737021329,
+      "loss": 2.055,
+      "step": 311870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014058645816215996,
+      "loss": 1.9749,
+      "step": 311875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014058476894024153,
+      "loss": 1.9608,
+      "step": 311880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014058307970445864,
+      "loss": 1.85,
+      "step": 311885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001405813904548118,
+      "loss": 2.3322,
+      "step": 311890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014057970119130163,
+      "loss": 1.966,
+      "step": 311895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014057801191392875,
+      "loss": 1.9992,
+      "step": 311900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014057632262269362,
+      "loss": 1.9568,
+      "step": 311905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014057463331759688,
+      "loss": 2.1665,
+      "step": 311910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014057294399863913,
+      "loss": 2.0122,
+      "step": 311915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014057125466582096,
+      "loss": 2.1264,
+      "step": 311920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001405695653191429,
+      "loss": 2.0864,
+      "step": 311925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014056787595860553,
+      "loss": 2.0736,
+      "step": 311930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014056618658420938,
+      "loss": 1.9496,
+      "step": 311935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00014056449719595517,
+      "loss": 2.0372,
+      "step": 311940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014056280779384337,
+      "loss": 1.9992,
+      "step": 311945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014056111837787454,
+      "loss": 2.2241,
+      "step": 311950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014055942894804934,
+      "loss": 1.9529,
+      "step": 311955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014055773950436828,
+      "loss": 1.9354,
+      "step": 311960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014055605004683198,
+      "loss": 2.065,
+      "step": 311965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000140554360575441,
+      "loss": 2.032,
+      "step": 311970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001405526710901959,
+      "loss": 1.9719,
+      "step": 311975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00014055098159109725,
+      "loss": 2.0001,
+      "step": 311980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001405492920781457,
+      "loss": 2.1347,
+      "step": 311985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014054760255134173,
+      "loss": 2.0949,
+      "step": 311990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000140545913010686,
+      "loss": 2.0332,
+      "step": 311995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014054422345617903,
+      "loss": 1.9674,
+      "step": 312000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014054253388782141,
+      "loss": 2.1024,
+      "step": 312005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014054084430561374,
+      "loss": 2.0017,
+      "step": 312010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001405391547095566,
+      "loss": 2.0368,
+      "step": 312015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001405374650996505,
+      "loss": 2.0388,
+      "step": 312020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014053577547589613,
+      "loss": 1.987,
+      "step": 312025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014053408583829396,
+      "loss": 2.1168,
+      "step": 312030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001405323961868446,
+      "loss": 2.0276,
+      "step": 312035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014053070652154866,
+      "loss": 2.1134,
+      "step": 312040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001405290168424067,
+      "loss": 2.1148,
+      "step": 312045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014052732714941928,
+      "loss": 1.9146,
+      "step": 312050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.000140525637442587,
+      "loss": 2.05,
+      "step": 312055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014052394772191045,
+      "loss": 1.9831,
+      "step": 312060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014052225798739015,
+      "loss": 1.999,
+      "step": 312065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001405205682390267,
+      "loss": 2.1413,
+      "step": 312070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00014051887847682073,
+      "loss": 2.0204,
+      "step": 312075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014051718870077275,
+      "loss": 1.9951,
+      "step": 312080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014051549891088338,
+      "loss": 2.1769,
+      "step": 312085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00014051380910715317,
+      "loss": 2.0164,
+      "step": 312090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001405121192895827,
+      "loss": 2.0201,
+      "step": 312095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00014051042945817258,
+      "loss": 2.0068,
+      "step": 312100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014050873961292333,
+      "loss": 1.917,
+      "step": 312105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014050704975383557,
+      "loss": 2.1057,
+      "step": 312110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001405053598809099,
+      "loss": 2.2134,
+      "step": 312115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014050366999414683,
+      "loss": 2.0821,
+      "step": 312120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014050198009354697,
+      "loss": 1.8858,
+      "step": 312125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001405002901791109,
+      "loss": 2.1362,
+      "step": 312130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014049860025083924,
+      "loss": 2.0453,
+      "step": 312135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014049691030873246,
+      "loss": 2.1956,
+      "step": 312140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014049522035279125,
+      "loss": 2.0151,
+      "step": 312145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014049353038301612,
+      "loss": 2.0936,
+      "step": 312150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014049184039940764,
+      "loss": 1.9126,
+      "step": 312155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014049015040196644,
+      "loss": 2.0582,
+      "step": 312160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014048846039069307,
+      "loss": 2.1001,
+      "step": 312165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014048677036558813,
+      "loss": 2.1094,
+      "step": 312170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014048508032665213,
+      "loss": 2.1804,
+      "step": 312175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001404833902738857,
+      "loss": 1.8663,
+      "step": 312180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014048170020728945,
+      "loss": 1.8988,
+      "step": 312185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014048001012686388,
+      "loss": 2.0792,
+      "step": 312190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014047832003260963,
+      "loss": 2.1726,
+      "step": 312195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001404766299245272,
+      "loss": 2.1858,
+      "step": 312200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014047493980261726,
+      "loss": 2.0444,
+      "step": 312205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014047324966688033,
+      "loss": 1.9588,
+      "step": 312210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00014047155951731706,
+      "loss": 1.9745,
+      "step": 312215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014046986935392793,
+      "loss": 2.0182,
+      "step": 312220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014046817917671354,
+      "loss": 2.1107,
+      "step": 312225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014046648898567447,
+      "loss": 1.9296,
+      "step": 312230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014046479878081138,
+      "loss": 1.8954,
+      "step": 312235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00014046310856212472,
+      "loss": 1.9768,
+      "step": 312240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014046141832961518,
+      "loss": 2.0274,
+      "step": 312245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014045972808328326,
+      "loss": 2.145,
+      "step": 312250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014045803782312955,
+      "loss": 2.066,
+      "step": 312255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014045634754915464,
+      "loss": 2.0586,
+      "step": 312260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014045465726135914,
+      "loss": 2.0962,
+      "step": 312265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014045296695974356,
+      "loss": 1.9921,
+      "step": 312270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014045127664430852,
+      "loss": 2.0573,
+      "step": 312275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001404495863150546,
+      "loss": 2.0269,
+      "step": 312280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014044789597198238,
+      "loss": 1.9732,
+      "step": 312285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001404462056150924,
+      "loss": 2.0074,
+      "step": 312290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014044451524438528,
+      "loss": 2.0741,
+      "step": 312295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014044282485986158,
+      "loss": 1.9177,
+      "step": 312300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00014044113446152188,
+      "loss": 2.1475,
+      "step": 312305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014043944404936674,
+      "loss": 2.0221,
+      "step": 312310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014043775362339674,
+      "loss": 1.8899,
+      "step": 312315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001404360631836125,
+      "loss": 2.1414,
+      "step": 312320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014043437273001457,
+      "loss": 1.94,
+      "step": 312325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.515625,
+      "learning_rate": 0.00014043268226260352,
+      "loss": 1.8897,
+      "step": 312330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001404309917813799,
+      "loss": 2.1275,
+      "step": 312335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014042930128634436,
+      "loss": 1.9628,
+      "step": 312340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014042761077749741,
+      "loss": 2.153,
+      "step": 312345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014042592025483968,
+      "loss": 1.9513,
+      "step": 312350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00014042422971837173,
+      "loss": 2.1037,
+      "step": 312355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001404225391680941,
+      "loss": 2.054,
+      "step": 312360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001404208486040074,
+      "loss": 2.0746,
+      "step": 312365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001404191580261122,
+      "loss": 1.9522,
+      "step": 312370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001404174674344091,
+      "loss": 2.1171,
+      "step": 312375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014041577682889867,
+      "loss": 2.0965,
+      "step": 312380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014041408620958144,
+      "loss": 2.1547,
+      "step": 312385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014041239557645803,
+      "loss": 1.9255,
+      "step": 312390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014041070492952905,
+      "loss": 2.0039,
+      "step": 312395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014040901426879505,
+      "loss": 2.0633,
+      "step": 312400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014040732359425653,
+      "loss": 2.1808,
+      "step": 312405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001404056329059142,
+      "loss": 2.0215,
+      "step": 312410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014040394220376853,
+      "loss": 2.0815,
+      "step": 312415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014040225148782018,
+      "loss": 2.1964,
+      "step": 312420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00014040056075806967,
+      "loss": 1.9954,
+      "step": 312425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014039887001451756,
+      "loss": 1.9893,
+      "step": 312430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001403971792571645,
+      "loss": 2.1127,
+      "step": 312435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014039548848601103,
+      "loss": 2.0701,
+      "step": 312440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014039379770105774,
+      "loss": 1.9469,
+      "step": 312445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00014039210690230518,
+      "loss": 2.264,
+      "step": 312450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014039041608975394,
+      "loss": 2.1374,
+      "step": 312455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014038872526340462,
+      "loss": 2.1204,
+      "step": 312460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014038703442325777,
+      "loss": 2.0804,
+      "step": 312465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00014038534356931395,
+      "loss": 1.9279,
+      "step": 312470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00014038365270157378,
+      "loss": 1.9665,
+      "step": 312475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001403819618200378,
+      "loss": 1.9452,
+      "step": 312480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014038027092470667,
+      "loss": 1.9861,
+      "step": 312485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001403785800155809,
+      "loss": 2.0643,
+      "step": 312490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000140376889092661,
+      "loss": 2.0422,
+      "step": 312495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001403751981559477,
+      "loss": 2.1028,
+      "step": 312500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014037350720544144,
+      "loss": 2.1087,
+      "step": 312505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014037181624114293,
+      "loss": 2.1124,
+      "step": 312510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00014037012526305263,
+      "loss": 2.1141,
+      "step": 312515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014036843427117117,
+      "loss": 1.9035,
+      "step": 312520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001403667432654991,
+      "loss": 2.0428,
+      "step": 312525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014036505224603705,
+      "loss": 2.0792,
+      "step": 312530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014036336121278555,
+      "loss": 2.1084,
+      "step": 312535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001403616701657452,
+      "loss": 2.0357,
+      "step": 312540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014035997910491656,
+      "loss": 1.9395,
+      "step": 312545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001403582880303002,
+      "loss": 2.0202,
+      "step": 312550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014035659694189677,
+      "loss": 2.058,
+      "step": 312555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014035490583970677,
+      "loss": 2.0859,
+      "step": 312560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001403532147237308,
+      "loss": 1.9122,
+      "step": 312565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014035152359396944,
+      "loss": 2.0424,
+      "step": 312570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014034983245042324,
+      "loss": 2.1108,
+      "step": 312575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014034814129309286,
+      "loss": 1.9044,
+      "step": 312580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001403464501219788,
+      "loss": 2.208,
+      "step": 312585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014034475893708165,
+      "loss": 2.0517,
+      "step": 312590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.000140343067738402,
+      "loss": 2.0437,
+      "step": 312595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014034137652594043,
+      "loss": 1.9969,
+      "step": 312600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001403396852996975,
+      "loss": 1.9782,
+      "step": 312605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014033799405967386,
+      "loss": 2.0565,
+      "step": 312610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014033630280586997,
+      "loss": 2.083,
+      "step": 312615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001403346115382865,
+      "loss": 2.1442,
+      "step": 312620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.000140332920256924,
+      "loss": 1.9074,
+      "step": 312625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000140331228961783,
+      "loss": 2.2662,
+      "step": 312630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014032953765286414,
+      "loss": 2.087,
+      "step": 312635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.000140327846330168,
+      "loss": 2.0689,
+      "step": 312640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014032615499369512,
+      "loss": 2.101,
+      "step": 312645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014032446364344608,
+      "loss": 2.073,
+      "step": 312650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001403227722794215,
+      "loss": 2.0987,
+      "step": 312655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001403210809016219,
+      "loss": 2.0869,
+      "step": 312660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001403193895100479,
+      "loss": 1.966,
+      "step": 312665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014031769810470007,
+      "loss": 2.0896,
+      "step": 312670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014031600668557901,
+      "loss": 2.0868,
+      "step": 312675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001403143152526852,
+      "loss": 1.9057,
+      "step": 312680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014031262380601934,
+      "loss": 2.0622,
+      "step": 312685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014031093234558197,
+      "loss": 2.0003,
+      "step": 312690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014030924087137363,
+      "loss": 2.0294,
+      "step": 312695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014030754938339492,
+      "loss": 1.9236,
+      "step": 312700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014030585788164644,
+      "loss": 2.1508,
+      "step": 312705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014030416636612872,
+      "loss": 2.0618,
+      "step": 312710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014030247483684237,
+      "loss": 2.1755,
+      "step": 312715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014030078329378798,
+      "loss": 2.0146,
+      "step": 312720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001402990917369661,
+      "loss": 2.1186,
+      "step": 312725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001402974001663773,
+      "loss": 2.0382,
+      "step": 312730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014029570858202224,
+      "loss": 1.9733,
+      "step": 312735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014029401698390135,
+      "loss": 1.9928,
+      "step": 312740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001402923253720154,
+      "loss": 2.1487,
+      "step": 312745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014029063374636474,
+      "loss": 1.9821,
+      "step": 312750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014028894210695017,
+      "loss": 2.1997,
+      "step": 312755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014028725045377212,
+      "loss": 2.1083,
+      "step": 312760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001402855587868312,
+      "loss": 2.0421,
+      "step": 312765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014028386710612803,
+      "loss": 2.1893,
+      "step": 312770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014028217541166315,
+      "loss": 2.0568,
+      "step": 312775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014028048370343715,
+      "loss": 1.8704,
+      "step": 312780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001402787919814506,
+      "loss": 2.1324,
+      "step": 312785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014027710024570408,
+      "loss": 1.9734,
+      "step": 312790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001402754084961982,
+      "loss": 2.112,
+      "step": 312795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014027371673293348,
+      "loss": 1.9841,
+      "step": 312800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00014027202495591056,
+      "loss": 2.0964,
+      "step": 312805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014027033316512996,
+      "loss": 2.1198,
+      "step": 312810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001402686413605923,
+      "loss": 2.1592,
+      "step": 312815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00014026694954229812,
+      "loss": 2.167,
+      "step": 312820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014026525771024807,
+      "loss": 2.0929,
+      "step": 312825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014026356586444263,
+      "loss": 2.1734,
+      "step": 312830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014026187400488241,
+      "loss": 2.1551,
+      "step": 312835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014026018213156804,
+      "loss": 2.0228,
+      "step": 312840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014025849024450004,
+      "loss": 2.055,
+      "step": 312845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014025679834367904,
+      "loss": 2.086,
+      "step": 312850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014025510642910556,
+      "loss": 2.0722,
+      "step": 312855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014025341450078023,
+      "loss": 1.9982,
+      "step": 312860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014025172255870358,
+      "loss": 2.1073,
+      "step": 312865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001402500306028762,
+      "loss": 2.0293,
+      "step": 312870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014024833863329872,
+      "loss": 2.2458,
+      "step": 312875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014024664664997165,
+      "loss": 2.0627,
+      "step": 312880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001402449546528956,
+      "loss": 2.1068,
+      "step": 312885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014024326264207114,
+      "loss": 1.9209,
+      "step": 312890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014024157061749885,
+      "loss": 1.9399,
+      "step": 312895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00014023987857917934,
+      "loss": 2.0432,
+      "step": 312900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001402381865271131,
+      "loss": 2.19,
+      "step": 312905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014023649446130083,
+      "loss": 2.0245,
+      "step": 312910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000140234802381743,
+      "loss": 2.2104,
+      "step": 312915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014023311028844027,
+      "loss": 1.9822,
+      "step": 312920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014023141818139313,
+      "loss": 1.9011,
+      "step": 312925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014022972606060223,
+      "loss": 2.0542,
+      "step": 312930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014022803392606814,
+      "loss": 2.0843,
+      "step": 312935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014022634177779142,
+      "loss": 2.0627,
+      "step": 312940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014022464961577265,
+      "loss": 1.8436,
+      "step": 312945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001402229574400124,
+      "loss": 1.8681,
+      "step": 312950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014022126525051125,
+      "loss": 2.0911,
+      "step": 312955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001402195730472698,
+      "loss": 2.1245,
+      "step": 312960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001402178808302886,
+      "loss": 2.0141,
+      "step": 312965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014021618859956828,
+      "loss": 2.0739,
+      "step": 312970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014021449635510938,
+      "loss": 2.023,
+      "step": 312975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014021280409691242,
+      "loss": 2.053,
+      "step": 312980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014021111182497807,
+      "loss": 2.0583,
+      "step": 312985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001402094195393069,
+      "loss": 2.0138,
+      "step": 312990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014020772723989943,
+      "loss": 1.868,
+      "step": 312995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001402060349267563,
+      "loss": 1.9877,
+      "step": 313000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00014020434259987802,
+      "loss": 1.9986,
+      "step": 313005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014020265025926524,
+      "loss": 1.9699,
+      "step": 313010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001402009579049185,
+      "loss": 2.0221,
+      "step": 313015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014019926553683837,
+      "loss": 2.0388,
+      "step": 313020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014019757315502547,
+      "loss": 1.8873,
+      "step": 313025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014019588075948032,
+      "loss": 2.1656,
+      "step": 313030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014019418835020352,
+      "loss": 1.9583,
+      "step": 313035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001401924959271957,
+      "loss": 2.0732,
+      "step": 313040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00014019080349045738,
+      "loss": 2.1857,
+      "step": 313045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00014018911103998912,
+      "loss": 2.1229,
+      "step": 313050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014018741857579157,
+      "loss": 2.0277,
+      "step": 313055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001401857260978652,
+      "loss": 2.0659,
+      "step": 313060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014018403360621076,
+      "loss": 2.1322,
+      "step": 313065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014018234110082863,
+      "loss": 2.1446,
+      "step": 313070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014018064858171956,
+      "loss": 2.0343,
+      "step": 313075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014017895604888402,
+      "loss": 2.1415,
+      "step": 313080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014017726350232262,
+      "loss": 2.2303,
+      "step": 313085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014017557094203592,
+      "loss": 2.0111,
+      "step": 313090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014017387836802453,
+      "loss": 2.172,
+      "step": 313095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00014017218578028903,
+      "loss": 1.7543,
+      "step": 313100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014017049317883,
+      "loss": 2.0292,
+      "step": 313105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00014016880056364796,
+      "loss": 2.0369,
+      "step": 313110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00014016710793474351,
+      "loss": 2.0782,
+      "step": 313115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001401654152921173,
+      "loss": 1.8096,
+      "step": 313120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014016372263576983,
+      "loss": 2.2998,
+      "step": 313125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001401620299657017,
+      "loss": 1.8842,
+      "step": 313130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001401603372819135,
+      "loss": 1.9707,
+      "step": 313135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001401586445844058,
+      "loss": 2.2389,
+      "step": 313140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014015695187317921,
+      "loss": 2.0178,
+      "step": 313145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014015525914823425,
+      "loss": 2.0391,
+      "step": 313150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014015356640957152,
+      "loss": 2.1745,
+      "step": 313155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014015187365719157,
+      "loss": 2.1091,
+      "step": 313160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014015018089109506,
+      "loss": 2.1041,
+      "step": 313165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014014848811128253,
+      "loss": 2.0675,
+      "step": 313170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001401467953177545,
+      "loss": 2.038,
+      "step": 313175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00014014510251051163,
+      "loss": 2.127,
+      "step": 313180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00014014340968955446,
+      "loss": 2.0868,
+      "step": 313185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00014014171685488357,
+      "loss": 1.9824,
+      "step": 313190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00014014002400649953,
+      "loss": 1.9973,
+      "step": 313195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014013833114440296,
+      "loss": 2.0931,
+      "step": 313200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014013663826859438,
+      "loss": 2.137,
+      "step": 313205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001401349453790744,
+      "loss": 1.973,
+      "step": 313210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00014013325247584364,
+      "loss": 2.0949,
+      "step": 313215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014013155955890257,
+      "loss": 1.8513,
+      "step": 313220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014012986662825184,
+      "loss": 2.1607,
+      "step": 313225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014012817368389203,
+      "loss": 1.9641,
+      "step": 313230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00014012648072582375,
+      "loss": 2.1066,
+      "step": 313235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001401247877540475,
+      "loss": 2.1741,
+      "step": 313240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001401230947685639,
+      "loss": 1.9266,
+      "step": 313245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001401214017693735,
+      "loss": 2.0776,
+      "step": 313250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014011970875647692,
+      "loss": 2.129,
+      "step": 313255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00014011801572987473,
+      "loss": 1.9461,
+      "step": 313260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014011632268956747,
+      "loss": 2.0272,
+      "step": 313265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00014011462963555576,
+      "loss": 2.0919,
+      "step": 313270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014011293656784014,
+      "loss": 2.0681,
+      "step": 313275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001401112434864213,
+      "loss": 1.8803,
+      "step": 313280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014010955039129964,
+      "loss": 1.8602,
+      "step": 313285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014010785728247588,
+      "loss": 2.094,
+      "step": 313290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014010616415995053,
+      "loss": 2.1274,
+      "step": 313295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001401044710237242,
+      "loss": 2.1351,
+      "step": 313300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014010277787379743,
+      "loss": 2.1861,
+      "step": 313305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014010108471017085,
+      "loss": 2.0692,
+      "step": 313310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000140099391532845,
+      "loss": 1.9614,
+      "step": 313315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00014009769834182048,
+      "loss": 2.0078,
+      "step": 313320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014009600513709784,
+      "loss": 1.8866,
+      "step": 313325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014009431191867769,
+      "loss": 1.9175,
+      "step": 313330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001400926186865606,
+      "loss": 1.9294,
+      "step": 313335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00014009092544074714,
+      "loss": 2.2064,
+      "step": 313340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.0001400892321812379,
+      "loss": 2.1082,
+      "step": 313345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014008753890803344,
+      "loss": 1.9794,
+      "step": 313350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00014008584562113434,
+      "loss": 2.0722,
+      "step": 313355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001400841523205412,
+      "loss": 2.1601,
+      "step": 313360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001400824590062546,
+      "loss": 1.9771,
+      "step": 313365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001400807656782751,
+      "loss": 2.0729,
+      "step": 313370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014007907233660327,
+      "loss": 2.0765,
+      "step": 313375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00014007737898123968,
+      "loss": 2.0954,
+      "step": 313380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00014007568561218495,
+      "loss": 2.1821,
+      "step": 313385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014007399222943967,
+      "loss": 1.7763,
+      "step": 313390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014007229883300436,
+      "loss": 2.1034,
+      "step": 313395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014007060542287964,
+      "loss": 1.9684,
+      "step": 313400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014006891199906606,
+      "loss": 2.1443,
+      "step": 313405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001400672185615642,
+      "loss": 2.1975,
+      "step": 313410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001400655251103747,
+      "loss": 1.9406,
+      "step": 313415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014006383164549805,
+      "loss": 2.1813,
+      "step": 313420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014006213816693489,
+      "loss": 2.0219,
+      "step": 313425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014006044467468576,
+      "loss": 2.0792,
+      "step": 313430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014005875116875123,
+      "loss": 2.2115,
+      "step": 313435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014005705764913196,
+      "loss": 2.028,
+      "step": 313440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014005536411582846,
+      "loss": 2.1391,
+      "step": 313445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001400536705688413,
+      "loss": 1.8622,
+      "step": 313450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014005197700817107,
+      "loss": 2.1769,
+      "step": 313455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001400502834338184,
+      "loss": 2.0014,
+      "step": 313460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001400485898457838,
+      "loss": 2.1277,
+      "step": 313465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014004689624406787,
+      "loss": 2.1285,
+      "step": 313470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001400452026286712,
+      "loss": 1.9428,
+      "step": 313475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00014004350899959436,
+      "loss": 1.9515,
+      "step": 313480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014004181535683794,
+      "loss": 2.1589,
+      "step": 313485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014004012170040252,
+      "loss": 2.0059,
+      "step": 313490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00014003842803028864,
+      "loss": 2.2311,
+      "step": 313495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001400367343464969,
+      "loss": 1.9741,
+      "step": 313500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014003504064902792,
+      "loss": 2.0913,
+      "step": 313505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00014003334693788225,
+      "loss": 1.9161,
+      "step": 313510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014003165321306043,
+      "loss": 2.0513,
+      "step": 313515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00014002995947456307,
+      "loss": 1.985,
+      "step": 313520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00014002826572239078,
+      "loss": 2.0666,
+      "step": 313525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014002657195654413,
+      "loss": 1.9619,
+      "step": 313530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001400248781770236,
+      "loss": 1.9512,
+      "step": 313535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001400231843838299,
+      "loss": 2.0893,
+      "step": 313540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001400214905769635,
+      "loss": 2.1436,
+      "step": 313545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001400197967564251,
+      "loss": 2.0474,
+      "step": 313550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001400181029222152,
+      "loss": 1.9573,
+      "step": 313555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014001640907433438,
+      "loss": 1.9812,
+      "step": 313560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00014001471521278322,
+      "loss": 2.1125,
+      "step": 313565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014001302133756228,
+      "loss": 2.1568,
+      "step": 313570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014001132744867223,
+      "loss": 2.1455,
+      "step": 313575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00014000963354611354,
+      "loss": 2.153,
+      "step": 313580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014000793962988685,
+      "loss": 2.0273,
+      "step": 313585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014000624569999271,
+      "loss": 2.2196,
+      "step": 313590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014000455175643172,
+      "loss": 1.8045,
+      "step": 313595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014000285779920446,
+      "loss": 1.9229,
+      "step": 313600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001400011638283115,
+      "loss": 1.9668,
+      "step": 313605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001399994698437534,
+      "loss": 1.9956,
+      "step": 313610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013999777584553075,
+      "loss": 2.0268,
+      "step": 313615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013999608183364414,
+      "loss": 2.0452,
+      "step": 313620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013999438780809415,
+      "loss": 1.94,
+      "step": 313625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013999269376888134,
+      "loss": 1.9379,
+      "step": 313630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013999099971600631,
+      "loss": 2.3153,
+      "step": 313635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013998930564946963,
+      "loss": 2.2154,
+      "step": 313640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001399876115692719,
+      "loss": 2.1086,
+      "step": 313645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013998591747541362,
+      "loss": 2.0073,
+      "step": 313650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013998422336789548,
+      "loss": 1.9732,
+      "step": 313655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013998252924671799,
+      "loss": 2.284,
+      "step": 313660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013998083511188172,
+      "loss": 1.9536,
+      "step": 313665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013997914096338732,
+      "loss": 2.0578,
+      "step": 313670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013997744680123526,
+      "loss": 2.0165,
+      "step": 313675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.0001399757526254262,
+      "loss": 2.1024,
+      "step": 313680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001399740584359607,
+      "loss": 1.8011,
+      "step": 313685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013997236423283937,
+      "loss": 2.0995,
+      "step": 313690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001399706700160627,
+      "loss": 2.0192,
+      "step": 313695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013996897578563139,
+      "loss": 2.0573,
+      "step": 313700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001399672815415459,
+      "loss": 2.0886,
+      "step": 313705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013996558728380688,
+      "loss": 2.2194,
+      "step": 313710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013996389301241488,
+      "loss": 1.9694,
+      "step": 313715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013996219872737048,
+      "loss": 1.9903,
+      "step": 313720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001399605044286743,
+      "loss": 1.9312,
+      "step": 313725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013995881011632686,
+      "loss": 2.1133,
+      "step": 313730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013995711579032878,
+      "loss": 2.029,
+      "step": 313735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00013995542145068063,
+      "loss": 2.197,
+      "step": 313740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013995372709738296,
+      "loss": 2.0656,
+      "step": 313745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001399520327304364,
+      "loss": 2.0695,
+      "step": 313750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013995033834984148,
+      "loss": 2.0674,
+      "step": 313755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013994864395559881,
+      "loss": 2.191,
+      "step": 313760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013994694954770894,
+      "loss": 2.0162,
+      "step": 313765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013994525512617248,
+      "loss": 2.1247,
+      "step": 313770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013994356069099001,
+      "loss": 2.0223,
+      "step": 313775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001399418662421621,
+      "loss": 2.0319,
+      "step": 313780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001399401717796893,
+      "loss": 2.0114,
+      "step": 313785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00013993847730357222,
+      "loss": 2.0603,
+      "step": 313790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013993678281381145,
+      "loss": 1.9283,
+      "step": 313795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013993508831040754,
+      "loss": 1.9392,
+      "step": 313800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013993339379336108,
+      "loss": 2.2613,
+      "step": 313805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013993169926267263,
+      "loss": 2.1119,
+      "step": 313810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001399300047183428,
+      "loss": 2.1239,
+      "step": 313815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013992831016037215,
+      "loss": 2.1857,
+      "step": 313820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001399266155887613,
+      "loss": 2.1572,
+      "step": 313825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013992492100351074,
+      "loss": 1.8742,
+      "step": 313830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013992322640462112,
+      "loss": 2.1125,
+      "step": 313835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013992153179209302,
+      "loss": 2.0013,
+      "step": 313840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013991983716592698,
+      "loss": 2.0695,
+      "step": 313845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013991814252612366,
+      "loss": 1.9703,
+      "step": 313850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013991644787268349,
+      "loss": 1.9969,
+      "step": 313855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001399147532056072,
+      "loss": 1.8635,
+      "step": 313860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013991305852489524,
+      "loss": 2.0233,
+      "step": 313865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013991136383054834,
+      "loss": 2.1323,
+      "step": 313870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013990966912256694,
+      "loss": 1.9536,
+      "step": 313875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001399079744009517,
+      "loss": 1.8464,
+      "step": 313880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013990627966570316,
+      "loss": 2.0209,
+      "step": 313885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001399045849168219,
+      "loss": 1.9495,
+      "step": 313890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013990289015430854,
+      "loss": 1.9503,
+      "step": 313895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001399011953781636,
+      "loss": 2.0021,
+      "step": 313900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001398995005883877,
+      "loss": 1.8722,
+      "step": 313905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001398978057849814,
+      "loss": 2.0861,
+      "step": 313910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013989611096794529,
+      "loss": 2.0635,
+      "step": 313915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013989441613727993,
+      "loss": 2.0665,
+      "step": 313920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013989272129298595,
+      "loss": 1.9682,
+      "step": 313925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013989102643506386,
+      "loss": 1.929,
+      "step": 313930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013988933156351428,
+      "loss": 2.0634,
+      "step": 313935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001398876366783378,
+      "loss": 1.9099,
+      "step": 313940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013988594177953494,
+      "loss": 2.2076,
+      "step": 313945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013988424686710635,
+      "loss": 2.2455,
+      "step": 313950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013988255194105257,
+      "loss": 1.9822,
+      "step": 313955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013988085700137419,
+      "loss": 1.9191,
+      "step": 313960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001398791620480718,
+      "loss": 2.119,
+      "step": 313965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013987746708114593,
+      "loss": 1.9746,
+      "step": 313970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013987577210059721,
+      "loss": 2.0094,
+      "step": 313975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001398740771064262,
+      "loss": 2.059,
+      "step": 313980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013987238209863353,
+      "loss": 2.0788,
+      "step": 313985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001398706870772197,
+      "loss": 2.1805,
+      "step": 313990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001398689920421853,
+      "loss": 2.175,
+      "step": 313995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001398672969935309,
+      "loss": 2.1906,
+      "step": 314000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013986560193125717,
+      "loss": 2.163,
+      "step": 314005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013986390685536462,
+      "loss": 1.9125,
+      "step": 314010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001398622117658538,
+      "loss": 1.9854,
+      "step": 314015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013986051666272536,
+      "loss": 2.1075,
+      "step": 314020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001398588215459798,
+      "loss": 2.1242,
+      "step": 314025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001398571264156178,
+      "loss": 2.1067,
+      "step": 314030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013985543127163986,
+      "loss": 1.8131,
+      "step": 314035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013985373611404657,
+      "loss": 2.0777,
+      "step": 314040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013985204094283852,
+      "loss": 1.912,
+      "step": 314045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001398503457580163,
+      "loss": 2.0165,
+      "step": 314050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001398486505595805,
+      "loss": 2.0494,
+      "step": 314055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013984695534753164,
+      "loss": 2.0728,
+      "step": 314060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013984526012187035,
+      "loss": 1.9224,
+      "step": 314065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013984356488259722,
+      "loss": 2.0683,
+      "step": 314070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001398418696297128,
+      "loss": 2.1094,
+      "step": 314075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013984017436321765,
+      "loss": 1.9236,
+      "step": 314080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013983847908311239,
+      "loss": 2.0429,
+      "step": 314085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013983678378939757,
+      "loss": 2.0222,
+      "step": 314090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001398350884820738,
+      "loss": 2.1696,
+      "step": 314095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013983339316114162,
+      "loss": 2.01,
+      "step": 314100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013983169782660164,
+      "loss": 1.9622,
+      "step": 314105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013983000247845445,
+      "loss": 2.0412,
+      "step": 314110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013982830711670058,
+      "loss": 2.012,
+      "step": 314115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013982661174134067,
+      "loss": 1.9221,
+      "step": 314120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013982491635237525,
+      "loss": 2.2102,
+      "step": 314125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013982322094980489,
+      "loss": 1.9848,
+      "step": 314130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00013982152553363022,
+      "loss": 1.9333,
+      "step": 314135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001398198301038518,
+      "loss": 1.9317,
+      "step": 314140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001398181346604702,
+      "loss": 2.1481,
+      "step": 314145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013981643920348602,
+      "loss": 2.0453,
+      "step": 314150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013981474373289977,
+      "loss": 1.9858,
+      "step": 314155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001398130482487121,
+      "loss": 2.1378,
+      "step": 314160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001398113527509236,
+      "loss": 1.9046,
+      "step": 314165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013980965723953484,
+      "loss": 1.94,
+      "step": 314170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013980796171454634,
+      "loss": 2.1915,
+      "step": 314175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001398062661759587,
+      "loss": 2.1093,
+      "step": 314180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013980457062377254,
+      "loss": 2.0002,
+      "step": 314185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001398028750579884,
+      "loss": 1.9389,
+      "step": 314190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013980117947860692,
+      "loss": 1.8632,
+      "step": 314195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001397994838856286,
+      "loss": 2.0259,
+      "step": 314200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013979778827905405,
+      "loss": 1.9318,
+      "step": 314205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013979609265888388,
+      "loss": 2.1204,
+      "step": 314210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013979439702511862,
+      "loss": 2.0346,
+      "step": 314215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001397927013777589,
+      "loss": 2.1593,
+      "step": 314220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013979100571680523,
+      "loss": 1.9798,
+      "step": 314225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013978931004225825,
+      "loss": 2.1925,
+      "step": 314230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013978761435411854,
+      "loss": 1.961,
+      "step": 314235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013978591865238663,
+      "loss": 1.9341,
+      "step": 314240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013978422293706315,
+      "loss": 1.9665,
+      "step": 314245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013978252720814864,
+      "loss": 2.0537,
+      "step": 314250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013978083146564372,
+      "loss": 2.1746,
+      "step": 314255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013977913570954894,
+      "loss": 2.2409,
+      "step": 314260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013977743993986484,
+      "loss": 1.9311,
+      "step": 314265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001397757441565921,
+      "loss": 2.1198,
+      "step": 314270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001397740483597312,
+      "loss": 2.0212,
+      "step": 314275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013977235254928285,
+      "loss": 1.8878,
+      "step": 314280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013977065672524748,
+      "loss": 2.0428,
+      "step": 314285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013976896088762573,
+      "loss": 2.0502,
+      "step": 314290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013976726503641816,
+      "loss": 1.8989,
+      "step": 314295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013976556917162544,
+      "loss": 1.7787,
+      "step": 314300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013976387329324802,
+      "loss": 2.144,
+      "step": 314305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013976217740128657,
+      "loss": 1.9948,
+      "step": 314310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013976048149574164,
+      "loss": 1.8543,
+      "step": 314315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013975878557661377,
+      "loss": 1.9034,
+      "step": 314320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001397570896439036,
+      "loss": 2.1288,
+      "step": 314325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001397553936976117,
+      "loss": 2.0271,
+      "step": 314330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013975369773773863,
+      "loss": 2.2458,
+      "step": 314335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.000139752001764285,
+      "loss": 2.1337,
+      "step": 314340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013975030577725132,
+      "loss": 2.1891,
+      "step": 314345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013974860977663823,
+      "loss": 1.9703,
+      "step": 314350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001397469137624463,
+      "loss": 2.1441,
+      "step": 314355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013974521773467614,
+      "loss": 2.0462,
+      "step": 314360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013974352169332822,
+      "loss": 2.1202,
+      "step": 314365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013974182563840324,
+      "loss": 2.0575,
+      "step": 314370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013974012956990172,
+      "loss": 2.0944,
+      "step": 314375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013973843348782426,
+      "loss": 2.2356,
+      "step": 314380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.25,
+      "learning_rate": 0.00013973673739217142,
+      "loss": 2.184,
+      "step": 314385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013973504128294377,
+      "loss": 2.0109,
+      "step": 314390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013973334516014196,
+      "loss": 2.0907,
+      "step": 314395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001397316490237665,
+      "loss": 2.0668,
+      "step": 314400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013972995287381794,
+      "loss": 2.04,
+      "step": 314405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013972825671029694,
+      "loss": 2.0727,
+      "step": 314410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013972656053320406,
+      "loss": 2.0683,
+      "step": 314415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013972486434253988,
+      "loss": 1.9314,
+      "step": 314420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013972316813830492,
+      "loss": 2.1087,
+      "step": 314425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013972147192049982,
+      "loss": 2.209,
+      "step": 314430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013971977568912516,
+      "loss": 1.9951,
+      "step": 314435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001397180794441815,
+      "loss": 1.9143,
+      "step": 314440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013971638318566944,
+      "loss": 1.9732,
+      "step": 314445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001397146869135895,
+      "loss": 2.1261,
+      "step": 314450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013971299062794231,
+      "loss": 2.0774,
+      "step": 314455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013971129432872848,
+      "loss": 2.1416,
+      "step": 314460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013970959801594854,
+      "loss": 2.0838,
+      "step": 314465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001397079016896031,
+      "loss": 1.9121,
+      "step": 314470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013970620534969267,
+      "loss": 2.154,
+      "step": 314475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013970450899621788,
+      "loss": 2.0071,
+      "step": 314480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013970281262917935,
+      "loss": 1.8741,
+      "step": 314485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001397011162485776,
+      "loss": 2.0016,
+      "step": 314490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013969941985441325,
+      "loss": 2.0844,
+      "step": 314495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013969772344668682,
+      "loss": 2.1758,
+      "step": 314500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013969602702539895,
+      "loss": 2.0232,
+      "step": 314505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013969433059055018,
+      "loss": 2.1213,
+      "step": 314510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013969263414214113,
+      "loss": 2.2223,
+      "step": 314515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013969093768017237,
+      "loss": 1.9846,
+      "step": 314520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013968924120464442,
+      "loss": 2.1647,
+      "step": 314525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013968754471555794,
+      "loss": 1.9769,
+      "step": 314530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013968584821291347,
+      "loss": 2.0596,
+      "step": 314535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013968415169671156,
+      "loss": 1.8695,
+      "step": 314540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013968245516695287,
+      "loss": 2.0496,
+      "step": 314545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013968075862363788,
+      "loss": 2.0411,
+      "step": 314550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001396790620667673,
+      "loss": 2.076,
+      "step": 314555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013967736549634158,
+      "loss": 2.0146,
+      "step": 314560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013967566891236137,
+      "loss": 2.139,
+      "step": 314565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00013967397231482722,
+      "loss": 1.8887,
+      "step": 314570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013967227570373973,
+      "loss": 2.1034,
+      "step": 314575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013967057907909947,
+      "loss": 1.9658,
+      "step": 314580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000139668882440907,
+      "loss": 2.127,
+      "step": 314585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013966718578916295,
+      "loss": 2.0771,
+      "step": 314590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013966548912386787,
+      "loss": 2.0594,
+      "step": 314595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013966379244502233,
+      "loss": 2.0177,
+      "step": 314600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013966209575262692,
+      "loss": 1.8708,
+      "step": 314605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013966039904668223,
+      "loss": 2.0614,
+      "step": 314610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013965870232718878,
+      "loss": 1.9389,
+      "step": 314615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013965700559414725,
+      "loss": 2.049,
+      "step": 314620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013965530884755815,
+      "loss": 2.0635,
+      "step": 314625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013965361208742206,
+      "loss": 2.0543,
+      "step": 314630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013965191531373962,
+      "loss": 1.9461,
+      "step": 314635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013965021852651135,
+      "loss": 1.9777,
+      "step": 314640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.0001396485217257378,
+      "loss": 2.1752,
+      "step": 314645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013964682491141966,
+      "loss": 2.2704,
+      "step": 314650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013964512808355744,
+      "loss": 1.9545,
+      "step": 314655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001396434312421517,
+      "loss": 2.0234,
+      "step": 314660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013964173438720301,
+      "loss": 1.9115,
+      "step": 314665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013964003751871205,
+      "loss": 2.1756,
+      "step": 314670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001396383406366793,
+      "loss": 2.0899,
+      "step": 314675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001396366437411054,
+      "loss": 2.0531,
+      "step": 314680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013963494683199087,
+      "loss": 2.006,
+      "step": 314685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013963324990933637,
+      "loss": 1.9208,
+      "step": 314690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013963155297314238,
+      "loss": 2.1718,
+      "step": 314695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013962985602340955,
+      "loss": 1.9322,
+      "step": 314700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013962815906013845,
+      "loss": 2.065,
+      "step": 314705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013962646208332967,
+      "loss": 2.1818,
+      "step": 314710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013962476509298374,
+      "loss": 2.2466,
+      "step": 314715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013962306808910127,
+      "loss": 2.09,
+      "step": 314720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013962137107168286,
+      "loss": 1.9053,
+      "step": 314725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013961967404072905,
+      "loss": 1.9092,
+      "step": 314730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013961797699624048,
+      "loss": 1.9739,
+      "step": 314735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013961627993821767,
+      "loss": 2.1788,
+      "step": 314740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013961458286666122,
+      "loss": 2.0317,
+      "step": 314745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013961288578157168,
+      "loss": 1.887,
+      "step": 314750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001396111886829497,
+      "loss": 2.1182,
+      "step": 314755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013960949157079583,
+      "loss": 1.936,
+      "step": 314760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013960779444511062,
+      "loss": 2.0762,
+      "step": 314765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013960609730589464,
+      "loss": 1.8449,
+      "step": 314770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013960440015314853,
+      "loss": 1.9138,
+      "step": 314775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013960270298687282,
+      "loss": 2.0483,
+      "step": 314780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013960100580706814,
+      "loss": 2.093,
+      "step": 314785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000139599308613735,
+      "loss": 2.0868,
+      "step": 314790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013959761140687403,
+      "loss": 2.1219,
+      "step": 314795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013959591418648582,
+      "loss": 2.191,
+      "step": 314800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013959421695257094,
+      "loss": 2.0228,
+      "step": 314805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001395925197051299,
+      "loss": 2.1035,
+      "step": 314810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013959082244416337,
+      "loss": 2.0285,
+      "step": 314815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001395891251696719,
+      "loss": 2.1068,
+      "step": 314820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001395874278816561,
+      "loss": 2.0419,
+      "step": 314825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013958573058011644,
+      "loss": 1.9987,
+      "step": 314830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013958403326505363,
+      "loss": 2.0928,
+      "step": 314835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013958233593646817,
+      "loss": 2.1372,
+      "step": 314840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013958063859436067,
+      "loss": 2.0133,
+      "step": 314845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013957894123873174,
+      "loss": 2.0618,
+      "step": 314850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013957724386958186,
+      "loss": 2.0129,
+      "step": 314855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013957554648691173,
+      "loss": 2.0593,
+      "step": 314860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013957384909072187,
+      "loss": 2.164,
+      "step": 314865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013957215168101285,
+      "loss": 2.083,
+      "step": 314870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013957045425778527,
+      "loss": 2.0168,
+      "step": 314875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001395687568210397,
+      "loss": 2.0736,
+      "step": 314880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013956705937077671,
+      "loss": 1.8651,
+      "step": 314885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013956536190699696,
+      "loss": 1.9436,
+      "step": 314890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013956366442970092,
+      "loss": 2.2618,
+      "step": 314895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013956196693888922,
+      "loss": 2.139,
+      "step": 314900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013956026943456241,
+      "loss": 2.117,
+      "step": 314905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013955857191672112,
+      "loss": 2.0124,
+      "step": 314910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013955687438536591,
+      "loss": 1.8469,
+      "step": 314915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013955517684049733,
+      "loss": 2.2393,
+      "step": 314920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000139553479282116,
+      "loss": 2.1253,
+      "step": 314925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001395517817102225,
+      "loss": 2.0219,
+      "step": 314930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013955008412481736,
+      "loss": 2.1967,
+      "step": 314935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001395483865259012,
+      "loss": 2.0986,
+      "step": 314940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013954668891347463,
+      "loss": 2.0738,
+      "step": 314945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013954499128753816,
+      "loss": 2.2301,
+      "step": 314950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001395432936480924,
+      "loss": 2.1177,
+      "step": 314955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013954159599513795,
+      "loss": 2.0039,
+      "step": 314960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013953989832867536,
+      "loss": 2.0167,
+      "step": 314965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013953820064870523,
+      "loss": 2.1491,
+      "step": 314970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013953650295522814,
+      "loss": 2.2768,
+      "step": 314975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013953480524824468,
+      "loss": 2.0108,
+      "step": 314980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001395331075277554,
+      "loss": 1.9892,
+      "step": 314985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013953140979376086,
+      "loss": 1.866,
+      "step": 314990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013952971204626171,
+      "loss": 2.0812,
+      "step": 314995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013952801428525848,
+      "loss": 2.0969,
+      "step": 315000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013952631651075178,
+      "loss": 2.0958,
+      "step": 315005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013952461872274215,
+      "loss": 1.9733,
+      "step": 315010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001395229209212302,
+      "loss": 2.3071,
+      "step": 315015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013952122310621648,
+      "loss": 2.1198,
+      "step": 315020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013951952527770166,
+      "loss": 1.9093,
+      "step": 315025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001395178274356862,
+      "loss": 1.877,
+      "step": 315030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013951612958017074,
+      "loss": 2.3177,
+      "step": 315035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00013951443171115587,
+      "loss": 2.2049,
+      "step": 315040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013951273382864214,
+      "loss": 1.9602,
+      "step": 315045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013951103593263015,
+      "loss": 2.0363,
+      "step": 315050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001395093380231205,
+      "loss": 1.8837,
+      "step": 315055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001395076401001137,
+      "loss": 1.9458,
+      "step": 315060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001395059421636104,
+      "loss": 2.1441,
+      "step": 315065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001395042442136111,
+      "loss": 2.1082,
+      "step": 315070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001395025462501165,
+      "loss": 1.8648,
+      "step": 315075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001395008482731271,
+      "loss": 2.0243,
+      "step": 315080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0,
+      "learning_rate": 0.0001394991502826435,
+      "loss": 2.0439,
+      "step": 315085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013949745227866624,
+      "loss": 2.1568,
+      "step": 315090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013949575426119593,
+      "loss": 1.976,
+      "step": 315095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001394940562302332,
+      "loss": 1.8357,
+      "step": 315100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013949235818577855,
+      "loss": 2.2017,
+      "step": 315105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001394906601278326,
+      "loss": 2.1111,
+      "step": 315110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013948896205639595,
+      "loss": 2.1258,
+      "step": 315115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013948726397146913,
+      "loss": 2.0993,
+      "step": 315120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00013948556587305272,
+      "loss": 2.061,
+      "step": 315125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00013948386776114737,
+      "loss": 2.2526,
+      "step": 315130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001394821696357536,
+      "loss": 1.9022,
+      "step": 315135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000139480471496872,
+      "loss": 2.0156,
+      "step": 315140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013947877334450316,
+      "loss": 1.9036,
+      "step": 315145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013947707517864764,
+      "loss": 2.2255,
+      "step": 315150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013947537699930603,
+      "loss": 2.1094,
+      "step": 315155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013947367880647895,
+      "loss": 2.1068,
+      "step": 315160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013947198060016696,
+      "loss": 1.9713,
+      "step": 315165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001394702823803706,
+      "loss": 2.102,
+      "step": 315170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013946858414709044,
+      "loss": 2.0824,
+      "step": 315175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001394668859003271,
+      "loss": 2.0141,
+      "step": 315180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001394651876400812,
+      "loss": 1.9078,
+      "step": 315185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013946348936635326,
+      "loss": 1.9366,
+      "step": 315190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013946179107914388,
+      "loss": 1.9278,
+      "step": 315195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013946009277845362,
+      "loss": 2.0512,
+      "step": 315200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013945839446428306,
+      "loss": 1.874,
+      "step": 315205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013945669613663286,
+      "loss": 2.0148,
+      "step": 315210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013945499779550349,
+      "loss": 2.0729,
+      "step": 315215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001394532994408956,
+      "loss": 2.0293,
+      "step": 315220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013945160107280973,
+      "loss": 2.2021,
+      "step": 315225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013944990269124647,
+      "loss": 2.0547,
+      "step": 315230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013944820429620641,
+      "loss": 2.2381,
+      "step": 315235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013944650588769016,
+      "loss": 2.1912,
+      "step": 315240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013944480746569824,
+      "loss": 2.0468,
+      "step": 315245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013944310903023127,
+      "loss": 2.1185,
+      "step": 315250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001394414105812898,
+      "loss": 2.2475,
+      "step": 315255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013943971211887447,
+      "loss": 2.1287,
+      "step": 315260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013943801364298578,
+      "loss": 2.1956,
+      "step": 315265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013943631515362433,
+      "loss": 1.9657,
+      "step": 315270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013943461665079075,
+      "loss": 2.0123,
+      "step": 315275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013943291813448564,
+      "loss": 1.9851,
+      "step": 315280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013943121960470944,
+      "loss": 2.1371,
+      "step": 315285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013942952106146285,
+      "loss": 1.8978,
+      "step": 315290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013942782250474643,
+      "loss": 2.1034,
+      "step": 315295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013942612393456078,
+      "loss": 2.2184,
+      "step": 315300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013942442535090643,
+      "loss": 1.9673,
+      "step": 315305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013942272675378394,
+      "loss": 2.184,
+      "step": 315310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013942102814319397,
+      "loss": 2.0787,
+      "step": 315315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013941932951913705,
+      "loss": 1.9061,
+      "step": 315320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013941763088161382,
+      "loss": 2.0452,
+      "step": 315325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013941593223062477,
+      "loss": 2.1896,
+      "step": 315330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001394142335661705,
+      "loss": 2.093,
+      "step": 315335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013941253488825164,
+      "loss": 2.1197,
+      "step": 315340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013941083619686874,
+      "loss": 1.9908,
+      "step": 315345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001394091374920224,
+      "loss": 2.0457,
+      "step": 315350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013940743877371317,
+      "loss": 2.1311,
+      "step": 315355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013940574004194165,
+      "loss": 2.0926,
+      "step": 315360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013940404129670836,
+      "loss": 1.9385,
+      "step": 315365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013940234253801402,
+      "loss": 1.8896,
+      "step": 315370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001394006437658591,
+      "loss": 1.9616,
+      "step": 315375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013939894498024417,
+      "loss": 2.0896,
+      "step": 315380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013939724618116988,
+      "loss": 1.949,
+      "step": 315385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013939554736863675,
+      "loss": 1.9946,
+      "step": 315390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001393938485426454,
+      "loss": 2.2353,
+      "step": 315395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013939214970319638,
+      "loss": 1.9866,
+      "step": 315400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013939045085029032,
+      "loss": 2.1228,
+      "step": 315405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013938875198392776,
+      "loss": 2.1075,
+      "step": 315410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013938705310410928,
+      "loss": 2.1405,
+      "step": 315415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013938535421083548,
+      "loss": 2.1389,
+      "step": 315420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001393836553041069,
+      "loss": 2.1343,
+      "step": 315425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013938195638392418,
+      "loss": 2.0621,
+      "step": 315430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013938025745028785,
+      "loss": 1.9646,
+      "step": 315435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013937855850319853,
+      "loss": 2.029,
+      "step": 315440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013937685954265672,
+      "loss": 2.0478,
+      "step": 315445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013937516056866314,
+      "loss": 2.0273,
+      "step": 315450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00013937346158121824,
+      "loss": 2.1401,
+      "step": 315455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001393717625803227,
+      "loss": 1.8775,
+      "step": 315460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013937006356597702,
+      "loss": 2.0684,
+      "step": 315465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013936836453818177,
+      "loss": 2.0758,
+      "step": 315470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013936666549693765,
+      "loss": 1.9324,
+      "step": 315475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013936496644224514,
+      "loss": 1.9682,
+      "step": 315480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013936326737410483,
+      "loss": 2.127,
+      "step": 315485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013936156829251733,
+      "loss": 1.9885,
+      "step": 315490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013935986919748317,
+      "loss": 1.9388,
+      "step": 315495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013935817008900298,
+      "loss": 2.1552,
+      "step": 315500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013935647096707733,
+      "loss": 2.0863,
+      "step": 315505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001393547718317068,
+      "loss": 2.1312,
+      "step": 315510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013935307268289198,
+      "loss": 1.9673,
+      "step": 315515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001393513735206334,
+      "loss": 2.1459,
+      "step": 315520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001393496743449317,
+      "loss": 2.1605,
+      "step": 315525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013934797515578742,
+      "loss": 2.1476,
+      "step": 315530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001393462759532012,
+      "loss": 1.9487,
+      "step": 315535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013934457673717355,
+      "loss": 2.1248,
+      "step": 315540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013934287750770507,
+      "loss": 2.0494,
+      "step": 315545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013934117826479636,
+      "loss": 2.1585,
+      "step": 315550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000139339479008448,
+      "loss": 1.9951,
+      "step": 315555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013933777973866053,
+      "loss": 2.0247,
+      "step": 315560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001393360804554346,
+      "loss": 2.1088,
+      "step": 315565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013933438115877074,
+      "loss": 2.0677,
+      "step": 315570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013933268184866955,
+      "loss": 2.008,
+      "step": 315575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013933098252513158,
+      "loss": 1.9474,
+      "step": 315580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001393292831881574,
+      "loss": 1.958,
+      "step": 315585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001393275838377477,
+      "loss": 2.045,
+      "step": 315590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013932588447390296,
+      "loss": 2.0961,
+      "step": 315595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013932418509662376,
+      "loss": 2.1137,
+      "step": 315600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001393224857059107,
+      "loss": 2.0209,
+      "step": 315605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013932078630176438,
+      "loss": 2.054,
+      "step": 315610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001393190868841854,
+      "loss": 2.0003,
+      "step": 315615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013931738745317428,
+      "loss": 2.0062,
+      "step": 315620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013931568800873163,
+      "loss": 2.0653,
+      "step": 315625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000139313988550858,
+      "loss": 2.0978,
+      "step": 315630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.000139312289079554,
+      "loss": 1.9378,
+      "step": 315635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013931058959482025,
+      "loss": 2.0373,
+      "step": 315640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013930889009665728,
+      "loss": 2.1094,
+      "step": 315645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013930719058506566,
+      "loss": 2.0908,
+      "step": 315650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.000139305491060046,
+      "loss": 1.9403,
+      "step": 315655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013930379152159885,
+      "loss": 1.9834,
+      "step": 315660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013930209196972486,
+      "loss": 1.8003,
+      "step": 315665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013930039240442454,
+      "loss": 2.004,
+      "step": 315670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001392986928256985,
+      "loss": 2.1475,
+      "step": 315675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001392969932335473,
+      "loss": 1.9874,
+      "step": 315680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013929529362797155,
+      "loss": 1.9749,
+      "step": 315685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013929359400897178,
+      "loss": 2.1691,
+      "step": 315690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013929189437654864,
+      "loss": 1.742,
+      "step": 315695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013929019473070265,
+      "loss": 2.1102,
+      "step": 315700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013928849507143445,
+      "loss": 2.1623,
+      "step": 315705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013928679539874456,
+      "loss": 2.0589,
+      "step": 315710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001392850957126336,
+      "loss": 1.9673,
+      "step": 315715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013928339601310214,
+      "loss": 1.8089,
+      "step": 315720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013928169630015073,
+      "loss": 1.9133,
+      "step": 315725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013927999657378,
+      "loss": 2.086,
+      "step": 315730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013927829683399055,
+      "loss": 2.2008,
+      "step": 315735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013927659708078287,
+      "loss": 1.984,
+      "step": 315740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013927489731415762,
+      "loss": 2.1221,
+      "step": 315745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013927319753411535,
+      "loss": 2.0527,
+      "step": 315750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013927149774065664,
+      "loss": 2.0521,
+      "step": 315755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001392697979337821,
+      "loss": 1.843,
+      "step": 315760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013926809811349222,
+      "loss": 2.268,
+      "step": 315765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013926639827978768,
+      "loss": 1.7456,
+      "step": 315770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000139264698432669,
+      "loss": 2.119,
+      "step": 315775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013926299857213684,
+      "loss": 2.1571,
+      "step": 315780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001392612986981917,
+      "loss": 1.9243,
+      "step": 315785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001392595988108342,
+      "loss": 1.9797,
+      "step": 315790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001392578989100649,
+      "loss": 1.9491,
+      "step": 315795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001392561989958844,
+      "loss": 2.0346,
+      "step": 315800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013925449906829326,
+      "loss": 2.0837,
+      "step": 315805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013925279912729206,
+      "loss": 2.054,
+      "step": 315810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001392510991728814,
+      "loss": 2.0093,
+      "step": 315815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013924939920506186,
+      "loss": 2.1789,
+      "step": 315820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000139247699223834,
+      "loss": 2.0648,
+      "step": 315825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013924599922919843,
+      "loss": 1.9335,
+      "step": 315830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001392442992211557,
+      "loss": 1.9979,
+      "step": 315835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001392425991997064,
+      "loss": 2.2048,
+      "step": 315840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013924089916485114,
+      "loss": 2.2043,
+      "step": 315845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001392391991165905,
+      "loss": 2.0745,
+      "step": 315850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013923749905492497,
+      "loss": 2.1054,
+      "step": 315855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013923579897985526,
+      "loss": 1.6906,
+      "step": 315860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013923409889138183,
+      "loss": 2.1495,
+      "step": 315865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001392323987895054,
+      "loss": 2.0138,
+      "step": 315870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0001392306986742264,
+      "loss": 2.1011,
+      "step": 315875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001392289985455455,
+      "loss": 2.01,
+      "step": 315880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013922729840346328,
+      "loss": 2.0033,
+      "step": 315885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001392255982479803,
+      "loss": 1.979,
+      "step": 315890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013922389807909715,
+      "loss": 2.1229,
+      "step": 315895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013922219789681438,
+      "loss": 2.1141,
+      "step": 315900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001392204977011326,
+      "loss": 1.9996,
+      "step": 315905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013921879749205238,
+      "loss": 1.9678,
+      "step": 315910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013921709726957436,
+      "loss": 1.9303,
+      "step": 315915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000139215397033699,
+      "loss": 2.1592,
+      "step": 315920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013921369678442696,
+      "loss": 2.0805,
+      "step": 315925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013921199652175883,
+      "loss": 2.1463,
+      "step": 315930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001392102962456952,
+      "loss": 2.2038,
+      "step": 315935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013920859595623658,
+      "loss": 2.0508,
+      "step": 315940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013920689565338363,
+      "loss": 2.0095,
+      "step": 315945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013920519533713684,
+      "loss": 2.0682,
+      "step": 315950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00013920349500749687,
+      "loss": 1.8277,
+      "step": 315955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001392017946644643,
+      "loss": 2.1436,
+      "step": 315960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013920009430803964,
+      "loss": 2.2354,
+      "step": 315965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013919839393822356,
+      "loss": 2.1968,
+      "step": 315970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013919669355501657,
+      "loss": 1.9707,
+      "step": 315975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013919499315841928,
+      "loss": 2.0979,
+      "step": 315980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001391932927484323,
+      "loss": 2.0286,
+      "step": 315985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013919159232505615,
+      "loss": 2.0141,
+      "step": 315990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013918989188829142,
+      "loss": 2.0362,
+      "step": 315995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013918819143813875,
+      "loss": 2.029,
+      "step": 316000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013918649097459867,
+      "loss": 1.7999,
+      "step": 316005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013918479049767178,
+      "loss": 2.0764,
+      "step": 316010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013918309000735866,
+      "loss": 1.8564,
+      "step": 316015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013918138950365987,
+      "loss": 1.9746,
+      "step": 316020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000139179688986576,
+      "loss": 2.2395,
+      "step": 316025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001391779884561077,
+      "loss": 1.988,
+      "step": 316030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013917628791225542,
+      "loss": 2.1088,
+      "step": 316035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013917458735501984,
+      "loss": 2.0341,
+      "step": 316040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001391728867844015,
+      "loss": 2.132,
+      "step": 316045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000139171186200401,
+      "loss": 1.8746,
+      "step": 316050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001391694856030189,
+      "loss": 2.1894,
+      "step": 316055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013916778499225579,
+      "loss": 2.0643,
+      "step": 316060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013916608436811224,
+      "loss": 2.1085,
+      "step": 316065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013916438373058886,
+      "loss": 2.0595,
+      "step": 316070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013916268307968626,
+      "loss": 2.0386,
+      "step": 316075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013916098241540493,
+      "loss": 2.0899,
+      "step": 316080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001391592817377455,
+      "loss": 1.98,
+      "step": 316085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013915758104670855,
+      "loss": 1.9416,
+      "step": 316090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001391558803422947,
+      "loss": 1.9844,
+      "step": 316095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013915417962450446,
+      "loss": 2.0733,
+      "step": 316100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013915247889333842,
+      "loss": 2.0596,
+      "step": 316105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001391507781487972,
+      "loss": 2.1638,
+      "step": 316110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001391490773908814,
+      "loss": 2.0841,
+      "step": 316115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013914737661959153,
+      "loss": 2.003,
+      "step": 316120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001391456758349282,
+      "loss": 2.074,
+      "step": 316125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000139143975036892,
+      "loss": 2.0774,
+      "step": 316130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013914227422548352,
+      "loss": 1.9178,
+      "step": 316135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013914057340070333,
+      "loss": 2.2371,
+      "step": 316140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000139138872562552,
+      "loss": 2.1349,
+      "step": 316145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013913717171103014,
+      "loss": 1.9333,
+      "step": 316150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013913547084613828,
+      "loss": 1.9923,
+      "step": 316155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013913376996787705,
+      "loss": 2.0615,
+      "step": 316160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013913206907624704,
+      "loss": 2.0499,
+      "step": 316165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013913036817124876,
+      "loss": 1.9598,
+      "step": 316170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013912866725288288,
+      "loss": 2.0177,
+      "step": 316175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001391269663211499,
+      "loss": 1.7614,
+      "step": 316180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013912526537605046,
+      "loss": 2.1663,
+      "step": 316185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013912356441758513,
+      "loss": 2.1345,
+      "step": 316190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013912186344575443,
+      "loss": 2.1053,
+      "step": 316195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013912016246055904,
+      "loss": 2.1117,
+      "step": 316200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013911846146199948,
+      "loss": 1.9235,
+      "step": 316205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013911676045007636,
+      "loss": 2.1242,
+      "step": 316210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013911505942479021,
+      "loss": 2.0367,
+      "step": 316215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013911335838614165,
+      "loss": 1.9718,
+      "step": 316220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001391116573341313,
+      "loss": 2.2537,
+      "step": 316225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013910995626875968,
+      "loss": 2.0181,
+      "step": 316230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001391082551900274,
+      "loss": 2.0732,
+      "step": 316235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.000139106554097935,
+      "loss": 1.949,
+      "step": 316240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001391048529924831,
+      "loss": 2.1342,
+      "step": 316245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013910315187367223,
+      "loss": 2.1391,
+      "step": 316250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001391014507415031,
+      "loss": 2.0198,
+      "step": 316255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013909974959597617,
+      "loss": 1.9632,
+      "step": 316260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013909804843709205,
+      "loss": 1.8683,
+      "step": 316265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013909634726485135,
+      "loss": 2.1695,
+      "step": 316270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001390946460792546,
+      "loss": 2.0439,
+      "step": 316275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013909294488030242,
+      "loss": 2.0876,
+      "step": 316280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013909124366799538,
+      "loss": 2.0898,
+      "step": 316285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013908954244233406,
+      "loss": 2.0102,
+      "step": 316290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013908784120331903,
+      "loss": 1.901,
+      "step": 316295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001390861399509509,
+      "loss": 2.14,
+      "step": 316300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013908443868523024,
+      "loss": 2.0668,
+      "step": 316305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013908273740615763,
+      "loss": 1.9429,
+      "step": 316310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013908103611373363,
+      "loss": 2.1515,
+      "step": 316315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013907933480795888,
+      "loss": 2.2328,
+      "step": 316320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013907763348883388,
+      "loss": 2.0226,
+      "step": 316325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013907593215635927,
+      "loss": 1.905,
+      "step": 316330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013907423081053558,
+      "loss": 1.8658,
+      "step": 316335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013907252945136346,
+      "loss": 2.0055,
+      "step": 316340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013907082807884346,
+      "loss": 2.1869,
+      "step": 316345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013906912669297614,
+      "loss": 2.1037,
+      "step": 316350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013906742529376208,
+      "loss": 2.1432,
+      "step": 316355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013906572388120193,
+      "loss": 2.2124,
+      "step": 316360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013906402245529616,
+      "loss": 1.954,
+      "step": 316365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013906232101604547,
+      "loss": 2.0774,
+      "step": 316370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013906061956345033,
+      "loss": 2.1366,
+      "step": 316375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00013905891809751138,
+      "loss": 2.0223,
+      "step": 316380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001390572166182292,
+      "loss": 1.9915,
+      "step": 316385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001390555151256044,
+      "loss": 1.9747,
+      "step": 316390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013905381361963755,
+      "loss": 2.0388,
+      "step": 316395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013905211210032915,
+      "loss": 1.9941,
+      "step": 316400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013905041056767983,
+      "loss": 1.9722,
+      "step": 316405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013904870902169022,
+      "loss": 2.1387,
+      "step": 316410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013904700746236086,
+      "loss": 2.1175,
+      "step": 316415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013904530588969231,
+      "loss": 2.2265,
+      "step": 316420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013904360430368518,
+      "loss": 2.1757,
+      "step": 316425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013904190270434008,
+      "loss": 2.1211,
+      "step": 316430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001390402010916575,
+      "loss": 2.0956,
+      "step": 316435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013903849946563814,
+      "loss": 2.0713,
+      "step": 316440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013903679782628249,
+      "loss": 1.9866,
+      "step": 316445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013903509617359117,
+      "loss": 2.0784,
+      "step": 316450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013903339450756474,
+      "loss": 2.1612,
+      "step": 316455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013903169282820382,
+      "loss": 2.0512,
+      "step": 316460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013902999113550894,
+      "loss": 2.0099,
+      "step": 316465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001390282894294807,
+      "loss": 2.0709,
+      "step": 316470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001390265877101197,
+      "loss": 2.0683,
+      "step": 316475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013902488597742653,
+      "loss": 1.9658,
+      "step": 316480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013902318423140178,
+      "loss": 2.0673,
+      "step": 316485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013902148247204593,
+      "loss": 2.0186,
+      "step": 316490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013901978069935967,
+      "loss": 1.9797,
+      "step": 316495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001390180789133435,
+      "loss": 2.1156,
+      "step": 316500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013901637711399812,
+      "loss": 2.1188,
+      "step": 316505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000139014675301324,
+      "loss": 2.0307,
+      "step": 316510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013901297347532175,
+      "loss": 2.2283,
+      "step": 316515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013901127163599197,
+      "loss": 2.2101,
+      "step": 316520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013900956978333526,
+      "loss": 1.9316,
+      "step": 316525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013900786791735213,
+      "loss": 2.1053,
+      "step": 316530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013900616603804324,
+      "loss": 1.8825,
+      "step": 316535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013900446414540912,
+      "loss": 2.1188,
+      "step": 316540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013900276223945035,
+      "loss": 2.0936,
+      "step": 316545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013900106032016754,
+      "loss": 2.0844,
+      "step": 316550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013899935838756127,
+      "loss": 2.13,
+      "step": 316555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013899765644163212,
+      "loss": 2.2154,
+      "step": 316560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013899595448238066,
+      "loss": 2.167,
+      "step": 316565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013899425250980745,
+      "loss": 2.1845,
+      "step": 316570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001389925505239131,
+      "loss": 2.103,
+      "step": 316575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013899084852469823,
+      "loss": 2.0808,
+      "step": 316580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013898914651216333,
+      "loss": 1.8303,
+      "step": 316585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013898744448630904,
+      "loss": 1.9683,
+      "step": 316590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013898574244713597,
+      "loss": 2.0691,
+      "step": 316595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013898404039464462,
+      "loss": 1.9708,
+      "step": 316600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013898233832883563,
+      "loss": 1.9077,
+      "step": 316605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013898063624970955,
+      "loss": 2.0098,
+      "step": 316610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013897893415726703,
+      "loss": 2.0773,
+      "step": 316615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013897723205150858,
+      "loss": 2.0938,
+      "step": 316620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013897552993243474,
+      "loss": 2.1138,
+      "step": 316625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001389738278000462,
+      "loss": 2.1869,
+      "step": 316630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013897212565434347,
+      "loss": 2.0545,
+      "step": 316635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013897042349532722,
+      "loss": 1.9861,
+      "step": 316640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001389687213229979,
+      "loss": 2.1374,
+      "step": 316645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013896701913735618,
+      "loss": 2.0546,
+      "step": 316650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013896531693840261,
+      "loss": 1.8563,
+      "step": 316655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013896361472613778,
+      "loss": 2.0803,
+      "step": 316660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001389619125005623,
+      "loss": 2.0351,
+      "step": 316665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001389602102616767,
+      "loss": 1.9298,
+      "step": 316670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013895850800948157,
+      "loss": 1.837,
+      "step": 316675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013895680574397754,
+      "loss": 2.1504,
+      "step": 316680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013895510346516516,
+      "loss": 1.9509,
+      "step": 316685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013895340117304502,
+      "loss": 2.175,
+      "step": 316690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013895169886761764,
+      "loss": 2.245,
+      "step": 316695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001389499965488837,
+      "loss": 2.1019,
+      "step": 316700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013894829421684372,
+      "loss": 2.1974,
+      "step": 316705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013894659187149827,
+      "loss": 2.174,
+      "step": 316710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013894488951284799,
+      "loss": 2.1208,
+      "step": 316715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013894318714089343,
+      "loss": 2.2079,
+      "step": 316720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013894148475563517,
+      "loss": 2.2999,
+      "step": 316725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013893978235707377,
+      "loss": 2.0852,
+      "step": 316730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013893807994520984,
+      "loss": 2.0331,
+      "step": 316735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000138936377520044,
+      "loss": 1.9475,
+      "step": 316740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013893467508157672,
+      "loss": 2.0546,
+      "step": 316745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001389329726298087,
+      "loss": 1.888,
+      "step": 316750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013893127016474045,
+      "loss": 1.9384,
+      "step": 316755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013892956768637256,
+      "loss": 2.1004,
+      "step": 316760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013892786519470565,
+      "loss": 2.0537,
+      "step": 316765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00013892616268974022,
+      "loss": 1.9304,
+      "step": 316770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013892446017147697,
+      "loss": 1.972,
+      "step": 316775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013892275763991642,
+      "loss": 2.2111,
+      "step": 316780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001389210550950591,
+      "loss": 2.0782,
+      "step": 316785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013891935253690566,
+      "loss": 2.2416,
+      "step": 316790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013891764996545668,
+      "loss": 2.0455,
+      "step": 316795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00013891594738071273,
+      "loss": 1.981,
+      "step": 316800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00013891424478267437,
+      "loss": 2.0419,
+      "step": 316805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001389125421713422,
+      "loss": 1.8724,
+      "step": 316810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001389108395467168,
+      "loss": 2.0298,
+      "step": 316815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013890913690879877,
+      "loss": 2.0468,
+      "step": 316820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013890743425758865,
+      "loss": 2.2007,
+      "step": 316825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013890573159308707,
+      "loss": 1.9932,
+      "step": 316830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013890402891529453,
+      "loss": 1.746,
+      "step": 316835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001389023262242117,
+      "loss": 2.0964,
+      "step": 316840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013890062351983916,
+      "loss": 2.1516,
+      "step": 316845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013889892080217746,
+      "loss": 2.0315,
+      "step": 316850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013889721807122714,
+      "loss": 1.84,
+      "step": 316855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013889551532698886,
+      "loss": 1.963,
+      "step": 316860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013889381256946315,
+      "loss": 1.9971,
+      "step": 316865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013889210979865058,
+      "loss": 2.1185,
+      "step": 316870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001388904070145518,
+      "loss": 2.0252,
+      "step": 316875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013888870421716734,
+      "loss": 2.2292,
+      "step": 316880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001388870014064978,
+      "loss": 1.9784,
+      "step": 316885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013888529858254375,
+      "loss": 2.1194,
+      "step": 316890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013888359574530578,
+      "loss": 2.0486,
+      "step": 316895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013888189289478446,
+      "loss": 1.959,
+      "step": 316900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001388801900309804,
+      "loss": 2.036,
+      "step": 316905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013887848715389412,
+      "loss": 2.0172,
+      "step": 316910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013887678426352632,
+      "loss": 1.9825,
+      "step": 316915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013887508135987744,
+      "loss": 2.2337,
+      "step": 316920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013887337844294815,
+      "loss": 1.923,
+      "step": 316925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013887167551273897,
+      "loss": 2.056,
+      "step": 316930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013886997256925057,
+      "loss": 2.2229,
+      "step": 316935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001388682696124835,
+      "loss": 2.0645,
+      "step": 316940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013886656664243823,
+      "loss": 2.1449,
+      "step": 316945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013886486365911554,
+      "loss": 2.0011,
+      "step": 316950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013886316066251585,
+      "loss": 2.1457,
+      "step": 316955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013886145765263982,
+      "loss": 2.1601,
+      "step": 316960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013885975462948801,
+      "loss": 2.2699,
+      "step": 316965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000138858051593061,
+      "loss": 2.1433,
+      "step": 316970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013885634854335935,
+      "loss": 2.128,
+      "step": 316975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001388546454803837,
+      "loss": 2.1265,
+      "step": 316980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001388529424041346,
+      "loss": 1.8645,
+      "step": 316985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013885123931461262,
+      "loss": 1.9728,
+      "step": 316990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013884953621181835,
+      "loss": 2.3124,
+      "step": 316995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00013884783309575237,
+      "loss": 1.9776,
+      "step": 317000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013884612996641527,
+      "loss": 2.123,
+      "step": 317005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013884442682380764,
+      "loss": 2.1136,
+      "step": 317010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013884272366793002,
+      "loss": 1.8866,
+      "step": 317015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013884102049878304,
+      "loss": 1.9845,
+      "step": 317020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013883931731636725,
+      "loss": 2.0884,
+      "step": 317025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013883761412068326,
+      "loss": 1.9436,
+      "step": 317030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013883591091173165,
+      "loss": 2.0157,
+      "step": 317035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013883420768951297,
+      "loss": 2.0766,
+      "step": 317040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013883250445402781,
+      "loss": 1.9534,
+      "step": 317045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013883080120527676,
+      "loss": 2.054,
+      "step": 317050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00013882909794326042,
+      "loss": 2.227,
+      "step": 317055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013882739466797937,
+      "loss": 2.1147,
+      "step": 317060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013882569137943414,
+      "loss": 1.9919,
+      "step": 317065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013882398807762538,
+      "loss": 1.8527,
+      "step": 317070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013882228476255366,
+      "loss": 2.1346,
+      "step": 317075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013882058143421949,
+      "loss": 1.8181,
+      "step": 317080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013881887809262354,
+      "loss": 2.1036,
+      "step": 317085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001388171747377663,
+      "loss": 1.9452,
+      "step": 317090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013881547136964853,
+      "loss": 2.1165,
+      "step": 317095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013881376798827058,
+      "loss": 2.068,
+      "step": 317100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001388120645936332,
+      "loss": 1.9953,
+      "step": 317105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013881036118573687,
+      "loss": 1.9464,
+      "step": 317110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013880865776458228,
+      "loss": 2.0217,
+      "step": 317115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001388069543301699,
+      "loss": 1.9806,
+      "step": 317120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013880525088250036,
+      "loss": 1.9682,
+      "step": 317125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013880354742157425,
+      "loss": 2.1195,
+      "step": 317130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00013880184394739214,
+      "loss": 2.0211,
+      "step": 317135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013880014045995463,
+      "loss": 2.0742,
+      "step": 317140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013879843695926232,
+      "loss": 2.023,
+      "step": 317145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001387967334453157,
+      "loss": 2.1795,
+      "step": 317150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013879502991811545,
+      "loss": 2.1466,
+      "step": 317155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001387933263776621,
+      "loss": 2.1975,
+      "step": 317160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013879162282395623,
+      "loss": 2.033,
+      "step": 317165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013878991925699846,
+      "loss": 1.8387,
+      "step": 317170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013878821567678935,
+      "loss": 2.0774,
+      "step": 317175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013878651208332947,
+      "loss": 2.0895,
+      "step": 317180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001387848084766194,
+      "loss": 2.0093,
+      "step": 317185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013878310485665978,
+      "loss": 2.1236,
+      "step": 317190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013878140122345112,
+      "loss": 1.9883,
+      "step": 317195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013877969757699403,
+      "loss": 2.1804,
+      "step": 317200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001387779939172891,
+      "loss": 1.835,
+      "step": 317205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013877629024433688,
+      "loss": 2.21,
+      "step": 317210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000138774586558138,
+      "loss": 2.0559,
+      "step": 317215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013877288285869303,
+      "loss": 1.9306,
+      "step": 317220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013877117914600252,
+      "loss": 2.1345,
+      "step": 317225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001387694754200671,
+      "loss": 2.0266,
+      "step": 317230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001387677716808873,
+      "loss": 2.0276,
+      "step": 317235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001387660679284637,
+      "loss": 2.1208,
+      "step": 317240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013876436416279694,
+      "loss": 2.0044,
+      "step": 317245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013876266038388759,
+      "loss": 1.8857,
+      "step": 317250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001387609565917362,
+      "loss": 2.1172,
+      "step": 317255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013875925278634333,
+      "loss": 2.0844,
+      "step": 317260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013875754896770963,
+      "loss": 1.9893,
+      "step": 317265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013875584513583562,
+      "loss": 2.0848,
+      "step": 317270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013875414129072195,
+      "loss": 1.9644,
+      "step": 317275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013875243743236915,
+      "loss": 2.0445,
+      "step": 317280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013875073356077778,
+      "loss": 1.9782,
+      "step": 317285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.0001387490296759485,
+      "loss": 2.0648,
+      "step": 317290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013874732577788182,
+      "loss": 2.1915,
+      "step": 317295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013874562186657835,
+      "loss": 2.182,
+      "step": 317300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001387439179420387,
+      "loss": 2.0657,
+      "step": 317305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001387422140042634,
+      "loss": 1.9357,
+      "step": 317310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013874051005325307,
+      "loss": 2.2048,
+      "step": 317315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013873880608900827,
+      "loss": 2.1056,
+      "step": 317320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001387371021115296,
+      "loss": 2.3252,
+      "step": 317325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001387353981208176,
+      "loss": 1.9647,
+      "step": 317330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013873369411687294,
+      "loss": 2.0688,
+      "step": 317335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013873199009969614,
+      "loss": 2.0764,
+      "step": 317340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013873028606928775,
+      "loss": 1.9801,
+      "step": 317345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001387285820256484,
+      "loss": 2.1203,
+      "step": 317350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001387268779687787,
+      "loss": 2.2275,
+      "step": 317355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013872517389867917,
+      "loss": 2.1762,
+      "step": 317360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013872346981535043,
+      "loss": 1.9096,
+      "step": 317365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013872176571879305,
+      "loss": 1.9065,
+      "step": 317370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013872006160900758,
+      "loss": 1.9721,
+      "step": 317375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013871835748599466,
+      "loss": 2.0815,
+      "step": 317380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013871665334975485,
+      "loss": 2.0882,
+      "step": 317385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00013871494920028872,
+      "loss": 2.1949,
+      "step": 317390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013871324503759687,
+      "loss": 2.1304,
+      "step": 317395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013871154086167987,
+      "loss": 2.1682,
+      "step": 317400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001387098366725383,
+      "loss": 2.1131,
+      "step": 317405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013870813247017277,
+      "loss": 2.0017,
+      "step": 317410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013870642825458384,
+      "loss": 1.9117,
+      "step": 317415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013870472402577208,
+      "loss": 2.0879,
+      "step": 317420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013870301978373805,
+      "loss": 2.0976,
+      "step": 317425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001387013155284824,
+      "loss": 2.1143,
+      "step": 317430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.21875,
+      "learning_rate": 0.0001386996112600057,
+      "loss": 2.0487,
+      "step": 317435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001386979069783085,
+      "loss": 2.0214,
+      "step": 317440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013869620268339137,
+      "loss": 2.0744,
+      "step": 317445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001386944983752549,
+      "loss": 2.0602,
+      "step": 317450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013869279405389972,
+      "loss": 2.0265,
+      "step": 317455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013869108971932637,
+      "loss": 1.7865,
+      "step": 317460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013868938537153546,
+      "loss": 1.97,
+      "step": 317465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013868768101052753,
+      "loss": 2.1675,
+      "step": 317470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001386859766363032,
+      "loss": 2.0509,
+      "step": 317475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013868427224886302,
+      "loss": 1.9905,
+      "step": 317480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001386825678482076,
+      "loss": 1.9731,
+      "step": 317485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013868086343433753,
+      "loss": 2.2212,
+      "step": 317490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013867915900725336,
+      "loss": 2.1144,
+      "step": 317495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001386774545669557,
+      "loss": 2.0202,
+      "step": 317500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013867575011344512,
+      "loss": 1.8506,
+      "step": 317505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013867404564672217,
+      "loss": 2.1337,
+      "step": 317510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013867234116678748,
+      "loss": 2.138,
+      "step": 317515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013867063667364165,
+      "loss": 1.9195,
+      "step": 317520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013866893216728522,
+      "loss": 2.0587,
+      "step": 317525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013866722764771874,
+      "loss": 1.9931,
+      "step": 317530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013866552311494288,
+      "loss": 2.2054,
+      "step": 317535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013866381856895814,
+      "loss": 2.1571,
+      "step": 317540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013866211400976515,
+      "loss": 2.2192,
+      "step": 317545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013866040943736452,
+      "loss": 2.0642,
+      "step": 317550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013865870485175674,
+      "loss": 2.0134,
+      "step": 317555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013865700025294245,
+      "loss": 2.0118,
+      "step": 317560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013865529564092223,
+      "loss": 2.1207,
+      "step": 317565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001386535910156967,
+      "loss": 2.2306,
+      "step": 317570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013865188637726637,
+      "loss": 1.8087,
+      "step": 317575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013865018172563187,
+      "loss": 2.2047,
+      "step": 317580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013864847706079373,
+      "loss": 2.0438,
+      "step": 317585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013864677238275262,
+      "loss": 1.9669,
+      "step": 317590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.000138645067691509,
+      "loss": 2.0202,
+      "step": 317595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001386433629870636,
+      "loss": 2.0662,
+      "step": 317600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001386416582694169,
+      "loss": 2.1706,
+      "step": 317605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.0001386399535385695,
+      "loss": 2.1224,
+      "step": 317610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.000138638248794522,
+      "loss": 1.973,
+      "step": 317615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013863654403727498,
+      "loss": 2.2069,
+      "step": 317620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000138634839266829,
+      "loss": 2.0361,
+      "step": 317625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013863313448318464,
+      "loss": 2.0813,
+      "step": 317630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013863142968634256,
+      "loss": 2.0996,
+      "step": 317635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013862972487630325,
+      "loss": 2.1112,
+      "step": 317640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001386280200530673,
+      "loss": 2.0318,
+      "step": 317645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013862631521663533,
+      "loss": 2.0392,
+      "step": 317650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013862461036700795,
+      "loss": 2.1042,
+      "step": 317655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013862290550418566,
+      "loss": 2.0562,
+      "step": 317660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013862120062816912,
+      "loss": 2.0858,
+      "step": 317665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013861949573895882,
+      "loss": 2.2178,
+      "step": 317670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00013861779083655543,
+      "loss": 1.9561,
+      "step": 317675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013861608592095952,
+      "loss": 2.19,
+      "step": 317680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00013861438099217166,
+      "loss": 2.1558,
+      "step": 317685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013861267605019242,
+      "loss": 1.9152,
+      "step": 317690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013861097109502235,
+      "loss": 2.0675,
+      "step": 317695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001386092661266621,
+      "loss": 1.92,
+      "step": 317700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013860756114511224,
+      "loss": 2.1675,
+      "step": 317705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001386058561503733,
+      "loss": 1.8531,
+      "step": 317710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013860415114244593,
+      "loss": 2.1271,
+      "step": 317715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013860244612133067,
+      "loss": 2.126,
+      "step": 317720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013860074108702812,
+      "loss": 2.1411,
+      "step": 317725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013859903603953889,
+      "loss": 2.035,
+      "step": 317730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013859733097886347,
+      "loss": 1.9872,
+      "step": 317735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013859562590500253,
+      "loss": 1.9374,
+      "step": 317740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001385939208179566,
+      "loss": 1.8371,
+      "step": 317745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013859221571772633,
+      "loss": 2.2557,
+      "step": 317750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013859051060431224,
+      "loss": 2.0367,
+      "step": 317755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013858880547771492,
+      "loss": 2.0,
+      "step": 317760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013858710033793498,
+      "loss": 2.1003,
+      "step": 317765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013858539518497298,
+      "loss": 2.1315,
+      "step": 317770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001385836900188295,
+      "loss": 2.096,
+      "step": 317775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013858198483950512,
+      "loss": 2.1372,
+      "step": 317780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013858027964700045,
+      "loss": 1.8471,
+      "step": 317785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001385785744413161,
+      "loss": 2.1003,
+      "step": 317790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013857686922245254,
+      "loss": 2.0179,
+      "step": 317795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013857516399041046,
+      "loss": 2.1322,
+      "step": 317800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001385734587451904,
+      "loss": 1.7456,
+      "step": 317805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013857175348679294,
+      "loss": 2.1089,
+      "step": 317810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013857004821521868,
+      "loss": 2.0292,
+      "step": 317815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013856834293046822,
+      "loss": 1.9199,
+      "step": 317820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013856663763254208,
+      "loss": 1.9573,
+      "step": 317825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013856493232144084,
+      "loss": 1.999,
+      "step": 317830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013856322699716515,
+      "loss": 2.193,
+      "step": 317835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013856152165971558,
+      "loss": 2.1662,
+      "step": 317840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013855981630909268,
+      "loss": 1.9857,
+      "step": 317845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013855811094529708,
+      "loss": 2.0123,
+      "step": 317850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013855640556832927,
+      "loss": 1.9876,
+      "step": 317855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001385547001781899,
+      "loss": 2.2317,
+      "step": 317860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001385529947748796,
+      "loss": 2.1368,
+      "step": 317865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013855128935839886,
+      "loss": 1.9063,
+      "step": 317870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013854958392874834,
+      "loss": 2.0874,
+      "step": 317875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013854787848592852,
+      "loss": 1.9904,
+      "step": 317880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001385461730299401,
+      "loss": 2.2234,
+      "step": 317885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013854446756078357,
+      "loss": 1.9563,
+      "step": 317890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013854276207845954,
+      "loss": 1.8418,
+      "step": 317895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00013854105658296864,
+      "loss": 2.1225,
+      "step": 317900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001385393510743114,
+      "loss": 2.1857,
+      "step": 317905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013853764555248844,
+      "loss": 2.0665,
+      "step": 317910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001385359400175003,
+      "loss": 2.0824,
+      "step": 317915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013853423446934755,
+      "loss": 2.0119,
+      "step": 317920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001385325289080309,
+      "loss": 1.966,
+      "step": 317925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013853082333355077,
+      "loss": 2.0578,
+      "step": 317930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013852911774590783,
+      "loss": 2.0004,
+      "step": 317935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013852741214510263,
+      "loss": 2.1673,
+      "step": 317940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013852570653113578,
+      "loss": 2.1706,
+      "step": 317945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013852400090400783,
+      "loss": 2.1153,
+      "step": 317950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001385222952637194,
+      "loss": 2.1854,
+      "step": 317955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013852058961027108,
+      "loss": 2.0686,
+      "step": 317960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001385188839436634,
+      "loss": 2.1313,
+      "step": 317965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013851717826389697,
+      "loss": 1.8341,
+      "step": 317970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013851547257097237,
+      "loss": 2.0078,
+      "step": 317975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001385137668648902,
+      "loss": 1.9002,
+      "step": 317980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013851206114565102,
+      "loss": 2.0386,
+      "step": 317985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001385103554132554,
+      "loss": 2.1533,
+      "step": 317990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000138508649667704,
+      "loss": 2.3078,
+      "step": 317995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013850694390899732,
+      "loss": 1.9827,
+      "step": 318000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013850523813713596,
+      "loss": 2.2097,
+      "step": 318005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013850353235212053,
+      "loss": 2.0922,
+      "step": 318010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013850182655395155,
+      "loss": 2.0487,
+      "step": 318015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013850012074262969,
+      "loss": 2.0614,
+      "step": 318020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001384984149181555,
+      "loss": 2.0094,
+      "step": 318025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001384967090805295,
+      "loss": 2.2159,
+      "step": 318030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013849500322975238,
+      "loss": 2.1334,
+      "step": 318035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013849329736582465,
+      "loss": 1.9564,
+      "step": 318040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001384915914887469,
+      "loss": 2.1807,
+      "step": 318045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013848988559851975,
+      "loss": 2.0922,
+      "step": 318050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013848817969514373,
+      "loss": 1.9175,
+      "step": 318055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013848647377861947,
+      "loss": 2.1604,
+      "step": 318060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013848476784894754,
+      "loss": 1.9563,
+      "step": 318065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013848306190612848,
+      "loss": 2.0717,
+      "step": 318070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001384813559501629,
+      "loss": 1.9844,
+      "step": 318075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013847964998105146,
+      "loss": 2.0258,
+      "step": 318080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001384779439987946,
+      "loss": 1.9278,
+      "step": 318085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013847623800339302,
+      "loss": 2.0004,
+      "step": 318090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013847453199484724,
+      "loss": 2.1593,
+      "step": 318095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00013847282597315787,
+      "loss": 2.0258,
+      "step": 318100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00013847111993832548,
+      "loss": 1.9866,
+      "step": 318105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013846941389035065,
+      "loss": 2.1177,
+      "step": 318110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013846770782923403,
+      "loss": 1.946,
+      "step": 318115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001384660017549761,
+      "loss": 1.981,
+      "step": 318120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013846429566757745,
+      "loss": 2.1219,
+      "step": 318125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013846258956703872,
+      "loss": 2.0925,
+      "step": 318130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001384608834533605,
+      "loss": 2.0827,
+      "step": 318135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013845917732654335,
+      "loss": 2.0407,
+      "step": 318140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001384574711865878,
+      "loss": 2.0689,
+      "step": 318145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001384557650334945,
+      "loss": 2.1043,
+      "step": 318150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013845405886726402,
+      "loss": 2.0402,
+      "step": 318155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013845235268789696,
+      "loss": 2.1804,
+      "step": 318160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013845064649539384,
+      "loss": 2.091,
+      "step": 318165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001384489402897553,
+      "loss": 1.9544,
+      "step": 318170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013844723407098187,
+      "loss": 2.0141,
+      "step": 318175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00013844552783907424,
+      "loss": 2.1105,
+      "step": 318180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013844382159403285,
+      "loss": 1.9513,
+      "step": 318185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001384421153358584,
+      "loss": 2.1272,
+      "step": 318190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001384404090645514,
+      "loss": 2.0499,
+      "step": 318195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013843870278011248,
+      "loss": 2.1205,
+      "step": 318200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013843699648254218,
+      "loss": 2.1011,
+      "step": 318205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001384352901718411,
+      "loss": 2.1703,
+      "step": 318210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013843358384800985,
+      "loss": 2.003,
+      "step": 318215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000138431877511049,
+      "loss": 2.1327,
+      "step": 318220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001384301711609591,
+      "loss": 1.8119,
+      "step": 318225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013842846479774075,
+      "loss": 1.9827,
+      "step": 318230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013842675842139457,
+      "loss": 2.0111,
+      "step": 318235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013842505203192108,
+      "loss": 2.1757,
+      "step": 318240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001384233456293209,
+      "loss": 2.278,
+      "step": 318245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013842163921359463,
+      "loss": 2.2035,
+      "step": 318250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013841993278474282,
+      "loss": 2.1811,
+      "step": 318255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.00013841822634276608,
+      "loss": 1.9101,
+      "step": 318260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013841651988766494,
+      "loss": 2.1333,
+      "step": 318265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013841481341944005,
+      "loss": 2.24,
+      "step": 318270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013841310693809197,
+      "loss": 1.9648,
+      "step": 318275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013841140044362126,
+      "loss": 1.9709,
+      "step": 318280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001384096939360285,
+      "loss": 1.9658,
+      "step": 318285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001384079874153143,
+      "loss": 2.0171,
+      "step": 318290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013840628088147928,
+      "loss": 1.8423,
+      "step": 318295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013840457433452393,
+      "loss": 2.0421,
+      "step": 318300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013840286777444888,
+      "loss": 1.9461,
+      "step": 318305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013840116120125473,
+      "loss": 1.8834,
+      "step": 318310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013839945461494203,
+      "loss": 2.0081,
+      "step": 318315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001383977480155114,
+      "loss": 1.9661,
+      "step": 318320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013839604140296344,
+      "loss": 1.8342,
+      "step": 318325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013839433477729865,
+      "loss": 2.1288,
+      "step": 318330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013839262813851767,
+      "loss": 2.0489,
+      "step": 318335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013839092148662104,
+      "loss": 2.0996,
+      "step": 318340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013838921482160941,
+      "loss": 1.9158,
+      "step": 318345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001383875081434833,
+      "loss": 2.0101,
+      "step": 318350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013838580145224337,
+      "loss": 2.1227,
+      "step": 318355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013838409474789008,
+      "loss": 2.1693,
+      "step": 318360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013838238803042416,
+      "loss": 2.204,
+      "step": 318365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001383806812998461,
+      "loss": 2.016,
+      "step": 318370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013837897455615646,
+      "loss": 1.9447,
+      "step": 318375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013837726779935593,
+      "loss": 2.1334,
+      "step": 318380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013837556102944498,
+      "loss": 2.0582,
+      "step": 318385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013837385424642426,
+      "loss": 2.1205,
+      "step": 318390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013837214745029435,
+      "loss": 1.7432,
+      "step": 318395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001383704406410558,
+      "loss": 2.1725,
+      "step": 318400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001383687338187092,
+      "loss": 1.9749,
+      "step": 318405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001383670269832552,
+      "loss": 2.1955,
+      "step": 318410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013836532013469427,
+      "loss": 1.9412,
+      "step": 318415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013836361327302705,
+      "loss": 2.0904,
+      "step": 318420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013836190639825415,
+      "loss": 2.2316,
+      "step": 318425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00013836019951037613,
+      "loss": 2.1608,
+      "step": 318430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013835849260939356,
+      "loss": 2.0408,
+      "step": 318435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013835678569530706,
+      "loss": 1.9904,
+      "step": 318440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013835507876811714,
+      "loss": 1.8573,
+      "step": 318445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013835337182782444,
+      "loss": 1.9193,
+      "step": 318450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013835166487442956,
+      "loss": 2.1091,
+      "step": 318455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013834995790793303,
+      "loss": 2.0739,
+      "step": 318460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013834825092833548,
+      "loss": 1.9448,
+      "step": 318465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013834654393563744,
+      "loss": 2.2009,
+      "step": 318470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013834483692983955,
+      "loss": 2.0936,
+      "step": 318475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013834312991094238,
+      "loss": 2.113,
+      "step": 318480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013834142287894648,
+      "loss": 2.0506,
+      "step": 318485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013833971583385244,
+      "loss": 2.1353,
+      "step": 318490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001383380087756609,
+      "loss": 2.2414,
+      "step": 318495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013833630170437237,
+      "loss": 2.0656,
+      "step": 318500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013833459461998747,
+      "loss": 2.2656,
+      "step": 318505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013833288752250678,
+      "loss": 2.1197,
+      "step": 318510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001383311804119309,
+      "loss": 2.0764,
+      "step": 318515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013832947328826036,
+      "loss": 1.9672,
+      "step": 318520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013832776615149578,
+      "loss": 1.8989,
+      "step": 318525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013832605900163775,
+      "loss": 1.9773,
+      "step": 318530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013832435183868685,
+      "loss": 1.9141,
+      "step": 318535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013832264466264363,
+      "loss": 2.1038,
+      "step": 318540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013832093747350875,
+      "loss": 2.1511,
+      "step": 318545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001383192302712827,
+      "loss": 2.0235,
+      "step": 318550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013831752305596613,
+      "loss": 2.2055,
+      "step": 318555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013831581582755958,
+      "loss": 2.0329,
+      "step": 318560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013831410858606365,
+      "loss": 2.1101,
+      "step": 318565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00013831240133147895,
+      "loss": 1.9406,
+      "step": 318570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00013831069406380602,
+      "loss": 2.0938,
+      "step": 318575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013830898678304546,
+      "loss": 2.1094,
+      "step": 318580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013830727948919785,
+      "loss": 2.0772,
+      "step": 318585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001383055721822638,
+      "loss": 2.1595,
+      "step": 318590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013830386486224388,
+      "loss": 2.0825,
+      "step": 318595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013830215752913864,
+      "loss": 1.9523,
+      "step": 318600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013830045018294867,
+      "loss": 2.2017,
+      "step": 318605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001382987428236746,
+      "loss": 2.1229,
+      "step": 318610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000138297035451317,
+      "loss": 2.0518,
+      "step": 318615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013829532806587642,
+      "loss": 2.0516,
+      "step": 318620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013829362066735348,
+      "loss": 2.0879,
+      "step": 318625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001382919132557487,
+      "loss": 2.0643,
+      "step": 318630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013829020583106273,
+      "loss": 2.0672,
+      "step": 318635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013828849839329614,
+      "loss": 2.07,
+      "step": 318640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001382867909424495,
+      "loss": 2.0047,
+      "step": 318645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001382850834785234,
+      "loss": 2.0154,
+      "step": 318650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013828337600151842,
+      "loss": 1.9683,
+      "step": 318655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013828166851143515,
+      "loss": 1.8491,
+      "step": 318660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013827996100827417,
+      "loss": 1.9504,
+      "step": 318665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013827825349203602,
+      "loss": 1.8281,
+      "step": 318670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013827654596272138,
+      "loss": 2.0009,
+      "step": 318675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013827483842033075,
+      "loss": 1.8745,
+      "step": 318680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013827313086486474,
+      "loss": 2.0289,
+      "step": 318685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013827142329632394,
+      "loss": 2.2325,
+      "step": 318690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013826971571470893,
+      "loss": 2.0055,
+      "step": 318695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013826800812002027,
+      "loss": 2.0224,
+      "step": 318700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001382663005122586,
+      "loss": 2.1821,
+      "step": 318705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013826459289142446,
+      "loss": 2.1042,
+      "step": 318710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013826288525751844,
+      "loss": 2.0231,
+      "step": 318715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013826117761054108,
+      "loss": 2.0111,
+      "step": 318720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013825946995049305,
+      "loss": 2.1284,
+      "step": 318725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001382577622773749,
+      "loss": 2.0302,
+      "step": 318730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001382560545911872,
+      "loss": 2.0889,
+      "step": 318735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001382543468919305,
+      "loss": 2.1489,
+      "step": 318740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013825263917960546,
+      "loss": 2.0214,
+      "step": 318745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013825093145421262,
+      "loss": 2.0021,
+      "step": 318750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013824922371575259,
+      "loss": 2.1251,
+      "step": 318755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013824751596422586,
+      "loss": 2.1591,
+      "step": 318760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013824580819963314,
+      "loss": 1.976,
+      "step": 318765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013824410042197495,
+      "loss": 2.1862,
+      "step": 318770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013824239263125187,
+      "loss": 2.0558,
+      "step": 318775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013824068482746452,
+      "loss": 1.841,
+      "step": 318780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013823897701061344,
+      "loss": 1.8765,
+      "step": 318785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013823726918069923,
+      "loss": 2.0366,
+      "step": 318790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013823556133772246,
+      "loss": 2.0623,
+      "step": 318795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013823385348168377,
+      "loss": 2.166,
+      "step": 318800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013823214561258366,
+      "loss": 2.0745,
+      "step": 318805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001382304377304228,
+      "loss": 2.0927,
+      "step": 318810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001382287298352017,
+      "loss": 2.1488,
+      "step": 318815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013822702192692096,
+      "loss": 1.9412,
+      "step": 318820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001382253140055812,
+      "loss": 2.0689,
+      "step": 318825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00013822360607118297,
+      "loss": 2.1018,
+      "step": 318830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013822189812372687,
+      "loss": 1.9371,
+      "step": 318835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013822019016321345,
+      "loss": 2.0033,
+      "step": 318840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013821848218964337,
+      "loss": 2.013,
+      "step": 318845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013821677420301713,
+      "loss": 2.1278,
+      "step": 318850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013821506620333537,
+      "loss": 2.0949,
+      "step": 318855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013821335819059861,
+      "loss": 2.0574,
+      "step": 318860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001382116501648075,
+      "loss": 2.0092,
+      "step": 318865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013820994212596262,
+      "loss": 2.1322,
+      "step": 318870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013820823407406448,
+      "loss": 2.0091,
+      "step": 318875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013820652600911372,
+      "loss": 2.1914,
+      "step": 318880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013820481793111098,
+      "loss": 2.129,
+      "step": 318885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013820310984005672,
+      "loss": 1.9623,
+      "step": 318890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001382014017359516,
+      "loss": 2.0108,
+      "step": 318895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001381996936187962,
+      "loss": 2.1383,
+      "step": 318900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013819798548859107,
+      "loss": 2.0637,
+      "step": 318905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013819627734533684,
+      "loss": 2.0763,
+      "step": 318910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013819456918903405,
+      "loss": 2.0456,
+      "step": 318915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013819286101968332,
+      "loss": 2.268,
+      "step": 318920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001381911528372852,
+      "loss": 2.0806,
+      "step": 318925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001381894446418403,
+      "loss": 2.0762,
+      "step": 318930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013818773643334918,
+      "loss": 2.0549,
+      "step": 318935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013818602821181247,
+      "loss": 2.2012,
+      "step": 318940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013818431997723068,
+      "loss": 2.1529,
+      "step": 318945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013818261172960443,
+      "loss": 1.8279,
+      "step": 318950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013818090346893433,
+      "loss": 2.0629,
+      "step": 318955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013817919519522093,
+      "loss": 2.0512,
+      "step": 318960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013817748690846484,
+      "loss": 2.0377,
+      "step": 318965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013817577860866662,
+      "loss": 2.0061,
+      "step": 318970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4375,
+      "learning_rate": 0.00013817407029582688,
+      "loss": 2.0434,
+      "step": 318975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013817236196994617,
+      "loss": 2.1159,
+      "step": 318980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013817065363102505,
+      "loss": 1.8943,
+      "step": 318985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013816894527906418,
+      "loss": 2.1331,
+      "step": 318990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013816723691406413,
+      "loss": 1.9816,
+      "step": 318995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013816552853602545,
+      "loss": 2.1519,
+      "step": 319000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013816382014494868,
+      "loss": 2.1005,
+      "step": 319005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013816211174083452,
+      "loss": 1.9958,
+      "step": 319010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013816040332368344,
+      "loss": 2.0203,
+      "step": 319015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001381586948934961,
+      "loss": 2.0929,
+      "step": 319020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001381569864502731,
+      "loss": 1.9591,
+      "step": 319025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001381552779940149,
+      "loss": 1.9367,
+      "step": 319030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001381535695247222,
+      "loss": 2.1413,
+      "step": 319035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013815186104239555,
+      "loss": 1.9818,
+      "step": 319040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013815015254703555,
+      "loss": 2.0507,
+      "step": 319045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013814844403864276,
+      "loss": 2.0749,
+      "step": 319050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013814673551721776,
+      "loss": 1.9719,
+      "step": 319055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001381450269827611,
+      "loss": 2.0045,
+      "step": 319060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013814331843527348,
+      "loss": 2.2596,
+      "step": 319065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001381416098747554,
+      "loss": 2.2431,
+      "step": 319070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013813990130120743,
+      "loss": 2.0204,
+      "step": 319075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013813819271463017,
+      "loss": 2.0807,
+      "step": 319080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001381364841150242,
+      "loss": 2.0338,
+      "step": 319085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013813477550239019,
+      "loss": 1.9019,
+      "step": 319090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013813306687672857,
+      "loss": 2.0419,
+      "step": 319095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013813135823804003,
+      "loss": 2.11,
+      "step": 319100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013812964958632512,
+      "loss": 2.1409,
+      "step": 319105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013812794092158445,
+      "loss": 1.9814,
+      "step": 319110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013812623224381858,
+      "loss": 2.2027,
+      "step": 319115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001381245235530281,
+      "loss": 1.9276,
+      "step": 319120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013812281484921357,
+      "loss": 1.9355,
+      "step": 319125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013812110613237562,
+      "loss": 2.098,
+      "step": 319130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013811939740251478,
+      "loss": 2.1929,
+      "step": 319135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013811768865963167,
+      "loss": 2.0082,
+      "step": 319140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001381159799037269,
+      "loss": 2.1707,
+      "step": 319145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013811427113480097,
+      "loss": 1.9655,
+      "step": 319150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013811256235285454,
+      "loss": 2.0392,
+      "step": 319155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001381108535578882,
+      "loss": 2.0477,
+      "step": 319160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013810914474990247,
+      "loss": 1.9349,
+      "step": 319165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013810743592889793,
+      "loss": 2.167,
+      "step": 319170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013810572709487521,
+      "loss": 1.9428,
+      "step": 319175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013810401824783494,
+      "loss": 2.1224,
+      "step": 319180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001381023093877776,
+      "loss": 2.0226,
+      "step": 319185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013810060051470385,
+      "loss": 2.1163,
+      "step": 319190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013809889162861417,
+      "loss": 2.0462,
+      "step": 319195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001380971827295093,
+      "loss": 2.1127,
+      "step": 319200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013809547381738973,
+      "loss": 1.9319,
+      "step": 319205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013809376489225603,
+      "loss": 1.9913,
+      "step": 319210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001380920559541088,
+      "loss": 2.1658,
+      "step": 319215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013809034700294865,
+      "loss": 2.1713,
+      "step": 319220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013808863803877616,
+      "loss": 2.0313,
+      "step": 319225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001380869290615919,
+      "loss": 2.1838,
+      "step": 319230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001380852200713964,
+      "loss": 2.1282,
+      "step": 319235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013808351106819034,
+      "loss": 2.0424,
+      "step": 319240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001380818020519743,
+      "loss": 2.1204,
+      "step": 319245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013808009302274877,
+      "loss": 1.9087,
+      "step": 319250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001380783839805144,
+      "loss": 2.1735,
+      "step": 319255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013807667492527176,
+      "loss": 2.0389,
+      "step": 319260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013807496585702147,
+      "loss": 2.0386,
+      "step": 319265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013807325677576407,
+      "loss": 1.9908,
+      "step": 319270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001380715476815001,
+      "loss": 1.889,
+      "step": 319275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013806983857423026,
+      "loss": 2.2165,
+      "step": 319280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013806812945395508,
+      "loss": 2.1393,
+      "step": 319285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001380664203206751,
+      "loss": 2.1557,
+      "step": 319290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013806471117439094,
+      "loss": 2.0561,
+      "step": 319295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001380630020151032,
+      "loss": 2.1088,
+      "step": 319300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013806129284281245,
+      "loss": 2.1845,
+      "step": 319305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00013805958365751926,
+      "loss": 2.0074,
+      "step": 319310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001380578744592242,
+      "loss": 2.0887,
+      "step": 319315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013805616524792798,
+      "loss": 2.025,
+      "step": 319320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.000138054456023631,
+      "loss": 2.0617,
+      "step": 319325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013805274678633392,
+      "loss": 2.0576,
+      "step": 319330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013805103753603736,
+      "loss": 2.1035,
+      "step": 319335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001380493282727419,
+      "loss": 2.0273,
+      "step": 319340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013804761899644804,
+      "loss": 2.0854,
+      "step": 319345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.734375,
+      "learning_rate": 0.00013804590970715647,
+      "loss": 2.0809,
+      "step": 319350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001380442004048677,
+      "loss": 2.177,
+      "step": 319355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013804249108958236,
+      "loss": 1.9278,
+      "step": 319360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013804078176130098,
+      "loss": 2.1551,
+      "step": 319365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013803907242002423,
+      "loss": 2.0816,
+      "step": 319370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001380373630657526,
+      "loss": 2.0021,
+      "step": 319375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013803565369848673,
+      "loss": 2.0554,
+      "step": 319380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013803394431822724,
+      "loss": 2.1472,
+      "step": 319385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001380322349249746,
+      "loss": 1.8729,
+      "step": 319390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013803052551872948,
+      "loss": 2.0511,
+      "step": 319395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013802881609949243,
+      "loss": 1.9807,
+      "step": 319400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001380271066672641,
+      "loss": 1.8149,
+      "step": 319405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013802539722204495,
+      "loss": 2.187,
+      "step": 319410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013802368776383569,
+      "loss": 1.9173,
+      "step": 319415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001380219782926368,
+      "loss": 2.0777,
+      "step": 319420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013802026880844893,
+      "loss": 1.8601,
+      "step": 319425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013801855931127265,
+      "loss": 1.9145,
+      "step": 319430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013801684980110855,
+      "loss": 2.079,
+      "step": 319435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001380151402779572,
+      "loss": 2.0776,
+      "step": 319440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013801343074181917,
+      "loss": 2.0372,
+      "step": 319445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013801172119269509,
+      "loss": 1.9351,
+      "step": 319450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.00013801001163058553,
+      "loss": 2.0277,
+      "step": 319455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013800830205549104,
+      "loss": 2.0171,
+      "step": 319460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001380065924674122,
+      "loss": 2.1002,
+      "step": 319465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013800488286634964,
+      "loss": 2.0601,
+      "step": 319470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013800317325230393,
+      "loss": 1.9443,
+      "step": 319475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013800146362527566,
+      "loss": 2.0785,
+      "step": 319480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013799975398526537,
+      "loss": 1.9641,
+      "step": 319485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013799804433227365,
+      "loss": 1.8688,
+      "step": 319490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00013799633466630117,
+      "loss": 1.981,
+      "step": 319495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013799462498734845,
+      "loss": 1.9519,
+      "step": 319500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013799291529541604,
+      "loss": 1.9437,
+      "step": 319505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00013799120559050457,
+      "loss": 2.0512,
+      "step": 319510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001379894958726146,
+      "loss": 2.0459,
+      "step": 319515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.00013798778614174678,
+      "loss": 2.0332,
+      "step": 319520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001379860763979016,
+      "loss": 2.0824,
+      "step": 319525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013798436664107973,
+      "loss": 2.0607,
+      "step": 319530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013798265687128166,
+      "loss": 2.2402,
+      "step": 319535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013798094708850807,
+      "loss": 2.0914,
+      "step": 319540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013797923729275946,
+      "loss": 2.0016,
+      "step": 319545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013797752748403647,
+      "loss": 2.2192,
+      "step": 319550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013797581766233967,
+      "loss": 2.0461,
+      "step": 319555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013797410782766964,
+      "loss": 2.0374,
+      "step": 319560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013797239798002696,
+      "loss": 2.1244,
+      "step": 319565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013797068811941224,
+      "loss": 2.0155,
+      "step": 319570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000137968978245826,
+      "loss": 1.9737,
+      "step": 319575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013796726835926893,
+      "loss": 1.8929,
+      "step": 319580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013796555845974152,
+      "loss": 2.1634,
+      "step": 319585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001379638485472444,
+      "loss": 2.1519,
+      "step": 319590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001379621386217781,
+      "loss": 2.1483,
+      "step": 319595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013796042868334329,
+      "loss": 2.0438,
+      "step": 319600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013795871873194048,
+      "loss": 2.0066,
+      "step": 319605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013795700876757032,
+      "loss": 2.1803,
+      "step": 319610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013795529879023335,
+      "loss": 2.0101,
+      "step": 319615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013795358879993016,
+      "loss": 2.1003,
+      "step": 319620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001379518787966613,
+      "loss": 1.8627,
+      "step": 319625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001379501687804274,
+      "loss": 2.0869,
+      "step": 319630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013794845875122905,
+      "loss": 2.0954,
+      "step": 319635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013794674870906684,
+      "loss": 2.0979,
+      "step": 319640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001379450386539413,
+      "loss": 2.1211,
+      "step": 319645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013794332858585304,
+      "loss": 1.9911,
+      "step": 319650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013794161850480267,
+      "loss": 1.9096,
+      "step": 319655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013793990841079077,
+      "loss": 2.1018,
+      "step": 319660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013793819830381786,
+      "loss": 1.9979,
+      "step": 319665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001379364881838846,
+      "loss": 2.0435,
+      "step": 319670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013793477805099157,
+      "loss": 2.0494,
+      "step": 319675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001379330679051393,
+      "loss": 2.0241,
+      "step": 319680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013793135774632843,
+      "loss": 2.1732,
+      "step": 319685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001379296475745595,
+      "loss": 1.9122,
+      "step": 319690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001379279373898331,
+      "loss": 1.9294,
+      "step": 319695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013792622719214988,
+      "loss": 1.9288,
+      "step": 319700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013792451698151035,
+      "loss": 2.0994,
+      "step": 319705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001379228067579151,
+      "loss": 1.9875,
+      "step": 319710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013792109652136474,
+      "loss": 1.9278,
+      "step": 319715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013791938627185984,
+      "loss": 1.9397,
+      "step": 319720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000137917676009401,
+      "loss": 2.0985,
+      "step": 319725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001379159657339888,
+      "loss": 2.0606,
+      "step": 319730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001379142554456238,
+      "loss": 2.2689,
+      "step": 319735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013791254514430664,
+      "loss": 2.1666,
+      "step": 319740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013791083483003782,
+      "loss": 2.0355,
+      "step": 319745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013790912450281802,
+      "loss": 2.0606,
+      "step": 319750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013790741416264772,
+      "loss": 1.9541,
+      "step": 319755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013790570380952757,
+      "loss": 2.0102,
+      "step": 319760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013790399344345818,
+      "loss": 2.0339,
+      "step": 319765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013790228306444007,
+      "loss": 2.1188,
+      "step": 319770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013790057267247389,
+      "loss": 2.0678,
+      "step": 319775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013789886226756016,
+      "loss": 2.0352,
+      "step": 319780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013789715184969948,
+      "loss": 2.1972,
+      "step": 319785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013789544141889243,
+      "loss": 2.0438,
+      "step": 319790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013789373097513967,
+      "loss": 1.9849,
+      "step": 319795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001378920205184417,
+      "loss": 2.1098,
+      "step": 319800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001378903100487991,
+      "loss": 2.1397,
+      "step": 319805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001378885995662125,
+      "loss": 2.1166,
+      "step": 319810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001378868890706825,
+      "loss": 2.1671,
+      "step": 319815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013788517856220959,
+      "loss": 2.1821,
+      "step": 319820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013788346804079446,
+      "loss": 2.0807,
+      "step": 319825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013788175750643764,
+      "loss": 2.1227,
+      "step": 319830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013788004695913973,
+      "loss": 2.0198,
+      "step": 319835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001378783363989013,
+      "loss": 2.0672,
+      "step": 319840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013787662582572292,
+      "loss": 2.0569,
+      "step": 319845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013787491523960523,
+      "loss": 2.0761,
+      "step": 319850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013787320464054878,
+      "loss": 2.1572,
+      "step": 319855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013787149402855414,
+      "loss": 2.0158,
+      "step": 319860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001378697834036219,
+      "loss": 1.8954,
+      "step": 319865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013786807276575268,
+      "loss": 2.0535,
+      "step": 319870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013786636211494705,
+      "loss": 2.0417,
+      "step": 319875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013786465145120558,
+      "loss": 1.9827,
+      "step": 319880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013786294077452885,
+      "loss": 2.1667,
+      "step": 319885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00013786123008491745,
+      "loss": 2.0316,
+      "step": 319890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013785951938237196,
+      "loss": 2.0474,
+      "step": 319895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00013785780866689296,
+      "loss": 2.0636,
+      "step": 319900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013785609793848109,
+      "loss": 1.9017,
+      "step": 319905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013785438719713687,
+      "loss": 2.0562,
+      "step": 319910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001378526764428609,
+      "loss": 1.9887,
+      "step": 319915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013785096567565377,
+      "loss": 2.035,
+      "step": 319920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013784925489551605,
+      "loss": 2.1453,
+      "step": 319925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013784754410244837,
+      "loss": 2.0263,
+      "step": 319930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013784583329645126,
+      "loss": 2.0834,
+      "step": 319935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013784412247752533,
+      "loss": 2.0806,
+      "step": 319940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013784241164567116,
+      "loss": 1.9421,
+      "step": 319945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013784070080088935,
+      "loss": 1.9642,
+      "step": 319950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013783898994318045,
+      "loss": 2.0441,
+      "step": 319955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013783727907254508,
+      "loss": 2.0018,
+      "step": 319960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001378355681889838,
+      "loss": 2.1243,
+      "step": 319965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013783385729249718,
+      "loss": 2.0664,
+      "step": 319970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001378321463830859,
+      "loss": 1.904,
+      "step": 319975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001378304354607504,
+      "loss": 2.0623,
+      "step": 319980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013782872452549138,
+      "loss": 1.9234,
+      "step": 319985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013782701357730937,
+      "loss": 2.0157,
+      "step": 319990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013782530261620494,
+      "loss": 2.1985,
+      "step": 319995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013782359164217874,
+      "loss": 1.8921,
+      "step": 320000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001378218806552313,
+      "loss": 2.1171,
+      "step": 320005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013782016965536322,
+      "loss": 2.1368,
+      "step": 320010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013781845864257506,
+      "loss": 2.1106,
+      "step": 320015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013781674761686747,
+      "loss": 1.965,
+      "step": 320020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000137815036578241,
+      "loss": 2.0495,
+      "step": 320025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013781332552669618,
+      "loss": 2.0396,
+      "step": 320030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013781161446223364,
+      "loss": 2.0229,
+      "step": 320035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000137809903384854,
+      "loss": 2.0236,
+      "step": 320040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001378081922945578,
+      "loss": 2.1404,
+      "step": 320045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013780648119134564,
+      "loss": 2.0946,
+      "step": 320050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001378047700752181,
+      "loss": 2.0714,
+      "step": 320055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013780305894617575,
+      "loss": 2.1347,
+      "step": 320060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013780134780421924,
+      "loss": 2.0365,
+      "step": 320065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013779963664934907,
+      "loss": 2.1065,
+      "step": 320070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013779792548156584,
+      "loss": 2.0113,
+      "step": 320075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013779621430087016,
+      "loss": 2.0138,
+      "step": 320080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001377945031072626,
+      "loss": 2.1362,
+      "step": 320085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001377927919007438,
+      "loss": 1.9901,
+      "step": 320090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013779108068131426,
+      "loss": 2.0131,
+      "step": 320095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001377893694489746,
+      "loss": 2.0974,
+      "step": 320100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001377876582037254,
+      "loss": 2.3394,
+      "step": 320105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013778594694556728,
+      "loss": 2.0122,
+      "step": 320110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013778423567450078,
+      "loss": 2.0192,
+      "step": 320115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001377825243905265,
+      "loss": 2.0635,
+      "step": 320120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000137780813093645,
+      "loss": 2.1731,
+      "step": 320125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013777910178385693,
+      "loss": 2.0581,
+      "step": 320130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013777739046116284,
+      "loss": 2.1059,
+      "step": 320135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013777567912556327,
+      "loss": 2.0172,
+      "step": 320140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013777396777705885,
+      "loss": 2.0394,
+      "step": 320145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013777225641565014,
+      "loss": 2.0756,
+      "step": 320150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001377705450413378,
+      "loss": 2.106,
+      "step": 320155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001377688336541223,
+      "loss": 2.1548,
+      "step": 320160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013776712225400432,
+      "loss": 2.0737,
+      "step": 320165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013776541084098437,
+      "loss": 2.1827,
+      "step": 320170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001377636994150631,
+      "loss": 2.0465,
+      "step": 320175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013776198797624107,
+      "loss": 2.0526,
+      "step": 320180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013776027652451884,
+      "loss": 2.0331,
+      "step": 320185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013775856505989702,
+      "loss": 2.0409,
+      "step": 320190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001377568535823762,
+      "loss": 1.9298,
+      "step": 320195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013775514209195695,
+      "loss": 2.152,
+      "step": 320200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013775343058863986,
+      "loss": 2.0754,
+      "step": 320205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001377517190724255,
+      "loss": 2.1889,
+      "step": 320210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013775000754331446,
+      "loss": 2.0453,
+      "step": 320215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00013774829600130735,
+      "loss": 2.1094,
+      "step": 320220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013774658444640477,
+      "loss": 2.1821,
+      "step": 320225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013774487287860723,
+      "loss": 2.2008,
+      "step": 320230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013774316129791536,
+      "loss": 2.1677,
+      "step": 320235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013774144970432972,
+      "loss": 2.2784,
+      "step": 320240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013773973809785096,
+      "loss": 1.9794,
+      "step": 320245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001377380264784796,
+      "loss": 2.2254,
+      "step": 320250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013773631484621625,
+      "loss": 2.0075,
+      "step": 320255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013773460320106148,
+      "loss": 2.181,
+      "step": 320260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013773289154301587,
+      "loss": 2.1019,
+      "step": 320265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013773117987208008,
+      "loss": 2.0096,
+      "step": 320270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001377294681882546,
+      "loss": 2.0581,
+      "step": 320275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013772775649154007,
+      "loss": 2.1924,
+      "step": 320280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00013772604478193702,
+      "loss": 2.2016,
+      "step": 320285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013772433305944608,
+      "loss": 1.8552,
+      "step": 320290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001377226213240678,
+      "loss": 1.9848,
+      "step": 320295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00013772090957580284,
+      "loss": 1.9077,
+      "step": 320300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.0001377191978146517,
+      "loss": 2.0817,
+      "step": 320305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013771748604061502,
+      "loss": 1.9923,
+      "step": 320310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013771577425369334,
+      "loss": 2.0529,
+      "step": 320315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013771406245388727,
+      "loss": 2.0376,
+      "step": 320320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013771235064119738,
+      "loss": 2.045,
+      "step": 320325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001377106388156243,
+      "loss": 2.225,
+      "step": 320330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013770892697716859,
+      "loss": 1.9828,
+      "step": 320335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001377072151258308,
+      "loss": 1.9178,
+      "step": 320340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001377055032616115,
+      "loss": 2.0459,
+      "step": 320345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001377037913845114,
+      "loss": 2.1144,
+      "step": 320350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013770207949453096,
+      "loss": 2.0108,
+      "step": 320355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013770036759167082,
+      "loss": 1.9748,
+      "step": 320360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013769865567593153,
+      "loss": 2.0862,
+      "step": 320365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001376969437473137,
+      "loss": 1.9887,
+      "step": 320370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013769523180581788,
+      "loss": 2.1046,
+      "step": 320375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013769351985144475,
+      "loss": 2.1295,
+      "step": 320380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001376918078841948,
+      "loss": 2.2,
+      "step": 320385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013769009590406864,
+      "loss": 2.1754,
+      "step": 320390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013768838391106688,
+      "loss": 2.2428,
+      "step": 320395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013768667190519005,
+      "loss": 2.0384,
+      "step": 320400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013768495988643879,
+      "loss": 2.0979,
+      "step": 320405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013768324785481368,
+      "loss": 2.2275,
+      "step": 320410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013768153581031526,
+      "loss": 2.1649,
+      "step": 320415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013767982375294415,
+      "loss": 2.0503,
+      "step": 320420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013767811168270095,
+      "loss": 2.0143,
+      "step": 320425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001376763995995862,
+      "loss": 2.0349,
+      "step": 320430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013767468750360053,
+      "loss": 1.8837,
+      "step": 320435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.0001376729753947445,
+      "loss": 2.0333,
+      "step": 320440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013767126327301866,
+      "loss": 1.921,
+      "step": 320445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00013766955113842365,
+      "loss": 2.0262,
+      "step": 320450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013766783899096007,
+      "loss": 2.1043,
+      "step": 320455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013766612683062846,
+      "loss": 1.9687,
+      "step": 320460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013766441465742942,
+      "loss": 2.0633,
+      "step": 320465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001376627024713635,
+      "loss": 2.0749,
+      "step": 320470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001376609902724314,
+      "loss": 2.0455,
+      "step": 320475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001376592780606335,
+      "loss": 1.9083,
+      "step": 320480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001376575658359706,
+      "loss": 1.9159,
+      "step": 320485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013765585359844316,
+      "loss": 2.0172,
+      "step": 320490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001376541413480518,
+      "loss": 2.1123,
+      "step": 320495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013765242908479712,
+      "loss": 1.707,
+      "step": 320500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013765071680867964,
+      "loss": 1.999,
+      "step": 320505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001376490045197,
+      "loss": 1.8681,
+      "step": 320510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013764729221785884,
+      "loss": 2.2371,
+      "step": 320515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013764557990315665,
+      "loss": 2.1088,
+      "step": 320520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013764386757559403,
+      "loss": 2.0234,
+      "step": 320525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001376421552351716,
+      "loss": 1.9503,
+      "step": 320530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001376404428818899,
+      "loss": 2.1106,
+      "step": 320535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013763873051574958,
+      "loss": 1.8967,
+      "step": 320540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013763701813675119,
+      "loss": 1.902,
+      "step": 320545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013763530574489527,
+      "loss": 2.1302,
+      "step": 320550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013763359334018248,
+      "loss": 2.1649,
+      "step": 320555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00013763188092261332,
+      "loss": 1.9687,
+      "step": 320560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001376301684921885,
+      "loss": 2.126,
+      "step": 320565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001376284560489085,
+      "loss": 2.1818,
+      "step": 320570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013762674359277393,
+      "loss": 2.0715,
+      "step": 320575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001376250311237854,
+      "loss": 2.0801,
+      "step": 320580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013762331864194347,
+      "loss": 2.046,
+      "step": 320585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001376216061472487,
+      "loss": 2.0389,
+      "step": 320590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013761989363970175,
+      "loss": 2.254,
+      "step": 320595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013761818111930315,
+      "loss": 2.0666,
+      "step": 320600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013761646858605347,
+      "loss": 2.0959,
+      "step": 320605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013761475603995334,
+      "loss": 1.8921,
+      "step": 320610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013761304348100336,
+      "loss": 2.1687,
+      "step": 320615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013761133090920406,
+      "loss": 2.0635,
+      "step": 320620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013760961832455604,
+      "loss": 1.953,
+      "step": 320625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013760790572705988,
+      "loss": 2.2266,
+      "step": 320630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013760619311671622,
+      "loss": 2.0528,
+      "step": 320635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013760448049352555,
+      "loss": 2.1056,
+      "step": 320640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013760276785748855,
+      "loss": 1.7276,
+      "step": 320645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013760105520860573,
+      "loss": 1.8968,
+      "step": 320650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013759934254687774,
+      "loss": 2.1753,
+      "step": 320655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001375976298723051,
+      "loss": 1.7129,
+      "step": 320660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013759591718488844,
+      "loss": 1.955,
+      "step": 320665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001375942044846283,
+      "loss": 2.1198,
+      "step": 320670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013759249177152535,
+      "loss": 1.9572,
+      "step": 320675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013759077904558012,
+      "loss": 1.9927,
+      "step": 320680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001375890663067932,
+      "loss": 2.103,
+      "step": 320685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013758735355516515,
+      "loss": 1.833,
+      "step": 320690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001375856407906966,
+      "loss": 2.1103,
+      "step": 320695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013758392801338807,
+      "loss": 1.9645,
+      "step": 320700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013758221522324024,
+      "loss": 2.1812,
+      "step": 320705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013758050242025362,
+      "loss": 1.9885,
+      "step": 320710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013757878960442881,
+      "loss": 2.029,
+      "step": 320715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013757707677576638,
+      "loss": 1.8888,
+      "step": 320720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000137575363934267,
+      "loss": 2.0852,
+      "step": 320725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001375736510799312,
+      "loss": 2.0533,
+      "step": 320730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001375719382127595,
+      "loss": 2.1489,
+      "step": 320735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013757022533275256,
+      "loss": 1.9415,
+      "step": 320740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013756851243991097,
+      "loss": 2.0847,
+      "step": 320745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001375667995342353,
+      "loss": 2.1251,
+      "step": 320750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013756508661572612,
+      "loss": 2.0687,
+      "step": 320755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013756337368438403,
+      "loss": 1.8498,
+      "step": 320760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013756166074020958,
+      "loss": 2.0532,
+      "step": 320765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013755994778320342,
+      "loss": 2.2224,
+      "step": 320770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013755823481336607,
+      "loss": 2.0445,
+      "step": 320775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013755652183069817,
+      "loss": 2.1248,
+      "step": 320780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013755480883520028,
+      "loss": 1.9722,
+      "step": 320785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013755309582687297,
+      "loss": 2.0651,
+      "step": 320790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013755138280571687,
+      "loss": 2.2227,
+      "step": 320795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013754966977173248,
+      "loss": 1.9315,
+      "step": 320800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001375479567249205,
+      "loss": 1.9596,
+      "step": 320805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013754624366528145,
+      "loss": 2.151,
+      "step": 320810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001375445305928159,
+      "loss": 2.1701,
+      "step": 320815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001375428175075245,
+      "loss": 1.7625,
+      "step": 320820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013754110440940775,
+      "loss": 2.1134,
+      "step": 320825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013753939129846626,
+      "loss": 1.8056,
+      "step": 320830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013753767817470066,
+      "loss": 2.0717,
+      "step": 320835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013753596503811153,
+      "loss": 2.0964,
+      "step": 320840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001375342518886994,
+      "loss": 2.0399,
+      "step": 320845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001375325387264649,
+      "loss": 2.0083,
+      "step": 320850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013753082555140859,
+      "loss": 2.2864,
+      "step": 320855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001375291123635311,
+      "loss": 2.0066,
+      "step": 320860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013752739916283297,
+      "loss": 2.0218,
+      "step": 320865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013752568594931481,
+      "loss": 1.9826,
+      "step": 320870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00013752397272297716,
+      "loss": 2.0984,
+      "step": 320875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013752225948382066,
+      "loss": 2.0847,
+      "step": 320880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013752054623184589,
+      "loss": 2.1493,
+      "step": 320885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013751883296705343,
+      "loss": 1.9687,
+      "step": 320890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013751711968944383,
+      "loss": 1.9832,
+      "step": 320895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001375154063990177,
+      "loss": 2.1227,
+      "step": 320900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013751369309577565,
+      "loss": 1.8871,
+      "step": 320905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001375119797797182,
+      "loss": 2.1564,
+      "step": 320910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013751026645084604,
+      "loss": 1.9514,
+      "step": 320915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013750855310915965,
+      "loss": 2.0971,
+      "step": 320920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013750683975465966,
+      "loss": 1.8963,
+      "step": 320925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013750512638734667,
+      "loss": 1.9475,
+      "step": 320930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013750341300722122,
+      "loss": 2.1222,
+      "step": 320935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013750169961428393,
+      "loss": 2.0389,
+      "step": 320940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001374999862085354,
+      "loss": 2.0357,
+      "step": 320945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001374982727899762,
+      "loss": 2.1171,
+      "step": 320950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001374965593586069,
+      "loss": 2.1302,
+      "step": 320955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001374948459144281,
+      "loss": 2.1587,
+      "step": 320960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013749313245744037,
+      "loss": 2.187,
+      "step": 320965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013749141898764428,
+      "loss": 2.1212,
+      "step": 320970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001374897055050405,
+      "loss": 1.9866,
+      "step": 320975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013748799200962952,
+      "loss": 2.0184,
+      "step": 320980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000137486278501412,
+      "loss": 2.14,
+      "step": 320985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001374845649803884,
+      "loss": 2.1121,
+      "step": 320990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013748285144655947,
+      "loss": 1.9806,
+      "step": 320995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001374811378999257,
+      "loss": 2.0842,
+      "step": 321000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.0001374794243404877,
+      "loss": 2.094,
+      "step": 321005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013747771076824602,
+      "loss": 2.1545,
+      "step": 321010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013747599718320127,
+      "loss": 2.1117,
+      "step": 321015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0001374742835853541,
+      "loss": 2.2203,
+      "step": 321020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000137472569974705,
+      "loss": 2.0489,
+      "step": 321025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013747085635125455,
+      "loss": 2.1539,
+      "step": 321030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013746914271500345,
+      "loss": 2.1429,
+      "step": 321035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013746742906595215,
+      "loss": 1.943,
+      "step": 321040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013746571540410133,
+      "loss": 2.129,
+      "step": 321045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013746400172945153,
+      "loss": 2.1701,
+      "step": 321050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013746228804200337,
+      "loss": 2.217,
+      "step": 321055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001374605743417574,
+      "loss": 2.0114,
+      "step": 321060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001374588606287142,
+      "loss": 1.8611,
+      "step": 321065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.875,
+      "learning_rate": 0.0001374571469028744,
+      "loss": 2.0997,
+      "step": 321070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013745543316423853,
+      "loss": 2.0759,
+      "step": 321075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001374537194128072,
+      "loss": 1.7896,
+      "step": 321080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013745200564858103,
+      "loss": 2.0873,
+      "step": 321085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001374502918715606,
+      "loss": 2.1834,
+      "step": 321090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.0001374485780817464,
+      "loss": 1.9863,
+      "step": 321095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013744686427913914,
+      "loss": 2.1169,
+      "step": 321100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013744515046373932,
+      "loss": 1.9799,
+      "step": 321105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013744343663554763,
+      "loss": 2.0223,
+      "step": 321110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013744172279456452,
+      "loss": 2.0019,
+      "step": 321115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013744000894079064,
+      "loss": 2.1575,
+      "step": 321120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013743829507422654,
+      "loss": 2.0454,
+      "step": 321125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.0001374365811948729,
+      "loss": 1.8668,
+      "step": 321130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013743486730273023,
+      "loss": 2.1503,
+      "step": 321135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001374331533977991,
+      "loss": 1.9964,
+      "step": 321140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013743143948008018,
+      "loss": 1.9969,
+      "step": 321145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013742972554957393,
+      "loss": 1.9979,
+      "step": 321150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013742801160628108,
+      "loss": 1.9164,
+      "step": 321155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013742629765020213,
+      "loss": 1.9416,
+      "step": 321160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013742458368133764,
+      "loss": 2.0493,
+      "step": 321165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013742286969968824,
+      "loss": 2.0528,
+      "step": 321170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013742115570525453,
+      "loss": 2.0519,
+      "step": 321175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013741944169803707,
+      "loss": 2.0411,
+      "step": 321180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013741772767803646,
+      "loss": 2.1985,
+      "step": 321185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013741601364525326,
+      "loss": 2.215,
+      "step": 321190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013741429959968806,
+      "loss": 1.8511,
+      "step": 321195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013741258554134147,
+      "loss": 2.1449,
+      "step": 321200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013741087147021405,
+      "loss": 2.0026,
+      "step": 321205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013740915738630642,
+      "loss": 2.2351,
+      "step": 321210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013740744328961913,
+      "loss": 2.2662,
+      "step": 321215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013740572918015279,
+      "loss": 2.1153,
+      "step": 321220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013740401505790798,
+      "loss": 1.8895,
+      "step": 321225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013740230092288525,
+      "loss": 2.0517,
+      "step": 321230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013740058677508523,
+      "loss": 2.0884,
+      "step": 321235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013739887261450848,
+      "loss": 1.9734,
+      "step": 321240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013739715844115563,
+      "loss": 1.8819,
+      "step": 321245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013739544425502722,
+      "loss": 1.9931,
+      "step": 321250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013739373005612382,
+      "loss": 2.1233,
+      "step": 321255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013739201584444608,
+      "loss": 1.9353,
+      "step": 321260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001373903016199945,
+      "loss": 2.1086,
+      "step": 321265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001373885873827698,
+      "loss": 2.0375,
+      "step": 321270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001373868731327724,
+      "loss": 2.0369,
+      "step": 321275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000137385158870003,
+      "loss": 2.0161,
+      "step": 321280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013738344459446214,
+      "loss": 2.14,
+      "step": 321285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013738173030615044,
+      "loss": 2.0576,
+      "step": 321290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013738001600506843,
+      "loss": 2.052,
+      "step": 321295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013737830169121678,
+      "loss": 1.9822,
+      "step": 321300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000137376587364596,
+      "loss": 2.0131,
+      "step": 321305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013737487302520663,
+      "loss": 2.0807,
+      "step": 321310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013737315867304943,
+      "loss": 1.9992,
+      "step": 321315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013737144430812483,
+      "loss": 2.0219,
+      "step": 321320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013736972993043346,
+      "loss": 1.944,
+      "step": 321325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013736801553997593,
+      "loss": 2.1611,
+      "step": 321330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001373663011367528,
+      "loss": 1.9345,
+      "step": 321335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013736458672076467,
+      "loss": 1.9864,
+      "step": 321340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013736287229201212,
+      "loss": 1.9027,
+      "step": 321345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013736115785049574,
+      "loss": 1.9175,
+      "step": 321350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001373594433962161,
+      "loss": 2.0666,
+      "step": 321355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013735772892917384,
+      "loss": 2.1382,
+      "step": 321360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001373560144493694,
+      "loss": 1.9762,
+      "step": 321365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013735429995680355,
+      "loss": 1.9619,
+      "step": 321370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013735258545147678,
+      "loss": 1.9308,
+      "step": 321375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6875,
+      "learning_rate": 0.00013735087093338968,
+      "loss": 1.9549,
+      "step": 321380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013734915640254287,
+      "loss": 2.3312,
+      "step": 321385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013734744185893687,
+      "loss": 2.1771,
+      "step": 321390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013734572730257231,
+      "loss": 2.1597,
+      "step": 321395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013734401273344978,
+      "loss": 2.0732,
+      "step": 321400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013734229815156988,
+      "loss": 2.0918,
+      "step": 321405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013734058355693317,
+      "loss": 2.0763,
+      "step": 321410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013733886894954026,
+      "loss": 2.1891,
+      "step": 321415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013733715432939165,
+      "loss": 1.8824,
+      "step": 321420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.000137335439696488,
+      "loss": 2.1087,
+      "step": 321425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013733372505082996,
+      "loss": 1.9722,
+      "step": 321430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013733201039241798,
+      "loss": 1.9684,
+      "step": 321435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013733029572125273,
+      "loss": 1.9529,
+      "step": 321440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013732858103733474,
+      "loss": 2.0482,
+      "step": 321445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013732686634066469,
+      "loss": 2.084,
+      "step": 321450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013732515163124306,
+      "loss": 1.9993,
+      "step": 321455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013732343690907049,
+      "loss": 1.9983,
+      "step": 321460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013732172217414756,
+      "loss": 1.9136,
+      "step": 321465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013732000742647487,
+      "loss": 2.1449,
+      "step": 321470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013731829266605295,
+      "loss": 2.0422,
+      "step": 321475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013731657789288246,
+      "loss": 2.0829,
+      "step": 321480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013731486310696392,
+      "loss": 2.0106,
+      "step": 321485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013731314830829798,
+      "loss": 2.0104,
+      "step": 321490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013731143349688516,
+      "loss": 2.0724,
+      "step": 321495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013730971867272608,
+      "loss": 2.0835,
+      "step": 321500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013730800383582133,
+      "loss": 2.1756,
+      "step": 321505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013730628898617152,
+      "loss": 2.0742,
+      "step": 321510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013730457412377718,
+      "loss": 1.7845,
+      "step": 321515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013730285924863894,
+      "loss": 2.1035,
+      "step": 321520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001373011443607573,
+      "loss": 1.9823,
+      "step": 321525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000137299429460133,
+      "loss": 1.9318,
+      "step": 321530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013729771454676647,
+      "loss": 1.9554,
+      "step": 321535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013729599962065842,
+      "loss": 2.1192,
+      "step": 321540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013729428468180937,
+      "loss": 2.0561,
+      "step": 321545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013729256973021986,
+      "loss": 2.071,
+      "step": 321550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013729085476589056,
+      "loss": 1.9493,
+      "step": 321555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013728913978882206,
+      "loss": 2.1172,
+      "step": 321560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001372874247990149,
+      "loss": 2.0217,
+      "step": 321565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013728570979646966,
+      "loss": 2.1586,
+      "step": 321570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013728399478118696,
+      "loss": 2.0631,
+      "step": 321575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013728227975316737,
+      "loss": 2.2205,
+      "step": 321580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013728056471241148,
+      "loss": 2.2129,
+      "step": 321585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013727884965891986,
+      "loss": 2.296,
+      "step": 321590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013727713459269313,
+      "loss": 1.8303,
+      "step": 321595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013727541951373185,
+      "loss": 2.0647,
+      "step": 321600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013727370442203658,
+      "loss": 2.1039,
+      "step": 321605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013727198931760796,
+      "loss": 2.0669,
+      "step": 321610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013727027420044655,
+      "loss": 2.1526,
+      "step": 321615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013726855907055297,
+      "loss": 2.0036,
+      "step": 321620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001372668439279277,
+      "loss": 1.9883,
+      "step": 321625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013726512877257149,
+      "loss": 2.047,
+      "step": 321630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013726341360448477,
+      "loss": 2.0613,
+      "step": 321635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001372616984236682,
+      "loss": 1.9424,
+      "step": 321640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013725998323012236,
+      "loss": 1.9956,
+      "step": 321645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013725826802384787,
+      "loss": 1.8482,
+      "step": 321650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013725655280484526,
+      "loss": 1.9584,
+      "step": 321655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013725483757311512,
+      "loss": 1.8678,
+      "step": 321660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013725312232865805,
+      "loss": 2.1016,
+      "step": 321665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013725140707147466,
+      "loss": 2.0471,
+      "step": 321670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001372496918015655,
+      "loss": 1.9761,
+      "step": 321675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001372479765189312,
+      "loss": 1.9402,
+      "step": 321680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013724626122357228,
+      "loss": 2.0978,
+      "step": 321685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013724454591548937,
+      "loss": 2.1433,
+      "step": 321690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013724283059468302,
+      "loss": 2.0017,
+      "step": 321695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013724111526115392,
+      "loss": 2.1788,
+      "step": 321700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001372393999149025,
+      "loss": 2.1986,
+      "step": 321705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013723768455592948,
+      "loss": 2.178,
+      "step": 321710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013723596918423535,
+      "loss": 2.1646,
+      "step": 321715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013723425379982075,
+      "loss": 2.2036,
+      "step": 321720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013723253840268627,
+      "loss": 2.0652,
+      "step": 321725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013723082299283248,
+      "loss": 2.1624,
+      "step": 321730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013722910757025996,
+      "loss": 2.0738,
+      "step": 321735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013722739213496926,
+      "loss": 2.1004,
+      "step": 321740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013722567668696106,
+      "loss": 1.97,
+      "step": 321745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013722396122623588,
+      "loss": 2.1035,
+      "step": 321750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013722224575279433,
+      "loss": 2.1586,
+      "step": 321755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013722053026663696,
+      "loss": 2.0369,
+      "step": 321760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013721881476776442,
+      "loss": 1.862,
+      "step": 321765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013721709925617722,
+      "loss": 2.1141,
+      "step": 321770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013721538373187602,
+      "loss": 1.9768,
+      "step": 321775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013721366819486133,
+      "loss": 1.7181,
+      "step": 321780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013721195264513377,
+      "loss": 1.8729,
+      "step": 321785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000137210237082694,
+      "loss": 2.0263,
+      "step": 321790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013720852150754246,
+      "loss": 1.9096,
+      "step": 321795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013720680591967988,
+      "loss": 2.0953,
+      "step": 321800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013720509031910674,
+      "loss": 2.0151,
+      "step": 321805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013720337470582368,
+      "loss": 2.0131,
+      "step": 321810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013720165907983126,
+      "loss": 1.9724,
+      "step": 321815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013719994344113008,
+      "loss": 2.0586,
+      "step": 321820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001371982277897207,
+      "loss": 2.0492,
+      "step": 321825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013719651212560377,
+      "loss": 1.8386,
+      "step": 321830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013719479644877983,
+      "loss": 2.0894,
+      "step": 321835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001371930807592495,
+      "loss": 2.0346,
+      "step": 321840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013719136505701329,
+      "loss": 1.9482,
+      "step": 321845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013718964934207186,
+      "loss": 2.0387,
+      "step": 321850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013718793361442577,
+      "loss": 1.9958,
+      "step": 321855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013718621787407563,
+      "loss": 2.0765,
+      "step": 321860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013718450212102197,
+      "loss": 2.1682,
+      "step": 321865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013718278635526542,
+      "loss": 1.948,
+      "step": 321870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00013718107057680652,
+      "loss": 1.9342,
+      "step": 321875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013717935478564595,
+      "loss": 2.0077,
+      "step": 321880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013717763898178422,
+      "loss": 2.1662,
+      "step": 321885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013717592316522193,
+      "loss": 2.1495,
+      "step": 321890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013717420733595968,
+      "loss": 2.1362,
+      "step": 321895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000137172491493998,
+      "loss": 2.0997,
+      "step": 321900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013717077563933758,
+      "loss": 2.201,
+      "step": 321905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013716905977197893,
+      "loss": 2.0113,
+      "step": 321910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013716734389192264,
+      "loss": 2.0903,
+      "step": 321915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013716562799916931,
+      "loss": 2.1844,
+      "step": 321920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013716391209371955,
+      "loss": 1.9622,
+      "step": 321925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013716219617557394,
+      "loss": 1.9922,
+      "step": 321930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000137160480244733,
+      "loss": 2.1362,
+      "step": 321935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013715876430119738,
+      "loss": 1.9644,
+      "step": 321940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013715704834496768,
+      "loss": 2.0505,
+      "step": 321945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00013715533237604446,
+      "loss": 2.0554,
+      "step": 321950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013715361639442825,
+      "loss": 2.0433,
+      "step": 321955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00013715190040011974,
+      "loss": 2.0244,
+      "step": 321960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013715018439311945,
+      "loss": 2.0958,
+      "step": 321965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013714846837342798,
+      "loss": 2.0823,
+      "step": 321970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013714675234104595,
+      "loss": 2.1349,
+      "step": 321975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00013714503629597385,
+      "loss": 2.0664,
+      "step": 321980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013714332023821238,
+      "loss": 2.0518,
+      "step": 321985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001371416041677621,
+      "loss": 2.1079,
+      "step": 321990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013713988808462354,
+      "loss": 2.2346,
+      "step": 321995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013713817198879734,
+      "loss": 2.0544,
+      "step": 322000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013713645588028404,
+      "loss": 2.0881,
+      "step": 322005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013713473975908428,
+      "loss": 2.112,
+      "step": 322010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001371330236251986,
+      "loss": 2.1287,
+      "step": 322015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001371313074786276,
+      "loss": 2.136,
+      "step": 322020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013712959131937188,
+      "loss": 1.9907,
+      "step": 322025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013712787514743205,
+      "loss": 2.1875,
+      "step": 322030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001371261589628086,
+      "loss": 2.0272,
+      "step": 322035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013712444276550224,
+      "loss": 2.0997,
+      "step": 322040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013712272655551347,
+      "loss": 2.1318,
+      "step": 322045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001371210103328429,
+      "loss": 2.0078,
+      "step": 322050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013711929409749112,
+      "loss": 2.1265,
+      "step": 322055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013711757784945873,
+      "loss": 2.0405,
+      "step": 322060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001371158615887463,
+      "loss": 2.0841,
+      "step": 322065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013711414531535442,
+      "loss": 2.2387,
+      "step": 322070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013711242902928364,
+      "loss": 2.0792,
+      "step": 322075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013711071273053463,
+      "loss": 2.1299,
+      "step": 322080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001371089964191079,
+      "loss": 2.0522,
+      "step": 322085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013710728009500408,
+      "loss": 1.982,
+      "step": 322090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013710556375822375,
+      "loss": 2.0962,
+      "step": 322095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013710384740876744,
+      "loss": 1.8861,
+      "step": 322100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013710213104663584,
+      "loss": 2.1682,
+      "step": 322105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013710041467182945,
+      "loss": 1.9793,
+      "step": 322110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013709869828434888,
+      "loss": 1.9911,
+      "step": 322115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001370969818841947,
+      "loss": 2.0413,
+      "step": 322120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013709526547136755,
+      "loss": 2.0455,
+      "step": 322125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013709354904586798,
+      "loss": 2.0326,
+      "step": 322130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001370918326076966,
+      "loss": 2.1407,
+      "step": 322135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013709011615685394,
+      "loss": 2.0958,
+      "step": 322140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013708839969334065,
+      "loss": 2.0314,
+      "step": 322145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013708668321715727,
+      "loss": 2.1452,
+      "step": 322150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013708496672830444,
+      "loss": 1.9863,
+      "step": 322155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001370832502267827,
+      "loss": 2.0619,
+      "step": 322160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013708153371259262,
+      "loss": 2.0381,
+      "step": 322165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013707981718573485,
+      "loss": 2.1321,
+      "step": 322170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013707810064620993,
+      "loss": 1.9641,
+      "step": 322175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013707638409401847,
+      "loss": 2.1092,
+      "step": 322180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013707466752916104,
+      "loss": 1.8891,
+      "step": 322185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001370729509516382,
+      "loss": 2.081,
+      "step": 322190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013707123436145058,
+      "loss": 2.1579,
+      "step": 322195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013706951775859877,
+      "loss": 1.8796,
+      "step": 322200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013706780114308332,
+      "loss": 1.9407,
+      "step": 322205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013706608451490485,
+      "loss": 2.0733,
+      "step": 322210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013706436787406395,
+      "loss": 1.9701,
+      "step": 322215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013706265122056117,
+      "loss": 2.0383,
+      "step": 322220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013706093455439713,
+      "loss": 1.9555,
+      "step": 322225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013705921787557238,
+      "loss": 1.9786,
+      "step": 322230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013705750118408756,
+      "loss": 2.081,
+      "step": 322235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001370557844799432,
+      "loss": 1.755,
+      "step": 322240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013705406776313994,
+      "loss": 1.9954,
+      "step": 322245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00013705235103367828,
+      "loss": 1.9945,
+      "step": 322250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013705063429155892,
+      "loss": 2.0351,
+      "step": 322255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013704891753678238,
+      "loss": 2.1214,
+      "step": 322260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013704720076934926,
+      "loss": 2.03,
+      "step": 322265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013704548398926013,
+      "loss": 2.0411,
+      "step": 322270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001370437671965156,
+      "loss": 1.9141,
+      "step": 322275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013704205039111623,
+      "loss": 2.0179,
+      "step": 322280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00013704033357306262,
+      "loss": 1.9805,
+      "step": 322285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001370386167423554,
+      "loss": 2.0006,
+      "step": 322290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001370368998989951,
+      "loss": 1.9541,
+      "step": 322295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001370351830429823,
+      "loss": 2.0691,
+      "step": 322300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013703346617431763,
+      "loss": 1.9467,
+      "step": 322305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013703174929300164,
+      "loss": 2.2653,
+      "step": 322310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.000137030032399035,
+      "loss": 2.1412,
+      "step": 322315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013702831549241816,
+      "loss": 2.0149,
+      "step": 322320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001370265985731518,
+      "loss": 1.9887,
+      "step": 322325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013702488164123647,
+      "loss": 2.1265,
+      "step": 322330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013702316469667278,
+      "loss": 1.9979,
+      "step": 322335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013702144773946129,
+      "loss": 2.1012,
+      "step": 322340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013701973076960262,
+      "loss": 2.2177,
+      "step": 322345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013701801378709734,
+      "loss": 1.9174,
+      "step": 322350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000137016296791946,
+      "loss": 1.9924,
+      "step": 322355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013701457978414924,
+      "loss": 2.0084,
+      "step": 322360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013701286276370764,
+      "loss": 2.2101,
+      "step": 322365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001370111457306218,
+      "loss": 2.0392,
+      "step": 322370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013700942868489225,
+      "loss": 1.8422,
+      "step": 322375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001370077116265196,
+      "loss": 2.0607,
+      "step": 322380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013700599455550443,
+      "loss": 2.0768,
+      "step": 322385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013700427747184737,
+      "loss": 1.9989,
+      "step": 322390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013700256037554897,
+      "loss": 2.0628,
+      "step": 322395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013700084326660986,
+      "loss": 2.0249,
+      "step": 322400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013699912614503057,
+      "loss": 2.0099,
+      "step": 322405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013699740901081166,
+      "loss": 1.9647,
+      "step": 322410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001369956918639538,
+      "loss": 2.048,
+      "step": 322415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013699397470445755,
+      "loss": 1.9485,
+      "step": 322420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001369922575323235,
+      "loss": 2.1254,
+      "step": 322425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013699054034755219,
+      "loss": 2.025,
+      "step": 322430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013698882315014424,
+      "loss": 1.9513,
+      "step": 322435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013698710594010023,
+      "loss": 2.0205,
+      "step": 322440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001369853887174208,
+      "loss": 2.0159,
+      "step": 322445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013698367148210646,
+      "loss": 1.9579,
+      "step": 322450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013698195423415783,
+      "loss": 2.2388,
+      "step": 322455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013698023697357548,
+      "loss": 1.8657,
+      "step": 322460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013697851970036003,
+      "loss": 2.061,
+      "step": 322465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013697680241451206,
+      "loss": 2.142,
+      "step": 322470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013697508511603212,
+      "loss": 2.1184,
+      "step": 322475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013697336780492083,
+      "loss": 2.0317,
+      "step": 322480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013697165048117874,
+      "loss": 1.9949,
+      "step": 322485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013696993314480648,
+      "loss": 1.9279,
+      "step": 322490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013696821579580465,
+      "loss": 2.0243,
+      "step": 322495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013696649843417377,
+      "loss": 2.013,
+      "step": 322500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013696478105991447,
+      "loss": 1.8687,
+      "step": 322505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013696306367302736,
+      "loss": 2.0902,
+      "step": 322510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.000136961346273513,
+      "loss": 2.0263,
+      "step": 322515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001369596288613719,
+      "loss": 1.9694,
+      "step": 322520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001369579114366048,
+      "loss": 1.9543,
+      "step": 322525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013695619399921216,
+      "loss": 2.0505,
+      "step": 322530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013695447654919464,
+      "loss": 1.8995,
+      "step": 322535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00013695275908655278,
+      "loss": 2.0044,
+      "step": 322540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001369510416112872,
+      "loss": 2.0548,
+      "step": 322545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013694932412339848,
+      "loss": 1.8683,
+      "step": 322550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001369476066228872,
+      "loss": 2.0833,
+      "step": 322555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013694588910975392,
+      "loss": 1.934,
+      "step": 322560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001369441715839993,
+      "loss": 2.011,
+      "step": 322565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013694245404562386,
+      "loss": 2.0762,
+      "step": 322570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001369407364946282,
+      "loss": 2.0677,
+      "step": 322575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001369390189310129,
+      "loss": 2.1899,
+      "step": 322580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001369373013547786,
+      "loss": 1.9981,
+      "step": 322585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013693558376592585,
+      "loss": 1.6322,
+      "step": 322590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013693386616445515,
+      "loss": 2.0558,
+      "step": 322595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013693214855036728,
+      "loss": 2.049,
+      "step": 322600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013693043092366264,
+      "loss": 1.971,
+      "step": 322605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013692871328434197,
+      "loss": 2.1592,
+      "step": 322610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013692699563240573,
+      "loss": 2.1308,
+      "step": 322615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013692527796785458,
+      "loss": 2.036,
+      "step": 322620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.000136923560290689,
+      "loss": 1.7691,
+      "step": 322625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013692184260090978,
+      "loss": 1.8554,
+      "step": 322630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013692012489851735,
+      "loss": 2.0558,
+      "step": 322635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001369184071835123,
+      "loss": 2.0272,
+      "step": 322640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013691668945589532,
+      "loss": 1.9478,
+      "step": 322645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013691497171566689,
+      "loss": 1.9625,
+      "step": 322650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013691325396282764,
+      "loss": 2.2362,
+      "step": 322655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013691153619737814,
+      "loss": 2.1404,
+      "step": 322660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000136909818419319,
+      "loss": 2.0754,
+      "step": 322665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013690810062865077,
+      "loss": 1.9141,
+      "step": 322670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001369063828253741,
+      "loss": 2.0887,
+      "step": 322675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013690466500948954,
+      "loss": 2.0503,
+      "step": 322680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013690294718099767,
+      "loss": 1.942,
+      "step": 322685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013690122933989908,
+      "loss": 2.0463,
+      "step": 322690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00013689951148619436,
+      "loss": 1.9802,
+      "step": 322695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013689779361988414,
+      "loss": 1.7974,
+      "step": 322700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013689607574096887,
+      "loss": 2.0573,
+      "step": 322705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001368943578494493,
+      "loss": 2.1852,
+      "step": 322710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013689263994532596,
+      "loss": 2.1178,
+      "step": 322715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013689092202859938,
+      "loss": 1.9853,
+      "step": 322720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013688920409927024,
+      "loss": 2.0337,
+      "step": 322725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013688748615733903,
+      "loss": 2.0511,
+      "step": 322730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001368857682028064,
+      "loss": 2.0899,
+      "step": 322735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013688405023567292,
+      "loss": 2.0655,
+      "step": 322740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013688233225593921,
+      "loss": 1.7745,
+      "step": 322745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001368806142636058,
+      "loss": 2.0681,
+      "step": 322750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013687889625867332,
+      "loss": 1.9719,
+      "step": 322755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001368771782411423,
+      "loss": 2.1242,
+      "step": 322760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013687546021101343,
+      "loss": 2.0498,
+      "step": 322765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001368737421682872,
+      "loss": 2.0691,
+      "step": 322770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013687202411296422,
+      "loss": 1.9745,
+      "step": 322775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013687030604504508,
+      "loss": 1.9033,
+      "step": 322780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001368685879645304,
+      "loss": 2.2233,
+      "step": 322785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013686686987142074,
+      "loss": 1.7821,
+      "step": 322790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013686515176571668,
+      "loss": 2.0706,
+      "step": 322795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013686343364741882,
+      "loss": 2.0445,
+      "step": 322800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013686171551652776,
+      "loss": 1.9458,
+      "step": 322805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013685999737304402,
+      "loss": 1.8961,
+      "step": 322810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013685827921696827,
+      "loss": 2.0545,
+      "step": 322815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013685656104830106,
+      "loss": 1.7797,
+      "step": 322820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000136854842867043,
+      "loss": 2.1536,
+      "step": 322825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013685312467319463,
+      "loss": 2.1261,
+      "step": 322830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00013685140646675658,
+      "loss": 2.0349,
+      "step": 322835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001368496882477294,
+      "loss": 1.9127,
+      "step": 322840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001368479700161137,
+      "loss": 1.9568,
+      "step": 322845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013684625177191008,
+      "loss": 1.992,
+      "step": 322850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013684453351511913,
+      "loss": 2.0157,
+      "step": 322855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013684281524574141,
+      "loss": 2.081,
+      "step": 322860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001368410969637775,
+      "loss": 2.0348,
+      "step": 322865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013683937866922798,
+      "loss": 2.1399,
+      "step": 322870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001368376603620935,
+      "loss": 2.0483,
+      "step": 322875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001368359420423746,
+      "loss": 1.9935,
+      "step": 322880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013683422371007188,
+      "loss": 2.1062,
+      "step": 322885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013683250536518588,
+      "loss": 1.8679,
+      "step": 322890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013683078700771727,
+      "loss": 2.101,
+      "step": 322895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013682906863766658,
+      "loss": 2.0178,
+      "step": 322900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013682735025503443,
+      "loss": 1.9944,
+      "step": 322905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001368256318598214,
+      "loss": 2.018,
+      "step": 322910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00013682391345202802,
+      "loss": 1.9998,
+      "step": 322915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013682219503165493,
+      "loss": 1.9769,
+      "step": 322920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013682047659870273,
+      "loss": 2.1037,
+      "step": 322925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.000136818758153172,
+      "loss": 1.9826,
+      "step": 322930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001368170396950633,
+      "loss": 2.146,
+      "step": 322935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013681532122437721,
+      "loss": 1.9678,
+      "step": 322940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013681360274111438,
+      "loss": 2.1285,
+      "step": 322945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001368118842452753,
+      "loss": 2.1888,
+      "step": 322950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013681016573686066,
+      "loss": 2.0053,
+      "step": 322955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013680844721587098,
+      "loss": 1.9898,
+      "step": 322960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013680672868230688,
+      "loss": 1.8542,
+      "step": 322965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001368050101361689,
+      "loss": 2.1288,
+      "step": 322970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001368032915774577,
+      "loss": 1.7506,
+      "step": 322975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013680157300617382,
+      "loss": 1.9544,
+      "step": 322980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013679985442231787,
+      "loss": 2.1208,
+      "step": 322985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013679813582589037,
+      "loss": 1.9308,
+      "step": 322990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000136796417216892,
+      "loss": 1.9952,
+      "step": 322995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001367946985953233,
+      "loss": 2.0697,
+      "step": 323000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013679297996118485,
+      "loss": 2.0372,
+      "step": 323005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013679126131447728,
+      "loss": 2.15,
+      "step": 323010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001367895426552011,
+      "loss": 2.0232,
+      "step": 323015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000136787823983357,
+      "loss": 1.9165,
+      "step": 323020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001367861052989455,
+      "loss": 1.9867,
+      "step": 323025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013678438660196717,
+      "loss": 2.163,
+      "step": 323030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013678266789242263,
+      "loss": 2.0037,
+      "step": 323035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013678094917031253,
+      "loss": 2.0683,
+      "step": 323040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013677923043563733,
+      "loss": 2.0912,
+      "step": 323045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013677751168839766,
+      "loss": 2.0372,
+      "step": 323050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013677579292859415,
+      "loss": 2.0152,
+      "step": 323055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013677407415622737,
+      "loss": 2.0868,
+      "step": 323060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001367723553712979,
+      "loss": 2.1293,
+      "step": 323065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013677063657380634,
+      "loss": 1.9573,
+      "step": 323070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013676891776375323,
+      "loss": 2.0637,
+      "step": 323075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013676719894113918,
+      "loss": 1.9784,
+      "step": 323080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013676548010596484,
+      "loss": 2.1741,
+      "step": 323085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001367637612582307,
+      "loss": 2.1776,
+      "step": 323090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013676204239793742,
+      "loss": 1.863,
+      "step": 323095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00013676032352508553,
+      "loss": 2.0479,
+      "step": 323100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013675860463967568,
+      "loss": 1.9564,
+      "step": 323105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001367568857417084,
+      "loss": 2.1269,
+      "step": 323110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013675516683118433,
+      "loss": 2.0228,
+      "step": 323115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000136753447908104,
+      "loss": 2.0493,
+      "step": 323120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000136751728972468,
+      "loss": 1.9412,
+      "step": 323125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000136750010024277,
+      "loss": 1.9517,
+      "step": 323130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001367482910635315,
+      "loss": 2.0952,
+      "step": 323135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013674657209023213,
+      "loss": 2.0047,
+      "step": 323140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001367448531043795,
+      "loss": 1.9721,
+      "step": 323145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001367431341059741,
+      "loss": 2.0325,
+      "step": 323150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001367414150950166,
+      "loss": 1.9115,
+      "step": 323155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013673969607150757,
+      "loss": 2.0549,
+      "step": 323160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013673797703544758,
+      "loss": 2.125,
+      "step": 323165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013673625798683726,
+      "loss": 2.0705,
+      "step": 323170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013673453892567715,
+      "loss": 2.1653,
+      "step": 323175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013673281985196787,
+      "loss": 1.9867,
+      "step": 323180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013673110076570995,
+      "loss": 1.9673,
+      "step": 323185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013672938166690404,
+      "loss": 2.1672,
+      "step": 323190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001367276625555507,
+      "loss": 1.9955,
+      "step": 323195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013672594343165057,
+      "loss": 2.0635,
+      "step": 323200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013672422429520414,
+      "loss": 2.0762,
+      "step": 323205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001367225051462121,
+      "loss": 2.1005,
+      "step": 323210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001367207859846749,
+      "loss": 2.152,
+      "step": 323215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001367190668105933,
+      "loss": 1.8852,
+      "step": 323220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013671734762396776,
+      "loss": 2.0361,
+      "step": 323225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013671562842479891,
+      "loss": 2.0049,
+      "step": 323230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013671390921308736,
+      "loss": 1.9481,
+      "step": 323235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013671218998883365,
+      "loss": 1.9818,
+      "step": 323240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001367104707520384,
+      "loss": 2.0411,
+      "step": 323245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001367087515027022,
+      "loss": 1.9531,
+      "step": 323250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001367070322408256,
+      "loss": 2.0018,
+      "step": 323255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001367053129664092,
+      "loss": 2.1192,
+      "step": 323260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013670359367945362,
+      "loss": 2.0583,
+      "step": 323265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013670187437995942,
+      "loss": 2.1003,
+      "step": 323270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001367001550679272,
+      "loss": 1.9781,
+      "step": 323275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013669843574335756,
+      "loss": 1.9481,
+      "step": 323280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013669671640625106,
+      "loss": 2.0565,
+      "step": 323285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013669499705660829,
+      "loss": 2.0673,
+      "step": 323290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001366932776944298,
+      "loss": 1.9783,
+      "step": 323295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013669155831971626,
+      "loss": 1.9362,
+      "step": 323300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013668983893246826,
+      "loss": 2.0066,
+      "step": 323305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.0001366881195326863,
+      "loss": 2.1769,
+      "step": 323310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.000136686400120371,
+      "loss": 2.136,
+      "step": 323315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013668468069552299,
+      "loss": 2.1083,
+      "step": 323320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001366829612581428,
+      "loss": 2.0091,
+      "step": 323325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001366812418082311,
+      "loss": 1.9502,
+      "step": 323330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00013667952234578838,
+      "loss": 1.9041,
+      "step": 323335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013667780287081527,
+      "loss": 2.102,
+      "step": 323340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013667608338331232,
+      "loss": 2.0809,
+      "step": 323345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001366743638832802,
+      "loss": 1.7681,
+      "step": 323350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013667264437071947,
+      "loss": 2.0821,
+      "step": 323355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001366709248456307,
+      "loss": 2.216,
+      "step": 323360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013666920530801445,
+      "loss": 2.2091,
+      "step": 323365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013666748575787133,
+      "loss": 2.0434,
+      "step": 323370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001366657661952019,
+      "loss": 2.134,
+      "step": 323375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013666404662000686,
+      "loss": 2.2622,
+      "step": 323380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013666232703228668,
+      "loss": 2.1257,
+      "step": 323385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013666060743204196,
+      "loss": 2.0878,
+      "step": 323390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013665888781927334,
+      "loss": 1.884,
+      "step": 323395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013665716819398138,
+      "loss": 2.0473,
+      "step": 323400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013665544855616664,
+      "loss": 2.0051,
+      "step": 323405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013665372890582977,
+      "loss": 2.1571,
+      "step": 323410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001366520092429713,
+      "loss": 1.9801,
+      "step": 323415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013665028956759184,
+      "loss": 1.9955,
+      "step": 323420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000136648569879692,
+      "loss": 2.2271,
+      "step": 323425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001366468501792723,
+      "loss": 1.9059,
+      "step": 323430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001366451304663334,
+      "loss": 2.269,
+      "step": 323435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013664341074087587,
+      "loss": 1.9058,
+      "step": 323440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013664169100290025,
+      "loss": 2.1443,
+      "step": 323445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013663997125240723,
+      "loss": 1.9791,
+      "step": 323450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013663825148939725,
+      "loss": 2.0497,
+      "step": 323455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000136636531713871,
+      "loss": 2.1124,
+      "step": 323460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013663481192582907,
+      "loss": 1.9461,
+      "step": 323465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013663309212527203,
+      "loss": 2.1705,
+      "step": 323470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013663137231220045,
+      "loss": 2.0576,
+      "step": 323475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001366296524866149,
+      "loss": 2.1117,
+      "step": 323480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000136627932648516,
+      "loss": 2.1302,
+      "step": 323485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013662621279790437,
+      "loss": 1.992,
+      "step": 323490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013662449293478057,
+      "loss": 2.0161,
+      "step": 323495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00013662277305914516,
+      "loss": 2.0651,
+      "step": 323500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013662105317099874,
+      "loss": 1.9635,
+      "step": 323505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013661933327034187,
+      "loss": 2.1743,
+      "step": 323510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013661761335717523,
+      "loss": 2.0038,
+      "step": 323515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001366158934314993,
+      "loss": 1.9689,
+      "step": 323520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013661417349331476,
+      "loss": 2.113,
+      "step": 323525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00013661245354262215,
+      "loss": 2.0011,
+      "step": 323530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013661073357942204,
+      "loss": 2.0093,
+      "step": 323535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013660901360371504,
+      "loss": 2.001,
+      "step": 323540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013660729361550175,
+      "loss": 1.806,
+      "step": 323545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013660557361478275,
+      "loss": 1.9455,
+      "step": 323550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001366038536015586,
+      "loss": 2.1808,
+      "step": 323555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013660213357582994,
+      "loss": 1.9889,
+      "step": 323560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.78125,
+      "learning_rate": 0.0001366004135375973,
+      "loss": 2.2451,
+      "step": 323565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001365986934868613,
+      "loss": 2.1408,
+      "step": 323570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013659697342362253,
+      "loss": 2.0747,
+      "step": 323575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013659525334788154,
+      "loss": 2.0152,
+      "step": 323580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.000136593533259639,
+      "loss": 1.9615,
+      "step": 323585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013659181315889542,
+      "loss": 2.0215,
+      "step": 323590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001365900930456514,
+      "loss": 1.8751,
+      "step": 323595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013658837291990756,
+      "loss": 2.1004,
+      "step": 323600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013658665278166448,
+      "loss": 2.1084,
+      "step": 323605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013658493263092272,
+      "loss": 2.0413,
+      "step": 323610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013658321246768287,
+      "loss": 2.0961,
+      "step": 323615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013658149229194553,
+      "loss": 1.9662,
+      "step": 323620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001365797721037113,
+      "loss": 2.0212,
+      "step": 323625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001365780519029808,
+      "loss": 2.0882,
+      "step": 323630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001365763316897545,
+      "loss": 1.8475,
+      "step": 323635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013657461146403306,
+      "loss": 2.0917,
+      "step": 323640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.0001365728912258171,
+      "loss": 2.2899,
+      "step": 323645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013657117097510725,
+      "loss": 2.0915,
+      "step": 323650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013656945071190393,
+      "loss": 2.1957,
+      "step": 323655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013656773043620787,
+      "loss": 2.153,
+      "step": 323660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013656601014801954,
+      "loss": 1.9534,
+      "step": 323665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013656428984733964,
+      "loss": 2.1629,
+      "step": 323670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013656256953416875,
+      "loss": 1.9586,
+      "step": 323675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001365608492085074,
+      "loss": 1.858,
+      "step": 323680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001365591288703562,
+      "loss": 2.0874,
+      "step": 323685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013655740851971574,
+      "loss": 1.8747,
+      "step": 323690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013655568815658657,
+      "loss": 2.0059,
+      "step": 323695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013655396778096933,
+      "loss": 2.19,
+      "step": 323700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013655224739286462,
+      "loss": 1.9412,
+      "step": 323705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013655052699227297,
+      "loss": 1.9738,
+      "step": 323710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013654880657919502,
+      "loss": 2.0166,
+      "step": 323715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001365470861536313,
+      "loss": 2.0087,
+      "step": 323720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001365453657155825,
+      "loss": 2.2131,
+      "step": 323725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013654364526504909,
+      "loss": 2.0688,
+      "step": 323730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013654192480203173,
+      "loss": 2.2557,
+      "step": 323735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013654020432653097,
+      "loss": 2.1001,
+      "step": 323740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013653848383854742,
+      "loss": 2.149,
+      "step": 323745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013653676333808164,
+      "loss": 1.9575,
+      "step": 323750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013653504282513425,
+      "loss": 1.8919,
+      "step": 323755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013653332229970585,
+      "loss": 1.915,
+      "step": 323760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000136531601761797,
+      "loss": 1.9356,
+      "step": 323765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001365298812114083,
+      "loss": 2.0665,
+      "step": 323770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013652816064854028,
+      "loss": 2.082,
+      "step": 323775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013652644007319358,
+      "loss": 1.9467,
+      "step": 323780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013652471948536883,
+      "loss": 1.9753,
+      "step": 323785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013652299888506657,
+      "loss": 2.1746,
+      "step": 323790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001365212782722874,
+      "loss": 2.1859,
+      "step": 323795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013651955764703183,
+      "loss": 1.9945,
+      "step": 323800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013651783700930058,
+      "loss": 1.9206,
+      "step": 323805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013651611635909417,
+      "loss": 2.0478,
+      "step": 323810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013651439569641318,
+      "loss": 2.0702,
+      "step": 323815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013651267502125822,
+      "loss": 2.0662,
+      "step": 323820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013651095433362987,
+      "loss": 1.9911,
+      "step": 323825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001365092336335287,
+      "loss": 1.8892,
+      "step": 323830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001365075129209553,
+      "loss": 1.9736,
+      "step": 323835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001365057921959103,
+      "loss": 2.0144,
+      "step": 323840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013650407145839423,
+      "loss": 2.0612,
+      "step": 323845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013650235070840772,
+      "loss": 1.9693,
+      "step": 323850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001365006299459514,
+      "loss": 1.919,
+      "step": 323855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013649890917102572,
+      "loss": 1.8929,
+      "step": 323860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013649718838363138,
+      "loss": 2.2077,
+      "step": 323865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013649546758376895,
+      "loss": 2.0786,
+      "step": 323870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000136493746771439,
+      "loss": 1.9401,
+      "step": 323875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013649202594664215,
+      "loss": 2.1649,
+      "step": 323880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001364903051093789,
+      "loss": 2.0835,
+      "step": 323885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013648858425964995,
+      "loss": 1.9396,
+      "step": 323890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013648686339745583,
+      "loss": 2.003,
+      "step": 323895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013648514252279713,
+      "loss": 2.0527,
+      "step": 323900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013648342163567447,
+      "loss": 2.1819,
+      "step": 323905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013648170073608836,
+      "loss": 2.0081,
+      "step": 323910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013647997982403946,
+      "loss": 2.0308,
+      "step": 323915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013647825889952835,
+      "loss": 2.0862,
+      "step": 323920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013647653796255562,
+      "loss": 1.9551,
+      "step": 323925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013647481701312182,
+      "loss": 1.9121,
+      "step": 323930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013647309605122754,
+      "loss": 2.1015,
+      "step": 323935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001364713750768734,
+      "loss": 1.9953,
+      "step": 323940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013646965409006,
+      "loss": 1.8526,
+      "step": 323945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013646793309078788,
+      "loss": 2.0585,
+      "step": 323950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013646621207905768,
+      "loss": 1.8597,
+      "step": 323955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013646449105486993,
+      "loss": 2.0165,
+      "step": 323960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013646277001822524,
+      "loss": 2.1251,
+      "step": 323965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013646104896912427,
+      "loss": 2.0787,
+      "step": 323970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013645932790756752,
+      "loss": 1.9013,
+      "step": 323975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013645760683355558,
+      "loss": 2.0962,
+      "step": 323980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013645588574708904,
+      "loss": 2.0269,
+      "step": 323985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013645416464816857,
+      "loss": 2.0945,
+      "step": 323990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013645244353679465,
+      "loss": 2.0481,
+      "step": 323995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013645072241296792,
+      "loss": 1.9309,
+      "step": 324000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013644900127668895,
+      "loss": 2.0878,
+      "step": 324005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001364472801279584,
+      "loss": 1.9371,
+      "step": 324010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013644555896677674,
+      "loss": 1.8159,
+      "step": 324015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013644383779314463,
+      "loss": 2.157,
+      "step": 324020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013644211660706262,
+      "loss": 2.1891,
+      "step": 324025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013644039540853137,
+      "loss": 2.2136,
+      "step": 324030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001364386741975514,
+      "loss": 2.1987,
+      "step": 324035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013643695297412334,
+      "loss": 2.1495,
+      "step": 324040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001364352317382477,
+      "loss": 2.1212,
+      "step": 324045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013643351048992517,
+      "loss": 2.2711,
+      "step": 324050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013643178922915627,
+      "loss": 2.0207,
+      "step": 324055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013643006795594165,
+      "loss": 2.0313,
+      "step": 324060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013642834667028183,
+      "loss": 1.9722,
+      "step": 324065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.0001364266253721774,
+      "loss": 2.0155,
+      "step": 324070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013642490406162898,
+      "loss": 2.0832,
+      "step": 324075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013642318273863717,
+      "loss": 2.0348,
+      "step": 324080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013642146140320255,
+      "loss": 2.0137,
+      "step": 324085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001364197400553257,
+      "loss": 1.9277,
+      "step": 324090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013641801869500718,
+      "loss": 2.1481,
+      "step": 324095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013641629732224762,
+      "loss": 1.9119,
+      "step": 324100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001364145759370476,
+      "loss": 2.1785,
+      "step": 324105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001364128545394077,
+      "loss": 2.065,
+      "step": 324110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013641113312932846,
+      "loss": 2.0173,
+      "step": 324115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013640941170681059,
+      "loss": 2.0988,
+      "step": 324120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013640769027185455,
+      "loss": 2.1024,
+      "step": 324125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000136405968824461,
+      "loss": 2.0436,
+      "step": 324130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013640424736463053,
+      "loss": 1.9063,
+      "step": 324135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013640252589236369,
+      "loss": 1.998,
+      "step": 324140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001364008044076611,
+      "loss": 1.9043,
+      "step": 324145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00013639908291052333,
+      "loss": 1.8844,
+      "step": 324150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013639736140095096,
+      "loss": 1.8192,
+      "step": 324155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001363956398789446,
+      "loss": 1.9701,
+      "step": 324160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013639391834450483,
+      "loss": 2.0212,
+      "step": 324165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013639219679763222,
+      "loss": 2.1567,
+      "step": 324170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001363904752383274,
+      "loss": 2.0165,
+      "step": 324175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013638875366659092,
+      "loss": 2.0888,
+      "step": 324180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013638703208242338,
+      "loss": 2.0256,
+      "step": 324185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013638531048582538,
+      "loss": 2.1613,
+      "step": 324190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001363835888767975,
+      "loss": 2.0175,
+      "step": 324195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013638186725534036,
+      "loss": 2.1266,
+      "step": 324200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013638014562145445,
+      "loss": 2.1422,
+      "step": 324205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013637842397514044,
+      "loss": 1.933,
+      "step": 324210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013637670231639892,
+      "loss": 2.0464,
+      "step": 324215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013637498064523044,
+      "loss": 2.0246,
+      "step": 324220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013637325896163565,
+      "loss": 2.0368,
+      "step": 324225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000136371537265615,
+      "loss": 2.2101,
+      "step": 324230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013636981555716925,
+      "loss": 2.0927,
+      "step": 324235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001363680938362989,
+      "loss": 2.146,
+      "step": 324240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013636637210300457,
+      "loss": 2.2074,
+      "step": 324245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001363646503572868,
+      "loss": 2.0592,
+      "step": 324250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001363629285991462,
+      "loss": 2.1548,
+      "step": 324255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013636120682858336,
+      "loss": 2.228,
+      "step": 324260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001363594850455989,
+      "loss": 2.0415,
+      "step": 324265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013635776325019336,
+      "loss": 1.8727,
+      "step": 324270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013635604144236735,
+      "loss": 2.1575,
+      "step": 324275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013635431962212146,
+      "loss": 1.9858,
+      "step": 324280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001363525977894563,
+      "loss": 2.0602,
+      "step": 324285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013635087594437238,
+      "loss": 2.0862,
+      "step": 324290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013634915408687037,
+      "loss": 2.1287,
+      "step": 324295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013634743221695083,
+      "loss": 2.1982,
+      "step": 324300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013634571033461435,
+      "loss": 2.1352,
+      "step": 324305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013634398843986151,
+      "loss": 1.9936,
+      "step": 324310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001363422665326929,
+      "loss": 2.0874,
+      "step": 324315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013634054461310915,
+      "loss": 2.062,
+      "step": 324320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013633882268111077,
+      "loss": 2.2664,
+      "step": 324325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013633710073669838,
+      "loss": 2.0214,
+      "step": 324330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013633537877987264,
+      "loss": 2.1573,
+      "step": 324335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013633365681063402,
+      "loss": 2.1872,
+      "step": 324340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013633193482898316,
+      "loss": 1.9724,
+      "step": 324345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013633021283492066,
+      "loss": 1.9975,
+      "step": 324350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013632849082844715,
+      "loss": 2.0803,
+      "step": 324355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001363267688095631,
+      "loss": 1.8616,
+      "step": 324360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001363250467782692,
+      "loss": 2.0324,
+      "step": 324365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000136323324734566,
+      "loss": 2.0652,
+      "step": 324370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013632160267845408,
+      "loss": 2.0396,
+      "step": 324375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013631988060993407,
+      "loss": 1.9979,
+      "step": 324380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013631815852900654,
+      "loss": 2.1029,
+      "step": 324385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.000136316436435672,
+      "loss": 2.0567,
+      "step": 324390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013631471432993118,
+      "loss": 1.804,
+      "step": 324395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013631299221178458,
+      "loss": 1.923,
+      "step": 324400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013631127008123279,
+      "loss": 2.0113,
+      "step": 324405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013630954793827642,
+      "loss": 2.0941,
+      "step": 324410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013630782578291604,
+      "loss": 1.943,
+      "step": 324415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013630610361515223,
+      "loss": 1.9505,
+      "step": 324420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00013630438143498564,
+      "loss": 2.1476,
+      "step": 324425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001363026592424168,
+      "loss": 1.9975,
+      "step": 324430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013630093703744629,
+      "loss": 2.126,
+      "step": 324435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013629921482007472,
+      "loss": 2.0196,
+      "step": 324440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001362974925903027,
+      "loss": 2.0846,
+      "step": 324445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013629577034813078,
+      "loss": 2.1047,
+      "step": 324450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001362940480935596,
+      "loss": 2.0496,
+      "step": 324455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001362923258265897,
+      "loss": 2.1519,
+      "step": 324460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013629060354722168,
+      "loss": 2.0998,
+      "step": 324465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013628888125545615,
+      "loss": 2.1906,
+      "step": 324470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013628715895129366,
+      "loss": 2.068,
+      "step": 324475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013628543663473478,
+      "loss": 2.0334,
+      "step": 324480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001362837143057802,
+      "loss": 2.0255,
+      "step": 324485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013628199196443042,
+      "loss": 1.9494,
+      "step": 324490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013628026961068606,
+      "loss": 2.0271,
+      "step": 324495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013627854724454768,
+      "loss": 1.9079,
+      "step": 324500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001362768248660159,
+      "loss": 2.0297,
+      "step": 324505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001362751024750913,
+      "loss": 2.1798,
+      "step": 324510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001362733800717745,
+      "loss": 2.174,
+      "step": 324515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013627165765606604,
+      "loss": 2.1409,
+      "step": 324520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013626993522796647,
+      "loss": 1.9948,
+      "step": 324525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001362682127874765,
+      "loss": 2.2069,
+      "step": 324530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013626649033459664,
+      "loss": 2.1142,
+      "step": 324535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013626476786932746,
+      "loss": 2.042,
+      "step": 324540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001362630453916696,
+      "loss": 2.0376,
+      "step": 324545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001362613229016236,
+      "loss": 2.1757,
+      "step": 324550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001362596003991901,
+      "loss": 2.0361,
+      "step": 324555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013625787788436967,
+      "loss": 2.1195,
+      "step": 324560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00013625615535716286,
+      "loss": 1.9081,
+      "step": 324565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013625443281757032,
+      "loss": 2.2242,
+      "step": 324570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013625271026559258,
+      "loss": 2.0999,
+      "step": 324575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013625098770123027,
+      "loss": 2.07,
+      "step": 324580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013624926512448396,
+      "loss": 1.9275,
+      "step": 324585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013624754253535425,
+      "loss": 2.0356,
+      "step": 324590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013624581993384172,
+      "loss": 2.2796,
+      "step": 324595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.000136244097319947,
+      "loss": 2.0499,
+      "step": 324600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013624237469367058,
+      "loss": 2.0366,
+      "step": 324605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013624065205501313,
+      "loss": 2.0635,
+      "step": 324610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013623892940397523,
+      "loss": 1.9449,
+      "step": 324615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013623720674055742,
+      "loss": 1.8571,
+      "step": 324620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013623548406476035,
+      "loss": 2.1099,
+      "step": 324625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001362337613765846,
+      "loss": 1.8834,
+      "step": 324630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001362320386760307,
+      "loss": 2.0047,
+      "step": 324635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001362303159630993,
+      "loss": 2.051,
+      "step": 324640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013622859323779093,
+      "loss": 2.0934,
+      "step": 324645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013622687050010628,
+      "loss": 2.0841,
+      "step": 324650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013622514775004586,
+      "loss": 2.1582,
+      "step": 324655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001362234249876102,
+      "loss": 2.0591,
+      "step": 324660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013622170221280002,
+      "loss": 1.9943,
+      "step": 324665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013621997942561584,
+      "loss": 2.1328,
+      "step": 324670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013621825662605828,
+      "loss": 2.1331,
+      "step": 324675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001362165338141279,
+      "loss": 2.1386,
+      "step": 324680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.40625,
+      "learning_rate": 0.00013621481098982524,
+      "loss": 1.9159,
+      "step": 324685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000136213088153151,
+      "loss": 2.2214,
+      "step": 324690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013621136530410572,
+      "loss": 2.156,
+      "step": 324695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013620964244268996,
+      "loss": 2.0734,
+      "step": 324700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013620791956890432,
+      "loss": 2.1835,
+      "step": 324705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013620619668274942,
+      "loss": 2.0557,
+      "step": 324710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001362044737842258,
+      "loss": 2.1614,
+      "step": 324715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013620275087333409,
+      "loss": 2.1053,
+      "step": 324720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013620102795007486,
+      "loss": 2.016,
+      "step": 324725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00013619930501444872,
+      "loss": 2.021,
+      "step": 324730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013619758206645622,
+      "loss": 1.9448,
+      "step": 324735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013619585910609798,
+      "loss": 2.0766,
+      "step": 324740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013619413613337456,
+      "loss": 1.977,
+      "step": 324745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013619241314828656,
+      "loss": 2.0425,
+      "step": 324750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001361906901508346,
+      "loss": 1.9604,
+      "step": 324755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013618896714101927,
+      "loss": 2.0327,
+      "step": 324760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013618724411884107,
+      "loss": 2.083,
+      "step": 324765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001361855210843007,
+      "loss": 2.0697,
+      "step": 324770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001361837980373987,
+      "loss": 2.1585,
+      "step": 324775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013618207497813562,
+      "loss": 1.9741,
+      "step": 324780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001361803519065121,
+      "loss": 2.1063,
+      "step": 324785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013617862882252875,
+      "loss": 2.1238,
+      "step": 324790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001361769057261861,
+      "loss": 1.9914,
+      "step": 324795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013617518261748474,
+      "loss": 1.8335,
+      "step": 324800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001361734594964253,
+      "loss": 2.0747,
+      "step": 324805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013617173636300836,
+      "loss": 1.9058,
+      "step": 324810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001361700132172345,
+      "loss": 2.1615,
+      "step": 324815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013616829005910427,
+      "loss": 2.2113,
+      "step": 324820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013616656688861833,
+      "loss": 1.9144,
+      "step": 324825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013616484370577723,
+      "loss": 1.9042,
+      "step": 324830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001361631205105816,
+      "loss": 2.0292,
+      "step": 324835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013616139730303196,
+      "loss": 2.2444,
+      "step": 324840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001361596740831289,
+      "loss": 2.1539,
+      "step": 324845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013615795085087306,
+      "loss": 2.1957,
+      "step": 324850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013615622760626504,
+      "loss": 2.1332,
+      "step": 324855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013615450434930536,
+      "loss": 2.1179,
+      "step": 324860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013615278107999466,
+      "loss": 1.9881,
+      "step": 324865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001361510577983335,
+      "loss": 1.9755,
+      "step": 324870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001361493345043225,
+      "loss": 2.0998,
+      "step": 324875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013614761119796223,
+      "loss": 2.1146,
+      "step": 324880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013614588787925328,
+      "loss": 2.145,
+      "step": 324885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013614416454819624,
+      "loss": 2.0117,
+      "step": 324890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013614244120479168,
+      "loss": 2.0817,
+      "step": 324895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013614071784904023,
+      "loss": 2.078,
+      "step": 324900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013613899448094242,
+      "loss": 1.9294,
+      "step": 324905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013613727110049892,
+      "loss": 1.9885,
+      "step": 324910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013613554770771025,
+      "loss": 1.8491,
+      "step": 324915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00013613382430257702,
+      "loss": 1.9539,
+      "step": 324920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013613210088509982,
+      "loss": 2.0472,
+      "step": 324925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013613037745527926,
+      "loss": 2.14,
+      "step": 324930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013612865401311586,
+      "loss": 1.9934,
+      "step": 324935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001361269305586103,
+      "loss": 2.0407,
+      "step": 324940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013612520709176313,
+      "loss": 2.0683,
+      "step": 324945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001361234836125749,
+      "loss": 2.0789,
+      "step": 324950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013612176012104625,
+      "loss": 2.1221,
+      "step": 324955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013612003661717776,
+      "loss": 2.1159,
+      "step": 324960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013611831310097,
+      "loss": 2.1126,
+      "step": 324965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013611658957242358,
+      "loss": 2.097,
+      "step": 324970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013611486603153908,
+      "loss": 2.2772,
+      "step": 324975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013611314247831703,
+      "loss": 2.0614,
+      "step": 324980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00013611141891275815,
+      "loss": 1.9271,
+      "step": 324985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013610969533486294,
+      "loss": 2.1745,
+      "step": 324990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.000136107971744632,
+      "loss": 2.0511,
+      "step": 324995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013610624814206588,
+      "loss": 2.0689,
+      "step": 325000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013610452452716526,
+      "loss": 1.9004,
+      "step": 325005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013610280089993066,
+      "loss": 2.133,
+      "step": 325010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013610107726036268,
+      "loss": 2.0598,
+      "step": 325015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013609935360846194,
+      "loss": 2.2012,
+      "step": 325020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013609762994422898,
+      "loss": 2.0288,
+      "step": 325025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001360959062676644,
+      "loss": 1.9626,
+      "step": 325030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013609418257876888,
+      "loss": 2.0341,
+      "step": 325035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013609245887754288,
+      "loss": 1.9689,
+      "step": 325040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013609073516398704,
+      "loss": 1.8944,
+      "step": 325045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013608901143810195,
+      "loss": 1.9932,
+      "step": 325050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001360872876998882,
+      "loss": 2.0129,
+      "step": 325055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001360855639493464,
+      "loss": 2.0407,
+      "step": 325060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001360838401864771,
+      "loss": 2.035,
+      "step": 325065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013608211641128089,
+      "loss": 2.041,
+      "step": 325070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001360803926237584,
+      "loss": 2.1204,
+      "step": 325075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001360786688239102,
+      "loss": 2.086,
+      "step": 325080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013607694501173684,
+      "loss": 2.2307,
+      "step": 325085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013607522118723895,
+      "loss": 2.0228,
+      "step": 325090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013607349735041713,
+      "loss": 2.16,
+      "step": 325095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001360717735012719,
+      "loss": 1.9951,
+      "step": 325100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00013607004963980398,
+      "loss": 2.0687,
+      "step": 325105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001360683257660138,
+      "loss": 2.162,
+      "step": 325110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013606660187990207,
+      "loss": 2.0448,
+      "step": 325115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013606487798146932,
+      "loss": 1.9325,
+      "step": 325120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013606315407071616,
+      "loss": 1.9843,
+      "step": 325125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013606143014764318,
+      "loss": 2.1809,
+      "step": 325130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00013605970621225096,
+      "loss": 2.0262,
+      "step": 325135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013605798226454005,
+      "loss": 1.9935,
+      "step": 325140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013605625830451112,
+      "loss": 2.0215,
+      "step": 325145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013605453433216473,
+      "loss": 2.0972,
+      "step": 325150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013605281034750143,
+      "loss": 1.9392,
+      "step": 325155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013605108635052183,
+      "loss": 2.0612,
+      "step": 325160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013604936234122653,
+      "loss": 1.9372,
+      "step": 325165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013604763831961614,
+      "loss": 2.1126,
+      "step": 325170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001360459142856912,
+      "loss": 2.0888,
+      "step": 325175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013604419023945232,
+      "loss": 2.0958,
+      "step": 325180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001360424661809001,
+      "loss": 2.1048,
+      "step": 325185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001360407421100351,
+      "loss": 2.0608,
+      "step": 325190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013603901802685796,
+      "loss": 2.0154,
+      "step": 325195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013603729393136922,
+      "loss": 2.0365,
+      "step": 325200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001360355698235695,
+      "loss": 2.1935,
+      "step": 325205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013603384570345936,
+      "loss": 2.0316,
+      "step": 325210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013603212157103938,
+      "loss": 2.1128,
+      "step": 325215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013603039742631024,
+      "loss": 1.9318,
+      "step": 325220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013602867326927242,
+      "loss": 1.9107,
+      "step": 325225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013602694909992655,
+      "loss": 2.1539,
+      "step": 325230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013602522491827323,
+      "loss": 1.9453,
+      "step": 325235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013602350072431307,
+      "loss": 1.9413,
+      "step": 325240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001360217765180466,
+      "loss": 2.0606,
+      "step": 325245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013602005229947442,
+      "loss": 2.034,
+      "step": 325250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013601832806859712,
+      "loss": 2.0464,
+      "step": 325255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013601660382541538,
+      "loss": 2.0601,
+      "step": 325260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013601487956992967,
+      "loss": 2.1224,
+      "step": 325265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013601315530214063,
+      "loss": 2.0931,
+      "step": 325270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013601143102204882,
+      "loss": 2.1331,
+      "step": 325275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013600970672965486,
+      "loss": 2.1507,
+      "step": 325280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001360079824249594,
+      "loss": 2.1304,
+      "step": 325285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013600625810796288,
+      "loss": 2.1636,
+      "step": 325290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.000136004533778666,
+      "loss": 2.1241,
+      "step": 325295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001360028094370693,
+      "loss": 2.2981,
+      "step": 325300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001360010850831734,
+      "loss": 2.1318,
+      "step": 325305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013599936071697887,
+      "loss": 2.1049,
+      "step": 325310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013599763633848633,
+      "loss": 2.0813,
+      "step": 325315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013599591194769631,
+      "loss": 2.0511,
+      "step": 325320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013599418754460944,
+      "loss": 2.1205,
+      "step": 325325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013599246312922633,
+      "loss": 2.1824,
+      "step": 325330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013599073870154753,
+      "loss": 2.029,
+      "step": 325335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013598901426157363,
+      "loss": 2.1176,
+      "step": 325340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00013598728980930523,
+      "loss": 1.9683,
+      "step": 325345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013598556534474292,
+      "loss": 1.9667,
+      "step": 325350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013598384086788728,
+      "loss": 2.1446,
+      "step": 325355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013598211637873893,
+      "loss": 2.0304,
+      "step": 325360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013598039187729844,
+      "loss": 1.9738,
+      "step": 325365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013597866736356638,
+      "loss": 2.1528,
+      "step": 325370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013597694283754334,
+      "loss": 2.035,
+      "step": 325375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00013597521829922997,
+      "loss": 2.1267,
+      "step": 325380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013597349374862676,
+      "loss": 2.1688,
+      "step": 325385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013597176918573437,
+      "loss": 1.9544,
+      "step": 325390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013597004461055338,
+      "loss": 1.9744,
+      "step": 325395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.0001359683200230844,
+      "loss": 2.0239,
+      "step": 325400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013596659542332794,
+      "loss": 1.9878,
+      "step": 325405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013596487081128463,
+      "loss": 2.0339,
+      "step": 325410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001359631461869551,
+      "loss": 1.9417,
+      "step": 325415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013596142155033992,
+      "loss": 1.9991,
+      "step": 325420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013595969690143964,
+      "loss": 1.9436,
+      "step": 325425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013595797224025488,
+      "loss": 2.1122,
+      "step": 325430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001359562475667862,
+      "loss": 2.0047,
+      "step": 325435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013595452288103424,
+      "loss": 2.0069,
+      "step": 325440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001359527981829996,
+      "loss": 1.8845,
+      "step": 325445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013595107347268279,
+      "loss": 2.0964,
+      "step": 325450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013594934875008446,
+      "loss": 2.1278,
+      "step": 325455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013594762401520515,
+      "loss": 2.1096,
+      "step": 325460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013594589926804547,
+      "loss": 2.0487,
+      "step": 325465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001359441745086061,
+      "loss": 2.2051,
+      "step": 325470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00013594244973688747,
+      "loss": 1.8307,
+      "step": 325475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013594072495289026,
+      "loss": 1.9932,
+      "step": 325480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013593900015661508,
+      "loss": 1.9379,
+      "step": 325485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013593727534806245,
+      "loss": 2.0221,
+      "step": 325490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013593555052723302,
+      "loss": 1.8123,
+      "step": 325495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013593382569412733,
+      "loss": 2.002,
+      "step": 325500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013593210084874598,
+      "loss": 2.0256,
+      "step": 325505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013593037599108962,
+      "loss": 2.0992,
+      "step": 325510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013592865112115878,
+      "loss": 2.0432,
+      "step": 325515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013592692623895405,
+      "loss": 2.0845,
+      "step": 325520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013592520134447601,
+      "loss": 2.1279,
+      "step": 325525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001359234764377253,
+      "loss": 2.1934,
+      "step": 325530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013592175151870247,
+      "loss": 2.0866,
+      "step": 325535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013592002658740815,
+      "loss": 2.0352,
+      "step": 325540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013591830164384283,
+      "loss": 2.1174,
+      "step": 325545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013591657668800722,
+      "loss": 2.0489,
+      "step": 325550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013591485171990183,
+      "loss": 1.9433,
+      "step": 325555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001359131267395273,
+      "loss": 1.9811,
+      "step": 325560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001359114017468842,
+      "loss": 2.1458,
+      "step": 325565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013590967674197306,
+      "loss": 2.0828,
+      "step": 325570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013590795172479454,
+      "loss": 2.1474,
+      "step": 325575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013590622669534924,
+      "loss": 1.969,
+      "step": 325580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013590450165363772,
+      "loss": 2.0258,
+      "step": 325585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013590277659966056,
+      "loss": 2.0823,
+      "step": 325590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013590105153341833,
+      "loss": 1.9534,
+      "step": 325595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013589932645491172,
+      "loss": 2.0445,
+      "step": 325600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001358976013641412,
+      "loss": 1.9743,
+      "step": 325605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013589587626110743,
+      "loss": 2.3149,
+      "step": 325610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013589415114581097,
+      "loss": 1.965,
+      "step": 325615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001358924260182524,
+      "loss": 2.1743,
+      "step": 325620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013589070087843233,
+      "loss": 2.0208,
+      "step": 325625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013588897572635138,
+      "loss": 2.0415,
+      "step": 325630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001358872505620101,
+      "loss": 2.0664,
+      "step": 325635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013588552538540904,
+      "loss": 2.0945,
+      "step": 325640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013588380019654886,
+      "loss": 2.1762,
+      "step": 325645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001358820749954301,
+      "loss": 2.0621,
+      "step": 325650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013588034978205343,
+      "loss": 1.9372,
+      "step": 325655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013587862455641933,
+      "loss": 2.0106,
+      "step": 325660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013587689931852844,
+      "loss": 2.0545,
+      "step": 325665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013587517406838138,
+      "loss": 2.0098,
+      "step": 325670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001358734488059787,
+      "loss": 2.198,
+      "step": 325675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013587172353132098,
+      "loss": 2.0555,
+      "step": 325680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013586999824440884,
+      "loss": 2.0742,
+      "step": 325685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013586827294524286,
+      "loss": 2.1204,
+      "step": 325690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00013586654763382368,
+      "loss": 1.8114,
+      "step": 325695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013586482231015177,
+      "loss": 1.8934,
+      "step": 325700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013586309697422778,
+      "loss": 2.17,
+      "step": 325705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013586137162605234,
+      "loss": 2.0193,
+      "step": 325710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000135859646265626,
+      "loss": 1.9272,
+      "step": 325715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013585792089294935,
+      "loss": 2.0571,
+      "step": 325720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000135856195508023,
+      "loss": 2.0269,
+      "step": 325725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013585447011084747,
+      "loss": 1.8911,
+      "step": 325730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013585274470142344,
+      "loss": 2.1408,
+      "step": 325735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013585101927975147,
+      "loss": 2.0325,
+      "step": 325740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013584929384583214,
+      "loss": 1.9142,
+      "step": 325745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013584756839966604,
+      "loss": 1.9604,
+      "step": 325750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013584584294125374,
+      "loss": 2.0169,
+      "step": 325755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001358441174705959,
+      "loss": 2.1834,
+      "step": 325760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.000135842391987693,
+      "loss": 2.0485,
+      "step": 325765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013584066649254572,
+      "loss": 2.1201,
+      "step": 325770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001358389409851546,
+      "loss": 1.9832,
+      "step": 325775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013583721546552027,
+      "loss": 1.909,
+      "step": 325780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.00013583548993364327,
+      "loss": 1.9838,
+      "step": 325785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013583376438952425,
+      "loss": 2.1111,
+      "step": 325790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013583203883316375,
+      "loss": 2.0942,
+      "step": 325795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013583031326456238,
+      "loss": 2.1418,
+      "step": 325800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001358285876837207,
+      "loss": 1.9898,
+      "step": 325805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013582686209063935,
+      "loss": 1.7625,
+      "step": 325810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001358251364853189,
+      "loss": 2.0496,
+      "step": 325815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001358234108677599,
+      "loss": 2.0346,
+      "step": 325820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013582168523796302,
+      "loss": 2.147,
+      "step": 325825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00013581995959592878,
+      "loss": 2.2766,
+      "step": 325830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013581823394165777,
+      "loss": 2.0651,
+      "step": 325835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013581650827515062,
+      "loss": 2.1317,
+      "step": 325840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013581478259640792,
+      "loss": 2.2203,
+      "step": 325845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001358130569054302,
+      "loss": 2.019,
+      "step": 325850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013581133120221814,
+      "loss": 2.1446,
+      "step": 325855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013580960548677227,
+      "loss": 2.037,
+      "step": 325860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00013580787975909315,
+      "loss": 1.9365,
+      "step": 325865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013580615401918142,
+      "loss": 1.8891,
+      "step": 325870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001358044282670377,
+      "loss": 2.0926,
+      "step": 325875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013580270250266249,
+      "loss": 1.9898,
+      "step": 325880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013580097672605645,
+      "loss": 1.926,
+      "step": 325885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013579925093722012,
+      "loss": 1.9245,
+      "step": 325890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013579752513615415,
+      "loss": 2.0167,
+      "step": 325895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013579579932285912,
+      "loss": 1.9888,
+      "step": 325900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013579407349733556,
+      "loss": 1.9222,
+      "step": 325905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001357923476595841,
+      "loss": 2.146,
+      "step": 325910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013579062180960532,
+      "loss": 2.1543,
+      "step": 325915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013578889594739982,
+      "loss": 2.0618,
+      "step": 325920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013578717007296817,
+      "loss": 2.1158,
+      "step": 325925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013578544418631097,
+      "loss": 2.1245,
+      "step": 325930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013578371828742883,
+      "loss": 2.1881,
+      "step": 325935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013578199237632233,
+      "loss": 2.0515,
+      "step": 325940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013578026645299205,
+      "loss": 1.9292,
+      "step": 325945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013577854051743853,
+      "loss": 2.22,
+      "step": 325950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013577681456966248,
+      "loss": 2.2213,
+      "step": 325955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001357750886096644,
+      "loss": 2.0271,
+      "step": 325960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001357733626374449,
+      "loss": 1.9872,
+      "step": 325965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013577163665300456,
+      "loss": 2.226,
+      "step": 325970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000135769910656344,
+      "loss": 2.0048,
+      "step": 325975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013576818464746377,
+      "loss": 2.0416,
+      "step": 325980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001357664586263645,
+      "loss": 2.157,
+      "step": 325985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013576473259304677,
+      "loss": 2.0697,
+      "step": 325990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013576300654751113,
+      "loss": 2.1463,
+      "step": 325995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001357612804897582,
+      "loss": 2.1213,
+      "step": 326000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013575955441978856,
+      "loss": 2.0619,
+      "step": 326005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013575782833760282,
+      "loss": 2.0394,
+      "step": 326010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013575610224320157,
+      "loss": 2.0808,
+      "step": 326015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013575437613658537,
+      "loss": 2.0749,
+      "step": 326020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013575265001775482,
+      "loss": 2.1514,
+      "step": 326025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001357509238867105,
+      "loss": 2.1439,
+      "step": 326030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001357491977434531,
+      "loss": 2.0116,
+      "step": 326035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013574747158798305,
+      "loss": 2.2415,
+      "step": 326040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013574574542030106,
+      "loss": 1.9819,
+      "step": 326045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013574401924040761,
+      "loss": 1.9423,
+      "step": 326050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013574229304830343,
+      "loss": 2.0538,
+      "step": 326055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013574056684398898,
+      "loss": 1.8603,
+      "step": 326060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013573884062746493,
+      "loss": 2.0951,
+      "step": 326065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001357371143987318,
+      "loss": 1.9718,
+      "step": 326070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001357353881577903,
+      "loss": 2.1041,
+      "step": 326075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001357336619046409,
+      "loss": 1.9978,
+      "step": 326080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013573193563928425,
+      "loss": 2.0197,
+      "step": 326085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001357302093617209,
+      "loss": 2.0282,
+      "step": 326090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013572848307195147,
+      "loss": 2.1434,
+      "step": 326095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013572675676997654,
+      "loss": 2.0226,
+      "step": 326100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001357250304557967,
+      "loss": 2.0761,
+      "step": 326105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00013572330412941255,
+      "loss": 1.8581,
+      "step": 326110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013572157779082468,
+      "loss": 2.2203,
+      "step": 326115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013571985144003365,
+      "loss": 2.1146,
+      "step": 326120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013571812507704012,
+      "loss": 1.9125,
+      "step": 326125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013571639870184457,
+      "loss": 1.9834,
+      "step": 326130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013571467231444765,
+      "loss": 2.078,
+      "step": 326135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013571294591484997,
+      "loss": 2.1256,
+      "step": 326140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013571121950305213,
+      "loss": 2.0974,
+      "step": 326145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013570949307905466,
+      "loss": 1.968,
+      "step": 326150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001357077666428582,
+      "loss": 1.9582,
+      "step": 326155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013570604019446327,
+      "loss": 2.0532,
+      "step": 326160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013570431373387053,
+      "loss": 2.0746,
+      "step": 326165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001357025872610806,
+      "loss": 2.0943,
+      "step": 326170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013570086077609398,
+      "loss": 2.1035,
+      "step": 326175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001356991342789113,
+      "loss": 2.2169,
+      "step": 326180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013569740776953312,
+      "loss": 1.7846,
+      "step": 326185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001356956812479601,
+      "loss": 2.0153,
+      "step": 326190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013569395471419278,
+      "loss": 2.0075,
+      "step": 326195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013569222816823176,
+      "loss": 1.9103,
+      "step": 326200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013569050161007758,
+      "loss": 2.014,
+      "step": 326205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013568877503973093,
+      "loss": 2.0915,
+      "step": 326210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013568704845719234,
+      "loss": 1.9827,
+      "step": 326215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013568532186246238,
+      "loss": 1.9526,
+      "step": 326220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013568359525554168,
+      "loss": 1.9847,
+      "step": 326225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013568186863643081,
+      "loss": 2.1087,
+      "step": 326230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001356801420051304,
+      "loss": 2.212,
+      "step": 326235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000135678415361641,
+      "loss": 2.0268,
+      "step": 326240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013567668870596318,
+      "loss": 2.0258,
+      "step": 326245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013567496203809758,
+      "loss": 2.0989,
+      "step": 326250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001356732353580447,
+      "loss": 1.9757,
+      "step": 326255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013567150866580527,
+      "loss": 2.0783,
+      "step": 326260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001356697819613798,
+      "loss": 1.9894,
+      "step": 326265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013566805524476886,
+      "loss": 1.8124,
+      "step": 326270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013566632851597308,
+      "loss": 1.9219,
+      "step": 326275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013566460177499303,
+      "loss": 2.2138,
+      "step": 326280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013566287502182932,
+      "loss": 1.9694,
+      "step": 326285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0,
+      "learning_rate": 0.0001356611482564825,
+      "loss": 2.036,
+      "step": 326290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001356594214789532,
+      "loss": 1.9612,
+      "step": 326295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.000135657694689242,
+      "loss": 2.1318,
+      "step": 326300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001356559678873495,
+      "loss": 2.0029,
+      "step": 326305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013565424107327622,
+      "loss": 2.1174,
+      "step": 326310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013565251424702286,
+      "loss": 2.0528,
+      "step": 326315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001356507874085899,
+      "loss": 2.0921,
+      "step": 326320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000135649060557978,
+      "loss": 2.1297,
+      "step": 326325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013564733369518777,
+      "loss": 2.0741,
+      "step": 326330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013564560682021974,
+      "loss": 1.973,
+      "step": 326335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013564387993307455,
+      "loss": 1.8746,
+      "step": 326340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013564215303375272,
+      "loss": 1.8854,
+      "step": 326345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001356404261222549,
+      "loss": 2.0413,
+      "step": 326350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013563869919858167,
+      "loss": 2.086,
+      "step": 326355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001356369722627336,
+      "loss": 1.9946,
+      "step": 326360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013563524531471133,
+      "loss": 1.9343,
+      "step": 326365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001356335183545154,
+      "loss": 2.195,
+      "step": 326370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001356317913821464,
+      "loss": 1.9983,
+      "step": 326375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00013563006439760495,
+      "loss": 1.9688,
+      "step": 326380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001356283374008916,
+      "loss": 2.0468,
+      "step": 326385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013562661039200698,
+      "loss": 2.0047,
+      "step": 326390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013562488337095165,
+      "loss": 2.12,
+      "step": 326395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00013562315633772624,
+      "loss": 2.257,
+      "step": 326400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013562142929233128,
+      "loss": 2.2096,
+      "step": 326405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00013561970223476742,
+      "loss": 2.1611,
+      "step": 326410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013561797516503521,
+      "loss": 2.1214,
+      "step": 326415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013561624808313528,
+      "loss": 2.0844,
+      "step": 326420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013561452098906815,
+      "loss": 1.957,
+      "step": 326425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001356127938828345,
+      "loss": 1.9685,
+      "step": 326430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013561106676443485,
+      "loss": 2.164,
+      "step": 326435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001356093396338698,
+      "loss": 1.9371,
+      "step": 326440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013560761249114,
+      "loss": 1.8786,
+      "step": 326445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00013560588533624595,
+      "loss": 2.2007,
+      "step": 326450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013560415816918828,
+      "loss": 2.0784,
+      "step": 326455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013560243098996759,
+      "loss": 2.1259,
+      "step": 326460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001356007037985845,
+      "loss": 2.2063,
+      "step": 326465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013559897659503952,
+      "loss": 2.1841,
+      "step": 326470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013559724937933331,
+      "loss": 1.9976,
+      "step": 326475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001355955221514664,
+      "loss": 1.9569,
+      "step": 326480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013559379491143948,
+      "loss": 2.1132,
+      "step": 326485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013559206765925304,
+      "loss": 1.8713,
+      "step": 326490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013559034039490768,
+      "loss": 2.1139,
+      "step": 326495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013558861311840402,
+      "loss": 2.0727,
+      "step": 326500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013558688582974266,
+      "loss": 1.8911,
+      "step": 326505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013558515852892417,
+      "loss": 2.2295,
+      "step": 326510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013558343121594914,
+      "loss": 2.0131,
+      "step": 326515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013558170389081818,
+      "loss": 1.8346,
+      "step": 326520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013557997655353184,
+      "loss": 2.1708,
+      "step": 326525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013557824920409076,
+      "loss": 2.2608,
+      "step": 326530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001355765218424955,
+      "loss": 2.0374,
+      "step": 326535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013557479446874664,
+      "loss": 2.1104,
+      "step": 326540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013557306708284478,
+      "loss": 2.003,
+      "step": 326545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013557133968479054,
+      "loss": 2.0557,
+      "step": 326550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013556961227458446,
+      "loss": 1.9983,
+      "step": 326555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013556788485222717,
+      "loss": 1.9475,
+      "step": 326560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013556615741771926,
+      "loss": 2.0597,
+      "step": 326565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001355644299710613,
+      "loss": 2.0415,
+      "step": 326570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013556270251225387,
+      "loss": 1.8487,
+      "step": 326575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001355609750412976,
+      "loss": 2.0091,
+      "step": 326580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013555924755819304,
+      "loss": 2.0057,
+      "step": 326585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00013555752006294078,
+      "loss": 2.024,
+      "step": 326590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013555579255554142,
+      "loss": 2.0776,
+      "step": 326595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001355540650359956,
+      "loss": 1.836,
+      "step": 326600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013555233750430385,
+      "loss": 2.1515,
+      "step": 326605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013555060996046678,
+      "loss": 1.9913,
+      "step": 326610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013554888240448494,
+      "loss": 2.1724,
+      "step": 326615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00013554715483635898,
+      "loss": 2.0835,
+      "step": 326620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.0001355454272560895,
+      "loss": 2.0435,
+      "step": 326625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013554369966367702,
+      "loss": 2.0508,
+      "step": 326630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013554197205912217,
+      "loss": 2.1205,
+      "step": 326635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013554024444242555,
+      "loss": 2.0841,
+      "step": 326640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013553851681358774,
+      "loss": 2.0326,
+      "step": 326645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00013553678917260933,
+      "loss": 2.0197,
+      "step": 326650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013553506151949089,
+      "loss": 2.03,
+      "step": 326655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013553333385423304,
+      "loss": 1.9887,
+      "step": 326660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013553160617683636,
+      "loss": 2.1054,
+      "step": 326665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001355298784873014,
+      "loss": 2.2566,
+      "step": 326670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013552815078562885,
+      "loss": 1.935,
+      "step": 326675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001355264230718192,
+      "loss": 1.9683,
+      "step": 326680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.0001355246953458731,
+      "loss": 1.8304,
+      "step": 326685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001355229676077911,
+      "loss": 1.97,
+      "step": 326690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013552123985757384,
+      "loss": 2.1799,
+      "step": 326695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013551951209522187,
+      "loss": 2.0996,
+      "step": 326700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013551778432073576,
+      "loss": 2.029,
+      "step": 326705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013551605653411617,
+      "loss": 2.1246,
+      "step": 326710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013551432873536366,
+      "loss": 2.2044,
+      "step": 326715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013551260092447876,
+      "loss": 2.0314,
+      "step": 326720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013551087310146212,
+      "loss": 2.1531,
+      "step": 326725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013550914526631435,
+      "loss": 2.0463,
+      "step": 326730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.000135507417419036,
+      "loss": 2.0488,
+      "step": 326735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001355056895596277,
+      "loss": 2.0461,
+      "step": 326740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013550396168808998,
+      "loss": 2.0945,
+      "step": 326745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013550223380442343,
+      "loss": 2.1023,
+      "step": 326750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00013550050590862872,
+      "loss": 2.1536,
+      "step": 326755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013549877800070643,
+      "loss": 2.0211,
+      "step": 326760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013549705008065706,
+      "loss": 2.1107,
+      "step": 326765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013549532214848127,
+      "loss": 2.0683,
+      "step": 326770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001354935942041796,
+      "loss": 2.0624,
+      "step": 326775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001354918662477527,
+      "loss": 2.1175,
+      "step": 326780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013549013827920115,
+      "loss": 2.0871,
+      "step": 326785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013548841029852553,
+      "loss": 1.9869,
+      "step": 326790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001354866823057264,
+      "loss": 2.117,
+      "step": 326795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013548495430080437,
+      "loss": 2.0447,
+      "step": 326800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013548322628376006,
+      "loss": 2.1441,
+      "step": 326805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000135481498254594,
+      "loss": 2.1977,
+      "step": 326810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013547977021330688,
+      "loss": 2.0181,
+      "step": 326815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013547804215989915,
+      "loss": 2.0849,
+      "step": 326820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013547631409437154,
+      "loss": 1.8653,
+      "step": 326825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013547458601672452,
+      "loss": 1.895,
+      "step": 326830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001354728579269588,
+      "loss": 2.0402,
+      "step": 326835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001354711298250749,
+      "loss": 2.0518,
+      "step": 326840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00013546940171107338,
+      "loss": 2.093,
+      "step": 326845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013546767358495489,
+      "loss": 1.9902,
+      "step": 326850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013546594544672,
+      "loss": 2.0297,
+      "step": 326855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013546421729636928,
+      "loss": 1.9027,
+      "step": 326860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013546248913390337,
+      "loss": 1.9985,
+      "step": 326865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001354607609593228,
+      "loss": 2.1296,
+      "step": 326870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013545903277262827,
+      "loss": 1.971,
+      "step": 326875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001354573045738202,
+      "loss": 2.1734,
+      "step": 326880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013545557636289928,
+      "loss": 2.0255,
+      "step": 326885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00013545384813986612,
+      "loss": 2.1632,
+      "step": 326890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013545211990472131,
+      "loss": 2.2336,
+      "step": 326895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013545039165746537,
+      "loss": 1.9237,
+      "step": 326900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013544866339809894,
+      "loss": 2.0115,
+      "step": 326905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001354469351266226,
+      "loss": 2.007,
+      "step": 326910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013544520684303694,
+      "loss": 2.1553,
+      "step": 326915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013544347854734258,
+      "loss": 2.0573,
+      "step": 326920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013544175023954007,
+      "loss": 1.9618,
+      "step": 326925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00013544002191963003,
+      "loss": 1.9997,
+      "step": 326930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000135438293587613,
+      "loss": 1.7491,
+      "step": 326935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013543656524348964,
+      "loss": 2.1919,
+      "step": 326940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013543483688726048,
+      "loss": 2.108,
+      "step": 326945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013543310851892617,
+      "loss": 2.2433,
+      "step": 326950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00013543138013848724,
+      "loss": 2.2084,
+      "step": 326955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001354296517459443,
+      "loss": 2.0217,
+      "step": 326960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013542792334129798,
+      "loss": 1.8986,
+      "step": 326965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00013542619492454882,
+      "loss": 2.2219,
+      "step": 326970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00013542446649569744,
+      "loss": 2.0713,
+      "step": 326975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013542273805474441,
+      "loss": 2.0172,
+      "step": 326980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013542100960169034,
+      "loss": 1.9612,
+      "step": 326985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001354192811365358,
+      "loss": 1.8921,
+      "step": 326990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001354175526592814,
+      "loss": 2.1884,
+      "step": 326995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001354158241699277,
+      "loss": 2.1095,
+      "step": 327000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013541409566847533,
+      "loss": 2.0897,
+      "step": 327005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013541236715492488,
+      "loss": 1.8019,
+      "step": 327010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001354106386292769,
+      "loss": 2.0191,
+      "step": 327015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000135408910091532,
+      "loss": 2.005,
+      "step": 327020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001354071815416908,
+      "loss": 2.0119,
+      "step": 327025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013540545297975384,
+      "loss": 2.0457,
+      "step": 327030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00013540372440572175,
+      "loss": 2.0518,
+      "step": 327035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001354019958195951,
+      "loss": 2.0504,
+      "step": 327040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00013540026722137446,
+      "loss": 2.1219,
+      "step": 327045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001353985386110605,
+      "loss": 1.9177,
+      "step": 327050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00013539680998865373,
+      "loss": 2.1058,
+      "step": 327055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013539508135415476,
+      "loss": 1.8445,
+      "step": 327060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0,
+      "learning_rate": 0.0001353933527075642,
+      "loss": 2.0609,
+      "step": 327065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00013539162404888262,
+      "loss": 2.0236,
+      "step": 327070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001353898953781106,
+      "loss": 2.113,
+      "step": 327075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001353881666952488,
+      "loss": 2.0175,
+      "step": 327080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00013538643800029774,
+      "loss": 1.9991,
+      "step": 327085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013538470929325803,
+      "loss": 1.9605,
+      "step": 327090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013538298057413022,
+      "loss": 2.032,
+      "step": 327095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.000135381251842915,
+      "loss": 2.0907,
+      "step": 327100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001353795230996129,
+      "loss": 2.1103,
+      "step": 327105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013537779434422446,
+      "loss": 2.0408,
+      "step": 327110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00013537606557675037,
+      "loss": 1.9827,
+      "step": 327115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013537433679719116,
+      "loss": 1.9975,
+      "step": 327120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00013537260800554744,
+      "loss": 2.0377,
+      "step": 327125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001353708792018198,
+      "loss": 2.1673,
+      "step": 327130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013536915038600878,
+      "loss": 2.1222,
+      "step": 327135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013536742155811508,
+      "loss": 2.0292,
+      "step": 327140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013536569271813923,
+      "loss": 2.088,
+      "step": 327145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013536396386608176,
+      "loss": 2.1617,
+      "step": 327150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013536223500194337,
+      "loss": 1.9908,
+      "step": 327155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013536050612572456,
+      "loss": 1.9125,
+      "step": 327160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013535877723742598,
+      "loss": 2.0293,
+      "step": 327165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00013535704833704824,
+      "loss": 2.2507,
+      "step": 327170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013535531942459182,
+      "loss": 2.1311,
+      "step": 327175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001353535905000574,
+      "loss": 2.0037,
+      "step": 327180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00013535186156344556,
+      "loss": 1.7844,
+      "step": 327185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013535013261475691,
+      "loss": 2.0179,
+      "step": 327190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000135348403653992,
+      "loss": 2.1549,
+      "step": 327195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013534667468115142,
+      "loss": 2.0306,
+      "step": 327200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013534494569623576,
+      "loss": 2.0568,
+      "step": 327205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013534321669924563,
+      "loss": 2.0708,
+      "step": 327210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00013534148769018165,
+      "loss": 2.1081,
+      "step": 327215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.125,
+      "learning_rate": 0.00013533975866904436,
+      "loss": 2.2165,
+      "step": 327220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013533802963583438,
+      "loss": 2.0132,
+      "step": 327225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013533630059055224,
+      "loss": 1.8638,
+      "step": 327230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013533457153319865,
+      "loss": 1.9573,
+      "step": 327235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001353328424637741,
+      "loss": 1.8982,
+      "step": 327240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00013533111338227916,
+      "loss": 1.8444,
+      "step": 327245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013532938428871454,
+      "loss": 2.0668,
+      "step": 327250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00013532765518308072,
+      "loss": 1.9384,
+      "step": 327255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013532592606537835,
+      "loss": 2.0799,
+      "step": 327260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.000135324196935608,
+      "loss": 1.9764,
+      "step": 327265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013532246779377028,
+      "loss": 2.0286,
+      "step": 327270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013532073863986574,
+      "loss": 2.1748,
+      "step": 327275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00013531900947389499,
+      "loss": 2.2375,
+      "step": 327280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013531728029585866,
+      "loss": 2.0472,
+      "step": 327285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013531555110575728,
+      "loss": 2.1504,
+      "step": 327290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001353138219035915,
+      "loss": 2.0508,
+      "step": 327295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013531209268936184,
+      "loss": 2.0644,
+      "step": 327300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00013531036346306894,
+      "loss": 2.1351,
+      "step": 327305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013530863422471336,
+      "loss": 2.0763,
+      "step": 327310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013530690497429572,
+      "loss": 2.1816,
+      "step": 327315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00013530517571181663,
+      "loss": 1.9257,
+      "step": 327320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001353034464372766,
+      "loss": 1.7969,
+      "step": 327325
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013530171715067635,
+      "loss": 2.242,
+      "step": 327330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013529998785201635,
+      "loss": 2.0714,
+      "step": 327335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001352982585412972,
+      "loss": 1.9164,
+      "step": 327340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00013529652921851957,
+      "loss": 2.0107,
+      "step": 327345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000135294799883684,
+      "loss": 1.9781,
+      "step": 327350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00013529307053679107,
+      "loss": 1.9621,
+      "step": 327355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001352913411778414,
+      "loss": 2.0518,
+      "step": 327360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013528961180683555,
+      "loss": 2.1946,
+      "step": 327365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00013528788242377413,
+      "loss": 2.3454,
+      "step": 327370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013528615302865776,
+      "loss": 2.0168,
+      "step": 327375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013528442362148698,
+      "loss": 2.0516,
+      "step": 327380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001352826942022624,
+      "loss": 2.1455,
+      "step": 327385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013528096477098458,
+      "loss": 2.0668,
+      "step": 327390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00013527923532765423,
+      "loss": 2.2093,
+      "step": 327395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001352775058722718,
+      "loss": 2.1564,
+      "step": 327400
     }
   ],
   "logging_steps": 5,
@@ -312494,7 +458374,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 200,
-  "total_flos": 1.3479703781281943e+20,
+  "total_flos": 1.976116670110095e+20,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null