diff --git "a/checkpoint-23845/trainer_state.json" "b/checkpoint-23845/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-23845/trainer_state.json"
@@ -0,0 +1,16721 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 23845,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0020968756552736424,
+      "grad_norm": 16.383289337158203,
+      "learning_rate": 0.0002,
+      "loss": 5.0295,
+      "step": 10
+    },
+    {
+      "epoch": 0.004193751310547285,
+      "grad_norm": 15.346772193908691,
+      "learning_rate": 0.0002,
+      "loss": 1.344,
+      "step": 20
+    },
+    {
+      "epoch": 0.006290626965820926,
+      "grad_norm": 5.9838738441467285,
+      "learning_rate": 0.0002,
+      "loss": 0.7367,
+      "step": 30
+    },
+    {
+      "epoch": 0.00838750262109457,
+      "grad_norm": 1.9119038581848145,
+      "learning_rate": 0.0002,
+      "loss": 0.4024,
+      "step": 40
+    },
+    {
+      "epoch": 0.010484378276368212,
+      "grad_norm": 1.0427629947662354,
+      "learning_rate": 0.0002,
+      "loss": 0.393,
+      "step": 50
+    },
+    {
+      "epoch": 0.012581253931641853,
+      "grad_norm": 3.237114429473877,
+      "learning_rate": 0.0002,
+      "loss": 0.2799,
+      "step": 60
+    },
+    {
+      "epoch": 0.014678129586915495,
+      "grad_norm": 2.2672061920166016,
+      "learning_rate": 0.0002,
+      "loss": 0.3716,
+      "step": 70
+    },
+    {
+      "epoch": 0.01677500524218914,
+      "grad_norm": 0.4691790044307709,
+      "learning_rate": 0.0002,
+      "loss": 0.3543,
+      "step": 80
+    },
+    {
+      "epoch": 0.018871880897462782,
+      "grad_norm": 4.190754413604736,
+      "learning_rate": 0.0002,
+      "loss": 0.2702,
+      "step": 90
+    },
+    {
+      "epoch": 0.020968756552736424,
+      "grad_norm": 1.0056904554367065,
+      "learning_rate": 0.0002,
+      "loss": 0.347,
+      "step": 100
+    },
+    {
+      "epoch": 0.023065632208010067,
+      "grad_norm": 1.437057375907898,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 110
+    },
+    {
+      "epoch": 0.025162507863283706,
+      "grad_norm": 0.36117708683013916,
+      "learning_rate": 0.0002,
+      "loss": 0.3003,
+      "step": 120
+    },
+    {
+      "epoch": 0.027259383518557348,
+      "grad_norm": 8.518256187438965,
+      "learning_rate": 0.0002,
+      "loss": 0.3501,
+      "step": 130
+    },
+    {
+      "epoch": 0.02935625917383099,
+      "grad_norm": 0.6031225323677063,
+      "learning_rate": 0.0002,
+      "loss": 0.2877,
+      "step": 140
+    },
+    {
+      "epoch": 0.03145313482910463,
+      "grad_norm": 0.909845232963562,
+      "learning_rate": 0.0002,
+      "loss": 0.2695,
+      "step": 150
+    },
+    {
+      "epoch": 0.03355001048437828,
+      "grad_norm": 0.46850165724754333,
+      "learning_rate": 0.0002,
+      "loss": 0.2755,
+      "step": 160
+    },
+    {
+      "epoch": 0.03564688613965192,
+      "grad_norm": 1.2919809818267822,
+      "learning_rate": 0.0002,
+      "loss": 0.381,
+      "step": 170
+    },
+    {
+      "epoch": 0.037743761794925564,
+      "grad_norm": 0.36922141909599304,
+      "learning_rate": 0.0002,
+      "loss": 0.3034,
+      "step": 180
+    },
+    {
+      "epoch": 0.0398406374501992,
+      "grad_norm": 0.41144391894340515,
+      "learning_rate": 0.0002,
+      "loss": 0.3339,
+      "step": 190
+    },
+    {
+      "epoch": 0.04193751310547285,
+      "grad_norm": 0.5304984450340271,
+      "learning_rate": 0.0002,
+      "loss": 0.2819,
+      "step": 200
+    },
+    {
+      "epoch": 0.04403438876074649,
+      "grad_norm": 0.36329638957977295,
+      "learning_rate": 0.0002,
+      "loss": 0.2943,
+      "step": 210
+    },
+    {
+      "epoch": 0.04613126441602013,
+      "grad_norm": 5.069764614105225,
+      "learning_rate": 0.0002,
+      "loss": 0.326,
+      "step": 220
+    },
+    {
+      "epoch": 0.04822814007129377,
+      "grad_norm": 5.163594722747803,
+      "learning_rate": 0.0002,
+      "loss": 0.2818,
+      "step": 230
+    },
+    {
+      "epoch": 0.05032501572656741,
+      "grad_norm": 2.1980459690093994,
+      "learning_rate": 0.0002,
+      "loss": 0.2522,
+      "step": 240
+    },
+    {
+      "epoch": 0.05242189138184106,
+      "grad_norm": 2.0721938610076904,
+      "learning_rate": 0.0002,
+      "loss": 0.2838,
+      "step": 250
+    },
+    {
+      "epoch": 0.054518767037114696,
+      "grad_norm": 0.21942710876464844,
+      "learning_rate": 0.0002,
+      "loss": 0.2509,
+      "step": 260
+    },
+    {
+      "epoch": 0.05661564269238834,
+      "grad_norm": 1.9733415842056274,
+      "learning_rate": 0.0002,
+      "loss": 0.2499,
+      "step": 270
+    },
+    {
+      "epoch": 0.05871251834766198,
+      "grad_norm": 0.1733933985233307,
+      "learning_rate": 0.0002,
+      "loss": 0.2478,
+      "step": 280
+    },
+    {
+      "epoch": 0.06080939400293563,
+      "grad_norm": 5.54453706741333,
+      "learning_rate": 0.0002,
+      "loss": 0.2746,
+      "step": 290
+    },
+    {
+      "epoch": 0.06290626965820927,
+      "grad_norm": 0.30464133620262146,
+      "learning_rate": 0.0002,
+      "loss": 0.3196,
+      "step": 300
+    },
+    {
+      "epoch": 0.0650031453134829,
+      "grad_norm": 5.397165298461914,
+      "learning_rate": 0.0002,
+      "loss": 0.3544,
+      "step": 310
+    },
+    {
+      "epoch": 0.06710002096875656,
+      "grad_norm": 0.21483393013477325,
+      "learning_rate": 0.0002,
+      "loss": 0.2509,
+      "step": 320
+    },
+    {
+      "epoch": 0.0691968966240302,
+      "grad_norm": 0.24638743698596954,
+      "learning_rate": 0.0002,
+      "loss": 0.2819,
+      "step": 330
+    },
+    {
+      "epoch": 0.07129377227930384,
+      "grad_norm": 9.33228874206543,
+      "learning_rate": 0.0002,
+      "loss": 0.3348,
+      "step": 340
+    },
+    {
+      "epoch": 0.07339064793457747,
+      "grad_norm": 2.1489148139953613,
+      "learning_rate": 0.0002,
+      "loss": 0.2708,
+      "step": 350
+    },
+    {
+      "epoch": 0.07548752358985113,
+      "grad_norm": 0.23243466019630432,
+      "learning_rate": 0.0002,
+      "loss": 0.249,
+      "step": 360
+    },
+    {
+      "epoch": 0.07758439924512477,
+      "grad_norm": 0.13301965594291687,
+      "learning_rate": 0.0002,
+      "loss": 0.2739,
+      "step": 370
+    },
+    {
+      "epoch": 0.0796812749003984,
+      "grad_norm": 0.2184901237487793,
+      "learning_rate": 0.0002,
+      "loss": 0.2718,
+      "step": 380
+    },
+    {
+      "epoch": 0.08177815055567204,
+      "grad_norm": 5.580859661102295,
+      "learning_rate": 0.0002,
+      "loss": 0.2882,
+      "step": 390
+    },
+    {
+      "epoch": 0.0838750262109457,
+      "grad_norm": 0.2310085892677307,
+      "learning_rate": 0.0002,
+      "loss": 0.2837,
+      "step": 400
+    },
+    {
+      "epoch": 0.08597190186621934,
+      "grad_norm": 0.16177065670490265,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 410
+    },
+    {
+      "epoch": 0.08806877752149297,
+      "grad_norm": 3.8961215019226074,
+      "learning_rate": 0.0002,
+      "loss": 0.3048,
+      "step": 420
+    },
+    {
+      "epoch": 0.09016565317676661,
+      "grad_norm": 0.09581358730792999,
+      "learning_rate": 0.0002,
+      "loss": 0.3008,
+      "step": 430
+    },
+    {
+      "epoch": 0.09226252883204027,
+      "grad_norm": 0.12218118458986282,
+      "learning_rate": 0.0002,
+      "loss": 0.2672,
+      "step": 440
+    },
+    {
+      "epoch": 0.0943594044873139,
+      "grad_norm": 0.11212728172540665,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 450
+    },
+    {
+      "epoch": 0.09645628014258754,
+      "grad_norm": 0.16998110711574554,
+      "learning_rate": 0.0002,
+      "loss": 0.268,
+      "step": 460
+    },
+    {
+      "epoch": 0.09855315579786118,
+      "grad_norm": 1.1045947074890137,
+      "learning_rate": 0.0002,
+      "loss": 0.249,
+      "step": 470
+    },
+    {
+      "epoch": 0.10065003145313482,
+      "grad_norm": 0.0940975770354271,
+      "learning_rate": 0.0002,
+      "loss": 0.2994,
+      "step": 480
+    },
+    {
+      "epoch": 0.10274690710840848,
+      "grad_norm": 0.1036498099565506,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 490
+    },
+    {
+      "epoch": 0.10484378276368211,
+      "grad_norm": 0.42365631461143494,
+      "learning_rate": 0.0002,
+      "loss": 0.2984,
+      "step": 500
+    },
+    {
+      "epoch": 0.10694065841895575,
+      "grad_norm": 0.08997116982936859,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 510
+    },
+    {
+      "epoch": 0.10903753407422939,
+      "grad_norm": 0.15138301253318787,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 520
+    },
+    {
+      "epoch": 0.11113440972950304,
+      "grad_norm": 2.1935176849365234,
+      "learning_rate": 0.0002,
+      "loss": 0.334,
+      "step": 530
+    },
+    {
+      "epoch": 0.11323128538477668,
+      "grad_norm": 0.15546467900276184,
+      "learning_rate": 0.0002,
+      "loss": 0.2875,
+      "step": 540
+    },
+    {
+      "epoch": 0.11532816104005032,
+      "grad_norm": 0.15564927458763123,
+      "learning_rate": 0.0002,
+      "loss": 0.2735,
+      "step": 550
+    },
+    {
+      "epoch": 0.11742503669532396,
+      "grad_norm": 0.1129130944609642,
+      "learning_rate": 0.0002,
+      "loss": 0.3484,
+      "step": 560
+    },
+    {
+      "epoch": 0.11952191235059761,
+      "grad_norm": 0.2688151001930237,
+      "learning_rate": 0.0002,
+      "loss": 0.28,
+      "step": 570
+    },
+    {
+      "epoch": 0.12161878800587125,
+      "grad_norm": 0.13618075847625732,
+      "learning_rate": 0.0002,
+      "loss": 0.3393,
+      "step": 580
+    },
+    {
+      "epoch": 0.12371566366114489,
+      "grad_norm": 0.1488252729177475,
+      "learning_rate": 0.0002,
+      "loss": 0.3001,
+      "step": 590
+    },
+    {
+      "epoch": 0.12581253931641853,
+      "grad_norm": 0.10563705116510391,
+      "learning_rate": 0.0002,
+      "loss": 0.3055,
+      "step": 600
+    },
+    {
+      "epoch": 0.12790941497169217,
+      "grad_norm": 0.07457701861858368,
+      "learning_rate": 0.0002,
+      "loss": 0.3068,
+      "step": 610
+    },
+    {
+      "epoch": 0.1300062906269658,
+      "grad_norm": 0.5072939991950989,
+      "learning_rate": 0.0002,
+      "loss": 0.2483,
+      "step": 620
+    },
+    {
+      "epoch": 0.13210316628223948,
+      "grad_norm": 0.20537391304969788,
+      "learning_rate": 0.0002,
+      "loss": 0.307,
+      "step": 630
+    },
+    {
+      "epoch": 0.13420004193751311,
+      "grad_norm": 0.144338920712471,
+      "learning_rate": 0.0002,
+      "loss": 0.275,
+      "step": 640
+    },
+    {
+      "epoch": 0.13629691759278675,
+      "grad_norm": 0.1316433846950531,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 650
+    },
+    {
+      "epoch": 0.1383937932480604,
+      "grad_norm": 3.318430185317993,
+      "learning_rate": 0.0002,
+      "loss": 0.3049,
+      "step": 660
+    },
+    {
+      "epoch": 0.14049066890333403,
+      "grad_norm": 0.08888725936412811,
+      "learning_rate": 0.0002,
+      "loss": 0.2817,
+      "step": 670
+    },
+    {
+      "epoch": 0.14258754455860767,
+      "grad_norm": 0.13618157804012299,
+      "learning_rate": 0.0002,
+      "loss": 0.2855,
+      "step": 680
+    },
+    {
+      "epoch": 0.1446844202138813,
+      "grad_norm": 0.1266484558582306,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 690
+    },
+    {
+      "epoch": 0.14678129586915495,
+      "grad_norm": 0.11817031353712082,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 700
+    },
+    {
+      "epoch": 0.14887817152442862,
+      "grad_norm": 0.12560369074344635,
+      "learning_rate": 0.0002,
+      "loss": 0.2685,
+      "step": 710
+    },
+    {
+      "epoch": 0.15097504717970225,
+      "grad_norm": 0.1666841357946396,
+      "learning_rate": 0.0002,
+      "loss": 0.2776,
+      "step": 720
+    },
+    {
+      "epoch": 0.1530719228349759,
+      "grad_norm": 0.16758698225021362,
+      "learning_rate": 0.0002,
+      "loss": 0.2887,
+      "step": 730
+    },
+    {
+      "epoch": 0.15516879849024953,
+      "grad_norm": 0.821657121181488,
+      "learning_rate": 0.0002,
+      "loss": 0.3388,
+      "step": 740
+    },
+    {
+      "epoch": 0.15726567414552317,
+      "grad_norm": 0.11897563934326172,
+      "learning_rate": 0.0002,
+      "loss": 0.2833,
+      "step": 750
+    },
+    {
+      "epoch": 0.1593625498007968,
+      "grad_norm": 2.6172780990600586,
+      "learning_rate": 0.0002,
+      "loss": 0.2832,
+      "step": 760
+    },
+    {
+      "epoch": 0.16145942545607045,
+      "grad_norm": 2.524484157562256,
+      "learning_rate": 0.0002,
+      "loss": 0.3135,
+      "step": 770
+    },
+    {
+      "epoch": 0.1635563011113441,
+      "grad_norm": 1.6356699466705322,
+      "learning_rate": 0.0002,
+      "loss": 0.267,
+      "step": 780
+    },
+    {
+      "epoch": 0.16565317676661773,
+      "grad_norm": 1.7804538011550903,
+      "learning_rate": 0.0002,
+      "loss": 0.3106,
+      "step": 790
+    },
+    {
+      "epoch": 0.1677500524218914,
+      "grad_norm": 0.07052750140428543,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 800
+    },
+    {
+      "epoch": 0.16984692807716503,
+      "grad_norm": 0.11844762414693832,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 810
+    },
+    {
+      "epoch": 0.17194380373243867,
+      "grad_norm": 0.0921357050538063,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 820
+    },
+    {
+      "epoch": 0.1740406793877123,
+      "grad_norm": 0.1175508201122284,
+      "learning_rate": 0.0002,
+      "loss": 0.2725,
+      "step": 830
+    },
+    {
+      "epoch": 0.17613755504298595,
+      "grad_norm": 0.11097151786088943,
+      "learning_rate": 0.0002,
+      "loss": 0.2694,
+      "step": 840
+    },
+    {
+      "epoch": 0.1782344306982596,
+      "grad_norm": 2.3500242233276367,
+      "learning_rate": 0.0002,
+      "loss": 0.3577,
+      "step": 850
+    },
+    {
+      "epoch": 0.18033130635353323,
+      "grad_norm": 0.46957358717918396,
+      "learning_rate": 0.0002,
+      "loss": 0.3253,
+      "step": 860
+    },
+    {
+      "epoch": 0.18242818200880687,
+      "grad_norm": 0.24576787650585175,
+      "learning_rate": 0.0002,
+      "loss": 0.2529,
+      "step": 870
+    },
+    {
+      "epoch": 0.18452505766408053,
+      "grad_norm": 0.21132692694664001,
+      "learning_rate": 0.0002,
+      "loss": 0.2894,
+      "step": 880
+    },
+    {
+      "epoch": 0.18662193331935417,
+      "grad_norm": 0.16808678209781647,
+      "learning_rate": 0.0002,
+      "loss": 0.2559,
+      "step": 890
+    },
+    {
+      "epoch": 0.1887188089746278,
+      "grad_norm": 0.1904708743095398,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 900
+    },
+    {
+      "epoch": 0.19081568462990145,
+      "grad_norm": 0.0702681764960289,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 910
+    },
+    {
+      "epoch": 0.1929125602851751,
+      "grad_norm": 0.16173458099365234,
+      "learning_rate": 0.0002,
+      "loss": 0.2731,
+      "step": 920
+    },
+    {
+      "epoch": 0.19500943594044873,
+      "grad_norm": 0.1556364893913269,
+      "learning_rate": 0.0002,
+      "loss": 0.2642,
+      "step": 930
+    },
+    {
+      "epoch": 0.19710631159572237,
+      "grad_norm": 0.09365347027778625,
+      "learning_rate": 0.0002,
+      "loss": 0.2828,
+      "step": 940
+    },
+    {
+      "epoch": 0.199203187250996,
+      "grad_norm": 0.12134792655706406,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 950
+    },
+    {
+      "epoch": 0.20130006290626964,
+      "grad_norm": 0.14340221881866455,
+      "learning_rate": 0.0002,
+      "loss": 0.3118,
+      "step": 960
+    },
+    {
+      "epoch": 0.2033969385615433,
+      "grad_norm": 0.15696048736572266,
+      "learning_rate": 0.0002,
+      "loss": 0.2679,
+      "step": 970
+    },
+    {
+      "epoch": 0.20549381421681695,
+      "grad_norm": 0.33497804403305054,
+      "learning_rate": 0.0002,
+      "loss": 0.3172,
+      "step": 980
+    },
+    {
+      "epoch": 0.2075906898720906,
+      "grad_norm": 0.15141892433166504,
+      "learning_rate": 0.0002,
+      "loss": 0.2592,
+      "step": 990
+    },
+    {
+      "epoch": 0.20968756552736423,
+      "grad_norm": 0.10774557292461395,
+      "learning_rate": 0.0002,
+      "loss": 0.3013,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21178444118263787,
+      "grad_norm": 0.09470309317111969,
+      "learning_rate": 0.0002,
+      "loss": 0.2705,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2138813168379115,
+      "grad_norm": 0.12090291827917099,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21597819249318514,
+      "grad_norm": 0.12572237849235535,
+      "learning_rate": 0.0002,
+      "loss": 0.2671,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21807506814845878,
+      "grad_norm": 0.19516214728355408,
+      "learning_rate": 0.0002,
+      "loss": 0.2878,
+      "step": 1040
+    },
+    {
+      "epoch": 0.22017194380373245,
+      "grad_norm": 0.08513014763593674,
+      "learning_rate": 0.0002,
+      "loss": 0.3277,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2222688194590061,
+      "grad_norm": 0.14394588768482208,
+      "learning_rate": 0.0002,
+      "loss": 0.3265,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22436569511427973,
+      "grad_norm": 3.9625415802001953,
+      "learning_rate": 0.0002,
+      "loss": 0.2838,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22646257076955337,
+      "grad_norm": 1.3177988529205322,
+      "learning_rate": 0.0002,
+      "loss": 0.2482,
+      "step": 1080
+    },
+    {
+      "epoch": 0.228559446424827,
+      "grad_norm": 0.1534273624420166,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 1090
+    },
+    {
+      "epoch": 0.23065632208010065,
+      "grad_norm": 0.18359704315662384,
+      "learning_rate": 0.0002,
+      "loss": 0.2471,
+      "step": 1100
+    },
+    {
+      "epoch": 0.23275319773537428,
+      "grad_norm": 0.1258287876844406,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23485007339064792,
+      "grad_norm": 0.19109375774860382,
+      "learning_rate": 0.0002,
+      "loss": 0.3102,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2369469490459216,
+      "grad_norm": 2.0408198833465576,
+      "learning_rate": 0.0002,
+      "loss": 0.2702,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23904382470119523,
+      "grad_norm": 0.06613194197416306,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 1140
+    },
+    {
+      "epoch": 0.24114070035646887,
+      "grad_norm": 0.11054085940122604,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2432375760117425,
+      "grad_norm": 1.8500365018844604,
+      "learning_rate": 0.0002,
+      "loss": 0.2886,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24533445166701615,
+      "grad_norm": 0.12011546641588211,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24743132732228978,
+      "grad_norm": 0.18363837897777557,
+      "learning_rate": 0.0002,
+      "loss": 0.3159,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24952820297756342,
+      "grad_norm": 0.07881256937980652,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 1190
+    },
+    {
+      "epoch": 0.25162507863283706,
+      "grad_norm": 0.08021403849124908,
+      "learning_rate": 0.0002,
+      "loss": 0.2977,
+      "step": 1200
+    },
+    {
+      "epoch": 0.25372195428811073,
+      "grad_norm": 1.153460144996643,
+      "learning_rate": 0.0002,
+      "loss": 0.2741,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25581882994338434,
+      "grad_norm": 0.8089906573295593,
+      "learning_rate": 0.0002,
+      "loss": 0.3158,
+      "step": 1220
+    },
+    {
+      "epoch": 0.257915705598658,
+      "grad_norm": 0.533995509147644,
+      "learning_rate": 0.0002,
+      "loss": 0.2876,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2600125812539316,
+      "grad_norm": 2.0046756267547607,
+      "learning_rate": 0.0002,
+      "loss": 0.3035,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2621094569092053,
+      "grad_norm": 0.13666591048240662,
+      "learning_rate": 0.0002,
+      "loss": 0.2692,
+      "step": 1250
+    },
+    {
+      "epoch": 0.26420633256447895,
+      "grad_norm": 3.1030569076538086,
+      "learning_rate": 0.0002,
+      "loss": 0.2813,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26630320821975256,
+      "grad_norm": 0.08713892847299576,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 1270
+    },
+    {
+      "epoch": 0.26840008387502623,
+      "grad_norm": 1.8841831684112549,
+      "learning_rate": 0.0002,
+      "loss": 0.3362,
+      "step": 1280
+    },
+    {
+      "epoch": 0.27049695953029984,
+      "grad_norm": 0.3094707429409027,
+      "learning_rate": 0.0002,
+      "loss": 0.293,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2725938351855735,
+      "grad_norm": 0.10224305093288422,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2746907108408471,
+      "grad_norm": 0.14384222030639648,
+      "learning_rate": 0.0002,
+      "loss": 0.3165,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2767875864961208,
+      "grad_norm": 2.114215850830078,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2788844621513944,
+      "grad_norm": 0.0842517539858818,
+      "learning_rate": 0.0002,
+      "loss": 0.2698,
+      "step": 1330
+    },
+    {
+      "epoch": 0.28098133780666806,
+      "grad_norm": 0.06867516785860062,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 1340
+    },
+    {
+      "epoch": 0.28307821346194173,
+      "grad_norm": 0.0686967521905899,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 1350
+    },
+    {
+      "epoch": 0.28517508911721534,
+      "grad_norm": 1.7461965084075928,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 1360
+    },
+    {
+      "epoch": 0.287271964772489,
+      "grad_norm": 0.09826788306236267,
+      "learning_rate": 0.0002,
+      "loss": 0.2804,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2893688404277626,
+      "grad_norm": 0.09541603922843933,
+      "learning_rate": 0.0002,
+      "loss": 0.2871,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2914657160830363,
+      "grad_norm": 0.12282253801822662,
+      "learning_rate": 0.0002,
+      "loss": 0.3452,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2935625917383099,
+      "grad_norm": 0.09135173261165619,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 1400
+    },
+    {
+      "epoch": 0.29565946739358356,
+      "grad_norm": 1.8899190425872803,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29775634304885723,
+      "grad_norm": 0.13118763267993927,
+      "learning_rate": 0.0002,
+      "loss": 0.3037,
+      "step": 1420
+    },
+    {
+      "epoch": 0.29985321870413084,
+      "grad_norm": 1.688429355621338,
+      "learning_rate": 0.0002,
+      "loss": 0.299,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3019500943594045,
+      "grad_norm": 0.11612743139266968,
+      "learning_rate": 0.0002,
+      "loss": 0.2699,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3040469700146781,
+      "grad_norm": 0.14019401371479034,
+      "learning_rate": 0.0002,
+      "loss": 0.2503,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3061438456699518,
+      "grad_norm": 0.8129562735557556,
+      "learning_rate": 0.0002,
+      "loss": 0.2853,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3082407213252254,
+      "grad_norm": 5.135369777679443,
+      "learning_rate": 0.0002,
+      "loss": 0.2949,
+      "step": 1470
+    },
+    {
+      "epoch": 0.31033759698049906,
+      "grad_norm": 0.11974827200174332,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3124344726357727,
+      "grad_norm": 0.10319698601961136,
+      "learning_rate": 0.0002,
+      "loss": 0.2879,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31453134829104634,
+      "grad_norm": 0.09822767227888107,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31662822394632,
+      "grad_norm": 0.09973836690187454,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3187250996015936,
+      "grad_norm": 16.39473533630371,
+      "learning_rate": 0.0002,
+      "loss": 0.2979,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3208219752568673,
+      "grad_norm": 0.07360737025737762,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3229188509121409,
+      "grad_norm": 0.11402890831232071,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 1540
+    },
+    {
+      "epoch": 0.32501572656741456,
+      "grad_norm": 0.09771336615085602,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3271126022226882,
+      "grad_norm": 1.3607581853866577,
+      "learning_rate": 0.0002,
+      "loss": 0.3091,
+      "step": 1560
+    },
+    {
+      "epoch": 0.32920947787796184,
+      "grad_norm": 0.09807518124580383,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 1570
+    },
+    {
+      "epoch": 0.33130635353323545,
+      "grad_norm": 0.098614901304245,
+      "learning_rate": 0.0002,
+      "loss": 0.344,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3334032291885091,
+      "grad_norm": 0.08970670402050018,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3355001048437828,
+      "grad_norm": 0.11069203168153763,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3375969804990564,
+      "grad_norm": 0.1457228660583496,
+      "learning_rate": 0.0002,
+      "loss": 0.3231,
+      "step": 1610
+    },
+    {
+      "epoch": 0.33969385615433006,
+      "grad_norm": 0.12434522062540054,
+      "learning_rate": 0.0002,
+      "loss": 0.2892,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3417907318096037,
+      "grad_norm": 0.14981336891651154,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 1630
+    },
+    {
+      "epoch": 0.34388760746487734,
+      "grad_norm": 1.7012590169906616,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 1640
+    },
+    {
+      "epoch": 0.34598448312015095,
+      "grad_norm": 0.10405845195055008,
+      "learning_rate": 0.0002,
+      "loss": 0.3362,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3480813587754246,
+      "grad_norm": 0.19532528519630432,
+      "learning_rate": 0.0002,
+      "loss": 0.2668,
+      "step": 1660
+    },
+    {
+      "epoch": 0.35017823443069823,
+      "grad_norm": 0.08479733765125275,
+      "learning_rate": 0.0002,
+      "loss": 0.2911,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3522751100859719,
+      "grad_norm": 1.8583425283432007,
+      "learning_rate": 0.0002,
+      "loss": 0.2697,
+      "step": 1680
+    },
+    {
+      "epoch": 0.35437198574124557,
+      "grad_norm": 0.08247324079275131,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3564688613965192,
+      "grad_norm": 2.080596923828125,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 1700
+    },
+    {
+      "epoch": 0.35856573705179284,
+      "grad_norm": 0.07867144048213959,
+      "learning_rate": 0.0002,
+      "loss": 0.2624,
+      "step": 1710
+    },
+    {
+      "epoch": 0.36066261270706645,
+      "grad_norm": 0.15009847283363342,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3627594883623401,
+      "grad_norm": 0.0880853533744812,
+      "learning_rate": 0.0002,
+      "loss": 0.3024,
+      "step": 1730
+    },
+    {
+      "epoch": 0.36485636401761373,
+      "grad_norm": 0.1102822870016098,
+      "learning_rate": 0.0002,
+      "loss": 0.2654,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3669532396728874,
+      "grad_norm": 0.20041993260383606,
+      "learning_rate": 0.0002,
+      "loss": 0.3052,
+      "step": 1750
+    },
+    {
+      "epoch": 0.36905011532816107,
+      "grad_norm": 0.07399024814367294,
+      "learning_rate": 0.0002,
+      "loss": 0.3312,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3711469909834347,
+      "grad_norm": 0.2129296362400055,
+      "learning_rate": 0.0002,
+      "loss": 0.2788,
+      "step": 1770
+    },
+    {
+      "epoch": 0.37324386663870834,
+      "grad_norm": 0.09153830260038376,
+      "learning_rate": 0.0002,
+      "loss": 0.2726,
+      "step": 1780
+    },
+    {
+      "epoch": 0.37534074229398195,
+      "grad_norm": 0.09638633579015732,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3774376179492556,
+      "grad_norm": 0.12968820333480835,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 1800
+    },
+    {
+      "epoch": 0.37953449360452923,
+      "grad_norm": 0.10243703424930573,
+      "learning_rate": 0.0002,
+      "loss": 0.287,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3816313692598029,
+      "grad_norm": 0.09311744570732117,
+      "learning_rate": 0.0002,
+      "loss": 0.2971,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3837282449150765,
+      "grad_norm": 0.09674180299043655,
+      "learning_rate": 0.0002,
+      "loss": 0.2549,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3858251205703502,
+      "grad_norm": 0.13742677867412567,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 1840
+    },
+    {
+      "epoch": 0.38792199622562384,
+      "grad_norm": 1.1147104501724243,
+      "learning_rate": 0.0002,
+      "loss": 0.3029,
+      "step": 1850
+    },
+    {
+      "epoch": 0.39001887188089746,
+      "grad_norm": 0.06606106460094452,
+      "learning_rate": 0.0002,
+      "loss": 0.2708,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3921157475361711,
+      "grad_norm": 0.06749361008405685,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 1870
+    },
+    {
+      "epoch": 0.39421262319144473,
+      "grad_norm": 0.07357250154018402,
+      "learning_rate": 0.0002,
+      "loss": 0.2621,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3963094988467184,
+      "grad_norm": 0.10349666327238083,
+      "learning_rate": 0.0002,
+      "loss": 0.2466,
+      "step": 1890
+    },
+    {
+      "epoch": 0.398406374501992,
+      "grad_norm": 0.14224931597709656,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4005032501572657,
+      "grad_norm": 0.10312917083501816,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4026001258125393,
+      "grad_norm": 0.10063254833221436,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 1920
+    },
+    {
+      "epoch": 0.40469700146781296,
+      "grad_norm": 0.1020762100815773,
+      "learning_rate": 0.0002,
+      "loss": 0.2549,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4067938771230866,
+      "grad_norm": 0.09656281769275665,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 1940
+    },
+    {
+      "epoch": 0.40889075277836023,
+      "grad_norm": 0.14976176619529724,
+      "learning_rate": 0.0002,
+      "loss": 0.27,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4109876284336339,
+      "grad_norm": 1.410688042640686,
+      "learning_rate": 0.0002,
+      "loss": 0.2953,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4130845040889075,
+      "grad_norm": 0.14357423782348633,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4151813797441812,
+      "grad_norm": 0.08045023679733276,
+      "learning_rate": 0.0002,
+      "loss": 0.2645,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4172782553994548,
+      "grad_norm": 1.9619622230529785,
+      "learning_rate": 0.0002,
+      "loss": 0.2757,
+      "step": 1990
+    },
+    {
+      "epoch": 0.41937513105472846,
+      "grad_norm": 0.06940364837646484,
+      "learning_rate": 0.0002,
+      "loss": 0.2685,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4214720067100021,
+      "grad_norm": 0.11344420164823532,
+      "learning_rate": 0.0002,
+      "loss": 0.2834,
+      "step": 2010
+    },
+    {
+      "epoch": 0.42356888236527573,
+      "grad_norm": 0.08228524029254913,
+      "learning_rate": 0.0002,
+      "loss": 0.2635,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4256657580205494,
+      "grad_norm": 0.07158046960830688,
+      "learning_rate": 0.0002,
+      "loss": 0.2528,
+      "step": 2030
+    },
+    {
+      "epoch": 0.427762633675823,
+      "grad_norm": 4.314914703369141,
+      "learning_rate": 0.0002,
+      "loss": 0.2649,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4298595093310967,
+      "grad_norm": 0.16018514335155487,
+      "learning_rate": 0.0002,
+      "loss": 0.2586,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4319563849863703,
+      "grad_norm": 1.0571707487106323,
+      "learning_rate": 0.0002,
+      "loss": 0.2592,
+      "step": 2060
+    },
+    {
+      "epoch": 0.43405326064164396,
+      "grad_norm": 0.48940908908843994,
+      "learning_rate": 0.0002,
+      "loss": 0.2692,
+      "step": 2070
+    },
+    {
+      "epoch": 0.43615013629691757,
+      "grad_norm": 0.07459184527397156,
+      "learning_rate": 0.0002,
+      "loss": 0.2498,
+      "step": 2080
+    },
+    {
+      "epoch": 0.43824701195219123,
+      "grad_norm": 0.07219909131526947,
+      "learning_rate": 0.0002,
+      "loss": 0.2599,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4403438876074649,
+      "grad_norm": 0.06879916042089462,
+      "learning_rate": 0.0002,
+      "loss": 0.2565,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4424407632627385,
+      "grad_norm": 1.8906136751174927,
+      "learning_rate": 0.0002,
+      "loss": 0.2874,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4445376389180122,
+      "grad_norm": 0.10700809955596924,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4466345145732858,
+      "grad_norm": 0.06732141971588135,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 2130
+    },
+    {
+      "epoch": 0.44873139022855946,
+      "grad_norm": 0.11520029604434967,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 2140
+    },
+    {
+      "epoch": 0.45082826588383307,
+      "grad_norm": 0.08251114189624786,
+      "learning_rate": 0.0002,
+      "loss": 0.2658,
+      "step": 2150
+    },
+    {
+      "epoch": 0.45292514153910673,
+      "grad_norm": 0.10146190226078033,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 2160
+    },
+    {
+      "epoch": 0.45502201719438035,
+      "grad_norm": 0.31207767128944397,
+      "learning_rate": 0.0002,
+      "loss": 0.2574,
+      "step": 2170
+    },
+    {
+      "epoch": 0.457118892849654,
+      "grad_norm": 0.14245358109474182,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4592157685049277,
+      "grad_norm": 0.08076608926057816,
+      "learning_rate": 0.0002,
+      "loss": 0.2912,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4613126441602013,
+      "grad_norm": 0.09804554283618927,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 2200
+    },
+    {
+      "epoch": 0.46340951981547496,
+      "grad_norm": 0.09204811602830887,
+      "learning_rate": 0.0002,
+      "loss": 0.2421,
+      "step": 2210
+    },
+    {
+      "epoch": 0.46550639547074857,
+      "grad_norm": 0.07955732196569443,
+      "learning_rate": 0.0002,
+      "loss": 0.3149,
+      "step": 2220
+    },
+    {
+      "epoch": 0.46760327112602224,
+      "grad_norm": 0.1081770807504654,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 2230
+    },
+    {
+      "epoch": 0.46970014678129585,
+      "grad_norm": 0.1168019101023674,
+      "learning_rate": 0.0002,
+      "loss": 0.303,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4717970224365695,
+      "grad_norm": 0.11181426793336868,
+      "learning_rate": 0.0002,
+      "loss": 0.2832,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4738938980918432,
+      "grad_norm": 0.11993461847305298,
+      "learning_rate": 0.0002,
+      "loss": 0.2953,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4759907737471168,
+      "grad_norm": 0.0672910287976265,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 2270
+    },
+    {
+      "epoch": 0.47808764940239046,
+      "grad_norm": 0.5338168144226074,
+      "learning_rate": 0.0002,
+      "loss": 0.2712,
+      "step": 2280
+    },
+    {
+      "epoch": 0.48018452505766407,
+      "grad_norm": 0.08533361554145813,
+      "learning_rate": 0.0002,
+      "loss": 0.2579,
+      "step": 2290
+    },
+    {
+      "epoch": 0.48228140071293774,
+      "grad_norm": 2.0753543376922607,
+      "learning_rate": 0.0002,
+      "loss": 0.2879,
+      "step": 2300
+    },
+    {
+      "epoch": 0.48437827636821135,
+      "grad_norm": 0.12475968152284622,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 2310
+    },
+    {
+      "epoch": 0.486475152023485,
+      "grad_norm": 0.07769046723842621,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4885720276787586,
+      "grad_norm": 0.45657530426979065,
+      "learning_rate": 0.0002,
+      "loss": 0.2415,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4906689033340323,
+      "grad_norm": 2.8291752338409424,
+      "learning_rate": 0.0002,
+      "loss": 0.2837,
+      "step": 2340
+    },
+    {
+      "epoch": 0.49276577898930596,
+      "grad_norm": 0.16259396076202393,
+      "learning_rate": 0.0002,
+      "loss": 0.2737,
+      "step": 2350
+    },
+    {
+      "epoch": 0.49486265464457957,
+      "grad_norm": 0.27614033222198486,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 2360
+    },
+    {
+      "epoch": 0.49695953029985324,
+      "grad_norm": 0.20377802848815918,
+      "learning_rate": 0.0002,
+      "loss": 0.2555,
+      "step": 2370
+    },
+    {
+      "epoch": 0.49905640595512685,
+      "grad_norm": 0.1173492893576622,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5011532816104005,
+      "grad_norm": 0.2819477617740631,
+      "learning_rate": 0.0002,
+      "loss": 0.3102,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5032501572656741,
+      "grad_norm": 0.11425253748893738,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5053470329209477,
+      "grad_norm": 1.2937549352645874,
+      "learning_rate": 0.0002,
+      "loss": 0.329,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5074439085762215,
+      "grad_norm": 0.16594870388507843,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5095407842314951,
+      "grad_norm": 1.6648545265197754,
+      "learning_rate": 0.0002,
+      "loss": 0.2893,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5116376598867687,
+      "grad_norm": 0.06415042281150818,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5137345355420424,
+      "grad_norm": 0.0840664952993393,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 2450
+    },
+    {
+      "epoch": 0.515831411197316,
+      "grad_norm": 1.8631309270858765,
+      "learning_rate": 0.0002,
+      "loss": 0.27,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5179282868525896,
+      "grad_norm": 0.08405490219593048,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5200251625078632,
+      "grad_norm": 0.10409712791442871,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 2480
+    },
+    {
+      "epoch": 0.522122038163137,
+      "grad_norm": 0.08199995011091232,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5242189138184106,
+      "grad_norm": 0.0819941982626915,
+      "learning_rate": 0.0002,
+      "loss": 0.2859,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.07962234318256378,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5284126651289579,
+      "grad_norm": 0.09645619243383408,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5305095407842315,
+      "grad_norm": 0.10941855609416962,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5326064164395051,
+      "grad_norm": 0.11371888220310211,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5347032920947787,
+      "grad_norm": 1.2199335098266602,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5368001677500525,
+      "grad_norm": 0.054858073592185974,
+      "learning_rate": 0.0002,
+      "loss": 0.2899,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5388970434053261,
+      "grad_norm": 0.0723327100276947,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5409939190605997,
+      "grad_norm": 0.11272138357162476,
+      "learning_rate": 0.0002,
+      "loss": 0.2828,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5430907947158734,
+      "grad_norm": 0.07356975972652435,
+      "learning_rate": 0.0002,
+      "loss": 0.2623,
+      "step": 2590
+    },
+    {
+      "epoch": 0.545187670371147,
+      "grad_norm": 0.12150203436613083,
+      "learning_rate": 0.0002,
+      "loss": 0.2496,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5472845460264206,
+      "grad_norm": 0.1342150866985321,
+      "learning_rate": 0.0002,
+      "loss": 0.2574,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5493814216816942,
+      "grad_norm": 0.10773837566375732,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 2620
+    },
+    {
+      "epoch": 0.551478297336968,
+      "grad_norm": 0.12423650175333023,
+      "learning_rate": 0.0002,
+      "loss": 0.2621,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5535751729922416,
+      "grad_norm": 0.28494974970817566,
+      "learning_rate": 0.0002,
+      "loss": 0.2926,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5556720486475152,
+      "grad_norm": 0.08630665391683578,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5577689243027888,
+      "grad_norm": 0.08209650963544846,
+      "learning_rate": 0.0002,
+      "loss": 0.3102,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5598657999580625,
+      "grad_norm": 0.1205248087644577,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5619626756133361,
+      "grad_norm": 1.9188755750656128,
+      "learning_rate": 0.0002,
+      "loss": 0.265,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5640595512686097,
+      "grad_norm": 0.07110995054244995,
+      "learning_rate": 0.0002,
+      "loss": 0.2709,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5661564269238835,
+      "grad_norm": 0.1682436764240265,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5682533025791571,
+      "grad_norm": 1.8173266649246216,
+      "learning_rate": 0.0002,
+      "loss": 0.2608,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5703501782344307,
+      "grad_norm": 0.07287939637899399,
+      "learning_rate": 0.0002,
+      "loss": 0.2699,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5724470538897043,
+      "grad_norm": 0.08551536500453949,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 2730
+    },
+    {
+      "epoch": 0.574543929544978,
+      "grad_norm": 0.17491847276687622,
+      "learning_rate": 0.0002,
+      "loss": 0.2873,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5766408052002516,
+      "grad_norm": 0.06950732320547104,
+      "learning_rate": 0.0002,
+      "loss": 0.3286,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5787376808555252,
+      "grad_norm": 0.1365327090024948,
+      "learning_rate": 0.0002,
+      "loss": 0.255,
+      "step": 2760
+    },
+    {
+      "epoch": 0.580834556510799,
+      "grad_norm": 1.6514012813568115,
+      "learning_rate": 0.0002,
+      "loss": 0.3005,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5829314321660726,
+      "grad_norm": 0.10234736651182175,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5850283078213462,
+      "grad_norm": 0.11370868235826492,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5871251834766198,
+      "grad_norm": 0.3800756633281708,
+      "learning_rate": 0.0002,
+      "loss": 0.2567,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5892220591318935,
+      "grad_norm": 0.0628165528178215,
+      "learning_rate": 0.0002,
+      "loss": 0.2713,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5913189347871671,
+      "grad_norm": 0.11104313284158707,
+      "learning_rate": 0.0002,
+      "loss": 0.2699,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5934158104424407,
+      "grad_norm": 0.10503487288951874,
+      "learning_rate": 0.0002,
+      "loss": 0.3043,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5955126860977145,
+      "grad_norm": 0.09854432940483093,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5976095617529881,
+      "grad_norm": 0.06774982064962387,
+      "learning_rate": 0.0002,
+      "loss": 0.2625,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5997064374082617,
+      "grad_norm": 0.09751401841640472,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6018033130635353,
+      "grad_norm": 0.07544960081577301,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 2870
+    },
+    {
+      "epoch": 0.603900188718809,
+      "grad_norm": 0.5726602673530579,
+      "learning_rate": 0.0002,
+      "loss": 0.2875,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6059970643740826,
+      "grad_norm": 0.08737173676490784,
+      "learning_rate": 0.0002,
+      "loss": 0.2657,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6080939400293562,
+      "grad_norm": 3.507925033569336,
+      "learning_rate": 0.0002,
+      "loss": 0.309,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6101908156846299,
+      "grad_norm": 2.278453826904297,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6122876913399036,
+      "grad_norm": 1.2136021852493286,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6143845669951772,
+      "grad_norm": 0.09733498841524124,
+      "learning_rate": 0.0002,
+      "loss": 0.2508,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6164814426504508,
+      "grad_norm": 0.0754965990781784,
+      "learning_rate": 0.0002,
+      "loss": 0.279,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6185783183057245,
+      "grad_norm": 0.8958814740180969,
+      "learning_rate": 0.0002,
+      "loss": 0.2697,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6206751939609981,
+      "grad_norm": 0.07990656793117523,
+      "learning_rate": 0.0002,
+      "loss": 0.2661,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6227720696162717,
+      "grad_norm": 2.7301337718963623,
+      "learning_rate": 0.0002,
+      "loss": 0.3001,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6248689452715454,
+      "grad_norm": 0.09214778244495392,
+      "learning_rate": 0.0002,
+      "loss": 0.3344,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6269658209268191,
+      "grad_norm": 3.18005633354187,
+      "learning_rate": 0.0002,
+      "loss": 0.328,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6290626965820927,
+      "grad_norm": 0.11004272103309631,
+      "learning_rate": 0.0002,
+      "loss": 0.2492,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6311595722373663,
+      "grad_norm": 5.838320732116699,
+      "learning_rate": 0.0002,
+      "loss": 0.2496,
+      "step": 3010
+    },
+    {
+      "epoch": 0.63325644789264,
+      "grad_norm": 0.18130512535572052,
+      "learning_rate": 0.0002,
+      "loss": 0.2471,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6353533235479136,
+      "grad_norm": 0.09574416279792786,
+      "learning_rate": 0.0002,
+      "loss": 0.2726,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6374501992031872,
+      "grad_norm": 0.128231942653656,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6395470748584609,
+      "grad_norm": 0.16388727724552155,
+      "learning_rate": 0.0002,
+      "loss": 0.2746,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6416439505137346,
+      "grad_norm": 0.10822491347789764,
+      "learning_rate": 0.0002,
+      "loss": 0.2904,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6437408261690082,
+      "grad_norm": 0.15372033417224884,
+      "learning_rate": 0.0002,
+      "loss": 0.2527,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6458377018242818,
+      "grad_norm": 0.1321578323841095,
+      "learning_rate": 0.0002,
+      "loss": 0.2474,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6479345774795554,
+      "grad_norm": 0.1597270965576172,
+      "learning_rate": 0.0002,
+      "loss": 0.2998,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6500314531348291,
+      "grad_norm": 0.07458209246397018,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6521283287901027,
+      "grad_norm": 2.394174575805664,
+      "learning_rate": 0.0002,
+      "loss": 0.275,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6542252044453764,
+      "grad_norm": 0.12575379014015198,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6563220801006501,
+      "grad_norm": 0.8199005722999573,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6584189557559237,
+      "grad_norm": 0.15044209361076355,
+      "learning_rate": 0.0002,
+      "loss": 0.2937,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6605158314111973,
+      "grad_norm": 1.6776069402694702,
+      "learning_rate": 0.0002,
+      "loss": 0.2731,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6626127070664709,
+      "grad_norm": 0.12253274023532867,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6647095827217446,
+      "grad_norm": 0.050848811864852905,
+      "learning_rate": 0.0002,
+      "loss": 0.2688,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6668064583770182,
+      "grad_norm": 0.09614066779613495,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6689033340322919,
+      "grad_norm": 0.09169217199087143,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6710002096875656,
+      "grad_norm": 0.13360008597373962,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6730970853428392,
+      "grad_norm": 0.11396951973438263,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6751939609981128,
+      "grad_norm": 0.09393135458230972,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6772908366533864,
+      "grad_norm": 0.08728672564029694,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6793877123086601,
+      "grad_norm": 0.07470987737178802,
+      "learning_rate": 0.0002,
+      "loss": 0.3351,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6814845879639337,
+      "grad_norm": 0.13528290390968323,
+      "learning_rate": 0.0002,
+      "loss": 0.2898,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6835814636192074,
+      "grad_norm": 3.9980695247650146,
+      "learning_rate": 0.0002,
+      "loss": 0.2946,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6856783392744811,
+      "grad_norm": 0.09749490767717361,
+      "learning_rate": 0.0002,
+      "loss": 0.304,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6877752149297547,
+      "grad_norm": 0.36559662222862244,
+      "learning_rate": 0.0002,
+      "loss": 0.2947,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6898720905850283,
+      "grad_norm": 1.9381011724472046,
+      "learning_rate": 0.0002,
+      "loss": 0.2889,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6919689662403019,
+      "grad_norm": 0.16328813135623932,
+      "learning_rate": 0.0002,
+      "loss": 0.2509,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6940658418955756,
+      "grad_norm": 0.09724120050668716,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6961627175508492,
+      "grad_norm": 0.08283833414316177,
+      "learning_rate": 0.0002,
+      "loss": 0.2574,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6982595932061229,
+      "grad_norm": 0.12056735903024673,
+      "learning_rate": 0.0002,
+      "loss": 0.2604,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7003564688613965,
+      "grad_norm": 0.10020340979099274,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7024533445166702,
+      "grad_norm": 0.10376536846160889,
+      "learning_rate": 0.0002,
+      "loss": 0.2585,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7045502201719438,
+      "grad_norm": 0.12203848361968994,
+      "learning_rate": 0.0002,
+      "loss": 0.2989,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7066470958272174,
+      "grad_norm": 1.848594069480896,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7087439714824911,
+      "grad_norm": 0.3774644136428833,
+      "learning_rate": 0.0002,
+      "loss": 0.308,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7108408471377647,
+      "grad_norm": 0.09510187804698944,
+      "learning_rate": 0.0002,
+      "loss": 0.2577,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7129377227930384,
+      "grad_norm": 3.617866277694702,
+      "learning_rate": 0.0002,
+      "loss": 0.2625,
+      "step": 3400
+    },
+    {
+      "epoch": 0.715034598448312,
+      "grad_norm": 0.09932883083820343,
+      "learning_rate": 0.0002,
+      "loss": 0.2765,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7171314741035857,
+      "grad_norm": 0.39475202560424805,
+      "learning_rate": 0.0002,
+      "loss": 0.3117,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7192283497588593,
+      "grad_norm": 1.8625463247299194,
+      "learning_rate": 0.0002,
+      "loss": 0.2932,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7213252254141329,
+      "grad_norm": 0.11761441826820374,
+      "learning_rate": 0.0002,
+      "loss": 0.2646,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7234221010694066,
+      "grad_norm": 0.11470779776573181,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7255189767246802,
+      "grad_norm": 0.10077525675296783,
+      "learning_rate": 0.0002,
+      "loss": 0.2416,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7276158523799539,
+      "grad_norm": 3.409393072128296,
+      "learning_rate": 0.0002,
+      "loss": 0.3278,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7297127280352275,
+      "grad_norm": 0.0658923089504242,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7318096036905012,
+      "grad_norm": 0.10743657499551773,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7339064793457748,
+      "grad_norm": 0.10818490386009216,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7360033550010484,
+      "grad_norm": 0.09132494032382965,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7381002306563221,
+      "grad_norm": 0.09728691726922989,
+      "learning_rate": 0.0002,
+      "loss": 0.2841,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7401971063115957,
+      "grad_norm": 0.10897944122552872,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7422939819668694,
+      "grad_norm": 0.143765389919281,
+      "learning_rate": 0.0002,
+      "loss": 0.2645,
+      "step": 3540
+    },
+    {
+      "epoch": 0.744390857622143,
+      "grad_norm": 0.07788211852312088,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7464877332774167,
+      "grad_norm": 3.2369561195373535,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7485846089326903,
+      "grad_norm": 0.0991063192486763,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7506814845879639,
+      "grad_norm": 0.07748661190271378,
+      "learning_rate": 0.0002,
+      "loss": 0.2694,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7527783602432375,
+      "grad_norm": 1.6730282306671143,
+      "learning_rate": 0.0002,
+      "loss": 0.2765,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7548752358985112,
+      "grad_norm": 0.10425275564193726,
+      "learning_rate": 0.0002,
+      "loss": 0.267,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7569721115537849,
+      "grad_norm": 0.08449215441942215,
+      "learning_rate": 0.0002,
+      "loss": 0.3076,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7590689872090585,
+      "grad_norm": 0.09143321216106415,
+      "learning_rate": 0.0002,
+      "loss": 0.2721,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7611658628643322,
+      "grad_norm": 0.0790523886680603,
+      "learning_rate": 0.0002,
+      "loss": 0.2613,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7632627385196058,
+      "grad_norm": 1.5371196269989014,
+      "learning_rate": 0.0002,
+      "loss": 0.2739,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7653596141748794,
+      "grad_norm": 0.06781799346208572,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 3650
+    },
+    {
+      "epoch": 0.767456489830153,
+      "grad_norm": 0.08765111118555069,
+      "learning_rate": 0.0002,
+      "loss": 0.2636,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7695533654854267,
+      "grad_norm": 0.9966158866882324,
+      "learning_rate": 0.0002,
+      "loss": 0.2556,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7716502411407004,
+      "grad_norm": 1.4500865936279297,
+      "learning_rate": 0.0002,
+      "loss": 0.2672,
+      "step": 3680
+    },
+    {
+      "epoch": 0.773747116795974,
+      "grad_norm": 0.1110655665397644,
+      "learning_rate": 0.0002,
+      "loss": 0.2743,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7758439924512477,
+      "grad_norm": 0.0707395151257515,
+      "learning_rate": 0.0002,
+      "loss": 0.2632,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7779408681065213,
+      "grad_norm": 2.7958946228027344,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7800377437617949,
+      "grad_norm": 0.07658112794160843,
+      "learning_rate": 0.0002,
+      "loss": 0.3528,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7821346194170685,
+      "grad_norm": 0.05015365406870842,
+      "learning_rate": 0.0002,
+      "loss": 0.2567,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7842314950723422,
+      "grad_norm": 0.0651109591126442,
+      "learning_rate": 0.0002,
+      "loss": 0.3063,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7863283707276159,
+      "grad_norm": 0.06813528388738632,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7884252463828895,
+      "grad_norm": 0.07487435638904572,
+      "learning_rate": 0.0002,
+      "loss": 0.2611,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7905221220381632,
+      "grad_norm": 0.14289982616901398,
+      "learning_rate": 0.0002,
+      "loss": 0.2834,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7926189976934368,
+      "grad_norm": 3.4994521141052246,
+      "learning_rate": 0.0002,
+      "loss": 0.3402,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7947158733487104,
+      "grad_norm": 0.06324627995491028,
+      "learning_rate": 0.0002,
+      "loss": 0.2671,
+      "step": 3790
+    },
+    {
+      "epoch": 0.796812749003984,
+      "grad_norm": 2.5252187252044678,
+      "learning_rate": 0.0002,
+      "loss": 0.294,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7989096246592577,
+      "grad_norm": 0.07738825678825378,
+      "learning_rate": 0.0002,
+      "loss": 0.2859,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8010065003145314,
+      "grad_norm": 0.13051818311214447,
+      "learning_rate": 0.0002,
+      "loss": 0.2495,
+      "step": 3820
+    },
+    {
+      "epoch": 0.803103375969805,
+      "grad_norm": 0.5131255388259888,
+      "learning_rate": 0.0002,
+      "loss": 0.2471,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8052002516250786,
+      "grad_norm": 0.103961281478405,
+      "learning_rate": 0.0002,
+      "loss": 0.2948,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8072971272803523,
+      "grad_norm": 0.09874695539474487,
+      "learning_rate": 0.0002,
+      "loss": 0.2747,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8093940029356259,
+      "grad_norm": 0.10877745598554611,
+      "learning_rate": 0.0002,
+      "loss": 0.2673,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8114908785908995,
+      "grad_norm": 0.07221799343824387,
+      "learning_rate": 0.0002,
+      "loss": 0.2812,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8135877542461732,
+      "grad_norm": 0.07352789491415024,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8156846299014469,
+      "grad_norm": 0.09816540032625198,
+      "learning_rate": 0.0002,
+      "loss": 0.2962,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8177815055567205,
+      "grad_norm": 0.12398573011159897,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8198783812119941,
+      "grad_norm": 0.05769842490553856,
+      "learning_rate": 0.0002,
+      "loss": 0.2656,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8219752568672678,
+      "grad_norm": 2.771360397338867,
+      "learning_rate": 0.0002,
+      "loss": 0.2693,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8240721325225414,
+      "grad_norm": 0.07756297290325165,
+      "learning_rate": 0.0002,
+      "loss": 0.2973,
+      "step": 3930
+    },
+    {
+      "epoch": 0.826169008177815,
+      "grad_norm": 0.9152072072029114,
+      "learning_rate": 0.0002,
+      "loss": 0.2702,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8282658838330887,
+      "grad_norm": 0.10475791990756989,
+      "learning_rate": 0.0002,
+      "loss": 0.2577,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8303627594883624,
+      "grad_norm": 2.7981786727905273,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 3960
+    },
+    {
+      "epoch": 0.832459635143636,
+      "grad_norm": 0.10307871550321579,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8345565107989096,
+      "grad_norm": 2.470330238342285,
+      "learning_rate": 0.0002,
+      "loss": 0.2901,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8366533864541833,
+      "grad_norm": 0.0711425393819809,
+      "learning_rate": 0.0002,
+      "loss": 0.266,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8387502621094569,
+      "grad_norm": 0.05242660269141197,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8408471377647305,
+      "grad_norm": 0.049969688057899475,
+      "learning_rate": 0.0002,
+      "loss": 0.3441,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8429440134200042,
+      "grad_norm": 1.7452404499053955,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8450408890752779,
+      "grad_norm": 0.0709633007645607,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8471377647305515,
+      "grad_norm": 1.0136632919311523,
+      "learning_rate": 0.0002,
+      "loss": 0.2799,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8492346403858251,
+      "grad_norm": 0.14160498976707458,
+      "learning_rate": 0.0002,
+      "loss": 0.2753,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8513315160410988,
+      "grad_norm": 0.08907493203878403,
+      "learning_rate": 0.0002,
+      "loss": 0.2594,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8534283916963724,
+      "grad_norm": 0.06804440915584564,
+      "learning_rate": 0.0002,
+      "loss": 0.2597,
+      "step": 4070
+    },
+    {
+      "epoch": 0.855525267351646,
+      "grad_norm": 0.10351166874170303,
+      "learning_rate": 0.0002,
+      "loss": 0.2582,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8576221430069196,
+      "grad_norm": 0.09024465084075928,
+      "learning_rate": 0.0002,
+      "loss": 0.2422,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8597190186621934,
+      "grad_norm": 0.09236204624176025,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 4100
+    },
+    {
+      "epoch": 0.861815894317467,
+      "grad_norm": 0.08826452493667603,
+      "learning_rate": 0.0002,
+      "loss": 0.2725,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8639127699727406,
+      "grad_norm": 0.08384134620428085,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8660096456280143,
+      "grad_norm": 0.06862927228212357,
+      "learning_rate": 0.0002,
+      "loss": 0.2904,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8681065212832879,
+      "grad_norm": 0.06906203180551529,
+      "learning_rate": 0.0002,
+      "loss": 0.2589,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8702033969385615,
+      "grad_norm": 0.11079169809818268,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8723002725938351,
+      "grad_norm": 0.05974194407463074,
+      "learning_rate": 0.0002,
+      "loss": 0.2743,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8743971482491089,
+      "grad_norm": 0.08867297321557999,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8764940239043825,
+      "grad_norm": 0.0735638365149498,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8785908995596561,
+      "grad_norm": 0.051879338920116425,
+      "learning_rate": 0.0002,
+      "loss": 0.2431,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8806877752149298,
+      "grad_norm": 0.0882013738155365,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8827846508702034,
+      "grad_norm": 0.09889431297779083,
+      "learning_rate": 0.0002,
+      "loss": 0.2955,
+      "step": 4210
+    },
+    {
+      "epoch": 0.884881526525477,
+      "grad_norm": 0.09547612071037292,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8869784021807506,
+      "grad_norm": 0.1306789666414261,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8890752778360244,
+      "grad_norm": 0.07582208514213562,
+      "learning_rate": 0.0002,
+      "loss": 0.2995,
+      "step": 4240
+    },
+    {
+      "epoch": 0.891172153491298,
+      "grad_norm": 0.30311834812164307,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8932690291465716,
+      "grad_norm": 0.07074534893035889,
+      "learning_rate": 0.0002,
+      "loss": 0.272,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8953659048018453,
+      "grad_norm": 0.12505321204662323,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8974627804571189,
+      "grad_norm": 1.3733842372894287,
+      "learning_rate": 0.0002,
+      "loss": 0.3084,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8995596561123925,
+      "grad_norm": 2.174771785736084,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9016565317676661,
+      "grad_norm": 0.05621308833360672,
+      "learning_rate": 0.0002,
+      "loss": 0.3032,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9037534074229399,
+      "grad_norm": 0.1187891885638237,
+      "learning_rate": 0.0002,
+      "loss": 0.2606,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9058502830782135,
+      "grad_norm": 7.58337926864624,
+      "learning_rate": 0.0002,
+      "loss": 0.3478,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9079471587334871,
+      "grad_norm": 0.07003562152385712,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9100440343887607,
+      "grad_norm": 0.08889756351709366,
+      "learning_rate": 0.0002,
+      "loss": 0.2636,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9121409100440344,
+      "grad_norm": 0.07013625651597977,
+      "learning_rate": 0.0002,
+      "loss": 0.2724,
+      "step": 4350
+    },
+    {
+      "epoch": 0.914237785699308,
+      "grad_norm": 0.06386245042085648,
+      "learning_rate": 0.0002,
+      "loss": 0.2429,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9163346613545816,
+      "grad_norm": 0.09225624054670334,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9184315370098554,
+      "grad_norm": 0.07157222181558609,
+      "learning_rate": 0.0002,
+      "loss": 0.2636,
+      "step": 4380
+    },
+    {
+      "epoch": 0.920528412665129,
+      "grad_norm": 0.05457375571131706,
+      "learning_rate": 0.0002,
+      "loss": 0.2875,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9226252883204026,
+      "grad_norm": 0.06286972016096115,
+      "learning_rate": 0.0002,
+      "loss": 0.2682,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9247221639756762,
+      "grad_norm": 0.12173102796077728,
+      "learning_rate": 0.0002,
+      "loss": 0.287,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9268190396309499,
+      "grad_norm": 0.10016096383333206,
+      "learning_rate": 0.0002,
+      "loss": 0.2683,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9289159152862235,
+      "grad_norm": 0.3565249741077423,
+      "learning_rate": 0.0002,
+      "loss": 0.2688,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9310127909414971,
+      "grad_norm": 1.8416194915771484,
+      "learning_rate": 0.0002,
+      "loss": 0.2896,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9331096665967709,
+      "grad_norm": 0.07465513795614243,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9352065422520445,
+      "grad_norm": 0.12205296009778976,
+      "learning_rate": 0.0002,
+      "loss": 0.2596,
+      "step": 4460
+    },
+    {
+      "epoch": 0.9373034179073181,
+      "grad_norm": 0.07248952239751816,
+      "learning_rate": 0.0002,
+      "loss": 0.2653,
+      "step": 4470
+    },
+    {
+      "epoch": 0.9394002935625917,
+      "grad_norm": 0.6678045392036438,
+      "learning_rate": 0.0002,
+      "loss": 0.3125,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9414971692178654,
+      "grad_norm": 0.08333967626094818,
+      "learning_rate": 0.0002,
+      "loss": 0.2988,
+      "step": 4490
+    },
+    {
+      "epoch": 0.943594044873139,
+      "grad_norm": 1.3239233493804932,
+      "learning_rate": 0.0002,
+      "loss": 0.2533,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9456909205284126,
+      "grad_norm": 0.10770973563194275,
+      "learning_rate": 0.0002,
+      "loss": 0.264,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9477877961836864,
+      "grad_norm": 0.07700380682945251,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 4520
+    },
+    {
+      "epoch": 0.94988467183896,
+      "grad_norm": 0.05980914458632469,
+      "learning_rate": 0.0002,
+      "loss": 0.291,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9519815474942336,
+      "grad_norm": 0.05780167877674103,
+      "learning_rate": 0.0002,
+      "loss": 0.3009,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9540784231495072,
+      "grad_norm": 0.0757979080080986,
+      "learning_rate": 0.0002,
+      "loss": 0.2625,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9561752988047809,
+      "grad_norm": 0.06890128552913666,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9582721744600545,
+      "grad_norm": 0.07870949804782867,
+      "learning_rate": 0.0002,
+      "loss": 0.271,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9603690501153281,
+      "grad_norm": 0.08925454318523407,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9624659257706017,
+      "grad_norm": 0.08406192064285278,
+      "learning_rate": 0.0002,
+      "loss": 0.2521,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9645628014258755,
+      "grad_norm": 0.10540168732404709,
+      "learning_rate": 0.0002,
+      "loss": 0.2424,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9666596770811491,
+      "grad_norm": 0.08675917983055115,
+      "learning_rate": 0.0002,
+      "loss": 0.2872,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9687565527364227,
+      "grad_norm": 2.3604719638824463,
+      "learning_rate": 0.0002,
+      "loss": 0.2933,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9708534283916964,
+      "grad_norm": 0.1408311426639557,
+      "learning_rate": 0.0002,
+      "loss": 0.2604,
+      "step": 4630
+    },
+    {
+      "epoch": 0.97295030404697,
+      "grad_norm": 0.1045566201210022,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9750471797022436,
+      "grad_norm": 0.08125372976064682,
+      "learning_rate": 0.0002,
+      "loss": 0.2542,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9771440553575172,
+      "grad_norm": 3.7775802612304688,
+      "learning_rate": 0.0002,
+      "loss": 0.2831,
+      "step": 4660
+    },
+    {
+      "epoch": 0.979240931012791,
+      "grad_norm": 0.12283279001712799,
+      "learning_rate": 0.0002,
+      "loss": 0.2473,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9813378066680646,
+      "grad_norm": 0.11770537495613098,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9834346823233382,
+      "grad_norm": 0.07334215193986893,
+      "learning_rate": 0.0002,
+      "loss": 0.2601,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9855315579786119,
+      "grad_norm": 0.0814523994922638,
+      "learning_rate": 0.0002,
+      "loss": 0.2918,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9876284336338855,
+      "grad_norm": 0.054461974650621414,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9897253092891591,
+      "grad_norm": 0.06933179497718811,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9918221849444327,
+      "grad_norm": 0.16758371889591217,
+      "learning_rate": 0.0002,
+      "loss": 0.2685,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9939190605997065,
+      "grad_norm": 0.06895585358142853,
+      "learning_rate": 0.0002,
+      "loss": 0.2709,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9960159362549801,
+      "grad_norm": 0.05912035331130028,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9981128119102537,
+      "grad_norm": 0.13841500878334045,
+      "learning_rate": 0.0002,
+      "loss": 0.2673,
+      "step": 4760
+    },
+    {
+      "epoch": 1.0002096875655273,
+      "grad_norm": 0.10441073030233383,
+      "learning_rate": 0.0002,
+      "loss": 0.27,
+      "step": 4770
+    },
+    {
+      "epoch": 1.002306563220801,
+      "grad_norm": 0.16520650684833527,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 4780
+    },
+    {
+      "epoch": 1.0044034388760748,
+      "grad_norm": 1.2790181636810303,
+      "learning_rate": 0.0002,
+      "loss": 0.2502,
+      "step": 4790
+    },
+    {
+      "epoch": 1.0065003145313483,
+      "grad_norm": 0.08104977756738663,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 4800
+    },
+    {
+      "epoch": 1.008597190186622,
+      "grad_norm": 0.07370474934577942,
+      "learning_rate": 0.0002,
+      "loss": 0.2611,
+      "step": 4810
+    },
+    {
+      "epoch": 1.0106940658418955,
+      "grad_norm": 0.07946150749921799,
+      "learning_rate": 0.0002,
+      "loss": 0.2761,
+      "step": 4820
+    },
+    {
+      "epoch": 1.0127909414971692,
+      "grad_norm": 0.1244131550192833,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 4830
+    },
+    {
+      "epoch": 1.014887817152443,
+      "grad_norm": 0.08558488637208939,
+      "learning_rate": 0.0002,
+      "loss": 0.2958,
+      "step": 4840
+    },
+    {
+      "epoch": 1.0169846928077164,
+      "grad_norm": 2.46805477142334,
+      "learning_rate": 0.0002,
+      "loss": 0.265,
+      "step": 4850
+    },
+    {
+      "epoch": 1.0190815684629901,
+      "grad_norm": 0.06383222341537476,
+      "learning_rate": 0.0002,
+      "loss": 0.3791,
+      "step": 4860
+    },
+    {
+      "epoch": 1.0211784441182639,
+      "grad_norm": 3.0055954456329346,
+      "learning_rate": 0.0002,
+      "loss": 0.3129,
+      "step": 4870
+    },
+    {
+      "epoch": 1.0232753197735374,
+      "grad_norm": 2.8449673652648926,
+      "learning_rate": 0.0002,
+      "loss": 0.3186,
+      "step": 4880
+    },
+    {
+      "epoch": 1.025372195428811,
+      "grad_norm": 0.07504308223724365,
+      "learning_rate": 0.0002,
+      "loss": 0.2656,
+      "step": 4890
+    },
+    {
+      "epoch": 1.0274690710840848,
+      "grad_norm": 0.9263021945953369,
+      "learning_rate": 0.0002,
+      "loss": 0.2822,
+      "step": 4900
+    },
+    {
+      "epoch": 1.0295659467393583,
+      "grad_norm": 0.07075469195842743,
+      "learning_rate": 0.0002,
+      "loss": 0.2842,
+      "step": 4910
+    },
+    {
+      "epoch": 1.031662822394632,
+      "grad_norm": 0.36625659465789795,
+      "learning_rate": 0.0002,
+      "loss": 0.2499,
+      "step": 4920
+    },
+    {
+      "epoch": 1.0337596980499058,
+      "grad_norm": 0.07981095463037491,
+      "learning_rate": 0.0002,
+      "loss": 0.2609,
+      "step": 4930
+    },
+    {
+      "epoch": 1.0358565737051793,
+      "grad_norm": 0.09062483161687851,
+      "learning_rate": 0.0002,
+      "loss": 0.2947,
+      "step": 4940
+    },
+    {
+      "epoch": 1.037953449360453,
+      "grad_norm": 0.5152995586395264,
+      "learning_rate": 0.0002,
+      "loss": 0.2673,
+      "step": 4950
+    },
+    {
+      "epoch": 1.0400503250157265,
+      "grad_norm": 0.09733402729034424,
+      "learning_rate": 0.0002,
+      "loss": 0.2574,
+      "step": 4960
+    },
+    {
+      "epoch": 1.0421472006710002,
+      "grad_norm": 0.107651486992836,
+      "learning_rate": 0.0002,
+      "loss": 0.3059,
+      "step": 4970
+    },
+    {
+      "epoch": 1.044244076326274,
+      "grad_norm": 0.22185586392879486,
+      "learning_rate": 0.0002,
+      "loss": 0.3173,
+      "step": 4980
+    },
+    {
+      "epoch": 1.0463409519815474,
+      "grad_norm": 0.15324990451335907,
+      "learning_rate": 0.0002,
+      "loss": 0.2735,
+      "step": 4990
+    },
+    {
+      "epoch": 1.0484378276368211,
+      "grad_norm": 0.12578018009662628,
+      "learning_rate": 0.0002,
+      "loss": 0.2551,
+      "step": 5000
+    },
+    {
+      "epoch": 1.0505347032920949,
+      "grad_norm": 0.06980187445878983,
+      "learning_rate": 0.0002,
+      "loss": 0.2488,
+      "step": 5010
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.11194098740816116,
+      "learning_rate": 0.0002,
+      "loss": 0.2662,
+      "step": 5020
+    },
+    {
+      "epoch": 1.054728454602642,
+      "grad_norm": 0.0770438089966774,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 5030
+    },
+    {
+      "epoch": 1.0568253302579158,
+      "grad_norm": 0.10786891728639603,
+      "learning_rate": 0.0002,
+      "loss": 0.2821,
+      "step": 5040
+    },
+    {
+      "epoch": 1.0589222059131893,
+      "grad_norm": 0.09369146078824997,
+      "learning_rate": 0.0002,
+      "loss": 0.2591,
+      "step": 5050
+    },
+    {
+      "epoch": 1.061019081568463,
+      "grad_norm": 0.07498877495527267,
+      "learning_rate": 0.0002,
+      "loss": 0.2595,
+      "step": 5060
+    },
+    {
+      "epoch": 1.0631159572237365,
+      "grad_norm": 0.226748526096344,
+      "learning_rate": 0.0002,
+      "loss": 0.2828,
+      "step": 5070
+    },
+    {
+      "epoch": 1.0652128328790103,
+      "grad_norm": 0.09331674873828888,
+      "learning_rate": 0.0002,
+      "loss": 0.2564,
+      "step": 5080
+    },
+    {
+      "epoch": 1.067309708534284,
+      "grad_norm": 0.10891097784042358,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 5090
+    },
+    {
+      "epoch": 1.0694065841895575,
+      "grad_norm": 0.0820799469947815,
+      "learning_rate": 0.0002,
+      "loss": 0.2548,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0715034598448312,
+      "grad_norm": 0.052631497383117676,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 5110
+    },
+    {
+      "epoch": 1.073600335500105,
+      "grad_norm": 0.12196387350559235,
+      "learning_rate": 0.0002,
+      "loss": 0.2594,
+      "step": 5120
+    },
+    {
+      "epoch": 1.0756972111553784,
+      "grad_norm": 0.07985110580921173,
+      "learning_rate": 0.0002,
+      "loss": 0.2743,
+      "step": 5130
+    },
+    {
+      "epoch": 1.0777940868106521,
+      "grad_norm": 0.05758510157465935,
+      "learning_rate": 0.0002,
+      "loss": 0.2919,
+      "step": 5140
+    },
+    {
+      "epoch": 1.0798909624659259,
+      "grad_norm": 0.10148131102323532,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 5150
+    },
+    {
+      "epoch": 1.0819878381211994,
+      "grad_norm": 0.07871562242507935,
+      "learning_rate": 0.0002,
+      "loss": 0.2798,
+      "step": 5160
+    },
+    {
+      "epoch": 1.084084713776473,
+      "grad_norm": 0.06986663490533829,
+      "learning_rate": 0.0002,
+      "loss": 0.2966,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0861815894317468,
+      "grad_norm": 0.05995866283774376,
+      "learning_rate": 0.0002,
+      "loss": 0.2569,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0882784650870203,
+      "grad_norm": 0.12248244881629944,
+      "learning_rate": 0.0002,
+      "loss": 0.2906,
+      "step": 5190
+    },
+    {
+      "epoch": 1.090375340742294,
+      "grad_norm": 0.052593328058719635,
+      "learning_rate": 0.0002,
+      "loss": 0.2561,
+      "step": 5200
+    },
+    {
+      "epoch": 1.0924722163975675,
+      "grad_norm": 0.05631803721189499,
+      "learning_rate": 0.0002,
+      "loss": 0.2943,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0945690920528413,
+      "grad_norm": 2.102130889892578,
+      "learning_rate": 0.0002,
+      "loss": 0.2991,
+      "step": 5220
+    },
+    {
+      "epoch": 1.096665967708115,
+      "grad_norm": 2.2242186069488525,
+      "learning_rate": 0.0002,
+      "loss": 0.258,
+      "step": 5230
+    },
+    {
+      "epoch": 1.0987628433633885,
+      "grad_norm": 0.06405676156282425,
+      "learning_rate": 0.0002,
+      "loss": 0.2845,
+      "step": 5240
+    },
+    {
+      "epoch": 1.1008597190186622,
+      "grad_norm": 0.05279644578695297,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 5250
+    },
+    {
+      "epoch": 1.102956594673936,
+      "grad_norm": 0.06458443403244019,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 5260
+    },
+    {
+      "epoch": 1.1050534703292094,
+      "grad_norm": 0.07742719352245331,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 5270
+    },
+    {
+      "epoch": 1.1071503459844831,
+      "grad_norm": 0.0937500149011612,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 5280
+    },
+    {
+      "epoch": 1.1092472216397569,
+      "grad_norm": 0.05128967761993408,
+      "learning_rate": 0.0002,
+      "loss": 0.2623,
+      "step": 5290
+    },
+    {
+      "epoch": 1.1113440972950304,
+      "grad_norm": 0.053862564265728,
+      "learning_rate": 0.0002,
+      "loss": 0.26,
+      "step": 5300
+    },
+    {
+      "epoch": 1.113440972950304,
+      "grad_norm": 0.07308442145586014,
+      "learning_rate": 0.0002,
+      "loss": 0.2653,
+      "step": 5310
+    },
+    {
+      "epoch": 1.1155378486055776,
+      "grad_norm": 0.08775884658098221,
+      "learning_rate": 0.0002,
+      "loss": 0.257,
+      "step": 5320
+    },
+    {
+      "epoch": 1.1176347242608513,
+      "grad_norm": 0.06515777856111526,
+      "learning_rate": 0.0002,
+      "loss": 0.2857,
+      "step": 5330
+    },
+    {
+      "epoch": 1.119731599916125,
+      "grad_norm": 0.07291428744792938,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 5340
+    },
+    {
+      "epoch": 1.1218284755713985,
+      "grad_norm": 0.10684103518724442,
+      "learning_rate": 0.0002,
+      "loss": 0.2425,
+      "step": 5350
+    },
+    {
+      "epoch": 1.1239253512266723,
+      "grad_norm": 2.470691680908203,
+      "learning_rate": 0.0002,
+      "loss": 0.2873,
+      "step": 5360
+    },
+    {
+      "epoch": 1.126022226881946,
+      "grad_norm": 1.8290586471557617,
+      "learning_rate": 0.0002,
+      "loss": 0.2646,
+      "step": 5370
+    },
+    {
+      "epoch": 1.1281191025372195,
+      "grad_norm": 0.08278166502714157,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 5380
+    },
+    {
+      "epoch": 1.1302159781924932,
+      "grad_norm": 0.07949317991733551,
+      "learning_rate": 0.0002,
+      "loss": 0.2741,
+      "step": 5390
+    },
+    {
+      "epoch": 1.132312853847767,
+      "grad_norm": 0.11934222280979156,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 5400
+    },
+    {
+      "epoch": 1.1344097295030404,
+      "grad_norm": 0.09001024812459946,
+      "learning_rate": 0.0002,
+      "loss": 0.2533,
+      "step": 5410
+    },
+    {
+      "epoch": 1.1365066051583141,
+      "grad_norm": 0.0852462500333786,
+      "learning_rate": 0.0002,
+      "loss": 0.2657,
+      "step": 5420
+    },
+    {
+      "epoch": 1.1386034808135879,
+      "grad_norm": 0.08970583975315094,
+      "learning_rate": 0.0002,
+      "loss": 0.2831,
+      "step": 5430
+    },
+    {
+      "epoch": 1.1407003564688614,
+      "grad_norm": 0.10680624097585678,
+      "learning_rate": 0.0002,
+      "loss": 0.2702,
+      "step": 5440
+    },
+    {
+      "epoch": 1.142797232124135,
+      "grad_norm": 0.1428230106830597,
+      "learning_rate": 0.0002,
+      "loss": 0.2779,
+      "step": 5450
+    },
+    {
+      "epoch": 1.1448941077794086,
+      "grad_norm": 0.0725833997130394,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 5460
+    },
+    {
+      "epoch": 1.1469909834346823,
+      "grad_norm": 1.3643620014190674,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 5470
+    },
+    {
+      "epoch": 1.149087859089956,
+      "grad_norm": 0.07396101206541061,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 5480
+    },
+    {
+      "epoch": 1.1511847347452295,
+      "grad_norm": 0.07980525493621826,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 5490
+    },
+    {
+      "epoch": 1.1532816104005033,
+      "grad_norm": 0.056196991354227066,
+      "learning_rate": 0.0002,
+      "loss": 0.2878,
+      "step": 5500
+    },
+    {
+      "epoch": 1.155378486055777,
+      "grad_norm": 0.057810667902231216,
+      "learning_rate": 0.0002,
+      "loss": 0.2955,
+      "step": 5510
+    },
+    {
+      "epoch": 1.1574753617110505,
+      "grad_norm": 0.15831854939460754,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 5520
+    },
+    {
+      "epoch": 1.1595722373663242,
+      "grad_norm": 0.07265540957450867,
+      "learning_rate": 0.0002,
+      "loss": 0.2538,
+      "step": 5530
+    },
+    {
+      "epoch": 1.1616691130215977,
+      "grad_norm": 0.21104565262794495,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 5540
+    },
+    {
+      "epoch": 1.1637659886768714,
+      "grad_norm": 0.0715293288230896,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 5550
+    },
+    {
+      "epoch": 1.1658628643321451,
+      "grad_norm": 0.10274804383516312,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 5560
+    },
+    {
+      "epoch": 1.1679597399874186,
+      "grad_norm": 4.549091339111328,
+      "learning_rate": 0.0002,
+      "loss": 0.2638,
+      "step": 5570
+    },
+    {
+      "epoch": 1.1700566156426924,
+      "grad_norm": 0.6761789917945862,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 5580
+    },
+    {
+      "epoch": 1.172153491297966,
+      "grad_norm": 0.10144541412591934,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 5590
+    },
+    {
+      "epoch": 1.1742503669532396,
+      "grad_norm": 1.6135568618774414,
+      "learning_rate": 0.0002,
+      "loss": 0.3067,
+      "step": 5600
+    },
+    {
+      "epoch": 1.1763472426085133,
+      "grad_norm": 0.07532897591590881,
+      "learning_rate": 0.0002,
+      "loss": 0.2809,
+      "step": 5610
+    },
+    {
+      "epoch": 1.178444118263787,
+      "grad_norm": 0.05869752913713455,
+      "learning_rate": 0.0002,
+      "loss": 0.2423,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1805409939190605,
+      "grad_norm": 0.10139182209968567,
+      "learning_rate": 0.0002,
+      "loss": 0.2565,
+      "step": 5630
+    },
+    {
+      "epoch": 1.1826378695743343,
+      "grad_norm": 0.25902649760246277,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 5640
+    },
+    {
+      "epoch": 1.184734745229608,
+      "grad_norm": 0.12678462266921997,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 5650
+    },
+    {
+      "epoch": 1.1868316208848815,
+      "grad_norm": 0.06784678995609283,
+      "learning_rate": 0.0002,
+      "loss": 0.2678,
+      "step": 5660
+    },
+    {
+      "epoch": 1.1889284965401552,
+      "grad_norm": 3.252060651779175,
+      "learning_rate": 0.0002,
+      "loss": 0.3044,
+      "step": 5670
+    },
+    {
+      "epoch": 1.191025372195429,
+      "grad_norm": 0.07103057205677032,
+      "learning_rate": 0.0002,
+      "loss": 0.2694,
+      "step": 5680
+    },
+    {
+      "epoch": 1.1931222478507024,
+      "grad_norm": 0.08140352368354797,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 5690
+    },
+    {
+      "epoch": 1.1952191235059761,
+      "grad_norm": 1.4584933519363403,
+      "learning_rate": 0.0002,
+      "loss": 0.2636,
+      "step": 5700
+    },
+    {
+      "epoch": 1.1973159991612496,
+      "grad_norm": 1.501371145248413,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 5710
+    },
+    {
+      "epoch": 1.1994128748165234,
+      "grad_norm": 0.0723869800567627,
+      "learning_rate": 0.0002,
+      "loss": 0.2591,
+      "step": 5720
+    },
+    {
+      "epoch": 1.201509750471797,
+      "grad_norm": 1.0655723810195923,
+      "learning_rate": 0.0002,
+      "loss": 0.297,
+      "step": 5730
+    },
+    {
+      "epoch": 1.2036066261270706,
+      "grad_norm": 5.795892715454102,
+      "learning_rate": 0.0002,
+      "loss": 0.2696,
+      "step": 5740
+    },
+    {
+      "epoch": 1.2057035017823443,
+      "grad_norm": 0.08217945694923401,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 5750
+    },
+    {
+      "epoch": 1.207800377437618,
+      "grad_norm": 0.1614740788936615,
+      "learning_rate": 0.0002,
+      "loss": 0.2606,
+      "step": 5760
+    },
+    {
+      "epoch": 1.2098972530928915,
+      "grad_norm": 0.07378556579351425,
+      "learning_rate": 0.0002,
+      "loss": 0.3326,
+      "step": 5770
+    },
+    {
+      "epoch": 1.2119941287481653,
+      "grad_norm": 0.07546745985746384,
+      "learning_rate": 0.0002,
+      "loss": 0.2596,
+      "step": 5780
+    },
+    {
+      "epoch": 1.2140910044034388,
+      "grad_norm": 2.203601360321045,
+      "learning_rate": 0.0002,
+      "loss": 0.2825,
+      "step": 5790
+    },
+    {
+      "epoch": 1.2161878800587125,
+      "grad_norm": 0.0542774498462677,
+      "learning_rate": 0.0002,
+      "loss": 0.2657,
+      "step": 5800
+    },
+    {
+      "epoch": 1.2182847557139862,
+      "grad_norm": 1.252591848373413,
+      "learning_rate": 0.0002,
+      "loss": 0.2573,
+      "step": 5810
+    },
+    {
+      "epoch": 1.2203816313692597,
+      "grad_norm": 0.1051403135061264,
+      "learning_rate": 0.0002,
+      "loss": 0.3223,
+      "step": 5820
+    },
+    {
+      "epoch": 1.2224785070245334,
+      "grad_norm": 0.13013513386249542,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 5830
+    },
+    {
+      "epoch": 1.2245753826798071,
+      "grad_norm": 0.11686304211616516,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 5840
+    },
+    {
+      "epoch": 1.2266722583350806,
+      "grad_norm": 2.8360345363616943,
+      "learning_rate": 0.0002,
+      "loss": 0.2861,
+      "step": 5850
+    },
+    {
+      "epoch": 1.2287691339903544,
+      "grad_norm": 0.06085509806871414,
+      "learning_rate": 0.0002,
+      "loss": 0.2602,
+      "step": 5860
+    },
+    {
+      "epoch": 1.230866009645628,
+      "grad_norm": 0.07168704271316528,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 5870
+    },
+    {
+      "epoch": 1.2329628853009016,
+      "grad_norm": 0.05347032472491264,
+      "learning_rate": 0.0002,
+      "loss": 0.2706,
+      "step": 5880
+    },
+    {
+      "epoch": 1.2350597609561753,
+      "grad_norm": 0.0719766765832901,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 5890
+    },
+    {
+      "epoch": 1.237156636611449,
+      "grad_norm": 2.9147186279296875,
+      "learning_rate": 0.0002,
+      "loss": 0.3168,
+      "step": 5900
+    },
+    {
+      "epoch": 1.2392535122667225,
+      "grad_norm": 0.0879439190030098,
+      "learning_rate": 0.0002,
+      "loss": 0.2585,
+      "step": 5910
+    },
+    {
+      "epoch": 1.2413503879219963,
+      "grad_norm": 0.09813595563173294,
+      "learning_rate": 0.0002,
+      "loss": 0.2606,
+      "step": 5920
+    },
+    {
+      "epoch": 1.24344726357727,
+      "grad_norm": 1.3518685102462769,
+      "learning_rate": 0.0002,
+      "loss": 0.2705,
+      "step": 5930
+    },
+    {
+      "epoch": 1.2455441392325435,
+      "grad_norm": 1.7133814096450806,
+      "learning_rate": 0.0002,
+      "loss": 0.2633,
+      "step": 5940
+    },
+    {
+      "epoch": 1.2476410148878172,
+      "grad_norm": 0.09435052424669266,
+      "learning_rate": 0.0002,
+      "loss": 0.2657,
+      "step": 5950
+    },
+    {
+      "epoch": 1.2497378905430907,
+      "grad_norm": 0.9879856705665588,
+      "learning_rate": 0.0002,
+      "loss": 0.2667,
+      "step": 5960
+    },
+    {
+      "epoch": 1.2518347661983644,
+      "grad_norm": 1.1095253229141235,
+      "learning_rate": 0.0002,
+      "loss": 0.2978,
+      "step": 5970
+    },
+    {
+      "epoch": 1.2539316418536381,
+      "grad_norm": 0.12190214544534683,
+      "learning_rate": 0.0002,
+      "loss": 0.2556,
+      "step": 5980
+    },
+    {
+      "epoch": 1.2560285175089116,
+      "grad_norm": 0.05829515680670738,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 5990
+    },
+    {
+      "epoch": 1.2581253931641854,
+      "grad_norm": 0.0911063402891159,
+      "learning_rate": 0.0002,
+      "loss": 0.2427,
+      "step": 6000
+    },
+    {
+      "epoch": 1.2602222688194589,
+      "grad_norm": 0.08175632357597351,
+      "learning_rate": 0.0002,
+      "loss": 0.2714,
+      "step": 6010
+    },
+    {
+      "epoch": 1.2623191444747326,
+      "grad_norm": 0.0865393579006195,
+      "learning_rate": 0.0002,
+      "loss": 0.3005,
+      "step": 6020
+    },
+    {
+      "epoch": 1.2644160201300063,
+      "grad_norm": 0.08705785870552063,
+      "learning_rate": 0.0002,
+      "loss": 0.2793,
+      "step": 6030
+    },
+    {
+      "epoch": 1.2665128957852798,
+      "grad_norm": 0.06791888922452927,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 6040
+    },
+    {
+      "epoch": 1.2686097714405535,
+      "grad_norm": 0.10770615190267563,
+      "learning_rate": 0.0002,
+      "loss": 0.3023,
+      "step": 6050
+    },
+    {
+      "epoch": 1.2707066470958273,
+      "grad_norm": 0.06609726697206497,
+      "learning_rate": 0.0002,
+      "loss": 0.2593,
+      "step": 6060
+    },
+    {
+      "epoch": 1.2728035227511008,
+      "grad_norm": 0.06439893692731857,
+      "learning_rate": 0.0002,
+      "loss": 0.2482,
+      "step": 6070
+    },
+    {
+      "epoch": 1.2749003984063745,
+      "grad_norm": 0.07232308387756348,
+      "learning_rate": 0.0002,
+      "loss": 0.2696,
+      "step": 6080
+    },
+    {
+      "epoch": 1.2769972740616482,
+      "grad_norm": 0.09587598592042923,
+      "learning_rate": 0.0002,
+      "loss": 0.2738,
+      "step": 6090
+    },
+    {
+      "epoch": 1.2790941497169217,
+      "grad_norm": 0.06788215786218643,
+      "learning_rate": 0.0002,
+      "loss": 0.2758,
+      "step": 6100
+    },
+    {
+      "epoch": 1.2811910253721954,
+      "grad_norm": 0.09348220378160477,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 6110
+    },
+    {
+      "epoch": 1.2832879010274691,
+      "grad_norm": 0.058364514261484146,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 6120
+    },
+    {
+      "epoch": 1.2853847766827426,
+      "grad_norm": 0.06461839377880096,
+      "learning_rate": 0.0002,
+      "loss": 0.2731,
+      "step": 6130
+    },
+    {
+      "epoch": 1.2874816523380164,
+      "grad_norm": 2.521012306213379,
+      "learning_rate": 0.0002,
+      "loss": 0.2755,
+      "step": 6140
+    },
+    {
+      "epoch": 1.28957852799329,
+      "grad_norm": 7.787893772125244,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 6150
+    },
+    {
+      "epoch": 1.2916754036485636,
+      "grad_norm": 0.07614123821258545,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 6160
+    },
+    {
+      "epoch": 1.2937722793038373,
+      "grad_norm": 0.06463972479104996,
+      "learning_rate": 0.0002,
+      "loss": 0.2627,
+      "step": 6170
+    },
+    {
+      "epoch": 1.295869154959111,
+      "grad_norm": 0.07085216790437698,
+      "learning_rate": 0.0002,
+      "loss": 0.2832,
+      "step": 6180
+    },
+    {
+      "epoch": 1.2979660306143845,
+      "grad_norm": 0.6012230515480042,
+      "learning_rate": 0.0002,
+      "loss": 0.2546,
+      "step": 6190
+    },
+    {
+      "epoch": 1.3000629062696583,
+      "grad_norm": 1.896254062652588,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 6200
+    },
+    {
+      "epoch": 1.302159781924932,
+      "grad_norm": 0.06487879157066345,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 6210
+    },
+    {
+      "epoch": 1.3042566575802055,
+      "grad_norm": 0.0652952492237091,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 6220
+    },
+    {
+      "epoch": 1.3063535332354792,
+      "grad_norm": 0.0889468789100647,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 6230
+    },
+    {
+      "epoch": 1.3084504088907527,
+      "grad_norm": 0.07459229975938797,
+      "learning_rate": 0.0002,
+      "loss": 0.257,
+      "step": 6240
+    },
+    {
+      "epoch": 1.3105472845460264,
+      "grad_norm": 0.40443506836891174,
+      "learning_rate": 0.0002,
+      "loss": 0.2907,
+      "step": 6250
+    },
+    {
+      "epoch": 1.3126441602013,
+      "grad_norm": 0.0493333600461483,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 6260
+    },
+    {
+      "epoch": 1.3147410358565736,
+      "grad_norm": 0.06762096285820007,
+      "learning_rate": 0.0002,
+      "loss": 0.2752,
+      "step": 6270
+    },
+    {
+      "epoch": 1.3168379115118474,
+      "grad_norm": 0.08447722345590591,
+      "learning_rate": 0.0002,
+      "loss": 0.2925,
+      "step": 6280
+    },
+    {
+      "epoch": 1.3189347871671209,
+      "grad_norm": 0.1290348619222641,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 6290
+    },
+    {
+      "epoch": 1.3210316628223946,
+      "grad_norm": 1.5261108875274658,
+      "learning_rate": 0.0002,
+      "loss": 0.258,
+      "step": 6300
+    },
+    {
+      "epoch": 1.3231285384776683,
+      "grad_norm": 0.07950276881456375,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 6310
+    },
+    {
+      "epoch": 1.3252254141329418,
+      "grad_norm": 0.054064471274614334,
+      "learning_rate": 0.0002,
+      "loss": 0.2655,
+      "step": 6320
+    },
+    {
+      "epoch": 1.3273222897882155,
+      "grad_norm": 1.496674656867981,
+      "learning_rate": 0.0002,
+      "loss": 0.2555,
+      "step": 6330
+    },
+    {
+      "epoch": 1.3294191654434893,
+      "grad_norm": 0.0939754843711853,
+      "learning_rate": 0.0002,
+      "loss": 0.2475,
+      "step": 6340
+    },
+    {
+      "epoch": 1.3315160410987628,
+      "grad_norm": 0.060312144458293915,
+      "learning_rate": 0.0002,
+      "loss": 0.2626,
+      "step": 6350
+    },
+    {
+      "epoch": 1.3336129167540365,
+      "grad_norm": 0.05656959488987923,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 6360
+    },
+    {
+      "epoch": 1.3357097924093102,
+      "grad_norm": 0.054611966013908386,
+      "learning_rate": 0.0002,
+      "loss": 0.2695,
+      "step": 6370
+    },
+    {
+      "epoch": 1.3378066680645837,
+      "grad_norm": 0.10933967679738998,
+      "learning_rate": 0.0002,
+      "loss": 0.2848,
+      "step": 6380
+    },
+    {
+      "epoch": 1.3399035437198574,
+      "grad_norm": 0.0762239620089531,
+      "learning_rate": 0.0002,
+      "loss": 0.2553,
+      "step": 6390
+    },
+    {
+      "epoch": 1.3420004193751311,
+      "grad_norm": 0.06087252497673035,
+      "learning_rate": 0.0002,
+      "loss": 0.2708,
+      "step": 6400
+    },
+    {
+      "epoch": 1.3440972950304046,
+      "grad_norm": 1.4033501148223877,
+      "learning_rate": 0.0002,
+      "loss": 0.2925,
+      "step": 6410
+    },
+    {
+      "epoch": 1.3461941706856784,
+      "grad_norm": 3.9681739807128906,
+      "learning_rate": 0.0002,
+      "loss": 0.2727,
+      "step": 6420
+    },
+    {
+      "epoch": 1.348291046340952,
+      "grad_norm": 0.07361237704753876,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 6430
+    },
+    {
+      "epoch": 1.3503879219962256,
+      "grad_norm": 0.07740651816129684,
+      "learning_rate": 0.0002,
+      "loss": 0.2779,
+      "step": 6440
+    },
+    {
+      "epoch": 1.3524847976514993,
+      "grad_norm": 0.09559556841850281,
+      "learning_rate": 0.0002,
+      "loss": 0.2537,
+      "step": 6450
+    },
+    {
+      "epoch": 1.354581673306773,
+      "grad_norm": 3.6172733306884766,
+      "learning_rate": 0.0002,
+      "loss": 0.2645,
+      "step": 6460
+    },
+    {
+      "epoch": 1.3566785489620465,
+      "grad_norm": 0.06684648990631104,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 6470
+    },
+    {
+      "epoch": 1.3587754246173203,
+      "grad_norm": 0.05803783982992172,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 6480
+    },
+    {
+      "epoch": 1.3608723002725938,
+      "grad_norm": 0.07825397700071335,
+      "learning_rate": 0.0002,
+      "loss": 0.2634,
+      "step": 6490
+    },
+    {
+      "epoch": 1.3629691759278675,
+      "grad_norm": 0.13001962006092072,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 6500
+    },
+    {
+      "epoch": 1.365066051583141,
+      "grad_norm": 0.05548061430454254,
+      "learning_rate": 0.0002,
+      "loss": 0.2701,
+      "step": 6510
+    },
+    {
+      "epoch": 1.3671629272384147,
+      "grad_norm": 0.07405369728803635,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 6520
+    },
+    {
+      "epoch": 1.3692598028936884,
+      "grad_norm": 0.059938061982393265,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 6530
+    },
+    {
+      "epoch": 1.371356678548962,
+      "grad_norm": 0.09082169085741043,
+      "learning_rate": 0.0002,
+      "loss": 0.2602,
+      "step": 6540
+    },
+    {
+      "epoch": 1.3734535542042356,
+      "grad_norm": 0.06905685365200043,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 6550
+    },
+    {
+      "epoch": 1.3755504298595094,
+      "grad_norm": 1.4017658233642578,
+      "learning_rate": 0.0002,
+      "loss": 0.2773,
+      "step": 6560
+    },
+    {
+      "epoch": 1.3776473055147829,
+      "grad_norm": 0.0469370074570179,
+      "learning_rate": 0.0002,
+      "loss": 0.2582,
+      "step": 6570
+    },
+    {
+      "epoch": 1.3797441811700566,
+      "grad_norm": 3.311746835708618,
+      "learning_rate": 0.0002,
+      "loss": 0.2984,
+      "step": 6580
+    },
+    {
+      "epoch": 1.3818410568253303,
+      "grad_norm": 0.062438882887363434,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 6590
+    },
+    {
+      "epoch": 1.3839379324806038,
+      "grad_norm": 0.05996204912662506,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 6600
+    },
+    {
+      "epoch": 1.3860348081358775,
+      "grad_norm": 0.0879199281334877,
+      "learning_rate": 0.0002,
+      "loss": 0.3161,
+      "step": 6610
+    },
+    {
+      "epoch": 1.3881316837911513,
+      "grad_norm": 0.09142164140939713,
+      "learning_rate": 0.0002,
+      "loss": 0.2952,
+      "step": 6620
+    },
+    {
+      "epoch": 1.3902285594464248,
+      "grad_norm": 0.10828060656785965,
+      "learning_rate": 0.0002,
+      "loss": 0.2692,
+      "step": 6630
+    },
+    {
+      "epoch": 1.3923254351016985,
+      "grad_norm": 0.0749066174030304,
+      "learning_rate": 0.0002,
+      "loss": 0.2895,
+      "step": 6640
+    },
+    {
+      "epoch": 1.3944223107569722,
+      "grad_norm": 0.05128903314471245,
+      "learning_rate": 0.0002,
+      "loss": 0.242,
+      "step": 6650
+    },
+    {
+      "epoch": 1.3965191864122457,
+      "grad_norm": 0.07985924929380417,
+      "learning_rate": 0.0002,
+      "loss": 0.2753,
+      "step": 6660
+    },
+    {
+      "epoch": 1.3986160620675194,
+      "grad_norm": 0.10848308354616165,
+      "learning_rate": 0.0002,
+      "loss": 0.3211,
+      "step": 6670
+    },
+    {
+      "epoch": 1.4007129377227931,
+      "grad_norm": 0.07220837473869324,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 6680
+    },
+    {
+      "epoch": 1.4028098133780667,
+      "grad_norm": 8.65963363647461,
+      "learning_rate": 0.0002,
+      "loss": 0.2821,
+      "step": 6690
+    },
+    {
+      "epoch": 1.4049066890333404,
+      "grad_norm": 0.10323558002710342,
+      "learning_rate": 0.0002,
+      "loss": 0.2568,
+      "step": 6700
+    },
+    {
+      "epoch": 1.407003564688614,
+      "grad_norm": 0.08332827687263489,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 6710
+    },
+    {
+      "epoch": 1.4091004403438876,
+      "grad_norm": 0.07475430518388748,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 6720
+    },
+    {
+      "epoch": 1.4111973159991613,
+      "grad_norm": 0.09305676817893982,
+      "learning_rate": 0.0002,
+      "loss": 0.2805,
+      "step": 6730
+    },
+    {
+      "epoch": 1.4132941916544348,
+      "grad_norm": 1.4678685665130615,
+      "learning_rate": 0.0002,
+      "loss": 0.3179,
+      "step": 6740
+    },
+    {
+      "epoch": 1.4153910673097085,
+      "grad_norm": 0.07024402916431427,
+      "learning_rate": 0.0002,
+      "loss": 0.2688,
+      "step": 6750
+    },
+    {
+      "epoch": 1.417487942964982,
+      "grad_norm": 0.3857756555080414,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 6760
+    },
+    {
+      "epoch": 1.4195848186202558,
+      "grad_norm": 0.06822605431079865,
+      "learning_rate": 0.0002,
+      "loss": 0.2819,
+      "step": 6770
+    },
+    {
+      "epoch": 1.4216816942755295,
+      "grad_norm": 0.12242255359888077,
+      "learning_rate": 0.0002,
+      "loss": 0.2546,
+      "step": 6780
+    },
+    {
+      "epoch": 1.423778569930803,
+      "grad_norm": 0.6426852345466614,
+      "learning_rate": 0.0002,
+      "loss": 0.2707,
+      "step": 6790
+    },
+    {
+      "epoch": 1.4258754455860767,
+      "grad_norm": 0.084357351064682,
+      "learning_rate": 0.0002,
+      "loss": 0.2605,
+      "step": 6800
+    },
+    {
+      "epoch": 1.4279723212413504,
+      "grad_norm": 0.06686043739318848,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 6810
+    },
+    {
+      "epoch": 1.430069196896624,
+      "grad_norm": 0.12673796713352203,
+      "learning_rate": 0.0002,
+      "loss": 0.2622,
+      "step": 6820
+    },
+    {
+      "epoch": 1.4321660725518977,
+      "grad_norm": 0.11757438629865646,
+      "learning_rate": 0.0002,
+      "loss": 0.3659,
+      "step": 6830
+    },
+    {
+      "epoch": 1.4342629482071714,
+      "grad_norm": 0.05200193449854851,
+      "learning_rate": 0.0002,
+      "loss": 0.3081,
+      "step": 6840
+    },
+    {
+      "epoch": 1.4363598238624449,
+      "grad_norm": 0.5917229652404785,
+      "learning_rate": 0.0002,
+      "loss": 0.2994,
+      "step": 6850
+    },
+    {
+      "epoch": 1.4384566995177186,
+      "grad_norm": 0.10604757815599442,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 6860
+    },
+    {
+      "epoch": 1.4405535751729923,
+      "grad_norm": 0.06061761453747749,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 6870
+    },
+    {
+      "epoch": 1.4426504508282658,
+      "grad_norm": 1.4297677278518677,
+      "learning_rate": 0.0002,
+      "loss": 0.3055,
+      "step": 6880
+    },
+    {
+      "epoch": 1.4447473264835395,
+      "grad_norm": 0.5999753475189209,
+      "learning_rate": 0.0002,
+      "loss": 0.3373,
+      "step": 6890
+    },
+    {
+      "epoch": 1.4468442021388133,
+      "grad_norm": 3.725703001022339,
+      "learning_rate": 0.0002,
+      "loss": 0.2953,
+      "step": 6900
+    },
+    {
+      "epoch": 1.4489410777940868,
+      "grad_norm": 0.12710486352443695,
+      "learning_rate": 0.0002,
+      "loss": 0.3168,
+      "step": 6910
+    },
+    {
+      "epoch": 1.4510379534493605,
+      "grad_norm": 0.10325577110052109,
+      "learning_rate": 0.0002,
+      "loss": 0.2799,
+      "step": 6920
+    },
+    {
+      "epoch": 1.4531348291046342,
+      "grad_norm": 0.07545780390501022,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 6930
+    },
+    {
+      "epoch": 1.4552317047599077,
+      "grad_norm": 0.07704131305217743,
+      "learning_rate": 0.0002,
+      "loss": 0.289,
+      "step": 6940
+    },
+    {
+      "epoch": 1.4573285804151814,
+      "grad_norm": 0.06797541677951813,
+      "learning_rate": 0.0002,
+      "loss": 0.2999,
+      "step": 6950
+    },
+    {
+      "epoch": 1.4594254560704552,
+      "grad_norm": 1.3126064538955688,
+      "learning_rate": 0.0002,
+      "loss": 0.2766,
+      "step": 6960
+    },
+    {
+      "epoch": 1.4615223317257287,
+      "grad_norm": 0.0826839804649353,
+      "learning_rate": 0.0002,
+      "loss": 0.2506,
+      "step": 6970
+    },
+    {
+      "epoch": 1.4636192073810024,
+      "grad_norm": 2.723210334777832,
+      "learning_rate": 0.0002,
+      "loss": 0.3051,
+      "step": 6980
+    },
+    {
+      "epoch": 1.4657160830362759,
+      "grad_norm": 0.30324873328208923,
+      "learning_rate": 0.0002,
+      "loss": 0.2619,
+      "step": 6990
+    },
+    {
+      "epoch": 1.4678129586915496,
+      "grad_norm": 0.0709717720746994,
+      "learning_rate": 0.0002,
+      "loss": 0.2588,
+      "step": 7000
+    },
+    {
+      "epoch": 1.469909834346823,
+      "grad_norm": 0.07733996957540512,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 7010
+    },
+    {
+      "epoch": 1.4720067100020968,
+      "grad_norm": 0.33669042587280273,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 7020
+    },
+    {
+      "epoch": 1.4741035856573705,
+      "grad_norm": 2.0644760131835938,
+      "learning_rate": 0.0002,
+      "loss": 0.2853,
+      "step": 7030
+    },
+    {
+      "epoch": 1.476200461312644,
+      "grad_norm": 0.09015507996082306,
+      "learning_rate": 0.0002,
+      "loss": 0.2632,
+      "step": 7040
+    },
+    {
+      "epoch": 1.4782973369679178,
+      "grad_norm": 3.7130730152130127,
+      "learning_rate": 0.0002,
+      "loss": 0.2879,
+      "step": 7050
+    },
+    {
+      "epoch": 1.4803942126231915,
+      "grad_norm": 0.05841604992747307,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 7060
+    },
+    {
+      "epoch": 1.482491088278465,
+      "grad_norm": 0.33909642696380615,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 7070
+    },
+    {
+      "epoch": 1.4845879639337387,
+      "grad_norm": 0.09259719401597977,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 7080
+    },
+    {
+      "epoch": 1.4866848395890124,
+      "grad_norm": 0.11720839887857437,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 7090
+    },
+    {
+      "epoch": 1.488781715244286,
+      "grad_norm": 0.1009572297334671,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 7100
+    },
+    {
+      "epoch": 1.4908785908995597,
+      "grad_norm": 0.33339136838912964,
+      "learning_rate": 0.0002,
+      "loss": 0.2724,
+      "step": 7110
+    },
+    {
+      "epoch": 1.4929754665548334,
+      "grad_norm": 0.10810337960720062,
+      "learning_rate": 0.0002,
+      "loss": 0.2732,
+      "step": 7120
+    },
+    {
+      "epoch": 1.4950723422101069,
+      "grad_norm": 0.08828122913837433,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 7130
+    },
+    {
+      "epoch": 1.4971692178653806,
+      "grad_norm": 0.1396791934967041,
+      "learning_rate": 0.0002,
+      "loss": 0.2538,
+      "step": 7140
+    },
+    {
+      "epoch": 1.4992660935206543,
+      "grad_norm": 2.327631950378418,
+      "learning_rate": 0.0002,
+      "loss": 0.2808,
+      "step": 7150
+    },
+    {
+      "epoch": 1.5013629691759278,
+      "grad_norm": 0.10392621159553528,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 7160
+    },
+    {
+      "epoch": 1.5034598448312015,
+      "grad_norm": 2.2027974128723145,
+      "learning_rate": 0.0002,
+      "loss": 0.2753,
+      "step": 7170
+    },
+    {
+      "epoch": 1.5055567204864753,
+      "grad_norm": 0.060588303953409195,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 7180
+    },
+    {
+      "epoch": 1.5076535961417488,
+      "grad_norm": 0.08776184171438217,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 7190
+    },
+    {
+      "epoch": 1.5097504717970225,
+      "grad_norm": 0.10150139033794403,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 7200
+    },
+    {
+      "epoch": 1.5118473474522962,
+      "grad_norm": 0.12122219055891037,
+      "learning_rate": 0.0002,
+      "loss": 0.2416,
+      "step": 7210
+    },
+    {
+      "epoch": 1.5139442231075697,
+      "grad_norm": 0.05999595671892166,
+      "learning_rate": 0.0002,
+      "loss": 0.3042,
+      "step": 7220
+    },
+    {
+      "epoch": 1.5160410987628432,
+      "grad_norm": 0.06926919519901276,
+      "learning_rate": 0.0002,
+      "loss": 0.2628,
+      "step": 7230
+    },
+    {
+      "epoch": 1.5181379744181172,
+      "grad_norm": 3.6650075912475586,
+      "learning_rate": 0.0002,
+      "loss": 0.2884,
+      "step": 7240
+    },
+    {
+      "epoch": 1.5202348500733907,
+      "grad_norm": 0.0633549615740776,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 7250
+    },
+    {
+      "epoch": 1.5223317257286642,
+      "grad_norm": 0.07859442383050919,
+      "learning_rate": 0.0002,
+      "loss": 0.3138,
+      "step": 7260
+    },
+    {
+      "epoch": 1.524428601383938,
+      "grad_norm": 1.6131170988082886,
+      "learning_rate": 0.0002,
+      "loss": 0.2494,
+      "step": 7270
+    },
+    {
+      "epoch": 1.5265254770392116,
+      "grad_norm": 0.0931943953037262,
+      "learning_rate": 0.0002,
+      "loss": 0.2499,
+      "step": 7280
+    },
+    {
+      "epoch": 1.528622352694485,
+      "grad_norm": 0.09297845512628555,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 7290
+    },
+    {
+      "epoch": 1.5307192283497588,
+      "grad_norm": 0.06294408440589905,
+      "learning_rate": 0.0002,
+      "loss": 0.2568,
+      "step": 7300
+    },
+    {
+      "epoch": 1.5328161040050325,
+      "grad_norm": 0.08879171311855316,
+      "learning_rate": 0.0002,
+      "loss": 0.2588,
+      "step": 7310
+    },
+    {
+      "epoch": 1.534912979660306,
+      "grad_norm": 0.05376000702381134,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 7320
+    },
+    {
+      "epoch": 1.5370098553155798,
+      "grad_norm": 3.216952323913574,
+      "learning_rate": 0.0002,
+      "loss": 0.3338,
+      "step": 7330
+    },
+    {
+      "epoch": 1.5391067309708535,
+      "grad_norm": 0.07937444746494293,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 7340
+    },
+    {
+      "epoch": 1.541203606626127,
+      "grad_norm": 1.347196340560913,
+      "learning_rate": 0.0002,
+      "loss": 0.3316,
+      "step": 7350
+    },
+    {
+      "epoch": 1.5433004822814007,
+      "grad_norm": 0.0543079674243927,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 7360
+    },
+    {
+      "epoch": 1.5453973579366744,
+      "grad_norm": 0.13290783762931824,
+      "learning_rate": 0.0002,
+      "loss": 0.2417,
+      "step": 7370
+    },
+    {
+      "epoch": 1.547494233591948,
+      "grad_norm": 2.971339225769043,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 7380
+    },
+    {
+      "epoch": 1.5495911092472217,
+      "grad_norm": 0.08533356338739395,
+      "learning_rate": 0.0002,
+      "loss": 0.2965,
+      "step": 7390
+    },
+    {
+      "epoch": 1.5516879849024954,
+      "grad_norm": 0.06842143833637238,
+      "learning_rate": 0.0002,
+      "loss": 0.2577,
+      "step": 7400
+    },
+    {
+      "epoch": 1.5537848605577689,
+      "grad_norm": 0.08902810513973236,
+      "learning_rate": 0.0002,
+      "loss": 0.2698,
+      "step": 7410
+    },
+    {
+      "epoch": 1.5558817362130426,
+      "grad_norm": 0.07102424651384354,
+      "learning_rate": 0.0002,
+      "loss": 0.319,
+      "step": 7420
+    },
+    {
+      "epoch": 1.5579786118683163,
+      "grad_norm": 0.06909465044736862,
+      "learning_rate": 0.0002,
+      "loss": 0.2628,
+      "step": 7430
+    },
+    {
+      "epoch": 1.5600754875235898,
+      "grad_norm": 0.06648514419794083,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 7440
+    },
+    {
+      "epoch": 1.5621723631788635,
+      "grad_norm": 0.09439200162887573,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 7450
+    },
+    {
+      "epoch": 1.5642692388341373,
+      "grad_norm": 0.09378870576620102,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 7460
+    },
+    {
+      "epoch": 1.5663661144894108,
+      "grad_norm": 0.07520859688520432,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 7470
+    },
+    {
+      "epoch": 1.5684629901446843,
+      "grad_norm": 0.10274530947208405,
+      "learning_rate": 0.0002,
+      "loss": 0.2897,
+      "step": 7480
+    },
+    {
+      "epoch": 1.5705598657999582,
+      "grad_norm": 0.07290078699588776,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 7490
+    },
+    {
+      "epoch": 1.5726567414552317,
+      "grad_norm": 0.057594846934080124,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 7500
+    },
+    {
+      "epoch": 1.5747536171105052,
+      "grad_norm": 0.080964095890522,
+      "learning_rate": 0.0002,
+      "loss": 0.2876,
+      "step": 7510
+    },
+    {
+      "epoch": 1.5768504927657792,
+      "grad_norm": 0.08604875952005386,
+      "learning_rate": 0.0002,
+      "loss": 0.2595,
+      "step": 7520
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.06626357138156891,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 7530
+    },
+    {
+      "epoch": 1.5810442440763262,
+      "grad_norm": 0.06062173843383789,
+      "learning_rate": 0.0002,
+      "loss": 0.3133,
+      "step": 7540
+    },
+    {
+      "epoch": 1.5831411197315999,
+      "grad_norm": 0.0724383071064949,
+      "learning_rate": 0.0002,
+      "loss": 0.2705,
+      "step": 7550
+    },
+    {
+      "epoch": 1.5852379953868736,
+      "grad_norm": 0.09347286075353622,
+      "learning_rate": 0.0002,
+      "loss": 0.2544,
+      "step": 7560
+    },
+    {
+      "epoch": 1.587334871042147,
+      "grad_norm": 0.0953749343752861,
+      "learning_rate": 0.0002,
+      "loss": 0.2632,
+      "step": 7570
+    },
+    {
+      "epoch": 1.5894317466974208,
+      "grad_norm": 0.10970430821180344,
+      "learning_rate": 0.0002,
+      "loss": 0.2791,
+      "step": 7580
+    },
+    {
+      "epoch": 1.5915286223526945,
+      "grad_norm": 2.3808388710021973,
+      "learning_rate": 0.0002,
+      "loss": 0.2915,
+      "step": 7590
+    },
+    {
+      "epoch": 1.593625498007968,
+      "grad_norm": 0.05185944214463234,
+      "learning_rate": 0.0002,
+      "loss": 0.2729,
+      "step": 7600
+    },
+    {
+      "epoch": 1.5957223736632418,
+      "grad_norm": 0.07438002526760101,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 7610
+    },
+    {
+      "epoch": 1.5978192493185155,
+      "grad_norm": 0.08762764185667038,
+      "learning_rate": 0.0002,
+      "loss": 0.2747,
+      "step": 7620
+    },
+    {
+      "epoch": 1.599916124973789,
+      "grad_norm": 0.6587502956390381,
+      "learning_rate": 0.0002,
+      "loss": 0.2635,
+      "step": 7630
+    },
+    {
+      "epoch": 1.6020130006290627,
+      "grad_norm": 0.05637750402092934,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 7640
+    },
+    {
+      "epoch": 1.6041098762843364,
+      "grad_norm": 0.05925126373767853,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 7650
+    },
+    {
+      "epoch": 1.60620675193961,
+      "grad_norm": 2.566734552383423,
+      "learning_rate": 0.0002,
+      "loss": 0.2935,
+      "step": 7660
+    },
+    {
+      "epoch": 1.6083036275948837,
+      "grad_norm": 0.06364738196134567,
+      "learning_rate": 0.0002,
+      "loss": 0.32,
+      "step": 7670
+    },
+    {
+      "epoch": 1.6104005032501574,
+      "grad_norm": 0.11035982519388199,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 7680
+    },
+    {
+      "epoch": 1.6124973789054309,
+      "grad_norm": 0.050208210945129395,
+      "learning_rate": 0.0002,
+      "loss": 0.2894,
+      "step": 7690
+    },
+    {
+      "epoch": 1.6145942545607046,
+      "grad_norm": 0.1036805585026741,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 7700
+    },
+    {
+      "epoch": 1.6166911302159783,
+      "grad_norm": 0.07067201286554337,
+      "learning_rate": 0.0002,
+      "loss": 0.2688,
+      "step": 7710
+    },
+    {
+      "epoch": 1.6187880058712518,
+      "grad_norm": 0.0834813043475151,
+      "learning_rate": 0.0002,
+      "loss": 0.2896,
+      "step": 7720
+    },
+    {
+      "epoch": 1.6208848815265253,
+      "grad_norm": 0.13986825942993164,
+      "learning_rate": 0.0002,
+      "loss": 0.2951,
+      "step": 7730
+    },
+    {
+      "epoch": 1.6229817571817993,
+      "grad_norm": 3.069458484649658,
+      "learning_rate": 0.0002,
+      "loss": 0.3001,
+      "step": 7740
+    },
+    {
+      "epoch": 1.6250786328370728,
+      "grad_norm": 0.07894471287727356,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 7750
+    },
+    {
+      "epoch": 1.6271755084923463,
+      "grad_norm": 0.0820389837026596,
+      "learning_rate": 0.0002,
+      "loss": 0.2813,
+      "step": 7760
+    },
+    {
+      "epoch": 1.6292723841476202,
+      "grad_norm": 0.05259540304541588,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 7770
+    },
+    {
+      "epoch": 1.6313692598028937,
+      "grad_norm": 0.07753746956586838,
+      "learning_rate": 0.0002,
+      "loss": 0.242,
+      "step": 7780
+    },
+    {
+      "epoch": 1.6334661354581672,
+      "grad_norm": 0.0717770904302597,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 7790
+    },
+    {
+      "epoch": 1.635563011113441,
+      "grad_norm": 0.05608302354812622,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 7800
+    },
+    {
+      "epoch": 1.6376598867687147,
+      "grad_norm": 4.554051399230957,
+      "learning_rate": 0.0002,
+      "loss": 0.2656,
+      "step": 7810
+    },
+    {
+      "epoch": 1.6397567624239882,
+      "grad_norm": 0.07997504621744156,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 7820
+    },
+    {
+      "epoch": 1.6418536380792619,
+      "grad_norm": 0.08014536648988724,
+      "learning_rate": 0.0002,
+      "loss": 0.2685,
+      "step": 7830
+    },
+    {
+      "epoch": 1.6439505137345356,
+      "grad_norm": 0.09674616903066635,
+      "learning_rate": 0.0002,
+      "loss": 0.2594,
+      "step": 7840
+    },
+    {
+      "epoch": 1.646047389389809,
+      "grad_norm": 0.9341862797737122,
+      "learning_rate": 0.0002,
+      "loss": 0.3198,
+      "step": 7850
+    },
+    {
+      "epoch": 1.6481442650450828,
+      "grad_norm": 0.09305543452501297,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 7860
+    },
+    {
+      "epoch": 1.6502411407003565,
+      "grad_norm": 0.07002803683280945,
+      "learning_rate": 0.0002,
+      "loss": 0.2493,
+      "step": 7870
+    },
+    {
+      "epoch": 1.65233801635563,
+      "grad_norm": 1.9571644067764282,
+      "learning_rate": 0.0002,
+      "loss": 0.2759,
+      "step": 7880
+    },
+    {
+      "epoch": 1.6544348920109038,
+      "grad_norm": 0.15317511558532715,
+      "learning_rate": 0.0002,
+      "loss": 0.2539,
+      "step": 7890
+    },
+    {
+      "epoch": 1.6565317676661775,
+      "grad_norm": 0.09483631700277328,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 7900
+    },
+    {
+      "epoch": 1.658628643321451,
+      "grad_norm": 0.1098591759800911,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 7910
+    },
+    {
+      "epoch": 1.6607255189767247,
+      "grad_norm": 0.05958917737007141,
+      "learning_rate": 0.0002,
+      "loss": 0.2752,
+      "step": 7920
+    },
+    {
+      "epoch": 1.6628223946319984,
+      "grad_norm": 0.07363269478082657,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 7930
+    },
+    {
+      "epoch": 1.664919270287272,
+      "grad_norm": 0.06663299351930618,
+      "learning_rate": 0.0002,
+      "loss": 0.2682,
+      "step": 7940
+    },
+    {
+      "epoch": 1.6670161459425457,
+      "grad_norm": 0.06834384053945541,
+      "learning_rate": 0.0002,
+      "loss": 0.2699,
+      "step": 7950
+    },
+    {
+      "epoch": 1.6691130215978194,
+      "grad_norm": 0.06692808866500854,
+      "learning_rate": 0.0002,
+      "loss": 0.2617,
+      "step": 7960
+    },
+    {
+      "epoch": 1.6712098972530929,
+      "grad_norm": 2.3505663871765137,
+      "learning_rate": 0.0002,
+      "loss": 0.271,
+      "step": 7970
+    },
+    {
+      "epoch": 1.6733067729083664,
+      "grad_norm": 0.08932004123926163,
+      "learning_rate": 0.0002,
+      "loss": 0.2759,
+      "step": 7980
+    },
+    {
+      "epoch": 1.6754036485636403,
+      "grad_norm": 0.058079443871974945,
+      "learning_rate": 0.0002,
+      "loss": 0.2748,
+      "step": 7990
+    },
+    {
+      "epoch": 1.6775005242189138,
+      "grad_norm": 0.08751317113637924,
+      "learning_rate": 0.0002,
+      "loss": 0.2944,
+      "step": 8000
+    },
+    {
+      "epoch": 1.6795973998741873,
+      "grad_norm": 2.1509130001068115,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 8010
+    },
+    {
+      "epoch": 1.6816942755294613,
+      "grad_norm": 0.045726533979177475,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 8020
+    },
+    {
+      "epoch": 1.6837911511847348,
+      "grad_norm": 0.1127922460436821,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 8030
+    },
+    {
+      "epoch": 1.6858880268400083,
+      "grad_norm": 8.662659645080566,
+      "learning_rate": 0.0002,
+      "loss": 0.2642,
+      "step": 8040
+    },
+    {
+      "epoch": 1.687984902495282,
+      "grad_norm": 0.6873268485069275,
+      "learning_rate": 0.0002,
+      "loss": 0.3018,
+      "step": 8050
+    },
+    {
+      "epoch": 1.6900817781505557,
+      "grad_norm": 0.07165851444005966,
+      "learning_rate": 0.0002,
+      "loss": 0.2429,
+      "step": 8060
+    },
+    {
+      "epoch": 1.6921786538058292,
+      "grad_norm": 0.1486668735742569,
+      "learning_rate": 0.0002,
+      "loss": 0.3148,
+      "step": 8070
+    },
+    {
+      "epoch": 1.694275529461103,
+      "grad_norm": 1.325584053993225,
+      "learning_rate": 0.0002,
+      "loss": 0.2655,
+      "step": 8080
+    },
+    {
+      "epoch": 1.6963724051163767,
+      "grad_norm": 2.7720537185668945,
+      "learning_rate": 0.0002,
+      "loss": 0.2934,
+      "step": 8090
+    },
+    {
+      "epoch": 1.6984692807716502,
+      "grad_norm": 0.21440936625003815,
+      "learning_rate": 0.0002,
+      "loss": 0.2749,
+      "step": 8100
+    },
+    {
+      "epoch": 1.7005661564269239,
+      "grad_norm": 0.9134746789932251,
+      "learning_rate": 0.0002,
+      "loss": 0.2626,
+      "step": 8110
+    },
+    {
+      "epoch": 1.7026630320821976,
+      "grad_norm": 0.07481775432825089,
+      "learning_rate": 0.0002,
+      "loss": 0.249,
+      "step": 8120
+    },
+    {
+      "epoch": 1.704759907737471,
+      "grad_norm": 0.45117613673210144,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 8130
+    },
+    {
+      "epoch": 1.7068567833927448,
+      "grad_norm": 2.4827306270599365,
+      "learning_rate": 0.0002,
+      "loss": 0.288,
+      "step": 8140
+    },
+    {
+      "epoch": 1.7089536590480185,
+      "grad_norm": 0.1967839002609253,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 8150
+    },
+    {
+      "epoch": 1.711050534703292,
+      "grad_norm": 1.2610207796096802,
+      "learning_rate": 0.0002,
+      "loss": 0.2692,
+      "step": 8160
+    },
+    {
+      "epoch": 1.7131474103585658,
+      "grad_norm": 0.09038415551185608,
+      "learning_rate": 0.0002,
+      "loss": 0.2565,
+      "step": 8170
+    },
+    {
+      "epoch": 1.7152442860138395,
+      "grad_norm": 0.524497389793396,
+      "learning_rate": 0.0002,
+      "loss": 0.2742,
+      "step": 8180
+    },
+    {
+      "epoch": 1.717341161669113,
+      "grad_norm": 0.12098541110754013,
+      "learning_rate": 0.0002,
+      "loss": 0.2724,
+      "step": 8190
+    },
+    {
+      "epoch": 1.7194380373243867,
+      "grad_norm": 0.109773188829422,
+      "learning_rate": 0.0002,
+      "loss": 0.254,
+      "step": 8200
+    },
+    {
+      "epoch": 1.7215349129796604,
+      "grad_norm": 0.2105139046907425,
+      "learning_rate": 0.0002,
+      "loss": 0.3159,
+      "step": 8210
+    },
+    {
+      "epoch": 1.723631788634934,
+      "grad_norm": 0.05839413404464722,
+      "learning_rate": 0.0002,
+      "loss": 0.263,
+      "step": 8220
+    },
+    {
+      "epoch": 1.7257286642902074,
+      "grad_norm": 1.7193588018417358,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 8230
+    },
+    {
+      "epoch": 1.7278255399454814,
+      "grad_norm": 0.11846201121807098,
+      "learning_rate": 0.0002,
+      "loss": 0.2539,
+      "step": 8240
+    },
+    {
+      "epoch": 1.7299224156007549,
+      "grad_norm": 0.10171257704496384,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 8250
+    },
+    {
+      "epoch": 1.7320192912560284,
+      "grad_norm": 2.8296456336975098,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 8260
+    },
+    {
+      "epoch": 1.7341161669113023,
+      "grad_norm": 0.08424999564886093,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 8270
+    },
+    {
+      "epoch": 1.7362130425665758,
+      "grad_norm": 0.08350451290607452,
+      "learning_rate": 0.0002,
+      "loss": 0.2823,
+      "step": 8280
+    },
+    {
+      "epoch": 1.7383099182218493,
+      "grad_norm": 0.1168193444609642,
+      "learning_rate": 0.0002,
+      "loss": 0.2566,
+      "step": 8290
+    },
+    {
+      "epoch": 1.740406793877123,
+      "grad_norm": 0.09300057590007782,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 8300
+    },
+    {
+      "epoch": 1.7425036695323968,
+      "grad_norm": 0.05919062718749046,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 8310
+    },
+    {
+      "epoch": 1.7446005451876703,
+      "grad_norm": 6.351716995239258,
+      "learning_rate": 0.0002,
+      "loss": 0.3258,
+      "step": 8320
+    },
+    {
+      "epoch": 1.746697420842944,
+      "grad_norm": 0.08938953280448914,
+      "learning_rate": 0.0002,
+      "loss": 0.2976,
+      "step": 8330
+    },
+    {
+      "epoch": 1.7487942964982177,
+      "grad_norm": 0.17226704955101013,
+      "learning_rate": 0.0002,
+      "loss": 0.2789,
+      "step": 8340
+    },
+    {
+      "epoch": 1.7508911721534912,
+      "grad_norm": 0.13070009648799896,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 8350
+    },
+    {
+      "epoch": 1.752988047808765,
+      "grad_norm": 1.1679989099502563,
+      "learning_rate": 0.0002,
+      "loss": 0.3043,
+      "step": 8360
+    },
+    {
+      "epoch": 1.7550849234640387,
+      "grad_norm": 0.09552992135286331,
+      "learning_rate": 0.0002,
+      "loss": 0.3402,
+      "step": 8370
+    },
+    {
+      "epoch": 1.7571817991193122,
+      "grad_norm": 0.18032927811145782,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 8380
+    },
+    {
+      "epoch": 1.7592786747745859,
+      "grad_norm": 4.900735378265381,
+      "learning_rate": 0.0002,
+      "loss": 0.5324,
+      "step": 8390
+    },
+    {
+      "epoch": 1.7613755504298596,
+      "grad_norm": 2.462291955947876,
+      "learning_rate": 0.0002,
+      "loss": 0.2466,
+      "step": 8400
+    },
+    {
+      "epoch": 1.763472426085133,
+      "grad_norm": 0.7509610056877136,
+      "learning_rate": 0.0002,
+      "loss": 0.2514,
+      "step": 8410
+    },
+    {
+      "epoch": 1.7655693017404068,
+      "grad_norm": 0.07756595313549042,
+      "learning_rate": 0.0002,
+      "loss": 0.2746,
+      "step": 8420
+    },
+    {
+      "epoch": 1.7676661773956805,
+      "grad_norm": 0.08571748435497284,
+      "learning_rate": 0.0002,
+      "loss": 0.2912,
+      "step": 8430
+    },
+    {
+      "epoch": 1.769763053050954,
+      "grad_norm": 0.07590015232563019,
+      "learning_rate": 0.0002,
+      "loss": 0.2677,
+      "step": 8440
+    },
+    {
+      "epoch": 1.7718599287062278,
+      "grad_norm": 0.11724907904863358,
+      "learning_rate": 0.0002,
+      "loss": 0.2674,
+      "step": 8450
+    },
+    {
+      "epoch": 1.7739568043615015,
+      "grad_norm": 2.972846269607544,
+      "learning_rate": 0.0002,
+      "loss": 0.3069,
+      "step": 8460
+    },
+    {
+      "epoch": 1.776053680016775,
+      "grad_norm": 0.057090677320957184,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 8470
+    },
+    {
+      "epoch": 1.7781505556720485,
+      "grad_norm": 0.09555861353874207,
+      "learning_rate": 0.0002,
+      "loss": 0.295,
+      "step": 8480
+    },
+    {
+      "epoch": 1.7802474313273224,
+      "grad_norm": 0.08062522113323212,
+      "learning_rate": 0.0002,
+      "loss": 0.2774,
+      "step": 8490
+    },
+    {
+      "epoch": 1.782344306982596,
+      "grad_norm": 0.06056349724531174,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 8500
+    },
+    {
+      "epoch": 1.7844411826378694,
+      "grad_norm": 2.523094415664673,
+      "learning_rate": 0.0002,
+      "loss": 0.3411,
+      "step": 8510
+    },
+    {
+      "epoch": 1.7865380582931434,
+      "grad_norm": 0.061809323728084564,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 8520
+    },
+    {
+      "epoch": 1.7886349339484169,
+      "grad_norm": 0.17487798631191254,
+      "learning_rate": 0.0002,
+      "loss": 0.2666,
+      "step": 8530
+    },
+    {
+      "epoch": 1.7907318096036904,
+      "grad_norm": 0.05645996332168579,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 8540
+    },
+    {
+      "epoch": 1.792828685258964,
+      "grad_norm": 0.0833887979388237,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 8550
+    },
+    {
+      "epoch": 1.7949255609142378,
+      "grad_norm": 0.08759808540344238,
+      "learning_rate": 0.0002,
+      "loss": 0.2932,
+      "step": 8560
+    },
+    {
+      "epoch": 1.7970224365695113,
+      "grad_norm": 0.07473572343587875,
+      "learning_rate": 0.0002,
+      "loss": 0.2739,
+      "step": 8570
+    },
+    {
+      "epoch": 1.799119312224785,
+      "grad_norm": 0.07650966942310333,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 8580
+    },
+    {
+      "epoch": 1.8012161878800588,
+      "grad_norm": 0.06564634293317795,
+      "learning_rate": 0.0002,
+      "loss": 0.2775,
+      "step": 8590
+    },
+    {
+      "epoch": 1.8033130635353323,
+      "grad_norm": 0.12063200771808624,
+      "learning_rate": 0.0002,
+      "loss": 0.2608,
+      "step": 8600
+    },
+    {
+      "epoch": 1.805409939190606,
+      "grad_norm": 0.08069991320371628,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 8610
+    },
+    {
+      "epoch": 1.8075068148458797,
+      "grad_norm": 0.07421161234378815,
+      "learning_rate": 0.0002,
+      "loss": 0.2749,
+      "step": 8620
+    },
+    {
+      "epoch": 1.8096036905011532,
+      "grad_norm": 0.0516473613679409,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 8630
+    },
+    {
+      "epoch": 1.811700566156427,
+      "grad_norm": 0.07442668080329895,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 8640
+    },
+    {
+      "epoch": 1.8137974418117007,
+      "grad_norm": 0.07932697981595993,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 8650
+    },
+    {
+      "epoch": 1.8158943174669742,
+      "grad_norm": 0.09281004965305328,
+      "learning_rate": 0.0002,
+      "loss": 0.2556,
+      "step": 8660
+    },
+    {
+      "epoch": 1.8179911931222479,
+      "grad_norm": 0.06372177600860596,
+      "learning_rate": 0.0002,
+      "loss": 0.3174,
+      "step": 8670
+    },
+    {
+      "epoch": 1.8200880687775216,
+      "grad_norm": 0.07562694698572159,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 8680
+    },
+    {
+      "epoch": 1.822184944432795,
+      "grad_norm": 0.07504847645759583,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 8690
+    },
+    {
+      "epoch": 1.8242818200880688,
+      "grad_norm": 0.10201393067836761,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 8700
+    },
+    {
+      "epoch": 1.8263786957433426,
+      "grad_norm": 0.07061925530433655,
+      "learning_rate": 0.0002,
+      "loss": 0.2587,
+      "step": 8710
+    },
+    {
+      "epoch": 1.828475571398616,
+      "grad_norm": 2.701824188232422,
+      "learning_rate": 0.0002,
+      "loss": 0.2845,
+      "step": 8720
+    },
+    {
+      "epoch": 1.8305724470538896,
+      "grad_norm": 0.20525291562080383,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 8730
+    },
+    {
+      "epoch": 1.8326693227091635,
+      "grad_norm": 0.05398187041282654,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 8740
+    },
+    {
+      "epoch": 1.834766198364437,
+      "grad_norm": 0.06944382935762405,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 8750
+    },
+    {
+      "epoch": 1.8368630740197105,
+      "grad_norm": 5.459568023681641,
+      "learning_rate": 0.0002,
+      "loss": 0.2816,
+      "step": 8760
+    },
+    {
+      "epoch": 1.8389599496749844,
+      "grad_norm": 0.06237971410155296,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 8770
+    },
+    {
+      "epoch": 1.841056825330258,
+      "grad_norm": 0.06791366636753082,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 8780
+    },
+    {
+      "epoch": 1.8431537009855314,
+      "grad_norm": 0.06800734996795654,
+      "learning_rate": 0.0002,
+      "loss": 0.2681,
+      "step": 8790
+    },
+    {
+      "epoch": 1.8452505766408052,
+      "grad_norm": 4.149272918701172,
+      "learning_rate": 0.0002,
+      "loss": 0.2848,
+      "step": 8800
+    },
+    {
+      "epoch": 1.8473474522960789,
+      "grad_norm": 0.06702340394258499,
+      "learning_rate": 0.0002,
+      "loss": 0.2555,
+      "step": 8810
+    },
+    {
+      "epoch": 1.8494443279513524,
+      "grad_norm": 0.048510730266571045,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 8820
+    },
+    {
+      "epoch": 1.851541203606626,
+      "grad_norm": 0.055737774819135666,
+      "learning_rate": 0.0002,
+      "loss": 0.2624,
+      "step": 8830
+    },
+    {
+      "epoch": 1.8536380792618998,
+      "grad_norm": 0.09368737787008286,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 8840
+    },
+    {
+      "epoch": 1.8557349549171733,
+      "grad_norm": 0.06682492047548294,
+      "learning_rate": 0.0002,
+      "loss": 0.3114,
+      "step": 8850
+    },
+    {
+      "epoch": 1.857831830572447,
+      "grad_norm": 0.1425323784351349,
+      "learning_rate": 0.0002,
+      "loss": 0.2614,
+      "step": 8860
+    },
+    {
+      "epoch": 1.8599287062277208,
+      "grad_norm": 0.08993595838546753,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 8870
+    },
+    {
+      "epoch": 1.8620255818829943,
+      "grad_norm": 0.06886916607618332,
+      "learning_rate": 0.0002,
+      "loss": 0.2788,
+      "step": 8880
+    },
+    {
+      "epoch": 1.864122457538268,
+      "grad_norm": 0.21538451313972473,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 8890
+    },
+    {
+      "epoch": 1.8662193331935417,
+      "grad_norm": 0.06627918034791946,
+      "learning_rate": 0.0002,
+      "loss": 0.2864,
+      "step": 8900
+    },
+    {
+      "epoch": 1.8683162088488152,
+      "grad_norm": 0.06920845806598663,
+      "learning_rate": 0.0002,
+      "loss": 0.26,
+      "step": 8910
+    },
+    {
+      "epoch": 1.870413084504089,
+      "grad_norm": 0.05559088662266731,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 8920
+    },
+    {
+      "epoch": 1.8725099601593627,
+      "grad_norm": 1.7723051309585571,
+      "learning_rate": 0.0002,
+      "loss": 0.2817,
+      "step": 8930
+    },
+    {
+      "epoch": 1.8746068358146362,
+      "grad_norm": 0.09814611822366714,
+      "learning_rate": 0.0002,
+      "loss": 0.2847,
+      "step": 8940
+    },
+    {
+      "epoch": 1.8767037114699099,
+      "grad_norm": 0.06993977725505829,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 8950
+    },
+    {
+      "epoch": 1.8788005871251836,
+      "grad_norm": 4.195693016052246,
+      "learning_rate": 0.0002,
+      "loss": 0.3096,
+      "step": 8960
+    },
+    {
+      "epoch": 1.880897462780457,
+      "grad_norm": 0.06579772382974625,
+      "learning_rate": 0.0002,
+      "loss": 0.2924,
+      "step": 8970
+    },
+    {
+      "epoch": 1.8829943384357306,
+      "grad_norm": 0.0795520544052124,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 8980
+    },
+    {
+      "epoch": 1.8850912140910046,
+      "grad_norm": 0.09593985229730606,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 8990
+    },
+    {
+      "epoch": 1.887188089746278,
+      "grad_norm": 0.06778479367494583,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 9000
+    },
+    {
+      "epoch": 1.8892849654015516,
+      "grad_norm": 0.06771530956029892,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 9010
+    },
+    {
+      "epoch": 1.8913818410568253,
+      "grad_norm": 0.0755813866853714,
+      "learning_rate": 0.0002,
+      "loss": 0.2909,
+      "step": 9020
+    },
+    {
+      "epoch": 1.893478716712099,
+      "grad_norm": 0.07563014328479767,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 9030
+    },
+    {
+      "epoch": 1.8955755923673725,
+      "grad_norm": 0.08678235858678818,
+      "learning_rate": 0.0002,
+      "loss": 0.2764,
+      "step": 9040
+    },
+    {
+      "epoch": 1.8976724680226462,
+      "grad_norm": 0.14399957656860352,
+      "learning_rate": 0.0002,
+      "loss": 0.2894,
+      "step": 9050
+    },
+    {
+      "epoch": 1.89976934367792,
+      "grad_norm": 0.07211213558912277,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 9060
+    },
+    {
+      "epoch": 1.9018662193331934,
+      "grad_norm": 0.0553593710064888,
+      "learning_rate": 0.0002,
+      "loss": 0.2857,
+      "step": 9070
+    },
+    {
+      "epoch": 1.9039630949884672,
+      "grad_norm": 0.057425178587436676,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 9080
+    },
+    {
+      "epoch": 1.9060599706437409,
+      "grad_norm": 0.06174889951944351,
+      "learning_rate": 0.0002,
+      "loss": 0.2811,
+      "step": 9090
+    },
+    {
+      "epoch": 1.9081568462990144,
+      "grad_norm": 0.045786961913108826,
+      "learning_rate": 0.0002,
+      "loss": 0.257,
+      "step": 9100
+    },
+    {
+      "epoch": 1.910253721954288,
+      "grad_norm": 0.09140277653932571,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 9110
+    },
+    {
+      "epoch": 1.9123505976095618,
+      "grad_norm": 0.05962507426738739,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 9120
+    },
+    {
+      "epoch": 1.9144474732648353,
+      "grad_norm": 0.07254903763532639,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 9130
+    },
+    {
+      "epoch": 1.916544348920109,
+      "grad_norm": 0.0659898966550827,
+      "learning_rate": 0.0002,
+      "loss": 0.2649,
+      "step": 9140
+    },
+    {
+      "epoch": 1.9186412245753828,
+      "grad_norm": 2.222851514816284,
+      "learning_rate": 0.0002,
+      "loss": 0.2714,
+      "step": 9150
+    },
+    {
+      "epoch": 1.9207381002306563,
+      "grad_norm": 2.2575175762176514,
+      "learning_rate": 0.0002,
+      "loss": 0.2615,
+      "step": 9160
+    },
+    {
+      "epoch": 1.92283497588593,
+      "grad_norm": 0.1300947368144989,
+      "learning_rate": 0.0002,
+      "loss": 0.3135,
+      "step": 9170
+    },
+    {
+      "epoch": 1.9249318515412037,
+      "grad_norm": 2.992399215698242,
+      "learning_rate": 0.0002,
+      "loss": 0.2685,
+      "step": 9180
+    },
+    {
+      "epoch": 1.9270287271964772,
+      "grad_norm": 4.321889400482178,
+      "learning_rate": 0.0002,
+      "loss": 0.2793,
+      "step": 9190
+    },
+    {
+      "epoch": 1.929125602851751,
+      "grad_norm": 0.09589577466249466,
+      "learning_rate": 0.0002,
+      "loss": 0.2665,
+      "step": 9200
+    },
+    {
+      "epoch": 1.9312224785070247,
+      "grad_norm": 0.06917759776115417,
+      "learning_rate": 0.0002,
+      "loss": 0.2601,
+      "step": 9210
+    },
+    {
+      "epoch": 1.9333193541622982,
+      "grad_norm": 0.06501258164644241,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 9220
+    },
+    {
+      "epoch": 1.9354162298175717,
+      "grad_norm": 3.5356762409210205,
+      "learning_rate": 0.0002,
+      "loss": 0.2917,
+      "step": 9230
+    },
+    {
+      "epoch": 1.9375131054728456,
+      "grad_norm": 0.08394140750169754,
+      "learning_rate": 0.0002,
+      "loss": 0.2707,
+      "step": 9240
+    },
+    {
+      "epoch": 1.939609981128119,
+      "grad_norm": 0.052391644567251205,
+      "learning_rate": 0.0002,
+      "loss": 0.2675,
+      "step": 9250
+    },
+    {
+      "epoch": 1.9417068567833926,
+      "grad_norm": 0.09241612255573273,
+      "learning_rate": 0.0002,
+      "loss": 0.2575,
+      "step": 9260
+    },
+    {
+      "epoch": 1.9438037324386663,
+      "grad_norm": 0.0741746574640274,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 9270
+    },
+    {
+      "epoch": 1.94590060809394,
+      "grad_norm": 0.08500709384679794,
+      "learning_rate": 0.0002,
+      "loss": 0.267,
+      "step": 9280
+    },
+    {
+      "epoch": 1.9479974837492136,
+      "grad_norm": 0.10579940676689148,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 9290
+    },
+    {
+      "epoch": 1.9500943594044873,
+      "grad_norm": 0.07684406638145447,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 9300
+    },
+    {
+      "epoch": 1.952191235059761,
+      "grad_norm": 0.06794850528240204,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 9310
+    },
+    {
+      "epoch": 1.9542881107150345,
+      "grad_norm": 0.08876547962427139,
+      "learning_rate": 0.0002,
+      "loss": 0.243,
+      "step": 9320
+    },
+    {
+      "epoch": 1.9563849863703082,
+      "grad_norm": 0.051869798451662064,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 9330
+    },
+    {
+      "epoch": 1.958481862025582,
+      "grad_norm": 0.10372921824455261,
+      "learning_rate": 0.0002,
+      "loss": 0.2917,
+      "step": 9340
+    },
+    {
+      "epoch": 1.9605787376808554,
+      "grad_norm": 0.0588776133954525,
+      "learning_rate": 0.0002,
+      "loss": 0.2619,
+      "step": 9350
+    },
+    {
+      "epoch": 1.9626756133361292,
+      "grad_norm": 0.08060908317565918,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 9360
+    },
+    {
+      "epoch": 1.9647724889914029,
+      "grad_norm": 0.06950705498456955,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 9370
+    },
+    {
+      "epoch": 1.9668693646466764,
+      "grad_norm": 0.07383579760789871,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 9380
+    },
+    {
+      "epoch": 1.96896624030195,
+      "grad_norm": 0.06754910200834274,
+      "learning_rate": 0.0002,
+      "loss": 0.2596,
+      "step": 9390
+    },
+    {
+      "epoch": 1.9710631159572238,
+      "grad_norm": 0.0853751078248024,
+      "learning_rate": 0.0002,
+      "loss": 0.309,
+      "step": 9400
+    },
+    {
+      "epoch": 1.9731599916124973,
+      "grad_norm": 0.0764593631029129,
+      "learning_rate": 0.0002,
+      "loss": 0.2483,
+      "step": 9410
+    },
+    {
+      "epoch": 1.975256867267771,
+      "grad_norm": 5.169427871704102,
+      "learning_rate": 0.0002,
+      "loss": 0.3071,
+      "step": 9420
+    },
+    {
+      "epoch": 1.9773537429230448,
+      "grad_norm": 0.19787247478961945,
+      "learning_rate": 0.0002,
+      "loss": 0.3134,
+      "step": 9430
+    },
+    {
+      "epoch": 1.9794506185783183,
+      "grad_norm": 0.10594961792230606,
+      "learning_rate": 0.0002,
+      "loss": 0.3063,
+      "step": 9440
+    },
+    {
+      "epoch": 1.981547494233592,
+      "grad_norm": 0.060998719185590744,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 9450
+    },
+    {
+      "epoch": 1.9836443698888657,
+      "grad_norm": 0.10632716119289398,
+      "learning_rate": 0.0002,
+      "loss": 0.2706,
+      "step": 9460
+    },
+    {
+      "epoch": 1.9857412455441392,
+      "grad_norm": 0.09421350806951523,
+      "learning_rate": 0.0002,
+      "loss": 0.271,
+      "step": 9470
+    },
+    {
+      "epoch": 1.9878381211994127,
+      "grad_norm": 0.12134144455194473,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 9480
+    },
+    {
+      "epoch": 1.9899349968546867,
+      "grad_norm": 0.05131775140762329,
+      "learning_rate": 0.0002,
+      "loss": 0.2712,
+      "step": 9490
+    },
+    {
+      "epoch": 1.9920318725099602,
+      "grad_norm": 0.08443418890237808,
+      "learning_rate": 0.0002,
+      "loss": 0.2905,
+      "step": 9500
+    },
+    {
+      "epoch": 1.9941287481652337,
+      "grad_norm": 0.055085159838199615,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 9510
+    },
+    {
+      "epoch": 1.9962256238205074,
+      "grad_norm": 0.07487735152244568,
+      "learning_rate": 0.0002,
+      "loss": 0.2739,
+      "step": 9520
+    },
+    {
+      "epoch": 1.9983224994757811,
+      "grad_norm": 0.06377065926790237,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 9530
+    },
+    {
+      "epoch": 2.0004193751310546,
+      "grad_norm": 2.2102627754211426,
+      "learning_rate": 0.0002,
+      "loss": 0.3525,
+      "step": 9540
+    },
+    {
+      "epoch": 2.0025162507863286,
+      "grad_norm": 0.14670875668525696,
+      "learning_rate": 0.0002,
+      "loss": 0.3185,
+      "step": 9550
+    },
+    {
+      "epoch": 2.004613126441602,
+      "grad_norm": 0.049855392426252365,
+      "learning_rate": 0.0002,
+      "loss": 0.2661,
+      "step": 9560
+    },
+    {
+      "epoch": 2.0067100020968756,
+      "grad_norm": 0.0794902816414833,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 9570
+    },
+    {
+      "epoch": 2.0088068777521495,
+      "grad_norm": 0.07908865809440613,
+      "learning_rate": 0.0002,
+      "loss": 0.2656,
+      "step": 9580
+    },
+    {
+      "epoch": 2.010903753407423,
+      "grad_norm": 0.07637414336204529,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 9590
+    },
+    {
+      "epoch": 2.0130006290626965,
+      "grad_norm": 0.07523901760578156,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 9600
+    },
+    {
+      "epoch": 2.0150975047179704,
+      "grad_norm": 0.05918925628066063,
+      "learning_rate": 0.0002,
+      "loss": 0.2853,
+      "step": 9610
+    },
+    {
+      "epoch": 2.017194380373244,
+      "grad_norm": 0.05923960730433464,
+      "learning_rate": 0.0002,
+      "loss": 0.2718,
+      "step": 9620
+    },
+    {
+      "epoch": 2.0192912560285174,
+      "grad_norm": 0.10112644731998444,
+      "learning_rate": 0.0002,
+      "loss": 0.2652,
+      "step": 9630
+    },
+    {
+      "epoch": 2.021388131683791,
+      "grad_norm": 0.07971510291099548,
+      "learning_rate": 0.0002,
+      "loss": 0.2928,
+      "step": 9640
+    },
+    {
+      "epoch": 2.023485007339065,
+      "grad_norm": 0.04727751389145851,
+      "learning_rate": 0.0002,
+      "loss": 0.2591,
+      "step": 9650
+    },
+    {
+      "epoch": 2.0255818829943384,
+      "grad_norm": 0.0745919719338417,
+      "learning_rate": 0.0002,
+      "loss": 0.2425,
+      "step": 9660
+    },
+    {
+      "epoch": 2.027678758649612,
+      "grad_norm": 2.6715872287750244,
+      "learning_rate": 0.0002,
+      "loss": 0.261,
+      "step": 9670
+    },
+    {
+      "epoch": 2.029775634304886,
+      "grad_norm": 0.06958667933940887,
+      "learning_rate": 0.0002,
+      "loss": 0.2612,
+      "step": 9680
+    },
+    {
+      "epoch": 2.0318725099601593,
+      "grad_norm": 0.056604884564876556,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 9690
+    },
+    {
+      "epoch": 2.033969385615433,
+      "grad_norm": 0.07836975902318954,
+      "learning_rate": 0.0002,
+      "loss": 0.2568,
+      "step": 9700
+    },
+    {
+      "epoch": 2.036066261270707,
+      "grad_norm": 3.000828742980957,
+      "learning_rate": 0.0002,
+      "loss": 0.2926,
+      "step": 9710
+    },
+    {
+      "epoch": 2.0381631369259803,
+      "grad_norm": 0.08067448437213898,
+      "learning_rate": 0.0002,
+      "loss": 0.2551,
+      "step": 9720
+    },
+    {
+      "epoch": 2.040260012581254,
+      "grad_norm": 0.057211246341466904,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 9730
+    },
+    {
+      "epoch": 2.0423568882365277,
+      "grad_norm": 1.01911461353302,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 9740
+    },
+    {
+      "epoch": 2.0444537638918012,
+      "grad_norm": 0.10453460365533829,
+      "learning_rate": 0.0002,
+      "loss": 0.2917,
+      "step": 9750
+    },
+    {
+      "epoch": 2.0465506395470747,
+      "grad_norm": 0.17895422875881195,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 9760
+    },
+    {
+      "epoch": 2.0486475152023487,
+      "grad_norm": 0.10618945211172104,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 9770
+    },
+    {
+      "epoch": 2.050744390857622,
+      "grad_norm": 0.10148117691278458,
+      "learning_rate": 0.0002,
+      "loss": 0.2517,
+      "step": 9780
+    },
+    {
+      "epoch": 2.0528412665128957,
+      "grad_norm": 0.11543615907430649,
+      "learning_rate": 0.0002,
+      "loss": 0.2431,
+      "step": 9790
+    },
+    {
+      "epoch": 2.0549381421681696,
+      "grad_norm": 0.05872245505452156,
+      "learning_rate": 0.0002,
+      "loss": 0.2701,
+      "step": 9800
+    },
+    {
+      "epoch": 2.057035017823443,
+      "grad_norm": 0.08900139480829239,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 9810
+    },
+    {
+      "epoch": 2.0591318934787166,
+      "grad_norm": 0.08121102303266525,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 9820
+    },
+    {
+      "epoch": 2.0612287691339906,
+      "grad_norm": 0.07904442399740219,
+      "learning_rate": 0.0002,
+      "loss": 0.2526,
+      "step": 9830
+    },
+    {
+      "epoch": 2.063325644789264,
+      "grad_norm": 8.165481567382812,
+      "learning_rate": 0.0002,
+      "loss": 0.2719,
+      "step": 9840
+    },
+    {
+      "epoch": 2.0654225204445376,
+      "grad_norm": 0.07967832684516907,
+      "learning_rate": 0.0002,
+      "loss": 0.2536,
+      "step": 9850
+    },
+    {
+      "epoch": 2.0675193960998115,
+      "grad_norm": 3.6327831745147705,
+      "learning_rate": 0.0002,
+      "loss": 0.2661,
+      "step": 9860
+    },
+    {
+      "epoch": 2.069616271755085,
+      "grad_norm": 0.0879238173365593,
+      "learning_rate": 0.0002,
+      "loss": 0.2562,
+      "step": 9870
+    },
+    {
+      "epoch": 2.0717131474103585,
+      "grad_norm": 0.11624127626419067,
+      "learning_rate": 0.0002,
+      "loss": 0.2569,
+      "step": 9880
+    },
+    {
+      "epoch": 2.073810023065632,
+      "grad_norm": 0.05972060561180115,
+      "learning_rate": 0.0002,
+      "loss": 0.2419,
+      "step": 9890
+    },
+    {
+      "epoch": 2.075906898720906,
+      "grad_norm": 0.09197453409433365,
+      "learning_rate": 0.0002,
+      "loss": 0.26,
+      "step": 9900
+    },
+    {
+      "epoch": 2.0780037743761794,
+      "grad_norm": 2.677229404449463,
+      "learning_rate": 0.0002,
+      "loss": 0.2719,
+      "step": 9910
+    },
+    {
+      "epoch": 2.080100650031453,
+      "grad_norm": 0.07371783256530762,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 9920
+    },
+    {
+      "epoch": 2.082197525686727,
+      "grad_norm": 0.08582104742527008,
+      "learning_rate": 0.0002,
+      "loss": 0.2652,
+      "step": 9930
+    },
+    {
+      "epoch": 2.0842944013420004,
+      "grad_norm": 0.08578525483608246,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 9940
+    },
+    {
+      "epoch": 2.086391276997274,
+      "grad_norm": 0.4039681851863861,
+      "learning_rate": 0.0002,
+      "loss": 0.3084,
+      "step": 9950
+    },
+    {
+      "epoch": 2.088488152652548,
+      "grad_norm": 0.33731213212013245,
+      "learning_rate": 0.0002,
+      "loss": 0.2818,
+      "step": 9960
+    },
+    {
+      "epoch": 2.0905850283078213,
+      "grad_norm": 0.13896989822387695,
+      "learning_rate": 0.0002,
+      "loss": 0.2617,
+      "step": 9970
+    },
+    {
+      "epoch": 2.092681903963095,
+      "grad_norm": 0.060974400490522385,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 9980
+    },
+    {
+      "epoch": 2.094778779618369,
+      "grad_norm": 0.06085608899593353,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 9990
+    },
+    {
+      "epoch": 2.0968756552736423,
+      "grad_norm": 0.07415081560611725,
+      "learning_rate": 0.0002,
+      "loss": 0.2908,
+      "step": 10000
+    },
+    {
+      "epoch": 2.098972530928916,
+      "grad_norm": 1.3254259824752808,
+      "learning_rate": 0.0002,
+      "loss": 0.2585,
+      "step": 10010
+    },
+    {
+      "epoch": 2.1010694065841897,
+      "grad_norm": 0.06385322660207748,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 10020
+    },
+    {
+      "epoch": 2.1031662822394632,
+      "grad_norm": 0.07257821410894394,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 10030
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.07844216376543045,
+      "learning_rate": 0.0002,
+      "loss": 0.2934,
+      "step": 10040
+    },
+    {
+      "epoch": 2.1073600335500107,
+      "grad_norm": 0.07658874243497849,
+      "learning_rate": 0.0002,
+      "loss": 0.3154,
+      "step": 10050
+    },
+    {
+      "epoch": 2.109456909205284,
+      "grad_norm": 0.07644514739513397,
+      "learning_rate": 0.0002,
+      "loss": 0.2876,
+      "step": 10060
+    },
+    {
+      "epoch": 2.1115537848605577,
+      "grad_norm": 0.06457582861185074,
+      "learning_rate": 0.0002,
+      "loss": 0.2656,
+      "step": 10070
+    },
+    {
+      "epoch": 2.1136506605158316,
+      "grad_norm": 3.1369924545288086,
+      "learning_rate": 0.0002,
+      "loss": 0.2947,
+      "step": 10080
+    },
+    {
+      "epoch": 2.115747536171105,
+      "grad_norm": 5.061370849609375,
+      "learning_rate": 0.0002,
+      "loss": 0.2607,
+      "step": 10090
+    },
+    {
+      "epoch": 2.1178444118263786,
+      "grad_norm": 0.0855194702744484,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 10100
+    },
+    {
+      "epoch": 2.1199412874816526,
+      "grad_norm": 0.8303119540214539,
+      "learning_rate": 0.0002,
+      "loss": 0.322,
+      "step": 10110
+    },
+    {
+      "epoch": 2.122038163136926,
+      "grad_norm": 0.05734449625015259,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 10120
+    },
+    {
+      "epoch": 2.1241350387921996,
+      "grad_norm": 0.08011757582426071,
+      "learning_rate": 0.0002,
+      "loss": 0.2963,
+      "step": 10130
+    },
+    {
+      "epoch": 2.126231914447473,
+      "grad_norm": 0.09854753315448761,
+      "learning_rate": 0.0002,
+      "loss": 0.2799,
+      "step": 10140
+    },
+    {
+      "epoch": 2.128328790102747,
+      "grad_norm": 0.07053971290588379,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 10150
+    },
+    {
+      "epoch": 2.1304256657580205,
+      "grad_norm": 0.07749179005622864,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 10160
+    },
+    {
+      "epoch": 2.132522541413294,
+      "grad_norm": 0.09702768921852112,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 10170
+    },
+    {
+      "epoch": 2.134619417068568,
+      "grad_norm": 0.15351717174053192,
+      "learning_rate": 0.0002,
+      "loss": 0.2431,
+      "step": 10180
+    },
+    {
+      "epoch": 2.1367162927238414,
+      "grad_norm": 0.05788223445415497,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 10190
+    },
+    {
+      "epoch": 2.138813168379115,
+      "grad_norm": 0.052700772881507874,
+      "learning_rate": 0.0002,
+      "loss": 0.2615,
+      "step": 10200
+    },
+    {
+      "epoch": 2.140910044034389,
+      "grad_norm": 0.053113095462322235,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 10210
+    },
+    {
+      "epoch": 2.1430069196896624,
+      "grad_norm": 3.7025339603424072,
+      "learning_rate": 0.0002,
+      "loss": 0.2842,
+      "step": 10220
+    },
+    {
+      "epoch": 2.145103795344936,
+      "grad_norm": 0.28933843970298767,
+      "learning_rate": 0.0002,
+      "loss": 0.2887,
+      "step": 10230
+    },
+    {
+      "epoch": 2.14720067100021,
+      "grad_norm": 3.0457377433776855,
+      "learning_rate": 0.0002,
+      "loss": 0.2622,
+      "step": 10240
+    },
+    {
+      "epoch": 2.1492975466554833,
+      "grad_norm": 0.3718264102935791,
+      "learning_rate": 0.0002,
+      "loss": 0.2572,
+      "step": 10250
+    },
+    {
+      "epoch": 2.151394422310757,
+      "grad_norm": 0.1321519911289215,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 10260
+    },
+    {
+      "epoch": 2.153491297966031,
+      "grad_norm": 0.1818964034318924,
+      "learning_rate": 0.0002,
+      "loss": 0.2604,
+      "step": 10270
+    },
+    {
+      "epoch": 2.1555881736213043,
+      "grad_norm": 4.632304668426514,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 10280
+    },
+    {
+      "epoch": 2.157685049276578,
+      "grad_norm": 0.26138266921043396,
+      "learning_rate": 0.0002,
+      "loss": 0.2507,
+      "step": 10290
+    },
+    {
+      "epoch": 2.1597819249318517,
+      "grad_norm": 11.915481567382812,
+      "learning_rate": 0.0002,
+      "loss": 0.2785,
+      "step": 10300
+    },
+    {
+      "epoch": 2.1618788005871252,
+      "grad_norm": 0.5993645191192627,
+      "learning_rate": 0.0002,
+      "loss": 0.268,
+      "step": 10310
+    },
+    {
+      "epoch": 2.1639756762423987,
+      "grad_norm": 0.1072174459695816,
+      "learning_rate": 0.0002,
+      "loss": 0.2469,
+      "step": 10320
+    },
+    {
+      "epoch": 2.1660725518976727,
+      "grad_norm": 0.20691901445388794,
+      "learning_rate": 0.0002,
+      "loss": 0.2753,
+      "step": 10330
+    },
+    {
+      "epoch": 2.168169427552946,
+      "grad_norm": 0.5400320887565613,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 10340
+    },
+    {
+      "epoch": 2.1702663032082197,
+      "grad_norm": 4.566557884216309,
+      "learning_rate": 0.0002,
+      "loss": 0.3557,
+      "step": 10350
+    },
+    {
+      "epoch": 2.1723631788634936,
+      "grad_norm": 0.29634520411491394,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 10360
+    },
+    {
+      "epoch": 2.174460054518767,
+      "grad_norm": 0.07772216945886612,
+      "learning_rate": 0.0002,
+      "loss": 0.2554,
+      "step": 10370
+    },
+    {
+      "epoch": 2.1765569301740406,
+      "grad_norm": 0.897849440574646,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 10380
+    },
+    {
+      "epoch": 2.178653805829314,
+      "grad_norm": 2.097255229949951,
+      "learning_rate": 0.0002,
+      "loss": 0.2715,
+      "step": 10390
+    },
+    {
+      "epoch": 2.180750681484588,
+      "grad_norm": 3.3917410373687744,
+      "learning_rate": 0.0002,
+      "loss": 0.3093,
+      "step": 10400
+    },
+    {
+      "epoch": 2.1828475571398616,
+      "grad_norm": 0.07134232670068741,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 10410
+    },
+    {
+      "epoch": 2.184944432795135,
+      "grad_norm": 0.11076632142066956,
+      "learning_rate": 0.0002,
+      "loss": 0.3185,
+      "step": 10420
+    },
+    {
+      "epoch": 2.187041308450409,
+      "grad_norm": 0.08823218941688538,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 10430
+    },
+    {
+      "epoch": 2.1891381841056825,
+      "grad_norm": 0.10004838556051254,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 10440
+    },
+    {
+      "epoch": 2.191235059760956,
+      "grad_norm": 0.10475442558526993,
+      "learning_rate": 0.0002,
+      "loss": 0.3018,
+      "step": 10450
+    },
+    {
+      "epoch": 2.19333193541623,
+      "grad_norm": 0.14505670964717865,
+      "learning_rate": 0.0002,
+      "loss": 0.3324,
+      "step": 10460
+    },
+    {
+      "epoch": 2.1954288110715034,
+      "grad_norm": 0.11367640644311905,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 10470
+    },
+    {
+      "epoch": 2.197525686726777,
+      "grad_norm": 0.08633550256490707,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 10480
+    },
+    {
+      "epoch": 2.199622562382051,
+      "grad_norm": 0.05888749659061432,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 10490
+    },
+    {
+      "epoch": 2.2017194380373244,
+      "grad_norm": 0.0671803280711174,
+      "learning_rate": 0.0002,
+      "loss": 0.2423,
+      "step": 10500
+    },
+    {
+      "epoch": 2.203816313692598,
+      "grad_norm": 0.10740163177251816,
+      "learning_rate": 0.0002,
+      "loss": 0.2593,
+      "step": 10510
+    },
+    {
+      "epoch": 2.205913189347872,
+      "grad_norm": 3.762645959854126,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 10520
+    },
+    {
+      "epoch": 2.2080100650031453,
+      "grad_norm": 0.05283233895897865,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 10530
+    },
+    {
+      "epoch": 2.210106940658419,
+      "grad_norm": 4.323790550231934,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 10540
+    },
+    {
+      "epoch": 2.212203816313693,
+      "grad_norm": 0.412309467792511,
+      "learning_rate": 0.0002,
+      "loss": 0.2884,
+      "step": 10550
+    },
+    {
+      "epoch": 2.2143006919689663,
+      "grad_norm": 0.07542133331298828,
+      "learning_rate": 0.0002,
+      "loss": 0.2772,
+      "step": 10560
+    },
+    {
+      "epoch": 2.21639756762424,
+      "grad_norm": 2.28047776222229,
+      "learning_rate": 0.0002,
+      "loss": 0.3163,
+      "step": 10570
+    },
+    {
+      "epoch": 2.2184944432795137,
+      "grad_norm": 0.09531054645776749,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 10580
+    },
+    {
+      "epoch": 2.2205913189347872,
+      "grad_norm": 0.08148845285177231,
+      "learning_rate": 0.0002,
+      "loss": 0.2787,
+      "step": 10590
+    },
+    {
+      "epoch": 2.2226881945900607,
+      "grad_norm": 3.0185651779174805,
+      "learning_rate": 0.0002,
+      "loss": 0.3192,
+      "step": 10600
+    },
+    {
+      "epoch": 2.2247850702453347,
+      "grad_norm": 3.6113274097442627,
+      "learning_rate": 0.0002,
+      "loss": 0.3345,
+      "step": 10610
+    },
+    {
+      "epoch": 2.226881945900608,
+      "grad_norm": 2.7463948726654053,
+      "learning_rate": 0.0002,
+      "loss": 0.3242,
+      "step": 10620
+    },
+    {
+      "epoch": 2.2289788215558817,
+      "grad_norm": 0.14245733618736267,
+      "learning_rate": 0.0002,
+      "loss": 0.268,
+      "step": 10630
+    },
+    {
+      "epoch": 2.231075697211155,
+      "grad_norm": 0.07894956320524216,
+      "learning_rate": 0.0002,
+      "loss": 0.2615,
+      "step": 10640
+    },
+    {
+      "epoch": 2.233172572866429,
+      "grad_norm": 4.841274261474609,
+      "learning_rate": 0.0002,
+      "loss": 0.271,
+      "step": 10650
+    },
+    {
+      "epoch": 2.2352694485217026,
+      "grad_norm": 0.27884334325790405,
+      "learning_rate": 0.0002,
+      "loss": 0.2677,
+      "step": 10660
+    },
+    {
+      "epoch": 2.237366324176976,
+      "grad_norm": 4.750665664672852,
+      "learning_rate": 0.0002,
+      "loss": 0.3232,
+      "step": 10670
+    },
+    {
+      "epoch": 2.23946319983225,
+      "grad_norm": 0.05985650420188904,
+      "learning_rate": 0.0002,
+      "loss": 0.27,
+      "step": 10680
+    },
+    {
+      "epoch": 2.2415600754875236,
+      "grad_norm": 5.086830139160156,
+      "learning_rate": 0.0002,
+      "loss": 0.3163,
+      "step": 10690
+    },
+    {
+      "epoch": 2.243656951142797,
+      "grad_norm": 0.7766889929771423,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 10700
+    },
+    {
+      "epoch": 2.245753826798071,
+      "grad_norm": 0.10954524576663971,
+      "learning_rate": 0.0002,
+      "loss": 0.2748,
+      "step": 10710
+    },
+    {
+      "epoch": 2.2478507024533445,
+      "grad_norm": 0.09740043431520462,
+      "learning_rate": 0.0002,
+      "loss": 0.2736,
+      "step": 10720
+    },
+    {
+      "epoch": 2.249947578108618,
+      "grad_norm": 0.09562437236309052,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 10730
+    },
+    {
+      "epoch": 2.252044453763892,
+      "grad_norm": 0.09752257168292999,
+      "learning_rate": 0.0002,
+      "loss": 0.2735,
+      "step": 10740
+    },
+    {
+      "epoch": 2.2541413294191655,
+      "grad_norm": 0.09222041815519333,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 10750
+    },
+    {
+      "epoch": 2.256238205074439,
+      "grad_norm": 2.323028802871704,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 10760
+    },
+    {
+      "epoch": 2.258335080729713,
+      "grad_norm": 0.05767745524644852,
+      "learning_rate": 0.0002,
+      "loss": 0.2518,
+      "step": 10770
+    },
+    {
+      "epoch": 2.2604319563849864,
+      "grad_norm": 0.060549937188625336,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 10780
+    },
+    {
+      "epoch": 2.26252883204026,
+      "grad_norm": 0.06989111751317978,
+      "learning_rate": 0.0002,
+      "loss": 0.2623,
+      "step": 10790
+    },
+    {
+      "epoch": 2.264625707695534,
+      "grad_norm": 1.974779725074768,
+      "learning_rate": 0.0002,
+      "loss": 0.2831,
+      "step": 10800
+    },
+    {
+      "epoch": 2.2667225833508073,
+      "grad_norm": 0.13178583979606628,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 10810
+    },
+    {
+      "epoch": 2.268819459006081,
+      "grad_norm": 5.346510410308838,
+      "learning_rate": 0.0002,
+      "loss": 0.3102,
+      "step": 10820
+    },
+    {
+      "epoch": 2.2709163346613543,
+      "grad_norm": 0.09758991748094559,
+      "learning_rate": 0.0002,
+      "loss": 0.2635,
+      "step": 10830
+    },
+    {
+      "epoch": 2.2730132103166283,
+      "grad_norm": 0.11416572332382202,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 10840
+    },
+    {
+      "epoch": 2.275110085971902,
+      "grad_norm": 0.09764792770147324,
+      "learning_rate": 0.0002,
+      "loss": 0.3057,
+      "step": 10850
+    },
+    {
+      "epoch": 2.2772069616271757,
+      "grad_norm": 0.07572513073682785,
+      "learning_rate": 0.0002,
+      "loss": 0.2745,
+      "step": 10860
+    },
+    {
+      "epoch": 2.2793038372824492,
+      "grad_norm": 0.08429969102144241,
+      "learning_rate": 0.0002,
+      "loss": 0.3372,
+      "step": 10870
+    },
+    {
+      "epoch": 2.2814007129377227,
+      "grad_norm": 0.09187424182891846,
+      "learning_rate": 0.0002,
+      "loss": 0.2887,
+      "step": 10880
+    },
+    {
+      "epoch": 2.2834975885929962,
+      "grad_norm": 0.10824614018201828,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 10890
+    },
+    {
+      "epoch": 2.28559446424827,
+      "grad_norm": 0.06268329918384552,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 10900
+    },
+    {
+      "epoch": 2.2876913399035437,
+      "grad_norm": 2.6071887016296387,
+      "learning_rate": 0.0002,
+      "loss": 0.2885,
+      "step": 10910
+    },
+    {
+      "epoch": 2.289788215558817,
+      "grad_norm": 0.33325842022895813,
+      "learning_rate": 0.0002,
+      "loss": 0.3228,
+      "step": 10920
+    },
+    {
+      "epoch": 2.291885091214091,
+      "grad_norm": 0.10186664015054703,
+      "learning_rate": 0.0002,
+      "loss": 0.2773,
+      "step": 10930
+    },
+    {
+      "epoch": 2.2939819668693646,
+      "grad_norm": 0.1469857096672058,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 10940
+    },
+    {
+      "epoch": 2.296078842524638,
+      "grad_norm": 0.23534300923347473,
+      "learning_rate": 0.0002,
+      "loss": 0.2654,
+      "step": 10950
+    },
+    {
+      "epoch": 2.298175718179912,
+      "grad_norm": 0.09361186623573303,
+      "learning_rate": 0.0002,
+      "loss": 0.2612,
+      "step": 10960
+    },
+    {
+      "epoch": 2.3002725938351856,
+      "grad_norm": 3.4896860122680664,
+      "learning_rate": 0.0002,
+      "loss": 0.2827,
+      "step": 10970
+    },
+    {
+      "epoch": 2.302369469490459,
+      "grad_norm": 0.07228028774261475,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 10980
+    },
+    {
+      "epoch": 2.304466345145733,
+      "grad_norm": 0.10563217848539352,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 10990
+    },
+    {
+      "epoch": 2.3065632208010065,
+      "grad_norm": 0.10082051157951355,
+      "learning_rate": 0.0002,
+      "loss": 0.2989,
+      "step": 11000
+    },
+    {
+      "epoch": 2.30866009645628,
+      "grad_norm": 0.09514462947845459,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 11010
+    },
+    {
+      "epoch": 2.310756972111554,
+      "grad_norm": 0.05637963116168976,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 11020
+    },
+    {
+      "epoch": 2.3128538477668275,
+      "grad_norm": 0.0820513516664505,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 11030
+    },
+    {
+      "epoch": 2.314950723422101,
+      "grad_norm": 0.49234119057655334,
+      "learning_rate": 0.0002,
+      "loss": 0.2607,
+      "step": 11040
+    },
+    {
+      "epoch": 2.317047599077375,
+      "grad_norm": 0.056676942855119705,
+      "learning_rate": 0.0002,
+      "loss": 0.2876,
+      "step": 11050
+    },
+    {
+      "epoch": 2.3191444747326484,
+      "grad_norm": 0.17538276314735413,
+      "learning_rate": 0.0002,
+      "loss": 0.256,
+      "step": 11060
+    },
+    {
+      "epoch": 2.321241350387922,
+      "grad_norm": 1.0409997701644897,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 11070
+    },
+    {
+      "epoch": 2.3233382260431954,
+      "grad_norm": 3.7242491245269775,
+      "learning_rate": 0.0002,
+      "loss": 0.2733,
+      "step": 11080
+    },
+    {
+      "epoch": 2.3254351016984693,
+      "grad_norm": 0.08921827375888824,
+      "learning_rate": 0.0002,
+      "loss": 0.2972,
+      "step": 11090
+    },
+    {
+      "epoch": 2.327531977353743,
+      "grad_norm": 0.083032988011837,
+      "learning_rate": 0.0002,
+      "loss": 0.3251,
+      "step": 11100
+    },
+    {
+      "epoch": 2.329628853009017,
+      "grad_norm": 0.10303895175457001,
+      "learning_rate": 0.0002,
+      "loss": 0.2617,
+      "step": 11110
+    },
+    {
+      "epoch": 2.3317257286642903,
+      "grad_norm": 0.1020461916923523,
+      "learning_rate": 0.0002,
+      "loss": 0.2736,
+      "step": 11120
+    },
+    {
+      "epoch": 2.333822604319564,
+      "grad_norm": 0.09795411676168442,
+      "learning_rate": 0.0002,
+      "loss": 0.2948,
+      "step": 11130
+    },
+    {
+      "epoch": 2.3359194799748373,
+      "grad_norm": 0.05664985999464989,
+      "learning_rate": 0.0002,
+      "loss": 0.2932,
+      "step": 11140
+    },
+    {
+      "epoch": 2.3380163556301112,
+      "grad_norm": 0.06109614670276642,
+      "learning_rate": 0.0002,
+      "loss": 0.3056,
+      "step": 11150
+    },
+    {
+      "epoch": 2.3401132312853847,
+      "grad_norm": 0.0780511125922203,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 11160
+    },
+    {
+      "epoch": 2.3422101069406582,
+      "grad_norm": 0.06719834357500076,
+      "learning_rate": 0.0002,
+      "loss": 0.2926,
+      "step": 11170
+    },
+    {
+      "epoch": 2.344306982595932,
+      "grad_norm": 0.11557465046644211,
+      "learning_rate": 0.0002,
+      "loss": 0.2482,
+      "step": 11180
+    },
+    {
+      "epoch": 2.3464038582512057,
+      "grad_norm": 0.06901777535676956,
+      "learning_rate": 0.0002,
+      "loss": 0.2958,
+      "step": 11190
+    },
+    {
+      "epoch": 2.348500733906479,
+      "grad_norm": 0.13302722573280334,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 11200
+    },
+    {
+      "epoch": 2.350597609561753,
+      "grad_norm": 0.060133375227451324,
+      "learning_rate": 0.0002,
+      "loss": 0.2639,
+      "step": 11210
+    },
+    {
+      "epoch": 2.3526944852170266,
+      "grad_norm": 0.062260959297418594,
+      "learning_rate": 0.0002,
+      "loss": 0.2538,
+      "step": 11220
+    },
+    {
+      "epoch": 2.3547913608723,
+      "grad_norm": 3.3237318992614746,
+      "learning_rate": 0.0002,
+      "loss": 0.2868,
+      "step": 11230
+    },
+    {
+      "epoch": 2.356888236527574,
+      "grad_norm": 0.08286066353321075,
+      "learning_rate": 0.0002,
+      "loss": 0.2731,
+      "step": 11240
+    },
+    {
+      "epoch": 2.3589851121828476,
+      "grad_norm": 0.07192427664995193,
+      "learning_rate": 0.0002,
+      "loss": 0.2618,
+      "step": 11250
+    },
+    {
+      "epoch": 2.361081987838121,
+      "grad_norm": 0.09143181145191193,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 11260
+    },
+    {
+      "epoch": 2.363178863493395,
+      "grad_norm": 0.09221427142620087,
+      "learning_rate": 0.0002,
+      "loss": 0.2411,
+      "step": 11270
+    },
+    {
+      "epoch": 2.3652757391486685,
+      "grad_norm": 0.06351318210363388,
+      "learning_rate": 0.0002,
+      "loss": 0.2427,
+      "step": 11280
+    },
+    {
+      "epoch": 2.367372614803942,
+      "grad_norm": 0.09457197785377502,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 11290
+    },
+    {
+      "epoch": 2.369469490459216,
+      "grad_norm": 0.05240204185247421,
+      "learning_rate": 0.0002,
+      "loss": 0.2893,
+      "step": 11300
+    },
+    {
+      "epoch": 2.3715663661144895,
+      "grad_norm": 0.08396653831005096,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 11310
+    },
+    {
+      "epoch": 2.373663241769763,
+      "grad_norm": 0.17449672520160675,
+      "learning_rate": 0.0002,
+      "loss": 0.3056,
+      "step": 11320
+    },
+    {
+      "epoch": 2.3757601174250365,
+      "grad_norm": 0.07481683790683746,
+      "learning_rate": 0.0002,
+      "loss": 0.2709,
+      "step": 11330
+    },
+    {
+      "epoch": 2.3778569930803104,
+      "grad_norm": 0.07266558706760406,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 11340
+    },
+    {
+      "epoch": 2.379953868735584,
+      "grad_norm": 0.059189122170209885,
+      "learning_rate": 0.0002,
+      "loss": 0.2421,
+      "step": 11350
+    },
+    {
+      "epoch": 2.382050744390858,
+      "grad_norm": 0.08741532266139984,
+      "learning_rate": 0.0002,
+      "loss": 0.2607,
+      "step": 11360
+    },
+    {
+      "epoch": 2.3841476200461313,
+      "grad_norm": 0.14015233516693115,
+      "learning_rate": 0.0002,
+      "loss": 0.3175,
+      "step": 11370
+    },
+    {
+      "epoch": 2.386244495701405,
+      "grad_norm": 10.488336563110352,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 11380
+    },
+    {
+      "epoch": 2.3883413713566783,
+      "grad_norm": 0.06913798302412033,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 11390
+    },
+    {
+      "epoch": 2.3904382470119523,
+      "grad_norm": 0.29002469778060913,
+      "learning_rate": 0.0002,
+      "loss": 0.256,
+      "step": 11400
+    },
+    {
+      "epoch": 2.392535122667226,
+      "grad_norm": 0.11059385538101196,
+      "learning_rate": 0.0002,
+      "loss": 0.2908,
+      "step": 11410
+    },
+    {
+      "epoch": 2.3946319983224993,
+      "grad_norm": 0.08097716420888901,
+      "learning_rate": 0.0002,
+      "loss": 0.2495,
+      "step": 11420
+    },
+    {
+      "epoch": 2.3967288739777732,
+      "grad_norm": 0.09082842618227005,
+      "learning_rate": 0.0002,
+      "loss": 0.3123,
+      "step": 11430
+    },
+    {
+      "epoch": 2.3988257496330467,
+      "grad_norm": 4.789621353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.3085,
+      "step": 11440
+    },
+    {
+      "epoch": 2.4009226252883202,
+      "grad_norm": 0.05977439135313034,
+      "learning_rate": 0.0002,
+      "loss": 0.2743,
+      "step": 11450
+    },
+    {
+      "epoch": 2.403019500943594,
+      "grad_norm": 3.4836881160736084,
+      "learning_rate": 0.0002,
+      "loss": 0.2806,
+      "step": 11460
+    },
+    {
+      "epoch": 2.4051163765988677,
+      "grad_norm": 0.07663311809301376,
+      "learning_rate": 0.0002,
+      "loss": 0.2564,
+      "step": 11470
+    },
+    {
+      "epoch": 2.407213252254141,
+      "grad_norm": 0.053984083235263824,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 11480
+    },
+    {
+      "epoch": 2.409310127909415,
+      "grad_norm": 0.059521205723285675,
+      "learning_rate": 0.0002,
+      "loss": 0.2712,
+      "step": 11490
+    },
+    {
+      "epoch": 2.4114070035646886,
+      "grad_norm": 2.755619764328003,
+      "learning_rate": 0.0002,
+      "loss": 0.3518,
+      "step": 11500
+    },
+    {
+      "epoch": 2.413503879219962,
+      "grad_norm": 0.06272760033607483,
+      "learning_rate": 0.0002,
+      "loss": 0.2542,
+      "step": 11510
+    },
+    {
+      "epoch": 2.415600754875236,
+      "grad_norm": 0.06255768239498138,
+      "learning_rate": 0.0002,
+      "loss": 0.2765,
+      "step": 11520
+    },
+    {
+      "epoch": 2.4176976305305096,
+      "grad_norm": 0.0611007884144783,
+      "learning_rate": 0.0002,
+      "loss": 0.2723,
+      "step": 11530
+    },
+    {
+      "epoch": 2.419794506185783,
+      "grad_norm": 0.0764545425772667,
+      "learning_rate": 0.0002,
+      "loss": 0.2426,
+      "step": 11540
+    },
+    {
+      "epoch": 2.421891381841057,
+      "grad_norm": 0.0603049099445343,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 11550
+    },
+    {
+      "epoch": 2.4239882574963305,
+      "grad_norm": 0.05984706059098244,
+      "learning_rate": 0.0002,
+      "loss": 0.2613,
+      "step": 11560
+    },
+    {
+      "epoch": 2.426085133151604,
+      "grad_norm": 0.09332500398159027,
+      "learning_rate": 0.0002,
+      "loss": 0.2595,
+      "step": 11570
+    },
+    {
+      "epoch": 2.4281820088068775,
+      "grad_norm": 2.8640525341033936,
+      "learning_rate": 0.0002,
+      "loss": 0.2837,
+      "step": 11580
+    },
+    {
+      "epoch": 2.4302788844621515,
+      "grad_norm": 0.06720197945833206,
+      "learning_rate": 0.0002,
+      "loss": 0.264,
+      "step": 11590
+    },
+    {
+      "epoch": 2.432375760117425,
+      "grad_norm": 0.053041424602270126,
+      "learning_rate": 0.0002,
+      "loss": 0.2623,
+      "step": 11600
+    },
+    {
+      "epoch": 2.434472635772699,
+      "grad_norm": 0.06075757369399071,
+      "learning_rate": 0.0002,
+      "loss": 0.3068,
+      "step": 11610
+    },
+    {
+      "epoch": 2.4365695114279724,
+      "grad_norm": 0.09649069607257843,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 11620
+    },
+    {
+      "epoch": 2.438666387083246,
+      "grad_norm": 0.10784143209457397,
+      "learning_rate": 0.0002,
+      "loss": 0.2532,
+      "step": 11630
+    },
+    {
+      "epoch": 2.4407632627385194,
+      "grad_norm": 2.704793930053711,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 11640
+    },
+    {
+      "epoch": 2.4428601383937933,
+      "grad_norm": 0.09455972164869308,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 11650
+    },
+    {
+      "epoch": 2.444957014049067,
+      "grad_norm": 0.05830010399222374,
+      "learning_rate": 0.0002,
+      "loss": 0.2727,
+      "step": 11660
+    },
+    {
+      "epoch": 2.4470538897043403,
+      "grad_norm": 3.8376691341400146,
+      "learning_rate": 0.0002,
+      "loss": 0.2645,
+      "step": 11670
+    },
+    {
+      "epoch": 2.4491507653596143,
+      "grad_norm": 0.07729987800121307,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 11680
+    },
+    {
+      "epoch": 2.451247641014888,
+      "grad_norm": 0.07105562835931778,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 11690
+    },
+    {
+      "epoch": 2.4533445166701613,
+      "grad_norm": 0.07121610641479492,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 11700
+    },
+    {
+      "epoch": 2.4554413923254352,
+      "grad_norm": 0.07170971482992172,
+      "learning_rate": 0.0002,
+      "loss": 0.2578,
+      "step": 11710
+    },
+    {
+      "epoch": 2.4575382679807087,
+      "grad_norm": 0.047798383980989456,
+      "learning_rate": 0.0002,
+      "loss": 0.2844,
+      "step": 11720
+    },
+    {
+      "epoch": 2.4596351436359822,
+      "grad_norm": 0.07498259842395782,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 11730
+    },
+    {
+      "epoch": 2.461732019291256,
+      "grad_norm": 0.08825545012950897,
+      "learning_rate": 0.0002,
+      "loss": 0.2914,
+      "step": 11740
+    },
+    {
+      "epoch": 2.4638288949465297,
+      "grad_norm": 0.09498555958271027,
+      "learning_rate": 0.0002,
+      "loss": 0.2938,
+      "step": 11750
+    },
+    {
+      "epoch": 2.465925770601803,
+      "grad_norm": 0.13061000406742096,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 11760
+    },
+    {
+      "epoch": 2.468022646257077,
+      "grad_norm": 0.11089792102575302,
+      "learning_rate": 0.0002,
+      "loss": 0.2754,
+      "step": 11770
+    },
+    {
+      "epoch": 2.4701195219123506,
+      "grad_norm": 0.15551145374774933,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 11780
+    },
+    {
+      "epoch": 2.472216397567624,
+      "grad_norm": 0.06919834762811661,
+      "learning_rate": 0.0002,
+      "loss": 0.264,
+      "step": 11790
+    },
+    {
+      "epoch": 2.474313273222898,
+      "grad_norm": 0.09006861597299576,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 11800
+    },
+    {
+      "epoch": 2.4764101488781716,
+      "grad_norm": 0.05835675820708275,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 11810
+    },
+    {
+      "epoch": 2.478507024533445,
+      "grad_norm": 0.061936892569065094,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 11820
+    },
+    {
+      "epoch": 2.4806039001887186,
+      "grad_norm": 0.07322103530168533,
+      "learning_rate": 0.0002,
+      "loss": 0.2708,
+      "step": 11830
+    },
+    {
+      "epoch": 2.4827007758439925,
+      "grad_norm": 0.07008501887321472,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 11840
+    },
+    {
+      "epoch": 2.484797651499266,
+      "grad_norm": 0.042795274406671524,
+      "learning_rate": 0.0002,
+      "loss": 0.243,
+      "step": 11850
+    },
+    {
+      "epoch": 2.48689452715454,
+      "grad_norm": 0.0587536059319973,
+      "learning_rate": 0.0002,
+      "loss": 0.2427,
+      "step": 11860
+    },
+    {
+      "epoch": 2.4889914028098135,
+      "grad_norm": 2.3064823150634766,
+      "learning_rate": 0.0002,
+      "loss": 0.2805,
+      "step": 11870
+    },
+    {
+      "epoch": 2.491088278465087,
+      "grad_norm": 0.07847301661968231,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 11880
+    },
+    {
+      "epoch": 2.4931851541203605,
+      "grad_norm": 0.05622512847185135,
+      "learning_rate": 0.0002,
+      "loss": 0.2738,
+      "step": 11890
+    },
+    {
+      "epoch": 2.4952820297756344,
+      "grad_norm": 4.064495086669922,
+      "learning_rate": 0.0002,
+      "loss": 0.346,
+      "step": 11900
+    },
+    {
+      "epoch": 2.497378905430908,
+      "grad_norm": 0.09369252622127533,
+      "learning_rate": 0.0002,
+      "loss": 0.2805,
+      "step": 11910
+    },
+    {
+      "epoch": 2.4994757810861814,
+      "grad_norm": 0.313037633895874,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 11920
+    },
+    {
+      "epoch": 2.5015726567414553,
+      "grad_norm": 0.8791367411613464,
+      "learning_rate": 0.0002,
+      "loss": 0.2752,
+      "step": 11930
+    },
+    {
+      "epoch": 2.503669532396729,
+      "grad_norm": 0.10336744785308838,
+      "learning_rate": 0.0002,
+      "loss": 0.3267,
+      "step": 11940
+    },
+    {
+      "epoch": 2.5057664080520023,
+      "grad_norm": 0.31161123514175415,
+      "learning_rate": 0.0002,
+      "loss": 0.2941,
+      "step": 11950
+    },
+    {
+      "epoch": 2.5078632837072763,
+      "grad_norm": 1.464455246925354,
+      "learning_rate": 0.0002,
+      "loss": 0.3281,
+      "step": 11960
+    },
+    {
+      "epoch": 2.50996015936255,
+      "grad_norm": 6.812768936157227,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 11970
+    },
+    {
+      "epoch": 2.5120570350178233,
+      "grad_norm": 0.17166270315647125,
+      "learning_rate": 0.0002,
+      "loss": 0.2806,
+      "step": 11980
+    },
+    {
+      "epoch": 2.5141539106730972,
+      "grad_norm": 0.09918298572301865,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 11990
+    },
+    {
+      "epoch": 2.5162507863283707,
+      "grad_norm": 0.21109609305858612,
+      "learning_rate": 0.0002,
+      "loss": 0.2637,
+      "step": 12000
+    },
+    {
+      "epoch": 2.5183476619836442,
+      "grad_norm": 0.16542023420333862,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 12010
+    },
+    {
+      "epoch": 2.5204445376389177,
+      "grad_norm": 0.10262157768011093,
+      "learning_rate": 0.0002,
+      "loss": 0.2483,
+      "step": 12020
+    },
+    {
+      "epoch": 2.5225414132941917,
+      "grad_norm": 0.08214957267045975,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 12030
+    },
+    {
+      "epoch": 2.524638288949465,
+      "grad_norm": 0.07718580961227417,
+      "learning_rate": 0.0002,
+      "loss": 0.2776,
+      "step": 12040
+    },
+    {
+      "epoch": 2.526735164604739,
+      "grad_norm": 0.09089051187038422,
+      "learning_rate": 0.0002,
+      "loss": 0.2864,
+      "step": 12050
+    },
+    {
+      "epoch": 2.5288320402600126,
+      "grad_norm": 0.08136362582445145,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 12060
+    },
+    {
+      "epoch": 2.530928915915286,
+      "grad_norm": 0.10573694854974747,
+      "learning_rate": 0.0002,
+      "loss": 0.2758,
+      "step": 12070
+    },
+    {
+      "epoch": 2.5330257915705596,
+      "grad_norm": 0.12686964869499207,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 12080
+    },
+    {
+      "epoch": 2.5351226672258336,
+      "grad_norm": 0.1271832436323166,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 12090
+    },
+    {
+      "epoch": 2.537219542881107,
+      "grad_norm": 0.06386271864175797,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 12100
+    },
+    {
+      "epoch": 2.539316418536381,
+      "grad_norm": 0.07435470819473267,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 12110
+    },
+    {
+      "epoch": 2.5414132941916545,
+      "grad_norm": 0.06934060901403427,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 12120
+    },
+    {
+      "epoch": 2.543510169846928,
+      "grad_norm": 0.08146541565656662,
+      "learning_rate": 0.0002,
+      "loss": 0.2833,
+      "step": 12130
+    },
+    {
+      "epoch": 2.5456070455022015,
+      "grad_norm": 0.06254391372203827,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 12140
+    },
+    {
+      "epoch": 2.5477039211574755,
+      "grad_norm": 0.057234663516283035,
+      "learning_rate": 0.0002,
+      "loss": 0.306,
+      "step": 12150
+    },
+    {
+      "epoch": 2.549800796812749,
+      "grad_norm": 0.08865554630756378,
+      "learning_rate": 0.0002,
+      "loss": 0.2638,
+      "step": 12160
+    },
+    {
+      "epoch": 2.551897672468023,
+      "grad_norm": 0.076085664331913,
+      "learning_rate": 0.0002,
+      "loss": 0.2732,
+      "step": 12170
+    },
+    {
+      "epoch": 2.5539945481232964,
+      "grad_norm": 0.06641250103712082,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 12180
+    },
+    {
+      "epoch": 2.55609142377857,
+      "grad_norm": 0.5051395297050476,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 12190
+    },
+    {
+      "epoch": 2.5581882994338434,
+      "grad_norm": 0.056382421404123306,
+      "learning_rate": 0.0002,
+      "loss": 0.2768,
+      "step": 12200
+    },
+    {
+      "epoch": 2.5602851750891173,
+      "grad_norm": 3.523155689239502,
+      "learning_rate": 0.0002,
+      "loss": 0.3345,
+      "step": 12210
+    },
+    {
+      "epoch": 2.562382050744391,
+      "grad_norm": 0.06870336085557938,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 12220
+    },
+    {
+      "epoch": 2.5644789263996643,
+      "grad_norm": 5.934406757354736,
+      "learning_rate": 0.0002,
+      "loss": 0.2691,
+      "step": 12230
+    },
+    {
+      "epoch": 2.5665758020549383,
+      "grad_norm": 0.09467433393001556,
+      "learning_rate": 0.0002,
+      "loss": 0.2741,
+      "step": 12240
+    },
+    {
+      "epoch": 2.568672677710212,
+      "grad_norm": 1.3006917238235474,
+      "learning_rate": 0.0002,
+      "loss": 0.2645,
+      "step": 12250
+    },
+    {
+      "epoch": 2.5707695533654853,
+      "grad_norm": 0.0707492008805275,
+      "learning_rate": 0.0002,
+      "loss": 0.2871,
+      "step": 12260
+    },
+    {
+      "epoch": 2.572866429020759,
+      "grad_norm": 0.05547978729009628,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 12270
+    },
+    {
+      "epoch": 2.5749633046760327,
+      "grad_norm": 0.060534629970788956,
+      "learning_rate": 0.0002,
+      "loss": 0.3156,
+      "step": 12280
+    },
+    {
+      "epoch": 2.5770601803313062,
+      "grad_norm": 0.7947044372558594,
+      "learning_rate": 0.0002,
+      "loss": 0.2579,
+      "step": 12290
+    },
+    {
+      "epoch": 2.57915705598658,
+      "grad_norm": 0.11533192545175552,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 12300
+    },
+    {
+      "epoch": 2.5812539316418537,
+      "grad_norm": 0.10663612186908722,
+      "learning_rate": 0.0002,
+      "loss": 0.2523,
+      "step": 12310
+    },
+    {
+      "epoch": 2.583350807297127,
+      "grad_norm": 0.0963052287697792,
+      "learning_rate": 0.0002,
+      "loss": 0.2856,
+      "step": 12320
+    },
+    {
+      "epoch": 2.5854476829524007,
+      "grad_norm": 0.08930831402540207,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 12330
+    },
+    {
+      "epoch": 2.5875445586076746,
+      "grad_norm": 0.10064201056957245,
+      "learning_rate": 0.0002,
+      "loss": 0.3087,
+      "step": 12340
+    },
+    {
+      "epoch": 2.589641434262948,
+      "grad_norm": 0.06038425862789154,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 12350
+    },
+    {
+      "epoch": 2.591738309918222,
+      "grad_norm": 0.07416463643312454,
+      "learning_rate": 0.0002,
+      "loss": 0.2586,
+      "step": 12360
+    },
+    {
+      "epoch": 2.5938351855734956,
+      "grad_norm": 2.1738131046295166,
+      "learning_rate": 0.0002,
+      "loss": 0.2933,
+      "step": 12370
+    },
+    {
+      "epoch": 2.595932061228769,
+      "grad_norm": 0.06500290334224701,
+      "learning_rate": 0.0002,
+      "loss": 0.2422,
+      "step": 12380
+    },
+    {
+      "epoch": 2.5980289368840426,
+      "grad_norm": 0.06393703818321228,
+      "learning_rate": 0.0002,
+      "loss": 0.2721,
+      "step": 12390
+    },
+    {
+      "epoch": 2.6001258125393165,
+      "grad_norm": 0.06458470970392227,
+      "learning_rate": 0.0002,
+      "loss": 0.2637,
+      "step": 12400
+    },
+    {
+      "epoch": 2.60222268819459,
+      "grad_norm": 0.08135922253131866,
+      "learning_rate": 0.0002,
+      "loss": 0.263,
+      "step": 12410
+    },
+    {
+      "epoch": 2.604319563849864,
+      "grad_norm": 0.052150268107652664,
+      "learning_rate": 0.0002,
+      "loss": 0.2408,
+      "step": 12420
+    },
+    {
+      "epoch": 2.6064164395051375,
+      "grad_norm": 0.08048977702856064,
+      "learning_rate": 0.0002,
+      "loss": 0.2547,
+      "step": 12430
+    },
+    {
+      "epoch": 2.608513315160411,
+      "grad_norm": 9.111763000488281,
+      "learning_rate": 0.0002,
+      "loss": 0.3519,
+      "step": 12440
+    },
+    {
+      "epoch": 2.6106101908156845,
+      "grad_norm": 0.1035270020365715,
+      "learning_rate": 0.0002,
+      "loss": 0.272,
+      "step": 12450
+    },
+    {
+      "epoch": 2.6127070664709584,
+      "grad_norm": 3.41133189201355,
+      "learning_rate": 0.0002,
+      "loss": 0.3269,
+      "step": 12460
+    },
+    {
+      "epoch": 2.614803942126232,
+      "grad_norm": 0.12809054553508759,
+      "learning_rate": 0.0002,
+      "loss": 0.2616,
+      "step": 12470
+    },
+    {
+      "epoch": 2.6169008177815054,
+      "grad_norm": 0.17681822180747986,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 12480
+    },
+    {
+      "epoch": 2.6189976934367794,
+      "grad_norm": 0.10416349768638611,
+      "learning_rate": 0.0002,
+      "loss": 0.305,
+      "step": 12490
+    },
+    {
+      "epoch": 2.621094569092053,
+      "grad_norm": 2.7841596603393555,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 12500
+    },
+    {
+      "epoch": 2.6231914447473264,
+      "grad_norm": 0.08688453584909439,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 12510
+    },
+    {
+      "epoch": 2.6252883204026,
+      "grad_norm": 1.3929383754730225,
+      "learning_rate": 0.0002,
+      "loss": 0.2929,
+      "step": 12520
+    },
+    {
+      "epoch": 2.627385196057874,
+      "grad_norm": 0.6145888566970825,
+      "learning_rate": 0.0002,
+      "loss": 0.2848,
+      "step": 12530
+    },
+    {
+      "epoch": 2.6294820717131473,
+      "grad_norm": 0.0767504870891571,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 12540
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.06284788995981216,
+      "learning_rate": 0.0002,
+      "loss": 0.3128,
+      "step": 12550
+    },
+    {
+      "epoch": 2.6336758230236947,
+      "grad_norm": 1.2423477172851562,
+      "learning_rate": 0.0002,
+      "loss": 0.27,
+      "step": 12560
+    },
+    {
+      "epoch": 2.6357726986789682,
+      "grad_norm": 0.07569570094347,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 12570
+    },
+    {
+      "epoch": 2.6378695743342417,
+      "grad_norm": 0.07165157049894333,
+      "learning_rate": 0.0002,
+      "loss": 0.2519,
+      "step": 12580
+    },
+    {
+      "epoch": 2.6399664499895157,
+      "grad_norm": 5.022453308105469,
+      "learning_rate": 0.0002,
+      "loss": 0.2629,
+      "step": 12590
+    },
+    {
+      "epoch": 2.642063325644789,
+      "grad_norm": 7.140848159790039,
+      "learning_rate": 0.0002,
+      "loss": 0.3051,
+      "step": 12600
+    },
+    {
+      "epoch": 2.644160201300063,
+      "grad_norm": 2.355073928833008,
+      "learning_rate": 0.0002,
+      "loss": 0.2495,
+      "step": 12610
+    },
+    {
+      "epoch": 2.6462570769553366,
+      "grad_norm": 0.0591590479016304,
+      "learning_rate": 0.0002,
+      "loss": 0.3141,
+      "step": 12620
+    },
+    {
+      "epoch": 2.64835395261061,
+      "grad_norm": 0.10556333512067795,
+      "learning_rate": 0.0002,
+      "loss": 0.2489,
+      "step": 12630
+    },
+    {
+      "epoch": 2.6504508282658836,
+      "grad_norm": 0.09041710197925568,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 12640
+    },
+    {
+      "epoch": 2.6525477039211576,
+      "grad_norm": 5.497735977172852,
+      "learning_rate": 0.0002,
+      "loss": 0.295,
+      "step": 12650
+    },
+    {
+      "epoch": 2.654644579576431,
+      "grad_norm": 0.15341436862945557,
+      "learning_rate": 0.0002,
+      "loss": 0.2849,
+      "step": 12660
+    },
+    {
+      "epoch": 2.656741455231705,
+      "grad_norm": 0.15221437811851501,
+      "learning_rate": 0.0002,
+      "loss": 0.2732,
+      "step": 12670
+    },
+    {
+      "epoch": 2.6588383308869785,
+      "grad_norm": 0.06831462681293488,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 12680
+    },
+    {
+      "epoch": 2.660935206542252,
+      "grad_norm": 1.100542426109314,
+      "learning_rate": 0.0002,
+      "loss": 0.2926,
+      "step": 12690
+    },
+    {
+      "epoch": 2.6630320821975255,
+      "grad_norm": 0.08846069127321243,
+      "learning_rate": 0.0002,
+      "loss": 0.264,
+      "step": 12700
+    },
+    {
+      "epoch": 2.6651289578527995,
+      "grad_norm": 0.06772548705339432,
+      "learning_rate": 0.0002,
+      "loss": 0.3007,
+      "step": 12710
+    },
+    {
+      "epoch": 2.667225833508073,
+      "grad_norm": 0.06996563822031021,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 12720
+    },
+    {
+      "epoch": 2.6693227091633465,
+      "grad_norm": 0.1149798259139061,
+      "learning_rate": 0.0002,
+      "loss": 0.2862,
+      "step": 12730
+    },
+    {
+      "epoch": 2.6714195848186204,
+      "grad_norm": 0.08013114333152771,
+      "learning_rate": 0.0002,
+      "loss": 0.2716,
+      "step": 12740
+    },
+    {
+      "epoch": 2.673516460473894,
+      "grad_norm": 0.0764973983168602,
+      "learning_rate": 0.0002,
+      "loss": 0.2576,
+      "step": 12750
+    },
+    {
+      "epoch": 2.6756133361291674,
+      "grad_norm": 0.05453988537192345,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 12760
+    },
+    {
+      "epoch": 2.677710211784441,
+      "grad_norm": 0.07638183236122131,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 12770
+    },
+    {
+      "epoch": 2.679807087439715,
+      "grad_norm": 0.07708717882633209,
+      "learning_rate": 0.0002,
+      "loss": 0.2424,
+      "step": 12780
+    },
+    {
+      "epoch": 2.6819039630949884,
+      "grad_norm": 0.8045825362205505,
+      "learning_rate": 0.0002,
+      "loss": 0.3402,
+      "step": 12790
+    },
+    {
+      "epoch": 2.6840008387502623,
+      "grad_norm": 0.057965174317359924,
+      "learning_rate": 0.0002,
+      "loss": 0.2758,
+      "step": 12800
+    },
+    {
+      "epoch": 2.686097714405536,
+      "grad_norm": 4.041475772857666,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 12810
+    },
+    {
+      "epoch": 2.6881945900608093,
+      "grad_norm": 0.2805425822734833,
+      "learning_rate": 0.0002,
+      "loss": 0.2633,
+      "step": 12820
+    },
+    {
+      "epoch": 2.690291465716083,
+      "grad_norm": 1.1750140190124512,
+      "learning_rate": 0.0002,
+      "loss": 0.312,
+      "step": 12830
+    },
+    {
+      "epoch": 2.6923883413713567,
+      "grad_norm": 0.18456393480300903,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 12840
+    },
+    {
+      "epoch": 2.6944852170266302,
+      "grad_norm": 0.09328345954418182,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 12850
+    },
+    {
+      "epoch": 2.696582092681904,
+      "grad_norm": 0.05224505066871643,
+      "learning_rate": 0.0002,
+      "loss": 0.2803,
+      "step": 12860
+    },
+    {
+      "epoch": 2.6986789683371777,
+      "grad_norm": 0.07151336967945099,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 12870
+    },
+    {
+      "epoch": 2.700775843992451,
+      "grad_norm": 0.07107602059841156,
+      "learning_rate": 0.0002,
+      "loss": 0.279,
+      "step": 12880
+    },
+    {
+      "epoch": 2.7028727196477247,
+      "grad_norm": 0.06088366359472275,
+      "learning_rate": 0.0002,
+      "loss": 0.2707,
+      "step": 12890
+    },
+    {
+      "epoch": 2.7049695953029986,
+      "grad_norm": 0.08219806104898453,
+      "learning_rate": 0.0002,
+      "loss": 0.3286,
+      "step": 12900
+    },
+    {
+      "epoch": 2.707066470958272,
+      "grad_norm": 0.4720647633075714,
+      "learning_rate": 0.0002,
+      "loss": 0.2994,
+      "step": 12910
+    },
+    {
+      "epoch": 2.709163346613546,
+      "grad_norm": 0.078771211206913,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 12920
+    },
+    {
+      "epoch": 2.7112602222688196,
+      "grad_norm": 3.5635182857513428,
+      "learning_rate": 0.0002,
+      "loss": 0.2775,
+      "step": 12930
+    },
+    {
+      "epoch": 2.713357097924093,
+      "grad_norm": 0.09435427188873291,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 12940
+    },
+    {
+      "epoch": 2.7154539735793666,
+      "grad_norm": 3.500767946243286,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 12950
+    },
+    {
+      "epoch": 2.7175508492346405,
+      "grad_norm": 0.08721265196800232,
+      "learning_rate": 0.0002,
+      "loss": 0.2891,
+      "step": 12960
+    },
+    {
+      "epoch": 2.719647724889914,
+      "grad_norm": 2.141937255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.2708,
+      "step": 12970
+    },
+    {
+      "epoch": 2.7217446005451875,
+      "grad_norm": 0.10957151651382446,
+      "learning_rate": 0.0002,
+      "loss": 0.3073,
+      "step": 12980
+    },
+    {
+      "epoch": 2.7238414762004615,
+      "grad_norm": 0.2539743185043335,
+      "learning_rate": 0.0002,
+      "loss": 0.2466,
+      "step": 12990
+    },
+    {
+      "epoch": 2.725938351855735,
+      "grad_norm": 5.60904598236084,
+      "learning_rate": 0.0002,
+      "loss": 0.2514,
+      "step": 13000
+    },
+    {
+      "epoch": 2.7280352275110085,
+      "grad_norm": 0.9625059962272644,
+      "learning_rate": 0.0002,
+      "loss": 0.2513,
+      "step": 13010
+    },
+    {
+      "epoch": 2.730132103166282,
+      "grad_norm": 10.193999290466309,
+      "learning_rate": 0.0002,
+      "loss": 0.3065,
+      "step": 13020
+    },
+    {
+      "epoch": 2.732228978821556,
+      "grad_norm": 0.7805821299552917,
+      "learning_rate": 0.0002,
+      "loss": 0.2819,
+      "step": 13030
+    },
+    {
+      "epoch": 2.7343258544768294,
+      "grad_norm": 0.08317407965660095,
+      "learning_rate": 0.0002,
+      "loss": 0.2425,
+      "step": 13040
+    },
+    {
+      "epoch": 2.7364227301321034,
+      "grad_norm": 0.12909261882305145,
+      "learning_rate": 0.0002,
+      "loss": 0.3156,
+      "step": 13050
+    },
+    {
+      "epoch": 2.738519605787377,
+      "grad_norm": 0.16666528582572937,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 13060
+    },
+    {
+      "epoch": 2.7406164814426504,
+      "grad_norm": 0.3270515501499176,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 13070
+    },
+    {
+      "epoch": 2.742713357097924,
+      "grad_norm": 0.08127190172672272,
+      "learning_rate": 0.0002,
+      "loss": 0.2737,
+      "step": 13080
+    },
+    {
+      "epoch": 2.744810232753198,
+      "grad_norm": 0.09699930250644684,
+      "learning_rate": 0.0002,
+      "loss": 0.2668,
+      "step": 13090
+    },
+    {
+      "epoch": 2.7469071084084713,
+      "grad_norm": 2.5660605430603027,
+      "learning_rate": 0.0002,
+      "loss": 0.2529,
+      "step": 13100
+    },
+    {
+      "epoch": 2.7490039840637452,
+      "grad_norm": 0.35261228680610657,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 13110
+    },
+    {
+      "epoch": 2.7511008597190187,
+      "grad_norm": 0.1015433669090271,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 13120
+    },
+    {
+      "epoch": 2.7531977353742922,
+      "grad_norm": 1.886006236076355,
+      "learning_rate": 0.0002,
+      "loss": 0.2649,
+      "step": 13130
+    },
+    {
+      "epoch": 2.7552946110295657,
+      "grad_norm": 0.1277870535850525,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 13140
+    },
+    {
+      "epoch": 2.7573914866848397,
+      "grad_norm": 0.0975937694311142,
+      "learning_rate": 0.0002,
+      "loss": 0.3081,
+      "step": 13150
+    },
+    {
+      "epoch": 2.759488362340113,
+      "grad_norm": 0.08423218876123428,
+      "learning_rate": 0.0002,
+      "loss": 0.2688,
+      "step": 13160
+    },
+    {
+      "epoch": 2.761585237995387,
+      "grad_norm": 4.316064834594727,
+      "learning_rate": 0.0002,
+      "loss": 0.303,
+      "step": 13170
+    },
+    {
+      "epoch": 2.7636821136506606,
+      "grad_norm": 0.2076009064912796,
+      "learning_rate": 0.0002,
+      "loss": 0.2469,
+      "step": 13180
+    },
+    {
+      "epoch": 2.765778989305934,
+      "grad_norm": 0.5169063806533813,
+      "learning_rate": 0.0002,
+      "loss": 0.2599,
+      "step": 13190
+    },
+    {
+      "epoch": 2.7678758649612076,
+      "grad_norm": 0.09089035540819168,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 13200
+    },
+    {
+      "epoch": 2.7699727406164816,
+      "grad_norm": 0.40360260009765625,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 13210
+    },
+    {
+      "epoch": 2.772069616271755,
+      "grad_norm": 0.09568203240633011,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 13220
+    },
+    {
+      "epoch": 2.7741664919270286,
+      "grad_norm": 0.08937239646911621,
+      "learning_rate": 0.0002,
+      "loss": 0.3118,
+      "step": 13230
+    },
+    {
+      "epoch": 2.7762633675823025,
+      "grad_norm": 0.1122610941529274,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 13240
+    },
+    {
+      "epoch": 2.778360243237576,
+      "grad_norm": 0.07057496905326843,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 13250
+    },
+    {
+      "epoch": 2.7804571188928495,
+      "grad_norm": 8.3790864944458,
+      "learning_rate": 0.0002,
+      "loss": 0.3083,
+      "step": 13260
+    },
+    {
+      "epoch": 2.782553994548123,
+      "grad_norm": 5.793963432312012,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 13270
+    },
+    {
+      "epoch": 2.784650870203397,
+      "grad_norm": 0.05257759243249893,
+      "learning_rate": 0.0002,
+      "loss": 0.2607,
+      "step": 13280
+    },
+    {
+      "epoch": 2.7867477458586705,
+      "grad_norm": 0.08371423929929733,
+      "learning_rate": 0.0002,
+      "loss": 0.2806,
+      "step": 13290
+    },
+    {
+      "epoch": 2.7888446215139444,
+      "grad_norm": 3.869866371154785,
+      "learning_rate": 0.0002,
+      "loss": 0.3398,
+      "step": 13300
+    },
+    {
+      "epoch": 2.790941497169218,
+      "grad_norm": 4.551204681396484,
+      "learning_rate": 0.0002,
+      "loss": 0.2745,
+      "step": 13310
+    },
+    {
+      "epoch": 2.7930383728244914,
+      "grad_norm": 0.10620396584272385,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 13320
+    },
+    {
+      "epoch": 2.795135248479765,
+      "grad_norm": 0.09897986054420471,
+      "learning_rate": 0.0002,
+      "loss": 0.3427,
+      "step": 13330
+    },
+    {
+      "epoch": 2.797232124135039,
+      "grad_norm": 0.07497262209653854,
+      "learning_rate": 0.0002,
+      "loss": 0.2584,
+      "step": 13340
+    },
+    {
+      "epoch": 2.7993289997903124,
+      "grad_norm": 0.09139002114534378,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 13350
+    },
+    {
+      "epoch": 2.8014258754455863,
+      "grad_norm": 1.2047020196914673,
+      "learning_rate": 0.0002,
+      "loss": 0.2744,
+      "step": 13360
+    },
+    {
+      "epoch": 2.80352275110086,
+      "grad_norm": 0.08489953726530075,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 13370
+    },
+    {
+      "epoch": 2.8056196267561333,
+      "grad_norm": 0.1599905788898468,
+      "learning_rate": 0.0002,
+      "loss": 0.2751,
+      "step": 13380
+    },
+    {
+      "epoch": 2.807716502411407,
+      "grad_norm": 0.06880758702754974,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 13390
+    },
+    {
+      "epoch": 2.8098133780666807,
+      "grad_norm": 0.08114618808031082,
+      "learning_rate": 0.0002,
+      "loss": 0.274,
+      "step": 13400
+    },
+    {
+      "epoch": 2.8119102537219542,
+      "grad_norm": 0.060263942927122116,
+      "learning_rate": 0.0002,
+      "loss": 0.2719,
+      "step": 13410
+    },
+    {
+      "epoch": 2.814007129377228,
+      "grad_norm": 0.1302911937236786,
+      "learning_rate": 0.0002,
+      "loss": 0.2825,
+      "step": 13420
+    },
+    {
+      "epoch": 2.8161040050325017,
+      "grad_norm": 0.04744647815823555,
+      "learning_rate": 0.0002,
+      "loss": 0.2431,
+      "step": 13430
+    },
+    {
+      "epoch": 2.818200880687775,
+      "grad_norm": 0.05303800106048584,
+      "learning_rate": 0.0002,
+      "loss": 0.2431,
+      "step": 13440
+    },
+    {
+      "epoch": 2.8202977563430487,
+      "grad_norm": 4.2985687255859375,
+      "learning_rate": 0.0002,
+      "loss": 0.2639,
+      "step": 13450
+    },
+    {
+      "epoch": 2.8223946319983226,
+      "grad_norm": 0.41262736916542053,
+      "learning_rate": 0.0002,
+      "loss": 0.2827,
+      "step": 13460
+    },
+    {
+      "epoch": 2.824491507653596,
+      "grad_norm": 5.396834850311279,
+      "learning_rate": 0.0002,
+      "loss": 0.2976,
+      "step": 13470
+    },
+    {
+      "epoch": 2.8265883833088696,
+      "grad_norm": 0.10336003452539444,
+      "learning_rate": 0.0002,
+      "loss": 0.2737,
+      "step": 13480
+    },
+    {
+      "epoch": 2.8286852589641436,
+      "grad_norm": 0.05884670093655586,
+      "learning_rate": 0.0002,
+      "loss": 0.2598,
+      "step": 13490
+    },
+    {
+      "epoch": 2.830782134619417,
+      "grad_norm": 4.840988636016846,
+      "learning_rate": 0.0002,
+      "loss": 0.259,
+      "step": 13500
+    },
+    {
+      "epoch": 2.8328790102746906,
+      "grad_norm": 0.30140170454978943,
+      "learning_rate": 0.0002,
+      "loss": 0.2752,
+      "step": 13510
+    },
+    {
+      "epoch": 2.834975885929964,
+      "grad_norm": 0.1117088794708252,
+      "learning_rate": 0.0002,
+      "loss": 0.2429,
+      "step": 13520
+    },
+    {
+      "epoch": 2.837072761585238,
+      "grad_norm": 0.054389581084251404,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 13530
+    },
+    {
+      "epoch": 2.8391696372405115,
+      "grad_norm": 0.07770632952451706,
+      "learning_rate": 0.0002,
+      "loss": 0.2841,
+      "step": 13540
+    },
+    {
+      "epoch": 2.8412665128957855,
+      "grad_norm": 0.10165666043758392,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 13550
+    },
+    {
+      "epoch": 2.843363388551059,
+      "grad_norm": 0.12631669640541077,
+      "learning_rate": 0.0002,
+      "loss": 0.2715,
+      "step": 13560
+    },
+    {
+      "epoch": 2.8454602642063325,
+      "grad_norm": 0.08221318572759628,
+      "learning_rate": 0.0002,
+      "loss": 0.2434,
+      "step": 13570
+    },
+    {
+      "epoch": 2.847557139861606,
+      "grad_norm": 0.07879089564085007,
+      "learning_rate": 0.0002,
+      "loss": 0.2644,
+      "step": 13580
+    },
+    {
+      "epoch": 2.84965401551688,
+      "grad_norm": 0.18776120245456696,
+      "learning_rate": 0.0002,
+      "loss": 0.2761,
+      "step": 13590
+    },
+    {
+      "epoch": 2.8517508911721534,
+      "grad_norm": 0.10581996291875839,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 13600
+    },
+    {
+      "epoch": 2.8538477668274274,
+      "grad_norm": 0.08558280020952225,
+      "learning_rate": 0.0002,
+      "loss": 0.2767,
+      "step": 13610
+    },
+    {
+      "epoch": 2.855944642482701,
+      "grad_norm": 0.06981467455625534,
+      "learning_rate": 0.0002,
+      "loss": 0.2767,
+      "step": 13620
+    },
+    {
+      "epoch": 2.8580415181379744,
+      "grad_norm": 0.08355915546417236,
+      "learning_rate": 0.0002,
+      "loss": 0.3006,
+      "step": 13630
+    },
+    {
+      "epoch": 2.860138393793248,
+      "grad_norm": 0.06184757128357887,
+      "learning_rate": 0.0002,
+      "loss": 0.2846,
+      "step": 13640
+    },
+    {
+      "epoch": 2.862235269448522,
+      "grad_norm": 0.07668596506118774,
+      "learning_rate": 0.0002,
+      "loss": 0.2558,
+      "step": 13650
+    },
+    {
+      "epoch": 2.8643321451037953,
+      "grad_norm": 0.3981834352016449,
+      "learning_rate": 0.0002,
+      "loss": 0.275,
+      "step": 13660
+    },
+    {
+      "epoch": 2.8664290207590692,
+      "grad_norm": 0.9413365125656128,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 13670
+    },
+    {
+      "epoch": 2.8685258964143427,
+      "grad_norm": 0.19572269916534424,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 13680
+    },
+    {
+      "epoch": 2.8706227720696162,
+      "grad_norm": 0.15002276003360748,
+      "learning_rate": 0.0002,
+      "loss": 0.3071,
+      "step": 13690
+    },
+    {
+      "epoch": 2.8727196477248897,
+      "grad_norm": 0.06685584038496017,
+      "learning_rate": 0.0002,
+      "loss": 0.3044,
+      "step": 13700
+    },
+    {
+      "epoch": 2.8748165233801637,
+      "grad_norm": 4.113365650177002,
+      "learning_rate": 0.0002,
+      "loss": 0.3452,
+      "step": 13710
+    },
+    {
+      "epoch": 2.876913399035437,
+      "grad_norm": 0.12345223128795624,
+      "learning_rate": 0.0002,
+      "loss": 0.3401,
+      "step": 13720
+    },
+    {
+      "epoch": 2.8790102746907107,
+      "grad_norm": 1.0276050567626953,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 13730
+    },
+    {
+      "epoch": 2.8811071503459846,
+      "grad_norm": 0.27919915318489075,
+      "learning_rate": 0.0002,
+      "loss": 0.3792,
+      "step": 13740
+    },
+    {
+      "epoch": 2.883204026001258,
+      "grad_norm": 1.4898934364318848,
+      "learning_rate": 0.0002,
+      "loss": 0.2625,
+      "step": 13750
+    },
+    {
+      "epoch": 2.8853009016565316,
+      "grad_norm": 0.1877644956111908,
+      "learning_rate": 0.0002,
+      "loss": 0.3044,
+      "step": 13760
+    },
+    {
+      "epoch": 2.887397777311805,
+      "grad_norm": 0.20642811059951782,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 13770
+    },
+    {
+      "epoch": 2.889494652967079,
+      "grad_norm": 1.3547812700271606,
+      "learning_rate": 0.0002,
+      "loss": 0.3656,
+      "step": 13780
+    },
+    {
+      "epoch": 2.8915915286223526,
+      "grad_norm": 0.8433927297592163,
+      "learning_rate": 0.0002,
+      "loss": 0.3086,
+      "step": 13790
+    },
+    {
+      "epoch": 2.8936884042776265,
+      "grad_norm": 2.5968425273895264,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 13800
+    },
+    {
+      "epoch": 2.8957852799329,
+      "grad_norm": 9.476542472839355,
+      "learning_rate": 0.0002,
+      "loss": 0.3393,
+      "step": 13810
+    },
+    {
+      "epoch": 2.8978821555881735,
+      "grad_norm": 1.2320115566253662,
+      "learning_rate": 0.0002,
+      "loss": 0.2803,
+      "step": 13820
+    },
+    {
+      "epoch": 2.899979031243447,
+      "grad_norm": 0.13934971392154694,
+      "learning_rate": 0.0002,
+      "loss": 0.248,
+      "step": 13830
+    },
+    {
+      "epoch": 2.902075906898721,
+      "grad_norm": 0.42801135778427124,
+      "learning_rate": 0.0002,
+      "loss": 0.2858,
+      "step": 13840
+    },
+    {
+      "epoch": 2.9041727825539945,
+      "grad_norm": 0.4410918056964874,
+      "learning_rate": 0.0002,
+      "loss": 0.3026,
+      "step": 13850
+    },
+    {
+      "epoch": 2.9062696582092684,
+      "grad_norm": 0.08050935715436935,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 13860
+    },
+    {
+      "epoch": 2.908366533864542,
+      "grad_norm": 0.38791462779045105,
+      "learning_rate": 0.0002,
+      "loss": 0.2902,
+      "step": 13870
+    },
+    {
+      "epoch": 2.9104634095198154,
+      "grad_norm": 7.857952117919922,
+      "learning_rate": 0.0002,
+      "loss": 0.3109,
+      "step": 13880
+    },
+    {
+      "epoch": 2.912560285175089,
+      "grad_norm": 0.7213678359985352,
+      "learning_rate": 0.0002,
+      "loss": 0.2754,
+      "step": 13890
+    },
+    {
+      "epoch": 2.914657160830363,
+      "grad_norm": 0.2854587435722351,
+      "learning_rate": 0.0002,
+      "loss": 0.3071,
+      "step": 13900
+    },
+    {
+      "epoch": 2.9167540364856364,
+      "grad_norm": 0.7336795330047607,
+      "learning_rate": 0.0002,
+      "loss": 0.278,
+      "step": 13910
+    },
+    {
+      "epoch": 2.9188509121409103,
+      "grad_norm": 0.42611584067344666,
+      "learning_rate": 0.0002,
+      "loss": 0.2723,
+      "step": 13920
+    },
+    {
+      "epoch": 2.920947787796184,
+      "grad_norm": 0.2687622606754303,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 13930
+    },
+    {
+      "epoch": 2.9230446634514573,
+      "grad_norm": 0.10501446574926376,
+      "learning_rate": 0.0002,
+      "loss": 0.2889,
+      "step": 13940
+    },
+    {
+      "epoch": 2.925141539106731,
+      "grad_norm": 6.071291923522949,
+      "learning_rate": 0.0002,
+      "loss": 0.2843,
+      "step": 13950
+    },
+    {
+      "epoch": 2.9272384147620047,
+      "grad_norm": 5.702052116394043,
+      "learning_rate": 0.0002,
+      "loss": 0.2853,
+      "step": 13960
+    },
+    {
+      "epoch": 2.9293352904172782,
+      "grad_norm": 0.8838995099067688,
+      "learning_rate": 0.0002,
+      "loss": 0.2863,
+      "step": 13970
+    },
+    {
+      "epoch": 2.9314321660725517,
+      "grad_norm": 0.11934137344360352,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 13980
+    },
+    {
+      "epoch": 2.9335290417278257,
+      "grad_norm": 1.4622691869735718,
+      "learning_rate": 0.0002,
+      "loss": 0.2466,
+      "step": 13990
+    },
+    {
+      "epoch": 2.935625917383099,
+      "grad_norm": 0.2274940013885498,
+      "learning_rate": 0.0002,
+      "loss": 0.2745,
+      "step": 14000
+    },
+    {
+      "epoch": 2.9377227930383727,
+      "grad_norm": 0.06870028376579285,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 14010
+    },
+    {
+      "epoch": 2.939819668693646,
+      "grad_norm": 13.201187133789062,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 14020
+    },
+    {
+      "epoch": 2.94191654434892,
+      "grad_norm": 0.0958949476480484,
+      "learning_rate": 0.0002,
+      "loss": 0.2621,
+      "step": 14030
+    },
+    {
+      "epoch": 2.9440134200041936,
+      "grad_norm": 1.1030700206756592,
+      "learning_rate": 0.0002,
+      "loss": 0.2474,
+      "step": 14040
+    },
+    {
+      "epoch": 2.9461102956594676,
+      "grad_norm": 3.825788736343384,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 14050
+    },
+    {
+      "epoch": 2.948207171314741,
+      "grad_norm": 0.2587536871433258,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 14060
+    },
+    {
+      "epoch": 2.9503040469700146,
+      "grad_norm": 6.049500465393066,
+      "learning_rate": 0.0002,
+      "loss": 0.3027,
+      "step": 14070
+    },
+    {
+      "epoch": 2.952400922625288,
+      "grad_norm": 0.11885102093219757,
+      "learning_rate": 0.0002,
+      "loss": 0.2429,
+      "step": 14080
+    },
+    {
+      "epoch": 2.954497798280562,
+      "grad_norm": 0.6359782814979553,
+      "learning_rate": 0.0002,
+      "loss": 0.2868,
+      "step": 14090
+    },
+    {
+      "epoch": 2.9565946739358355,
+      "grad_norm": 0.06189330294728279,
+      "learning_rate": 0.0002,
+      "loss": 0.3036,
+      "step": 14100
+    },
+    {
+      "epoch": 2.9586915495911095,
+      "grad_norm": 0.07890813052654266,
+      "learning_rate": 0.0002,
+      "loss": 0.2765,
+      "step": 14110
+    },
+    {
+      "epoch": 2.960788425246383,
+      "grad_norm": 4.272983074188232,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 14120
+    },
+    {
+      "epoch": 2.9628853009016565,
+      "grad_norm": 0.09933266788721085,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 14130
+    },
+    {
+      "epoch": 2.96498217655693,
+      "grad_norm": 0.1176319271326065,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 14140
+    },
+    {
+      "epoch": 2.967079052212204,
+      "grad_norm": 0.09006159752607346,
+      "learning_rate": 0.0002,
+      "loss": 0.2641,
+      "step": 14150
+    },
+    {
+      "epoch": 2.9691759278674774,
+      "grad_norm": 0.11632592976093292,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 14160
+    },
+    {
+      "epoch": 2.9712728035227514,
+      "grad_norm": 2.242246627807617,
+      "learning_rate": 0.0002,
+      "loss": 0.2692,
+      "step": 14170
+    },
+    {
+      "epoch": 2.973369679178025,
+      "grad_norm": 0.8853824138641357,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 14180
+    },
+    {
+      "epoch": 2.9754665548332984,
+      "grad_norm": 0.1101587638258934,
+      "learning_rate": 0.0002,
+      "loss": 0.242,
+      "step": 14190
+    },
+    {
+      "epoch": 2.977563430488572,
+      "grad_norm": 0.06838403642177582,
+      "learning_rate": 0.0002,
+      "loss": 0.3337,
+      "step": 14200
+    },
+    {
+      "epoch": 2.979660306143846,
+      "grad_norm": 0.46546003222465515,
+      "learning_rate": 0.0002,
+      "loss": 0.3376,
+      "step": 14210
+    },
+    {
+      "epoch": 2.9817571817991193,
+      "grad_norm": 0.08868476748466492,
+      "learning_rate": 0.0002,
+      "loss": 0.2678,
+      "step": 14220
+    },
+    {
+      "epoch": 2.983854057454393,
+      "grad_norm": 0.08028911799192429,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 14230
+    },
+    {
+      "epoch": 2.9859509331096667,
+      "grad_norm": 0.06891479343175888,
+      "learning_rate": 0.0002,
+      "loss": 0.2433,
+      "step": 14240
+    },
+    {
+      "epoch": 2.9880478087649402,
+      "grad_norm": 0.09133801609277725,
+      "learning_rate": 0.0002,
+      "loss": 0.2846,
+      "step": 14250
+    },
+    {
+      "epoch": 2.9901446844202137,
+      "grad_norm": 0.058840859681367874,
+      "learning_rate": 0.0002,
+      "loss": 0.266,
+      "step": 14260
+    },
+    {
+      "epoch": 2.9922415600754873,
+      "grad_norm": 0.08534104377031326,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 14270
+    },
+    {
+      "epoch": 2.994338435730761,
+      "grad_norm": 4.0780720710754395,
+      "learning_rate": 0.0002,
+      "loss": 0.4004,
+      "step": 14280
+    },
+    {
+      "epoch": 2.9964353113860347,
+      "grad_norm": 0.25896960496902466,
+      "learning_rate": 0.0002,
+      "loss": 0.3645,
+      "step": 14290
+    },
+    {
+      "epoch": 2.9985321870413086,
+      "grad_norm": 0.08144643902778625,
+      "learning_rate": 0.0002,
+      "loss": 0.2427,
+      "step": 14300
+    },
+    {
+      "epoch": 3.000629062696582,
+      "grad_norm": 0.06053667888045311,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 14310
+    },
+    {
+      "epoch": 3.0027259383518556,
+      "grad_norm": 0.12006381154060364,
+      "learning_rate": 0.0002,
+      "loss": 0.3749,
+      "step": 14320
+    },
+    {
+      "epoch": 3.0048228140071296,
+      "grad_norm": 0.09352512657642365,
+      "learning_rate": 0.0002,
+      "loss": 0.2993,
+      "step": 14330
+    },
+    {
+      "epoch": 3.006919689662403,
+      "grad_norm": 0.06361488997936249,
+      "learning_rate": 0.0002,
+      "loss": 0.2438,
+      "step": 14340
+    },
+    {
+      "epoch": 3.0090165653176766,
+      "grad_norm": 0.11830562353134155,
+      "learning_rate": 0.0002,
+      "loss": 0.3046,
+      "step": 14350
+    },
+    {
+      "epoch": 3.01111344097295,
+      "grad_norm": 0.053154319524765015,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 14360
+    },
+    {
+      "epoch": 3.013210316628224,
+      "grad_norm": 0.05998919904232025,
+      "learning_rate": 0.0002,
+      "loss": 0.2995,
+      "step": 14370
+    },
+    {
+      "epoch": 3.0153071922834975,
+      "grad_norm": 0.10642017424106598,
+      "learning_rate": 0.0002,
+      "loss": 0.2877,
+      "step": 14380
+    },
+    {
+      "epoch": 3.017404067938771,
+      "grad_norm": 0.13320499658584595,
+      "learning_rate": 0.0002,
+      "loss": 0.2906,
+      "step": 14390
+    },
+    {
+      "epoch": 3.019500943594045,
+      "grad_norm": 6.584969997406006,
+      "learning_rate": 0.0002,
+      "loss": 0.3778,
+      "step": 14400
+    },
+    {
+      "epoch": 3.0215978192493185,
+      "grad_norm": 7.320637226104736,
+      "learning_rate": 0.0002,
+      "loss": 0.2791,
+      "step": 14410
+    },
+    {
+      "epoch": 3.023694694904592,
+      "grad_norm": 0.06931141018867493,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 14420
+    },
+    {
+      "epoch": 3.025791570559866,
+      "grad_norm": 0.1134222224354744,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 14430
+    },
+    {
+      "epoch": 3.0278884462151394,
+      "grad_norm": 0.057289376854896545,
+      "learning_rate": 0.0002,
+      "loss": 0.2887,
+      "step": 14440
+    },
+    {
+      "epoch": 3.029985321870413,
+      "grad_norm": 0.1048545092344284,
+      "learning_rate": 0.0002,
+      "loss": 0.2821,
+      "step": 14450
+    },
+    {
+      "epoch": 3.032082197525687,
+      "grad_norm": 0.09596660733222961,
+      "learning_rate": 0.0002,
+      "loss": 0.2749,
+      "step": 14460
+    },
+    {
+      "epoch": 3.0341790731809604,
+      "grad_norm": 0.5900943279266357,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 14470
+    },
+    {
+      "epoch": 3.036275948836234,
+      "grad_norm": 6.8857951164245605,
+      "learning_rate": 0.0002,
+      "loss": 0.2466,
+      "step": 14480
+    },
+    {
+      "epoch": 3.038372824491508,
+      "grad_norm": 0.08555381000041962,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 14490
+    },
+    {
+      "epoch": 3.0404697001467813,
+      "grad_norm": 0.07776056975126266,
+      "learning_rate": 0.0002,
+      "loss": 0.3113,
+      "step": 14500
+    },
+    {
+      "epoch": 3.042566575802055,
+      "grad_norm": 10.093559265136719,
+      "learning_rate": 0.0002,
+      "loss": 0.3165,
+      "step": 14510
+    },
+    {
+      "epoch": 3.0446634514573288,
+      "grad_norm": 0.07289782166481018,
+      "learning_rate": 0.0002,
+      "loss": 0.2988,
+      "step": 14520
+    },
+    {
+      "epoch": 3.0467603271126023,
+      "grad_norm": 0.048002567142248154,
+      "learning_rate": 0.0002,
+      "loss": 0.2748,
+      "step": 14530
+    },
+    {
+      "epoch": 3.0488572027678758,
+      "grad_norm": 5.174429416656494,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 14540
+    },
+    {
+      "epoch": 3.0509540784231497,
+      "grad_norm": 0.0720122829079628,
+      "learning_rate": 0.0002,
+      "loss": 0.2625,
+      "step": 14550
+    },
+    {
+      "epoch": 3.053050954078423,
+      "grad_norm": 0.08649764955043793,
+      "learning_rate": 0.0002,
+      "loss": 0.2573,
+      "step": 14560
+    },
+    {
+      "epoch": 3.0551478297336967,
+      "grad_norm": 0.045956335961818695,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 14570
+    },
+    {
+      "epoch": 3.0572447053889706,
+      "grad_norm": 0.07197452336549759,
+      "learning_rate": 0.0002,
+      "loss": 0.2651,
+      "step": 14580
+    },
+    {
+      "epoch": 3.059341581044244,
+      "grad_norm": 0.0696483701467514,
+      "learning_rate": 0.0002,
+      "loss": 0.2435,
+      "step": 14590
+    },
+    {
+      "epoch": 3.0614384566995176,
+      "grad_norm": 0.054920099675655365,
+      "learning_rate": 0.0002,
+      "loss": 0.3082,
+      "step": 14600
+    },
+    {
+      "epoch": 3.0635353323547916,
+      "grad_norm": 0.0675140768289566,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 14610
+    },
+    {
+      "epoch": 3.065632208010065,
+      "grad_norm": 0.11363296955823898,
+      "learning_rate": 0.0002,
+      "loss": 0.2874,
+      "step": 14620
+    },
+    {
+      "epoch": 3.0677290836653386,
+      "grad_norm": 0.051395099610090256,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 14630
+    },
+    {
+      "epoch": 3.069825959320612,
+      "grad_norm": 0.06727784126996994,
+      "learning_rate": 0.0002,
+      "loss": 0.2772,
+      "step": 14640
+    },
+    {
+      "epoch": 3.071922834975886,
+      "grad_norm": 0.08824780583381653,
+      "learning_rate": 0.0002,
+      "loss": 0.2637,
+      "step": 14650
+    },
+    {
+      "epoch": 3.0740197106311595,
+      "grad_norm": 0.06606782227754593,
+      "learning_rate": 0.0002,
+      "loss": 0.2681,
+      "step": 14660
+    },
+    {
+      "epoch": 3.076116586286433,
+      "grad_norm": 0.05145835131406784,
+      "learning_rate": 0.0002,
+      "loss": 0.2646,
+      "step": 14670
+    },
+    {
+      "epoch": 3.078213461941707,
+      "grad_norm": 0.16860048472881317,
+      "learning_rate": 0.0002,
+      "loss": 0.2554,
+      "step": 14680
+    },
+    {
+      "epoch": 3.0803103375969805,
+      "grad_norm": 0.08995959162712097,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 14690
+    },
+    {
+      "epoch": 3.082407213252254,
+      "grad_norm": 0.08270885050296783,
+      "learning_rate": 0.0002,
+      "loss": 0.2884,
+      "step": 14700
+    },
+    {
+      "epoch": 3.084504088907528,
+      "grad_norm": 5.329971790313721,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 14710
+    },
+    {
+      "epoch": 3.0866009645628014,
+      "grad_norm": 0.37981072068214417,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 14720
+    },
+    {
+      "epoch": 3.088697840218075,
+      "grad_norm": 0.09584033489227295,
+      "learning_rate": 0.0002,
+      "loss": 0.2905,
+      "step": 14730
+    },
+    {
+      "epoch": 3.090794715873349,
+      "grad_norm": 0.0819215402007103,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 14740
+    },
+    {
+      "epoch": 3.0928915915286224,
+      "grad_norm": 0.11679327487945557,
+      "learning_rate": 0.0002,
+      "loss": 0.2586,
+      "step": 14750
+    },
+    {
+      "epoch": 3.094988467183896,
+      "grad_norm": 0.08995963633060455,
+      "learning_rate": 0.0002,
+      "loss": 0.2579,
+      "step": 14760
+    },
+    {
+      "epoch": 3.09708534283917,
+      "grad_norm": 0.04532779008150101,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 14770
+    },
+    {
+      "epoch": 3.0991822184944433,
+      "grad_norm": 0.09251581877470016,
+      "learning_rate": 0.0002,
+      "loss": 0.2437,
+      "step": 14780
+    },
+    {
+      "epoch": 3.101279094149717,
+      "grad_norm": 0.06317158788442612,
+      "learning_rate": 0.0002,
+      "loss": 0.2427,
+      "step": 14790
+    },
+    {
+      "epoch": 3.1033759698049908,
+      "grad_norm": 0.07175248861312866,
+      "learning_rate": 0.0002,
+      "loss": 0.2825,
+      "step": 14800
+    },
+    {
+      "epoch": 3.1054728454602643,
+      "grad_norm": 0.05058152601122856,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 14810
+    },
+    {
+      "epoch": 3.1075697211155378,
+      "grad_norm": 0.07968831807374954,
+      "learning_rate": 0.0002,
+      "loss": 0.2439,
+      "step": 14820
+    },
+    {
+      "epoch": 3.1096665967708117,
+      "grad_norm": 0.07660433650016785,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 14830
+    },
+    {
+      "epoch": 3.111763472426085,
+      "grad_norm": 0.055322594940662384,
+      "learning_rate": 0.0002,
+      "loss": 0.2937,
+      "step": 14840
+    },
+    {
+      "epoch": 3.1138603480813587,
+      "grad_norm": 0.15317735075950623,
+      "learning_rate": 0.0002,
+      "loss": 0.2693,
+      "step": 14850
+    },
+    {
+      "epoch": 3.1159572237366326,
+      "grad_norm": 0.0666760727763176,
+      "learning_rate": 0.0002,
+      "loss": 0.244,
+      "step": 14860
+    },
+    {
+      "epoch": 3.118054099391906,
+      "grad_norm": 0.07330338656902313,
+      "learning_rate": 0.0002,
+      "loss": 0.2583,
+      "step": 14870
+    },
+    {
+      "epoch": 3.1201509750471796,
+      "grad_norm": 4.008168697357178,
+      "learning_rate": 0.0002,
+      "loss": 0.2798,
+      "step": 14880
+    },
+    {
+      "epoch": 3.122247850702453,
+      "grad_norm": 11.84869384765625,
+      "learning_rate": 0.0002,
+      "loss": 0.3852,
+      "step": 14890
+    },
+    {
+      "epoch": 3.124344726357727,
+      "grad_norm": 1.1625456809997559,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 14900
+    },
+    {
+      "epoch": 3.1264416020130006,
+      "grad_norm": 3.9291110038757324,
+      "learning_rate": 0.0002,
+      "loss": 0.3082,
+      "step": 14910
+    },
+    {
+      "epoch": 3.128538477668274,
+      "grad_norm": 0.0719500258564949,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 14920
+    },
+    {
+      "epoch": 3.130635353323548,
+      "grad_norm": 8.359219551086426,
+      "learning_rate": 0.0002,
+      "loss": 0.2796,
+      "step": 14930
+    },
+    {
+      "epoch": 3.1327322289788215,
+      "grad_norm": 0.14066196978092194,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 14940
+    },
+    {
+      "epoch": 3.134829104634095,
+      "grad_norm": 0.09063920378684998,
+      "learning_rate": 0.0002,
+      "loss": 0.2424,
+      "step": 14950
+    },
+    {
+      "epoch": 3.136925980289369,
+      "grad_norm": 0.15819169580936432,
+      "learning_rate": 0.0002,
+      "loss": 0.2681,
+      "step": 14960
+    },
+    {
+      "epoch": 3.1390228559446425,
+      "grad_norm": 0.06944765895605087,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 14970
+    },
+    {
+      "epoch": 3.141119731599916,
+      "grad_norm": 0.1226600632071495,
+      "learning_rate": 0.0002,
+      "loss": 0.2826,
+      "step": 14980
+    },
+    {
+      "epoch": 3.14321660725519,
+      "grad_norm": 0.2014627605676651,
+      "learning_rate": 0.0002,
+      "loss": 0.2977,
+      "step": 14990
+    },
+    {
+      "epoch": 3.1453134829104634,
+      "grad_norm": 4.710205078125,
+      "learning_rate": 0.0002,
+      "loss": 0.2617,
+      "step": 15000
+    },
+    {
+      "epoch": 3.147410358565737,
+      "grad_norm": 0.10994257032871246,
+      "learning_rate": 0.0002,
+      "loss": 0.2661,
+      "step": 15010
+    },
+    {
+      "epoch": 3.149507234221011,
+      "grad_norm": 0.8386690616607666,
+      "learning_rate": 0.0002,
+      "loss": 0.248,
+      "step": 15020
+    },
+    {
+      "epoch": 3.1516041098762844,
+      "grad_norm": 0.20938342809677124,
+      "learning_rate": 0.0002,
+      "loss": 0.31,
+      "step": 15030
+    },
+    {
+      "epoch": 3.153700985531558,
+      "grad_norm": 1.402302622795105,
+      "learning_rate": 0.0002,
+      "loss": 0.2876,
+      "step": 15040
+    },
+    {
+      "epoch": 3.155797861186832,
+      "grad_norm": 1.6246540546417236,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 15050
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.1873074173927307,
+      "learning_rate": 0.0002,
+      "loss": 0.3074,
+      "step": 15060
+    },
+    {
+      "epoch": 3.159991612497379,
+      "grad_norm": 6.368509769439697,
+      "learning_rate": 0.0002,
+      "loss": 0.3141,
+      "step": 15070
+    },
+    {
+      "epoch": 3.1620884881526528,
+      "grad_norm": 0.2556443512439728,
+      "learning_rate": 0.0002,
+      "loss": 0.2615,
+      "step": 15080
+    },
+    {
+      "epoch": 3.1641853638079263,
+      "grad_norm": 3.358315944671631,
+      "learning_rate": 0.0002,
+      "loss": 0.272,
+      "step": 15090
+    },
+    {
+      "epoch": 3.1662822394631998,
+      "grad_norm": 0.13668504357337952,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 15100
+    },
+    {
+      "epoch": 3.1683791151184737,
+      "grad_norm": 1.0130856037139893,
+      "learning_rate": 0.0002,
+      "loss": 0.2415,
+      "step": 15110
+    },
+    {
+      "epoch": 3.170475990773747,
+      "grad_norm": 0.08739043772220612,
+      "learning_rate": 0.0002,
+      "loss": 0.2918,
+      "step": 15120
+    },
+    {
+      "epoch": 3.1725728664290207,
+      "grad_norm": 0.7321306467056274,
+      "learning_rate": 0.0002,
+      "loss": 0.2831,
+      "step": 15130
+    },
+    {
+      "epoch": 3.174669742084294,
+      "grad_norm": 0.18466271460056305,
+      "learning_rate": 0.0002,
+      "loss": 0.2803,
+      "step": 15140
+    },
+    {
+      "epoch": 3.176766617739568,
+      "grad_norm": 0.37699639797210693,
+      "learning_rate": 0.0002,
+      "loss": 0.2594,
+      "step": 15150
+    },
+    {
+      "epoch": 3.1788634933948416,
+      "grad_norm": 5.0264763832092285,
+      "learning_rate": 0.0002,
+      "loss": 0.3041,
+      "step": 15160
+    },
+    {
+      "epoch": 3.180960369050115,
+      "grad_norm": 0.1934100240468979,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 15170
+    },
+    {
+      "epoch": 3.183057244705389,
+      "grad_norm": 0.09424150735139847,
+      "learning_rate": 0.0002,
+      "loss": 0.3119,
+      "step": 15180
+    },
+    {
+      "epoch": 3.1851541203606626,
+      "grad_norm": 0.05102267488837242,
+      "learning_rate": 0.0002,
+      "loss": 0.243,
+      "step": 15190
+    },
+    {
+      "epoch": 3.187250996015936,
+      "grad_norm": 0.09043313562870026,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 15200
+    },
+    {
+      "epoch": 3.18934787167121,
+      "grad_norm": 10.404960632324219,
+      "learning_rate": 0.0002,
+      "loss": 0.3076,
+      "step": 15210
+    },
+    {
+      "epoch": 3.1914447473264835,
+      "grad_norm": 9.10566234588623,
+      "learning_rate": 0.0002,
+      "loss": 0.352,
+      "step": 15220
+    },
+    {
+      "epoch": 3.193541622981757,
+      "grad_norm": 0.09223295748233795,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 15230
+    },
+    {
+      "epoch": 3.195638498637031,
+      "grad_norm": 0.12465586513280869,
+      "learning_rate": 0.0002,
+      "loss": 0.245,
+      "step": 15240
+    },
+    {
+      "epoch": 3.1977353742923045,
+      "grad_norm": 9.97639274597168,
+      "learning_rate": 0.0002,
+      "loss": 0.2801,
+      "step": 15250
+    },
+    {
+      "epoch": 3.199832249947578,
+      "grad_norm": 0.1685313880443573,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 15260
+    },
+    {
+      "epoch": 3.201929125602852,
+      "grad_norm": 0.2190469652414322,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 15270
+    },
+    {
+      "epoch": 3.2040260012581254,
+      "grad_norm": 7.2927775382995605,
+      "learning_rate": 0.0002,
+      "loss": 0.2772,
+      "step": 15280
+    },
+    {
+      "epoch": 3.206122876913399,
+      "grad_norm": 5.700809478759766,
+      "learning_rate": 0.0002,
+      "loss": 0.3131,
+      "step": 15290
+    },
+    {
+      "epoch": 3.208219752568673,
+      "grad_norm": 1.4845271110534668,
+      "learning_rate": 0.0002,
+      "loss": 0.3096,
+      "step": 15300
+    },
+    {
+      "epoch": 3.2103166282239464,
+      "grad_norm": 0.08886919170618057,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 15310
+    },
+    {
+      "epoch": 3.21241350387922,
+      "grad_norm": 0.11333826929330826,
+      "learning_rate": 0.0002,
+      "loss": 0.2849,
+      "step": 15320
+    },
+    {
+      "epoch": 3.214510379534494,
+      "grad_norm": 1.2613693475723267,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 15330
+    },
+    {
+      "epoch": 3.2166072551897673,
+      "grad_norm": 0.1572788953781128,
+      "learning_rate": 0.0002,
+      "loss": 0.28,
+      "step": 15340
+    },
+    {
+      "epoch": 3.218704130845041,
+      "grad_norm": 0.9039869904518127,
+      "learning_rate": 0.0002,
+      "loss": 0.2768,
+      "step": 15350
+    },
+    {
+      "epoch": 3.2208010065003148,
+      "grad_norm": 0.3126716613769531,
+      "learning_rate": 0.0002,
+      "loss": 0.2663,
+      "step": 15360
+    },
+    {
+      "epoch": 3.2228978821555883,
+      "grad_norm": 0.44652101397514343,
+      "learning_rate": 0.0002,
+      "loss": 0.3121,
+      "step": 15370
+    },
+    {
+      "epoch": 3.2249947578108618,
+      "grad_norm": 0.6988358497619629,
+      "learning_rate": 0.0002,
+      "loss": 0.2734,
+      "step": 15380
+    },
+    {
+      "epoch": 3.2270916334661353,
+      "grad_norm": 7.399543285369873,
+      "learning_rate": 0.0002,
+      "loss": 0.2515,
+      "step": 15390
+    },
+    {
+      "epoch": 3.229188509121409,
+      "grad_norm": 0.9261417984962463,
+      "learning_rate": 0.0002,
+      "loss": 0.2469,
+      "step": 15400
+    },
+    {
+      "epoch": 3.2312853847766827,
+      "grad_norm": 0.5016912817955017,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 15410
+    },
+    {
+      "epoch": 3.233382260431956,
+      "grad_norm": 16.135902404785156,
+      "learning_rate": 0.0002,
+      "loss": 0.275,
+      "step": 15420
+    },
+    {
+      "epoch": 3.23547913608723,
+      "grad_norm": 23.701261520385742,
+      "learning_rate": 0.0002,
+      "loss": 0.2501,
+      "step": 15430
+    },
+    {
+      "epoch": 3.2375760117425036,
+      "grad_norm": 5.819489479064941,
+      "learning_rate": 0.0002,
+      "loss": 0.3584,
+      "step": 15440
+    },
+    {
+      "epoch": 3.239672887397777,
+      "grad_norm": 5.641313076019287,
+      "learning_rate": 0.0002,
+      "loss": 0.3476,
+      "step": 15450
+    },
+    {
+      "epoch": 3.241769763053051,
+      "grad_norm": 0.2730180025100708,
+      "learning_rate": 0.0002,
+      "loss": 0.3157,
+      "step": 15460
+    },
+    {
+      "epoch": 3.2438666387083246,
+      "grad_norm": 2.8253672122955322,
+      "learning_rate": 0.0002,
+      "loss": 0.2499,
+      "step": 15470
+    },
+    {
+      "epoch": 3.245963514363598,
+      "grad_norm": 0.17094270884990692,
+      "learning_rate": 0.0002,
+      "loss": 0.2804,
+      "step": 15480
+    },
+    {
+      "epoch": 3.248060390018872,
+      "grad_norm": 0.3280426263809204,
+      "learning_rate": 0.0002,
+      "loss": 0.2534,
+      "step": 15490
+    },
+    {
+      "epoch": 3.2501572656741455,
+      "grad_norm": 2.55387806892395,
+      "learning_rate": 0.0002,
+      "loss": 0.329,
+      "step": 15500
+    },
+    {
+      "epoch": 3.252254141329419,
+      "grad_norm": 1.651965856552124,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 15510
+    },
+    {
+      "epoch": 3.254351016984693,
+      "grad_norm": 4.471637725830078,
+      "learning_rate": 0.0002,
+      "loss": 0.2505,
+      "step": 15520
+    },
+    {
+      "epoch": 3.2564478926399665,
+      "grad_norm": 0.9797324538230896,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 15530
+    },
+    {
+      "epoch": 3.25854476829524,
+      "grad_norm": 1.2143648862838745,
+      "learning_rate": 0.0002,
+      "loss": 0.3634,
+      "step": 15540
+    },
+    {
+      "epoch": 3.260641643950514,
+      "grad_norm": 7.796568393707275,
+      "learning_rate": 0.0002,
+      "loss": 0.3131,
+      "step": 15550
+    },
+    {
+      "epoch": 3.2627385196057874,
+      "grad_norm": 2.383347511291504,
+      "learning_rate": 0.0002,
+      "loss": 0.2947,
+      "step": 15560
+    },
+    {
+      "epoch": 3.264835395261061,
+      "grad_norm": 0.2774340808391571,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 15570
+    },
+    {
+      "epoch": 3.2669322709163344,
+      "grad_norm": 0.2396259754896164,
+      "learning_rate": 0.0002,
+      "loss": 0.2837,
+      "step": 15580
+    },
+    {
+      "epoch": 3.2690291465716084,
+      "grad_norm": 2.832759141921997,
+      "learning_rate": 0.0002,
+      "loss": 0.2479,
+      "step": 15590
+    },
+    {
+      "epoch": 3.271126022226882,
+      "grad_norm": 2.0031135082244873,
+      "learning_rate": 0.0002,
+      "loss": 0.2445,
+      "step": 15600
+    },
+    {
+      "epoch": 3.273222897882156,
+      "grad_norm": 0.11969097703695297,
+      "learning_rate": 0.0002,
+      "loss": 0.2664,
+      "step": 15610
+    },
+    {
+      "epoch": 3.2753197735374293,
+      "grad_norm": 0.3683199882507324,
+      "learning_rate": 0.0002,
+      "loss": 0.248,
+      "step": 15620
+    },
+    {
+      "epoch": 3.277416649192703,
+      "grad_norm": 0.9202993512153625,
+      "learning_rate": 0.0002,
+      "loss": 0.2736,
+      "step": 15630
+    },
+    {
+      "epoch": 3.2795135248479763,
+      "grad_norm": 0.07227743417024612,
+      "learning_rate": 0.0002,
+      "loss": 0.2428,
+      "step": 15640
+    },
+    {
+      "epoch": 3.2816104005032503,
+      "grad_norm": 0.3796677887439728,
+      "learning_rate": 0.0002,
+      "loss": 0.2834,
+      "step": 15650
+    },
+    {
+      "epoch": 3.2837072761585238,
+      "grad_norm": 0.09739989042282104,
+      "learning_rate": 0.0002,
+      "loss": 0.3535,
+      "step": 15660
+    },
+    {
+      "epoch": 3.2858041518137973,
+      "grad_norm": 3.952669143676758,
+      "learning_rate": 0.0002,
+      "loss": 0.2781,
+      "step": 15670
+    },
+    {
+      "epoch": 3.287901027469071,
+      "grad_norm": 0.35927778482437134,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 15680
+    },
+    {
+      "epoch": 3.2899979031243447,
+      "grad_norm": 0.08492504060268402,
+      "learning_rate": 0.0002,
+      "loss": 0.2626,
+      "step": 15690
+    },
+    {
+      "epoch": 3.292094778779618,
+      "grad_norm": 0.12667356431484222,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 15700
+    },
+    {
+      "epoch": 3.294191654434892,
+      "grad_norm": 1.5572481155395508,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 15710
+    },
+    {
+      "epoch": 3.2962885300901656,
+      "grad_norm": 0.07433910667896271,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 15720
+    },
+    {
+      "epoch": 3.298385405745439,
+      "grad_norm": 0.09573189914226532,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 15730
+    },
+    {
+      "epoch": 3.300482281400713,
+      "grad_norm": 0.10412845760583878,
+      "learning_rate": 0.0002,
+      "loss": 0.2611,
+      "step": 15740
+    },
+    {
+      "epoch": 3.3025791570559866,
+      "grad_norm": 0.09497421979904175,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 15750
+    },
+    {
+      "epoch": 3.30467603271126,
+      "grad_norm": 0.09060948342084885,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 15760
+    },
+    {
+      "epoch": 3.306772908366534,
+      "grad_norm": 0.056995928287506104,
+      "learning_rate": 0.0002,
+      "loss": 0.2687,
+      "step": 15770
+    },
+    {
+      "epoch": 3.3088697840218075,
+      "grad_norm": 0.10010065883398056,
+      "learning_rate": 0.0002,
+      "loss": 0.2856,
+      "step": 15780
+    },
+    {
+      "epoch": 3.310966659677081,
+      "grad_norm": 0.09349839389324188,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 15790
+    },
+    {
+      "epoch": 3.313063535332355,
+      "grad_norm": 0.9555671811103821,
+      "learning_rate": 0.0002,
+      "loss": 0.2865,
+      "step": 15800
+    },
+    {
+      "epoch": 3.3151604109876285,
+      "grad_norm": 0.15668243169784546,
+      "learning_rate": 0.0002,
+      "loss": 0.2914,
+      "step": 15810
+    },
+    {
+      "epoch": 3.317257286642902,
+      "grad_norm": 0.0925045758485794,
+      "learning_rate": 0.0002,
+      "loss": 0.2436,
+      "step": 15820
+    },
+    {
+      "epoch": 3.3193541622981755,
+      "grad_norm": 0.11761250346899033,
+      "learning_rate": 0.0002,
+      "loss": 0.2842,
+      "step": 15830
+    },
+    {
+      "epoch": 3.3214510379534494,
+      "grad_norm": 0.1782289296388626,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 15840
+    },
+    {
+      "epoch": 3.323547913608723,
+      "grad_norm": 26.115440368652344,
+      "learning_rate": 0.0002,
+      "loss": 0.3119,
+      "step": 15850
+    },
+    {
+      "epoch": 3.325644789263997,
+      "grad_norm": 8.95678424835205,
+      "learning_rate": 0.0002,
+      "loss": 0.2799,
+      "step": 15860
+    },
+    {
+      "epoch": 3.3277416649192704,
+      "grad_norm": 0.21364077925682068,
+      "learning_rate": 0.0002,
+      "loss": 0.2725,
+      "step": 15870
+    },
+    {
+      "epoch": 3.329838540574544,
+      "grad_norm": 0.23057472705841064,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 15880
+    },
+    {
+      "epoch": 3.3319354162298174,
+      "grad_norm": 6.40324592590332,
+      "learning_rate": 0.0002,
+      "loss": 0.3034,
+      "step": 15890
+    },
+    {
+      "epoch": 3.3340322918850913,
+      "grad_norm": 4.420671463012695,
+      "learning_rate": 0.0002,
+      "loss": 0.3232,
+      "step": 15900
+    },
+    {
+      "epoch": 3.336129167540365,
+      "grad_norm": 0.1368934065103531,
+      "learning_rate": 0.0002,
+      "loss": 0.3034,
+      "step": 15910
+    },
+    {
+      "epoch": 3.3382260431956383,
+      "grad_norm": 0.18694345653057098,
+      "learning_rate": 0.0002,
+      "loss": 0.3256,
+      "step": 15920
+    },
+    {
+      "epoch": 3.3403229188509123,
+      "grad_norm": 7.2182297706604,
+      "learning_rate": 0.0002,
+      "loss": 0.3175,
+      "step": 15930
+    },
+    {
+      "epoch": 3.3424197945061858,
+      "grad_norm": 0.5887989401817322,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 15940
+    },
+    {
+      "epoch": 3.3445166701614593,
+      "grad_norm": 0.1084941029548645,
+      "learning_rate": 0.0002,
+      "loss": 0.2706,
+      "step": 15950
+    },
+    {
+      "epoch": 3.346613545816733,
+      "grad_norm": 0.13692770898342133,
+      "learning_rate": 0.0002,
+      "loss": 0.2443,
+      "step": 15960
+    },
+    {
+      "epoch": 3.3487104214720067,
+      "grad_norm": 0.1195460706949234,
+      "learning_rate": 0.0002,
+      "loss": 0.2975,
+      "step": 15970
+    },
+    {
+      "epoch": 3.35080729712728,
+      "grad_norm": 0.12102661281824112,
+      "learning_rate": 0.0002,
+      "loss": 0.2478,
+      "step": 15980
+    },
+    {
+      "epoch": 3.352904172782554,
+      "grad_norm": 0.11354684084653854,
+      "learning_rate": 0.0002,
+      "loss": 0.2766,
+      "step": 15990
+    },
+    {
+      "epoch": 3.3550010484378276,
+      "grad_norm": 0.300896018743515,
+      "learning_rate": 0.0002,
+      "loss": 0.2848,
+      "step": 16000
+    },
+    {
+      "epoch": 3.357097924093101,
+      "grad_norm": 0.09630642086267471,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 16010
+    },
+    {
+      "epoch": 3.359194799748375,
+      "grad_norm": 0.09073801338672638,
+      "learning_rate": 0.0002,
+      "loss": 0.2474,
+      "step": 16020
+    },
+    {
+      "epoch": 3.3612916754036486,
+      "grad_norm": 0.2423640638589859,
+      "learning_rate": 0.0002,
+      "loss": 0.3038,
+      "step": 16030
+    },
+    {
+      "epoch": 3.363388551058922,
+      "grad_norm": 0.34864917397499084,
+      "learning_rate": 0.0002,
+      "loss": 0.2697,
+      "step": 16040
+    },
+    {
+      "epoch": 3.365485426714196,
+      "grad_norm": 1.5239778757095337,
+      "learning_rate": 0.0002,
+      "loss": 0.2771,
+      "step": 16050
+    },
+    {
+      "epoch": 3.3675823023694695,
+      "grad_norm": 2.1417646408081055,
+      "learning_rate": 0.0002,
+      "loss": 0.3491,
+      "step": 16060
+    },
+    {
+      "epoch": 3.369679178024743,
+      "grad_norm": 0.2309965044260025,
+      "learning_rate": 0.0002,
+      "loss": 0.303,
+      "step": 16070
+    },
+    {
+      "epoch": 3.3717760536800165,
+      "grad_norm": 0.5179856419563293,
+      "learning_rate": 0.0002,
+      "loss": 0.2761,
+      "step": 16080
+    },
+    {
+      "epoch": 3.3738729293352905,
+      "grad_norm": 0.28771182894706726,
+      "learning_rate": 0.0002,
+      "loss": 0.2579,
+      "step": 16090
+    },
+    {
+      "epoch": 3.375969804990564,
+      "grad_norm": 0.08401656895875931,
+      "learning_rate": 0.0002,
+      "loss": 0.2695,
+      "step": 16100
+    },
+    {
+      "epoch": 3.378066680645838,
+      "grad_norm": 0.15731945633888245,
+      "learning_rate": 0.0002,
+      "loss": 0.306,
+      "step": 16110
+    },
+    {
+      "epoch": 3.3801635563011114,
+      "grad_norm": 0.10326915979385376,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 16120
+    },
+    {
+      "epoch": 3.382260431956385,
+      "grad_norm": 0.46936795115470886,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 16130
+    },
+    {
+      "epoch": 3.3843573076116584,
+      "grad_norm": 0.10452701896429062,
+      "learning_rate": 0.0002,
+      "loss": 0.3082,
+      "step": 16140
+    },
+    {
+      "epoch": 3.3864541832669324,
+      "grad_norm": 11.49814510345459,
+      "learning_rate": 0.0002,
+      "loss": 0.2666,
+      "step": 16150
+    },
+    {
+      "epoch": 3.388551058922206,
+      "grad_norm": 0.7889800667762756,
+      "learning_rate": 0.0002,
+      "loss": 0.2711,
+      "step": 16160
+    },
+    {
+      "epoch": 3.3906479345774794,
+      "grad_norm": 0.21009521186351776,
+      "learning_rate": 0.0002,
+      "loss": 0.2719,
+      "step": 16170
+    },
+    {
+      "epoch": 3.3927448102327533,
+      "grad_norm": 0.09990337491035461,
+      "learning_rate": 0.0002,
+      "loss": 0.295,
+      "step": 16180
+    },
+    {
+      "epoch": 3.394841685888027,
+      "grad_norm": 1.3079146146774292,
+      "learning_rate": 0.0002,
+      "loss": 0.2758,
+      "step": 16190
+    },
+    {
+      "epoch": 3.3969385615433003,
+      "grad_norm": 2.922199010848999,
+      "learning_rate": 0.0002,
+      "loss": 0.2836,
+      "step": 16200
+    },
+    {
+      "epoch": 3.3990354371985743,
+      "grad_norm": 0.090868279337883,
+      "learning_rate": 0.0002,
+      "loss": 0.2932,
+      "step": 16210
+    },
+    {
+      "epoch": 3.4011323128538478,
+      "grad_norm": 2.9474756717681885,
+      "learning_rate": 0.0002,
+      "loss": 0.2778,
+      "step": 16220
+    },
+    {
+      "epoch": 3.4032291885091213,
+      "grad_norm": 0.11677554249763489,
+      "learning_rate": 0.0002,
+      "loss": 0.3101,
+      "step": 16230
+    },
+    {
+      "epoch": 3.405326064164395,
+      "grad_norm": 1.1826450824737549,
+      "learning_rate": 0.0002,
+      "loss": 0.3257,
+      "step": 16240
+    },
+    {
+      "epoch": 3.4074229398196687,
+      "grad_norm": 0.3995857536792755,
+      "learning_rate": 0.0002,
+      "loss": 0.3077,
+      "step": 16250
+    },
+    {
+      "epoch": 3.409519815474942,
+      "grad_norm": 5.6643266677856445,
+      "learning_rate": 0.0002,
+      "loss": 0.2657,
+      "step": 16260
+    },
+    {
+      "epoch": 3.411616691130216,
+      "grad_norm": 10.34603214263916,
+      "learning_rate": 0.0002,
+      "loss": 0.2948,
+      "step": 16270
+    },
+    {
+      "epoch": 3.4137135667854897,
+      "grad_norm": 0.4758007228374481,
+      "learning_rate": 0.0002,
+      "loss": 0.2976,
+      "step": 16280
+    },
+    {
+      "epoch": 3.415810442440763,
+      "grad_norm": 15.302112579345703,
+      "learning_rate": 0.0002,
+      "loss": 0.3524,
+      "step": 16290
+    },
+    {
+      "epoch": 3.417907318096037,
+      "grad_norm": 0.20163509249687195,
+      "learning_rate": 0.0002,
+      "loss": 0.2469,
+      "step": 16300
+    },
+    {
+      "epoch": 3.4200041937513106,
+      "grad_norm": 0.3203161656856537,
+      "learning_rate": 0.0002,
+      "loss": 0.301,
+      "step": 16310
+    },
+    {
+      "epoch": 3.422101069406584,
+      "grad_norm": 4.417267799377441,
+      "learning_rate": 0.0002,
+      "loss": 0.355,
+      "step": 16320
+    },
+    {
+      "epoch": 3.4241979450618576,
+      "grad_norm": 6.401717662811279,
+      "learning_rate": 0.0002,
+      "loss": 0.3086,
+      "step": 16330
+    },
+    {
+      "epoch": 3.4262948207171315,
+      "grad_norm": 0.4465663731098175,
+      "learning_rate": 0.0002,
+      "loss": 0.2818,
+      "step": 16340
+    },
+    {
+      "epoch": 3.428391696372405,
+      "grad_norm": 0.6320145726203918,
+      "learning_rate": 0.0002,
+      "loss": 0.3339,
+      "step": 16350
+    },
+    {
+      "epoch": 3.430488572027679,
+      "grad_norm": 0.0976993590593338,
+      "learning_rate": 0.0002,
+      "loss": 0.2778,
+      "step": 16360
+    },
+    {
+      "epoch": 3.4325854476829525,
+      "grad_norm": 0.10189173370599747,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 16370
+    },
+    {
+      "epoch": 3.434682323338226,
+      "grad_norm": 5.137659072875977,
+      "learning_rate": 0.0002,
+      "loss": 0.3291,
+      "step": 16380
+    },
+    {
+      "epoch": 3.4367791989934995,
+      "grad_norm": 61.313507080078125,
+      "learning_rate": 0.0002,
+      "loss": 0.291,
+      "step": 16390
+    },
+    {
+      "epoch": 3.4388760746487734,
+      "grad_norm": 0.3099044859409332,
+      "learning_rate": 0.0002,
+      "loss": 0.3097,
+      "step": 16400
+    },
+    {
+      "epoch": 3.440972950304047,
+      "grad_norm": 0.15905895829200745,
+      "learning_rate": 0.0002,
+      "loss": 0.2659,
+      "step": 16410
+    },
+    {
+      "epoch": 3.4430698259593204,
+      "grad_norm": 0.0829874575138092,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 16420
+    },
+    {
+      "epoch": 3.4451667016145944,
+      "grad_norm": 0.1382455974817276,
+      "learning_rate": 0.0002,
+      "loss": 0.2753,
+      "step": 16430
+    },
+    {
+      "epoch": 3.447263577269868,
+      "grad_norm": 1.0949729681015015,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 16440
+    },
+    {
+      "epoch": 3.4493604529251414,
+      "grad_norm": 0.46144360303878784,
+      "learning_rate": 0.0002,
+      "loss": 0.2449,
+      "step": 16450
+    },
+    {
+      "epoch": 3.4514573285804153,
+      "grad_norm": 0.08408217877149582,
+      "learning_rate": 0.0002,
+      "loss": 0.2811,
+      "step": 16460
+    },
+    {
+      "epoch": 3.453554204235689,
+      "grad_norm": 13.272506713867188,
+      "learning_rate": 0.0002,
+      "loss": 0.2794,
+      "step": 16470
+    },
+    {
+      "epoch": 3.4556510798909623,
+      "grad_norm": 0.13884180784225464,
+      "learning_rate": 0.0002,
+      "loss": 0.3177,
+      "step": 16480
+    },
+    {
+      "epoch": 3.4577479555462363,
+      "grad_norm": 0.12016461789608002,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 16490
+    },
+    {
+      "epoch": 3.4598448312015098,
+      "grad_norm": 0.13663773238658905,
+      "learning_rate": 0.0002,
+      "loss": 0.2805,
+      "step": 16500
+    },
+    {
+      "epoch": 3.4619417068567833,
+      "grad_norm": 0.6070636510848999,
+      "learning_rate": 0.0002,
+      "loss": 0.2444,
+      "step": 16510
+    },
+    {
+      "epoch": 3.464038582512057,
+      "grad_norm": 28.876049041748047,
+      "learning_rate": 0.0002,
+      "loss": 0.427,
+      "step": 16520
+    },
+    {
+      "epoch": 3.4661354581673307,
+      "grad_norm": 0.09710516780614853,
+      "learning_rate": 0.0002,
+      "loss": 0.2669,
+      "step": 16530
+    },
+    {
+      "epoch": 3.468232333822604,
+      "grad_norm": 0.1740933507680893,
+      "learning_rate": 0.0002,
+      "loss": 0.2508,
+      "step": 16540
+    },
+    {
+      "epoch": 3.470329209477878,
+      "grad_norm": 0.11242100596427917,
+      "learning_rate": 0.0002,
+      "loss": 0.2622,
+      "step": 16550
+    },
+    {
+      "epoch": 3.4724260851331517,
+      "grad_norm": 3.6925201416015625,
+      "learning_rate": 0.0002,
+      "loss": 0.3327,
+      "step": 16560
+    },
+    {
+      "epoch": 3.474522960788425,
+      "grad_norm": 0.13877654075622559,
+      "learning_rate": 0.0002,
+      "loss": 0.2746,
+      "step": 16570
+    },
+    {
+      "epoch": 3.4766198364436987,
+      "grad_norm": 0.12891237437725067,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 16580
+    },
+    {
+      "epoch": 3.4787167120989726,
+      "grad_norm": 1.3704564571380615,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 16590
+    },
+    {
+      "epoch": 3.480813587754246,
+      "grad_norm": 14.254056930541992,
+      "learning_rate": 0.0002,
+      "loss": 0.3183,
+      "step": 16600
+    },
+    {
+      "epoch": 3.48291046340952,
+      "grad_norm": 0.10901372879743576,
+      "learning_rate": 0.0002,
+      "loss": 0.2473,
+      "step": 16610
+    },
+    {
+      "epoch": 3.4850073390647935,
+      "grad_norm": 0.0792684555053711,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 16620
+    },
+    {
+      "epoch": 3.487104214720067,
+      "grad_norm": 0.08512654900550842,
+      "learning_rate": 0.0002,
+      "loss": 0.2864,
+      "step": 16630
+    },
+    {
+      "epoch": 3.4892010903753405,
+      "grad_norm": 0.2178574949502945,
+      "learning_rate": 0.0002,
+      "loss": 0.2715,
+      "step": 16640
+    },
+    {
+      "epoch": 3.4912979660306145,
+      "grad_norm": 0.10082219541072845,
+      "learning_rate": 0.0002,
+      "loss": 0.316,
+      "step": 16650
+    },
+    {
+      "epoch": 3.493394841685888,
+      "grad_norm": 2.8539416790008545,
+      "learning_rate": 0.0002,
+      "loss": 0.2639,
+      "step": 16660
+    },
+    {
+      "epoch": 3.4954917173411615,
+      "grad_norm": 0.11633438616991043,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 16670
+    },
+    {
+      "epoch": 3.4975885929964354,
+      "grad_norm": 0.08961481600999832,
+      "learning_rate": 0.0002,
+      "loss": 0.2778,
+      "step": 16680
+    },
+    {
+      "epoch": 3.499685468651709,
+      "grad_norm": 0.08329380303621292,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 16690
+    },
+    {
+      "epoch": 3.5017823443069824,
+      "grad_norm": 0.2124902606010437,
+      "learning_rate": 0.0002,
+      "loss": 0.2787,
+      "step": 16700
+    },
+    {
+      "epoch": 3.5038792199622564,
+      "grad_norm": 0.07479745149612427,
+      "learning_rate": 0.0002,
+      "loss": 0.2424,
+      "step": 16710
+    },
+    {
+      "epoch": 3.50597609561753,
+      "grad_norm": 0.8046960830688477,
+      "learning_rate": 0.0002,
+      "loss": 0.3099,
+      "step": 16720
+    },
+    {
+      "epoch": 3.5080729712728034,
+      "grad_norm": 0.17260153591632843,
+      "learning_rate": 0.0002,
+      "loss": 0.2451,
+      "step": 16730
+    },
+    {
+      "epoch": 3.5101698469280773,
+      "grad_norm": 0.4296531677246094,
+      "learning_rate": 0.0002,
+      "loss": 0.3065,
+      "step": 16740
+    },
+    {
+      "epoch": 3.512266722583351,
+      "grad_norm": 1.1957885026931763,
+      "learning_rate": 0.0002,
+      "loss": 0.2784,
+      "step": 16750
+    },
+    {
+      "epoch": 3.5143635982386243,
+      "grad_norm": 0.08505085110664368,
+      "learning_rate": 0.0002,
+      "loss": 0.2827,
+      "step": 16760
+    },
+    {
+      "epoch": 3.516460473893898,
+      "grad_norm": 1.681307315826416,
+      "learning_rate": 0.0002,
+      "loss": 0.2906,
+      "step": 16770
+    },
+    {
+      "epoch": 3.5185573495491718,
+      "grad_norm": 0.19510099291801453,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 16780
+    },
+    {
+      "epoch": 3.5206542252044453,
+      "grad_norm": 0.40013596415519714,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 16790
+    },
+    {
+      "epoch": 3.522751100859719,
+      "grad_norm": 20.94623565673828,
+      "learning_rate": 0.0002,
+      "loss": 0.3363,
+      "step": 16800
+    },
+    {
+      "epoch": 3.5248479765149927,
+      "grad_norm": 0.11770687252283096,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 16810
+    },
+    {
+      "epoch": 3.526944852170266,
+      "grad_norm": 0.47683051228523254,
+      "learning_rate": 0.0002,
+      "loss": 0.2424,
+      "step": 16820
+    },
+    {
+      "epoch": 3.5290417278255397,
+      "grad_norm": 0.3234413266181946,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 16830
+    },
+    {
+      "epoch": 3.5311386034808137,
+      "grad_norm": 66.1975326538086,
+      "learning_rate": 0.0002,
+      "loss": 0.3623,
+      "step": 16840
+    },
+    {
+      "epoch": 3.533235479136087,
+      "grad_norm": 10.475239753723145,
+      "learning_rate": 0.0002,
+      "loss": 1.6228,
+      "step": 16850
+    },
+    {
+      "epoch": 3.535332354791361,
+      "grad_norm": 0.19146008789539337,
+      "learning_rate": 0.0002,
+      "loss": 0.3097,
+      "step": 16860
+    },
+    {
+      "epoch": 3.5374292304466346,
+      "grad_norm": 0.10526818037033081,
+      "learning_rate": 0.0002,
+      "loss": 0.3286,
+      "step": 16870
+    },
+    {
+      "epoch": 3.539526106101908,
+      "grad_norm": 0.2106499820947647,
+      "learning_rate": 0.0002,
+      "loss": 0.326,
+      "step": 16880
+    },
+    {
+      "epoch": 3.5416229817571816,
+      "grad_norm": 0.16286778450012207,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 16890
+    },
+    {
+      "epoch": 3.5437198574124555,
+      "grad_norm": 0.2097437083721161,
+      "learning_rate": 0.0002,
+      "loss": 0.3212,
+      "step": 16900
+    },
+    {
+      "epoch": 3.545816733067729,
+      "grad_norm": 0.8928031921386719,
+      "learning_rate": 0.0002,
+      "loss": 0.2455,
+      "step": 16910
+    },
+    {
+      "epoch": 3.547913608723003,
+      "grad_norm": 0.20158305764198303,
+      "learning_rate": 0.0002,
+      "loss": 0.2946,
+      "step": 16920
+    },
+    {
+      "epoch": 3.5500104843782765,
+      "grad_norm": 8.960654258728027,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 16930
+    },
+    {
+      "epoch": 3.55210736003355,
+      "grad_norm": 3.4482638835906982,
+      "learning_rate": 0.0002,
+      "loss": 0.2985,
+      "step": 16940
+    },
+    {
+      "epoch": 3.5542042356888235,
+      "grad_norm": 1.4937551021575928,
+      "learning_rate": 0.0002,
+      "loss": 0.2759,
+      "step": 16950
+    },
+    {
+      "epoch": 3.5563011113440974,
+      "grad_norm": 3.2123825550079346,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 16960
+    },
+    {
+      "epoch": 3.558397986999371,
+      "grad_norm": 0.10020950436592102,
+      "learning_rate": 0.0002,
+      "loss": 0.2486,
+      "step": 16970
+    },
+    {
+      "epoch": 3.5604948626546444,
+      "grad_norm": 0.6148474812507629,
+      "learning_rate": 0.0002,
+      "loss": 0.3961,
+      "step": 16980
+    },
+    {
+      "epoch": 3.5625917383099184,
+      "grad_norm": 0.11368487775325775,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 16990
+    },
+    {
+      "epoch": 3.564688613965192,
+      "grad_norm": 0.4417993724346161,
+      "learning_rate": 0.0002,
+      "loss": 0.4003,
+      "step": 17000
+    },
+    {
+      "epoch": 3.5667854896204654,
+      "grad_norm": 0.10857246071100235,
+      "learning_rate": 0.0002,
+      "loss": 0.2453,
+      "step": 17010
+    },
+    {
+      "epoch": 3.568882365275739,
+      "grad_norm": 0.4231739044189453,
+      "learning_rate": 0.0002,
+      "loss": 0.2452,
+      "step": 17020
+    },
+    {
+      "epoch": 3.570979240931013,
+      "grad_norm": 0.21850062906742096,
+      "learning_rate": 0.0002,
+      "loss": 0.3663,
+      "step": 17030
+    },
+    {
+      "epoch": 3.5730761165862863,
+      "grad_norm": 0.10159339010715485,
+      "learning_rate": 0.0002,
+      "loss": 0.3287,
+      "step": 17040
+    },
+    {
+      "epoch": 3.5751729922415603,
+      "grad_norm": 0.2898092269897461,
+      "learning_rate": 0.0002,
+      "loss": 0.2447,
+      "step": 17050
+    },
+    {
+      "epoch": 3.5772698678968338,
+      "grad_norm": 7.484994411468506,
+      "learning_rate": 0.0002,
+      "loss": 0.2693,
+      "step": 17060
+    },
+    {
+      "epoch": 3.5793667435521073,
+      "grad_norm": 27.404441833496094,
+      "learning_rate": 0.0002,
+      "loss": 0.3023,
+      "step": 17070
+    },
+    {
+      "epoch": 3.5814636192073808,
+      "grad_norm": 3.085254430770874,
+      "learning_rate": 0.0002,
+      "loss": 0.2913,
+      "step": 17080
+    },
+    {
+      "epoch": 3.5835604948626547,
+      "grad_norm": 2.5266666412353516,
+      "learning_rate": 0.0002,
+      "loss": 0.3504,
+      "step": 17090
+    },
+    {
+      "epoch": 3.585657370517928,
+      "grad_norm": 0.17346085608005524,
+      "learning_rate": 0.0002,
+      "loss": 0.2483,
+      "step": 17100
+    },
+    {
+      "epoch": 3.587754246173202,
+      "grad_norm": 5.260568618774414,
+      "learning_rate": 0.0002,
+      "loss": 0.2502,
+      "step": 17110
+    },
+    {
+      "epoch": 3.5898511218284757,
+      "grad_norm": 8.863641738891602,
+      "learning_rate": 0.0002,
+      "loss": 0.2695,
+      "step": 17120
+    },
+    {
+      "epoch": 3.591947997483749,
+      "grad_norm": 0.36519744992256165,
+      "learning_rate": 0.0002,
+      "loss": 0.2684,
+      "step": 17130
+    },
+    {
+      "epoch": 3.5940448731390227,
+      "grad_norm": 1.6287387609481812,
+      "learning_rate": 0.0002,
+      "loss": 0.2827,
+      "step": 17140
+    },
+    {
+      "epoch": 3.5961417487942966,
+      "grad_norm": 15.245902061462402,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 17150
+    },
+    {
+      "epoch": 3.59823862444957,
+      "grad_norm": 26.536678314208984,
+      "learning_rate": 0.0002,
+      "loss": 0.3645,
+      "step": 17160
+    },
+    {
+      "epoch": 3.600335500104844,
+      "grad_norm": 10.872307777404785,
+      "learning_rate": 0.0002,
+      "loss": 0.4323,
+      "step": 17170
+    },
+    {
+      "epoch": 3.6024323757601175,
+      "grad_norm": 14.617993354797363,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 17180
+    },
+    {
+      "epoch": 3.604529251415391,
+      "grad_norm": 3.0631062984466553,
+      "learning_rate": 0.0002,
+      "loss": 0.3932,
+      "step": 17190
+    },
+    {
+      "epoch": 3.6066261270706645,
+      "grad_norm": 7.422130107879639,
+      "learning_rate": 0.0002,
+      "loss": 0.3115,
+      "step": 17200
+    },
+    {
+      "epoch": 3.6087230027259385,
+      "grad_norm": 2.1507389545440674,
+      "learning_rate": 0.0002,
+      "loss": 0.2567,
+      "step": 17210
+    },
+    {
+      "epoch": 3.610819878381212,
+      "grad_norm": 3.0334174633026123,
+      "learning_rate": 0.0002,
+      "loss": 0.251,
+      "step": 17220
+    },
+    {
+      "epoch": 3.6129167540364855,
+      "grad_norm": 3.9418137073516846,
+      "learning_rate": 0.0002,
+      "loss": 0.3277,
+      "step": 17230
+    },
+    {
+      "epoch": 3.6150136296917594,
+      "grad_norm": 14.20264720916748,
+      "learning_rate": 0.0002,
+      "loss": 0.2983,
+      "step": 17240
+    },
+    {
+      "epoch": 3.617110505347033,
+      "grad_norm": 5.825155735015869,
+      "learning_rate": 0.0002,
+      "loss": 0.3208,
+      "step": 17250
+    },
+    {
+      "epoch": 3.6192073810023064,
+      "grad_norm": 15.464000701904297,
+      "learning_rate": 0.0002,
+      "loss": 0.2542,
+      "step": 17260
+    },
+    {
+      "epoch": 3.62130425665758,
+      "grad_norm": 3.840942621231079,
+      "learning_rate": 0.0002,
+      "loss": 0.2642,
+      "step": 17270
+    },
+    {
+      "epoch": 3.623401132312854,
+      "grad_norm": 1.0902684926986694,
+      "learning_rate": 0.0002,
+      "loss": 0.345,
+      "step": 17280
+    },
+    {
+      "epoch": 3.6254980079681274,
+      "grad_norm": 10.194125175476074,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 17290
+    },
+    {
+      "epoch": 3.6275948836234013,
+      "grad_norm": 0.5578778982162476,
+      "learning_rate": 0.0002,
+      "loss": 0.2488,
+      "step": 17300
+    },
+    {
+      "epoch": 3.629691759278675,
+      "grad_norm": 0.55025714635849,
+      "learning_rate": 0.0002,
+      "loss": 0.2786,
+      "step": 17310
+    },
+    {
+      "epoch": 3.6317886349339483,
+      "grad_norm": 2.2328052520751953,
+      "learning_rate": 0.0002,
+      "loss": 0.301,
+      "step": 17320
+    },
+    {
+      "epoch": 3.633885510589222,
+      "grad_norm": 21.651777267456055,
+      "learning_rate": 0.0002,
+      "loss": 0.3668,
+      "step": 17330
+    },
+    {
+      "epoch": 3.6359823862444958,
+      "grad_norm": 10.565069198608398,
+      "learning_rate": 0.0002,
+      "loss": 0.2922,
+      "step": 17340
+    },
+    {
+      "epoch": 3.6380792618997693,
+      "grad_norm": 12.717290878295898,
+      "learning_rate": 0.0002,
+      "loss": 0.3583,
+      "step": 17350
+    },
+    {
+      "epoch": 3.640176137555043,
+      "grad_norm": 8.592275619506836,
+      "learning_rate": 0.0002,
+      "loss": 0.3302,
+      "step": 17360
+    },
+    {
+      "epoch": 3.6422730132103167,
+      "grad_norm": 1.2025078535079956,
+      "learning_rate": 0.0002,
+      "loss": 0.2525,
+      "step": 17370
+    },
+    {
+      "epoch": 3.64436988886559,
+      "grad_norm": 0.9852961301803589,
+      "learning_rate": 0.0002,
+      "loss": 0.3788,
+      "step": 17380
+    },
+    {
+      "epoch": 3.6464667645208637,
+      "grad_norm": 11.788804054260254,
+      "learning_rate": 0.0002,
+      "loss": 0.2989,
+      "step": 17390
+    },
+    {
+      "epoch": 3.6485636401761377,
+      "grad_norm": 2.911071300506592,
+      "learning_rate": 0.0002,
+      "loss": 0.3112,
+      "step": 17400
+    },
+    {
+      "epoch": 3.650660515831411,
+      "grad_norm": 1.0645408630371094,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 17410
+    },
+    {
+      "epoch": 3.652757391486685,
+      "grad_norm": 0.17559315264225006,
+      "learning_rate": 0.0002,
+      "loss": 0.2792,
+      "step": 17420
+    },
+    {
+      "epoch": 3.6548542671419586,
+      "grad_norm": 0.39271262288093567,
+      "learning_rate": 0.0002,
+      "loss": 0.2963,
+      "step": 17430
+    },
+    {
+      "epoch": 3.656951142797232,
+      "grad_norm": 0.6976293921470642,
+      "learning_rate": 0.0002,
+      "loss": 0.3541,
+      "step": 17440
+    },
+    {
+      "epoch": 3.6590480184525056,
+      "grad_norm": 0.7792041897773743,
+      "learning_rate": 0.0002,
+      "loss": 0.2937,
+      "step": 17450
+    },
+    {
+      "epoch": 3.6611448941077795,
+      "grad_norm": 0.5158128142356873,
+      "learning_rate": 0.0002,
+      "loss": 0.3441,
+      "step": 17460
+    },
+    {
+      "epoch": 3.663241769763053,
+      "grad_norm": 2.1796839237213135,
+      "learning_rate": 0.0002,
+      "loss": 0.3525,
+      "step": 17470
+    },
+    {
+      "epoch": 3.6653386454183265,
+      "grad_norm": 7.255102157592773,
+      "learning_rate": 0.0002,
+      "loss": 0.4197,
+      "step": 17480
+    },
+    {
+      "epoch": 3.6674355210736005,
+      "grad_norm": 1.5092166662216187,
+      "learning_rate": 0.0002,
+      "loss": 0.3176,
+      "step": 17490
+    },
+    {
+      "epoch": 3.669532396728874,
+      "grad_norm": 0.32293596863746643,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 17500
+    },
+    {
+      "epoch": 3.6716292723841475,
+      "grad_norm": 1.2418729066848755,
+      "learning_rate": 0.0002,
+      "loss": 0.2499,
+      "step": 17510
+    },
+    {
+      "epoch": 3.673726148039421,
+      "grad_norm": 8.73947811126709,
+      "learning_rate": 0.0002,
+      "loss": 0.3851,
+      "step": 17520
+    },
+    {
+      "epoch": 3.675823023694695,
+      "grad_norm": 0.43778735399246216,
+      "learning_rate": 0.0002,
+      "loss": 0.2479,
+      "step": 17530
+    },
+    {
+      "epoch": 3.6779198993499684,
+      "grad_norm": 5.211206436157227,
+      "learning_rate": 0.0002,
+      "loss": 0.3464,
+      "step": 17540
+    },
+    {
+      "epoch": 3.6800167750052424,
+      "grad_norm": 0.3823220133781433,
+      "learning_rate": 0.0002,
+      "loss": 0.3327,
+      "step": 17550
+    },
+    {
+      "epoch": 3.682113650660516,
+      "grad_norm": 1.5433018207550049,
+      "learning_rate": 0.0002,
+      "loss": 0.2942,
+      "step": 17560
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.5892255306243896,
+      "learning_rate": 0.0002,
+      "loss": 0.2552,
+      "step": 17570
+    },
+    {
+      "epoch": 3.686307401971063,
+      "grad_norm": 0.17415273189544678,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 17580
+    },
+    {
+      "epoch": 3.688404277626337,
+      "grad_norm": 0.13931488990783691,
+      "learning_rate": 0.0002,
+      "loss": 0.2599,
+      "step": 17590
+    },
+    {
+      "epoch": 3.6905011532816103,
+      "grad_norm": 0.3451101779937744,
+      "learning_rate": 0.0002,
+      "loss": 0.3056,
+      "step": 17600
+    },
+    {
+      "epoch": 3.6925980289368843,
+      "grad_norm": 5.913544654846191,
+      "learning_rate": 0.0002,
+      "loss": 0.2928,
+      "step": 17610
+    },
+    {
+      "epoch": 3.6946949045921578,
+      "grad_norm": 1.0163569450378418,
+      "learning_rate": 0.0002,
+      "loss": 0.2797,
+      "step": 17620
+    },
+    {
+      "epoch": 3.6967917802474313,
+      "grad_norm": 0.27711400389671326,
+      "learning_rate": 0.0002,
+      "loss": 0.3853,
+      "step": 17630
+    },
+    {
+      "epoch": 3.6988886559027048,
+      "grad_norm": 0.5875184535980225,
+      "learning_rate": 0.0002,
+      "loss": 0.2511,
+      "step": 17640
+    },
+    {
+      "epoch": 3.7009855315579787,
+      "grad_norm": 0.15635895729064941,
+      "learning_rate": 0.0002,
+      "loss": 0.2719,
+      "step": 17650
+    },
+    {
+      "epoch": 3.703082407213252,
+      "grad_norm": 0.15206418931484222,
+      "learning_rate": 0.0002,
+      "loss": 0.325,
+      "step": 17660
+    },
+    {
+      "epoch": 3.705179282868526,
+      "grad_norm": 0.08338475227355957,
+      "learning_rate": 0.0002,
+      "loss": 0.281,
+      "step": 17670
+    },
+    {
+      "epoch": 3.7072761585237997,
+      "grad_norm": 0.4060516059398651,
+      "learning_rate": 0.0002,
+      "loss": 0.3038,
+      "step": 17680
+    },
+    {
+      "epoch": 3.709373034179073,
+      "grad_norm": 0.23129555583000183,
+      "learning_rate": 0.0002,
+      "loss": 0.2948,
+      "step": 17690
+    },
+    {
+      "epoch": 3.7114699098343467,
+      "grad_norm": 0.19247643649578094,
+      "learning_rate": 0.0002,
+      "loss": 0.319,
+      "step": 17700
+    },
+    {
+      "epoch": 3.7135667854896206,
+      "grad_norm": 1.2437766790390015,
+      "learning_rate": 0.0002,
+      "loss": 0.274,
+      "step": 17710
+    },
+    {
+      "epoch": 3.715663661144894,
+      "grad_norm": 0.10187529772520065,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 17720
+    },
+    {
+      "epoch": 3.7177605368001676,
+      "grad_norm": 0.1905500441789627,
+      "learning_rate": 0.0002,
+      "loss": 0.277,
+      "step": 17730
+    },
+    {
+      "epoch": 3.7198574124554415,
+      "grad_norm": 0.18983417749404907,
+      "learning_rate": 0.0002,
+      "loss": 0.3274,
+      "step": 17740
+    },
+    {
+      "epoch": 3.721954288110715,
+      "grad_norm": 12.874968528747559,
+      "learning_rate": 0.0002,
+      "loss": 0.2807,
+      "step": 17750
+    },
+    {
+      "epoch": 3.7240511637659885,
+      "grad_norm": 0.11451554298400879,
+      "learning_rate": 0.0002,
+      "loss": 0.3368,
+      "step": 17760
+    },
+    {
+      "epoch": 3.726148039421262,
+      "grad_norm": 0.10391887277364731,
+      "learning_rate": 0.0002,
+      "loss": 0.2853,
+      "step": 17770
+    },
+    {
+      "epoch": 3.728244915076536,
+      "grad_norm": 0.11095970869064331,
+      "learning_rate": 0.0002,
+      "loss": 0.32,
+      "step": 17780
+    },
+    {
+      "epoch": 3.7303417907318095,
+      "grad_norm": 0.09512672573328018,
+      "learning_rate": 0.0002,
+      "loss": 0.2787,
+      "step": 17790
+    },
+    {
+      "epoch": 3.7324386663870834,
+      "grad_norm": 0.19171316921710968,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 17800
+    },
+    {
+      "epoch": 3.734535542042357,
+      "grad_norm": 0.12946048378944397,
+      "learning_rate": 0.0002,
+      "loss": 0.2843,
+      "step": 17810
+    },
+    {
+      "epoch": 3.7366324176976304,
+      "grad_norm": 0.06607726216316223,
+      "learning_rate": 0.0002,
+      "loss": 0.2844,
+      "step": 17820
+    },
+    {
+      "epoch": 3.738729293352904,
+      "grad_norm": 0.40193358063697815,
+      "learning_rate": 0.0002,
+      "loss": 0.3293,
+      "step": 17830
+    },
+    {
+      "epoch": 3.740826169008178,
+      "grad_norm": 0.14429262280464172,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 17840
+    },
+    {
+      "epoch": 3.7429230446634514,
+      "grad_norm": 0.11837374418973923,
+      "learning_rate": 0.0002,
+      "loss": 0.2532,
+      "step": 17850
+    },
+    {
+      "epoch": 3.7450199203187253,
+      "grad_norm": 0.182389035820961,
+      "learning_rate": 0.0002,
+      "loss": 0.2528,
+      "step": 17860
+    },
+    {
+      "epoch": 3.747116795973999,
+      "grad_norm": 2.132552146911621,
+      "learning_rate": 0.0002,
+      "loss": 0.3393,
+      "step": 17870
+    },
+    {
+      "epoch": 3.7492136716292723,
+      "grad_norm": 0.14405368268489838,
+      "learning_rate": 0.0002,
+      "loss": 0.2848,
+      "step": 17880
+    },
+    {
+      "epoch": 3.751310547284546,
+      "grad_norm": 20.570039749145508,
+      "learning_rate": 0.0002,
+      "loss": 0.277,
+      "step": 17890
+    },
+    {
+      "epoch": 3.7534074229398198,
+      "grad_norm": 8.199270248413086,
+      "learning_rate": 0.0002,
+      "loss": 0.2777,
+      "step": 17900
+    },
+    {
+      "epoch": 3.7555042985950933,
+      "grad_norm": 1.1704336404800415,
+      "learning_rate": 0.0002,
+      "loss": 0.25,
+      "step": 17910
+    },
+    {
+      "epoch": 3.757601174250367,
+      "grad_norm": 4.142645835876465,
+      "learning_rate": 0.0002,
+      "loss": 0.3166,
+      "step": 17920
+    },
+    {
+      "epoch": 3.7596980499056407,
+      "grad_norm": 1.2840096950531006,
+      "learning_rate": 0.0002,
+      "loss": 0.2479,
+      "step": 17930
+    },
+    {
+      "epoch": 3.761794925560914,
+      "grad_norm": 0.2658231854438782,
+      "learning_rate": 0.0002,
+      "loss": 0.295,
+      "step": 17940
+    },
+    {
+      "epoch": 3.7638918012161877,
+      "grad_norm": 10.568625450134277,
+      "learning_rate": 0.0002,
+      "loss": 0.2504,
+      "step": 17950
+    },
+    {
+      "epoch": 3.7659886768714617,
+      "grad_norm": 0.7367125153541565,
+      "learning_rate": 0.0002,
+      "loss": 0.2729,
+      "step": 17960
+    },
+    {
+      "epoch": 3.768085552526735,
+      "grad_norm": 15.142430305480957,
+      "learning_rate": 0.0002,
+      "loss": 0.3903,
+      "step": 17970
+    },
+    {
+      "epoch": 3.7701824281820087,
+      "grad_norm": 0.8723947405815125,
+      "learning_rate": 0.0002,
+      "loss": 0.2495,
+      "step": 17980
+    },
+    {
+      "epoch": 3.7722793038372826,
+      "grad_norm": 0.542462944984436,
+      "learning_rate": 0.0002,
+      "loss": 0.2764,
+      "step": 17990
+    },
+    {
+      "epoch": 3.774376179492556,
+      "grad_norm": 0.5069801807403564,
+      "learning_rate": 0.0002,
+      "loss": 0.2774,
+      "step": 18000
+    },
+    {
+      "epoch": 3.7764730551478296,
+      "grad_norm": 28.867746353149414,
+      "learning_rate": 0.0002,
+      "loss": 0.3051,
+      "step": 18010
+    },
+    {
+      "epoch": 3.778569930803103,
+      "grad_norm": 19.830551147460938,
+      "learning_rate": 0.0002,
+      "loss": 0.3139,
+      "step": 18020
+    },
+    {
+      "epoch": 3.780666806458377,
+      "grad_norm": 13.429841041564941,
+      "learning_rate": 0.0002,
+      "loss": 0.2492,
+      "step": 18030
+    },
+    {
+      "epoch": 3.7827636821136505,
+      "grad_norm": 0.7679717540740967,
+      "learning_rate": 0.0002,
+      "loss": 0.2486,
+      "step": 18040
+    },
+    {
+      "epoch": 3.7848605577689245,
+      "grad_norm": 0.280867338180542,
+      "learning_rate": 0.0002,
+      "loss": 0.2875,
+      "step": 18050
+    },
+    {
+      "epoch": 3.786957433424198,
+      "grad_norm": 0.576727032661438,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 18060
+    },
+    {
+      "epoch": 3.7890543090794715,
+      "grad_norm": 12.595909118652344,
+      "learning_rate": 0.0002,
+      "loss": 0.3076,
+      "step": 18070
+    },
+    {
+      "epoch": 3.791151184734745,
+      "grad_norm": 8.194901466369629,
+      "learning_rate": 0.0002,
+      "loss": 0.3268,
+      "step": 18080
+    },
+    {
+      "epoch": 3.793248060390019,
+      "grad_norm": 0.9795236587524414,
+      "learning_rate": 0.0002,
+      "loss": 0.2486,
+      "step": 18090
+    },
+    {
+      "epoch": 3.7953449360452924,
+      "grad_norm": 0.21155227720737457,
+      "learning_rate": 0.0002,
+      "loss": 0.2967,
+      "step": 18100
+    },
+    {
+      "epoch": 3.7974418117005664,
+      "grad_norm": 0.17290417850017548,
+      "learning_rate": 0.0002,
+      "loss": 0.2474,
+      "step": 18110
+    },
+    {
+      "epoch": 3.79953868735584,
+      "grad_norm": 0.1753161996603012,
+      "learning_rate": 0.0002,
+      "loss": 0.2472,
+      "step": 18120
+    },
+    {
+      "epoch": 3.8016355630111134,
+      "grad_norm": 0.16531455516815186,
+      "learning_rate": 0.0002,
+      "loss": 0.3316,
+      "step": 18130
+    },
+    {
+      "epoch": 3.803732438666387,
+      "grad_norm": 0.27308961749076843,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 18140
+    },
+    {
+      "epoch": 3.805829314321661,
+      "grad_norm": 0.885614812374115,
+      "learning_rate": 0.0002,
+      "loss": 0.3144,
+      "step": 18150
+    },
+    {
+      "epoch": 3.8079261899769343,
+      "grad_norm": 0.23859460651874542,
+      "learning_rate": 0.0002,
+      "loss": 0.3965,
+      "step": 18160
+    },
+    {
+      "epoch": 3.8100230656322083,
+      "grad_norm": 0.8167296051979065,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 18170
+    },
+    {
+      "epoch": 3.8121199412874818,
+      "grad_norm": 14.719027519226074,
+      "learning_rate": 0.0002,
+      "loss": 0.3193,
+      "step": 18180
+    },
+    {
+      "epoch": 3.8142168169427553,
+      "grad_norm": 18.75009536743164,
+      "learning_rate": 0.0002,
+      "loss": 0.3505,
+      "step": 18190
+    },
+    {
+      "epoch": 3.8163136925980288,
+      "grad_norm": 0.7599632740020752,
+      "learning_rate": 0.0002,
+      "loss": 0.2432,
+      "step": 18200
+    },
+    {
+      "epoch": 3.8184105682533027,
+      "grad_norm": 2.993929862976074,
+      "learning_rate": 0.0002,
+      "loss": 0.4737,
+      "step": 18210
+    },
+    {
+      "epoch": 3.820507443908576,
+      "grad_norm": 0.14531955122947693,
+      "learning_rate": 0.0002,
+      "loss": 0.2779,
+      "step": 18220
+    },
+    {
+      "epoch": 3.8226043195638497,
+      "grad_norm": 0.21003879606723785,
+      "learning_rate": 0.0002,
+      "loss": 0.2883,
+      "step": 18230
+    },
+    {
+      "epoch": 3.8247011952191237,
+      "grad_norm": 15.457707405090332,
+      "learning_rate": 0.0002,
+      "loss": 0.3807,
+      "step": 18240
+    },
+    {
+      "epoch": 3.826798070874397,
+      "grad_norm": 0.12080878019332886,
+      "learning_rate": 0.0002,
+      "loss": 0.269,
+      "step": 18250
+    },
+    {
+      "epoch": 3.8288949465296707,
+      "grad_norm": 0.20637743175029755,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 18260
+    },
+    {
+      "epoch": 3.830991822184944,
+      "grad_norm": 16.86208724975586,
+      "learning_rate": 0.0002,
+      "loss": 0.3,
+      "step": 18270
+    },
+    {
+      "epoch": 3.833088697840218,
+      "grad_norm": 0.15844117105007172,
+      "learning_rate": 0.0002,
+      "loss": 0.2729,
+      "step": 18280
+    },
+    {
+      "epoch": 3.8351855734954916,
+      "grad_norm": 1.4949979782104492,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 18290
+    },
+    {
+      "epoch": 3.8372824491507656,
+      "grad_norm": 0.2570766806602478,
+      "learning_rate": 0.0002,
+      "loss": 0.2507,
+      "step": 18300
+    },
+    {
+      "epoch": 3.839379324806039,
+      "grad_norm": 0.27064013481140137,
+      "learning_rate": 0.0002,
+      "loss": 0.2641,
+      "step": 18310
+    },
+    {
+      "epoch": 3.8414762004613126,
+      "grad_norm": 0.10055670887231827,
+      "learning_rate": 0.0002,
+      "loss": 0.2928,
+      "step": 18320
+    },
+    {
+      "epoch": 3.843573076116586,
+      "grad_norm": 2.022714614868164,
+      "learning_rate": 0.0002,
+      "loss": 0.2814,
+      "step": 18330
+    },
+    {
+      "epoch": 3.84566995177186,
+      "grad_norm": 0.17389263212680817,
+      "learning_rate": 0.0002,
+      "loss": 0.3323,
+      "step": 18340
+    },
+    {
+      "epoch": 3.8477668274271335,
+      "grad_norm": 1.4782214164733887,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 18350
+    },
+    {
+      "epoch": 3.8498637030824074,
+      "grad_norm": 1.1519814729690552,
+      "learning_rate": 0.0002,
+      "loss": 0.297,
+      "step": 18360
+    },
+    {
+      "epoch": 3.851960578737681,
+      "grad_norm": 0.8425706028938293,
+      "learning_rate": 0.0002,
+      "loss": 0.2739,
+      "step": 18370
+    },
+    {
+      "epoch": 3.8540574543929544,
+      "grad_norm": 0.17960014939308167,
+      "learning_rate": 0.0002,
+      "loss": 0.3024,
+      "step": 18380
+    },
+    {
+      "epoch": 3.856154330048228,
+      "grad_norm": 0.7193502187728882,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 18390
+    },
+    {
+      "epoch": 3.858251205703502,
+      "grad_norm": 0.5273101925849915,
+      "learning_rate": 0.0002,
+      "loss": 0.3255,
+      "step": 18400
+    },
+    {
+      "epoch": 3.8603480813587754,
+      "grad_norm": 9.503673553466797,
+      "learning_rate": 0.0002,
+      "loss": 0.3206,
+      "step": 18410
+    },
+    {
+      "epoch": 3.8624449570140493,
+      "grad_norm": 0.15235988795757294,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 18420
+    },
+    {
+      "epoch": 3.864541832669323,
+      "grad_norm": 0.24428336322307587,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 18430
+    },
+    {
+      "epoch": 3.8666387083245963,
+      "grad_norm": 12.147161483764648,
+      "learning_rate": 0.0002,
+      "loss": 0.2859,
+      "step": 18440
+    },
+    {
+      "epoch": 3.86873558397987,
+      "grad_norm": 0.24032670259475708,
+      "learning_rate": 0.0002,
+      "loss": 0.2585,
+      "step": 18450
+    },
+    {
+      "epoch": 3.8708324596351438,
+      "grad_norm": 0.626478374004364,
+      "learning_rate": 0.0002,
+      "loss": 0.2903,
+      "step": 18460
+    },
+    {
+      "epoch": 3.8729293352904173,
+      "grad_norm": 0.7303033471107483,
+      "learning_rate": 0.0002,
+      "loss": 0.3426,
+      "step": 18470
+    },
+    {
+      "epoch": 3.8750262109456908,
+      "grad_norm": 1.397939920425415,
+      "learning_rate": 0.0002,
+      "loss": 0.3911,
+      "step": 18480
+    },
+    {
+      "epoch": 3.8771230866009647,
+      "grad_norm": 0.2924603223800659,
+      "learning_rate": 0.0002,
+      "loss": 0.3147,
+      "step": 18490
+    },
+    {
+      "epoch": 3.879219962256238,
+      "grad_norm": 11.158350944519043,
+      "learning_rate": 0.0002,
+      "loss": 0.2983,
+      "step": 18500
+    },
+    {
+      "epoch": 3.8813168379115117,
+      "grad_norm": 0.773551344871521,
+      "learning_rate": 0.0002,
+      "loss": 0.2881,
+      "step": 18510
+    },
+    {
+      "epoch": 3.883413713566785,
+      "grad_norm": 0.7801753282546997,
+      "learning_rate": 0.0002,
+      "loss": 0.2908,
+      "step": 18520
+    },
+    {
+      "epoch": 3.885510589222059,
+      "grad_norm": 0.9729545712471008,
+      "learning_rate": 0.0002,
+      "loss": 0.3189,
+      "step": 18530
+    },
+    {
+      "epoch": 3.8876074648773327,
+      "grad_norm": 7.370116233825684,
+      "learning_rate": 0.0002,
+      "loss": 0.2981,
+      "step": 18540
+    },
+    {
+      "epoch": 3.8897043405326066,
+      "grad_norm": 1.4260889291763306,
+      "learning_rate": 0.0002,
+      "loss": 0.3241,
+      "step": 18550
+    },
+    {
+      "epoch": 3.89180121618788,
+      "grad_norm": 1.2304461002349854,
+      "learning_rate": 0.0002,
+      "loss": 0.2543,
+      "step": 18560
+    },
+    {
+      "epoch": 3.8938980918431536,
+      "grad_norm": 10.699807167053223,
+      "learning_rate": 0.0002,
+      "loss": 0.3361,
+      "step": 18570
+    },
+    {
+      "epoch": 3.895994967498427,
+      "grad_norm": 5.950390338897705,
+      "learning_rate": 0.0002,
+      "loss": 0.3532,
+      "step": 18580
+    },
+    {
+      "epoch": 3.898091843153701,
+      "grad_norm": 0.6864115595817566,
+      "learning_rate": 0.0002,
+      "loss": 0.2665,
+      "step": 18590
+    },
+    {
+      "epoch": 3.9001887188089746,
+      "grad_norm": 1.4243322610855103,
+      "learning_rate": 0.0002,
+      "loss": 0.2468,
+      "step": 18600
+    },
+    {
+      "epoch": 3.9022855944642485,
+      "grad_norm": 15.98333740234375,
+      "learning_rate": 0.0002,
+      "loss": 0.3061,
+      "step": 18610
+    },
+    {
+      "epoch": 3.904382470119522,
+      "grad_norm": 0.6285650730133057,
+      "learning_rate": 0.0002,
+      "loss": 0.3156,
+      "step": 18620
+    },
+    {
+      "epoch": 3.9064793457747955,
+      "grad_norm": 10.524168014526367,
+      "learning_rate": 0.0002,
+      "loss": 0.3351,
+      "step": 18630
+    },
+    {
+      "epoch": 3.908576221430069,
+      "grad_norm": 0.1589958518743515,
+      "learning_rate": 0.0002,
+      "loss": 0.2456,
+      "step": 18640
+    },
+    {
+      "epoch": 3.910673097085343,
+      "grad_norm": 0.1514245867729187,
+      "learning_rate": 0.0002,
+      "loss": 0.2736,
+      "step": 18650
+    },
+    {
+      "epoch": 3.9127699727406164,
+      "grad_norm": 0.2135051190853119,
+      "learning_rate": 0.0002,
+      "loss": 0.2448,
+      "step": 18660
+    },
+    {
+      "epoch": 3.9148668483958904,
+      "grad_norm": 0.35259103775024414,
+      "learning_rate": 0.0002,
+      "loss": 0.2454,
+      "step": 18670
+    },
+    {
+      "epoch": 3.916963724051164,
+      "grad_norm": 0.27322497963905334,
+      "learning_rate": 0.0002,
+      "loss": 0.3172,
+      "step": 18680
+    },
+    {
+      "epoch": 3.9190605997064374,
+      "grad_norm": 7.8673906326293945,
+      "learning_rate": 0.0002,
+      "loss": 0.323,
+      "step": 18690
+    },
+    {
+      "epoch": 3.921157475361711,
+      "grad_norm": 4.626621723175049,
+      "learning_rate": 0.0002,
+      "loss": 0.2822,
+      "step": 18700
+    },
+    {
+      "epoch": 3.923254351016985,
+      "grad_norm": 0.13362011313438416,
+      "learning_rate": 0.0002,
+      "loss": 0.2921,
+      "step": 18710
+    },
+    {
+      "epoch": 3.9253512266722583,
+      "grad_norm": 0.2543129622936249,
+      "learning_rate": 0.0002,
+      "loss": 0.3575,
+      "step": 18720
+    },
+    {
+      "epoch": 3.927448102327532,
+      "grad_norm": 0.21031315624713898,
+      "learning_rate": 0.0002,
+      "loss": 0.3077,
+      "step": 18730
+    },
+    {
+      "epoch": 3.9295449779828058,
+      "grad_norm": 1.1506986618041992,
+      "learning_rate": 0.0002,
+      "loss": 0.2742,
+      "step": 18740
+    },
+    {
+      "epoch": 3.9316418536380793,
+      "grad_norm": 9.54204273223877,
+      "learning_rate": 0.0002,
+      "loss": 0.3883,
+      "step": 18750
+    },
+    {
+      "epoch": 3.9337387292933528,
+      "grad_norm": 0.9476527571678162,
+      "learning_rate": 0.0002,
+      "loss": 0.2547,
+      "step": 18760
+    },
+    {
+      "epoch": 3.9358356049486263,
+      "grad_norm": 0.33886954188346863,
+      "learning_rate": 0.0002,
+      "loss": 0.3091,
+      "step": 18770
+    },
+    {
+      "epoch": 3.9379324806039,
+      "grad_norm": 0.41044721007347107,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 18780
+    },
+    {
+      "epoch": 3.9400293562591737,
+      "grad_norm": 13.303047180175781,
+      "learning_rate": 0.0002,
+      "loss": 0.3628,
+      "step": 18790
+    },
+    {
+      "epoch": 3.9421262319144477,
+      "grad_norm": 0.8790237307548523,
+      "learning_rate": 0.0002,
+      "loss": 0.2819,
+      "step": 18800
+    },
+    {
+      "epoch": 3.944223107569721,
+      "grad_norm": 0.29523730278015137,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 18810
+    },
+    {
+      "epoch": 3.9463199832249947,
+      "grad_norm": 1.0903148651123047,
+      "learning_rate": 0.0002,
+      "loss": 0.317,
+      "step": 18820
+    },
+    {
+      "epoch": 3.948416858880268,
+      "grad_norm": 2.2300570011138916,
+      "learning_rate": 0.0002,
+      "loss": 0.3695,
+      "step": 18830
+    },
+    {
+      "epoch": 3.950513734535542,
+      "grad_norm": 9.108022689819336,
+      "learning_rate": 0.0002,
+      "loss": 0.253,
+      "step": 18840
+    },
+    {
+      "epoch": 3.9526106101908156,
+      "grad_norm": 0.383888304233551,
+      "learning_rate": 0.0002,
+      "loss": 0.2647,
+      "step": 18850
+    },
+    {
+      "epoch": 3.9547074858460896,
+      "grad_norm": 9.491477012634277,
+      "learning_rate": 0.0002,
+      "loss": 0.2457,
+      "step": 18860
+    },
+    {
+      "epoch": 3.956804361501363,
+      "grad_norm": 0.9953404068946838,
+      "learning_rate": 0.0002,
+      "loss": 0.3413,
+      "step": 18870
+    },
+    {
+      "epoch": 3.9589012371566366,
+      "grad_norm": 0.2776716649532318,
+      "learning_rate": 0.0002,
+      "loss": 0.4267,
+      "step": 18880
+    },
+    {
+      "epoch": 3.96099811281191,
+      "grad_norm": 2.3032448291778564,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 18890
+    },
+    {
+      "epoch": 3.963094988467184,
+      "grad_norm": 0.35753384232521057,
+      "learning_rate": 0.0002,
+      "loss": 0.3361,
+      "step": 18900
+    },
+    {
+      "epoch": 3.9651918641224575,
+      "grad_norm": 2.1207659244537354,
+      "learning_rate": 0.0002,
+      "loss": 0.366,
+      "step": 18910
+    },
+    {
+      "epoch": 3.9672887397777314,
+      "grad_norm": 16.24346923828125,
+      "learning_rate": 0.0002,
+      "loss": 0.323,
+      "step": 18920
+    },
+    {
+      "epoch": 3.969385615433005,
+      "grad_norm": 0.1958165466785431,
+      "learning_rate": 0.0002,
+      "loss": 0.2467,
+      "step": 18930
+    },
+    {
+      "epoch": 3.9714824910882784,
+      "grad_norm": 0.3013644218444824,
+      "learning_rate": 0.0002,
+      "loss": 0.315,
+      "step": 18940
+    },
+    {
+      "epoch": 3.973579366743552,
+      "grad_norm": 9.605280876159668,
+      "learning_rate": 0.0002,
+      "loss": 0.2968,
+      "step": 18950
+    },
+    {
+      "epoch": 3.975676242398826,
+      "grad_norm": 4.225147724151611,
+      "learning_rate": 0.0002,
+      "loss": 0.2526,
+      "step": 18960
+    },
+    {
+      "epoch": 3.9777731180540994,
+      "grad_norm": 3.812077045440674,
+      "learning_rate": 0.0002,
+      "loss": 0.2869,
+      "step": 18970
+    },
+    {
+      "epoch": 3.979869993709373,
+      "grad_norm": 1.702895164489746,
+      "learning_rate": 0.0002,
+      "loss": 0.2474,
+      "step": 18980
+    },
+    {
+      "epoch": 3.981966869364647,
+      "grad_norm": 0.22785484790802002,
+      "learning_rate": 0.0002,
+      "loss": 0.335,
+      "step": 18990
+    },
+    {
+      "epoch": 3.9840637450199203,
+      "grad_norm": 2.7882556915283203,
+      "learning_rate": 0.0002,
+      "loss": 0.248,
+      "step": 19000
+    },
+    {
+      "epoch": 3.986160620675194,
+      "grad_norm": 7.80819034576416,
+      "learning_rate": 0.0002,
+      "loss": 0.3786,
+      "step": 19010
+    },
+    {
+      "epoch": 3.9882574963304673,
+      "grad_norm": 0.829639732837677,
+      "learning_rate": 0.0002,
+      "loss": 0.3423,
+      "step": 19020
+    },
+    {
+      "epoch": 3.9903543719857413,
+      "grad_norm": 4.340249538421631,
+      "learning_rate": 0.0002,
+      "loss": 0.2621,
+      "step": 19030
+    },
+    {
+      "epoch": 3.9924512476410148,
+      "grad_norm": 3.492180824279785,
+      "learning_rate": 0.0002,
+      "loss": 0.2488,
+      "step": 19040
+    },
+    {
+      "epoch": 3.9945481232962887,
+      "grad_norm": 1.187342882156372,
+      "learning_rate": 0.0002,
+      "loss": 0.2897,
+      "step": 19050
+    },
+    {
+      "epoch": 3.9966449989515622,
+      "grad_norm": 0.2959229052066803,
+      "learning_rate": 0.0002,
+      "loss": 0.3287,
+      "step": 19060
+    },
+    {
+      "epoch": 3.9987418746068357,
+      "grad_norm": 2.1524133682250977,
+      "learning_rate": 0.0002,
+      "loss": 0.2873,
+      "step": 19070
+    },
+    {
+      "epoch": 4.000838750262109,
+      "grad_norm": 0.7680649757385254,
+      "learning_rate": 0.0002,
+      "loss": 0.2441,
+      "step": 19080
+    },
+    {
+      "epoch": 4.002935625917383,
+      "grad_norm": 3.3124094009399414,
+      "learning_rate": 0.0002,
+      "loss": 0.246,
+      "step": 19090
+    },
+    {
+      "epoch": 4.005032501572657,
+      "grad_norm": 8.343110084533691,
+      "learning_rate": 0.0002,
+      "loss": 0.2892,
+      "step": 19100
+    },
+    {
+      "epoch": 4.007129377227931,
+      "grad_norm": 0.46529117226600647,
+      "learning_rate": 0.0002,
+      "loss": 0.3646,
+      "step": 19110
+    },
+    {
+      "epoch": 4.009226252883204,
+      "grad_norm": 1.364998698234558,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 19120
+    },
+    {
+      "epoch": 4.011323128538478,
+      "grad_norm": 1.5017606019973755,
+      "learning_rate": 0.0002,
+      "loss": 0.4226,
+      "step": 19130
+    },
+    {
+      "epoch": 4.013420004193751,
+      "grad_norm": 4.142571926116943,
+      "learning_rate": 0.0002,
+      "loss": 0.3087,
+      "step": 19140
+    },
+    {
+      "epoch": 4.015516879849025,
+      "grad_norm": 8.241575241088867,
+      "learning_rate": 0.0002,
+      "loss": 0.2812,
+      "step": 19150
+    },
+    {
+      "epoch": 4.017613755504299,
+      "grad_norm": 3.2250149250030518,
+      "learning_rate": 0.0002,
+      "loss": 0.3269,
+      "step": 19160
+    },
+    {
+      "epoch": 4.0197106311595725,
+      "grad_norm": 0.6055905222892761,
+      "learning_rate": 0.0002,
+      "loss": 0.2535,
+      "step": 19170
+    },
+    {
+      "epoch": 4.021807506814846,
+      "grad_norm": 0.2831921875476837,
+      "learning_rate": 0.0002,
+      "loss": 0.2572,
+      "step": 19180
+    },
+    {
+      "epoch": 4.0239043824701195,
+      "grad_norm": 0.3150438964366913,
+      "learning_rate": 0.0002,
+      "loss": 0.3007,
+      "step": 19190
+    },
+    {
+      "epoch": 4.026001258125393,
+      "grad_norm": 12.150964736938477,
+      "learning_rate": 0.0002,
+      "loss": 0.281,
+      "step": 19200
+    },
+    {
+      "epoch": 4.0280981337806665,
+      "grad_norm": 13.298822402954102,
+      "learning_rate": 0.0002,
+      "loss": 0.3141,
+      "step": 19210
+    },
+    {
+      "epoch": 4.030195009435941,
+      "grad_norm": 0.38828134536743164,
+      "learning_rate": 0.0002,
+      "loss": 0.254,
+      "step": 19220
+    },
+    {
+      "epoch": 4.032291885091214,
+      "grad_norm": 0.5457937717437744,
+      "learning_rate": 0.0002,
+      "loss": 0.2458,
+      "step": 19230
+    },
+    {
+      "epoch": 4.034388760746488,
+      "grad_norm": 0.26707398891448975,
+      "learning_rate": 0.0002,
+      "loss": 0.249,
+      "step": 19240
+    },
+    {
+      "epoch": 4.036485636401761,
+      "grad_norm": 26.078208923339844,
+      "learning_rate": 0.0002,
+      "loss": 0.3605,
+      "step": 19250
+    },
+    {
+      "epoch": 4.038582512057035,
+      "grad_norm": 0.1831461787223816,
+      "learning_rate": 0.0002,
+      "loss": 0.2754,
+      "step": 19260
+    },
+    {
+      "epoch": 4.040679387712308,
+      "grad_norm": 3.2851014137268066,
+      "learning_rate": 0.0002,
+      "loss": 0.3133,
+      "step": 19270
+    },
+    {
+      "epoch": 4.042776263367582,
+      "grad_norm": 6.421793460845947,
+      "learning_rate": 0.0002,
+      "loss": 0.3104,
+      "step": 19280
+    },
+    {
+      "epoch": 4.044873139022856,
+      "grad_norm": 22.904735565185547,
+      "learning_rate": 0.0002,
+      "loss": 0.3006,
+      "step": 19290
+    },
+    {
+      "epoch": 4.04697001467813,
+      "grad_norm": 9.35460090637207,
+      "learning_rate": 0.0002,
+      "loss": 0.3001,
+      "step": 19300
+    },
+    {
+      "epoch": 4.049066890333403,
+      "grad_norm": 2.4891185760498047,
+      "learning_rate": 0.0002,
+      "loss": 0.3267,
+      "step": 19310
+    },
+    {
+      "epoch": 4.051163765988677,
+      "grad_norm": 3.316619396209717,
+      "learning_rate": 0.0002,
+      "loss": 0.248,
+      "step": 19320
+    },
+    {
+      "epoch": 4.05326064164395,
+      "grad_norm": 8.631692886352539,
+      "learning_rate": 0.0002,
+      "loss": 0.3136,
+      "step": 19330
+    },
+    {
+      "epoch": 4.055357517299224,
+      "grad_norm": 1.2316213846206665,
+      "learning_rate": 0.0002,
+      "loss": 0.4906,
+      "step": 19340
+    },
+    {
+      "epoch": 4.057454392954498,
+      "grad_norm": 0.4751741588115692,
+      "learning_rate": 0.0002,
+      "loss": 0.2471,
+      "step": 19350
+    },
+    {
+      "epoch": 4.059551268609772,
+      "grad_norm": 2.4287467002868652,
+      "learning_rate": 0.0002,
+      "loss": 0.3005,
+      "step": 19360
+    },
+    {
+      "epoch": 4.061648144265045,
+      "grad_norm": 5.299201488494873,
+      "learning_rate": 0.0002,
+      "loss": 0.2505,
+      "step": 19370
+    },
+    {
+      "epoch": 4.063745019920319,
+      "grad_norm": 0.45759066939353943,
+      "learning_rate": 0.0002,
+      "loss": 0.2545,
+      "step": 19380
+    },
+    {
+      "epoch": 4.065841895575592,
+      "grad_norm": 8.147871017456055,
+      "learning_rate": 0.0002,
+      "loss": 0.3042,
+      "step": 19390
+    },
+    {
+      "epoch": 4.067938771230866,
+      "grad_norm": 3.946077823638916,
+      "learning_rate": 0.0002,
+      "loss": 0.3312,
+      "step": 19400
+    },
+    {
+      "epoch": 4.07003564688614,
+      "grad_norm": 24.572731018066406,
+      "learning_rate": 0.0002,
+      "loss": 0.4547,
+      "step": 19410
+    },
+    {
+      "epoch": 4.072132522541414,
+      "grad_norm": 3.766494035720825,
+      "learning_rate": 0.0002,
+      "loss": 0.2459,
+      "step": 19420
+    },
+    {
+      "epoch": 4.074229398196687,
+      "grad_norm": 7.838687419891357,
+      "learning_rate": 0.0002,
+      "loss": 0.4842,
+      "step": 19430
+    },
+    {
+      "epoch": 4.076326273851961,
+      "grad_norm": 0.370241641998291,
+      "learning_rate": 0.0002,
+      "loss": 0.2907,
+      "step": 19440
+    },
+    {
+      "epoch": 4.078423149507234,
+      "grad_norm": 0.26800379157066345,
+      "learning_rate": 0.0002,
+      "loss": 0.2482,
+      "step": 19450
+    },
+    {
+      "epoch": 4.080520025162508,
+      "grad_norm": 0.5113723874092102,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 19460
+    },
+    {
+      "epoch": 4.082616900817781,
+      "grad_norm": 0.381949245929718,
+      "learning_rate": 0.0002,
+      "loss": 0.3141,
+      "step": 19470
+    },
+    {
+      "epoch": 4.0847137764730554,
+      "grad_norm": 0.7983457446098328,
+      "learning_rate": 0.0002,
+      "loss": 0.2473,
+      "step": 19480
+    },
+    {
+      "epoch": 4.086810652128329,
+      "grad_norm": 0.41288188099861145,
+      "learning_rate": 0.0002,
+      "loss": 0.2446,
+      "step": 19490
+    },
+    {
+      "epoch": 4.0889075277836024,
+      "grad_norm": 0.24723954498767853,
+      "learning_rate": 0.0002,
+      "loss": 0.2868,
+      "step": 19500
+    },
+    {
+      "epoch": 4.091004403438876,
+      "grad_norm": 0.1693434864282608,
+      "learning_rate": 0.0002,
+      "loss": 0.2461,
+      "step": 19510
+    },
+    {
+      "epoch": 4.0931012790941494,
+      "grad_norm": 0.557501494884491,
+      "learning_rate": 0.0002,
+      "loss": 0.3037,
+      "step": 19520
+    },
+    {
+      "epoch": 4.095198154749423,
+      "grad_norm": 0.6941953897476196,
+      "learning_rate": 0.0002,
+      "loss": 0.2497,
+      "step": 19530
+    },
+    {
+      "epoch": 4.097295030404697,
+      "grad_norm": 0.5740838050842285,
+      "learning_rate": 0.0002,
+      "loss": 0.2511,
+      "step": 19540
+    },
+    {
+      "epoch": 4.099391906059971,
+      "grad_norm": 1.7676140069961548,
+      "learning_rate": 0.0002,
+      "loss": 0.2954,
+      "step": 19550
+    },
+    {
+      "epoch": 4.101488781715244,
+      "grad_norm": 5.540597915649414,
+      "learning_rate": 0.0002,
+      "loss": 0.3015,
+      "step": 19560
+    },
+    {
+      "epoch": 4.103585657370518,
+      "grad_norm": 5.951892375946045,
+      "learning_rate": 0.0002,
+      "loss": 0.3055,
+      "step": 19570
+    },
+    {
+      "epoch": 4.105682533025791,
+      "grad_norm": 11.446410179138184,
+      "learning_rate": 0.0002,
+      "loss": 0.2508,
+      "step": 19580
+    },
+    {
+      "epoch": 4.107779408681065,
+      "grad_norm": 0.43041107058525085,
+      "learning_rate": 0.0002,
+      "loss": 0.3027,
+      "step": 19590
+    },
+    {
+      "epoch": 4.109876284336339,
+      "grad_norm": 0.48323339223861694,
+      "learning_rate": 0.0002,
+      "loss": 0.2867,
+      "step": 19600
+    },
+    {
+      "epoch": 4.111973159991613,
+      "grad_norm": 0.5927427411079407,
+      "learning_rate": 0.0002,
+      "loss": 0.3368,
+      "step": 19610
+    },
+    {
+      "epoch": 4.114070035646886,
+      "grad_norm": 2.952925205230713,
+      "learning_rate": 0.0002,
+      "loss": 0.3301,
+      "step": 19620
+    },
+    {
+      "epoch": 4.11616691130216,
+      "grad_norm": 4.207739353179932,
+      "learning_rate": 0.0002,
+      "loss": 0.3181,
+      "step": 19630
+    },
+    {
+      "epoch": 4.118263786957433,
+      "grad_norm": 2.4143640995025635,
+      "learning_rate": 0.0002,
+      "loss": 0.3088,
+      "step": 19640
+    },
+    {
+      "epoch": 4.120360662612707,
+      "grad_norm": 0.6868274211883545,
+      "learning_rate": 0.0002,
+      "loss": 0.2837,
+      "step": 19650
+    },
+    {
+      "epoch": 4.122457538267981,
+      "grad_norm": 1.6815471649169922,
+      "learning_rate": 0.0002,
+      "loss": 0.2477,
+      "step": 19660
+    },
+    {
+      "epoch": 4.124554413923255,
+      "grad_norm": 0.845425009727478,
+      "learning_rate": 0.0002,
+      "loss": 0.3732,
+      "step": 19670
+    },
+    {
+      "epoch": 4.126651289578528,
+      "grad_norm": 2.966221332550049,
+      "learning_rate": 0.0002,
+      "loss": 0.25,
+      "step": 19680
+    },
+    {
+      "epoch": 4.128748165233802,
+      "grad_norm": 0.26147690415382385,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 19690
+    },
+    {
+      "epoch": 4.130845040889075,
+      "grad_norm": 1.215938925743103,
+      "learning_rate": 0.0002,
+      "loss": 0.3892,
+      "step": 19700
+    },
+    {
+      "epoch": 4.132941916544349,
+      "grad_norm": 0.2903278172016144,
+      "learning_rate": 0.0002,
+      "loss": 0.308,
+      "step": 19710
+    },
+    {
+      "epoch": 4.135038792199623,
+      "grad_norm": 0.33962321281433105,
+      "learning_rate": 0.0002,
+      "loss": 0.3393,
+      "step": 19720
+    },
+    {
+      "epoch": 4.1371356678548965,
+      "grad_norm": 2.0429956912994385,
+      "learning_rate": 0.0002,
+      "loss": 0.2933,
+      "step": 19730
+    },
+    {
+      "epoch": 4.13923254351017,
+      "grad_norm": 3.423198938369751,
+      "learning_rate": 0.0002,
+      "loss": 0.3338,
+      "step": 19740
+    },
+    {
+      "epoch": 4.1413294191654435,
+      "grad_norm": 0.21673530340194702,
+      "learning_rate": 0.0002,
+      "loss": 0.3316,
+      "step": 19750
+    },
+    {
+      "epoch": 4.143426294820717,
+      "grad_norm": 23.13585090637207,
+      "learning_rate": 0.0002,
+      "loss": 0.2715,
+      "step": 19760
+    },
+    {
+      "epoch": 4.1455231704759905,
+      "grad_norm": 18.557872772216797,
+      "learning_rate": 0.0002,
+      "loss": 0.3716,
+      "step": 19770
+    },
+    {
+      "epoch": 4.147620046131264,
+      "grad_norm": 32.08924865722656,
+      "learning_rate": 0.0002,
+      "loss": 0.37,
+      "step": 19780
+    },
+    {
+      "epoch": 4.149716921786538,
+      "grad_norm": 0.26837587356567383,
+      "learning_rate": 0.0002,
+      "loss": 0.2617,
+      "step": 19790
+    },
+    {
+      "epoch": 4.151813797441812,
+      "grad_norm": 0.33404791355133057,
+      "learning_rate": 0.0002,
+      "loss": 0.2472,
+      "step": 19800
+    },
+    {
+      "epoch": 4.153910673097085,
+      "grad_norm": 0.5519550442695618,
+      "learning_rate": 0.0002,
+      "loss": 0.4107,
+      "step": 19810
+    },
+    {
+      "epoch": 4.156007548752359,
+      "grad_norm": 2.7054030895233154,
+      "learning_rate": 0.0002,
+      "loss": 0.2485,
+      "step": 19820
+    },
+    {
+      "epoch": 4.158104424407632,
+      "grad_norm": 0.9935354590415955,
+      "learning_rate": 0.0002,
+      "loss": 0.3529,
+      "step": 19830
+    },
+    {
+      "epoch": 4.160201300062906,
+      "grad_norm": 0.3448370397090912,
+      "learning_rate": 0.0002,
+      "loss": 0.3026,
+      "step": 19840
+    },
+    {
+      "epoch": 4.16229817571818,
+      "grad_norm": 21.877988815307617,
+      "learning_rate": 0.0002,
+      "loss": 0.3752,
+      "step": 19850
+    },
+    {
+      "epoch": 4.164395051373454,
+      "grad_norm": 1.063585638999939,
+      "learning_rate": 0.0002,
+      "loss": 0.2481,
+      "step": 19860
+    },
+    {
+      "epoch": 4.166491927028727,
+      "grad_norm": 0.8071320056915283,
+      "learning_rate": 0.0002,
+      "loss": 0.3006,
+      "step": 19870
+    },
+    {
+      "epoch": 4.168588802684001,
+      "grad_norm": 0.8395910859107971,
+      "learning_rate": 0.0002,
+      "loss": 0.2501,
+      "step": 19880
+    },
+    {
+      "epoch": 4.170685678339274,
+      "grad_norm": 0.6836159825325012,
+      "learning_rate": 0.0002,
+      "loss": 0.2925,
+      "step": 19890
+    },
+    {
+      "epoch": 4.172782553994548,
+      "grad_norm": 0.48117589950561523,
+      "learning_rate": 0.0002,
+      "loss": 0.3058,
+      "step": 19900
+    },
+    {
+      "epoch": 4.174879429649822,
+      "grad_norm": 0.6200854778289795,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 19910
+    },
+    {
+      "epoch": 4.176976305305096,
+      "grad_norm": 15.760970115661621,
+      "learning_rate": 0.0002,
+      "loss": 0.3091,
+      "step": 19920
+    },
+    {
+      "epoch": 4.179073180960369,
+      "grad_norm": 27.362468719482422,
+      "learning_rate": 0.0002,
+      "loss": 0.445,
+      "step": 19930
+    },
+    {
+      "epoch": 4.181170056615643,
+      "grad_norm": 1.0544220209121704,
+      "learning_rate": 0.0002,
+      "loss": 0.3394,
+      "step": 19940
+    },
+    {
+      "epoch": 4.183266932270916,
+      "grad_norm": 8.582993507385254,
+      "learning_rate": 0.0002,
+      "loss": 0.2508,
+      "step": 19950
+    },
+    {
+      "epoch": 4.18536380792619,
+      "grad_norm": 0.3823242485523224,
+      "learning_rate": 0.0002,
+      "loss": 0.5184,
+      "step": 19960
+    },
+    {
+      "epoch": 4.187460683581463,
+      "grad_norm": 0.1922634392976761,
+      "learning_rate": 0.0002,
+      "loss": 0.2462,
+      "step": 19970
+    },
+    {
+      "epoch": 4.189557559236738,
+      "grad_norm": 0.3688085377216339,
+      "learning_rate": 0.0002,
+      "loss": 0.2484,
+      "step": 19980
+    },
+    {
+      "epoch": 4.191654434892011,
+      "grad_norm": 0.38985419273376465,
+      "learning_rate": 0.0002,
+      "loss": 0.247,
+      "step": 19990
+    },
+    {
+      "epoch": 4.193751310547285,
+      "grad_norm": 14.897371292114258,
+      "learning_rate": 0.0002,
+      "loss": 0.2517,
+      "step": 20000
+    },
+    {
+      "epoch": 4.195848186202558,
+      "grad_norm": 0.27810803055763245,
+      "learning_rate": 0.0002,
+      "loss": 0.2945,
+      "step": 20010
+    },
+    {
+      "epoch": 4.197945061857832,
+      "grad_norm": 2.055335283279419,
+      "learning_rate": 0.0002,
+      "loss": 0.3384,
+      "step": 20020
+    },
+    {
+      "epoch": 4.200041937513105,
+      "grad_norm": 35.41399383544922,
+      "learning_rate": 0.0002,
+      "loss": 0.2872,
+      "step": 20030
+    },
+    {
+      "epoch": 4.2021388131683794,
+      "grad_norm": 0.3352704644203186,
+      "learning_rate": 0.0002,
+      "loss": 0.25,
+      "step": 20040
+    },
+    {
+      "epoch": 4.204235688823653,
+      "grad_norm": 10.434800148010254,
+      "learning_rate": 0.0002,
+      "loss": 0.3184,
+      "step": 20050
+    },
+    {
+      "epoch": 4.2063325644789265,
+      "grad_norm": 0.7769026756286621,
+      "learning_rate": 0.0002,
+      "loss": 0.4273,
+      "step": 20060
+    },
+    {
+      "epoch": 4.2084294401342,
+      "grad_norm": 0.48685234785079956,
+      "learning_rate": 0.0002,
+      "loss": 0.3596,
+      "step": 20070
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 8.412596702575684,
+      "learning_rate": 0.0002,
+      "loss": 0.2924,
+      "step": 20080
+    },
+    {
+      "epoch": 4.212623191444747,
+      "grad_norm": 7.435706615447998,
+      "learning_rate": 0.0002,
+      "loss": 0.3786,
+      "step": 20090
+    },
+    {
+      "epoch": 4.214720067100021,
+      "grad_norm": 0.4827675223350525,
+      "learning_rate": 0.0002,
+      "loss": 0.3096,
+      "step": 20100
+    },
+    {
+      "epoch": 4.216816942755295,
+      "grad_norm": 2.754267692565918,
+      "learning_rate": 0.0002,
+      "loss": 0.3012,
+      "step": 20110
+    },
+    {
+      "epoch": 4.218913818410568,
+      "grad_norm": 0.3863971531391144,
+      "learning_rate": 0.0002,
+      "loss": 0.3547,
+      "step": 20120
+    },
+    {
+      "epoch": 4.221010694065842,
+      "grad_norm": 0.3589477837085724,
+      "learning_rate": 0.0002,
+      "loss": 0.2464,
+      "step": 20130
+    },
+    {
+      "epoch": 4.223107569721115,
+      "grad_norm": 9.28115177154541,
+      "learning_rate": 0.0002,
+      "loss": 0.3512,
+      "step": 20140
+    },
+    {
+      "epoch": 4.225204445376389,
+      "grad_norm": 1.1233867406845093,
+      "learning_rate": 0.0002,
+      "loss": 0.2903,
+      "step": 20150
+    },
+    {
+      "epoch": 4.227301321031663,
+      "grad_norm": 2.778885841369629,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 20160
+    },
+    {
+      "epoch": 4.229398196686937,
+      "grad_norm": 0.5360897779464722,
+      "learning_rate": 0.0002,
+      "loss": 0.2699,
+      "step": 20170
+    },
+    {
+      "epoch": 4.23149507234221,
+      "grad_norm": 16.425745010375977,
+      "learning_rate": 0.0002,
+      "loss": 0.3357,
+      "step": 20180
+    },
+    {
+      "epoch": 4.233591947997484,
+      "grad_norm": 20.68720054626465,
+      "learning_rate": 0.0002,
+      "loss": 0.3336,
+      "step": 20190
+    },
+    {
+      "epoch": 4.235688823652757,
+      "grad_norm": 8.801888465881348,
+      "learning_rate": 0.0002,
+      "loss": 0.2671,
+      "step": 20200
+    },
+    {
+      "epoch": 4.237785699308031,
+      "grad_norm": 5.202653884887695,
+      "learning_rate": 0.0002,
+      "loss": 0.3124,
+      "step": 20210
+    },
+    {
+      "epoch": 4.239882574963305,
+      "grad_norm": 20.8682804107666,
+      "learning_rate": 0.0002,
+      "loss": 0.2606,
+      "step": 20220
+    },
+    {
+      "epoch": 4.241979450618579,
+      "grad_norm": 2.6714205741882324,
+      "learning_rate": 0.0002,
+      "loss": 0.2463,
+      "step": 20230
+    },
+    {
+      "epoch": 4.244076326273852,
+      "grad_norm": 20.23837661743164,
+      "learning_rate": 0.0002,
+      "loss": 0.3427,
+      "step": 20240
+    },
+    {
+      "epoch": 4.246173201929126,
+      "grad_norm": 2.1679625511169434,
+      "learning_rate": 0.0002,
+      "loss": 0.3103,
+      "step": 20250
+    },
+    {
+      "epoch": 4.248270077584399,
+      "grad_norm": 27.004417419433594,
+      "learning_rate": 0.0002,
+      "loss": 0.4334,
+      "step": 20260
+    },
+    {
+      "epoch": 4.250366953239673,
+      "grad_norm": 0.6717422008514404,
+      "learning_rate": 0.0002,
+      "loss": 0.2503,
+      "step": 20270
+    },
+    {
+      "epoch": 4.252463828894946,
+      "grad_norm": 0.48832905292510986,
+      "learning_rate": 0.0002,
+      "loss": 0.3939,
+      "step": 20280
+    },
+    {
+      "epoch": 4.2545607045502205,
+      "grad_norm": 19.908430099487305,
+      "learning_rate": 0.0002,
+      "loss": 0.312,
+      "step": 20290
+    },
+    {
+      "epoch": 4.256657580205494,
+      "grad_norm": 6.089963912963867,
+      "learning_rate": 0.0002,
+      "loss": 0.2523,
+      "step": 20300
+    },
+    {
+      "epoch": 4.2587544558607675,
+      "grad_norm": 6.796069622039795,
+      "learning_rate": 0.0002,
+      "loss": 0.2714,
+      "step": 20310
+    },
+    {
+      "epoch": 4.260851331516041,
+      "grad_norm": 9.671527862548828,
+      "learning_rate": 0.0002,
+      "loss": 0.3746,
+      "step": 20320
+    },
+    {
+      "epoch": 4.2629482071713145,
+      "grad_norm": 10.420218467712402,
+      "learning_rate": 0.0002,
+      "loss": 0.299,
+      "step": 20330
+    },
+    {
+      "epoch": 4.265045082826588,
+      "grad_norm": 6.379571437835693,
+      "learning_rate": 0.0002,
+      "loss": 0.2959,
+      "step": 20340
+    },
+    {
+      "epoch": 4.267141958481862,
+      "grad_norm": 0.3346550464630127,
+      "learning_rate": 0.0002,
+      "loss": 0.3061,
+      "step": 20350
+    },
+    {
+      "epoch": 4.269238834137136,
+      "grad_norm": 2.433529853820801,
+      "learning_rate": 0.0002,
+      "loss": 0.3757,
+      "step": 20360
+    },
+    {
+      "epoch": 4.271335709792409,
+      "grad_norm": 56.37220764160156,
+      "learning_rate": 0.0002,
+      "loss": 0.3016,
+      "step": 20370
+    },
+    {
+      "epoch": 4.273432585447683,
+      "grad_norm": 43.44303512573242,
+      "learning_rate": 0.0002,
+      "loss": 0.2864,
+      "step": 20380
+    },
+    {
+      "epoch": 4.275529461102956,
+      "grad_norm": 3.851576328277588,
+      "learning_rate": 0.0002,
+      "loss": 0.3883,
+      "step": 20390
+    },
+    {
+      "epoch": 4.27762633675823,
+      "grad_norm": 41.155879974365234,
+      "learning_rate": 0.0002,
+      "loss": 0.3486,
+      "step": 20400
+    },
+    {
+      "epoch": 4.279723212413504,
+      "grad_norm": 7.531223297119141,
+      "learning_rate": 0.0002,
+      "loss": 0.2479,
+      "step": 20410
+    },
+    {
+      "epoch": 4.281820088068778,
+      "grad_norm": 16.209299087524414,
+      "learning_rate": 0.0002,
+      "loss": 0.279,
+      "step": 20420
+    },
+    {
+      "epoch": 4.283916963724051,
+      "grad_norm": 15.871269226074219,
+      "learning_rate": 0.0002,
+      "loss": 0.3702,
+      "step": 20430
+    },
+    {
+      "epoch": 4.286013839379325,
+      "grad_norm": 26.545909881591797,
+      "learning_rate": 0.0002,
+      "loss": 0.2684,
+      "step": 20440
+    },
+    {
+      "epoch": 4.288110715034598,
+      "grad_norm": 17.186182022094727,
+      "learning_rate": 0.0002,
+      "loss": 0.2698,
+      "step": 20450
+    },
+    {
+      "epoch": 4.290207590689872,
+      "grad_norm": 18.780088424682617,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 20460
+    },
+    {
+      "epoch": 4.292304466345145,
+      "grad_norm": 624.7259521484375,
+      "learning_rate": 0.0002,
+      "loss": 0.9436,
+      "step": 20470
+    },
+    {
+      "epoch": 4.29440134200042,
+      "grad_norm": 16.724163055419922,
+      "learning_rate": 0.0002,
+      "loss": 1.1358,
+      "step": 20480
+    },
+    {
+      "epoch": 4.296498217655693,
+      "grad_norm": 4.588830471038818,
+      "learning_rate": 0.0002,
+      "loss": 0.249,
+      "step": 20490
+    },
+    {
+      "epoch": 4.298595093310967,
+      "grad_norm": 83.1984634399414,
+      "learning_rate": 0.0002,
+      "loss": 0.3602,
+      "step": 20500
+    },
+    {
+      "epoch": 4.30069196896624,
+      "grad_norm": 26.37981605529785,
+      "learning_rate": 0.0002,
+      "loss": 0.3221,
+      "step": 20510
+    },
+    {
+      "epoch": 4.302788844621514,
+      "grad_norm": 47.30278396606445,
+      "learning_rate": 0.0002,
+      "loss": 0.2829,
+      "step": 20520
+    },
+    {
+      "epoch": 4.304885720276787,
+      "grad_norm": 33.844642639160156,
+      "learning_rate": 0.0002,
+      "loss": 0.4587,
+      "step": 20530
+    },
+    {
+      "epoch": 4.306982595932062,
+      "grad_norm": 8.160712242126465,
+      "learning_rate": 0.0002,
+      "loss": 0.2647,
+      "step": 20540
+    },
+    {
+      "epoch": 4.309079471587335,
+      "grad_norm": 55.612396240234375,
+      "learning_rate": 0.0002,
+      "loss": 0.3565,
+      "step": 20550
+    },
+    {
+      "epoch": 4.311176347242609,
+      "grad_norm": 10.851349830627441,
+      "learning_rate": 0.0002,
+      "loss": 0.3704,
+      "step": 20560
+    },
+    {
+      "epoch": 4.313273222897882,
+      "grad_norm": 18.54105567932129,
+      "learning_rate": 0.0002,
+      "loss": 0.3417,
+      "step": 20570
+    },
+    {
+      "epoch": 4.315370098553156,
+      "grad_norm": 5.300592422485352,
+      "learning_rate": 0.0002,
+      "loss": 0.2883,
+      "step": 20580
+    },
+    {
+      "epoch": 4.317466974208429,
+      "grad_norm": 22.01258659362793,
+      "learning_rate": 0.0002,
+      "loss": 0.3597,
+      "step": 20590
+    },
+    {
+      "epoch": 4.3195638498637035,
+      "grad_norm": 50.189697265625,
+      "learning_rate": 0.0002,
+      "loss": 0.3421,
+      "step": 20600
+    },
+    {
+      "epoch": 4.321660725518977,
+      "grad_norm": 13.125137329101562,
+      "learning_rate": 0.0002,
+      "loss": 0.3829,
+      "step": 20610
+    },
+    {
+      "epoch": 4.3237576011742505,
+      "grad_norm": 64.3811264038086,
+      "learning_rate": 0.0002,
+      "loss": 0.391,
+      "step": 20620
+    },
+    {
+      "epoch": 4.325854476829524,
+      "grad_norm": 1.811645746231079,
+      "learning_rate": 0.0002,
+      "loss": 0.2639,
+      "step": 20630
+    },
+    {
+      "epoch": 4.3279513524847975,
+      "grad_norm": 12.494913101196289,
+      "learning_rate": 0.0002,
+      "loss": 0.2465,
+      "step": 20640
+    },
+    {
+      "epoch": 4.330048228140071,
+      "grad_norm": 7.9811201095581055,
+      "learning_rate": 0.0002,
+      "loss": 0.2627,
+      "step": 20650
+    },
+    {
+      "epoch": 4.332145103795345,
+      "grad_norm": 5.287240505218506,
+      "learning_rate": 0.0002,
+      "loss": 0.2506,
+      "step": 20660
+    },
+    {
+      "epoch": 4.334241979450619,
+      "grad_norm": 1.0475136041641235,
+      "learning_rate": 0.0002,
+      "loss": 0.2512,
+      "step": 20670
+    },
+    {
+      "epoch": 4.336338855105892,
+      "grad_norm": 10.19491195678711,
+      "learning_rate": 0.0002,
+      "loss": 0.3017,
+      "step": 20680
+    },
+    {
+      "epoch": 4.338435730761166,
+      "grad_norm": 67.91783142089844,
+      "learning_rate": 0.0002,
+      "loss": 0.3486,
+      "step": 20690
+    },
+    {
+      "epoch": 4.340532606416439,
+      "grad_norm": 3.169609785079956,
+      "learning_rate": 0.0002,
+      "loss": 0.3006,
+      "step": 20700
+    },
+    {
+      "epoch": 4.342629482071713,
+      "grad_norm": 8.942724227905273,
+      "learning_rate": 0.0002,
+      "loss": 0.2486,
+      "step": 20710
+    },
+    {
+      "epoch": 4.344726357726987,
+      "grad_norm": 7.17462682723999,
+      "learning_rate": 0.0002,
+      "loss": 0.2686,
+      "step": 20720
+    },
+    {
+      "epoch": 4.346823233382261,
+      "grad_norm": 6.14696741104126,
+      "learning_rate": 0.0002,
+      "loss": 0.3343,
+      "step": 20730
+    },
+    {
+      "epoch": 4.348920109037534,
+      "grad_norm": 16.166717529296875,
+      "learning_rate": 0.0002,
+      "loss": 0.3094,
+      "step": 20740
+    },
+    {
+      "epoch": 4.351016984692808,
+      "grad_norm": 36.18352508544922,
+      "learning_rate": 0.0002,
+      "loss": 0.4342,
+      "step": 20750
+    },
+    {
+      "epoch": 4.353113860348081,
+      "grad_norm": 3.9650731086730957,
+      "learning_rate": 0.0002,
+      "loss": 0.2763,
+      "step": 20760
+    },
+    {
+      "epoch": 4.355210736003355,
+      "grad_norm": 4.956878185272217,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 20770
+    },
+    {
+      "epoch": 4.357307611658628,
+      "grad_norm": 33.45060729980469,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 20780
+    },
+    {
+      "epoch": 4.359404487313903,
+      "grad_norm": 6.082530498504639,
+      "learning_rate": 0.0002,
+      "loss": 0.3175,
+      "step": 20790
+    },
+    {
+      "epoch": 4.361501362969176,
+      "grad_norm": 34.4377555847168,
+      "learning_rate": 0.0002,
+      "loss": 0.3072,
+      "step": 20800
+    },
+    {
+      "epoch": 4.36359823862445,
+      "grad_norm": 11.42460823059082,
+      "learning_rate": 0.0002,
+      "loss": 0.2476,
+      "step": 20810
+    },
+    {
+      "epoch": 4.365695114279723,
+      "grad_norm": 2.2365314960479736,
+      "learning_rate": 0.0002,
+      "loss": 0.2976,
+      "step": 20820
+    },
+    {
+      "epoch": 4.367791989934997,
+      "grad_norm": 12.357730865478516,
+      "learning_rate": 0.0002,
+      "loss": 0.3729,
+      "step": 20830
+    },
+    {
+      "epoch": 4.36988886559027,
+      "grad_norm": 1.3517647981643677,
+      "learning_rate": 0.0002,
+      "loss": 0.2908,
+      "step": 20840
+    },
+    {
+      "epoch": 4.3719857412455445,
+      "grad_norm": 2.2728474140167236,
+      "learning_rate": 0.0002,
+      "loss": 0.3596,
+      "step": 20850
+    },
+    {
+      "epoch": 4.374082616900818,
+      "grad_norm": 0.8363596200942993,
+      "learning_rate": 0.0002,
+      "loss": 0.2544,
+      "step": 20860
+    },
+    {
+      "epoch": 4.3761794925560915,
+      "grad_norm": 2.689927816390991,
+      "learning_rate": 0.0002,
+      "loss": 0.2475,
+      "step": 20870
+    },
+    {
+      "epoch": 4.378276368211365,
+      "grad_norm": 2.3501954078674316,
+      "learning_rate": 0.0002,
+      "loss": 0.3222,
+      "step": 20880
+    },
+    {
+      "epoch": 4.3803732438666385,
+      "grad_norm": 0.6565678715705872,
+      "learning_rate": 0.0002,
+      "loss": 0.2511,
+      "step": 20890
+    },
+    {
+      "epoch": 4.382470119521912,
+      "grad_norm": 12.648674011230469,
+      "learning_rate": 0.0002,
+      "loss": 0.2525,
+      "step": 20900
+    },
+    {
+      "epoch": 4.384566995177186,
+      "grad_norm": 60.73137283325195,
+      "learning_rate": 0.0002,
+      "loss": 0.435,
+      "step": 20910
+    },
+    {
+      "epoch": 4.38666387083246,
+      "grad_norm": 34.06174087524414,
+      "learning_rate": 0.0002,
+      "loss": 0.3655,
+      "step": 20920
+    },
+    {
+      "epoch": 4.388760746487733,
+      "grad_norm": 11.02375316619873,
+      "learning_rate": 0.0002,
+      "loss": 0.3604,
+      "step": 20930
+    },
+    {
+      "epoch": 4.390857622143007,
+      "grad_norm": 6.379631042480469,
+      "learning_rate": 0.0002,
+      "loss": 0.3405,
+      "step": 20940
+    },
+    {
+      "epoch": 4.39295449779828,
+      "grad_norm": 2.8621158599853516,
+      "learning_rate": 0.0002,
+      "loss": 0.304,
+      "step": 20950
+    },
+    {
+      "epoch": 4.395051373453554,
+      "grad_norm": 80.70250701904297,
+      "learning_rate": 0.0002,
+      "loss": 0.3652,
+      "step": 20960
+    },
+    {
+      "epoch": 4.397148249108827,
+      "grad_norm": 2.1578187942504883,
+      "learning_rate": 0.0002,
+      "loss": 0.3582,
+      "step": 20970
+    },
+    {
+      "epoch": 4.399245124764102,
+      "grad_norm": 6.415958404541016,
+      "learning_rate": 0.0002,
+      "loss": 0.3253,
+      "step": 20980
+    },
+    {
+      "epoch": 4.401342000419375,
+      "grad_norm": 21.651079177856445,
+      "learning_rate": 0.0002,
+      "loss": 0.3046,
+      "step": 20990
+    },
+    {
+      "epoch": 4.403438876074649,
+      "grad_norm": 9.651481628417969,
+      "learning_rate": 0.0002,
+      "loss": 0.2596,
+      "step": 21000
+    },
+    {
+      "epoch": 4.405535751729922,
+      "grad_norm": 14.337310791015625,
+      "learning_rate": 0.0002,
+      "loss": 0.4149,
+      "step": 21010
+    },
+    {
+      "epoch": 4.407632627385196,
+      "grad_norm": 43.099578857421875,
+      "learning_rate": 0.0002,
+      "loss": 0.2976,
+      "step": 21020
+    },
+    {
+      "epoch": 4.409729503040469,
+      "grad_norm": 417.2384948730469,
+      "learning_rate": 0.0002,
+      "loss": 0.689,
+      "step": 21030
+    },
+    {
+      "epoch": 4.411826378695744,
+      "grad_norm": 7.477783203125,
+      "learning_rate": 0.0002,
+      "loss": 0.2786,
+      "step": 21040
+    },
+    {
+      "epoch": 4.413923254351017,
+      "grad_norm": 17.863725662231445,
+      "learning_rate": 0.0002,
+      "loss": 0.3768,
+      "step": 21050
+    },
+    {
+      "epoch": 4.416020130006291,
+      "grad_norm": 56.08100509643555,
+      "learning_rate": 0.0002,
+      "loss": 0.3009,
+      "step": 21060
+    },
+    {
+      "epoch": 4.418117005661564,
+      "grad_norm": 22.835973739624023,
+      "learning_rate": 0.0002,
+      "loss": 0.3914,
+      "step": 21070
+    },
+    {
+      "epoch": 4.420213881316838,
+      "grad_norm": 39.187904357910156,
+      "learning_rate": 0.0002,
+      "loss": 0.3252,
+      "step": 21080
+    },
+    {
+      "epoch": 4.422310756972111,
+      "grad_norm": 18.0650634765625,
+      "learning_rate": 0.0002,
+      "loss": 0.268,
+      "step": 21090
+    },
+    {
+      "epoch": 4.424407632627386,
+      "grad_norm": 9.460448265075684,
+      "learning_rate": 0.0002,
+      "loss": 0.258,
+      "step": 21100
+    },
+    {
+      "epoch": 4.426504508282659,
+      "grad_norm": 3.0662641525268555,
+      "learning_rate": 0.0002,
+      "loss": 0.2507,
+      "step": 21110
+    },
+    {
+      "epoch": 4.428601383937933,
+      "grad_norm": 12.422566413879395,
+      "learning_rate": 0.0002,
+      "loss": 0.3051,
+      "step": 21120
+    },
+    {
+      "epoch": 4.430698259593206,
+      "grad_norm": 27.117107391357422,
+      "learning_rate": 0.0002,
+      "loss": 0.2727,
+      "step": 21130
+    },
+    {
+      "epoch": 4.43279513524848,
+      "grad_norm": 41.62672805786133,
+      "learning_rate": 0.0002,
+      "loss": 0.3167,
+      "step": 21140
+    },
+    {
+      "epoch": 4.434892010903753,
+      "grad_norm": 191.19381713867188,
+      "learning_rate": 0.0002,
+      "loss": 0.2917,
+      "step": 21150
+    },
+    {
+      "epoch": 4.4369888865590275,
+      "grad_norm": 7.5268120765686035,
+      "learning_rate": 0.0002,
+      "loss": 0.3105,
+      "step": 21160
+    },
+    {
+      "epoch": 4.439085762214301,
+      "grad_norm": 6.784496307373047,
+      "learning_rate": 0.0002,
+      "loss": 0.3932,
+      "step": 21170
+    },
+    {
+      "epoch": 4.4411826378695745,
+      "grad_norm": 3.4503252506256104,
+      "learning_rate": 0.0002,
+      "loss": 0.2953,
+      "step": 21180
+    },
+    {
+      "epoch": 4.443279513524848,
+      "grad_norm": 1.6427088975906372,
+      "learning_rate": 0.0002,
+      "loss": 0.3064,
+      "step": 21190
+    },
+    {
+      "epoch": 4.4453763891801215,
+      "grad_norm": 4.4671807289123535,
+      "learning_rate": 0.0002,
+      "loss": 0.3886,
+      "step": 21200
+    },
+    {
+      "epoch": 4.447473264835395,
+      "grad_norm": 10.051387786865234,
+      "learning_rate": 0.0002,
+      "loss": 0.3087,
+      "step": 21210
+    },
+    {
+      "epoch": 4.449570140490669,
+      "grad_norm": 19.033756256103516,
+      "learning_rate": 0.0002,
+      "loss": 0.4619,
+      "step": 21220
+    },
+    {
+      "epoch": 4.451667016145943,
+      "grad_norm": 15.37178897857666,
+      "learning_rate": 0.0002,
+      "loss": 0.4557,
+      "step": 21230
+    },
+    {
+      "epoch": 4.453763891801216,
+      "grad_norm": 5.2867279052734375,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 21240
+    },
+    {
+      "epoch": 4.45586076745649,
+      "grad_norm": 2.558028221130371,
+      "learning_rate": 0.0002,
+      "loss": 0.2647,
+      "step": 21250
+    },
+    {
+      "epoch": 4.457957643111763,
+      "grad_norm": 35.25192642211914,
+      "learning_rate": 0.0002,
+      "loss": 0.4654,
+      "step": 21260
+    },
+    {
+      "epoch": 4.460054518767037,
+      "grad_norm": 27.573955535888672,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 21270
+    },
+    {
+      "epoch": 4.46215139442231,
+      "grad_norm": 3.5311503410339355,
+      "learning_rate": 0.0002,
+      "loss": 0.4097,
+      "step": 21280
+    },
+    {
+      "epoch": 4.464248270077585,
+      "grad_norm": 10.815498352050781,
+      "learning_rate": 0.0002,
+      "loss": 0.4622,
+      "step": 21290
+    },
+    {
+      "epoch": 4.466345145732858,
+      "grad_norm": 8.656516075134277,
+      "learning_rate": 0.0002,
+      "loss": 0.2548,
+      "step": 21300
+    },
+    {
+      "epoch": 4.468442021388132,
+      "grad_norm": 47.45362091064453,
+      "learning_rate": 0.0002,
+      "loss": 0.3418,
+      "step": 21310
+    },
+    {
+      "epoch": 4.470538897043405,
+      "grad_norm": 15.242376327514648,
+      "learning_rate": 0.0002,
+      "loss": 0.3816,
+      "step": 21320
+    },
+    {
+      "epoch": 4.472635772698679,
+      "grad_norm": 13.443343162536621,
+      "learning_rate": 0.0002,
+      "loss": 0.3601,
+      "step": 21330
+    },
+    {
+      "epoch": 4.474732648353952,
+      "grad_norm": 94.393798828125,
+      "learning_rate": 0.0002,
+      "loss": 0.4557,
+      "step": 21340
+    },
+    {
+      "epoch": 4.476829524009227,
+      "grad_norm": 207.5814666748047,
+      "learning_rate": 0.0002,
+      "loss": 0.9471,
+      "step": 21350
+    },
+    {
+      "epoch": 4.4789263996645,
+      "grad_norm": 42.534278869628906,
+      "learning_rate": 0.0002,
+      "loss": 0.8521,
+      "step": 21360
+    },
+    {
+      "epoch": 4.481023275319774,
+      "grad_norm": 15.869437217712402,
+      "learning_rate": 0.0002,
+      "loss": 0.4873,
+      "step": 21370
+    },
+    {
+      "epoch": 4.483120150975047,
+      "grad_norm": 2.4193897247314453,
+      "learning_rate": 0.0002,
+      "loss": 0.3548,
+      "step": 21380
+    },
+    {
+      "epoch": 4.485217026630321,
+      "grad_norm": 17.73488426208496,
+      "learning_rate": 0.0002,
+      "loss": 0.2896,
+      "step": 21390
+    },
+    {
+      "epoch": 4.487313902285594,
+      "grad_norm": 48.30131149291992,
+      "learning_rate": 0.0002,
+      "loss": 0.3949,
+      "step": 21400
+    },
+    {
+      "epoch": 4.4894107779408685,
+      "grad_norm": 5.021493911743164,
+      "learning_rate": 0.0002,
+      "loss": 0.3309,
+      "step": 21410
+    },
+    {
+      "epoch": 4.491507653596142,
+      "grad_norm": 28.720203399658203,
+      "learning_rate": 0.0002,
+      "loss": 0.3268,
+      "step": 21420
+    },
+    {
+      "epoch": 4.4936045292514155,
+      "grad_norm": 2.697554111480713,
+      "learning_rate": 0.0002,
+      "loss": 0.3162,
+      "step": 21430
+    },
+    {
+      "epoch": 4.495701404906689,
+      "grad_norm": 57.5301513671875,
+      "learning_rate": 0.0002,
+      "loss": 0.3567,
+      "step": 21440
+    },
+    {
+      "epoch": 4.4977982805619625,
+      "grad_norm": 6.8275299072265625,
+      "learning_rate": 0.0002,
+      "loss": 0.3677,
+      "step": 21450
+    },
+    {
+      "epoch": 4.499895156217236,
+      "grad_norm": 46.234619140625,
+      "learning_rate": 0.0002,
+      "loss": 0.4583,
+      "step": 21460
+    },
+    {
+      "epoch": 4.5019920318725095,
+      "grad_norm": 6.242780685424805,
+      "learning_rate": 0.0002,
+      "loss": 0.2637,
+      "step": 21470
+    },
+    {
+      "epoch": 4.504088907527784,
+      "grad_norm": 2.4962878227233887,
+      "learning_rate": 0.0002,
+      "loss": 0.2564,
+      "step": 21480
+    },
+    {
+      "epoch": 4.506185783183057,
+      "grad_norm": 17.033349990844727,
+      "learning_rate": 0.0002,
+      "loss": 0.3303,
+      "step": 21490
+    },
+    {
+      "epoch": 4.508282658838331,
+      "grad_norm": 8.312446594238281,
+      "learning_rate": 0.0002,
+      "loss": 0.3764,
+      "step": 21500
+    },
+    {
+      "epoch": 4.510379534493604,
+      "grad_norm": 6.448761940002441,
+      "learning_rate": 0.0002,
+      "loss": 0.2504,
+      "step": 21510
+    },
+    {
+      "epoch": 4.512476410148878,
+      "grad_norm": 64.10653686523438,
+      "learning_rate": 0.0002,
+      "loss": 0.3032,
+      "step": 21520
+    },
+    {
+      "epoch": 4.514573285804152,
+      "grad_norm": 21.183940887451172,
+      "learning_rate": 0.0002,
+      "loss": 0.3583,
+      "step": 21530
+    },
+    {
+      "epoch": 4.516670161459426,
+      "grad_norm": 18.349613189697266,
+      "learning_rate": 0.0002,
+      "loss": 0.3541,
+      "step": 21540
+    },
+    {
+      "epoch": 4.518767037114699,
+      "grad_norm": 21.80953025817871,
+      "learning_rate": 0.0002,
+      "loss": 0.3975,
+      "step": 21550
+    },
+    {
+      "epoch": 4.520863912769973,
+      "grad_norm": 38.548728942871094,
+      "learning_rate": 0.0002,
+      "loss": 0.4618,
+      "step": 21560
+    },
+    {
+      "epoch": 4.522960788425246,
+      "grad_norm": 5.460489273071289,
+      "learning_rate": 0.0002,
+      "loss": 0.3965,
+      "step": 21570
+    },
+    {
+      "epoch": 4.52505766408052,
+      "grad_norm": 52.827781677246094,
+      "learning_rate": 0.0002,
+      "loss": 0.3507,
+      "step": 21580
+    },
+    {
+      "epoch": 4.527154539735793,
+      "grad_norm": 3.8467347621917725,
+      "learning_rate": 0.0002,
+      "loss": 0.3747,
+      "step": 21590
+    },
+    {
+      "epoch": 4.529251415391068,
+      "grad_norm": 7.897753715515137,
+      "learning_rate": 0.0002,
+      "loss": 0.2542,
+      "step": 21600
+    },
+    {
+      "epoch": 4.531348291046341,
+      "grad_norm": 11.481263160705566,
+      "learning_rate": 0.0002,
+      "loss": 0.3945,
+      "step": 21610
+    },
+    {
+      "epoch": 4.533445166701615,
+      "grad_norm": 1.1834608316421509,
+      "learning_rate": 0.0002,
+      "loss": 0.3417,
+      "step": 21620
+    },
+    {
+      "epoch": 4.535542042356888,
+      "grad_norm": 3.0790064334869385,
+      "learning_rate": 0.0002,
+      "loss": 0.5012,
+      "step": 21630
+    },
+    {
+      "epoch": 4.537638918012162,
+      "grad_norm": 5.346984386444092,
+      "learning_rate": 0.0002,
+      "loss": 0.3821,
+      "step": 21640
+    },
+    {
+      "epoch": 4.539735793667435,
+      "grad_norm": 12.952540397644043,
+      "learning_rate": 0.0002,
+      "loss": 0.4092,
+      "step": 21650
+    },
+    {
+      "epoch": 4.541832669322709,
+      "grad_norm": 39.88877487182617,
+      "learning_rate": 0.0002,
+      "loss": 0.4586,
+      "step": 21660
+    },
+    {
+      "epoch": 4.543929544977983,
+      "grad_norm": 41.47772216796875,
+      "learning_rate": 0.0002,
+      "loss": 0.3667,
+      "step": 21670
+    },
+    {
+      "epoch": 4.546026420633257,
+      "grad_norm": 15.996908187866211,
+      "learning_rate": 0.0002,
+      "loss": 0.308,
+      "step": 21680
+    },
+    {
+      "epoch": 4.54812329628853,
+      "grad_norm": 4.2574381828308105,
+      "learning_rate": 0.0002,
+      "loss": 0.3494,
+      "step": 21690
+    },
+    {
+      "epoch": 4.550220171943804,
+      "grad_norm": 16.708751678466797,
+      "learning_rate": 0.0002,
+      "loss": 0.4887,
+      "step": 21700
+    },
+    {
+      "epoch": 4.552317047599077,
+      "grad_norm": 2.954166889190674,
+      "learning_rate": 0.0002,
+      "loss": 0.3128,
+      "step": 21710
+    },
+    {
+      "epoch": 4.5544139232543515,
+      "grad_norm": 4.792957305908203,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 21720
+    },
+    {
+      "epoch": 4.556510798909625,
+      "grad_norm": 1.6357908248901367,
+      "learning_rate": 0.0002,
+      "loss": 0.279,
+      "step": 21730
+    },
+    {
+      "epoch": 4.5586076745648985,
+      "grad_norm": 13.55954647064209,
+      "learning_rate": 0.0002,
+      "loss": 0.4901,
+      "step": 21740
+    },
+    {
+      "epoch": 4.560704550220172,
+      "grad_norm": 0.8995484113693237,
+      "learning_rate": 0.0002,
+      "loss": 0.3749,
+      "step": 21750
+    },
+    {
+      "epoch": 4.5628014258754455,
+      "grad_norm": 5.566488265991211,
+      "learning_rate": 0.0002,
+      "loss": 0.3651,
+      "step": 21760
+    },
+    {
+      "epoch": 4.564898301530719,
+      "grad_norm": 17.96256446838379,
+      "learning_rate": 0.0002,
+      "loss": 0.2538,
+      "step": 21770
+    },
+    {
+      "epoch": 4.5669951771859925,
+      "grad_norm": 10.625880241394043,
+      "learning_rate": 0.0002,
+      "loss": 0.2498,
+      "step": 21780
+    },
+    {
+      "epoch": 4.569092052841267,
+      "grad_norm": 3.7621958255767822,
+      "learning_rate": 0.0002,
+      "loss": 0.3333,
+      "step": 21790
+    },
+    {
+      "epoch": 4.57118892849654,
+      "grad_norm": 5.985695838928223,
+      "learning_rate": 0.0002,
+      "loss": 0.2501,
+      "step": 21800
+    },
+    {
+      "epoch": 4.573285804151814,
+      "grad_norm": 18.665172576904297,
+      "learning_rate": 0.0002,
+      "loss": 0.4385,
+      "step": 21810
+    },
+    {
+      "epoch": 4.575382679807087,
+      "grad_norm": 11.269116401672363,
+      "learning_rate": 0.0002,
+      "loss": 0.2882,
+      "step": 21820
+    },
+    {
+      "epoch": 4.577479555462361,
+      "grad_norm": 125.940185546875,
+      "learning_rate": 0.0002,
+      "loss": 0.4505,
+      "step": 21830
+    },
+    {
+      "epoch": 4.579576431117634,
+      "grad_norm": 3.6832833290100098,
+      "learning_rate": 0.0002,
+      "loss": 0.5584,
+      "step": 21840
+    },
+    {
+      "epoch": 4.581673306772909,
+      "grad_norm": 26.39314079284668,
+      "learning_rate": 0.0002,
+      "loss": 0.4416,
+      "step": 21850
+    },
+    {
+      "epoch": 4.583770182428182,
+      "grad_norm": 3.046283006668091,
+      "learning_rate": 0.0002,
+      "loss": 0.2492,
+      "step": 21860
+    },
+    {
+      "epoch": 4.585867058083456,
+      "grad_norm": 4.790574073791504,
+      "learning_rate": 0.0002,
+      "loss": 0.2998,
+      "step": 21870
+    },
+    {
+      "epoch": 4.587963933738729,
+      "grad_norm": 21.731502532958984,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 21880
+    },
+    {
+      "epoch": 4.590060809394003,
+      "grad_norm": 1.8135603666305542,
+      "learning_rate": 0.0002,
+      "loss": 0.2608,
+      "step": 21890
+    },
+    {
+      "epoch": 4.592157685049276,
+      "grad_norm": 0.8779616951942444,
+      "learning_rate": 0.0002,
+      "loss": 0.2493,
+      "step": 21900
+    },
+    {
+      "epoch": 4.594254560704551,
+      "grad_norm": 6.280771732330322,
+      "learning_rate": 0.0002,
+      "loss": 0.2793,
+      "step": 21910
+    },
+    {
+      "epoch": 4.596351436359824,
+      "grad_norm": 7.2261152267456055,
+      "learning_rate": 0.0002,
+      "loss": 0.3184,
+      "step": 21920
+    },
+    {
+      "epoch": 4.598448312015098,
+      "grad_norm": 16.766836166381836,
+      "learning_rate": 0.0002,
+      "loss": 0.3586,
+      "step": 21930
+    },
+    {
+      "epoch": 4.600545187670371,
+      "grad_norm": 1.0573318004608154,
+      "learning_rate": 0.0002,
+      "loss": 0.2479,
+      "step": 21940
+    },
+    {
+      "epoch": 4.602642063325645,
+      "grad_norm": 19.868959426879883,
+      "learning_rate": 0.0002,
+      "loss": 0.392,
+      "step": 21950
+    },
+    {
+      "epoch": 4.604738938980918,
+      "grad_norm": 19.53751564025879,
+      "learning_rate": 0.0002,
+      "loss": 0.5002,
+      "step": 21960
+    },
+    {
+      "epoch": 4.606835814636192,
+      "grad_norm": 1.9004204273223877,
+      "learning_rate": 0.0002,
+      "loss": 0.376,
+      "step": 21970
+    },
+    {
+      "epoch": 4.608932690291466,
+      "grad_norm": 13.868071556091309,
+      "learning_rate": 0.0002,
+      "loss": 0.2562,
+      "step": 21980
+    },
+    {
+      "epoch": 4.6110295659467395,
+      "grad_norm": 3.889951467514038,
+      "learning_rate": 0.0002,
+      "loss": 0.3608,
+      "step": 21990
+    },
+    {
+      "epoch": 4.613126441602013,
+      "grad_norm": 6.130751609802246,
+      "learning_rate": 0.0002,
+      "loss": 0.3432,
+      "step": 22000
+    },
+    {
+      "epoch": 4.6152233172572865,
+      "grad_norm": 2.158506155014038,
+      "learning_rate": 0.0002,
+      "loss": 0.2507,
+      "step": 22010
+    },
+    {
+      "epoch": 4.61732019291256,
+      "grad_norm": 5.004857540130615,
+      "learning_rate": 0.0002,
+      "loss": 0.3078,
+      "step": 22020
+    },
+    {
+      "epoch": 4.619417068567834,
+      "grad_norm": 2.55324649810791,
+      "learning_rate": 0.0002,
+      "loss": 0.4217,
+      "step": 22030
+    },
+    {
+      "epoch": 4.621513944223108,
+      "grad_norm": 2.718355655670166,
+      "learning_rate": 0.0002,
+      "loss": 0.2676,
+      "step": 22040
+    },
+    {
+      "epoch": 4.623610819878381,
+      "grad_norm": 11.014877319335938,
+      "learning_rate": 0.0002,
+      "loss": 0.2641,
+      "step": 22050
+    },
+    {
+      "epoch": 4.625707695533655,
+      "grad_norm": 7.405698299407959,
+      "learning_rate": 0.0002,
+      "loss": 0.2586,
+      "step": 22060
+    },
+    {
+      "epoch": 4.627804571188928,
+      "grad_norm": 21.379192352294922,
+      "learning_rate": 0.0002,
+      "loss": 0.4185,
+      "step": 22070
+    },
+    {
+      "epoch": 4.629901446844202,
+      "grad_norm": 8.769160270690918,
+      "learning_rate": 0.0002,
+      "loss": 0.2551,
+      "step": 22080
+    },
+    {
+      "epoch": 4.631998322499475,
+      "grad_norm": 7.6808600425720215,
+      "learning_rate": 0.0002,
+      "loss": 0.2705,
+      "step": 22090
+    },
+    {
+      "epoch": 4.63409519815475,
+      "grad_norm": 5.105208873748779,
+      "learning_rate": 0.0002,
+      "loss": 0.2525,
+      "step": 22100
+    },
+    {
+      "epoch": 4.636192073810023,
+      "grad_norm": 12.7372407913208,
+      "learning_rate": 0.0002,
+      "loss": 0.3947,
+      "step": 22110
+    },
+    {
+      "epoch": 4.638288949465297,
+      "grad_norm": 11.61864185333252,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 22120
+    },
+    {
+      "epoch": 4.64038582512057,
+      "grad_norm": 16.5443172454834,
+      "learning_rate": 0.0002,
+      "loss": 0.5547,
+      "step": 22130
+    },
+    {
+      "epoch": 4.642482700775844,
+      "grad_norm": 18.34696388244629,
+      "learning_rate": 0.0002,
+      "loss": 0.4231,
+      "step": 22140
+    },
+    {
+      "epoch": 4.644579576431117,
+      "grad_norm": 25.033620834350586,
+      "learning_rate": 0.0002,
+      "loss": 0.2918,
+      "step": 22150
+    },
+    {
+      "epoch": 4.646676452086391,
+      "grad_norm": 18.237457275390625,
+      "learning_rate": 0.0002,
+      "loss": 0.3403,
+      "step": 22160
+    },
+    {
+      "epoch": 4.648773327741665,
+      "grad_norm": 4.656020641326904,
+      "learning_rate": 0.0002,
+      "loss": 0.446,
+      "step": 22170
+    },
+    {
+      "epoch": 4.650870203396939,
+      "grad_norm": 27.86395263671875,
+      "learning_rate": 0.0002,
+      "loss": 0.4859,
+      "step": 22180
+    },
+    {
+      "epoch": 4.652967079052212,
+      "grad_norm": 2.5194180011749268,
+      "learning_rate": 0.0002,
+      "loss": 0.2589,
+      "step": 22190
+    },
+    {
+      "epoch": 4.655063954707486,
+      "grad_norm": 64.11674499511719,
+      "learning_rate": 0.0002,
+      "loss": 0.2603,
+      "step": 22200
+    },
+    {
+      "epoch": 4.657160830362759,
+      "grad_norm": 12.19118881225586,
+      "learning_rate": 0.0002,
+      "loss": 0.2726,
+      "step": 22210
+    },
+    {
+      "epoch": 4.659257706018034,
+      "grad_norm": 5.304041862487793,
+      "learning_rate": 0.0002,
+      "loss": 0.3687,
+      "step": 22220
+    },
+    {
+      "epoch": 4.661354581673307,
+      "grad_norm": 19.514848709106445,
+      "learning_rate": 0.0002,
+      "loss": 0.2648,
+      "step": 22230
+    },
+    {
+      "epoch": 4.663451457328581,
+      "grad_norm": 50.096492767333984,
+      "learning_rate": 0.0002,
+      "loss": 0.2644,
+      "step": 22240
+    },
+    {
+      "epoch": 4.665548332983854,
+      "grad_norm": 1.945943832397461,
+      "learning_rate": 0.0002,
+      "loss": 0.4599,
+      "step": 22250
+    },
+    {
+      "epoch": 4.667645208639128,
+      "grad_norm": 15.48399543762207,
+      "learning_rate": 0.0002,
+      "loss": 0.3748,
+      "step": 22260
+    },
+    {
+      "epoch": 4.669742084294401,
+      "grad_norm": 23.655168533325195,
+      "learning_rate": 0.0002,
+      "loss": 0.3264,
+      "step": 22270
+    },
+    {
+      "epoch": 4.671838959949675,
+      "grad_norm": 5.6444478034973145,
+      "learning_rate": 0.0002,
+      "loss": 0.3308,
+      "step": 22280
+    },
+    {
+      "epoch": 4.673935835604949,
+      "grad_norm": 1.5966367721557617,
+      "learning_rate": 0.0002,
+      "loss": 0.3293,
+      "step": 22290
+    },
+    {
+      "epoch": 4.6760327112602225,
+      "grad_norm": 37.680545806884766,
+      "learning_rate": 0.0002,
+      "loss": 0.4031,
+      "step": 22300
+    },
+    {
+      "epoch": 4.678129586915496,
+      "grad_norm": 3.1248462200164795,
+      "learning_rate": 0.0002,
+      "loss": 0.4096,
+      "step": 22310
+    },
+    {
+      "epoch": 4.6802264625707695,
+      "grad_norm": 2.226181745529175,
+      "learning_rate": 0.0002,
+      "loss": 0.3701,
+      "step": 22320
+    },
+    {
+      "epoch": 4.682323338226043,
+      "grad_norm": 4.016665458679199,
+      "learning_rate": 0.0002,
+      "loss": 0.401,
+      "step": 22330
+    },
+    {
+      "epoch": 4.6844202138813165,
+      "grad_norm": 17.831260681152344,
+      "learning_rate": 0.0002,
+      "loss": 0.3257,
+      "step": 22340
+    },
+    {
+      "epoch": 4.686517089536591,
+      "grad_norm": 17.144725799560547,
+      "learning_rate": 0.0002,
+      "loss": 0.3335,
+      "step": 22350
+    },
+    {
+      "epoch": 4.688613965191864,
+      "grad_norm": 7.547942638397217,
+      "learning_rate": 0.0002,
+      "loss": 0.4015,
+      "step": 22360
+    },
+    {
+      "epoch": 4.690710840847138,
+      "grad_norm": 6.964625358581543,
+      "learning_rate": 0.0002,
+      "loss": 0.3709,
+      "step": 22370
+    },
+    {
+      "epoch": 4.692807716502411,
+      "grad_norm": 5.216811656951904,
+      "learning_rate": 0.0002,
+      "loss": 0.4521,
+      "step": 22380
+    },
+    {
+      "epoch": 4.694904592157685,
+      "grad_norm": 47.28529739379883,
+      "learning_rate": 0.0002,
+      "loss": 0.3236,
+      "step": 22390
+    },
+    {
+      "epoch": 4.697001467812958,
+      "grad_norm": 1.253324031829834,
+      "learning_rate": 0.0002,
+      "loss": 0.3612,
+      "step": 22400
+    },
+    {
+      "epoch": 4.699098343468233,
+      "grad_norm": 7.6197733879089355,
+      "learning_rate": 0.0002,
+      "loss": 0.2537,
+      "step": 22410
+    },
+    {
+      "epoch": 4.701195219123506,
+      "grad_norm": 72.7631607055664,
+      "learning_rate": 0.0002,
+      "loss": 0.9054,
+      "step": 22420
+    },
+    {
+      "epoch": 4.70329209477878,
+      "grad_norm": 35.460487365722656,
+      "learning_rate": 0.0002,
+      "loss": 0.3916,
+      "step": 22430
+    },
+    {
+      "epoch": 4.705388970434053,
+      "grad_norm": 31.78118324279785,
+      "learning_rate": 0.0002,
+      "loss": 0.4392,
+      "step": 22440
+    },
+    {
+      "epoch": 4.707485846089327,
+      "grad_norm": 59.87028121948242,
+      "learning_rate": 0.0002,
+      "loss": 0.4103,
+      "step": 22450
+    },
+    {
+      "epoch": 4.7095827217446,
+      "grad_norm": 1.4908517599105835,
+      "learning_rate": 0.0002,
+      "loss": 0.2769,
+      "step": 22460
+    },
+    {
+      "epoch": 4.711679597399874,
+      "grad_norm": 3.2404863834381104,
+      "learning_rate": 0.0002,
+      "loss": 0.3169,
+      "step": 22470
+    },
+    {
+      "epoch": 4.713776473055148,
+      "grad_norm": 5.952626705169678,
+      "learning_rate": 0.0002,
+      "loss": 0.4722,
+      "step": 22480
+    },
+    {
+      "epoch": 4.715873348710422,
+      "grad_norm": 9.007226943969727,
+      "learning_rate": 0.0002,
+      "loss": 0.4048,
+      "step": 22490
+    },
+    {
+      "epoch": 4.717970224365695,
+      "grad_norm": 4.316738128662109,
+      "learning_rate": 0.0002,
+      "loss": 0.3234,
+      "step": 22500
+    },
+    {
+      "epoch": 4.720067100020969,
+      "grad_norm": 7.632235050201416,
+      "learning_rate": 0.0002,
+      "loss": 0.3282,
+      "step": 22510
+    },
+    {
+      "epoch": 4.722163975676242,
+      "grad_norm": 4.759393215179443,
+      "learning_rate": 0.0002,
+      "loss": 0.3523,
+      "step": 22520
+    },
+    {
+      "epoch": 4.7242608513315165,
+      "grad_norm": 2.0390775203704834,
+      "learning_rate": 0.0002,
+      "loss": 0.3229,
+      "step": 22530
+    },
+    {
+      "epoch": 4.72635772698679,
+      "grad_norm": 12.438055992126465,
+      "learning_rate": 0.0002,
+      "loss": 0.272,
+      "step": 22540
+    },
+    {
+      "epoch": 4.7284546026420635,
+      "grad_norm": 13.05179500579834,
+      "learning_rate": 0.0002,
+      "loss": 0.3151,
+      "step": 22550
+    },
+    {
+      "epoch": 4.730551478297337,
+      "grad_norm": 4.340366840362549,
+      "learning_rate": 0.0002,
+      "loss": 0.4242,
+      "step": 22560
+    },
+    {
+      "epoch": 4.7326483539526105,
+      "grad_norm": 7.640782833099365,
+      "learning_rate": 0.0002,
+      "loss": 0.4112,
+      "step": 22570
+    },
+    {
+      "epoch": 4.734745229607884,
+      "grad_norm": 1.2433539628982544,
+      "learning_rate": 0.0002,
+      "loss": 0.2546,
+      "step": 22580
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 4.08035945892334,
+      "learning_rate": 0.0002,
+      "loss": 0.3997,
+      "step": 22590
+    },
+    {
+      "epoch": 4.738938980918432,
+      "grad_norm": 7.260676383972168,
+      "learning_rate": 0.0002,
+      "loss": 0.4803,
+      "step": 22600
+    },
+    {
+      "epoch": 4.741035856573705,
+      "grad_norm": 5.27596378326416,
+      "learning_rate": 0.0002,
+      "loss": 0.3236,
+      "step": 22610
+    },
+    {
+      "epoch": 4.743132732228979,
+      "grad_norm": 8.552403450012207,
+      "learning_rate": 0.0002,
+      "loss": 0.3111,
+      "step": 22620
+    },
+    {
+      "epoch": 4.745229607884252,
+      "grad_norm": 41.805084228515625,
+      "learning_rate": 0.0002,
+      "loss": 0.5692,
+      "step": 22630
+    },
+    {
+      "epoch": 4.747326483539526,
+      "grad_norm": 45.03173828125,
+      "learning_rate": 0.0002,
+      "loss": 0.4797,
+      "step": 22640
+    },
+    {
+      "epoch": 4.749423359194799,
+      "grad_norm": 11.940930366516113,
+      "learning_rate": 0.0002,
+      "loss": 0.4627,
+      "step": 22650
+    },
+    {
+      "epoch": 4.751520234850073,
+      "grad_norm": 17.953216552734375,
+      "learning_rate": 0.0002,
+      "loss": 0.5131,
+      "step": 22660
+    },
+    {
+      "epoch": 4.753617110505347,
+      "grad_norm": 9.348217010498047,
+      "learning_rate": 0.0002,
+      "loss": 0.3437,
+      "step": 22670
+    },
+    {
+      "epoch": 4.755713986160621,
+      "grad_norm": 2.931450128555298,
+      "learning_rate": 0.0002,
+      "loss": 0.2577,
+      "step": 22680
+    },
+    {
+      "epoch": 4.757810861815894,
+      "grad_norm": 3.602579355239868,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 22690
+    },
+    {
+      "epoch": 4.759907737471168,
+      "grad_norm": 41.81117630004883,
+      "learning_rate": 0.0002,
+      "loss": 0.5134,
+      "step": 22700
+    },
+    {
+      "epoch": 4.762004613126441,
+      "grad_norm": 5.8748650550842285,
+      "learning_rate": 0.0002,
+      "loss": 0.2563,
+      "step": 22710
+    },
+    {
+      "epoch": 4.764101488781716,
+      "grad_norm": 5.086737155914307,
+      "learning_rate": 0.0002,
+      "loss": 0.4383,
+      "step": 22720
+    },
+    {
+      "epoch": 4.766198364436989,
+      "grad_norm": 53.62387466430664,
+      "learning_rate": 0.0002,
+      "loss": 0.4937,
+      "step": 22730
+    },
+    {
+      "epoch": 4.768295240092263,
+      "grad_norm": 12.729607582092285,
+      "learning_rate": 0.0002,
+      "loss": 0.2615,
+      "step": 22740
+    },
+    {
+      "epoch": 4.770392115747536,
+      "grad_norm": 89.39778137207031,
+      "learning_rate": 0.0002,
+      "loss": 0.3657,
+      "step": 22750
+    },
+    {
+      "epoch": 4.77248899140281,
+      "grad_norm": 32.737648010253906,
+      "learning_rate": 0.0002,
+      "loss": 0.337,
+      "step": 22760
+    },
+    {
+      "epoch": 4.774585867058083,
+      "grad_norm": 10.218364715576172,
+      "learning_rate": 0.0002,
+      "loss": 0.2745,
+      "step": 22770
+    },
+    {
+      "epoch": 4.776682742713357,
+      "grad_norm": 9.050323486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.4922,
+      "step": 22780
+    },
+    {
+      "epoch": 4.778779618368631,
+      "grad_norm": 5.656611442565918,
+      "learning_rate": 0.0002,
+      "loss": 0.2631,
+      "step": 22790
+    },
+    {
+      "epoch": 4.780876494023905,
+      "grad_norm": 52.178993225097656,
+      "learning_rate": 0.0002,
+      "loss": 0.2844,
+      "step": 22800
+    },
+    {
+      "epoch": 4.782973369679178,
+      "grad_norm": 47.777408599853516,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 22810
+    },
+    {
+      "epoch": 4.785070245334452,
+      "grad_norm": 4.114084243774414,
+      "learning_rate": 0.0002,
+      "loss": 0.2857,
+      "step": 22820
+    },
+    {
+      "epoch": 4.787167120989725,
+      "grad_norm": 12.685486793518066,
+      "learning_rate": 0.0002,
+      "loss": 0.2551,
+      "step": 22830
+    },
+    {
+      "epoch": 4.789263996644999,
+      "grad_norm": 3.0896663665771484,
+      "learning_rate": 0.0002,
+      "loss": 0.3513,
+      "step": 22840
+    },
+    {
+      "epoch": 4.791360872300273,
+      "grad_norm": 8.13403606414795,
+      "learning_rate": 0.0002,
+      "loss": 0.3306,
+      "step": 22850
+    },
+    {
+      "epoch": 4.7934577479555465,
+      "grad_norm": 138.98486328125,
+      "learning_rate": 0.0002,
+      "loss": 0.3493,
+      "step": 22860
+    },
+    {
+      "epoch": 4.79555462361082,
+      "grad_norm": 3.541754722595215,
+      "learning_rate": 0.0002,
+      "loss": 0.262,
+      "step": 22870
+    },
+    {
+      "epoch": 4.7976514992660935,
+      "grad_norm": 21.595823287963867,
+      "learning_rate": 0.0002,
+      "loss": 0.259,
+      "step": 22880
+    },
+    {
+      "epoch": 4.799748374921367,
+      "grad_norm": 3.3957481384277344,
+      "learning_rate": 0.0002,
+      "loss": 0.2567,
+      "step": 22890
+    },
+    {
+      "epoch": 4.8018452505766405,
+      "grad_norm": 15.020285606384277,
+      "learning_rate": 0.0002,
+      "loss": 0.3227,
+      "step": 22900
+    },
+    {
+      "epoch": 4.803942126231915,
+      "grad_norm": 17.61827278137207,
+      "learning_rate": 0.0002,
+      "loss": 0.4037,
+      "step": 22910
+    },
+    {
+      "epoch": 4.806039001887188,
+      "grad_norm": 2.315112352371216,
+      "learning_rate": 0.0002,
+      "loss": 0.3231,
+      "step": 22920
+    },
+    {
+      "epoch": 4.808135877542462,
+      "grad_norm": 4.932908058166504,
+      "learning_rate": 0.0002,
+      "loss": 0.3858,
+      "step": 22930
+    },
+    {
+      "epoch": 4.810232753197735,
+      "grad_norm": 4.863265514373779,
+      "learning_rate": 0.0002,
+      "loss": 0.3371,
+      "step": 22940
+    },
+    {
+      "epoch": 4.812329628853009,
+      "grad_norm": 53.41450881958008,
+      "learning_rate": 0.0002,
+      "loss": 0.3507,
+      "step": 22950
+    },
+    {
+      "epoch": 4.814426504508282,
+      "grad_norm": 47.964725494384766,
+      "learning_rate": 0.0002,
+      "loss": 0.5097,
+      "step": 22960
+    },
+    {
+      "epoch": 4.816523380163556,
+      "grad_norm": 37.04306411743164,
+      "learning_rate": 0.0002,
+      "loss": 0.4286,
+      "step": 22970
+    },
+    {
+      "epoch": 4.81862025581883,
+      "grad_norm": 1.4727418422698975,
+      "learning_rate": 0.0002,
+      "loss": 0.312,
+      "step": 22980
+    },
+    {
+      "epoch": 4.820717131474104,
+      "grad_norm": 54.622379302978516,
+      "learning_rate": 0.0002,
+      "loss": 0.3793,
+      "step": 22990
+    },
+    {
+      "epoch": 4.822814007129377,
+      "grad_norm": 68.0690689086914,
+      "learning_rate": 0.0002,
+      "loss": 0.5235,
+      "step": 23000
+    },
+    {
+      "epoch": 4.824910882784651,
+      "grad_norm": 55.60499954223633,
+      "learning_rate": 0.0002,
+      "loss": 0.3695,
+      "step": 23010
+    },
+    {
+      "epoch": 4.827007758439924,
+      "grad_norm": 2.116057872772217,
+      "learning_rate": 0.0002,
+      "loss": 0.2591,
+      "step": 23020
+    },
+    {
+      "epoch": 4.829104634095199,
+      "grad_norm": 27.25665283203125,
+      "learning_rate": 0.0002,
+      "loss": 0.258,
+      "step": 23030
+    },
+    {
+      "epoch": 4.831201509750472,
+      "grad_norm": 1.7732720375061035,
+      "learning_rate": 0.0002,
+      "loss": 0.3434,
+      "step": 23040
+    },
+    {
+      "epoch": 4.833298385405746,
+      "grad_norm": 1.4641029834747314,
+      "learning_rate": 0.0002,
+      "loss": 0.2559,
+      "step": 23050
+    },
+    {
+      "epoch": 4.835395261061019,
+      "grad_norm": 15.68821907043457,
+      "learning_rate": 0.0002,
+      "loss": 0.4408,
+      "step": 23060
+    },
+    {
+      "epoch": 4.837492136716293,
+      "grad_norm": 50.2274169921875,
+      "learning_rate": 0.0002,
+      "loss": 0.2794,
+      "step": 23070
+    },
+    {
+      "epoch": 4.839589012371566,
+      "grad_norm": 13.899438858032227,
+      "learning_rate": 0.0002,
+      "loss": 0.4018,
+      "step": 23080
+    },
+    {
+      "epoch": 4.84168588802684,
+      "grad_norm": 24.751676559448242,
+      "learning_rate": 0.0002,
+      "loss": 0.3897,
+      "step": 23090
+    },
+    {
+      "epoch": 4.843782763682114,
+      "grad_norm": 70.79833221435547,
+      "learning_rate": 0.0002,
+      "loss": 0.7469,
+      "step": 23100
+    },
+    {
+      "epoch": 4.8458796393373875,
+      "grad_norm": 2.478081464767456,
+      "learning_rate": 0.0002,
+      "loss": 0.2668,
+      "step": 23110
+    },
+    {
+      "epoch": 4.847976514992661,
+      "grad_norm": 3.0055885314941406,
+      "learning_rate": 0.0002,
+      "loss": 0.2569,
+      "step": 23120
+    },
+    {
+      "epoch": 4.8500733906479345,
+      "grad_norm": 105.44467163085938,
+      "learning_rate": 0.0002,
+      "loss": 0.299,
+      "step": 23130
+    },
+    {
+      "epoch": 4.852170266303208,
+      "grad_norm": 172.33485412597656,
+      "learning_rate": 0.0002,
+      "loss": 0.4352,
+      "step": 23140
+    },
+    {
+      "epoch": 4.8542671419584815,
+      "grad_norm": 2.8913474082946777,
+      "learning_rate": 0.0002,
+      "loss": 0.2581,
+      "step": 23150
+    },
+    {
+      "epoch": 4.856364017613755,
+      "grad_norm": 63.7279052734375,
+      "learning_rate": 0.0002,
+      "loss": 0.328,
+      "step": 23160
+    },
+    {
+      "epoch": 4.858460893269029,
+      "grad_norm": 16.1322021484375,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 23170
+    },
+    {
+      "epoch": 4.860557768924303,
+      "grad_norm": 4.814330101013184,
+      "learning_rate": 0.0002,
+      "loss": 0.4127,
+      "step": 23180
+    },
+    {
+      "epoch": 4.862654644579576,
+      "grad_norm": 6.94169807434082,
+      "learning_rate": 0.0002,
+      "loss": 0.3144,
+      "step": 23190
+    },
+    {
+      "epoch": 4.86475152023485,
+      "grad_norm": 43.06448745727539,
+      "learning_rate": 0.0002,
+      "loss": 0.3599,
+      "step": 23200
+    },
+    {
+      "epoch": 4.866848395890123,
+      "grad_norm": 100.02835845947266,
+      "learning_rate": 0.0002,
+      "loss": 0.3969,
+      "step": 23210
+    },
+    {
+      "epoch": 4.868945271545398,
+      "grad_norm": 173.16786193847656,
+      "learning_rate": 0.0002,
+      "loss": 0.4545,
+      "step": 23220
+    },
+    {
+      "epoch": 4.871042147200671,
+      "grad_norm": 81.39131927490234,
+      "learning_rate": 0.0002,
+      "loss": 0.7088,
+      "step": 23230
+    },
+    {
+      "epoch": 4.873139022855945,
+      "grad_norm": 114.65706634521484,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 23240
+    },
+    {
+      "epoch": 4.875235898511218,
+      "grad_norm": 79.38748931884766,
+      "learning_rate": 0.0002,
+      "loss": 0.3835,
+      "step": 23250
+    },
+    {
+      "epoch": 4.877332774166492,
+      "grad_norm": 21.65077781677246,
+      "learning_rate": 0.0002,
+      "loss": 0.3005,
+      "step": 23260
+    },
+    {
+      "epoch": 4.879429649821765,
+      "grad_norm": 20.43645477294922,
+      "learning_rate": 0.0002,
+      "loss": 0.3529,
+      "step": 23270
+    },
+    {
+      "epoch": 4.881526525477039,
+      "grad_norm": 13.03095531463623,
+      "learning_rate": 0.0002,
+      "loss": 0.4077,
+      "step": 23280
+    },
+    {
+      "epoch": 4.883623401132313,
+      "grad_norm": 0.7234331965446472,
+      "learning_rate": 0.0002,
+      "loss": 0.4606,
+      "step": 23290
+    },
+    {
+      "epoch": 4.885720276787587,
+      "grad_norm": 3.002551555633545,
+      "learning_rate": 0.0002,
+      "loss": 0.2611,
+      "step": 23300
+    },
+    {
+      "epoch": 4.88781715244286,
+      "grad_norm": 48.173095703125,
+      "learning_rate": 0.0002,
+      "loss": 0.4378,
+      "step": 23310
+    },
+    {
+      "epoch": 4.889914028098134,
+      "grad_norm": 30.906227111816406,
+      "learning_rate": 0.0002,
+      "loss": 0.4116,
+      "step": 23320
+    },
+    {
+      "epoch": 4.892010903753407,
+      "grad_norm": 52.0058479309082,
+      "learning_rate": 0.0002,
+      "loss": 0.4907,
+      "step": 23330
+    },
+    {
+      "epoch": 4.894107779408681,
+      "grad_norm": 3.250602960586548,
+      "learning_rate": 0.0002,
+      "loss": 0.4432,
+      "step": 23340
+    },
+    {
+      "epoch": 4.896204655063955,
+      "grad_norm": 7.294270992279053,
+      "learning_rate": 0.0002,
+      "loss": 0.268,
+      "step": 23350
+    },
+    {
+      "epoch": 4.898301530719229,
+      "grad_norm": 16.525890350341797,
+      "learning_rate": 0.0002,
+      "loss": 0.3601,
+      "step": 23360
+    },
+    {
+      "epoch": 4.900398406374502,
+      "grad_norm": 29.520700454711914,
+      "learning_rate": 0.0002,
+      "loss": 0.276,
+      "step": 23370
+    },
+    {
+      "epoch": 4.902495282029776,
+      "grad_norm": 13.231327056884766,
+      "learning_rate": 0.0002,
+      "loss": 0.2752,
+      "step": 23380
+    },
+    {
+      "epoch": 4.904592157685049,
+      "grad_norm": 8.970483779907227,
+      "learning_rate": 0.0002,
+      "loss": 0.3489,
+      "step": 23390
+    },
+    {
+      "epoch": 4.906689033340323,
+      "grad_norm": 14.564881324768066,
+      "learning_rate": 0.0002,
+      "loss": 0.3565,
+      "step": 23400
+    },
+    {
+      "epoch": 4.908785908995597,
+      "grad_norm": 47.08441925048828,
+      "learning_rate": 0.0002,
+      "loss": 0.5337,
+      "step": 23410
+    },
+    {
+      "epoch": 4.9108827846508705,
+      "grad_norm": 10.598335266113281,
+      "learning_rate": 0.0002,
+      "loss": 0.3706,
+      "step": 23420
+    },
+    {
+      "epoch": 4.912979660306144,
+      "grad_norm": 7.135160446166992,
+      "learning_rate": 0.0002,
+      "loss": 0.3633,
+      "step": 23430
+    },
+    {
+      "epoch": 4.9150765359614175,
+      "grad_norm": 4.119414806365967,
+      "learning_rate": 0.0002,
+      "loss": 0.3192,
+      "step": 23440
+    },
+    {
+      "epoch": 4.917173411616691,
+      "grad_norm": 61.618770599365234,
+      "learning_rate": 0.0002,
+      "loss": 0.2827,
+      "step": 23450
+    },
+    {
+      "epoch": 4.9192702872719645,
+      "grad_norm": 2.349839687347412,
+      "learning_rate": 0.0002,
+      "loss": 0.4235,
+      "step": 23460
+    },
+    {
+      "epoch": 4.921367162927238,
+      "grad_norm": 4.726789951324463,
+      "learning_rate": 0.0002,
+      "loss": 0.392,
+      "step": 23470
+    },
+    {
+      "epoch": 4.923464038582512,
+      "grad_norm": 100.8683853149414,
+      "learning_rate": 0.0002,
+      "loss": 0.2534,
+      "step": 23480
+    },
+    {
+      "epoch": 4.925560914237786,
+      "grad_norm": 5.879467964172363,
+      "learning_rate": 0.0002,
+      "loss": 0.3332,
+      "step": 23490
+    },
+    {
+      "epoch": 4.927657789893059,
+      "grad_norm": 11.780795097351074,
+      "learning_rate": 0.0002,
+      "loss": 0.2696,
+      "step": 23500
+    },
+    {
+      "epoch": 4.929754665548333,
+      "grad_norm": 9.69931697845459,
+      "learning_rate": 0.0002,
+      "loss": 0.3792,
+      "step": 23510
+    },
+    {
+      "epoch": 4.931851541203606,
+      "grad_norm": 109.73184204101562,
+      "learning_rate": 0.0002,
+      "loss": 6.2043,
+      "step": 23520
+    },
+    {
+      "epoch": 4.933948416858881,
+      "grad_norm": 46.461143493652344,
+      "learning_rate": 0.0002,
+      "loss": 6.4032,
+      "step": 23530
+    },
+    {
+      "epoch": 4.936045292514154,
+      "grad_norm": 108.99391174316406,
+      "learning_rate": 0.0002,
+      "loss": 2.7763,
+      "step": 23540
+    },
+    {
+      "epoch": 4.938142168169428,
+      "grad_norm": 285.75054931640625,
+      "learning_rate": 0.0002,
+      "loss": 1.7655,
+      "step": 23550
+    },
+    {
+      "epoch": 4.940239043824701,
+      "grad_norm": 128.00816345214844,
+      "learning_rate": 0.0002,
+      "loss": 0.9253,
+      "step": 23560
+    },
+    {
+      "epoch": 4.942335919479975,
+      "grad_norm": 126.41734313964844,
+      "learning_rate": 0.0002,
+      "loss": 0.7302,
+      "step": 23570
+    },
+    {
+      "epoch": 4.944432795135248,
+      "grad_norm": 62.741641998291016,
+      "learning_rate": 0.0002,
+      "loss": 0.5692,
+      "step": 23580
+    },
+    {
+      "epoch": 4.946529670790522,
+      "grad_norm": 33.307369232177734,
+      "learning_rate": 0.0002,
+      "loss": 0.4376,
+      "step": 23590
+    },
+    {
+      "epoch": 4.948626546445796,
+      "grad_norm": 26.6583251953125,
+      "learning_rate": 0.0002,
+      "loss": 0.4938,
+      "step": 23600
+    },
+    {
+      "epoch": 4.95072342210107,
+      "grad_norm": 14.1249361038208,
+      "learning_rate": 0.0002,
+      "loss": 0.4565,
+      "step": 23610
+    },
+    {
+      "epoch": 4.952820297756343,
+      "grad_norm": 45.58442687988281,
+      "learning_rate": 0.0002,
+      "loss": 0.5171,
+      "step": 23620
+    },
+    {
+      "epoch": 4.954917173411617,
+      "grad_norm": 26.00278091430664,
+      "learning_rate": 0.0002,
+      "loss": 0.5916,
+      "step": 23630
+    },
+    {
+      "epoch": 4.95701404906689,
+      "grad_norm": 12.711030960083008,
+      "learning_rate": 0.0002,
+      "loss": 0.4861,
+      "step": 23640
+    },
+    {
+      "epoch": 4.959110924722164,
+      "grad_norm": 46.92948532104492,
+      "learning_rate": 0.0002,
+      "loss": 0.3547,
+      "step": 23650
+    },
+    {
+      "epoch": 4.961207800377437,
+      "grad_norm": 73.09407043457031,
+      "learning_rate": 0.0002,
+      "loss": 0.3132,
+      "step": 23660
+    },
+    {
+      "epoch": 4.9633046760327115,
+      "grad_norm": 64.90545654296875,
+      "learning_rate": 0.0002,
+      "loss": 0.9628,
+      "step": 23670
+    },
+    {
+      "epoch": 4.965401551687985,
+      "grad_norm": 17.83898162841797,
+      "learning_rate": 0.0002,
+      "loss": 0.3811,
+      "step": 23680
+    },
+    {
+      "epoch": 4.9674984273432585,
+      "grad_norm": 55.39151382446289,
+      "learning_rate": 0.0002,
+      "loss": 0.2989,
+      "step": 23690
+    },
+    {
+      "epoch": 4.969595302998532,
+      "grad_norm": 24.683879852294922,
+      "learning_rate": 0.0002,
+      "loss": 0.7712,
+      "step": 23700
+    },
+    {
+      "epoch": 4.9716921786538055,
+      "grad_norm": 106.67837524414062,
+      "learning_rate": 0.0002,
+      "loss": 0.3724,
+      "step": 23710
+    },
+    {
+      "epoch": 4.97378905430908,
+      "grad_norm": 10.091436386108398,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 23720
+    },
+    {
+      "epoch": 4.975885929964353,
+      "grad_norm": 55.912933349609375,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 23730
+    },
+    {
+      "epoch": 4.977982805619627,
+      "grad_norm": 303.874755859375,
+      "learning_rate": 0.0002,
+      "loss": 0.9055,
+      "step": 23740
+    },
+    {
+      "epoch": 4.9800796812749,
+      "grad_norm": 53.40581512451172,
+      "learning_rate": 0.0002,
+      "loss": 0.8064,
+      "step": 23750
+    },
+    {
+      "epoch": 4.982176556930174,
+      "grad_norm": 65.19920349121094,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 23760
+    },
+    {
+      "epoch": 4.984273432585447,
+      "grad_norm": 183.5748291015625,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 23770
+    },
+    {
+      "epoch": 4.986370308240721,
+      "grad_norm": 17.69962501525879,
+      "learning_rate": 0.0002,
+      "loss": 0.4914,
+      "step": 23780
+    },
+    {
+      "epoch": 4.988467183895995,
+      "grad_norm": 20.91483497619629,
+      "learning_rate": 0.0002,
+      "loss": 0.4115,
+      "step": 23790
+    },
+    {
+      "epoch": 4.990564059551269,
+      "grad_norm": 38.18173599243164,
+      "learning_rate": 0.0002,
+      "loss": 0.4986,
+      "step": 23800
+    },
+    {
+      "epoch": 4.992660935206542,
+      "grad_norm": 59.649051666259766,
+      "learning_rate": 0.0002,
+      "loss": 0.7222,
+      "step": 23810
+    },
+    {
+      "epoch": 4.994757810861816,
+      "grad_norm": 8.023183822631836,
+      "learning_rate": 0.0002,
+      "loss": 0.3814,
+      "step": 23820
+    },
+    {
+      "epoch": 4.996854686517089,
+      "grad_norm": 5.922271728515625,
+      "learning_rate": 0.0002,
+      "loss": 0.4391,
+      "step": 23830
+    },
+    {
+      "epoch": 4.998951562172363,
+      "grad_norm": 52.8021240234375,
+      "learning_rate": 0.0002,
+      "loss": 0.461,
+      "step": 23840
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 23845,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.8878877907124224e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}