diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,63034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.1264584434721976,
+  "eval_steps": 500,
+  "global_step": 90000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.4050938163577509e-05,
+      "grad_norm": 2.420112133026123,
+      "learning_rate": 3.793733314598848e-08,
+      "loss": 10.5041,
+      "step": 10
+    },
+    {
+      "epoch": 2.8101876327155018e-05,
+      "grad_norm": 2.226503849029541,
+      "learning_rate": 8.008992553042011e-08,
+      "loss": 10.4971,
+      "step": 20
+    },
+    {
+      "epoch": 4.215281449073253e-05,
+      "grad_norm": 2.459202527999878,
+      "learning_rate": 1.2224251791485176e-07,
+      "loss": 10.5017,
+      "step": 30
+    },
+    {
+      "epoch": 5.6203752654310035e-05,
+      "grad_norm": 2.5928807258605957,
+      "learning_rate": 1.6439511029928338e-07,
+      "loss": 10.5044,
+      "step": 40
+    },
+    {
+      "epoch": 7.025469081788755e-05,
+      "grad_norm": 2.4456496238708496,
+      "learning_rate": 2.0654770268371502e-07,
+      "loss": 10.4912,
+      "step": 50
+    },
+    {
+      "epoch": 8.430562898146506e-05,
+      "grad_norm": 3.4608006477355957,
+      "learning_rate": 2.4870029506814667e-07,
+      "loss": 10.5029,
+      "step": 60
+    },
+    {
+      "epoch": 9.835656714504256e-05,
+      "grad_norm": 2.387044668197632,
+      "learning_rate": 2.908528874525783e-07,
+      "loss": 10.4957,
+      "step": 70
+    },
+    {
+      "epoch": 0.00011240750530862007,
+      "grad_norm": 2.4745988845825195,
+      "learning_rate": 3.3300547983700995e-07,
+      "loss": 10.503,
+      "step": 80
+    },
+    {
+      "epoch": 0.00012645844347219758,
+      "grad_norm": 3.1475226879119873,
+      "learning_rate": 3.751580722214416e-07,
+      "loss": 10.4968,
+      "step": 90
+    },
+    {
+      "epoch": 0.0001405093816357751,
+      "grad_norm": 2.5408034324645996,
+      "learning_rate": 4.173106646058732e-07,
+      "loss": 10.4923,
+      "step": 100
+    },
+    {
+      "epoch": 0.0001545603197993526,
+      "grad_norm": 2.2912540435791016,
+      "learning_rate": 4.594632569903049e-07,
+      "loss": 10.4942,
+      "step": 110
+    },
+    {
+      "epoch": 0.0001686112579629301,
+      "grad_norm": 2.3079867362976074,
+      "learning_rate": 5.016158493747365e-07,
+      "loss": 10.4897,
+      "step": 120
+    },
+    {
+      "epoch": 0.00018266219612650763,
+      "grad_norm": 2.3974714279174805,
+      "learning_rate": 5.437684417591682e-07,
+      "loss": 10.4806,
+      "step": 130
+    },
+    {
+      "epoch": 0.00019671313429008513,
+      "grad_norm": 2.180088758468628,
+      "learning_rate": 5.859210341435998e-07,
+      "loss": 10.4768,
+      "step": 140
+    },
+    {
+      "epoch": 0.00021076407245366265,
+      "grad_norm": 2.415879249572754,
+      "learning_rate": 6.280736265280315e-07,
+      "loss": 10.4835,
+      "step": 150
+    },
+    {
+      "epoch": 0.00022481501061724014,
+      "grad_norm": 2.506563425064087,
+      "learning_rate": 6.70226218912463e-07,
+      "loss": 10.4767,
+      "step": 160
+    },
+    {
+      "epoch": 0.00023886594878081766,
+      "grad_norm": 2.7735683917999268,
+      "learning_rate": 7.123788112968946e-07,
+      "loss": 10.4738,
+      "step": 170
+    },
+    {
+      "epoch": 0.00025291688694439516,
+      "grad_norm": 2.5280656814575195,
+      "learning_rate": 7.545314036813263e-07,
+      "loss": 10.4746,
+      "step": 180
+    },
+    {
+      "epoch": 0.0002669678251079727,
+      "grad_norm": 2.373305559158325,
+      "learning_rate": 7.966839960657579e-07,
+      "loss": 10.4709,
+      "step": 190
+    },
+    {
+      "epoch": 0.0002810187632715502,
+      "grad_norm": 2.3311121463775635,
+      "learning_rate": 8.388365884501895e-07,
+      "loss": 10.4571,
+      "step": 200
+    },
+    {
+      "epoch": 0.0002950697014351277,
+      "grad_norm": 2.61584210395813,
+      "learning_rate": 8.809891808346213e-07,
+      "loss": 10.4502,
+      "step": 210
+    },
+    {
+      "epoch": 0.0003091206395987052,
+      "grad_norm": 2.391726016998291,
+      "learning_rate": 9.231417732190529e-07,
+      "loss": 10.4474,
+      "step": 220
+    },
+    {
+      "epoch": 0.0003231715777622827,
+      "grad_norm": 2.519890785217285,
+      "learning_rate": 9.652943656034844e-07,
+      "loss": 10.4378,
+      "step": 230
+    },
+    {
+      "epoch": 0.0003372225159258602,
+      "grad_norm": 2.557464599609375,
+      "learning_rate": 1.0074469579879162e-06,
+      "loss": 10.4381,
+      "step": 240
+    },
+    {
+      "epoch": 0.00035127345408943775,
+      "grad_norm": 2.5819175243377686,
+      "learning_rate": 1.0495995503723478e-06,
+      "loss": 10.4289,
+      "step": 250
+    },
+    {
+      "epoch": 0.00036532439225301527,
+      "grad_norm": 2.536975622177124,
+      "learning_rate": 1.0917521427567794e-06,
+      "loss": 10.4208,
+      "step": 260
+    },
+    {
+      "epoch": 0.00037937533041659273,
+      "grad_norm": 2.575305700302124,
+      "learning_rate": 1.1339047351412112e-06,
+      "loss": 10.4069,
+      "step": 270
+    },
+    {
+      "epoch": 0.00039342626858017025,
+      "grad_norm": 2.553637981414795,
+      "learning_rate": 1.1760573275256428e-06,
+      "loss": 10.3983,
+      "step": 280
+    },
+    {
+      "epoch": 0.0004074772067437478,
+      "grad_norm": 2.40779709815979,
+      "learning_rate": 1.2182099199100744e-06,
+      "loss": 10.3975,
+      "step": 290
+    },
+    {
+      "epoch": 0.0004215281449073253,
+      "grad_norm": 2.8702962398529053,
+      "learning_rate": 1.260362512294506e-06,
+      "loss": 10.3814,
+      "step": 300
+    },
+    {
+      "epoch": 0.0004355790830709028,
+      "grad_norm": 2.869860887527466,
+      "learning_rate": 1.3025151046789375e-06,
+      "loss": 10.3902,
+      "step": 310
+    },
+    {
+      "epoch": 0.0004496300212344803,
+      "grad_norm": 2.448800563812256,
+      "learning_rate": 1.3446676970633693e-06,
+      "loss": 10.3575,
+      "step": 320
+    },
+    {
+      "epoch": 0.0004636809593980578,
+      "grad_norm": 2.3154518604278564,
+      "learning_rate": 1.386820289447801e-06,
+      "loss": 10.3371,
+      "step": 330
+    },
+    {
+      "epoch": 0.0004777318975616353,
+      "grad_norm": 2.516812562942505,
+      "learning_rate": 1.4289728818322327e-06,
+      "loss": 10.3448,
+      "step": 340
+    },
+    {
+      "epoch": 0.0004917828357252128,
+      "grad_norm": 2.7383711338043213,
+      "learning_rate": 1.4711254742166643e-06,
+      "loss": 10.3181,
+      "step": 350
+    },
+    {
+      "epoch": 0.0005058337738887903,
+      "grad_norm": 2.346808910369873,
+      "learning_rate": 1.513278066601096e-06,
+      "loss": 10.2987,
+      "step": 360
+    },
+    {
+      "epoch": 0.0005198847120523679,
+      "grad_norm": 2.4467155933380127,
+      "learning_rate": 1.5554306589855275e-06,
+      "loss": 10.2587,
+      "step": 370
+    },
+    {
+      "epoch": 0.0005339356502159454,
+      "grad_norm": 2.229548454284668,
+      "learning_rate": 1.597583251369959e-06,
+      "loss": 10.2929,
+      "step": 380
+    },
+    {
+      "epoch": 0.0005479865883795228,
+      "grad_norm": 2.3183753490448,
+      "learning_rate": 1.6397358437543907e-06,
+      "loss": 10.231,
+      "step": 390
+    },
+    {
+      "epoch": 0.0005620375265431004,
+      "grad_norm": 2.497016668319702,
+      "learning_rate": 1.6818884361388223e-06,
+      "loss": 10.2457,
+      "step": 400
+    },
+    {
+      "epoch": 0.0005760884647066779,
+      "grad_norm": 2.5405783653259277,
+      "learning_rate": 1.724041028523254e-06,
+      "loss": 10.2305,
+      "step": 410
+    },
+    {
+      "epoch": 0.0005901394028702554,
+      "grad_norm": 2.377484083175659,
+      "learning_rate": 1.7661936209076857e-06,
+      "loss": 10.176,
+      "step": 420
+    },
+    {
+      "epoch": 0.0006041903410338329,
+      "grad_norm": 1.8428674936294556,
+      "learning_rate": 1.8083462132921172e-06,
+      "loss": 10.1656,
+      "step": 430
+    },
+    {
+      "epoch": 0.0006182412791974104,
+      "grad_norm": 2.1964094638824463,
+      "learning_rate": 1.8504988056765488e-06,
+      "loss": 10.1992,
+      "step": 440
+    },
+    {
+      "epoch": 0.0006322922173609879,
+      "grad_norm": 2.0467209815979004,
+      "learning_rate": 1.8926513980609804e-06,
+      "loss": 10.1928,
+      "step": 450
+    },
+    {
+      "epoch": 0.0006463431555245654,
+      "grad_norm": 2.0101869106292725,
+      "learning_rate": 1.9348039904454122e-06,
+      "loss": 10.1389,
+      "step": 460
+    },
+    {
+      "epoch": 0.000660394093688143,
+      "grad_norm": 1.808905005455017,
+      "learning_rate": 1.976956582829844e-06,
+      "loss": 10.1005,
+      "step": 470
+    },
+    {
+      "epoch": 0.0006744450318517205,
+      "grad_norm": 2.215162515640259,
+      "learning_rate": 2.0191091752142754e-06,
+      "loss": 10.0665,
+      "step": 480
+    },
+    {
+      "epoch": 0.0006884959700152979,
+      "grad_norm": 1.7877708673477173,
+      "learning_rate": 2.061261767598707e-06,
+      "loss": 10.0854,
+      "step": 490
+    },
+    {
+      "epoch": 0.0007025469081788755,
+      "grad_norm": 2.141770124435425,
+      "learning_rate": 2.1034143599831386e-06,
+      "loss": 10.04,
+      "step": 500
+    },
+    {
+      "epoch": 0.000716597846342453,
+      "grad_norm": 1.706669807434082,
+      "learning_rate": 2.14556695236757e-06,
+      "loss": 10.0788,
+      "step": 510
+    },
+    {
+      "epoch": 0.0007306487845060305,
+      "grad_norm": 1.7067891359329224,
+      "learning_rate": 2.1877195447520018e-06,
+      "loss": 9.9621,
+      "step": 520
+    },
+    {
+      "epoch": 0.000744699722669608,
+      "grad_norm": 1.784010410308838,
+      "learning_rate": 2.2298721371364338e-06,
+      "loss": 9.97,
+      "step": 530
+    },
+    {
+      "epoch": 0.0007587506608331855,
+      "grad_norm": 2.3449370861053467,
+      "learning_rate": 2.2720247295208654e-06,
+      "loss": 9.9688,
+      "step": 540
+    },
+    {
+      "epoch": 0.000772801598996763,
+      "grad_norm": 1.652551293373108,
+      "learning_rate": 2.314177321905297e-06,
+      "loss": 9.9469,
+      "step": 550
+    },
+    {
+      "epoch": 0.0007868525371603405,
+      "grad_norm": 1.5717960596084595,
+      "learning_rate": 2.356329914289729e-06,
+      "loss": 9.9077,
+      "step": 560
+    },
+    {
+      "epoch": 0.0008009034753239181,
+      "grad_norm": 1.4391708374023438,
+      "learning_rate": 2.3984825066741606e-06,
+      "loss": 9.9488,
+      "step": 570
+    },
+    {
+      "epoch": 0.0008149544134874955,
+      "grad_norm": 1.4368293285369873,
+      "learning_rate": 2.440635099058592e-06,
+      "loss": 9.848,
+      "step": 580
+    },
+    {
+      "epoch": 0.000829005351651073,
+      "grad_norm": 1.9761817455291748,
+      "learning_rate": 2.4827876914430237e-06,
+      "loss": 9.8789,
+      "step": 590
+    },
+    {
+      "epoch": 0.0008430562898146506,
+      "grad_norm": 1.1829458475112915,
+      "learning_rate": 2.5249402838274553e-06,
+      "loss": 9.8254,
+      "step": 600
+    },
+    {
+      "epoch": 0.0008571072279782281,
+      "grad_norm": 1.5284178256988525,
+      "learning_rate": 2.567092876211887e-06,
+      "loss": 9.8105,
+      "step": 610
+    },
+    {
+      "epoch": 0.0008711581661418056,
+      "grad_norm": 1.3892128467559814,
+      "learning_rate": 2.6092454685963185e-06,
+      "loss": 9.8078,
+      "step": 620
+    },
+    {
+      "epoch": 0.0008852091043053831,
+      "grad_norm": 1.2729390859603882,
+      "learning_rate": 2.65139806098075e-06,
+      "loss": 9.7646,
+      "step": 630
+    },
+    {
+      "epoch": 0.0008992600424689606,
+      "grad_norm": 1.1548229455947876,
+      "learning_rate": 2.6935506533651817e-06,
+      "loss": 9.7455,
+      "step": 640
+    },
+    {
+      "epoch": 0.0009133109806325381,
+      "grad_norm": 1.2774633169174194,
+      "learning_rate": 2.7357032457496133e-06,
+      "loss": 9.74,
+      "step": 650
+    },
+    {
+      "epoch": 0.0009273619187961156,
+      "grad_norm": 1.1594111919403076,
+      "learning_rate": 2.777855838134045e-06,
+      "loss": 9.7456,
+      "step": 660
+    },
+    {
+      "epoch": 0.0009414128569596932,
+      "grad_norm": 1.22251296043396,
+      "learning_rate": 2.8200084305184765e-06,
+      "loss": 9.7187,
+      "step": 670
+    },
+    {
+      "epoch": 0.0009554637951232706,
+      "grad_norm": 1.4419326782226562,
+      "learning_rate": 2.8621610229029085e-06,
+      "loss": 9.6953,
+      "step": 680
+    },
+    {
+      "epoch": 0.0009695147332868481,
+      "grad_norm": 1.0429575443267822,
+      "learning_rate": 2.90431361528734e-06,
+      "loss": 9.608,
+      "step": 690
+    },
+    {
+      "epoch": 0.0009835656714504257,
+      "grad_norm": 1.1225723028182983,
+      "learning_rate": 2.9464662076717717e-06,
+      "loss": 9.604,
+      "step": 700
+    },
+    {
+      "epoch": 0.0009976166096140032,
+      "grad_norm": 1.0744000673294067,
+      "learning_rate": 2.9886188000562033e-06,
+      "loss": 9.61,
+      "step": 710
+    },
+    {
+      "epoch": 0.0010116675477775806,
+      "grad_norm": 1.0817335844039917,
+      "learning_rate": 3.030771392440635e-06,
+      "loss": 9.533,
+      "step": 720
+    },
+    {
+      "epoch": 0.001025718485941158,
+      "grad_norm": 1.5660817623138428,
+      "learning_rate": 3.0729239848250664e-06,
+      "loss": 9.5771,
+      "step": 730
+    },
+    {
+      "epoch": 0.0010397694241047358,
+      "grad_norm": 1.141122579574585,
+      "learning_rate": 3.115076577209498e-06,
+      "loss": 9.5357,
+      "step": 740
+    },
+    {
+      "epoch": 0.0010538203622683132,
+      "grad_norm": 0.990804135799408,
+      "learning_rate": 3.1572291695939296e-06,
+      "loss": 9.5077,
+      "step": 750
+    },
+    {
+      "epoch": 0.0010678713004318907,
+      "grad_norm": 0.9177122116088867,
+      "learning_rate": 3.199381761978361e-06,
+      "loss": 9.4705,
+      "step": 760
+    },
+    {
+      "epoch": 0.0010819222385954682,
+      "grad_norm": 1.214892864227295,
+      "learning_rate": 3.241534354362793e-06,
+      "loss": 9.4865,
+      "step": 770
+    },
+    {
+      "epoch": 0.0010959731767590456,
+      "grad_norm": 0.943670392036438,
+      "learning_rate": 3.2836869467472244e-06,
+      "loss": 9.4527,
+      "step": 780
+    },
+    {
+      "epoch": 0.0011100241149226233,
+      "grad_norm": 1.3754514455795288,
+      "learning_rate": 3.325839539131656e-06,
+      "loss": 9.4666,
+      "step": 790
+    },
+    {
+      "epoch": 0.0011240750530862008,
+      "grad_norm": 1.0014880895614624,
+      "learning_rate": 3.3679921315160876e-06,
+      "loss": 9.4266,
+      "step": 800
+    },
+    {
+      "epoch": 0.0011381259912497783,
+      "grad_norm": 1.1103196144104004,
+      "learning_rate": 3.41014472390052e-06,
+      "loss": 9.4199,
+      "step": 810
+    },
+    {
+      "epoch": 0.0011521769294133557,
+      "grad_norm": 0.8306254148483276,
+      "learning_rate": 3.4522973162849516e-06,
+      "loss": 9.396,
+      "step": 820
+    },
+    {
+      "epoch": 0.0011662278675769332,
+      "grad_norm": 1.3354564905166626,
+      "learning_rate": 3.494449908669383e-06,
+      "loss": 9.4007,
+      "step": 830
+    },
+    {
+      "epoch": 0.0011802788057405109,
+      "grad_norm": 1.1535601615905762,
+      "learning_rate": 3.5366025010538148e-06,
+      "loss": 9.3241,
+      "step": 840
+    },
+    {
+      "epoch": 0.0011943297439040883,
+      "grad_norm": 0.8198082447052002,
+      "learning_rate": 3.5787550934382464e-06,
+      "loss": 9.3004,
+      "step": 850
+    },
+    {
+      "epoch": 0.0012083806820676658,
+      "grad_norm": 0.9945477247238159,
+      "learning_rate": 3.620907685822678e-06,
+      "loss": 9.2414,
+      "step": 860
+    },
+    {
+      "epoch": 0.0012224316202312433,
+      "grad_norm": 0.842752993106842,
+      "learning_rate": 3.6630602782071095e-06,
+      "loss": 9.2984,
+      "step": 870
+    },
+    {
+      "epoch": 0.0012364825583948207,
+      "grad_norm": 0.8242266774177551,
+      "learning_rate": 3.705212870591541e-06,
+      "loss": 9.2572,
+      "step": 880
+    },
+    {
+      "epoch": 0.0012505334965583984,
+      "grad_norm": 0.8884344696998596,
+      "learning_rate": 3.7473654629759727e-06,
+      "loss": 9.2445,
+      "step": 890
+    },
+    {
+      "epoch": 0.0012645844347219759,
+      "grad_norm": 0.8368833065032959,
+      "learning_rate": 3.7895180553604043e-06,
+      "loss": 9.254,
+      "step": 900
+    },
+    {
+      "epoch": 0.0012786353728855534,
+      "grad_norm": 0.9276847243309021,
+      "learning_rate": 3.831670647744836e-06,
+      "loss": 9.2425,
+      "step": 910
+    },
+    {
+      "epoch": 0.0012926863110491308,
+      "grad_norm": 0.827470064163208,
+      "learning_rate": 3.873823240129267e-06,
+      "loss": 9.2212,
+      "step": 920
+    },
+    {
+      "epoch": 0.0013067372492127083,
+      "grad_norm": 0.9183015823364258,
+      "learning_rate": 3.915975832513699e-06,
+      "loss": 9.1258,
+      "step": 930
+    },
+    {
+      "epoch": 0.001320788187376286,
+      "grad_norm": 0.8526141047477722,
+      "learning_rate": 3.958128424898131e-06,
+      "loss": 9.1636,
+      "step": 940
+    },
+    {
+      "epoch": 0.0013348391255398634,
+      "grad_norm": 0.9103283286094666,
+      "learning_rate": 4.000281017282562e-06,
+      "loss": 9.1574,
+      "step": 950
+    },
+    {
+      "epoch": 0.001348890063703441,
+      "grad_norm": 0.8416949510574341,
+      "learning_rate": 4.042433609666994e-06,
+      "loss": 9.1176,
+      "step": 960
+    },
+    {
+      "epoch": 0.0013629410018670184,
+      "grad_norm": 0.8522294759750366,
+      "learning_rate": 4.084586202051426e-06,
+      "loss": 9.0875,
+      "step": 970
+    },
+    {
+      "epoch": 0.0013769919400305958,
+      "grad_norm": 0.8377172946929932,
+      "learning_rate": 4.1267387944358575e-06,
+      "loss": 9.0616,
+      "step": 980
+    },
+    {
+      "epoch": 0.0013910428781941735,
+      "grad_norm": 0.7883292436599731,
+      "learning_rate": 4.1688913868202895e-06,
+      "loss": 9.0931,
+      "step": 990
+    },
+    {
+      "epoch": 0.001405093816357751,
+      "grad_norm": 1.2264714241027832,
+      "learning_rate": 4.211043979204721e-06,
+      "loss": 9.0459,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0014191447545213285,
+      "grad_norm": 0.8207995295524597,
+      "learning_rate": 4.253196571589153e-06,
+      "loss": 9.0147,
+      "step": 1010
+    },
+    {
+      "epoch": 0.001433195692684906,
+      "grad_norm": 0.751477062702179,
+      "learning_rate": 4.295349163973584e-06,
+      "loss": 9.0429,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0014472466308484834,
+      "grad_norm": 0.7454262971878052,
+      "learning_rate": 4.337501756358016e-06,
+      "loss": 9.0418,
+      "step": 1030
+    },
+    {
+      "epoch": 0.001461297569012061,
+      "grad_norm": 0.7283434271812439,
+      "learning_rate": 4.379654348742447e-06,
+      "loss": 9.0124,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0014753485071756385,
+      "grad_norm": 0.8492335677146912,
+      "learning_rate": 4.421806941126879e-06,
+      "loss": 9.014,
+      "step": 1050
+    },
+    {
+      "epoch": 0.001489399445339216,
+      "grad_norm": 0.7394980192184448,
+      "learning_rate": 4.463959533511311e-06,
+      "loss": 9.0262,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0015034503835027935,
+      "grad_norm": 0.8477874398231506,
+      "learning_rate": 4.506112125895742e-06,
+      "loss": 8.9394,
+      "step": 1070
+    },
+    {
+      "epoch": 0.001517501321666371,
+      "grad_norm": 0.7641156315803528,
+      "learning_rate": 4.548264718280174e-06,
+      "loss": 8.9173,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0015315522598299486,
+      "grad_norm": 0.7645962834358215,
+      "learning_rate": 4.590417310664605e-06,
+      "loss": 8.897,
+      "step": 1090
+    },
+    {
+      "epoch": 0.001545603197993526,
+      "grad_norm": 0.7267779111862183,
+      "learning_rate": 4.632569903049037e-06,
+      "loss": 8.9216,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0015596541361571035,
+      "grad_norm": 0.8410118818283081,
+      "learning_rate": 4.6747224954334686e-06,
+      "loss": 8.8682,
+      "step": 1110
+    },
+    {
+      "epoch": 0.001573705074320681,
+      "grad_norm": 0.7551319003105164,
+      "learning_rate": 4.7168750878179006e-06,
+      "loss": 8.8911,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0015877560124842585,
+      "grad_norm": 1.4869343042373657,
+      "learning_rate": 4.759027680202332e-06,
+      "loss": 8.8781,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0016018069506478362,
+      "grad_norm": 0.6931923627853394,
+      "learning_rate": 4.801180272586764e-06,
+      "loss": 8.8261,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0016158578888114136,
+      "grad_norm": 0.7531590461730957,
+      "learning_rate": 4.843332864971195e-06,
+      "loss": 8.8656,
+      "step": 1150
+    },
+    {
+      "epoch": 0.001629908826974991,
+      "grad_norm": 0.862003743648529,
+      "learning_rate": 4.885485457355627e-06,
+      "loss": 8.8532,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0016439597651385686,
+      "grad_norm": 0.6877137422561646,
+      "learning_rate": 4.927638049740058e-06,
+      "loss": 8.8643,
+      "step": 1170
+    },
+    {
+      "epoch": 0.001658010703302146,
+      "grad_norm": 0.6912798881530762,
+      "learning_rate": 4.96979064212449e-06,
+      "loss": 8.8212,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0016720616414657237,
+      "grad_norm": 0.7850083708763123,
+      "learning_rate": 5.011943234508921e-06,
+      "loss": 8.8194,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0016861125796293012,
+      "grad_norm": 0.7854230999946594,
+      "learning_rate": 5.054095826893353e-06,
+      "loss": 8.7891,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0017001635177928786,
+      "grad_norm": 0.6913483142852783,
+      "learning_rate": 5.0962484192777845e-06,
+      "loss": 8.78,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0017142144559564561,
+      "grad_norm": 0.881175696849823,
+      "learning_rate": 5.1384010116622165e-06,
+      "loss": 8.7337,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0017282653941200336,
+      "grad_norm": 2.36761736869812,
+      "learning_rate": 5.180553604046648e-06,
+      "loss": 8.7539,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0017423163322836113,
+      "grad_norm": 0.7986196875572205,
+      "learning_rate": 5.2227061964310805e-06,
+      "loss": 8.7183,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0017563672704471887,
+      "grad_norm": 0.6862300038337708,
+      "learning_rate": 5.2648587888155125e-06,
+      "loss": 8.6992,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0017704182086107662,
+      "grad_norm": 0.727844774723053,
+      "learning_rate": 5.307011381199944e-06,
+      "loss": 8.6963,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0017844691467743437,
+      "grad_norm": 0.6336658596992493,
+      "learning_rate": 5.349163973584376e-06,
+      "loss": 8.7178,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0017985200849379211,
+      "grad_norm": 0.6945951581001282,
+      "learning_rate": 5.391316565968807e-06,
+      "loss": 8.6921,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0018125710231014988,
+      "grad_norm": 1.3838273286819458,
+      "learning_rate": 5.433469158353239e-06,
+      "loss": 8.6852,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0018266219612650763,
+      "grad_norm": 1.8115507364273071,
+      "learning_rate": 5.47562175073767e-06,
+      "loss": 8.7111,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0018406728994286537,
+      "grad_norm": 0.7600476741790771,
+      "learning_rate": 5.517774343122102e-06,
+      "loss": 8.6884,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0018547238375922312,
+      "grad_norm": 0.6664397716522217,
+      "learning_rate": 5.559926935506533e-06,
+      "loss": 8.636,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0018687747757558087,
+      "grad_norm": 0.6572439074516296,
+      "learning_rate": 5.602079527890965e-06,
+      "loss": 8.61,
+      "step": 1330
+    },
+    {
+      "epoch": 0.0018828257139193864,
+      "grad_norm": 0.6541756987571716,
+      "learning_rate": 5.644232120275396e-06,
+      "loss": 8.613,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0018968766520829638,
+      "grad_norm": 0.7731262445449829,
+      "learning_rate": 5.686384712659828e-06,
+      "loss": 8.5664,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0019109275902465413,
+      "grad_norm": 0.8053380846977234,
+      "learning_rate": 5.72853730504426e-06,
+      "loss": 8.6225,
+      "step": 1360
+    },
+    {
+      "epoch": 0.0019249785284101188,
+      "grad_norm": 0.6197734475135803,
+      "learning_rate": 5.770689897428692e-06,
+      "loss": 8.564,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0019390294665736962,
+      "grad_norm": 0.623519241809845,
+      "learning_rate": 5.812842489813123e-06,
+      "loss": 8.5342,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0019530804047372737,
+      "grad_norm": 0.9381340742111206,
+      "learning_rate": 5.854995082197555e-06,
+      "loss": 8.5497,
+      "step": 1390
+    },
+    {
+      "epoch": 0.0019671313429008514,
+      "grad_norm": 0.677198052406311,
+      "learning_rate": 5.897147674581986e-06,
+      "loss": 8.5417,
+      "step": 1400
+    },
+    {
+      "epoch": 0.001981182281064429,
+      "grad_norm": 0.6278719305992126,
+      "learning_rate": 5.939300266966418e-06,
+      "loss": 8.5709,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0019952332192280063,
+      "grad_norm": 0.6248645782470703,
+      "learning_rate": 5.981452859350849e-06,
+      "loss": 8.5463,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0020092841573915838,
+      "grad_norm": 0.6436347365379333,
+      "learning_rate": 6.023605451735281e-06,
+      "loss": 8.4893,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0020233350955551612,
+      "grad_norm": 0.6340057849884033,
+      "learning_rate": 6.065758044119712e-06,
+      "loss": 8.5257,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0020373860337187387,
+      "grad_norm": 0.715142548084259,
+      "learning_rate": 6.107910636504144e-06,
+      "loss": 8.4914,
+      "step": 1450
+    },
+    {
+      "epoch": 0.002051436971882316,
+      "grad_norm": 0.7200644612312317,
+      "learning_rate": 6.1500632288885755e-06,
+      "loss": 8.4823,
+      "step": 1460
+    },
+    {
+      "epoch": 0.002065487910045894,
+      "grad_norm": 0.8068726062774658,
+      "learning_rate": 6.1922158212730075e-06,
+      "loss": 8.4923,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0020795388482094715,
+      "grad_norm": 0.6273617744445801,
+      "learning_rate": 6.2343684136574395e-06,
+      "loss": 8.4483,
+      "step": 1480
+    },
+    {
+      "epoch": 0.002093589786373049,
+      "grad_norm": 0.6604838967323303,
+      "learning_rate": 6.2765210060418715e-06,
+      "loss": 8.4753,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0021076407245366265,
+      "grad_norm": 0.6091654300689697,
+      "learning_rate": 6.318673598426303e-06,
+      "loss": 8.4524,
+      "step": 1500
+    },
+    {
+      "epoch": 0.002121691662700204,
+      "grad_norm": 0.7857375144958496,
+      "learning_rate": 6.360826190810735e-06,
+      "loss": 8.4062,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0021357426008637814,
+      "grad_norm": 0.6728093028068542,
+      "learning_rate": 6.402978783195166e-06,
+      "loss": 8.4559,
+      "step": 1520
+    },
+    {
+      "epoch": 0.002149793539027359,
+      "grad_norm": 0.6304994821548462,
+      "learning_rate": 6.445131375579598e-06,
+      "loss": 8.4665,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0021638444771909363,
+      "grad_norm": 0.6075509786605835,
+      "learning_rate": 6.48728396796403e-06,
+      "loss": 8.4142,
+      "step": 1540
+    },
+    {
+      "epoch": 0.002177895415354514,
+      "grad_norm": 0.6211352348327637,
+      "learning_rate": 6.529436560348461e-06,
+      "loss": 8.4188,
+      "step": 1550
+    },
+    {
+      "epoch": 0.0021919463535180913,
+      "grad_norm": 0.7409493327140808,
+      "learning_rate": 6.571589152732893e-06,
+      "loss": 8.3903,
+      "step": 1560
+    },
+    {
+      "epoch": 0.002205997291681669,
+      "grad_norm": 0.6649135947227478,
+      "learning_rate": 6.613741745117324e-06,
+      "loss": 8.3788,
+      "step": 1570
+    },
+    {
+      "epoch": 0.0022200482298452466,
+      "grad_norm": 0.6305505037307739,
+      "learning_rate": 6.655894337501756e-06,
+      "loss": 8.3711,
+      "step": 1580
+    },
+    {
+      "epoch": 0.002234099168008824,
+      "grad_norm": 0.5979957580566406,
+      "learning_rate": 6.698046929886187e-06,
+      "loss": 8.3786,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0022481501061724016,
+      "grad_norm": 0.6189422607421875,
+      "learning_rate": 6.7401995222706194e-06,
+      "loss": 8.3388,
+      "step": 1600
+    },
+    {
+      "epoch": 0.002262201044335979,
+      "grad_norm": 0.6126115322113037,
+      "learning_rate": 6.782352114655051e-06,
+      "loss": 8.382,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0022762519824995565,
+      "grad_norm": 0.6059759855270386,
+      "learning_rate": 6.824504707039483e-06,
+      "loss": 8.3826,
+      "step": 1620
+    },
+    {
+      "epoch": 0.002290302920663134,
+      "grad_norm": 0.6775907874107361,
+      "learning_rate": 6.866657299423914e-06,
+      "loss": 8.3403,
+      "step": 1630
+    },
+    {
+      "epoch": 0.0023043538588267114,
+      "grad_norm": 0.7798088788986206,
+      "learning_rate": 6.908809891808346e-06,
+      "loss": 8.3181,
+      "step": 1640
+    },
+    {
+      "epoch": 0.002318404796990289,
+      "grad_norm": 0.8462784886360168,
+      "learning_rate": 6.950962484192777e-06,
+      "loss": 8.3389,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0023324557351538664,
+      "grad_norm": 0.5759812593460083,
+      "learning_rate": 6.993115076577209e-06,
+      "loss": 8.3245,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0023465066733174443,
+      "grad_norm": 0.8392948508262634,
+      "learning_rate": 7.03526766896164e-06,
+      "loss": 8.3226,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0023605576114810217,
+      "grad_norm": 0.711455762386322,
+      "learning_rate": 7.077420261346072e-06,
+      "loss": 8.2694,
+      "step": 1680
+    },
+    {
+      "epoch": 0.002374608549644599,
+      "grad_norm": 0.6649680137634277,
+      "learning_rate": 7.119572853730503e-06,
+      "loss": 8.3165,
+      "step": 1690
+    },
+    {
+      "epoch": 0.0023886594878081767,
+      "grad_norm": 0.5890215635299683,
+      "learning_rate": 7.161725446114935e-06,
+      "loss": 8.3204,
+      "step": 1700
+    },
+    {
+      "epoch": 0.002402710425971754,
+      "grad_norm": 0.6956003904342651,
+      "learning_rate": 7.2038780384993665e-06,
+      "loss": 8.2913,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0024167613641353316,
+      "grad_norm": 0.6579176187515259,
+      "learning_rate": 7.2460306308837985e-06,
+      "loss": 8.2964,
+      "step": 1720
+    },
+    {
+      "epoch": 0.002430812302298909,
+      "grad_norm": 0.9485933184623718,
+      "learning_rate": 7.2881832232682305e-06,
+      "loss": 8.304,
+      "step": 1730
+    },
+    {
+      "epoch": 0.0024448632404624865,
+      "grad_norm": 0.6445179581642151,
+      "learning_rate": 7.3303358156526625e-06,
+      "loss": 8.2677,
+      "step": 1740
+    },
+    {
+      "epoch": 0.002458914178626064,
+      "grad_norm": 0.6118063926696777,
+      "learning_rate": 7.372488408037094e-06,
+      "loss": 8.3056,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0024729651167896415,
+      "grad_norm": 0.5871592164039612,
+      "learning_rate": 7.414641000421526e-06,
+      "loss": 8.2554,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0024870160549532194,
+      "grad_norm": 0.6152786612510681,
+      "learning_rate": 7.456793592805957e-06,
+      "loss": 8.2661,
+      "step": 1770
+    },
+    {
+      "epoch": 0.002501066993116797,
+      "grad_norm": 0.6126775145530701,
+      "learning_rate": 7.498946185190389e-06,
+      "loss": 8.2189,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0025151179312803743,
+      "grad_norm": 0.5303098559379578,
+      "learning_rate": 7.54109877757482e-06,
+      "loss": 8.2553,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0025291688694439518,
+      "grad_norm": 0.668865978717804,
+      "learning_rate": 7.583251369959252e-06,
+      "loss": 8.225,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0025432198076075292,
+      "grad_norm": 0.6807597875595093,
+      "learning_rate": 7.625403962343683e-06,
+      "loss": 8.2616,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0025572707457711067,
+      "grad_norm": 0.5899785757064819,
+      "learning_rate": 7.667556554728115e-06,
+      "loss": 8.2212,
+      "step": 1820
+    },
+    {
+      "epoch": 0.002571321683934684,
+      "grad_norm": 0.7344907522201538,
+      "learning_rate": 7.709709147112546e-06,
+      "loss": 8.2091,
+      "step": 1830
+    },
+    {
+      "epoch": 0.0025853726220982616,
+      "grad_norm": 0.6710782051086426,
+      "learning_rate": 7.751861739496978e-06,
+      "loss": 8.2347,
+      "step": 1840
+    },
+    {
+      "epoch": 0.002599423560261839,
+      "grad_norm": 0.5794618725776672,
+      "learning_rate": 7.79401433188141e-06,
+      "loss": 8.1866,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0026134744984254166,
+      "grad_norm": 0.6005648970603943,
+      "learning_rate": 7.836166924265842e-06,
+      "loss": 8.2057,
+      "step": 1860
+    },
+    {
+      "epoch": 0.002627525436588994,
+      "grad_norm": 0.6351447701454163,
+      "learning_rate": 7.878319516650273e-06,
+      "loss": 8.2069,
+      "step": 1870
+    },
+    {
+      "epoch": 0.002641576374752572,
+      "grad_norm": 2.6223857402801514,
+      "learning_rate": 7.920472109034706e-06,
+      "loss": 8.2184,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0026556273129161494,
+      "grad_norm": 0.6811882257461548,
+      "learning_rate": 7.962624701419137e-06,
+      "loss": 8.197,
+      "step": 1890
+    },
+    {
+      "epoch": 0.002669678251079727,
+      "grad_norm": 0.6392279863357544,
+      "learning_rate": 8.004777293803568e-06,
+      "loss": 8.212,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0026837291892433043,
+      "grad_norm": 0.6159211993217468,
+      "learning_rate": 8.046929886187999e-06,
+      "loss": 8.2124,
+      "step": 1910
+    },
+    {
+      "epoch": 0.002697780127406882,
+      "grad_norm": 0.9476580619812012,
+      "learning_rate": 8.089082478572432e-06,
+      "loss": 8.1537,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0027118310655704593,
+      "grad_norm": 0.645937442779541,
+      "learning_rate": 8.131235070956863e-06,
+      "loss": 8.1634,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0027258820037340367,
+      "grad_norm": 0.6065277457237244,
+      "learning_rate": 8.173387663341294e-06,
+      "loss": 8.1557,
+      "step": 1940
+    },
+    {
+      "epoch": 0.002739932941897614,
+      "grad_norm": 0.6352872848510742,
+      "learning_rate": 8.215540255725726e-06,
+      "loss": 8.1722,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0027539838800611917,
+      "grad_norm": 0.5992732644081116,
+      "learning_rate": 8.257692848110158e-06,
+      "loss": 8.1705,
+      "step": 1960
+    },
+    {
+      "epoch": 0.002768034818224769,
+      "grad_norm": 0.8603838682174683,
+      "learning_rate": 8.29984544049459e-06,
+      "loss": 8.1483,
+      "step": 1970
+    },
+    {
+      "epoch": 0.002782085756388347,
+      "grad_norm": 0.7166236042976379,
+      "learning_rate": 8.341998032879022e-06,
+      "loss": 8.1939,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0027961366945519245,
+      "grad_norm": 0.6886283755302429,
+      "learning_rate": 8.384150625263454e-06,
+      "loss": 8.1459,
+      "step": 1990
+    },
+    {
+      "epoch": 0.002810187632715502,
+      "grad_norm": 0.6554070115089417,
+      "learning_rate": 8.426303217647885e-06,
+      "loss": 8.1696,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0028242385708790794,
+      "grad_norm": 0.6008890271186829,
+      "learning_rate": 8.468455810032316e-06,
+      "loss": 8.0969,
+      "step": 2010
+    },
+    {
+      "epoch": 0.002838289509042657,
+      "grad_norm": 0.5523585081100464,
+      "learning_rate": 8.510608402416749e-06,
+      "loss": 8.1298,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0028523404472062344,
+      "grad_norm": 0.616106390953064,
+      "learning_rate": 8.55276099480118e-06,
+      "loss": 8.1633,
+      "step": 2030
+    },
+    {
+      "epoch": 0.002866391385369812,
+      "grad_norm": 0.7677391767501831,
+      "learning_rate": 8.594913587185611e-06,
+      "loss": 8.1168,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0028804423235333893,
+      "grad_norm": 1.135625958442688,
+      "learning_rate": 8.637066179570042e-06,
+      "loss": 8.1114,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0028944932616969668,
+      "grad_norm": 0.6018761396408081,
+      "learning_rate": 8.679218771954475e-06,
+      "loss": 8.125,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0029085441998605442,
+      "grad_norm": 0.6487704515457153,
+      "learning_rate": 8.721371364338906e-06,
+      "loss": 8.1119,
+      "step": 2070
+    },
+    {
+      "epoch": 0.002922595138024122,
+      "grad_norm": 0.6532260775566101,
+      "learning_rate": 8.763523956723337e-06,
+      "loss": 8.0926,
+      "step": 2080
+    },
+    {
+      "epoch": 0.0029366460761876996,
+      "grad_norm": 0.7966541051864624,
+      "learning_rate": 8.805676549107769e-06,
+      "loss": 8.1293,
+      "step": 2090
+    },
+    {
+      "epoch": 0.002950697014351277,
+      "grad_norm": 0.5987772345542908,
+      "learning_rate": 8.847829141492201e-06,
+      "loss": 8.0792,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0029647479525148545,
+      "grad_norm": 0.5964518189430237,
+      "learning_rate": 8.889981733876633e-06,
+      "loss": 8.1097,
+      "step": 2110
+    },
+    {
+      "epoch": 0.002978798890678432,
+      "grad_norm": 0.6664417386054993,
+      "learning_rate": 8.932134326261064e-06,
+      "loss": 8.0914,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0029928498288420095,
+      "grad_norm": 0.7204576730728149,
+      "learning_rate": 8.974286918645495e-06,
+      "loss": 8.0853,
+      "step": 2130
+    },
+    {
+      "epoch": 0.003006900767005587,
+      "grad_norm": 0.698284924030304,
+      "learning_rate": 9.016439511029928e-06,
+      "loss": 8.1351,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0030209517051691644,
+      "grad_norm": 0.6838796138763428,
+      "learning_rate": 9.058592103414359e-06,
+      "loss": 8.0672,
+      "step": 2150
+    },
+    {
+      "epoch": 0.003035002643332742,
+      "grad_norm": 0.6164630651473999,
+      "learning_rate": 9.10074469579879e-06,
+      "loss": 8.1009,
+      "step": 2160
+    },
+    {
+      "epoch": 0.0030490535814963193,
+      "grad_norm": 0.7379197478294373,
+      "learning_rate": 9.142897288183221e-06,
+      "loss": 8.05,
+      "step": 2170
+    },
+    {
+      "epoch": 0.0030631045196598972,
+      "grad_norm": 0.6812073588371277,
+      "learning_rate": 9.185049880567654e-06,
+      "loss": 8.0716,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0030771554578234747,
+      "grad_norm": 0.6673164963722229,
+      "learning_rate": 9.227202472952085e-06,
+      "loss": 8.1383,
+      "step": 2190
+    },
+    {
+      "epoch": 0.003091206395987052,
+      "grad_norm": 0.7186004519462585,
+      "learning_rate": 9.269355065336517e-06,
+      "loss": 8.0792,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0031052573341506296,
+      "grad_norm": 0.9671910405158997,
+      "learning_rate": 9.311507657720948e-06,
+      "loss": 8.0954,
+      "step": 2210
+    },
+    {
+      "epoch": 0.003119308272314207,
+      "grad_norm": 0.6637594103813171,
+      "learning_rate": 9.35366025010538e-06,
+      "loss": 8.0612,
+      "step": 2220
+    },
+    {
+      "epoch": 0.0031333592104777846,
+      "grad_norm": 0.6269683837890625,
+      "learning_rate": 9.395812842489812e-06,
+      "loss": 8.0422,
+      "step": 2230
+    },
+    {
+      "epoch": 0.003147410148641362,
+      "grad_norm": 0.6617890000343323,
+      "learning_rate": 9.437965434874245e-06,
+      "loss": 8.0337,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0031614610868049395,
+      "grad_norm": 0.6194980144500732,
+      "learning_rate": 9.480118027258676e-06,
+      "loss": 8.099,
+      "step": 2250
+    },
+    {
+      "epoch": 0.003175512024968517,
+      "grad_norm": 0.6944716572761536,
+      "learning_rate": 9.522270619643107e-06,
+      "loss": 8.0398,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0031895629631320944,
+      "grad_norm": 0.6326583623886108,
+      "learning_rate": 9.564423212027538e-06,
+      "loss": 8.0368,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0032036139012956723,
+      "grad_norm": 0.583436906337738,
+      "learning_rate": 9.606575804411971e-06,
+      "loss": 8.0594,
+      "step": 2280
+    },
+    {
+      "epoch": 0.00321766483945925,
+      "grad_norm": 0.6804009079933167,
+      "learning_rate": 9.648728396796402e-06,
+      "loss": 8.0141,
+      "step": 2290
+    },
+    {
+      "epoch": 0.0032317157776228273,
+      "grad_norm": 0.6234678030014038,
+      "learning_rate": 9.690880989180833e-06,
+      "loss": 7.9864,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0032457667157864047,
+      "grad_norm": 0.6890107989311218,
+      "learning_rate": 9.733033581565264e-06,
+      "loss": 8.0474,
+      "step": 2310
+    },
+    {
+      "epoch": 0.003259817653949982,
+      "grad_norm": 0.7134827375411987,
+      "learning_rate": 9.775186173949697e-06,
+      "loss": 8.0306,
+      "step": 2320
+    },
+    {
+      "epoch": 0.0032738685921135597,
+      "grad_norm": 0.6207192540168762,
+      "learning_rate": 9.817338766334128e-06,
+      "loss": 7.9592,
+      "step": 2330
+    },
+    {
+      "epoch": 0.003287919530277137,
+      "grad_norm": 1.1250572204589844,
+      "learning_rate": 9.85949135871856e-06,
+      "loss": 8.0243,
+      "step": 2340
+    },
+    {
+      "epoch": 0.0033019704684407146,
+      "grad_norm": 0.7023635506629944,
+      "learning_rate": 9.901643951102993e-06,
+      "loss": 8.017,
+      "step": 2350
+    },
+    {
+      "epoch": 0.003316021406604292,
+      "grad_norm": 0.6336594820022583,
+      "learning_rate": 9.943796543487424e-06,
+      "loss": 8.0084,
+      "step": 2360
+    },
+    {
+      "epoch": 0.0033300723447678695,
+      "grad_norm": 0.7274377942085266,
+      "learning_rate": 9.985949135871857e-06,
+      "loss": 8.0152,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0033441232829314474,
+      "grad_norm": 0.8217412829399109,
+      "learning_rate": 1.0028101728256286e-05,
+      "loss": 7.9932,
+      "step": 2380
+    },
+    {
+      "epoch": 0.003358174221095025,
+      "grad_norm": 0.7048346400260925,
+      "learning_rate": 1.0070254320640719e-05,
+      "loss": 7.9944,
+      "step": 2390
+    },
+    {
+      "epoch": 0.0033722251592586024,
+      "grad_norm": 0.6946651935577393,
+      "learning_rate": 1.011240691302515e-05,
+      "loss": 7.9447,
+      "step": 2400
+    },
+    {
+      "epoch": 0.00338627609742218,
+      "grad_norm": 0.7006087899208069,
+      "learning_rate": 1.0154559505409583e-05,
+      "loss": 8.0093,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0034003270355857573,
+      "grad_norm": 0.8661535382270813,
+      "learning_rate": 1.0196712097794012e-05,
+      "loss": 8.0699,
+      "step": 2420
+    },
+    {
+      "epoch": 0.0034143779737493348,
+      "grad_norm": 0.8086657524108887,
+      "learning_rate": 1.0238864690178445e-05,
+      "loss": 7.9983,
+      "step": 2430
+    },
+    {
+      "epoch": 0.0034284289119129122,
+      "grad_norm": 0.6144962906837463,
+      "learning_rate": 1.0281017282562876e-05,
+      "loss": 8.0364,
+      "step": 2440
+    },
+    {
+      "epoch": 0.0034424798500764897,
+      "grad_norm": 0.7119608521461487,
+      "learning_rate": 1.032316987494731e-05,
+      "loss": 7.9961,
+      "step": 2450
+    },
+    {
+      "epoch": 0.003456530788240067,
+      "grad_norm": 0.7629860043525696,
+      "learning_rate": 1.0365322467331739e-05,
+      "loss": 7.9789,
+      "step": 2460
+    },
+    {
+      "epoch": 0.0034705817264036446,
+      "grad_norm": 0.6489969491958618,
+      "learning_rate": 1.0407475059716172e-05,
+      "loss": 7.9433,
+      "step": 2470
+    },
+    {
+      "epoch": 0.0034846326645672225,
+      "grad_norm": 0.6559635400772095,
+      "learning_rate": 1.0449627652100603e-05,
+      "loss": 7.9848,
+      "step": 2480
+    },
+    {
+      "epoch": 0.0034986836027308,
+      "grad_norm": 0.6898662447929382,
+      "learning_rate": 1.0491780244485036e-05,
+      "loss": 7.9388,
+      "step": 2490
+    },
+    {
+      "epoch": 0.0035127345408943775,
+      "grad_norm": 0.7968945503234863,
+      "learning_rate": 1.0533932836869465e-05,
+      "loss": 7.987,
+      "step": 2500
+    },
+    {
+      "epoch": 0.003526785479057955,
+      "grad_norm": 0.8573049306869507,
+      "learning_rate": 1.0576085429253898e-05,
+      "loss": 7.937,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0035408364172215324,
+      "grad_norm": 0.7501293420791626,
+      "learning_rate": 1.0618238021638329e-05,
+      "loss": 7.9119,
+      "step": 2520
+    },
+    {
+      "epoch": 0.00355488735538511,
+      "grad_norm": 0.6564654111862183,
+      "learning_rate": 1.0660390614022762e-05,
+      "loss": 7.9062,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0035689382935486873,
+      "grad_norm": 0.7637408971786499,
+      "learning_rate": 1.0702543206407191e-05,
+      "loss": 7.9529,
+      "step": 2540
+    },
+    {
+      "epoch": 0.003582989231712265,
+      "grad_norm": 0.6853562593460083,
+      "learning_rate": 1.0744695798791624e-05,
+      "loss": 7.9617,
+      "step": 2550
+    },
+    {
+      "epoch": 0.0035970401698758423,
+      "grad_norm": 0.9160966873168945,
+      "learning_rate": 1.0786848391176055e-05,
+      "loss": 7.95,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0036110911080394197,
+      "grad_norm": 0.7625723481178284,
+      "learning_rate": 1.0829000983560488e-05,
+      "loss": 7.9144,
+      "step": 2570
+    },
+    {
+      "epoch": 0.0036251420462029976,
+      "grad_norm": 0.6952919363975525,
+      "learning_rate": 1.0871153575944918e-05,
+      "loss": 7.9539,
+      "step": 2580
+    },
+    {
+      "epoch": 0.003639192984366575,
+      "grad_norm": 0.8298730254173279,
+      "learning_rate": 1.091330616832935e-05,
+      "loss": 7.9589,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0036532439225301526,
+      "grad_norm": 0.8768827319145203,
+      "learning_rate": 1.0955458760713784e-05,
+      "loss": 7.9266,
+      "step": 2600
+    },
+    {
+      "epoch": 0.00366729486069373,
+      "grad_norm": 0.6922865509986877,
+      "learning_rate": 1.0997611353098215e-05,
+      "loss": 7.9015,
+      "step": 2610
+    },
+    {
+      "epoch": 0.0036813457988573075,
+      "grad_norm": 0.7246437668800354,
+      "learning_rate": 1.1039763945482648e-05,
+      "loss": 7.9234,
+      "step": 2620
+    },
+    {
+      "epoch": 0.003695396737020885,
+      "grad_norm": 0.8041806817054749,
+      "learning_rate": 1.1081916537867077e-05,
+      "loss": 7.9054,
+      "step": 2630
+    },
+    {
+      "epoch": 0.0037094476751844624,
+      "grad_norm": 0.7150896191596985,
+      "learning_rate": 1.112406913025151e-05,
+      "loss": 7.9446,
+      "step": 2640
+    },
+    {
+      "epoch": 0.00372349861334804,
+      "grad_norm": 0.8792434930801392,
+      "learning_rate": 1.1166221722635941e-05,
+      "loss": 7.9111,
+      "step": 2650
+    },
+    {
+      "epoch": 0.0037375495515116174,
+      "grad_norm": 0.7889289259910583,
+      "learning_rate": 1.1208374315020374e-05,
+      "loss": 7.9393,
+      "step": 2660
+    },
+    {
+      "epoch": 0.003751600489675195,
+      "grad_norm": 0.8603889346122742,
+      "learning_rate": 1.1250526907404803e-05,
+      "loss": 7.8837,
+      "step": 2670
+    },
+    {
+      "epoch": 0.0037656514278387727,
+      "grad_norm": 0.8094549179077148,
+      "learning_rate": 1.1292679499789236e-05,
+      "loss": 7.9193,
+      "step": 2680
+    },
+    {
+      "epoch": 0.00377970236600235,
+      "grad_norm": 0.5979373455047607,
+      "learning_rate": 1.1334832092173667e-05,
+      "loss": 7.8922,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0037937533041659277,
+      "grad_norm": 0.8888489007949829,
+      "learning_rate": 1.13769846845581e-05,
+      "loss": 7.8884,
+      "step": 2700
+    },
+    {
+      "epoch": 0.003807804242329505,
+      "grad_norm": 0.7355663180351257,
+      "learning_rate": 1.141913727694253e-05,
+      "loss": 7.9256,
+      "step": 2710
+    },
+    {
+      "epoch": 0.0038218551804930826,
+      "grad_norm": 0.7639088034629822,
+      "learning_rate": 1.1461289869326963e-05,
+      "loss": 7.8711,
+      "step": 2720
+    },
+    {
+      "epoch": 0.00383590611865666,
+      "grad_norm": 1.0683914422988892,
+      "learning_rate": 1.1503442461711394e-05,
+      "loss": 7.8724,
+      "step": 2730
+    },
+    {
+      "epoch": 0.0038499570568202375,
+      "grad_norm": 0.99086594581604,
+      "learning_rate": 1.1545595054095827e-05,
+      "loss": 7.8734,
+      "step": 2740
+    },
+    {
+      "epoch": 0.003864007994983815,
+      "grad_norm": 0.7015689015388489,
+      "learning_rate": 1.1587747646480256e-05,
+      "loss": 7.893,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0038780589331473925,
+      "grad_norm": 0.7968684434890747,
+      "learning_rate": 1.1629900238864689e-05,
+      "loss": 7.8584,
+      "step": 2760
+    },
+    {
+      "epoch": 0.00389210987131097,
+      "grad_norm": 0.7174790501594543,
+      "learning_rate": 1.167205283124912e-05,
+      "loss": 7.8578,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0039061608094745474,
+      "grad_norm": 0.818223237991333,
+      "learning_rate": 1.1714205423633553e-05,
+      "loss": 7.8421,
+      "step": 2780
+    },
+    {
+      "epoch": 0.003920211747638125,
+      "grad_norm": 0.7619180083274841,
+      "learning_rate": 1.1756358016017982e-05,
+      "loss": 7.8614,
+      "step": 2790
+    },
+    {
+      "epoch": 0.003934262685801703,
+      "grad_norm": 1.0655688047409058,
+      "learning_rate": 1.1798510608402415e-05,
+      "loss": 7.8461,
+      "step": 2800
+    },
+    {
+      "epoch": 0.00394831362396528,
+      "grad_norm": 0.7509090304374695,
+      "learning_rate": 1.1840663200786846e-05,
+      "loss": 7.857,
+      "step": 2810
+    },
+    {
+      "epoch": 0.003962364562128858,
+      "grad_norm": 0.7447671890258789,
+      "learning_rate": 1.188281579317128e-05,
+      "loss": 7.8801,
+      "step": 2820
+    },
+    {
+      "epoch": 0.003976415500292435,
+      "grad_norm": 0.9486709237098694,
+      "learning_rate": 1.1924968385555709e-05,
+      "loss": 7.7945,
+      "step": 2830
+    },
+    {
+      "epoch": 0.003990466438456013,
+      "grad_norm": 0.6755268573760986,
+      "learning_rate": 1.1967120977940142e-05,
+      "loss": 7.8919,
+      "step": 2840
+    },
+    {
+      "epoch": 0.0040045173766195905,
+      "grad_norm": 0.8210179805755615,
+      "learning_rate": 1.2009273570324575e-05,
+      "loss": 7.8437,
+      "step": 2850
+    },
+    {
+      "epoch": 0.0040185683147831676,
+      "grad_norm": 0.7738075256347656,
+      "learning_rate": 1.2051426162709006e-05,
+      "loss": 7.7473,
+      "step": 2860
+    },
+    {
+      "epoch": 0.0040326192529467455,
+      "grad_norm": 1.1072646379470825,
+      "learning_rate": 1.2093578755093439e-05,
+      "loss": 7.8089,
+      "step": 2870
+    },
+    {
+      "epoch": 0.0040466701911103225,
+      "grad_norm": 0.7715216279029846,
+      "learning_rate": 1.2135731347477868e-05,
+      "loss": 7.8353,
+      "step": 2880
+    },
+    {
+      "epoch": 0.0040607211292739,
+      "grad_norm": 0.8774746656417847,
+      "learning_rate": 1.2177883939862301e-05,
+      "loss": 7.7942,
+      "step": 2890
+    },
+    {
+      "epoch": 0.004074772067437477,
+      "grad_norm": 0.8812255263328552,
+      "learning_rate": 1.2220036532246732e-05,
+      "loss": 7.8712,
+      "step": 2900
+    },
+    {
+      "epoch": 0.004088823005601055,
+      "grad_norm": 0.8199585676193237,
+      "learning_rate": 1.2262189124631165e-05,
+      "loss": 7.8129,
+      "step": 2910
+    },
+    {
+      "epoch": 0.004102873943764632,
+      "grad_norm": 0.6967678666114807,
+      "learning_rate": 1.2304341717015594e-05,
+      "loss": 7.7619,
+      "step": 2920
+    },
+    {
+      "epoch": 0.00411692488192821,
+      "grad_norm": 0.744611918926239,
+      "learning_rate": 1.2346494309400027e-05,
+      "loss": 7.8447,
+      "step": 2930
+    },
+    {
+      "epoch": 0.004130975820091788,
+      "grad_norm": 0.8083895444869995,
+      "learning_rate": 1.2388646901784458e-05,
+      "loss": 7.8054,
+      "step": 2940
+    },
+    {
+      "epoch": 0.004145026758255365,
+      "grad_norm": 0.8558597564697266,
+      "learning_rate": 1.2430799494168891e-05,
+      "loss": 7.756,
+      "step": 2950
+    },
+    {
+      "epoch": 0.004159077696418943,
+      "grad_norm": 0.9740859866142273,
+      "learning_rate": 1.247295208655332e-05,
+      "loss": 7.7429,
+      "step": 2960
+    },
+    {
+      "epoch": 0.00417312863458252,
+      "grad_norm": 0.7745287418365479,
+      "learning_rate": 1.2515104678937754e-05,
+      "loss": 7.7885,
+      "step": 2970
+    },
+    {
+      "epoch": 0.004187179572746098,
+      "grad_norm": 0.8137255907058716,
+      "learning_rate": 1.2557257271322185e-05,
+      "loss": 7.7817,
+      "step": 2980
+    },
+    {
+      "epoch": 0.004201230510909675,
+      "grad_norm": 0.8644803166389465,
+      "learning_rate": 1.2599409863706618e-05,
+      "loss": 7.8588,
+      "step": 2990
+    },
+    {
+      "epoch": 0.004215281449073253,
+      "grad_norm": 0.8625914454460144,
+      "learning_rate": 1.2641562456091047e-05,
+      "loss": 7.728,
+      "step": 3000
+    },
+    {
+      "epoch": 0.00422933238723683,
+      "grad_norm": 1.1142210960388184,
+      "learning_rate": 1.268371504847548e-05,
+      "loss": 7.7437,
+      "step": 3010
+    },
+    {
+      "epoch": 0.004243383325400408,
+      "grad_norm": 0.852684736251831,
+      "learning_rate": 1.2725867640859911e-05,
+      "loss": 7.7905,
+      "step": 3020
+    },
+    {
+      "epoch": 0.004257434263563985,
+      "grad_norm": 0.8908416628837585,
+      "learning_rate": 1.2768020233244344e-05,
+      "loss": 7.6894,
+      "step": 3030
+    },
+    {
+      "epoch": 0.004271485201727563,
+      "grad_norm": 0.8012714982032776,
+      "learning_rate": 1.2810172825628773e-05,
+      "loss": 7.8086,
+      "step": 3040
+    },
+    {
+      "epoch": 0.004285536139891141,
+      "grad_norm": 0.6924638748168945,
+      "learning_rate": 1.2852325418013206e-05,
+      "loss": 7.8487,
+      "step": 3050
+    },
+    {
+      "epoch": 0.004299587078054718,
+      "grad_norm": 0.8051576614379883,
+      "learning_rate": 1.2894478010397638e-05,
+      "loss": 7.7666,
+      "step": 3060
+    },
+    {
+      "epoch": 0.004313638016218296,
+      "grad_norm": 1.3711603879928589,
+      "learning_rate": 1.293663060278207e-05,
+      "loss": 7.7913,
+      "step": 3070
+    },
+    {
+      "epoch": 0.004327688954381873,
+      "grad_norm": 0.8338873982429504,
+      "learning_rate": 1.2978783195166503e-05,
+      "loss": 7.7112,
+      "step": 3080
+    },
+    {
+      "epoch": 0.004341739892545451,
+      "grad_norm": 0.7337360978126526,
+      "learning_rate": 1.3020935787550933e-05,
+      "loss": 7.753,
+      "step": 3090
+    },
+    {
+      "epoch": 0.004355790830709028,
+      "grad_norm": 0.888911247253418,
+      "learning_rate": 1.3063088379935366e-05,
+      "loss": 7.6947,
+      "step": 3100
+    },
+    {
+      "epoch": 0.0043698417688726055,
+      "grad_norm": 0.852385401725769,
+      "learning_rate": 1.3105240972319797e-05,
+      "loss": 7.7105,
+      "step": 3110
+    },
+    {
+      "epoch": 0.0043838927070361825,
+      "grad_norm": 0.840452253818512,
+      "learning_rate": 1.314739356470423e-05,
+      "loss": 7.8018,
+      "step": 3120
+    },
+    {
+      "epoch": 0.0043979436451997604,
+      "grad_norm": 0.8020384311676025,
+      "learning_rate": 1.3189546157088659e-05,
+      "loss": 7.7558,
+      "step": 3130
+    },
+    {
+      "epoch": 0.004411994583363338,
+      "grad_norm": 0.8536344170570374,
+      "learning_rate": 1.3231698749473092e-05,
+      "loss": 7.6887,
+      "step": 3140
+    },
+    {
+      "epoch": 0.004426045521526915,
+      "grad_norm": 0.8429573178291321,
+      "learning_rate": 1.3273851341857523e-05,
+      "loss": 7.7193,
+      "step": 3150
+    },
+    {
+      "epoch": 0.004440096459690493,
+      "grad_norm": 0.9258587956428528,
+      "learning_rate": 1.3316003934241956e-05,
+      "loss": 7.7103,
+      "step": 3160
+    },
+    {
+      "epoch": 0.00445414739785407,
+      "grad_norm": 0.7705568671226501,
+      "learning_rate": 1.3358156526626385e-05,
+      "loss": 7.707,
+      "step": 3170
+    },
+    {
+      "epoch": 0.004468198336017648,
+      "grad_norm": 1.0604978799819946,
+      "learning_rate": 1.3400309119010818e-05,
+      "loss": 7.7287,
+      "step": 3180
+    },
+    {
+      "epoch": 0.004482249274181225,
+      "grad_norm": 0.8479204177856445,
+      "learning_rate": 1.344246171139525e-05,
+      "loss": 7.6984,
+      "step": 3190
+    },
+    {
+      "epoch": 0.004496300212344803,
+      "grad_norm": 0.8553823232650757,
+      "learning_rate": 1.3484614303779682e-05,
+      "loss": 7.6764,
+      "step": 3200
+    },
+    {
+      "epoch": 0.00451035115050838,
+      "grad_norm": 0.9012032151222229,
+      "learning_rate": 1.3526766896164112e-05,
+      "loss": 7.7013,
+      "step": 3210
+    },
+    {
+      "epoch": 0.004524402088671958,
+      "grad_norm": 0.8775386810302734,
+      "learning_rate": 1.3568919488548545e-05,
+      "loss": 7.7516,
+      "step": 3220
+    },
+    {
+      "epoch": 0.004538453026835535,
+      "grad_norm": 1.1585534811019897,
+      "learning_rate": 1.3611072080932976e-05,
+      "loss": 7.6667,
+      "step": 3230
+    },
+    {
+      "epoch": 0.004552503964999113,
+      "grad_norm": 0.9868244528770447,
+      "learning_rate": 1.3653224673317409e-05,
+      "loss": 7.6702,
+      "step": 3240
+    },
+    {
+      "epoch": 0.004566554903162691,
+      "grad_norm": 0.9672613143920898,
+      "learning_rate": 1.3695377265701838e-05,
+      "loss": 7.6212,
+      "step": 3250
+    },
+    {
+      "epoch": 0.004580605841326268,
+      "grad_norm": 1.0549743175506592,
+      "learning_rate": 1.3737529858086271e-05,
+      "loss": 7.6721,
+      "step": 3260
+    },
+    {
+      "epoch": 0.004594656779489846,
+      "grad_norm": 0.8549468517303467,
+      "learning_rate": 1.3779682450470702e-05,
+      "loss": 7.809,
+      "step": 3270
+    },
+    {
+      "epoch": 0.004608707717653423,
+      "grad_norm": 1.5186914205551147,
+      "learning_rate": 1.3821835042855135e-05,
+      "loss": 7.6878,
+      "step": 3280
+    },
+    {
+      "epoch": 0.004622758655817001,
+      "grad_norm": 1.0272847414016724,
+      "learning_rate": 1.3863987635239565e-05,
+      "loss": 7.6362,
+      "step": 3290
+    },
+    {
+      "epoch": 0.004636809593980578,
+      "grad_norm": 0.8402895927429199,
+      "learning_rate": 1.3906140227623997e-05,
+      "loss": 7.6875,
+      "step": 3300
+    },
+    {
+      "epoch": 0.004650860532144156,
+      "grad_norm": 0.8236798644065857,
+      "learning_rate": 1.3948292820008429e-05,
+      "loss": 7.6046,
+      "step": 3310
+    },
+    {
+      "epoch": 0.004664911470307733,
+      "grad_norm": 1.09159517288208,
+      "learning_rate": 1.3990445412392861e-05,
+      "loss": 7.7144,
+      "step": 3320
+    },
+    {
+      "epoch": 0.004678962408471311,
+      "grad_norm": 0.9624311327934265,
+      "learning_rate": 1.4032598004777294e-05,
+      "loss": 7.6603,
+      "step": 3330
+    },
+    {
+      "epoch": 0.0046930133466348885,
+      "grad_norm": 0.9037488698959351,
+      "learning_rate": 1.4074750597161724e-05,
+      "loss": 7.6887,
+      "step": 3340
+    },
+    {
+      "epoch": 0.004707064284798466,
+      "grad_norm": 0.8625498414039612,
+      "learning_rate": 1.4116903189546157e-05,
+      "loss": 7.6297,
+      "step": 3350
+    },
+    {
+      "epoch": 0.0047211152229620435,
+      "grad_norm": 0.9417233467102051,
+      "learning_rate": 1.4159055781930588e-05,
+      "loss": 7.6359,
+      "step": 3360
+    },
+    {
+      "epoch": 0.0047351661611256205,
+      "grad_norm": 0.9447644948959351,
+      "learning_rate": 1.420120837431502e-05,
+      "loss": 7.6173,
+      "step": 3370
+    },
+    {
+      "epoch": 0.004749217099289198,
+      "grad_norm": 0.9068923592567444,
+      "learning_rate": 1.424336096669945e-05,
+      "loss": 7.6694,
+      "step": 3380
+    },
+    {
+      "epoch": 0.0047632680374527754,
+      "grad_norm": 0.9849382042884827,
+      "learning_rate": 1.4285513559083883e-05,
+      "loss": 7.6411,
+      "step": 3390
+    },
+    {
+      "epoch": 0.004777318975616353,
+      "grad_norm": 0.8907585740089417,
+      "learning_rate": 1.4327666151468314e-05,
+      "loss": 7.6669,
+      "step": 3400
+    },
+    {
+      "epoch": 0.00479136991377993,
+      "grad_norm": 1.1062923669815063,
+      "learning_rate": 1.4369818743852747e-05,
+      "loss": 7.6615,
+      "step": 3410
+    },
+    {
+      "epoch": 0.004805420851943508,
+      "grad_norm": 0.9381359219551086,
+      "learning_rate": 1.4411971336237176e-05,
+      "loss": 7.6943,
+      "step": 3420
+    },
+    {
+      "epoch": 0.004819471790107085,
+      "grad_norm": 1.3278305530548096,
+      "learning_rate": 1.445412392862161e-05,
+      "loss": 7.6195,
+      "step": 3430
+    },
+    {
+      "epoch": 0.004833522728270663,
+      "grad_norm": 0.8325286507606506,
+      "learning_rate": 1.449627652100604e-05,
+      "loss": 7.6601,
+      "step": 3440
+    },
+    {
+      "epoch": 0.004847573666434241,
+      "grad_norm": 0.9755851030349731,
+      "learning_rate": 1.4538429113390473e-05,
+      "loss": 7.6727,
+      "step": 3450
+    },
+    {
+      "epoch": 0.004861624604597818,
+      "grad_norm": 0.8660493493080139,
+      "learning_rate": 1.4580581705774903e-05,
+      "loss": 7.5598,
+      "step": 3460
+    },
+    {
+      "epoch": 0.004875675542761396,
+      "grad_norm": 0.8259555101394653,
+      "learning_rate": 1.4622734298159336e-05,
+      "loss": 7.5946,
+      "step": 3470
+    },
+    {
+      "epoch": 0.004889726480924973,
+      "grad_norm": 0.9008260369300842,
+      "learning_rate": 1.4664886890543767e-05,
+      "loss": 7.5731,
+      "step": 3480
+    },
+    {
+      "epoch": 0.004903777419088551,
+      "grad_norm": 0.8646583557128906,
+      "learning_rate": 1.47070394829282e-05,
+      "loss": 7.602,
+      "step": 3490
+    },
+    {
+      "epoch": 0.004917828357252128,
+      "grad_norm": 0.8853713870048523,
+      "learning_rate": 1.474919207531263e-05,
+      "loss": 7.5559,
+      "step": 3500
+    },
+    {
+      "epoch": 0.004931879295415706,
+      "grad_norm": 0.8749480247497559,
+      "learning_rate": 1.4791344667697062e-05,
+      "loss": 7.5826,
+      "step": 3510
+    },
+    {
+      "epoch": 0.004945930233579283,
+      "grad_norm": 0.8559566736221313,
+      "learning_rate": 1.4833497260081493e-05,
+      "loss": 7.6301,
+      "step": 3520
+    },
+    {
+      "epoch": 0.004959981171742861,
+      "grad_norm": 0.8258165121078491,
+      "learning_rate": 1.4875649852465926e-05,
+      "loss": 7.5909,
+      "step": 3530
+    },
+    {
+      "epoch": 0.004974032109906439,
+      "grad_norm": 0.9734731912612915,
+      "learning_rate": 1.4917802444850356e-05,
+      "loss": 7.5878,
+      "step": 3540
+    },
+    {
+      "epoch": 0.004988083048070016,
+      "grad_norm": 0.9722371697425842,
+      "learning_rate": 1.4959955037234788e-05,
+      "loss": 7.5181,
+      "step": 3550
+    },
+    {
+      "epoch": 0.005002133986233594,
+      "grad_norm": 1.2670505046844482,
+      "learning_rate": 1.500210762961922e-05,
+      "loss": 7.6094,
+      "step": 3560
+    },
+    {
+      "epoch": 0.005016184924397171,
+      "grad_norm": 0.8584160804748535,
+      "learning_rate": 1.5044260222003652e-05,
+      "loss": 7.5955,
+      "step": 3570
+    },
+    {
+      "epoch": 0.005030235862560749,
+      "grad_norm": 1.0112240314483643,
+      "learning_rate": 1.5086412814388085e-05,
+      "loss": 7.6168,
+      "step": 3580
+    },
+    {
+      "epoch": 0.005044286800724326,
+      "grad_norm": 4.073171138763428,
+      "learning_rate": 1.5128565406772515e-05,
+      "loss": 7.5669,
+      "step": 3590
+    },
+    {
+      "epoch": 0.0050583377388879035,
+      "grad_norm": 0.9480034112930298,
+      "learning_rate": 1.5170717999156948e-05,
+      "loss": 7.505,
+      "step": 3600
+    },
+    {
+      "epoch": 0.005072388677051481,
+      "grad_norm": 0.95200115442276,
+      "learning_rate": 1.5212870591541379e-05,
+      "loss": 7.6224,
+      "step": 3610
+    },
+    {
+      "epoch": 0.0050864396152150585,
+      "grad_norm": 1.117475986480713,
+      "learning_rate": 1.5255023183925812e-05,
+      "loss": 7.5023,
+      "step": 3620
+    },
+    {
+      "epoch": 0.0051004905533786355,
+      "grad_norm": 0.9365076422691345,
+      "learning_rate": 1.529717577631024e-05,
+      "loss": 7.5295,
+      "step": 3630
+    },
+    {
+      "epoch": 0.005114541491542213,
+      "grad_norm": 0.9868768453598022,
+      "learning_rate": 1.5339328368694676e-05,
+      "loss": 7.5347,
+      "step": 3640
+    },
+    {
+      "epoch": 0.005128592429705791,
+      "grad_norm": 0.860265851020813,
+      "learning_rate": 1.5381480961079103e-05,
+      "loss": 7.5199,
+      "step": 3650
+    },
+    {
+      "epoch": 0.005142643367869368,
+      "grad_norm": 0.9999997615814209,
+      "learning_rate": 1.5423633553463538e-05,
+      "loss": 7.5793,
+      "step": 3660
+    },
+    {
+      "epoch": 0.005156694306032946,
+      "grad_norm": 0.8633819222450256,
+      "learning_rate": 1.546578614584797e-05,
+      "loss": 7.6475,
+      "step": 3670
+    },
+    {
+      "epoch": 0.005170745244196523,
+      "grad_norm": 0.902273416519165,
+      "learning_rate": 1.55079387382324e-05,
+      "loss": 7.5576,
+      "step": 3680
+    },
+    {
+      "epoch": 0.005184796182360101,
+      "grad_norm": 1.04349946975708,
+      "learning_rate": 1.555009133061683e-05,
+      "loss": 7.5067,
+      "step": 3690
+    },
+    {
+      "epoch": 0.005198847120523678,
+      "grad_norm": 0.8825836181640625,
+      "learning_rate": 1.5592243923001263e-05,
+      "loss": 7.5545,
+      "step": 3700
+    },
+    {
+      "epoch": 0.005212898058687256,
+      "grad_norm": 0.853202760219574,
+      "learning_rate": 1.5634396515385694e-05,
+      "loss": 7.615,
+      "step": 3710
+    },
+    {
+      "epoch": 0.005226948996850833,
+      "grad_norm": 1.0415570735931396,
+      "learning_rate": 1.567654910777013e-05,
+      "loss": 7.5865,
+      "step": 3720
+    },
+    {
+      "epoch": 0.005240999935014411,
+      "grad_norm": 0.943265974521637,
+      "learning_rate": 1.5718701700154556e-05,
+      "loss": 7.5048,
+      "step": 3730
+    },
+    {
+      "epoch": 0.005255050873177988,
+      "grad_norm": 1.0362399816513062,
+      "learning_rate": 1.576085429253899e-05,
+      "loss": 7.5796,
+      "step": 3740
+    },
+    {
+      "epoch": 0.005269101811341566,
+      "grad_norm": 1.0052247047424316,
+      "learning_rate": 1.5803006884923422e-05,
+      "loss": 7.4803,
+      "step": 3750
+    },
+    {
+      "epoch": 0.005283152749505144,
+      "grad_norm": 1.0194339752197266,
+      "learning_rate": 1.5845159477307853e-05,
+      "loss": 7.5259,
+      "step": 3760
+    },
+    {
+      "epoch": 0.005297203687668721,
+      "grad_norm": 1.3323463201522827,
+      "learning_rate": 1.5887312069692284e-05,
+      "loss": 7.5403,
+      "step": 3770
+    },
+    {
+      "epoch": 0.005311254625832299,
+      "grad_norm": 0.9728202223777771,
+      "learning_rate": 1.5929464662076715e-05,
+      "loss": 7.526,
+      "step": 3780
+    },
+    {
+      "epoch": 0.005325305563995876,
+      "grad_norm": 0.979168176651001,
+      "learning_rate": 1.5971617254461147e-05,
+      "loss": 7.5024,
+      "step": 3790
+    },
+    {
+      "epoch": 0.005339356502159454,
+      "grad_norm": 1.0642094612121582,
+      "learning_rate": 1.601376984684558e-05,
+      "loss": 7.548,
+      "step": 3800
+    },
+    {
+      "epoch": 0.005353407440323031,
+      "grad_norm": 0.8981004953384399,
+      "learning_rate": 1.605592243923001e-05,
+      "loss": 7.5229,
+      "step": 3810
+    },
+    {
+      "epoch": 0.005367458378486609,
+      "grad_norm": 0.9074918031692505,
+      "learning_rate": 1.6098075031614443e-05,
+      "loss": 7.5607,
+      "step": 3820
+    },
+    {
+      "epoch": 0.005381509316650186,
+      "grad_norm": 0.9015301465988159,
+      "learning_rate": 1.6140227623998875e-05,
+      "loss": 7.5874,
+      "step": 3830
+    },
+    {
+      "epoch": 0.005395560254813764,
+      "grad_norm": 0.8846139907836914,
+      "learning_rate": 1.6182380216383306e-05,
+      "loss": 7.4998,
+      "step": 3840
+    },
+    {
+      "epoch": 0.0054096111929773415,
+      "grad_norm": 1.1710816621780396,
+      "learning_rate": 1.622453280876774e-05,
+      "loss": 7.4512,
+      "step": 3850
+    },
+    {
+      "epoch": 0.0054236621311409185,
+      "grad_norm": 1.1861531734466553,
+      "learning_rate": 1.6266685401152168e-05,
+      "loss": 7.4586,
+      "step": 3860
+    },
+    {
+      "epoch": 0.0054377130693044964,
+      "grad_norm": 1.260762095451355,
+      "learning_rate": 1.6308837993536603e-05,
+      "loss": 7.4841,
+      "step": 3870
+    },
+    {
+      "epoch": 0.0054517640074680735,
+      "grad_norm": 0.8817400932312012,
+      "learning_rate": 1.6350990585921034e-05,
+      "loss": 7.4975,
+      "step": 3880
+    },
+    {
+      "epoch": 0.005465814945631651,
+      "grad_norm": 1.0645872354507446,
+      "learning_rate": 1.6393143178305465e-05,
+      "loss": 7.5029,
+      "step": 3890
+    },
+    {
+      "epoch": 0.005479865883795228,
+      "grad_norm": 0.9742002487182617,
+      "learning_rate": 1.6435295770689896e-05,
+      "loss": 7.482,
+      "step": 3900
+    },
+    {
+      "epoch": 0.005493916821958806,
+      "grad_norm": 0.9298391342163086,
+      "learning_rate": 1.6477448363074327e-05,
+      "loss": 7.5051,
+      "step": 3910
+    },
+    {
+      "epoch": 0.005507967760122383,
+      "grad_norm": 0.9993727207183838,
+      "learning_rate": 1.651960095545876e-05,
+      "loss": 7.4972,
+      "step": 3920
+    },
+    {
+      "epoch": 0.005522018698285961,
+      "grad_norm": 1.3055437803268433,
+      "learning_rate": 1.6561753547843193e-05,
+      "loss": 7.4705,
+      "step": 3930
+    },
+    {
+      "epoch": 0.005536069636449538,
+      "grad_norm": 1.0684734582901,
+      "learning_rate": 1.660390614022762e-05,
+      "loss": 7.45,
+      "step": 3940
+    },
+    {
+      "epoch": 0.005550120574613116,
+      "grad_norm": 1.1004630327224731,
+      "learning_rate": 1.6646058732612055e-05,
+      "loss": 7.4706,
+      "step": 3950
+    },
+    {
+      "epoch": 0.005564171512776694,
+      "grad_norm": 1.3142606019973755,
+      "learning_rate": 1.6688211324996487e-05,
+      "loss": 7.4427,
+      "step": 3960
+    },
+    {
+      "epoch": 0.005578222450940271,
+      "grad_norm": 0.9538849592208862,
+      "learning_rate": 1.6730363917380918e-05,
+      "loss": 7.4318,
+      "step": 3970
+    },
+    {
+      "epoch": 0.005592273389103849,
+      "grad_norm": 1.0903221368789673,
+      "learning_rate": 1.677251650976535e-05,
+      "loss": 7.5066,
+      "step": 3980
+    },
+    {
+      "epoch": 0.005606324327267426,
+      "grad_norm": 1.7417656183242798,
+      "learning_rate": 1.681466910214978e-05,
+      "loss": 7.5136,
+      "step": 3990
+    },
+    {
+      "epoch": 0.005620375265431004,
+      "grad_norm": 1.4083300828933716,
+      "learning_rate": 1.685682169453421e-05,
+      "loss": 7.5295,
+      "step": 4000
+    },
+    {
+      "epoch": 0.005634426203594581,
+      "grad_norm": 0.9326221942901611,
+      "learning_rate": 1.6898974286918646e-05,
+      "loss": 7.4544,
+      "step": 4010
+    },
+    {
+      "epoch": 0.005648477141758159,
+      "grad_norm": 0.936113715171814,
+      "learning_rate": 1.6941126879303074e-05,
+      "loss": 7.4514,
+      "step": 4020
+    },
+    {
+      "epoch": 0.005662528079921736,
+      "grad_norm": 0.8883742690086365,
+      "learning_rate": 1.6983279471687508e-05,
+      "loss": 7.4411,
+      "step": 4030
+    },
+    {
+      "epoch": 0.005676579018085314,
+      "grad_norm": 1.0120148658752441,
+      "learning_rate": 1.702543206407194e-05,
+      "loss": 7.4305,
+      "step": 4040
+    },
+    {
+      "epoch": 0.005690629956248892,
+      "grad_norm": 1.0194205045700073,
+      "learning_rate": 1.706758465645637e-05,
+      "loss": 7.4132,
+      "step": 4050
+    },
+    {
+      "epoch": 0.005704680894412469,
+      "grad_norm": 0.8495796918869019,
+      "learning_rate": 1.71097372488408e-05,
+      "loss": 7.4617,
+      "step": 4060
+    },
+    {
+      "epoch": 0.005718731832576047,
+      "grad_norm": 1.0692280530929565,
+      "learning_rate": 1.7151889841225233e-05,
+      "loss": 7.4526,
+      "step": 4070
+    },
+    {
+      "epoch": 0.005732782770739624,
+      "grad_norm": 0.9888425469398499,
+      "learning_rate": 1.7194042433609667e-05,
+      "loss": 7.4095,
+      "step": 4080
+    },
+    {
+      "epoch": 0.005746833708903202,
+      "grad_norm": 1.1134666204452515,
+      "learning_rate": 1.72361950259941e-05,
+      "loss": 7.4779,
+      "step": 4090
+    },
+    {
+      "epoch": 0.005760884647066779,
+      "grad_norm": 0.9555155634880066,
+      "learning_rate": 1.727834761837853e-05,
+      "loss": 7.4583,
+      "step": 4100
+    },
+    {
+      "epoch": 0.0057749355852303565,
+      "grad_norm": 1.1139003038406372,
+      "learning_rate": 1.732050021076296e-05,
+      "loss": 7.3599,
+      "step": 4110
+    },
+    {
+      "epoch": 0.0057889865233939335,
+      "grad_norm": 0.9886006116867065,
+      "learning_rate": 1.7362652803147392e-05,
+      "loss": 7.4064,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0058030374615575114,
+      "grad_norm": 1.1849690675735474,
+      "learning_rate": 1.7404805395531823e-05,
+      "loss": 7.4462,
+      "step": 4130
+    },
+    {
+      "epoch": 0.0058170883997210885,
+      "grad_norm": 1.1874699592590332,
+      "learning_rate": 1.7446957987916258e-05,
+      "loss": 7.4456,
+      "step": 4140
+    },
+    {
+      "epoch": 0.005831139337884666,
+      "grad_norm": 1.0019316673278809,
+      "learning_rate": 1.7489110580300686e-05,
+      "loss": 7.4241,
+      "step": 4150
+    },
+    {
+      "epoch": 0.005845190276048244,
+      "grad_norm": 1.0301438570022583,
+      "learning_rate": 1.753126317268512e-05,
+      "loss": 7.3951,
+      "step": 4160
+    },
+    {
+      "epoch": 0.005859241214211821,
+      "grad_norm": 1.0591559410095215,
+      "learning_rate": 1.757341576506955e-05,
+      "loss": 7.4101,
+      "step": 4170
+    },
+    {
+      "epoch": 0.005873292152375399,
+      "grad_norm": 1.3416529893875122,
+      "learning_rate": 1.7615568357453982e-05,
+      "loss": 7.4434,
+      "step": 4180
+    },
+    {
+      "epoch": 0.005887343090538976,
+      "grad_norm": 1.153102159500122,
+      "learning_rate": 1.7657720949838414e-05,
+      "loss": 7.4831,
+      "step": 4190
+    },
+    {
+      "epoch": 0.005901394028702554,
+      "grad_norm": 1.0388206243515015,
+      "learning_rate": 1.7699873542222845e-05,
+      "loss": 7.4446,
+      "step": 4200
+    },
+    {
+      "epoch": 0.005915444966866131,
+      "grad_norm": 0.9760084748268127,
+      "learning_rate": 1.7742026134607276e-05,
+      "loss": 7.4868,
+      "step": 4210
+    },
+    {
+      "epoch": 0.005929495905029709,
+      "grad_norm": 1.2405914068222046,
+      "learning_rate": 1.778417872699171e-05,
+      "loss": 7.429,
+      "step": 4220
+    },
+    {
+      "epoch": 0.005943546843193286,
+      "grad_norm": 0.9897080659866333,
+      "learning_rate": 1.7826331319376138e-05,
+      "loss": 7.3724,
+      "step": 4230
+    },
+    {
+      "epoch": 0.005957597781356864,
+      "grad_norm": 0.85589200258255,
+      "learning_rate": 1.7868483911760573e-05,
+      "loss": 7.451,
+      "step": 4240
+    },
+    {
+      "epoch": 0.005971648719520442,
+      "grad_norm": 1.0381278991699219,
+      "learning_rate": 1.7910636504145004e-05,
+      "loss": 7.3856,
+      "step": 4250
+    },
+    {
+      "epoch": 0.005985699657684019,
+      "grad_norm": 1.2100075483322144,
+      "learning_rate": 1.7952789096529435e-05,
+      "loss": 7.4264,
+      "step": 4260
+    },
+    {
+      "epoch": 0.005999750595847597,
+      "grad_norm": 0.9492905139923096,
+      "learning_rate": 1.7994941688913866e-05,
+      "loss": 7.4114,
+      "step": 4270
+    },
+    {
+      "epoch": 0.006013801534011174,
+      "grad_norm": 0.92913419008255,
+      "learning_rate": 1.8037094281298297e-05,
+      "loss": 7.4022,
+      "step": 4280
+    },
+    {
+      "epoch": 0.006027852472174752,
+      "grad_norm": 1.1783087253570557,
+      "learning_rate": 1.807924687368273e-05,
+      "loss": 7.3281,
+      "step": 4290
+    },
+    {
+      "epoch": 0.006041903410338329,
+      "grad_norm": 1.008894681930542,
+      "learning_rate": 1.8121399466067163e-05,
+      "loss": 7.39,
+      "step": 4300
+    },
+    {
+      "epoch": 0.006055954348501907,
+      "grad_norm": 0.9797470569610596,
+      "learning_rate": 1.816355205845159e-05,
+      "loss": 7.4415,
+      "step": 4310
+    },
+    {
+      "epoch": 0.006070005286665484,
+      "grad_norm": 1.3586163520812988,
+      "learning_rate": 1.8205704650836025e-05,
+      "loss": 7.3472,
+      "step": 4320
+    },
+    {
+      "epoch": 0.006084056224829062,
+      "grad_norm": 1.0115594863891602,
+      "learning_rate": 1.8247857243220457e-05,
+      "loss": 7.4058,
+      "step": 4330
+    },
+    {
+      "epoch": 0.006098107162992639,
+      "grad_norm": 1.0275291204452515,
+      "learning_rate": 1.8290009835604888e-05,
+      "loss": 7.2664,
+      "step": 4340
+    },
+    {
+      "epoch": 0.0061121581011562166,
+      "grad_norm": 0.9744628667831421,
+      "learning_rate": 1.833216242798932e-05,
+      "loss": 7.3819,
+      "step": 4350
+    },
+    {
+      "epoch": 0.0061262090393197945,
+      "grad_norm": 1.1602356433868408,
+      "learning_rate": 1.837431502037375e-05,
+      "loss": 7.3576,
+      "step": 4360
+    },
+    {
+      "epoch": 0.0061402599774833715,
+      "grad_norm": 1.18546462059021,
+      "learning_rate": 1.8416467612758185e-05,
+      "loss": 7.3605,
+      "step": 4370
+    },
+    {
+      "epoch": 0.006154310915646949,
+      "grad_norm": 0.9172095060348511,
+      "learning_rate": 1.8458620205142616e-05,
+      "loss": 7.3156,
+      "step": 4380
+    },
+    {
+      "epoch": 0.006168361853810526,
+      "grad_norm": 1.1265218257904053,
+      "learning_rate": 1.8500772797527047e-05,
+      "loss": 7.3078,
+      "step": 4390
+    },
+    {
+      "epoch": 0.006182412791974104,
+      "grad_norm": 0.9350690841674805,
+      "learning_rate": 1.8542925389911478e-05,
+      "loss": 7.3547,
+      "step": 4400
+    },
+    {
+      "epoch": 0.006196463730137681,
+      "grad_norm": 0.9449189901351929,
+      "learning_rate": 1.858507798229591e-05,
+      "loss": 7.3303,
+      "step": 4410
+    },
+    {
+      "epoch": 0.006210514668301259,
+      "grad_norm": 1.1331335306167603,
+      "learning_rate": 1.862723057468034e-05,
+      "loss": 7.3691,
+      "step": 4420
+    },
+    {
+      "epoch": 0.006224565606464836,
+      "grad_norm": 1.1100194454193115,
+      "learning_rate": 1.8669383167064772e-05,
+      "loss": 7.3479,
+      "step": 4430
+    },
+    {
+      "epoch": 0.006238616544628414,
+      "grad_norm": 1.2014113664627075,
+      "learning_rate": 1.8711535759449203e-05,
+      "loss": 7.2566,
+      "step": 4440
+    },
+    {
+      "epoch": 0.006252667482791992,
+      "grad_norm": 1.2027499675750732,
+      "learning_rate": 1.8753688351833634e-05,
+      "loss": 7.3306,
+      "step": 4450
+    },
+    {
+      "epoch": 0.006266718420955569,
+      "grad_norm": 1.2113648653030396,
+      "learning_rate": 1.879584094421807e-05,
+      "loss": 7.2312,
+      "step": 4460
+    },
+    {
+      "epoch": 0.006280769359119147,
+      "grad_norm": 1.14470636844635,
+      "learning_rate": 1.88379935366025e-05,
+      "loss": 7.3603,
+      "step": 4470
+    },
+    {
+      "epoch": 0.006294820297282724,
+      "grad_norm": 1.0293289422988892,
+      "learning_rate": 1.888014612898693e-05,
+      "loss": 7.3057,
+      "step": 4480
+    },
+    {
+      "epoch": 0.006308871235446302,
+      "grad_norm": 1.1022226810455322,
+      "learning_rate": 1.8922298721371362e-05,
+      "loss": 7.2761,
+      "step": 4490
+    },
+    {
+      "epoch": 0.006322922173609879,
+      "grad_norm": 1.1800730228424072,
+      "learning_rate": 1.8964451313755793e-05,
+      "loss": 7.2929,
+      "step": 4500
+    },
+    {
+      "epoch": 0.006336973111773457,
+      "grad_norm": 0.8201963901519775,
+      "learning_rate": 1.9006603906140228e-05,
+      "loss": 7.3127,
+      "step": 4510
+    },
+    {
+      "epoch": 0.006351024049937034,
+      "grad_norm": 1.0580271482467651,
+      "learning_rate": 1.904875649852466e-05,
+      "loss": 7.2965,
+      "step": 4520
+    },
+    {
+      "epoch": 0.006365074988100612,
+      "grad_norm": 1.135973572731018,
+      "learning_rate": 1.9090909090909087e-05,
+      "loss": 7.3115,
+      "step": 4530
+    },
+    {
+      "epoch": 0.006379125926264189,
+      "grad_norm": 1.0042225122451782,
+      "learning_rate": 1.913306168329352e-05,
+      "loss": 7.326,
+      "step": 4540
+    },
+    {
+      "epoch": 0.006393176864427767,
+      "grad_norm": 1.073927879333496,
+      "learning_rate": 1.9175214275677952e-05,
+      "loss": 7.2742,
+      "step": 4550
+    },
+    {
+      "epoch": 0.006407227802591345,
+      "grad_norm": 1.384734034538269,
+      "learning_rate": 1.9217366868062384e-05,
+      "loss": 7.2784,
+      "step": 4560
+    },
+    {
+      "epoch": 0.006421278740754922,
+      "grad_norm": 1.0155476331710815,
+      "learning_rate": 1.9259519460446818e-05,
+      "loss": 7.2515,
+      "step": 4570
+    },
+    {
+      "epoch": 0.0064353296789185,
+      "grad_norm": 1.076328992843628,
+      "learning_rate": 1.9301672052831246e-05,
+      "loss": 7.3195,
+      "step": 4580
+    },
+    {
+      "epoch": 0.006449380617082077,
+      "grad_norm": 1.0640712976455688,
+      "learning_rate": 1.934382464521568e-05,
+      "loss": 7.2912,
+      "step": 4590
+    },
+    {
+      "epoch": 0.0064634315552456545,
+      "grad_norm": 0.9461340308189392,
+      "learning_rate": 1.9385977237600112e-05,
+      "loss": 7.3366,
+      "step": 4600
+    },
+    {
+      "epoch": 0.0064774824934092316,
+      "grad_norm": 0.9778033494949341,
+      "learning_rate": 1.9428129829984543e-05,
+      "loss": 7.2975,
+      "step": 4610
+    },
+    {
+      "epoch": 0.0064915334315728095,
+      "grad_norm": 1.0978727340698242,
+      "learning_rate": 1.9470282422368974e-05,
+      "loss": 7.271,
+      "step": 4620
+    },
+    {
+      "epoch": 0.0065055843697363865,
+      "grad_norm": 1.1852045059204102,
+      "learning_rate": 1.9512435014753405e-05,
+      "loss": 7.2987,
+      "step": 4630
+    },
+    {
+      "epoch": 0.006519635307899964,
+      "grad_norm": 1.0794644355773926,
+      "learning_rate": 1.9554587607137836e-05,
+      "loss": 7.2204,
+      "step": 4640
+    },
+    {
+      "epoch": 0.006533686246063541,
+      "grad_norm": 1.1216827630996704,
+      "learning_rate": 1.959674019952227e-05,
+      "loss": 7.2575,
+      "step": 4650
+    },
+    {
+      "epoch": 0.006547737184227119,
+      "grad_norm": 0.9647300839424133,
+      "learning_rate": 1.96388927919067e-05,
+      "loss": 7.3514,
+      "step": 4660
+    },
+    {
+      "epoch": 0.006561788122390697,
+      "grad_norm": 1.1785202026367188,
+      "learning_rate": 1.9681045384291133e-05,
+      "loss": 7.2722,
+      "step": 4670
+    },
+    {
+      "epoch": 0.006575839060554274,
+      "grad_norm": 1.2579607963562012,
+      "learning_rate": 1.9723197976675564e-05,
+      "loss": 7.2684,
+      "step": 4680
+    },
+    {
+      "epoch": 0.006589889998717852,
+      "grad_norm": 1.069290280342102,
+      "learning_rate": 1.9765350569059996e-05,
+      "loss": 7.2563,
+      "step": 4690
+    },
+    {
+      "epoch": 0.006603940936881429,
+      "grad_norm": 1.1216323375701904,
+      "learning_rate": 1.9807503161444427e-05,
+      "loss": 7.1934,
+      "step": 4700
+    },
+    {
+      "epoch": 0.006617991875045007,
+      "grad_norm": 1.0338348150253296,
+      "learning_rate": 1.9849655753828858e-05,
+      "loss": 7.2595,
+      "step": 4710
+    },
+    {
+      "epoch": 0.006632042813208584,
+      "grad_norm": 1.0657542943954468,
+      "learning_rate": 1.989180834621329e-05,
+      "loss": 7.2447,
+      "step": 4720
+    },
+    {
+      "epoch": 0.006646093751372162,
+      "grad_norm": 1.0090150833129883,
+      "learning_rate": 1.9933960938597724e-05,
+      "loss": 7.2676,
+      "step": 4730
+    },
+    {
+      "epoch": 0.006660144689535739,
+      "grad_norm": 1.244844913482666,
+      "learning_rate": 1.997611353098215e-05,
+      "loss": 7.2931,
+      "step": 4740
+    },
+    {
+      "epoch": 0.006674195627699317,
+      "grad_norm": 1.4052555561065674,
+      "learning_rate": 2.0018266123366586e-05,
+      "loss": 7.3218,
+      "step": 4750
+    },
+    {
+      "epoch": 0.006688246565862895,
+      "grad_norm": 1.0389999151229858,
+      "learning_rate": 2.0060418715751017e-05,
+      "loss": 7.3679,
+      "step": 4760
+    },
+    {
+      "epoch": 0.006702297504026472,
+      "grad_norm": 1.2123712301254272,
+      "learning_rate": 2.010257130813545e-05,
+      "loss": 7.3174,
+      "step": 4770
+    },
+    {
+      "epoch": 0.00671634844219005,
+      "grad_norm": 1.31826651096344,
+      "learning_rate": 2.014472390051988e-05,
+      "loss": 7.2676,
+      "step": 4780
+    },
+    {
+      "epoch": 0.006730399380353627,
+      "grad_norm": 1.0967237949371338,
+      "learning_rate": 2.018687649290431e-05,
+      "loss": 7.2652,
+      "step": 4790
+    },
+    {
+      "epoch": 0.006744450318517205,
+      "grad_norm": 1.1359235048294067,
+      "learning_rate": 2.0229029085288742e-05,
+      "loss": 7.2868,
+      "step": 4800
+    },
+    {
+      "epoch": 0.006758501256680782,
+      "grad_norm": 1.052154779434204,
+      "learning_rate": 2.0271181677673176e-05,
+      "loss": 7.288,
+      "step": 4810
+    },
+    {
+      "epoch": 0.00677255219484436,
+      "grad_norm": 1.0705679655075073,
+      "learning_rate": 2.0313334270057608e-05,
+      "loss": 7.2405,
+      "step": 4820
+    },
+    {
+      "epoch": 0.006786603133007937,
+      "grad_norm": 0.9692743420600891,
+      "learning_rate": 2.035548686244204e-05,
+      "loss": 7.1738,
+      "step": 4830
+    },
+    {
+      "epoch": 0.006800654071171515,
+      "grad_norm": 1.2184343338012695,
+      "learning_rate": 2.039763945482647e-05,
+      "loss": 7.2294,
+      "step": 4840
+    },
+    {
+      "epoch": 0.006814705009335092,
+      "grad_norm": 1.4309422969818115,
+      "learning_rate": 2.04397920472109e-05,
+      "loss": 7.2412,
+      "step": 4850
+    },
+    {
+      "epoch": 0.0068287559474986695,
+      "grad_norm": 1.538222312927246,
+      "learning_rate": 2.0481944639595336e-05,
+      "loss": 7.1677,
+      "step": 4860
+    },
+    {
+      "epoch": 0.006842806885662247,
+      "grad_norm": 1.1325163841247559,
+      "learning_rate": 2.0524097231979763e-05,
+      "loss": 7.2538,
+      "step": 4870
+    },
+    {
+      "epoch": 0.0068568578238258245,
+      "grad_norm": 1.0051361322402954,
+      "learning_rate": 2.0566249824364198e-05,
+      "loss": 7.2056,
+      "step": 4880
+    },
+    {
+      "epoch": 0.006870908761989402,
+      "grad_norm": 1.08148193359375,
+      "learning_rate": 2.060840241674863e-05,
+      "loss": 7.1734,
+      "step": 4890
+    },
+    {
+      "epoch": 0.006884959700152979,
+      "grad_norm": 1.2696799039840698,
+      "learning_rate": 2.065055500913306e-05,
+      "loss": 7.2288,
+      "step": 4900
+    },
+    {
+      "epoch": 0.006899010638316557,
+      "grad_norm": 1.0847587585449219,
+      "learning_rate": 2.069270760151749e-05,
+      "loss": 7.2802,
+      "step": 4910
+    },
+    {
+      "epoch": 0.006913061576480134,
+      "grad_norm": 1.2094314098358154,
+      "learning_rate": 2.0734860193901923e-05,
+      "loss": 7.2064,
+      "step": 4920
+    },
+    {
+      "epoch": 0.006927112514643712,
+      "grad_norm": 1.1414620876312256,
+      "learning_rate": 2.0777012786286354e-05,
+      "loss": 7.1366,
+      "step": 4930
+    },
+    {
+      "epoch": 0.006941163452807289,
+      "grad_norm": 1.2508331537246704,
+      "learning_rate": 2.081916537867079e-05,
+      "loss": 7.2477,
+      "step": 4940
+    },
+    {
+      "epoch": 0.006955214390970867,
+      "grad_norm": 1.1051875352859497,
+      "learning_rate": 2.0861317971055216e-05,
+      "loss": 7.2439,
+      "step": 4950
+    },
+    {
+      "epoch": 0.006969265329134445,
+      "grad_norm": 0.9596846103668213,
+      "learning_rate": 2.090347056343965e-05,
+      "loss": 7.1859,
+      "step": 4960
+    },
+    {
+      "epoch": 0.006983316267298022,
+      "grad_norm": 1.0876295566558838,
+      "learning_rate": 2.0945623155824082e-05,
+      "loss": 7.2189,
+      "step": 4970
+    },
+    {
+      "epoch": 0.0069973672054616,
+      "grad_norm": 1.1610974073410034,
+      "learning_rate": 2.0987775748208513e-05,
+      "loss": 7.1999,
+      "step": 4980
+    },
+    {
+      "epoch": 0.007011418143625177,
+      "grad_norm": 1.023998498916626,
+      "learning_rate": 2.1029928340592944e-05,
+      "loss": 7.1728,
+      "step": 4990
+    },
+    {
+      "epoch": 0.007025469081788755,
+      "grad_norm": 1.176242470741272,
+      "learning_rate": 2.1072080932977375e-05,
+      "loss": 7.162,
+      "step": 5000
+    },
+    {
+      "epoch": 0.007039520019952332,
+      "grad_norm": 1.1266804933547974,
+      "learning_rate": 2.1114233525361806e-05,
+      "loss": 7.2674,
+      "step": 5010
+    },
+    {
+      "epoch": 0.00705357095811591,
+      "grad_norm": 1.0107471942901611,
+      "learning_rate": 2.115638611774624e-05,
+      "loss": 7.168,
+      "step": 5020
+    },
+    {
+      "epoch": 0.007067621896279487,
+      "grad_norm": 0.9729942679405212,
+      "learning_rate": 2.119853871013067e-05,
+      "loss": 7.2244,
+      "step": 5030
+    },
+    {
+      "epoch": 0.007081672834443065,
+      "grad_norm": 1.117859125137329,
+      "learning_rate": 2.1240691302515103e-05,
+      "loss": 7.1621,
+      "step": 5040
+    },
+    {
+      "epoch": 0.007095723772606642,
+      "grad_norm": 1.206984281539917,
+      "learning_rate": 2.1282843894899535e-05,
+      "loss": 7.2517,
+      "step": 5050
+    },
+    {
+      "epoch": 0.00710977471077022,
+      "grad_norm": 1.356512188911438,
+      "learning_rate": 2.1324996487283966e-05,
+      "loss": 7.2202,
+      "step": 5060
+    },
+    {
+      "epoch": 0.007123825648933798,
+      "grad_norm": 1.0092519521713257,
+      "learning_rate": 2.13671490796684e-05,
+      "loss": 7.2292,
+      "step": 5070
+    },
+    {
+      "epoch": 0.007137876587097375,
+      "grad_norm": 1.4104299545288086,
+      "learning_rate": 2.1409301672052828e-05,
+      "loss": 7.1438,
+      "step": 5080
+    },
+    {
+      "epoch": 0.0071519275252609526,
+      "grad_norm": 1.2122313976287842,
+      "learning_rate": 2.145145426443726e-05,
+      "loss": 7.1562,
+      "step": 5090
+    },
+    {
+      "epoch": 0.00716597846342453,
+      "grad_norm": 1.1602479219436646,
+      "learning_rate": 2.1493606856821694e-05,
+      "loss": 7.166,
+      "step": 5100
+    },
+    {
+      "epoch": 0.0071800294015881075,
+      "grad_norm": 1.1590441465377808,
+      "learning_rate": 2.1535759449206125e-05,
+      "loss": 7.155,
+      "step": 5110
+    },
+    {
+      "epoch": 0.0071940803397516845,
+      "grad_norm": 0.9332759380340576,
+      "learning_rate": 2.1577912041590556e-05,
+      "loss": 7.1189,
+      "step": 5120
+    },
+    {
+      "epoch": 0.007208131277915262,
+      "grad_norm": 1.1661341190338135,
+      "learning_rate": 2.1620064633974987e-05,
+      "loss": 7.1424,
+      "step": 5130
+    },
+    {
+      "epoch": 0.0072221822160788395,
+      "grad_norm": 1.2383019924163818,
+      "learning_rate": 2.166221722635942e-05,
+      "loss": 7.1516,
+      "step": 5140
+    },
+    {
+      "epoch": 0.007236233154242417,
+      "grad_norm": 1.1211152076721191,
+      "learning_rate": 2.1704369818743853e-05,
+      "loss": 7.0884,
+      "step": 5150
+    },
+    {
+      "epoch": 0.007250284092405995,
+      "grad_norm": 1.0935217142105103,
+      "learning_rate": 2.174652241112828e-05,
+      "loss": 7.1914,
+      "step": 5160
+    },
+    {
+      "epoch": 0.007264335030569572,
+      "grad_norm": 1.0398094654083252,
+      "learning_rate": 2.1788675003512712e-05,
+      "loss": 7.1011,
+      "step": 5170
+    },
+    {
+      "epoch": 0.00727838596873315,
+      "grad_norm": 1.0619064569473267,
+      "learning_rate": 2.1830827595897146e-05,
+      "loss": 7.1611,
+      "step": 5180
+    },
+    {
+      "epoch": 0.007292436906896727,
+      "grad_norm": 1.0957516431808472,
+      "learning_rate": 2.1872980188281578e-05,
+      "loss": 7.0981,
+      "step": 5190
+    },
+    {
+      "epoch": 0.007306487845060305,
+      "grad_norm": 1.1831027269363403,
+      "learning_rate": 2.191513278066601e-05,
+      "loss": 7.1843,
+      "step": 5200
+    },
+    {
+      "epoch": 0.007320538783223882,
+      "grad_norm": 1.323491096496582,
+      "learning_rate": 2.195728537305044e-05,
+      "loss": 7.1189,
+      "step": 5210
+    },
+    {
+      "epoch": 0.00733458972138746,
+      "grad_norm": 1.022131085395813,
+      "learning_rate": 2.199943796543487e-05,
+      "loss": 7.1855,
+      "step": 5220
+    },
+    {
+      "epoch": 0.007348640659551037,
+      "grad_norm": 1.081374168395996,
+      "learning_rate": 2.2041590557819306e-05,
+      "loss": 7.1447,
+      "step": 5230
+    },
+    {
+      "epoch": 0.007362691597714615,
+      "grad_norm": 1.0144031047821045,
+      "learning_rate": 2.2083743150203733e-05,
+      "loss": 7.1438,
+      "step": 5240
+    },
+    {
+      "epoch": 0.007376742535878192,
+      "grad_norm": 1.2261079549789429,
+      "learning_rate": 2.2125895742588168e-05,
+      "loss": 7.2035,
+      "step": 5250
+    },
+    {
+      "epoch": 0.00739079347404177,
+      "grad_norm": 1.2673313617706299,
+      "learning_rate": 2.21680483349726e-05,
+      "loss": 7.1767,
+      "step": 5260
+    },
+    {
+      "epoch": 0.007404844412205348,
+      "grad_norm": 1.063643455505371,
+      "learning_rate": 2.221020092735703e-05,
+      "loss": 7.1446,
+      "step": 5270
+    },
+    {
+      "epoch": 0.007418895350368925,
+      "grad_norm": 1.092332124710083,
+      "learning_rate": 2.225235351974146e-05,
+      "loss": 7.132,
+      "step": 5280
+    },
+    {
+      "epoch": 0.007432946288532503,
+      "grad_norm": 1.2160872220993042,
+      "learning_rate": 2.2294506112125893e-05,
+      "loss": 7.2127,
+      "step": 5290
+    },
+    {
+      "epoch": 0.00744699722669608,
+      "grad_norm": 1.1070367097854614,
+      "learning_rate": 2.2336658704510324e-05,
+      "loss": 7.1532,
+      "step": 5300
+    },
+    {
+      "epoch": 0.007461048164859658,
+      "grad_norm": 1.136398196220398,
+      "learning_rate": 2.237881129689476e-05,
+      "loss": 7.1131,
+      "step": 5310
+    },
+    {
+      "epoch": 0.007475099103023235,
+      "grad_norm": 1.0606321096420288,
+      "learning_rate": 2.242096388927919e-05,
+      "loss": 7.0342,
+      "step": 5320
+    },
+    {
+      "epoch": 0.007489150041186813,
+      "grad_norm": 1.138594150543213,
+      "learning_rate": 2.246311648166362e-05,
+      "loss": 7.0824,
+      "step": 5330
+    },
+    {
+      "epoch": 0.00750320097935039,
+      "grad_norm": 1.3768428564071655,
+      "learning_rate": 2.2505269074048052e-05,
+      "loss": 7.1352,
+      "step": 5340
+    },
+    {
+      "epoch": 0.0075172519175139675,
+      "grad_norm": 1.2241921424865723,
+      "learning_rate": 2.2547421666432483e-05,
+      "loss": 7.0765,
+      "step": 5350
+    },
+    {
+      "epoch": 0.0075313028556775454,
+      "grad_norm": 1.0443278551101685,
+      "learning_rate": 2.2589574258816918e-05,
+      "loss": 7.0452,
+      "step": 5360
+    },
+    {
+      "epoch": 0.0075453537938411225,
+      "grad_norm": 1.295654296875,
+      "learning_rate": 2.2631726851201345e-05,
+      "loss": 7.0193,
+      "step": 5370
+    },
+    {
+      "epoch": 0.0075594047320047,
+      "grad_norm": 1.6669979095458984,
+      "learning_rate": 2.2673879443585777e-05,
+      "loss": 7.0958,
+      "step": 5380
+    },
+    {
+      "epoch": 0.007573455670168277,
+      "grad_norm": 1.039218544960022,
+      "learning_rate": 2.271603203597021e-05,
+      "loss": 7.1377,
+      "step": 5390
+    },
+    {
+      "epoch": 0.007587506608331855,
+      "grad_norm": 1.0409983396530151,
+      "learning_rate": 2.2758184628354642e-05,
+      "loss": 7.0415,
+      "step": 5400
+    },
+    {
+      "epoch": 0.007601557546495432,
+      "grad_norm": 1.0646721124649048,
+      "learning_rate": 2.2800337220739073e-05,
+      "loss": 7.0761,
+      "step": 5410
+    },
+    {
+      "epoch": 0.00761560848465901,
+      "grad_norm": 1.143831729888916,
+      "learning_rate": 2.2842489813123505e-05,
+      "loss": 7.083,
+      "step": 5420
+    },
+    {
+      "epoch": 0.007629659422822587,
+      "grad_norm": 1.0598105192184448,
+      "learning_rate": 2.2884642405507936e-05,
+      "loss": 7.0908,
+      "step": 5430
+    },
+    {
+      "epoch": 0.007643710360986165,
+      "grad_norm": 1.2153172492980957,
+      "learning_rate": 2.292679499789237e-05,
+      "loss": 7.0801,
+      "step": 5440
+    },
+    {
+      "epoch": 0.007657761299149742,
+      "grad_norm": 1.0428032875061035,
+      "learning_rate": 2.2968947590276798e-05,
+      "loss": 7.1072,
+      "step": 5450
+    },
+    {
+      "epoch": 0.00767181223731332,
+      "grad_norm": 1.0740203857421875,
+      "learning_rate": 2.301110018266123e-05,
+      "loss": 7.101,
+      "step": 5460
+    },
+    {
+      "epoch": 0.007685863175476898,
+      "grad_norm": 1.193075180053711,
+      "learning_rate": 2.3053252775045664e-05,
+      "loss": 7.1157,
+      "step": 5470
+    },
+    {
+      "epoch": 0.007699914113640475,
+      "grad_norm": 1.1059104204177856,
+      "learning_rate": 2.3095405367430095e-05,
+      "loss": 7.1466,
+      "step": 5480
+    },
+    {
+      "epoch": 0.007713965051804053,
+      "grad_norm": 1.121106743812561,
+      "learning_rate": 2.3137557959814526e-05,
+      "loss": 7.1207,
+      "step": 5490
+    },
+    {
+      "epoch": 0.00772801598996763,
+      "grad_norm": 1.1866158246994019,
+      "learning_rate": 2.3179710552198957e-05,
+      "loss": 7.0526,
+      "step": 5500
+    },
+    {
+      "epoch": 0.007742066928131208,
+      "grad_norm": 1.1638514995574951,
+      "learning_rate": 2.322186314458339e-05,
+      "loss": 7.0261,
+      "step": 5510
+    },
+    {
+      "epoch": 0.007756117866294785,
+      "grad_norm": 1.1464370489120483,
+      "learning_rate": 2.3264015736967823e-05,
+      "loss": 7.1148,
+      "step": 5520
+    },
+    {
+      "epoch": 0.007770168804458363,
+      "grad_norm": 1.263282060623169,
+      "learning_rate": 2.330616832935225e-05,
+      "loss": 7.1217,
+      "step": 5530
+    },
+    {
+      "epoch": 0.00778421974262194,
+      "grad_norm": 1.1215978860855103,
+      "learning_rate": 2.3348320921736685e-05,
+      "loss": 7.1119,
+      "step": 5540
+    },
+    {
+      "epoch": 0.007798270680785518,
+      "grad_norm": 1.093409776687622,
+      "learning_rate": 2.3390473514121117e-05,
+      "loss": 7.0892,
+      "step": 5550
+    },
+    {
+      "epoch": 0.007812321618949095,
+      "grad_norm": 1.2007728815078735,
+      "learning_rate": 2.3432626106505548e-05,
+      "loss": 7.051,
+      "step": 5560
+    },
+    {
+      "epoch": 0.007826372557112674,
+      "grad_norm": 1.0814827680587769,
+      "learning_rate": 2.3474778698889982e-05,
+      "loss": 7.1414,
+      "step": 5570
+    },
+    {
+      "epoch": 0.00784042349527625,
+      "grad_norm": 1.2719553709030151,
+      "learning_rate": 2.351693129127441e-05,
+      "loss": 7.0283,
+      "step": 5580
+    },
+    {
+      "epoch": 0.007854474433439828,
+      "grad_norm": 1.3586655855178833,
+      "learning_rate": 2.355908388365884e-05,
+      "loss": 7.0621,
+      "step": 5590
+    },
+    {
+      "epoch": 0.007868525371603406,
+      "grad_norm": 1.0803841352462769,
+      "learning_rate": 2.3601236476043276e-05,
+      "loss": 7.0976,
+      "step": 5600
+    },
+    {
+      "epoch": 0.007882576309766983,
+      "grad_norm": 1.3272629976272583,
+      "learning_rate": 2.3643389068427707e-05,
+      "loss": 7.0699,
+      "step": 5610
+    },
+    {
+      "epoch": 0.00789662724793056,
+      "grad_norm": 1.1906023025512695,
+      "learning_rate": 2.3685541660812138e-05,
+      "loss": 7.1024,
+      "step": 5620
+    },
+    {
+      "epoch": 0.007910678186094137,
+      "grad_norm": 1.0353652238845825,
+      "learning_rate": 2.372769425319657e-05,
+      "loss": 7.1339,
+      "step": 5630
+    },
+    {
+      "epoch": 0.007924729124257715,
+      "grad_norm": 1.1890437602996826,
+      "learning_rate": 2.3769846845581e-05,
+      "loss": 7.1143,
+      "step": 5640
+    },
+    {
+      "epoch": 0.007938780062421293,
+      "grad_norm": 1.1110581159591675,
+      "learning_rate": 2.3811999437965435e-05,
+      "loss": 7.0033,
+      "step": 5650
+    },
+    {
+      "epoch": 0.00795283100058487,
+      "grad_norm": 1.1018701791763306,
+      "learning_rate": 2.3854152030349863e-05,
+      "loss": 7.0789,
+      "step": 5660
+    },
+    {
+      "epoch": 0.007966881938748447,
+      "grad_norm": 1.2645286321640015,
+      "learning_rate": 2.3896304622734294e-05,
+      "loss": 7.0673,
+      "step": 5670
+    },
+    {
+      "epoch": 0.007980932876912025,
+      "grad_norm": 1.2000459432601929,
+      "learning_rate": 2.393845721511873e-05,
+      "loss": 7.0056,
+      "step": 5680
+    },
+    {
+      "epoch": 0.007994983815075603,
+      "grad_norm": 1.5909388065338135,
+      "learning_rate": 2.398060980750316e-05,
+      "loss": 7.0763,
+      "step": 5690
+    },
+    {
+      "epoch": 0.008009034753239181,
+      "grad_norm": 1.1488021612167358,
+      "learning_rate": 2.402276239988759e-05,
+      "loss": 7.0547,
+      "step": 5700
+    },
+    {
+      "epoch": 0.008023085691402757,
+      "grad_norm": 1.6401759386062622,
+      "learning_rate": 2.4064914992272022e-05,
+      "loss": 7.0965,
+      "step": 5710
+    },
+    {
+      "epoch": 0.008037136629566335,
+      "grad_norm": 1.1876939535140991,
+      "learning_rate": 2.4107067584656453e-05,
+      "loss": 7.0763,
+      "step": 5720
+    },
+    {
+      "epoch": 0.008051187567729913,
+      "grad_norm": 1.2102808952331543,
+      "learning_rate": 2.4149220177040888e-05,
+      "loss": 6.9282,
+      "step": 5730
+    },
+    {
+      "epoch": 0.008065238505893491,
+      "grad_norm": 1.1134921312332153,
+      "learning_rate": 2.4191372769425316e-05,
+      "loss": 7.0256,
+      "step": 5740
+    },
+    {
+      "epoch": 0.008079289444057067,
+      "grad_norm": 1.1920967102050781,
+      "learning_rate": 2.4233525361809747e-05,
+      "loss": 7.0314,
+      "step": 5750
+    },
+    {
+      "epoch": 0.008093340382220645,
+      "grad_norm": 1.218900442123413,
+      "learning_rate": 2.427567795419418e-05,
+      "loss": 7.0789,
+      "step": 5760
+    },
+    {
+      "epoch": 0.008107391320384223,
+      "grad_norm": 1.160243034362793,
+      "learning_rate": 2.4317830546578612e-05,
+      "loss": 6.9588,
+      "step": 5770
+    },
+    {
+      "epoch": 0.0081214422585478,
+      "grad_norm": 1.1619845628738403,
+      "learning_rate": 2.4359983138963044e-05,
+      "loss": 7.0157,
+      "step": 5780
+    },
+    {
+      "epoch": 0.008135493196711379,
+      "grad_norm": 1.2614120244979858,
+      "learning_rate": 2.4402135731347475e-05,
+      "loss": 6.9232,
+      "step": 5790
+    },
+    {
+      "epoch": 0.008149544134874955,
+      "grad_norm": 1.1834042072296143,
+      "learning_rate": 2.4444288323731906e-05,
+      "loss": 7.0801,
+      "step": 5800
+    },
+    {
+      "epoch": 0.008163595073038533,
+      "grad_norm": 1.147611141204834,
+      "learning_rate": 2.448644091611634e-05,
+      "loss": 7.0824,
+      "step": 5810
+    },
+    {
+      "epoch": 0.00817764601120211,
+      "grad_norm": 1.121339201927185,
+      "learning_rate": 2.452859350850077e-05,
+      "loss": 6.9956,
+      "step": 5820
+    },
+    {
+      "epoch": 0.008191696949365689,
+      "grad_norm": 1.15691077709198,
+      "learning_rate": 2.45707461008852e-05,
+      "loss": 6.977,
+      "step": 5830
+    },
+    {
+      "epoch": 0.008205747887529265,
+      "grad_norm": 1.2041352987289429,
+      "learning_rate": 2.4612898693269634e-05,
+      "loss": 6.9522,
+      "step": 5840
+    },
+    {
+      "epoch": 0.008219798825692843,
+      "grad_norm": 1.0645039081573486,
+      "learning_rate": 2.4655051285654065e-05,
+      "loss": 7.0521,
+      "step": 5850
+    },
+    {
+      "epoch": 0.00823384976385642,
+      "grad_norm": 1.0398435592651367,
+      "learning_rate": 2.46972038780385e-05,
+      "loss": 6.9993,
+      "step": 5860
+    },
+    {
+      "epoch": 0.008247900702019998,
+      "grad_norm": 1.1055554151535034,
+      "learning_rate": 2.4739356470422927e-05,
+      "loss": 7.0278,
+      "step": 5870
+    },
+    {
+      "epoch": 0.008261951640183576,
+      "grad_norm": 1.3027024269104004,
+      "learning_rate": 2.478150906280736e-05,
+      "loss": 7.0749,
+      "step": 5880
+    },
+    {
+      "epoch": 0.008276002578347152,
+      "grad_norm": 1.2701828479766846,
+      "learning_rate": 2.4823661655191793e-05,
+      "loss": 6.9991,
+      "step": 5890
+    },
+    {
+      "epoch": 0.00829005351651073,
+      "grad_norm": 1.2022544145584106,
+      "learning_rate": 2.4865814247576224e-05,
+      "loss": 6.957,
+      "step": 5900
+    },
+    {
+      "epoch": 0.008304104454674308,
+      "grad_norm": 1.1316558122634888,
+      "learning_rate": 2.4907966839960656e-05,
+      "loss": 6.9934,
+      "step": 5910
+    },
+    {
+      "epoch": 0.008318155392837886,
+      "grad_norm": 1.539913535118103,
+      "learning_rate": 2.4950119432345087e-05,
+      "loss": 7.0327,
+      "step": 5920
+    },
+    {
+      "epoch": 0.008332206331001462,
+      "grad_norm": 1.088806390762329,
+      "learning_rate": 2.4992272024729518e-05,
+      "loss": 6.9959,
+      "step": 5930
+    },
+    {
+      "epoch": 0.00834625726916504,
+      "grad_norm": 1.0958342552185059,
+      "learning_rate": 2.5034424617113952e-05,
+      "loss": 7.0824,
+      "step": 5940
+    },
+    {
+      "epoch": 0.008360308207328618,
+      "grad_norm": 1.241213321685791,
+      "learning_rate": 2.507657720949838e-05,
+      "loss": 7.007,
+      "step": 5950
+    },
+    {
+      "epoch": 0.008374359145492196,
+      "grad_norm": 1.1807730197906494,
+      "learning_rate": 2.511872980188281e-05,
+      "loss": 7.0675,
+      "step": 5960
+    },
+    {
+      "epoch": 0.008388410083655774,
+      "grad_norm": 1.1051346063613892,
+      "learning_rate": 2.5160882394267246e-05,
+      "loss": 7.0387,
+      "step": 5970
+    },
+    {
+      "epoch": 0.00840246102181935,
+      "grad_norm": 1.1446903944015503,
+      "learning_rate": 2.5203034986651677e-05,
+      "loss": 6.9076,
+      "step": 5980
+    },
+    {
+      "epoch": 0.008416511959982928,
+      "grad_norm": 1.0661766529083252,
+      "learning_rate": 2.5245187579036108e-05,
+      "loss": 6.9638,
+      "step": 5990
+    },
+    {
+      "epoch": 0.008430562898146506,
+      "grad_norm": 1.3106430768966675,
+      "learning_rate": 2.528734017142054e-05,
+      "loss": 6.9139,
+      "step": 6000
+    },
+    {
+      "epoch": 0.008444613836310084,
+      "grad_norm": 1.0190626382827759,
+      "learning_rate": 2.532949276380497e-05,
+      "loss": 7.0121,
+      "step": 6010
+    },
+    {
+      "epoch": 0.00845866477447366,
+      "grad_norm": 1.2163209915161133,
+      "learning_rate": 2.5371645356189405e-05,
+      "loss": 6.9522,
+      "step": 6020
+    },
+    {
+      "epoch": 0.008472715712637238,
+      "grad_norm": 1.3020004034042358,
+      "learning_rate": 2.5413797948573833e-05,
+      "loss": 7.0786,
+      "step": 6030
+    },
+    {
+      "epoch": 0.008486766650800816,
+      "grad_norm": 1.3286876678466797,
+      "learning_rate": 2.5455950540958264e-05,
+      "loss": 6.9549,
+      "step": 6040
+    },
+    {
+      "epoch": 0.008500817588964394,
+      "grad_norm": 1.3438900709152222,
+      "learning_rate": 2.54981031333427e-05,
+      "loss": 6.9829,
+      "step": 6050
+    },
+    {
+      "epoch": 0.00851486852712797,
+      "grad_norm": 1.0740140676498413,
+      "learning_rate": 2.554025572572713e-05,
+      "loss": 6.9899,
+      "step": 6060
+    },
+    {
+      "epoch": 0.008528919465291548,
+      "grad_norm": 1.1501201391220093,
+      "learning_rate": 2.5582408318111564e-05,
+      "loss": 7.0018,
+      "step": 6070
+    },
+    {
+      "epoch": 0.008542970403455126,
+      "grad_norm": 1.1902880668640137,
+      "learning_rate": 2.5624560910495992e-05,
+      "loss": 7.0088,
+      "step": 6080
+    },
+    {
+      "epoch": 0.008557021341618704,
+      "grad_norm": 1.0988359451293945,
+      "learning_rate": 2.5666713502880423e-05,
+      "loss": 6.9861,
+      "step": 6090
+    },
+    {
+      "epoch": 0.008571072279782281,
+      "grad_norm": 1.1356360912322998,
+      "learning_rate": 2.5708866095264858e-05,
+      "loss": 7.0025,
+      "step": 6100
+    },
+    {
+      "epoch": 0.008585123217945858,
+      "grad_norm": 1.2396526336669922,
+      "learning_rate": 2.575101868764929e-05,
+      "loss": 6.9691,
+      "step": 6110
+    },
+    {
+      "epoch": 0.008599174156109436,
+      "grad_norm": 1.1919825077056885,
+      "learning_rate": 2.5793171280033717e-05,
+      "loss": 6.968,
+      "step": 6120
+    },
+    {
+      "epoch": 0.008613225094273013,
+      "grad_norm": 1.2087693214416504,
+      "learning_rate": 2.583532387241815e-05,
+      "loss": 6.9081,
+      "step": 6130
+    },
+    {
+      "epoch": 0.008627276032436591,
+      "grad_norm": 1.0366860628128052,
+      "learning_rate": 2.5877476464802583e-05,
+      "loss": 6.9781,
+      "step": 6140
+    },
+    {
+      "epoch": 0.008641326970600167,
+      "grad_norm": 1.0888359546661377,
+      "learning_rate": 2.5919629057187017e-05,
+      "loss": 7.0216,
+      "step": 6150
+    },
+    {
+      "epoch": 0.008655377908763745,
+      "grad_norm": 1.055550217628479,
+      "learning_rate": 2.5961781649571445e-05,
+      "loss": 6.9755,
+      "step": 6160
+    },
+    {
+      "epoch": 0.008669428846927323,
+      "grad_norm": 1.300152063369751,
+      "learning_rate": 2.6003934241955876e-05,
+      "loss": 6.9486,
+      "step": 6170
+    },
+    {
+      "epoch": 0.008683479785090901,
+      "grad_norm": 1.2063759565353394,
+      "learning_rate": 2.604608683434031e-05,
+      "loss": 6.9448,
+      "step": 6180
+    },
+    {
+      "epoch": 0.008697530723254479,
+      "grad_norm": 1.3412996530532837,
+      "learning_rate": 2.6088239426724742e-05,
+      "loss": 6.9007,
+      "step": 6190
+    },
+    {
+      "epoch": 0.008711581661418055,
+      "grad_norm": 1.2316458225250244,
+      "learning_rate": 2.613039201910917e-05,
+      "loss": 6.842,
+      "step": 6200
+    },
+    {
+      "epoch": 0.008725632599581633,
+      "grad_norm": 1.3386321067810059,
+      "learning_rate": 2.6172544611493604e-05,
+      "loss": 6.9621,
+      "step": 6210
+    },
+    {
+      "epoch": 0.008739683537745211,
+      "grad_norm": 1.2841423749923706,
+      "learning_rate": 2.6214697203878035e-05,
+      "loss": 6.9877,
+      "step": 6220
+    },
+    {
+      "epoch": 0.008753734475908789,
+      "grad_norm": 1.238442301750183,
+      "learning_rate": 2.625684979626247e-05,
+      "loss": 6.9483,
+      "step": 6230
+    },
+    {
+      "epoch": 0.008767785414072365,
+      "grad_norm": 1.2642581462860107,
+      "learning_rate": 2.6299002388646898e-05,
+      "loss": 6.974,
+      "step": 6240
+    },
+    {
+      "epoch": 0.008781836352235943,
+      "grad_norm": 1.0157082080841064,
+      "learning_rate": 2.634115498103133e-05,
+      "loss": 6.9708,
+      "step": 6250
+    },
+    {
+      "epoch": 0.008795887290399521,
+      "grad_norm": 1.1940841674804688,
+      "learning_rate": 2.6383307573415763e-05,
+      "loss": 6.8949,
+      "step": 6260
+    },
+    {
+      "epoch": 0.008809938228563099,
+      "grad_norm": 1.1627519130706787,
+      "learning_rate": 2.6425460165800194e-05,
+      "loss": 6.9352,
+      "step": 6270
+    },
+    {
+      "epoch": 0.008823989166726677,
+      "grad_norm": 1.2034648656845093,
+      "learning_rate": 2.6467612758184626e-05,
+      "loss": 6.8588,
+      "step": 6280
+    },
+    {
+      "epoch": 0.008838040104890253,
+      "grad_norm": 1.2993476390838623,
+      "learning_rate": 2.6509765350569057e-05,
+      "loss": 6.8985,
+      "step": 6290
+    },
+    {
+      "epoch": 0.00885209104305383,
+      "grad_norm": 1.2249815464019775,
+      "learning_rate": 2.6551917942953488e-05,
+      "loss": 6.9706,
+      "step": 6300
+    },
+    {
+      "epoch": 0.008866141981217409,
+      "grad_norm": 1.2293840646743774,
+      "learning_rate": 2.6594070535337923e-05,
+      "loss": 6.8822,
+      "step": 6310
+    },
+    {
+      "epoch": 0.008880192919380987,
+      "grad_norm": 1.242686152458191,
+      "learning_rate": 2.6636223127722354e-05,
+      "loss": 6.8955,
+      "step": 6320
+    },
+    {
+      "epoch": 0.008894243857544563,
+      "grad_norm": 1.4114511013031006,
+      "learning_rate": 2.667837572010678e-05,
+      "loss": 6.8935,
+      "step": 6330
+    },
+    {
+      "epoch": 0.00890829479570814,
+      "grad_norm": 1.3672587871551514,
+      "learning_rate": 2.6720528312491216e-05,
+      "loss": 6.8912,
+      "step": 6340
+    },
+    {
+      "epoch": 0.008922345733871719,
+      "grad_norm": 1.2746614217758179,
+      "learning_rate": 2.6762680904875647e-05,
+      "loss": 6.8686,
+      "step": 6350
+    },
+    {
+      "epoch": 0.008936396672035296,
+      "grad_norm": 1.19154691696167,
+      "learning_rate": 2.6804833497260082e-05,
+      "loss": 6.9352,
+      "step": 6360
+    },
+    {
+      "epoch": 0.008950447610198874,
+      "grad_norm": 1.0989205837249756,
+      "learning_rate": 2.684698608964451e-05,
+      "loss": 6.9164,
+      "step": 6370
+    },
+    {
+      "epoch": 0.00896449854836245,
+      "grad_norm": 1.2119799852371216,
+      "learning_rate": 2.688913868202894e-05,
+      "loss": 6.9376,
+      "step": 6380
+    },
+    {
+      "epoch": 0.008978549486526028,
+      "grad_norm": 1.1754270792007446,
+      "learning_rate": 2.6931291274413375e-05,
+      "loss": 7.0092,
+      "step": 6390
+    },
+    {
+      "epoch": 0.008992600424689606,
+      "grad_norm": 1.3046258687973022,
+      "learning_rate": 2.6973443866797806e-05,
+      "loss": 6.9119,
+      "step": 6400
+    },
+    {
+      "epoch": 0.009006651362853184,
+      "grad_norm": 1.185444712638855,
+      "learning_rate": 2.7015596459182234e-05,
+      "loss": 6.9164,
+      "step": 6410
+    },
+    {
+      "epoch": 0.00902070230101676,
+      "grad_norm": 1.089029312133789,
+      "learning_rate": 2.705774905156667e-05,
+      "loss": 6.9452,
+      "step": 6420
+    },
+    {
+      "epoch": 0.009034753239180338,
+      "grad_norm": 1.1562166213989258,
+      "learning_rate": 2.70999016439511e-05,
+      "loss": 6.9627,
+      "step": 6430
+    },
+    {
+      "epoch": 0.009048804177343916,
+      "grad_norm": 1.1243866682052612,
+      "learning_rate": 2.7142054236335534e-05,
+      "loss": 6.9085,
+      "step": 6440
+    },
+    {
+      "epoch": 0.009062855115507494,
+      "grad_norm": 1.2206236124038696,
+      "learning_rate": 2.7184206828719962e-05,
+      "loss": 6.9762,
+      "step": 6450
+    },
+    {
+      "epoch": 0.00907690605367107,
+      "grad_norm": 1.1540404558181763,
+      "learning_rate": 2.7226359421104393e-05,
+      "loss": 6.9419,
+      "step": 6460
+    },
+    {
+      "epoch": 0.009090956991834648,
+      "grad_norm": 1.1724482774734497,
+      "learning_rate": 2.7268512013488828e-05,
+      "loss": 6.8162,
+      "step": 6470
+    },
+    {
+      "epoch": 0.009105007929998226,
+      "grad_norm": 1.1414722204208374,
+      "learning_rate": 2.731066460587326e-05,
+      "loss": 6.85,
+      "step": 6480
+    },
+    {
+      "epoch": 0.009119058868161804,
+      "grad_norm": 1.3120932579040527,
+      "learning_rate": 2.7352817198257687e-05,
+      "loss": 6.883,
+      "step": 6490
+    },
+    {
+      "epoch": 0.009133109806325382,
+      "grad_norm": 1.4635417461395264,
+      "learning_rate": 2.739496979064212e-05,
+      "loss": 6.8511,
+      "step": 6500
+    },
+    {
+      "epoch": 0.009147160744488958,
+      "grad_norm": 1.1401208639144897,
+      "learning_rate": 2.7437122383026553e-05,
+      "loss": 6.8325,
+      "step": 6510
+    },
+    {
+      "epoch": 0.009161211682652536,
+      "grad_norm": 1.2455246448516846,
+      "learning_rate": 2.7479274975410987e-05,
+      "loss": 6.9031,
+      "step": 6520
+    },
+    {
+      "epoch": 0.009175262620816114,
+      "grad_norm": 1.1095499992370605,
+      "learning_rate": 2.752142756779542e-05,
+      "loss": 6.9847,
+      "step": 6530
+    },
+    {
+      "epoch": 0.009189313558979692,
+      "grad_norm": 1.1827448606491089,
+      "learning_rate": 2.7563580160179846e-05,
+      "loss": 6.8022,
+      "step": 6540
+    },
+    {
+      "epoch": 0.009203364497143268,
+      "grad_norm": 1.260500431060791,
+      "learning_rate": 2.760573275256428e-05,
+      "loss": 6.8418,
+      "step": 6550
+    },
+    {
+      "epoch": 0.009217415435306846,
+      "grad_norm": 1.2838935852050781,
+      "learning_rate": 2.7647885344948712e-05,
+      "loss": 6.7921,
+      "step": 6560
+    },
+    {
+      "epoch": 0.009231466373470424,
+      "grad_norm": 1.320967674255371,
+      "learning_rate": 2.7690037937333146e-05,
+      "loss": 6.9222,
+      "step": 6570
+    },
+    {
+      "epoch": 0.009245517311634002,
+      "grad_norm": 1.126835823059082,
+      "learning_rate": 2.7732190529717574e-05,
+      "loss": 6.8736,
+      "step": 6580
+    },
+    {
+      "epoch": 0.00925956824979758,
+      "grad_norm": 1.314828634262085,
+      "learning_rate": 2.7774343122102005e-05,
+      "loss": 6.9342,
+      "step": 6590
+    },
+    {
+      "epoch": 0.009273619187961156,
+      "grad_norm": 1.155680775642395,
+      "learning_rate": 2.781649571448644e-05,
+      "loss": 6.8753,
+      "step": 6600
+    },
+    {
+      "epoch": 0.009287670126124734,
+      "grad_norm": 1.1509275436401367,
+      "learning_rate": 2.785864830687087e-05,
+      "loss": 6.9379,
+      "step": 6610
+    },
+    {
+      "epoch": 0.009301721064288311,
+      "grad_norm": 1.267380952835083,
+      "learning_rate": 2.79008008992553e-05,
+      "loss": 6.729,
+      "step": 6620
+    },
+    {
+      "epoch": 0.00931577200245189,
+      "grad_norm": 1.0873631238937378,
+      "learning_rate": 2.7942953491639733e-05,
+      "loss": 6.9163,
+      "step": 6630
+    },
+    {
+      "epoch": 0.009329822940615465,
+      "grad_norm": 1.1794565916061401,
+      "learning_rate": 2.7985106084024165e-05,
+      "loss": 6.8978,
+      "step": 6640
+    },
+    {
+      "epoch": 0.009343873878779043,
+      "grad_norm": 1.2664817571640015,
+      "learning_rate": 2.80272586764086e-05,
+      "loss": 6.8032,
+      "step": 6650
+    },
+    {
+      "epoch": 0.009357924816942621,
+      "grad_norm": 1.2982505559921265,
+      "learning_rate": 2.8069411268793027e-05,
+      "loss": 6.8812,
+      "step": 6660
+    },
+    {
+      "epoch": 0.0093719757551062,
+      "grad_norm": 1.0752772092819214,
+      "learning_rate": 2.8111563861177458e-05,
+      "loss": 6.8571,
+      "step": 6670
+    },
+    {
+      "epoch": 0.009386026693269777,
+      "grad_norm": 1.1509788036346436,
+      "learning_rate": 2.8153716453561893e-05,
+      "loss": 6.8439,
+      "step": 6680
+    },
+    {
+      "epoch": 0.009400077631433353,
+      "grad_norm": 1.1949745416641235,
+      "learning_rate": 2.8195869045946324e-05,
+      "loss": 6.8121,
+      "step": 6690
+    },
+    {
+      "epoch": 0.009414128569596931,
+      "grad_norm": 1.280206561088562,
+      "learning_rate": 2.823802163833075e-05,
+      "loss": 6.8162,
+      "step": 6700
+    },
+    {
+      "epoch": 0.009428179507760509,
+      "grad_norm": 1.1214113235473633,
+      "learning_rate": 2.8280174230715186e-05,
+      "loss": 6.9153,
+      "step": 6710
+    },
+    {
+      "epoch": 0.009442230445924087,
+      "grad_norm": 1.1382800340652466,
+      "learning_rate": 2.8322326823099617e-05,
+      "loss": 6.7568,
+      "step": 6720
+    },
+    {
+      "epoch": 0.009456281384087663,
+      "grad_norm": 1.1620829105377197,
+      "learning_rate": 2.8364479415484052e-05,
+      "loss": 6.8636,
+      "step": 6730
+    },
+    {
+      "epoch": 0.009470332322251241,
+      "grad_norm": 1.1612575054168701,
+      "learning_rate": 2.840663200786848e-05,
+      "loss": 6.8372,
+      "step": 6740
+    },
+    {
+      "epoch": 0.009484383260414819,
+      "grad_norm": 1.1702032089233398,
+      "learning_rate": 2.844878460025291e-05,
+      "loss": 6.9363,
+      "step": 6750
+    },
+    {
+      "epoch": 0.009498434198578397,
+      "grad_norm": 1.1699141263961792,
+      "learning_rate": 2.8490937192637345e-05,
+      "loss": 6.8231,
+      "step": 6760
+    },
+    {
+      "epoch": 0.009512485136741973,
+      "grad_norm": 1.2124075889587402,
+      "learning_rate": 2.8533089785021777e-05,
+      "loss": 6.9048,
+      "step": 6770
+    },
+    {
+      "epoch": 0.009526536074905551,
+      "grad_norm": 1.189448595046997,
+      "learning_rate": 2.857524237740621e-05,
+      "loss": 6.8002,
+      "step": 6780
+    },
+    {
+      "epoch": 0.009540587013069129,
+      "grad_norm": 1.2542109489440918,
+      "learning_rate": 2.861739496979064e-05,
+      "loss": 6.7874,
+      "step": 6790
+    },
+    {
+      "epoch": 0.009554637951232707,
+      "grad_norm": 1.1644352674484253,
+      "learning_rate": 2.865954756217507e-05,
+      "loss": 6.8212,
+      "step": 6800
+    },
+    {
+      "epoch": 0.009568688889396285,
+      "grad_norm": 1.0751014947891235,
+      "learning_rate": 2.8701700154559505e-05,
+      "loss": 6.8566,
+      "step": 6810
+    },
+    {
+      "epoch": 0.00958273982755986,
+      "grad_norm": 1.3088459968566895,
+      "learning_rate": 2.8743852746943936e-05,
+      "loss": 6.8754,
+      "step": 6820
+    },
+    {
+      "epoch": 0.009596790765723439,
+      "grad_norm": 1.2002370357513428,
+      "learning_rate": 2.8786005339328364e-05,
+      "loss": 6.7577,
+      "step": 6830
+    },
+    {
+      "epoch": 0.009610841703887017,
+      "grad_norm": 1.3408019542694092,
+      "learning_rate": 2.8828157931712798e-05,
+      "loss": 6.7329,
+      "step": 6840
+    },
+    {
+      "epoch": 0.009624892642050594,
+      "grad_norm": 1.1012219190597534,
+      "learning_rate": 2.887031052409723e-05,
+      "loss": 6.8618,
+      "step": 6850
+    },
+    {
+      "epoch": 0.00963894358021417,
+      "grad_norm": 1.610071063041687,
+      "learning_rate": 2.8912463116481664e-05,
+      "loss": 6.8109,
+      "step": 6860
+    },
+    {
+      "epoch": 0.009652994518377749,
+      "grad_norm": 1.3259707689285278,
+      "learning_rate": 2.895461570886609e-05,
+      "loss": 6.8621,
+      "step": 6870
+    },
+    {
+      "epoch": 0.009667045456541326,
+      "grad_norm": 1.3494741916656494,
+      "learning_rate": 2.8996768301250523e-05,
+      "loss": 6.8059,
+      "step": 6880
+    },
+    {
+      "epoch": 0.009681096394704904,
+      "grad_norm": 1.2349156141281128,
+      "learning_rate": 2.9038920893634957e-05,
+      "loss": 6.7744,
+      "step": 6890
+    },
+    {
+      "epoch": 0.009695147332868482,
+      "grad_norm": 1.1511788368225098,
+      "learning_rate": 2.908107348601939e-05,
+      "loss": 6.8519,
+      "step": 6900
+    },
+    {
+      "epoch": 0.009709198271032058,
+      "grad_norm": 1.3229706287384033,
+      "learning_rate": 2.9123226078403816e-05,
+      "loss": 6.9277,
+      "step": 6910
+    },
+    {
+      "epoch": 0.009723249209195636,
+      "grad_norm": 1.3774425983428955,
+      "learning_rate": 2.916537867078825e-05,
+      "loss": 6.821,
+      "step": 6920
+    },
+    {
+      "epoch": 0.009737300147359214,
+      "grad_norm": 1.2297898530960083,
+      "learning_rate": 2.9207531263172682e-05,
+      "loss": 6.7814,
+      "step": 6930
+    },
+    {
+      "epoch": 0.009751351085522792,
+      "grad_norm": 1.274950385093689,
+      "learning_rate": 2.9249683855557117e-05,
+      "loss": 6.8029,
+      "step": 6940
+    },
+    {
+      "epoch": 0.009765402023686368,
+      "grad_norm": 1.3338391780853271,
+      "learning_rate": 2.9291836447941544e-05,
+      "loss": 6.772,
+      "step": 6950
+    },
+    {
+      "epoch": 0.009779452961849946,
+      "grad_norm": 1.3144457340240479,
+      "learning_rate": 2.9333989040325975e-05,
+      "loss": 6.8667,
+      "step": 6960
+    },
+    {
+      "epoch": 0.009793503900013524,
+      "grad_norm": 1.1216480731964111,
+      "learning_rate": 2.937614163271041e-05,
+      "loss": 6.8348,
+      "step": 6970
+    },
+    {
+      "epoch": 0.009807554838177102,
+      "grad_norm": 1.1729443073272705,
+      "learning_rate": 2.941829422509484e-05,
+      "loss": 6.8038,
+      "step": 6980
+    },
+    {
+      "epoch": 0.00982160577634068,
+      "grad_norm": 1.1787049770355225,
+      "learning_rate": 2.946044681747927e-05,
+      "loss": 6.8645,
+      "step": 6990
+    },
+    {
+      "epoch": 0.009835656714504256,
+      "grad_norm": 1.3254892826080322,
+      "learning_rate": 2.9502599409863704e-05,
+      "loss": 6.7894,
+      "step": 7000
+    },
+    {
+      "epoch": 0.009849707652667834,
+      "grad_norm": 1.362343668937683,
+      "learning_rate": 2.9544752002248135e-05,
+      "loss": 6.7402,
+      "step": 7010
+    },
+    {
+      "epoch": 0.009863758590831412,
+      "grad_norm": 1.3607126474380493,
+      "learning_rate": 2.958690459463257e-05,
+      "loss": 6.7859,
+      "step": 7020
+    },
+    {
+      "epoch": 0.00987780952899499,
+      "grad_norm": 1.1887863874435425,
+      "learning_rate": 2.9629057187017e-05,
+      "loss": 6.847,
+      "step": 7030
+    },
+    {
+      "epoch": 0.009891860467158566,
+      "grad_norm": 3.412855625152588,
+      "learning_rate": 2.9671209779401428e-05,
+      "loss": 6.7606,
+      "step": 7040
+    },
+    {
+      "epoch": 0.009905911405322144,
+      "grad_norm": 1.3983832597732544,
+      "learning_rate": 2.9713362371785863e-05,
+      "loss": 6.8176,
+      "step": 7050
+    },
+    {
+      "epoch": 0.009919962343485722,
+      "grad_norm": 1.4330317974090576,
+      "learning_rate": 2.9755514964170294e-05,
+      "loss": 6.845,
+      "step": 7060
+    },
+    {
+      "epoch": 0.0099340132816493,
+      "grad_norm": 2.303821325302124,
+      "learning_rate": 2.979766755655473e-05,
+      "loss": 6.6737,
+      "step": 7070
+    },
+    {
+      "epoch": 0.009948064219812877,
+      "grad_norm": 1.079118013381958,
+      "learning_rate": 2.9839820148939156e-05,
+      "loss": 6.8816,
+      "step": 7080
+    },
+    {
+      "epoch": 0.009962115157976454,
+      "grad_norm": 1.1327508687973022,
+      "learning_rate": 2.9881972741323587e-05,
+      "loss": 6.8724,
+      "step": 7090
+    },
+    {
+      "epoch": 0.009976166096140032,
+      "grad_norm": 1.1694974899291992,
+      "learning_rate": 2.9924125333708022e-05,
+      "loss": 6.779,
+      "step": 7100
+    },
+    {
+      "epoch": 0.00999021703430361,
+      "grad_norm": 1.2839218378067017,
+      "learning_rate": 2.9966277926092453e-05,
+      "loss": 6.8217,
+      "step": 7110
+    },
+    {
+      "epoch": 0.010004267972467187,
+      "grad_norm": 1.1525654792785645,
+      "learning_rate": 3.000843051847688e-05,
+      "loss": 6.798,
+      "step": 7120
+    },
+    {
+      "epoch": 0.010018318910630764,
+      "grad_norm": 1.0679171085357666,
+      "learning_rate": 3.0050583110861315e-05,
+      "loss": 6.7938,
+      "step": 7130
+    },
+    {
+      "epoch": 0.010032369848794341,
+      "grad_norm": 1.440910816192627,
+      "learning_rate": 3.0092735703245747e-05,
+      "loss": 6.6892,
+      "step": 7140
+    },
+    {
+      "epoch": 0.01004642078695792,
+      "grad_norm": 1.1955666542053223,
+      "learning_rate": 3.013488829563018e-05,
+      "loss": 6.7765,
+      "step": 7150
+    },
+    {
+      "epoch": 0.010060471725121497,
+      "grad_norm": 1.318375825881958,
+      "learning_rate": 3.017704088801461e-05,
+      "loss": 6.8462,
+      "step": 7160
+    },
+    {
+      "epoch": 0.010074522663285073,
+      "grad_norm": 1.328094244003296,
+      "learning_rate": 3.021919348039904e-05,
+      "loss": 6.7747,
+      "step": 7170
+    },
+    {
+      "epoch": 0.010088573601448651,
+      "grad_norm": 1.5209085941314697,
+      "learning_rate": 3.0261346072783475e-05,
+      "loss": 6.852,
+      "step": 7180
+    },
+    {
+      "epoch": 0.01010262453961223,
+      "grad_norm": 1.3846906423568726,
+      "learning_rate": 3.0303498665167906e-05,
+      "loss": 6.7635,
+      "step": 7190
+    },
+    {
+      "epoch": 0.010116675477775807,
+      "grad_norm": 1.2889553308486938,
+      "learning_rate": 3.0345651257552334e-05,
+      "loss": 6.7899,
+      "step": 7200
+    },
+    {
+      "epoch": 0.010130726415939385,
+      "grad_norm": 1.244968056678772,
+      "learning_rate": 3.0387803849936768e-05,
+      "loss": 6.8504,
+      "step": 7210
+    },
+    {
+      "epoch": 0.010144777354102961,
+      "grad_norm": 1.5130319595336914,
+      "learning_rate": 3.04299564423212e-05,
+      "loss": 6.7033,
+      "step": 7220
+    },
+    {
+      "epoch": 0.010158828292266539,
+      "grad_norm": 1.188735008239746,
+      "learning_rate": 3.0472109034705634e-05,
+      "loss": 6.7656,
+      "step": 7230
+    },
+    {
+      "epoch": 0.010172879230430117,
+      "grad_norm": 1.2299892902374268,
+      "learning_rate": 3.051426162709006e-05,
+      "loss": 6.7203,
+      "step": 7240
+    },
+    {
+      "epoch": 0.010186930168593695,
+      "grad_norm": 1.3426145315170288,
+      "learning_rate": 3.055641421947449e-05,
+      "loss": 6.7867,
+      "step": 7250
+    },
+    {
+      "epoch": 0.010200981106757271,
+      "grad_norm": 1.440065860748291,
+      "learning_rate": 3.0598566811858924e-05,
+      "loss": 6.6989,
+      "step": 7260
+    },
+    {
+      "epoch": 0.010215032044920849,
+      "grad_norm": 1.7373557090759277,
+      "learning_rate": 3.064071940424336e-05,
+      "loss": 6.762,
+      "step": 7270
+    },
+    {
+      "epoch": 0.010229082983084427,
+      "grad_norm": 1.1723196506500244,
+      "learning_rate": 3.068287199662779e-05,
+      "loss": 6.7585,
+      "step": 7280
+    },
+    {
+      "epoch": 0.010243133921248005,
+      "grad_norm": 1.1131891012191772,
+      "learning_rate": 3.072502458901222e-05,
+      "loss": 6.6601,
+      "step": 7290
+    },
+    {
+      "epoch": 0.010257184859411583,
+      "grad_norm": 1.3221385478973389,
+      "learning_rate": 3.0767177181396655e-05,
+      "loss": 6.798,
+      "step": 7300
+    },
+    {
+      "epoch": 0.010271235797575159,
+      "grad_norm": 1.1797014474868774,
+      "learning_rate": 3.080932977378109e-05,
+      "loss": 6.7014,
+      "step": 7310
+    },
+    {
+      "epoch": 0.010285286735738737,
+      "grad_norm": 1.2688157558441162,
+      "learning_rate": 3.085148236616552e-05,
+      "loss": 6.7095,
+      "step": 7320
+    },
+    {
+      "epoch": 0.010299337673902315,
+      "grad_norm": 1.3802592754364014,
+      "learning_rate": 3.089363495854995e-05,
+      "loss": 6.6367,
+      "step": 7330
+    },
+    {
+      "epoch": 0.010313388612065892,
+      "grad_norm": 1.2973464727401733,
+      "learning_rate": 3.093578755093438e-05,
+      "loss": 6.6415,
+      "step": 7340
+    },
+    {
+      "epoch": 0.010327439550229469,
+      "grad_norm": 1.2183587551116943,
+      "learning_rate": 3.097794014331881e-05,
+      "loss": 6.7612,
+      "step": 7350
+    },
+    {
+      "epoch": 0.010341490488393047,
+      "grad_norm": 1.2998961210250854,
+      "learning_rate": 3.102009273570324e-05,
+      "loss": 6.7186,
+      "step": 7360
+    },
+    {
+      "epoch": 0.010355541426556624,
+      "grad_norm": 1.2842721939086914,
+      "learning_rate": 3.1062245328087674e-05,
+      "loss": 6.796,
+      "step": 7370
+    },
+    {
+      "epoch": 0.010369592364720202,
+      "grad_norm": 1.5530544519424438,
+      "learning_rate": 3.1104397920472105e-05,
+      "loss": 6.7277,
+      "step": 7380
+    },
+    {
+      "epoch": 0.01038364330288378,
+      "grad_norm": 1.4051449298858643,
+      "learning_rate": 3.1146550512856536e-05,
+      "loss": 6.6586,
+      "step": 7390
+    },
+    {
+      "epoch": 0.010397694241047356,
+      "grad_norm": 1.4194215536117554,
+      "learning_rate": 3.1188703105240974e-05,
+      "loss": 6.7337,
+      "step": 7400
+    },
+    {
+      "epoch": 0.010411745179210934,
+      "grad_norm": 1.2050557136535645,
+      "learning_rate": 3.12308556976254e-05,
+      "loss": 6.7198,
+      "step": 7410
+    },
+    {
+      "epoch": 0.010425796117374512,
+      "grad_norm": 1.340126872062683,
+      "learning_rate": 3.127300829000983e-05,
+      "loss": 6.6974,
+      "step": 7420
+    },
+    {
+      "epoch": 0.01043984705553809,
+      "grad_norm": 1.363870620727539,
+      "learning_rate": 3.131516088239427e-05,
+      "loss": 6.653,
+      "step": 7430
+    },
+    {
+      "epoch": 0.010453897993701666,
+      "grad_norm": 1.3523595333099365,
+      "learning_rate": 3.13573134747787e-05,
+      "loss": 6.6777,
+      "step": 7440
+    },
+    {
+      "epoch": 0.010467948931865244,
+      "grad_norm": 1.3135426044464111,
+      "learning_rate": 3.139946606716312e-05,
+      "loss": 6.7118,
+      "step": 7450
+    },
+    {
+      "epoch": 0.010481999870028822,
+      "grad_norm": 1.7983295917510986,
+      "learning_rate": 3.144161865954756e-05,
+      "loss": 6.5969,
+      "step": 7460
+    },
+    {
+      "epoch": 0.0104960508081924,
+      "grad_norm": 1.22897469997406,
+      "learning_rate": 3.148377125193199e-05,
+      "loss": 6.6821,
+      "step": 7470
+    },
+    {
+      "epoch": 0.010510101746355976,
+      "grad_norm": 1.2474480867385864,
+      "learning_rate": 3.152592384431642e-05,
+      "loss": 6.6954,
+      "step": 7480
+    },
+    {
+      "epoch": 0.010524152684519554,
+      "grad_norm": 1.1647069454193115,
+      "learning_rate": 3.1568076436700854e-05,
+      "loss": 6.7303,
+      "step": 7490
+    },
+    {
+      "epoch": 0.010538203622683132,
+      "grad_norm": 1.2561836242675781,
+      "learning_rate": 3.1610229029085286e-05,
+      "loss": 6.7037,
+      "step": 7500
+    },
+    {
+      "epoch": 0.01055225456084671,
+      "grad_norm": 1.7592111825942993,
+      "learning_rate": 3.165238162146972e-05,
+      "loss": 6.6393,
+      "step": 7510
+    },
+    {
+      "epoch": 0.010566305499010288,
+      "grad_norm": 1.4957785606384277,
+      "learning_rate": 3.169453421385415e-05,
+      "loss": 6.7021,
+      "step": 7520
+    },
+    {
+      "epoch": 0.010580356437173864,
+      "grad_norm": 1.2453256845474243,
+      "learning_rate": 3.1736686806238586e-05,
+      "loss": 6.7024,
+      "step": 7530
+    },
+    {
+      "epoch": 0.010594407375337442,
+      "grad_norm": 1.2314335107803345,
+      "learning_rate": 3.177883939862301e-05,
+      "loss": 6.6214,
+      "step": 7540
+    },
+    {
+      "epoch": 0.01060845831350102,
+      "grad_norm": 1.3570759296417236,
+      "learning_rate": 3.182099199100744e-05,
+      "loss": 6.6482,
+      "step": 7550
+    },
+    {
+      "epoch": 0.010622509251664598,
+      "grad_norm": 1.2058477401733398,
+      "learning_rate": 3.186314458339188e-05,
+      "loss": 6.7454,
+      "step": 7560
+    },
+    {
+      "epoch": 0.010636560189828174,
+      "grad_norm": 1.3977819681167603,
+      "learning_rate": 3.190529717577631e-05,
+      "loss": 6.7017,
+      "step": 7570
+    },
+    {
+      "epoch": 0.010650611127991752,
+      "grad_norm": 1.2504565715789795,
+      "learning_rate": 3.1947449768160735e-05,
+      "loss": 6.7633,
+      "step": 7580
+    },
+    {
+      "epoch": 0.01066466206615533,
+      "grad_norm": 1.3335305452346802,
+      "learning_rate": 3.198960236054517e-05,
+      "loss": 6.6588,
+      "step": 7590
+    },
+    {
+      "epoch": 0.010678713004318907,
+      "grad_norm": 1.2901426553726196,
+      "learning_rate": 3.2031754952929604e-05,
+      "loss": 6.6741,
+      "step": 7600
+    },
+    {
+      "epoch": 0.010692763942482485,
+      "grad_norm": 1.484226107597351,
+      "learning_rate": 3.2073907545314035e-05,
+      "loss": 6.684,
+      "step": 7610
+    },
+    {
+      "epoch": 0.010706814880646062,
+      "grad_norm": 1.2085685729980469,
+      "learning_rate": 3.2116060137698466e-05,
+      "loss": 6.7001,
+      "step": 7620
+    },
+    {
+      "epoch": 0.01072086581880964,
+      "grad_norm": 1.306835412979126,
+      "learning_rate": 3.21582127300829e-05,
+      "loss": 6.7055,
+      "step": 7630
+    },
+    {
+      "epoch": 0.010734916756973217,
+      "grad_norm": 1.3770405054092407,
+      "learning_rate": 3.220036532246733e-05,
+      "loss": 6.7178,
+      "step": 7640
+    },
+    {
+      "epoch": 0.010748967695136795,
+      "grad_norm": 1.4259955883026123,
+      "learning_rate": 3.224251791485176e-05,
+      "loss": 6.6999,
+      "step": 7650
+    },
+    {
+      "epoch": 0.010763018633300371,
+      "grad_norm": 1.2039424180984497,
+      "learning_rate": 3.228467050723619e-05,
+      "loss": 6.6817,
+      "step": 7660
+    },
+    {
+      "epoch": 0.01077706957146395,
+      "grad_norm": 1.3195371627807617,
+      "learning_rate": 3.232682309962062e-05,
+      "loss": 6.6196,
+      "step": 7670
+    },
+    {
+      "epoch": 0.010791120509627527,
+      "grad_norm": 1.228158950805664,
+      "learning_rate": 3.236897569200505e-05,
+      "loss": 6.6338,
+      "step": 7680
+    },
+    {
+      "epoch": 0.010805171447791105,
+      "grad_norm": 1.4088062047958374,
+      "learning_rate": 3.241112828438949e-05,
+      "loss": 6.6519,
+      "step": 7690
+    },
+    {
+      "epoch": 0.010819222385954683,
+      "grad_norm": 1.2354357242584229,
+      "learning_rate": 3.2453280876773916e-05,
+      "loss": 6.6397,
+      "step": 7700
+    },
+    {
+      "epoch": 0.01083327332411826,
+      "grad_norm": 1.2311620712280273,
+      "learning_rate": 3.249543346915835e-05,
+      "loss": 6.6254,
+      "step": 7710
+    },
+    {
+      "epoch": 0.010847324262281837,
+      "grad_norm": 1.1504223346710205,
+      "learning_rate": 3.2537586061542785e-05,
+      "loss": 6.7407,
+      "step": 7720
+    },
+    {
+      "epoch": 0.010861375200445415,
+      "grad_norm": 1.3598875999450684,
+      "learning_rate": 3.2579738653927216e-05,
+      "loss": 6.7246,
+      "step": 7730
+    },
+    {
+      "epoch": 0.010875426138608993,
+      "grad_norm": 1.3000996112823486,
+      "learning_rate": 3.262189124631164e-05,
+      "loss": 6.7343,
+      "step": 7740
+    },
+    {
+      "epoch": 0.010889477076772569,
+      "grad_norm": 1.4005444049835205,
+      "learning_rate": 3.266404383869608e-05,
+      "loss": 6.6171,
+      "step": 7750
+    },
+    {
+      "epoch": 0.010903528014936147,
+      "grad_norm": 1.3075186014175415,
+      "learning_rate": 3.270619643108051e-05,
+      "loss": 6.7218,
+      "step": 7760
+    },
+    {
+      "epoch": 0.010917578953099725,
+      "grad_norm": 1.383267879486084,
+      "learning_rate": 3.274834902346494e-05,
+      "loss": 6.6246,
+      "step": 7770
+    },
+    {
+      "epoch": 0.010931629891263303,
+      "grad_norm": 1.3374590873718262,
+      "learning_rate": 3.279050161584937e-05,
+      "loss": 6.757,
+      "step": 7780
+    },
+    {
+      "epoch": 0.01094568082942688,
+      "grad_norm": 1.3575273752212524,
+      "learning_rate": 3.28326542082338e-05,
+      "loss": 6.7405,
+      "step": 7790
+    },
+    {
+      "epoch": 0.010959731767590457,
+      "grad_norm": 1.5748355388641357,
+      "learning_rate": 3.2874806800618234e-05,
+      "loss": 6.6514,
+      "step": 7800
+    },
+    {
+      "epoch": 0.010973782705754035,
+      "grad_norm": 1.6142196655273438,
+      "learning_rate": 3.2916959393002665e-05,
+      "loss": 6.5779,
+      "step": 7810
+    },
+    {
+      "epoch": 0.010987833643917613,
+      "grad_norm": 1.1203367710113525,
+      "learning_rate": 3.29591119853871e-05,
+      "loss": 6.7072,
+      "step": 7820
+    },
+    {
+      "epoch": 0.01100188458208119,
+      "grad_norm": 1.2387946844100952,
+      "learning_rate": 3.300126457777153e-05,
+      "loss": 6.5173,
+      "step": 7830
+    },
+    {
+      "epoch": 0.011015935520244767,
+      "grad_norm": 1.3735456466674805,
+      "learning_rate": 3.304341717015596e-05,
+      "loss": 6.7404,
+      "step": 7840
+    },
+    {
+      "epoch": 0.011029986458408345,
+      "grad_norm": 1.399306297302246,
+      "learning_rate": 3.30855697625404e-05,
+      "loss": 6.6327,
+      "step": 7850
+    },
+    {
+      "epoch": 0.011044037396571922,
+      "grad_norm": 1.3363549709320068,
+      "learning_rate": 3.312772235492483e-05,
+      "loss": 6.6128,
+      "step": 7860
+    },
+    {
+      "epoch": 0.0110580883347355,
+      "grad_norm": 1.1382472515106201,
+      "learning_rate": 3.316987494730925e-05,
+      "loss": 6.6297,
+      "step": 7870
+    },
+    {
+      "epoch": 0.011072139272899077,
+      "grad_norm": 1.407041072845459,
+      "learning_rate": 3.321202753969369e-05,
+      "loss": 6.5615,
+      "step": 7880
+    },
+    {
+      "epoch": 0.011086190211062654,
+      "grad_norm": 1.3194184303283691,
+      "learning_rate": 3.325418013207812e-05,
+      "loss": 6.7404,
+      "step": 7890
+    },
+    {
+      "epoch": 0.011100241149226232,
+      "grad_norm": 1.298169732093811,
+      "learning_rate": 3.329633272446255e-05,
+      "loss": 6.6713,
+      "step": 7900
+    },
+    {
+      "epoch": 0.01111429208738981,
+      "grad_norm": 1.251718282699585,
+      "learning_rate": 3.3338485316846984e-05,
+      "loss": 6.623,
+      "step": 7910
+    },
+    {
+      "epoch": 0.011128343025553388,
+      "grad_norm": 1.2488762140274048,
+      "learning_rate": 3.3380637909231415e-05,
+      "loss": 6.5778,
+      "step": 7920
+    },
+    {
+      "epoch": 0.011142393963716964,
+      "grad_norm": 1.156110405921936,
+      "learning_rate": 3.3422790501615846e-05,
+      "loss": 6.66,
+      "step": 7930
+    },
+    {
+      "epoch": 0.011156444901880542,
+      "grad_norm": 1.4375478029251099,
+      "learning_rate": 3.346494309400028e-05,
+      "loss": 6.5979,
+      "step": 7940
+    },
+    {
+      "epoch": 0.01117049584004412,
+      "grad_norm": 1.3894151449203491,
+      "learning_rate": 3.350709568638471e-05,
+      "loss": 6.6524,
+      "step": 7950
+    },
+    {
+      "epoch": 0.011184546778207698,
+      "grad_norm": 1.1708953380584717,
+      "learning_rate": 3.354924827876914e-05,
+      "loss": 6.6713,
+      "step": 7960
+    },
+    {
+      "epoch": 0.011198597716371274,
+      "grad_norm": 1.281219482421875,
+      "learning_rate": 3.359140087115357e-05,
+      "loss": 6.6162,
+      "step": 7970
+    },
+    {
+      "epoch": 0.011212648654534852,
+      "grad_norm": 1.3841021060943604,
+      "learning_rate": 3.363355346353801e-05,
+      "loss": 6.5542,
+      "step": 7980
+    },
+    {
+      "epoch": 0.01122669959269843,
+      "grad_norm": 1.1785082817077637,
+      "learning_rate": 3.367570605592243e-05,
+      "loss": 6.5345,
+      "step": 7990
+    },
+    {
+      "epoch": 0.011240750530862008,
+      "grad_norm": 1.361771821975708,
+      "learning_rate": 3.3717858648306864e-05,
+      "loss": 6.6326,
+      "step": 8000
+    },
+    {
+      "epoch": 0.011254801469025586,
+      "grad_norm": 1.6702193021774292,
+      "learning_rate": 3.37600112406913e-05,
+      "loss": 6.6394,
+      "step": 8010
+    },
+    {
+      "epoch": 0.011268852407189162,
+      "grad_norm": 1.1941388845443726,
+      "learning_rate": 3.380216383307573e-05,
+      "loss": 6.6372,
+      "step": 8020
+    },
+    {
+      "epoch": 0.01128290334535274,
+      "grad_norm": 1.274591326713562,
+      "learning_rate": 3.3844316425460164e-05,
+      "loss": 6.6055,
+      "step": 8030
+    },
+    {
+      "epoch": 0.011296954283516318,
+      "grad_norm": 1.3808879852294922,
+      "learning_rate": 3.3886469017844596e-05,
+      "loss": 6.5508,
+      "step": 8040
+    },
+    {
+      "epoch": 0.011311005221679896,
+      "grad_norm": 1.2351412773132324,
+      "learning_rate": 3.392862161022903e-05,
+      "loss": 6.6163,
+      "step": 8050
+    },
+    {
+      "epoch": 0.011325056159843472,
+      "grad_norm": 1.38816237449646,
+      "learning_rate": 3.397077420261346e-05,
+      "loss": 6.6613,
+      "step": 8060
+    },
+    {
+      "epoch": 0.01133910709800705,
+      "grad_norm": 1.3823046684265137,
+      "learning_rate": 3.401292679499789e-05,
+      "loss": 6.6828,
+      "step": 8070
+    },
+    {
+      "epoch": 0.011353158036170628,
+      "grad_norm": 1.2434042692184448,
+      "learning_rate": 3.405507938738232e-05,
+      "loss": 6.5608,
+      "step": 8080
+    },
+    {
+      "epoch": 0.011367208974334206,
+      "grad_norm": 1.3518249988555908,
+      "learning_rate": 3.409723197976675e-05,
+      "loss": 6.6077,
+      "step": 8090
+    },
+    {
+      "epoch": 0.011381259912497783,
+      "grad_norm": 1.5739810466766357,
+      "learning_rate": 3.413938457215118e-05,
+      "loss": 6.7089,
+      "step": 8100
+    },
+    {
+      "epoch": 0.01139531085066136,
+      "grad_norm": 1.2564321756362915,
+      "learning_rate": 3.418153716453562e-05,
+      "loss": 6.6499,
+      "step": 8110
+    },
+    {
+      "epoch": 0.011409361788824937,
+      "grad_norm": 1.2737637758255005,
+      "learning_rate": 3.4223689756920045e-05,
+      "loss": 6.5284,
+      "step": 8120
+    },
+    {
+      "epoch": 0.011423412726988515,
+      "grad_norm": 1.3004697561264038,
+      "learning_rate": 3.4265842349304476e-05,
+      "loss": 6.6387,
+      "step": 8130
+    },
+    {
+      "epoch": 0.011437463665152093,
+      "grad_norm": 1.3879871368408203,
+      "learning_rate": 3.4307994941688914e-05,
+      "loss": 6.5927,
+      "step": 8140
+    },
+    {
+      "epoch": 0.01145151460331567,
+      "grad_norm": 1.338653802871704,
+      "learning_rate": 3.4350147534073345e-05,
+      "loss": 6.5564,
+      "step": 8150
+    },
+    {
+      "epoch": 0.011465565541479247,
+      "grad_norm": 1.4665417671203613,
+      "learning_rate": 3.439230012645777e-05,
+      "loss": 6.5217,
+      "step": 8160
+    },
+    {
+      "epoch": 0.011479616479642825,
+      "grad_norm": 1.3037580251693726,
+      "learning_rate": 3.443445271884221e-05,
+      "loss": 6.6446,
+      "step": 8170
+    },
+    {
+      "epoch": 0.011493667417806403,
+      "grad_norm": 1.2073763608932495,
+      "learning_rate": 3.447660531122664e-05,
+      "loss": 6.646,
+      "step": 8180
+    },
+    {
+      "epoch": 0.011507718355969981,
+      "grad_norm": 1.291408896446228,
+      "learning_rate": 3.451875790361107e-05,
+      "loss": 6.672,
+      "step": 8190
+    },
+    {
+      "epoch": 0.011521769294133557,
+      "grad_norm": 1.1927132606506348,
+      "learning_rate": 3.45609104959955e-05,
+      "loss": 6.5918,
+      "step": 8200
+    },
+    {
+      "epoch": 0.011535820232297135,
+      "grad_norm": 1.3004413843154907,
+      "learning_rate": 3.460306308837993e-05,
+      "loss": 6.4986,
+      "step": 8210
+    },
+    {
+      "epoch": 0.011549871170460713,
+      "grad_norm": 1.3016691207885742,
+      "learning_rate": 3.4645215680764363e-05,
+      "loss": 6.5268,
+      "step": 8220
+    },
+    {
+      "epoch": 0.011563922108624291,
+      "grad_norm": 1.3215467929840088,
+      "learning_rate": 3.4687368273148795e-05,
+      "loss": 6.5952,
+      "step": 8230
+    },
+    {
+      "epoch": 0.011577973046787867,
+      "grad_norm": 1.1828099489212036,
+      "learning_rate": 3.4729520865533226e-05,
+      "loss": 6.6606,
+      "step": 8240
+    },
+    {
+      "epoch": 0.011592023984951445,
+      "grad_norm": 1.3190696239471436,
+      "learning_rate": 3.477167345791766e-05,
+      "loss": 6.5955,
+      "step": 8250
+    },
+    {
+      "epoch": 0.011606074923115023,
+      "grad_norm": 1.2753030061721802,
+      "learning_rate": 3.481382605030209e-05,
+      "loss": 6.6255,
+      "step": 8260
+    },
+    {
+      "epoch": 0.0116201258612786,
+      "grad_norm": 1.3981962203979492,
+      "learning_rate": 3.4855978642686526e-05,
+      "loss": 6.4703,
+      "step": 8270
+    },
+    {
+      "epoch": 0.011634176799442177,
+      "grad_norm": 1.2518624067306519,
+      "learning_rate": 3.489813123507096e-05,
+      "loss": 6.5757,
+      "step": 8280
+    },
+    {
+      "epoch": 0.011648227737605755,
+      "grad_norm": 1.1934725046157837,
+      "learning_rate": 3.494028382745538e-05,
+      "loss": 6.5563,
+      "step": 8290
+    },
+    {
+      "epoch": 0.011662278675769333,
+      "grad_norm": 1.5110435485839844,
+      "learning_rate": 3.498243641983982e-05,
+      "loss": 6.565,
+      "step": 8300
+    },
+    {
+      "epoch": 0.01167632961393291,
+      "grad_norm": 1.2714707851409912,
+      "learning_rate": 3.502458901222425e-05,
+      "loss": 6.6103,
+      "step": 8310
+    },
+    {
+      "epoch": 0.011690380552096489,
+      "grad_norm": 1.3573921918869019,
+      "learning_rate": 3.506674160460868e-05,
+      "loss": 6.5801,
+      "step": 8320
+    },
+    {
+      "epoch": 0.011704431490260065,
+      "grad_norm": 1.2519656419754028,
+      "learning_rate": 3.510889419699311e-05,
+      "loss": 6.503,
+      "step": 8330
+    },
+    {
+      "epoch": 0.011718482428423643,
+      "grad_norm": 1.3184814453125,
+      "learning_rate": 3.5151046789377544e-05,
+      "loss": 6.6102,
+      "step": 8340
+    },
+    {
+      "epoch": 0.01173253336658722,
+      "grad_norm": 1.2902504205703735,
+      "learning_rate": 3.5193199381761975e-05,
+      "loss": 6.4634,
+      "step": 8350
+    },
+    {
+      "epoch": 0.011746584304750798,
+      "grad_norm": 1.3389567136764526,
+      "learning_rate": 3.5235351974146407e-05,
+      "loss": 6.5529,
+      "step": 8360
+    },
+    {
+      "epoch": 0.011760635242914375,
+      "grad_norm": 1.684384822845459,
+      "learning_rate": 3.527750456653084e-05,
+      "loss": 6.4738,
+      "step": 8370
+    },
+    {
+      "epoch": 0.011774686181077952,
+      "grad_norm": 1.1865743398666382,
+      "learning_rate": 3.531965715891527e-05,
+      "loss": 6.4601,
+      "step": 8380
+    },
+    {
+      "epoch": 0.01178873711924153,
+      "grad_norm": 1.2686684131622314,
+      "learning_rate": 3.53618097512997e-05,
+      "loss": 6.5644,
+      "step": 8390
+    },
+    {
+      "epoch": 0.011802788057405108,
+      "grad_norm": 1.3220882415771484,
+      "learning_rate": 3.540396234368414e-05,
+      "loss": 6.5406,
+      "step": 8400
+    },
+    {
+      "epoch": 0.011816838995568686,
+      "grad_norm": 1.302966833114624,
+      "learning_rate": 3.544611493606856e-05,
+      "loss": 6.5661,
+      "step": 8410
+    },
+    {
+      "epoch": 0.011830889933732262,
+      "grad_norm": 1.270646572113037,
+      "learning_rate": 3.5488267528452994e-05,
+      "loss": 6.4866,
+      "step": 8420
+    },
+    {
+      "epoch": 0.01184494087189584,
+      "grad_norm": 1.3006125688552856,
+      "learning_rate": 3.553042012083743e-05,
+      "loss": 6.5404,
+      "step": 8430
+    },
+    {
+      "epoch": 0.011858991810059418,
+      "grad_norm": 1.5760924816131592,
+      "learning_rate": 3.557257271322186e-05,
+      "loss": 6.6259,
+      "step": 8440
+    },
+    {
+      "epoch": 0.011873042748222996,
+      "grad_norm": 1.215240240097046,
+      "learning_rate": 3.561472530560629e-05,
+      "loss": 6.4761,
+      "step": 8450
+    },
+    {
+      "epoch": 0.011887093686386572,
+      "grad_norm": 1.2556153535842896,
+      "learning_rate": 3.5656877897990725e-05,
+      "loss": 6.53,
+      "step": 8460
+    },
+    {
+      "epoch": 0.01190114462455015,
+      "grad_norm": 1.2683357000350952,
+      "learning_rate": 3.5699030490375156e-05,
+      "loss": 6.4377,
+      "step": 8470
+    },
+    {
+      "epoch": 0.011915195562713728,
+      "grad_norm": 1.187314748764038,
+      "learning_rate": 3.574118308275959e-05,
+      "loss": 6.5894,
+      "step": 8480
+    },
+    {
+      "epoch": 0.011929246500877306,
+      "grad_norm": 1.3034703731536865,
+      "learning_rate": 3.578333567514402e-05,
+      "loss": 6.5753,
+      "step": 8490
+    },
+    {
+      "epoch": 0.011943297439040884,
+      "grad_norm": 1.4902338981628418,
+      "learning_rate": 3.582548826752845e-05,
+      "loss": 6.5297,
+      "step": 8500
+    },
+    {
+      "epoch": 0.01195734837720446,
+      "grad_norm": 1.3260395526885986,
+      "learning_rate": 3.586764085991288e-05,
+      "loss": 6.6046,
+      "step": 8510
+    },
+    {
+      "epoch": 0.011971399315368038,
+      "grad_norm": 1.3407644033432007,
+      "learning_rate": 3.590979345229731e-05,
+      "loss": 6.4248,
+      "step": 8520
+    },
+    {
+      "epoch": 0.011985450253531616,
+      "grad_norm": 1.1548830270767212,
+      "learning_rate": 3.595194604468175e-05,
+      "loss": 6.5296,
+      "step": 8530
+    },
+    {
+      "epoch": 0.011999501191695194,
+      "grad_norm": 1.3393534421920776,
+      "learning_rate": 3.5994098637066174e-05,
+      "loss": 6.5889,
+      "step": 8540
+    },
+    {
+      "epoch": 0.01201355212985877,
+      "grad_norm": 1.3542684316635132,
+      "learning_rate": 3.6036251229450605e-05,
+      "loss": 6.5508,
+      "step": 8550
+    },
+    {
+      "epoch": 0.012027603068022348,
+      "grad_norm": 1.5744121074676514,
+      "learning_rate": 3.6078403821835043e-05,
+      "loss": 6.5049,
+      "step": 8560
+    },
+    {
+      "epoch": 0.012041654006185926,
+      "grad_norm": 1.1964699029922485,
+      "learning_rate": 3.6120556414219475e-05,
+      "loss": 6.538,
+      "step": 8570
+    },
+    {
+      "epoch": 0.012055704944349504,
+      "grad_norm": 1.3729437589645386,
+      "learning_rate": 3.61627090066039e-05,
+      "loss": 6.5814,
+      "step": 8580
+    },
+    {
+      "epoch": 0.01206975588251308,
+      "grad_norm": 1.190537691116333,
+      "learning_rate": 3.620486159898834e-05,
+      "loss": 6.503,
+      "step": 8590
+    },
+    {
+      "epoch": 0.012083806820676658,
+      "grad_norm": 1.2392871379852295,
+      "learning_rate": 3.624701419137277e-05,
+      "loss": 6.4875,
+      "step": 8600
+    },
+    {
+      "epoch": 0.012097857758840235,
+      "grad_norm": 1.3132319450378418,
+      "learning_rate": 3.62891667837572e-05,
+      "loss": 6.4879,
+      "step": 8610
+    },
+    {
+      "epoch": 0.012111908697003813,
+      "grad_norm": 1.3270124197006226,
+      "learning_rate": 3.633131937614163e-05,
+      "loss": 6.415,
+      "step": 8620
+    },
+    {
+      "epoch": 0.012125959635167391,
+      "grad_norm": 1.2737349271774292,
+      "learning_rate": 3.637347196852606e-05,
+      "loss": 6.5448,
+      "step": 8630
+    },
+    {
+      "epoch": 0.012140010573330967,
+      "grad_norm": 1.1706392765045166,
+      "learning_rate": 3.641562456091049e-05,
+      "loss": 6.4907,
+      "step": 8640
+    },
+    {
+      "epoch": 0.012154061511494545,
+      "grad_norm": 1.3053278923034668,
+      "learning_rate": 3.6457777153294924e-05,
+      "loss": 6.6704,
+      "step": 8650
+    },
+    {
+      "epoch": 0.012168112449658123,
+      "grad_norm": 1.3305950164794922,
+      "learning_rate": 3.6499929745679355e-05,
+      "loss": 6.4386,
+      "step": 8660
+    },
+    {
+      "epoch": 0.012182163387821701,
+      "grad_norm": 1.307633638381958,
+      "learning_rate": 3.6542082338063786e-05,
+      "loss": 6.4467,
+      "step": 8670
+    },
+    {
+      "epoch": 0.012196214325985277,
+      "grad_norm": 1.3092464208602905,
+      "learning_rate": 3.658423493044822e-05,
+      "loss": 6.4597,
+      "step": 8680
+    },
+    {
+      "epoch": 0.012210265264148855,
+      "grad_norm": 1.368456244468689,
+      "learning_rate": 3.6626387522832655e-05,
+      "loss": 6.5467,
+      "step": 8690
+    },
+    {
+      "epoch": 0.012224316202312433,
+      "grad_norm": 1.2031217813491821,
+      "learning_rate": 3.666854011521708e-05,
+      "loss": 6.4379,
+      "step": 8700
+    },
+    {
+      "epoch": 0.012238367140476011,
+      "grad_norm": 1.300486445426941,
+      "learning_rate": 3.671069270760151e-05,
+      "loss": 6.4787,
+      "step": 8710
+    },
+    {
+      "epoch": 0.012252418078639589,
+      "grad_norm": 1.6316710710525513,
+      "learning_rate": 3.675284529998595e-05,
+      "loss": 6.4383,
+      "step": 8720
+    },
+    {
+      "epoch": 0.012266469016803165,
+      "grad_norm": 1.327189564704895,
+      "learning_rate": 3.679499789237038e-05,
+      "loss": 6.4901,
+      "step": 8730
+    },
+    {
+      "epoch": 0.012280519954966743,
+      "grad_norm": 1.4074703454971313,
+      "learning_rate": 3.6837150484754804e-05,
+      "loss": 6.5223,
+      "step": 8740
+    },
+    {
+      "epoch": 0.012294570893130321,
+      "grad_norm": 1.3575947284698486,
+      "learning_rate": 3.687930307713924e-05,
+      "loss": 6.5039,
+      "step": 8750
+    },
+    {
+      "epoch": 0.012308621831293899,
+      "grad_norm": 1.2397525310516357,
+      "learning_rate": 3.6921455669523674e-05,
+      "loss": 6.3608,
+      "step": 8760
+    },
+    {
+      "epoch": 0.012322672769457475,
+      "grad_norm": 1.3196090459823608,
+      "learning_rate": 3.6963608261908105e-05,
+      "loss": 6.4237,
+      "step": 8770
+    },
+    {
+      "epoch": 0.012336723707621053,
+      "grad_norm": 1.4720513820648193,
+      "learning_rate": 3.7005760854292536e-05,
+      "loss": 6.4594,
+      "step": 8780
+    },
+    {
+      "epoch": 0.01235077464578463,
+      "grad_norm": 1.3891912698745728,
+      "learning_rate": 3.704791344667697e-05,
+      "loss": 6.4615,
+      "step": 8790
+    },
+    {
+      "epoch": 0.012364825583948209,
+      "grad_norm": 1.2116029262542725,
+      "learning_rate": 3.70900660390614e-05,
+      "loss": 6.4994,
+      "step": 8800
+    },
+    {
+      "epoch": 0.012378876522111787,
+      "grad_norm": 1.3657456636428833,
+      "learning_rate": 3.713221863144583e-05,
+      "loss": 6.5264,
+      "step": 8810
+    },
+    {
+      "epoch": 0.012392927460275363,
+      "grad_norm": 1.2680433988571167,
+      "learning_rate": 3.717437122383027e-05,
+      "loss": 6.4179,
+      "step": 8820
+    },
+    {
+      "epoch": 0.01240697839843894,
+      "grad_norm": 1.2071449756622314,
+      "learning_rate": 3.721652381621469e-05,
+      "loss": 6.6144,
+      "step": 8830
+    },
+    {
+      "epoch": 0.012421029336602519,
+      "grad_norm": 1.2629899978637695,
+      "learning_rate": 3.725867640859912e-05,
+      "loss": 6.5407,
+      "step": 8840
+    },
+    {
+      "epoch": 0.012435080274766096,
+      "grad_norm": 1.296008586883545,
+      "learning_rate": 3.730082900098356e-05,
+      "loss": 6.4203,
+      "step": 8850
+    },
+    {
+      "epoch": 0.012449131212929673,
+      "grad_norm": 1.4418439865112305,
+      "learning_rate": 3.734298159336799e-05,
+      "loss": 6.4575,
+      "step": 8860
+    },
+    {
+      "epoch": 0.01246318215109325,
+      "grad_norm": 1.3441669940948486,
+      "learning_rate": 3.7385134185752416e-05,
+      "loss": 6.4932,
+      "step": 8870
+    },
+    {
+      "epoch": 0.012477233089256828,
+      "grad_norm": 1.339735507965088,
+      "learning_rate": 3.7427286778136854e-05,
+      "loss": 6.5467,
+      "step": 8880
+    },
+    {
+      "epoch": 0.012491284027420406,
+      "grad_norm": 1.3307244777679443,
+      "learning_rate": 3.7469439370521285e-05,
+      "loss": 6.5179,
+      "step": 8890
+    },
+    {
+      "epoch": 0.012505334965583984,
+      "grad_norm": 1.3968795537948608,
+      "learning_rate": 3.751159196290571e-05,
+      "loss": 6.5239,
+      "step": 8900
+    },
+    {
+      "epoch": 0.01251938590374756,
+      "grad_norm": 1.3690930604934692,
+      "learning_rate": 3.755374455529015e-05,
+      "loss": 6.585,
+      "step": 8910
+    },
+    {
+      "epoch": 0.012533436841911138,
+      "grad_norm": 1.4065380096435547,
+      "learning_rate": 3.759589714767458e-05,
+      "loss": 6.5333,
+      "step": 8920
+    },
+    {
+      "epoch": 0.012547487780074716,
+      "grad_norm": 1.351392388343811,
+      "learning_rate": 3.7638049740059e-05,
+      "loss": 6.4869,
+      "step": 8930
+    },
+    {
+      "epoch": 0.012561538718238294,
+      "grad_norm": 1.3816670179367065,
+      "learning_rate": 3.768020233244344e-05,
+      "loss": 6.4295,
+      "step": 8940
+    },
+    {
+      "epoch": 0.01257558965640187,
+      "grad_norm": 1.369028091430664,
+      "learning_rate": 3.772235492482787e-05,
+      "loss": 6.504,
+      "step": 8950
+    },
+    {
+      "epoch": 0.012589640594565448,
+      "grad_norm": 1.4562000036239624,
+      "learning_rate": 3.776450751721231e-05,
+      "loss": 6.4471,
+      "step": 8960
+    },
+    {
+      "epoch": 0.012603691532729026,
+      "grad_norm": 1.8173776865005493,
+      "learning_rate": 3.7806660109596735e-05,
+      "loss": 6.5009,
+      "step": 8970
+    },
+    {
+      "epoch": 0.012617742470892604,
+      "grad_norm": 1.3316607475280762,
+      "learning_rate": 3.7848812701981166e-05,
+      "loss": 6.3643,
+      "step": 8980
+    },
+    {
+      "epoch": 0.01263179340905618,
+      "grad_norm": 1.2292101383209229,
+      "learning_rate": 3.7890965294365604e-05,
+      "loss": 6.3989,
+      "step": 8990
+    },
+    {
+      "epoch": 0.012645844347219758,
+      "grad_norm": 1.34901762008667,
+      "learning_rate": 3.793311788675003e-05,
+      "loss": 6.4566,
+      "step": 9000
+    },
+    {
+      "epoch": 0.012659895285383336,
+      "grad_norm": 1.2581123113632202,
+      "learning_rate": 3.7975270479134466e-05,
+      "loss": 6.4794,
+      "step": 9010
+    },
+    {
+      "epoch": 0.012673946223546914,
+      "grad_norm": 1.2883479595184326,
+      "learning_rate": 3.80174230715189e-05,
+      "loss": 6.4325,
+      "step": 9020
+    },
+    {
+      "epoch": 0.012687997161710492,
+      "grad_norm": 1.5478203296661377,
+      "learning_rate": 3.805957566390332e-05,
+      "loss": 6.3938,
+      "step": 9030
+    },
+    {
+      "epoch": 0.012702048099874068,
+      "grad_norm": 1.318039059638977,
+      "learning_rate": 3.810172825628776e-05,
+      "loss": 6.5348,
+      "step": 9040
+    },
+    {
+      "epoch": 0.012716099038037646,
+      "grad_norm": 1.3280646800994873,
+      "learning_rate": 3.814388084867219e-05,
+      "loss": 6.5114,
+      "step": 9050
+    },
+    {
+      "epoch": 0.012730149976201224,
+      "grad_norm": 1.2793512344360352,
+      "learning_rate": 3.8186033441056615e-05,
+      "loss": 6.483,
+      "step": 9060
+    },
+    {
+      "epoch": 0.012744200914364802,
+      "grad_norm": 1.3596251010894775,
+      "learning_rate": 3.822818603344105e-05,
+      "loss": 6.449,
+      "step": 9070
+    },
+    {
+      "epoch": 0.012758251852528378,
+      "grad_norm": 1.2819275856018066,
+      "learning_rate": 3.8270338625825484e-05,
+      "loss": 6.554,
+      "step": 9080
+    },
+    {
+      "epoch": 0.012772302790691956,
+      "grad_norm": 1.3615672588348389,
+      "learning_rate": 3.831249121820992e-05,
+      "loss": 6.4931,
+      "step": 9090
+    },
+    {
+      "epoch": 0.012786353728855534,
+      "grad_norm": 1.944034218788147,
+      "learning_rate": 3.835464381059435e-05,
+      "loss": 6.4004,
+      "step": 9100
+    },
+    {
+      "epoch": 0.012800404667019111,
+      "grad_norm": 1.4013776779174805,
+      "learning_rate": 3.839679640297878e-05,
+      "loss": 6.3946,
+      "step": 9110
+    },
+    {
+      "epoch": 0.01281445560518269,
+      "grad_norm": 1.3120478391647339,
+      "learning_rate": 3.8438948995363216e-05,
+      "loss": 6.3586,
+      "step": 9120
+    },
+    {
+      "epoch": 0.012828506543346265,
+      "grad_norm": 1.2097243070602417,
+      "learning_rate": 3.848110158774764e-05,
+      "loss": 6.3219,
+      "step": 9130
+    },
+    {
+      "epoch": 0.012842557481509843,
+      "grad_norm": 1.2938241958618164,
+      "learning_rate": 3.852325418013207e-05,
+      "loss": 6.5543,
+      "step": 9140
+    },
+    {
+      "epoch": 0.012856608419673421,
+      "grad_norm": 1.4525402784347534,
+      "learning_rate": 3.856540677251651e-05,
+      "loss": 6.4622,
+      "step": 9150
+    },
+    {
+      "epoch": 0.012870659357837,
+      "grad_norm": 1.2755237817764282,
+      "learning_rate": 3.8607559364900934e-05,
+      "loss": 6.3855,
+      "step": 9160
+    },
+    {
+      "epoch": 0.012884710296000575,
+      "grad_norm": 1.2028285264968872,
+      "learning_rate": 3.864971195728537e-05,
+      "loss": 6.4834,
+      "step": 9170
+    },
+    {
+      "epoch": 0.012898761234164153,
+      "grad_norm": 1.3781890869140625,
+      "learning_rate": 3.86918645496698e-05,
+      "loss": 6.4325,
+      "step": 9180
+    },
+    {
+      "epoch": 0.012912812172327731,
+      "grad_norm": 1.3861724138259888,
+      "learning_rate": 3.873401714205423e-05,
+      "loss": 6.3593,
+      "step": 9190
+    },
+    {
+      "epoch": 0.012926863110491309,
+      "grad_norm": 1.4471116065979004,
+      "learning_rate": 3.8776169734438665e-05,
+      "loss": 6.4511,
+      "step": 9200
+    },
+    {
+      "epoch": 0.012940914048654887,
+      "grad_norm": 1.250730037689209,
+      "learning_rate": 3.8818322326823096e-05,
+      "loss": 6.4343,
+      "step": 9210
+    },
+    {
+      "epoch": 0.012954964986818463,
+      "grad_norm": 1.2801191806793213,
+      "learning_rate": 3.886047491920752e-05,
+      "loss": 6.3752,
+      "step": 9220
+    },
+    {
+      "epoch": 0.012969015924982041,
+      "grad_norm": 1.210339069366455,
+      "learning_rate": 3.890262751159196e-05,
+      "loss": 6.4224,
+      "step": 9230
+    },
+    {
+      "epoch": 0.012983066863145619,
+      "grad_norm": 1.3136581182479858,
+      "learning_rate": 3.894478010397639e-05,
+      "loss": 6.3563,
+      "step": 9240
+    },
+    {
+      "epoch": 0.012997117801309197,
+      "grad_norm": 1.5053766965866089,
+      "learning_rate": 3.898693269636083e-05,
+      "loss": 6.3402,
+      "step": 9250
+    },
+    {
+      "epoch": 0.013011168739472773,
+      "grad_norm": 1.661244511604309,
+      "learning_rate": 3.902908528874525e-05,
+      "loss": 6.3517,
+      "step": 9260
+    },
+    {
+      "epoch": 0.013025219677636351,
+      "grad_norm": 1.3023381233215332,
+      "learning_rate": 3.907123788112968e-05,
+      "loss": 6.5525,
+      "step": 9270
+    },
+    {
+      "epoch": 0.013039270615799929,
+      "grad_norm": 1.444373369216919,
+      "learning_rate": 3.911339047351412e-05,
+      "loss": 6.5172,
+      "step": 9280
+    },
+    {
+      "epoch": 0.013053321553963507,
+      "grad_norm": 1.375860333442688,
+      "learning_rate": 3.9155543065898546e-05,
+      "loss": 6.4256,
+      "step": 9290
+    },
+    {
+      "epoch": 0.013067372492127083,
+      "grad_norm": 1.346632719039917,
+      "learning_rate": 3.9197695658282984e-05,
+      "loss": 6.3385,
+      "step": 9300
+    },
+    {
+      "epoch": 0.01308142343029066,
+      "grad_norm": 1.2784466743469238,
+      "learning_rate": 3.9239848250667415e-05,
+      "loss": 6.3262,
+      "step": 9310
+    },
+    {
+      "epoch": 0.013095474368454239,
+      "grad_norm": 1.4308885335922241,
+      "learning_rate": 3.928200084305184e-05,
+      "loss": 6.3977,
+      "step": 9320
+    },
+    {
+      "epoch": 0.013109525306617817,
+      "grad_norm": 1.379128336906433,
+      "learning_rate": 3.932415343543628e-05,
+      "loss": 6.3745,
+      "step": 9330
+    },
+    {
+      "epoch": 0.013123576244781394,
+      "grad_norm": 1.3541892766952515,
+      "learning_rate": 3.936630602782071e-05,
+      "loss": 6.5073,
+      "step": 9340
+    },
+    {
+      "epoch": 0.01313762718294497,
+      "grad_norm": 1.5439261198043823,
+      "learning_rate": 3.940845862020513e-05,
+      "loss": 6.4435,
+      "step": 9350
+    },
+    {
+      "epoch": 0.013151678121108549,
+      "grad_norm": 1.257303237915039,
+      "learning_rate": 3.945061121258957e-05,
+      "loss": 6.3503,
+      "step": 9360
+    },
+    {
+      "epoch": 0.013165729059272126,
+      "grad_norm": 1.358400583267212,
+      "learning_rate": 3.9492763804974e-05,
+      "loss": 6.3759,
+      "step": 9370
+    },
+    {
+      "epoch": 0.013179779997435704,
+      "grad_norm": 1.232875108718872,
+      "learning_rate": 3.953491639735844e-05,
+      "loss": 6.3632,
+      "step": 9380
+    },
+    {
+      "epoch": 0.01319383093559928,
+      "grad_norm": 1.3150238990783691,
+      "learning_rate": 3.9577068989742864e-05,
+      "loss": 6.4415,
+      "step": 9390
+    },
+    {
+      "epoch": 0.013207881873762858,
+      "grad_norm": 1.2941501140594482,
+      "learning_rate": 3.9619221582127295e-05,
+      "loss": 6.5106,
+      "step": 9400
+    },
+    {
+      "epoch": 0.013221932811926436,
+      "grad_norm": 1.3233859539031982,
+      "learning_rate": 3.966137417451173e-05,
+      "loss": 6.4211,
+      "step": 9410
+    },
+    {
+      "epoch": 0.013235983750090014,
+      "grad_norm": 1.309694528579712,
+      "learning_rate": 3.970352676689616e-05,
+      "loss": 6.4375,
+      "step": 9420
+    },
+    {
+      "epoch": 0.013250034688253592,
+      "grad_norm": 1.3564949035644531,
+      "learning_rate": 3.974567935928059e-05,
+      "loss": 6.3014,
+      "step": 9430
+    },
+    {
+      "epoch": 0.013264085626417168,
+      "grad_norm": 1.276855230331421,
+      "learning_rate": 3.978783195166503e-05,
+      "loss": 6.3429,
+      "step": 9440
+    },
+    {
+      "epoch": 0.013278136564580746,
+      "grad_norm": 1.356063961982727,
+      "learning_rate": 3.982998454404945e-05,
+      "loss": 6.4402,
+      "step": 9450
+    },
+    {
+      "epoch": 0.013292187502744324,
+      "grad_norm": 1.242181658744812,
+      "learning_rate": 3.987213713643389e-05,
+      "loss": 6.4431,
+      "step": 9460
+    },
+    {
+      "epoch": 0.013306238440907902,
+      "grad_norm": 1.3604857921600342,
+      "learning_rate": 3.991428972881832e-05,
+      "loss": 6.3414,
+      "step": 9470
+    },
+    {
+      "epoch": 0.013320289379071478,
+      "grad_norm": 1.2341139316558838,
+      "learning_rate": 3.9956442321202745e-05,
+      "loss": 6.4268,
+      "step": 9480
+    },
+    {
+      "epoch": 0.013334340317235056,
+      "grad_norm": 1.4619051218032837,
+      "learning_rate": 3.999859491358718e-05,
+      "loss": 6.4312,
+      "step": 9490
+    },
+    {
+      "epoch": 0.013348391255398634,
+      "grad_norm": 1.2710881233215332,
+      "learning_rate": 4.0040747505971614e-05,
+      "loss": 6.3772,
+      "step": 9500
+    },
+    {
+      "epoch": 0.013362442193562212,
+      "grad_norm": 1.3365315198898315,
+      "learning_rate": 4.008290009835605e-05,
+      "loss": 6.3996,
+      "step": 9510
+    },
+    {
+      "epoch": 0.01337649313172579,
+      "grad_norm": 1.246775507926941,
+      "learning_rate": 4.0125052690740476e-05,
+      "loss": 6.4799,
+      "step": 9520
+    },
+    {
+      "epoch": 0.013390544069889366,
+      "grad_norm": 1.3018437623977661,
+      "learning_rate": 4.016720528312491e-05,
+      "loss": 6.3685,
+      "step": 9530
+    },
+    {
+      "epoch": 0.013404595008052944,
+      "grad_norm": 1.413640022277832,
+      "learning_rate": 4.0209357875509345e-05,
+      "loss": 6.3522,
+      "step": 9540
+    },
+    {
+      "epoch": 0.013418645946216522,
+      "grad_norm": 1.3587901592254639,
+      "learning_rate": 4.025151046789377e-05,
+      "loss": 6.3307,
+      "step": 9550
+    },
+    {
+      "epoch": 0.0134326968843801,
+      "grad_norm": 1.3375356197357178,
+      "learning_rate": 4.02936630602782e-05,
+      "loss": 6.423,
+      "step": 9560
+    },
+    {
+      "epoch": 0.013446747822543676,
+      "grad_norm": 1.5337927341461182,
+      "learning_rate": 4.033581565266264e-05,
+      "loss": 6.3995,
+      "step": 9570
+    },
+    {
+      "epoch": 0.013460798760707254,
+      "grad_norm": 1.299627661705017,
+      "learning_rate": 4.037796824504706e-05,
+      "loss": 6.3389,
+      "step": 9580
+    },
+    {
+      "epoch": 0.013474849698870832,
+      "grad_norm": 1.1940332651138306,
+      "learning_rate": 4.04201208374315e-05,
+      "loss": 6.3567,
+      "step": 9590
+    },
+    {
+      "epoch": 0.01348890063703441,
+      "grad_norm": 1.4569729566574097,
+      "learning_rate": 4.046227342981593e-05,
+      "loss": 6.3624,
+      "step": 9600
+    },
+    {
+      "epoch": 0.013502951575197987,
+      "grad_norm": 1.3093900680541992,
+      "learning_rate": 4.0504426022200357e-05,
+      "loss": 6.4537,
+      "step": 9610
+    },
+    {
+      "epoch": 0.013517002513361564,
+      "grad_norm": 1.5371428728103638,
+      "learning_rate": 4.0546578614584795e-05,
+      "loss": 6.414,
+      "step": 9620
+    },
+    {
+      "epoch": 0.013531053451525141,
+      "grad_norm": 1.2927149534225464,
+      "learning_rate": 4.0588731206969226e-05,
+      "loss": 6.3938,
+      "step": 9630
+    },
+    {
+      "epoch": 0.01354510438968872,
+      "grad_norm": 1.2334342002868652,
+      "learning_rate": 4.063088379935365e-05,
+      "loss": 6.36,
+      "step": 9640
+    },
+    {
+      "epoch": 0.013559155327852297,
+      "grad_norm": 1.3164347410202026,
+      "learning_rate": 4.067303639173809e-05,
+      "loss": 6.324,
+      "step": 9650
+    },
+    {
+      "epoch": 0.013573206266015873,
+      "grad_norm": 1.3269118070602417,
+      "learning_rate": 4.071518898412252e-05,
+      "loss": 6.3407,
+      "step": 9660
+    },
+    {
+      "epoch": 0.013587257204179451,
+      "grad_norm": 1.3465888500213623,
+      "learning_rate": 4.075734157650696e-05,
+      "loss": 6.3591,
+      "step": 9670
+    },
+    {
+      "epoch": 0.01360130814234303,
+      "grad_norm": 1.2967374324798584,
+      "learning_rate": 4.079949416889138e-05,
+      "loss": 6.3319,
+      "step": 9680
+    },
+    {
+      "epoch": 0.013615359080506607,
+      "grad_norm": 1.2995193004608154,
+      "learning_rate": 4.084164676127581e-05,
+      "loss": 6.3171,
+      "step": 9690
+    },
+    {
+      "epoch": 0.013629410018670183,
+      "grad_norm": 1.4583133459091187,
+      "learning_rate": 4.088379935366025e-05,
+      "loss": 6.296,
+      "step": 9700
+    },
+    {
+      "epoch": 0.013643460956833761,
+      "grad_norm": 1.3236838579177856,
+      "learning_rate": 4.0925951946044675e-05,
+      "loss": 6.4108,
+      "step": 9710
+    },
+    {
+      "epoch": 0.013657511894997339,
+      "grad_norm": 1.3856712579727173,
+      "learning_rate": 4.0968104538429106e-05,
+      "loss": 6.3383,
+      "step": 9720
+    },
+    {
+      "epoch": 0.013671562833160917,
+      "grad_norm": 1.4215503931045532,
+      "learning_rate": 4.1010257130813544e-05,
+      "loss": 6.3143,
+      "step": 9730
+    },
+    {
+      "epoch": 0.013685613771324495,
+      "grad_norm": 1.3027349710464478,
+      "learning_rate": 4.105240972319797e-05,
+      "loss": 6.3573,
+      "step": 9740
+    },
+    {
+      "epoch": 0.013699664709488071,
+      "grad_norm": 1.5820553302764893,
+      "learning_rate": 4.1094562315582406e-05,
+      "loss": 6.3765,
+      "step": 9750
+    },
+    {
+      "epoch": 0.013713715647651649,
+      "grad_norm": 1.3663697242736816,
+      "learning_rate": 4.113671490796684e-05,
+      "loss": 6.3362,
+      "step": 9760
+    },
+    {
+      "epoch": 0.013727766585815227,
+      "grad_norm": 1.3366793394088745,
+      "learning_rate": 4.117886750035126e-05,
+      "loss": 6.3561,
+      "step": 9770
+    },
+    {
+      "epoch": 0.013741817523978805,
+      "grad_norm": 1.2957160472869873,
+      "learning_rate": 4.12210200927357e-05,
+      "loss": 6.402,
+      "step": 9780
+    },
+    {
+      "epoch": 0.013755868462142381,
+      "grad_norm": 1.2606065273284912,
+      "learning_rate": 4.126317268512013e-05,
+      "loss": 6.3366,
+      "step": 9790
+    },
+    {
+      "epoch": 0.013769919400305959,
+      "grad_norm": 1.3806688785552979,
+      "learning_rate": 4.130532527750457e-05,
+      "loss": 6.3261,
+      "step": 9800
+    },
+    {
+      "epoch": 0.013783970338469537,
+      "grad_norm": 1.4678658246994019,
+      "learning_rate": 4.1347477869888993e-05,
+      "loss": 6.3808,
+      "step": 9810
+    },
+    {
+      "epoch": 0.013798021276633115,
+      "grad_norm": 1.2760154008865356,
+      "learning_rate": 4.1389630462273425e-05,
+      "loss": 6.3305,
+      "step": 9820
+    },
+    {
+      "epoch": 0.013812072214796692,
+      "grad_norm": 1.4611419439315796,
+      "learning_rate": 4.143178305465786e-05,
+      "loss": 6.2193,
+      "step": 9830
+    },
+    {
+      "epoch": 0.013826123152960269,
+      "grad_norm": 1.3268877267837524,
+      "learning_rate": 4.147393564704229e-05,
+      "loss": 6.3524,
+      "step": 9840
+    },
+    {
+      "epoch": 0.013840174091123847,
+      "grad_norm": 1.3382116556167603,
+      "learning_rate": 4.151608823942672e-05,
+      "loss": 6.4381,
+      "step": 9850
+    },
+    {
+      "epoch": 0.013854225029287424,
+      "grad_norm": 1.4644103050231934,
+      "learning_rate": 4.1558240831811156e-05,
+      "loss": 6.2919,
+      "step": 9860
+    },
+    {
+      "epoch": 0.013868275967451002,
+      "grad_norm": 1.427890658378601,
+      "learning_rate": 4.160039342419558e-05,
+      "loss": 6.3372,
+      "step": 9870
+    },
+    {
+      "epoch": 0.013882326905614579,
+      "grad_norm": 1.189791202545166,
+      "learning_rate": 4.164254601658002e-05,
+      "loss": 6.3238,
+      "step": 9880
+    },
+    {
+      "epoch": 0.013896377843778156,
+      "grad_norm": 1.2701431512832642,
+      "learning_rate": 4.168469860896445e-05,
+      "loss": 6.3215,
+      "step": 9890
+    },
+    {
+      "epoch": 0.013910428781941734,
+      "grad_norm": 1.2288933992385864,
+      "learning_rate": 4.1726851201348874e-05,
+      "loss": 6.2862,
+      "step": 9900
+    },
+    {
+      "epoch": 0.013924479720105312,
+      "grad_norm": 1.2237406969070435,
+      "learning_rate": 4.176900379373331e-05,
+      "loss": 6.47,
+      "step": 9910
+    },
+    {
+      "epoch": 0.01393853065826889,
+      "grad_norm": 1.292963981628418,
+      "learning_rate": 4.181115638611774e-05,
+      "loss": 6.4382,
+      "step": 9920
+    },
+    {
+      "epoch": 0.013952581596432466,
+      "grad_norm": 1.2853929996490479,
+      "learning_rate": 4.185330897850217e-05,
+      "loss": 6.3467,
+      "step": 9930
+    },
+    {
+      "epoch": 0.013966632534596044,
+      "grad_norm": 1.2729337215423584,
+      "learning_rate": 4.1895461570886605e-05,
+      "loss": 6.3551,
+      "step": 9940
+    },
+    {
+      "epoch": 0.013980683472759622,
+      "grad_norm": 1.2991608381271362,
+      "learning_rate": 4.1937614163271037e-05,
+      "loss": 6.3629,
+      "step": 9950
+    },
+    {
+      "epoch": 0.0139947344109232,
+      "grad_norm": 1.286419153213501,
+      "learning_rate": 4.1979766755655475e-05,
+      "loss": 6.4067,
+      "step": 9960
+    },
+    {
+      "epoch": 0.014008785349086776,
+      "grad_norm": 1.2718861103057861,
+      "learning_rate": 4.20219193480399e-05,
+      "loss": 6.3202,
+      "step": 9970
+    },
+    {
+      "epoch": 0.014022836287250354,
+      "grad_norm": 1.3436089754104614,
+      "learning_rate": 4.206407194042433e-05,
+      "loss": 6.219,
+      "step": 9980
+    },
+    {
+      "epoch": 0.014036887225413932,
+      "grad_norm": 1.2512117624282837,
+      "learning_rate": 4.210622453280877e-05,
+      "loss": 6.2775,
+      "step": 9990
+    },
+    {
+      "epoch": 0.01405093816357751,
+      "grad_norm": 1.3441169261932373,
+      "learning_rate": 4.214837712519319e-05,
+      "loss": 6.3914,
+      "step": 10000
+    },
+    {
+      "epoch": 0.014064989101741088,
+      "grad_norm": 1.3099346160888672,
+      "learning_rate": 4.219052971757763e-05,
+      "loss": 6.3451,
+      "step": 10010
+    },
+    {
+      "epoch": 0.014079040039904664,
+      "grad_norm": 1.3395233154296875,
+      "learning_rate": 4.223268230996206e-05,
+      "loss": 6.3383,
+      "step": 10020
+    },
+    {
+      "epoch": 0.014093090978068242,
+      "grad_norm": 1.3763487339019775,
+      "learning_rate": 4.2274834902346486e-05,
+      "loss": 6.289,
+      "step": 10030
+    },
+    {
+      "epoch": 0.01410714191623182,
+      "grad_norm": 1.429871916770935,
+      "learning_rate": 4.2316987494730924e-05,
+      "loss": 6.2917,
+      "step": 10040
+    },
+    {
+      "epoch": 0.014121192854395398,
+      "grad_norm": 1.394494891166687,
+      "learning_rate": 4.235492482787691e-05,
+      "loss": 6.3043,
+      "step": 10050
+    },
+    {
+      "epoch": 0.014135243792558974,
+      "grad_norm": 1.4214140176773071,
+      "learning_rate": 4.239707742026134e-05,
+      "loss": 6.3363,
+      "step": 10060
+    },
+    {
+      "epoch": 0.014149294730722552,
+      "grad_norm": 1.2173434495925903,
+      "learning_rate": 4.2439230012645776e-05,
+      "loss": 6.3484,
+      "step": 10070
+    },
+    {
+      "epoch": 0.01416334566888613,
+      "grad_norm": 1.3638588190078735,
+      "learning_rate": 4.248138260503021e-05,
+      "loss": 6.3495,
+      "step": 10080
+    },
+    {
+      "epoch": 0.014177396607049707,
+      "grad_norm": 1.300630807876587,
+      "learning_rate": 4.252353519741464e-05,
+      "loss": 6.3795,
+      "step": 10090
+    },
+    {
+      "epoch": 0.014191447545213284,
+      "grad_norm": 1.279421329498291,
+      "learning_rate": 4.256568778979907e-05,
+      "loss": 6.3479,
+      "step": 10100
+    },
+    {
+      "epoch": 0.014205498483376862,
+      "grad_norm": 1.2923552989959717,
+      "learning_rate": 4.26078403821835e-05,
+      "loss": 6.3892,
+      "step": 10110
+    },
+    {
+      "epoch": 0.01421954942154044,
+      "grad_norm": 1.2862358093261719,
+      "learning_rate": 4.264999297456793e-05,
+      "loss": 6.3228,
+      "step": 10120
+    },
+    {
+      "epoch": 0.014233600359704017,
+      "grad_norm": 1.3364365100860596,
+      "learning_rate": 4.269214556695236e-05,
+      "loss": 6.4024,
+      "step": 10130
+    },
+    {
+      "epoch": 0.014247651297867595,
+      "grad_norm": 1.2868720293045044,
+      "learning_rate": 4.27342981593368e-05,
+      "loss": 6.2095,
+      "step": 10140
+    },
+    {
+      "epoch": 0.014261702236031171,
+      "grad_norm": 1.306137204170227,
+      "learning_rate": 4.2776450751721225e-05,
+      "loss": 6.3935,
+      "step": 10150
+    },
+    {
+      "epoch": 0.01427575317419475,
+      "grad_norm": 1.35478675365448,
+      "learning_rate": 4.2818603344105656e-05,
+      "loss": 6.3092,
+      "step": 10160
+    },
+    {
+      "epoch": 0.014289804112358327,
+      "grad_norm": 1.340928316116333,
+      "learning_rate": 4.2860755936490094e-05,
+      "loss": 6.2858,
+      "step": 10170
+    },
+    {
+      "epoch": 0.014303855050521905,
+      "grad_norm": 1.4878336191177368,
+      "learning_rate": 4.290290852887452e-05,
+      "loss": 6.297,
+      "step": 10180
+    },
+    {
+      "epoch": 0.014317905988685481,
+      "grad_norm": 1.3728898763656616,
+      "learning_rate": 4.294506112125895e-05,
+      "loss": 6.2908,
+      "step": 10190
+    },
+    {
+      "epoch": 0.01433195692684906,
+      "grad_norm": 1.3865792751312256,
+      "learning_rate": 4.298721371364339e-05,
+      "loss": 6.3898,
+      "step": 10200
+    },
+    {
+      "epoch": 0.014346007865012637,
+      "grad_norm": 1.3288946151733398,
+      "learning_rate": 4.302936630602782e-05,
+      "loss": 6.2661,
+      "step": 10210
+    },
+    {
+      "epoch": 0.014360058803176215,
+      "grad_norm": 1.5158910751342773,
+      "learning_rate": 4.306730363917381e-05,
+      "loss": 6.2648,
+      "step": 10220
+    },
+    {
+      "epoch": 0.014374109741339793,
+      "grad_norm": 1.41054368019104,
+      "learning_rate": 4.310945623155824e-05,
+      "loss": 6.3832,
+      "step": 10230
+    },
+    {
+      "epoch": 0.014388160679503369,
+      "grad_norm": 1.5252634286880493,
+      "learning_rate": 4.3151608823942664e-05,
+      "loss": 6.3621,
+      "step": 10240
+    },
+    {
+      "epoch": 0.014402211617666947,
+      "grad_norm": 1.2399297952651978,
+      "learning_rate": 4.31937614163271e-05,
+      "loss": 6.3802,
+      "step": 10250
+    },
+    {
+      "epoch": 0.014416262555830525,
+      "grad_norm": 1.2246840000152588,
+      "learning_rate": 4.323591400871153e-05,
+      "loss": 6.3442,
+      "step": 10260
+    },
+    {
+      "epoch": 0.014430313493994103,
+      "grad_norm": 1.349806785583496,
+      "learning_rate": 4.327806660109597e-05,
+      "loss": 6.2469,
+      "step": 10270
+    },
+    {
+      "epoch": 0.014444364432157679,
+      "grad_norm": 1.199388027191162,
+      "learning_rate": 4.3320219193480395e-05,
+      "loss": 6.3044,
+      "step": 10280
+    },
+    {
+      "epoch": 0.014458415370321257,
+      "grad_norm": 1.371137261390686,
+      "learning_rate": 4.3362371785864826e-05,
+      "loss": 6.2964,
+      "step": 10290
+    },
+    {
+      "epoch": 0.014472466308484835,
+      "grad_norm": 1.3206599950790405,
+      "learning_rate": 4.3404524378249264e-05,
+      "loss": 6.3184,
+      "step": 10300
+    },
+    {
+      "epoch": 0.014486517246648413,
+      "grad_norm": 1.4087791442871094,
+      "learning_rate": 4.344667697063369e-05,
+      "loss": 6.2876,
+      "step": 10310
+    },
+    {
+      "epoch": 0.01450056818481199,
+      "grad_norm": 1.4454388618469238,
+      "learning_rate": 4.348882956301812e-05,
+      "loss": 6.2846,
+      "step": 10320
+    },
+    {
+      "epoch": 0.014514619122975567,
+      "grad_norm": 1.414621353149414,
+      "learning_rate": 4.353098215540256e-05,
+      "loss": 6.2277,
+      "step": 10330
+    },
+    {
+      "epoch": 0.014528670061139145,
+      "grad_norm": 1.3868155479431152,
+      "learning_rate": 4.357313474778698e-05,
+      "loss": 6.2959,
+      "step": 10340
+    },
+    {
+      "epoch": 0.014542720999302722,
+      "grad_norm": 1.3480900526046753,
+      "learning_rate": 4.361528734017142e-05,
+      "loss": 6.2991,
+      "step": 10350
+    },
+    {
+      "epoch": 0.0145567719374663,
+      "grad_norm": 1.3821889162063599,
+      "learning_rate": 4.365743993255585e-05,
+      "loss": 6.2858,
+      "step": 10360
+    },
+    {
+      "epoch": 0.014570822875629877,
+      "grad_norm": 1.3947319984436035,
+      "learning_rate": 4.3699592524940276e-05,
+      "loss": 6.2764,
+      "step": 10370
+    },
+    {
+      "epoch": 0.014584873813793454,
+      "grad_norm": 1.4054744243621826,
+      "learning_rate": 4.3741745117324714e-05,
+      "loss": 6.3224,
+      "step": 10380
+    },
+    {
+      "epoch": 0.014598924751957032,
+      "grad_norm": 1.2518731355667114,
+      "learning_rate": 4.3783897709709145e-05,
+      "loss": 6.3417,
+      "step": 10390
+    },
+    {
+      "epoch": 0.01461297569012061,
+      "grad_norm": 1.4203547239303589,
+      "learning_rate": 4.382605030209357e-05,
+      "loss": 6.3121,
+      "step": 10400
+    },
+    {
+      "epoch": 0.014627026628284186,
+      "grad_norm": 1.4533650875091553,
+      "learning_rate": 4.386820289447801e-05,
+      "loss": 6.2781,
+      "step": 10410
+    },
+    {
+      "epoch": 0.014641077566447764,
+      "grad_norm": 1.4527699947357178,
+      "learning_rate": 4.391035548686244e-05,
+      "loss": 6.3558,
+      "step": 10420
+    },
+    {
+      "epoch": 0.014655128504611342,
+      "grad_norm": 1.324060082435608,
+      "learning_rate": 4.3952508079246876e-05,
+      "loss": 6.3507,
+      "step": 10430
+    },
+    {
+      "epoch": 0.01466917944277492,
+      "grad_norm": 1.4087752103805542,
+      "learning_rate": 4.39946606716313e-05,
+      "loss": 6.3084,
+      "step": 10440
+    },
+    {
+      "epoch": 0.014683230380938498,
+      "grad_norm": 1.2102359533309937,
+      "learning_rate": 4.403681326401573e-05,
+      "loss": 6.3034,
+      "step": 10450
+    },
+    {
+      "epoch": 0.014697281319102074,
+      "grad_norm": 1.3251889944076538,
+      "learning_rate": 4.407896585640017e-05,
+      "loss": 6.2459,
+      "step": 10460
+    },
+    {
+      "epoch": 0.014711332257265652,
+      "grad_norm": 1.3922446966171265,
+      "learning_rate": 4.4121118448784594e-05,
+      "loss": 6.3356,
+      "step": 10470
+    },
+    {
+      "epoch": 0.01472538319542923,
+      "grad_norm": 1.269550085067749,
+      "learning_rate": 4.416327104116903e-05,
+      "loss": 6.2387,
+      "step": 10480
+    },
+    {
+      "epoch": 0.014739434133592808,
+      "grad_norm": 1.3506200313568115,
+      "learning_rate": 4.420542363355346e-05,
+      "loss": 6.2655,
+      "step": 10490
+    },
+    {
+      "epoch": 0.014753485071756384,
+      "grad_norm": 1.3239065408706665,
+      "learning_rate": 4.424757622593789e-05,
+      "loss": 6.267,
+      "step": 10500
+    },
+    {
+      "epoch": 0.014767536009919962,
+      "grad_norm": 1.2084630727767944,
+      "learning_rate": 4.4289728818322325e-05,
+      "loss": 6.3075,
+      "step": 10510
+    },
+    {
+      "epoch": 0.01478158694808354,
+      "grad_norm": 1.3695273399353027,
+      "learning_rate": 4.433188141070676e-05,
+      "loss": 6.3784,
+      "step": 10520
+    },
+    {
+      "epoch": 0.014795637886247118,
+      "grad_norm": 1.4278502464294434,
+      "learning_rate": 4.437403400309118e-05,
+      "loss": 6.2445,
+      "step": 10530
+    },
+    {
+      "epoch": 0.014809688824410696,
+      "grad_norm": 1.606279969215393,
+      "learning_rate": 4.441618659547562e-05,
+      "loss": 6.2251,
+      "step": 10540
+    },
+    {
+      "epoch": 0.014823739762574272,
+      "grad_norm": 1.283535122871399,
+      "learning_rate": 4.445833918786005e-05,
+      "loss": 6.2737,
+      "step": 10550
+    },
+    {
+      "epoch": 0.01483779070073785,
+      "grad_norm": 1.2936784029006958,
+      "learning_rate": 4.450049178024449e-05,
+      "loss": 6.3372,
+      "step": 10560
+    },
+    {
+      "epoch": 0.014851841638901428,
+      "grad_norm": 1.4122823476791382,
+      "learning_rate": 4.454264437262891e-05,
+      "loss": 6.3038,
+      "step": 10570
+    },
+    {
+      "epoch": 0.014865892577065005,
+      "grad_norm": 1.2987902164459229,
+      "learning_rate": 4.4584796965013344e-05,
+      "loss": 6.2672,
+      "step": 10580
+    },
+    {
+      "epoch": 0.014879943515228582,
+      "grad_norm": 1.2851322889328003,
+      "learning_rate": 4.462694955739778e-05,
+      "loss": 6.2606,
+      "step": 10590
+    },
+    {
+      "epoch": 0.01489399445339216,
+      "grad_norm": 1.359154224395752,
+      "learning_rate": 4.4669102149782206e-05,
+      "loss": 6.3153,
+      "step": 10600
+    },
+    {
+      "epoch": 0.014908045391555737,
+      "grad_norm": 1.3480968475341797,
+      "learning_rate": 4.471125474216664e-05,
+      "loss": 6.3336,
+      "step": 10610
+    },
+    {
+      "epoch": 0.014922096329719315,
+      "grad_norm": 1.3098258972167969,
+      "learning_rate": 4.4753407334551075e-05,
+      "loss": 6.302,
+      "step": 10620
+    },
+    {
+      "epoch": 0.014936147267882893,
+      "grad_norm": 1.3076599836349487,
+      "learning_rate": 4.47955599269355e-05,
+      "loss": 6.3044,
+      "step": 10630
+    },
+    {
+      "epoch": 0.01495019820604647,
+      "grad_norm": 1.2845678329467773,
+      "learning_rate": 4.483771251931994e-05,
+      "loss": 6.2815,
+      "step": 10640
+    },
+    {
+      "epoch": 0.014964249144210047,
+      "grad_norm": 1.2279037237167358,
+      "learning_rate": 4.487986511170437e-05,
+      "loss": 6.2752,
+      "step": 10650
+    },
+    {
+      "epoch": 0.014978300082373625,
+      "grad_norm": 1.3851795196533203,
+      "learning_rate": 4.492201770408879e-05,
+      "loss": 6.2336,
+      "step": 10660
+    },
+    {
+      "epoch": 0.014992351020537203,
+      "grad_norm": 1.5091443061828613,
+      "learning_rate": 4.496417029647323e-05,
+      "loss": 6.1372,
+      "step": 10670
+    },
+    {
+      "epoch": 0.01500640195870078,
+      "grad_norm": 1.263030767440796,
+      "learning_rate": 4.500632288885766e-05,
+      "loss": 6.2136,
+      "step": 10680
+    },
+    {
+      "epoch": 0.015020452896864357,
+      "grad_norm": 1.508408546447754,
+      "learning_rate": 4.5048475481242086e-05,
+      "loss": 6.2105,
+      "step": 10690
+    },
+    {
+      "epoch": 0.015034503835027935,
+      "grad_norm": 1.5163711309432983,
+      "learning_rate": 4.5090628073626524e-05,
+      "loss": 6.1942,
+      "step": 10700
+    },
+    {
+      "epoch": 0.015048554773191513,
+      "grad_norm": 1.3551746606826782,
+      "learning_rate": 4.5132780666010956e-05,
+      "loss": 6.1646,
+      "step": 10710
+    },
+    {
+      "epoch": 0.015062605711355091,
+      "grad_norm": 1.3781139850616455,
+      "learning_rate": 4.5174933258395394e-05,
+      "loss": 6.2178,
+      "step": 10720
+    },
+    {
+      "epoch": 0.015076656649518667,
+      "grad_norm": 1.3036412000656128,
+      "learning_rate": 4.521708585077982e-05,
+      "loss": 6.2511,
+      "step": 10730
+    },
+    {
+      "epoch": 0.015090707587682245,
+      "grad_norm": 1.3681645393371582,
+      "learning_rate": 4.525923844316425e-05,
+      "loss": 6.1301,
+      "step": 10740
+    },
+    {
+      "epoch": 0.015104758525845823,
+      "grad_norm": 1.3642160892486572,
+      "learning_rate": 4.530139103554869e-05,
+      "loss": 6.228,
+      "step": 10750
+    },
+    {
+      "epoch": 0.0151188094640094,
+      "grad_norm": 1.410559058189392,
+      "learning_rate": 4.534354362793311e-05,
+      "loss": 6.2981,
+      "step": 10760
+    },
+    {
+      "epoch": 0.015132860402172977,
+      "grad_norm": 1.3881288766860962,
+      "learning_rate": 4.538569622031755e-05,
+      "loss": 6.1561,
+      "step": 10770
+    },
+    {
+      "epoch": 0.015146911340336555,
+      "grad_norm": 1.231331467628479,
+      "learning_rate": 4.542784881270198e-05,
+      "loss": 6.2037,
+      "step": 10780
+    },
+    {
+      "epoch": 0.015160962278500133,
+      "grad_norm": 1.5481162071228027,
+      "learning_rate": 4.5470001405086405e-05,
+      "loss": 6.1547,
+      "step": 10790
+    },
+    {
+      "epoch": 0.01517501321666371,
+      "grad_norm": 1.3833677768707275,
+      "learning_rate": 4.551215399747084e-05,
+      "loss": 6.2681,
+      "step": 10800
+    },
+    {
+      "epoch": 0.015189064154827287,
+      "grad_norm": 1.297667145729065,
+      "learning_rate": 4.5554306589855274e-05,
+      "loss": 6.1601,
+      "step": 10810
+    },
+    {
+      "epoch": 0.015203115092990865,
+      "grad_norm": 1.2731614112854004,
+      "learning_rate": 4.55964591822397e-05,
+      "loss": 6.2338,
+      "step": 10820
+    },
+    {
+      "epoch": 0.015217166031154443,
+      "grad_norm": 1.392951488494873,
+      "learning_rate": 4.5638611774624136e-05,
+      "loss": 6.223,
+      "step": 10830
+    },
+    {
+      "epoch": 0.01523121696931802,
+      "grad_norm": 1.152622103691101,
+      "learning_rate": 4.568076436700857e-05,
+      "loss": 6.1884,
+      "step": 10840
+    },
+    {
+      "epoch": 0.015245267907481598,
+      "grad_norm": 1.287348985671997,
+      "learning_rate": 4.5722916959393005e-05,
+      "loss": 6.1756,
+      "step": 10850
+    },
+    {
+      "epoch": 0.015259318845645175,
+      "grad_norm": 1.3075429201126099,
+      "learning_rate": 4.576506955177743e-05,
+      "loss": 6.1748,
+      "step": 10860
+    },
+    {
+      "epoch": 0.015273369783808752,
+      "grad_norm": 1.3519240617752075,
+      "learning_rate": 4.580722214416186e-05,
+      "loss": 6.2611,
+      "step": 10870
+    },
+    {
+      "epoch": 0.01528742072197233,
+      "grad_norm": 1.257500410079956,
+      "learning_rate": 4.58493747365463e-05,
+      "loss": 6.1883,
+      "step": 10880
+    },
+    {
+      "epoch": 0.015301471660135908,
+      "grad_norm": 1.276843786239624,
+      "learning_rate": 4.589152732893072e-05,
+      "loss": 6.1289,
+      "step": 10890
+    },
+    {
+      "epoch": 0.015315522598299484,
+      "grad_norm": 1.397799015045166,
+      "learning_rate": 4.5933679921315155e-05,
+      "loss": 6.1006,
+      "step": 10900
+    },
+    {
+      "epoch": 0.015329573536463062,
+      "grad_norm": 1.3642454147338867,
+      "learning_rate": 4.597583251369959e-05,
+      "loss": 6.179,
+      "step": 10910
+    },
+    {
+      "epoch": 0.01534362447462664,
+      "grad_norm": 1.406509280204773,
+      "learning_rate": 4.601798510608402e-05,
+      "loss": 6.1252,
+      "step": 10920
+    },
+    {
+      "epoch": 0.015357675412790218,
+      "grad_norm": 1.4512871503829956,
+      "learning_rate": 4.6060137698468455e-05,
+      "loss": 6.2659,
+      "step": 10930
+    },
+    {
+      "epoch": 0.015371726350953796,
+      "grad_norm": 1.332694172859192,
+      "learning_rate": 4.6102290290852886e-05,
+      "loss": 6.2391,
+      "step": 10940
+    },
+    {
+      "epoch": 0.015385777289117372,
+      "grad_norm": 1.3527559041976929,
+      "learning_rate": 4.614444288323731e-05,
+      "loss": 6.2161,
+      "step": 10950
+    },
+    {
+      "epoch": 0.01539982822728095,
+      "grad_norm": 1.3626435995101929,
+      "learning_rate": 4.618659547562175e-05,
+      "loss": 6.1904,
+      "step": 10960
+    },
+    {
+      "epoch": 0.015413879165444528,
+      "grad_norm": 1.4977718591690063,
+      "learning_rate": 4.622874806800618e-05,
+      "loss": 6.1898,
+      "step": 10970
+    },
+    {
+      "epoch": 0.015427930103608106,
+      "grad_norm": 1.512209415435791,
+      "learning_rate": 4.627090066039062e-05,
+      "loss": 6.139,
+      "step": 10980
+    },
+    {
+      "epoch": 0.015441981041771682,
+      "grad_norm": 1.363409161567688,
+      "learning_rate": 4.631305325277504e-05,
+      "loss": 6.1603,
+      "step": 10990
+    },
+    {
+      "epoch": 0.01545603197993526,
+      "grad_norm": 1.5099941492080688,
+      "learning_rate": 4.635520584515947e-05,
+      "loss": 6.1706,
+      "step": 11000
+    },
+    {
+      "epoch": 0.015470082918098838,
+      "grad_norm": 1.9168140888214111,
+      "learning_rate": 4.639735843754391e-05,
+      "loss": 6.2084,
+      "step": 11010
+    },
+    {
+      "epoch": 0.015484133856262416,
+      "grad_norm": 1.2205970287322998,
+      "learning_rate": 4.6439511029928335e-05,
+      "loss": 6.252,
+      "step": 11020
+    },
+    {
+      "epoch": 0.015498184794425994,
+      "grad_norm": 1.5673115253448486,
+      "learning_rate": 4.6481663622312766e-05,
+      "loss": 6.0759,
+      "step": 11030
+    },
+    {
+      "epoch": 0.01551223573258957,
+      "grad_norm": 1.4149396419525146,
+      "learning_rate": 4.6523816214697204e-05,
+      "loss": 6.2689,
+      "step": 11040
+    },
+    {
+      "epoch": 0.015526286670753148,
+      "grad_norm": 1.3954936265945435,
+      "learning_rate": 4.656596880708163e-05,
+      "loss": 6.2501,
+      "step": 11050
+    },
+    {
+      "epoch": 0.015540337608916726,
+      "grad_norm": 1.5006779432296753,
+      "learning_rate": 4.660812139946607e-05,
+      "loss": 6.1785,
+      "step": 11060
+    },
+    {
+      "epoch": 0.015554388547080304,
+      "grad_norm": 1.3999886512756348,
+      "learning_rate": 4.66502739918505e-05,
+      "loss": 6.15,
+      "step": 11070
+    },
+    {
+      "epoch": 0.01556843948524388,
+      "grad_norm": 2.2167131900787354,
+      "learning_rate": 4.669242658423492e-05,
+      "loss": 6.132,
+      "step": 11080
+    },
+    {
+      "epoch": 0.015582490423407458,
+      "grad_norm": 1.401848316192627,
+      "learning_rate": 4.673457917661936e-05,
+      "loss": 6.281,
+      "step": 11090
+    },
+    {
+      "epoch": 0.015596541361571035,
+      "grad_norm": 1.3416310548782349,
+      "learning_rate": 4.677673176900379e-05,
+      "loss": 6.1515,
+      "step": 11100
+    },
+    {
+      "epoch": 0.015610592299734613,
+      "grad_norm": 1.5161608457565308,
+      "learning_rate": 4.6818884361388216e-05,
+      "loss": 6.1189,
+      "step": 11110
+    },
+    {
+      "epoch": 0.01562464323789819,
+      "grad_norm": 1.393735408782959,
+      "learning_rate": 4.6861036953772654e-05,
+      "loss": 6.2588,
+      "step": 11120
+    },
+    {
+      "epoch": 0.015638694176061767,
+      "grad_norm": 1.3404139280319214,
+      "learning_rate": 4.6903189546157085e-05,
+      "loss": 6.234,
+      "step": 11130
+    },
+    {
+      "epoch": 0.015652745114225347,
+      "grad_norm": 1.372836947441101,
+      "learning_rate": 4.694534213854152e-05,
+      "loss": 6.2106,
+      "step": 11140
+    },
+    {
+      "epoch": 0.015666796052388923,
+      "grad_norm": 1.3898221254348755,
+      "learning_rate": 4.698749473092595e-05,
+      "loss": 6.0849,
+      "step": 11150
+    },
+    {
+      "epoch": 0.0156808469905525,
+      "grad_norm": 1.336380124092102,
+      "learning_rate": 4.702964732331038e-05,
+      "loss": 6.1996,
+      "step": 11160
+    },
+    {
+      "epoch": 0.01569489792871608,
+      "grad_norm": 1.478607177734375,
+      "learning_rate": 4.7071799915694816e-05,
+      "loss": 6.1785,
+      "step": 11170
+    },
+    {
+      "epoch": 0.015708948866879655,
+      "grad_norm": 1.5594801902770996,
+      "learning_rate": 4.711395250807924e-05,
+      "loss": 6.2302,
+      "step": 11180
+    },
+    {
+      "epoch": 0.01572299980504323,
+      "grad_norm": 1.377120852470398,
+      "learning_rate": 4.715610510046367e-05,
+      "loss": 6.2416,
+      "step": 11190
+    },
+    {
+      "epoch": 0.01573705074320681,
+      "grad_norm": 1.5377556085586548,
+      "learning_rate": 4.719825769284811e-05,
+      "loss": 6.1337,
+      "step": 11200
+    },
+    {
+      "epoch": 0.015751101681370387,
+      "grad_norm": 1.7049307823181152,
+      "learning_rate": 4.7240410285232534e-05,
+      "loss": 6.1986,
+      "step": 11210
+    },
+    {
+      "epoch": 0.015765152619533967,
+      "grad_norm": 1.3457587957382202,
+      "learning_rate": 4.728256287761697e-05,
+      "loss": 6.0877,
+      "step": 11220
+    },
+    {
+      "epoch": 0.015779203557697543,
+      "grad_norm": 1.3759392499923706,
+      "learning_rate": 4.73247154700014e-05,
+      "loss": 6.1598,
+      "step": 11230
+    },
+    {
+      "epoch": 0.01579325449586112,
+      "grad_norm": 1.3922202587127686,
+      "learning_rate": 4.736686806238583e-05,
+      "loss": 6.2017,
+      "step": 11240
+    },
+    {
+      "epoch": 0.0158073054340247,
+      "grad_norm": 1.3885234594345093,
+      "learning_rate": 4.7409020654770266e-05,
+      "loss": 6.2302,
+      "step": 11250
+    },
+    {
+      "epoch": 0.015821356372188275,
+      "grad_norm": 1.4035016298294067,
+      "learning_rate": 4.74511732471547e-05,
+      "loss": 6.204,
+      "step": 11260
+    },
+    {
+      "epoch": 0.015835407310351855,
+      "grad_norm": 1.3315571546554565,
+      "learning_rate": 4.749332583953913e-05,
+      "loss": 6.1677,
+      "step": 11270
+    },
+    {
+      "epoch": 0.01584945824851543,
+      "grad_norm": 1.2895710468292236,
+      "learning_rate": 4.753547843192356e-05,
+      "loss": 6.2274,
+      "step": 11280
+    },
+    {
+      "epoch": 0.015863509186679007,
+      "grad_norm": 1.3065532445907593,
+      "learning_rate": 4.757763102430799e-05,
+      "loss": 6.0836,
+      "step": 11290
+    },
+    {
+      "epoch": 0.015877560124842587,
+      "grad_norm": 1.392031192779541,
+      "learning_rate": 4.761978361669243e-05,
+      "loss": 6.1599,
+      "step": 11300
+    },
+    {
+      "epoch": 0.015891611063006163,
+      "grad_norm": 1.3048975467681885,
+      "learning_rate": 4.766193620907685e-05,
+      "loss": 6.1467,
+      "step": 11310
+    },
+    {
+      "epoch": 0.01590566200116974,
+      "grad_norm": 1.5152511596679688,
+      "learning_rate": 4.7704088801461284e-05,
+      "loss": 6.0913,
+      "step": 11320
+    },
+    {
+      "epoch": 0.01591971293933332,
+      "grad_norm": 1.3078099489212036,
+      "learning_rate": 4.774624139384572e-05,
+      "loss": 6.0885,
+      "step": 11330
+    },
+    {
+      "epoch": 0.015933763877496895,
+      "grad_norm": 1.2487436532974243,
+      "learning_rate": 4.7788393986230146e-05,
+      "loss": 6.2689,
+      "step": 11340
+    },
+    {
+      "epoch": 0.015947814815660474,
+      "grad_norm": 1.5608822107315063,
+      "learning_rate": 4.7830546578614584e-05,
+      "loss": 6.1847,
+      "step": 11350
+    },
+    {
+      "epoch": 0.01596186575382405,
+      "grad_norm": 1.7839996814727783,
+      "learning_rate": 4.7872699170999015e-05,
+      "loss": 6.0849,
+      "step": 11360
+    },
+    {
+      "epoch": 0.015975916691987627,
+      "grad_norm": 1.5817362070083618,
+      "learning_rate": 4.791485176338344e-05,
+      "loss": 6.2027,
+      "step": 11370
+    },
+    {
+      "epoch": 0.015989967630151206,
+      "grad_norm": 1.7094651460647583,
+      "learning_rate": 4.795700435576788e-05,
+      "loss": 6.1865,
+      "step": 11380
+    },
+    {
+      "epoch": 0.016004018568314782,
+      "grad_norm": 1.3029943704605103,
+      "learning_rate": 4.799915694815231e-05,
+      "loss": 6.1732,
+      "step": 11390
+    },
+    {
+      "epoch": 0.016018069506478362,
+      "grad_norm": 1.6077370643615723,
+      "learning_rate": 4.804130954053673e-05,
+      "loss": 6.2421,
+      "step": 11400
+    },
+    {
+      "epoch": 0.016032120444641938,
+      "grad_norm": 1.7310693264007568,
+      "learning_rate": 4.808346213292117e-05,
+      "loss": 6.0574,
+      "step": 11410
+    },
+    {
+      "epoch": 0.016046171382805514,
+      "grad_norm": 1.3837238550186157,
+      "learning_rate": 4.81256147253056e-05,
+      "loss": 6.1435,
+      "step": 11420
+    },
+    {
+      "epoch": 0.016060222320969094,
+      "grad_norm": 1.3638036251068115,
+      "learning_rate": 4.816776731769004e-05,
+      "loss": 6.0729,
+      "step": 11430
+    },
+    {
+      "epoch": 0.01607427325913267,
+      "grad_norm": 1.391371726989746,
+      "learning_rate": 4.8209919910074465e-05,
+      "loss": 6.0995,
+      "step": 11440
+    },
+    {
+      "epoch": 0.01608832419729625,
+      "grad_norm": 1.3891973495483398,
+      "learning_rate": 4.8252072502458896e-05,
+      "loss": 6.0128,
+      "step": 11450
+    },
+    {
+      "epoch": 0.016102375135459826,
+      "grad_norm": 1.384169578552246,
+      "learning_rate": 4.8294225094843334e-05,
+      "loss": 6.2458,
+      "step": 11460
+    },
+    {
+      "epoch": 0.016116426073623402,
+      "grad_norm": 1.3138374090194702,
+      "learning_rate": 4.833637768722776e-05,
+      "loss": 6.1883,
+      "step": 11470
+    },
+    {
+      "epoch": 0.016130477011786982,
+      "grad_norm": 1.3187780380249023,
+      "learning_rate": 4.8378530279612196e-05,
+      "loss": 6.128,
+      "step": 11480
+    },
+    {
+      "epoch": 0.016144527949950558,
+      "grad_norm": 1.3277584314346313,
+      "learning_rate": 4.842068287199663e-05,
+      "loss": 6.2424,
+      "step": 11490
+    },
+    {
+      "epoch": 0.016158578888114134,
+      "grad_norm": 1.2720946073532104,
+      "learning_rate": 4.846283546438105e-05,
+      "loss": 6.2401,
+      "step": 11500
+    },
+    {
+      "epoch": 0.016172629826277714,
+      "grad_norm": 1.4264894723892212,
+      "learning_rate": 4.850498805676549e-05,
+      "loss": 6.1807,
+      "step": 11510
+    },
+    {
+      "epoch": 0.01618668076444129,
+      "grad_norm": 1.3776253461837769,
+      "learning_rate": 4.854714064914992e-05,
+      "loss": 6.2413,
+      "step": 11520
+    },
+    {
+      "epoch": 0.01620073170260487,
+      "grad_norm": 1.2899820804595947,
+      "learning_rate": 4.8589293241534345e-05,
+      "loss": 6.1676,
+      "step": 11530
+    },
+    {
+      "epoch": 0.016214782640768446,
+      "grad_norm": 1.3107612133026123,
+      "learning_rate": 4.863144583391878e-05,
+      "loss": 6.1658,
+      "step": 11540
+    },
+    {
+      "epoch": 0.016228833578932022,
+      "grad_norm": 1.4616020917892456,
+      "learning_rate": 4.8673598426303214e-05,
+      "loss": 6.1274,
+      "step": 11550
+    },
+    {
+      "epoch": 0.0162428845170956,
+      "grad_norm": 1.3054602146148682,
+      "learning_rate": 4.8715751018687645e-05,
+      "loss": 6.119,
+      "step": 11560
+    },
+    {
+      "epoch": 0.016256935455259178,
+      "grad_norm": 1.3119100332260132,
+      "learning_rate": 4.8757903611072077e-05,
+      "loss": 6.0736,
+      "step": 11570
+    },
+    {
+      "epoch": 0.016270986393422757,
+      "grad_norm": 1.3544719219207764,
+      "learning_rate": 4.880005620345651e-05,
+      "loss": 6.1392,
+      "step": 11580
+    },
+    {
+      "epoch": 0.016285037331586334,
+      "grad_norm": 1.3637272119522095,
+      "learning_rate": 4.8842208795840946e-05,
+      "loss": 6.0982,
+      "step": 11590
+    },
+    {
+      "epoch": 0.01629908826974991,
+      "grad_norm": 1.431994915008545,
+      "learning_rate": 4.888436138822537e-05,
+      "loss": 6.07,
+      "step": 11600
+    },
+    {
+      "epoch": 0.01631313920791349,
+      "grad_norm": 1.322766661643982,
+      "learning_rate": 4.89265139806098e-05,
+      "loss": 6.2458,
+      "step": 11610
+    },
+    {
+      "epoch": 0.016327190146077065,
+      "grad_norm": 1.5207664966583252,
+      "learning_rate": 4.896866657299424e-05,
+      "loss": 6.0894,
+      "step": 11620
+    },
+    {
+      "epoch": 0.016341241084240645,
+      "grad_norm": 1.950318455696106,
+      "learning_rate": 4.9010819165378664e-05,
+      "loss": 6.112,
+      "step": 11630
+    },
+    {
+      "epoch": 0.01635529202240422,
+      "grad_norm": 1.8058013916015625,
+      "learning_rate": 4.90529717577631e-05,
+      "loss": 6.1553,
+      "step": 11640
+    },
+    {
+      "epoch": 0.016369342960567797,
+      "grad_norm": 1.2825206518173218,
+      "learning_rate": 4.909512435014753e-05,
+      "loss": 6.185,
+      "step": 11650
+    },
+    {
+      "epoch": 0.016383393898731377,
+      "grad_norm": 1.4293164014816284,
+      "learning_rate": 4.913727694253196e-05,
+      "loss": 6.2412,
+      "step": 11660
+    },
+    {
+      "epoch": 0.016397444836894953,
+      "grad_norm": 1.2992738485336304,
+      "learning_rate": 4.9179429534916395e-05,
+      "loss": 6.1107,
+      "step": 11670
+    },
+    {
+      "epoch": 0.01641149577505853,
+      "grad_norm": 1.3839462995529175,
+      "learning_rate": 4.9221582127300826e-05,
+      "loss": 6.0917,
+      "step": 11680
+    },
+    {
+      "epoch": 0.01642554671322211,
+      "grad_norm": 1.3559750318527222,
+      "learning_rate": 4.926373471968525e-05,
+      "loss": 6.1816,
+      "step": 11690
+    },
+    {
+      "epoch": 0.016439597651385685,
+      "grad_norm": 1.3381038904190063,
+      "learning_rate": 4.930588731206969e-05,
+      "loss": 6.2072,
+      "step": 11700
+    },
+    {
+      "epoch": 0.016453648589549265,
+      "grad_norm": 1.3606219291687012,
+      "learning_rate": 4.934803990445412e-05,
+      "loss": 6.1239,
+      "step": 11710
+    },
+    {
+      "epoch": 0.01646769952771284,
+      "grad_norm": 1.2954307794570923,
+      "learning_rate": 4.939019249683856e-05,
+      "loss": 6.1274,
+      "step": 11720
+    },
+    {
+      "epoch": 0.016481750465876417,
+      "grad_norm": 1.3133169412612915,
+      "learning_rate": 4.943234508922298e-05,
+      "loss": 6.0767,
+      "step": 11730
+    },
+    {
+      "epoch": 0.016495801404039997,
+      "grad_norm": 1.675938367843628,
+      "learning_rate": 4.947449768160741e-05,
+      "loss": 6.1603,
+      "step": 11740
+    },
+    {
+      "epoch": 0.016509852342203573,
+      "grad_norm": 1.3132497072219849,
+      "learning_rate": 4.951665027399185e-05,
+      "loss": 6.1483,
+      "step": 11750
+    },
+    {
+      "epoch": 0.016523903280367153,
+      "grad_norm": 1.491671085357666,
+      "learning_rate": 4.9558802866376276e-05,
+      "loss": 6.0745,
+      "step": 11760
+    },
+    {
+      "epoch": 0.01653795421853073,
+      "grad_norm": 1.487144112586975,
+      "learning_rate": 4.9600955458760713e-05,
+      "loss": 6.0128,
+      "step": 11770
+    },
+    {
+      "epoch": 0.016552005156694305,
+      "grad_norm": 1.572177767753601,
+      "learning_rate": 4.9643108051145145e-05,
+      "loss": 6.1975,
+      "step": 11780
+    },
+    {
+      "epoch": 0.016566056094857885,
+      "grad_norm": 1.355470061302185,
+      "learning_rate": 4.968526064352957e-05,
+      "loss": 6.0389,
+      "step": 11790
+    },
+    {
+      "epoch": 0.01658010703302146,
+      "grad_norm": 1.437559962272644,
+      "learning_rate": 4.972741323591401e-05,
+      "loss": 6.0687,
+      "step": 11800
+    },
+    {
+      "epoch": 0.016594157971185037,
+      "grad_norm": 1.2300171852111816,
+      "learning_rate": 4.976956582829844e-05,
+      "loss": 6.119,
+      "step": 11810
+    },
+    {
+      "epoch": 0.016608208909348617,
+      "grad_norm": 1.3405951261520386,
+      "learning_rate": 4.981171842068286e-05,
+      "loss": 6.2048,
+      "step": 11820
+    },
+    {
+      "epoch": 0.016622259847512193,
+      "grad_norm": 1.3226433992385864,
+      "learning_rate": 4.98538710130673e-05,
+      "loss": 6.0897,
+      "step": 11830
+    },
+    {
+      "epoch": 0.016636310785675772,
+      "grad_norm": 1.3560805320739746,
+      "learning_rate": 4.989602360545173e-05,
+      "loss": 6.0075,
+      "step": 11840
+    },
+    {
+      "epoch": 0.01665036172383935,
+      "grad_norm": 1.4106473922729492,
+      "learning_rate": 4.993817619783616e-05,
+      "loss": 5.9296,
+      "step": 11850
+    },
+    {
+      "epoch": 0.016664412662002925,
+      "grad_norm": 1.2530484199523926,
+      "learning_rate": 4.9980328790220594e-05,
+      "loss": 6.0896,
+      "step": 11860
+    },
+    {
+      "epoch": 0.016678463600166504,
+      "grad_norm": 1.28843092918396,
+      "learning_rate": 5.0022481382605025e-05,
+      "loss": 6.1235,
+      "step": 11870
+    },
+    {
+      "epoch": 0.01669251453833008,
+      "grad_norm": 1.4062579870224,
+      "learning_rate": 5.006463397498946e-05,
+      "loss": 6.1136,
+      "step": 11880
+    },
+    {
+      "epoch": 0.01670656547649366,
+      "grad_norm": 1.275445818901062,
+      "learning_rate": 5.010678656737389e-05,
+      "loss": 6.0155,
+      "step": 11890
+    },
+    {
+      "epoch": 0.016720616414657236,
+      "grad_norm": 1.9027255773544312,
+      "learning_rate": 5.014893915975832e-05,
+      "loss": 6.1248,
+      "step": 11900
+    },
+    {
+      "epoch": 0.016734667352820812,
+      "grad_norm": 1.4187486171722412,
+      "learning_rate": 5.0191091752142757e-05,
+      "loss": 6.0323,
+      "step": 11910
+    },
+    {
+      "epoch": 0.016748718290984392,
+      "grad_norm": 1.3106961250305176,
+      "learning_rate": 5.023324434452718e-05,
+      "loss": 6.0743,
+      "step": 11920
+    },
+    {
+      "epoch": 0.016762769229147968,
+      "grad_norm": 1.2284607887268066,
+      "learning_rate": 5.027539693691162e-05,
+      "loss": 6.1276,
+      "step": 11930
+    },
+    {
+      "epoch": 0.016776820167311548,
+      "grad_norm": 1.403949499130249,
+      "learning_rate": 5.031754952929605e-05,
+      "loss": 6.0686,
+      "step": 11940
+    },
+    {
+      "epoch": 0.016790871105475124,
+      "grad_norm": 1.421419382095337,
+      "learning_rate": 5.0359702121680474e-05,
+      "loss": 6.241,
+      "step": 11950
+    },
+    {
+      "epoch": 0.0168049220436387,
+      "grad_norm": 1.5890074968338013,
+      "learning_rate": 5.040185471406491e-05,
+      "loss": 6.1183,
+      "step": 11960
+    },
+    {
+      "epoch": 0.01681897298180228,
+      "grad_norm": 1.38971745967865,
+      "learning_rate": 5.0444007306449344e-05,
+      "loss": 6.0829,
+      "step": 11970
+    },
+    {
+      "epoch": 0.016833023919965856,
+      "grad_norm": 1.4102171659469604,
+      "learning_rate": 5.0486159898833775e-05,
+      "loss": 6.1236,
+      "step": 11980
+    },
+    {
+      "epoch": 0.016847074858129432,
+      "grad_norm": 1.258589506149292,
+      "learning_rate": 5.0528312491218206e-05,
+      "loss": 6.1806,
+      "step": 11990
+    },
+    {
+      "epoch": 0.016861125796293012,
+      "grad_norm": 1.3471759557724,
+      "learning_rate": 5.057046508360264e-05,
+      "loss": 6.0465,
+      "step": 12000
+    },
+    {
+      "epoch": 0.016875176734456588,
+      "grad_norm": 1.5169402360916138,
+      "learning_rate": 5.0612617675987075e-05,
+      "loss": 6.1002,
+      "step": 12010
+    },
+    {
+      "epoch": 0.016889227672620168,
+      "grad_norm": 1.5497729778289795,
+      "learning_rate": 5.06547702683715e-05,
+      "loss": 6.1487,
+      "step": 12020
+    },
+    {
+      "epoch": 0.016903278610783744,
+      "grad_norm": 1.364227294921875,
+      "learning_rate": 5.069692286075593e-05,
+      "loss": 6.0852,
+      "step": 12030
+    },
+    {
+      "epoch": 0.01691732954894732,
+      "grad_norm": 1.3978562355041504,
+      "learning_rate": 5.073907545314037e-05,
+      "loss": 6.0646,
+      "step": 12040
+    },
+    {
+      "epoch": 0.0169313804871109,
+      "grad_norm": 1.2657595872879028,
+      "learning_rate": 5.078122804552479e-05,
+      "loss": 6.0511,
+      "step": 12050
+    },
+    {
+      "epoch": 0.016945431425274476,
+      "grad_norm": 1.472236156463623,
+      "learning_rate": 5.082338063790923e-05,
+      "loss": 6.1274,
+      "step": 12060
+    },
+    {
+      "epoch": 0.016959482363438055,
+      "grad_norm": 1.381701946258545,
+      "learning_rate": 5.086553323029366e-05,
+      "loss": 6.0854,
+      "step": 12070
+    },
+    {
+      "epoch": 0.01697353330160163,
+      "grad_norm": 1.3113288879394531,
+      "learning_rate": 5.0907685822678086e-05,
+      "loss": 6.1267,
+      "step": 12080
+    },
+    {
+      "epoch": 0.016987584239765208,
+      "grad_norm": 1.3406379222869873,
+      "learning_rate": 5.0949838415062524e-05,
+      "loss": 6.0556,
+      "step": 12090
+    },
+    {
+      "epoch": 0.017001635177928787,
+      "grad_norm": 1.3727220296859741,
+      "learning_rate": 5.0991991007446956e-05,
+      "loss": 6.0826,
+      "step": 12100
+    },
+    {
+      "epoch": 0.017015686116092364,
+      "grad_norm": 1.3808116912841797,
+      "learning_rate": 5.103414359983138e-05,
+      "loss": 6.2429,
+      "step": 12110
+    },
+    {
+      "epoch": 0.01702973705425594,
+      "grad_norm": 1.4226034879684448,
+      "learning_rate": 5.107629619221582e-05,
+      "loss": 5.9922,
+      "step": 12120
+    },
+    {
+      "epoch": 0.01704378799241952,
+      "grad_norm": 1.3345751762390137,
+      "learning_rate": 5.111844878460025e-05,
+      "loss": 6.1254,
+      "step": 12130
+    },
+    {
+      "epoch": 0.017057838930583095,
+      "grad_norm": 1.391182780265808,
+      "learning_rate": 5.116060137698468e-05,
+      "loss": 5.9991,
+      "step": 12140
+    },
+    {
+      "epoch": 0.017071889868746675,
+      "grad_norm": 1.3782682418823242,
+      "learning_rate": 5.120275396936911e-05,
+      "loss": 6.1427,
+      "step": 12150
+    },
+    {
+      "epoch": 0.01708594080691025,
+      "grad_norm": 1.4560561180114746,
+      "learning_rate": 5.124490656175354e-05,
+      "loss": 6.0616,
+      "step": 12160
+    },
+    {
+      "epoch": 0.017099991745073827,
+      "grad_norm": 1.3621957302093506,
+      "learning_rate": 5.128705915413798e-05,
+      "loss": 6.099,
+      "step": 12170
+    },
+    {
+      "epoch": 0.017114042683237407,
+      "grad_norm": 1.2872947454452515,
+      "learning_rate": 5.1329211746522405e-05,
+      "loss": 6.068,
+      "step": 12180
+    },
+    {
+      "epoch": 0.017128093621400983,
+      "grad_norm": 1.3031105995178223,
+      "learning_rate": 5.1371364338906836e-05,
+      "loss": 6.1765,
+      "step": 12190
+    },
+    {
+      "epoch": 0.017142144559564563,
+      "grad_norm": 1.4918880462646484,
+      "learning_rate": 5.1413516931291274e-05,
+      "loss": 6.0744,
+      "step": 12200
+    },
+    {
+      "epoch": 0.01715619549772814,
+      "grad_norm": 1.66692316532135,
+      "learning_rate": 5.14556695236757e-05,
+      "loss": 6.1172,
+      "step": 12210
+    },
+    {
+      "epoch": 0.017170246435891715,
+      "grad_norm": 1.3032985925674438,
+      "learning_rate": 5.1497822116060136e-05,
+      "loss": 6.0353,
+      "step": 12220
+    },
+    {
+      "epoch": 0.017184297374055295,
+      "grad_norm": 1.3920645713806152,
+      "learning_rate": 5.153997470844457e-05,
+      "loss": 6.1914,
+      "step": 12230
+    },
+    {
+      "epoch": 0.01719834831221887,
+      "grad_norm": 1.3095813989639282,
+      "learning_rate": 5.158212730082899e-05,
+      "loss": 6.0989,
+      "step": 12240
+    },
+    {
+      "epoch": 0.01721239925038245,
+      "grad_norm": 1.347722053527832,
+      "learning_rate": 5.162427989321343e-05,
+      "loss": 5.9789,
+      "step": 12250
+    },
+    {
+      "epoch": 0.017226450188546027,
+      "grad_norm": 1.4624723196029663,
+      "learning_rate": 5.166643248559786e-05,
+      "loss": 6.0599,
+      "step": 12260
+    },
+    {
+      "epoch": 0.017240501126709603,
+      "grad_norm": 1.4272193908691406,
+      "learning_rate": 5.170858507798229e-05,
+      "loss": 6.1182,
+      "step": 12270
+    },
+    {
+      "epoch": 0.017254552064873183,
+      "grad_norm": 1.3339524269104004,
+      "learning_rate": 5.175073767036672e-05,
+      "loss": 6.0278,
+      "step": 12280
+    },
+    {
+      "epoch": 0.01726860300303676,
+      "grad_norm": 1.505544900894165,
+      "learning_rate": 5.1792890262751154e-05,
+      "loss": 6.089,
+      "step": 12290
+    },
+    {
+      "epoch": 0.017282653941200335,
+      "grad_norm": 1.400678277015686,
+      "learning_rate": 5.1835042855135586e-05,
+      "loss": 6.1141,
+      "step": 12300
+    },
+    {
+      "epoch": 0.017296704879363915,
+      "grad_norm": 1.316159963607788,
+      "learning_rate": 5.187719544752002e-05,
+      "loss": 6.1014,
+      "step": 12310
+    },
+    {
+      "epoch": 0.01731075581752749,
+      "grad_norm": 1.41743803024292,
+      "learning_rate": 5.191934803990445e-05,
+      "loss": 6.0915,
+      "step": 12320
+    },
+    {
+      "epoch": 0.01732480675569107,
+      "grad_norm": 1.7660351991653442,
+      "learning_rate": 5.1961500632288886e-05,
+      "loss": 6.0905,
+      "step": 12330
+    },
+    {
+      "epoch": 0.017338857693854647,
+      "grad_norm": 1.3657212257385254,
+      "learning_rate": 5.200365322467331e-05,
+      "loss": 6.0799,
+      "step": 12340
+    },
+    {
+      "epoch": 0.017352908632018223,
+      "grad_norm": 1.3705576658248901,
+      "learning_rate": 5.204580581705775e-05,
+      "loss": 6.013,
+      "step": 12350
+    },
+    {
+      "epoch": 0.017366959570181802,
+      "grad_norm": 1.4308192729949951,
+      "learning_rate": 5.208795840944218e-05,
+      "loss": 6.1158,
+      "step": 12360
+    },
+    {
+      "epoch": 0.01738101050834538,
+      "grad_norm": 1.3916292190551758,
+      "learning_rate": 5.2130111001826604e-05,
+      "loss": 5.9591,
+      "step": 12370
+    },
+    {
+      "epoch": 0.017395061446508958,
+      "grad_norm": 1.3628098964691162,
+      "learning_rate": 5.217226359421104e-05,
+      "loss": 6.1443,
+      "step": 12380
+    },
+    {
+      "epoch": 0.017409112384672534,
+      "grad_norm": 1.2999829053878784,
+      "learning_rate": 5.221441618659547e-05,
+      "loss": 6.0797,
+      "step": 12390
+    },
+    {
+      "epoch": 0.01742316332283611,
+      "grad_norm": 1.4970372915267944,
+      "learning_rate": 5.22565687789799e-05,
+      "loss": 6.0399,
+      "step": 12400
+    },
+    {
+      "epoch": 0.01743721426099969,
+      "grad_norm": 1.466609239578247,
+      "learning_rate": 5.2298721371364335e-05,
+      "loss": 6.0547,
+      "step": 12410
+    },
+    {
+      "epoch": 0.017451265199163266,
+      "grad_norm": 1.4185417890548706,
+      "learning_rate": 5.2340873963748766e-05,
+      "loss": 6.0352,
+      "step": 12420
+    },
+    {
+      "epoch": 0.017465316137326842,
+      "grad_norm": 1.6092323064804077,
+      "learning_rate": 5.23830265561332e-05,
+      "loss": 6.0051,
+      "step": 12430
+    },
+    {
+      "epoch": 0.017479367075490422,
+      "grad_norm": 1.5270646810531616,
+      "learning_rate": 5.242517914851763e-05,
+      "loss": 6.1214,
+      "step": 12440
+    },
+    {
+      "epoch": 0.017493418013653998,
+      "grad_norm": 1.3804599046707153,
+      "learning_rate": 5.246733174090206e-05,
+      "loss": 6.0227,
+      "step": 12450
+    },
+    {
+      "epoch": 0.017507468951817578,
+      "grad_norm": 1.4650616645812988,
+      "learning_rate": 5.25094843332865e-05,
+      "loss": 6.1227,
+      "step": 12460
+    },
+    {
+      "epoch": 0.017521519889981154,
+      "grad_norm": 1.3534270524978638,
+      "learning_rate": 5.255163692567092e-05,
+      "loss": 6.0883,
+      "step": 12470
+    },
+    {
+      "epoch": 0.01753557082814473,
+      "grad_norm": 1.298676609992981,
+      "learning_rate": 5.259378951805536e-05,
+      "loss": 6.047,
+      "step": 12480
+    },
+    {
+      "epoch": 0.01754962176630831,
+      "grad_norm": 1.342682123184204,
+      "learning_rate": 5.263594211043979e-05,
+      "loss": 6.0541,
+      "step": 12490
+    },
+    {
+      "epoch": 0.017563672704471886,
+      "grad_norm": 1.375024676322937,
+      "learning_rate": 5.2678094702824216e-05,
+      "loss": 6.0099,
+      "step": 12500
+    },
+    {
+      "epoch": 0.017577723642635466,
+      "grad_norm": 1.4682930707931519,
+      "learning_rate": 5.2720247295208654e-05,
+      "loss": 6.1075,
+      "step": 12510
+    },
+    {
+      "epoch": 0.017591774580799042,
+      "grad_norm": 1.4236594438552856,
+      "learning_rate": 5.2762399887593085e-05,
+      "loss": 6.052,
+      "step": 12520
+    },
+    {
+      "epoch": 0.017605825518962618,
+      "grad_norm": 1.344035267829895,
+      "learning_rate": 5.280455247997751e-05,
+      "loss": 5.9737,
+      "step": 12530
+    },
+    {
+      "epoch": 0.017619876457126198,
+      "grad_norm": 1.3492854833602905,
+      "learning_rate": 5.284670507236195e-05,
+      "loss": 6.1154,
+      "step": 12540
+    },
+    {
+      "epoch": 0.017633927395289774,
+      "grad_norm": 1.4145216941833496,
+      "learning_rate": 5.288885766474638e-05,
+      "loss": 6.0093,
+      "step": 12550
+    },
+    {
+      "epoch": 0.017647978333453353,
+      "grad_norm": 1.3583636283874512,
+      "learning_rate": 5.293101025713081e-05,
+      "loss": 6.0203,
+      "step": 12560
+    },
+    {
+      "epoch": 0.01766202927161693,
+      "grad_norm": 1.3437408208847046,
+      "learning_rate": 5.297316284951524e-05,
+      "loss": 6.0749,
+      "step": 12570
+    },
+    {
+      "epoch": 0.017676080209780506,
+      "grad_norm": 1.3554259538650513,
+      "learning_rate": 5.301531544189967e-05,
+      "loss": 5.97,
+      "step": 12580
+    },
+    {
+      "epoch": 0.017690131147944085,
+      "grad_norm": 1.4989122152328491,
+      "learning_rate": 5.30574680342841e-05,
+      "loss": 6.0707,
+      "step": 12590
+    },
+    {
+      "epoch": 0.01770418208610766,
+      "grad_norm": 1.3016396760940552,
+      "learning_rate": 5.3099620626668534e-05,
+      "loss": 6.0922,
+      "step": 12600
+    },
+    {
+      "epoch": 0.017718233024271238,
+      "grad_norm": 1.3750371932983398,
+      "learning_rate": 5.3141773219052965e-05,
+      "loss": 5.9968,
+      "step": 12610
+    },
+    {
+      "epoch": 0.017732283962434817,
+      "grad_norm": 1.3117624521255493,
+      "learning_rate": 5.31839258114374e-05,
+      "loss": 5.9331,
+      "step": 12620
+    },
+    {
+      "epoch": 0.017746334900598393,
+      "grad_norm": 1.3047480583190918,
+      "learning_rate": 5.322607840382183e-05,
+      "loss": 6.1537,
+      "step": 12630
+    },
+    {
+      "epoch": 0.017760385838761973,
+      "grad_norm": 1.4629182815551758,
+      "learning_rate": 5.3268230996206266e-05,
+      "loss": 6.1317,
+      "step": 12640
+    },
+    {
+      "epoch": 0.01777443677692555,
+      "grad_norm": 1.5809321403503418,
+      "learning_rate": 5.33103835885907e-05,
+      "loss": 6.0821,
+      "step": 12650
+    },
+    {
+      "epoch": 0.017788487715089125,
+      "grad_norm": 1.3663504123687744,
+      "learning_rate": 5.335253618097512e-05,
+      "loss": 6.0406,
+      "step": 12660
+    },
+    {
+      "epoch": 0.017802538653252705,
+      "grad_norm": 1.3569718599319458,
+      "learning_rate": 5.339468877335956e-05,
+      "loss": 6.0533,
+      "step": 12670
+    },
+    {
+      "epoch": 0.01781658959141628,
+      "grad_norm": 1.3136539459228516,
+      "learning_rate": 5.343684136574399e-05,
+      "loss": 5.9933,
+      "step": 12680
+    },
+    {
+      "epoch": 0.01783064052957986,
+      "grad_norm": 1.3268357515335083,
+      "learning_rate": 5.347899395812842e-05,
+      "loss": 6.0975,
+      "step": 12690
+    },
+    {
+      "epoch": 0.017844691467743437,
+      "grad_norm": 1.3501018285751343,
+      "learning_rate": 5.352114655051285e-05,
+      "loss": 6.0835,
+      "step": 12700
+    },
+    {
+      "epoch": 0.017858742405907013,
+      "grad_norm": 1.4120509624481201,
+      "learning_rate": 5.3563299142897284e-05,
+      "loss": 6.0582,
+      "step": 12710
+    },
+    {
+      "epoch": 0.017872793344070593,
+      "grad_norm": 1.6523367166519165,
+      "learning_rate": 5.3605451735281715e-05,
+      "loss": 5.9865,
+      "step": 12720
+    },
+    {
+      "epoch": 0.01788684428223417,
+      "grad_norm": 1.4812240600585938,
+      "learning_rate": 5.3647604327666146e-05,
+      "loss": 6.0202,
+      "step": 12730
+    },
+    {
+      "epoch": 0.01790089522039775,
+      "grad_norm": 1.4424370527267456,
+      "learning_rate": 5.368975692005058e-05,
+      "loss": 6.0952,
+      "step": 12740
+    },
+    {
+      "epoch": 0.017914946158561325,
+      "grad_norm": 1.420494556427002,
+      "learning_rate": 5.3731909512435015e-05,
+      "loss": 6.1514,
+      "step": 12750
+    },
+    {
+      "epoch": 0.0179289970967249,
+      "grad_norm": 1.4344902038574219,
+      "learning_rate": 5.377406210481944e-05,
+      "loss": 6.0853,
+      "step": 12760
+    },
+    {
+      "epoch": 0.01794304803488848,
+      "grad_norm": 1.3586077690124512,
+      "learning_rate": 5.381621469720388e-05,
+      "loss": 6.0114,
+      "step": 12770
+    },
+    {
+      "epoch": 0.017957098973052057,
+      "grad_norm": 1.310850977897644,
+      "learning_rate": 5.385836728958831e-05,
+      "loss": 5.9627,
+      "step": 12780
+    },
+    {
+      "epoch": 0.017971149911215633,
+      "grad_norm": 1.4529310464859009,
+      "learning_rate": 5.390051988197273e-05,
+      "loss": 5.9186,
+      "step": 12790
+    },
+    {
+      "epoch": 0.017985200849379213,
+      "grad_norm": 1.4991663694381714,
+      "learning_rate": 5.394267247435717e-05,
+      "loss": 5.9598,
+      "step": 12800
+    },
+    {
+      "epoch": 0.01799925178754279,
+      "grad_norm": 1.3883793354034424,
+      "learning_rate": 5.39848250667416e-05,
+      "loss": 6.0834,
+      "step": 12810
+    },
+    {
+      "epoch": 0.01801330272570637,
+      "grad_norm": 1.35647451877594,
+      "learning_rate": 5.4026977659126027e-05,
+      "loss": 6.1214,
+      "step": 12820
+    },
+    {
+      "epoch": 0.018027353663869945,
+      "grad_norm": 1.3526296615600586,
+      "learning_rate": 5.4069130251510465e-05,
+      "loss": 6.0049,
+      "step": 12830
+    },
+    {
+      "epoch": 0.01804140460203352,
+      "grad_norm": 1.364187240600586,
+      "learning_rate": 5.4111282843894896e-05,
+      "loss": 5.9091,
+      "step": 12840
+    },
+    {
+      "epoch": 0.0180554555401971,
+      "grad_norm": 1.277514100074768,
+      "learning_rate": 5.415343543627933e-05,
+      "loss": 6.0787,
+      "step": 12850
+    },
+    {
+      "epoch": 0.018069506478360677,
+      "grad_norm": 1.7247467041015625,
+      "learning_rate": 5.419558802866376e-05,
+      "loss": 6.0347,
+      "step": 12860
+    },
+    {
+      "epoch": 0.018083557416524256,
+      "grad_norm": 1.2670753002166748,
+      "learning_rate": 5.423774062104819e-05,
+      "loss": 6.0311,
+      "step": 12870
+    },
+    {
+      "epoch": 0.018097608354687832,
+      "grad_norm": 1.3634384870529175,
+      "learning_rate": 5.427989321343262e-05,
+      "loss": 6.0208,
+      "step": 12880
+    },
+    {
+      "epoch": 0.01811165929285141,
+      "grad_norm": 1.5839406251907349,
+      "learning_rate": 5.432204580581705e-05,
+      "loss": 6.0122,
+      "step": 12890
+    },
+    {
+      "epoch": 0.018125710231014988,
+      "grad_norm": 1.3542689085006714,
+      "learning_rate": 5.436419839820148e-05,
+      "loss": 5.9979,
+      "step": 12900
+    },
+    {
+      "epoch": 0.018139761169178564,
+      "grad_norm": 1.4818419218063354,
+      "learning_rate": 5.440635099058592e-05,
+      "loss": 6.0396,
+      "step": 12910
+    },
+    {
+      "epoch": 0.01815381210734214,
+      "grad_norm": 1.4347730875015259,
+      "learning_rate": 5.4448503582970345e-05,
+      "loss": 6.0252,
+      "step": 12920
+    },
+    {
+      "epoch": 0.01816786304550572,
+      "grad_norm": 1.463624119758606,
+      "learning_rate": 5.449065617535478e-05,
+      "loss": 6.0238,
+      "step": 12930
+    },
+    {
+      "epoch": 0.018181913983669296,
+      "grad_norm": 1.5159509181976318,
+      "learning_rate": 5.4532808767739214e-05,
+      "loss": 6.0117,
+      "step": 12940
+    },
+    {
+      "epoch": 0.018195964921832876,
+      "grad_norm": 1.306830644607544,
+      "learning_rate": 5.457496136012364e-05,
+      "loss": 6.1917,
+      "step": 12950
+    },
+    {
+      "epoch": 0.018210015859996452,
+      "grad_norm": 1.3079791069030762,
+      "learning_rate": 5.4617113952508076e-05,
+      "loss": 6.0244,
+      "step": 12960
+    },
+    {
+      "epoch": 0.018224066798160028,
+      "grad_norm": 1.3655420541763306,
+      "learning_rate": 5.465926654489251e-05,
+      "loss": 5.9953,
+      "step": 12970
+    },
+    {
+      "epoch": 0.018238117736323608,
+      "grad_norm": 1.5014837980270386,
+      "learning_rate": 5.470141913727694e-05,
+      "loss": 6.0585,
+      "step": 12980
+    },
+    {
+      "epoch": 0.018252168674487184,
+      "grad_norm": 1.595160722732544,
+      "learning_rate": 5.474357172966137e-05,
+      "loss": 5.9723,
+      "step": 12990
+    },
+    {
+      "epoch": 0.018266219612650764,
+      "grad_norm": 1.2678043842315674,
+      "learning_rate": 5.47857243220458e-05,
+      "loss": 6.03,
+      "step": 13000
+    },
+    {
+      "epoch": 0.01828027055081434,
+      "grad_norm": 1.397788405418396,
+      "learning_rate": 5.482787691443023e-05,
+      "loss": 5.9627,
+      "step": 13010
+    },
+    {
+      "epoch": 0.018294321488977916,
+      "grad_norm": 1.4957337379455566,
+      "learning_rate": 5.4870029506814663e-05,
+      "loss": 5.9015,
+      "step": 13020
+    },
+    {
+      "epoch": 0.018308372427141496,
+      "grad_norm": 1.3790498971939087,
+      "learning_rate": 5.4912182099199095e-05,
+      "loss": 5.9694,
+      "step": 13030
+    },
+    {
+      "epoch": 0.018322423365305072,
+      "grad_norm": 1.4078073501586914,
+      "learning_rate": 5.4954334691583526e-05,
+      "loss": 6.0007,
+      "step": 13040
+    },
+    {
+      "epoch": 0.01833647430346865,
+      "grad_norm": 1.4309850931167603,
+      "learning_rate": 5.499648728396796e-05,
+      "loss": 5.8825,
+      "step": 13050
+    },
+    {
+      "epoch": 0.018350525241632228,
+      "grad_norm": 1.3303261995315552,
+      "learning_rate": 5.5038639876352395e-05,
+      "loss": 5.96,
+      "step": 13060
+    },
+    {
+      "epoch": 0.018364576179795804,
+      "grad_norm": 1.2958881855010986,
+      "learning_rate": 5.5080792468736826e-05,
+      "loss": 6.1374,
+      "step": 13070
+    },
+    {
+      "epoch": 0.018378627117959383,
+      "grad_norm": 1.2930759191513062,
+      "learning_rate": 5.512294506112125e-05,
+      "loss": 6.0608,
+      "step": 13080
+    },
+    {
+      "epoch": 0.01839267805612296,
+      "grad_norm": 1.3812742233276367,
+      "learning_rate": 5.516509765350569e-05,
+      "loss": 6.0503,
+      "step": 13090
+    },
+    {
+      "epoch": 0.018406728994286536,
+      "grad_norm": 1.5456600189208984,
+      "learning_rate": 5.520725024589012e-05,
+      "loss": 5.9067,
+      "step": 13100
+    },
+    {
+      "epoch": 0.018420779932450115,
+      "grad_norm": 1.3641170263290405,
+      "learning_rate": 5.5249402838274544e-05,
+      "loss": 6.0021,
+      "step": 13110
+    },
+    {
+      "epoch": 0.01843483087061369,
+      "grad_norm": 1.5792025327682495,
+      "learning_rate": 5.529155543065898e-05,
+      "loss": 5.9245,
+      "step": 13120
+    },
+    {
+      "epoch": 0.01844888180877727,
+      "grad_norm": 1.3822730779647827,
+      "learning_rate": 5.533370802304341e-05,
+      "loss": 5.9526,
+      "step": 13130
+    },
+    {
+      "epoch": 0.018462932746940847,
+      "grad_norm": 1.744850754737854,
+      "learning_rate": 5.5375860615427844e-05,
+      "loss": 6.0086,
+      "step": 13140
+    },
+    {
+      "epoch": 0.018476983685104423,
+      "grad_norm": 1.3281506299972534,
+      "learning_rate": 5.5418013207812275e-05,
+      "loss": 5.9487,
+      "step": 13150
+    },
+    {
+      "epoch": 0.018491034623268003,
+      "grad_norm": 1.401555061340332,
+      "learning_rate": 5.5460165800196707e-05,
+      "loss": 6.0319,
+      "step": 13160
+    },
+    {
+      "epoch": 0.01850508556143158,
+      "grad_norm": 1.3791970014572144,
+      "learning_rate": 5.550231839258114e-05,
+      "loss": 5.9704,
+      "step": 13170
+    },
+    {
+      "epoch": 0.01851913649959516,
+      "grad_norm": 1.3170675039291382,
+      "learning_rate": 5.554447098496557e-05,
+      "loss": 5.9452,
+      "step": 13180
+    },
+    {
+      "epoch": 0.018533187437758735,
+      "grad_norm": 1.3928661346435547,
+      "learning_rate": 5.558662357735001e-05,
+      "loss": 5.9074,
+      "step": 13190
+    },
+    {
+      "epoch": 0.01854723837592231,
+      "grad_norm": 1.7144209146499634,
+      "learning_rate": 5.562877616973444e-05,
+      "loss": 5.9106,
+      "step": 13200
+    },
+    {
+      "epoch": 0.01856128931408589,
+      "grad_norm": 1.4539275169372559,
+      "learning_rate": 5.567092876211886e-05,
+      "loss": 6.0976,
+      "step": 13210
+    },
+    {
+      "epoch": 0.018575340252249467,
+      "grad_norm": 1.3310813903808594,
+      "learning_rate": 5.57130813545033e-05,
+      "loss": 6.0462,
+      "step": 13220
+    },
+    {
+      "epoch": 0.018589391190413043,
+      "grad_norm": 1.4120519161224365,
+      "learning_rate": 5.575523394688773e-05,
+      "loss": 5.9199,
+      "step": 13230
+    },
+    {
+      "epoch": 0.018603442128576623,
+      "grad_norm": 1.307513952255249,
+      "learning_rate": 5.5797386539272156e-05,
+      "loss": 5.9939,
+      "step": 13240
+    },
+    {
+      "epoch": 0.0186174930667402,
+      "grad_norm": 1.2844830751419067,
+      "learning_rate": 5.5839539131656594e-05,
+      "loss": 5.9712,
+      "step": 13250
+    },
+    {
+      "epoch": 0.01863154400490378,
+      "grad_norm": 1.4797707796096802,
+      "learning_rate": 5.5881691724041025e-05,
+      "loss": 6.0282,
+      "step": 13260
+    },
+    {
+      "epoch": 0.018645594943067355,
+      "grad_norm": 1.3347375392913818,
+      "learning_rate": 5.5923844316425456e-05,
+      "loss": 5.928,
+      "step": 13270
+    },
+    {
+      "epoch": 0.01865964588123093,
+      "grad_norm": 1.4141706228256226,
+      "learning_rate": 5.596599690880989e-05,
+      "loss": 5.965,
+      "step": 13280
+    },
+    {
+      "epoch": 0.01867369681939451,
+      "grad_norm": 1.3178112506866455,
+      "learning_rate": 5.600814950119432e-05,
+      "loss": 5.9698,
+      "step": 13290
+    },
+    {
+      "epoch": 0.018687747757558087,
+      "grad_norm": 1.3804385662078857,
+      "learning_rate": 5.605030209357875e-05,
+      "loss": 5.9723,
+      "step": 13300
+    },
+    {
+      "epoch": 0.018701798695721666,
+      "grad_norm": 1.3772902488708496,
+      "learning_rate": 5.609245468596318e-05,
+      "loss": 5.9814,
+      "step": 13310
+    },
+    {
+      "epoch": 0.018715849633885243,
+      "grad_norm": 1.3835526704788208,
+      "learning_rate": 5.613460727834761e-05,
+      "loss": 6.0272,
+      "step": 13320
+    },
+    {
+      "epoch": 0.01872990057204882,
+      "grad_norm": 1.3807588815689087,
+      "learning_rate": 5.617675987073204e-05,
+      "loss": 6.0179,
+      "step": 13330
+    },
+    {
+      "epoch": 0.0187439515102124,
+      "grad_norm": 1.3371533155441284,
+      "learning_rate": 5.6218912463116474e-05,
+      "loss": 6.0128,
+      "step": 13340
+    },
+    {
+      "epoch": 0.018758002448375975,
+      "grad_norm": 1.4686000347137451,
+      "learning_rate": 5.626106505550091e-05,
+      "loss": 5.9024,
+      "step": 13350
+    },
+    {
+      "epoch": 0.018772053386539554,
+      "grad_norm": 1.7148891687393188,
+      "learning_rate": 5.6303217647885343e-05,
+      "loss": 6.0377,
+      "step": 13360
+    },
+    {
+      "epoch": 0.01878610432470313,
+      "grad_norm": 1.4156179428100586,
+      "learning_rate": 5.634537024026977e-05,
+      "loss": 6.0123,
+      "step": 13370
+    },
+    {
+      "epoch": 0.018800155262866707,
+      "grad_norm": 1.2572938203811646,
+      "learning_rate": 5.6387522832654206e-05,
+      "loss": 5.8405,
+      "step": 13380
+    },
+    {
+      "epoch": 0.018814206201030286,
+      "grad_norm": 1.3694168329238892,
+      "learning_rate": 5.642967542503864e-05,
+      "loss": 6.0598,
+      "step": 13390
+    },
+    {
+      "epoch": 0.018828257139193862,
+      "grad_norm": 1.4427056312561035,
+      "learning_rate": 5.647182801742306e-05,
+      "loss": 5.91,
+      "step": 13400
+    },
+    {
+      "epoch": 0.01884230807735744,
+      "grad_norm": 1.3949912786483765,
+      "learning_rate": 5.65139806098075e-05,
+      "loss": 5.941,
+      "step": 13410
+    },
+    {
+      "epoch": 0.018856359015521018,
+      "grad_norm": 2.0856781005859375,
+      "learning_rate": 5.655613320219193e-05,
+      "loss": 5.8934,
+      "step": 13420
+    },
+    {
+      "epoch": 0.018870409953684594,
+      "grad_norm": 1.295962929725647,
+      "learning_rate": 5.659828579457636e-05,
+      "loss": 6.0145,
+      "step": 13430
+    },
+    {
+      "epoch": 0.018884460891848174,
+      "grad_norm": 1.350246787071228,
+      "learning_rate": 5.664043838696079e-05,
+      "loss": 5.9026,
+      "step": 13440
+    },
+    {
+      "epoch": 0.01889851183001175,
+      "grad_norm": 1.334167718887329,
+      "learning_rate": 5.6682590979345224e-05,
+      "loss": 6.0541,
+      "step": 13450
+    },
+    {
+      "epoch": 0.018912562768175326,
+      "grad_norm": 1.4749501943588257,
+      "learning_rate": 5.6724743571729655e-05,
+      "loss": 5.9257,
+      "step": 13460
+    },
+    {
+      "epoch": 0.018926613706338906,
+      "grad_norm": 1.4598968029022217,
+      "learning_rate": 5.6766896164114086e-05,
+      "loss": 5.9241,
+      "step": 13470
+    },
+    {
+      "epoch": 0.018940664644502482,
+      "grad_norm": 1.5799657106399536,
+      "learning_rate": 5.6809048756498524e-05,
+      "loss": 5.8404,
+      "step": 13480
+    },
+    {
+      "epoch": 0.01895471558266606,
+      "grad_norm": 1.455440878868103,
+      "learning_rate": 5.6851201348882955e-05,
+      "loss": 5.8688,
+      "step": 13490
+    },
+    {
+      "epoch": 0.018968766520829638,
+      "grad_norm": 1.424730896949768,
+      "learning_rate": 5.689335394126738e-05,
+      "loss": 5.9688,
+      "step": 13500
+    },
+    {
+      "epoch": 0.018982817458993214,
+      "grad_norm": 1.2700331211090088,
+      "learning_rate": 5.693550653365182e-05,
+      "loss": 5.9306,
+      "step": 13510
+    },
+    {
+      "epoch": 0.018996868397156794,
+      "grad_norm": 1.3546868562698364,
+      "learning_rate": 5.697765912603625e-05,
+      "loss": 5.9353,
+      "step": 13520
+    },
+    {
+      "epoch": 0.01901091933532037,
+      "grad_norm": 1.417958378791809,
+      "learning_rate": 5.701981171842067e-05,
+      "loss": 5.9517,
+      "step": 13530
+    },
+    {
+      "epoch": 0.019024970273483946,
+      "grad_norm": 1.3052090406417847,
+      "learning_rate": 5.706196431080511e-05,
+      "loss": 5.8374,
+      "step": 13540
+    },
+    {
+      "epoch": 0.019039021211647526,
+      "grad_norm": 1.298235297203064,
+      "learning_rate": 5.710411690318954e-05,
+      "loss": 5.9565,
+      "step": 13550
+    },
+    {
+      "epoch": 0.019053072149811102,
+      "grad_norm": 1.4889936447143555,
+      "learning_rate": 5.7146269495573974e-05,
+      "loss": 5.9748,
+      "step": 13560
+    },
+    {
+      "epoch": 0.01906712308797468,
+      "grad_norm": 1.4061089754104614,
+      "learning_rate": 5.7188422087958405e-05,
+      "loss": 5.9171,
+      "step": 13570
+    },
+    {
+      "epoch": 0.019081174026138258,
+      "grad_norm": 1.4704773426055908,
+      "learning_rate": 5.7230574680342836e-05,
+      "loss": 6.0023,
+      "step": 13580
+    },
+    {
+      "epoch": 0.019095224964301834,
+      "grad_norm": 1.344780683517456,
+      "learning_rate": 5.727272727272727e-05,
+      "loss": 5.9356,
+      "step": 13590
+    },
+    {
+      "epoch": 0.019109275902465413,
+      "grad_norm": 1.3091164827346802,
+      "learning_rate": 5.73148798651117e-05,
+      "loss": 5.9137,
+      "step": 13600
+    },
+    {
+      "epoch": 0.01912332684062899,
+      "grad_norm": 1.3960378170013428,
+      "learning_rate": 5.735703245749613e-05,
+      "loss": 5.9779,
+      "step": 13610
+    },
+    {
+      "epoch": 0.01913737777879257,
+      "grad_norm": 1.5556848049163818,
+      "learning_rate": 5.739918504988056e-05,
+      "loss": 5.9581,
+      "step": 13620
+    },
+    {
+      "epoch": 0.019151428716956145,
+      "grad_norm": 1.3245570659637451,
+      "learning_rate": 5.744133764226499e-05,
+      "loss": 5.9721,
+      "step": 13630
+    },
+    {
+      "epoch": 0.01916547965511972,
+      "grad_norm": 1.578223705291748,
+      "learning_rate": 5.748349023464943e-05,
+      "loss": 5.9285,
+      "step": 13640
+    },
+    {
+      "epoch": 0.0191795305932833,
+      "grad_norm": 2.247985363006592,
+      "learning_rate": 5.752564282703386e-05,
+      "loss": 5.9884,
+      "step": 13650
+    },
+    {
+      "epoch": 0.019193581531446877,
+      "grad_norm": 1.6837748289108276,
+      "learning_rate": 5.7567795419418285e-05,
+      "loss": 6.0128,
+      "step": 13660
+    },
+    {
+      "epoch": 0.019207632469610457,
+      "grad_norm": 1.367041826248169,
+      "learning_rate": 5.760994801180272e-05,
+      "loss": 5.9926,
+      "step": 13670
+    },
+    {
+      "epoch": 0.019221683407774033,
+      "grad_norm": 1.4417003393173218,
+      "learning_rate": 5.7652100604187154e-05,
+      "loss": 5.9076,
+      "step": 13680
+    },
+    {
+      "epoch": 0.01923573434593761,
+      "grad_norm": 1.3726568222045898,
+      "learning_rate": 5.7694253196571586e-05,
+      "loss": 5.875,
+      "step": 13690
+    },
+    {
+      "epoch": 0.01924978528410119,
+      "grad_norm": 1.3321431875228882,
+      "learning_rate": 5.773640578895602e-05,
+      "loss": 5.9552,
+      "step": 13700
+    },
+    {
+      "epoch": 0.019263836222264765,
+      "grad_norm": 1.372376561164856,
+      "learning_rate": 5.777855838134045e-05,
+      "loss": 5.955,
+      "step": 13710
+    },
+    {
+      "epoch": 0.01927788716042834,
+      "grad_norm": 1.4178439378738403,
+      "learning_rate": 5.782071097372488e-05,
+      "loss": 5.8555,
+      "step": 13720
+    },
+    {
+      "epoch": 0.01929193809859192,
+      "grad_norm": 1.3034266233444214,
+      "learning_rate": 5.786286356610931e-05,
+      "loss": 6.0031,
+      "step": 13730
+    },
+    {
+      "epoch": 0.019305989036755497,
+      "grad_norm": 1.7195543050765991,
+      "learning_rate": 5.790501615849374e-05,
+      "loss": 5.8395,
+      "step": 13740
+    },
+    {
+      "epoch": 0.019320039974919077,
+      "grad_norm": 1.3207449913024902,
+      "learning_rate": 5.794716875087817e-05,
+      "loss": 5.9347,
+      "step": 13750
+    },
+    {
+      "epoch": 0.019334090913082653,
+      "grad_norm": 1.3822189569473267,
+      "learning_rate": 5.7989321343262604e-05,
+      "loss": 5.8751,
+      "step": 13760
+    },
+    {
+      "epoch": 0.01934814185124623,
+      "grad_norm": 1.5890324115753174,
+      "learning_rate": 5.803147393564704e-05,
+      "loss": 5.8387,
+      "step": 13770
+    },
+    {
+      "epoch": 0.01936219278940981,
+      "grad_norm": 1.2956420183181763,
+      "learning_rate": 5.807362652803147e-05,
+      "loss": 5.8924,
+      "step": 13780
+    },
+    {
+      "epoch": 0.019376243727573385,
+      "grad_norm": 1.572546362876892,
+      "learning_rate": 5.81157791204159e-05,
+      "loss": 5.9325,
+      "step": 13790
+    },
+    {
+      "epoch": 0.019390294665736964,
+      "grad_norm": 1.4117943048477173,
+      "learning_rate": 5.8157931712800335e-05,
+      "loss": 5.8553,
+      "step": 13800
+    },
+    {
+      "epoch": 0.01940434560390054,
+      "grad_norm": 1.48485267162323,
+      "learning_rate": 5.8200084305184766e-05,
+      "loss": 5.8948,
+      "step": 13810
+    },
+    {
+      "epoch": 0.019418396542064117,
+      "grad_norm": 1.3272455930709839,
+      "learning_rate": 5.824223689756919e-05,
+      "loss": 5.8358,
+      "step": 13820
+    },
+    {
+      "epoch": 0.019432447480227696,
+      "grad_norm": 1.4525662660598755,
+      "learning_rate": 5.828438948995363e-05,
+      "loss": 5.937,
+      "step": 13830
+    },
+    {
+      "epoch": 0.019446498418391273,
+      "grad_norm": 1.495578646659851,
+      "learning_rate": 5.832654208233806e-05,
+      "loss": 5.9296,
+      "step": 13840
+    },
+    {
+      "epoch": 0.01946054935655485,
+      "grad_norm": 1.4523495435714722,
+      "learning_rate": 5.836869467472249e-05,
+      "loss": 5.7838,
+      "step": 13850
+    },
+    {
+      "epoch": 0.01947460029471843,
+      "grad_norm": 1.3701502084732056,
+      "learning_rate": 5.841084726710692e-05,
+      "loss": 5.9265,
+      "step": 13860
+    },
+    {
+      "epoch": 0.019488651232882005,
+      "grad_norm": 1.435974359512329,
+      "learning_rate": 5.845299985949135e-05,
+      "loss": 5.7779,
+      "step": 13870
+    },
+    {
+      "epoch": 0.019502702171045584,
+      "grad_norm": 1.507411003112793,
+      "learning_rate": 5.8495152451875784e-05,
+      "loss": 6.0682,
+      "step": 13880
+    },
+    {
+      "epoch": 0.01951675310920916,
+      "grad_norm": 1.71627938747406,
+      "learning_rate": 5.8537305044260216e-05,
+      "loss": 5.8261,
+      "step": 13890
+    },
+    {
+      "epoch": 0.019530804047372737,
+      "grad_norm": 1.3306772708892822,
+      "learning_rate": 5.857945763664465e-05,
+      "loss": 5.879,
+      "step": 13900
+    },
+    {
+      "epoch": 0.019544854985536316,
+      "grad_norm": 1.433265209197998,
+      "learning_rate": 5.862161022902908e-05,
+      "loss": 5.8721,
+      "step": 13910
+    },
+    {
+      "epoch": 0.019558905923699892,
+      "grad_norm": 1.2996599674224854,
+      "learning_rate": 5.866376282141351e-05,
+      "loss": 6.0022,
+      "step": 13920
+    },
+    {
+      "epoch": 0.019572956861863472,
+      "grad_norm": 1.3266264200210571,
+      "learning_rate": 5.870591541379795e-05,
+      "loss": 5.8978,
+      "step": 13930
+    },
+    {
+      "epoch": 0.019587007800027048,
+      "grad_norm": 1.3822417259216309,
+      "learning_rate": 5.874806800618238e-05,
+      "loss": 5.9197,
+      "step": 13940
+    },
+    {
+      "epoch": 0.019601058738190624,
+      "grad_norm": 1.4425207376480103,
+      "learning_rate": 5.87902205985668e-05,
+      "loss": 5.8509,
+      "step": 13950
+    },
+    {
+      "epoch": 0.019615109676354204,
+      "grad_norm": 1.3024529218673706,
+      "learning_rate": 5.883237319095124e-05,
+      "loss": 5.9852,
+      "step": 13960
+    },
+    {
+      "epoch": 0.01962916061451778,
+      "grad_norm": 1.3837809562683105,
+      "learning_rate": 5.887452578333567e-05,
+      "loss": 5.9071,
+      "step": 13970
+    },
+    {
+      "epoch": 0.01964321155268136,
+      "grad_norm": 1.37079918384552,
+      "learning_rate": 5.89166783757201e-05,
+      "loss": 5.9889,
+      "step": 13980
+    },
+    {
+      "epoch": 0.019657262490844936,
+      "grad_norm": 1.5206080675125122,
+      "learning_rate": 5.8958830968104534e-05,
+      "loss": 5.7997,
+      "step": 13990
+    },
+    {
+      "epoch": 0.019671313429008512,
+      "grad_norm": 1.3415592908859253,
+      "learning_rate": 5.9000983560488965e-05,
+      "loss": 5.9298,
+      "step": 14000
+    },
+    {
+      "epoch": 0.01968536436717209,
+      "grad_norm": 1.7559236288070679,
+      "learning_rate": 5.9043136152873396e-05,
+      "loss": 6.0002,
+      "step": 14010
+    },
+    {
+      "epoch": 0.019699415305335668,
+      "grad_norm": 1.3365631103515625,
+      "learning_rate": 5.908528874525783e-05,
+      "loss": 5.8709,
+      "step": 14020
+    },
+    {
+      "epoch": 0.019713466243499244,
+      "grad_norm": 1.3177831172943115,
+      "learning_rate": 5.912744133764226e-05,
+      "loss": 5.8005,
+      "step": 14030
+    },
+    {
+      "epoch": 0.019727517181662824,
+      "grad_norm": 1.3259848356246948,
+      "learning_rate": 5.916959393002669e-05,
+      "loss": 5.9384,
+      "step": 14040
+    },
+    {
+      "epoch": 0.0197415681198264,
+      "grad_norm": 1.5154435634613037,
+      "learning_rate": 5.921174652241112e-05,
+      "loss": 5.9317,
+      "step": 14050
+    },
+    {
+      "epoch": 0.01975561905798998,
+      "grad_norm": 1.3720771074295044,
+      "learning_rate": 5.925389911479556e-05,
+      "loss": 5.8961,
+      "step": 14060
+    },
+    {
+      "epoch": 0.019769669996153556,
+      "grad_norm": 1.6641851663589478,
+      "learning_rate": 5.9296051707179983e-05,
+      "loss": 5.9034,
+      "step": 14070
+    },
+    {
+      "epoch": 0.019783720934317132,
+      "grad_norm": 1.3994899988174438,
+      "learning_rate": 5.9338204299564415e-05,
+      "loss": 5.8884,
+      "step": 14080
+    },
+    {
+      "epoch": 0.01979777187248071,
+      "grad_norm": 1.3691990375518799,
+      "learning_rate": 5.938035689194885e-05,
+      "loss": 5.8671,
+      "step": 14090
+    },
+    {
+      "epoch": 0.019811822810644288,
+      "grad_norm": 1.7563625574111938,
+      "learning_rate": 5.9422509484333284e-05,
+      "loss": 5.8736,
+      "step": 14100
+    },
+    {
+      "epoch": 0.019825873748807867,
+      "grad_norm": 1.3554989099502563,
+      "learning_rate": 5.946466207671771e-05,
+      "loss": 5.9775,
+      "step": 14110
+    },
+    {
+      "epoch": 0.019839924686971443,
+      "grad_norm": 1.4109904766082764,
+      "learning_rate": 5.9506814669102146e-05,
+      "loss": 5.9443,
+      "step": 14120
+    },
+    {
+      "epoch": 0.01985397562513502,
+      "grad_norm": 1.383750081062317,
+      "learning_rate": 5.954896726148658e-05,
+      "loss": 5.855,
+      "step": 14130
+    },
+    {
+      "epoch": 0.0198680265632986,
+      "grad_norm": 1.4050294160842896,
+      "learning_rate": 5.959111985387101e-05,
+      "loss": 5.9805,
+      "step": 14140
+    },
+    {
+      "epoch": 0.019882077501462175,
+      "grad_norm": 1.4362891912460327,
+      "learning_rate": 5.963327244625544e-05,
+      "loss": 5.8024,
+      "step": 14150
+    },
+    {
+      "epoch": 0.019896128439625755,
+      "grad_norm": 1.4630895853042603,
+      "learning_rate": 5.967542503863987e-05,
+      "loss": 5.9194,
+      "step": 14160
+    },
+    {
+      "epoch": 0.01991017937778933,
+      "grad_norm": 1.3018559217453003,
+      "learning_rate": 5.97175776310243e-05,
+      "loss": 5.9572,
+      "step": 14170
+    },
+    {
+      "epoch": 0.019924230315952907,
+      "grad_norm": 1.3787480592727661,
+      "learning_rate": 5.975973022340873e-05,
+      "loss": 5.9111,
+      "step": 14180
+    },
+    {
+      "epoch": 0.019938281254116487,
+      "grad_norm": 1.5770620107650757,
+      "learning_rate": 5.980188281579317e-05,
+      "loss": 5.949,
+      "step": 14190
+    },
+    {
+      "epoch": 0.019952332192280063,
+      "grad_norm": 1.32578706741333,
+      "learning_rate": 5.9844035408177595e-05,
+      "loss": 5.8478,
+      "step": 14200
+    },
+    {
+      "epoch": 0.01996638313044364,
+      "grad_norm": 1.3311973810195923,
+      "learning_rate": 5.9886188000562027e-05,
+      "loss": 5.7895,
+      "step": 14210
+    },
+    {
+      "epoch": 0.01998043406860722,
+      "grad_norm": 1.4675605297088623,
+      "learning_rate": 5.9928340592946464e-05,
+      "loss": 5.8426,
+      "step": 14220
+    },
+    {
+      "epoch": 0.019994485006770795,
+      "grad_norm": 1.389006495475769,
+      "learning_rate": 5.9970493185330896e-05,
+      "loss": 5.8199,
+      "step": 14230
+    },
+    {
+      "epoch": 0.020008535944934375,
+      "grad_norm": 1.7180744409561157,
+      "learning_rate": 6.001264577771532e-05,
+      "loss": 5.8948,
+      "step": 14240
+    },
+    {
+      "epoch": 0.02002258688309795,
+      "grad_norm": 1.4792721271514893,
+      "learning_rate": 6.005479837009976e-05,
+      "loss": 5.8289,
+      "step": 14250
+    },
+    {
+      "epoch": 0.020036637821261527,
+      "grad_norm": 1.3706181049346924,
+      "learning_rate": 6.009695096248419e-05,
+      "loss": 5.9004,
+      "step": 14260
+    },
+    {
+      "epoch": 0.020050688759425107,
+      "grad_norm": 1.3976632356643677,
+      "learning_rate": 6.013910355486862e-05,
+      "loss": 5.837,
+      "step": 14270
+    },
+    {
+      "epoch": 0.020064739697588683,
+      "grad_norm": 1.3973368406295776,
+      "learning_rate": 6.018125614725305e-05,
+      "loss": 5.9126,
+      "step": 14280
+    },
+    {
+      "epoch": 0.020078790635752262,
+      "grad_norm": 1.5715432167053223,
+      "learning_rate": 6.022340873963748e-05,
+      "loss": 6.0113,
+      "step": 14290
+    },
+    {
+      "epoch": 0.02009284157391584,
+      "grad_norm": 1.3540730476379395,
+      "learning_rate": 6.0265561332021914e-05,
+      "loss": 5.9477,
+      "step": 14300
+    },
+    {
+      "epoch": 0.020106892512079415,
+      "grad_norm": 1.3370450735092163,
+      "learning_rate": 6.0307713924406345e-05,
+      "loss": 5.8446,
+      "step": 14310
+    },
+    {
+      "epoch": 0.020120943450242994,
+      "grad_norm": 1.3597146272659302,
+      "learning_rate": 6.0349866516790776e-05,
+      "loss": 5.8455,
+      "step": 14320
+    },
+    {
+      "epoch": 0.02013499438840657,
+      "grad_norm": 1.4030274152755737,
+      "learning_rate": 6.039201910917521e-05,
+      "loss": 5.8089,
+      "step": 14330
+    },
+    {
+      "epoch": 0.020149045326570147,
+      "grad_norm": 1.470107913017273,
+      "learning_rate": 6.043417170155964e-05,
+      "loss": 5.8458,
+      "step": 14340
+    },
+    {
+      "epoch": 0.020163096264733726,
+      "grad_norm": 1.3324445486068726,
+      "learning_rate": 6.0476324293944076e-05,
+      "loss": 5.9478,
+      "step": 14350
+    },
+    {
+      "epoch": 0.020177147202897303,
+      "grad_norm": 1.3293122053146362,
+      "learning_rate": 6.05184768863285e-05,
+      "loss": 5.8075,
+      "step": 14360
+    },
+    {
+      "epoch": 0.020191198141060882,
+      "grad_norm": 1.3335965871810913,
+      "learning_rate": 6.056062947871293e-05,
+      "loss": 6.0072,
+      "step": 14370
+    },
+    {
+      "epoch": 0.02020524907922446,
+      "grad_norm": 1.4222981929779053,
+      "learning_rate": 6.060278207109737e-05,
+      "loss": 5.9405,
+      "step": 14380
+    },
+    {
+      "epoch": 0.020219300017388035,
+      "grad_norm": 1.292315125465393,
+      "learning_rate": 6.06449346634818e-05,
+      "loss": 5.959,
+      "step": 14390
+    },
+    {
+      "epoch": 0.020233350955551614,
+      "grad_norm": 1.3387219905853271,
+      "learning_rate": 6.0687087255866225e-05,
+      "loss": 5.9308,
+      "step": 14400
+    },
+    {
+      "epoch": 0.02024740189371519,
+      "grad_norm": 1.4554129838943481,
+      "learning_rate": 6.0729239848250663e-05,
+      "loss": 5.8909,
+      "step": 14410
+    },
+    {
+      "epoch": 0.02026145283187877,
+      "grad_norm": Infinity,
+      "learning_rate": 6.0771392440635095e-05,
+      "loss": 5.927,
+      "step": 14420
+    },
+    {
+      "epoch": 0.020275503770042346,
+      "grad_norm": 1.4156428575515747,
+      "learning_rate": 6.0809329773781084e-05,
+      "loss": 5.964,
+      "step": 14430
+    },
+    {
+      "epoch": 0.020289554708205922,
+      "grad_norm": 1.5934478044509888,
+      "learning_rate": 6.0851482366165515e-05,
+      "loss": 5.7974,
+      "step": 14440
+    },
+    {
+      "epoch": 0.020303605646369502,
+      "grad_norm": 1.6974800825119019,
+      "learning_rate": 6.089363495854994e-05,
+      "loss": 5.913,
+      "step": 14450
+    },
+    {
+      "epoch": 0.020317656584533078,
+      "grad_norm": 1.4129645824432373,
+      "learning_rate": 6.093578755093438e-05,
+      "loss": 5.8084,
+      "step": 14460
+    },
+    {
+      "epoch": 0.020331707522696658,
+      "grad_norm": 1.3150914907455444,
+      "learning_rate": 6.097794014331881e-05,
+      "loss": 5.8431,
+      "step": 14470
+    },
+    {
+      "epoch": 0.020345758460860234,
+      "grad_norm": 1.4848755598068237,
+      "learning_rate": 6.1020092735703247e-05,
+      "loss": 5.8406,
+      "step": 14480
+    },
+    {
+      "epoch": 0.02035980939902381,
+      "grad_norm": 1.348905324935913,
+      "learning_rate": 6.106224532808767e-05,
+      "loss": 5.9378,
+      "step": 14490
+    },
+    {
+      "epoch": 0.02037386033718739,
+      "grad_norm": 1.266818881034851,
+      "learning_rate": 6.11043979204721e-05,
+      "loss": 5.8756,
+      "step": 14500
+    },
+    {
+      "epoch": 0.020387911275350966,
+      "grad_norm": 1.3369197845458984,
+      "learning_rate": 6.114655051285653e-05,
+      "loss": 5.8134,
+      "step": 14510
+    },
+    {
+      "epoch": 0.020401962213514542,
+      "grad_norm": 1.5117855072021484,
+      "learning_rate": 6.118870310524096e-05,
+      "loss": 5.8981,
+      "step": 14520
+    },
+    {
+      "epoch": 0.02041601315167812,
+      "grad_norm": 1.425666332244873,
+      "learning_rate": 6.12308556976254e-05,
+      "loss": 5.9648,
+      "step": 14530
+    },
+    {
+      "epoch": 0.020430064089841698,
+      "grad_norm": 1.3908796310424805,
+      "learning_rate": 6.127300829000983e-05,
+      "loss": 5.7853,
+      "step": 14540
+    },
+    {
+      "epoch": 0.020444115028005277,
+      "grad_norm": 1.3612520694732666,
+      "learning_rate": 6.131516088239426e-05,
+      "loss": 5.7987,
+      "step": 14550
+    },
+    {
+      "epoch": 0.020458165966168854,
+      "grad_norm": 1.5223125219345093,
+      "learning_rate": 6.13573134747787e-05,
+      "loss": 5.9002,
+      "step": 14560
+    },
+    {
+      "epoch": 0.02047221690433243,
+      "grad_norm": 1.5592213869094849,
+      "learning_rate": 6.139946606716312e-05,
+      "loss": 5.867,
+      "step": 14570
+    },
+    {
+      "epoch": 0.02048626784249601,
+      "grad_norm": 1.5414535999298096,
+      "learning_rate": 6.144161865954755e-05,
+      "loss": 5.7891,
+      "step": 14580
+    },
+    {
+      "epoch": 0.020500318780659586,
+      "grad_norm": 1.3643969297409058,
+      "learning_rate": 6.1483771251932e-05,
+      "loss": 5.8342,
+      "step": 14590
+    },
+    {
+      "epoch": 0.020514369718823165,
+      "grad_norm": 1.3756287097930908,
+      "learning_rate": 6.152592384431641e-05,
+      "loss": 5.9926,
+      "step": 14600
+    },
+    {
+      "epoch": 0.02052842065698674,
+      "grad_norm": 1.461431622505188,
+      "learning_rate": 6.156807643670086e-05,
+      "loss": 5.8282,
+      "step": 14610
+    },
+    {
+      "epoch": 0.020542471595150318,
+      "grad_norm": 1.5052664279937744,
+      "learning_rate": 6.161022902908529e-05,
+      "loss": 5.867,
+      "step": 14620
+    },
+    {
+      "epoch": 0.020556522533313897,
+      "grad_norm": 1.815929651260376,
+      "learning_rate": 6.165238162146971e-05,
+      "loss": 5.875,
+      "step": 14630
+    },
+    {
+      "epoch": 0.020570573471477473,
+      "grad_norm": 1.991647481918335,
+      "learning_rate": 6.169453421385415e-05,
+      "loss": 5.767,
+      "step": 14640
+    },
+    {
+      "epoch": 0.02058462440964105,
+      "grad_norm": 1.380648136138916,
+      "learning_rate": 6.173668680623858e-05,
+      "loss": 5.9249,
+      "step": 14650
+    },
+    {
+      "epoch": 0.02059867534780463,
+      "grad_norm": 1.3880946636199951,
+      "learning_rate": 6.1778839398623e-05,
+      "loss": 6.0072,
+      "step": 14660
+    },
+    {
+      "epoch": 0.020612726285968205,
+      "grad_norm": 1.4181009531021118,
+      "learning_rate": 6.182099199100745e-05,
+      "loss": 5.7029,
+      "step": 14670
+    },
+    {
+      "epoch": 0.020626777224131785,
+      "grad_norm": 1.4614509344100952,
+      "learning_rate": 6.186314458339188e-05,
+      "loss": 5.778,
+      "step": 14680
+    },
+    {
+      "epoch": 0.02064082816229536,
+      "grad_norm": 1.4017057418823242,
+      "learning_rate": 6.190529717577631e-05,
+      "loss": 5.7973,
+      "step": 14690
+    },
+    {
+      "epoch": 0.020654879100458937,
+      "grad_norm": 1.4342793226242065,
+      "learning_rate": 6.194744976816074e-05,
+      "loss": 5.8359,
+      "step": 14700
+    },
+    {
+      "epoch": 0.020668930038622517,
+      "grad_norm": 1.4066189527511597,
+      "learning_rate": 6.198960236054517e-05,
+      "loss": 5.8003,
+      "step": 14710
+    },
+    {
+      "epoch": 0.020682980976786093,
+      "grad_norm": 1.3695813417434692,
+      "learning_rate": 6.20317549529296e-05,
+      "loss": 5.9072,
+      "step": 14720
+    },
+    {
+      "epoch": 0.020697031914949673,
+      "grad_norm": 1.7797951698303223,
+      "learning_rate": 6.207390754531403e-05,
+      "loss": 5.772,
+      "step": 14730
+    },
+    {
+      "epoch": 0.02071108285311325,
+      "grad_norm": 1.340478539466858,
+      "learning_rate": 6.211606013769846e-05,
+      "loss": 5.9631,
+      "step": 14740
+    },
+    {
+      "epoch": 0.020725133791276825,
+      "grad_norm": 1.3329967260360718,
+      "learning_rate": 6.21582127300829e-05,
+      "loss": 5.899,
+      "step": 14750
+    },
+    {
+      "epoch": 0.020739184729440405,
+      "grad_norm": 1.4106431007385254,
+      "learning_rate": 6.220036532246733e-05,
+      "loss": 5.9145,
+      "step": 14760
+    },
+    {
+      "epoch": 0.02075323566760398,
+      "grad_norm": 1.4221521615982056,
+      "learning_rate": 6.224251791485176e-05,
+      "loss": 5.9208,
+      "step": 14770
+    },
+    {
+      "epoch": 0.02076728660576756,
+      "grad_norm": 1.3601226806640625,
+      "learning_rate": 6.228467050723619e-05,
+      "loss": 5.8966,
+      "step": 14780
+    },
+    {
+      "epoch": 0.020781337543931137,
+      "grad_norm": 1.3591755628585815,
+      "learning_rate": 6.232682309962062e-05,
+      "loss": 5.7644,
+      "step": 14790
+    },
+    {
+      "epoch": 0.020795388482094713,
+      "grad_norm": 1.4168270826339722,
+      "learning_rate": 6.236897569200505e-05,
+      "loss": 5.8457,
+      "step": 14800
+    },
+    {
+      "epoch": 0.020809439420258292,
+      "grad_norm": 1.443789005279541,
+      "learning_rate": 6.241112828438948e-05,
+      "loss": 5.8221,
+      "step": 14810
+    },
+    {
+      "epoch": 0.02082349035842187,
+      "grad_norm": 1.4544609785079956,
+      "learning_rate": 6.245328087677393e-05,
+      "loss": 5.8429,
+      "step": 14820
+    },
+    {
+      "epoch": 0.020837541296585445,
+      "grad_norm": 1.2243056297302246,
+      "learning_rate": 6.249543346915834e-05,
+      "loss": 5.8578,
+      "step": 14830
+    },
+    {
+      "epoch": 0.020851592234749024,
+      "grad_norm": 1.3254730701446533,
+      "learning_rate": 6.253758606154278e-05,
+      "loss": 5.9038,
+      "step": 14840
+    },
+    {
+      "epoch": 0.0208656431729126,
+      "grad_norm": 1.5144115686416626,
+      "learning_rate": 6.257973865392722e-05,
+      "loss": 5.81,
+      "step": 14850
+    },
+    {
+      "epoch": 0.02087969411107618,
+      "grad_norm": 1.35421621799469,
+      "learning_rate": 6.262189124631164e-05,
+      "loss": 5.8403,
+      "step": 14860
+    },
+    {
+      "epoch": 0.020893745049239756,
+      "grad_norm": 1.3334990739822388,
+      "learning_rate": 6.266404383869607e-05,
+      "loss": 5.7432,
+      "step": 14870
+    },
+    {
+      "epoch": 0.020907795987403333,
+      "grad_norm": 1.3845977783203125,
+      "learning_rate": 6.270619643108051e-05,
+      "loss": 5.7457,
+      "step": 14880
+    },
+    {
+      "epoch": 0.020921846925566912,
+      "grad_norm": 1.606353521347046,
+      "learning_rate": 6.274834902346493e-05,
+      "loss": 5.7664,
+      "step": 14890
+    },
+    {
+      "epoch": 0.02093589786373049,
+      "grad_norm": 1.4412970542907715,
+      "learning_rate": 6.279050161584938e-05,
+      "loss": 5.8228,
+      "step": 14900
+    },
+    {
+      "epoch": 0.020949948801894068,
+      "grad_norm": 1.57476007938385,
+      "learning_rate": 6.283265420823381e-05,
+      "loss": 5.8189,
+      "step": 14910
+    },
+    {
+      "epoch": 0.020963999740057644,
+      "grad_norm": 1.4712239503860474,
+      "learning_rate": 6.287480680061822e-05,
+      "loss": 5.8303,
+      "step": 14920
+    },
+    {
+      "epoch": 0.02097805067822122,
+      "grad_norm": 1.3110334873199463,
+      "learning_rate": 6.291695939300267e-05,
+      "loss": 5.8904,
+      "step": 14930
+    },
+    {
+      "epoch": 0.0209921016163848,
+      "grad_norm": 1.3854535818099976,
+      "learning_rate": 6.29591119853871e-05,
+      "loss": 5.7378,
+      "step": 14940
+    },
+    {
+      "epoch": 0.021006152554548376,
+      "grad_norm": 1.5063308477401733,
+      "learning_rate": 6.300126457777152e-05,
+      "loss": 5.8302,
+      "step": 14950
+    },
+    {
+      "epoch": 0.021020203492711952,
+      "grad_norm": 1.347753643989563,
+      "learning_rate": 6.304341717015596e-05,
+      "loss": 5.8284,
+      "step": 14960
+    },
+    {
+      "epoch": 0.021034254430875532,
+      "grad_norm": 1.7087639570236206,
+      "learning_rate": 6.30855697625404e-05,
+      "loss": 5.9404,
+      "step": 14970
+    },
+    {
+      "epoch": 0.021048305369039108,
+      "grad_norm": 1.4151090383529663,
+      "learning_rate": 6.312772235492483e-05,
+      "loss": 5.8551,
+      "step": 14980
+    },
+    {
+      "epoch": 0.021062356307202688,
+      "grad_norm": 1.3254867792129517,
+      "learning_rate": 6.316987494730926e-05,
+      "loss": 5.8843,
+      "step": 14990
+    },
+    {
+      "epoch": 0.021076407245366264,
+      "grad_norm": 1.3656538724899292,
+      "learning_rate": 6.321202753969369e-05,
+      "loss": 5.8179,
+      "step": 15000
+    },
+    {
+      "epoch": 0.02109045818352984,
+      "grad_norm": 1.3731048107147217,
+      "learning_rate": 6.325418013207812e-05,
+      "loss": 5.776,
+      "step": 15010
+    },
+    {
+      "epoch": 0.02110450912169342,
+      "grad_norm": 1.6539822816848755,
+      "learning_rate": 6.329633272446255e-05,
+      "loss": 5.796,
+      "step": 15020
+    },
+    {
+      "epoch": 0.021118560059856996,
+      "grad_norm": 1.6321409940719604,
+      "learning_rate": 6.333848531684698e-05,
+      "loss": 5.7292,
+      "step": 15030
+    },
+    {
+      "epoch": 0.021132610998020575,
+      "grad_norm": 1.516080379486084,
+      "learning_rate": 6.338063790923141e-05,
+      "loss": 5.7586,
+      "step": 15040
+    },
+    {
+      "epoch": 0.02114666193618415,
+      "grad_norm": 1.3451189994812012,
+      "learning_rate": 6.342279050161584e-05,
+      "loss": 5.7364,
+      "step": 15050
+    },
+    {
+      "epoch": 0.021160712874347728,
+      "grad_norm": 1.4270164966583252,
+      "learning_rate": 6.346494309400027e-05,
+      "loss": 5.7612,
+      "step": 15060
+    },
+    {
+      "epoch": 0.021174763812511307,
+      "grad_norm": 1.353387475013733,
+      "learning_rate": 6.35070956863847e-05,
+      "loss": 5.8736,
+      "step": 15070
+    },
+    {
+      "epoch": 0.021188814750674884,
+      "grad_norm": 1.4181205034255981,
+      "learning_rate": 6.354924827876914e-05,
+      "loss": 5.9262,
+      "step": 15080
+    },
+    {
+      "epoch": 0.021202865688838463,
+      "grad_norm": 1.3350406885147095,
+      "learning_rate": 6.359140087115357e-05,
+      "loss": 5.843,
+      "step": 15090
+    },
+    {
+      "epoch": 0.02121691662700204,
+      "grad_norm": 1.676595687866211,
+      "learning_rate": 6.3633553463538e-05,
+      "loss": 5.806,
+      "step": 15100
+    },
+    {
+      "epoch": 0.021230967565165616,
+      "grad_norm": 1.579698920249939,
+      "learning_rate": 6.367570605592244e-05,
+      "loss": 5.8444,
+      "step": 15110
+    },
+    {
+      "epoch": 0.021245018503329195,
+      "grad_norm": 1.5201925039291382,
+      "learning_rate": 6.371785864830686e-05,
+      "loss": 5.948,
+      "step": 15120
+    },
+    {
+      "epoch": 0.02125906944149277,
+      "grad_norm": 1.3935141563415527,
+      "learning_rate": 6.376001124069129e-05,
+      "loss": 5.8822,
+      "step": 15130
+    },
+    {
+      "epoch": 0.021273120379656348,
+      "grad_norm": 1.358180284500122,
+      "learning_rate": 6.380216383307574e-05,
+      "loss": 5.8469,
+      "step": 15140
+    },
+    {
+      "epoch": 0.021287171317819927,
+      "grad_norm": 1.4216656684875488,
+      "learning_rate": 6.384431642546016e-05,
+      "loss": 5.8851,
+      "step": 15150
+    },
+    {
+      "epoch": 0.021301222255983503,
+      "grad_norm": 1.40565025806427,
+      "learning_rate": 6.388646901784459e-05,
+      "loss": 5.7583,
+      "step": 15160
+    },
+    {
+      "epoch": 0.021315273194147083,
+      "grad_norm": 1.30914306640625,
+      "learning_rate": 6.392862161022903e-05,
+      "loss": 5.9395,
+      "step": 15170
+    },
+    {
+      "epoch": 0.02132932413231066,
+      "grad_norm": 1.4431228637695312,
+      "learning_rate": 6.397077420261345e-05,
+      "loss": 5.776,
+      "step": 15180
+    },
+    {
+      "epoch": 0.021343375070474235,
+      "grad_norm": 1.397368311882019,
+      "learning_rate": 6.40129267949979e-05,
+      "loss": 5.7714,
+      "step": 15190
+    },
+    {
+      "epoch": 0.021357426008637815,
+      "grad_norm": 1.303511142730713,
+      "learning_rate": 6.405507938738232e-05,
+      "loss": 5.8594,
+      "step": 15200
+    },
+    {
+      "epoch": 0.02137147694680139,
+      "grad_norm": 1.359408974647522,
+      "learning_rate": 6.409723197976674e-05,
+      "loss": 5.7992,
+      "step": 15210
+    },
+    {
+      "epoch": 0.02138552788496497,
+      "grad_norm": 1.2777255773544312,
+      "learning_rate": 6.413938457215119e-05,
+      "loss": 5.8816,
+      "step": 15220
+    },
+    {
+      "epoch": 0.021399578823128547,
+      "grad_norm": 1.7330981492996216,
+      "learning_rate": 6.418153716453562e-05,
+      "loss": 5.7651,
+      "step": 15230
+    },
+    {
+      "epoch": 0.021413629761292123,
+      "grad_norm": 1.324683427810669,
+      "learning_rate": 6.42194744976816e-05,
+      "loss": 5.7404,
+      "step": 15240
+    },
+    {
+      "epoch": 0.021427680699455703,
+      "grad_norm": 1.3045562505722046,
+      "learning_rate": 6.426162709006603e-05,
+      "loss": 5.7965,
+      "step": 15250
+    },
+    {
+      "epoch": 0.02144173163761928,
+      "grad_norm": 1.3371610641479492,
+      "learning_rate": 6.430377968245046e-05,
+      "loss": 5.8492,
+      "step": 15260
+    },
+    {
+      "epoch": 0.02145578257578286,
+      "grad_norm": 1.61046302318573,
+      "learning_rate": 6.43459322748349e-05,
+      "loss": 5.8089,
+      "step": 15270
+    },
+    {
+      "epoch": 0.021469833513946435,
+      "grad_norm": 1.5446206331253052,
+      "learning_rate": 6.438808486721933e-05,
+      "loss": 5.8823,
+      "step": 15280
+    },
+    {
+      "epoch": 0.02148388445211001,
+      "grad_norm": 1.337916612625122,
+      "learning_rate": 6.443023745960376e-05,
+      "loss": 5.8088,
+      "step": 15290
+    },
+    {
+      "epoch": 0.02149793539027359,
+      "grad_norm": 1.5150680541992188,
+      "learning_rate": 6.447239005198819e-05,
+      "loss": 5.7181,
+      "step": 15300
+    },
+    {
+      "epoch": 0.021511986328437167,
+      "grad_norm": 1.3614176511764526,
+      "learning_rate": 6.451454264437262e-05,
+      "loss": 5.8309,
+      "step": 15310
+    },
+    {
+      "epoch": 0.021526037266600743,
+      "grad_norm": 1.357839584350586,
+      "learning_rate": 6.455669523675706e-05,
+      "loss": 5.8946,
+      "step": 15320
+    },
+    {
+      "epoch": 0.021540088204764322,
+      "grad_norm": 1.4024747610092163,
+      "learning_rate": 6.459884782914148e-05,
+      "loss": 5.8192,
+      "step": 15330
+    },
+    {
+      "epoch": 0.0215541391429279,
+      "grad_norm": 1.4322363138198853,
+      "learning_rate": 6.464100042152591e-05,
+      "loss": 5.6945,
+      "step": 15340
+    },
+    {
+      "epoch": 0.021568190081091478,
+      "grad_norm": 1.4403985738754272,
+      "learning_rate": 6.468315301391036e-05,
+      "loss": 5.926,
+      "step": 15350
+    },
+    {
+      "epoch": 0.021582241019255054,
+      "grad_norm": 1.593833565711975,
+      "learning_rate": 6.472530560629477e-05,
+      "loss": 5.8134,
+      "step": 15360
+    },
+    {
+      "epoch": 0.02159629195741863,
+      "grad_norm": 1.4765543937683105,
+      "learning_rate": 6.47674581986792e-05,
+      "loss": 5.8854,
+      "step": 15370
+    },
+    {
+      "epoch": 0.02161034289558221,
+      "grad_norm": 1.3382881879806519,
+      "learning_rate": 6.480961079106365e-05,
+      "loss": 5.8777,
+      "step": 15380
+    },
+    {
+      "epoch": 0.021624393833745786,
+      "grad_norm": 1.6510887145996094,
+      "learning_rate": 6.485176338344807e-05,
+      "loss": 5.7864,
+      "step": 15390
+    },
+    {
+      "epoch": 0.021638444771909366,
+      "grad_norm": 1.3946115970611572,
+      "learning_rate": 6.489391597583251e-05,
+      "loss": 5.7041,
+      "step": 15400
+    },
+    {
+      "epoch": 0.021652495710072942,
+      "grad_norm": 1.3984819650650024,
+      "learning_rate": 6.493606856821694e-05,
+      "loss": 5.7186,
+      "step": 15410
+    },
+    {
+      "epoch": 0.02166654664823652,
+      "grad_norm": 1.9340324401855469,
+      "learning_rate": 6.497822116060136e-05,
+      "loss": 5.7089,
+      "step": 15420
+    },
+    {
+      "epoch": 0.021680597586400098,
+      "grad_norm": 1.4083871841430664,
+      "learning_rate": 6.50203737529858e-05,
+      "loss": 5.8155,
+      "step": 15430
+    },
+    {
+      "epoch": 0.021694648524563674,
+      "grad_norm": 1.389730453491211,
+      "learning_rate": 6.506252634537024e-05,
+      "loss": 5.7463,
+      "step": 15440
+    },
+    {
+      "epoch": 0.02170869946272725,
+      "grad_norm": 1.2647806406021118,
+      "learning_rate": 6.510467893775467e-05,
+      "loss": 5.8177,
+      "step": 15450
+    },
+    {
+      "epoch": 0.02172275040089083,
+      "grad_norm": 1.4326426982879639,
+      "learning_rate": 6.51468315301391e-05,
+      "loss": 5.7414,
+      "step": 15460
+    },
+    {
+      "epoch": 0.021736801339054406,
+      "grad_norm": 1.3881163597106934,
+      "learning_rate": 6.518898412252353e-05,
+      "loss": 5.624,
+      "step": 15470
+    },
+    {
+      "epoch": 0.021750852277217986,
+      "grad_norm": 1.3433114290237427,
+      "learning_rate": 6.523113671490796e-05,
+      "loss": 5.7344,
+      "step": 15480
+    },
+    {
+      "epoch": 0.021764903215381562,
+      "grad_norm": 1.3703086376190186,
+      "learning_rate": 6.52732893072924e-05,
+      "loss": 5.7321,
+      "step": 15490
+    },
+    {
+      "epoch": 0.021778954153545138,
+      "grad_norm": 1.3004043102264404,
+      "learning_rate": 6.531544189967682e-05,
+      "loss": 5.7776,
+      "step": 15500
+    },
+    {
+      "epoch": 0.021793005091708718,
+      "grad_norm": 1.3912044763565063,
+      "learning_rate": 6.535759449206126e-05,
+      "loss": 5.75,
+      "step": 15510
+    },
+    {
+      "epoch": 0.021807056029872294,
+      "grad_norm": 1.3019267320632935,
+      "learning_rate": 6.539974708444569e-05,
+      "loss": 5.8134,
+      "step": 15520
+    },
+    {
+      "epoch": 0.021821106968035874,
+      "grad_norm": 1.2605361938476562,
+      "learning_rate": 6.544189967683012e-05,
+      "loss": 5.7915,
+      "step": 15530
+    },
+    {
+      "epoch": 0.02183515790619945,
+      "grad_norm": 1.4486380815505981,
+      "learning_rate": 6.548405226921455e-05,
+      "loss": 5.8089,
+      "step": 15540
+    },
+    {
+      "epoch": 0.021849208844363026,
+      "grad_norm": 1.4889944791793823,
+      "learning_rate": 6.552620486159898e-05,
+      "loss": 5.8072,
+      "step": 15550
+    },
+    {
+      "epoch": 0.021863259782526605,
+      "grad_norm": 1.389483094215393,
+      "learning_rate": 6.556835745398341e-05,
+      "loss": 5.8426,
+      "step": 15560
+    },
+    {
+      "epoch": 0.02187731072069018,
+      "grad_norm": 1.465795874595642,
+      "learning_rate": 6.561051004636784e-05,
+      "loss": 5.7026,
+      "step": 15570
+    },
+    {
+      "epoch": 0.02189136165885376,
+      "grad_norm": 1.4371973276138306,
+      "learning_rate": 6.565266263875227e-05,
+      "loss": 5.8007,
+      "step": 15580
+    },
+    {
+      "epoch": 0.021905412597017337,
+      "grad_norm": 1.3100478649139404,
+      "learning_rate": 6.56948152311367e-05,
+      "loss": 5.7882,
+      "step": 15590
+    },
+    {
+      "epoch": 0.021919463535180914,
+      "grad_norm": 1.3720271587371826,
+      "learning_rate": 6.573696782352114e-05,
+      "loss": 5.7386,
+      "step": 15600
+    },
+    {
+      "epoch": 0.021933514473344493,
+      "grad_norm": 1.6056287288665771,
+      "learning_rate": 6.577912041590558e-05,
+      "loss": 5.7643,
+      "step": 15610
+    },
+    {
+      "epoch": 0.02194756541150807,
+      "grad_norm": 1.4117242097854614,
+      "learning_rate": 6.582127300829e-05,
+      "loss": 5.8054,
+      "step": 15620
+    },
+    {
+      "epoch": 0.021961616349671646,
+      "grad_norm": 1.3101048469543457,
+      "learning_rate": 6.586342560067443e-05,
+      "loss": 5.8407,
+      "step": 15630
+    },
+    {
+      "epoch": 0.021975667287835225,
+      "grad_norm": 1.535385251045227,
+      "learning_rate": 6.590557819305887e-05,
+      "loss": 5.7486,
+      "step": 15640
+    },
+    {
+      "epoch": 0.0219897182259988,
+      "grad_norm": 1.3539336919784546,
+      "learning_rate": 6.594773078544329e-05,
+      "loss": 5.7473,
+      "step": 15650
+    },
+    {
+      "epoch": 0.02200376916416238,
+      "grad_norm": 1.5434645414352417,
+      "learning_rate": 6.598988337782774e-05,
+      "loss": 5.8255,
+      "step": 15660
+    },
+    {
+      "epoch": 0.022017820102325957,
+      "grad_norm": 1.3859068155288696,
+      "learning_rate": 6.603203597021217e-05,
+      "loss": 5.8002,
+      "step": 15670
+    },
+    {
+      "epoch": 0.022031871040489533,
+      "grad_norm": 1.397512674331665,
+      "learning_rate": 6.607418856259659e-05,
+      "loss": 5.8374,
+      "step": 15680
+    },
+    {
+      "epoch": 0.022045921978653113,
+      "grad_norm": 1.4506967067718506,
+      "learning_rate": 6.611634115498103e-05,
+      "loss": 5.8459,
+      "step": 15690
+    },
+    {
+      "epoch": 0.02205997291681669,
+      "grad_norm": 1.3882638216018677,
+      "learning_rate": 6.615849374736546e-05,
+      "loss": 5.8247,
+      "step": 15700
+    },
+    {
+      "epoch": 0.02207402385498027,
+      "grad_norm": 1.308134913444519,
+      "learning_rate": 6.620064633974988e-05,
+      "loss": 5.7157,
+      "step": 15710
+    },
+    {
+      "epoch": 0.022088074793143845,
+      "grad_norm": 1.4445452690124512,
+      "learning_rate": 6.624279893213432e-05,
+      "loss": 5.7608,
+      "step": 15720
+    },
+    {
+      "epoch": 0.02210212573130742,
+      "grad_norm": 1.362390398979187,
+      "learning_rate": 6.628495152451875e-05,
+      "loss": 5.9047,
+      "step": 15730
+    },
+    {
+      "epoch": 0.022116176669471,
+      "grad_norm": 1.2991571426391602,
+      "learning_rate": 6.632710411690319e-05,
+      "loss": 5.7554,
+      "step": 15740
+    },
+    {
+      "epoch": 0.022130227607634577,
+      "grad_norm": 1.328895926475525,
+      "learning_rate": 6.636925670928762e-05,
+      "loss": 5.7317,
+      "step": 15750
+    },
+    {
+      "epoch": 0.022144278545798153,
+      "grad_norm": 1.303283929824829,
+      "learning_rate": 6.641140930167205e-05,
+      "loss": 5.7671,
+      "step": 15760
+    },
+    {
+      "epoch": 0.022158329483961733,
+      "grad_norm": 1.358275055885315,
+      "learning_rate": 6.645356189405648e-05,
+      "loss": 5.7322,
+      "step": 15770
+    },
+    {
+      "epoch": 0.02217238042212531,
+      "grad_norm": 1.4214792251586914,
+      "learning_rate": 6.649571448644091e-05,
+      "loss": 5.6888,
+      "step": 15780
+    },
+    {
+      "epoch": 0.02218643136028889,
+      "grad_norm": 1.3279885053634644,
+      "learning_rate": 6.653786707882534e-05,
+      "loss": 5.7591,
+      "step": 15790
+    },
+    {
+      "epoch": 0.022200482298452465,
+      "grad_norm": 1.3507026433944702,
+      "learning_rate": 6.658001967120977e-05,
+      "loss": 5.7571,
+      "step": 15800
+    },
+    {
+      "epoch": 0.02221453323661604,
+      "grad_norm": 1.323520541191101,
+      "learning_rate": 6.66221722635942e-05,
+      "loss": 5.8356,
+      "step": 15810
+    },
+    {
+      "epoch": 0.02222858417477962,
+      "grad_norm": 1.2954471111297607,
+      "learning_rate": 6.666432485597864e-05,
+      "loss": 5.9235,
+      "step": 15820
+    },
+    {
+      "epoch": 0.022242635112943197,
+      "grad_norm": 1.3912978172302246,
+      "learning_rate": 6.670647744836307e-05,
+      "loss": 5.7872,
+      "step": 15830
+    },
+    {
+      "epoch": 0.022256686051106776,
+      "grad_norm": 1.5537766218185425,
+      "learning_rate": 6.67486300407475e-05,
+      "loss": 5.7668,
+      "step": 15840
+    },
+    {
+      "epoch": 0.022270736989270352,
+      "grad_norm": 1.3550326824188232,
+      "learning_rate": 6.679078263313193e-05,
+      "loss": 5.7059,
+      "step": 15850
+    },
+    {
+      "epoch": 0.02228478792743393,
+      "grad_norm": 1.3379523754119873,
+      "learning_rate": 6.683293522551636e-05,
+      "loss": 5.7736,
+      "step": 15860
+    },
+    {
+      "epoch": 0.022298838865597508,
+      "grad_norm": 1.4727954864501953,
+      "learning_rate": 6.687508781790079e-05,
+      "loss": 5.687,
+      "step": 15870
+    },
+    {
+      "epoch": 0.022312889803761084,
+      "grad_norm": 1.5783716440200806,
+      "learning_rate": 6.691724041028522e-05,
+      "loss": 5.6273,
+      "step": 15880
+    },
+    {
+      "epoch": 0.022326940741924664,
+      "grad_norm": 1.2996593713760376,
+      "learning_rate": 6.695939300266965e-05,
+      "loss": 5.809,
+      "step": 15890
+    },
+    {
+      "epoch": 0.02234099168008824,
+      "grad_norm": 1.3592562675476074,
+      "learning_rate": 6.70015455950541e-05,
+      "loss": 5.7186,
+      "step": 15900
+    },
+    {
+      "epoch": 0.022355042618251816,
+      "grad_norm": 1.363147258758545,
+      "learning_rate": 6.704369818743852e-05,
+      "loss": 5.7087,
+      "step": 15910
+    },
+    {
+      "epoch": 0.022369093556415396,
+      "grad_norm": 1.3323581218719482,
+      "learning_rate": 6.708585077982295e-05,
+      "loss": 5.7382,
+      "step": 15920
+    },
+    {
+      "epoch": 0.022383144494578972,
+      "grad_norm": 1.3843733072280884,
+      "learning_rate": 6.712800337220739e-05,
+      "loss": 5.7236,
+      "step": 15930
+    },
+    {
+      "epoch": 0.02239719543274255,
+      "grad_norm": 1.336782455444336,
+      "learning_rate": 6.717015596459181e-05,
+      "loss": 5.737,
+      "step": 15940
+    },
+    {
+      "epoch": 0.022411246370906128,
+      "grad_norm": 1.3597850799560547,
+      "learning_rate": 6.721230855697625e-05,
+      "loss": 5.6848,
+      "step": 15950
+    },
+    {
+      "epoch": 0.022425297309069704,
+      "grad_norm": 1.360641360282898,
+      "learning_rate": 6.725446114936069e-05,
+      "loss": 5.8756,
+      "step": 15960
+    },
+    {
+      "epoch": 0.022439348247233284,
+      "grad_norm": 1.4128172397613525,
+      "learning_rate": 6.72966137417451e-05,
+      "loss": 5.7612,
+      "step": 15970
+    },
+    {
+      "epoch": 0.02245339918539686,
+      "grad_norm": 1.4090365171432495,
+      "learning_rate": 6.733876633412955e-05,
+      "loss": 5.6862,
+      "step": 15980
+    },
+    {
+      "epoch": 0.022467450123560436,
+      "grad_norm": 1.555239200592041,
+      "learning_rate": 6.738091892651398e-05,
+      "loss": 5.6615,
+      "step": 15990
+    },
+    {
+      "epoch": 0.022481501061724016,
+      "grad_norm": 1.3207861185073853,
+      "learning_rate": 6.74230715188984e-05,
+      "loss": 5.7863,
+      "step": 16000
+    },
+    {
+      "epoch": 0.022495551999887592,
+      "grad_norm": 1.3086305856704712,
+      "learning_rate": 6.746522411128284e-05,
+      "loss": 5.7663,
+      "step": 16010
+    },
+    {
+      "epoch": 0.02250960293805117,
+      "grad_norm": 1.5462536811828613,
+      "learning_rate": 6.750737670366727e-05,
+      "loss": 5.7121,
+      "step": 16020
+    },
+    {
+      "epoch": 0.022523653876214748,
+      "grad_norm": 1.4813846349716187,
+      "learning_rate": 6.75495292960517e-05,
+      "loss": 5.7485,
+      "step": 16030
+    },
+    {
+      "epoch": 0.022537704814378324,
+      "grad_norm": 1.3974696397781372,
+      "learning_rate": 6.759168188843613e-05,
+      "loss": 5.6604,
+      "step": 16040
+    },
+    {
+      "epoch": 0.022551755752541904,
+      "grad_norm": 1.3554956912994385,
+      "learning_rate": 6.763383448082057e-05,
+      "loss": 5.5456,
+      "step": 16050
+    },
+    {
+      "epoch": 0.02256580669070548,
+      "grad_norm": 1.4472044706344604,
+      "learning_rate": 6.7675987073205e-05,
+      "loss": 5.6876,
+      "step": 16060
+    },
+    {
+      "epoch": 0.022579857628869056,
+      "grad_norm": 1.4171758890151978,
+      "learning_rate": 6.771813966558943e-05,
+      "loss": 5.7768,
+      "step": 16070
+    },
+    {
+      "epoch": 0.022593908567032635,
+      "grad_norm": 1.4032776355743408,
+      "learning_rate": 6.776029225797386e-05,
+      "loss": 5.6466,
+      "step": 16080
+    },
+    {
+      "epoch": 0.02260795950519621,
+      "grad_norm": 1.2590081691741943,
+      "learning_rate": 6.780244485035829e-05,
+      "loss": 5.6616,
+      "step": 16090
+    },
+    {
+      "epoch": 0.02262201044335979,
+      "grad_norm": 1.3556309938430786,
+      "learning_rate": 6.784459744274272e-05,
+      "loss": 5.797,
+      "step": 16100
+    },
+    {
+      "epoch": 0.022636061381523367,
+      "grad_norm": 1.6029269695281982,
+      "learning_rate": 6.788675003512715e-05,
+      "loss": 5.7445,
+      "step": 16110
+    },
+    {
+      "epoch": 0.022650112319686944,
+      "grad_norm": 1.6440356969833374,
+      "learning_rate": 6.792890262751158e-05,
+      "loss": 5.6423,
+      "step": 16120
+    },
+    {
+      "epoch": 0.022664163257850523,
+      "grad_norm": 1.447268009185791,
+      "learning_rate": 6.797105521989602e-05,
+      "loss": 5.8162,
+      "step": 16130
+    },
+    {
+      "epoch": 0.0226782141960141,
+      "grad_norm": 1.3291443586349487,
+      "learning_rate": 6.801320781228045e-05,
+      "loss": 5.822,
+      "step": 16140
+    },
+    {
+      "epoch": 0.02269226513417768,
+      "grad_norm": 1.4091272354125977,
+      "learning_rate": 6.805536040466488e-05,
+      "loss": 5.7244,
+      "step": 16150
+    },
+    {
+      "epoch": 0.022706316072341255,
+      "grad_norm": 1.579410433769226,
+      "learning_rate": 6.809751299704932e-05,
+      "loss": 5.7525,
+      "step": 16160
+    },
+    {
+      "epoch": 0.02272036701050483,
+      "grad_norm": 1.368188500404358,
+      "learning_rate": 6.813966558943374e-05,
+      "loss": 5.7373,
+      "step": 16170
+    },
+    {
+      "epoch": 0.02273441794866841,
+      "grad_norm": 1.3030613660812378,
+      "learning_rate": 6.818181818181817e-05,
+      "loss": 5.7726,
+      "step": 16180
+    },
+    {
+      "epoch": 0.022748468886831987,
+      "grad_norm": 1.3315412998199463,
+      "learning_rate": 6.822397077420262e-05,
+      "loss": 5.7665,
+      "step": 16190
+    },
+    {
+      "epoch": 0.022762519824995567,
+      "grad_norm": 1.265251636505127,
+      "learning_rate": 6.826612336658703e-05,
+      "loss": 5.7164,
+      "step": 16200
+    },
+    {
+      "epoch": 0.022776570763159143,
+      "grad_norm": 1.3779250383377075,
+      "learning_rate": 6.830827595897146e-05,
+      "loss": 5.8774,
+      "step": 16210
+    },
+    {
+      "epoch": 0.02279062170132272,
+      "grad_norm": 1.3391507863998413,
+      "learning_rate": 6.835042855135591e-05,
+      "loss": 5.6763,
+      "step": 16220
+    },
+    {
+      "epoch": 0.0228046726394863,
+      "grad_norm": 1.2613348960876465,
+      "learning_rate": 6.839258114374033e-05,
+      "loss": 5.7929,
+      "step": 16230
+    },
+    {
+      "epoch": 0.022818723577649875,
+      "grad_norm": 1.3553537130355835,
+      "learning_rate": 6.843473373612477e-05,
+      "loss": 5.6879,
+      "step": 16240
+    },
+    {
+      "epoch": 0.02283277451581345,
+      "grad_norm": 1.3512293100357056,
+      "learning_rate": 6.84768863285092e-05,
+      "loss": 5.7638,
+      "step": 16250
+    },
+    {
+      "epoch": 0.02284682545397703,
+      "grad_norm": 1.3155027627944946,
+      "learning_rate": 6.851903892089362e-05,
+      "loss": 5.7267,
+      "step": 16260
+    },
+    {
+      "epoch": 0.022860876392140607,
+      "grad_norm": 1.4464770555496216,
+      "learning_rate": 6.856119151327806e-05,
+      "loss": 5.6172,
+      "step": 16270
+    },
+    {
+      "epoch": 0.022874927330304187,
+      "grad_norm": 1.4445135593414307,
+      "learning_rate": 6.86033441056625e-05,
+      "loss": 5.7,
+      "step": 16280
+    },
+    {
+      "epoch": 0.022888978268467763,
+      "grad_norm": 1.5342646837234497,
+      "learning_rate": 6.864549669804691e-05,
+      "loss": 5.6978,
+      "step": 16290
+    },
+    {
+      "epoch": 0.02290302920663134,
+      "grad_norm": 1.335935354232788,
+      "learning_rate": 6.868764929043136e-05,
+      "loss": 5.7446,
+      "step": 16300
+    },
+    {
+      "epoch": 0.02291708014479492,
+      "grad_norm": 1.326428771018982,
+      "learning_rate": 6.872980188281579e-05,
+      "loss": 5.6136,
+      "step": 16310
+    },
+    {
+      "epoch": 0.022931131082958495,
+      "grad_norm": 1.3541918992996216,
+      "learning_rate": 6.877195447520022e-05,
+      "loss": 5.7738,
+      "step": 16320
+    },
+    {
+      "epoch": 0.022945182021122074,
+      "grad_norm": 1.8092317581176758,
+      "learning_rate": 6.881410706758465e-05,
+      "loss": 5.6128,
+      "step": 16330
+    },
+    {
+      "epoch": 0.02295923295928565,
+      "grad_norm": 1.2965271472930908,
+      "learning_rate": 6.885625965996908e-05,
+      "loss": 5.8584,
+      "step": 16340
+    },
+    {
+      "epoch": 0.022973283897449227,
+      "grad_norm": 1.3288991451263428,
+      "learning_rate": 6.889841225235351e-05,
+      "loss": 5.7374,
+      "step": 16350
+    },
+    {
+      "epoch": 0.022987334835612806,
+      "grad_norm": 1.3659766912460327,
+      "learning_rate": 6.894056484473795e-05,
+      "loss": 5.6953,
+      "step": 16360
+    },
+    {
+      "epoch": 0.023001385773776382,
+      "grad_norm": 1.3702479600906372,
+      "learning_rate": 6.898271743712238e-05,
+      "loss": 5.7156,
+      "step": 16370
+    },
+    {
+      "epoch": 0.023015436711939962,
+      "grad_norm": 1.3132061958312988,
+      "learning_rate": 6.902487002950681e-05,
+      "loss": 5.6502,
+      "step": 16380
+    },
+    {
+      "epoch": 0.023029487650103538,
+      "grad_norm": 1.3793644905090332,
+      "learning_rate": 6.906702262189124e-05,
+      "loss": 5.756,
+      "step": 16390
+    },
+    {
+      "epoch": 0.023043538588267114,
+      "grad_norm": 1.364650011062622,
+      "learning_rate": 6.910917521427567e-05,
+      "loss": 5.7081,
+      "step": 16400
+    },
+    {
+      "epoch": 0.023057589526430694,
+      "grad_norm": 1.303212285041809,
+      "learning_rate": 6.91513278066601e-05,
+      "loss": 5.7405,
+      "step": 16410
+    },
+    {
+      "epoch": 0.02307164046459427,
+      "grad_norm": 1.382996678352356,
+      "learning_rate": 6.919348039904453e-05,
+      "loss": 5.6833,
+      "step": 16420
+    },
+    {
+      "epoch": 0.023085691402757846,
+      "grad_norm": 1.4374713897705078,
+      "learning_rate": 6.923563299142896e-05,
+      "loss": 5.7699,
+      "step": 16430
+    },
+    {
+      "epoch": 0.023099742340921426,
+      "grad_norm": 1.286665439605713,
+      "learning_rate": 6.92777855838134e-05,
+      "loss": 5.7886,
+      "step": 16440
+    },
+    {
+      "epoch": 0.023113793279085002,
+      "grad_norm": 1.403110384941101,
+      "learning_rate": 6.931993817619784e-05,
+      "loss": 5.7172,
+      "step": 16450
+    },
+    {
+      "epoch": 0.023127844217248582,
+      "grad_norm": 1.7290191650390625,
+      "learning_rate": 6.936209076858226e-05,
+      "loss": 5.7493,
+      "step": 16460
+    },
+    {
+      "epoch": 0.023141895155412158,
+      "grad_norm": 1.2815217971801758,
+      "learning_rate": 6.940424336096669e-05,
+      "loss": 5.7261,
+      "step": 16470
+    },
+    {
+      "epoch": 0.023155946093575734,
+      "grad_norm": 1.5143299102783203,
+      "learning_rate": 6.944639595335113e-05,
+      "loss": 5.7154,
+      "step": 16480
+    },
+    {
+      "epoch": 0.023169997031739314,
+      "grad_norm": 1.3418786525726318,
+      "learning_rate": 6.948854854573555e-05,
+      "loss": 5.7851,
+      "step": 16490
+    },
+    {
+      "epoch": 0.02318404796990289,
+      "grad_norm": 1.296385407447815,
+      "learning_rate": 6.953070113811998e-05,
+      "loss": 5.7819,
+      "step": 16500
+    },
+    {
+      "epoch": 0.02319809890806647,
+      "grad_norm": 1.3354800939559937,
+      "learning_rate": 6.957285373050443e-05,
+      "loss": 5.7586,
+      "step": 16510
+    },
+    {
+      "epoch": 0.023212149846230046,
+      "grad_norm": 1.6433796882629395,
+      "learning_rate": 6.961500632288884e-05,
+      "loss": 5.6559,
+      "step": 16520
+    },
+    {
+      "epoch": 0.023226200784393622,
+      "grad_norm": 1.4096791744232178,
+      "learning_rate": 6.965715891527329e-05,
+      "loss": 5.7111,
+      "step": 16530
+    },
+    {
+      "epoch": 0.0232402517225572,
+      "grad_norm": 1.3959141969680786,
+      "learning_rate": 6.969931150765772e-05,
+      "loss": 5.7255,
+      "step": 16540
+    },
+    {
+      "epoch": 0.023254302660720778,
+      "grad_norm": 1.2790894508361816,
+      "learning_rate": 6.974146410004214e-05,
+      "loss": 5.6613,
+      "step": 16550
+    },
+    {
+      "epoch": 0.023268353598884354,
+      "grad_norm": 1.3713908195495605,
+      "learning_rate": 6.978361669242658e-05,
+      "loss": 5.6093,
+      "step": 16560
+    },
+    {
+      "epoch": 0.023282404537047934,
+      "grad_norm": 1.4066718816757202,
+      "learning_rate": 6.982576928481101e-05,
+      "loss": 5.5383,
+      "step": 16570
+    },
+    {
+      "epoch": 0.02329645547521151,
+      "grad_norm": 1.5075857639312744,
+      "learning_rate": 6.986792187719543e-05,
+      "loss": 5.5824,
+      "step": 16580
+    },
+    {
+      "epoch": 0.02331050641337509,
+      "grad_norm": 1.4346543550491333,
+      "learning_rate": 6.991007446957988e-05,
+      "loss": 5.8348,
+      "step": 16590
+    },
+    {
+      "epoch": 0.023324557351538665,
+      "grad_norm": 1.4327467679977417,
+      "learning_rate": 6.995222706196431e-05,
+      "loss": 5.5361,
+      "step": 16600
+    },
+    {
+      "epoch": 0.02333860828970224,
+      "grad_norm": 1.403082251548767,
+      "learning_rate": 6.999437965434874e-05,
+      "loss": 5.6565,
+      "step": 16610
+    },
+    {
+      "epoch": 0.02335265922786582,
+      "grad_norm": 1.7867377996444702,
+      "learning_rate": 7.003653224673317e-05,
+      "loss": 5.7467,
+      "step": 16620
+    },
+    {
+      "epoch": 0.023366710166029397,
+      "grad_norm": 1.3568111658096313,
+      "learning_rate": 7.00786848391176e-05,
+      "loss": 5.624,
+      "step": 16630
+    },
+    {
+      "epoch": 0.023380761104192977,
+      "grad_norm": 1.4141385555267334,
+      "learning_rate": 7.012083743150203e-05,
+      "loss": 5.7333,
+      "step": 16640
+    },
+    {
+      "epoch": 0.023394812042356553,
+      "grad_norm": 1.526979923248291,
+      "learning_rate": 7.016299002388646e-05,
+      "loss": 5.6305,
+      "step": 16650
+    },
+    {
+      "epoch": 0.02340886298052013,
+      "grad_norm": 1.2974737882614136,
+      "learning_rate": 7.02051426162709e-05,
+      "loss": 5.7538,
+      "step": 16660
+    },
+    {
+      "epoch": 0.02342291391868371,
+      "grad_norm": 1.3604271411895752,
+      "learning_rate": 7.024729520865533e-05,
+      "loss": 5.7628,
+      "step": 16670
+    },
+    {
+      "epoch": 0.023436964856847285,
+      "grad_norm": 1.2945364713668823,
+      "learning_rate": 7.028944780103976e-05,
+      "loss": 5.6513,
+      "step": 16680
+    },
+    {
+      "epoch": 0.023451015795010865,
+      "grad_norm": 1.8330632448196411,
+      "learning_rate": 7.033160039342419e-05,
+      "loss": 5.7871,
+      "step": 16690
+    },
+    {
+      "epoch": 0.02346506673317444,
+      "grad_norm": 1.4950273036956787,
+      "learning_rate": 7.037375298580862e-05,
+      "loss": 5.7615,
+      "step": 16700
+    },
+    {
+      "epoch": 0.023479117671338017,
+      "grad_norm": 1.4327980279922485,
+      "learning_rate": 7.041590557819305e-05,
+      "loss": 5.6334,
+      "step": 16710
+    },
+    {
+      "epoch": 0.023493168609501597,
+      "grad_norm": 1.2713170051574707,
+      "learning_rate": 7.045805817057748e-05,
+      "loss": 5.8051,
+      "step": 16720
+    },
+    {
+      "epoch": 0.023507219547665173,
+      "grad_norm": 1.3698358535766602,
+      "learning_rate": 7.050021076296191e-05,
+      "loss": 5.6809,
+      "step": 16730
+    },
+    {
+      "epoch": 0.02352127048582875,
+      "grad_norm": 1.40662682056427,
+      "learning_rate": 7.054236335534636e-05,
+      "loss": 5.6715,
+      "step": 16740
+    },
+    {
+      "epoch": 0.02353532142399233,
+      "grad_norm": 1.3832154273986816,
+      "learning_rate": 7.058451594773077e-05,
+      "loss": 5.7687,
+      "step": 16750
+    },
+    {
+      "epoch": 0.023549372362155905,
+      "grad_norm": 1.2708238363265991,
+      "learning_rate": 7.06266685401152e-05,
+      "loss": 5.7387,
+      "step": 16760
+    },
+    {
+      "epoch": 0.023563423300319485,
+      "grad_norm": 1.5633302927017212,
+      "learning_rate": 7.066882113249965e-05,
+      "loss": 5.6757,
+      "step": 16770
+    },
+    {
+      "epoch": 0.02357747423848306,
+      "grad_norm": 1.4342371225357056,
+      "learning_rate": 7.071097372488407e-05,
+      "loss": 5.7106,
+      "step": 16780
+    },
+    {
+      "epoch": 0.023591525176646637,
+      "grad_norm": 1.3297955989837646,
+      "learning_rate": 7.07531263172685e-05,
+      "loss": 5.664,
+      "step": 16790
+    },
+    {
+      "epoch": 0.023605576114810217,
+      "grad_norm": 1.9332789182662964,
+      "learning_rate": 7.079527890965294e-05,
+      "loss": 5.6539,
+      "step": 16800
+    },
+    {
+      "epoch": 0.023619627052973793,
+      "grad_norm": 1.3228834867477417,
+      "learning_rate": 7.083743150203736e-05,
+      "loss": 5.6064,
+      "step": 16810
+    },
+    {
+      "epoch": 0.023633677991137372,
+      "grad_norm": 1.4686253070831299,
+      "learning_rate": 7.08795840944218e-05,
+      "loss": 5.6387,
+      "step": 16820
+    },
+    {
+      "epoch": 0.02364772892930095,
+      "grad_norm": 1.4611742496490479,
+      "learning_rate": 7.092173668680624e-05,
+      "loss": 5.7412,
+      "step": 16830
+    },
+    {
+      "epoch": 0.023661779867464525,
+      "grad_norm": 1.2942790985107422,
+      "learning_rate": 7.096388927919066e-05,
+      "loss": 5.6693,
+      "step": 16840
+    },
+    {
+      "epoch": 0.023675830805628104,
+      "grad_norm": 1.254348635673523,
+      "learning_rate": 7.10060418715751e-05,
+      "loss": 5.7693,
+      "step": 16850
+    },
+    {
+      "epoch": 0.02368988174379168,
+      "grad_norm": 1.426842212677002,
+      "learning_rate": 7.104819446395953e-05,
+      "loss": 5.6919,
+      "step": 16860
+    },
+    {
+      "epoch": 0.023703932681955257,
+      "grad_norm": 1.349927544593811,
+      "learning_rate": 7.109034705634395e-05,
+      "loss": 5.62,
+      "step": 16870
+    },
+    {
+      "epoch": 0.023717983620118836,
+      "grad_norm": 1.4106230735778809,
+      "learning_rate": 7.11324996487284e-05,
+      "loss": 5.8021,
+      "step": 16880
+    },
+    {
+      "epoch": 0.023732034558282412,
+      "grad_norm": 1.3730113506317139,
+      "learning_rate": 7.117465224111282e-05,
+      "loss": 5.6808,
+      "step": 16890
+    },
+    {
+      "epoch": 0.023746085496445992,
+      "grad_norm": 1.320567011833191,
+      "learning_rate": 7.121680483349726e-05,
+      "loss": 5.6301,
+      "step": 16900
+    },
+    {
+      "epoch": 0.023760136434609568,
+      "grad_norm": 1.3973300457000732,
+      "learning_rate": 7.125895742588169e-05,
+      "loss": 5.6409,
+      "step": 16910
+    },
+    {
+      "epoch": 0.023774187372773144,
+      "grad_norm": 1.354918122291565,
+      "learning_rate": 7.130111001826612e-05,
+      "loss": 5.6952,
+      "step": 16920
+    },
+    {
+      "epoch": 0.023788238310936724,
+      "grad_norm": 1.419696569442749,
+      "learning_rate": 7.134326261065055e-05,
+      "loss": 5.6368,
+      "step": 16930
+    },
+    {
+      "epoch": 0.0238022892491003,
+      "grad_norm": 1.403704285621643,
+      "learning_rate": 7.138541520303498e-05,
+      "loss": 5.6484,
+      "step": 16940
+    },
+    {
+      "epoch": 0.02381634018726388,
+      "grad_norm": 1.366337776184082,
+      "learning_rate": 7.142756779541941e-05,
+      "loss": 5.7407,
+      "step": 16950
+    },
+    {
+      "epoch": 0.023830391125427456,
+      "grad_norm": 1.469281554222107,
+      "learning_rate": 7.146972038780384e-05,
+      "loss": 5.719,
+      "step": 16960
+    },
+    {
+      "epoch": 0.023844442063591032,
+      "grad_norm": 1.321475625038147,
+      "learning_rate": 7.151187298018827e-05,
+      "loss": 5.517,
+      "step": 16970
+    },
+    {
+      "epoch": 0.023858493001754612,
+      "grad_norm": 1.3294295072555542,
+      "learning_rate": 7.15540255725727e-05,
+      "loss": 5.6871,
+      "step": 16980
+    },
+    {
+      "epoch": 0.023872543939918188,
+      "grad_norm": 1.3294674158096313,
+      "learning_rate": 7.159617816495714e-05,
+      "loss": 5.6561,
+      "step": 16990
+    },
+    {
+      "epoch": 0.023886594878081768,
+      "grad_norm": 1.4403409957885742,
+      "learning_rate": 7.163833075734157e-05,
+      "loss": 5.5484,
+      "step": 17000
+    },
+    {
+      "epoch": 0.023900645816245344,
+      "grad_norm": 1.3805781602859497,
+      "learning_rate": 7.1680483349726e-05,
+      "loss": 5.7715,
+      "step": 17010
+    },
+    {
+      "epoch": 0.02391469675440892,
+      "grad_norm": 1.2874103784561157,
+      "learning_rate": 7.172263594211043e-05,
+      "loss": 5.6586,
+      "step": 17020
+    },
+    {
+      "epoch": 0.0239287476925725,
+      "grad_norm": 1.2773886919021606,
+      "learning_rate": 7.176478853449487e-05,
+      "loss": 5.7795,
+      "step": 17030
+    },
+    {
+      "epoch": 0.023942798630736076,
+      "grad_norm": 1.4012190103530884,
+      "learning_rate": 7.180694112687929e-05,
+      "loss": 5.686,
+      "step": 17040
+    },
+    {
+      "epoch": 0.023956849568899652,
+      "grad_norm": 1.3619203567504883,
+      "learning_rate": 7.184909371926372e-05,
+      "loss": 5.7779,
+      "step": 17050
+    },
+    {
+      "epoch": 0.02397090050706323,
+      "grad_norm": 1.6374523639678955,
+      "learning_rate": 7.189124631164817e-05,
+      "loss": 5.7293,
+      "step": 17060
+    },
+    {
+      "epoch": 0.023984951445226808,
+      "grad_norm": 1.365061640739441,
+      "learning_rate": 7.193339890403259e-05,
+      "loss": 5.6493,
+      "step": 17070
+    },
+    {
+      "epoch": 0.023999002383390387,
+      "grad_norm": 1.458706021308899,
+      "learning_rate": 7.197555149641702e-05,
+      "loss": 5.6065,
+      "step": 17080
+    },
+    {
+      "epoch": 0.024013053321553963,
+      "grad_norm": 1.3380498886108398,
+      "learning_rate": 7.201770408880146e-05,
+      "loss": 5.575,
+      "step": 17090
+    },
+    {
+      "epoch": 0.02402710425971754,
+      "grad_norm": 1.2822099924087524,
+      "learning_rate": 7.205985668118588e-05,
+      "loss": 5.6609,
+      "step": 17100
+    },
+    {
+      "epoch": 0.02404115519788112,
+      "grad_norm": 1.3940728902816772,
+      "learning_rate": 7.210200927357032e-05,
+      "loss": 5.6604,
+      "step": 17110
+    },
+    {
+      "epoch": 0.024055206136044695,
+      "grad_norm": 1.3046749830245972,
+      "learning_rate": 7.214416186595475e-05,
+      "loss": 5.6748,
+      "step": 17120
+    },
+    {
+      "epoch": 0.024069257074208275,
+      "grad_norm": 1.3955029249191284,
+      "learning_rate": 7.218631445833917e-05,
+      "loss": 5.6872,
+      "step": 17130
+    },
+    {
+      "epoch": 0.02408330801237185,
+      "grad_norm": 1.2928085327148438,
+      "learning_rate": 7.222846705072362e-05,
+      "loss": 5.6749,
+      "step": 17140
+    },
+    {
+      "epoch": 0.024097358950535427,
+      "grad_norm": 1.4071931838989258,
+      "learning_rate": 7.227061964310805e-05,
+      "loss": 5.6209,
+      "step": 17150
+    },
+    {
+      "epoch": 0.024111409888699007,
+      "grad_norm": 1.4092259407043457,
+      "learning_rate": 7.231277223549248e-05,
+      "loss": 5.616,
+      "step": 17160
+    },
+    {
+      "epoch": 0.024125460826862583,
+      "grad_norm": 1.3400150537490845,
+      "learning_rate": 7.235492482787691e-05,
+      "loss": 5.7729,
+      "step": 17170
+    },
+    {
+      "epoch": 0.02413951176502616,
+      "grad_norm": 1.7592744827270508,
+      "learning_rate": 7.239707742026134e-05,
+      "loss": 5.6002,
+      "step": 17180
+    },
+    {
+      "epoch": 0.02415356270318974,
+      "grad_norm": 1.4428009986877441,
+      "learning_rate": 7.243923001264577e-05,
+      "loss": 5.5269,
+      "step": 17190
+    },
+    {
+      "epoch": 0.024167613641353315,
+      "grad_norm": 1.266127347946167,
+      "learning_rate": 7.24813826050302e-05,
+      "loss": 5.6309,
+      "step": 17200
+    },
+    {
+      "epoch": 0.024181664579516895,
+      "grad_norm": 1.3543121814727783,
+      "learning_rate": 7.252353519741464e-05,
+      "loss": 5.5518,
+      "step": 17210
+    },
+    {
+      "epoch": 0.02419571551768047,
+      "grad_norm": 1.3485888242721558,
+      "learning_rate": 7.256568778979907e-05,
+      "loss": 5.6967,
+      "step": 17220
+    },
+    {
+      "epoch": 0.024209766455844047,
+      "grad_norm": 1.4433296918869019,
+      "learning_rate": 7.26078403821835e-05,
+      "loss": 5.7349,
+      "step": 17230
+    },
+    {
+      "epoch": 0.024223817394007627,
+      "grad_norm": 1.47990882396698,
+      "learning_rate": 7.264999297456793e-05,
+      "loss": 5.7185,
+      "step": 17240
+    },
+    {
+      "epoch": 0.024237868332171203,
+      "grad_norm": 1.3784239292144775,
+      "learning_rate": 7.269214556695236e-05,
+      "loss": 5.6906,
+      "step": 17250
+    },
+    {
+      "epoch": 0.024251919270334783,
+      "grad_norm": 1.329148530960083,
+      "learning_rate": 7.273429815933679e-05,
+      "loss": 5.6781,
+      "step": 17260
+    },
+    {
+      "epoch": 0.02426597020849836,
+      "grad_norm": 1.3315861225128174,
+      "learning_rate": 7.277645075172122e-05,
+      "loss": 5.6335,
+      "step": 17270
+    },
+    {
+      "epoch": 0.024280021146661935,
+      "grad_norm": 1.4295711517333984,
+      "learning_rate": 7.281860334410565e-05,
+      "loss": 5.6599,
+      "step": 17280
+    },
+    {
+      "epoch": 0.024294072084825515,
+      "grad_norm": 1.3033784627914429,
+      "learning_rate": 7.286075593649008e-05,
+      "loss": 5.6773,
+      "step": 17290
+    },
+    {
+      "epoch": 0.02430812302298909,
+      "grad_norm": 1.5363476276397705,
+      "learning_rate": 7.290290852887452e-05,
+      "loss": 5.7043,
+      "step": 17300
+    },
+    {
+      "epoch": 0.02432217396115267,
+      "grad_norm": 1.4859110116958618,
+      "learning_rate": 7.294506112125895e-05,
+      "loss": 5.5239,
+      "step": 17310
+    },
+    {
+      "epoch": 0.024336224899316247,
+      "grad_norm": 1.3148119449615479,
+      "learning_rate": 7.298721371364339e-05,
+      "loss": 5.7742,
+      "step": 17320
+    },
+    {
+      "epoch": 0.024350275837479823,
+      "grad_norm": 1.4127575159072876,
+      "learning_rate": 7.302936630602781e-05,
+      "loss": 5.6235,
+      "step": 17330
+    },
+    {
+      "epoch": 0.024364326775643402,
+      "grad_norm": 1.7024136781692505,
+      "learning_rate": 7.307151889841224e-05,
+      "loss": 5.6536,
+      "step": 17340
+    },
+    {
+      "epoch": 0.02437837771380698,
+      "grad_norm": 1.4915293455123901,
+      "learning_rate": 7.311367149079669e-05,
+      "loss": 5.5354,
+      "step": 17350
+    },
+    {
+      "epoch": 0.024392428651970555,
+      "grad_norm": 1.431097149848938,
+      "learning_rate": 7.31558240831811e-05,
+      "loss": 5.626,
+      "step": 17360
+    },
+    {
+      "epoch": 0.024406479590134134,
+      "grad_norm": 1.411797285079956,
+      "learning_rate": 7.319797667556555e-05,
+      "loss": 5.6458,
+      "step": 17370
+    },
+    {
+      "epoch": 0.02442053052829771,
+      "grad_norm": 1.374850869178772,
+      "learning_rate": 7.324012926794998e-05,
+      "loss": 5.6726,
+      "step": 17380
+    },
+    {
+      "epoch": 0.02443458146646129,
+      "grad_norm": 1.3310719728469849,
+      "learning_rate": 7.32822818603344e-05,
+      "loss": 5.6913,
+      "step": 17390
+    },
+    {
+      "epoch": 0.024448632404624866,
+      "grad_norm": 1.5239213705062866,
+      "learning_rate": 7.332443445271884e-05,
+      "loss": 5.7034,
+      "step": 17400
+    },
+    {
+      "epoch": 0.024462683342788442,
+      "grad_norm": 1.3302022218704224,
+      "learning_rate": 7.336658704510327e-05,
+      "loss": 5.6239,
+      "step": 17410
+    },
+    {
+      "epoch": 0.024476734280952022,
+      "grad_norm": 1.499558687210083,
+      "learning_rate": 7.340873963748769e-05,
+      "loss": 5.7574,
+      "step": 17420
+    },
+    {
+      "epoch": 0.024490785219115598,
+      "grad_norm": 1.323960304260254,
+      "learning_rate": 7.345089222987213e-05,
+      "loss": 5.6001,
+      "step": 17430
+    },
+    {
+      "epoch": 0.024504836157279178,
+      "grad_norm": 1.2899998426437378,
+      "learning_rate": 7.349304482225657e-05,
+      "loss": 5.5648,
+      "step": 17440
+    },
+    {
+      "epoch": 0.024518887095442754,
+      "grad_norm": 1.3837937116622925,
+      "learning_rate": 7.3535197414641e-05,
+      "loss": 5.6718,
+      "step": 17450
+    },
+    {
+      "epoch": 0.02453293803360633,
+      "grad_norm": 1.2842652797698975,
+      "learning_rate": 7.357735000702543e-05,
+      "loss": 5.7037,
+      "step": 17460
+    },
+    {
+      "epoch": 0.02454698897176991,
+      "grad_norm": 1.2789689302444458,
+      "learning_rate": 7.361950259940986e-05,
+      "loss": 5.6116,
+      "step": 17470
+    },
+    {
+      "epoch": 0.024561039909933486,
+      "grad_norm": 1.3312838077545166,
+      "learning_rate": 7.366165519179429e-05,
+      "loss": 5.7832,
+      "step": 17480
+    },
+    {
+      "epoch": 0.024575090848097062,
+      "grad_norm": 1.3789297342300415,
+      "learning_rate": 7.370380778417872e-05,
+      "loss": 5.6628,
+      "step": 17490
+    },
+    {
+      "epoch": 0.024589141786260642,
+      "grad_norm": 1.552386999130249,
+      "learning_rate": 7.374596037656315e-05,
+      "loss": 5.7062,
+      "step": 17500
+    },
+    {
+      "epoch": 0.024603192724424218,
+      "grad_norm": 1.3529428243637085,
+      "learning_rate": 7.378811296894758e-05,
+      "loss": 5.5797,
+      "step": 17510
+    },
+    {
+      "epoch": 0.024617243662587798,
+      "grad_norm": 1.3707754611968994,
+      "learning_rate": 7.383026556133202e-05,
+      "loss": 5.6016,
+      "step": 17520
+    },
+    {
+      "epoch": 0.024631294600751374,
+      "grad_norm": 1.6769342422485352,
+      "learning_rate": 7.387241815371645e-05,
+      "loss": 5.6126,
+      "step": 17530
+    },
+    {
+      "epoch": 0.02464534553891495,
+      "grad_norm": 1.3697290420532227,
+      "learning_rate": 7.391457074610088e-05,
+      "loss": 5.5616,
+      "step": 17540
+    },
+    {
+      "epoch": 0.02465939647707853,
+      "grad_norm": 1.4338105916976929,
+      "learning_rate": 7.395672333848531e-05,
+      "loss": 5.6251,
+      "step": 17550
+    },
+    {
+      "epoch": 0.024673447415242106,
+      "grad_norm": 1.338873028755188,
+      "learning_rate": 7.399887593086974e-05,
+      "loss": 5.5995,
+      "step": 17560
+    },
+    {
+      "epoch": 0.024687498353405685,
+      "grad_norm": 1.6078602075576782,
+      "learning_rate": 7.404102852325417e-05,
+      "loss": 5.7184,
+      "step": 17570
+    },
+    {
+      "epoch": 0.02470154929156926,
+      "grad_norm": 1.374680995941162,
+      "learning_rate": 7.40831811156386e-05,
+      "loss": 5.54,
+      "step": 17580
+    },
+    {
+      "epoch": 0.024715600229732838,
+      "grad_norm": 1.3792024850845337,
+      "learning_rate": 7.41211184487846e-05,
+      "loss": 5.6917,
+      "step": 17590
+    },
+    {
+      "epoch": 0.024729651167896417,
+      "grad_norm": 1.3895095586776733,
+      "learning_rate": 7.416327104116903e-05,
+      "loss": 5.5612,
+      "step": 17600
+    },
+    {
+      "epoch": 0.024743702106059993,
+      "grad_norm": 1.251436710357666,
+      "learning_rate": 7.420542363355346e-05,
+      "loss": 5.6356,
+      "step": 17610
+    },
+    {
+      "epoch": 0.024757753044223573,
+      "grad_norm": 1.6639325618743896,
+      "learning_rate": 7.424757622593789e-05,
+      "loss": 5.613,
+      "step": 17620
+    },
+    {
+      "epoch": 0.02477180398238715,
+      "grad_norm": 1.2654144763946533,
+      "learning_rate": 7.428972881832232e-05,
+      "loss": 5.6399,
+      "step": 17630
+    },
+    {
+      "epoch": 0.024785854920550725,
+      "grad_norm": 1.3583436012268066,
+      "learning_rate": 7.433188141070675e-05,
+      "loss": 5.7586,
+      "step": 17640
+    },
+    {
+      "epoch": 0.024799905858714305,
+      "grad_norm": 1.3544237613677979,
+      "learning_rate": 7.437403400309119e-05,
+      "loss": 5.7153,
+      "step": 17650
+    },
+    {
+      "epoch": 0.02481395679687788,
+      "grad_norm": 1.3555983304977417,
+      "learning_rate": 7.441618659547562e-05,
+      "loss": 5.5901,
+      "step": 17660
+    },
+    {
+      "epoch": 0.024828007735041457,
+      "grad_norm": 1.5493931770324707,
+      "learning_rate": 7.445833918786005e-05,
+      "loss": 5.5702,
+      "step": 17670
+    },
+    {
+      "epoch": 0.024842058673205037,
+      "grad_norm": 1.4650983810424805,
+      "learning_rate": 7.450049178024448e-05,
+      "loss": 5.6018,
+      "step": 17680
+    },
+    {
+      "epoch": 0.024856109611368613,
+      "grad_norm": 1.2635927200317383,
+      "learning_rate": 7.454264437262891e-05,
+      "loss": 5.5942,
+      "step": 17690
+    },
+    {
+      "epoch": 0.024870160549532193,
+      "grad_norm": 1.5327472686767578,
+      "learning_rate": 7.458479696501334e-05,
+      "loss": 5.5997,
+      "step": 17700
+    },
+    {
+      "epoch": 0.02488421148769577,
+      "grad_norm": 1.350309133529663,
+      "learning_rate": 7.462694955739777e-05,
+      "loss": 5.582,
+      "step": 17710
+    },
+    {
+      "epoch": 0.024898262425859345,
+      "grad_norm": 1.321396827697754,
+      "learning_rate": 7.46691021497822e-05,
+      "loss": 5.582,
+      "step": 17720
+    },
+    {
+      "epoch": 0.024912313364022925,
+      "grad_norm": 1.3941783905029297,
+      "learning_rate": 7.471125474216663e-05,
+      "loss": 5.7505,
+      "step": 17730
+    },
+    {
+      "epoch": 0.0249263643021865,
+      "grad_norm": 1.573265790939331,
+      "learning_rate": 7.475340733455107e-05,
+      "loss": 5.4622,
+      "step": 17740
+    },
+    {
+      "epoch": 0.02494041524035008,
+      "grad_norm": 1.3758049011230469,
+      "learning_rate": 7.47955599269355e-05,
+      "loss": 5.5887,
+      "step": 17750
+    },
+    {
+      "epoch": 0.024954466178513657,
+      "grad_norm": 1.3547019958496094,
+      "learning_rate": 7.483771251931993e-05,
+      "loss": 5.603,
+      "step": 17760
+    },
+    {
+      "epoch": 0.024968517116677233,
+      "grad_norm": 1.464343786239624,
+      "learning_rate": 7.487986511170436e-05,
+      "loss": 5.5724,
+      "step": 17770
+    },
+    {
+      "epoch": 0.024982568054840813,
+      "grad_norm": 1.409480333328247,
+      "learning_rate": 7.492201770408879e-05,
+      "loss": 5.6456,
+      "step": 17780
+    },
+    {
+      "epoch": 0.02499661899300439,
+      "grad_norm": 1.4030753374099731,
+      "learning_rate": 7.496417029647324e-05,
+      "loss": 5.6821,
+      "step": 17790
+    },
+    {
+      "epoch": 0.02501066993116797,
+      "grad_norm": 1.4202769994735718,
+      "learning_rate": 7.500632288885765e-05,
+      "loss": 5.6878,
+      "step": 17800
+    },
+    {
+      "epoch": 0.025024720869331545,
+      "grad_norm": 1.3402527570724487,
+      "learning_rate": 7.504847548124208e-05,
+      "loss": 5.5967,
+      "step": 17810
+    },
+    {
+      "epoch": 0.02503877180749512,
+      "grad_norm": 1.4031054973602295,
+      "learning_rate": 7.509062807362653e-05,
+      "loss": 5.6123,
+      "step": 17820
+    },
+    {
+      "epoch": 0.0250528227456587,
+      "grad_norm": 1.5110526084899902,
+      "learning_rate": 7.513278066601096e-05,
+      "loss": 5.6823,
+      "step": 17830
+    },
+    {
+      "epoch": 0.025066873683822277,
+      "grad_norm": 1.3663294315338135,
+      "learning_rate": 7.517493325839538e-05,
+      "loss": 5.6382,
+      "step": 17840
+    },
+    {
+      "epoch": 0.025080924621985853,
+      "grad_norm": 1.2902328968048096,
+      "learning_rate": 7.521708585077982e-05,
+      "loss": 5.6346,
+      "step": 17850
+    },
+    {
+      "epoch": 0.025094975560149432,
+      "grad_norm": 1.4534049034118652,
+      "learning_rate": 7.525923844316425e-05,
+      "loss": 5.5818,
+      "step": 17860
+    },
+    {
+      "epoch": 0.02510902649831301,
+      "grad_norm": 1.3165909051895142,
+      "learning_rate": 7.530139103554867e-05,
+      "loss": 5.6324,
+      "step": 17870
+    },
+    {
+      "epoch": 0.025123077436476588,
+      "grad_norm": 1.443505048751831,
+      "learning_rate": 7.534354362793312e-05,
+      "loss": 5.6841,
+      "step": 17880
+    },
+    {
+      "epoch": 0.025137128374640164,
+      "grad_norm": 1.364423155784607,
+      "learning_rate": 7.538569622031755e-05,
+      "loss": 5.6508,
+      "step": 17890
+    },
+    {
+      "epoch": 0.02515117931280374,
+      "grad_norm": 1.3083369731903076,
+      "learning_rate": 7.542784881270196e-05,
+      "loss": 5.7182,
+      "step": 17900
+    },
+    {
+      "epoch": 0.02516523025096732,
+      "grad_norm": 1.2806565761566162,
+      "learning_rate": 7.547000140508641e-05,
+      "loss": 5.6075,
+      "step": 17910
+    },
+    {
+      "epoch": 0.025179281189130896,
+      "grad_norm": 1.3733397722244263,
+      "learning_rate": 7.551215399747084e-05,
+      "loss": 5.6118,
+      "step": 17920
+    },
+    {
+      "epoch": 0.025193332127294476,
+      "grad_norm": 1.3024393320083618,
+      "learning_rate": 7.555430658985526e-05,
+      "loss": 5.628,
+      "step": 17930
+    },
+    {
+      "epoch": 0.025207383065458052,
+      "grad_norm": 1.447492003440857,
+      "learning_rate": 7.55964591822397e-05,
+      "loss": 5.6351,
+      "step": 17940
+    },
+    {
+      "epoch": 0.025221434003621628,
+      "grad_norm": 1.2635152339935303,
+      "learning_rate": 7.563861177462413e-05,
+      "loss": 5.5923,
+      "step": 17950
+    },
+    {
+      "epoch": 0.025235484941785208,
+      "grad_norm": 1.2946640253067017,
+      "learning_rate": 7.568076436700855e-05,
+      "loss": 5.6809,
+      "step": 17960
+    },
+    {
+      "epoch": 0.025249535879948784,
+      "grad_norm": 1.283579707145691,
+      "learning_rate": 7.5722916959393e-05,
+      "loss": 5.7118,
+      "step": 17970
+    },
+    {
+      "epoch": 0.02526358681811236,
+      "grad_norm": 1.259277582168579,
+      "learning_rate": 7.576506955177743e-05,
+      "loss": 5.6226,
+      "step": 17980
+    },
+    {
+      "epoch": 0.02527763775627594,
+      "grad_norm": 1.519494891166687,
+      "learning_rate": 7.580722214416187e-05,
+      "loss": 5.6467,
+      "step": 17990
+    },
+    {
+      "epoch": 0.025291688694439516,
+      "grad_norm": 1.3044480085372925,
+      "learning_rate": 7.584937473654629e-05,
+      "loss": 5.7049,
+      "step": 18000
+    },
+    {
+      "epoch": 0.025305739632603096,
+      "grad_norm": 1.3547728061676025,
+      "learning_rate": 7.589152732893072e-05,
+      "loss": 5.5962,
+      "step": 18010
+    },
+    {
+      "epoch": 0.025319790570766672,
+      "grad_norm": 1.3373003005981445,
+      "learning_rate": 7.593367992131517e-05,
+      "loss": 5.6107,
+      "step": 18020
+    },
+    {
+      "epoch": 0.025333841508930248,
+      "grad_norm": 1.3276454210281372,
+      "learning_rate": 7.597583251369958e-05,
+      "loss": 5.5887,
+      "step": 18030
+    },
+    {
+      "epoch": 0.025347892447093828,
+      "grad_norm": 1.3012672662734985,
+      "learning_rate": 7.601798510608401e-05,
+      "loss": 5.6691,
+      "step": 18040
+    },
+    {
+      "epoch": 0.025361943385257404,
+      "grad_norm": 1.3194389343261719,
+      "learning_rate": 7.606013769846846e-05,
+      "loss": 5.6852,
+      "step": 18050
+    },
+    {
+      "epoch": 0.025375994323420983,
+      "grad_norm": 1.2897201776504517,
+      "learning_rate": 7.610229029085288e-05,
+      "loss": 5.6437,
+      "step": 18060
+    },
+    {
+      "epoch": 0.02539004526158456,
+      "grad_norm": 2.91089129447937,
+      "learning_rate": 7.614444288323731e-05,
+      "loss": 5.6226,
+      "step": 18070
+    },
+    {
+      "epoch": 0.025404096199748136,
+      "grad_norm": 1.2783751487731934,
+      "learning_rate": 7.618659547562175e-05,
+      "loss": 5.5288,
+      "step": 18080
+    },
+    {
+      "epoch": 0.025418147137911715,
+      "grad_norm": 1.5145020484924316,
+      "learning_rate": 7.622874806800617e-05,
+      "loss": 5.611,
+      "step": 18090
+    },
+    {
+      "epoch": 0.02543219807607529,
+      "grad_norm": 1.3364498615264893,
+      "learning_rate": 7.62709006603906e-05,
+      "loss": 5.5694,
+      "step": 18100
+    },
+    {
+      "epoch": 0.02544624901423887,
+      "grad_norm": 1.3756706714630127,
+      "learning_rate": 7.631305325277505e-05,
+      "loss": 5.5867,
+      "step": 18110
+    },
+    {
+      "epoch": 0.025460299952402447,
+      "grad_norm": 1.4030925035476685,
+      "learning_rate": 7.635520584515948e-05,
+      "loss": 5.5534,
+      "step": 18120
+    },
+    {
+      "epoch": 0.025474350890566023,
+      "grad_norm": 1.3178571462631226,
+      "learning_rate": 7.63973584375439e-05,
+      "loss": 5.6151,
+      "step": 18130
+    },
+    {
+      "epoch": 0.025488401828729603,
+      "grad_norm": 1.3697510957717896,
+      "learning_rate": 7.643951102992834e-05,
+      "loss": 5.648,
+      "step": 18140
+    },
+    {
+      "epoch": 0.02550245276689318,
+      "grad_norm": 1.3539316654205322,
+      "learning_rate": 7.648166362231277e-05,
+      "loss": 5.5328,
+      "step": 18150
+    },
+    {
+      "epoch": 0.025516503705056755,
+      "grad_norm": 1.3074488639831543,
+      "learning_rate": 7.652381621469719e-05,
+      "loss": 5.6203,
+      "step": 18160
+    },
+    {
+      "epoch": 0.025530554643220335,
+      "grad_norm": 1.3524091243743896,
+      "learning_rate": 7.656596880708163e-05,
+      "loss": 5.6306,
+      "step": 18170
+    },
+    {
+      "epoch": 0.02554460558138391,
+      "grad_norm": 1.3374627828598022,
+      "learning_rate": 7.660812139946606e-05,
+      "loss": 5.6288,
+      "step": 18180
+    },
+    {
+      "epoch": 0.02555865651954749,
+      "grad_norm": 1.351073145866394,
+      "learning_rate": 7.665027399185048e-05,
+      "loss": 5.606,
+      "step": 18190
+    },
+    {
+      "epoch": 0.025572707457711067,
+      "grad_norm": 1.4369218349456787,
+      "learning_rate": 7.669242658423493e-05,
+      "loss": 5.5697,
+      "step": 18200
+    },
+    {
+      "epoch": 0.025586758395874643,
+      "grad_norm": 1.3648687601089478,
+      "learning_rate": 7.673457917661936e-05,
+      "loss": 5.5595,
+      "step": 18210
+    },
+    {
+      "epoch": 0.025600809334038223,
+      "grad_norm": 1.3496894836425781,
+      "learning_rate": 7.677673176900378e-05,
+      "loss": 5.528,
+      "step": 18220
+    },
+    {
+      "epoch": 0.0256148602722018,
+      "grad_norm": 1.3187439441680908,
+      "learning_rate": 7.681888436138822e-05,
+      "loss": 5.6865,
+      "step": 18230
+    },
+    {
+      "epoch": 0.02562891121036538,
+      "grad_norm": 1.4901741743087769,
+      "learning_rate": 7.686103695377265e-05,
+      "loss": 5.5185,
+      "step": 18240
+    },
+    {
+      "epoch": 0.025642962148528955,
+      "grad_norm": 1.4721450805664062,
+      "learning_rate": 7.69031895461571e-05,
+      "loss": 5.5971,
+      "step": 18250
+    },
+    {
+      "epoch": 0.02565701308669253,
+      "grad_norm": 1.3747284412384033,
+      "learning_rate": 7.694534213854151e-05,
+      "loss": 5.6086,
+      "step": 18260
+    },
+    {
+      "epoch": 0.02567106402485611,
+      "grad_norm": 1.4034130573272705,
+      "learning_rate": 7.698749473092594e-05,
+      "loss": 5.585,
+      "step": 18270
+    },
+    {
+      "epoch": 0.025685114963019687,
+      "grad_norm": 1.4254305362701416,
+      "learning_rate": 7.702964732331039e-05,
+      "loss": 5.5106,
+      "step": 18280
+    },
+    {
+      "epoch": 0.025699165901183263,
+      "grad_norm": 1.4483678340911865,
+      "learning_rate": 7.707179991569481e-05,
+      "loss": 5.5116,
+      "step": 18290
+    },
+    {
+      "epoch": 0.025713216839346843,
+      "grad_norm": 1.3134675025939941,
+      "learning_rate": 7.711395250807924e-05,
+      "loss": 5.5165,
+      "step": 18300
+    },
+    {
+      "epoch": 0.02572726777751042,
+      "grad_norm": 1.3608453273773193,
+      "learning_rate": 7.715610510046368e-05,
+      "loss": 5.4691,
+      "step": 18310
+    },
+    {
+      "epoch": 0.025741318715674,
+      "grad_norm": 1.3626971244812012,
+      "learning_rate": 7.71982576928481e-05,
+      "loss": 5.5634,
+      "step": 18320
+    },
+    {
+      "epoch": 0.025755369653837575,
+      "grad_norm": 1.3774510622024536,
+      "learning_rate": 7.724041028523253e-05,
+      "loss": 5.5621,
+      "step": 18330
+    },
+    {
+      "epoch": 0.02576942059200115,
+      "grad_norm": 1.361126184463501,
+      "learning_rate": 7.728256287761698e-05,
+      "loss": 5.5653,
+      "step": 18340
+    },
+    {
+      "epoch": 0.02578347153016473,
+      "grad_norm": 1.440781593322754,
+      "learning_rate": 7.73247154700014e-05,
+      "loss": 5.5529,
+      "step": 18350
+    },
+    {
+      "epoch": 0.025797522468328307,
+      "grad_norm": 1.3827389478683472,
+      "learning_rate": 7.736686806238583e-05,
+      "loss": 5.4872,
+      "step": 18360
+    },
+    {
+      "epoch": 0.025811573406491886,
+      "grad_norm": 1.3901987075805664,
+      "learning_rate": 7.740902065477027e-05,
+      "loss": 5.6825,
+      "step": 18370
+    },
+    {
+      "epoch": 0.025825624344655462,
+      "grad_norm": 1.549964189529419,
+      "learning_rate": 7.745117324715469e-05,
+      "loss": 5.5766,
+      "step": 18380
+    },
+    {
+      "epoch": 0.02583967528281904,
+      "grad_norm": 1.3051047325134277,
+      "learning_rate": 7.749332583953912e-05,
+      "loss": 5.5934,
+      "step": 18390
+    },
+    {
+      "epoch": 0.025853726220982618,
+      "grad_norm": 1.3041412830352783,
+      "learning_rate": 7.753547843192356e-05,
+      "loss": 5.5017,
+      "step": 18400
+    },
+    {
+      "epoch": 0.025867777159146194,
+      "grad_norm": 1.3146135807037354,
+      "learning_rate": 7.7577631024308e-05,
+      "loss": 5.5809,
+      "step": 18410
+    },
+    {
+      "epoch": 0.025881828097309774,
+      "grad_norm": 1.3756433725357056,
+      "learning_rate": 7.761978361669241e-05,
+      "loss": 5.553,
+      "step": 18420
+    },
+    {
+      "epoch": 0.02589587903547335,
+      "grad_norm": 1.5293617248535156,
+      "learning_rate": 7.766193620907686e-05,
+      "loss": 5.6161,
+      "step": 18430
+    },
+    {
+      "epoch": 0.025909929973636926,
+      "grad_norm": 1.3322265148162842,
+      "learning_rate": 7.770408880146129e-05,
+      "loss": 5.6277,
+      "step": 18440
+    },
+    {
+      "epoch": 0.025923980911800506,
+      "grad_norm": 1.5310215950012207,
+      "learning_rate": 7.77462413938457e-05,
+      "loss": 5.7149,
+      "step": 18450
+    },
+    {
+      "epoch": 0.025938031849964082,
+      "grad_norm": 1.3711066246032715,
+      "learning_rate": 7.778839398623015e-05,
+      "loss": 5.5819,
+      "step": 18460
+    },
+    {
+      "epoch": 0.025952082788127658,
+      "grad_norm": 1.3732609748840332,
+      "learning_rate": 7.783054657861458e-05,
+      "loss": 5.557,
+      "step": 18470
+    },
+    {
+      "epoch": 0.025966133726291238,
+      "grad_norm": 1.3032290935516357,
+      "learning_rate": 7.7872699170999e-05,
+      "loss": 5.6042,
+      "step": 18480
+    },
+    {
+      "epoch": 0.025980184664454814,
+      "grad_norm": 1.4029079675674438,
+      "learning_rate": 7.791485176338344e-05,
+      "loss": 5.6057,
+      "step": 18490
+    },
+    {
+      "epoch": 0.025994235602618394,
+      "grad_norm": 1.2981117963790894,
+      "learning_rate": 7.795700435576788e-05,
+      "loss": 5.572,
+      "step": 18500
+    },
+    {
+      "epoch": 0.02600828654078197,
+      "grad_norm": 1.346895456314087,
+      "learning_rate": 7.799915694815229e-05,
+      "loss": 5.5795,
+      "step": 18510
+    },
+    {
+      "epoch": 0.026022337478945546,
+      "grad_norm": 1.3572524785995483,
+      "learning_rate": 7.804130954053674e-05,
+      "loss": 5.5995,
+      "step": 18520
+    },
+    {
+      "epoch": 0.026036388417109126,
+      "grad_norm": 1.4479949474334717,
+      "learning_rate": 7.808346213292117e-05,
+      "loss": 5.6112,
+      "step": 18530
+    },
+    {
+      "epoch": 0.026050439355272702,
+      "grad_norm": 1.5404094457626343,
+      "learning_rate": 7.812561472530561e-05,
+      "loss": 5.6727,
+      "step": 18540
+    },
+    {
+      "epoch": 0.02606449029343628,
+      "grad_norm": 1.4202011823654175,
+      "learning_rate": 7.816776731769003e-05,
+      "loss": 5.61,
+      "step": 18550
+    },
+    {
+      "epoch": 0.026078541231599858,
+      "grad_norm": 1.4051263332366943,
+      "learning_rate": 7.820991991007446e-05,
+      "loss": 5.5492,
+      "step": 18560
+    },
+    {
+      "epoch": 0.026092592169763434,
+      "grad_norm": 1.3520148992538452,
+      "learning_rate": 7.82520725024589e-05,
+      "loss": 5.6385,
+      "step": 18570
+    },
+    {
+      "epoch": 0.026106643107927013,
+      "grad_norm": 1.389690637588501,
+      "learning_rate": 7.829422509484332e-05,
+      "loss": 5.482,
+      "step": 18580
+    },
+    {
+      "epoch": 0.02612069404609059,
+      "grad_norm": 1.3874398469924927,
+      "learning_rate": 7.833637768722776e-05,
+      "loss": 5.5605,
+      "step": 18590
+    },
+    {
+      "epoch": 0.026134744984254166,
+      "grad_norm": 1.4007757902145386,
+      "learning_rate": 7.83785302796122e-05,
+      "loss": 5.5494,
+      "step": 18600
+    },
+    {
+      "epoch": 0.026148795922417745,
+      "grad_norm": 1.5084589719772339,
+      "learning_rate": 7.842068287199662e-05,
+      "loss": 5.5122,
+      "step": 18610
+    },
+    {
+      "epoch": 0.02616284686058132,
+      "grad_norm": 1.4278961420059204,
+      "learning_rate": 7.846283546438105e-05,
+      "loss": 5.5855,
+      "step": 18620
+    },
+    {
+      "epoch": 0.0261768977987449,
+      "grad_norm": 1.4229931831359863,
+      "learning_rate": 7.85049880567655e-05,
+      "loss": 5.4738,
+      "step": 18630
+    },
+    {
+      "epoch": 0.026190948736908477,
+      "grad_norm": 1.4617115259170532,
+      "learning_rate": 7.854714064914991e-05,
+      "loss": 5.6064,
+      "step": 18640
+    },
+    {
+      "epoch": 0.026204999675072053,
+      "grad_norm": 1.335037350654602,
+      "learning_rate": 7.858929324153434e-05,
+      "loss": 5.5919,
+      "step": 18650
+    },
+    {
+      "epoch": 0.026219050613235633,
+      "grad_norm": 1.3201597929000854,
+      "learning_rate": 7.863144583391879e-05,
+      "loss": 5.6363,
+      "step": 18660
+    },
+    {
+      "epoch": 0.02623310155139921,
+      "grad_norm": 1.5056530237197876,
+      "learning_rate": 7.86735984263032e-05,
+      "loss": 5.5251,
+      "step": 18670
+    },
+    {
+      "epoch": 0.02624715248956279,
+      "grad_norm": 1.3666268587112427,
+      "learning_rate": 7.871575101868764e-05,
+      "loss": 5.5708,
+      "step": 18680
+    },
+    {
+      "epoch": 0.026261203427726365,
+      "grad_norm": 1.4931973218917847,
+      "learning_rate": 7.875790361107208e-05,
+      "loss": 5.5645,
+      "step": 18690
+    },
+    {
+      "epoch": 0.02627525436588994,
+      "grad_norm": 1.3463979959487915,
+      "learning_rate": 7.880005620345651e-05,
+      "loss": 5.4477,
+      "step": 18700
+    },
+    {
+      "epoch": 0.02628930530405352,
+      "grad_norm": 1.3805086612701416,
+      "learning_rate": 7.884220879584093e-05,
+      "loss": 5.6309,
+      "step": 18710
+    },
+    {
+      "epoch": 0.026303356242217097,
+      "grad_norm": 1.3408585786819458,
+      "learning_rate": 7.888436138822537e-05,
+      "loss": 5.5943,
+      "step": 18720
+    },
+    {
+      "epoch": 0.026317407180380677,
+      "grad_norm": 1.3854987621307373,
+      "learning_rate": 7.89265139806098e-05,
+      "loss": 5.6367,
+      "step": 18730
+    },
+    {
+      "epoch": 0.026331458118544253,
+      "grad_norm": 1.4840235710144043,
+      "learning_rate": 7.896866657299422e-05,
+      "loss": 5.6203,
+      "step": 18740
+    },
+    {
+      "epoch": 0.02634550905670783,
+      "grad_norm": 1.4539438486099243,
+      "learning_rate": 7.901081916537867e-05,
+      "loss": 5.6533,
+      "step": 18750
+    },
+    {
+      "epoch": 0.02635955999487141,
+      "grad_norm": 1.6064668893814087,
+      "learning_rate": 7.90529717577631e-05,
+      "loss": 5.57,
+      "step": 18760
+    },
+    {
+      "epoch": 0.026373610933034985,
+      "grad_norm": 1.3408623933792114,
+      "learning_rate": 7.909512435014752e-05,
+      "loss": 5.4955,
+      "step": 18770
+    },
+    {
+      "epoch": 0.02638766187119856,
+      "grad_norm": 1.9005026817321777,
+      "learning_rate": 7.913727694253196e-05,
+      "loss": 5.6209,
+      "step": 18780
+    },
+    {
+      "epoch": 0.02640171280936214,
+      "grad_norm": 1.2707617282867432,
+      "learning_rate": 7.917942953491639e-05,
+      "loss": 5.6083,
+      "step": 18790
+    },
+    {
+      "epoch": 0.026415763747525717,
+      "grad_norm": 1.494612216949463,
+      "learning_rate": 7.922158212730081e-05,
+      "loss": 5.5413,
+      "step": 18800
+    },
+    {
+      "epoch": 0.026429814685689296,
+      "grad_norm": 1.4194583892822266,
+      "learning_rate": 7.926373471968525e-05,
+      "loss": 5.5375,
+      "step": 18810
+    },
+    {
+      "epoch": 0.026443865623852873,
+      "grad_norm": 1.5205409526824951,
+      "learning_rate": 7.930588731206969e-05,
+      "loss": 5.5377,
+      "step": 18820
+    },
+    {
+      "epoch": 0.02645791656201645,
+      "grad_norm": 1.2400825023651123,
+      "learning_rate": 7.934803990445413e-05,
+      "loss": 5.5744,
+      "step": 18830
+    },
+    {
+      "epoch": 0.02647196750018003,
+      "grad_norm": 1.3689380884170532,
+      "learning_rate": 7.939019249683855e-05,
+      "loss": 5.618,
+      "step": 18840
+    },
+    {
+      "epoch": 0.026486018438343605,
+      "grad_norm": 1.4133222103118896,
+      "learning_rate": 7.943234508922298e-05,
+      "loss": 5.7025,
+      "step": 18850
+    },
+    {
+      "epoch": 0.026500069376507184,
+      "grad_norm": 1.4041091203689575,
+      "learning_rate": 7.947449768160742e-05,
+      "loss": 5.5289,
+      "step": 18860
+    },
+    {
+      "epoch": 0.02651412031467076,
+      "grad_norm": 1.2778563499450684,
+      "learning_rate": 7.951665027399184e-05,
+      "loss": 5.504,
+      "step": 18870
+    },
+    {
+      "epoch": 0.026528171252834336,
+      "grad_norm": 1.4318342208862305,
+      "learning_rate": 7.955880286637627e-05,
+      "loss": 5.5416,
+      "step": 18880
+    },
+    {
+      "epoch": 0.026542222190997916,
+      "grad_norm": 1.3022536039352417,
+      "learning_rate": 7.960095545876072e-05,
+      "loss": 5.4591,
+      "step": 18890
+    },
+    {
+      "epoch": 0.026556273129161492,
+      "grad_norm": 1.3003450632095337,
+      "learning_rate": 7.964310805114514e-05,
+      "loss": 5.5049,
+      "step": 18900
+    },
+    {
+      "epoch": 0.026570324067325072,
+      "grad_norm": 1.4249634742736816,
+      "learning_rate": 7.968526064352957e-05,
+      "loss": 5.5521,
+      "step": 18910
+    },
+    {
+      "epoch": 0.026584375005488648,
+      "grad_norm": 2.4674744606018066,
+      "learning_rate": 7.972741323591401e-05,
+      "loss": 5.5821,
+      "step": 18920
+    },
+    {
+      "epoch": 0.026598425943652224,
+      "grad_norm": 1.4836219549179077,
+      "learning_rate": 7.976956582829843e-05,
+      "loss": 5.6154,
+      "step": 18930
+    },
+    {
+      "epoch": 0.026612476881815804,
+      "grad_norm": 1.395345687866211,
+      "learning_rate": 7.981171842068286e-05,
+      "loss": 5.589,
+      "step": 18940
+    },
+    {
+      "epoch": 0.02662652781997938,
+      "grad_norm": 1.3151682615280151,
+      "learning_rate": 7.98538710130673e-05,
+      "loss": 5.4916,
+      "step": 18950
+    },
+    {
+      "epoch": 0.026640578758142956,
+      "grad_norm": 1.6428303718566895,
+      "learning_rate": 7.989602360545172e-05,
+      "loss": 5.6574,
+      "step": 18960
+    },
+    {
+      "epoch": 0.026654629696306536,
+      "grad_norm": 1.3292784690856934,
+      "learning_rate": 7.993817619783615e-05,
+      "loss": 5.6062,
+      "step": 18970
+    },
+    {
+      "epoch": 0.026668680634470112,
+      "grad_norm": 1.4000933170318604,
+      "learning_rate": 7.99803287902206e-05,
+      "loss": 5.5415,
+      "step": 18980
+    },
+    {
+      "epoch": 0.02668273157263369,
+      "grad_norm": 1.428321361541748,
+      "learning_rate": 8.002248138260503e-05,
+      "loss": 5.5416,
+      "step": 18990
+    },
+    {
+      "epoch": 0.026696782510797268,
+      "grad_norm": 1.3596489429473877,
+      "learning_rate": 8.006463397498945e-05,
+      "loss": 5.6136,
+      "step": 19000
+    },
+    {
+      "epoch": 0.026710833448960844,
+      "grad_norm": 1.3439053297042847,
+      "learning_rate": 8.010678656737389e-05,
+      "loss": 5.5557,
+      "step": 19010
+    },
+    {
+      "epoch": 0.026724884387124424,
+      "grad_norm": 1.3456841707229614,
+      "learning_rate": 8.014893915975832e-05,
+      "loss": 5.484,
+      "step": 19020
+    },
+    {
+      "epoch": 0.026738935325288,
+      "grad_norm": 1.317115306854248,
+      "learning_rate": 8.019109175214274e-05,
+      "loss": 5.6083,
+      "step": 19030
+    },
+    {
+      "epoch": 0.02675298626345158,
+      "grad_norm": 1.845428466796875,
+      "learning_rate": 8.023324434452719e-05,
+      "loss": 5.5963,
+      "step": 19040
+    },
+    {
+      "epoch": 0.026767037201615156,
+      "grad_norm": 1.309800386428833,
+      "learning_rate": 8.027539693691162e-05,
+      "loss": 5.5722,
+      "step": 19050
+    },
+    {
+      "epoch": 0.026781088139778732,
+      "grad_norm": 1.389327883720398,
+      "learning_rate": 8.031754952929603e-05,
+      "loss": 5.4975,
+      "step": 19060
+    },
+    {
+      "epoch": 0.02679513907794231,
+      "grad_norm": 1.251827359199524,
+      "learning_rate": 8.035970212168048e-05,
+      "loss": 5.5743,
+      "step": 19070
+    },
+    {
+      "epoch": 0.026809190016105888,
+      "grad_norm": 1.2980040311813354,
+      "learning_rate": 8.040185471406491e-05,
+      "loss": 5.5124,
+      "step": 19080
+    },
+    {
+      "epoch": 0.026823240954269464,
+      "grad_norm": 1.3532140254974365,
+      "learning_rate": 8.044400730644933e-05,
+      "loss": 5.5087,
+      "step": 19090
+    },
+    {
+      "epoch": 0.026837291892433043,
+      "grad_norm": 1.3311684131622314,
+      "learning_rate": 8.048615989883377e-05,
+      "loss": 5.5773,
+      "step": 19100
+    },
+    {
+      "epoch": 0.02685134283059662,
+      "grad_norm": 1.3414361476898193,
+      "learning_rate": 8.05283124912182e-05,
+      "loss": 5.4767,
+      "step": 19110
+    },
+    {
+      "epoch": 0.0268653937687602,
+      "grad_norm": 1.47315514087677,
+      "learning_rate": 8.057046508360265e-05,
+      "loss": 5.5304,
+      "step": 19120
+    },
+    {
+      "epoch": 0.026879444706923775,
+      "grad_norm": 1.3902984857559204,
+      "learning_rate": 8.061261767598707e-05,
+      "loss": 5.5592,
+      "step": 19130
+    },
+    {
+      "epoch": 0.02689349564508735,
+      "grad_norm": 1.3104959726333618,
+      "learning_rate": 8.06547702683715e-05,
+      "loss": 5.5947,
+      "step": 19140
+    },
+    {
+      "epoch": 0.02690754658325093,
+      "grad_norm": 1.377226710319519,
+      "learning_rate": 8.069692286075594e-05,
+      "loss": 5.6055,
+      "step": 19150
+    },
+    {
+      "epoch": 0.026921597521414507,
+      "grad_norm": 1.3036943674087524,
+      "learning_rate": 8.073907545314036e-05,
+      "loss": 5.5622,
+      "step": 19160
+    },
+    {
+      "epoch": 0.026935648459578087,
+      "grad_norm": 1.395580768585205,
+      "learning_rate": 8.078122804552479e-05,
+      "loss": 5.4697,
+      "step": 19170
+    },
+    {
+      "epoch": 0.026949699397741663,
+      "grad_norm": 1.2811263799667358,
+      "learning_rate": 8.082338063790923e-05,
+      "loss": 5.5932,
+      "step": 19180
+    },
+    {
+      "epoch": 0.02696375033590524,
+      "grad_norm": 1.2819894552230835,
+      "learning_rate": 8.086553323029365e-05,
+      "loss": 5.5619,
+      "step": 19190
+    },
+    {
+      "epoch": 0.02697780127406882,
+      "grad_norm": 1.2503257989883423,
+      "learning_rate": 8.090768582267808e-05,
+      "loss": 5.6324,
+      "step": 19200
+    },
+    {
+      "epoch": 0.026991852212232395,
+      "grad_norm": 1.6546611785888672,
+      "learning_rate": 8.094983841506253e-05,
+      "loss": 5.451,
+      "step": 19210
+    },
+    {
+      "epoch": 0.027005903150395975,
+      "grad_norm": 1.3742631673812866,
+      "learning_rate": 8.099199100744695e-05,
+      "loss": 5.6103,
+      "step": 19220
+    },
+    {
+      "epoch": 0.02701995408855955,
+      "grad_norm": 1.5028876066207886,
+      "learning_rate": 8.103414359983138e-05,
+      "loss": 5.4252,
+      "step": 19230
+    },
+    {
+      "epoch": 0.027034005026723127,
+      "grad_norm": 1.4231295585632324,
+      "learning_rate": 8.107629619221582e-05,
+      "loss": 5.5802,
+      "step": 19240
+    },
+    {
+      "epoch": 0.027048055964886707,
+      "grad_norm": 1.3292512893676758,
+      "learning_rate": 8.111844878460025e-05,
+      "loss": 5.4812,
+      "step": 19250
+    },
+    {
+      "epoch": 0.027062106903050283,
+      "grad_norm": 1.4668748378753662,
+      "learning_rate": 8.116060137698467e-05,
+      "loss": 5.5739,
+      "step": 19260
+    },
+    {
+      "epoch": 0.02707615784121386,
+      "grad_norm": 1.3313186168670654,
+      "learning_rate": 8.120275396936912e-05,
+      "loss": 5.5789,
+      "step": 19270
+    },
+    {
+      "epoch": 0.02709020877937744,
+      "grad_norm": 1.3527790307998657,
+      "learning_rate": 8.124490656175355e-05,
+      "loss": 5.4486,
+      "step": 19280
+    },
+    {
+      "epoch": 0.027104259717541015,
+      "grad_norm": 1.3377543687820435,
+      "learning_rate": 8.128705915413796e-05,
+      "loss": 5.4651,
+      "step": 19290
+    },
+    {
+      "epoch": 0.027118310655704594,
+      "grad_norm": 1.3352538347244263,
+      "learning_rate": 8.132921174652241e-05,
+      "loss": 5.4763,
+      "step": 19300
+    },
+    {
+      "epoch": 0.02713236159386817,
+      "grad_norm": 1.3395129442214966,
+      "learning_rate": 8.137136433890684e-05,
+      "loss": 5.4819,
+      "step": 19310
+    },
+    {
+      "epoch": 0.027146412532031747,
+      "grad_norm": 1.3543874025344849,
+      "learning_rate": 8.141351693129126e-05,
+      "loss": 5.5129,
+      "step": 19320
+    },
+    {
+      "epoch": 0.027160463470195326,
+      "grad_norm": 1.3789215087890625,
+      "learning_rate": 8.14556695236757e-05,
+      "loss": 5.5339,
+      "step": 19330
+    },
+    {
+      "epoch": 0.027174514408358903,
+      "grad_norm": 1.388160228729248,
+      "learning_rate": 8.149782211606013e-05,
+      "loss": 5.5459,
+      "step": 19340
+    },
+    {
+      "epoch": 0.027188565346522482,
+      "grad_norm": 1.3371630907058716,
+      "learning_rate": 8.153997470844455e-05,
+      "loss": 5.6077,
+      "step": 19350
+    },
+    {
+      "epoch": 0.02720261628468606,
+      "grad_norm": 1.3276069164276123,
+      "learning_rate": 8.1582127300829e-05,
+      "loss": 5.5951,
+      "step": 19360
+    },
+    {
+      "epoch": 0.027216667222849635,
+      "grad_norm": 1.3663477897644043,
+      "learning_rate": 8.162427989321343e-05,
+      "loss": 5.581,
+      "step": 19370
+    },
+    {
+      "epoch": 0.027230718161013214,
+      "grad_norm": 1.48804771900177,
+      "learning_rate": 8.166643248559784e-05,
+      "loss": 5.6145,
+      "step": 19380
+    },
+    {
+      "epoch": 0.02724476909917679,
+      "grad_norm": 1.3625435829162598,
+      "learning_rate": 8.170858507798229e-05,
+      "loss": 5.4211,
+      "step": 19390
+    },
+    {
+      "epoch": 0.027258820037340366,
+      "grad_norm": 1.324641227722168,
+      "learning_rate": 8.175073767036672e-05,
+      "loss": 5.429,
+      "step": 19400
+    },
+    {
+      "epoch": 0.027272870975503946,
+      "grad_norm": 1.3358043432235718,
+      "learning_rate": 8.179289026275117e-05,
+      "loss": 5.4692,
+      "step": 19410
+    },
+    {
+      "epoch": 0.027286921913667522,
+      "grad_norm": 1.48837411403656,
+      "learning_rate": 8.183504285513558e-05,
+      "loss": 5.5711,
+      "step": 19420
+    },
+    {
+      "epoch": 0.027300972851831102,
+      "grad_norm": 1.3000835180282593,
+      "learning_rate": 8.187719544752001e-05,
+      "loss": 5.4646,
+      "step": 19430
+    },
+    {
+      "epoch": 0.027315023789994678,
+      "grad_norm": 1.4490911960601807,
+      "learning_rate": 8.191934803990446e-05,
+      "loss": 5.495,
+      "step": 19440
+    },
+    {
+      "epoch": 0.027329074728158254,
+      "grad_norm": 1.3494820594787598,
+      "learning_rate": 8.196150063228888e-05,
+      "loss": 5.573,
+      "step": 19450
+    },
+    {
+      "epoch": 0.027343125666321834,
+      "grad_norm": 1.3359415531158447,
+      "learning_rate": 8.200365322467331e-05,
+      "loss": 5.6712,
+      "step": 19460
+    },
+    {
+      "epoch": 0.02735717660448541,
+      "grad_norm": 1.3212742805480957,
+      "learning_rate": 8.204580581705775e-05,
+      "loss": 5.6058,
+      "step": 19470
+    },
+    {
+      "epoch": 0.02737122754264899,
+      "grad_norm": 1.3946962356567383,
+      "learning_rate": 8.208795840944217e-05,
+      "loss": 5.4978,
+      "step": 19480
+    },
+    {
+      "epoch": 0.027385278480812566,
+      "grad_norm": 1.3919817209243774,
+      "learning_rate": 8.21301110018266e-05,
+      "loss": 5.5974,
+      "step": 19490
+    },
+    {
+      "epoch": 0.027399329418976142,
+      "grad_norm": 1.320351481437683,
+      "learning_rate": 8.217226359421105e-05,
+      "loss": 5.5079,
+      "step": 19500
+    },
+    {
+      "epoch": 0.02741338035713972,
+      "grad_norm": 1.326079249382019,
+      "learning_rate": 8.221441618659546e-05,
+      "loss": 5.5104,
+      "step": 19510
+    },
+    {
+      "epoch": 0.027427431295303298,
+      "grad_norm": 1.6833606958389282,
+      "learning_rate": 8.22565687789799e-05,
+      "loss": 5.5943,
+      "step": 19520
+    },
+    {
+      "epoch": 0.027441482233466877,
+      "grad_norm": 1.373024344444275,
+      "learning_rate": 8.229872137136434e-05,
+      "loss": 5.5817,
+      "step": 19530
+    },
+    {
+      "epoch": 0.027455533171630454,
+      "grad_norm": 1.2997173070907593,
+      "learning_rate": 8.234087396374877e-05,
+      "loss": 5.5115,
+      "step": 19540
+    },
+    {
+      "epoch": 0.02746958410979403,
+      "grad_norm": 1.349067211151123,
+      "learning_rate": 8.238302655613319e-05,
+      "loss": 5.5306,
+      "step": 19550
+    },
+    {
+      "epoch": 0.02748363504795761,
+      "grad_norm": 1.379137396812439,
+      "learning_rate": 8.242517914851763e-05,
+      "loss": 5.5341,
+      "step": 19560
+    },
+    {
+      "epoch": 0.027497685986121186,
+      "grad_norm": 1.4701979160308838,
+      "learning_rate": 8.246733174090206e-05,
+      "loss": 5.4278,
+      "step": 19570
+    },
+    {
+      "epoch": 0.027511736924284762,
+      "grad_norm": 1.373063087463379,
+      "learning_rate": 8.250948433328648e-05,
+      "loss": 5.4814,
+      "step": 19580
+    },
+    {
+      "epoch": 0.02752578786244834,
+      "grad_norm": 1.3411657810211182,
+      "learning_rate": 8.255163692567093e-05,
+      "loss": 5.5641,
+      "step": 19590
+    },
+    {
+      "epoch": 0.027539838800611918,
+      "grad_norm": 1.2834957838058472,
+      "learning_rate": 8.259378951805536e-05,
+      "loss": 5.5242,
+      "step": 19600
+    },
+    {
+      "epoch": 0.027553889738775497,
+      "grad_norm": 1.3774769306182861,
+      "learning_rate": 8.263594211043978e-05,
+      "loss": 5.6078,
+      "step": 19610
+    },
+    {
+      "epoch": 0.027567940676939073,
+      "grad_norm": 1.3527755737304688,
+      "learning_rate": 8.267809470282422e-05,
+      "loss": 5.443,
+      "step": 19620
+    },
+    {
+      "epoch": 0.02758199161510265,
+      "grad_norm": 1.8241328001022339,
+      "learning_rate": 8.272024729520865e-05,
+      "loss": 5.62,
+      "step": 19630
+    },
+    {
+      "epoch": 0.02759604255326623,
+      "grad_norm": 1.3106392621994019,
+      "learning_rate": 8.276239988759307e-05,
+      "loss": 5.5135,
+      "step": 19640
+    },
+    {
+      "epoch": 0.027610093491429805,
+      "grad_norm": 1.3482909202575684,
+      "learning_rate": 8.280455247997751e-05,
+      "loss": 5.4549,
+      "step": 19650
+    },
+    {
+      "epoch": 0.027624144429593385,
+      "grad_norm": 1.3328837156295776,
+      "learning_rate": 8.284670507236194e-05,
+      "loss": 5.4759,
+      "step": 19660
+    },
+    {
+      "epoch": 0.02763819536775696,
+      "grad_norm": 1.3374454975128174,
+      "learning_rate": 8.288885766474636e-05,
+      "loss": 5.5537,
+      "step": 19670
+    },
+    {
+      "epoch": 0.027652246305920537,
+      "grad_norm": 1.3683466911315918,
+      "learning_rate": 8.293101025713081e-05,
+      "loss": 5.462,
+      "step": 19680
+    },
+    {
+      "epoch": 0.027666297244084117,
+      "grad_norm": 1.3449815511703491,
+      "learning_rate": 8.297316284951524e-05,
+      "loss": 5.5387,
+      "step": 19690
+    },
+    {
+      "epoch": 0.027680348182247693,
+      "grad_norm": 1.3451807498931885,
+      "learning_rate": 8.301531544189968e-05,
+      "loss": 5.5744,
+      "step": 19700
+    },
+    {
+      "epoch": 0.02769439912041127,
+      "grad_norm": 1.3333075046539307,
+      "learning_rate": 8.30574680342841e-05,
+      "loss": 5.5967,
+      "step": 19710
+    },
+    {
+      "epoch": 0.02770845005857485,
+      "grad_norm": 1.331404685974121,
+      "learning_rate": 8.309962062666853e-05,
+      "loss": 5.5404,
+      "step": 19720
+    },
+    {
+      "epoch": 0.027722500996738425,
+      "grad_norm": 2.0781617164611816,
+      "learning_rate": 8.314177321905298e-05,
+      "loss": 5.4125,
+      "step": 19730
+    },
+    {
+      "epoch": 0.027736551934902005,
+      "grad_norm": 1.341507077217102,
+      "learning_rate": 8.31839258114374e-05,
+      "loss": 5.5681,
+      "step": 19740
+    },
+    {
+      "epoch": 0.02775060287306558,
+      "grad_norm": 1.3700333833694458,
+      "learning_rate": 8.322607840382183e-05,
+      "loss": 5.5813,
+      "step": 19750
+    },
+    {
+      "epoch": 0.027764653811229157,
+      "grad_norm": 1.3675099611282349,
+      "learning_rate": 8.326823099620627e-05,
+      "loss": 5.4456,
+      "step": 19760
+    },
+    {
+      "epoch": 0.027778704749392737,
+      "grad_norm": 1.3649263381958008,
+      "learning_rate": 8.331038358859069e-05,
+      "loss": 5.5941,
+      "step": 19770
+    },
+    {
+      "epoch": 0.027792755687556313,
+      "grad_norm": 1.3281179666519165,
+      "learning_rate": 8.335253618097512e-05,
+      "loss": 5.4776,
+      "step": 19780
+    },
+    {
+      "epoch": 0.027806806625719892,
+      "grad_norm": 1.314491629600525,
+      "learning_rate": 8.339468877335956e-05,
+      "loss": 5.6428,
+      "step": 19790
+    },
+    {
+      "epoch": 0.02782085756388347,
+      "grad_norm": 1.4704593420028687,
+      "learning_rate": 8.343684136574398e-05,
+      "loss": 5.5351,
+      "step": 19800
+    },
+    {
+      "epoch": 0.027834908502047045,
+      "grad_norm": 1.364715337753296,
+      "learning_rate": 8.347899395812841e-05,
+      "loss": 5.6232,
+      "step": 19810
+    },
+    {
+      "epoch": 0.027848959440210624,
+      "grad_norm": 1.2975870370864868,
+      "learning_rate": 8.352114655051286e-05,
+      "loss": 5.5265,
+      "step": 19820
+    },
+    {
+      "epoch": 0.0278630103783742,
+      "grad_norm": 1.2619905471801758,
+      "learning_rate": 8.356329914289729e-05,
+      "loss": 5.5378,
+      "step": 19830
+    },
+    {
+      "epoch": 0.02787706131653778,
+      "grad_norm": 1.4168630838394165,
+      "learning_rate": 8.36054517352817e-05,
+      "loss": 5.5177,
+      "step": 19840
+    },
+    {
+      "epoch": 0.027891112254701356,
+      "grad_norm": 1.4585949182510376,
+      "learning_rate": 8.364760432766615e-05,
+      "loss": 5.4355,
+      "step": 19850
+    },
+    {
+      "epoch": 0.027905163192864933,
+      "grad_norm": 1.356927752494812,
+      "learning_rate": 8.368975692005058e-05,
+      "loss": 5.5808,
+      "step": 19860
+    },
+    {
+      "epoch": 0.027919214131028512,
+      "grad_norm": 1.495552659034729,
+      "learning_rate": 8.3731909512435e-05,
+      "loss": 5.5017,
+      "step": 19870
+    },
+    {
+      "epoch": 0.02793326506919209,
+      "grad_norm": 1.2983280420303345,
+      "learning_rate": 8.377406210481944e-05,
+      "loss": 5.4182,
+      "step": 19880
+    },
+    {
+      "epoch": 0.027947316007355665,
+      "grad_norm": 1.3719418048858643,
+      "learning_rate": 8.381621469720387e-05,
+      "loss": 5.5936,
+      "step": 19890
+    },
+    {
+      "epoch": 0.027961366945519244,
+      "grad_norm": 1.450944423675537,
+      "learning_rate": 8.385836728958829e-05,
+      "loss": 5.4513,
+      "step": 19900
+    },
+    {
+      "epoch": 0.02797541788368282,
+      "grad_norm": 1.314232587814331,
+      "learning_rate": 8.390051988197274e-05,
+      "loss": 5.4597,
+      "step": 19910
+    },
+    {
+      "epoch": 0.0279894688218464,
+      "grad_norm": 1.274675726890564,
+      "learning_rate": 8.394267247435717e-05,
+      "loss": 5.4734,
+      "step": 19920
+    },
+    {
+      "epoch": 0.028003519760009976,
+      "grad_norm": 1.3578556776046753,
+      "learning_rate": 8.398482506674159e-05,
+      "loss": 5.4294,
+      "step": 19930
+    },
+    {
+      "epoch": 0.028017570698173552,
+      "grad_norm": 1.3357514142990112,
+      "learning_rate": 8.402697765912603e-05,
+      "loss": 5.4595,
+      "step": 19940
+    },
+    {
+      "epoch": 0.028031621636337132,
+      "grad_norm": 1.3431594371795654,
+      "learning_rate": 8.406913025151046e-05,
+      "loss": 5.5151,
+      "step": 19950
+    },
+    {
+      "epoch": 0.028045672574500708,
+      "grad_norm": 1.5500048398971558,
+      "learning_rate": 8.41112828438949e-05,
+      "loss": 5.5227,
+      "step": 19960
+    },
+    {
+      "epoch": 0.028059723512664288,
+      "grad_norm": 1.3698660135269165,
+      "learning_rate": 8.415343543627932e-05,
+      "loss": 5.5025,
+      "step": 19970
+    },
+    {
+      "epoch": 0.028073774450827864,
+      "grad_norm": 1.332960844039917,
+      "learning_rate": 8.419558802866376e-05,
+      "loss": 5.5337,
+      "step": 19980
+    },
+    {
+      "epoch": 0.02808782538899144,
+      "grad_norm": 1.2651346921920776,
+      "learning_rate": 8.42377406210482e-05,
+      "loss": 5.5576,
+      "step": 19990
+    },
+    {
+      "epoch": 0.02810187632715502,
+      "grad_norm": 1.3758925199508667,
+      "learning_rate": 8.427989321343262e-05,
+      "loss": 5.3875,
+      "step": 20000
+    },
+    {
+      "epoch": 0.028115927265318596,
+      "grad_norm": 1.412123203277588,
+      "learning_rate": 8.432204580581705e-05,
+      "loss": 5.6327,
+      "step": 20010
+    },
+    {
+      "epoch": 0.028129978203482175,
+      "grad_norm": 1.3514657020568848,
+      "learning_rate": 8.43641983982015e-05,
+      "loss": 5.5142,
+      "step": 20020
+    },
+    {
+      "epoch": 0.02814402914164575,
+      "grad_norm": 1.4179363250732422,
+      "learning_rate": 8.440635099058591e-05,
+      "loss": 5.5414,
+      "step": 20030
+    },
+    {
+      "epoch": 0.028158080079809328,
+      "grad_norm": 1.3700520992279053,
+      "learning_rate": 8.444850358297034e-05,
+      "loss": 5.5191,
+      "step": 20040
+    },
+    {
+      "epoch": 0.028172131017972907,
+      "grad_norm": 1.3933957815170288,
+      "learning_rate": 8.449065617535479e-05,
+      "loss": 5.5648,
+      "step": 20050
+    },
+    {
+      "epoch": 0.028186181956136484,
+      "grad_norm": 1.2371602058410645,
+      "learning_rate": 8.45328087677392e-05,
+      "loss": 5.5392,
+      "step": 20060
+    },
+    {
+      "epoch": 0.02820023289430006,
+      "grad_norm": 1.3389537334442139,
+      "learning_rate": 8.457496136012364e-05,
+      "loss": 5.4112,
+      "step": 20070
+    },
+    {
+      "epoch": 0.02821428383246364,
+      "grad_norm": 1.3131827116012573,
+      "learning_rate": 8.461711395250808e-05,
+      "loss": 5.5364,
+      "step": 20080
+    },
+    {
+      "epoch": 0.028228334770627216,
+      "grad_norm": 1.3104963302612305,
+      "learning_rate": 8.46592665448925e-05,
+      "loss": 5.4475,
+      "step": 20090
+    },
+    {
+      "epoch": 0.028242385708790795,
+      "grad_norm": 1.3459300994873047,
+      "learning_rate": 8.470141913727693e-05,
+      "loss": 5.5386,
+      "step": 20100
+    },
+    {
+      "epoch": 0.02825643664695437,
+      "grad_norm": 1.3658233880996704,
+      "learning_rate": 8.474357172966137e-05,
+      "loss": 5.5473,
+      "step": 20110
+    },
+    {
+      "epoch": 0.028270487585117948,
+      "grad_norm": 1.3540921211242676,
+      "learning_rate": 8.47857243220458e-05,
+      "loss": 5.4997,
+      "step": 20120
+    },
+    {
+      "epoch": 0.028284538523281527,
+      "grad_norm": 1.2725752592086792,
+      "learning_rate": 8.482787691443022e-05,
+      "loss": 5.5258,
+      "step": 20130
+    },
+    {
+      "epoch": 0.028298589461445103,
+      "grad_norm": 1.3213703632354736,
+      "learning_rate": 8.487002950681467e-05,
+      "loss": 5.5548,
+      "step": 20140
+    },
+    {
+      "epoch": 0.028312640399608683,
+      "grad_norm": 1.304655909538269,
+      "learning_rate": 8.49121820991991e-05,
+      "loss": 5.648,
+      "step": 20150
+    },
+    {
+      "epoch": 0.02832669133777226,
+      "grad_norm": 1.2564914226531982,
+      "learning_rate": 8.495433469158352e-05,
+      "loss": 5.5097,
+      "step": 20160
+    },
+    {
+      "epoch": 0.028340742275935835,
+      "grad_norm": 1.2632144689559937,
+      "learning_rate": 8.499648728396796e-05,
+      "loss": 5.4913,
+      "step": 20170
+    },
+    {
+      "epoch": 0.028354793214099415,
+      "grad_norm": 1.2769023180007935,
+      "learning_rate": 8.503863987635239e-05,
+      "loss": 5.504,
+      "step": 20180
+    },
+    {
+      "epoch": 0.02836884415226299,
+      "grad_norm": 1.3045737743377686,
+      "learning_rate": 8.508079246873681e-05,
+      "loss": 5.3591,
+      "step": 20190
+    },
+    {
+      "epoch": 0.028382895090426567,
+      "grad_norm": 1.3489607572555542,
+      "learning_rate": 8.512294506112125e-05,
+      "loss": 5.4323,
+      "step": 20200
+    },
+    {
+      "epoch": 0.028396946028590147,
+      "grad_norm": 1.3132845163345337,
+      "learning_rate": 8.516509765350569e-05,
+      "loss": 5.5967,
+      "step": 20210
+    },
+    {
+      "epoch": 0.028410996966753723,
+      "grad_norm": 1.3848180770874023,
+      "learning_rate": 8.52072502458901e-05,
+      "loss": 5.5421,
+      "step": 20220
+    },
+    {
+      "epoch": 0.028425047904917303,
+      "grad_norm": 1.3170721530914307,
+      "learning_rate": 8.524940283827455e-05,
+      "loss": 5.5576,
+      "step": 20230
+    },
+    {
+      "epoch": 0.02843909884308088,
+      "grad_norm": 1.2676606178283691,
+      "learning_rate": 8.529155543065898e-05,
+      "loss": 5.5059,
+      "step": 20240
+    },
+    {
+      "epoch": 0.028453149781244455,
+      "grad_norm": 1.3508251905441284,
+      "learning_rate": 8.533370802304342e-05,
+      "loss": 5.3921,
+      "step": 20250
+    },
+    {
+      "epoch": 0.028467200719408035,
+      "grad_norm": 1.4981266260147095,
+      "learning_rate": 8.537586061542784e-05,
+      "loss": 5.4407,
+      "step": 20260
+    },
+    {
+      "epoch": 0.02848125165757161,
+      "grad_norm": 1.2797385454177856,
+      "learning_rate": 8.541801320781227e-05,
+      "loss": 5.4869,
+      "step": 20270
+    },
+    {
+      "epoch": 0.02849530259573519,
+      "grad_norm": 1.394260287284851,
+      "learning_rate": 8.546016580019672e-05,
+      "loss": 5.5474,
+      "step": 20280
+    },
+    {
+      "epoch": 0.028509353533898767,
+      "grad_norm": 1.4078445434570312,
+      "learning_rate": 8.550231839258114e-05,
+      "loss": 5.384,
+      "step": 20290
+    },
+    {
+      "epoch": 0.028523404472062343,
+      "grad_norm": 1.7833536863327026,
+      "learning_rate": 8.554447098496557e-05,
+      "loss": 5.4554,
+      "step": 20300
+    },
+    {
+      "epoch": 0.028537455410225922,
+      "grad_norm": 1.3692078590393066,
+      "learning_rate": 8.558662357735001e-05,
+      "loss": 5.4846,
+      "step": 20310
+    },
+    {
+      "epoch": 0.0285515063483895,
+      "grad_norm": 1.3532313108444214,
+      "learning_rate": 8.562877616973443e-05,
+      "loss": 5.3635,
+      "step": 20320
+    },
+    {
+      "epoch": 0.028565557286553078,
+      "grad_norm": 1.2888007164001465,
+      "learning_rate": 8.567092876211886e-05,
+      "loss": 5.5104,
+      "step": 20330
+    },
+    {
+      "epoch": 0.028579608224716654,
+      "grad_norm": 1.3509674072265625,
+      "learning_rate": 8.57130813545033e-05,
+      "loss": 5.5451,
+      "step": 20340
+    },
+    {
+      "epoch": 0.02859365916288023,
+      "grad_norm": 1.3917303085327148,
+      "learning_rate": 8.575523394688772e-05,
+      "loss": 5.4849,
+      "step": 20350
+    },
+    {
+      "epoch": 0.02860771010104381,
+      "grad_norm": 1.3092026710510254,
+      "learning_rate": 8.579738653927215e-05,
+      "loss": 5.3983,
+      "step": 20360
+    },
+    {
+      "epoch": 0.028621761039207386,
+      "grad_norm": 1.3765193223953247,
+      "learning_rate": 8.58395391316566e-05,
+      "loss": 5.5259,
+      "step": 20370
+    },
+    {
+      "epoch": 0.028635811977370963,
+      "grad_norm": 1.38758385181427,
+      "learning_rate": 8.588169172404102e-05,
+      "loss": 5.4927,
+      "step": 20380
+    },
+    {
+      "epoch": 0.028649862915534542,
+      "grad_norm": 1.3264268636703491,
+      "learning_rate": 8.592384431642545e-05,
+      "loss": 5.5557,
+      "step": 20390
+    },
+    {
+      "epoch": 0.02866391385369812,
+      "grad_norm": 1.3711224794387817,
+      "learning_rate": 8.596599690880989e-05,
+      "loss": 5.4997,
+      "step": 20400
+    },
+    {
+      "epoch": 0.028677964791861698,
+      "grad_norm": 1.430208683013916,
+      "learning_rate": 8.600814950119432e-05,
+      "loss": 5.3837,
+      "step": 20410
+    },
+    {
+      "epoch": 0.028692015730025274,
+      "grad_norm": 1.2916505336761475,
+      "learning_rate": 8.605030209357874e-05,
+      "loss": 5.5252,
+      "step": 20420
+    },
+    {
+      "epoch": 0.02870606666818885,
+      "grad_norm": 1.300502061843872,
+      "learning_rate": 8.609245468596319e-05,
+      "loss": 5.4593,
+      "step": 20430
+    },
+    {
+      "epoch": 0.02872011760635243,
+      "grad_norm": 1.2897007465362549,
+      "learning_rate": 8.613460727834762e-05,
+      "loss": 5.4996,
+      "step": 20440
+    },
+    {
+      "epoch": 0.028734168544516006,
+      "grad_norm": 1.3202191591262817,
+      "learning_rate": 8.617675987073203e-05,
+      "loss": 5.3855,
+      "step": 20450
+    },
+    {
+      "epoch": 0.028748219482679586,
+      "grad_norm": 1.3272267580032349,
+      "learning_rate": 8.621891246311648e-05,
+      "loss": 5.5036,
+      "step": 20460
+    },
+    {
+      "epoch": 0.028762270420843162,
+      "grad_norm": 1.308984398841858,
+      "learning_rate": 8.626106505550091e-05,
+      "loss": 5.5322,
+      "step": 20470
+    },
+    {
+      "epoch": 0.028776321359006738,
+      "grad_norm": 1.3932908773422241,
+      "learning_rate": 8.630321764788533e-05,
+      "loss": 5.524,
+      "step": 20480
+    },
+    {
+      "epoch": 0.028790372297170318,
+      "grad_norm": 1.3947856426239014,
+      "learning_rate": 8.634537024026977e-05,
+      "loss": 5.5781,
+      "step": 20490
+    },
+    {
+      "epoch": 0.028804423235333894,
+      "grad_norm": 1.2810063362121582,
+      "learning_rate": 8.63875228326542e-05,
+      "loss": 5.4543,
+      "step": 20500
+    },
+    {
+      "epoch": 0.02881847417349747,
+      "grad_norm": 1.3561229705810547,
+      "learning_rate": 8.642967542503862e-05,
+      "loss": 5.5105,
+      "step": 20510
+    },
+    {
+      "epoch": 0.02883252511166105,
+      "grad_norm": 1.3260747194290161,
+      "learning_rate": 8.647182801742307e-05,
+      "loss": 5.451,
+      "step": 20520
+    },
+    {
+      "epoch": 0.028846576049824626,
+      "grad_norm": 1.3854423761367798,
+      "learning_rate": 8.65139806098075e-05,
+      "loss": 5.4256,
+      "step": 20530
+    },
+    {
+      "epoch": 0.028860626987988205,
+      "grad_norm": 1.3165987730026245,
+      "learning_rate": 8.655613320219194e-05,
+      "loss": 5.3924,
+      "step": 20540
+    },
+    {
+      "epoch": 0.02887467792615178,
+      "grad_norm": 2.6763813495635986,
+      "learning_rate": 8.659828579457636e-05,
+      "loss": 5.524,
+      "step": 20550
+    },
+    {
+      "epoch": 0.028888728864315358,
+      "grad_norm": 1.459052324295044,
+      "learning_rate": 8.664043838696079e-05,
+      "loss": 5.4794,
+      "step": 20560
+    },
+    {
+      "epoch": 0.028902779802478937,
+      "grad_norm": 1.276800274848938,
+      "learning_rate": 8.668259097934523e-05,
+      "loss": 5.4635,
+      "step": 20570
+    },
+    {
+      "epoch": 0.028916830740642514,
+      "grad_norm": 1.3238308429718018,
+      "learning_rate": 8.672474357172965e-05,
+      "loss": 5.5032,
+      "step": 20580
+    },
+    {
+      "epoch": 0.028930881678806093,
+      "grad_norm": 1.308498740196228,
+      "learning_rate": 8.676689616411408e-05,
+      "loss": 5.5274,
+      "step": 20590
+    },
+    {
+      "epoch": 0.02894493261696967,
+      "grad_norm": 1.2564195394515991,
+      "learning_rate": 8.680904875649853e-05,
+      "loss": 5.5354,
+      "step": 20600
+    },
+    {
+      "epoch": 0.028958983555133246,
+      "grad_norm": 1.276263952255249,
+      "learning_rate": 8.685120134888295e-05,
+      "loss": 5.4732,
+      "step": 20610
+    },
+    {
+      "epoch": 0.028973034493296825,
+      "grad_norm": 1.257238745689392,
+      "learning_rate": 8.689335394126738e-05,
+      "loss": 5.5754,
+      "step": 20620
+    },
+    {
+      "epoch": 0.0289870854314604,
+      "grad_norm": 1.402646541595459,
+      "learning_rate": 8.693550653365182e-05,
+      "loss": 5.5832,
+      "step": 20630
+    },
+    {
+      "epoch": 0.02900113636962398,
+      "grad_norm": 1.2543669939041138,
+      "learning_rate": 8.697765912603624e-05,
+      "loss": 5.5227,
+      "step": 20640
+    },
+    {
+      "epoch": 0.029015187307787557,
+      "grad_norm": 1.3000344038009644,
+      "learning_rate": 8.701981171842067e-05,
+      "loss": 5.3653,
+      "step": 20650
+    },
+    {
+      "epoch": 0.029029238245951133,
+      "grad_norm": 1.3282537460327148,
+      "learning_rate": 8.706196431080512e-05,
+      "loss": 5.4161,
+      "step": 20660
+    },
+    {
+      "epoch": 0.029043289184114713,
+      "grad_norm": 1.3235125541687012,
+      "learning_rate": 8.710411690318953e-05,
+      "loss": 5.5692,
+      "step": 20670
+    },
+    {
+      "epoch": 0.02905734012227829,
+      "grad_norm": 1.3211613893508911,
+      "learning_rate": 8.714626949557396e-05,
+      "loss": 5.4011,
+      "step": 20680
+    },
+    {
+      "epoch": 0.029071391060441865,
+      "grad_norm": 1.3463915586471558,
+      "learning_rate": 8.718842208795841e-05,
+      "loss": 5.5313,
+      "step": 20690
+    },
+    {
+      "epoch": 0.029085441998605445,
+      "grad_norm": 1.304567813873291,
+      "learning_rate": 8.723057468034284e-05,
+      "loss": 5.3108,
+      "step": 20700
+    },
+    {
+      "epoch": 0.02909949293676902,
+      "grad_norm": 1.2378467321395874,
+      "learning_rate": 8.727272727272726e-05,
+      "loss": 5.5691,
+      "step": 20710
+    },
+    {
+      "epoch": 0.0291135438749326,
+      "grad_norm": 1.2816358804702759,
+      "learning_rate": 8.73148798651117e-05,
+      "loss": 5.4171,
+      "step": 20720
+    },
+    {
+      "epoch": 0.029127594813096177,
+      "grad_norm": 1.467879056930542,
+      "learning_rate": 8.735703245749613e-05,
+      "loss": 5.424,
+      "step": 20730
+    },
+    {
+      "epoch": 0.029141645751259753,
+      "grad_norm": 1.2690702676773071,
+      "learning_rate": 8.739918504988055e-05,
+      "loss": 5.5673,
+      "step": 20740
+    },
+    {
+      "epoch": 0.029155696689423333,
+      "grad_norm": 1.2710336446762085,
+      "learning_rate": 8.7441337642265e-05,
+      "loss": 5.47,
+      "step": 20750
+    },
+    {
+      "epoch": 0.02916974762758691,
+      "grad_norm": 1.2570821046829224,
+      "learning_rate": 8.748349023464943e-05,
+      "loss": 5.5539,
+      "step": 20760
+    },
+    {
+      "epoch": 0.02918379856575049,
+      "grad_norm": 1.3138800859451294,
+      "learning_rate": 8.752564282703384e-05,
+      "loss": 5.4017,
+      "step": 20770
+    },
+    {
+      "epoch": 0.029197849503914065,
+      "grad_norm": 1.3933358192443848,
+      "learning_rate": 8.756779541941829e-05,
+      "loss": 5.3666,
+      "step": 20780
+    },
+    {
+      "epoch": 0.02921190044207764,
+      "grad_norm": 1.499240517616272,
+      "learning_rate": 8.760994801180272e-05,
+      "loss": 5.4957,
+      "step": 20790
+    },
+    {
+      "epoch": 0.02922595138024122,
+      "grad_norm": 1.332919955253601,
+      "learning_rate": 8.765210060418714e-05,
+      "loss": 5.4886,
+      "step": 20800
+    },
+    {
+      "epoch": 0.029240002318404797,
+      "grad_norm": 1.2993764877319336,
+      "learning_rate": 8.769425319657158e-05,
+      "loss": 5.5195,
+      "step": 20810
+    },
+    {
+      "epoch": 0.029254053256568373,
+      "grad_norm": 1.3067516088485718,
+      "learning_rate": 8.773640578895601e-05,
+      "loss": 5.5169,
+      "step": 20820
+    },
+    {
+      "epoch": 0.029268104194731952,
+      "grad_norm": 1.28152596950531,
+      "learning_rate": 8.777855838134046e-05,
+      "loss": 5.4082,
+      "step": 20830
+    },
+    {
+      "epoch": 0.02928215513289553,
+      "grad_norm": 1.3164806365966797,
+      "learning_rate": 8.782071097372488e-05,
+      "loss": 5.5273,
+      "step": 20840
+    },
+    {
+      "epoch": 0.029296206071059108,
+      "grad_norm": 1.3631839752197266,
+      "learning_rate": 8.786286356610931e-05,
+      "loss": 5.4297,
+      "step": 20850
+    },
+    {
+      "epoch": 0.029310257009222684,
+      "grad_norm": 1.3792098760604858,
+      "learning_rate": 8.790501615849375e-05,
+      "loss": 5.4204,
+      "step": 20860
+    },
+    {
+      "epoch": 0.02932430794738626,
+      "grad_norm": 1.3812150955200195,
+      "learning_rate": 8.794716875087817e-05,
+      "loss": 5.3696,
+      "step": 20870
+    },
+    {
+      "epoch": 0.02933835888554984,
+      "grad_norm": 1.2828326225280762,
+      "learning_rate": 8.79893213432626e-05,
+      "loss": 5.5572,
+      "step": 20880
+    },
+    {
+      "epoch": 0.029352409823713416,
+      "grad_norm": 1.4241127967834473,
+      "learning_rate": 8.803147393564705e-05,
+      "loss": 5.3305,
+      "step": 20890
+    },
+    {
+      "epoch": 0.029366460761876996,
+      "grad_norm": 1.386412501335144,
+      "learning_rate": 8.807362652803146e-05,
+      "loss": 5.5092,
+      "step": 20900
+    },
+    {
+      "epoch": 0.029380511700040572,
+      "grad_norm": 1.2819172143936157,
+      "learning_rate": 8.81157791204159e-05,
+      "loss": 5.3611,
+      "step": 20910
+    },
+    {
+      "epoch": 0.02939456263820415,
+      "grad_norm": 1.305132269859314,
+      "learning_rate": 8.815793171280034e-05,
+      "loss": 5.5092,
+      "step": 20920
+    },
+    {
+      "epoch": 0.029408613576367728,
+      "grad_norm": 1.359618902206421,
+      "learning_rate": 8.820008430518476e-05,
+      "loss": 5.5222,
+      "step": 20930
+    },
+    {
+      "epoch": 0.029422664514531304,
+      "grad_norm": 1.3207499980926514,
+      "learning_rate": 8.824223689756919e-05,
+      "loss": 5.4073,
+      "step": 20940
+    },
+    {
+      "epoch": 0.029436715452694884,
+      "grad_norm": 1.4006059169769287,
+      "learning_rate": 8.828438948995363e-05,
+      "loss": 5.4006,
+      "step": 20950
+    },
+    {
+      "epoch": 0.02945076639085846,
+      "grad_norm": 1.3299939632415771,
+      "learning_rate": 8.832654208233806e-05,
+      "loss": 5.5144,
+      "step": 20960
+    },
+    {
+      "epoch": 0.029464817329022036,
+      "grad_norm": 1.2843600511550903,
+      "learning_rate": 8.836869467472248e-05,
+      "loss": 5.4639,
+      "step": 20970
+    },
+    {
+      "epoch": 0.029478868267185616,
+      "grad_norm": 1.2437927722930908,
+      "learning_rate": 8.841084726710693e-05,
+      "loss": 5.5152,
+      "step": 20980
+    },
+    {
+      "epoch": 0.029492919205349192,
+      "grad_norm": 1.3445243835449219,
+      "learning_rate": 8.845299985949136e-05,
+      "loss": 5.5151,
+      "step": 20990
+    },
+    {
+      "epoch": 0.029506970143512768,
+      "grad_norm": 1.3153207302093506,
+      "learning_rate": 8.849515245187578e-05,
+      "loss": 5.5303,
+      "step": 21000
+    },
+    {
+      "epoch": 0.029521021081676348,
+      "grad_norm": 1.3267221450805664,
+      "learning_rate": 8.853730504426022e-05,
+      "loss": 5.5593,
+      "step": 21010
+    },
+    {
+      "epoch": 0.029535072019839924,
+      "grad_norm": 1.3878754377365112,
+      "learning_rate": 8.857945763664465e-05,
+      "loss": 5.4331,
+      "step": 21020
+    },
+    {
+      "epoch": 0.029549122958003503,
+      "grad_norm": 1.3253564834594727,
+      "learning_rate": 8.862161022902907e-05,
+      "loss": 5.4197,
+      "step": 21030
+    },
+    {
+      "epoch": 0.02956317389616708,
+      "grad_norm": 1.2962226867675781,
+      "learning_rate": 8.866376282141351e-05,
+      "loss": 5.4463,
+      "step": 21040
+    },
+    {
+      "epoch": 0.029577224834330656,
+      "grad_norm": 1.2892351150512695,
+      "learning_rate": 8.870591541379794e-05,
+      "loss": 5.4884,
+      "step": 21050
+    },
+    {
+      "epoch": 0.029591275772494235,
+      "grad_norm": 1.435368537902832,
+      "learning_rate": 8.874806800618236e-05,
+      "loss": 5.5157,
+      "step": 21060
+    },
+    {
+      "epoch": 0.02960532671065781,
+      "grad_norm": 1.3492792844772339,
+      "learning_rate": 8.879022059856681e-05,
+      "loss": 5.498,
+      "step": 21070
+    },
+    {
+      "epoch": 0.02961937764882139,
+      "grad_norm": 1.3389153480529785,
+      "learning_rate": 8.883237319095124e-05,
+      "loss": 5.4208,
+      "step": 21080
+    },
+    {
+      "epoch": 0.029633428586984967,
+      "grad_norm": 1.3750202655792236,
+      "learning_rate": 8.887452578333566e-05,
+      "loss": 5.3824,
+      "step": 21090
+    },
+    {
+      "epoch": 0.029647479525148544,
+      "grad_norm": 1.3427931070327759,
+      "learning_rate": 8.89166783757201e-05,
+      "loss": 5.4627,
+      "step": 21100
+    },
+    {
+      "epoch": 0.029661530463312123,
+      "grad_norm": 1.3169935941696167,
+      "learning_rate": 8.895883096810453e-05,
+      "loss": 5.4553,
+      "step": 21110
+    },
+    {
+      "epoch": 0.0296755814014757,
+      "grad_norm": 1.3012065887451172,
+      "learning_rate": 8.900098356048898e-05,
+      "loss": 5.4464,
+      "step": 21120
+    },
+    {
+      "epoch": 0.029689632339639276,
+      "grad_norm": 1.3812952041625977,
+      "learning_rate": 8.90431361528734e-05,
+      "loss": 5.4815,
+      "step": 21130
+    },
+    {
+      "epoch": 0.029703683277802855,
+      "grad_norm": 1.336200475692749,
+      "learning_rate": 8.908528874525782e-05,
+      "loss": 5.3232,
+      "step": 21140
+    },
+    {
+      "epoch": 0.02971773421596643,
+      "grad_norm": 1.2766376733779907,
+      "learning_rate": 8.912744133764227e-05,
+      "loss": 5.4727,
+      "step": 21150
+    },
+    {
+      "epoch": 0.02973178515413001,
+      "grad_norm": 1.391106367111206,
+      "learning_rate": 8.916959393002669e-05,
+      "loss": 5.5038,
+      "step": 21160
+    },
+    {
+      "epoch": 0.029745836092293587,
+      "grad_norm": 1.3515926599502563,
+      "learning_rate": 8.921174652241112e-05,
+      "loss": 5.4136,
+      "step": 21170
+    },
+    {
+      "epoch": 0.029759887030457163,
+      "grad_norm": 1.3355499505996704,
+      "learning_rate": 8.925389911479556e-05,
+      "loss": 5.3543,
+      "step": 21180
+    },
+    {
+      "epoch": 0.029773937968620743,
+      "grad_norm": 1.2321969270706177,
+      "learning_rate": 8.929605170717998e-05,
+      "loss": 5.4438,
+      "step": 21190
+    },
+    {
+      "epoch": 0.02978798890678432,
+      "grad_norm": 1.378713607788086,
+      "learning_rate": 8.933820429956441e-05,
+      "loss": 5.3934,
+      "step": 21200
+    },
+    {
+      "epoch": 0.0298020398449479,
+      "grad_norm": 1.2393652200698853,
+      "learning_rate": 8.938035689194886e-05,
+      "loss": 5.5751,
+      "step": 21210
+    },
+    {
+      "epoch": 0.029816090783111475,
+      "grad_norm": 1.4586329460144043,
+      "learning_rate": 8.942250948433327e-05,
+      "loss": 5.2181,
+      "step": 21220
+    },
+    {
+      "epoch": 0.02983014172127505,
+      "grad_norm": 1.5103076696395874,
+      "learning_rate": 8.94646620767177e-05,
+      "loss": 5.4515,
+      "step": 21230
+    },
+    {
+      "epoch": 0.02984419265943863,
+      "grad_norm": 1.4935081005096436,
+      "learning_rate": 8.950681466910215e-05,
+      "loss": 5.4309,
+      "step": 21240
+    },
+    {
+      "epoch": 0.029858243597602207,
+      "grad_norm": 1.3122179508209229,
+      "learning_rate": 8.954896726148658e-05,
+      "loss": 5.3174,
+      "step": 21250
+    },
+    {
+      "epoch": 0.029872294535765787,
+      "grad_norm": 1.280525803565979,
+      "learning_rate": 8.9591119853871e-05,
+      "loss": 5.3517,
+      "step": 21260
+    },
+    {
+      "epoch": 0.029886345473929363,
+      "grad_norm": 1.3139269351959229,
+      "learning_rate": 8.963327244625544e-05,
+      "loss": 5.4784,
+      "step": 21270
+    },
+    {
+      "epoch": 0.02990039641209294,
+      "grad_norm": 1.2973021268844604,
+      "learning_rate": 8.967542503863987e-05,
+      "loss": 5.3769,
+      "step": 21280
+    },
+    {
+      "epoch": 0.02991444735025652,
+      "grad_norm": 1.4092621803283691,
+      "learning_rate": 8.971757763102429e-05,
+      "loss": 5.3972,
+      "step": 21290
+    },
+    {
+      "epoch": 0.029928498288420095,
+      "grad_norm": 1.364310383796692,
+      "learning_rate": 8.975973022340874e-05,
+      "loss": 5.4728,
+      "step": 21300
+    },
+    {
+      "epoch": 0.02994254922658367,
+      "grad_norm": 1.516514539718628,
+      "learning_rate": 8.980188281579317e-05,
+      "loss": 5.4473,
+      "step": 21310
+    },
+    {
+      "epoch": 0.02995660016474725,
+      "grad_norm": 1.413217306137085,
+      "learning_rate": 8.984403540817759e-05,
+      "loss": 5.5172,
+      "step": 21320
+    },
+    {
+      "epoch": 0.029970651102910827,
+      "grad_norm": 1.724693775177002,
+      "learning_rate": 8.988618800056203e-05,
+      "loss": 5.3085,
+      "step": 21330
+    },
+    {
+      "epoch": 0.029984702041074406,
+      "grad_norm": 1.2294498682022095,
+      "learning_rate": 8.992834059294646e-05,
+      "loss": 5.5309,
+      "step": 21340
+    },
+    {
+      "epoch": 0.029998752979237982,
+      "grad_norm": 1.2753736972808838,
+      "learning_rate": 8.997049318533088e-05,
+      "loss": 5.4041,
+      "step": 21350
+    },
+    {
+      "epoch": 0.03001280391740156,
+      "grad_norm": 1.3634488582611084,
+      "learning_rate": 9.001264577771532e-05,
+      "loss": 5.4316,
+      "step": 21360
+    },
+    {
+      "epoch": 0.030026854855565138,
+      "grad_norm": 1.3544750213623047,
+      "learning_rate": 9.005479837009976e-05,
+      "loss": 5.4193,
+      "step": 21370
+    },
+    {
+      "epoch": 0.030040905793728714,
+      "grad_norm": 1.3579610586166382,
+      "learning_rate": 9.009695096248417e-05,
+      "loss": 5.4289,
+      "step": 21380
+    },
+    {
+      "epoch": 0.030054956731892294,
+      "grad_norm": 1.3251279592514038,
+      "learning_rate": 9.013910355486862e-05,
+      "loss": 5.4037,
+      "step": 21390
+    },
+    {
+      "epoch": 0.03006900767005587,
+      "grad_norm": 1.4140822887420654,
+      "learning_rate": 9.018125614725305e-05,
+      "loss": 5.3694,
+      "step": 21400
+    },
+    {
+      "epoch": 0.030083058608219446,
+      "grad_norm": 1.3545302152633667,
+      "learning_rate": 9.02234087396375e-05,
+      "loss": 5.4191,
+      "step": 21410
+    },
+    {
+      "epoch": 0.030097109546383026,
+      "grad_norm": 1.3583157062530518,
+      "learning_rate": 9.026556133202191e-05,
+      "loss": 5.4124,
+      "step": 21420
+    },
+    {
+      "epoch": 0.030111160484546602,
+      "grad_norm": 1.3296844959259033,
+      "learning_rate": 9.030771392440634e-05,
+      "loss": 5.3846,
+      "step": 21430
+    },
+    {
+      "epoch": 0.030125211422710182,
+      "grad_norm": 1.287698745727539,
+      "learning_rate": 9.034986651679079e-05,
+      "loss": 5.3998,
+      "step": 21440
+    },
+    {
+      "epoch": 0.030139262360873758,
+      "grad_norm": 1.4231420755386353,
+      "learning_rate": 9.03920191091752e-05,
+      "loss": 5.2575,
+      "step": 21450
+    },
+    {
+      "epoch": 0.030153313299037334,
+      "grad_norm": 1.347367286682129,
+      "learning_rate": 9.043417170155964e-05,
+      "loss": 5.4645,
+      "step": 21460
+    },
+    {
+      "epoch": 0.030167364237200914,
+      "grad_norm": 1.2664035558700562,
+      "learning_rate": 9.047632429394408e-05,
+      "loss": 5.4775,
+      "step": 21470
+    },
+    {
+      "epoch": 0.03018141517536449,
+      "grad_norm": 1.3278405666351318,
+      "learning_rate": 9.05184768863285e-05,
+      "loss": 5.5146,
+      "step": 21480
+    },
+    {
+      "epoch": 0.030195466113528066,
+      "grad_norm": 1.3342235088348389,
+      "learning_rate": 9.056062947871293e-05,
+      "loss": 5.438,
+      "step": 21490
+    },
+    {
+      "epoch": 0.030209517051691646,
+      "grad_norm": 1.2799886465072632,
+      "learning_rate": 9.060278207109737e-05,
+      "loss": 5.4206,
+      "step": 21500
+    },
+    {
+      "epoch": 0.030223567989855222,
+      "grad_norm": 1.3722513914108276,
+      "learning_rate": 9.064493466348179e-05,
+      "loss": 5.2671,
+      "step": 21510
+    },
+    {
+      "epoch": 0.0302376189280188,
+      "grad_norm": 1.2341845035552979,
+      "learning_rate": 9.068708725586622e-05,
+      "loss": 5.4003,
+      "step": 21520
+    },
+    {
+      "epoch": 0.030251669866182378,
+      "grad_norm": 1.3256713151931763,
+      "learning_rate": 9.072923984825067e-05,
+      "loss": 5.4125,
+      "step": 21530
+    },
+    {
+      "epoch": 0.030265720804345954,
+      "grad_norm": 1.2737789154052734,
+      "learning_rate": 9.07713924406351e-05,
+      "loss": 5.493,
+      "step": 21540
+    },
+    {
+      "epoch": 0.030279771742509533,
+      "grad_norm": 1.3018547296524048,
+      "learning_rate": 9.081354503301952e-05,
+      "loss": 5.4518,
+      "step": 21550
+    },
+    {
+      "epoch": 0.03029382268067311,
+      "grad_norm": 1.2525168657302856,
+      "learning_rate": 9.085569762540396e-05,
+      "loss": 5.4166,
+      "step": 21560
+    },
+    {
+      "epoch": 0.03030787361883669,
+      "grad_norm": 1.5232125520706177,
+      "learning_rate": 9.089785021778839e-05,
+      "loss": 5.3974,
+      "step": 21570
+    },
+    {
+      "epoch": 0.030321924557000265,
+      "grad_norm": 1.332067608833313,
+      "learning_rate": 9.094000281017281e-05,
+      "loss": 5.2823,
+      "step": 21580
+    },
+    {
+      "epoch": 0.03033597549516384,
+      "grad_norm": 1.3199515342712402,
+      "learning_rate": 9.098215540255725e-05,
+      "loss": 5.5286,
+      "step": 21590
+    },
+    {
+      "epoch": 0.03035002643332742,
+      "grad_norm": 1.3310644626617432,
+      "learning_rate": 9.102430799494169e-05,
+      "loss": 5.3436,
+      "step": 21600
+    },
+    {
+      "epoch": 0.030364077371490997,
+      "grad_norm": 1.285497784614563,
+      "learning_rate": 9.10664605873261e-05,
+      "loss": 5.3442,
+      "step": 21610
+    },
+    {
+      "epoch": 0.030378128309654574,
+      "grad_norm": 1.5494871139526367,
+      "learning_rate": 9.110861317971055e-05,
+      "loss": 5.5381,
+      "step": 21620
+    },
+    {
+      "epoch": 0.030392179247818153,
+      "grad_norm": 1.3371659517288208,
+      "learning_rate": 9.114655051285653e-05,
+      "loss": 5.3604,
+      "step": 21630
+    },
+    {
+      "epoch": 0.03040623018598173,
+      "grad_norm": 1.322951078414917,
+      "learning_rate": 9.118870310524096e-05,
+      "loss": 5.4602,
+      "step": 21640
+    },
+    {
+      "epoch": 0.03042028112414531,
+      "grad_norm": 1.5651347637176514,
+      "learning_rate": 9.123085569762539e-05,
+      "loss": 5.4247,
+      "step": 21650
+    },
+    {
+      "epoch": 0.030434332062308885,
+      "grad_norm": 1.3033775091171265,
+      "learning_rate": 9.127300829000982e-05,
+      "loss": 5.4838,
+      "step": 21660
+    },
+    {
+      "epoch": 0.03044838300047246,
+      "grad_norm": 1.436387300491333,
+      "learning_rate": 9.131516088239427e-05,
+      "loss": 5.2139,
+      "step": 21670
+    },
+    {
+      "epoch": 0.03046243393863604,
+      "grad_norm": 1.3395622968673706,
+      "learning_rate": 9.135731347477869e-05,
+      "loss": 5.4355,
+      "step": 21680
+    },
+    {
+      "epoch": 0.030476484876799617,
+      "grad_norm": 1.457055926322937,
+      "learning_rate": 9.139946606716312e-05,
+      "loss": 5.3278,
+      "step": 21690
+    },
+    {
+      "epoch": 0.030490535814963197,
+      "grad_norm": 1.3071521520614624,
+      "learning_rate": 9.144161865954756e-05,
+      "loss": 5.2767,
+      "step": 21700
+    },
+    {
+      "epoch": 0.030504586753126773,
+      "grad_norm": 1.3810970783233643,
+      "learning_rate": 9.148377125193198e-05,
+      "loss": 5.4068,
+      "step": 21710
+    },
+    {
+      "epoch": 0.03051863769129035,
+      "grad_norm": 1.3929694890975952,
+      "learning_rate": 9.152592384431641e-05,
+      "loss": 5.4296,
+      "step": 21720
+    },
+    {
+      "epoch": 0.03053268862945393,
+      "grad_norm": 1.5036205053329468,
+      "learning_rate": 9.156807643670086e-05,
+      "loss": 5.4989,
+      "step": 21730
+    },
+    {
+      "epoch": 0.030546739567617505,
+      "grad_norm": 1.315721035003662,
+      "learning_rate": 9.161022902908527e-05,
+      "loss": 5.4409,
+      "step": 21740
+    },
+    {
+      "epoch": 0.030560790505781085,
+      "grad_norm": 1.323768138885498,
+      "learning_rate": 9.165238162146972e-05,
+      "loss": 5.4287,
+      "step": 21750
+    },
+    {
+      "epoch": 0.03057484144394466,
+      "grad_norm": 1.445924997329712,
+      "learning_rate": 9.169453421385415e-05,
+      "loss": 5.4339,
+      "step": 21760
+    },
+    {
+      "epoch": 0.030588892382108237,
+      "grad_norm": 1.287317156791687,
+      "learning_rate": 9.173668680623857e-05,
+      "loss": 5.3386,
+      "step": 21770
+    },
+    {
+      "epoch": 0.030602943320271817,
+      "grad_norm": 1.3343262672424316,
+      "learning_rate": 9.177883939862301e-05,
+      "loss": 5.4419,
+      "step": 21780
+    },
+    {
+      "epoch": 0.030616994258435393,
+      "grad_norm": 1.309372901916504,
+      "learning_rate": 9.182099199100744e-05,
+      "loss": 5.3939,
+      "step": 21790
+    },
+    {
+      "epoch": 0.03063104519659897,
+      "grad_norm": 1.3397578001022339,
+      "learning_rate": 9.186314458339186e-05,
+      "loss": 5.4396,
+      "step": 21800
+    },
+    {
+      "epoch": 0.03064509613476255,
+      "grad_norm": 1.3325626850128174,
+      "learning_rate": 9.19052971757763e-05,
+      "loss": 5.3904,
+      "step": 21810
+    },
+    {
+      "epoch": 0.030659147072926125,
+      "grad_norm": 1.2755528688430786,
+      "learning_rate": 9.194744976816074e-05,
+      "loss": 5.352,
+      "step": 21820
+    },
+    {
+      "epoch": 0.030673198011089704,
+      "grad_norm": 1.3710780143737793,
+      "learning_rate": 9.198960236054517e-05,
+      "loss": 5.3976,
+      "step": 21830
+    },
+    {
+      "epoch": 0.03068724894925328,
+      "grad_norm": 1.29194176197052,
+      "learning_rate": 9.20317549529296e-05,
+      "loss": 5.4736,
+      "step": 21840
+    },
+    {
+      "epoch": 0.030701299887416857,
+      "grad_norm": 1.2661114931106567,
+      "learning_rate": 9.207390754531403e-05,
+      "loss": 5.5116,
+      "step": 21850
+    },
+    {
+      "epoch": 0.030715350825580436,
+      "grad_norm": 1.337958574295044,
+      "learning_rate": 9.211606013769846e-05,
+      "loss": 5.4852,
+      "step": 21860
+    },
+    {
+      "epoch": 0.030729401763744012,
+      "grad_norm": 1.3310383558273315,
+      "learning_rate": 9.215821273008289e-05,
+      "loss": 5.2673,
+      "step": 21870
+    },
+    {
+      "epoch": 0.030743452701907592,
+      "grad_norm": 1.2530386447906494,
+      "learning_rate": 9.220036532246732e-05,
+      "loss": 5.4167,
+      "step": 21880
+    },
+    {
+      "epoch": 0.030757503640071168,
+      "grad_norm": 1.273057460784912,
+      "learning_rate": 9.224251791485175e-05,
+      "loss": 5.4284,
+      "step": 21890
+    },
+    {
+      "epoch": 0.030771554578234744,
+      "grad_norm": 1.264639973640442,
+      "learning_rate": 9.228467050723619e-05,
+      "loss": 5.3107,
+      "step": 21900
+    },
+    {
+      "epoch": 0.030785605516398324,
+      "grad_norm": 1.8401588201522827,
+      "learning_rate": 9.232682309962062e-05,
+      "loss": 5.3392,
+      "step": 21910
+    },
+    {
+      "epoch": 0.0307996564545619,
+      "grad_norm": 1.2661328315734863,
+      "learning_rate": 9.236897569200505e-05,
+      "loss": 5.4274,
+      "step": 21920
+    },
+    {
+      "epoch": 0.030813707392725476,
+      "grad_norm": 1.2905445098876953,
+      "learning_rate": 9.241112828438948e-05,
+      "loss": 5.4376,
+      "step": 21930
+    },
+    {
+      "epoch": 0.030827758330889056,
+      "grad_norm": 1.395332932472229,
+      "learning_rate": 9.245328087677391e-05,
+      "loss": 5.3368,
+      "step": 21940
+    },
+    {
+      "epoch": 0.030841809269052632,
+      "grad_norm": 1.3216753005981445,
+      "learning_rate": 9.249543346915834e-05,
+      "loss": 5.4493,
+      "step": 21950
+    },
+    {
+      "epoch": 0.030855860207216212,
+      "grad_norm": 1.528210997581482,
+      "learning_rate": 9.253758606154279e-05,
+      "loss": 5.3529,
+      "step": 21960
+    },
+    {
+      "epoch": 0.030869911145379788,
+      "grad_norm": 1.333594560623169,
+      "learning_rate": 9.25797386539272e-05,
+      "loss": 5.2689,
+      "step": 21970
+    },
+    {
+      "epoch": 0.030883962083543364,
+      "grad_norm": 1.2748163938522339,
+      "learning_rate": 9.262189124631164e-05,
+      "loss": 5.2231,
+      "step": 21980
+    },
+    {
+      "epoch": 0.030898013021706944,
+      "grad_norm": 1.4222743511199951,
+      "learning_rate": 9.266404383869608e-05,
+      "loss": 5.5653,
+      "step": 21990
+    },
+    {
+      "epoch": 0.03091206395987052,
+      "grad_norm": 1.2896723747253418,
+      "learning_rate": 9.27061964310805e-05,
+      "loss": 5.4378,
+      "step": 22000
+    },
+    {
+      "epoch": 0.0309261148980341,
+      "grad_norm": 1.3336145877838135,
+      "learning_rate": 9.274834902346493e-05,
+      "loss": 5.4317,
+      "step": 22010
+    },
+    {
+      "epoch": 0.030940165836197676,
+      "grad_norm": 1.4646224975585938,
+      "learning_rate": 9.279050161584937e-05,
+      "loss": 5.331,
+      "step": 22020
+    },
+    {
+      "epoch": 0.030954216774361252,
+      "grad_norm": 1.4956231117248535,
+      "learning_rate": 9.283265420823379e-05,
+      "loss": 5.4396,
+      "step": 22030
+    },
+    {
+      "epoch": 0.03096826771252483,
+      "grad_norm": 1.3492366075515747,
+      "learning_rate": 9.287480680061824e-05,
+      "loss": 5.3148,
+      "step": 22040
+    },
+    {
+      "epoch": 0.030982318650688408,
+      "grad_norm": 1.420286774635315,
+      "learning_rate": 9.291695939300267e-05,
+      "loss": 5.3644,
+      "step": 22050
+    },
+    {
+      "epoch": 0.030996369588851987,
+      "grad_norm": 1.3561815023422241,
+      "learning_rate": 9.295911198538708e-05,
+      "loss": 5.3762,
+      "step": 22060
+    },
+    {
+      "epoch": 0.031010420527015563,
+      "grad_norm": 1.2883050441741943,
+      "learning_rate": 9.300126457777153e-05,
+      "loss": 5.3012,
+      "step": 22070
+    },
+    {
+      "epoch": 0.03102447146517914,
+      "grad_norm": 1.2738019227981567,
+      "learning_rate": 9.304341717015596e-05,
+      "loss": 5.4794,
+      "step": 22080
+    },
+    {
+      "epoch": 0.03103852240334272,
+      "grad_norm": 1.4682445526123047,
+      "learning_rate": 9.308556976254039e-05,
+      "loss": 5.3829,
+      "step": 22090
+    },
+    {
+      "epoch": 0.031052573341506295,
+      "grad_norm": 1.2694358825683594,
+      "learning_rate": 9.312772235492482e-05,
+      "loss": 5.4007,
+      "step": 22100
+    },
+    {
+      "epoch": 0.03106662427966987,
+      "grad_norm": 1.3557790517807007,
+      "learning_rate": 9.316987494730925e-05,
+      "loss": 5.3658,
+      "step": 22110
+    },
+    {
+      "epoch": 0.03108067521783345,
+      "grad_norm": 1.38999342918396,
+      "learning_rate": 9.321202753969368e-05,
+      "loss": 5.3173,
+      "step": 22120
+    },
+    {
+      "epoch": 0.031094726155997027,
+      "grad_norm": 1.4154685735702515,
+      "learning_rate": 9.325418013207812e-05,
+      "loss": 5.296,
+      "step": 22130
+    },
+    {
+      "epoch": 0.031108777094160607,
+      "grad_norm": 1.3089263439178467,
+      "learning_rate": 9.329633272446255e-05,
+      "loss": 5.3943,
+      "step": 22140
+    },
+    {
+      "epoch": 0.031122828032324183,
+      "grad_norm": 1.258065104484558,
+      "learning_rate": 9.333848531684698e-05,
+      "loss": 5.4311,
+      "step": 22150
+    },
+    {
+      "epoch": 0.03113687897048776,
+      "grad_norm": 1.329082727432251,
+      "learning_rate": 9.338063790923141e-05,
+      "loss": 5.3696,
+      "step": 22160
+    },
+    {
+      "epoch": 0.03115092990865134,
+      "grad_norm": 1.3508650064468384,
+      "learning_rate": 9.342279050161584e-05,
+      "loss": 5.3758,
+      "step": 22170
+    },
+    {
+      "epoch": 0.031164980846814915,
+      "grad_norm": 1.3524917364120483,
+      "learning_rate": 9.346494309400027e-05,
+      "loss": 5.3318,
+      "step": 22180
+    },
+    {
+      "epoch": 0.031179031784978495,
+      "grad_norm": 1.2871861457824707,
+      "learning_rate": 9.35070956863847e-05,
+      "loss": 5.5478,
+      "step": 22190
+    },
+    {
+      "epoch": 0.03119308272314207,
+      "grad_norm": 1.3874597549438477,
+      "learning_rate": 9.354924827876913e-05,
+      "loss": 5.4872,
+      "step": 22200
+    },
+    {
+      "epoch": 0.031207133661305647,
+      "grad_norm": 1.3094698190689087,
+      "learning_rate": 9.359140087115357e-05,
+      "loss": 5.3514,
+      "step": 22210
+    },
+    {
+      "epoch": 0.031221184599469227,
+      "grad_norm": 1.498294711112976,
+      "learning_rate": 9.3633553463538e-05,
+      "loss": 5.2953,
+      "step": 22220
+    },
+    {
+      "epoch": 0.031235235537632803,
+      "grad_norm": 1.7673845291137695,
+      "learning_rate": 9.367570605592243e-05,
+      "loss": 5.3295,
+      "step": 22230
+    },
+    {
+      "epoch": 0.03124928647579638,
+      "grad_norm": 1.4196926355361938,
+      "learning_rate": 9.371785864830686e-05,
+      "loss": 5.3922,
+      "step": 22240
+    },
+    {
+      "epoch": 0.03126333741395996,
+      "grad_norm": 1.6076236963272095,
+      "learning_rate": 9.37600112406913e-05,
+      "loss": 5.425,
+      "step": 22250
+    },
+    {
+      "epoch": 0.031277388352123535,
+      "grad_norm": 1.2859355211257935,
+      "learning_rate": 9.380216383307572e-05,
+      "loss": 5.4363,
+      "step": 22260
+    },
+    {
+      "epoch": 0.03129143929028711,
+      "grad_norm": 1.3375812768936157,
+      "learning_rate": 9.384431642546015e-05,
+      "loss": 5.5041,
+      "step": 22270
+    },
+    {
+      "epoch": 0.031305490228450694,
+      "grad_norm": 1.2610328197479248,
+      "learning_rate": 9.38864690178446e-05,
+      "loss": 5.3793,
+      "step": 22280
+    },
+    {
+      "epoch": 0.03131954116661427,
+      "grad_norm": 1.3440093994140625,
+      "learning_rate": 9.392862161022901e-05,
+      "loss": 5.2878,
+      "step": 22290
+    },
+    {
+      "epoch": 0.031333592104777847,
+      "grad_norm": 1.3581554889678955,
+      "learning_rate": 9.397077420261345e-05,
+      "loss": 5.4093,
+      "step": 22300
+    },
+    {
+      "epoch": 0.03134764304294142,
+      "grad_norm": 1.3433340787887573,
+      "learning_rate": 9.401292679499789e-05,
+      "loss": 5.391,
+      "step": 22310
+    },
+    {
+      "epoch": 0.031361693981105,
+      "grad_norm": 1.3540128469467163,
+      "learning_rate": 9.405507938738231e-05,
+      "loss": 5.3798,
+      "step": 22320
+    },
+    {
+      "epoch": 0.031375744919268575,
+      "grad_norm": 1.2961390018463135,
+      "learning_rate": 9.409723197976674e-05,
+      "loss": 5.316,
+      "step": 22330
+    },
+    {
+      "epoch": 0.03138979585743216,
+      "grad_norm": 1.2970017194747925,
+      "learning_rate": 9.413938457215118e-05,
+      "loss": 5.3238,
+      "step": 22340
+    },
+    {
+      "epoch": 0.031403846795595734,
+      "grad_norm": 1.275976300239563,
+      "learning_rate": 9.41815371645356e-05,
+      "loss": 5.3903,
+      "step": 22350
+    },
+    {
+      "epoch": 0.03141789773375931,
+      "grad_norm": 1.329797625541687,
+      "learning_rate": 9.422368975692005e-05,
+      "loss": 5.3756,
+      "step": 22360
+    },
+    {
+      "epoch": 0.03143194867192289,
+      "grad_norm": 1.3486462831497192,
+      "learning_rate": 9.426584234930448e-05,
+      "loss": 5.4281,
+      "step": 22370
+    },
+    {
+      "epoch": 0.03144599961008646,
+      "grad_norm": 1.3495275974273682,
+      "learning_rate": 9.430799494168891e-05,
+      "loss": 5.3231,
+      "step": 22380
+    },
+    {
+      "epoch": 0.031460050548250046,
+      "grad_norm": 1.2206830978393555,
+      "learning_rate": 9.435014753407334e-05,
+      "loss": 5.3296,
+      "step": 22390
+    },
+    {
+      "epoch": 0.03147410148641362,
+      "grad_norm": 1.2643572092056274,
+      "learning_rate": 9.439230012645777e-05,
+      "loss": 5.3465,
+      "step": 22400
+    },
+    {
+      "epoch": 0.0314881524245772,
+      "grad_norm": 1.3389561176300049,
+      "learning_rate": 9.44344527188422e-05,
+      "loss": 5.3205,
+      "step": 22410
+    },
+    {
+      "epoch": 0.031502203362740774,
+      "grad_norm": 1.2590327262878418,
+      "learning_rate": 9.447660531122663e-05,
+      "loss": 5.3742,
+      "step": 22420
+    },
+    {
+      "epoch": 0.03151625430090435,
+      "grad_norm": 1.2709726095199585,
+      "learning_rate": 9.451875790361106e-05,
+      "loss": 5.4671,
+      "step": 22430
+    },
+    {
+      "epoch": 0.031530305239067934,
+      "grad_norm": 1.3207013607025146,
+      "learning_rate": 9.45609104959955e-05,
+      "loss": 5.2865,
+      "step": 22440
+    },
+    {
+      "epoch": 0.03154435617723151,
+      "grad_norm": 1.3502670526504517,
+      "learning_rate": 9.460306308837993e-05,
+      "loss": 5.2844,
+      "step": 22450
+    },
+    {
+      "epoch": 0.031558407115395086,
+      "grad_norm": 1.3033579587936401,
+      "learning_rate": 9.464521568076436e-05,
+      "loss": 5.5248,
+      "step": 22460
+    },
+    {
+      "epoch": 0.03157245805355866,
+      "grad_norm": 1.2336679697036743,
+      "learning_rate": 9.468736827314879e-05,
+      "loss": 5.4307,
+      "step": 22470
+    },
+    {
+      "epoch": 0.03158650899172224,
+      "grad_norm": 1.4275672435760498,
+      "learning_rate": 9.472952086553322e-05,
+      "loss": 5.4736,
+      "step": 22480
+    },
+    {
+      "epoch": 0.03160055992988582,
+      "grad_norm": 1.3759491443634033,
+      "learning_rate": 9.477167345791765e-05,
+      "loss": 5.38,
+      "step": 22490
+    },
+    {
+      "epoch": 0.0316146108680494,
+      "grad_norm": 1.3446354866027832,
+      "learning_rate": 9.481382605030208e-05,
+      "loss": 5.4072,
+      "step": 22500
+    },
+    {
+      "epoch": 0.031628661806212974,
+      "grad_norm": 1.3388227224349976,
+      "learning_rate": 9.485597864268651e-05,
+      "loss": 5.3462,
+      "step": 22510
+    },
+    {
+      "epoch": 0.03164271274437655,
+      "grad_norm": 1.3788104057312012,
+      "learning_rate": 9.489813123507095e-05,
+      "loss": 5.4326,
+      "step": 22520
+    },
+    {
+      "epoch": 0.031656763682540126,
+      "grad_norm": 1.424253225326538,
+      "learning_rate": 9.494028382745538e-05,
+      "loss": 5.4119,
+      "step": 22530
+    },
+    {
+      "epoch": 0.03167081462070371,
+      "grad_norm": 1.2999223470687866,
+      "learning_rate": 9.498243641983982e-05,
+      "loss": 5.3624,
+      "step": 22540
+    },
+    {
+      "epoch": 0.031684865558867285,
+      "grad_norm": 1.368601679801941,
+      "learning_rate": 9.502458901222424e-05,
+      "loss": 5.2332,
+      "step": 22550
+    },
+    {
+      "epoch": 0.03169891649703086,
+      "grad_norm": 1.4578156471252441,
+      "learning_rate": 9.506674160460867e-05,
+      "loss": 5.2938,
+      "step": 22560
+    },
+    {
+      "epoch": 0.03171296743519444,
+      "grad_norm": 1.274147391319275,
+      "learning_rate": 9.510889419699311e-05,
+      "loss": 5.3459,
+      "step": 22570
+    },
+    {
+      "epoch": 0.031727018373358014,
+      "grad_norm": 1.3094562292099,
+      "learning_rate": 9.515104678937753e-05,
+      "loss": 5.3795,
+      "step": 22580
+    },
+    {
+      "epoch": 0.0317410693115216,
+      "grad_norm": 1.2858974933624268,
+      "learning_rate": 9.519319938176196e-05,
+      "loss": 5.4598,
+      "step": 22590
+    },
+    {
+      "epoch": 0.03175512024968517,
+      "grad_norm": 1.267670750617981,
+      "learning_rate": 9.523535197414641e-05,
+      "loss": 5.3513,
+      "step": 22600
+    },
+    {
+      "epoch": 0.03176917118784875,
+      "grad_norm": 1.3018923997879028,
+      "learning_rate": 9.527750456653083e-05,
+      "loss": 5.3053,
+      "step": 22610
+    },
+    {
+      "epoch": 0.031783222126012325,
+      "grad_norm": 1.5751721858978271,
+      "learning_rate": 9.531965715891526e-05,
+      "loss": 5.5364,
+      "step": 22620
+    },
+    {
+      "epoch": 0.0317972730641759,
+      "grad_norm": 1.3210186958312988,
+      "learning_rate": 9.53618097512997e-05,
+      "loss": 5.3663,
+      "step": 22630
+    },
+    {
+      "epoch": 0.03181132400233948,
+      "grad_norm": 1.3259351253509521,
+      "learning_rate": 9.540396234368412e-05,
+      "loss": 5.4101,
+      "step": 22640
+    },
+    {
+      "epoch": 0.03182537494050306,
+      "grad_norm": 1.2493022680282593,
+      "learning_rate": 9.544611493606856e-05,
+      "loss": 5.409,
+      "step": 22650
+    },
+    {
+      "epoch": 0.03183942587866664,
+      "grad_norm": 1.3749110698699951,
+      "learning_rate": 9.5488267528453e-05,
+      "loss": 5.4479,
+      "step": 22660
+    },
+    {
+      "epoch": 0.03185347681683021,
+      "grad_norm": 1.2747150659561157,
+      "learning_rate": 9.553042012083743e-05,
+      "loss": 5.3817,
+      "step": 22670
+    },
+    {
+      "epoch": 0.03186752775499379,
+      "grad_norm": 1.2162046432495117,
+      "learning_rate": 9.557257271322186e-05,
+      "loss": 5.3368,
+      "step": 22680
+    },
+    {
+      "epoch": 0.031881578693157366,
+      "grad_norm": 1.301879644393921,
+      "learning_rate": 9.561472530560629e-05,
+      "loss": 5.4536,
+      "step": 22690
+    },
+    {
+      "epoch": 0.03189562963132095,
+      "grad_norm": 1.389109492301941,
+      "learning_rate": 9.565687789799072e-05,
+      "loss": 5.2986,
+      "step": 22700
+    },
+    {
+      "epoch": 0.031909680569484525,
+      "grad_norm": 1.2711687088012695,
+      "learning_rate": 9.569903049037515e-05,
+      "loss": 5.488,
+      "step": 22710
+    },
+    {
+      "epoch": 0.0319237315076481,
+      "grad_norm": 1.2965320348739624,
+      "learning_rate": 9.573696782352113e-05,
+      "loss": 5.2603,
+      "step": 22720
+    },
+    {
+      "epoch": 0.03193778244581168,
+      "grad_norm": 1.325437307357788,
+      "learning_rate": 9.577912041590558e-05,
+      "loss": 5.4919,
+      "step": 22730
+    },
+    {
+      "epoch": 0.03195183338397525,
+      "grad_norm": 1.2769930362701416,
+      "learning_rate": 9.582127300829e-05,
+      "loss": 5.329,
+      "step": 22740
+    },
+    {
+      "epoch": 0.031965884322138836,
+      "grad_norm": 1.3971630334854126,
+      "learning_rate": 9.586342560067443e-05,
+      "loss": 5.3242,
+      "step": 22750
+    },
+    {
+      "epoch": 0.03197993526030241,
+      "grad_norm": 1.2280455827713013,
+      "learning_rate": 9.590557819305887e-05,
+      "loss": 5.342,
+      "step": 22760
+    },
+    {
+      "epoch": 0.03199398619846599,
+      "grad_norm": 1.3401799201965332,
+      "learning_rate": 9.594773078544329e-05,
+      "loss": 5.4924,
+      "step": 22770
+    },
+    {
+      "epoch": 0.032008037136629565,
+      "grad_norm": 1.4017051458358765,
+      "learning_rate": 9.598988337782772e-05,
+      "loss": 5.277,
+      "step": 22780
+    },
+    {
+      "epoch": 0.03202208807479314,
+      "grad_norm": 1.3228023052215576,
+      "learning_rate": 9.603203597021217e-05,
+      "loss": 5.3887,
+      "step": 22790
+    },
+    {
+      "epoch": 0.032036139012956724,
+      "grad_norm": 1.2626895904541016,
+      "learning_rate": 9.60741885625966e-05,
+      "loss": 5.4167,
+      "step": 22800
+    },
+    {
+      "epoch": 0.0320501899511203,
+      "grad_norm": 1.3468395471572876,
+      "learning_rate": 9.611634115498101e-05,
+      "loss": 5.3809,
+      "step": 22810
+    },
+    {
+      "epoch": 0.032064240889283877,
+      "grad_norm": 1.2649286985397339,
+      "learning_rate": 9.615849374736546e-05,
+      "loss": 5.4244,
+      "step": 22820
+    },
+    {
+      "epoch": 0.03207829182744745,
+      "grad_norm": 1.2764794826507568,
+      "learning_rate": 9.620064633974989e-05,
+      "loss": 5.3179,
+      "step": 22830
+    },
+    {
+      "epoch": 0.03209234276561103,
+      "grad_norm": 1.2899025678634644,
+      "learning_rate": 9.624279893213431e-05,
+      "loss": 5.3826,
+      "step": 22840
+    },
+    {
+      "epoch": 0.03210639370377461,
+      "grad_norm": 1.3156379461288452,
+      "learning_rate": 9.628495152451875e-05,
+      "loss": 5.4528,
+      "step": 22850
+    },
+    {
+      "epoch": 0.03212044464193819,
+      "grad_norm": 1.3568313121795654,
+      "learning_rate": 9.632710411690318e-05,
+      "loss": 5.3015,
+      "step": 22860
+    },
+    {
+      "epoch": 0.032134495580101764,
+      "grad_norm": 1.3608930110931396,
+      "learning_rate": 9.63692567092876e-05,
+      "loss": 5.3645,
+      "step": 22870
+    },
+    {
+      "epoch": 0.03214854651826534,
+      "grad_norm": 1.3903223276138306,
+      "learning_rate": 9.641140930167205e-05,
+      "loss": 5.2674,
+      "step": 22880
+    },
+    {
+      "epoch": 0.03216259745642892,
+      "grad_norm": 1.372069001197815,
+      "learning_rate": 9.645356189405648e-05,
+      "loss": 5.3272,
+      "step": 22890
+    },
+    {
+      "epoch": 0.0321766483945925,
+      "grad_norm": 1.373261570930481,
+      "learning_rate": 9.64957144864409e-05,
+      "loss": 5.2367,
+      "step": 22900
+    },
+    {
+      "epoch": 0.032190699332756076,
+      "grad_norm": 1.2778481245040894,
+      "learning_rate": 9.653786707882534e-05,
+      "loss": 5.4686,
+      "step": 22910
+    },
+    {
+      "epoch": 0.03220475027091965,
+      "grad_norm": 1.4180645942687988,
+      "learning_rate": 9.658001967120977e-05,
+      "loss": 5.3619,
+      "step": 22920
+    },
+    {
+      "epoch": 0.03221880120908323,
+      "grad_norm": 1.3312855958938599,
+      "learning_rate": 9.662217226359422e-05,
+      "loss": 5.2804,
+      "step": 22930
+    },
+    {
+      "epoch": 0.032232852147246804,
+      "grad_norm": 1.3726269006729126,
+      "learning_rate": 9.666432485597863e-05,
+      "loss": 5.428,
+      "step": 22940
+    },
+    {
+      "epoch": 0.03224690308541038,
+      "grad_norm": 1.2660539150238037,
+      "learning_rate": 9.670647744836306e-05,
+      "loss": 5.3517,
+      "step": 22950
+    },
+    {
+      "epoch": 0.032260954023573964,
+      "grad_norm": 1.3094338178634644,
+      "learning_rate": 9.674863004074751e-05,
+      "loss": 5.3259,
+      "step": 22960
+    },
+    {
+      "epoch": 0.03227500496173754,
+      "grad_norm": 1.348771572113037,
+      "learning_rate": 9.679078263313193e-05,
+      "loss": 5.3373,
+      "step": 22970
+    },
+    {
+      "epoch": 0.032289055899901116,
+      "grad_norm": 1.3427238464355469,
+      "learning_rate": 9.683293522551636e-05,
+      "loss": 5.2462,
+      "step": 22980
+    },
+    {
+      "epoch": 0.03230310683806469,
+      "grad_norm": 1.2989392280578613,
+      "learning_rate": 9.68750878179008e-05,
+      "loss": 5.4482,
+      "step": 22990
+    },
+    {
+      "epoch": 0.03231715777622827,
+      "grad_norm": 1.3329780101776123,
+      "learning_rate": 9.691724041028522e-05,
+      "loss": 5.2809,
+      "step": 23000
+    },
+    {
+      "epoch": 0.03233120871439185,
+      "grad_norm": 1.3921016454696655,
+      "learning_rate": 9.695939300266965e-05,
+      "loss": 5.5561,
+      "step": 23010
+    },
+    {
+      "epoch": 0.03234525965255543,
+      "grad_norm": 1.3452013731002808,
+      "learning_rate": 9.70015455950541e-05,
+      "loss": 5.3734,
+      "step": 23020
+    },
+    {
+      "epoch": 0.032359310590719004,
+      "grad_norm": 1.2437107563018799,
+      "learning_rate": 9.704369818743851e-05,
+      "loss": 5.4029,
+      "step": 23030
+    },
+    {
+      "epoch": 0.03237336152888258,
+      "grad_norm": 1.2906579971313477,
+      "learning_rate": 9.708585077982294e-05,
+      "loss": 5.1958,
+      "step": 23040
+    },
+    {
+      "epoch": 0.032387412467046156,
+      "grad_norm": 1.290738821029663,
+      "learning_rate": 9.712800337220739e-05,
+      "loss": 5.3823,
+      "step": 23050
+    },
+    {
+      "epoch": 0.03240146340520974,
+      "grad_norm": 1.3592157363891602,
+      "learning_rate": 9.71701559645918e-05,
+      "loss": 5.3865,
+      "step": 23060
+    },
+    {
+      "epoch": 0.032415514343373315,
+      "grad_norm": 1.4341562986373901,
+      "learning_rate": 9.721230855697624e-05,
+      "loss": 5.1396,
+      "step": 23070
+    },
+    {
+      "epoch": 0.03242956528153689,
+      "grad_norm": 1.7886507511138916,
+      "learning_rate": 9.725446114936068e-05,
+      "loss": 5.2892,
+      "step": 23080
+    },
+    {
+      "epoch": 0.03244361621970047,
+      "grad_norm": 1.2833921909332275,
+      "learning_rate": 9.729661374174511e-05,
+      "loss": 5.408,
+      "step": 23090
+    },
+    {
+      "epoch": 0.032457667157864044,
+      "grad_norm": 1.2953739166259766,
+      "learning_rate": 9.733876633412953e-05,
+      "loss": 5.2775,
+      "step": 23100
+    },
+    {
+      "epoch": 0.03247171809602763,
+      "grad_norm": 1.264936923980713,
+      "learning_rate": 9.738091892651398e-05,
+      "loss": 5.3672,
+      "step": 23110
+    },
+    {
+      "epoch": 0.0324857690341912,
+      "grad_norm": 1.324508786201477,
+      "learning_rate": 9.742307151889841e-05,
+      "loss": 5.2859,
+      "step": 23120
+    },
+    {
+      "epoch": 0.03249981997235478,
+      "grad_norm": 1.2716864347457886,
+      "learning_rate": 9.746522411128282e-05,
+      "loss": 5.2861,
+      "step": 23130
+    },
+    {
+      "epoch": 0.032513870910518355,
+      "grad_norm": 1.2742117643356323,
+      "learning_rate": 9.750737670366727e-05,
+      "loss": 5.3845,
+      "step": 23140
+    },
+    {
+      "epoch": 0.03252792184868193,
+      "grad_norm": 1.3381208181381226,
+      "learning_rate": 9.75495292960517e-05,
+      "loss": 5.3024,
+      "step": 23150
+    },
+    {
+      "epoch": 0.032541972786845515,
+      "grad_norm": 1.4787967205047607,
+      "learning_rate": 9.759168188843612e-05,
+      "loss": 5.3082,
+      "step": 23160
+    },
+    {
+      "epoch": 0.03255602372500909,
+      "grad_norm": 1.301733136177063,
+      "learning_rate": 9.763383448082056e-05,
+      "loss": 5.4468,
+      "step": 23170
+    },
+    {
+      "epoch": 0.03257007466317267,
+      "grad_norm": 1.271909236907959,
+      "learning_rate": 9.7675987073205e-05,
+      "loss": 5.2869,
+      "step": 23180
+    },
+    {
+      "epoch": 0.03258412560133624,
+      "grad_norm": 1.2851873636245728,
+      "learning_rate": 9.771813966558941e-05,
+      "loss": 5.2482,
+      "step": 23190
+    },
+    {
+      "epoch": 0.03259817653949982,
+      "grad_norm": 1.307614803314209,
+      "learning_rate": 9.776029225797386e-05,
+      "loss": 5.36,
+      "step": 23200
+    },
+    {
+      "epoch": 0.0326122274776634,
+      "grad_norm": 1.25319504737854,
+      "learning_rate": 9.780244485035829e-05,
+      "loss": 5.409,
+      "step": 23210
+    },
+    {
+      "epoch": 0.03262627841582698,
+      "grad_norm": 1.4005918502807617,
+      "learning_rate": 9.784459744274273e-05,
+      "loss": 5.3335,
+      "step": 23220
+    },
+    {
+      "epoch": 0.032640329353990555,
+      "grad_norm": 1.2450411319732666,
+      "learning_rate": 9.788675003512715e-05,
+      "loss": 5.3397,
+      "step": 23230
+    },
+    {
+      "epoch": 0.03265438029215413,
+      "grad_norm": 1.2542985677719116,
+      "learning_rate": 9.792890262751158e-05,
+      "loss": 5.4002,
+      "step": 23240
+    },
+    {
+      "epoch": 0.03266843123031771,
+      "grad_norm": 1.2760860919952393,
+      "learning_rate": 9.797105521989603e-05,
+      "loss": 5.4205,
+      "step": 23250
+    },
+    {
+      "epoch": 0.03268248216848129,
+      "grad_norm": 1.3349649906158447,
+      "learning_rate": 9.801320781228044e-05,
+      "loss": 5.3124,
+      "step": 23260
+    },
+    {
+      "epoch": 0.032696533106644866,
+      "grad_norm": 1.2851274013519287,
+      "learning_rate": 9.805536040466487e-05,
+      "loss": 5.2979,
+      "step": 23270
+    },
+    {
+      "epoch": 0.03271058404480844,
+      "grad_norm": 1.28581702709198,
+      "learning_rate": 9.809751299704932e-05,
+      "loss": 5.3434,
+      "step": 23280
+    },
+    {
+      "epoch": 0.03272463498297202,
+      "grad_norm": 1.2845999002456665,
+      "learning_rate": 9.813966558943374e-05,
+      "loss": 5.3271,
+      "step": 23290
+    },
+    {
+      "epoch": 0.032738685921135595,
+      "grad_norm": 1.4002037048339844,
+      "learning_rate": 9.818181818181817e-05,
+      "loss": 5.3637,
+      "step": 23300
+    },
+    {
+      "epoch": 0.03275273685929917,
+      "grad_norm": 1.2308608293533325,
+      "learning_rate": 9.822397077420261e-05,
+      "loss": 5.3833,
+      "step": 23310
+    },
+    {
+      "epoch": 0.032766787797462754,
+      "grad_norm": 1.3694802522659302,
+      "learning_rate": 9.826612336658703e-05,
+      "loss": 5.386,
+      "step": 23320
+    },
+    {
+      "epoch": 0.03278083873562633,
+      "grad_norm": 1.294423222541809,
+      "learning_rate": 9.830827595897146e-05,
+      "loss": 5.2552,
+      "step": 23330
+    },
+    {
+      "epoch": 0.032794889673789906,
+      "grad_norm": 1.3576372861862183,
+      "learning_rate": 9.83504285513559e-05,
+      "loss": 5.3575,
+      "step": 23340
+    },
+    {
+      "epoch": 0.03280894061195348,
+      "grad_norm": 1.276759386062622,
+      "learning_rate": 9.839258114374032e-05,
+      "loss": 5.4282,
+      "step": 23350
+    },
+    {
+      "epoch": 0.03282299155011706,
+      "grad_norm": 1.290276288986206,
+      "learning_rate": 9.843473373612476e-05,
+      "loss": 5.2632,
+      "step": 23360
+    },
+    {
+      "epoch": 0.03283704248828064,
+      "grad_norm": 1.3245643377304077,
+      "learning_rate": 9.84768863285092e-05,
+      "loss": 5.3792,
+      "step": 23370
+    },
+    {
+      "epoch": 0.03285109342644422,
+      "grad_norm": 1.3463149070739746,
+      "learning_rate": 9.851903892089363e-05,
+      "loss": 5.3489,
+      "step": 23380
+    },
+    {
+      "epoch": 0.032865144364607794,
+      "grad_norm": 1.3925325870513916,
+      "learning_rate": 9.856119151327805e-05,
+      "loss": 5.2327,
+      "step": 23390
+    },
+    {
+      "epoch": 0.03287919530277137,
+      "grad_norm": 1.3809847831726074,
+      "learning_rate": 9.86033441056625e-05,
+      "loss": 5.44,
+      "step": 23400
+    },
+    {
+      "epoch": 0.03289324624093495,
+      "grad_norm": 1.2953121662139893,
+      "learning_rate": 9.864549669804692e-05,
+      "loss": 5.2537,
+      "step": 23410
+    },
+    {
+      "epoch": 0.03290729717909853,
+      "grad_norm": 1.2769551277160645,
+      "learning_rate": 9.868764929043134e-05,
+      "loss": 5.4532,
+      "step": 23420
+    },
+    {
+      "epoch": 0.032921348117262106,
+      "grad_norm": 1.3292839527130127,
+      "learning_rate": 9.872980188281579e-05,
+      "loss": 5.3824,
+      "step": 23430
+    },
+    {
+      "epoch": 0.03293539905542568,
+      "grad_norm": 1.2691352367401123,
+      "learning_rate": 9.877195447520022e-05,
+      "loss": 5.3559,
+      "step": 23440
+    },
+    {
+      "epoch": 0.03294944999358926,
+      "grad_norm": 1.4160397052764893,
+      "learning_rate": 9.881410706758464e-05,
+      "loss": 5.2673,
+      "step": 23450
+    },
+    {
+      "epoch": 0.032963500931752834,
+      "grad_norm": 1.3387774229049683,
+      "learning_rate": 9.885625965996908e-05,
+      "loss": 5.3447,
+      "step": 23460
+    },
+    {
+      "epoch": 0.03297755186991642,
+      "grad_norm": 1.3649985790252686,
+      "learning_rate": 9.889841225235351e-05,
+      "loss": 5.3222,
+      "step": 23470
+    },
+    {
+      "epoch": 0.032991602808079994,
+      "grad_norm": 1.2664110660552979,
+      "learning_rate": 9.894056484473793e-05,
+      "loss": 5.34,
+      "step": 23480
+    },
+    {
+      "epoch": 0.03300565374624357,
+      "grad_norm": 1.3583790063858032,
+      "learning_rate": 9.898271743712237e-05,
+      "loss": 5.2413,
+      "step": 23490
+    },
+    {
+      "epoch": 0.033019704684407146,
+      "grad_norm": 1.3465863466262817,
+      "learning_rate": 9.90248700295068e-05,
+      "loss": 5.245,
+      "step": 23500
+    },
+    {
+      "epoch": 0.03303375562257072,
+      "grad_norm": 1.3353192806243896,
+      "learning_rate": 9.906702262189125e-05,
+      "loss": 5.3347,
+      "step": 23510
+    },
+    {
+      "epoch": 0.033047806560734305,
+      "grad_norm": 1.3138058185577393,
+      "learning_rate": 9.910917521427567e-05,
+      "loss": 5.2903,
+      "step": 23520
+    },
+    {
+      "epoch": 0.03306185749889788,
+      "grad_norm": 1.308861494064331,
+      "learning_rate": 9.91513278066601e-05,
+      "loss": 5.2704,
+      "step": 23530
+    },
+    {
+      "epoch": 0.03307590843706146,
+      "grad_norm": 1.2975902557373047,
+      "learning_rate": 9.919348039904454e-05,
+      "loss": 5.3236,
+      "step": 23540
+    },
+    {
+      "epoch": 0.033089959375225034,
+      "grad_norm": 1.2685211896896362,
+      "learning_rate": 9.923563299142896e-05,
+      "loss": 5.3083,
+      "step": 23550
+    },
+    {
+      "epoch": 0.03310401031338861,
+      "grad_norm": 1.3065454959869385,
+      "learning_rate": 9.927778558381339e-05,
+      "loss": 5.3769,
+      "step": 23560
+    },
+    {
+      "epoch": 0.03311806125155219,
+      "grad_norm": 1.3842145204544067,
+      "learning_rate": 9.931993817619784e-05,
+      "loss": 5.3922,
+      "step": 23570
+    },
+    {
+      "epoch": 0.03313211218971577,
+      "grad_norm": 1.272464632987976,
+      "learning_rate": 9.936209076858225e-05,
+      "loss": 5.2546,
+      "step": 23580
+    },
+    {
+      "epoch": 0.033146163127879345,
+      "grad_norm": 1.3734886646270752,
+      "learning_rate": 9.940424336096669e-05,
+      "loss": 5.4634,
+      "step": 23590
+    },
+    {
+      "epoch": 0.03316021406604292,
+      "grad_norm": 1.3234834671020508,
+      "learning_rate": 9.944639595335113e-05,
+      "loss": 5.2879,
+      "step": 23600
+    },
+    {
+      "epoch": 0.0331742650042065,
+      "grad_norm": 1.3114765882492065,
+      "learning_rate": 9.948854854573555e-05,
+      "loss": 5.2448,
+      "step": 23610
+    },
+    {
+      "epoch": 0.033188315942370074,
+      "grad_norm": 1.239916443824768,
+      "learning_rate": 9.953070113811998e-05,
+      "loss": 5.1157,
+      "step": 23620
+    },
+    {
+      "epoch": 0.03320236688053366,
+      "grad_norm": 1.3152765035629272,
+      "learning_rate": 9.957285373050442e-05,
+      "loss": 5.2116,
+      "step": 23630
+    },
+    {
+      "epoch": 0.03321641781869723,
+      "grad_norm": 1.3076961040496826,
+      "learning_rate": 9.961500632288884e-05,
+      "loss": 5.3021,
+      "step": 23640
+    },
+    {
+      "epoch": 0.03323046875686081,
+      "grad_norm": 1.3413221836090088,
+      "learning_rate": 9.965715891527327e-05,
+      "loss": 5.3721,
+      "step": 23650
+    },
+    {
+      "epoch": 0.033244519695024385,
+      "grad_norm": 1.2938029766082764,
+      "learning_rate": 9.969931150765772e-05,
+      "loss": 5.313,
+      "step": 23660
+    },
+    {
+      "epoch": 0.03325857063318796,
+      "grad_norm": 1.2781319618225098,
+      "learning_rate": 9.974146410004215e-05,
+      "loss": 5.2119,
+      "step": 23670
+    },
+    {
+      "epoch": 0.033272621571351545,
+      "grad_norm": 1.3234959840774536,
+      "learning_rate": 9.978361669242657e-05,
+      "loss": 5.4493,
+      "step": 23680
+    },
+    {
+      "epoch": 0.03328667250951512,
+      "grad_norm": 1.4533023834228516,
+      "learning_rate": 9.982576928481101e-05,
+      "loss": 5.317,
+      "step": 23690
+    },
+    {
+      "epoch": 0.0333007234476787,
+      "grad_norm": 1.276374101638794,
+      "learning_rate": 9.986792187719544e-05,
+      "loss": 5.3605,
+      "step": 23700
+    },
+    {
+      "epoch": 0.03331477438584227,
+      "grad_norm": 1.292145013809204,
+      "learning_rate": 9.991007446957986e-05,
+      "loss": 5.3252,
+      "step": 23710
+    },
+    {
+      "epoch": 0.03332882532400585,
+      "grad_norm": 1.2901628017425537,
+      "learning_rate": 9.99522270619643e-05,
+      "loss": 5.2686,
+      "step": 23720
+    },
+    {
+      "epoch": 0.03334287626216943,
+      "grad_norm": 1.3100521564483643,
+      "learning_rate": 9.999437965434874e-05,
+      "loss": 5.1247,
+      "step": 23730
+    },
+    {
+      "epoch": 0.03335692720033301,
+      "grad_norm": 1.344598650932312,
+      "learning_rate": 0.00010003653224673315,
+      "loss": 5.2823,
+      "step": 23740
+    },
+    {
+      "epoch": 0.033370978138496585,
+      "grad_norm": 1.4314061403274536,
+      "learning_rate": 0.0001000786848391176,
+      "loss": 5.2714,
+      "step": 23750
+    },
+    {
+      "epoch": 0.03338502907666016,
+      "grad_norm": 1.2748030424118042,
+      "learning_rate": 0.00010012083743150203,
+      "loss": 5.3582,
+      "step": 23760
+    },
+    {
+      "epoch": 0.03339908001482374,
+      "grad_norm": 1.3783010244369507,
+      "learning_rate": 0.00010016299002388645,
+      "loss": 5.1891,
+      "step": 23770
+    },
+    {
+      "epoch": 0.03341313095298732,
+      "grad_norm": 1.3372441530227661,
+      "learning_rate": 0.00010020514261627089,
+      "loss": 5.3044,
+      "step": 23780
+    },
+    {
+      "epoch": 0.033427181891150896,
+      "grad_norm": 1.2305197715759277,
+      "learning_rate": 0.00010024729520865532,
+      "loss": 5.3672,
+      "step": 23790
+    },
+    {
+      "epoch": 0.03344123282931447,
+      "grad_norm": 1.28226900100708,
+      "learning_rate": 0.00010028944780103977,
+      "loss": 5.3432,
+      "step": 23800
+    },
+    {
+      "epoch": 0.03345528376747805,
+      "grad_norm": 1.2784920930862427,
+      "learning_rate": 0.00010033160039342418,
+      "loss": 5.2558,
+      "step": 23810
+    },
+    {
+      "epoch": 0.033469334705641625,
+      "grad_norm": 1.298294186592102,
+      "learning_rate": 0.00010037375298580862,
+      "loss": 5.296,
+      "step": 23820
+    },
+    {
+      "epoch": 0.03348338564380521,
+      "grad_norm": 1.3087866306304932,
+      "learning_rate": 0.00010041590557819306,
+      "loss": 5.2793,
+      "step": 23830
+    },
+    {
+      "epoch": 0.033497436581968784,
+      "grad_norm": 1.3522417545318604,
+      "learning_rate": 0.00010045805817057748,
+      "loss": 5.369,
+      "step": 23840
+    },
+    {
+      "epoch": 0.03351148752013236,
+      "grad_norm": 1.2974127531051636,
+      "learning_rate": 0.00010050021076296191,
+      "loss": 5.3775,
+      "step": 23850
+    },
+    {
+      "epoch": 0.033525538458295936,
+      "grad_norm": 1.3162837028503418,
+      "learning_rate": 0.00010054236335534635,
+      "loss": 5.3664,
+      "step": 23860
+    },
+    {
+      "epoch": 0.03353958939645951,
+      "grad_norm": 1.2961353063583374,
+      "learning_rate": 0.00010058451594773077,
+      "loss": 5.3546,
+      "step": 23870
+    },
+    {
+      "epoch": 0.033553640334623096,
+      "grad_norm": 1.3011044263839722,
+      "learning_rate": 0.0001006266685401152,
+      "loss": 5.2702,
+      "step": 23880
+    },
+    {
+      "epoch": 0.03356769127278667,
+      "grad_norm": 1.2920984029769897,
+      "learning_rate": 0.00010066882113249965,
+      "loss": 5.2699,
+      "step": 23890
+    },
+    {
+      "epoch": 0.03358174221095025,
+      "grad_norm": 1.2577017545700073,
+      "learning_rate": 0.00010071097372488407,
+      "loss": 5.3344,
+      "step": 23900
+    },
+    {
+      "epoch": 0.033595793149113824,
+      "grad_norm": 2.0838184356689453,
+      "learning_rate": 0.0001007531263172685,
+      "loss": 5.1927,
+      "step": 23910
+    },
+    {
+      "epoch": 0.0336098440872774,
+      "grad_norm": 1.5343761444091797,
+      "learning_rate": 0.00010079527890965294,
+      "loss": 5.3123,
+      "step": 23920
+    },
+    {
+      "epoch": 0.03362389502544098,
+      "grad_norm": 1.3116449117660522,
+      "learning_rate": 0.00010083743150203737,
+      "loss": 5.3366,
+      "step": 23930
+    },
+    {
+      "epoch": 0.03363794596360456,
+      "grad_norm": 1.2814358472824097,
+      "learning_rate": 0.00010087958409442179,
+      "loss": 5.2453,
+      "step": 23940
+    },
+    {
+      "epoch": 0.033651996901768136,
+      "grad_norm": 1.2302325963974,
+      "learning_rate": 0.00010092173668680623,
+      "loss": 5.2506,
+      "step": 23950
+    },
+    {
+      "epoch": 0.03366604783993171,
+      "grad_norm": 1.3629188537597656,
+      "learning_rate": 0.00010096388927919067,
+      "loss": 5.3369,
+      "step": 23960
+    },
+    {
+      "epoch": 0.03368009877809529,
+      "grad_norm": 1.241894245147705,
+      "learning_rate": 0.00010100604187157508,
+      "loss": 5.3224,
+      "step": 23970
+    },
+    {
+      "epoch": 0.033694149716258864,
+      "grad_norm": 1.3450833559036255,
+      "learning_rate": 0.00010104819446395953,
+      "loss": 5.2227,
+      "step": 23980
+    },
+    {
+      "epoch": 0.03370820065442245,
+      "grad_norm": 1.3539717197418213,
+      "learning_rate": 0.00010109034705634396,
+      "loss": 5.1751,
+      "step": 23990
+    },
+    {
+      "epoch": 0.033722251592586024,
+      "grad_norm": 1.5781745910644531,
+      "learning_rate": 0.00010113249964872838,
+      "loss": 5.2189,
+      "step": 24000
+    },
+    {
+      "epoch": 0.0337363025307496,
+      "grad_norm": 1.2791818380355835,
+      "learning_rate": 0.00010117465224111282,
+      "loss": 5.3397,
+      "step": 24010
+    },
+    {
+      "epoch": 0.033750353468913176,
+      "grad_norm": 1.318710207939148,
+      "learning_rate": 0.00010121680483349725,
+      "loss": 5.2535,
+      "step": 24020
+    },
+    {
+      "epoch": 0.03376440440707675,
+      "grad_norm": 1.4116847515106201,
+      "learning_rate": 0.00010125895742588167,
+      "loss": 5.246,
+      "step": 24030
+    },
+    {
+      "epoch": 0.033778455345240335,
+      "grad_norm": 1.352486491203308,
+      "learning_rate": 0.00010130111001826612,
+      "loss": 5.3818,
+      "step": 24040
+    },
+    {
+      "epoch": 0.03379250628340391,
+      "grad_norm": 1.3490184545516968,
+      "learning_rate": 0.00010134326261065055,
+      "loss": 5.1459,
+      "step": 24050
+    },
+    {
+      "epoch": 0.03380655722156749,
+      "grad_norm": 1.25826895236969,
+      "learning_rate": 0.00010138541520303496,
+      "loss": 5.2441,
+      "step": 24060
+    },
+    {
+      "epoch": 0.033820608159731064,
+      "grad_norm": 1.2731109857559204,
+      "learning_rate": 0.00010142756779541941,
+      "loss": 5.382,
+      "step": 24070
+    },
+    {
+      "epoch": 0.03383465909789464,
+      "grad_norm": 1.2558667659759521,
+      "learning_rate": 0.00010146972038780384,
+      "loss": 5.2548,
+      "step": 24080
+    },
+    {
+      "epoch": 0.03384871003605822,
+      "grad_norm": 1.3731662034988403,
+      "learning_rate": 0.00010151187298018828,
+      "loss": 5.1889,
+      "step": 24090
+    },
+    {
+      "epoch": 0.0338627609742218,
+      "grad_norm": 1.2584823369979858,
+      "learning_rate": 0.0001015540255725727,
+      "loss": 5.2309,
+      "step": 24100
+    },
+    {
+      "epoch": 0.033876811912385375,
+      "grad_norm": 1.3157720565795898,
+      "learning_rate": 0.00010159617816495713,
+      "loss": 5.147,
+      "step": 24110
+    },
+    {
+      "epoch": 0.03389086285054895,
+      "grad_norm": 1.3942573070526123,
+      "learning_rate": 0.00010163833075734158,
+      "loss": 5.1946,
+      "step": 24120
+    },
+    {
+      "epoch": 0.03390491378871253,
+      "grad_norm": 1.218231201171875,
+      "learning_rate": 0.000101680483349726,
+      "loss": 5.3256,
+      "step": 24130
+    },
+    {
+      "epoch": 0.03391896472687611,
+      "grad_norm": 1.3514090776443481,
+      "learning_rate": 0.00010172263594211043,
+      "loss": 5.2472,
+      "step": 24140
+    },
+    {
+      "epoch": 0.03393301566503969,
+      "grad_norm": 1.4769346714019775,
+      "learning_rate": 0.00010176478853449487,
+      "loss": 5.3211,
+      "step": 24150
+    },
+    {
+      "epoch": 0.03394706660320326,
+      "grad_norm": 1.4301344156265259,
+      "learning_rate": 0.00010180694112687929,
+      "loss": 5.3223,
+      "step": 24160
+    },
+    {
+      "epoch": 0.03396111754136684,
+      "grad_norm": 1.3295481204986572,
+      "learning_rate": 0.00010184909371926372,
+      "loss": 5.2742,
+      "step": 24170
+    },
+    {
+      "epoch": 0.033975168479530415,
+      "grad_norm": 1.34248948097229,
+      "learning_rate": 0.00010189124631164817,
+      "loss": 5.3578,
+      "step": 24180
+    },
+    {
+      "epoch": 0.033989219417694,
+      "grad_norm": 1.3010623455047607,
+      "learning_rate": 0.00010193339890403258,
+      "loss": 5.3045,
+      "step": 24190
+    },
+    {
+      "epoch": 0.034003270355857575,
+      "grad_norm": 1.3236461877822876,
+      "learning_rate": 0.00010197555149641701,
+      "loss": 5.2843,
+      "step": 24200
+    },
+    {
+      "epoch": 0.03401732129402115,
+      "grad_norm": 1.3258452415466309,
+      "learning_rate": 0.00010201770408880146,
+      "loss": 5.1747,
+      "step": 24210
+    },
+    {
+      "epoch": 0.03403137223218473,
+      "grad_norm": 1.411746621131897,
+      "learning_rate": 0.00010205985668118589,
+      "loss": 5.2508,
+      "step": 24220
+    },
+    {
+      "epoch": 0.0340454231703483,
+      "grad_norm": 1.399804949760437,
+      "learning_rate": 0.00010210200927357031,
+      "loss": 5.3071,
+      "step": 24230
+    },
+    {
+      "epoch": 0.03405947410851188,
+      "grad_norm": 1.3025681972503662,
+      "learning_rate": 0.00010214416186595475,
+      "loss": 5.1783,
+      "step": 24240
+    },
+    {
+      "epoch": 0.03407352504667546,
+      "grad_norm": 1.245466709136963,
+      "learning_rate": 0.00010218631445833918,
+      "loss": 5.2707,
+      "step": 24250
+    },
+    {
+      "epoch": 0.03408757598483904,
+      "grad_norm": 1.307504415512085,
+      "learning_rate": 0.0001022284670507236,
+      "loss": 5.418,
+      "step": 24260
+    },
+    {
+      "epoch": 0.034101626923002615,
+      "grad_norm": 1.331075668334961,
+      "learning_rate": 0.00010227061964310805,
+      "loss": 5.3077,
+      "step": 24270
+    },
+    {
+      "epoch": 0.03411567786116619,
+      "grad_norm": 1.3218508958816528,
+      "learning_rate": 0.00010231277223549248,
+      "loss": 5.4505,
+      "step": 24280
+    },
+    {
+      "epoch": 0.03412972879932977,
+      "grad_norm": 1.4919476509094238,
+      "learning_rate": 0.0001023549248278769,
+      "loss": 5.1643,
+      "step": 24290
+    },
+    {
+      "epoch": 0.03414377973749335,
+      "grad_norm": 1.511077642440796,
+      "learning_rate": 0.00010239707742026134,
+      "loss": 5.3138,
+      "step": 24300
+    },
+    {
+      "epoch": 0.034157830675656926,
+      "grad_norm": 1.2608944177627563,
+      "learning_rate": 0.00010243923001264577,
+      "loss": 5.353,
+      "step": 24310
+    },
+    {
+      "epoch": 0.0341718816138205,
+      "grad_norm": 1.2811615467071533,
+      "learning_rate": 0.00010248138260503019,
+      "loss": 5.326,
+      "step": 24320
+    },
+    {
+      "epoch": 0.03418593255198408,
+      "grad_norm": 1.3658688068389893,
+      "learning_rate": 0.00010252353519741463,
+      "loss": 5.2803,
+      "step": 24330
+    },
+    {
+      "epoch": 0.034199983490147655,
+      "grad_norm": 1.2968345880508423,
+      "learning_rate": 0.00010256568778979906,
+      "loss": 5.2007,
+      "step": 24340
+    },
+    {
+      "epoch": 0.03421403442831124,
+      "grad_norm": 1.3284701108932495,
+      "learning_rate": 0.00010260784038218348,
+      "loss": 5.2224,
+      "step": 24350
+    },
+    {
+      "epoch": 0.034228085366474814,
+      "grad_norm": 1.2858362197875977,
+      "learning_rate": 0.00010264999297456793,
+      "loss": 5.3118,
+      "step": 24360
+    },
+    {
+      "epoch": 0.03424213630463839,
+      "grad_norm": 1.242591381072998,
+      "learning_rate": 0.00010269214556695236,
+      "loss": 5.2242,
+      "step": 24370
+    },
+    {
+      "epoch": 0.034256187242801966,
+      "grad_norm": 1.282493233680725,
+      "learning_rate": 0.0001027342981593368,
+      "loss": 5.4394,
+      "step": 24380
+    },
+    {
+      "epoch": 0.03427023818096554,
+      "grad_norm": 1.2927570343017578,
+      "learning_rate": 0.00010277645075172122,
+      "loss": 5.3704,
+      "step": 24390
+    },
+    {
+      "epoch": 0.034284289119129126,
+      "grad_norm": 1.2820286750793457,
+      "learning_rate": 0.00010281860334410565,
+      "loss": 5.2751,
+      "step": 24400
+    },
+    {
+      "epoch": 0.0342983400572927,
+      "grad_norm": 1.3351532220840454,
+      "learning_rate": 0.0001028607559364901,
+      "loss": 5.2028,
+      "step": 24410
+    },
+    {
+      "epoch": 0.03431239099545628,
+      "grad_norm": 1.3240007162094116,
+      "learning_rate": 0.00010290290852887451,
+      "loss": 5.1208,
+      "step": 24420
+    },
+    {
+      "epoch": 0.034326441933619854,
+      "grad_norm": 1.3099170923233032,
+      "learning_rate": 0.00010294506112125894,
+      "loss": 5.3686,
+      "step": 24430
+    },
+    {
+      "epoch": 0.03434049287178343,
+      "grad_norm": 1.3434085845947266,
+      "learning_rate": 0.00010298721371364339,
+      "loss": 5.2695,
+      "step": 24440
+    },
+    {
+      "epoch": 0.034354543809947014,
+      "grad_norm": 1.2506746053695679,
+      "learning_rate": 0.0001030293663060278,
+      "loss": 5.4538,
+      "step": 24450
+    },
+    {
+      "epoch": 0.03436859474811059,
+      "grad_norm": 1.2607885599136353,
+      "learning_rate": 0.00010307151889841224,
+      "loss": 5.2082,
+      "step": 24460
+    },
+    {
+      "epoch": 0.034382645686274166,
+      "grad_norm": 1.2790932655334473,
+      "learning_rate": 0.00010311367149079668,
+      "loss": 5.3349,
+      "step": 24470
+    },
+    {
+      "epoch": 0.03439669662443774,
+      "grad_norm": 1.4397903680801392,
+      "learning_rate": 0.0001031558240831811,
+      "loss": 5.2953,
+      "step": 24480
+    },
+    {
+      "epoch": 0.03441074756260132,
+      "grad_norm": 1.254449725151062,
+      "learning_rate": 0.00010319797667556553,
+      "loss": 5.3391,
+      "step": 24490
+    },
+    {
+      "epoch": 0.0344247985007649,
+      "grad_norm": 1.3138560056686401,
+      "learning_rate": 0.00010324012926794998,
+      "loss": 5.2934,
+      "step": 24500
+    },
+    {
+      "epoch": 0.03443884943892848,
+      "grad_norm": 1.2442389726638794,
+      "learning_rate": 0.00010328228186033441,
+      "loss": 5.1722,
+      "step": 24510
+    },
+    {
+      "epoch": 0.034452900377092054,
+      "grad_norm": 1.3116623163223267,
+      "learning_rate": 0.00010332443445271882,
+      "loss": 5.2642,
+      "step": 24520
+    },
+    {
+      "epoch": 0.03446695131525563,
+      "grad_norm": 1.3116538524627686,
+      "learning_rate": 0.00010336658704510327,
+      "loss": 5.3113,
+      "step": 24530
+    },
+    {
+      "epoch": 0.034481002253419206,
+      "grad_norm": 1.3560738563537598,
+      "learning_rate": 0.0001034087396374877,
+      "loss": 5.2302,
+      "step": 24540
+    },
+    {
+      "epoch": 0.03449505319158278,
+      "grad_norm": 1.3531079292297363,
+      "learning_rate": 0.00010345089222987212,
+      "loss": 5.2305,
+      "step": 24550
+    },
+    {
+      "epoch": 0.034509104129746365,
+      "grad_norm": 1.26160728931427,
+      "learning_rate": 0.00010349304482225656,
+      "loss": 5.2376,
+      "step": 24560
+    },
+    {
+      "epoch": 0.03452315506790994,
+      "grad_norm": 1.4624823331832886,
+      "learning_rate": 0.000103535197414641,
+      "loss": 5.1941,
+      "step": 24570
+    },
+    {
+      "epoch": 0.03453720600607352,
+      "grad_norm": 1.3396354913711548,
+      "learning_rate": 0.00010357735000702541,
+      "loss": 5.1731,
+      "step": 24580
+    },
+    {
+      "epoch": 0.034551256944237094,
+      "grad_norm": 1.3305305242538452,
+      "learning_rate": 0.00010361950259940986,
+      "loss": 5.3613,
+      "step": 24590
+    },
+    {
+      "epoch": 0.03456530788240067,
+      "grad_norm": 1.2732019424438477,
+      "learning_rate": 0.00010366165519179429,
+      "loss": 5.2624,
+      "step": 24600
+    },
+    {
+      "epoch": 0.03457935882056425,
+      "grad_norm": 1.428470253944397,
+      "learning_rate": 0.0001037038077841787,
+      "loss": 5.2881,
+      "step": 24610
+    },
+    {
+      "epoch": 0.03459340975872783,
+      "grad_norm": 1.3105640411376953,
+      "learning_rate": 0.00010374596037656315,
+      "loss": 5.3397,
+      "step": 24620
+    },
+    {
+      "epoch": 0.034607460696891405,
+      "grad_norm": 1.4078952074050903,
+      "learning_rate": 0.00010378811296894758,
+      "loss": 5.4407,
+      "step": 24630
+    },
+    {
+      "epoch": 0.03462151163505498,
+      "grad_norm": 1.3400802612304688,
+      "learning_rate": 0.00010383026556133203,
+      "loss": 5.3244,
+      "step": 24640
+    },
+    {
+      "epoch": 0.03463556257321856,
+      "grad_norm": 1.315213918685913,
+      "learning_rate": 0.00010387241815371644,
+      "loss": 5.3574,
+      "step": 24650
+    },
+    {
+      "epoch": 0.03464961351138214,
+      "grad_norm": 1.2948580980300903,
+      "learning_rate": 0.00010391457074610087,
+      "loss": 5.3071,
+      "step": 24660
+    },
+    {
+      "epoch": 0.03466366444954572,
+      "grad_norm": 1.3167226314544678,
+      "learning_rate": 0.00010395672333848532,
+      "loss": 5.2615,
+      "step": 24670
+    },
+    {
+      "epoch": 0.03467771538770929,
+      "grad_norm": 1.3595393896102905,
+      "learning_rate": 0.00010399887593086974,
+      "loss": 5.3395,
+      "step": 24680
+    },
+    {
+      "epoch": 0.03469176632587287,
+      "grad_norm": 1.3089308738708496,
+      "learning_rate": 0.00010404102852325417,
+      "loss": 5.1544,
+      "step": 24690
+    },
+    {
+      "epoch": 0.034705817264036445,
+      "grad_norm": 1.2565276622772217,
+      "learning_rate": 0.00010408318111563861,
+      "loss": 5.3002,
+      "step": 24700
+    },
+    {
+      "epoch": 0.03471986820220003,
+      "grad_norm": 1.256977915763855,
+      "learning_rate": 0.00010412533370802303,
+      "loss": 5.287,
+      "step": 24710
+    },
+    {
+      "epoch": 0.034733919140363605,
+      "grad_norm": 1.2311729192733765,
+      "learning_rate": 0.00010416748630040746,
+      "loss": 5.39,
+      "step": 24720
+    },
+    {
+      "epoch": 0.03474797007852718,
+      "grad_norm": 1.2389764785766602,
+      "learning_rate": 0.0001042096388927919,
+      "loss": 5.3162,
+      "step": 24730
+    },
+    {
+      "epoch": 0.03476202101669076,
+      "grad_norm": 1.2789169549942017,
+      "learning_rate": 0.00010425179148517632,
+      "loss": 5.3211,
+      "step": 24740
+    },
+    {
+      "epoch": 0.03477607195485433,
+      "grad_norm": 1.225974202156067,
+      "learning_rate": 0.00010429394407756076,
+      "loss": 5.2767,
+      "step": 24750
+    },
+    {
+      "epoch": 0.034790122893017916,
+      "grad_norm": 1.2819294929504395,
+      "learning_rate": 0.0001043360966699452,
+      "loss": 5.335,
+      "step": 24760
+    },
+    {
+      "epoch": 0.03480417383118149,
+      "grad_norm": 1.329951524734497,
+      "learning_rate": 0.00010437824926232962,
+      "loss": 5.1979,
+      "step": 24770
+    },
+    {
+      "epoch": 0.03481822476934507,
+      "grad_norm": 1.2084249258041382,
+      "learning_rate": 0.00010442040185471405,
+      "loss": 5.3172,
+      "step": 24780
+    },
+    {
+      "epoch": 0.034832275707508645,
+      "grad_norm": 1.2773807048797607,
+      "learning_rate": 0.0001044625544470985,
+      "loss": 5.2737,
+      "step": 24790
+    },
+    {
+      "epoch": 0.03484632664567222,
+      "grad_norm": 1.302215576171875,
+      "learning_rate": 0.00010450470703948292,
+      "loss": 5.2025,
+      "step": 24800
+    },
+    {
+      "epoch": 0.034860377583835804,
+      "grad_norm": 1.317597508430481,
+      "learning_rate": 0.00010454685963186734,
+      "loss": 5.3506,
+      "step": 24810
+    },
+    {
+      "epoch": 0.03487442852199938,
+      "grad_norm": 1.2877779006958008,
+      "learning_rate": 0.00010458901222425179,
+      "loss": 5.2092,
+      "step": 24820
+    },
+    {
+      "epoch": 0.034888479460162956,
+      "grad_norm": 1.297938585281372,
+      "learning_rate": 0.00010463116481663622,
+      "loss": 5.3321,
+      "step": 24830
+    },
+    {
+      "epoch": 0.03490253039832653,
+      "grad_norm": 1.4365018606185913,
+      "learning_rate": 0.00010467331740902064,
+      "loss": 5.1708,
+      "step": 24840
+    },
+    {
+      "epoch": 0.03491658133649011,
+      "grad_norm": 1.2824749946594238,
+      "learning_rate": 0.00010471547000140508,
+      "loss": 5.2512,
+      "step": 24850
+    },
+    {
+      "epoch": 0.034930632274653685,
+      "grad_norm": 1.3265628814697266,
+      "learning_rate": 0.00010475762259378951,
+      "loss": 5.1575,
+      "step": 24860
+    },
+    {
+      "epoch": 0.03494468321281727,
+      "grad_norm": 1.360796332359314,
+      "learning_rate": 0.00010479977518617393,
+      "loss": 5.2298,
+      "step": 24870
+    },
+    {
+      "epoch": 0.034958734150980844,
+      "grad_norm": 1.2341952323913574,
+      "learning_rate": 0.00010484192777855837,
+      "loss": 5.1945,
+      "step": 24880
+    },
+    {
+      "epoch": 0.03497278508914442,
+      "grad_norm": 1.2878468036651611,
+      "learning_rate": 0.0001048840803709428,
+      "loss": 5.2382,
+      "step": 24890
+    },
+    {
+      "epoch": 0.034986836027307996,
+      "grad_norm": 1.248975157737732,
+      "learning_rate": 0.00010492623296332722,
+      "loss": 5.4503,
+      "step": 24900
+    },
+    {
+      "epoch": 0.03500088696547157,
+      "grad_norm": 1.2369763851165771,
+      "learning_rate": 0.00010496838555571167,
+      "loss": 5.2011,
+      "step": 24910
+    },
+    {
+      "epoch": 0.035014937903635156,
+      "grad_norm": 1.4231282472610474,
+      "learning_rate": 0.0001050105381480961,
+      "loss": 5.3073,
+      "step": 24920
+    },
+    {
+      "epoch": 0.03502898884179873,
+      "grad_norm": 1.3121232986450195,
+      "learning_rate": 0.00010505269074048054,
+      "loss": 5.2135,
+      "step": 24930
+    },
+    {
+      "epoch": 0.03504303977996231,
+      "grad_norm": 1.2677346467971802,
+      "learning_rate": 0.00010509484333286496,
+      "loss": 5.2371,
+      "step": 24940
+    },
+    {
+      "epoch": 0.035057090718125884,
+      "grad_norm": 1.359431266784668,
+      "learning_rate": 0.00010513699592524939,
+      "loss": 5.2875,
+      "step": 24950
+    },
+    {
+      "epoch": 0.03507114165628946,
+      "grad_norm": 1.2087974548339844,
+      "learning_rate": 0.00010517914851763384,
+      "loss": 5.1022,
+      "step": 24960
+    },
+    {
+      "epoch": 0.035085192594453044,
+      "grad_norm": 1.297215461730957,
+      "learning_rate": 0.00010522130111001825,
+      "loss": 5.3486,
+      "step": 24970
+    },
+    {
+      "epoch": 0.03509924353261662,
+      "grad_norm": 1.2546544075012207,
+      "learning_rate": 0.00010526345370240269,
+      "loss": 5.2505,
+      "step": 24980
+    },
+    {
+      "epoch": 0.035113294470780196,
+      "grad_norm": 1.3065543174743652,
+      "learning_rate": 0.00010530560629478713,
+      "loss": 5.1717,
+      "step": 24990
+    },
+    {
+      "epoch": 0.03512734540894377,
+      "grad_norm": 1.2734280824661255,
+      "learning_rate": 0.00010534775888717155,
+      "loss": 5.1809,
+      "step": 25000
+    },
+    {
+      "epoch": 0.03514139634710735,
+      "grad_norm": 1.3350660800933838,
+      "learning_rate": 0.00010538991147955598,
+      "loss": 5.3357,
+      "step": 25010
+    },
+    {
+      "epoch": 0.03515544728527093,
+      "grad_norm": 1.3462845087051392,
+      "learning_rate": 0.00010543206407194042,
+      "loss": 5.266,
+      "step": 25020
+    },
+    {
+      "epoch": 0.03516949822343451,
+      "grad_norm": 1.4067381620407104,
+      "learning_rate": 0.00010547000140508639,
+      "loss": 5.2521,
+      "step": 25030
+    },
+    {
+      "epoch": 0.035183549161598084,
+      "grad_norm": 1.3267967700958252,
+      "learning_rate": 0.00010551215399747084,
+      "loss": 5.2519,
+      "step": 25040
+    },
+    {
+      "epoch": 0.03519760009976166,
+      "grad_norm": 1.2769830226898193,
+      "learning_rate": 0.00010555430658985527,
+      "loss": 5.3735,
+      "step": 25050
+    },
+    {
+      "epoch": 0.035211651037925236,
+      "grad_norm": 1.4258005619049072,
+      "learning_rate": 0.0001055964591822397,
+      "loss": 5.3553,
+      "step": 25060
+    },
+    {
+      "epoch": 0.03522570197608882,
+      "grad_norm": 1.370233416557312,
+      "learning_rate": 0.00010563861177462413,
+      "loss": 5.2365,
+      "step": 25070
+    },
+    {
+      "epoch": 0.035239752914252395,
+      "grad_norm": 1.2431062459945679,
+      "learning_rate": 0.00010568076436700856,
+      "loss": 5.244,
+      "step": 25080
+    },
+    {
+      "epoch": 0.03525380385241597,
+      "grad_norm": 1.4300603866577148,
+      "learning_rate": 0.00010572291695939301,
+      "loss": 5.1916,
+      "step": 25090
+    },
+    {
+      "epoch": 0.03526785479057955,
+      "grad_norm": 1.3223384618759155,
+      "learning_rate": 0.00010576506955177742,
+      "loss": 5.1435,
+      "step": 25100
+    },
+    {
+      "epoch": 0.035281905728743124,
+      "grad_norm": 1.2834317684173584,
+      "learning_rate": 0.00010580722214416186,
+      "loss": 5.3326,
+      "step": 25110
+    },
+    {
+      "epoch": 0.03529595666690671,
+      "grad_norm": 1.2990752458572388,
+      "learning_rate": 0.0001058493747365463,
+      "loss": 5.1775,
+      "step": 25120
+    },
+    {
+      "epoch": 0.03531000760507028,
+      "grad_norm": 1.2824280261993408,
+      "learning_rate": 0.00010589152732893072,
+      "loss": 5.2595,
+      "step": 25130
+    },
+    {
+      "epoch": 0.03532405854323386,
+      "grad_norm": 1.3103786706924438,
+      "learning_rate": 0.00010593367992131515,
+      "loss": 5.2454,
+      "step": 25140
+    },
+    {
+      "epoch": 0.035338109481397435,
+      "grad_norm": 1.2637652158737183,
+      "learning_rate": 0.0001059758325136996,
+      "loss": 5.1544,
+      "step": 25150
+    },
+    {
+      "epoch": 0.03535216041956101,
+      "grad_norm": 1.2903791666030884,
+      "learning_rate": 0.00010601798510608401,
+      "loss": 5.2983,
+      "step": 25160
+    },
+    {
+      "epoch": 0.03536621135772459,
+      "grad_norm": 1.268822193145752,
+      "learning_rate": 0.00010606013769846844,
+      "loss": 5.1807,
+      "step": 25170
+    },
+    {
+      "epoch": 0.03538026229588817,
+      "grad_norm": 1.3194992542266846,
+      "learning_rate": 0.00010610229029085289,
+      "loss": 5.1897,
+      "step": 25180
+    },
+    {
+      "epoch": 0.03539431323405175,
+      "grad_norm": 1.3022356033325195,
+      "learning_rate": 0.0001061444428832373,
+      "loss": 5.1931,
+      "step": 25190
+    },
+    {
+      "epoch": 0.03540836417221532,
+      "grad_norm": 1.4322096109390259,
+      "learning_rate": 0.00010618659547562174,
+      "loss": 5.1337,
+      "step": 25200
+    },
+    {
+      "epoch": 0.0354224151103789,
+      "grad_norm": 1.285513162612915,
+      "learning_rate": 0.00010622874806800618,
+      "loss": 5.3514,
+      "step": 25210
+    },
+    {
+      "epoch": 0.035436466048542475,
+      "grad_norm": 1.3835152387619019,
+      "learning_rate": 0.00010627090066039061,
+      "loss": 5.2987,
+      "step": 25220
+    },
+    {
+      "epoch": 0.03545051698670606,
+      "grad_norm": 1.3201239109039307,
+      "learning_rate": 0.00010631305325277503,
+      "loss": 5.3481,
+      "step": 25230
+    },
+    {
+      "epoch": 0.035464567924869635,
+      "grad_norm": 1.3172633647918701,
+      "learning_rate": 0.00010635520584515947,
+      "loss": 5.2151,
+      "step": 25240
+    },
+    {
+      "epoch": 0.03547861886303321,
+      "grad_norm": 1.3208476305007935,
+      "learning_rate": 0.0001063973584375439,
+      "loss": 5.2665,
+      "step": 25250
+    },
+    {
+      "epoch": 0.03549266980119679,
+      "grad_norm": 1.2850348949432373,
+      "learning_rate": 0.00010643951102992832,
+      "loss": 5.2579,
+      "step": 25260
+    },
+    {
+      "epoch": 0.03550672073936036,
+      "grad_norm": 1.2918301820755005,
+      "learning_rate": 0.00010648166362231277,
+      "loss": 5.3202,
+      "step": 25270
+    },
+    {
+      "epoch": 0.035520771677523946,
+      "grad_norm": 1.3454853296279907,
+      "learning_rate": 0.0001065238162146972,
+      "loss": 5.302,
+      "step": 25280
+    },
+    {
+      "epoch": 0.03553482261568752,
+      "grad_norm": 1.23288893699646,
+      "learning_rate": 0.00010656596880708162,
+      "loss": 5.2014,
+      "step": 25290
+    },
+    {
+      "epoch": 0.0355488735538511,
+      "grad_norm": 1.2703696489334106,
+      "learning_rate": 0.00010660812139946606,
+      "loss": 5.2891,
+      "step": 25300
+    },
+    {
+      "epoch": 0.035562924492014675,
+      "grad_norm": 1.3757070302963257,
+      "learning_rate": 0.00010665027399185049,
+      "loss": 5.2369,
+      "step": 25310
+    },
+    {
+      "epoch": 0.03557697543017825,
+      "grad_norm": 1.3169059753417969,
+      "learning_rate": 0.00010669242658423491,
+      "loss": 5.2194,
+      "step": 25320
+    },
+    {
+      "epoch": 0.035591026368341834,
+      "grad_norm": 1.3210643529891968,
+      "learning_rate": 0.00010673457917661935,
+      "loss": 5.2344,
+      "step": 25330
+    },
+    {
+      "epoch": 0.03560507730650541,
+      "grad_norm": 1.2993870973587036,
+      "learning_rate": 0.00010677673176900379,
+      "loss": 5.2018,
+      "step": 25340
+    },
+    {
+      "epoch": 0.035619128244668986,
+      "grad_norm": 1.299553632736206,
+      "learning_rate": 0.00010681888436138822,
+      "loss": 5.2323,
+      "step": 25350
+    },
+    {
+      "epoch": 0.03563317918283256,
+      "grad_norm": 1.4303969144821167,
+      "learning_rate": 0.00010686103695377265,
+      "loss": 5.1968,
+      "step": 25360
+    },
+    {
+      "epoch": 0.03564723012099614,
+      "grad_norm": 1.2443745136260986,
+      "learning_rate": 0.00010690318954615708,
+      "loss": 5.3775,
+      "step": 25370
+    },
+    {
+      "epoch": 0.03566128105915972,
+      "grad_norm": 1.5378773212432861,
+      "learning_rate": 0.00010694534213854152,
+      "loss": 5.1952,
+      "step": 25380
+    },
+    {
+      "epoch": 0.0356753319973233,
+      "grad_norm": 1.2778575420379639,
+      "learning_rate": 0.00010698749473092594,
+      "loss": 5.1902,
+      "step": 25390
+    },
+    {
+      "epoch": 0.035689382935486874,
+      "grad_norm": 1.241885781288147,
+      "learning_rate": 0.00010702964732331037,
+      "loss": 5.3114,
+      "step": 25400
+    },
+    {
+      "epoch": 0.03570343387365045,
+      "grad_norm": 1.250803828239441,
+      "learning_rate": 0.00010707179991569482,
+      "loss": 5.2635,
+      "step": 25410
+    },
+    {
+      "epoch": 0.035717484811814026,
+      "grad_norm": 1.1884135007858276,
+      "learning_rate": 0.00010711395250807924,
+      "loss": 5.2907,
+      "step": 25420
+    },
+    {
+      "epoch": 0.03573153574997761,
+      "grad_norm": 1.169700026512146,
+      "learning_rate": 0.00010715610510046367,
+      "loss": 5.2049,
+      "step": 25430
+    },
+    {
+      "epoch": 0.035745586688141186,
+      "grad_norm": 1.2907392978668213,
+      "learning_rate": 0.00010719825769284811,
+      "loss": 5.1974,
+      "step": 25440
+    },
+    {
+      "epoch": 0.03575963762630476,
+      "grad_norm": 1.3100101947784424,
+      "learning_rate": 0.00010724041028523253,
+      "loss": 5.1541,
+      "step": 25450
+    },
+    {
+      "epoch": 0.03577368856446834,
+      "grad_norm": 1.3046437501907349,
+      "learning_rate": 0.00010728256287761696,
+      "loss": 5.2203,
+      "step": 25460
+    },
+    {
+      "epoch": 0.035787739502631914,
+      "grad_norm": 1.2555994987487793,
+      "learning_rate": 0.0001073247154700014,
+      "loss": 5.2891,
+      "step": 25470
+    },
+    {
+      "epoch": 0.0358017904407955,
+      "grad_norm": 1.315011978149414,
+      "learning_rate": 0.00010736686806238582,
+      "loss": 5.3418,
+      "step": 25480
+    },
+    {
+      "epoch": 0.035815841378959073,
+      "grad_norm": 1.3488926887512207,
+      "learning_rate": 0.00010740902065477025,
+      "loss": 5.082,
+      "step": 25490
+    },
+    {
+      "epoch": 0.03582989231712265,
+      "grad_norm": 1.2231671810150146,
+      "learning_rate": 0.0001074511732471547,
+      "loss": 5.2338,
+      "step": 25500
+    },
+    {
+      "epoch": 0.035843943255286226,
+      "grad_norm": 1.2394180297851562,
+      "learning_rate": 0.00010749332583953913,
+      "loss": 5.2079,
+      "step": 25510
+    },
+    {
+      "epoch": 0.0358579941934498,
+      "grad_norm": 1.2652291059494019,
+      "learning_rate": 0.00010753547843192355,
+      "loss": 5.09,
+      "step": 25520
+    },
+    {
+      "epoch": 0.03587204513161338,
+      "grad_norm": 1.3209153413772583,
+      "learning_rate": 0.00010757763102430799,
+      "loss": 5.1912,
+      "step": 25530
+    },
+    {
+      "epoch": 0.03588609606977696,
+      "grad_norm": 1.2716671228408813,
+      "learning_rate": 0.00010761978361669242,
+      "loss": 5.1782,
+      "step": 25540
+    },
+    {
+      "epoch": 0.03590014700794054,
+      "grad_norm": 1.2218382358551025,
+      "learning_rate": 0.00010766193620907684,
+      "loss": 5.3313,
+      "step": 25550
+    },
+    {
+      "epoch": 0.035914197946104114,
+      "grad_norm": 1.2553699016571045,
+      "learning_rate": 0.00010770408880146129,
+      "loss": 5.2144,
+      "step": 25560
+    },
+    {
+      "epoch": 0.03592824888426769,
+      "grad_norm": 1.3492262363433838,
+      "learning_rate": 0.00010774624139384572,
+      "loss": 5.2676,
+      "step": 25570
+    },
+    {
+      "epoch": 0.035942299822431266,
+      "grad_norm": 1.2544468641281128,
+      "learning_rate": 0.00010778839398623013,
+      "loss": 5.2419,
+      "step": 25580
+    },
+    {
+      "epoch": 0.03595635076059485,
+      "grad_norm": 1.228851079940796,
+      "learning_rate": 0.00010783054657861458,
+      "loss": 5.2684,
+      "step": 25590
+    },
+    {
+      "epoch": 0.035970401698758425,
+      "grad_norm": 1.2383124828338623,
+      "learning_rate": 0.00010787269917099901,
+      "loss": 5.2532,
+      "step": 25600
+    },
+    {
+      "epoch": 0.035984452636922,
+      "grad_norm": 1.3050917387008667,
+      "learning_rate": 0.00010791485176338343,
+      "loss": 5.2751,
+      "step": 25610
+    },
+    {
+      "epoch": 0.03599850357508558,
+      "grad_norm": 1.26146399974823,
+      "learning_rate": 0.00010795700435576787,
+      "loss": 5.1677,
+      "step": 25620
+    },
+    {
+      "epoch": 0.036012554513249154,
+      "grad_norm": 1.2697525024414062,
+      "learning_rate": 0.0001079991569481523,
+      "loss": 5.3691,
+      "step": 25630
+    },
+    {
+      "epoch": 0.03602660545141274,
+      "grad_norm": 1.299674153327942,
+      "learning_rate": 0.00010804130954053673,
+      "loss": 5.198,
+      "step": 25640
+    },
+    {
+      "epoch": 0.03604065638957631,
+      "grad_norm": 1.3521478176116943,
+      "learning_rate": 0.00010808346213292117,
+      "loss": 5.0404,
+      "step": 25650
+    },
+    {
+      "epoch": 0.03605470732773989,
+      "grad_norm": 1.270524501800537,
+      "learning_rate": 0.0001081256147253056,
+      "loss": 5.1391,
+      "step": 25660
+    },
+    {
+      "epoch": 0.036068758265903465,
+      "grad_norm": 1.2725433111190796,
+      "learning_rate": 0.00010816776731769004,
+      "loss": 5.3264,
+      "step": 25670
+    },
+    {
+      "epoch": 0.03608280920406704,
+      "grad_norm": 1.283799171447754,
+      "learning_rate": 0.00010820991991007446,
+      "loss": 5.2389,
+      "step": 25680
+    },
+    {
+      "epoch": 0.036096860142230625,
+      "grad_norm": 1.2880244255065918,
+      "learning_rate": 0.00010825207250245889,
+      "loss": 5.2285,
+      "step": 25690
+    },
+    {
+      "epoch": 0.0361109110803942,
+      "grad_norm": 1.3934433460235596,
+      "learning_rate": 0.00010829422509484334,
+      "loss": 5.2087,
+      "step": 25700
+    },
+    {
+      "epoch": 0.03612496201855778,
+      "grad_norm": 1.244972825050354,
+      "learning_rate": 0.00010833637768722775,
+      "loss": 5.3292,
+      "step": 25710
+    },
+    {
+      "epoch": 0.03613901295672135,
+      "grad_norm": 1.2489376068115234,
+      "learning_rate": 0.00010837853027961218,
+      "loss": 5.4028,
+      "step": 25720
+    },
+    {
+      "epoch": 0.03615306389488493,
+      "grad_norm": 1.2495671510696411,
+      "learning_rate": 0.00010842068287199663,
+      "loss": 5.1091,
+      "step": 25730
+    },
+    {
+      "epoch": 0.03616711483304851,
+      "grad_norm": 1.2367503643035889,
+      "learning_rate": 0.00010846283546438105,
+      "loss": 5.3419,
+      "step": 25740
+    },
+    {
+      "epoch": 0.03618116577121209,
+      "grad_norm": 1.247799038887024,
+      "learning_rate": 0.00010850498805676548,
+      "loss": 5.2247,
+      "step": 25750
+    },
+    {
+      "epoch": 0.036195216709375665,
+      "grad_norm": 1.3112616539001465,
+      "learning_rate": 0.00010854714064914992,
+      "loss": 5.3467,
+      "step": 25760
+    },
+    {
+      "epoch": 0.03620926764753924,
+      "grad_norm": 1.2683191299438477,
+      "learning_rate": 0.00010858929324153435,
+      "loss": 5.2628,
+      "step": 25770
+    },
+    {
+      "epoch": 0.03622331858570282,
+      "grad_norm": 1.295248031616211,
+      "learning_rate": 0.00010863144583391877,
+      "loss": 5.2255,
+      "step": 25780
+    },
+    {
+      "epoch": 0.0362373695238664,
+      "grad_norm": 1.3652722835540771,
+      "learning_rate": 0.00010867359842630322,
+      "loss": 5.1288,
+      "step": 25790
+    },
+    {
+      "epoch": 0.036251420462029976,
+      "grad_norm": 1.3258819580078125,
+      "learning_rate": 0.00010871575101868765,
+      "loss": 5.2706,
+      "step": 25800
+    },
+    {
+      "epoch": 0.03626547140019355,
+      "grad_norm": 1.213344931602478,
+      "learning_rate": 0.00010875790361107206,
+      "loss": 5.2729,
+      "step": 25810
+    },
+    {
+      "epoch": 0.03627952233835713,
+      "grad_norm": 1.3529800176620483,
+      "learning_rate": 0.00010880005620345651,
+      "loss": 5.1662,
+      "step": 25820
+    },
+    {
+      "epoch": 0.036293573276520705,
+      "grad_norm": 1.239091157913208,
+      "learning_rate": 0.00010884220879584094,
+      "loss": 5.1821,
+      "step": 25830
+    },
+    {
+      "epoch": 0.03630762421468428,
+      "grad_norm": 1.2374707460403442,
+      "learning_rate": 0.00010888436138822536,
+      "loss": 5.326,
+      "step": 25840
+    },
+    {
+      "epoch": 0.036321675152847864,
+      "grad_norm": 1.2616337537765503,
+      "learning_rate": 0.0001089265139806098,
+      "loss": 5.2543,
+      "step": 25850
+    },
+    {
+      "epoch": 0.03633572609101144,
+      "grad_norm": 1.256049633026123,
+      "learning_rate": 0.00010896866657299423,
+      "loss": 5.1544,
+      "step": 25860
+    },
+    {
+      "epoch": 0.036349777029175016,
+      "grad_norm": 1.4155590534210205,
+      "learning_rate": 0.00010901081916537865,
+      "loss": 5.3104,
+      "step": 25870
+    },
+    {
+      "epoch": 0.03636382796733859,
+      "grad_norm": 1.2644518613815308,
+      "learning_rate": 0.0001090529717577631,
+      "loss": 5.198,
+      "step": 25880
+    },
+    {
+      "epoch": 0.03637787890550217,
+      "grad_norm": 1.264570951461792,
+      "learning_rate": 0.00010909512435014753,
+      "loss": 5.0051,
+      "step": 25890
+    },
+    {
+      "epoch": 0.03639192984366575,
+      "grad_norm": 1.208848237991333,
+      "learning_rate": 0.00010913727694253195,
+      "loss": 5.2239,
+      "step": 25900
+    },
+    {
+      "epoch": 0.03640598078182933,
+      "grad_norm": 1.2104336023330688,
+      "learning_rate": 0.00010917942953491639,
+      "loss": 5.2732,
+      "step": 25910
+    },
+    {
+      "epoch": 0.036420031719992904,
+      "grad_norm": 1.298794150352478,
+      "learning_rate": 0.00010922158212730082,
+      "loss": 5.1825,
+      "step": 25920
+    },
+    {
+      "epoch": 0.03643408265815648,
+      "grad_norm": 1.3972251415252686,
+      "learning_rate": 0.00010926373471968525,
+      "loss": 5.4068,
+      "step": 25930
+    },
+    {
+      "epoch": 0.036448133596320056,
+      "grad_norm": 1.3141639232635498,
+      "learning_rate": 0.00010930588731206968,
+      "loss": 5.3439,
+      "step": 25940
+    },
+    {
+      "epoch": 0.03646218453448364,
+      "grad_norm": 1.4204273223876953,
+      "learning_rate": 0.00010934803990445411,
+      "loss": 5.1509,
+      "step": 25950
+    },
+    {
+      "epoch": 0.036476235472647216,
+      "grad_norm": 1.2282731533050537,
+      "learning_rate": 0.00010939019249683855,
+      "loss": 5.1782,
+      "step": 25960
+    },
+    {
+      "epoch": 0.03649028641081079,
+      "grad_norm": 1.2781927585601807,
+      "learning_rate": 0.00010943234508922298,
+      "loss": 5.1435,
+      "step": 25970
+    },
+    {
+      "epoch": 0.03650433734897437,
+      "grad_norm": 1.2722803354263306,
+      "learning_rate": 0.00010947449768160741,
+      "loss": 5.1179,
+      "step": 25980
+    },
+    {
+      "epoch": 0.036518388287137944,
+      "grad_norm": 1.3081459999084473,
+      "learning_rate": 0.00010951665027399185,
+      "loss": 5.0402,
+      "step": 25990
+    },
+    {
+      "epoch": 0.03653243922530153,
+      "grad_norm": 1.235568881034851,
+      "learning_rate": 0.00010955880286637627,
+      "loss": 5.22,
+      "step": 26000
+    },
+    {
+      "epoch": 0.036546490163465103,
+      "grad_norm": 1.3073358535766602,
+      "learning_rate": 0.0001096009554587607,
+      "loss": 5.2479,
+      "step": 26010
+    },
+    {
+      "epoch": 0.03656054110162868,
+      "grad_norm": 1.6806720495224,
+      "learning_rate": 0.00010964310805114515,
+      "loss": 5.1738,
+      "step": 26020
+    },
+    {
+      "epoch": 0.036574592039792256,
+      "grad_norm": 1.2512214183807373,
+      "learning_rate": 0.00010968526064352956,
+      "loss": 5.2327,
+      "step": 26030
+    },
+    {
+      "epoch": 0.03658864297795583,
+      "grad_norm": 1.2663812637329102,
+      "learning_rate": 0.000109727413235914,
+      "loss": 5.1307,
+      "step": 26040
+    },
+    {
+      "epoch": 0.036602693916119415,
+      "grad_norm": 1.2181764841079712,
+      "learning_rate": 0.00010976956582829844,
+      "loss": 5.2383,
+      "step": 26050
+    },
+    {
+      "epoch": 0.03661674485428299,
+      "grad_norm": 1.291024088859558,
+      "learning_rate": 0.00010981171842068287,
+      "loss": 5.1706,
+      "step": 26060
+    },
+    {
+      "epoch": 0.03663079579244657,
+      "grad_norm": 1.2037990093231201,
+      "learning_rate": 0.00010985387101306729,
+      "loss": 5.1271,
+      "step": 26070
+    },
+    {
+      "epoch": 0.036644846730610144,
+      "grad_norm": 1.1984660625457764,
+      "learning_rate": 0.00010989602360545173,
+      "loss": 5.291,
+      "step": 26080
+    },
+    {
+      "epoch": 0.03665889766877372,
+      "grad_norm": 1.301593542098999,
+      "learning_rate": 0.00010993817619783616,
+      "loss": 5.1974,
+      "step": 26090
+    },
+    {
+      "epoch": 0.0366729486069373,
+      "grad_norm": 1.249263882637024,
+      "learning_rate": 0.00010998032879022058,
+      "loss": 5.1413,
+      "step": 26100
+    },
+    {
+      "epoch": 0.03668699954510088,
+      "grad_norm": 1.3108259439468384,
+      "learning_rate": 0.00011002248138260503,
+      "loss": 5.3095,
+      "step": 26110
+    },
+    {
+      "epoch": 0.036701050483264455,
+      "grad_norm": 1.3736425638198853,
+      "learning_rate": 0.00011006463397498946,
+      "loss": 5.2484,
+      "step": 26120
+    },
+    {
+      "epoch": 0.03671510142142803,
+      "grad_norm": 1.2452633380889893,
+      "learning_rate": 0.00011010678656737388,
+      "loss": 5.2032,
+      "step": 26130
+    },
+    {
+      "epoch": 0.03672915235959161,
+      "grad_norm": 1.235262393951416,
+      "learning_rate": 0.00011014893915975832,
+      "loss": 5.2425,
+      "step": 26140
+    },
+    {
+      "epoch": 0.036743203297755184,
+      "grad_norm": 1.2423274517059326,
+      "learning_rate": 0.00011019109175214275,
+      "loss": 5.1752,
+      "step": 26150
+    },
+    {
+      "epoch": 0.03675725423591877,
+      "grad_norm": 1.3279187679290771,
+      "learning_rate": 0.00011023324434452717,
+      "loss": 5.232,
+      "step": 26160
+    },
+    {
+      "epoch": 0.03677130517408234,
+      "grad_norm": 1.2280527353286743,
+      "learning_rate": 0.00011027539693691161,
+      "loss": 5.3038,
+      "step": 26170
+    },
+    {
+      "epoch": 0.03678535611224592,
+      "grad_norm": 1.324863314628601,
+      "learning_rate": 0.00011031754952929604,
+      "loss": 5.131,
+      "step": 26180
+    },
+    {
+      "epoch": 0.036799407050409495,
+      "grad_norm": 1.3295210599899292,
+      "learning_rate": 0.00011035970212168046,
+      "loss": 5.2175,
+      "step": 26190
+    },
+    {
+      "epoch": 0.03681345798857307,
+      "grad_norm": 1.2481762170791626,
+      "learning_rate": 0.00011040185471406491,
+      "loss": 5.1572,
+      "step": 26200
+    },
+    {
+      "epoch": 0.036827508926736655,
+      "grad_norm": 1.2397472858428955,
+      "learning_rate": 0.00011044400730644934,
+      "loss": 5.1869,
+      "step": 26210
+    },
+    {
+      "epoch": 0.03684155986490023,
+      "grad_norm": 1.243333339691162,
+      "learning_rate": 0.00011048615989883377,
+      "loss": 5.1236,
+      "step": 26220
+    },
+    {
+      "epoch": 0.03685561080306381,
+      "grad_norm": 1.2801663875579834,
+      "learning_rate": 0.0001105283124912182,
+      "loss": 5.2453,
+      "step": 26230
+    },
+    {
+      "epoch": 0.03686966174122738,
+      "grad_norm": 1.3085575103759766,
+      "learning_rate": 0.00011057046508360263,
+      "loss": 5.131,
+      "step": 26240
+    },
+    {
+      "epoch": 0.03688371267939096,
+      "grad_norm": 1.2628742456436157,
+      "learning_rate": 0.00011061261767598706,
+      "loss": 5.2099,
+      "step": 26250
+    },
+    {
+      "epoch": 0.03689776361755454,
+      "grad_norm": 1.223900556564331,
+      "learning_rate": 0.0001106547702683715,
+      "loss": 5.234,
+      "step": 26260
+    },
+    {
+      "epoch": 0.03691181455571812,
+      "grad_norm": 1.3309193849563599,
+      "learning_rate": 0.00011069692286075593,
+      "loss": 5.3202,
+      "step": 26270
+    },
+    {
+      "epoch": 0.036925865493881695,
+      "grad_norm": 1.4030629396438599,
+      "learning_rate": 0.00011073907545314037,
+      "loss": 5.3066,
+      "step": 26280
+    },
+    {
+      "epoch": 0.03693991643204527,
+      "grad_norm": 1.2538514137268066,
+      "learning_rate": 0.00011078122804552479,
+      "loss": 5.157,
+      "step": 26290
+    },
+    {
+      "epoch": 0.03695396737020885,
+      "grad_norm": 1.26717209815979,
+      "learning_rate": 0.00011082338063790922,
+      "loss": 5.1852,
+      "step": 26300
+    },
+    {
+      "epoch": 0.03696801830837243,
+      "grad_norm": 1.2575275897979736,
+      "learning_rate": 0.00011086553323029366,
+      "loss": 5.3232,
+      "step": 26310
+    },
+    {
+      "epoch": 0.036982069246536006,
+      "grad_norm": 1.3039658069610596,
+      "learning_rate": 0.00011090768582267808,
+      "loss": 5.1781,
+      "step": 26320
+    },
+    {
+      "epoch": 0.03699612018469958,
+      "grad_norm": 1.3649189472198486,
+      "learning_rate": 0.00011094983841506251,
+      "loss": 5.2759,
+      "step": 26330
+    },
+    {
+      "epoch": 0.03701017112286316,
+      "grad_norm": 1.210462212562561,
+      "learning_rate": 0.00011099199100744696,
+      "loss": 5.0839,
+      "step": 26340
+    },
+    {
+      "epoch": 0.037024222061026735,
+      "grad_norm": 1.2349313497543335,
+      "learning_rate": 0.00011103414359983139,
+      "loss": 5.1527,
+      "step": 26350
+    },
+    {
+      "epoch": 0.03703827299919032,
+      "grad_norm": 1.326310634613037,
+      "learning_rate": 0.0001110762961922158,
+      "loss": 5.1428,
+      "step": 26360
+    },
+    {
+      "epoch": 0.037052323937353894,
+      "grad_norm": 1.2547760009765625,
+      "learning_rate": 0.00011111844878460025,
+      "loss": 5.1292,
+      "step": 26370
+    },
+    {
+      "epoch": 0.03706637487551747,
+      "grad_norm": 1.3274054527282715,
+      "learning_rate": 0.00011116060137698468,
+      "loss": 5.262,
+      "step": 26380
+    },
+    {
+      "epoch": 0.037080425813681046,
+      "grad_norm": 1.247420072555542,
+      "learning_rate": 0.0001112027539693691,
+      "loss": 5.3095,
+      "step": 26390
+    },
+    {
+      "epoch": 0.03709447675184462,
+      "grad_norm": 1.4132479429244995,
+      "learning_rate": 0.00011124490656175354,
+      "loss": 5.2025,
+      "step": 26400
+    },
+    {
+      "epoch": 0.037108527690008206,
+      "grad_norm": 1.262511968612671,
+      "learning_rate": 0.00011128705915413798,
+      "loss": 5.1409,
+      "step": 26410
+    },
+    {
+      "epoch": 0.03712257862817178,
+      "grad_norm": 1.2406529188156128,
+      "learning_rate": 0.00011132921174652239,
+      "loss": 5.1738,
+      "step": 26420
+    },
+    {
+      "epoch": 0.03713662956633536,
+      "grad_norm": 1.169571042060852,
+      "learning_rate": 0.00011137136433890684,
+      "loss": 5.2443,
+      "step": 26430
+    },
+    {
+      "epoch": 0.037150680504498934,
+      "grad_norm": 1.268477201461792,
+      "learning_rate": 0.00011141351693129127,
+      "loss": 5.1545,
+      "step": 26440
+    },
+    {
+      "epoch": 0.03716473144266251,
+      "grad_norm": 1.2400575876235962,
+      "learning_rate": 0.00011145566952367569,
+      "loss": 5.2186,
+      "step": 26450
+    },
+    {
+      "epoch": 0.037178782380826086,
+      "grad_norm": 1.244472622871399,
+      "learning_rate": 0.00011149782211606013,
+      "loss": 5.1656,
+      "step": 26460
+    },
+    {
+      "epoch": 0.03719283331898967,
+      "grad_norm": 1.3057937622070312,
+      "learning_rate": 0.00011153997470844456,
+      "loss": 5.0956,
+      "step": 26470
+    },
+    {
+      "epoch": 0.037206884257153246,
+      "grad_norm": 1.222322702407837,
+      "learning_rate": 0.00011158212730082898,
+      "loss": 5.2298,
+      "step": 26480
+    },
+    {
+      "epoch": 0.03722093519531682,
+      "grad_norm": 1.2240188121795654,
+      "learning_rate": 0.00011162427989321342,
+      "loss": 5.1473,
+      "step": 26490
+    },
+    {
+      "epoch": 0.0372349861334804,
+      "grad_norm": 1.2890682220458984,
+      "learning_rate": 0.00011166643248559786,
+      "loss": 5.0835,
+      "step": 26500
+    },
+    {
+      "epoch": 0.037249037071643974,
+      "grad_norm": 1.3182138204574585,
+      "learning_rate": 0.00011170858507798229,
+      "loss": 5.3088,
+      "step": 26510
+    },
+    {
+      "epoch": 0.03726308800980756,
+      "grad_norm": 1.2409218549728394,
+      "learning_rate": 0.00011175073767036672,
+      "loss": 5.2822,
+      "step": 26520
+    },
+    {
+      "epoch": 0.037277138947971133,
+      "grad_norm": 1.2367993593215942,
+      "learning_rate": 0.00011179289026275115,
+      "loss": 5.1691,
+      "step": 26530
+    },
+    {
+      "epoch": 0.03729118988613471,
+      "grad_norm": 1.2489789724349976,
+      "learning_rate": 0.00011183504285513558,
+      "loss": 5.2188,
+      "step": 26540
+    },
+    {
+      "epoch": 0.037305240824298286,
+      "grad_norm": 1.2142417430877686,
+      "learning_rate": 0.00011187719544752001,
+      "loss": 5.2871,
+      "step": 26550
+    },
+    {
+      "epoch": 0.03731929176246186,
+      "grad_norm": 1.2378629446029663,
+      "learning_rate": 0.00011191934803990444,
+      "loss": 5.0973,
+      "step": 26560
+    },
+    {
+      "epoch": 0.037333342700625445,
+      "grad_norm": 1.2806910276412964,
+      "learning_rate": 0.00011196150063228889,
+      "loss": 5.3434,
+      "step": 26570
+    },
+    {
+      "epoch": 0.03734739363878902,
+      "grad_norm": 1.3529186248779297,
+      "learning_rate": 0.0001120036532246733,
+      "loss": 5.1493,
+      "step": 26580
+    },
+    {
+      "epoch": 0.0373614445769526,
+      "grad_norm": 1.2549669742584229,
+      "learning_rate": 0.00011204580581705774,
+      "loss": 5.0873,
+      "step": 26590
+    },
+    {
+      "epoch": 0.037375495515116174,
+      "grad_norm": 1.2252702713012695,
+      "learning_rate": 0.00011208795840944218,
+      "loss": 5.1133,
+      "step": 26600
+    },
+    {
+      "epoch": 0.03738954645327975,
+      "grad_norm": 1.2540626525878906,
+      "learning_rate": 0.0001121301110018266,
+      "loss": 5.2449,
+      "step": 26610
+    },
+    {
+      "epoch": 0.03740359739144333,
+      "grad_norm": 1.178832769393921,
+      "learning_rate": 0.00011217226359421103,
+      "loss": 5.1955,
+      "step": 26620
+    },
+    {
+      "epoch": 0.03741764832960691,
+      "grad_norm": 1.2755051851272583,
+      "learning_rate": 0.00011221441618659547,
+      "loss": 5.0336,
+      "step": 26630
+    },
+    {
+      "epoch": 0.037431699267770485,
+      "grad_norm": 1.2864094972610474,
+      "learning_rate": 0.0001122565687789799,
+      "loss": 5.2242,
+      "step": 26640
+    },
+    {
+      "epoch": 0.03744575020593406,
+      "grad_norm": 1.2537206411361694,
+      "learning_rate": 0.00011229872137136432,
+      "loss": 5.3039,
+      "step": 26650
+    },
+    {
+      "epoch": 0.03745980114409764,
+      "grad_norm": 1.2298046350479126,
+      "learning_rate": 0.00011234087396374877,
+      "loss": 5.1417,
+      "step": 26660
+    },
+    {
+      "epoch": 0.03747385208226122,
+      "grad_norm": 1.2369047403335571,
+      "learning_rate": 0.0001123830265561332,
+      "loss": 5.2243,
+      "step": 26670
+    },
+    {
+      "epoch": 0.0374879030204248,
+      "grad_norm": 1.2635163068771362,
+      "learning_rate": 0.00011242517914851762,
+      "loss": 5.2349,
+      "step": 26680
+    },
+    {
+      "epoch": 0.03750195395858837,
+      "grad_norm": 1.3052091598510742,
+      "learning_rate": 0.00011246733174090206,
+      "loss": 5.1844,
+      "step": 26690
+    },
+    {
+      "epoch": 0.03751600489675195,
+      "grad_norm": 1.2509127855300903,
+      "learning_rate": 0.00011250948433328649,
+      "loss": 5.1712,
+      "step": 26700
+    },
+    {
+      "epoch": 0.037530055834915525,
+      "grad_norm": 1.2266826629638672,
+      "learning_rate": 0.00011255163692567091,
+      "loss": 5.1993,
+      "step": 26710
+    },
+    {
+      "epoch": 0.03754410677307911,
+      "grad_norm": 1.3366031646728516,
+      "learning_rate": 0.00011259378951805535,
+      "loss": 5.178,
+      "step": 26720
+    },
+    {
+      "epoch": 0.037558157711242685,
+      "grad_norm": 1.2389544248580933,
+      "learning_rate": 0.00011263594211043979,
+      "loss": 5.282,
+      "step": 26730
+    },
+    {
+      "epoch": 0.03757220864940626,
+      "grad_norm": 1.2713171243667603,
+      "learning_rate": 0.0001126780947028242,
+      "loss": 5.1055,
+      "step": 26740
+    },
+    {
+      "epoch": 0.03758625958756984,
+      "grad_norm": 1.268943190574646,
+      "learning_rate": 0.00011272024729520865,
+      "loss": 5.2751,
+      "step": 26750
+    },
+    {
+      "epoch": 0.03760031052573341,
+      "grad_norm": 1.302060842514038,
+      "learning_rate": 0.00011276239988759308,
+      "loss": 5.1239,
+      "step": 26760
+    },
+    {
+      "epoch": 0.03761436146389699,
+      "grad_norm": 1.2425678968429565,
+      "learning_rate": 0.00011280455247997751,
+      "loss": 5.2777,
+      "step": 26770
+    },
+    {
+      "epoch": 0.03762841240206057,
+      "grad_norm": 1.297904133796692,
+      "learning_rate": 0.00011284670507236194,
+      "loss": 5.3101,
+      "step": 26780
+    },
+    {
+      "epoch": 0.03764246334022415,
+      "grad_norm": 1.3182086944580078,
+      "learning_rate": 0.00011288885766474637,
+      "loss": 5.1886,
+      "step": 26790
+    },
+    {
+      "epoch": 0.037656514278387725,
+      "grad_norm": 1.1934064626693726,
+      "learning_rate": 0.0001129310102571308,
+      "loss": 5.1946,
+      "step": 26800
+    },
+    {
+      "epoch": 0.0376705652165513,
+      "grad_norm": 1.2613691091537476,
+      "learning_rate": 0.00011297316284951524,
+      "loss": 5.2152,
+      "step": 26810
+    },
+    {
+      "epoch": 0.03768461615471488,
+      "grad_norm": 1.2664833068847656,
+      "learning_rate": 0.00011301531544189967,
+      "loss": 5.3019,
+      "step": 26820
+    },
+    {
+      "epoch": 0.03769866709287846,
+      "grad_norm": 1.687343955039978,
+      "learning_rate": 0.0001130574680342841,
+      "loss": 5.081,
+      "step": 26830
+    },
+    {
+      "epoch": 0.037712718031042036,
+      "grad_norm": 1.2772016525268555,
+      "learning_rate": 0.00011309962062666853,
+      "loss": 5.1301,
+      "step": 26840
+    },
+    {
+      "epoch": 0.03772676896920561,
+      "grad_norm": 1.2881845235824585,
+      "learning_rate": 0.00011314177321905296,
+      "loss": 5.2073,
+      "step": 26850
+    },
+    {
+      "epoch": 0.03774081990736919,
+      "grad_norm": 1.4994240999221802,
+      "learning_rate": 0.0001131839258114374,
+      "loss": 5.2521,
+      "step": 26860
+    },
+    {
+      "epoch": 0.037754870845532765,
+      "grad_norm": 1.3068636655807495,
+      "learning_rate": 0.00011322607840382182,
+      "loss": 5.1688,
+      "step": 26870
+    },
+    {
+      "epoch": 0.03776892178369635,
+      "grad_norm": 1.2948259115219116,
+      "learning_rate": 0.00011326823099620625,
+      "loss": 5.1923,
+      "step": 26880
+    },
+    {
+      "epoch": 0.037782972721859924,
+      "grad_norm": 1.2383815050125122,
+      "learning_rate": 0.0001133103835885907,
+      "loss": 5.205,
+      "step": 26890
+    },
+    {
+      "epoch": 0.0377970236600235,
+      "grad_norm": 1.258082628250122,
+      "learning_rate": 0.00011335253618097512,
+      "loss": 5.1863,
+      "step": 26900
+    },
+    {
+      "epoch": 0.037811074598187076,
+      "grad_norm": 1.2574599981307983,
+      "learning_rate": 0.00011339468877335955,
+      "loss": 5.1459,
+      "step": 26910
+    },
+    {
+      "epoch": 0.03782512553635065,
+      "grad_norm": 1.3260610103607178,
+      "learning_rate": 0.00011343684136574399,
+      "loss": 5.2764,
+      "step": 26920
+    },
+    {
+      "epoch": 0.037839176474514236,
+      "grad_norm": 1.1936684846878052,
+      "learning_rate": 0.00011347899395812842,
+      "loss": 5.2587,
+      "step": 26930
+    },
+    {
+      "epoch": 0.03785322741267781,
+      "grad_norm": 1.3640984296798706,
+      "learning_rate": 0.00011352114655051284,
+      "loss": 5.0741,
+      "step": 26940
+    },
+    {
+      "epoch": 0.03786727835084139,
+      "grad_norm": 1.3140736818313599,
+      "learning_rate": 0.00011356329914289729,
+      "loss": 5.1392,
+      "step": 26950
+    },
+    {
+      "epoch": 0.037881329289004964,
+      "grad_norm": 1.268675446510315,
+      "learning_rate": 0.00011360545173528172,
+      "loss": 5.1514,
+      "step": 26960
+    },
+    {
+      "epoch": 0.03789538022716854,
+      "grad_norm": 1.4075084924697876,
+      "learning_rate": 0.00011364760432766613,
+      "loss": 5.1024,
+      "step": 26970
+    },
+    {
+      "epoch": 0.03790943116533212,
+      "grad_norm": 1.2317532300949097,
+      "learning_rate": 0.00011368975692005058,
+      "loss": 5.2554,
+      "step": 26980
+    },
+    {
+      "epoch": 0.0379234821034957,
+      "grad_norm": 1.3724669218063354,
+      "learning_rate": 0.00011373190951243501,
+      "loss": 5.1995,
+      "step": 26990
+    },
+    {
+      "epoch": 0.037937533041659276,
+      "grad_norm": 1.2524470090866089,
+      "learning_rate": 0.00011377406210481943,
+      "loss": 5.2892,
+      "step": 27000
+    },
+    {
+      "epoch": 0.03795158397982285,
+      "grad_norm": 1.231868863105774,
+      "learning_rate": 0.00011381621469720387,
+      "loss": 5.2432,
+      "step": 27010
+    },
+    {
+      "epoch": 0.03796563491798643,
+      "grad_norm": 1.224562406539917,
+      "learning_rate": 0.0001138583672895883,
+      "loss": 5.3197,
+      "step": 27020
+    },
+    {
+      "epoch": 0.03797968585615001,
+      "grad_norm": 1.390925407409668,
+      "learning_rate": 0.00011390051988197272,
+      "loss": 5.2181,
+      "step": 27030
+    },
+    {
+      "epoch": 0.03799373679431359,
+      "grad_norm": 1.3121964931488037,
+      "learning_rate": 0.00011394267247435717,
+      "loss": 5.1817,
+      "step": 27040
+    },
+    {
+      "epoch": 0.03800778773247716,
+      "grad_norm": 1.2776087522506714,
+      "learning_rate": 0.0001139848250667416,
+      "loss": 5.2207,
+      "step": 27050
+    },
+    {
+      "epoch": 0.03802183867064074,
+      "grad_norm": 1.2448629140853882,
+      "learning_rate": 0.00011402697765912603,
+      "loss": 5.1486,
+      "step": 27060
+    },
+    {
+      "epoch": 0.038035889608804316,
+      "grad_norm": 1.22585129737854,
+      "learning_rate": 0.00011406913025151046,
+      "loss": 5.1954,
+      "step": 27070
+    },
+    {
+      "epoch": 0.03804994054696789,
+      "grad_norm": 1.289219856262207,
+      "learning_rate": 0.00011411128284389489,
+      "loss": 5.1862,
+      "step": 27080
+    },
+    {
+      "epoch": 0.038063991485131475,
+      "grad_norm": 1.2818207740783691,
+      "learning_rate": 0.00011415343543627932,
+      "loss": 5.2999,
+      "step": 27090
+    },
+    {
+      "epoch": 0.03807804242329505,
+      "grad_norm": 1.237623691558838,
+      "learning_rate": 0.00011419558802866375,
+      "loss": 5.0843,
+      "step": 27100
+    },
+    {
+      "epoch": 0.03809209336145863,
+      "grad_norm": 1.233920931816101,
+      "learning_rate": 0.00011423774062104818,
+      "loss": 5.1438,
+      "step": 27110
+    },
+    {
+      "epoch": 0.038106144299622204,
+      "grad_norm": 1.782360315322876,
+      "learning_rate": 0.00011427989321343262,
+      "loss": 5.1636,
+      "step": 27120
+    },
+    {
+      "epoch": 0.03812019523778578,
+      "grad_norm": 1.2615609169006348,
+      "learning_rate": 0.00011432204580581705,
+      "loss": 5.1292,
+      "step": 27130
+    },
+    {
+      "epoch": 0.03813424617594936,
+      "grad_norm": 1.2505155801773071,
+      "learning_rate": 0.00011436419839820148,
+      "loss": 5.1816,
+      "step": 27140
+    },
+    {
+      "epoch": 0.03814829711411294,
+      "grad_norm": 1.231854796409607,
+      "learning_rate": 0.00011440635099058592,
+      "loss": 5.1576,
+      "step": 27150
+    },
+    {
+      "epoch": 0.038162348052276515,
+      "grad_norm": 1.2065684795379639,
+      "learning_rate": 0.00011444850358297034,
+      "loss": 5.1434,
+      "step": 27160
+    },
+    {
+      "epoch": 0.03817639899044009,
+      "grad_norm": 1.2832350730895996,
+      "learning_rate": 0.00011449065617535477,
+      "loss": 5.1549,
+      "step": 27170
+    },
+    {
+      "epoch": 0.03819044992860367,
+      "grad_norm": 1.2158983945846558,
+      "learning_rate": 0.00011453280876773922,
+      "loss": 5.1136,
+      "step": 27180
+    },
+    {
+      "epoch": 0.03820450086676725,
+      "grad_norm": 1.2421070337295532,
+      "learning_rate": 0.00011457496136012363,
+      "loss": 5.1758,
+      "step": 27190
+    },
+    {
+      "epoch": 0.03821855180493083,
+      "grad_norm": 1.2263178825378418,
+      "learning_rate": 0.00011461711395250806,
+      "loss": 5.1627,
+      "step": 27200
+    },
+    {
+      "epoch": 0.0382326027430944,
+      "grad_norm": 1.2281807661056519,
+      "learning_rate": 0.00011465926654489251,
+      "loss": 5.1769,
+      "step": 27210
+    },
+    {
+      "epoch": 0.03824665368125798,
+      "grad_norm": 1.2824562788009644,
+      "learning_rate": 0.00011470141913727694,
+      "loss": 5.0437,
+      "step": 27220
+    },
+    {
+      "epoch": 0.038260704619421555,
+      "grad_norm": 1.2487444877624512,
+      "learning_rate": 0.00011474357172966136,
+      "loss": 5.1845,
+      "step": 27230
+    },
+    {
+      "epoch": 0.03827475555758514,
+      "grad_norm": 1.223210096359253,
+      "learning_rate": 0.0001147857243220458,
+      "loss": 5.1532,
+      "step": 27240
+    },
+    {
+      "epoch": 0.038288806495748715,
+      "grad_norm": 1.2843011617660522,
+      "learning_rate": 0.00011482787691443023,
+      "loss": 5.2197,
+      "step": 27250
+    },
+    {
+      "epoch": 0.03830285743391229,
+      "grad_norm": 1.2639187574386597,
+      "learning_rate": 0.00011487002950681465,
+      "loss": 5.1132,
+      "step": 27260
+    },
+    {
+      "epoch": 0.03831690837207587,
+      "grad_norm": 1.2122303247451782,
+      "learning_rate": 0.0001149121820991991,
+      "loss": 5.255,
+      "step": 27270
+    },
+    {
+      "epoch": 0.03833095931023944,
+      "grad_norm": 1.2370988130569458,
+      "learning_rate": 0.00011495433469158353,
+      "loss": 5.1655,
+      "step": 27280
+    },
+    {
+      "epoch": 0.038345010248403026,
+      "grad_norm": 1.2647359371185303,
+      "learning_rate": 0.00011499648728396794,
+      "loss": 5.2058,
+      "step": 27290
+    },
+    {
+      "epoch": 0.0383590611865666,
+      "grad_norm": 1.2288408279418945,
+      "learning_rate": 0.00011503863987635239,
+      "loss": 5.2224,
+      "step": 27300
+    },
+    {
+      "epoch": 0.03837311212473018,
+      "grad_norm": 1.3430143594741821,
+      "learning_rate": 0.00011508079246873682,
+      "loss": 5.1902,
+      "step": 27310
+    },
+    {
+      "epoch": 0.038387163062893755,
+      "grad_norm": 1.2298165559768677,
+      "learning_rate": 0.00011512294506112124,
+      "loss": 5.2141,
+      "step": 27320
+    },
+    {
+      "epoch": 0.03840121400105733,
+      "grad_norm": 1.2337199449539185,
+      "learning_rate": 0.00011516509765350568,
+      "loss": 5.1845,
+      "step": 27330
+    },
+    {
+      "epoch": 0.038415264939220914,
+      "grad_norm": 1.1939727067947388,
+      "learning_rate": 0.00011520725024589011,
+      "loss": 5.2187,
+      "step": 27340
+    },
+    {
+      "epoch": 0.03842931587738449,
+      "grad_norm": 1.3627872467041016,
+      "learning_rate": 0.00011524940283827455,
+      "loss": 5.0741,
+      "step": 27350
+    },
+    {
+      "epoch": 0.038443366815548066,
+      "grad_norm": 1.3677321672439575,
+      "learning_rate": 0.00011529155543065898,
+      "loss": 5.1668,
+      "step": 27360
+    },
+    {
+      "epoch": 0.03845741775371164,
+      "grad_norm": 1.2773394584655762,
+      "learning_rate": 0.00011533370802304341,
+      "loss": 5.1311,
+      "step": 27370
+    },
+    {
+      "epoch": 0.03847146869187522,
+      "grad_norm": 1.2625548839569092,
+      "learning_rate": 0.00011537586061542784,
+      "loss": 5.2052,
+      "step": 27380
+    },
+    {
+      "epoch": 0.038485519630038795,
+      "grad_norm": 1.227333426475525,
+      "learning_rate": 0.00011541801320781227,
+      "loss": 5.3206,
+      "step": 27390
+    },
+    {
+      "epoch": 0.03849957056820238,
+      "grad_norm": 1.214938759803772,
+      "learning_rate": 0.0001154601658001967,
+      "loss": 5.0788,
+      "step": 27400
+    },
+    {
+      "epoch": 0.038513621506365954,
+      "grad_norm": 1.2600798606872559,
+      "learning_rate": 0.00011550231839258113,
+      "loss": 5.2539,
+      "step": 27410
+    },
+    {
+      "epoch": 0.03852767244452953,
+      "grad_norm": 1.2314510345458984,
+      "learning_rate": 0.00011554447098496556,
+      "loss": 5.2616,
+      "step": 27420
+    },
+    {
+      "epoch": 0.038541723382693106,
+      "grad_norm": 1.2492339611053467,
+      "learning_rate": 0.00011558662357735,
+      "loss": 5.1835,
+      "step": 27430
+    },
+    {
+      "epoch": 0.03855577432085668,
+      "grad_norm": 1.3038430213928223,
+      "learning_rate": 0.00011562877616973444,
+      "loss": 5.1959,
+      "step": 27440
+    },
+    {
+      "epoch": 0.038569825259020266,
+      "grad_norm": 1.2819589376449585,
+      "learning_rate": 0.00011567092876211886,
+      "loss": 5.2702,
+      "step": 27450
+    },
+    {
+      "epoch": 0.03858387619718384,
+      "grad_norm": 1.2621688842773438,
+      "learning_rate": 0.00011571308135450329,
+      "loss": 5.2331,
+      "step": 27460
+    },
+    {
+      "epoch": 0.03859792713534742,
+      "grad_norm": 1.2393717765808105,
+      "learning_rate": 0.00011575523394688773,
+      "loss": 5.2205,
+      "step": 27470
+    },
+    {
+      "epoch": 0.038611978073510994,
+      "grad_norm": 1.407251238822937,
+      "learning_rate": 0.00011579738653927216,
+      "loss": 5.0831,
+      "step": 27480
+    },
+    {
+      "epoch": 0.03862602901167457,
+      "grad_norm": 1.196433663368225,
+      "learning_rate": 0.00011583953913165658,
+      "loss": 5.2502,
+      "step": 27490
+    },
+    {
+      "epoch": 0.03864007994983815,
+      "grad_norm": 1.2640856504440308,
+      "learning_rate": 0.00011588169172404103,
+      "loss": 5.1203,
+      "step": 27500
+    },
+    {
+      "epoch": 0.03865413088800173,
+      "grad_norm": 1.2602075338363647,
+      "learning_rate": 0.00011592384431642546,
+      "loss": 5.1076,
+      "step": 27510
+    },
+    {
+      "epoch": 0.038668181826165306,
+      "grad_norm": 1.226286768913269,
+      "learning_rate": 0.00011596599690880988,
+      "loss": 5.1437,
+      "step": 27520
+    },
+    {
+      "epoch": 0.03868223276432888,
+      "grad_norm": 1.2031055688858032,
+      "learning_rate": 0.00011600814950119432,
+      "loss": 5.1661,
+      "step": 27530
+    },
+    {
+      "epoch": 0.03869628370249246,
+      "grad_norm": 1.2152131795883179,
+      "learning_rate": 0.00011605030209357875,
+      "loss": 5.143,
+      "step": 27540
+    },
+    {
+      "epoch": 0.03871033464065604,
+      "grad_norm": 1.2977949380874634,
+      "learning_rate": 0.00011609245468596317,
+      "loss": 5.2466,
+      "step": 27550
+    },
+    {
+      "epoch": 0.03872438557881962,
+      "grad_norm": 1.2086836099624634,
+      "learning_rate": 0.00011613460727834761,
+      "loss": 5.2369,
+      "step": 27560
+    },
+    {
+      "epoch": 0.03873843651698319,
+      "grad_norm": 1.1972341537475586,
+      "learning_rate": 0.00011617675987073204,
+      "loss": 5.2176,
+      "step": 27570
+    },
+    {
+      "epoch": 0.03875248745514677,
+      "grad_norm": 1.3312931060791016,
+      "learning_rate": 0.00011621891246311646,
+      "loss": 4.9767,
+      "step": 27580
+    },
+    {
+      "epoch": 0.038766538393310346,
+      "grad_norm": 1.2363582849502563,
+      "learning_rate": 0.00011626106505550091,
+      "loss": 5.1467,
+      "step": 27590
+    },
+    {
+      "epoch": 0.03878058933147393,
+      "grad_norm": 1.2445141077041626,
+      "learning_rate": 0.00011630321764788534,
+      "loss": 4.9751,
+      "step": 27600
+    },
+    {
+      "epoch": 0.038794640269637505,
+      "grad_norm": 1.3403300046920776,
+      "learning_rate": 0.00011634537024026976,
+      "loss": 5.1634,
+      "step": 27610
+    },
+    {
+      "epoch": 0.03880869120780108,
+      "grad_norm": 1.2853689193725586,
+      "learning_rate": 0.0001163875228326542,
+      "loss": 5.2343,
+      "step": 27620
+    },
+    {
+      "epoch": 0.03882274214596466,
+      "grad_norm": 1.2790160179138184,
+      "learning_rate": 0.00011642967542503863,
+      "loss": 5.3136,
+      "step": 27630
+    },
+    {
+      "epoch": 0.038836793084128234,
+      "grad_norm": 1.3508522510528564,
+      "learning_rate": 0.00011647182801742306,
+      "loss": 5.0613,
+      "step": 27640
+    },
+    {
+      "epoch": 0.03885084402229182,
+      "grad_norm": 1.274295687675476,
+      "learning_rate": 0.0001165139806098075,
+      "loss": 5.1176,
+      "step": 27650
+    },
+    {
+      "epoch": 0.03886489496045539,
+      "grad_norm": 1.3735030889511108,
+      "learning_rate": 0.00011655613320219193,
+      "loss": 5.2095,
+      "step": 27660
+    },
+    {
+      "epoch": 0.03887894589861897,
+      "grad_norm": 1.2918239831924438,
+      "learning_rate": 0.00011659828579457636,
+      "loss": 5.2064,
+      "step": 27670
+    },
+    {
+      "epoch": 0.038892996836782545,
+      "grad_norm": 1.2781789302825928,
+      "learning_rate": 0.00011664043838696079,
+      "loss": 5.0818,
+      "step": 27680
+    },
+    {
+      "epoch": 0.03890704777494612,
+      "grad_norm": 1.2447351217269897,
+      "learning_rate": 0.00011668259097934522,
+      "loss": 5.085,
+      "step": 27690
+    },
+    {
+      "epoch": 0.0389210987131097,
+      "grad_norm": 1.2342867851257324,
+      "learning_rate": 0.00011672474357172965,
+      "loss": 5.1657,
+      "step": 27700
+    },
+    {
+      "epoch": 0.03893514965127328,
+      "grad_norm": 1.2419430017471313,
+      "learning_rate": 0.00011676689616411408,
+      "loss": 5.1151,
+      "step": 27710
+    },
+    {
+      "epoch": 0.03894920058943686,
+      "grad_norm": 1.261952519416809,
+      "learning_rate": 0.00011680904875649851,
+      "loss": 4.978,
+      "step": 27720
+    },
+    {
+      "epoch": 0.03896325152760043,
+      "grad_norm": 1.2440228462219238,
+      "learning_rate": 0.00011685120134888294,
+      "loss": 5.079,
+      "step": 27730
+    },
+    {
+      "epoch": 0.03897730246576401,
+      "grad_norm": 1.2461313009262085,
+      "learning_rate": 0.00011689335394126737,
+      "loss": 5.074,
+      "step": 27740
+    },
+    {
+      "epoch": 0.038991353403927585,
+      "grad_norm": 1.3326444625854492,
+      "learning_rate": 0.0001169355065336518,
+      "loss": 5.1748,
+      "step": 27750
+    },
+    {
+      "epoch": 0.03900540434209117,
+      "grad_norm": 1.1846364736557007,
+      "learning_rate": 0.00011697765912603625,
+      "loss": 5.1285,
+      "step": 27760
+    },
+    {
+      "epoch": 0.039019455280254745,
+      "grad_norm": 1.3407005071640015,
+      "learning_rate": 0.00011701981171842068,
+      "loss": 5.1625,
+      "step": 27770
+    },
+    {
+      "epoch": 0.03903350621841832,
+      "grad_norm": 1.1774260997772217,
+      "learning_rate": 0.0001170619643108051,
+      "loss": 5.1255,
+      "step": 27780
+    },
+    {
+      "epoch": 0.0390475571565819,
+      "grad_norm": 1.2277191877365112,
+      "learning_rate": 0.00011710411690318954,
+      "loss": 5.0722,
+      "step": 27790
+    },
+    {
+      "epoch": 0.03906160809474547,
+      "grad_norm": 1.2346806526184082,
+      "learning_rate": 0.00011714626949557398,
+      "loss": 5.3759,
+      "step": 27800
+    },
+    {
+      "epoch": 0.039075659032909056,
+      "grad_norm": 1.2467949390411377,
+      "learning_rate": 0.00011718842208795839,
+      "loss": 5.0743,
+      "step": 27810
+    },
+    {
+      "epoch": 0.03908970997107263,
+      "grad_norm": 1.2584148645401,
+      "learning_rate": 0.00011723057468034284,
+      "loss": 5.2156,
+      "step": 27820
+    },
+    {
+      "epoch": 0.03910376090923621,
+      "grad_norm": 1.2830735445022583,
+      "learning_rate": 0.00011727272727272727,
+      "loss": 5.1202,
+      "step": 27830
+    },
+    {
+      "epoch": 0.039117811847399785,
+      "grad_norm": 1.2006511688232422,
+      "learning_rate": 0.00011731487986511169,
+      "loss": 5.1867,
+      "step": 27840
+    },
+    {
+      "epoch": 0.03913186278556336,
+      "grad_norm": 1.2319871187210083,
+      "learning_rate": 0.00011735703245749613,
+      "loss": 5.1403,
+      "step": 27850
+    },
+    {
+      "epoch": 0.039145913723726944,
+      "grad_norm": 1.2313419580459595,
+      "learning_rate": 0.00011739918504988056,
+      "loss": 5.1664,
+      "step": 27860
+    },
+    {
+      "epoch": 0.03915996466189052,
+      "grad_norm": 1.2943781614303589,
+      "learning_rate": 0.00011744133764226498,
+      "loss": 5.0664,
+      "step": 27870
+    },
+    {
+      "epoch": 0.039174015600054096,
+      "grad_norm": 1.2293378114700317,
+      "learning_rate": 0.00011748349023464942,
+      "loss": 5.0583,
+      "step": 27880
+    },
+    {
+      "epoch": 0.03918806653821767,
+      "grad_norm": 1.1826175451278687,
+      "learning_rate": 0.00011752564282703386,
+      "loss": 5.1246,
+      "step": 27890
+    },
+    {
+      "epoch": 0.03920211747638125,
+      "grad_norm": 1.1859021186828613,
+      "learning_rate": 0.00011756779541941827,
+      "loss": 5.166,
+      "step": 27900
+    },
+    {
+      "epoch": 0.03921616841454483,
+      "grad_norm": 1.2054551839828491,
+      "learning_rate": 0.00011760994801180272,
+      "loss": 5.0331,
+      "step": 27910
+    },
+    {
+      "epoch": 0.03923021935270841,
+      "grad_norm": 1.2477535009384155,
+      "learning_rate": 0.0001176478853449487,
+      "loss": 5.1473,
+      "step": 27920
+    },
+    {
+      "epoch": 0.039244270290871984,
+      "grad_norm": 1.1738812923431396,
+      "learning_rate": 0.00011769003793733315,
+      "loss": 5.2345,
+      "step": 27930
+    },
+    {
+      "epoch": 0.03925832122903556,
+      "grad_norm": 1.2477853298187256,
+      "learning_rate": 0.00011773219052971756,
+      "loss": 5.1245,
+      "step": 27940
+    },
+    {
+      "epoch": 0.039272372167199136,
+      "grad_norm": 1.1837849617004395,
+      "learning_rate": 0.000117774343122102,
+      "loss": 5.2802,
+      "step": 27950
+    },
+    {
+      "epoch": 0.03928642310536272,
+      "grad_norm": 1.2485941648483276,
+      "learning_rate": 0.00011781649571448644,
+      "loss": 5.1342,
+      "step": 27960
+    },
+    {
+      "epoch": 0.039300474043526296,
+      "grad_norm": 1.2612384557724,
+      "learning_rate": 0.00011785864830687086,
+      "loss": 5.2221,
+      "step": 27970
+    },
+    {
+      "epoch": 0.03931452498168987,
+      "grad_norm": 1.2098948955535889,
+      "learning_rate": 0.00011790080089925529,
+      "loss": 5.0834,
+      "step": 27980
+    },
+    {
+      "epoch": 0.03932857591985345,
+      "grad_norm": 1.2792184352874756,
+      "learning_rate": 0.00011794295349163973,
+      "loss": 5.1232,
+      "step": 27990
+    },
+    {
+      "epoch": 0.039342626858017024,
+      "grad_norm": 1.1822669506072998,
+      "learning_rate": 0.00011798510608402415,
+      "loss": 5.1502,
+      "step": 28000
+    },
+    {
+      "epoch": 0.03935667779618061,
+      "grad_norm": 1.237909197807312,
+      "learning_rate": 0.00011802725867640858,
+      "loss": 5.0886,
+      "step": 28010
+    },
+    {
+      "epoch": 0.03937072873434418,
+      "grad_norm": 1.4007855653762817,
+      "learning_rate": 0.00011806941126879303,
+      "loss": 5.1382,
+      "step": 28020
+    },
+    {
+      "epoch": 0.03938477967250776,
+      "grad_norm": 1.2554928064346313,
+      "learning_rate": 0.00011811156386117744,
+      "loss": 5.1784,
+      "step": 28030
+    },
+    {
+      "epoch": 0.039398830610671336,
+      "grad_norm": 1.2145435810089111,
+      "learning_rate": 0.00011815371645356187,
+      "loss": 5.2392,
+      "step": 28040
+    },
+    {
+      "epoch": 0.03941288154883491,
+      "grad_norm": 1.194926381111145,
+      "learning_rate": 0.00011819586904594632,
+      "loss": 5.1088,
+      "step": 28050
+    },
+    {
+      "epoch": 0.03942693248699849,
+      "grad_norm": 1.2417168617248535,
+      "learning_rate": 0.00011823802163833075,
+      "loss": 5.2147,
+      "step": 28060
+    },
+    {
+      "epoch": 0.03944098342516207,
+      "grad_norm": 1.171698808670044,
+      "learning_rate": 0.00011828017423071517,
+      "loss": 5.1953,
+      "step": 28070
+    },
+    {
+      "epoch": 0.03945503436332565,
+      "grad_norm": 1.2243210077285767,
+      "learning_rate": 0.00011832232682309961,
+      "loss": 5.243,
+      "step": 28080
+    },
+    {
+      "epoch": 0.03946908530148922,
+      "grad_norm": 1.2159587144851685,
+      "learning_rate": 0.00011836447941548404,
+      "loss": 5.0568,
+      "step": 28090
+    },
+    {
+      "epoch": 0.0394831362396528,
+      "grad_norm": 1.2428697347640991,
+      "learning_rate": 0.00011840663200786846,
+      "loss": 5.0406,
+      "step": 28100
+    },
+    {
+      "epoch": 0.039497187177816376,
+      "grad_norm": 1.2648569345474243,
+      "learning_rate": 0.0001184487846002529,
+      "loss": 5.222,
+      "step": 28110
+    },
+    {
+      "epoch": 0.03951123811597996,
+      "grad_norm": 1.2005157470703125,
+      "learning_rate": 0.00011849093719263734,
+      "loss": 5.1258,
+      "step": 28120
+    },
+    {
+      "epoch": 0.039525289054143535,
+      "grad_norm": 1.2685343027114868,
+      "learning_rate": 0.00011853308978502177,
+      "loss": 5.064,
+      "step": 28130
+    },
+    {
+      "epoch": 0.03953933999230711,
+      "grad_norm": 1.2965656518936157,
+      "learning_rate": 0.0001185752423774062,
+      "loss": 5.0134,
+      "step": 28140
+    },
+    {
+      "epoch": 0.03955339093047069,
+      "grad_norm": 1.264532208442688,
+      "learning_rate": 0.00011861739496979063,
+      "loss": 5.2289,
+      "step": 28150
+    },
+    {
+      "epoch": 0.039567441868634264,
+      "grad_norm": 1.3054733276367188,
+      "learning_rate": 0.00011865954756217506,
+      "loss": 5.1761,
+      "step": 28160
+    },
+    {
+      "epoch": 0.03958149280679785,
+      "grad_norm": 1.1878447532653809,
+      "learning_rate": 0.0001187017001545595,
+      "loss": 5.1396,
+      "step": 28170
+    },
+    {
+      "epoch": 0.03959554374496142,
+      "grad_norm": 1.3018479347229004,
+      "learning_rate": 0.00011874385274694392,
+      "loss": 5.1258,
+      "step": 28180
+    },
+    {
+      "epoch": 0.039609594683125,
+      "grad_norm": 1.2443947792053223,
+      "learning_rate": 0.00011878600533932837,
+      "loss": 5.1624,
+      "step": 28190
+    },
+    {
+      "epoch": 0.039623645621288575,
+      "grad_norm": 1.3728779554367065,
+      "learning_rate": 0.00011882815793171279,
+      "loss": 5.2995,
+      "step": 28200
+    },
+    {
+      "epoch": 0.03963769655945215,
+      "grad_norm": 1.1758084297180176,
+      "learning_rate": 0.00011887031052409722,
+      "loss": 5.1263,
+      "step": 28210
+    },
+    {
+      "epoch": 0.039651747497615734,
+      "grad_norm": 1.2095502614974976,
+      "learning_rate": 0.00011891246311648166,
+      "loss": 5.0737,
+      "step": 28220
+    },
+    {
+      "epoch": 0.03966579843577931,
+      "grad_norm": 1.2369705438613892,
+      "learning_rate": 0.00011895461570886608,
+      "loss": 5.2049,
+      "step": 28230
+    },
+    {
+      "epoch": 0.03967984937394289,
+      "grad_norm": 1.2536842823028564,
+      "learning_rate": 0.00011899676830125051,
+      "loss": 5.1596,
+      "step": 28240
+    },
+    {
+      "epoch": 0.03969390031210646,
+      "grad_norm": 1.2963687181472778,
+      "learning_rate": 0.00011903892089363496,
+      "loss": 5.0851,
+      "step": 28250
+    },
+    {
+      "epoch": 0.03970795125027004,
+      "grad_norm": 1.3036259412765503,
+      "learning_rate": 0.00011908107348601937,
+      "loss": 5.2885,
+      "step": 28260
+    },
+    {
+      "epoch": 0.03972200218843362,
+      "grad_norm": 1.2262283563613892,
+      "learning_rate": 0.0001191232260784038,
+      "loss": 5.2086,
+      "step": 28270
+    },
+    {
+      "epoch": 0.0397360531265972,
+      "grad_norm": 1.234108805656433,
+      "learning_rate": 0.00011916537867078825,
+      "loss": 5.185,
+      "step": 28280
+    },
+    {
+      "epoch": 0.039750104064760775,
+      "grad_norm": 1.2334399223327637,
+      "learning_rate": 0.00011920753126317267,
+      "loss": 5.1078,
+      "step": 28290
+    },
+    {
+      "epoch": 0.03976415500292435,
+      "grad_norm": 1.2662261724472046,
+      "learning_rate": 0.0001192496838555571,
+      "loss": 5.1132,
+      "step": 28300
+    },
+    {
+      "epoch": 0.03977820594108793,
+      "grad_norm": 1.2096099853515625,
+      "learning_rate": 0.00011929183644794154,
+      "loss": 5.2598,
+      "step": 28310
+    },
+    {
+      "epoch": 0.03979225687925151,
+      "grad_norm": 1.2106359004974365,
+      "learning_rate": 0.00011933398904032596,
+      "loss": 4.9739,
+      "step": 28320
+    },
+    {
+      "epoch": 0.039806307817415086,
+      "grad_norm": 1.268173336982727,
+      "learning_rate": 0.00011937614163271039,
+      "loss": 5.3448,
+      "step": 28330
+    },
+    {
+      "epoch": 0.03982035875557866,
+      "grad_norm": 1.2181192636489868,
+      "learning_rate": 0.00011941829422509484,
+      "loss": 5.0814,
+      "step": 28340
+    },
+    {
+      "epoch": 0.03983440969374224,
+      "grad_norm": 1.2973514795303345,
+      "learning_rate": 0.00011946044681747927,
+      "loss": 4.9829,
+      "step": 28350
+    },
+    {
+      "epoch": 0.039848460631905815,
+      "grad_norm": 1.1853080987930298,
+      "learning_rate": 0.00011950259940986369,
+      "loss": 5.1086,
+      "step": 28360
+    },
+    {
+      "epoch": 0.03986251157006939,
+      "grad_norm": 1.2887145280838013,
+      "learning_rate": 0.00011954475200224813,
+      "loss": 5.2579,
+      "step": 28370
+    },
+    {
+      "epoch": 0.039876562508232974,
+      "grad_norm": 1.3244291543960571,
+      "learning_rate": 0.00011958690459463256,
+      "loss": 5.0825,
+      "step": 28380
+    },
+    {
+      "epoch": 0.03989061344639655,
+      "grad_norm": 1.2618260383605957,
+      "learning_rate": 0.00011962905718701698,
+      "loss": 5.0181,
+      "step": 28390
+    },
+    {
+      "epoch": 0.039904664384560126,
+      "grad_norm": 1.179956078529358,
+      "learning_rate": 0.00011967120977940142,
+      "loss": 5.2237,
+      "step": 28400
+    },
+    {
+      "epoch": 0.0399187153227237,
+      "grad_norm": 1.2036820650100708,
+      "learning_rate": 0.00011971336237178585,
+      "loss": 5.0647,
+      "step": 28410
+    },
+    {
+      "epoch": 0.03993276626088728,
+      "grad_norm": 1.2143844366073608,
+      "learning_rate": 0.00011975551496417029,
+      "loss": 5.0621,
+      "step": 28420
+    },
+    {
+      "epoch": 0.03994681719905086,
+      "grad_norm": 1.1590876579284668,
+      "learning_rate": 0.00011979766755655472,
+      "loss": 5.0342,
+      "step": 28430
+    },
+    {
+      "epoch": 0.03996086813721444,
+      "grad_norm": 1.2688127756118774,
+      "learning_rate": 0.00011983982014893915,
+      "loss": 5.2687,
+      "step": 28440
+    },
+    {
+      "epoch": 0.039974919075378014,
+      "grad_norm": 1.236639142036438,
+      "learning_rate": 0.00011988197274132358,
+      "loss": 5.0454,
+      "step": 28450
+    },
+    {
+      "epoch": 0.03998897001354159,
+      "grad_norm": 1.2120410203933716,
+      "learning_rate": 0.00011992412533370801,
+      "loss": 5.1101,
+      "step": 28460
+    },
+    {
+      "epoch": 0.040003020951705166,
+      "grad_norm": 1.2371752262115479,
+      "learning_rate": 0.00011996627792609244,
+      "loss": 5.1434,
+      "step": 28470
+    },
+    {
+      "epoch": 0.04001707188986875,
+      "grad_norm": 1.2628980875015259,
+      "learning_rate": 0.00012000843051847689,
+      "loss": 5.058,
+      "step": 28480
+    },
+    {
+      "epoch": 0.040031122828032326,
+      "grad_norm": 1.2172374725341797,
+      "learning_rate": 0.0001200505831108613,
+      "loss": 5.1972,
+      "step": 28490
+    },
+    {
+      "epoch": 0.0400451737661959,
+      "grad_norm": 1.1844308376312256,
+      "learning_rate": 0.00012009273570324574,
+      "loss": 5.1624,
+      "step": 28500
+    },
+    {
+      "epoch": 0.04005922470435948,
+      "grad_norm": 1.2529436349868774,
+      "learning_rate": 0.00012013488829563018,
+      "loss": 5.0565,
+      "step": 28510
+    },
+    {
+      "epoch": 0.040073275642523054,
+      "grad_norm": 1.2237712144851685,
+      "learning_rate": 0.0001201770408880146,
+      "loss": 5.1909,
+      "step": 28520
+    },
+    {
+      "epoch": 0.04008732658068664,
+      "grad_norm": 1.2147319316864014,
+      "learning_rate": 0.00012021919348039903,
+      "loss": 5.0974,
+      "step": 28530
+    },
+    {
+      "epoch": 0.04010137751885021,
+      "grad_norm": 1.194334626197815,
+      "learning_rate": 0.00012026134607278347,
+      "loss": 5.2433,
+      "step": 28540
+    },
+    {
+      "epoch": 0.04011542845701379,
+      "grad_norm": 1.2386977672576904,
+      "learning_rate": 0.00012030349866516789,
+      "loss": 5.2399,
+      "step": 28550
+    },
+    {
+      "epoch": 0.040129479395177366,
+      "grad_norm": 1.2008346319198608,
+      "learning_rate": 0.00012034565125755232,
+      "loss": 5.0889,
+      "step": 28560
+    },
+    {
+      "epoch": 0.04014353033334094,
+      "grad_norm": 1.1994179487228394,
+      "learning_rate": 0.00012038780384993677,
+      "loss": 5.1955,
+      "step": 28570
+    },
+    {
+      "epoch": 0.040157581271504525,
+      "grad_norm": 1.251625895500183,
+      "learning_rate": 0.00012042995644232118,
+      "loss": 5.1563,
+      "step": 28580
+    },
+    {
+      "epoch": 0.0401716322096681,
+      "grad_norm": 1.3374724388122559,
+      "learning_rate": 0.00012047210903470562,
+      "loss": 5.2259,
+      "step": 28590
+    },
+    {
+      "epoch": 0.04018568314783168,
+      "grad_norm": 1.184990644454956,
+      "learning_rate": 0.00012051426162709006,
+      "loss": 5.1774,
+      "step": 28600
+    },
+    {
+      "epoch": 0.04019973408599525,
+      "grad_norm": 1.2657346725463867,
+      "learning_rate": 0.00012055641421947449,
+      "loss": 5.0466,
+      "step": 28610
+    },
+    {
+      "epoch": 0.04021378502415883,
+      "grad_norm": 1.301020860671997,
+      "learning_rate": 0.00012059435155262047,
+      "loss": 5.1848,
+      "step": 28620
+    },
+    {
+      "epoch": 0.04022783596232241,
+      "grad_norm": 1.2240586280822754,
+      "learning_rate": 0.0001206365041450049,
+      "loss": 5.1032,
+      "step": 28630
+    },
+    {
+      "epoch": 0.04024188690048599,
+      "grad_norm": 1.2186791896820068,
+      "learning_rate": 0.00012067865673738935,
+      "loss": 5.1814,
+      "step": 28640
+    },
+    {
+      "epoch": 0.040255937838649565,
+      "grad_norm": 1.2839056253433228,
+      "learning_rate": 0.00012072080932977377,
+      "loss": 5.2019,
+      "step": 28650
+    },
+    {
+      "epoch": 0.04026998877681314,
+      "grad_norm": 1.2120087146759033,
+      "learning_rate": 0.0001207629619221582,
+      "loss": 5.1894,
+      "step": 28660
+    },
+    {
+      "epoch": 0.04028403971497672,
+      "grad_norm": 1.3091609477996826,
+      "learning_rate": 0.00012080511451454264,
+      "loss": 5.2507,
+      "step": 28670
+    },
+    {
+      "epoch": 0.040298090653140294,
+      "grad_norm": 1.285378098487854,
+      "learning_rate": 0.00012084726710692706,
+      "loss": 5.1372,
+      "step": 28680
+    },
+    {
+      "epoch": 0.04031214159130388,
+      "grad_norm": 1.1991231441497803,
+      "learning_rate": 0.00012088941969931149,
+      "loss": 5.1167,
+      "step": 28690
+    },
+    {
+      "epoch": 0.04032619252946745,
+      "grad_norm": 1.2777354717254639,
+      "learning_rate": 0.00012093157229169594,
+      "loss": 5.0535,
+      "step": 28700
+    },
+    {
+      "epoch": 0.04034024346763103,
+      "grad_norm": 1.2229773998260498,
+      "learning_rate": 0.00012097372488408035,
+      "loss": 5.1478,
+      "step": 28710
+    },
+    {
+      "epoch": 0.040354294405794605,
+      "grad_norm": 1.2146847248077393,
+      "learning_rate": 0.00012101587747646479,
+      "loss": 5.1944,
+      "step": 28720
+    },
+    {
+      "epoch": 0.04036834534395818,
+      "grad_norm": 1.1953580379486084,
+      "learning_rate": 0.00012105803006884923,
+      "loss": 5.1611,
+      "step": 28730
+    },
+    {
+      "epoch": 0.040382396282121764,
+      "grad_norm": 1.2256001234054565,
+      "learning_rate": 0.00012110018266123366,
+      "loss": 5.316,
+      "step": 28740
+    },
+    {
+      "epoch": 0.04039644722028534,
+      "grad_norm": 1.2031221389770508,
+      "learning_rate": 0.00012114233525361808,
+      "loss": 5.0734,
+      "step": 28750
+    },
+    {
+      "epoch": 0.04041049815844892,
+      "grad_norm": 1.2694804668426514,
+      "learning_rate": 0.00012118448784600252,
+      "loss": 5.1379,
+      "step": 28760
+    },
+    {
+      "epoch": 0.04042454909661249,
+      "grad_norm": 1.2258470058441162,
+      "learning_rate": 0.00012122664043838696,
+      "loss": 5.0536,
+      "step": 28770
+    },
+    {
+      "epoch": 0.04043860003477607,
+      "grad_norm": 1.2044484615325928,
+      "learning_rate": 0.00012126879303077137,
+      "loss": 5.2435,
+      "step": 28780
+    },
+    {
+      "epoch": 0.04045265097293965,
+      "grad_norm": 1.27933931350708,
+      "learning_rate": 0.00012131094562315582,
+      "loss": 5.1742,
+      "step": 28790
+    },
+    {
+      "epoch": 0.04046670191110323,
+      "grad_norm": 1.2397048473358154,
+      "learning_rate": 0.00012135309821554025,
+      "loss": 5.1537,
+      "step": 28800
+    },
+    {
+      "epoch": 0.040480752849266805,
+      "grad_norm": 1.3826310634613037,
+      "learning_rate": 0.00012139525080792467,
+      "loss": 5.1767,
+      "step": 28810
+    },
+    {
+      "epoch": 0.04049480378743038,
+      "grad_norm": 1.2300126552581787,
+      "learning_rate": 0.00012143740340030911,
+      "loss": 5.0833,
+      "step": 28820
+    },
+    {
+      "epoch": 0.04050885472559396,
+      "grad_norm": 1.2584890127182007,
+      "learning_rate": 0.00012147955599269354,
+      "loss": 5.1247,
+      "step": 28830
+    },
+    {
+      "epoch": 0.04052290566375754,
+      "grad_norm": 1.433069109916687,
+      "learning_rate": 0.00012152170858507796,
+      "loss": 5.133,
+      "step": 28840
+    },
+    {
+      "epoch": 0.040536956601921116,
+      "grad_norm": 1.2440624237060547,
+      "learning_rate": 0.0001215638611774624,
+      "loss": 5.0757,
+      "step": 28850
+    },
+    {
+      "epoch": 0.04055100754008469,
+      "grad_norm": 1.216599702835083,
+      "learning_rate": 0.00012160601376984684,
+      "loss": 4.9827,
+      "step": 28860
+    },
+    {
+      "epoch": 0.04056505847824827,
+      "grad_norm": 1.3541535139083862,
+      "learning_rate": 0.00012164816636223125,
+      "loss": 5.112,
+      "step": 28870
+    },
+    {
+      "epoch": 0.040579109416411845,
+      "grad_norm": 1.1753549575805664,
+      "learning_rate": 0.0001216903189546157,
+      "loss": 5.1629,
+      "step": 28880
+    },
+    {
+      "epoch": 0.04059316035457543,
+      "grad_norm": 1.2143678665161133,
+      "learning_rate": 0.00012173247154700013,
+      "loss": 5.1621,
+      "step": 28890
+    },
+    {
+      "epoch": 0.040607211292739004,
+      "grad_norm": 1.2164275646209717,
+      "learning_rate": 0.00012177462413938457,
+      "loss": 5.1983,
+      "step": 28900
+    },
+    {
+      "epoch": 0.04062126223090258,
+      "grad_norm": 1.676349401473999,
+      "learning_rate": 0.00012181677673176899,
+      "loss": 5.1482,
+      "step": 28910
+    },
+    {
+      "epoch": 0.040635313169066156,
+      "grad_norm": 1.188805341720581,
+      "learning_rate": 0.00012185892932415342,
+      "loss": 5.1028,
+      "step": 28920
+    },
+    {
+      "epoch": 0.04064936410722973,
+      "grad_norm": 1.304349660873413,
+      "learning_rate": 0.00012190108191653787,
+      "loss": 5.1201,
+      "step": 28930
+    },
+    {
+      "epoch": 0.040663415045393315,
+      "grad_norm": 1.256514549255371,
+      "learning_rate": 0.00012194323450892229,
+      "loss": 5.1092,
+      "step": 28940
+    },
+    {
+      "epoch": 0.04067746598355689,
+      "grad_norm": 1.1946409940719604,
+      "learning_rate": 0.00012198538710130672,
+      "loss": 5.2572,
+      "step": 28950
+    },
+    {
+      "epoch": 0.04069151692172047,
+      "grad_norm": 1.2821025848388672,
+      "learning_rate": 0.00012202753969369116,
+      "loss": 5.16,
+      "step": 28960
+    },
+    {
+      "epoch": 0.040705567859884044,
+      "grad_norm": 1.2164089679718018,
+      "learning_rate": 0.00012206969228607558,
+      "loss": 5.1757,
+      "step": 28970
+    },
+    {
+      "epoch": 0.04071961879804762,
+      "grad_norm": 1.3499404191970825,
+      "learning_rate": 0.00012211184487846002,
+      "loss": 5.1214,
+      "step": 28980
+    },
+    {
+      "epoch": 0.040733669736211196,
+      "grad_norm": 1.1832070350646973,
+      "learning_rate": 0.00012215399747084445,
+      "loss": 5.1601,
+      "step": 28990
+    },
+    {
+      "epoch": 0.04074772067437478,
+      "grad_norm": 1.2199169397354126,
+      "learning_rate": 0.00012219615006322886,
+      "loss": 5.046,
+      "step": 29000
+    },
+    {
+      "epoch": 0.040761771612538356,
+      "grad_norm": 1.2331231832504272,
+      "learning_rate": 0.00012223830265561332,
+      "loss": 5.0809,
+      "step": 29010
+    },
+    {
+      "epoch": 0.04077582255070193,
+      "grad_norm": 1.2529569864273071,
+      "learning_rate": 0.00012228045524799775,
+      "loss": 5.0944,
+      "step": 29020
+    },
+    {
+      "epoch": 0.04078987348886551,
+      "grad_norm": 1.2748099565505981,
+      "learning_rate": 0.00012232260784038218,
+      "loss": 4.9984,
+      "step": 29030
+    },
+    {
+      "epoch": 0.040803924427029084,
+      "grad_norm": 1.151564121246338,
+      "learning_rate": 0.0001223647604327666,
+      "loss": 5.0823,
+      "step": 29040
+    },
+    {
+      "epoch": 0.04081797536519267,
+      "grad_norm": 1.1809443235397339,
+      "learning_rate": 0.00012240691302515104,
+      "loss": 5.1298,
+      "step": 29050
+    },
+    {
+      "epoch": 0.04083202630335624,
+      "grad_norm": 1.2853679656982422,
+      "learning_rate": 0.00012244906561753547,
+      "loss": 5.1489,
+      "step": 29060
+    },
+    {
+      "epoch": 0.04084607724151982,
+      "grad_norm": 1.2451461553573608,
+      "learning_rate": 0.0001224912182099199,
+      "loss": 5.2294,
+      "step": 29070
+    },
+    {
+      "epoch": 0.040860128179683396,
+      "grad_norm": 1.1953366994857788,
+      "learning_rate": 0.00012253337080230434,
+      "loss": 5.1022,
+      "step": 29080
+    },
+    {
+      "epoch": 0.04087417911784697,
+      "grad_norm": 1.1801906824111938,
+      "learning_rate": 0.00012257552339468877,
+      "loss": 5.0872,
+      "step": 29090
+    },
+    {
+      "epoch": 0.040888230056010555,
+      "grad_norm": 1.2454745769500732,
+      "learning_rate": 0.0001226176759870732,
+      "loss": 5.1487,
+      "step": 29100
+    },
+    {
+      "epoch": 0.04090228099417413,
+      "grad_norm": 1.5468188524246216,
+      "learning_rate": 0.00012265982857945763,
+      "loss": 5.094,
+      "step": 29110
+    },
+    {
+      "epoch": 0.04091633193233771,
+      "grad_norm": 1.2361538410186768,
+      "learning_rate": 0.00012270198117184206,
+      "loss": 5.1832,
+      "step": 29120
+    },
+    {
+      "epoch": 0.04093038287050128,
+      "grad_norm": 1.2000207901000977,
+      "learning_rate": 0.0001227441337642265,
+      "loss": 5.1678,
+      "step": 29130
+    },
+    {
+      "epoch": 0.04094443380866486,
+      "grad_norm": 1.1986608505249023,
+      "learning_rate": 0.00012278628635661092,
+      "loss": 5.1264,
+      "step": 29140
+    },
+    {
+      "epoch": 0.04095848474682844,
+      "grad_norm": 1.2702739238739014,
+      "learning_rate": 0.00012282843894899535,
+      "loss": 5.1283,
+      "step": 29150
+    },
+    {
+      "epoch": 0.04097253568499202,
+      "grad_norm": 1.256484031677246,
+      "learning_rate": 0.00012287059154137978,
+      "loss": 5.1021,
+      "step": 29160
+    },
+    {
+      "epoch": 0.040986586623155595,
+      "grad_norm": 1.2504194974899292,
+      "learning_rate": 0.00012291274413376422,
+      "loss": 5.1583,
+      "step": 29170
+    },
+    {
+      "epoch": 0.04100063756131917,
+      "grad_norm": 1.193787693977356,
+      "learning_rate": 0.00012295489672614865,
+      "loss": 5.1885,
+      "step": 29180
+    },
+    {
+      "epoch": 0.04101468849948275,
+      "grad_norm": 1.2166396379470825,
+      "learning_rate": 0.00012299704931853308,
+      "loss": 4.9795,
+      "step": 29190
+    },
+    {
+      "epoch": 0.04102873943764633,
+      "grad_norm": 1.3542882204055786,
+      "learning_rate": 0.0001230392019109175,
+      "loss": 5.0044,
+      "step": 29200
+    },
+    {
+      "epoch": 0.04104279037580991,
+      "grad_norm": 1.3186988830566406,
+      "learning_rate": 0.00012308135450330194,
+      "loss": 5.0079,
+      "step": 29210
+    },
+    {
+      "epoch": 0.04105684131397348,
+      "grad_norm": 1.174622893333435,
+      "learning_rate": 0.00012312350709568637,
+      "loss": 5.1067,
+      "step": 29220
+    },
+    {
+      "epoch": 0.04107089225213706,
+      "grad_norm": 1.1636348962783813,
+      "learning_rate": 0.0001231656596880708,
+      "loss": 5.1078,
+      "step": 29230
+    },
+    {
+      "epoch": 0.041084943190300635,
+      "grad_norm": 1.1850835084915161,
+      "learning_rate": 0.00012320781228045523,
+      "loss": 5.1331,
+      "step": 29240
+    },
+    {
+      "epoch": 0.04109899412846422,
+      "grad_norm": 1.2667229175567627,
+      "learning_rate": 0.00012324996487283966,
+      "loss": 5.0366,
+      "step": 29250
+    },
+    {
+      "epoch": 0.041113045066627794,
+      "grad_norm": 1.217487096786499,
+      "learning_rate": 0.0001232921174652241,
+      "loss": 5.1203,
+      "step": 29260
+    },
+    {
+      "epoch": 0.04112709600479137,
+      "grad_norm": 1.2223021984100342,
+      "learning_rate": 0.00012333427005760853,
+      "loss": 5.0865,
+      "step": 29270
+    },
+    {
+      "epoch": 0.04114114694295495,
+      "grad_norm": 1.1677926778793335,
+      "learning_rate": 0.00012337642264999296,
+      "loss": 5.1227,
+      "step": 29280
+    },
+    {
+      "epoch": 0.04115519788111852,
+      "grad_norm": 1.197870135307312,
+      "learning_rate": 0.0001234185752423774,
+      "loss": 5.1118,
+      "step": 29290
+    },
+    {
+      "epoch": 0.0411692488192821,
+      "grad_norm": 1.2114304304122925,
+      "learning_rate": 0.00012346072783476182,
+      "loss": 5.1161,
+      "step": 29300
+    },
+    {
+      "epoch": 0.04118329975744568,
+      "grad_norm": 1.266849160194397,
+      "learning_rate": 0.00012350288042714625,
+      "loss": 5.0399,
+      "step": 29310
+    },
+    {
+      "epoch": 0.04119735069560926,
+      "grad_norm": 1.3081083297729492,
+      "learning_rate": 0.0001235450330195307,
+      "loss": 5.1199,
+      "step": 29320
+    },
+    {
+      "epoch": 0.041211401633772834,
+      "grad_norm": 1.2739002704620361,
+      "learning_rate": 0.00012358718561191511,
+      "loss": 5.0112,
+      "step": 29330
+    },
+    {
+      "epoch": 0.04122545257193641,
+      "grad_norm": 1.150550127029419,
+      "learning_rate": 0.00012362933820429955,
+      "loss": 5.0035,
+      "step": 29340
+    },
+    {
+      "epoch": 0.04123950351009999,
+      "grad_norm": 1.1802303791046143,
+      "learning_rate": 0.000123671490796684,
+      "loss": 5.0593,
+      "step": 29350
+    },
+    {
+      "epoch": 0.04125355444826357,
+      "grad_norm": 1.209800124168396,
+      "learning_rate": 0.0001237136433890684,
+      "loss": 5.0386,
+      "step": 29360
+    },
+    {
+      "epoch": 0.041267605386427146,
+      "grad_norm": 1.1998001337051392,
+      "learning_rate": 0.00012375579598145284,
+      "loss": 5.1109,
+      "step": 29370
+    },
+    {
+      "epoch": 0.04128165632459072,
+      "grad_norm": 1.1864341497421265,
+      "learning_rate": 0.0001237979485738373,
+      "loss": 5.1374,
+      "step": 29380
+    },
+    {
+      "epoch": 0.0412957072627543,
+      "grad_norm": 1.1645110845565796,
+      "learning_rate": 0.0001238401011662217,
+      "loss": 5.0924,
+      "step": 29390
+    },
+    {
+      "epoch": 0.041309758200917875,
+      "grad_norm": 1.193508505821228,
+      "learning_rate": 0.00012388225375860613,
+      "loss": 5.1911,
+      "step": 29400
+    },
+    {
+      "epoch": 0.04132380913908146,
+      "grad_norm": 1.2677333354949951,
+      "learning_rate": 0.0001239244063509906,
+      "loss": 5.1231,
+      "step": 29410
+    },
+    {
+      "epoch": 0.041337860077245034,
+      "grad_norm": 1.1700749397277832,
+      "learning_rate": 0.000123966558943375,
+      "loss": 5.1528,
+      "step": 29420
+    },
+    {
+      "epoch": 0.04135191101540861,
+      "grad_norm": 1.1878379583358765,
+      "learning_rate": 0.00012400871153575943,
+      "loss": 5.2438,
+      "step": 29430
+    },
+    {
+      "epoch": 0.041365961953572186,
+      "grad_norm": 1.2088021039962769,
+      "learning_rate": 0.00012405086412814388,
+      "loss": 5.0927,
+      "step": 29440
+    },
+    {
+      "epoch": 0.04138001289173576,
+      "grad_norm": 1.279140830039978,
+      "learning_rate": 0.00012409301672052832,
+      "loss": 5.0007,
+      "step": 29450
+    },
+    {
+      "epoch": 0.041394063829899345,
+      "grad_norm": 1.22353196144104,
+      "learning_rate": 0.00012413516931291272,
+      "loss": 5.1005,
+      "step": 29460
+    },
+    {
+      "epoch": 0.04140811476806292,
+      "grad_norm": 1.2469463348388672,
+      "learning_rate": 0.00012417732190529718,
+      "loss": 5.0887,
+      "step": 29470
+    },
+    {
+      "epoch": 0.0414221657062265,
+      "grad_norm": 1.2041430473327637,
+      "learning_rate": 0.0001242194744976816,
+      "loss": 5.1795,
+      "step": 29480
+    },
+    {
+      "epoch": 0.041436216644390074,
+      "grad_norm": 1.2046658992767334,
+      "learning_rate": 0.000124261627090066,
+      "loss": 5.1795,
+      "step": 29490
+    },
+    {
+      "epoch": 0.04145026758255365,
+      "grad_norm": 1.1940624713897705,
+      "learning_rate": 0.00012430377968245047,
+      "loss": 5.1144,
+      "step": 29500
+    },
+    {
+      "epoch": 0.04146431852071723,
+      "grad_norm": 1.2514973878860474,
+      "learning_rate": 0.0001243459322748349,
+      "loss": 5.0636,
+      "step": 29510
+    },
+    {
+      "epoch": 0.04147836945888081,
+      "grad_norm": 1.166250228881836,
+      "learning_rate": 0.0001243880848672193,
+      "loss": 5.1468,
+      "step": 29520
+    },
+    {
+      "epoch": 0.041492420397044386,
+      "grad_norm": 1.201289176940918,
+      "learning_rate": 0.00012443023745960376,
+      "loss": 5.0477,
+      "step": 29530
+    },
+    {
+      "epoch": 0.04150647133520796,
+      "grad_norm": 1.2102603912353516,
+      "learning_rate": 0.0001244723900519882,
+      "loss": 4.9634,
+      "step": 29540
+    },
+    {
+      "epoch": 0.04152052227337154,
+      "grad_norm": 1.303805947303772,
+      "learning_rate": 0.0001245145426443726,
+      "loss": 5.0351,
+      "step": 29550
+    },
+    {
+      "epoch": 0.04153457321153512,
+      "grad_norm": 1.160770058631897,
+      "learning_rate": 0.00012455669523675706,
+      "loss": 5.224,
+      "step": 29560
+    },
+    {
+      "epoch": 0.0415486241496987,
+      "grad_norm": 1.1994746923446655,
+      "learning_rate": 0.0001245988478291415,
+      "loss": 5.1209,
+      "step": 29570
+    },
+    {
+      "epoch": 0.04156267508786227,
+      "grad_norm": 1.2357219457626343,
+      "learning_rate": 0.0001246410004215259,
+      "loss": 5.0569,
+      "step": 29580
+    },
+    {
+      "epoch": 0.04157672602602585,
+      "grad_norm": 1.1528397798538208,
+      "learning_rate": 0.00012468315301391035,
+      "loss": 5.1259,
+      "step": 29590
+    },
+    {
+      "epoch": 0.041590776964189426,
+      "grad_norm": 1.1492503881454468,
+      "learning_rate": 0.00012472530560629478,
+      "loss": 5.2507,
+      "step": 29600
+    },
+    {
+      "epoch": 0.041604827902353,
+      "grad_norm": 1.1577057838439941,
+      "learning_rate": 0.00012476745819867921,
+      "loss": 5.2534,
+      "step": 29610
+    },
+    {
+      "epoch": 0.041618878840516585,
+      "grad_norm": 1.2352550029754639,
+      "learning_rate": 0.00012480961079106365,
+      "loss": 5.0825,
+      "step": 29620
+    },
+    {
+      "epoch": 0.04163292977868016,
+      "grad_norm": 1.359626054763794,
+      "learning_rate": 0.00012485176338344808,
+      "loss": 5.0889,
+      "step": 29630
+    },
+    {
+      "epoch": 0.04164698071684374,
+      "grad_norm": 1.2673550844192505,
+      "learning_rate": 0.0001248939159758325,
+      "loss": 5.0325,
+      "step": 29640
+    },
+    {
+      "epoch": 0.04166103165500731,
+      "grad_norm": 1.1786859035491943,
+      "learning_rate": 0.00012493606856821694,
+      "loss": 5.1071,
+      "step": 29650
+    },
+    {
+      "epoch": 0.04167508259317089,
+      "grad_norm": 1.209380030632019,
+      "learning_rate": 0.00012497822116060137,
+      "loss": 5.172,
+      "step": 29660
+    },
+    {
+      "epoch": 0.04168913353133447,
+      "grad_norm": 1.3903775215148926,
+      "learning_rate": 0.0001250203737529858,
+      "loss": 5.128,
+      "step": 29670
+    },
+    {
+      "epoch": 0.04170318446949805,
+      "grad_norm": 1.2083300352096558,
+      "learning_rate": 0.00012506252634537023,
+      "loss": 5.2088,
+      "step": 29680
+    },
+    {
+      "epoch": 0.041717235407661625,
+      "grad_norm": 1.3227828741073608,
+      "learning_rate": 0.00012510467893775466,
+      "loss": 5.0573,
+      "step": 29690
+    },
+    {
+      "epoch": 0.0417312863458252,
+      "grad_norm": 1.19940185546875,
+      "learning_rate": 0.0001251468315301391,
+      "loss": 5.0194,
+      "step": 29700
+    },
+    {
+      "epoch": 0.04174533728398878,
+      "grad_norm": 1.232577919960022,
+      "learning_rate": 0.00012518898412252353,
+      "loss": 5.1452,
+      "step": 29710
+    },
+    {
+      "epoch": 0.04175938822215236,
+      "grad_norm": 1.2753794193267822,
+      "learning_rate": 0.00012523113671490796,
+      "loss": 4.9559,
+      "step": 29720
+    },
+    {
+      "epoch": 0.04177343916031594,
+      "grad_norm": 1.3034454584121704,
+      "learning_rate": 0.0001252732893072924,
+      "loss": 5.1681,
+      "step": 29730
+    },
+    {
+      "epoch": 0.04178749009847951,
+      "grad_norm": 1.1941345930099487,
+      "learning_rate": 0.00012531544189967682,
+      "loss": 5.0454,
+      "step": 29740
+    },
+    {
+      "epoch": 0.04180154103664309,
+      "grad_norm": 1.2708008289337158,
+      "learning_rate": 0.00012535759449206125,
+      "loss": 5.0656,
+      "step": 29750
+    },
+    {
+      "epoch": 0.041815591974806665,
+      "grad_norm": 1.1832579374313354,
+      "learning_rate": 0.00012539974708444568,
+      "loss": 5.0031,
+      "step": 29760
+    },
+    {
+      "epoch": 0.04182964291297025,
+      "grad_norm": 1.1988276243209839,
+      "learning_rate": 0.0001254418996768301,
+      "loss": 5.0703,
+      "step": 29770
+    },
+    {
+      "epoch": 0.041843693851133824,
+      "grad_norm": 1.2514029741287231,
+      "learning_rate": 0.00012548405226921454,
+      "loss": 5.1405,
+      "step": 29780
+    },
+    {
+      "epoch": 0.0418577447892974,
+      "grad_norm": 1.1804388761520386,
+      "learning_rate": 0.00012552620486159898,
+      "loss": 5.0957,
+      "step": 29790
+    },
+    {
+      "epoch": 0.04187179572746098,
+      "grad_norm": 1.2735486030578613,
+      "learning_rate": 0.0001255683574539834,
+      "loss": 5.0302,
+      "step": 29800
+    },
+    {
+      "epoch": 0.04188584666562455,
+      "grad_norm": 1.2295114994049072,
+      "learning_rate": 0.00012561051004636784,
+      "loss": 5.2056,
+      "step": 29810
+    },
+    {
+      "epoch": 0.041899897603788136,
+      "grad_norm": 1.1553417444229126,
+      "learning_rate": 0.00012565266263875227,
+      "loss": 5.0882,
+      "step": 29820
+    },
+    {
+      "epoch": 0.04191394854195171,
+      "grad_norm": 1.304266333580017,
+      "learning_rate": 0.0001256948152311367,
+      "loss": 5.2603,
+      "step": 29830
+    },
+    {
+      "epoch": 0.04192799948011529,
+      "grad_norm": 1.1928671598434448,
+      "learning_rate": 0.00012573696782352113,
+      "loss": 5.1606,
+      "step": 29840
+    },
+    {
+      "epoch": 0.041942050418278864,
+      "grad_norm": 1.2966556549072266,
+      "learning_rate": 0.00012577912041590556,
+      "loss": 5.1442,
+      "step": 29850
+    },
+    {
+      "epoch": 0.04195610135644244,
+      "grad_norm": 1.3299918174743652,
+      "learning_rate": 0.00012582127300829,
+      "loss": 5.0088,
+      "step": 29860
+    },
+    {
+      "epoch": 0.041970152294606024,
+      "grad_norm": 1.2035354375839233,
+      "learning_rate": 0.00012586342560067442,
+      "loss": 5.0605,
+      "step": 29870
+    },
+    {
+      "epoch": 0.0419842032327696,
+      "grad_norm": 1.1587860584259033,
+      "learning_rate": 0.00012590557819305886,
+      "loss": 5.262,
+      "step": 29880
+    },
+    {
+      "epoch": 0.041998254170933176,
+      "grad_norm": 1.212730884552002,
+      "learning_rate": 0.0001259477307854433,
+      "loss": 5.1807,
+      "step": 29890
+    },
+    {
+      "epoch": 0.04201230510909675,
+      "grad_norm": 1.2755563259124756,
+      "learning_rate": 0.00012598988337782774,
+      "loss": 5.1815,
+      "step": 29900
+    },
+    {
+      "epoch": 0.04202635604726033,
+      "grad_norm": 1.2221256494522095,
+      "learning_rate": 0.00012603203597021215,
+      "loss": 5.1726,
+      "step": 29910
+    },
+    {
+      "epoch": 0.042040406985423905,
+      "grad_norm": 1.246488332748413,
+      "learning_rate": 0.00012607418856259658,
+      "loss": 5.1286,
+      "step": 29920
+    },
+    {
+      "epoch": 0.04205445792358749,
+      "grad_norm": 1.2140997648239136,
+      "learning_rate": 0.00012611634115498104,
+      "loss": 5.1014,
+      "step": 29930
+    },
+    {
+      "epoch": 0.042068508861751064,
+      "grad_norm": 1.2002910375595093,
+      "learning_rate": 0.00012615849374736544,
+      "loss": 5.1418,
+      "step": 29940
+    },
+    {
+      "epoch": 0.04208255979991464,
+      "grad_norm": 1.1797535419464111,
+      "learning_rate": 0.00012620064633974987,
+      "loss": 5.2805,
+      "step": 29950
+    },
+    {
+      "epoch": 0.042096610738078216,
+      "grad_norm": 1.2059688568115234,
+      "learning_rate": 0.00012624279893213433,
+      "loss": 5.1886,
+      "step": 29960
+    },
+    {
+      "epoch": 0.04211066167624179,
+      "grad_norm": 1.22100031375885,
+      "learning_rate": 0.00012628495152451874,
+      "loss": 5.1398,
+      "step": 29970
+    },
+    {
+      "epoch": 0.042124712614405375,
+      "grad_norm": 1.2340439558029175,
+      "learning_rate": 0.00012632710411690317,
+      "loss": 5.1082,
+      "step": 29980
+    },
+    {
+      "epoch": 0.04213876355256895,
+      "grad_norm": 1.218373417854309,
+      "learning_rate": 0.00012636925670928763,
+      "loss": 5.1635,
+      "step": 29990
+    },
+    {
+      "epoch": 0.04215281449073253,
+      "grad_norm": 1.2156459093093872,
+      "learning_rate": 0.00012641140930167203,
+      "loss": 5.1422,
+      "step": 30000
+    },
+    {
+      "epoch": 0.042166865428896104,
+      "grad_norm": 1.1739648580551147,
+      "learning_rate": 0.00012645356189405646,
+      "loss": 5.2125,
+      "step": 30010
+    },
+    {
+      "epoch": 0.04218091636705968,
+      "grad_norm": 1.2138736248016357,
+      "learning_rate": 0.00012649571448644092,
+      "loss": 5.1365,
+      "step": 30020
+    },
+    {
+      "epoch": 0.04219496730522326,
+      "grad_norm": 1.1886487007141113,
+      "learning_rate": 0.00012653786707882535,
+      "loss": 5.0892,
+      "step": 30030
+    },
+    {
+      "epoch": 0.04220901824338684,
+      "grad_norm": 1.1931877136230469,
+      "learning_rate": 0.00012658001967120975,
+      "loss": 5.091,
+      "step": 30040
+    },
+    {
+      "epoch": 0.042223069181550416,
+      "grad_norm": 1.1976298093795776,
+      "learning_rate": 0.0001266221722635942,
+      "loss": 5.243,
+      "step": 30050
+    },
+    {
+      "epoch": 0.04223712011971399,
+      "grad_norm": 1.1765670776367188,
+      "learning_rate": 0.00012666432485597864,
+      "loss": 5.1197,
+      "step": 30060
+    },
+    {
+      "epoch": 0.04225117105787757,
+      "grad_norm": 1.182282567024231,
+      "learning_rate": 0.00012670647744836305,
+      "loss": 5.0915,
+      "step": 30070
+    },
+    {
+      "epoch": 0.04226522199604115,
+      "grad_norm": 1.2112866640090942,
+      "learning_rate": 0.0001267486300407475,
+      "loss": 5.1765,
+      "step": 30080
+    },
+    {
+      "epoch": 0.04227927293420473,
+      "grad_norm": 1.2055011987686157,
+      "learning_rate": 0.00012679078263313194,
+      "loss": 5.0592,
+      "step": 30090
+    },
+    {
+      "epoch": 0.0422933238723683,
+      "grad_norm": 1.2951544523239136,
+      "learning_rate": 0.00012683293522551634,
+      "loss": 5.0923,
+      "step": 30100
+    },
+    {
+      "epoch": 0.04230737481053188,
+      "grad_norm": 1.1470328569412231,
+      "learning_rate": 0.0001268750878179008,
+      "loss": 5.103,
+      "step": 30110
+    },
+    {
+      "epoch": 0.042321425748695456,
+      "grad_norm": 1.1909065246582031,
+      "learning_rate": 0.00012691724041028523,
+      "loss": 5.2148,
+      "step": 30120
+    },
+    {
+      "epoch": 0.04233547668685904,
+      "grad_norm": 1.2222651243209839,
+      "learning_rate": 0.00012695939300266963,
+      "loss": 5.07,
+      "step": 30130
+    },
+    {
+      "epoch": 0.042349527625022615,
+      "grad_norm": 1.1813335418701172,
+      "learning_rate": 0.0001270015455950541,
+      "loss": 5.0987,
+      "step": 30140
+    },
+    {
+      "epoch": 0.04236357856318619,
+      "grad_norm": 1.1861828565597534,
+      "learning_rate": 0.00012704369818743852,
+      "loss": 5.1389,
+      "step": 30150
+    },
+    {
+      "epoch": 0.04237762950134977,
+      "grad_norm": 1.2178916931152344,
+      "learning_rate": 0.00012708585077982293,
+      "loss": 5.1295,
+      "step": 30160
+    },
+    {
+      "epoch": 0.04239168043951334,
+      "grad_norm": 1.1351721286773682,
+      "learning_rate": 0.0001271280033722074,
+      "loss": 5.069,
+      "step": 30170
+    },
+    {
+      "epoch": 0.042405731377676927,
+      "grad_norm": 1.2155929803848267,
+      "learning_rate": 0.00012717015596459182,
+      "loss": 5.1474,
+      "step": 30180
+    },
+    {
+      "epoch": 0.0424197823158405,
+      "grad_norm": 1.1644866466522217,
+      "learning_rate": 0.00012721230855697625,
+      "loss": 5.1563,
+      "step": 30190
+    },
+    {
+      "epoch": 0.04243383325400408,
+      "grad_norm": 1.1552120447158813,
+      "learning_rate": 0.00012725446114936068,
+      "loss": 5.2732,
+      "step": 30200
+    },
+    {
+      "epoch": 0.042447884192167655,
+      "grad_norm": 1.1935938596725464,
+      "learning_rate": 0.0001272966137417451,
+      "loss": 5.0557,
+      "step": 30210
+    },
+    {
+      "epoch": 0.04246193513033123,
+      "grad_norm": 1.207800030708313,
+      "learning_rate": 0.00012733876633412954,
+      "loss": 5.0201,
+      "step": 30220
+    },
+    {
+      "epoch": 0.04247598606849481,
+      "grad_norm": 1.2109053134918213,
+      "learning_rate": 0.00012738091892651397,
+      "loss": 5.0742,
+      "step": 30230
+    },
+    {
+      "epoch": 0.04249003700665839,
+      "grad_norm": 1.20108962059021,
+      "learning_rate": 0.0001274230715188984,
+      "loss": 5.1348,
+      "step": 30240
+    },
+    {
+      "epoch": 0.04250408794482197,
+      "grad_norm": 1.1453440189361572,
+      "learning_rate": 0.00012746522411128284,
+      "loss": 5.1966,
+      "step": 30250
+    },
+    {
+      "epoch": 0.04251813888298554,
+      "grad_norm": 1.1746257543563843,
+      "learning_rate": 0.00012750737670366727,
+      "loss": 5.0678,
+      "step": 30260
+    },
+    {
+      "epoch": 0.04253218982114912,
+      "grad_norm": 1.5763996839523315,
+      "learning_rate": 0.0001275495292960517,
+      "loss": 5.0624,
+      "step": 30270
+    },
+    {
+      "epoch": 0.042546240759312695,
+      "grad_norm": 1.2596639394760132,
+      "learning_rate": 0.00012759168188843613,
+      "loss": 5.0564,
+      "step": 30280
+    },
+    {
+      "epoch": 0.04256029169747628,
+      "grad_norm": 1.1996608972549438,
+      "learning_rate": 0.00012763383448082056,
+      "loss": 4.9513,
+      "step": 30290
+    },
+    {
+      "epoch": 0.042574342635639854,
+      "grad_norm": 1.2479219436645508,
+      "learning_rate": 0.000127675987073205,
+      "loss": 5.0576,
+      "step": 30300
+    },
+    {
+      "epoch": 0.04258839357380343,
+      "grad_norm": 1.2073161602020264,
+      "learning_rate": 0.00012771813966558942,
+      "loss": 5.1667,
+      "step": 30310
+    },
+    {
+      "epoch": 0.04260244451196701,
+      "grad_norm": 1.2177608013153076,
+      "learning_rate": 0.00012776029225797385,
+      "loss": 4.9945,
+      "step": 30320
+    },
+    {
+      "epoch": 0.04261649545013058,
+      "grad_norm": 1.213683843612671,
+      "learning_rate": 0.00012780244485035829,
+      "loss": 5.1458,
+      "step": 30330
+    },
+    {
+      "epoch": 0.042630546388294166,
+      "grad_norm": 1.2280396223068237,
+      "learning_rate": 0.00012784459744274272,
+      "loss": 5.0936,
+      "step": 30340
+    },
+    {
+      "epoch": 0.04264459732645774,
+      "grad_norm": 1.2426551580429077,
+      "learning_rate": 0.00012788675003512715,
+      "loss": 5.0878,
+      "step": 30350
+    },
+    {
+      "epoch": 0.04265864826462132,
+      "grad_norm": 1.2090293169021606,
+      "learning_rate": 0.00012792890262751158,
+      "loss": 5.1355,
+      "step": 30360
+    },
+    {
+      "epoch": 0.042672699202784894,
+      "grad_norm": 1.1868354082107544,
+      "learning_rate": 0.000127971055219896,
+      "loss": 5.0465,
+      "step": 30370
+    },
+    {
+      "epoch": 0.04268675014094847,
+      "grad_norm": 1.1995588541030884,
+      "learning_rate": 0.00012801320781228044,
+      "loss": 5.1283,
+      "step": 30380
+    },
+    {
+      "epoch": 0.042700801079112054,
+      "grad_norm": 1.3185629844665527,
+      "learning_rate": 0.00012805536040466487,
+      "loss": 5.1045,
+      "step": 30390
+    },
+    {
+      "epoch": 0.04271485201727563,
+      "grad_norm": 1.1966207027435303,
+      "learning_rate": 0.0001280975129970493,
+      "loss": 4.9284,
+      "step": 30400
+    },
+    {
+      "epoch": 0.042728902955439206,
+      "grad_norm": 1.1545592546463013,
+      "learning_rate": 0.00012813966558943373,
+      "loss": 5.1675,
+      "step": 30410
+    },
+    {
+      "epoch": 0.04274295389360278,
+      "grad_norm": 1.1954106092453003,
+      "learning_rate": 0.00012818181818181817,
+      "loss": 5.1079,
+      "step": 30420
+    },
+    {
+      "epoch": 0.04275700483176636,
+      "grad_norm": 1.1693055629730225,
+      "learning_rate": 0.0001282239707742026,
+      "loss": 5.2517,
+      "step": 30430
+    },
+    {
+      "epoch": 0.04277105576992994,
+      "grad_norm": 1.2295222282409668,
+      "learning_rate": 0.00012826612336658703,
+      "loss": 5.117,
+      "step": 30440
+    },
+    {
+      "epoch": 0.04278510670809352,
+      "grad_norm": 1.2061803340911865,
+      "learning_rate": 0.00012830827595897149,
+      "loss": 5.0885,
+      "step": 30450
+    },
+    {
+      "epoch": 0.042799157646257094,
+      "grad_norm": 1.1537740230560303,
+      "learning_rate": 0.0001283504285513559,
+      "loss": 5.1159,
+      "step": 30460
+    },
+    {
+      "epoch": 0.04281320858442067,
+      "grad_norm": 1.150728464126587,
+      "learning_rate": 0.00012839258114374032,
+      "loss": 5.2855,
+      "step": 30470
+    },
+    {
+      "epoch": 0.042827259522584246,
+      "grad_norm": 1.178875207901001,
+      "learning_rate": 0.00012843473373612478,
+      "loss": 5.2431,
+      "step": 30480
+    },
+    {
+      "epoch": 0.04284131046074783,
+      "grad_norm": 1.1979405879974365,
+      "learning_rate": 0.00012847688632850918,
+      "loss": 5.0651,
+      "step": 30490
+    },
+    {
+      "epoch": 0.042855361398911405,
+      "grad_norm": 1.142595648765564,
+      "learning_rate": 0.00012851903892089361,
+      "loss": 5.0605,
+      "step": 30500
+    },
+    {
+      "epoch": 0.04286941233707498,
+      "grad_norm": 1.142235279083252,
+      "learning_rate": 0.00012856119151327807,
+      "loss": 5.1623,
+      "step": 30510
+    },
+    {
+      "epoch": 0.04288346327523856,
+      "grad_norm": 1.2095253467559814,
+      "learning_rate": 0.00012860334410566248,
+      "loss": 5.0609,
+      "step": 30520
+    },
+    {
+      "epoch": 0.042897514213402134,
+      "grad_norm": 1.1336437463760376,
+      "learning_rate": 0.0001286454966980469,
+      "loss": 5.0318,
+      "step": 30530
+    },
+    {
+      "epoch": 0.04291156515156572,
+      "grad_norm": 1.2415790557861328,
+      "learning_rate": 0.00012868764929043137,
+      "loss": 5.2027,
+      "step": 30540
+    },
+    {
+      "epoch": 0.04292561608972929,
+      "grad_norm": 1.2634472846984863,
+      "learning_rate": 0.00012872980188281577,
+      "loss": 5.1282,
+      "step": 30550
+    },
+    {
+      "epoch": 0.04293966702789287,
+      "grad_norm": 1.198366403579712,
+      "learning_rate": 0.0001287719544752002,
+      "loss": 5.1052,
+      "step": 30560
+    },
+    {
+      "epoch": 0.042953717966056446,
+      "grad_norm": 1.2169760465621948,
+      "learning_rate": 0.00012881410706758466,
+      "loss": 5.0278,
+      "step": 30570
+    },
+    {
+      "epoch": 0.04296776890422002,
+      "grad_norm": 1.2113124132156372,
+      "learning_rate": 0.00012885625965996906,
+      "loss": 5.0952,
+      "step": 30580
+    },
+    {
+      "epoch": 0.0429818198423836,
+      "grad_norm": 1.247658371925354,
+      "learning_rate": 0.0001288984122523535,
+      "loss": 5.1624,
+      "step": 30590
+    },
+    {
+      "epoch": 0.04299587078054718,
+      "grad_norm": 1.3335022926330566,
+      "learning_rate": 0.00012894056484473795,
+      "loss": 5.016,
+      "step": 30600
+    },
+    {
+      "epoch": 0.04300992171871076,
+      "grad_norm": 1.1578710079193115,
+      "learning_rate": 0.00012898271743712238,
+      "loss": 5.2555,
+      "step": 30610
+    },
+    {
+      "epoch": 0.04302397265687433,
+      "grad_norm": 1.1831419467926025,
+      "learning_rate": 0.0001290248700295068,
+      "loss": 5.2272,
+      "step": 30620
+    },
+    {
+      "epoch": 0.04303802359503791,
+      "grad_norm": 1.3762239217758179,
+      "learning_rate": 0.00012906702262189125,
+      "loss": 5.1915,
+      "step": 30630
+    },
+    {
+      "epoch": 0.043052074533201486,
+      "grad_norm": 1.204743504524231,
+      "learning_rate": 0.00012910917521427568,
+      "loss": 5.1658,
+      "step": 30640
+    },
+    {
+      "epoch": 0.04306612547136507,
+      "grad_norm": 1.2079994678497314,
+      "learning_rate": 0.00012915132780666008,
+      "loss": 5.0166,
+      "step": 30650
+    },
+    {
+      "epoch": 0.043080176409528645,
+      "grad_norm": 1.2461872100830078,
+      "learning_rate": 0.00012919348039904454,
+      "loss": 5.0831,
+      "step": 30660
+    },
+    {
+      "epoch": 0.04309422734769222,
+      "grad_norm": 1.1971070766448975,
+      "learning_rate": 0.00012923563299142897,
+      "loss": 5.1336,
+      "step": 30670
+    },
+    {
+      "epoch": 0.0431082782858558,
+      "grad_norm": 1.1303532123565674,
+      "learning_rate": 0.00012927778558381338,
+      "loss": 5.0732,
+      "step": 30680
+    },
+    {
+      "epoch": 0.04312232922401937,
+      "grad_norm": 1.1667498350143433,
+      "learning_rate": 0.00012931993817619783,
+      "loss": 5.2159,
+      "step": 30690
+    },
+    {
+      "epoch": 0.043136380162182957,
+      "grad_norm": 1.159986972808838,
+      "learning_rate": 0.00012936209076858227,
+      "loss": 5.1623,
+      "step": 30700
+    },
+    {
+      "epoch": 0.04315043110034653,
+      "grad_norm": 1.1590168476104736,
+      "learning_rate": 0.00012940424336096667,
+      "loss": 5.1168,
+      "step": 30710
+    },
+    {
+      "epoch": 0.04316448203851011,
+      "grad_norm": 1.156524658203125,
+      "learning_rate": 0.00012944639595335113,
+      "loss": 5.0177,
+      "step": 30720
+    },
+    {
+      "epoch": 0.043178532976673685,
+      "grad_norm": 1.2327642440795898,
+      "learning_rate": 0.00012948854854573556,
+      "loss": 5.1663,
+      "step": 30730
+    },
+    {
+      "epoch": 0.04319258391483726,
+      "grad_norm": 1.215132236480713,
+      "learning_rate": 0.00012953070113812,
+      "loss": 5.1787,
+      "step": 30740
+    },
+    {
+      "epoch": 0.043206634853000844,
+      "grad_norm": 1.198944091796875,
+      "learning_rate": 0.00012957285373050442,
+      "loss": 5.1268,
+      "step": 30750
+    },
+    {
+      "epoch": 0.04322068579116442,
+      "grad_norm": 1.1945573091506958,
+      "learning_rate": 0.00012961500632288885,
+      "loss": 5.1569,
+      "step": 30760
+    },
+    {
+      "epoch": 0.043234736729328,
+      "grad_norm": 1.1851938962936401,
+      "learning_rate": 0.00012965715891527328,
+      "loss": 5.0766,
+      "step": 30770
+    },
+    {
+      "epoch": 0.04324878766749157,
+      "grad_norm": 1.2323837280273438,
+      "learning_rate": 0.00012969931150765771,
+      "loss": 5.0651,
+      "step": 30780
+    },
+    {
+      "epoch": 0.04326283860565515,
+      "grad_norm": 1.1993993520736694,
+      "learning_rate": 0.00012974146410004215,
+      "loss": 4.9883,
+      "step": 30790
+    },
+    {
+      "epoch": 0.04327688954381873,
+      "grad_norm": 1.2613015174865723,
+      "learning_rate": 0.00012978361669242658,
+      "loss": 5.0884,
+      "step": 30800
+    },
+    {
+      "epoch": 0.04329094048198231,
+      "grad_norm": 1.2269909381866455,
+      "learning_rate": 0.000129825769284811,
+      "loss": 5.1133,
+      "step": 30810
+    },
+    {
+      "epoch": 0.043304991420145884,
+      "grad_norm": 1.2280735969543457,
+      "learning_rate": 0.00012986792187719544,
+      "loss": 4.8907,
+      "step": 30820
+    },
+    {
+      "epoch": 0.04331904235830946,
+      "grad_norm": 1.1935653686523438,
+      "learning_rate": 0.00012991007446957987,
+      "loss": 5.042,
+      "step": 30830
+    },
+    {
+      "epoch": 0.04333309329647304,
+      "grad_norm": 1.1728026866912842,
+      "learning_rate": 0.0001299522270619643,
+      "loss": 5.0976,
+      "step": 30840
+    },
+    {
+      "epoch": 0.04334714423463662,
+      "grad_norm": 1.1702702045440674,
+      "learning_rate": 0.00012999437965434873,
+      "loss": 5.0627,
+      "step": 30850
+    },
+    {
+      "epoch": 0.043361195172800196,
+      "grad_norm": 1.144583821296692,
+      "learning_rate": 0.00013003653224673316,
+      "loss": 4.9926,
+      "step": 30860
+    },
+    {
+      "epoch": 0.04337524611096377,
+      "grad_norm": 1.1992604732513428,
+      "learning_rate": 0.0001300786848391176,
+      "loss": 5.1402,
+      "step": 30870
+    },
+    {
+      "epoch": 0.04338929704912735,
+      "grad_norm": 1.2514524459838867,
+      "learning_rate": 0.00013012083743150203,
+      "loss": 5.2172,
+      "step": 30880
+    },
+    {
+      "epoch": 0.043403347987290924,
+      "grad_norm": 1.1470156908035278,
+      "learning_rate": 0.00013016299002388646,
+      "loss": 5.0893,
+      "step": 30890
+    },
+    {
+      "epoch": 0.0434173989254545,
+      "grad_norm": 1.1299411058425903,
+      "learning_rate": 0.0001302051426162709,
+      "loss": 5.1019,
+      "step": 30900
+    },
+    {
+      "epoch": 0.043431449863618084,
+      "grad_norm": 1.2057139873504639,
+      "learning_rate": 0.00013024729520865532,
+      "loss": 5.0939,
+      "step": 30910
+    },
+    {
+      "epoch": 0.04344550080178166,
+      "grad_norm": 1.108620285987854,
+      "learning_rate": 0.00013028944780103975,
+      "loss": 5.1344,
+      "step": 30920
+    },
+    {
+      "epoch": 0.043459551739945236,
+      "grad_norm": 1.22635817527771,
+      "learning_rate": 0.00013033160039342418,
+      "loss": 5.0597,
+      "step": 30930
+    },
+    {
+      "epoch": 0.04347360267810881,
+      "grad_norm": 1.2512584924697876,
+      "learning_rate": 0.0001303737529858086,
+      "loss": 4.9751,
+      "step": 30940
+    },
+    {
+      "epoch": 0.04348765361627239,
+      "grad_norm": 1.198045253753662,
+      "learning_rate": 0.00013041590557819304,
+      "loss": 5.0859,
+      "step": 30950
+    },
+    {
+      "epoch": 0.04350170455443597,
+      "grad_norm": 1.21504545211792,
+      "learning_rate": 0.00013045805817057748,
+      "loss": 5.1187,
+      "step": 30960
+    },
+    {
+      "epoch": 0.04351575549259955,
+      "grad_norm": 1.1842291355133057,
+      "learning_rate": 0.0001305002107629619,
+      "loss": 5.0685,
+      "step": 30970
+    },
+    {
+      "epoch": 0.043529806430763124,
+      "grad_norm": 1.1496257781982422,
+      "learning_rate": 0.00013054236335534634,
+      "loss": 5.1684,
+      "step": 30980
+    },
+    {
+      "epoch": 0.0435438573689267,
+      "grad_norm": 1.2007911205291748,
+      "learning_rate": 0.00013058451594773077,
+      "loss": 5.1206,
+      "step": 30990
+    },
+    {
+      "epoch": 0.043557908307090276,
+      "grad_norm": 1.1559113264083862,
+      "learning_rate": 0.0001306266685401152,
+      "loss": 5.1135,
+      "step": 31000
+    },
+    {
+      "epoch": 0.04357195924525386,
+      "grad_norm": 1.2255867719650269,
+      "learning_rate": 0.00013066882113249963,
+      "loss": 5.0647,
+      "step": 31010
+    },
+    {
+      "epoch": 0.043586010183417435,
+      "grad_norm": 1.1707844734191895,
+      "learning_rate": 0.00013071097372488406,
+      "loss": 5.0317,
+      "step": 31020
+    },
+    {
+      "epoch": 0.04360006112158101,
+      "grad_norm": 1.1722240447998047,
+      "learning_rate": 0.00013075312631726852,
+      "loss": 4.9861,
+      "step": 31030
+    },
+    {
+      "epoch": 0.04361411205974459,
+      "grad_norm": 1.164713740348816,
+      "learning_rate": 0.00013079527890965293,
+      "loss": 5.2333,
+      "step": 31040
+    },
+    {
+      "epoch": 0.043628162997908164,
+      "grad_norm": 1.2219550609588623,
+      "learning_rate": 0.00013083743150203736,
+      "loss": 5.118,
+      "step": 31050
+    },
+    {
+      "epoch": 0.04364221393607175,
+      "grad_norm": 1.173423171043396,
+      "learning_rate": 0.00013087958409442181,
+      "loss": 5.1398,
+      "step": 31060
+    },
+    {
+      "epoch": 0.04365626487423532,
+      "grad_norm": 1.227458119392395,
+      "learning_rate": 0.00013092173668680622,
+      "loss": 5.0945,
+      "step": 31070
+    },
+    {
+      "epoch": 0.0436703158123989,
+      "grad_norm": 1.2771947383880615,
+      "learning_rate": 0.00013096388927919065,
+      "loss": 5.0949,
+      "step": 31080
+    },
+    {
+      "epoch": 0.043684366750562476,
+      "grad_norm": 1.218270182609558,
+      "learning_rate": 0.0001310060418715751,
+      "loss": 5.1639,
+      "step": 31090
+    },
+    {
+      "epoch": 0.04369841768872605,
+      "grad_norm": 1.195109248161316,
+      "learning_rate": 0.0001310481944639595,
+      "loss": 5.1404,
+      "step": 31100
+    },
+    {
+      "epoch": 0.043712468626889635,
+      "grad_norm": 1.2469673156738281,
+      "learning_rate": 0.00013109034705634394,
+      "loss": 5.086,
+      "step": 31110
+    },
+    {
+      "epoch": 0.04372651956505321,
+      "grad_norm": 1.1276779174804688,
+      "learning_rate": 0.0001311324996487284,
+      "loss": 5.0138,
+      "step": 31120
+    },
+    {
+      "epoch": 0.04374057050321679,
+      "grad_norm": 1.1672612428665161,
+      "learning_rate": 0.0001311746522411128,
+      "loss": 5.0668,
+      "step": 31130
+    },
+    {
+      "epoch": 0.04375462144138036,
+      "grad_norm": 1.2316880226135254,
+      "learning_rate": 0.00013121680483349724,
+      "loss": 5.0723,
+      "step": 31140
+    },
+    {
+      "epoch": 0.04376867237954394,
+      "grad_norm": 1.3245317935943604,
+      "learning_rate": 0.0001312589574258817,
+      "loss": 5.0082,
+      "step": 31150
+    },
+    {
+      "epoch": 0.04378272331770752,
+      "grad_norm": 1.2602311372756958,
+      "learning_rate": 0.0001313011100182661,
+      "loss": 4.9338,
+      "step": 31160
+    },
+    {
+      "epoch": 0.0437967742558711,
+      "grad_norm": 1.1890308856964111,
+      "learning_rate": 0.00013134326261065053,
+      "loss": 5.1929,
+      "step": 31170
+    },
+    {
+      "epoch": 0.043810825194034675,
+      "grad_norm": 1.2277629375457764,
+      "learning_rate": 0.000131385415203035,
+      "loss": 4.9834,
+      "step": 31180
+    },
+    {
+      "epoch": 0.04382487613219825,
+      "grad_norm": 1.142812728881836,
+      "learning_rate": 0.00013142756779541942,
+      "loss": 5.1015,
+      "step": 31190
+    },
+    {
+      "epoch": 0.04383892707036183,
+      "grad_norm": 1.1541565656661987,
+      "learning_rate": 0.00013146972038780382,
+      "loss": 5.0397,
+      "step": 31200
+    },
+    {
+      "epoch": 0.0438529780085254,
+      "grad_norm": 1.1937357187271118,
+      "learning_rate": 0.00013151187298018828,
+      "loss": 5.0137,
+      "step": 31210
+    },
+    {
+      "epoch": 0.043867028946688987,
+      "grad_norm": 1.1668943166732788,
+      "learning_rate": 0.0001315540255725727,
+      "loss": 5.1796,
+      "step": 31220
+    },
+    {
+      "epoch": 0.04388107988485256,
+      "grad_norm": 1.1891443729400635,
+      "learning_rate": 0.00013159617816495712,
+      "loss": 5.059,
+      "step": 31230
+    },
+    {
+      "epoch": 0.04389513082301614,
+      "grad_norm": 1.1822489500045776,
+      "learning_rate": 0.00013163833075734158,
+      "loss": 5.1285,
+      "step": 31240
+    },
+    {
+      "epoch": 0.043909181761179715,
+      "grad_norm": 1.1516724824905396,
+      "learning_rate": 0.000131680483349726,
+      "loss": 5.1245,
+      "step": 31250
+    },
+    {
+      "epoch": 0.04392323269934329,
+      "grad_norm": 1.2096267938613892,
+      "learning_rate": 0.0001317226359421104,
+      "loss": 5.1178,
+      "step": 31260
+    },
+    {
+      "epoch": 0.043937283637506874,
+      "grad_norm": 1.2081648111343384,
+      "learning_rate": 0.00013176478853449487,
+      "loss": 5.1561,
+      "step": 31270
+    },
+    {
+      "epoch": 0.04395133457567045,
+      "grad_norm": 1.1660428047180176,
+      "learning_rate": 0.0001318069411268793,
+      "loss": 5.2086,
+      "step": 31280
+    },
+    {
+      "epoch": 0.04396538551383403,
+      "grad_norm": 1.153578519821167,
+      "learning_rate": 0.0001318490937192637,
+      "loss": 5.156,
+      "step": 31290
+    },
+    {
+      "epoch": 0.0439794364519976,
+      "grad_norm": 1.2928626537322998,
+      "learning_rate": 0.00013189124631164816,
+      "loss": 5.0846,
+      "step": 31300
+    },
+    {
+      "epoch": 0.04399348739016118,
+      "grad_norm": 1.1599982976913452,
+      "learning_rate": 0.0001319333989040326,
+      "loss": 5.1444,
+      "step": 31310
+    },
+    {
+      "epoch": 0.04400753832832476,
+      "grad_norm": 1.1558305025100708,
+      "learning_rate": 0.00013197555149641702,
+      "loss": 5.0282,
+      "step": 31320
+    },
+    {
+      "epoch": 0.04402158926648834,
+      "grad_norm": 1.168467402458191,
+      "learning_rate": 0.00013201770408880146,
+      "loss": 4.9011,
+      "step": 31330
+    },
+    {
+      "epoch": 0.044035640204651914,
+      "grad_norm": 1.1910271644592285,
+      "learning_rate": 0.0001320598566811859,
+      "loss": 5.0972,
+      "step": 31340
+    },
+    {
+      "epoch": 0.04404969114281549,
+      "grad_norm": 1.1661889553070068,
+      "learning_rate": 0.00013210200927357032,
+      "loss": 5.0018,
+      "step": 31350
+    },
+    {
+      "epoch": 0.04406374208097907,
+      "grad_norm": 1.1902189254760742,
+      "learning_rate": 0.00013214416186595475,
+      "loss": 5.1285,
+      "step": 31360
+    },
+    {
+      "epoch": 0.04407779301914265,
+      "grad_norm": 1.191560983657837,
+      "learning_rate": 0.00013218631445833918,
+      "loss": 5.0772,
+      "step": 31370
+    },
+    {
+      "epoch": 0.044091843957306226,
+      "grad_norm": 1.108267903327942,
+      "learning_rate": 0.0001322284670507236,
+      "loss": 5.1525,
+      "step": 31380
+    },
+    {
+      "epoch": 0.0441058948954698,
+      "grad_norm": 1.2221838235855103,
+      "learning_rate": 0.00013227061964310804,
+      "loss": 5.2337,
+      "step": 31390
+    },
+    {
+      "epoch": 0.04411994583363338,
+      "grad_norm": 1.19728684425354,
+      "learning_rate": 0.00013231277223549247,
+      "loss": 5.1011,
+      "step": 31400
+    },
+    {
+      "epoch": 0.044133996771796954,
+      "grad_norm": 1.1746065616607666,
+      "learning_rate": 0.0001323549248278769,
+      "loss": 5.1121,
+      "step": 31410
+    },
+    {
+      "epoch": 0.04414804770996054,
+      "grad_norm": 1.2161784172058105,
+      "learning_rate": 0.00013239707742026134,
+      "loss": 5.1285,
+      "step": 31420
+    },
+    {
+      "epoch": 0.044162098648124114,
+      "grad_norm": 1.2407193183898926,
+      "learning_rate": 0.00013243923001264577,
+      "loss": 5.0633,
+      "step": 31430
+    },
+    {
+      "epoch": 0.04417614958628769,
+      "grad_norm": 1.1342822313308716,
+      "learning_rate": 0.0001324813826050302,
+      "loss": 5.0568,
+      "step": 31440
+    },
+    {
+      "epoch": 0.044190200524451266,
+      "grad_norm": 1.2115027904510498,
+      "learning_rate": 0.00013252353519741463,
+      "loss": 4.9452,
+      "step": 31450
+    },
+    {
+      "epoch": 0.04420425146261484,
+      "grad_norm": 1.20106840133667,
+      "learning_rate": 0.00013256568778979906,
+      "loss": 5.0924,
+      "step": 31460
+    },
+    {
+      "epoch": 0.044218302400778425,
+      "grad_norm": 1.1099613904953003,
+      "learning_rate": 0.0001326078403821835,
+      "loss": 5.07,
+      "step": 31470
+    },
+    {
+      "epoch": 0.044232353338942,
+      "grad_norm": 1.1473101377487183,
+      "learning_rate": 0.00013264999297456792,
+      "loss": 5.069,
+      "step": 31480
+    },
+    {
+      "epoch": 0.04424640427710558,
+      "grad_norm": 1.1742898225784302,
+      "learning_rate": 0.00013269214556695235,
+      "loss": 5.1199,
+      "step": 31490
+    },
+    {
+      "epoch": 0.044260455215269154,
+      "grad_norm": 1.1407841444015503,
+      "learning_rate": 0.00013273429815933679,
+      "loss": 5.0929,
+      "step": 31500
+    },
+    {
+      "epoch": 0.04427450615343273,
+      "grad_norm": 1.17631196975708,
+      "learning_rate": 0.00013277645075172122,
+      "loss": 4.9963,
+      "step": 31510
+    },
+    {
+      "epoch": 0.044288557091596306,
+      "grad_norm": 1.1480594873428345,
+      "learning_rate": 0.00013281860334410565,
+      "loss": 5.11,
+      "step": 31520
+    },
+    {
+      "epoch": 0.04430260802975989,
+      "grad_norm": 1.2020182609558105,
+      "learning_rate": 0.00013286075593649008,
+      "loss": 5.0679,
+      "step": 31530
+    },
+    {
+      "epoch": 0.044316658967923465,
+      "grad_norm": 1.230470061302185,
+      "learning_rate": 0.0001329029085288745,
+      "loss": 4.9922,
+      "step": 31540
+    },
+    {
+      "epoch": 0.04433070990608704,
+      "grad_norm": 1.2092961072921753,
+      "learning_rate": 0.00013294506112125894,
+      "loss": 5.0695,
+      "step": 31550
+    },
+    {
+      "epoch": 0.04434476084425062,
+      "grad_norm": 1.2055702209472656,
+      "learning_rate": 0.00013298721371364337,
+      "loss": 5.1677,
+      "step": 31560
+    },
+    {
+      "epoch": 0.044358811782414194,
+      "grad_norm": 1.1955968141555786,
+      "learning_rate": 0.0001330293663060278,
+      "loss": 5.1062,
+      "step": 31570
+    },
+    {
+      "epoch": 0.04437286272057778,
+      "grad_norm": 1.1392838954925537,
+      "learning_rate": 0.00013307151889841224,
+      "loss": 5.0741,
+      "step": 31580
+    },
+    {
+      "epoch": 0.04438691365874135,
+      "grad_norm": 1.1732559204101562,
+      "learning_rate": 0.00013311367149079667,
+      "loss": 5.0425,
+      "step": 31590
+    },
+    {
+      "epoch": 0.04440096459690493,
+      "grad_norm": 1.1739134788513184,
+      "learning_rate": 0.0001331558240831811,
+      "loss": 5.0015,
+      "step": 31600
+    },
+    {
+      "epoch": 0.044415015535068506,
+      "grad_norm": 1.175411581993103,
+      "learning_rate": 0.00013319797667556556,
+      "loss": 5.0818,
+      "step": 31610
+    },
+    {
+      "epoch": 0.04442906647323208,
+      "grad_norm": 1.2318931818008423,
+      "learning_rate": 0.00013324012926794996,
+      "loss": 4.9621,
+      "step": 31620
+    },
+    {
+      "epoch": 0.044443117411395665,
+      "grad_norm": 1.146024227142334,
+      "learning_rate": 0.0001332822818603344,
+      "loss": 5.061,
+      "step": 31630
+    },
+    {
+      "epoch": 0.04445716834955924,
+      "grad_norm": 1.1564781665802002,
+      "learning_rate": 0.00013332443445271885,
+      "loss": 5.2572,
+      "step": 31640
+    },
+    {
+      "epoch": 0.04447121928772282,
+      "grad_norm": 1.1591886281967163,
+      "learning_rate": 0.00013336658704510325,
+      "loss": 5.1789,
+      "step": 31650
+    },
+    {
+      "epoch": 0.04448527022588639,
+      "grad_norm": 1.172905445098877,
+      "learning_rate": 0.00013340873963748768,
+      "loss": 5.0811,
+      "step": 31660
+    },
+    {
+      "epoch": 0.04449932116404997,
+      "grad_norm": 1.1963396072387695,
+      "learning_rate": 0.00013345089222987214,
+      "loss": 5.15,
+      "step": 31670
+    },
+    {
+      "epoch": 0.04451337210221355,
+      "grad_norm": 1.234837532043457,
+      "learning_rate": 0.00013349304482225655,
+      "loss": 5.1728,
+      "step": 31680
+    },
+    {
+      "epoch": 0.04452742304037713,
+      "grad_norm": 1.179028868675232,
+      "learning_rate": 0.00013353519741464098,
+      "loss": 5.0332,
+      "step": 31690
+    },
+    {
+      "epoch": 0.044541473978540705,
+      "grad_norm": 1.1915132999420166,
+      "learning_rate": 0.00013357735000702544,
+      "loss": 5.1575,
+      "step": 31700
+    },
+    {
+      "epoch": 0.04455552491670428,
+      "grad_norm": 1.248386263847351,
+      "learning_rate": 0.00013361950259940984,
+      "loss": 5.1046,
+      "step": 31710
+    },
+    {
+      "epoch": 0.04456957585486786,
+      "grad_norm": 1.134567141532898,
+      "learning_rate": 0.00013366165519179427,
+      "loss": 5.0822,
+      "step": 31720
+    },
+    {
+      "epoch": 0.04458362679303144,
+      "grad_norm": 1.2228316068649292,
+      "learning_rate": 0.00013370380778417873,
+      "loss": 4.9389,
+      "step": 31730
+    },
+    {
+      "epoch": 0.044597677731195016,
+      "grad_norm": 1.143936276435852,
+      "learning_rate": 0.00013374596037656316,
+      "loss": 5.0541,
+      "step": 31740
+    },
+    {
+      "epoch": 0.04461172866935859,
+      "grad_norm": 1.1601719856262207,
+      "learning_rate": 0.00013378811296894757,
+      "loss": 5.1395,
+      "step": 31750
+    },
+    {
+      "epoch": 0.04462577960752217,
+      "grad_norm": 1.1841803789138794,
+      "learning_rate": 0.00013383026556133202,
+      "loss": 5.0936,
+      "step": 31760
+    },
+    {
+      "epoch": 0.044639830545685745,
+      "grad_norm": 1.1612204313278198,
+      "learning_rate": 0.00013387241815371645,
+      "loss": 5.0203,
+      "step": 31770
+    },
+    {
+      "epoch": 0.04465388148384933,
+      "grad_norm": 1.2054232358932495,
+      "learning_rate": 0.00013391457074610086,
+      "loss": 5.0786,
+      "step": 31780
+    },
+    {
+      "epoch": 0.044667932422012904,
+      "grad_norm": 1.155108094215393,
+      "learning_rate": 0.00013395672333848532,
+      "loss": 4.9753,
+      "step": 31790
+    },
+    {
+      "epoch": 0.04468198336017648,
+      "grad_norm": 1.1976162195205688,
+      "learning_rate": 0.00013399887593086975,
+      "loss": 5.0861,
+      "step": 31800
+    },
+    {
+      "epoch": 0.04469603429834006,
+      "grad_norm": 1.157468557357788,
+      "learning_rate": 0.00013404102852325415,
+      "loss": 5.1192,
+      "step": 31810
+    },
+    {
+      "epoch": 0.04471008523650363,
+      "grad_norm": 1.1830439567565918,
+      "learning_rate": 0.0001340831811156386,
+      "loss": 5.1491,
+      "step": 31820
+    },
+    {
+      "epoch": 0.04472413617466721,
+      "grad_norm": 1.2510123252868652,
+      "learning_rate": 0.00013412533370802304,
+      "loss": 5.0752,
+      "step": 31830
+    },
+    {
+      "epoch": 0.04473818711283079,
+      "grad_norm": 1.151624083518982,
+      "learning_rate": 0.00013416748630040745,
+      "loss": 5.0527,
+      "step": 31840
+    },
+    {
+      "epoch": 0.04475223805099437,
+      "grad_norm": 1.1923450231552124,
+      "learning_rate": 0.0001342096388927919,
+      "loss": 5.0977,
+      "step": 31850
+    },
+    {
+      "epoch": 0.044766288989157944,
+      "grad_norm": 1.1659468412399292,
+      "learning_rate": 0.00013425179148517633,
+      "loss": 4.9261,
+      "step": 31860
+    },
+    {
+      "epoch": 0.04478033992732152,
+      "grad_norm": 1.139060378074646,
+      "learning_rate": 0.00013429394407756074,
+      "loss": 5.0291,
+      "step": 31870
+    },
+    {
+      "epoch": 0.0447943908654851,
+      "grad_norm": 1.134356141090393,
+      "learning_rate": 0.0001343360966699452,
+      "loss": 5.1025,
+      "step": 31880
+    },
+    {
+      "epoch": 0.04480844180364868,
+      "grad_norm": 1.1306984424591064,
+      "learning_rate": 0.00013437824926232963,
+      "loss": 5.0889,
+      "step": 31890
+    },
+    {
+      "epoch": 0.044822492741812256,
+      "grad_norm": 1.1675775051116943,
+      "learning_rate": 0.00013442040185471406,
+      "loss": 4.949,
+      "step": 31900
+    },
+    {
+      "epoch": 0.04483654367997583,
+      "grad_norm": 1.2445342540740967,
+      "learning_rate": 0.0001344625544470985,
+      "loss": 4.9771,
+      "step": 31910
+    },
+    {
+      "epoch": 0.04485059461813941,
+      "grad_norm": 1.1660617589950562,
+      "learning_rate": 0.00013450470703948292,
+      "loss": 5.0768,
+      "step": 31920
+    },
+    {
+      "epoch": 0.044864645556302984,
+      "grad_norm": 1.2834054231643677,
+      "learning_rate": 0.00013454685963186735,
+      "loss": 4.9632,
+      "step": 31930
+    },
+    {
+      "epoch": 0.04487869649446657,
+      "grad_norm": 1.2516728639602661,
+      "learning_rate": 0.00013458901222425178,
+      "loss": 5.091,
+      "step": 31940
+    },
+    {
+      "epoch": 0.044892747432630144,
+      "grad_norm": 1.1859095096588135,
+      "learning_rate": 0.00013463116481663622,
+      "loss": 4.9886,
+      "step": 31950
+    },
+    {
+      "epoch": 0.04490679837079372,
+      "grad_norm": 1.140558123588562,
+      "learning_rate": 0.00013467331740902065,
+      "loss": 5.1474,
+      "step": 31960
+    },
+    {
+      "epoch": 0.044920849308957296,
+      "grad_norm": 1.1825441122055054,
+      "learning_rate": 0.00013471547000140508,
+      "loss": 5.0819,
+      "step": 31970
+    },
+    {
+      "epoch": 0.04493490024712087,
+      "grad_norm": 1.1421507596969604,
+      "learning_rate": 0.0001347576225937895,
+      "loss": 5.0934,
+      "step": 31980
+    },
+    {
+      "epoch": 0.044948951185284455,
+      "grad_norm": 1.1730067729949951,
+      "learning_rate": 0.00013479977518617394,
+      "loss": 4.9571,
+      "step": 31990
+    },
+    {
+      "epoch": 0.04496300212344803,
+      "grad_norm": 1.1735901832580566,
+      "learning_rate": 0.00013484192777855837,
+      "loss": 5.0852,
+      "step": 32000
+    },
+    {
+      "epoch": 0.04497705306161161,
+      "grad_norm": 1.2650270462036133,
+      "learning_rate": 0.0001348840803709428,
+      "loss": 4.9872,
+      "step": 32010
+    },
+    {
+      "epoch": 0.044991103999775184,
+      "grad_norm": 1.2169119119644165,
+      "learning_rate": 0.00013492623296332723,
+      "loss": 5.0531,
+      "step": 32020
+    },
+    {
+      "epoch": 0.04500515493793876,
+      "grad_norm": 1.1884304285049438,
+      "learning_rate": 0.00013496838555571166,
+      "loss": 4.9522,
+      "step": 32030
+    },
+    {
+      "epoch": 0.04501920587610234,
+      "grad_norm": 1.131784439086914,
+      "learning_rate": 0.0001350105381480961,
+      "loss": 5.0842,
+      "step": 32040
+    },
+    {
+      "epoch": 0.04503325681426592,
+      "grad_norm": 1.1771255731582642,
+      "learning_rate": 0.00013505269074048053,
+      "loss": 5.1013,
+      "step": 32050
+    },
+    {
+      "epoch": 0.045047307752429495,
+      "grad_norm": 1.181799054145813,
+      "learning_rate": 0.00013509484333286496,
+      "loss": 5.0919,
+      "step": 32060
+    },
+    {
+      "epoch": 0.04506135869059307,
+      "grad_norm": 1.173443078994751,
+      "learning_rate": 0.0001351369959252494,
+      "loss": 4.9914,
+      "step": 32070
+    },
+    {
+      "epoch": 0.04507540962875665,
+      "grad_norm": 1.1408971548080444,
+      "learning_rate": 0.00013517914851763382,
+      "loss": 5.0476,
+      "step": 32080
+    },
+    {
+      "epoch": 0.04508946056692023,
+      "grad_norm": 1.1605677604675293,
+      "learning_rate": 0.00013522130111001825,
+      "loss": 5.1048,
+      "step": 32090
+    },
+    {
+      "epoch": 0.04510351150508381,
+      "grad_norm": 1.1556565761566162,
+      "learning_rate": 0.00013526345370240268,
+      "loss": 5.0675,
+      "step": 32100
+    },
+    {
+      "epoch": 0.04511756244324738,
+      "grad_norm": 1.1450446844100952,
+      "learning_rate": 0.00013530560629478711,
+      "loss": 5.0533,
+      "step": 32110
+    },
+    {
+      "epoch": 0.04513161338141096,
+      "grad_norm": 1.163759708404541,
+      "learning_rate": 0.00013534775888717155,
+      "loss": 5.1082,
+      "step": 32120
+    },
+    {
+      "epoch": 0.045145664319574536,
+      "grad_norm": 1.2103548049926758,
+      "learning_rate": 0.00013538991147955598,
+      "loss": 5.0448,
+      "step": 32130
+    },
+    {
+      "epoch": 0.04515971525773811,
+      "grad_norm": 1.3155592679977417,
+      "learning_rate": 0.0001354320640719404,
+      "loss": 5.0333,
+      "step": 32140
+    },
+    {
+      "epoch": 0.045173766195901695,
+      "grad_norm": 1.169601321220398,
+      "learning_rate": 0.00013547421666432484,
+      "loss": 5.0035,
+      "step": 32150
+    },
+    {
+      "epoch": 0.04518781713406527,
+      "grad_norm": 1.1616172790527344,
+      "learning_rate": 0.0001355163692567093,
+      "loss": 5.0827,
+      "step": 32160
+    },
+    {
+      "epoch": 0.04520186807222885,
+      "grad_norm": 1.1275705099105835,
+      "learning_rate": 0.0001355585218490937,
+      "loss": 5.0115,
+      "step": 32170
+    },
+    {
+      "epoch": 0.04521591901039242,
+      "grad_norm": 1.1657685041427612,
+      "learning_rate": 0.00013560067444147813,
+      "loss": 5.0359,
+      "step": 32180
+    },
+    {
+      "epoch": 0.045229969948556,
+      "grad_norm": 1.1922658681869507,
+      "learning_rate": 0.0001356428270338626,
+      "loss": 5.0735,
+      "step": 32190
+    },
+    {
+      "epoch": 0.04524402088671958,
+      "grad_norm": 1.262093186378479,
+      "learning_rate": 0.000135684979626247,
+      "loss": 5.0809,
+      "step": 32200
+    },
+    {
+      "epoch": 0.04525807182488316,
+      "grad_norm": 1.1995415687561035,
+      "learning_rate": 0.00013572713221863143,
+      "loss": 5.0154,
+      "step": 32210
+    },
+    {
+      "epoch": 0.045272122763046735,
+      "grad_norm": 1.1763862371444702,
+      "learning_rate": 0.00013576928481101588,
+      "loss": 5.0166,
+      "step": 32220
+    },
+    {
+      "epoch": 0.04528617370121031,
+      "grad_norm": 1.1900880336761475,
+      "learning_rate": 0.0001358114374034003,
+      "loss": 5.0693,
+      "step": 32230
+    },
+    {
+      "epoch": 0.04530022463937389,
+      "grad_norm": 1.1440578699111938,
+      "learning_rate": 0.00013585358999578472,
+      "loss": 5.0902,
+      "step": 32240
+    },
+    {
+      "epoch": 0.04531427557753747,
+      "grad_norm": 1.1583417654037476,
+      "learning_rate": 0.00013589574258816918,
+      "loss": 4.9809,
+      "step": 32250
+    },
+    {
+      "epoch": 0.045328326515701046,
+      "grad_norm": 1.2270220518112183,
+      "learning_rate": 0.00013593789518055358,
+      "loss": 5.099,
+      "step": 32260
+    },
+    {
+      "epoch": 0.04534237745386462,
+      "grad_norm": 1.1560698747634888,
+      "learning_rate": 0.000135980047772938,
+      "loss": 5.1224,
+      "step": 32270
+    },
+    {
+      "epoch": 0.0453564283920282,
+      "grad_norm": 1.1910443305969238,
+      "learning_rate": 0.00013602220036532247,
+      "loss": 5.0196,
+      "step": 32280
+    },
+    {
+      "epoch": 0.045370479330191775,
+      "grad_norm": 1.204832673072815,
+      "learning_rate": 0.00013606435295770688,
+      "loss": 5.0512,
+      "step": 32290
+    },
+    {
+      "epoch": 0.04538453026835536,
+      "grad_norm": 1.3012568950653076,
+      "learning_rate": 0.0001361065055500913,
+      "loss": 5.0248,
+      "step": 32300
+    },
+    {
+      "epoch": 0.045398581206518934,
+      "grad_norm": 1.1138745546340942,
+      "learning_rate": 0.00013614865814247576,
+      "loss": 5.1426,
+      "step": 32310
+    },
+    {
+      "epoch": 0.04541263214468251,
+      "grad_norm": 1.3530911207199097,
+      "learning_rate": 0.0001361908107348602,
+      "loss": 4.9985,
+      "step": 32320
+    },
+    {
+      "epoch": 0.04542668308284609,
+      "grad_norm": 1.1052865982055664,
+      "learning_rate": 0.0001362329633272446,
+      "loss": 5.0387,
+      "step": 32330
+    },
+    {
+      "epoch": 0.04544073402100966,
+      "grad_norm": 1.1460787057876587,
+      "learning_rate": 0.00013627511591962906,
+      "loss": 4.9708,
+      "step": 32340
+    },
+    {
+      "epoch": 0.045454784959173246,
+      "grad_norm": 1.1664546728134155,
+      "learning_rate": 0.0001363172685120135,
+      "loss": 4.9381,
+      "step": 32350
+    },
+    {
+      "epoch": 0.04546883589733682,
+      "grad_norm": 1.1577123403549194,
+      "learning_rate": 0.0001363594211043979,
+      "loss": 4.9815,
+      "step": 32360
+    },
+    {
+      "epoch": 0.0454828868355004,
+      "grad_norm": 1.1973512172698975,
+      "learning_rate": 0.00013640157369678235,
+      "loss": 5.1239,
+      "step": 32370
+    },
+    {
+      "epoch": 0.045496937773663974,
+      "grad_norm": 1.1406227350234985,
+      "learning_rate": 0.00013644372628916678,
+      "loss": 5.0569,
+      "step": 32380
+    },
+    {
+      "epoch": 0.04551098871182755,
+      "grad_norm": 1.1365989446640015,
+      "learning_rate": 0.0001364858788815512,
+      "loss": 4.9423,
+      "step": 32390
+    },
+    {
+      "epoch": 0.045525039649991134,
+      "grad_norm": 1.1738041639328003,
+      "learning_rate": 0.00013652803147393565,
+      "loss": 5.1111,
+      "step": 32400
+    },
+    {
+      "epoch": 0.04553909058815471,
+      "grad_norm": 1.2123348712921143,
+      "learning_rate": 0.00013657018406632008,
+      "loss": 5.1021,
+      "step": 32410
+    },
+    {
+      "epoch": 0.045553141526318286,
+      "grad_norm": 1.1308833360671997,
+      "learning_rate": 0.00013661233665870448,
+      "loss": 5.0533,
+      "step": 32420
+    },
+    {
+      "epoch": 0.04556719246448186,
+      "grad_norm": 1.2196406126022339,
+      "learning_rate": 0.00013665448925108894,
+      "loss": 4.9591,
+      "step": 32430
+    },
+    {
+      "epoch": 0.04558124340264544,
+      "grad_norm": 1.1016089916229248,
+      "learning_rate": 0.00013669664184347337,
+      "loss": 5.0497,
+      "step": 32440
+    },
+    {
+      "epoch": 0.045595294340809014,
+      "grad_norm": 1.248906135559082,
+      "learning_rate": 0.0001367387944358578,
+      "loss": 5.0935,
+      "step": 32450
+    },
+    {
+      "epoch": 0.0456093452789726,
+      "grad_norm": 1.1859537363052368,
+      "learning_rate": 0.00013678094702824223,
+      "loss": 5.1309,
+      "step": 32460
+    },
+    {
+      "epoch": 0.045623396217136174,
+      "grad_norm": 1.2062934637069702,
+      "learning_rate": 0.00013682309962062666,
+      "loss": 5.0923,
+      "step": 32470
+    },
+    {
+      "epoch": 0.04563744715529975,
+      "grad_norm": 1.4972939491271973,
+      "learning_rate": 0.0001368652522130111,
+      "loss": 5.009,
+      "step": 32480
+    },
+    {
+      "epoch": 0.045651498093463326,
+      "grad_norm": 1.1737431287765503,
+      "learning_rate": 0.00013690740480539553,
+      "loss": 5.1537,
+      "step": 32490
+    },
+    {
+      "epoch": 0.0456655490316269,
+      "grad_norm": 1.1139081716537476,
+      "learning_rate": 0.00013694955739777996,
+      "loss": 5.1087,
+      "step": 32500
+    },
+    {
+      "epoch": 0.045679599969790485,
+      "grad_norm": 1.2105811834335327,
+      "learning_rate": 0.0001369917099901644,
+      "loss": 5.0994,
+      "step": 32510
+    },
+    {
+      "epoch": 0.04569365090795406,
+      "grad_norm": 1.2313215732574463,
+      "learning_rate": 0.00013703386258254882,
+      "loss": 5.0346,
+      "step": 32520
+    },
+    {
+      "epoch": 0.04570770184611764,
+      "grad_norm": 1.1877297163009644,
+      "learning_rate": 0.00013707601517493325,
+      "loss": 5.132,
+      "step": 32530
+    },
+    {
+      "epoch": 0.045721752784281214,
+      "grad_norm": 1.1226810216903687,
+      "learning_rate": 0.00013711816776731768,
+      "loss": 5.0537,
+      "step": 32540
+    },
+    {
+      "epoch": 0.04573580372244479,
+      "grad_norm": 1.2374706268310547,
+      "learning_rate": 0.0001371603203597021,
+      "loss": 5.0868,
+      "step": 32550
+    },
+    {
+      "epoch": 0.04574985466060837,
+      "grad_norm": 1.1252933740615845,
+      "learning_rate": 0.00013720247295208654,
+      "loss": 5.1987,
+      "step": 32560
+    },
+    {
+      "epoch": 0.04576390559877195,
+      "grad_norm": 1.1599668264389038,
+      "learning_rate": 0.00013724462554447097,
+      "loss": 5.0369,
+      "step": 32570
+    },
+    {
+      "epoch": 0.045777956536935525,
+      "grad_norm": 1.2881426811218262,
+      "learning_rate": 0.0001372867781368554,
+      "loss": 5.0175,
+      "step": 32580
+    },
+    {
+      "epoch": 0.0457920074750991,
+      "grad_norm": 1.1907261610031128,
+      "learning_rate": 0.00013732893072923984,
+      "loss": 5.0749,
+      "step": 32590
+    },
+    {
+      "epoch": 0.04580605841326268,
+      "grad_norm": 1.1891196966171265,
+      "learning_rate": 0.00013737108332162427,
+      "loss": 5.1202,
+      "step": 32600
+    },
+    {
+      "epoch": 0.04582010935142626,
+      "grad_norm": 1.1137062311172485,
+      "learning_rate": 0.0001374132359140087,
+      "loss": 4.9954,
+      "step": 32610
+    },
+    {
+      "epoch": 0.04583416028958984,
+      "grad_norm": 1.1719475984573364,
+      "learning_rate": 0.00013745538850639313,
+      "loss": 4.9482,
+      "step": 32620
+    },
+    {
+      "epoch": 0.04584821122775341,
+      "grad_norm": 1.1451879739761353,
+      "learning_rate": 0.00013749754109877756,
+      "loss": 4.9943,
+      "step": 32630
+    },
+    {
+      "epoch": 0.04586226216591699,
+      "grad_norm": 1.144889235496521,
+      "learning_rate": 0.000137539693691162,
+      "loss": 5.0704,
+      "step": 32640
+    },
+    {
+      "epoch": 0.045876313104080566,
+      "grad_norm": 1.1692876815795898,
+      "learning_rate": 0.00013758184628354642,
+      "loss": 4.9091,
+      "step": 32650
+    },
+    {
+      "epoch": 0.04589036404224415,
+      "grad_norm": 1.1342777013778687,
+      "learning_rate": 0.00013762399887593086,
+      "loss": 5.0762,
+      "step": 32660
+    },
+    {
+      "epoch": 0.045904414980407725,
+      "grad_norm": 1.1293963193893433,
+      "learning_rate": 0.0001376661514683153,
+      "loss": 5.1474,
+      "step": 32670
+    },
+    {
+      "epoch": 0.0459184659185713,
+      "grad_norm": 1.1727898120880127,
+      "learning_rate": 0.00013770830406069972,
+      "loss": 5.0395,
+      "step": 32680
+    },
+    {
+      "epoch": 0.04593251685673488,
+      "grad_norm": 1.1893597841262817,
+      "learning_rate": 0.00013775045665308415,
+      "loss": 5.0505,
+      "step": 32690
+    },
+    {
+      "epoch": 0.04594656779489845,
+      "grad_norm": 1.1680330038070679,
+      "learning_rate": 0.00013779260924546858,
+      "loss": 5.0726,
+      "step": 32700
+    },
+    {
+      "epoch": 0.045960618733062036,
+      "grad_norm": 1.1110119819641113,
+      "learning_rate": 0.000137834761837853,
+      "loss": 4.9678,
+      "step": 32710
+    },
+    {
+      "epoch": 0.04597466967122561,
+      "grad_norm": 1.1456838846206665,
+      "learning_rate": 0.00013787691443023744,
+      "loss": 5.0045,
+      "step": 32720
+    },
+    {
+      "epoch": 0.04598872060938919,
+      "grad_norm": 1.1601930856704712,
+      "learning_rate": 0.00013791906702262187,
+      "loss": 5.044,
+      "step": 32730
+    },
+    {
+      "epoch": 0.046002771547552765,
+      "grad_norm": 1.1178258657455444,
+      "learning_rate": 0.00013796121961500633,
+      "loss": 4.9794,
+      "step": 32740
+    },
+    {
+      "epoch": 0.04601682248571634,
+      "grad_norm": 1.1255162954330444,
+      "learning_rate": 0.00013800337220739074,
+      "loss": 5.0032,
+      "step": 32750
+    },
+    {
+      "epoch": 0.046030873423879924,
+      "grad_norm": 1.1232777833938599,
+      "learning_rate": 0.00013804552479977517,
+      "loss": 5.0948,
+      "step": 32760
+    },
+    {
+      "epoch": 0.0460449243620435,
+      "grad_norm": 1.158825159072876,
+      "learning_rate": 0.00013808767739215963,
+      "loss": 5.0024,
+      "step": 32770
+    },
+    {
+      "epoch": 0.046058975300207076,
+      "grad_norm": 1.132293462753296,
+      "learning_rate": 0.00013812982998454403,
+      "loss": 5.0553,
+      "step": 32780
+    },
+    {
+      "epoch": 0.04607302623837065,
+      "grad_norm": 1.1477060317993164,
+      "learning_rate": 0.00013817198257692846,
+      "loss": 5.016,
+      "step": 32790
+    },
+    {
+      "epoch": 0.04608707717653423,
+      "grad_norm": 1.1272825002670288,
+      "learning_rate": 0.00013821413516931292,
+      "loss": 5.0562,
+      "step": 32800
+    },
+    {
+      "epoch": 0.046101128114697805,
+      "grad_norm": 1.1596038341522217,
+      "learning_rate": 0.00013825628776169732,
+      "loss": 5.0503,
+      "step": 32810
+    },
+    {
+      "epoch": 0.04611517905286139,
+      "grad_norm": 1.131719708442688,
+      "learning_rate": 0.00013829844035408175,
+      "loss": 4.9989,
+      "step": 32820
+    },
+    {
+      "epoch": 0.046129229991024964,
+      "grad_norm": 1.1962887048721313,
+      "learning_rate": 0.0001383405929464662,
+      "loss": 5.0298,
+      "step": 32830
+    },
+    {
+      "epoch": 0.04614328092918854,
+      "grad_norm": 1.133915901184082,
+      "learning_rate": 0.00013838274553885062,
+      "loss": 4.9067,
+      "step": 32840
+    },
+    {
+      "epoch": 0.04615733186735212,
+      "grad_norm": 1.2771695852279663,
+      "learning_rate": 0.00013842489813123505,
+      "loss": 5.0036,
+      "step": 32850
+    },
+    {
+      "epoch": 0.04617138280551569,
+      "grad_norm": 1.1963088512420654,
+      "learning_rate": 0.0001384670507236195,
+      "loss": 5.0803,
+      "step": 32860
+    },
+    {
+      "epoch": 0.046185433743679276,
+      "grad_norm": 1.0991750955581665,
+      "learning_rate": 0.0001385092033160039,
+      "loss": 5.0325,
+      "step": 32870
+    },
+    {
+      "epoch": 0.04619948468184285,
+      "grad_norm": 1.179234504699707,
+      "learning_rate": 0.00013855135590838834,
+      "loss": 5.0296,
+      "step": 32880
+    },
+    {
+      "epoch": 0.04621353562000643,
+      "grad_norm": 1.1082992553710938,
+      "learning_rate": 0.0001385935085007728,
+      "loss": 5.1175,
+      "step": 32890
+    },
+    {
+      "epoch": 0.046227586558170004,
+      "grad_norm": 1.2209382057189941,
+      "learning_rate": 0.00013863566109315723,
+      "loss": 5.0361,
+      "step": 32900
+    },
+    {
+      "epoch": 0.04624163749633358,
+      "grad_norm": 1.1588517427444458,
+      "learning_rate": 0.00013867781368554163,
+      "loss": 5.1443,
+      "step": 32910
+    },
+    {
+      "epoch": 0.046255688434497164,
+      "grad_norm": 1.0982693433761597,
+      "learning_rate": 0.0001387199662779261,
+      "loss": 5.0724,
+      "step": 32920
+    },
+    {
+      "epoch": 0.04626973937266074,
+      "grad_norm": 1.1611078977584839,
+      "learning_rate": 0.00013876211887031052,
+      "loss": 5.127,
+      "step": 32930
+    },
+    {
+      "epoch": 0.046283790310824316,
+      "grad_norm": 1.1593987941741943,
+      "learning_rate": 0.00013880427146269493,
+      "loss": 5.1295,
+      "step": 32940
+    },
+    {
+      "epoch": 0.04629784124898789,
+      "grad_norm": 1.1332095861434937,
+      "learning_rate": 0.00013884642405507939,
+      "loss": 5.0213,
+      "step": 32950
+    },
+    {
+      "epoch": 0.04631189218715147,
+      "grad_norm": 1.1835927963256836,
+      "learning_rate": 0.00013888857664746382,
+      "loss": 5.0062,
+      "step": 32960
+    },
+    {
+      "epoch": 0.04632594312531505,
+      "grad_norm": 1.1010524034500122,
+      "learning_rate": 0.00013893072923984822,
+      "loss": 5.1005,
+      "step": 32970
+    },
+    {
+      "epoch": 0.04633999406347863,
+      "grad_norm": 1.1981157064437866,
+      "learning_rate": 0.00013897288183223268,
+      "loss": 4.8705,
+      "step": 32980
+    },
+    {
+      "epoch": 0.046354045001642204,
+      "grad_norm": 1.2006856203079224,
+      "learning_rate": 0.0001390150344246171,
+      "loss": 5.0685,
+      "step": 32990
+    },
+    {
+      "epoch": 0.04636809593980578,
+      "grad_norm": 1.1585785150527954,
+      "learning_rate": 0.00013905718701700152,
+      "loss": 5.0302,
+      "step": 33000
+    },
+    {
+      "epoch": 0.046382146877969356,
+      "grad_norm": 1.1677348613739014,
+      "learning_rate": 0.00013909933960938597,
+      "loss": 4.9233,
+      "step": 33010
+    },
+    {
+      "epoch": 0.04639619781613294,
+      "grad_norm": 1.1076332330703735,
+      "learning_rate": 0.0001391414922017704,
+      "loss": 5.0191,
+      "step": 33020
+    },
+    {
+      "epoch": 0.046410248754296515,
+      "grad_norm": 1.0935347080230713,
+      "learning_rate": 0.00013918364479415484,
+      "loss": 5.1139,
+      "step": 33030
+    },
+    {
+      "epoch": 0.04642429969246009,
+      "grad_norm": 1.1202341318130493,
+      "learning_rate": 0.00013922579738653927,
+      "loss": 5.0648,
+      "step": 33040
+    },
+    {
+      "epoch": 0.04643835063062367,
+      "grad_norm": 1.1524226665496826,
+      "learning_rate": 0.0001392679499789237,
+      "loss": 4.8851,
+      "step": 33050
+    },
+    {
+      "epoch": 0.046452401568787244,
+      "grad_norm": 1.19994056224823,
+      "learning_rate": 0.00013931010257130813,
+      "loss": 5.0289,
+      "step": 33060
+    },
+    {
+      "epoch": 0.04646645250695083,
+      "grad_norm": 1.1779851913452148,
+      "learning_rate": 0.00013935225516369256,
+      "loss": 5.0094,
+      "step": 33070
+    },
+    {
+      "epoch": 0.0464805034451144,
+      "grad_norm": 1.186610460281372,
+      "learning_rate": 0.000139394407756077,
+      "loss": 5.0195,
+      "step": 33080
+    },
+    {
+      "epoch": 0.04649455438327798,
+      "grad_norm": 1.1642428636550903,
+      "learning_rate": 0.00013943656034846142,
+      "loss": 5.0686,
+      "step": 33090
+    },
+    {
+      "epoch": 0.046508605321441555,
+      "grad_norm": 1.1461211442947388,
+      "learning_rate": 0.00013947871294084585,
+      "loss": 5.1077,
+      "step": 33100
+    },
+    {
+      "epoch": 0.04652265625960513,
+      "grad_norm": 1.1319652795791626,
+      "learning_rate": 0.00013952086553323029,
+      "loss": 5.0228,
+      "step": 33110
+    },
+    {
+      "epoch": 0.04653670719776871,
+      "grad_norm": 1.2022271156311035,
+      "learning_rate": 0.00013956301812561472,
+      "loss": 5.0792,
+      "step": 33120
+    },
+    {
+      "epoch": 0.04655075813593229,
+      "grad_norm": 1.1919790506362915,
+      "learning_rate": 0.00013960517071799915,
+      "loss": 5.0433,
+      "step": 33130
+    },
+    {
+      "epoch": 0.04656480907409587,
+      "grad_norm": 1.1596547365188599,
+      "learning_rate": 0.00013964732331038358,
+      "loss": 5.0325,
+      "step": 33140
+    },
+    {
+      "epoch": 0.04657886001225944,
+      "grad_norm": 1.16820228099823,
+      "learning_rate": 0.000139689475902768,
+      "loss": 5.0048,
+      "step": 33150
+    },
+    {
+      "epoch": 0.04659291095042302,
+      "grad_norm": 1.422676682472229,
+      "learning_rate": 0.00013973162849515244,
+      "loss": 4.9945,
+      "step": 33160
+    },
+    {
+      "epoch": 0.046606961888586595,
+      "grad_norm": 1.1482213735580444,
+      "learning_rate": 0.00013977378108753687,
+      "loss": 4.9878,
+      "step": 33170
+    },
+    {
+      "epoch": 0.04662101282675018,
+      "grad_norm": 1.2206376791000366,
+      "learning_rate": 0.0001398159336799213,
+      "loss": 5.1302,
+      "step": 33180
+    },
+    {
+      "epoch": 0.046635063764913755,
+      "grad_norm": 1.1734592914581299,
+      "learning_rate": 0.00013985808627230573,
+      "loss": 4.9858,
+      "step": 33190
+    },
+    {
+      "epoch": 0.04664911470307733,
+      "grad_norm": 1.1573774814605713,
+      "learning_rate": 0.00013990023886469017,
+      "loss": 5.0833,
+      "step": 33200
+    },
+    {
+      "epoch": 0.04666316564124091,
+      "grad_norm": 1.1239194869995117,
+      "learning_rate": 0.0001399423914570746,
+      "loss": 4.9798,
+      "step": 33210
+    },
+    {
+      "epoch": 0.04667721657940448,
+      "grad_norm": 1.2762165069580078,
+      "learning_rate": 0.00013998454404945903,
+      "loss": 4.9813,
+      "step": 33220
+    },
+    {
+      "epoch": 0.046691267517568066,
+      "grad_norm": 1.1590991020202637,
+      "learning_rate": 0.00014002669664184346,
+      "loss": 5.0678,
+      "step": 33230
+    },
+    {
+      "epoch": 0.04670531845573164,
+      "grad_norm": 1.2096633911132812,
+      "learning_rate": 0.0001400688492342279,
+      "loss": 5.2786,
+      "step": 33240
+    },
+    {
+      "epoch": 0.04671936939389522,
+      "grad_norm": 1.1752020120620728,
+      "learning_rate": 0.00014011100182661232,
+      "loss": 5.097,
+      "step": 33250
+    },
+    {
+      "epoch": 0.046733420332058795,
+      "grad_norm": 1.1489932537078857,
+      "learning_rate": 0.00014015315441899675,
+      "loss": 5.1115,
+      "step": 33260
+    },
+    {
+      "epoch": 0.04674747127022237,
+      "grad_norm": 1.2402178049087524,
+      "learning_rate": 0.00014019530701138118,
+      "loss": 5.0294,
+      "step": 33270
+    },
+    {
+      "epoch": 0.046761522208385954,
+      "grad_norm": 1.109150767326355,
+      "learning_rate": 0.00014023745960376561,
+      "loss": 5.1305,
+      "step": 33280
+    },
+    {
+      "epoch": 0.04677557314654953,
+      "grad_norm": 1.1309486627578735,
+      "learning_rate": 0.00014027961219615005,
+      "loss": 4.9889,
+      "step": 33290
+    },
+    {
+      "epoch": 0.046789624084713106,
+      "grad_norm": 1.1985018253326416,
+      "learning_rate": 0.00014032176478853448,
+      "loss": 5.0472,
+      "step": 33300
+    },
+    {
+      "epoch": 0.04680367502287668,
+      "grad_norm": 1.231369972229004,
+      "learning_rate": 0.0001403639173809189,
+      "loss": 5.0311,
+      "step": 33310
+    },
+    {
+      "epoch": 0.04681772596104026,
+      "grad_norm": 1.1748765707015991,
+      "learning_rate": 0.00014040606997330337,
+      "loss": 4.9819,
+      "step": 33320
+    },
+    {
+      "epoch": 0.04683177689920384,
+      "grad_norm": 1.2126888036727905,
+      "learning_rate": 0.00014044822256568777,
+      "loss": 4.9947,
+      "step": 33330
+    },
+    {
+      "epoch": 0.04684582783736742,
+      "grad_norm": 1.2520742416381836,
+      "learning_rate": 0.0001404903751580722,
+      "loss": 4.9942,
+      "step": 33340
+    },
+    {
+      "epoch": 0.046859878775530994,
+      "grad_norm": 1.1721147298812866,
+      "learning_rate": 0.00014053252775045666,
+      "loss": 5.0119,
+      "step": 33350
+    },
+    {
+      "epoch": 0.04687392971369457,
+      "grad_norm": 1.1375186443328857,
+      "learning_rate": 0.00014057468034284106,
+      "loss": 5.1042,
+      "step": 33360
+    },
+    {
+      "epoch": 0.04688798065185815,
+      "grad_norm": 1.1971328258514404,
+      "learning_rate": 0.0001406168329352255,
+      "loss": 5.0726,
+      "step": 33370
+    },
+    {
+      "epoch": 0.04690203159002173,
+      "grad_norm": 1.4078949689865112,
+      "learning_rate": 0.00014065898552760995,
+      "loss": 5.0296,
+      "step": 33380
+    },
+    {
+      "epoch": 0.046916082528185306,
+      "grad_norm": 1.2646558284759521,
+      "learning_rate": 0.00014070113811999436,
+      "loss": 5.0794,
+      "step": 33390
+    },
+    {
+      "epoch": 0.04693013346634888,
+      "grad_norm": 1.107875943183899,
+      "learning_rate": 0.0001407432907123788,
+      "loss": 5.0813,
+      "step": 33400
+    },
+    {
+      "epoch": 0.04694418440451246,
+      "grad_norm": 1.1319985389709473,
+      "learning_rate": 0.00014078544330476325,
+      "loss": 5.1663,
+      "step": 33410
+    },
+    {
+      "epoch": 0.046958235342676034,
+      "grad_norm": 1.0848567485809326,
+      "learning_rate": 0.00014082759589714765,
+      "loss": 5.1029,
+      "step": 33420
+    },
+    {
+      "epoch": 0.04697228628083961,
+      "grad_norm": 1.1160449981689453,
+      "learning_rate": 0.00014086974848953208,
+      "loss": 4.9525,
+      "step": 33430
+    },
+    {
+      "epoch": 0.046986337219003194,
+      "grad_norm": 1.1369086503982544,
+      "learning_rate": 0.00014091190108191654,
+      "loss": 5.0147,
+      "step": 33440
+    },
+    {
+      "epoch": 0.04700038815716677,
+      "grad_norm": 1.1825052499771118,
+      "learning_rate": 0.00014095405367430097,
+      "loss": 5.0531,
+      "step": 33450
+    },
+    {
+      "epoch": 0.047014439095330346,
+      "grad_norm": 1.1444425582885742,
+      "learning_rate": 0.00014099620626668538,
+      "loss": 5.0372,
+      "step": 33460
+    },
+    {
+      "epoch": 0.04702849003349392,
+      "grad_norm": 1.1329076290130615,
+      "learning_rate": 0.00014103835885906983,
+      "loss": 5.189,
+      "step": 33470
+    },
+    {
+      "epoch": 0.0470425409716575,
+      "grad_norm": 1.1967828273773193,
+      "learning_rate": 0.00014108051145145427,
+      "loss": 5.0807,
+      "step": 33480
+    },
+    {
+      "epoch": 0.04705659190982108,
+      "grad_norm": 1.1986347436904907,
+      "learning_rate": 0.00014112266404383867,
+      "loss": 5.0648,
+      "step": 33490
+    },
+    {
+      "epoch": 0.04707064284798466,
+      "grad_norm": 1.131181001663208,
+      "learning_rate": 0.00014116481663622313,
+      "loss": 5.0826,
+      "step": 33500
+    },
+    {
+      "epoch": 0.047084693786148234,
+      "grad_norm": 1.1698004007339478,
+      "learning_rate": 0.00014120696922860756,
+      "loss": 5.0187,
+      "step": 33510
+    },
+    {
+      "epoch": 0.04709874472431181,
+      "grad_norm": 1.186985731124878,
+      "learning_rate": 0.00014124912182099196,
+      "loss": 5.071,
+      "step": 33520
+    },
+    {
+      "epoch": 0.047112795662475386,
+      "grad_norm": 1.1888569593429565,
+      "learning_rate": 0.00014129127441337642,
+      "loss": 4.9117,
+      "step": 33530
+    },
+    {
+      "epoch": 0.04712684660063897,
+      "grad_norm": 1.1342166662216187,
+      "learning_rate": 0.00014133342700576085,
+      "loss": 5.0269,
+      "step": 33540
+    },
+    {
+      "epoch": 0.047140897538802545,
+      "grad_norm": 1.2070249319076538,
+      "learning_rate": 0.00014137557959814526,
+      "loss": 4.9468,
+      "step": 33550
+    },
+    {
+      "epoch": 0.04715494847696612,
+      "grad_norm": 1.1142017841339111,
+      "learning_rate": 0.00014141773219052971,
+      "loss": 5.2021,
+      "step": 33560
+    },
+    {
+      "epoch": 0.0471689994151297,
+      "grad_norm": 1.092371940612793,
+      "learning_rate": 0.00014145988478291415,
+      "loss": 5.0756,
+      "step": 33570
+    },
+    {
+      "epoch": 0.047183050353293274,
+      "grad_norm": 1.1797840595245361,
+      "learning_rate": 0.00014150203737529855,
+      "loss": 5.0859,
+      "step": 33580
+    },
+    {
+      "epoch": 0.04719710129145686,
+      "grad_norm": 1.177193522453308,
+      "learning_rate": 0.000141544189967683,
+      "loss": 4.9599,
+      "step": 33590
+    },
+    {
+      "epoch": 0.04721115222962043,
+      "grad_norm": 1.1384317874908447,
+      "learning_rate": 0.00014158634256006744,
+      "loss": 5.0421,
+      "step": 33600
+    },
+    {
+      "epoch": 0.04722520316778401,
+      "grad_norm": 1.1723968982696533,
+      "learning_rate": 0.00014162849515245187,
+      "loss": 5.0571,
+      "step": 33610
+    },
+    {
+      "epoch": 0.047239254105947585,
+      "grad_norm": 1.1619266271591187,
+      "learning_rate": 0.0001416706477448363,
+      "loss": 5.114,
+      "step": 33620
+    },
+    {
+      "epoch": 0.04725330504411116,
+      "grad_norm": 1.1768392324447632,
+      "learning_rate": 0.00014171280033722073,
+      "loss": 4.9927,
+      "step": 33630
+    },
+    {
+      "epoch": 0.047267355982274745,
+      "grad_norm": 1.1438103914260864,
+      "learning_rate": 0.00014175495292960516,
+      "loss": 5.0783,
+      "step": 33640
+    },
+    {
+      "epoch": 0.04728140692043832,
+      "grad_norm": 1.165272831916809,
+      "learning_rate": 0.0001417971055219896,
+      "loss": 4.9889,
+      "step": 33650
+    },
+    {
+      "epoch": 0.0472954578586019,
+      "grad_norm": 1.2394903898239136,
+      "learning_rate": 0.00014183925811437403,
+      "loss": 5.0751,
+      "step": 33660
+    },
+    {
+      "epoch": 0.04730950879676547,
+      "grad_norm": 1.2130417823791504,
+      "learning_rate": 0.00014188141070675846,
+      "loss": 5.1452,
+      "step": 33670
+    },
+    {
+      "epoch": 0.04732355973492905,
+      "grad_norm": 1.100905418395996,
+      "learning_rate": 0.0001419235632991429,
+      "loss": 5.1668,
+      "step": 33680
+    },
+    {
+      "epoch": 0.04733761067309263,
+      "grad_norm": 1.1132055521011353,
+      "learning_rate": 0.00014196571589152732,
+      "loss": 4.9762,
+      "step": 33690
+    },
+    {
+      "epoch": 0.04735166161125621,
+      "grad_norm": 1.1489577293395996,
+      "learning_rate": 0.00014200786848391175,
+      "loss": 5.0032,
+      "step": 33700
+    },
+    {
+      "epoch": 0.047365712549419785,
+      "grad_norm": 1.1104612350463867,
+      "learning_rate": 0.00014205002107629618,
+      "loss": 5.0617,
+      "step": 33710
+    },
+    {
+      "epoch": 0.04737976348758336,
+      "grad_norm": 1.148146629333496,
+      "learning_rate": 0.0001420921736686806,
+      "loss": 5.0499,
+      "step": 33720
+    },
+    {
+      "epoch": 0.04739381442574694,
+      "grad_norm": 1.1576919555664062,
+      "learning_rate": 0.00014213432626106504,
+      "loss": 5.0133,
+      "step": 33730
+    },
+    {
+      "epoch": 0.04740786536391051,
+      "grad_norm": 1.1263247728347778,
+      "learning_rate": 0.00014217647885344948,
+      "loss": 4.8634,
+      "step": 33740
+    },
+    {
+      "epoch": 0.047421916302074096,
+      "grad_norm": 1.168540358543396,
+      "learning_rate": 0.0001422186314458339,
+      "loss": 4.9549,
+      "step": 33750
+    },
+    {
+      "epoch": 0.04743596724023767,
+      "grad_norm": 1.1476012468338013,
+      "learning_rate": 0.00014226078403821834,
+      "loss": 5.0192,
+      "step": 33760
+    },
+    {
+      "epoch": 0.04745001817840125,
+      "grad_norm": 1.1322742700576782,
+      "learning_rate": 0.00014230293663060277,
+      "loss": 5.1356,
+      "step": 33770
+    },
+    {
+      "epoch": 0.047464069116564825,
+      "grad_norm": 1.147298812866211,
+      "learning_rate": 0.0001423450892229872,
+      "loss": 5.0278,
+      "step": 33780
+    },
+    {
+      "epoch": 0.0474781200547284,
+      "grad_norm": 1.1303203105926514,
+      "learning_rate": 0.00014238724181537163,
+      "loss": 5.0996,
+      "step": 33790
+    },
+    {
+      "epoch": 0.047492170992891984,
+      "grad_norm": 1.0890021324157715,
+      "learning_rate": 0.00014242939440775606,
+      "loss": 5.0753,
+      "step": 33800
+    },
+    {
+      "epoch": 0.04750622193105556,
+      "grad_norm": 1.163889765739441,
+      "learning_rate": 0.0001424715470001405,
+      "loss": 5.0723,
+      "step": 33810
+    },
+    {
+      "epoch": 0.047520272869219136,
+      "grad_norm": 1.1471232175827026,
+      "learning_rate": 0.00014251369959252492,
+      "loss": 5.0793,
+      "step": 33820
+    },
+    {
+      "epoch": 0.04753432380738271,
+      "grad_norm": 1.135847568511963,
+      "learning_rate": 0.00014255585218490936,
+      "loss": 5.1346,
+      "step": 33830
+    },
+    {
+      "epoch": 0.04754837474554629,
+      "grad_norm": 1.1431607007980347,
+      "learning_rate": 0.0001425980047772938,
+      "loss": 5.1053,
+      "step": 33840
+    },
+    {
+      "epoch": 0.04756242568370987,
+      "grad_norm": 1.1063638925552368,
+      "learning_rate": 0.00014264015736967822,
+      "loss": 4.9736,
+      "step": 33850
+    },
+    {
+      "epoch": 0.04757647662187345,
+      "grad_norm": 1.1206468343734741,
+      "learning_rate": 0.00014268230996206265,
+      "loss": 4.9508,
+      "step": 33860
+    },
+    {
+      "epoch": 0.047590527560037024,
+      "grad_norm": 1.234711766242981,
+      "learning_rate": 0.00014272446255444708,
+      "loss": 4.9879,
+      "step": 33870
+    },
+    {
+      "epoch": 0.0476045784982006,
+      "grad_norm": 1.2123228311538696,
+      "learning_rate": 0.0001427666151468315,
+      "loss": 4.8591,
+      "step": 33880
+    },
+    {
+      "epoch": 0.04761862943636418,
+      "grad_norm": 1.330238938331604,
+      "learning_rate": 0.00014280876773921594,
+      "loss": 4.9129,
+      "step": 33890
+    },
+    {
+      "epoch": 0.04763268037452776,
+      "grad_norm": 1.1890876293182373,
+      "learning_rate": 0.0001428509203316004,
+      "loss": 5.018,
+      "step": 33900
+    },
+    {
+      "epoch": 0.047646731312691336,
+      "grad_norm": 1.2122526168823242,
+      "learning_rate": 0.0001428930729239848,
+      "loss": 5.1118,
+      "step": 33910
+    },
+    {
+      "epoch": 0.04766078225085491,
+      "grad_norm": 1.1024436950683594,
+      "learning_rate": 0.00014293522551636924,
+      "loss": 4.8631,
+      "step": 33920
+    },
+    {
+      "epoch": 0.04767483318901849,
+      "grad_norm": 1.12148118019104,
+      "learning_rate": 0.0001429773781087537,
+      "loss": 5.0444,
+      "step": 33930
+    },
+    {
+      "epoch": 0.047688884127182064,
+      "grad_norm": 1.1458978652954102,
+      "learning_rate": 0.0001430195307011381,
+      "loss": 5.1447,
+      "step": 33940
+    },
+    {
+      "epoch": 0.04770293506534565,
+      "grad_norm": 1.1465718746185303,
+      "learning_rate": 0.00014306168329352253,
+      "loss": 4.968,
+      "step": 33950
+    },
+    {
+      "epoch": 0.047716986003509224,
+      "grad_norm": 1.162516713142395,
+      "learning_rate": 0.000143103835885907,
+      "loss": 5.0728,
+      "step": 33960
+    },
+    {
+      "epoch": 0.0477310369416728,
+      "grad_norm": 1.1164641380310059,
+      "learning_rate": 0.0001431459884782914,
+      "loss": 4.9858,
+      "step": 33970
+    },
+    {
+      "epoch": 0.047745087879836376,
+      "grad_norm": 1.1258015632629395,
+      "learning_rate": 0.00014318814107067582,
+      "loss": 5.0594,
+      "step": 33980
+    },
+    {
+      "epoch": 0.04775913881799995,
+      "grad_norm": 1.1459895372390747,
+      "learning_rate": 0.00014323029366306028,
+      "loss": 4.987,
+      "step": 33990
+    },
+    {
+      "epoch": 0.047773189756163535,
+      "grad_norm": 1.16275155544281,
+      "learning_rate": 0.00014327244625544469,
+      "loss": 5.0785,
+      "step": 34000
+    },
+    {
+      "epoch": 0.04778724069432711,
+      "grad_norm": 1.1919983625411987,
+      "learning_rate": 0.00014331459884782912,
+      "loss": 5.0735,
+      "step": 34010
+    },
+    {
+      "epoch": 0.04780129163249069,
+      "grad_norm": 1.0925780534744263,
+      "learning_rate": 0.00014335675144021358,
+      "loss": 5.1679,
+      "step": 34020
+    },
+    {
+      "epoch": 0.047815342570654264,
+      "grad_norm": 1.1478568315505981,
+      "learning_rate": 0.000143398904032598,
+      "loss": 5.1252,
+      "step": 34030
+    },
+    {
+      "epoch": 0.04782939350881784,
+      "grad_norm": 1.1364628076553345,
+      "learning_rate": 0.0001434410566249824,
+      "loss": 4.9149,
+      "step": 34040
+    },
+    {
+      "epoch": 0.047843444446981416,
+      "grad_norm": 1.130101203918457,
+      "learning_rate": 0.00014348320921736687,
+      "loss": 4.909,
+      "step": 34050
+    },
+    {
+      "epoch": 0.047857495385145,
+      "grad_norm": 1.151053547859192,
+      "learning_rate": 0.0001435253618097513,
+      "loss": 5.0317,
+      "step": 34060
+    },
+    {
+      "epoch": 0.047871546323308575,
+      "grad_norm": 1.158276081085205,
+      "learning_rate": 0.0001435675144021357,
+      "loss": 5.0723,
+      "step": 34070
+    },
+    {
+      "epoch": 0.04788559726147215,
+      "grad_norm": 1.2227519750595093,
+      "learning_rate": 0.00014360966699452016,
+      "loss": 5.0276,
+      "step": 34080
+    },
+    {
+      "epoch": 0.04789964819963573,
+      "grad_norm": 1.1406713724136353,
+      "learning_rate": 0.0001436518195869046,
+      "loss": 5.0192,
+      "step": 34090
+    },
+    {
+      "epoch": 0.047913699137799304,
+      "grad_norm": 1.1668028831481934,
+      "learning_rate": 0.000143693972179289,
+      "loss": 4.9819,
+      "step": 34100
+    },
+    {
+      "epoch": 0.04792775007596289,
+      "grad_norm": 1.1254831552505493,
+      "learning_rate": 0.00014373612477167346,
+      "loss": 5.0085,
+      "step": 34110
+    },
+    {
+      "epoch": 0.04794180101412646,
+      "grad_norm": 1.1782492399215698,
+      "learning_rate": 0.0001437782773640579,
+      "loss": 5.087,
+      "step": 34120
+    },
+    {
+      "epoch": 0.04795585195229004,
+      "grad_norm": 1.125497817993164,
+      "learning_rate": 0.0001438204299564423,
+      "loss": 5.0525,
+      "step": 34130
+    },
+    {
+      "epoch": 0.047969902890453615,
+      "grad_norm": 1.1250324249267578,
+      "learning_rate": 0.00014386258254882675,
+      "loss": 5.0964,
+      "step": 34140
+    },
+    {
+      "epoch": 0.04798395382861719,
+      "grad_norm": 1.1316850185394287,
+      "learning_rate": 0.00014390473514121118,
+      "loss": 5.0596,
+      "step": 34150
+    },
+    {
+      "epoch": 0.047998004766780775,
+      "grad_norm": 1.158501386642456,
+      "learning_rate": 0.0001439468877335956,
+      "loss": 5.0526,
+      "step": 34160
+    },
+    {
+      "epoch": 0.04801205570494435,
+      "grad_norm": 1.1655758619308472,
+      "learning_rate": 0.00014398904032598004,
+      "loss": 5.0539,
+      "step": 34170
+    },
+    {
+      "epoch": 0.04802610664310793,
+      "grad_norm": 1.1360056400299072,
+      "learning_rate": 0.00014403119291836447,
+      "loss": 5.0661,
+      "step": 34180
+    },
+    {
+      "epoch": 0.0480401575812715,
+      "grad_norm": 1.1133544445037842,
+      "learning_rate": 0.0001440733455107489,
+      "loss": 5.0339,
+      "step": 34190
+    },
+    {
+      "epoch": 0.04805420851943508,
+      "grad_norm": 1.0935070514678955,
+      "learning_rate": 0.00014411549810313334,
+      "loss": 4.9833,
+      "step": 34200
+    },
+    {
+      "epoch": 0.04806825945759866,
+      "grad_norm": 1.0671685934066772,
+      "learning_rate": 0.00014415765069551777,
+      "loss": 4.9763,
+      "step": 34210
+    },
+    {
+      "epoch": 0.04808231039576224,
+      "grad_norm": 1.1430456638336182,
+      "learning_rate": 0.0001441998032879022,
+      "loss": 4.9482,
+      "step": 34220
+    },
+    {
+      "epoch": 0.048096361333925815,
+      "grad_norm": 1.1322462558746338,
+      "learning_rate": 0.00014424195588028663,
+      "loss": 4.9501,
+      "step": 34230
+    },
+    {
+      "epoch": 0.04811041227208939,
+      "grad_norm": 1.121475100517273,
+      "learning_rate": 0.00014428410847267106,
+      "loss": 5.0683,
+      "step": 34240
+    },
+    {
+      "epoch": 0.04812446321025297,
+      "grad_norm": 1.1033834218978882,
+      "learning_rate": 0.0001443262610650555,
+      "loss": 4.9089,
+      "step": 34250
+    },
+    {
+      "epoch": 0.04813851414841655,
+      "grad_norm": 1.1717206239700317,
+      "learning_rate": 0.00014436841365743992,
+      "loss": 4.9873,
+      "step": 34260
+    },
+    {
+      "epoch": 0.048152565086580126,
+      "grad_norm": 1.1643035411834717,
+      "learning_rate": 0.00014441056624982435,
+      "loss": 5.0396,
+      "step": 34270
+    },
+    {
+      "epoch": 0.0481666160247437,
+      "grad_norm": 1.109316110610962,
+      "learning_rate": 0.00014445271884220879,
+      "loss": 5.0065,
+      "step": 34280
+    },
+    {
+      "epoch": 0.04818066696290728,
+      "grad_norm": 1.0868771076202393,
+      "learning_rate": 0.00014449487143459322,
+      "loss": 5.0783,
+      "step": 34290
+    },
+    {
+      "epoch": 0.048194717901070855,
+      "grad_norm": 1.1115925312042236,
+      "learning_rate": 0.00014453702402697765,
+      "loss": 4.9573,
+      "step": 34300
+    },
+    {
+      "epoch": 0.04820876883923444,
+      "grad_norm": 1.14295494556427,
+      "learning_rate": 0.00014457496136012363,
+      "loss": 5.016,
+      "step": 34310
+    },
+    {
+      "epoch": 0.048222819777398014,
+      "grad_norm": 1.148398518562317,
+      "learning_rate": 0.00014461711395250806,
+      "loss": 5.147,
+      "step": 34320
+    },
+    {
+      "epoch": 0.04823687071556159,
+      "grad_norm": 1.1277004480361938,
+      "learning_rate": 0.0001446592665448925,
+      "loss": 5.1473,
+      "step": 34330
+    },
+    {
+      "epoch": 0.048250921653725166,
+      "grad_norm": 1.186437964439392,
+      "learning_rate": 0.00014470141913727692,
+      "loss": 4.9353,
+      "step": 34340
+    },
+    {
+      "epoch": 0.04826497259188874,
+      "grad_norm": 1.131174921989441,
+      "learning_rate": 0.00014474357172966136,
+      "loss": 5.0711,
+      "step": 34350
+    },
+    {
+      "epoch": 0.04827902353005232,
+      "grad_norm": 1.1172866821289062,
+      "learning_rate": 0.0001447857243220458,
+      "loss": 5.1223,
+      "step": 34360
+    },
+    {
+      "epoch": 0.0482930744682159,
+      "grad_norm": 1.179067850112915,
+      "learning_rate": 0.00014482787691443022,
+      "loss": 4.9628,
+      "step": 34370
+    },
+    {
+      "epoch": 0.04830712540637948,
+      "grad_norm": 1.1836308240890503,
+      "learning_rate": 0.00014487002950681465,
+      "loss": 5.0934,
+      "step": 34380
+    },
+    {
+      "epoch": 0.048321176344543054,
+      "grad_norm": 1.1305451393127441,
+      "learning_rate": 0.00014491218209919908,
+      "loss": 5.0188,
+      "step": 34390
+    },
+    {
+      "epoch": 0.04833522728270663,
+      "grad_norm": 1.113901138305664,
+      "learning_rate": 0.0001449543346915835,
+      "loss": 5.0124,
+      "step": 34400
+    },
+    {
+      "epoch": 0.04834927822087021,
+      "grad_norm": 1.1200542449951172,
+      "learning_rate": 0.00014499648728396794,
+      "loss": 5.1159,
+      "step": 34410
+    },
+    {
+      "epoch": 0.04836332915903379,
+      "grad_norm": 1.1390150785446167,
+      "learning_rate": 0.00014503863987635237,
+      "loss": 5.0467,
+      "step": 34420
+    },
+    {
+      "epoch": 0.048377380097197366,
+      "grad_norm": 1.1669032573699951,
+      "learning_rate": 0.0001450807924687368,
+      "loss": 5.0509,
+      "step": 34430
+    },
+    {
+      "epoch": 0.04839143103536094,
+      "grad_norm": 1.1723250150680542,
+      "learning_rate": 0.00014512294506112124,
+      "loss": 4.8566,
+      "step": 34440
+    },
+    {
+      "epoch": 0.04840548197352452,
+      "grad_norm": 1.2041908502578735,
+      "learning_rate": 0.0001451650976535057,
+      "loss": 5.0341,
+      "step": 34450
+    },
+    {
+      "epoch": 0.048419532911688094,
+      "grad_norm": 1.1292054653167725,
+      "learning_rate": 0.0001452072502458901,
+      "loss": 5.0321,
+      "step": 34460
+    },
+    {
+      "epoch": 0.04843358384985168,
+      "grad_norm": 1.12458074092865,
+      "learning_rate": 0.00014524940283827453,
+      "loss": 5.0766,
+      "step": 34470
+    },
+    {
+      "epoch": 0.048447634788015254,
+      "grad_norm": 1.1062582731246948,
+      "learning_rate": 0.000145291555430659,
+      "loss": 5.0649,
+      "step": 34480
+    },
+    {
+      "epoch": 0.04846168572617883,
+      "grad_norm": 1.124700665473938,
+      "learning_rate": 0.0001453337080230434,
+      "loss": 4.9559,
+      "step": 34490
+    },
+    {
+      "epoch": 0.048475736664342406,
+      "grad_norm": 1.1494004726409912,
+      "learning_rate": 0.00014537586061542782,
+      "loss": 5.0064,
+      "step": 34500
+    },
+    {
+      "epoch": 0.04848978760250598,
+      "grad_norm": 1.1343144178390503,
+      "learning_rate": 0.00014541801320781228,
+      "loss": 5.0244,
+      "step": 34510
+    },
+    {
+      "epoch": 0.048503838540669565,
+      "grad_norm": 1.0855003595352173,
+      "learning_rate": 0.00014546016580019669,
+      "loss": 5.0812,
+      "step": 34520
+    },
+    {
+      "epoch": 0.04851788947883314,
+      "grad_norm": 1.1069141626358032,
+      "learning_rate": 0.00014550231839258112,
+      "loss": 5.0764,
+      "step": 34530
+    },
+    {
+      "epoch": 0.04853194041699672,
+      "grad_norm": 1.470924735069275,
+      "learning_rate": 0.00014554447098496557,
+      "loss": 5.0076,
+      "step": 34540
+    },
+    {
+      "epoch": 0.048545991355160294,
+      "grad_norm": 1.1423304080963135,
+      "learning_rate": 0.00014558662357734998,
+      "loss": 5.008,
+      "step": 34550
+    },
+    {
+      "epoch": 0.04856004229332387,
+      "grad_norm": 1.164551854133606,
+      "learning_rate": 0.00014562877616973444,
+      "loss": 4.9289,
+      "step": 34560
+    },
+    {
+      "epoch": 0.04857409323148745,
+      "grad_norm": 1.0787816047668457,
+      "learning_rate": 0.00014567092876211887,
+      "loss": 5.0759,
+      "step": 34570
+    },
+    {
+      "epoch": 0.04858814416965103,
+      "grad_norm": 1.1306999921798706,
+      "learning_rate": 0.0001457130813545033,
+      "loss": 5.0986,
+      "step": 34580
+    },
+    {
+      "epoch": 0.048602195107814605,
+      "grad_norm": 1.1416428089141846,
+      "learning_rate": 0.00014575523394688773,
+      "loss": 5.0618,
+      "step": 34590
+    },
+    {
+      "epoch": 0.04861624604597818,
+      "grad_norm": 1.1224125623703003,
+      "learning_rate": 0.00014579738653927216,
+      "loss": 5.0685,
+      "step": 34600
+    },
+    {
+      "epoch": 0.04863029698414176,
+      "grad_norm": 1.1280869245529175,
+      "learning_rate": 0.0001458395391316566,
+      "loss": 4.9973,
+      "step": 34610
+    },
+    {
+      "epoch": 0.04864434792230534,
+      "grad_norm": 1.1328657865524292,
+      "learning_rate": 0.00014588169172404102,
+      "loss": 5.0089,
+      "step": 34620
+    },
+    {
+      "epoch": 0.04865839886046892,
+      "grad_norm": 1.183826208114624,
+      "learning_rate": 0.00014592384431642546,
+      "loss": 4.9232,
+      "step": 34630
+    },
+    {
+      "epoch": 0.04867244979863249,
+      "grad_norm": 1.1926147937774658,
+      "learning_rate": 0.00014596599690880989,
+      "loss": 4.9648,
+      "step": 34640
+    },
+    {
+      "epoch": 0.04868650073679607,
+      "grad_norm": 1.1898232698440552,
+      "learning_rate": 0.00014600814950119432,
+      "loss": 4.9844,
+      "step": 34650
+    },
+    {
+      "epoch": 0.048700551674959645,
+      "grad_norm": 1.1884058713912964,
+      "learning_rate": 0.00014605030209357875,
+      "loss": 5.0502,
+      "step": 34660
+    },
+    {
+      "epoch": 0.04871460261312322,
+      "grad_norm": 1.1177476644515991,
+      "learning_rate": 0.00014609245468596318,
+      "loss": 5.0569,
+      "step": 34670
+    },
+    {
+      "epoch": 0.048728653551286805,
+      "grad_norm": 1.3436133861541748,
+      "learning_rate": 0.0001461346072783476,
+      "loss": 5.0094,
+      "step": 34680
+    },
+    {
+      "epoch": 0.04874270448945038,
+      "grad_norm": 1.1321064233779907,
+      "learning_rate": 0.00014617675987073204,
+      "loss": 5.0142,
+      "step": 34690
+    },
+    {
+      "epoch": 0.04875675542761396,
+      "grad_norm": 1.104989767074585,
+      "learning_rate": 0.00014621891246311647,
+      "loss": 5.0279,
+      "step": 34700
+    },
+    {
+      "epoch": 0.04877080636577753,
+      "grad_norm": 1.1625908613204956,
+      "learning_rate": 0.0001462610650555009,
+      "loss": 4.9406,
+      "step": 34710
+    },
+    {
+      "epoch": 0.04878485730394111,
+      "grad_norm": 1.1442335844039917,
+      "learning_rate": 0.00014630321764788534,
+      "loss": 4.9705,
+      "step": 34720
+    },
+    {
+      "epoch": 0.04879890824210469,
+      "grad_norm": 1.1111855506896973,
+      "learning_rate": 0.00014634537024026977,
+      "loss": 5.003,
+      "step": 34730
+    },
+    {
+      "epoch": 0.04881295918026827,
+      "grad_norm": 1.1129965782165527,
+      "learning_rate": 0.0001463875228326542,
+      "loss": 4.8758,
+      "step": 34740
+    },
+    {
+      "epoch": 0.048827010118431845,
+      "grad_norm": 1.1186314821243286,
+      "learning_rate": 0.00014642967542503863,
+      "loss": 5.0139,
+      "step": 34750
+    },
+    {
+      "epoch": 0.04884106105659542,
+      "grad_norm": 1.150681972503662,
+      "learning_rate": 0.00014647182801742306,
+      "loss": 4.9994,
+      "step": 34760
+    },
+    {
+      "epoch": 0.048855111994759,
+      "grad_norm": 1.1509026288986206,
+      "learning_rate": 0.0001465139806098075,
+      "loss": 5.0786,
+      "step": 34770
+    },
+    {
+      "epoch": 0.04886916293292258,
+      "grad_norm": 1.1005713939666748,
+      "learning_rate": 0.00014655613320219192,
+      "loss": 4.9926,
+      "step": 34780
+    },
+    {
+      "epoch": 0.048883213871086156,
+      "grad_norm": 1.176209807395935,
+      "learning_rate": 0.00014659828579457635,
+      "loss": 5.0281,
+      "step": 34790
+    },
+    {
+      "epoch": 0.04889726480924973,
+      "grad_norm": 1.1285464763641357,
+      "learning_rate": 0.00014664043838696078,
+      "loss": 4.8861,
+      "step": 34800
+    },
+    {
+      "epoch": 0.04891131574741331,
+      "grad_norm": 1.1435015201568604,
+      "learning_rate": 0.00014668259097934522,
+      "loss": 4.9688,
+      "step": 34810
+    },
+    {
+      "epoch": 0.048925366685576885,
+      "grad_norm": 1.1347907781600952,
+      "learning_rate": 0.00014672474357172965,
+      "loss": 5.0083,
+      "step": 34820
+    },
+    {
+      "epoch": 0.04893941762374047,
+      "grad_norm": 1.1737812757492065,
+      "learning_rate": 0.00014676689616411408,
+      "loss": 5.0881,
+      "step": 34830
+    },
+    {
+      "epoch": 0.048953468561904044,
+      "grad_norm": 1.0986801385879517,
+      "learning_rate": 0.0001468090487564985,
+      "loss": 5.0823,
+      "step": 34840
+    },
+    {
+      "epoch": 0.04896751950006762,
+      "grad_norm": 1.218393325805664,
+      "learning_rate": 0.00014685120134888294,
+      "loss": 5.0719,
+      "step": 34850
+    },
+    {
+      "epoch": 0.048981570438231196,
+      "grad_norm": 1.0812214612960815,
+      "learning_rate": 0.00014689335394126737,
+      "loss": 5.0184,
+      "step": 34860
+    },
+    {
+      "epoch": 0.04899562137639477,
+      "grad_norm": 1.2031632661819458,
+      "learning_rate": 0.0001469355065336518,
+      "loss": 5.0449,
+      "step": 34870
+    },
+    {
+      "epoch": 0.049009672314558356,
+      "grad_norm": 1.109456181526184,
+      "learning_rate": 0.00014697765912603623,
+      "loss": 5.1284,
+      "step": 34880
+    },
+    {
+      "epoch": 0.04902372325272193,
+      "grad_norm": 1.1223009824752808,
+      "learning_rate": 0.00014701981171842067,
+      "loss": 4.9705,
+      "step": 34890
+    },
+    {
+      "epoch": 0.04903777419088551,
+      "grad_norm": 1.125307559967041,
+      "learning_rate": 0.0001470619643108051,
+      "loss": 4.9975,
+      "step": 34900
+    },
+    {
+      "epoch": 0.049051825129049084,
+      "grad_norm": 1.1051353216171265,
+      "learning_rate": 0.00014710411690318953,
+      "loss": 4.9682,
+      "step": 34910
+    },
+    {
+      "epoch": 0.04906587606721266,
+      "grad_norm": 1.0811420679092407,
+      "learning_rate": 0.00014714626949557396,
+      "loss": 5.0424,
+      "step": 34920
+    },
+    {
+      "epoch": 0.049079927005376243,
+      "grad_norm": 1.1198779344558716,
+      "learning_rate": 0.0001471884220879584,
+      "loss": 5.0246,
+      "step": 34930
+    },
+    {
+      "epoch": 0.04909397794353982,
+      "grad_norm": 1.1411025524139404,
+      "learning_rate": 0.00014723057468034282,
+      "loss": 5.0529,
+      "step": 34940
+    },
+    {
+      "epoch": 0.049108028881703396,
+      "grad_norm": 1.1323539018630981,
+      "learning_rate": 0.00014727272727272725,
+      "loss": 4.9568,
+      "step": 34950
+    },
+    {
+      "epoch": 0.04912207981986697,
+      "grad_norm": 1.1151742935180664,
+      "learning_rate": 0.00014731487986511168,
+      "loss": 4.9563,
+      "step": 34960
+    },
+    {
+      "epoch": 0.04913613075803055,
+      "grad_norm": 1.1359392404556274,
+      "learning_rate": 0.00014735703245749611,
+      "loss": 5.112,
+      "step": 34970
+    },
+    {
+      "epoch": 0.049150181696194124,
+      "grad_norm": 1.0861750841140747,
+      "learning_rate": 0.00014739918504988055,
+      "loss": 4.9891,
+      "step": 34980
+    },
+    {
+      "epoch": 0.04916423263435771,
+      "grad_norm": 1.2048242092132568,
+      "learning_rate": 0.00014744133764226498,
+      "loss": 5.0255,
+      "step": 34990
+    },
+    {
+      "epoch": 0.049178283572521284,
+      "grad_norm": 1.1281739473342896,
+      "learning_rate": 0.00014748349023464944,
+      "loss": 5.1152,
+      "step": 35000
+    },
+    {
+      "epoch": 0.04919233451068486,
+      "grad_norm": 1.1170549392700195,
+      "learning_rate": 0.00014752564282703384,
+      "loss": 4.8994,
+      "step": 35010
+    },
+    {
+      "epoch": 0.049206385448848436,
+      "grad_norm": 1.1060420274734497,
+      "learning_rate": 0.00014756779541941827,
+      "loss": 5.1206,
+      "step": 35020
+    },
+    {
+      "epoch": 0.04922043638701201,
+      "grad_norm": 1.105675458908081,
+      "learning_rate": 0.00014760994801180273,
+      "loss": 4.9421,
+      "step": 35030
+    },
+    {
+      "epoch": 0.049234487325175595,
+      "grad_norm": 1.1171681880950928,
+      "learning_rate": 0.00014765210060418713,
+      "loss": 4.9546,
+      "step": 35040
+    },
+    {
+      "epoch": 0.04924853826333917,
+      "grad_norm": 1.1521064043045044,
+      "learning_rate": 0.00014769425319657156,
+      "loss": 4.8988,
+      "step": 35050
+    },
+    {
+      "epoch": 0.04926258920150275,
+      "grad_norm": 1.110176920890808,
+      "learning_rate": 0.00014773640578895602,
+      "loss": 4.9492,
+      "step": 35060
+    },
+    {
+      "epoch": 0.049276640139666324,
+      "grad_norm": 1.0732078552246094,
+      "learning_rate": 0.00014777855838134043,
+      "loss": 5.0328,
+      "step": 35070
+    },
+    {
+      "epoch": 0.0492906910778299,
+      "grad_norm": 1.1150749921798706,
+      "learning_rate": 0.00014782071097372486,
+      "loss": 5.1412,
+      "step": 35080
+    },
+    {
+      "epoch": 0.04930474201599348,
+      "grad_norm": 1.1571887731552124,
+      "learning_rate": 0.00014786286356610932,
+      "loss": 4.9802,
+      "step": 35090
+    },
+    {
+      "epoch": 0.04931879295415706,
+      "grad_norm": 1.0949782133102417,
+      "learning_rate": 0.00014790501615849372,
+      "loss": 4.9798,
+      "step": 35100
+    },
+    {
+      "epoch": 0.049332843892320635,
+      "grad_norm": 1.1114603281021118,
+      "learning_rate": 0.00014794716875087815,
+      "loss": 4.9468,
+      "step": 35110
+    },
+    {
+      "epoch": 0.04934689483048421,
+      "grad_norm": 1.0757004022598267,
+      "learning_rate": 0.0001479893213432626,
+      "loss": 4.8796,
+      "step": 35120
+    },
+    {
+      "epoch": 0.04936094576864779,
+      "grad_norm": 1.1924103498458862,
+      "learning_rate": 0.000148031473935647,
+      "loss": 4.932,
+      "step": 35130
+    },
+    {
+      "epoch": 0.04937499670681137,
+      "grad_norm": 1.1277077198028564,
+      "learning_rate": 0.00014807362652803147,
+      "loss": 5.0101,
+      "step": 35140
+    },
+    {
+      "epoch": 0.04938904764497495,
+      "grad_norm": 1.1275465488433838,
+      "learning_rate": 0.0001481157791204159,
+      "loss": 4.9974,
+      "step": 35150
+    },
+    {
+      "epoch": 0.04940309858313852,
+      "grad_norm": 1.1476144790649414,
+      "learning_rate": 0.00014815793171280033,
+      "loss": 5.0018,
+      "step": 35160
+    },
+    {
+      "epoch": 0.0494171495213021,
+      "grad_norm": 1.1097756624221802,
+      "learning_rate": 0.00014820008430518477,
+      "loss": 5.0057,
+      "step": 35170
+    },
+    {
+      "epoch": 0.049431200459465675,
+      "grad_norm": 1.1443592309951782,
+      "learning_rate": 0.0001482422368975692,
+      "loss": 4.9678,
+      "step": 35180
+    },
+    {
+      "epoch": 0.04944525139762926,
+      "grad_norm": 1.1339356899261475,
+      "learning_rate": 0.00014828438948995363,
+      "loss": 4.9923,
+      "step": 35190
+    },
+    {
+      "epoch": 0.049459302335792835,
+      "grad_norm": 1.0916308164596558,
+      "learning_rate": 0.00014832654208233806,
+      "loss": 5.0241,
+      "step": 35200
+    },
+    {
+      "epoch": 0.04947335327395641,
+      "grad_norm": 1.1227270364761353,
+      "learning_rate": 0.0001483686946747225,
+      "loss": 4.9788,
+      "step": 35210
+    },
+    {
+      "epoch": 0.04948740421211999,
+      "grad_norm": 1.126926064491272,
+      "learning_rate": 0.00014841084726710692,
+      "loss": 4.8996,
+      "step": 35220
+    },
+    {
+      "epoch": 0.04950145515028356,
+      "grad_norm": 1.1398160457611084,
+      "learning_rate": 0.00014845299985949135,
+      "loss": 4.9572,
+      "step": 35230
+    },
+    {
+      "epoch": 0.049515506088447146,
+      "grad_norm": 1.1179039478302002,
+      "learning_rate": 0.00014849515245187578,
+      "loss": 4.9517,
+      "step": 35240
+    },
+    {
+      "epoch": 0.04952955702661072,
+      "grad_norm": 1.1064491271972656,
+      "learning_rate": 0.00014853730504426021,
+      "loss": 4.8927,
+      "step": 35250
+    },
+    {
+      "epoch": 0.0495436079647743,
+      "grad_norm": 1.128804087638855,
+      "learning_rate": 0.00014857945763664465,
+      "loss": 5.0498,
+      "step": 35260
+    },
+    {
+      "epoch": 0.049557658902937875,
+      "grad_norm": 1.1051520109176636,
+      "learning_rate": 0.00014862161022902908,
+      "loss": 5.0604,
+      "step": 35270
+    },
+    {
+      "epoch": 0.04957170984110145,
+      "grad_norm": 1.1551424264907837,
+      "learning_rate": 0.0001486637628214135,
+      "loss": 4.917,
+      "step": 35280
+    },
+    {
+      "epoch": 0.049585760779265034,
+      "grad_norm": 1.081084132194519,
+      "learning_rate": 0.00014870591541379794,
+      "loss": 5.2189,
+      "step": 35290
+    },
+    {
+      "epoch": 0.04959981171742861,
+      "grad_norm": 1.1549731492996216,
+      "learning_rate": 0.00014874806800618237,
+      "loss": 4.9709,
+      "step": 35300
+    },
+    {
+      "epoch": 0.049613862655592186,
+      "grad_norm": 1.1668425798416138,
+      "learning_rate": 0.0001487902205985668,
+      "loss": 5.0498,
+      "step": 35310
+    },
+    {
+      "epoch": 0.04962791359375576,
+      "grad_norm": 1.143610954284668,
+      "learning_rate": 0.00014883237319095123,
+      "loss": 5.0319,
+      "step": 35320
+    },
+    {
+      "epoch": 0.04964196453191934,
+      "grad_norm": 1.1000237464904785,
+      "learning_rate": 0.00014887452578333566,
+      "loss": 5.0029,
+      "step": 35330
+    },
+    {
+      "epoch": 0.049656015470082915,
+      "grad_norm": 1.1435863971710205,
+      "learning_rate": 0.0001489166783757201,
+      "loss": 4.8895,
+      "step": 35340
+    },
+    {
+      "epoch": 0.0496700664082465,
+      "grad_norm": 1.1478296518325806,
+      "learning_rate": 0.00014895883096810453,
+      "loss": 5.0575,
+      "step": 35350
+    },
+    {
+      "epoch": 0.049684117346410074,
+      "grad_norm": 1.0952467918395996,
+      "learning_rate": 0.00014900098356048896,
+      "loss": 5.0144,
+      "step": 35360
+    },
+    {
+      "epoch": 0.04969816828457365,
+      "grad_norm": 1.0802350044250488,
+      "learning_rate": 0.0001490431361528734,
+      "loss": 5.0557,
+      "step": 35370
+    },
+    {
+      "epoch": 0.049712219222737226,
+      "grad_norm": 1.1121940612792969,
+      "learning_rate": 0.00014908528874525782,
+      "loss": 5.0688,
+      "step": 35380
+    },
+    {
+      "epoch": 0.0497262701609008,
+      "grad_norm": 1.1207494735717773,
+      "learning_rate": 0.00014912744133764225,
+      "loss": 4.9295,
+      "step": 35390
+    },
+    {
+      "epoch": 0.049740321099064386,
+      "grad_norm": 1.1117680072784424,
+      "learning_rate": 0.00014916959393002668,
+      "loss": 5.0406,
+      "step": 35400
+    },
+    {
+      "epoch": 0.04975437203722796,
+      "grad_norm": 1.2448337078094482,
+      "learning_rate": 0.0001492117465224111,
+      "loss": 4.9559,
+      "step": 35410
+    },
+    {
+      "epoch": 0.04976842297539154,
+      "grad_norm": 1.0946520566940308,
+      "learning_rate": 0.00014925389911479554,
+      "loss": 4.9064,
+      "step": 35420
+    },
+    {
+      "epoch": 0.049782473913555114,
+      "grad_norm": 1.0836005210876465,
+      "learning_rate": 0.00014929605170717998,
+      "loss": 5.0171,
+      "step": 35430
+    },
+    {
+      "epoch": 0.04979652485171869,
+      "grad_norm": 1.1312141418457031,
+      "learning_rate": 0.0001493382042995644,
+      "loss": 5.103,
+      "step": 35440
+    },
+    {
+      "epoch": 0.04981057578988227,
+      "grad_norm": 1.2918745279312134,
+      "learning_rate": 0.00014938035689194884,
+      "loss": 5.056,
+      "step": 35450
+    },
+    {
+      "epoch": 0.04982462672804585,
+      "grad_norm": 1.1127320528030396,
+      "learning_rate": 0.00014942250948433327,
+      "loss": 4.988,
+      "step": 35460
+    },
+    {
+      "epoch": 0.049838677666209426,
+      "grad_norm": 1.1448490619659424,
+      "learning_rate": 0.0001494646620767177,
+      "loss": 5.141,
+      "step": 35470
+    },
+    {
+      "epoch": 0.049852728604373,
+      "grad_norm": 1.1456868648529053,
+      "learning_rate": 0.00014950681466910213,
+      "loss": 5.0021,
+      "step": 35480
+    },
+    {
+      "epoch": 0.04986677954253658,
+      "grad_norm": 1.1046836376190186,
+      "learning_rate": 0.00014954896726148656,
+      "loss": 5.0716,
+      "step": 35490
+    },
+    {
+      "epoch": 0.04988083048070016,
+      "grad_norm": 1.1139931678771973,
+      "learning_rate": 0.000149591119853871,
+      "loss": 5.1356,
+      "step": 35500
+    },
+    {
+      "epoch": 0.04989488141886374,
+      "grad_norm": 1.103615164756775,
+      "learning_rate": 0.00014963327244625542,
+      "loss": 4.9488,
+      "step": 35510
+    },
+    {
+      "epoch": 0.049908932357027314,
+      "grad_norm": 1.1086006164550781,
+      "learning_rate": 0.00014967542503863986,
+      "loss": 5.1068,
+      "step": 35520
+    },
+    {
+      "epoch": 0.04992298329519089,
+      "grad_norm": 1.2253687381744385,
+      "learning_rate": 0.0001497175776310243,
+      "loss": 4.9127,
+      "step": 35530
+    },
+    {
+      "epoch": 0.049937034233354466,
+      "grad_norm": 1.0812503099441528,
+      "learning_rate": 0.00014975973022340872,
+      "loss": 5.0393,
+      "step": 35540
+    },
+    {
+      "epoch": 0.04995108517151805,
+      "grad_norm": 1.108386516571045,
+      "learning_rate": 0.00014980188281579315,
+      "loss": 4.9698,
+      "step": 35550
+    },
+    {
+      "epoch": 0.049965136109681625,
+      "grad_norm": 1.108074426651001,
+      "learning_rate": 0.00014984403540817758,
+      "loss": 5.0325,
+      "step": 35560
+    },
+    {
+      "epoch": 0.0499791870478452,
+      "grad_norm": 1.1028242111206055,
+      "learning_rate": 0.000149886188000562,
+      "loss": 5.0655,
+      "step": 35570
+    },
+    {
+      "epoch": 0.04999323798600878,
+      "grad_norm": 1.1405538320541382,
+      "learning_rate": 0.00014992834059294647,
+      "loss": 5.0859,
+      "step": 35580
+    },
+    {
+      "epoch": 0.050007288924172354,
+      "grad_norm": 1.1215912103652954,
+      "learning_rate": 0.00014997049318533087,
+      "loss": 4.9616,
+      "step": 35590
+    },
+    {
+      "epoch": 0.05002133986233594,
+      "grad_norm": 1.2131357192993164,
+      "learning_rate": 0.0001500126457777153,
+      "loss": 4.9392,
+      "step": 35600
+    },
+    {
+      "epoch": 0.05003539080049951,
+      "grad_norm": 1.1143559217453003,
+      "learning_rate": 0.00015005479837009974,
+      "loss": 4.9206,
+      "step": 35610
+    },
+    {
+      "epoch": 0.05004944173866309,
+      "grad_norm": 1.0911166667938232,
+      "learning_rate": 0.00015009695096248417,
+      "loss": 5.0549,
+      "step": 35620
+    },
+    {
+      "epoch": 0.050063492676826665,
+      "grad_norm": 1.0939656496047974,
+      "learning_rate": 0.0001501391035548686,
+      "loss": 5.0054,
+      "step": 35630
+    },
+    {
+      "epoch": 0.05007754361499024,
+      "grad_norm": 1.123145341873169,
+      "learning_rate": 0.00015018125614725306,
+      "loss": 4.9684,
+      "step": 35640
+    },
+    {
+      "epoch": 0.05009159455315382,
+      "grad_norm": 1.072049856185913,
+      "learning_rate": 0.0001502234087396375,
+      "loss": 4.9857,
+      "step": 35650
+    },
+    {
+      "epoch": 0.0501056454913174,
+      "grad_norm": 1.1123992204666138,
+      "learning_rate": 0.00015026556133202192,
+      "loss": 5.0464,
+      "step": 35660
+    },
+    {
+      "epoch": 0.05011969642948098,
+      "grad_norm": 1.2838733196258545,
+      "learning_rate": 0.00015030771392440632,
+      "loss": 5.0244,
+      "step": 35670
+    },
+    {
+      "epoch": 0.05013374736764455,
+      "grad_norm": 1.0853863954544067,
+      "learning_rate": 0.00015034986651679075,
+      "loss": 5.0946,
+      "step": 35680
+    },
+    {
+      "epoch": 0.05014779830580813,
+      "grad_norm": 1.0802594423294067,
+      "learning_rate": 0.00015039201910917519,
+      "loss": 4.9547,
+      "step": 35690
+    },
+    {
+      "epoch": 0.050161849243971705,
+      "grad_norm": 1.1661789417266846,
+      "learning_rate": 0.00015043417170155964,
+      "loss": 5.0453,
+      "step": 35700
+    },
+    {
+      "epoch": 0.05017590018213529,
+      "grad_norm": 1.0816981792449951,
+      "learning_rate": 0.00015047632429394408,
+      "loss": 5.1652,
+      "step": 35710
+    },
+    {
+      "epoch": 0.050189951120298865,
+      "grad_norm": 1.1324831247329712,
+      "learning_rate": 0.0001505184768863285,
+      "loss": 4.9944,
+      "step": 35720
+    },
+    {
+      "epoch": 0.05020400205846244,
+      "grad_norm": 1.1340142488479614,
+      "learning_rate": 0.0001505606294787129,
+      "loss": 5.0676,
+      "step": 35730
+    },
+    {
+      "epoch": 0.05021805299662602,
+      "grad_norm": 1.0971640348434448,
+      "learning_rate": 0.00015060278207109734,
+      "loss": 4.9213,
+      "step": 35740
+    },
+    {
+      "epoch": 0.05023210393478959,
+      "grad_norm": 1.1202173233032227,
+      "learning_rate": 0.0001506449346634818,
+      "loss": 5.0313,
+      "step": 35750
+    },
+    {
+      "epoch": 0.050246154872953176,
+      "grad_norm": 1.1177239418029785,
+      "learning_rate": 0.00015068708725586623,
+      "loss": 5.0798,
+      "step": 35760
+    },
+    {
+      "epoch": 0.05026020581111675,
+      "grad_norm": 1.1042288541793823,
+      "learning_rate": 0.00015072923984825066,
+      "loss": 5.1434,
+      "step": 35770
+    },
+    {
+      "epoch": 0.05027425674928033,
+      "grad_norm": 1.0804433822631836,
+      "learning_rate": 0.0001507713924406351,
+      "loss": 5.0876,
+      "step": 35780
+    },
+    {
+      "epoch": 0.050288307687443905,
+      "grad_norm": 1.0773561000823975,
+      "learning_rate": 0.00015081354503301952,
+      "loss": 5.0045,
+      "step": 35790
+    },
+    {
+      "epoch": 0.05030235862560748,
+      "grad_norm": 1.1056697368621826,
+      "learning_rate": 0.00015085569762540393,
+      "loss": 5.1274,
+      "step": 35800
+    },
+    {
+      "epoch": 0.050316409563771064,
+      "grad_norm": 1.1184196472167969,
+      "learning_rate": 0.0001508978502177884,
+      "loss": 4.973,
+      "step": 35810
+    },
+    {
+      "epoch": 0.05033046050193464,
+      "grad_norm": 1.1502302885055542,
+      "learning_rate": 0.00015094000281017282,
+      "loss": 4.8589,
+      "step": 35820
+    },
+    {
+      "epoch": 0.050344511440098216,
+      "grad_norm": 1.110656499862671,
+      "learning_rate": 0.00015098215540255725,
+      "loss": 5.1146,
+      "step": 35830
+    },
+    {
+      "epoch": 0.05035856237826179,
+      "grad_norm": 1.0898936986923218,
+      "learning_rate": 0.00015102430799494168,
+      "loss": 5.0243,
+      "step": 35840
+    },
+    {
+      "epoch": 0.05037261331642537,
+      "grad_norm": 1.1786738634109497,
+      "learning_rate": 0.0001510664605873261,
+      "loss": 5.0675,
+      "step": 35850
+    },
+    {
+      "epoch": 0.05038666425458895,
+      "grad_norm": 1.1352401971817017,
+      "learning_rate": 0.00015110861317971052,
+      "loss": 5.0869,
+      "step": 35860
+    },
+    {
+      "epoch": 0.05040071519275253,
+      "grad_norm": 1.2441682815551758,
+      "learning_rate": 0.00015115076577209497,
+      "loss": 4.9753,
+      "step": 35870
+    },
+    {
+      "epoch": 0.050414766130916104,
+      "grad_norm": 1.062424898147583,
+      "learning_rate": 0.0001511929183644794,
+      "loss": 5.1051,
+      "step": 35880
+    },
+    {
+      "epoch": 0.05042881706907968,
+      "grad_norm": 1.0863795280456543,
+      "learning_rate": 0.00015123507095686384,
+      "loss": 5.0136,
+      "step": 35890
+    },
+    {
+      "epoch": 0.050442868007243256,
+      "grad_norm": 1.1157557964324951,
+      "learning_rate": 0.00015127722354924827,
+      "loss": 5.0345,
+      "step": 35900
+    },
+    {
+      "epoch": 0.05045691894540684,
+      "grad_norm": 1.0928047895431519,
+      "learning_rate": 0.0001513193761416327,
+      "loss": 5.0341,
+      "step": 35910
+    },
+    {
+      "epoch": 0.050470969883570416,
+      "grad_norm": 1.103150486946106,
+      "learning_rate": 0.0001513615287340171,
+      "loss": 5.0651,
+      "step": 35920
+    },
+    {
+      "epoch": 0.05048502082173399,
+      "grad_norm": 1.1346688270568848,
+      "learning_rate": 0.00015140368132640156,
+      "loss": 4.9009,
+      "step": 35930
+    },
+    {
+      "epoch": 0.05049907175989757,
+      "grad_norm": 1.1874549388885498,
+      "learning_rate": 0.000151445833918786,
+      "loss": 4.9587,
+      "step": 35940
+    },
+    {
+      "epoch": 0.050513122698061144,
+      "grad_norm": 1.10201895236969,
+      "learning_rate": 0.00015148798651117042,
+      "loss": 5.0994,
+      "step": 35950
+    },
+    {
+      "epoch": 0.05052717363622472,
+      "grad_norm": 1.1142152547836304,
+      "learning_rate": 0.00015153013910355485,
+      "loss": 4.9086,
+      "step": 35960
+    },
+    {
+      "epoch": 0.0505412245743883,
+      "grad_norm": 1.1207566261291504,
+      "learning_rate": 0.00015157229169593929,
+      "loss": 4.9371,
+      "step": 35970
+    },
+    {
+      "epoch": 0.05055527551255188,
+      "grad_norm": 1.146672248840332,
+      "learning_rate": 0.00015161444428832374,
+      "loss": 4.9947,
+      "step": 35980
+    },
+    {
+      "epoch": 0.050569326450715456,
+      "grad_norm": 1.1421736478805542,
+      "learning_rate": 0.00015165659688070815,
+      "loss": 5.1101,
+      "step": 35990
+    },
+    {
+      "epoch": 0.05058337738887903,
+      "grad_norm": 1.0799651145935059,
+      "learning_rate": 0.00015169874947309258,
+      "loss": 4.856,
+      "step": 36000
+    },
+    {
+      "epoch": 0.05059742832704261,
+      "grad_norm": 1.1083508729934692,
+      "learning_rate": 0.000151740902065477,
+      "loss": 4.9631,
+      "step": 36010
+    },
+    {
+      "epoch": 0.05061147926520619,
+      "grad_norm": 1.1054953336715698,
+      "learning_rate": 0.00015178305465786144,
+      "loss": 5.0091,
+      "step": 36020
+    },
+    {
+      "epoch": 0.05062553020336977,
+      "grad_norm": 1.219901204109192,
+      "learning_rate": 0.00015182520725024587,
+      "loss": 4.9669,
+      "step": 36030
+    },
+    {
+      "epoch": 0.050639581141533344,
+      "grad_norm": 1.1822065114974976,
+      "learning_rate": 0.00015186735984263033,
+      "loss": 4.9951,
+      "step": 36040
+    },
+    {
+      "epoch": 0.05065363207969692,
+      "grad_norm": 1.0649558305740356,
+      "learning_rate": 0.00015190951243501473,
+      "loss": 4.9975,
+      "step": 36050
+    },
+    {
+      "epoch": 0.050667683017860496,
+      "grad_norm": 1.1625783443450928,
+      "learning_rate": 0.00015195166502739917,
+      "loss": 4.9153,
+      "step": 36060
+    },
+    {
+      "epoch": 0.05068173395602408,
+      "grad_norm": 1.0902538299560547,
+      "learning_rate": 0.0001519938176197836,
+      "loss": 4.9411,
+      "step": 36070
+    },
+    {
+      "epoch": 0.050695784894187655,
+      "grad_norm": 1.12613046169281,
+      "learning_rate": 0.00015203597021216803,
+      "loss": 4.9862,
+      "step": 36080
+    },
+    {
+      "epoch": 0.05070983583235123,
+      "grad_norm": 1.1342322826385498,
+      "learning_rate": 0.00015207812280455246,
+      "loss": 5.0452,
+      "step": 36090
+    },
+    {
+      "epoch": 0.05072388677051481,
+      "grad_norm": 1.0751118659973145,
+      "learning_rate": 0.00015212027539693692,
+      "loss": 5.0084,
+      "step": 36100
+    },
+    {
+      "epoch": 0.050737937708678384,
+      "grad_norm": 1.1124134063720703,
+      "learning_rate": 0.00015216242798932135,
+      "loss": 4.9795,
+      "step": 36110
+    },
+    {
+      "epoch": 0.05075198864684197,
+      "grad_norm": 1.1308832168579102,
+      "learning_rate": 0.00015220458058170575,
+      "loss": 5.0751,
+      "step": 36120
+    },
+    {
+      "epoch": 0.05076603958500554,
+      "grad_norm": 1.1083581447601318,
+      "learning_rate": 0.00015224673317409018,
+      "loss": 5.0698,
+      "step": 36130
+    },
+    {
+      "epoch": 0.05078009052316912,
+      "grad_norm": 1.140774130821228,
+      "learning_rate": 0.00015228888576647462,
+      "loss": 5.041,
+      "step": 36140
+    },
+    {
+      "epoch": 0.050794141461332695,
+      "grad_norm": 1.0934237241744995,
+      "learning_rate": 0.00015233103835885905,
+      "loss": 5.0086,
+      "step": 36150
+    },
+    {
+      "epoch": 0.05080819239949627,
+      "grad_norm": 1.0771994590759277,
+      "learning_rate": 0.0001523731909512435,
+      "loss": 4.9932,
+      "step": 36160
+    },
+    {
+      "epoch": 0.050822243337659855,
+      "grad_norm": 1.076748251914978,
+      "learning_rate": 0.00015241534354362794,
+      "loss": 4.8132,
+      "step": 36170
+    },
+    {
+      "epoch": 0.05083629427582343,
+      "grad_norm": 1.0804756879806519,
+      "learning_rate": 0.00015245749613601234,
+      "loss": 4.9832,
+      "step": 36180
+    },
+    {
+      "epoch": 0.05085034521398701,
+      "grad_norm": 1.0959327220916748,
+      "learning_rate": 0.00015249964872839677,
+      "loss": 4.9817,
+      "step": 36190
+    },
+    {
+      "epoch": 0.05086439615215058,
+      "grad_norm": 1.185053825378418,
+      "learning_rate": 0.0001525418013207812,
+      "loss": 5.0076,
+      "step": 36200
+    },
+    {
+      "epoch": 0.05087844709031416,
+      "grad_norm": 1.1760470867156982,
+      "learning_rate": 0.00015258395391316563,
+      "loss": 4.9851,
+      "step": 36210
+    },
+    {
+      "epoch": 0.05089249802847774,
+      "grad_norm": 1.1177805662155151,
+      "learning_rate": 0.0001526261065055501,
+      "loss": 5.0122,
+      "step": 36220
+    },
+    {
+      "epoch": 0.05090654896664132,
+      "grad_norm": 1.0969743728637695,
+      "learning_rate": 0.00015266825909793452,
+      "loss": 4.8199,
+      "step": 36230
+    },
+    {
+      "epoch": 0.050920599904804895,
+      "grad_norm": 1.171985387802124,
+      "learning_rate": 0.00015271041169031895,
+      "loss": 4.9671,
+      "step": 36240
+    },
+    {
+      "epoch": 0.05093465084296847,
+      "grad_norm": 1.0644655227661133,
+      "learning_rate": 0.00015275256428270336,
+      "loss": 5.0531,
+      "step": 36250
+    },
+    {
+      "epoch": 0.05094870178113205,
+      "grad_norm": 1.0977424383163452,
+      "learning_rate": 0.0001527947168750878,
+      "loss": 4.9316,
+      "step": 36260
+    },
+    {
+      "epoch": 0.05096275271929562,
+      "grad_norm": 1.0816428661346436,
+      "learning_rate": 0.00015283686946747222,
+      "loss": 5.0497,
+      "step": 36270
+    },
+    {
+      "epoch": 0.050976803657459206,
+      "grad_norm": 1.1141808032989502,
+      "learning_rate": 0.00015287902205985668,
+      "loss": 4.8977,
+      "step": 36280
+    },
+    {
+      "epoch": 0.05099085459562278,
+      "grad_norm": 1.135312557220459,
+      "learning_rate": 0.0001529211746522411,
+      "loss": 5.0077,
+      "step": 36290
+    },
+    {
+      "epoch": 0.05100490553378636,
+      "grad_norm": 1.0765386819839478,
+      "learning_rate": 0.00015296332724462554,
+      "loss": 4.9414,
+      "step": 36300
+    },
+    {
+      "epoch": 0.051018956471949935,
+      "grad_norm": 1.0505393743515015,
+      "learning_rate": 0.00015300547983700995,
+      "loss": 4.8978,
+      "step": 36310
+    },
+    {
+      "epoch": 0.05103300741011351,
+      "grad_norm": 1.127390742301941,
+      "learning_rate": 0.00015304763242939438,
+      "loss": 5.0173,
+      "step": 36320
+    },
+    {
+      "epoch": 0.051047058348277094,
+      "grad_norm": 1.1132893562316895,
+      "learning_rate": 0.00015308978502177883,
+      "loss": 5.0465,
+      "step": 36330
+    },
+    {
+      "epoch": 0.05106110928644067,
+      "grad_norm": 1.0964865684509277,
+      "learning_rate": 0.00015313193761416327,
+      "loss": 5.009,
+      "step": 36340
+    },
+    {
+      "epoch": 0.051075160224604246,
+      "grad_norm": 1.0559228658676147,
+      "learning_rate": 0.0001531740902065477,
+      "loss": 5.136,
+      "step": 36350
+    },
+    {
+      "epoch": 0.05108921116276782,
+      "grad_norm": 1.0860519409179688,
+      "learning_rate": 0.00015321624279893213,
+      "loss": 5.0737,
+      "step": 36360
+    },
+    {
+      "epoch": 0.0511032621009314,
+      "grad_norm": 1.1100194454193115,
+      "learning_rate": 0.00015325839539131656,
+      "loss": 4.9814,
+      "step": 36370
+    },
+    {
+      "epoch": 0.05111731303909498,
+      "grad_norm": 1.0740619897842407,
+      "learning_rate": 0.00015330054798370096,
+      "loss": 4.9415,
+      "step": 36380
+    },
+    {
+      "epoch": 0.05113136397725856,
+      "grad_norm": 1.1333972215652466,
+      "learning_rate": 0.00015334270057608542,
+      "loss": 5.037,
+      "step": 36390
+    },
+    {
+      "epoch": 0.051145414915422134,
+      "grad_norm": 1.1923198699951172,
+      "learning_rate": 0.00015338485316846985,
+      "loss": 4.9674,
+      "step": 36400
+    },
+    {
+      "epoch": 0.05115946585358571,
+      "grad_norm": 1.0857388973236084,
+      "learning_rate": 0.00015342700576085428,
+      "loss": 5.0191,
+      "step": 36410
+    },
+    {
+      "epoch": 0.051173516791749286,
+      "grad_norm": 1.1336112022399902,
+      "learning_rate": 0.00015346494309400027,
+      "loss": 4.9396,
+      "step": 36420
+    },
+    {
+      "epoch": 0.05118756772991287,
+      "grad_norm": 1.1554714441299438,
+      "learning_rate": 0.0001535070956863847,
+      "loss": 4.9489,
+      "step": 36430
+    },
+    {
+      "epoch": 0.051201618668076446,
+      "grad_norm": 1.1246522665023804,
+      "learning_rate": 0.00015354924827876913,
+      "loss": 5.0433,
+      "step": 36440
+    },
+    {
+      "epoch": 0.05121566960624002,
+      "grad_norm": 1.1257320642471313,
+      "learning_rate": 0.00015359140087115356,
+      "loss": 4.9265,
+      "step": 36450
+    },
+    {
+      "epoch": 0.0512297205444036,
+      "grad_norm": 1.1546956300735474,
+      "learning_rate": 0.00015363355346353802,
+      "loss": 4.9957,
+      "step": 36460
+    },
+    {
+      "epoch": 0.051243771482567174,
+      "grad_norm": 1.1014739274978638,
+      "learning_rate": 0.00015367570605592242,
+      "loss": 4.9966,
+      "step": 36470
+    },
+    {
+      "epoch": 0.05125782242073076,
+      "grad_norm": 1.1065007448196411,
+      "learning_rate": 0.00015371785864830685,
+      "loss": 4.9691,
+      "step": 36480
+    },
+    {
+      "epoch": 0.05127187335889433,
+      "grad_norm": 1.0437864065170288,
+      "learning_rate": 0.00015376001124069128,
+      "loss": 5.0153,
+      "step": 36490
+    },
+    {
+      "epoch": 0.05128592429705791,
+      "grad_norm": 1.0865705013275146,
+      "learning_rate": 0.00015380216383307572,
+      "loss": 5.0422,
+      "step": 36500
+    },
+    {
+      "epoch": 0.051299975235221486,
+      "grad_norm": 1.0838688611984253,
+      "learning_rate": 0.00015384431642546015,
+      "loss": 4.9375,
+      "step": 36510
+    },
+    {
+      "epoch": 0.05131402617338506,
+      "grad_norm": 1.13599693775177,
+      "learning_rate": 0.0001538864690178446,
+      "loss": 4.9728,
+      "step": 36520
+    },
+    {
+      "epoch": 0.051328077111548645,
+      "grad_norm": 1.183531403541565,
+      "learning_rate": 0.00015392862161022904,
+      "loss": 5.066,
+      "step": 36530
+    },
+    {
+      "epoch": 0.05134212804971222,
+      "grad_norm": 1.1362754106521606,
+      "learning_rate": 0.00015397077420261344,
+      "loss": 5.0401,
+      "step": 36540
+    },
+    {
+      "epoch": 0.0513561789878758,
+      "grad_norm": 1.1068400144577026,
+      "learning_rate": 0.00015401292679499787,
+      "loss": 4.9703,
+      "step": 36550
+    },
+    {
+      "epoch": 0.051370229926039374,
+      "grad_norm": 1.1438322067260742,
+      "learning_rate": 0.0001540550793873823,
+      "loss": 5.0719,
+      "step": 36560
+    },
+    {
+      "epoch": 0.05138428086420295,
+      "grad_norm": 1.0228193998336792,
+      "learning_rate": 0.00015409723197976673,
+      "loss": 4.9806,
+      "step": 36570
+    },
+    {
+      "epoch": 0.051398331802366526,
+      "grad_norm": 1.0660607814788818,
+      "learning_rate": 0.0001541393845721512,
+      "loss": 5.1565,
+      "step": 36580
+    },
+    {
+      "epoch": 0.05141238274053011,
+      "grad_norm": 1.1452786922454834,
+      "learning_rate": 0.00015418153716453562,
+      "loss": 5.0448,
+      "step": 36590
+    },
+    {
+      "epoch": 0.051426433678693685,
+      "grad_norm": 1.1097620725631714,
+      "learning_rate": 0.00015422368975692003,
+      "loss": 5.046,
+      "step": 36600
+    },
+    {
+      "epoch": 0.05144048461685726,
+      "grad_norm": 1.2317959070205688,
+      "learning_rate": 0.00015426584234930446,
+      "loss": 4.9258,
+      "step": 36610
+    },
+    {
+      "epoch": 0.05145453555502084,
+      "grad_norm": 1.0700747966766357,
+      "learning_rate": 0.0001543079949416889,
+      "loss": 5.0544,
+      "step": 36620
+    },
+    {
+      "epoch": 0.051468586493184414,
+      "grad_norm": 1.0716843605041504,
+      "learning_rate": 0.00015435014753407332,
+      "loss": 5.0264,
+      "step": 36630
+    },
+    {
+      "epoch": 0.051482637431348,
+      "grad_norm": 1.0859941244125366,
+      "learning_rate": 0.00015439230012645778,
+      "loss": 4.989,
+      "step": 36640
+    },
+    {
+      "epoch": 0.05149668836951157,
+      "grad_norm": 1.0916000604629517,
+      "learning_rate": 0.0001544344527188422,
+      "loss": 4.971,
+      "step": 36650
+    },
+    {
+      "epoch": 0.05151073930767515,
+      "grad_norm": 1.139979362487793,
+      "learning_rate": 0.00015447660531122664,
+      "loss": 4.9515,
+      "step": 36660
+    },
+    {
+      "epoch": 0.051524790245838725,
+      "grad_norm": 1.1469497680664062,
+      "learning_rate": 0.00015451875790361105,
+      "loss": 4.9998,
+      "step": 36670
+    },
+    {
+      "epoch": 0.0515388411840023,
+      "grad_norm": 1.1082267761230469,
+      "learning_rate": 0.00015456091049599548,
+      "loss": 4.9369,
+      "step": 36680
+    },
+    {
+      "epoch": 0.051552892122165885,
+      "grad_norm": 1.1943933963775635,
+      "learning_rate": 0.0001546030630883799,
+      "loss": 4.9594,
+      "step": 36690
+    },
+    {
+      "epoch": 0.05156694306032946,
+      "grad_norm": 1.0887293815612793,
+      "learning_rate": 0.00015464521568076437,
+      "loss": 4.9823,
+      "step": 36700
+    },
+    {
+      "epoch": 0.05158099399849304,
+      "grad_norm": 1.1254092454910278,
+      "learning_rate": 0.0001546873682731488,
+      "loss": 4.9227,
+      "step": 36710
+    },
+    {
+      "epoch": 0.05159504493665661,
+      "grad_norm": 1.1450250148773193,
+      "learning_rate": 0.00015472952086553323,
+      "loss": 4.9345,
+      "step": 36720
+    },
+    {
+      "epoch": 0.05160909587482019,
+      "grad_norm": 1.0992549657821655,
+      "learning_rate": 0.00015477167345791763,
+      "loss": 4.9191,
+      "step": 36730
+    },
+    {
+      "epoch": 0.05162314681298377,
+      "grad_norm": 1.0670653581619263,
+      "learning_rate": 0.00015481382605030206,
+      "loss": 4.9883,
+      "step": 36740
+    },
+    {
+      "epoch": 0.05163719775114735,
+      "grad_norm": 1.074080467224121,
+      "learning_rate": 0.0001548559786426865,
+      "loss": 4.9899,
+      "step": 36750
+    },
+    {
+      "epoch": 0.051651248689310925,
+      "grad_norm": 1.1951225996017456,
+      "learning_rate": 0.00015489813123507095,
+      "loss": 5.037,
+      "step": 36760
+    },
+    {
+      "epoch": 0.0516652996274745,
+      "grad_norm": 1.0775877237319946,
+      "learning_rate": 0.00015494028382745538,
+      "loss": 5.0046,
+      "step": 36770
+    },
+    {
+      "epoch": 0.05167935056563808,
+      "grad_norm": 1.087863802909851,
+      "learning_rate": 0.00015498243641983982,
+      "loss": 4.9907,
+      "step": 36780
+    },
+    {
+      "epoch": 0.05169340150380166,
+      "grad_norm": 1.0674740076065063,
+      "learning_rate": 0.00015502458901222425,
+      "loss": 5.0147,
+      "step": 36790
+    },
+    {
+      "epoch": 0.051707452441965236,
+      "grad_norm": 1.069732904434204,
+      "learning_rate": 0.00015506674160460865,
+      "loss": 4.8633,
+      "step": 36800
+    },
+    {
+      "epoch": 0.05172150338012881,
+      "grad_norm": 1.0820239782333374,
+      "learning_rate": 0.00015510889419699308,
+      "loss": 5.0354,
+      "step": 36810
+    },
+    {
+      "epoch": 0.05173555431829239,
+      "grad_norm": 1.1347521543502808,
+      "learning_rate": 0.00015515104678937754,
+      "loss": 4.8265,
+      "step": 36820
+    },
+    {
+      "epoch": 0.051749605256455965,
+      "grad_norm": 1.1913303136825562,
+      "learning_rate": 0.00015519319938176197,
+      "loss": 4.9479,
+      "step": 36830
+    },
+    {
+      "epoch": 0.05176365619461955,
+      "grad_norm": 1.1147087812423706,
+      "learning_rate": 0.0001552353519741464,
+      "loss": 4.9907,
+      "step": 36840
+    },
+    {
+      "epoch": 0.051777707132783124,
+      "grad_norm": 1.076640009880066,
+      "learning_rate": 0.00015527750456653083,
+      "loss": 4.9654,
+      "step": 36850
+    },
+    {
+      "epoch": 0.0517917580709467,
+      "grad_norm": 1.1436219215393066,
+      "learning_rate": 0.00015531965715891524,
+      "loss": 4.9585,
+      "step": 36860
+    },
+    {
+      "epoch": 0.051805809009110276,
+      "grad_norm": 1.0725477933883667,
+      "learning_rate": 0.00015536180975129967,
+      "loss": 5.0113,
+      "step": 36870
+    },
+    {
+      "epoch": 0.05181985994727385,
+      "grad_norm": 1.087472677230835,
+      "learning_rate": 0.00015540396234368413,
+      "loss": 4.8823,
+      "step": 36880
+    },
+    {
+      "epoch": 0.05183391088543743,
+      "grad_norm": 1.0537265539169312,
+      "learning_rate": 0.00015544611493606856,
+      "loss": 4.8053,
+      "step": 36890
+    },
+    {
+      "epoch": 0.05184796182360101,
+      "grad_norm": 1.0963505506515503,
+      "learning_rate": 0.000155488267528453,
+      "loss": 5.0603,
+      "step": 36900
+    },
+    {
+      "epoch": 0.05186201276176459,
+      "grad_norm": 1.1273267269134521,
+      "learning_rate": 0.00015553042012083742,
+      "loss": 5.0182,
+      "step": 36910
+    },
+    {
+      "epoch": 0.051876063699928164,
+      "grad_norm": 1.0548876523971558,
+      "learning_rate": 0.00015557257271322188,
+      "loss": 5.1116,
+      "step": 36920
+    },
+    {
+      "epoch": 0.05189011463809174,
+      "grad_norm": 1.1563364267349243,
+      "learning_rate": 0.00015561472530560626,
+      "loss": 4.9696,
+      "step": 36930
+    },
+    {
+      "epoch": 0.051904165576255316,
+      "grad_norm": 1.0634068250656128,
+      "learning_rate": 0.00015565687789799071,
+      "loss": 4.9628,
+      "step": 36940
+    },
+    {
+      "epoch": 0.0519182165144189,
+      "grad_norm": 1.182931900024414,
+      "learning_rate": 0.00015569903049037515,
+      "loss": 5.021,
+      "step": 36950
+    },
+    {
+      "epoch": 0.051932267452582476,
+      "grad_norm": 1.0970754623413086,
+      "learning_rate": 0.00015574118308275958,
+      "loss": 5.0209,
+      "step": 36960
+    },
+    {
+      "epoch": 0.05194631839074605,
+      "grad_norm": 1.10125732421875,
+      "learning_rate": 0.000155783335675144,
+      "loss": 4.963,
+      "step": 36970
+    },
+    {
+      "epoch": 0.05196036932890963,
+      "grad_norm": 1.1242049932479858,
+      "learning_rate": 0.00015582548826752847,
+      "loss": 5.0308,
+      "step": 36980
+    },
+    {
+      "epoch": 0.051974420267073204,
+      "grad_norm": 1.114256739616394,
+      "learning_rate": 0.00015586764085991287,
+      "loss": 5.0974,
+      "step": 36990
+    },
+    {
+      "epoch": 0.05198847120523679,
+      "grad_norm": 1.080618143081665,
+      "learning_rate": 0.0001559097934522973,
+      "loss": 4.9911,
+      "step": 37000
+    },
+    {
+      "epoch": 0.05200252214340036,
+      "grad_norm": 1.097937822341919,
+      "learning_rate": 0.00015595194604468173,
+      "loss": 4.9756,
+      "step": 37010
+    },
+    {
+      "epoch": 0.05201657308156394,
+      "grad_norm": 1.2937126159667969,
+      "learning_rate": 0.00015599409863706616,
+      "loss": 5.0298,
+      "step": 37020
+    },
+    {
+      "epoch": 0.052030624019727516,
+      "grad_norm": 1.1030606031417847,
+      "learning_rate": 0.0001560362512294506,
+      "loss": 4.9976,
+      "step": 37030
+    },
+    {
+      "epoch": 0.05204467495789109,
+      "grad_norm": 1.1693426370620728,
+      "learning_rate": 0.00015607840382183505,
+      "loss": 4.9024,
+      "step": 37040
+    },
+    {
+      "epoch": 0.052058725896054675,
+      "grad_norm": 1.1187068223953247,
+      "learning_rate": 0.00015612055641421946,
+      "loss": 5.0483,
+      "step": 37050
+    },
+    {
+      "epoch": 0.05207277683421825,
+      "grad_norm": 1.0532891750335693,
+      "learning_rate": 0.0001561627090066039,
+      "loss": 4.9193,
+      "step": 37060
+    },
+    {
+      "epoch": 0.05208682777238183,
+      "grad_norm": 1.1207010746002197,
+      "learning_rate": 0.00015620486159898832,
+      "loss": 4.9848,
+      "step": 37070
+    },
+    {
+      "epoch": 0.052100878710545404,
+      "grad_norm": 1.0693873167037964,
+      "learning_rate": 0.00015624701419137275,
+      "loss": 4.9215,
+      "step": 37080
+    },
+    {
+      "epoch": 0.05211492964870898,
+      "grad_norm": 1.0835233926773071,
+      "learning_rate": 0.00015628916678375718,
+      "loss": 4.7552,
+      "step": 37090
+    },
+    {
+      "epoch": 0.05212898058687256,
+      "grad_norm": 1.124945878982544,
+      "learning_rate": 0.00015633131937614164,
+      "loss": 5.0005,
+      "step": 37100
+    },
+    {
+      "epoch": 0.05214303152503614,
+      "grad_norm": 1.0725970268249512,
+      "learning_rate": 0.00015637347196852607,
+      "loss": 4.9894,
+      "step": 37110
+    },
+    {
+      "epoch": 0.052157082463199715,
+      "grad_norm": 1.1235997676849365,
+      "learning_rate": 0.00015641562456091048,
+      "loss": 5.0398,
+      "step": 37120
+    },
+    {
+      "epoch": 0.05217113340136329,
+      "grad_norm": 1.1638944149017334,
+      "learning_rate": 0.0001564577771532949,
+      "loss": 4.8523,
+      "step": 37130
+    },
+    {
+      "epoch": 0.05218518433952687,
+      "grad_norm": 1.1262266635894775,
+      "learning_rate": 0.00015649992974567934,
+      "loss": 4.8287,
+      "step": 37140
+    },
+    {
+      "epoch": 0.05219923527769045,
+      "grad_norm": 1.1217573881149292,
+      "learning_rate": 0.00015654208233806377,
+      "loss": 5.0582,
+      "step": 37150
+    },
+    {
+      "epoch": 0.05221328621585403,
+      "grad_norm": 1.1533286571502686,
+      "learning_rate": 0.00015658423493044823,
+      "loss": 4.939,
+      "step": 37160
+    },
+    {
+      "epoch": 0.0522273371540176,
+      "grad_norm": 1.063332438468933,
+      "learning_rate": 0.00015662638752283266,
+      "loss": 5.0078,
+      "step": 37170
+    },
+    {
+      "epoch": 0.05224138809218118,
+      "grad_norm": 1.0666674375534058,
+      "learning_rate": 0.00015666854011521706,
+      "loss": 4.8933,
+      "step": 37180
+    },
+    {
+      "epoch": 0.052255439030344755,
+      "grad_norm": 1.1299502849578857,
+      "learning_rate": 0.0001567106927076015,
+      "loss": 4.9698,
+      "step": 37190
+    },
+    {
+      "epoch": 0.05226948996850833,
+      "grad_norm": 1.0899701118469238,
+      "learning_rate": 0.00015675284529998592,
+      "loss": 5.1382,
+      "step": 37200
+    },
+    {
+      "epoch": 0.052283540906671915,
+      "grad_norm": 1.0890862941741943,
+      "learning_rate": 0.00015679499789237036,
+      "loss": 4.9112,
+      "step": 37210
+    },
+    {
+      "epoch": 0.05229759184483549,
+      "grad_norm": 1.1074151992797852,
+      "learning_rate": 0.00015683715048475481,
+      "loss": 4.9958,
+      "step": 37220
+    },
+    {
+      "epoch": 0.05231164278299907,
+      "grad_norm": 1.047777771949768,
+      "learning_rate": 0.00015687930307713925,
+      "loss": 4.8429,
+      "step": 37230
+    },
+    {
+      "epoch": 0.05232569372116264,
+      "grad_norm": 1.0324103832244873,
+      "learning_rate": 0.00015692145566952368,
+      "loss": 4.9171,
+      "step": 37240
+    },
+    {
+      "epoch": 0.05233974465932622,
+      "grad_norm": 1.043428897857666,
+      "learning_rate": 0.00015696360826190808,
+      "loss": 5.0088,
+      "step": 37250
+    },
+    {
+      "epoch": 0.0523537955974898,
+      "grad_norm": 1.107131004333496,
+      "learning_rate": 0.0001570057608542925,
+      "loss": 5.0575,
+      "step": 37260
+    },
+    {
+      "epoch": 0.05236784653565338,
+      "grad_norm": 1.1451491117477417,
+      "learning_rate": 0.00015704791344667694,
+      "loss": 4.9536,
+      "step": 37270
+    },
+    {
+      "epoch": 0.052381897473816955,
+      "grad_norm": 1.1125915050506592,
+      "learning_rate": 0.0001570900660390614,
+      "loss": 4.987,
+      "step": 37280
+    },
+    {
+      "epoch": 0.05239594841198053,
+      "grad_norm": 1.1330432891845703,
+      "learning_rate": 0.00015713221863144583,
+      "loss": 4.9512,
+      "step": 37290
+    },
+    {
+      "epoch": 0.05240999935014411,
+      "grad_norm": 1.0728533267974854,
+      "learning_rate": 0.00015717437122383026,
+      "loss": 5.0056,
+      "step": 37300
+    },
+    {
+      "epoch": 0.05242405028830769,
+      "grad_norm": 1.1220287084579468,
+      "learning_rate": 0.00015721652381621467,
+      "loss": 4.8007,
+      "step": 37310
+    },
+    {
+      "epoch": 0.052438101226471266,
+      "grad_norm": 1.086573839187622,
+      "learning_rate": 0.0001572586764085991,
+      "loss": 4.8958,
+      "step": 37320
+    },
+    {
+      "epoch": 0.05245215216463484,
+      "grad_norm": 1.034942865371704,
+      "learning_rate": 0.00015730082900098353,
+      "loss": 4.9862,
+      "step": 37330
+    },
+    {
+      "epoch": 0.05246620310279842,
+      "grad_norm": 1.2076352834701538,
+      "learning_rate": 0.000157342981593368,
+      "loss": 5.0167,
+      "step": 37340
+    },
+    {
+      "epoch": 0.052480254040961995,
+      "grad_norm": 1.0755456686019897,
+      "learning_rate": 0.00015738513418575242,
+      "loss": 4.9192,
+      "step": 37350
+    },
+    {
+      "epoch": 0.05249430497912558,
+      "grad_norm": 1.0850402116775513,
+      "learning_rate": 0.00015742728677813685,
+      "loss": 5.0214,
+      "step": 37360
+    },
+    {
+      "epoch": 0.052508355917289154,
+      "grad_norm": 1.1084506511688232,
+      "learning_rate": 0.00015746943937052128,
+      "loss": 5.02,
+      "step": 37370
+    },
+    {
+      "epoch": 0.05252240685545273,
+      "grad_norm": 1.1227412223815918,
+      "learning_rate": 0.00015751159196290569,
+      "loss": 5.046,
+      "step": 37380
+    },
+    {
+      "epoch": 0.052536457793616306,
+      "grad_norm": 1.1785093545913696,
+      "learning_rate": 0.00015755374455529012,
+      "loss": 4.9965,
+      "step": 37390
+    },
+    {
+      "epoch": 0.05255050873177988,
+      "grad_norm": 1.1120465993881226,
+      "learning_rate": 0.00015759589714767458,
+      "loss": 4.9608,
+      "step": 37400
+    },
+    {
+      "epoch": 0.052564559669943466,
+      "grad_norm": 1.1275166273117065,
+      "learning_rate": 0.000157638049740059,
+      "loss": 4.9256,
+      "step": 37410
+    },
+    {
+      "epoch": 0.05257861060810704,
+      "grad_norm": 1.0952259302139282,
+      "learning_rate": 0.00015768020233244344,
+      "loss": 4.9027,
+      "step": 37420
+    },
+    {
+      "epoch": 0.05259266154627062,
+      "grad_norm": 1.084206223487854,
+      "learning_rate": 0.00015772235492482787,
+      "loss": 4.9846,
+      "step": 37430
+    },
+    {
+      "epoch": 0.052606712484434194,
+      "grad_norm": 1.1174198389053345,
+      "learning_rate": 0.00015776450751721227,
+      "loss": 4.8721,
+      "step": 37440
+    },
+    {
+      "epoch": 0.05262076342259777,
+      "grad_norm": 1.077741265296936,
+      "learning_rate": 0.0001578066601095967,
+      "loss": 4.9004,
+      "step": 37450
+    },
+    {
+      "epoch": 0.05263481436076135,
+      "grad_norm": 1.1274837255477905,
+      "learning_rate": 0.00015784881270198116,
+      "loss": 4.9527,
+      "step": 37460
+    },
+    {
+      "epoch": 0.05264886529892493,
+      "grad_norm": 1.098956823348999,
+      "learning_rate": 0.0001578909652943656,
+      "loss": 4.9924,
+      "step": 37470
+    },
+    {
+      "epoch": 0.052662916237088506,
+      "grad_norm": 1.2196323871612549,
+      "learning_rate": 0.00015793311788675002,
+      "loss": 4.9223,
+      "step": 37480
+    },
+    {
+      "epoch": 0.05267696717525208,
+      "grad_norm": 1.065298318862915,
+      "learning_rate": 0.00015797527047913446,
+      "loss": 4.9821,
+      "step": 37490
+    },
+    {
+      "epoch": 0.05269101811341566,
+      "grad_norm": 1.0735142230987549,
+      "learning_rate": 0.00015801742307151891,
+      "loss": 5.0825,
+      "step": 37500
+    },
+    {
+      "epoch": 0.052705069051579234,
+      "grad_norm": 1.075454831123352,
+      "learning_rate": 0.0001580595756639033,
+      "loss": 4.9954,
+      "step": 37510
+    },
+    {
+      "epoch": 0.05271911998974282,
+      "grad_norm": 1.0807503461837769,
+      "learning_rate": 0.00015810172825628775,
+      "loss": 4.9428,
+      "step": 37520
+    },
+    {
+      "epoch": 0.05273317092790639,
+      "grad_norm": 1.1390740871429443,
+      "learning_rate": 0.00015814388084867218,
+      "loss": 4.9437,
+      "step": 37530
+    },
+    {
+      "epoch": 0.05274722186606997,
+      "grad_norm": 1.0902714729309082,
+      "learning_rate": 0.0001581860334410566,
+      "loss": 5.0031,
+      "step": 37540
+    },
+    {
+      "epoch": 0.052761272804233546,
+      "grad_norm": 1.0692888498306274,
+      "learning_rate": 0.00015822818603344104,
+      "loss": 4.8607,
+      "step": 37550
+    },
+    {
+      "epoch": 0.05277532374239712,
+      "grad_norm": 1.0641309022903442,
+      "learning_rate": 0.0001582703386258255,
+      "loss": 5.0244,
+      "step": 37560
+    },
+    {
+      "epoch": 0.052789374680560705,
+      "grad_norm": 1.0614022016525269,
+      "learning_rate": 0.00015831249121820988,
+      "loss": 4.937,
+      "step": 37570
+    },
+    {
+      "epoch": 0.05280342561872428,
+      "grad_norm": 1.0764083862304688,
+      "learning_rate": 0.00015835464381059434,
+      "loss": 4.8395,
+      "step": 37580
+    },
+    {
+      "epoch": 0.05281747655688786,
+      "grad_norm": 1.091915488243103,
+      "learning_rate": 0.00015839679640297877,
+      "loss": 5.0053,
+      "step": 37590
+    },
+    {
+      "epoch": 0.052831527495051434,
+      "grad_norm": 1.3076285123825073,
+      "learning_rate": 0.0001584389489953632,
+      "loss": 4.8942,
+      "step": 37600
+    },
+    {
+      "epoch": 0.05284557843321501,
+      "grad_norm": 1.0910054445266724,
+      "learning_rate": 0.00015848110158774763,
+      "loss": 4.9173,
+      "step": 37610
+    },
+    {
+      "epoch": 0.05285962937137859,
+      "grad_norm": 1.0928317308425903,
+      "learning_rate": 0.0001585232541801321,
+      "loss": 5.0224,
+      "step": 37620
+    },
+    {
+      "epoch": 0.05287368030954217,
+      "grad_norm": 1.0733320713043213,
+      "learning_rate": 0.00015856540677251652,
+      "loss": 4.9854,
+      "step": 37630
+    },
+    {
+      "epoch": 0.052887731247705745,
+      "grad_norm": 1.0689090490341187,
+      "learning_rate": 0.00015860755936490092,
+      "loss": 4.9577,
+      "step": 37640
+    },
+    {
+      "epoch": 0.05290178218586932,
+      "grad_norm": 1.1283597946166992,
+      "learning_rate": 0.00015864971195728535,
+      "loss": 5.0415,
+      "step": 37650
+    },
+    {
+      "epoch": 0.0529158331240329,
+      "grad_norm": 1.1071979999542236,
+      "learning_rate": 0.00015869186454966979,
+      "loss": 5.0478,
+      "step": 37660
+    },
+    {
+      "epoch": 0.05292988406219648,
+      "grad_norm": 1.0887236595153809,
+      "learning_rate": 0.00015873401714205422,
+      "loss": 5.0003,
+      "step": 37670
+    },
+    {
+      "epoch": 0.05294393500036006,
+      "grad_norm": 1.1010711193084717,
+      "learning_rate": 0.00015877616973443868,
+      "loss": 4.8824,
+      "step": 37680
+    },
+    {
+      "epoch": 0.05295798593852363,
+      "grad_norm": 1.0766741037368774,
+      "learning_rate": 0.0001588183223268231,
+      "loss": 5.1363,
+      "step": 37690
+    },
+    {
+      "epoch": 0.05297203687668721,
+      "grad_norm": 1.0775905847549438,
+      "learning_rate": 0.0001588604749192075,
+      "loss": 4.9825,
+      "step": 37700
+    },
+    {
+      "epoch": 0.052986087814850785,
+      "grad_norm": 1.0823009014129639,
+      "learning_rate": 0.00015890262751159194,
+      "loss": 5.2041,
+      "step": 37710
+    },
+    {
+      "epoch": 0.05300013875301437,
+      "grad_norm": 1.0547107458114624,
+      "learning_rate": 0.00015894478010397637,
+      "loss": 5.036,
+      "step": 37720
+    },
+    {
+      "epoch": 0.053014189691177944,
+      "grad_norm": 1.0810598134994507,
+      "learning_rate": 0.0001589869326963608,
+      "loss": 4.7697,
+      "step": 37730
+    },
+    {
+      "epoch": 0.05302824062934152,
+      "grad_norm": 1.0776020288467407,
+      "learning_rate": 0.00015902908528874526,
+      "loss": 4.95,
+      "step": 37740
+    },
+    {
+      "epoch": 0.0530422915675051,
+      "grad_norm": 1.068782925605774,
+      "learning_rate": 0.0001590712378811297,
+      "loss": 4.9632,
+      "step": 37750
+    },
+    {
+      "epoch": 0.05305634250566867,
+      "grad_norm": 1.0810242891311646,
+      "learning_rate": 0.0001591133904735141,
+      "loss": 5.0325,
+      "step": 37760
+    },
+    {
+      "epoch": 0.053070393443832256,
+      "grad_norm": 1.0585830211639404,
+      "learning_rate": 0.00015915554306589853,
+      "loss": 4.9651,
+      "step": 37770
+    },
+    {
+      "epoch": 0.05308444438199583,
+      "grad_norm": 1.0817153453826904,
+      "learning_rate": 0.00015919769565828296,
+      "loss": 5.022,
+      "step": 37780
+    },
+    {
+      "epoch": 0.05309849532015941,
+      "grad_norm": 1.1465531587600708,
+      "learning_rate": 0.0001592398482506674,
+      "loss": 4.9242,
+      "step": 37790
+    },
+    {
+      "epoch": 0.053112546258322985,
+      "grad_norm": 1.1970518827438354,
+      "learning_rate": 0.00015928200084305185,
+      "loss": 5.0169,
+      "step": 37800
+    },
+    {
+      "epoch": 0.05312659719648656,
+      "grad_norm": 1.0447371006011963,
+      "learning_rate": 0.00015932415343543628,
+      "loss": 5.0174,
+      "step": 37810
+    },
+    {
+      "epoch": 0.053140648134650144,
+      "grad_norm": 1.117707371711731,
+      "learning_rate": 0.0001593663060278207,
+      "loss": 4.9709,
+      "step": 37820
+    },
+    {
+      "epoch": 0.05315469907281372,
+      "grad_norm": 1.101830005645752,
+      "learning_rate": 0.00015940845862020512,
+      "loss": 4.9622,
+      "step": 37830
+    },
+    {
+      "epoch": 0.053168750010977296,
+      "grad_norm": 1.0951014757156372,
+      "learning_rate": 0.00015945061121258955,
+      "loss": 5.005,
+      "step": 37840
+    },
+    {
+      "epoch": 0.05318280094914087,
+      "grad_norm": 1.0685999393463135,
+      "learning_rate": 0.00015949276380497398,
+      "loss": 4.9262,
+      "step": 37850
+    },
+    {
+      "epoch": 0.05319685188730445,
+      "grad_norm": 1.1599674224853516,
+      "learning_rate": 0.00015953491639735844,
+      "loss": 4.9453,
+      "step": 37860
+    },
+    {
+      "epoch": 0.053210902825468025,
+      "grad_norm": 1.1290907859802246,
+      "learning_rate": 0.00015957706898974287,
+      "loss": 4.8283,
+      "step": 37870
+    },
+    {
+      "epoch": 0.05322495376363161,
+      "grad_norm": 1.1260758638381958,
+      "learning_rate": 0.0001596192215821273,
+      "loss": 4.8132,
+      "step": 37880
+    },
+    {
+      "epoch": 0.053239004701795184,
+      "grad_norm": 1.4716039896011353,
+      "learning_rate": 0.0001596613741745117,
+      "loss": 5.0606,
+      "step": 37890
+    },
+    {
+      "epoch": 0.05325305563995876,
+      "grad_norm": 1.08821439743042,
+      "learning_rate": 0.00015970352676689613,
+      "loss": 5.085,
+      "step": 37900
+    },
+    {
+      "epoch": 0.053267106578122336,
+      "grad_norm": 1.0185142755508423,
+      "learning_rate": 0.00015974567935928056,
+      "loss": 4.9683,
+      "step": 37910
+    },
+    {
+      "epoch": 0.05328115751628591,
+      "grad_norm": 1.0454940795898438,
+      "learning_rate": 0.00015978783195166502,
+      "loss": 5.0988,
+      "step": 37920
+    },
+    {
+      "epoch": 0.053295208454449496,
+      "grad_norm": 1.085282325744629,
+      "learning_rate": 0.00015982998454404945,
+      "loss": 4.873,
+      "step": 37930
+    },
+    {
+      "epoch": 0.05330925939261307,
+      "grad_norm": 1.1121746301651,
+      "learning_rate": 0.00015987213713643389,
+      "loss": 4.97,
+      "step": 37940
+    },
+    {
+      "epoch": 0.05332331033077665,
+      "grad_norm": 1.072363257408142,
+      "learning_rate": 0.00015991428972881832,
+      "loss": 4.9087,
+      "step": 37950
+    },
+    {
+      "epoch": 0.053337361268940224,
+      "grad_norm": 1.0807424783706665,
+      "learning_rate": 0.00015995644232120272,
+      "loss": 4.9707,
+      "step": 37960
+    },
+    {
+      "epoch": 0.0533514122071038,
+      "grad_norm": 1.1155600547790527,
+      "learning_rate": 0.00015999859491358715,
+      "loss": 4.9794,
+      "step": 37970
+    },
+    {
+      "epoch": 0.05336546314526738,
+      "grad_norm": 1.1688952445983887,
+      "learning_rate": 0.0001600407475059716,
+      "loss": 4.8757,
+      "step": 37980
+    },
+    {
+      "epoch": 0.05337951408343096,
+      "grad_norm": 1.0695239305496216,
+      "learning_rate": 0.00016008290009835604,
+      "loss": 4.9425,
+      "step": 37990
+    },
+    {
+      "epoch": 0.053393565021594536,
+      "grad_norm": 1.1101523637771606,
+      "learning_rate": 0.00016012505269074047,
+      "loss": 4.9533,
+      "step": 38000
+    },
+    {
+      "epoch": 0.05340761595975811,
+      "grad_norm": 1.0731979608535767,
+      "learning_rate": 0.0001601672052831249,
+      "loss": 4.9356,
+      "step": 38010
+    },
+    {
+      "epoch": 0.05342166689792169,
+      "grad_norm": 1.0717893838882446,
+      "learning_rate": 0.0001602093578755093,
+      "loss": 4.9014,
+      "step": 38020
+    },
+    {
+      "epoch": 0.05343571783608527,
+      "grad_norm": 1.1218794584274292,
+      "learning_rate": 0.00016025151046789374,
+      "loss": 5.0567,
+      "step": 38030
+    },
+    {
+      "epoch": 0.05344976877424885,
+      "grad_norm": 1.0864797830581665,
+      "learning_rate": 0.0001602936630602782,
+      "loss": 4.964,
+      "step": 38040
+    },
+    {
+      "epoch": 0.05346381971241242,
+      "grad_norm": 1.1077324151992798,
+      "learning_rate": 0.00016033581565266263,
+      "loss": 5.0164,
+      "step": 38050
+    },
+    {
+      "epoch": 0.053477870650576,
+      "grad_norm": 1.0500108003616333,
+      "learning_rate": 0.00016037796824504706,
+      "loss": 4.9392,
+      "step": 38060
+    },
+    {
+      "epoch": 0.053491921588739576,
+      "grad_norm": 1.052202820777893,
+      "learning_rate": 0.0001604201208374315,
+      "loss": 5.1295,
+      "step": 38070
+    },
+    {
+      "epoch": 0.05350597252690316,
+      "grad_norm": 1.0734517574310303,
+      "learning_rate": 0.00016046227342981592,
+      "loss": 4.9639,
+      "step": 38080
+    },
+    {
+      "epoch": 0.053520023465066735,
+      "grad_norm": 1.0562039613723755,
+      "learning_rate": 0.00016050442602220033,
+      "loss": 5.0704,
+      "step": 38090
+    },
+    {
+      "epoch": 0.05353407440323031,
+      "grad_norm": 1.0819730758666992,
+      "learning_rate": 0.00016054657861458478,
+      "loss": 4.9824,
+      "step": 38100
+    },
+    {
+      "epoch": 0.05354812534139389,
+      "grad_norm": 1.0542058944702148,
+      "learning_rate": 0.00016058873120696922,
+      "loss": 5.0282,
+      "step": 38110
+    },
+    {
+      "epoch": 0.053562176279557464,
+      "grad_norm": 1.0968252420425415,
+      "learning_rate": 0.00016063088379935365,
+      "loss": 4.9974,
+      "step": 38120
+    },
+    {
+      "epoch": 0.05357622721772105,
+      "grad_norm": 1.1440109014511108,
+      "learning_rate": 0.00016067303639173808,
+      "loss": 5.0201,
+      "step": 38130
+    },
+    {
+      "epoch": 0.05359027815588462,
+      "grad_norm": 1.1223822832107544,
+      "learning_rate": 0.00016071518898412254,
+      "loss": 4.885,
+      "step": 38140
+    },
+    {
+      "epoch": 0.0536043290940482,
+      "grad_norm": 1.0383020639419556,
+      "learning_rate": 0.0001607573415765069,
+      "loss": 5.0449,
+      "step": 38150
+    },
+    {
+      "epoch": 0.053618380032211775,
+      "grad_norm": 1.1103941202163696,
+      "learning_rate": 0.00016079949416889137,
+      "loss": 5.1299,
+      "step": 38160
+    },
+    {
+      "epoch": 0.05363243097037535,
+      "grad_norm": 1.1310505867004395,
+      "learning_rate": 0.0001608416467612758,
+      "loss": 4.8367,
+      "step": 38170
+    },
+    {
+      "epoch": 0.05364648190853893,
+      "grad_norm": 1.1387687921524048,
+      "learning_rate": 0.00016088379935366023,
+      "loss": 4.9226,
+      "step": 38180
+    },
+    {
+      "epoch": 0.05366053284670251,
+      "grad_norm": 1.082069993019104,
+      "learning_rate": 0.00016092595194604466,
+      "loss": 5.0986,
+      "step": 38190
+    },
+    {
+      "epoch": 0.05367458378486609,
+      "grad_norm": 1.1030524969100952,
+      "learning_rate": 0.00016096810453842912,
+      "loss": 5.0223,
+      "step": 38200
+    },
+    {
+      "epoch": 0.05368863472302966,
+      "grad_norm": 1.0807095766067505,
+      "learning_rate": 0.00016101025713081355,
+      "loss": 4.998,
+      "step": 38210
+    },
+    {
+      "epoch": 0.05370268566119324,
+      "grad_norm": 1.0664376020431519,
+      "learning_rate": 0.00016105240972319796,
+      "loss": 5.0521,
+      "step": 38220
+    },
+    {
+      "epoch": 0.053716736599356815,
+      "grad_norm": 1.1031697988510132,
+      "learning_rate": 0.0001610945623155824,
+      "loss": 4.9556,
+      "step": 38230
+    },
+    {
+      "epoch": 0.0537307875375204,
+      "grad_norm": 1.0721416473388672,
+      "learning_rate": 0.00016113671490796682,
+      "loss": 5.0203,
+      "step": 38240
+    },
+    {
+      "epoch": 0.053744838475683974,
+      "grad_norm": 1.0892760753631592,
+      "learning_rate": 0.00016117886750035125,
+      "loss": 4.9218,
+      "step": 38250
+    },
+    {
+      "epoch": 0.05375888941384755,
+      "grad_norm": 1.0493428707122803,
+      "learning_rate": 0.0001612210200927357,
+      "loss": 5.0262,
+      "step": 38260
+    },
+    {
+      "epoch": 0.05377294035201113,
+      "grad_norm": 1.0725332498550415,
+      "learning_rate": 0.00016126317268512014,
+      "loss": 5.0425,
+      "step": 38270
+    },
+    {
+      "epoch": 0.0537869912901747,
+      "grad_norm": 1.0133819580078125,
+      "learning_rate": 0.00016130532527750455,
+      "loss": 5.1757,
+      "step": 38280
+    },
+    {
+      "epoch": 0.053801042228338286,
+      "grad_norm": 1.0614526271820068,
+      "learning_rate": 0.00016134747786988898,
+      "loss": 4.8729,
+      "step": 38290
+    },
+    {
+      "epoch": 0.05381509316650186,
+      "grad_norm": 1.0696353912353516,
+      "learning_rate": 0.0001613896304622734,
+      "loss": 4.9481,
+      "step": 38300
+    },
+    {
+      "epoch": 0.05382914410466544,
+      "grad_norm": 1.1202436685562134,
+      "learning_rate": 0.00016143178305465784,
+      "loss": 4.9642,
+      "step": 38310
+    },
+    {
+      "epoch": 0.053843195042829015,
+      "grad_norm": 1.1082916259765625,
+      "learning_rate": 0.0001614739356470423,
+      "loss": 4.9846,
+      "step": 38320
+    },
+    {
+      "epoch": 0.05385724598099259,
+      "grad_norm": 1.0403755903244019,
+      "learning_rate": 0.00016151608823942673,
+      "loss": 5.0774,
+      "step": 38330
+    },
+    {
+      "epoch": 0.053871296919156174,
+      "grad_norm": 1.1000362634658813,
+      "learning_rate": 0.00016155824083181116,
+      "loss": 4.9838,
+      "step": 38340
+    },
+    {
+      "epoch": 0.05388534785731975,
+      "grad_norm": 1.0491472482681274,
+      "learning_rate": 0.00016160039342419556,
+      "loss": 4.971,
+      "step": 38350
+    },
+    {
+      "epoch": 0.053899398795483326,
+      "grad_norm": 1.093478798866272,
+      "learning_rate": 0.00016164254601658,
+      "loss": 4.9692,
+      "step": 38360
+    },
+    {
+      "epoch": 0.0539134497336469,
+      "grad_norm": 1.0910667181015015,
+      "learning_rate": 0.00016168469860896443,
+      "loss": 4.9924,
+      "step": 38370
+    },
+    {
+      "epoch": 0.05392750067181048,
+      "grad_norm": 1.0395450592041016,
+      "learning_rate": 0.00016172685120134888,
+      "loss": 5.0483,
+      "step": 38380
+    },
+    {
+      "epoch": 0.05394155160997406,
+      "grad_norm": 1.1088807582855225,
+      "learning_rate": 0.00016176900379373331,
+      "loss": 5.0126,
+      "step": 38390
+    },
+    {
+      "epoch": 0.05395560254813764,
+      "grad_norm": 1.096949815750122,
+      "learning_rate": 0.00016181115638611775,
+      "loss": 4.8924,
+      "step": 38400
+    },
+    {
+      "epoch": 0.053969653486301214,
+      "grad_norm": 1.0980435609817505,
+      "learning_rate": 0.00016185330897850215,
+      "loss": 4.9031,
+      "step": 38410
+    },
+    {
+      "epoch": 0.05398370442446479,
+      "grad_norm": 1.1065305471420288,
+      "learning_rate": 0.00016189546157088658,
+      "loss": 4.9263,
+      "step": 38420
+    },
+    {
+      "epoch": 0.053997755362628366,
+      "grad_norm": 1.0549814701080322,
+      "learning_rate": 0.000161937614163271,
+      "loss": 4.9421,
+      "step": 38430
+    },
+    {
+      "epoch": 0.05401180630079195,
+      "grad_norm": 1.0844722986221313,
+      "learning_rate": 0.00016197976675565547,
+      "loss": 5.0158,
+      "step": 38440
+    },
+    {
+      "epoch": 0.054025857238955526,
+      "grad_norm": 1.0721558332443237,
+      "learning_rate": 0.0001620219193480399,
+      "loss": 4.9792,
+      "step": 38450
+    },
+    {
+      "epoch": 0.0540399081771191,
+      "grad_norm": 1.0752381086349487,
+      "learning_rate": 0.00016206407194042433,
+      "loss": 4.834,
+      "step": 38460
+    },
+    {
+      "epoch": 0.05405395911528268,
+      "grad_norm": 1.0965216159820557,
+      "learning_rate": 0.00016210622453280874,
+      "loss": 4.9676,
+      "step": 38470
+    },
+    {
+      "epoch": 0.054068010053446254,
+      "grad_norm": 1.094175934791565,
+      "learning_rate": 0.00016214837712519317,
+      "loss": 4.9671,
+      "step": 38480
+    },
+    {
+      "epoch": 0.05408206099160983,
+      "grad_norm": 1.0462054014205933,
+      "learning_rate": 0.0001621905297175776,
+      "loss": 5.075,
+      "step": 38490
+    },
+    {
+      "epoch": 0.05409611192977341,
+      "grad_norm": 1.0655064582824707,
+      "learning_rate": 0.00016223268230996206,
+      "loss": 4.9576,
+      "step": 38500
+    },
+    {
+      "epoch": 0.05411016286793699,
+      "grad_norm": 1.075216293334961,
+      "learning_rate": 0.0001622748349023465,
+      "loss": 5.0349,
+      "step": 38510
+    },
+    {
+      "epoch": 0.054124213806100566,
+      "grad_norm": 1.0411499738693237,
+      "learning_rate": 0.00016231698749473092,
+      "loss": 5.0538,
+      "step": 38520
+    },
+    {
+      "epoch": 0.05413826474426414,
+      "grad_norm": 1.0397356748580933,
+      "learning_rate": 0.00016235914008711535,
+      "loss": 5.0188,
+      "step": 38530
+    },
+    {
+      "epoch": 0.05415231568242772,
+      "grad_norm": 1.0994830131530762,
+      "learning_rate": 0.00016240129267949976,
+      "loss": 4.9885,
+      "step": 38540
+    },
+    {
+      "epoch": 0.0541663666205913,
+      "grad_norm": 1.059415578842163,
+      "learning_rate": 0.0001624434452718842,
+      "loss": 4.868,
+      "step": 38550
+    },
+    {
+      "epoch": 0.05418041755875488,
+      "grad_norm": 1.1089651584625244,
+      "learning_rate": 0.00016248559786426864,
+      "loss": 4.8518,
+      "step": 38560
+    },
+    {
+      "epoch": 0.05419446849691845,
+      "grad_norm": 1.064829707145691,
+      "learning_rate": 0.00016252775045665308,
+      "loss": 4.9925,
+      "step": 38570
+    },
+    {
+      "epoch": 0.05420851943508203,
+      "grad_norm": 1.0721412897109985,
+      "learning_rate": 0.0001625699030490375,
+      "loss": 4.9596,
+      "step": 38580
+    },
+    {
+      "epoch": 0.054222570373245606,
+      "grad_norm": 1.1417285203933716,
+      "learning_rate": 0.00016261205564142194,
+      "loss": 4.8477,
+      "step": 38590
+    },
+    {
+      "epoch": 0.05423662131140919,
+      "grad_norm": 1.2103323936462402,
+      "learning_rate": 0.00016265420823380634,
+      "loss": 5.13,
+      "step": 38600
+    },
+    {
+      "epoch": 0.054250672249572765,
+      "grad_norm": 1.1253100633621216,
+      "learning_rate": 0.00016269636082619077,
+      "loss": 4.9663,
+      "step": 38610
+    },
+    {
+      "epoch": 0.05426472318773634,
+      "grad_norm": 1.0597470998764038,
+      "learning_rate": 0.00016273851341857523,
+      "loss": 4.9221,
+      "step": 38620
+    },
+    {
+      "epoch": 0.05427877412589992,
+      "grad_norm": 1.0937083959579468,
+      "learning_rate": 0.00016278066601095966,
+      "loss": 4.8933,
+      "step": 38630
+    },
+    {
+      "epoch": 0.054292825064063494,
+      "grad_norm": 1.0493961572647095,
+      "learning_rate": 0.0001628228186033441,
+      "loss": 5.0548,
+      "step": 38640
+    },
+    {
+      "epoch": 0.05430687600222708,
+      "grad_norm": 1.0712189674377441,
+      "learning_rate": 0.00016286497119572853,
+      "loss": 4.8846,
+      "step": 38650
+    },
+    {
+      "epoch": 0.05432092694039065,
+      "grad_norm": 1.1186167001724243,
+      "learning_rate": 0.00016290712378811296,
+      "loss": 5.0001,
+      "step": 38660
+    },
+    {
+      "epoch": 0.05433497787855423,
+      "grad_norm": 1.1650162935256958,
+      "learning_rate": 0.00016294927638049736,
+      "loss": 4.8981,
+      "step": 38670
+    },
+    {
+      "epoch": 0.054349028816717805,
+      "grad_norm": 1.0327341556549072,
+      "learning_rate": 0.00016299142897288182,
+      "loss": 5.0457,
+      "step": 38680
+    },
+    {
+      "epoch": 0.05436307975488138,
+      "grad_norm": 1.051758050918579,
+      "learning_rate": 0.00016303358156526625,
+      "loss": 4.9659,
+      "step": 38690
+    },
+    {
+      "epoch": 0.054377130693044964,
+      "grad_norm": 1.1347402334213257,
+      "learning_rate": 0.00016307573415765068,
+      "loss": 4.9157,
+      "step": 38700
+    },
+    {
+      "epoch": 0.05439118163120854,
+      "grad_norm": 1.0630327463150024,
+      "learning_rate": 0.0001631178867500351,
+      "loss": 5.0936,
+      "step": 38710
+    },
+    {
+      "epoch": 0.05440523256937212,
+      "grad_norm": 1.104245662689209,
+      "learning_rate": 0.00016316003934241957,
+      "loss": 5.0357,
+      "step": 38720
+    },
+    {
+      "epoch": 0.05441928350753569,
+      "grad_norm": 1.0906322002410889,
+      "learning_rate": 0.00016320219193480395,
+      "loss": 4.9834,
+      "step": 38730
+    },
+    {
+      "epoch": 0.05443333444569927,
+      "grad_norm": 1.0565413236618042,
+      "learning_rate": 0.0001632443445271884,
+      "loss": 4.9378,
+      "step": 38740
+    },
+    {
+      "epoch": 0.05444738538386285,
+      "grad_norm": 1.0998698472976685,
+      "learning_rate": 0.00016328649711957284,
+      "loss": 5.0077,
+      "step": 38750
+    },
+    {
+      "epoch": 0.05446143632202643,
+      "grad_norm": 1.1021455526351929,
+      "learning_rate": 0.00016332864971195727,
+      "loss": 4.9545,
+      "step": 38760
+    },
+    {
+      "epoch": 0.054475487260190004,
+      "grad_norm": 1.0735493898391724,
+      "learning_rate": 0.0001633708023043417,
+      "loss": 4.9852,
+      "step": 38770
+    },
+    {
+      "epoch": 0.05448953819835358,
+      "grad_norm": 1.0596339702606201,
+      "learning_rate": 0.00016341295489672616,
+      "loss": 4.919,
+      "step": 38780
+    },
+    {
+      "epoch": 0.05450358913651716,
+      "grad_norm": 1.1908012628555298,
+      "learning_rate": 0.0001634551074891106,
+      "loss": 4.8531,
+      "step": 38790
+    },
+    {
+      "epoch": 0.05451764007468073,
+      "grad_norm": 1.1102972030639648,
+      "learning_rate": 0.000163497260081495,
+      "loss": 4.8526,
+      "step": 38800
+    },
+    {
+      "epoch": 0.054531691012844316,
+      "grad_norm": 1.1045284271240234,
+      "learning_rate": 0.00016353941267387942,
+      "loss": 4.9537,
+      "step": 38810
+    },
+    {
+      "epoch": 0.05454574195100789,
+      "grad_norm": 1.0729451179504395,
+      "learning_rate": 0.00016358156526626386,
+      "loss": 4.9667,
+      "step": 38820
+    },
+    {
+      "epoch": 0.05455979288917147,
+      "grad_norm": 1.0459548234939575,
+      "learning_rate": 0.00016362371785864829,
+      "loss": 4.913,
+      "step": 38830
+    },
+    {
+      "epoch": 0.054573843827335045,
+      "grad_norm": 1.0615431070327759,
+      "learning_rate": 0.00016366587045103274,
+      "loss": 4.9522,
+      "step": 38840
+    },
+    {
+      "epoch": 0.05458789476549862,
+      "grad_norm": 1.1227502822875977,
+      "learning_rate": 0.00016370802304341718,
+      "loss": 4.9627,
+      "step": 38850
+    },
+    {
+      "epoch": 0.054601945703662204,
+      "grad_norm": 1.0452730655670166,
+      "learning_rate": 0.00016375017563580158,
+      "loss": 5.0616,
+      "step": 38860
+    },
+    {
+      "epoch": 0.05461599664182578,
+      "grad_norm": 0.998740017414093,
+      "learning_rate": 0.000163792328228186,
+      "loss": 4.9609,
+      "step": 38870
+    },
+    {
+      "epoch": 0.054630047579989356,
+      "grad_norm": 1.1312004327774048,
+      "learning_rate": 0.00016383448082057044,
+      "loss": 4.9495,
+      "step": 38880
+    },
+    {
+      "epoch": 0.05464409851815293,
+      "grad_norm": 1.0797289609909058,
+      "learning_rate": 0.00016387663341295487,
+      "loss": 4.946,
+      "step": 38890
+    },
+    {
+      "epoch": 0.05465814945631651,
+      "grad_norm": 1.0702334642410278,
+      "learning_rate": 0.00016391878600533933,
+      "loss": 5.0044,
+      "step": 38900
+    },
+    {
+      "epoch": 0.05467220039448009,
+      "grad_norm": 1.0228101015090942,
+      "learning_rate": 0.00016396093859772376,
+      "loss": 4.8987,
+      "step": 38910
+    },
+    {
+      "epoch": 0.05468625133264367,
+      "grad_norm": 1.0430359840393066,
+      "learning_rate": 0.0001640030911901082,
+      "loss": 5.0504,
+      "step": 38920
+    },
+    {
+      "epoch": 0.054700302270807244,
+      "grad_norm": 1.065110206604004,
+      "learning_rate": 0.0001640452437824926,
+      "loss": 4.9139,
+      "step": 38930
+    },
+    {
+      "epoch": 0.05471435320897082,
+      "grad_norm": 1.1111505031585693,
+      "learning_rate": 0.00016408739637487703,
+      "loss": 4.9611,
+      "step": 38940
+    },
+    {
+      "epoch": 0.054728404147134396,
+      "grad_norm": 1.0788532495498657,
+      "learning_rate": 0.00016412954896726146,
+      "loss": 4.8959,
+      "step": 38950
+    },
+    {
+      "epoch": 0.05474245508529798,
+      "grad_norm": 1.0466620922088623,
+      "learning_rate": 0.00016417170155964592,
+      "loss": 5.1461,
+      "step": 38960
+    },
+    {
+      "epoch": 0.054756506023461556,
+      "grad_norm": 1.115582823753357,
+      "learning_rate": 0.00016421385415203035,
+      "loss": 4.9678,
+      "step": 38970
+    },
+    {
+      "epoch": 0.05477055696162513,
+      "grad_norm": 1.0731269121170044,
+      "learning_rate": 0.00016425600674441478,
+      "loss": 4.9063,
+      "step": 38980
+    },
+    {
+      "epoch": 0.05478460789978871,
+      "grad_norm": 1.081316351890564,
+      "learning_rate": 0.00016429815933679918,
+      "loss": 4.9726,
+      "step": 38990
+    },
+    {
+      "epoch": 0.054798658837952284,
+      "grad_norm": 1.0765340328216553,
+      "learning_rate": 0.00016434031192918362,
+      "loss": 4.8929,
+      "step": 39000
+    },
+    {
+      "epoch": 0.05481270977611587,
+      "grad_norm": 1.0560728311538696,
+      "learning_rate": 0.00016438246452156805,
+      "loss": 4.9395,
+      "step": 39010
+    },
+    {
+      "epoch": 0.05482676071427944,
+      "grad_norm": 1.069467306137085,
+      "learning_rate": 0.0001644246171139525,
+      "loss": 5.0583,
+      "step": 39020
+    },
+    {
+      "epoch": 0.05484081165244302,
+      "grad_norm": 1.0423463582992554,
+      "learning_rate": 0.00016446676970633694,
+      "loss": 4.993,
+      "step": 39030
+    },
+    {
+      "epoch": 0.054854862590606596,
+      "grad_norm": 1.064470648765564,
+      "learning_rate": 0.00016450892229872137,
+      "loss": 5.0129,
+      "step": 39040
+    },
+    {
+      "epoch": 0.05486891352877017,
+      "grad_norm": 1.0736603736877441,
+      "learning_rate": 0.0001645510748911058,
+      "loss": 5.0073,
+      "step": 39050
+    },
+    {
+      "epoch": 0.054882964466933755,
+      "grad_norm": 1.045296549797058,
+      "learning_rate": 0.0001645932274834902,
+      "loss": 4.9443,
+      "step": 39060
+    },
+    {
+      "epoch": 0.05489701540509733,
+      "grad_norm": 1.1380881071090698,
+      "learning_rate": 0.00016463538007587463,
+      "loss": 4.9244,
+      "step": 39070
+    },
+    {
+      "epoch": 0.05491106634326091,
+      "grad_norm": 1.0342426300048828,
+      "learning_rate": 0.0001646775326682591,
+      "loss": 4.9206,
+      "step": 39080
+    },
+    {
+      "epoch": 0.05492511728142448,
+      "grad_norm": 1.0200796127319336,
+      "learning_rate": 0.00016471968526064352,
+      "loss": 5.0105,
+      "step": 39090
+    },
+    {
+      "epoch": 0.05493916821958806,
+      "grad_norm": 1.0826342105865479,
+      "learning_rate": 0.00016476183785302795,
+      "loss": 5.0702,
+      "step": 39100
+    },
+    {
+      "epoch": 0.054953219157751636,
+      "grad_norm": 1.0333572626113892,
+      "learning_rate": 0.00016480399044541239,
+      "loss": 4.8559,
+      "step": 39110
+    },
+    {
+      "epoch": 0.05496727009591522,
+      "grad_norm": 1.1156673431396484,
+      "learning_rate": 0.0001648461430377968,
+      "loss": 5.0019,
+      "step": 39120
+    },
+    {
+      "epoch": 0.054981321034078795,
+      "grad_norm": 1.07420814037323,
+      "learning_rate": 0.00016488829563018122,
+      "loss": 5.0687,
+      "step": 39130
+    },
+    {
+      "epoch": 0.05499537197224237,
+      "grad_norm": 1.0014334917068481,
+      "learning_rate": 0.00016493044822256568,
+      "loss": 5.0476,
+      "step": 39140
+    },
+    {
+      "epoch": 0.05500942291040595,
+      "grad_norm": 1.0669128894805908,
+      "learning_rate": 0.0001649726008149501,
+      "loss": 4.8751,
+      "step": 39150
+    },
+    {
+      "epoch": 0.055023473848569524,
+      "grad_norm": 1.0790259838104248,
+      "learning_rate": 0.00016501475340733454,
+      "loss": 5.0306,
+      "step": 39160
+    },
+    {
+      "epoch": 0.05503752478673311,
+      "grad_norm": 1.0873439311981201,
+      "learning_rate": 0.00016505690599971897,
+      "loss": 4.8936,
+      "step": 39170
+    },
+    {
+      "epoch": 0.05505157572489668,
+      "grad_norm": 1.0448416471481323,
+      "learning_rate": 0.00016509905859210338,
+      "loss": 4.8749,
+      "step": 39180
+    },
+    {
+      "epoch": 0.05506562666306026,
+      "grad_norm": 1.0575209856033325,
+      "learning_rate": 0.0001651412111844878,
+      "loss": 4.9797,
+      "step": 39190
+    },
+    {
+      "epoch": 0.055079677601223835,
+      "grad_norm": 1.1213315725326538,
+      "learning_rate": 0.00016518336377687227,
+      "loss": 4.9521,
+      "step": 39200
+    },
+    {
+      "epoch": 0.05509372853938741,
+      "grad_norm": 1.0260167121887207,
+      "learning_rate": 0.0001652255163692567,
+      "loss": 5.0328,
+      "step": 39210
+    },
+    {
+      "epoch": 0.055107779477550994,
+      "grad_norm": 1.0591065883636475,
+      "learning_rate": 0.00016526766896164113,
+      "loss": 4.9218,
+      "step": 39220
+    },
+    {
+      "epoch": 0.05512183041571457,
+      "grad_norm": 1.0388381481170654,
+      "learning_rate": 0.00016530982155402556,
+      "loss": 4.9386,
+      "step": 39230
+    },
+    {
+      "epoch": 0.05513588135387815,
+      "grad_norm": 1.0415207147598267,
+      "learning_rate": 0.00016535197414641,
+      "loss": 4.8885,
+      "step": 39240
+    },
+    {
+      "epoch": 0.05514993229204172,
+      "grad_norm": 1.0508288145065308,
+      "learning_rate": 0.0001653941267387944,
+      "loss": 5.0216,
+      "step": 39250
+    },
+    {
+      "epoch": 0.0551639832302053,
+      "grad_norm": 1.0683437585830688,
+      "learning_rate": 0.00016543627933117885,
+      "loss": 4.9671,
+      "step": 39260
+    },
+    {
+      "epoch": 0.05517803416836888,
+      "grad_norm": 1.0968106985092163,
+      "learning_rate": 0.00016547843192356328,
+      "loss": 4.9947,
+      "step": 39270
+    },
+    {
+      "epoch": 0.05519208510653246,
+      "grad_norm": 1.0676803588867188,
+      "learning_rate": 0.00016552058451594772,
+      "loss": 4.973,
+      "step": 39280
+    },
+    {
+      "epoch": 0.055206136044696034,
+      "grad_norm": 1.1953974962234497,
+      "learning_rate": 0.00016556273710833215,
+      "loss": 5.0037,
+      "step": 39290
+    },
+    {
+      "epoch": 0.05522018698285961,
+      "grad_norm": 1.0898711681365967,
+      "learning_rate": 0.0001656048897007166,
+      "loss": 4.8778,
+      "step": 39300
+    },
+    {
+      "epoch": 0.05523423792102319,
+      "grad_norm": 1.0796436071395874,
+      "learning_rate": 0.00016564282703386256,
+      "loss": 5.0486,
+      "step": 39310
+    },
+    {
+      "epoch": 0.05524828885918677,
+      "grad_norm": 1.0733755826950073,
+      "learning_rate": 0.000165684979626247,
+      "loss": 4.9972,
+      "step": 39320
+    },
+    {
+      "epoch": 0.055262339797350346,
+      "grad_norm": 1.0305414199829102,
+      "learning_rate": 0.00016572713221863145,
+      "loss": 5.0652,
+      "step": 39330
+    },
+    {
+      "epoch": 0.05527639073551392,
+      "grad_norm": 1.0890687704086304,
+      "learning_rate": 0.00016576928481101588,
+      "loss": 5.0277,
+      "step": 39340
+    },
+    {
+      "epoch": 0.0552904416736775,
+      "grad_norm": 1.1217213869094849,
+      "learning_rate": 0.00016581143740340029,
+      "loss": 4.8927,
+      "step": 39350
+    },
+    {
+      "epoch": 0.055304492611841075,
+      "grad_norm": 1.1358542442321777,
+      "learning_rate": 0.00016585358999578472,
+      "loss": 4.9728,
+      "step": 39360
+    },
+    {
+      "epoch": 0.05531854355000466,
+      "grad_norm": 1.036201000213623,
+      "learning_rate": 0.00016589574258816915,
+      "loss": 4.9836,
+      "step": 39370
+    },
+    {
+      "epoch": 0.055332594488168234,
+      "grad_norm": 1.0400617122650146,
+      "learning_rate": 0.0001659378951805536,
+      "loss": 4.9628,
+      "step": 39380
+    },
+    {
+      "epoch": 0.05534664542633181,
+      "grad_norm": 1.030198097229004,
+      "learning_rate": 0.00016598004777293804,
+      "loss": 4.8906,
+      "step": 39390
+    },
+    {
+      "epoch": 0.055360696364495386,
+      "grad_norm": 1.0506401062011719,
+      "learning_rate": 0.00016602220036532247,
+      "loss": 4.8948,
+      "step": 39400
+    },
+    {
+      "epoch": 0.05537474730265896,
+      "grad_norm": 1.0474390983581543,
+      "learning_rate": 0.00016606435295770687,
+      "loss": 4.9137,
+      "step": 39410
+    },
+    {
+      "epoch": 0.05538879824082254,
+      "grad_norm": 1.0166150331497192,
+      "learning_rate": 0.0001661065055500913,
+      "loss": 5.0177,
+      "step": 39420
+    },
+    {
+      "epoch": 0.05540284917898612,
+      "grad_norm": 1.0387684106826782,
+      "learning_rate": 0.00016614865814247573,
+      "loss": 4.9143,
+      "step": 39430
+    },
+    {
+      "epoch": 0.0554169001171497,
+      "grad_norm": 1.0397957563400269,
+      "learning_rate": 0.0001661908107348602,
+      "loss": 4.9583,
+      "step": 39440
+    },
+    {
+      "epoch": 0.055430951055313274,
+      "grad_norm": 1.0370898246765137,
+      "learning_rate": 0.00016623296332724462,
+      "loss": 4.9651,
+      "step": 39450
+    },
+    {
+      "epoch": 0.05544500199347685,
+      "grad_norm": 1.0566965341567993,
+      "learning_rate": 0.00016627511591962906,
+      "loss": 5.0027,
+      "step": 39460
+    },
+    {
+      "epoch": 0.055459052931640426,
+      "grad_norm": 1.054910659790039,
+      "learning_rate": 0.0001663172685120135,
+      "loss": 5.0484,
+      "step": 39470
+    },
+    {
+      "epoch": 0.05547310386980401,
+      "grad_norm": 1.0622105598449707,
+      "learning_rate": 0.0001663594211043979,
+      "loss": 4.9111,
+      "step": 39480
+    },
+    {
+      "epoch": 0.055487154807967586,
+      "grad_norm": 1.0327736139297485,
+      "learning_rate": 0.00016640157369678232,
+      "loss": 5.0337,
+      "step": 39490
+    },
+    {
+      "epoch": 0.05550120574613116,
+      "grad_norm": 1.0651620626449585,
+      "learning_rate": 0.00016644372628916678,
+      "loss": 4.9398,
+      "step": 39500
+    },
+    {
+      "epoch": 0.05551525668429474,
+      "grad_norm": 1.0605753660202026,
+      "learning_rate": 0.0001664858788815512,
+      "loss": 5.0069,
+      "step": 39510
+    },
+    {
+      "epoch": 0.055529307622458314,
+      "grad_norm": 1.0352438688278198,
+      "learning_rate": 0.00016652803147393564,
+      "loss": 4.9148,
+      "step": 39520
+    },
+    {
+      "epoch": 0.0555433585606219,
+      "grad_norm": 1.1176104545593262,
+      "learning_rate": 0.00016657018406632007,
+      "loss": 4.9403,
+      "step": 39530
+    },
+    {
+      "epoch": 0.05555740949878547,
+      "grad_norm": 1.0405285358428955,
+      "learning_rate": 0.00016661233665870448,
+      "loss": 4.8954,
+      "step": 39540
+    },
+    {
+      "epoch": 0.05557146043694905,
+      "grad_norm": 1.0079467296600342,
+      "learning_rate": 0.0001666544892510889,
+      "loss": 5.0099,
+      "step": 39550
+    },
+    {
+      "epoch": 0.055585511375112626,
+      "grad_norm": 1.0810550451278687,
+      "learning_rate": 0.00016669664184347337,
+      "loss": 4.9254,
+      "step": 39560
+    },
+    {
+      "epoch": 0.0555995623132762,
+      "grad_norm": 1.0852543115615845,
+      "learning_rate": 0.0001667387944358578,
+      "loss": 4.9133,
+      "step": 39570
+    },
+    {
+      "epoch": 0.055613613251439785,
+      "grad_norm": 1.074148178100586,
+      "learning_rate": 0.00016678094702824223,
+      "loss": 4.8821,
+      "step": 39580
+    },
+    {
+      "epoch": 0.05562766418960336,
+      "grad_norm": 1.2053712606430054,
+      "learning_rate": 0.00016682309962062666,
+      "loss": 4.8644,
+      "step": 39590
+    },
+    {
+      "epoch": 0.05564171512776694,
+      "grad_norm": 1.3233577013015747,
+      "learning_rate": 0.00016686525221301106,
+      "loss": 4.9788,
+      "step": 39600
+    },
+    {
+      "epoch": 0.05565576606593051,
+      "grad_norm": 1.0858327150344849,
+      "learning_rate": 0.0001669074048053955,
+      "loss": 4.9615,
+      "step": 39610
+    },
+    {
+      "epoch": 0.05566981700409409,
+      "grad_norm": 1.0294164419174194,
+      "learning_rate": 0.00016694955739777995,
+      "loss": 4.8376,
+      "step": 39620
+    },
+    {
+      "epoch": 0.05568386794225767,
+      "grad_norm": 1.046635627746582,
+      "learning_rate": 0.00016699170999016439,
+      "loss": 4.9348,
+      "step": 39630
+    },
+    {
+      "epoch": 0.05569791888042125,
+      "grad_norm": 1.0553175210952759,
+      "learning_rate": 0.00016703386258254882,
+      "loss": 5.0127,
+      "step": 39640
+    },
+    {
+      "epoch": 0.055711969818584825,
+      "grad_norm": 1.0525593757629395,
+      "learning_rate": 0.00016707601517493325,
+      "loss": 4.9244,
+      "step": 39650
+    },
+    {
+      "epoch": 0.0557260207567484,
+      "grad_norm": 1.049413800239563,
+      "learning_rate": 0.00016711816776731768,
+      "loss": 4.9705,
+      "step": 39660
+    },
+    {
+      "epoch": 0.05574007169491198,
+      "grad_norm": 1.0601624250411987,
+      "learning_rate": 0.00016716032035970208,
+      "loss": 4.9637,
+      "step": 39670
+    },
+    {
+      "epoch": 0.05575412263307556,
+      "grad_norm": 1.0123969316482544,
+      "learning_rate": 0.00016720247295208654,
+      "loss": 4.9764,
+      "step": 39680
+    },
+    {
+      "epoch": 0.05576817357123914,
+      "grad_norm": 1.0407583713531494,
+      "learning_rate": 0.00016724462554447097,
+      "loss": 4.9696,
+      "step": 39690
+    },
+    {
+      "epoch": 0.05578222450940271,
+      "grad_norm": 1.0655407905578613,
+      "learning_rate": 0.0001672867781368554,
+      "loss": 4.955,
+      "step": 39700
+    },
+    {
+      "epoch": 0.05579627544756629,
+      "grad_norm": 1.0213630199432373,
+      "learning_rate": 0.00016732893072923983,
+      "loss": 4.8745,
+      "step": 39710
+    },
+    {
+      "epoch": 0.055810326385729865,
+      "grad_norm": 1.0620572566986084,
+      "learning_rate": 0.00016737108332162427,
+      "loss": 4.925,
+      "step": 39720
+    },
+    {
+      "epoch": 0.05582437732389344,
+      "grad_norm": 1.0573869943618774,
+      "learning_rate": 0.00016741323591400867,
+      "loss": 4.9061,
+      "step": 39730
+    },
+    {
+      "epoch": 0.055838428262057024,
+      "grad_norm": 1.0326967239379883,
+      "learning_rate": 0.00016745538850639313,
+      "loss": 5.0449,
+      "step": 39740
+    },
+    {
+      "epoch": 0.0558524792002206,
+      "grad_norm": 1.082334280014038,
+      "learning_rate": 0.00016749754109877756,
+      "loss": 4.9577,
+      "step": 39750
+    },
+    {
+      "epoch": 0.05586653013838418,
+      "grad_norm": 1.0817333459854126,
+      "learning_rate": 0.000167539693691162,
+      "loss": 4.9172,
+      "step": 39760
+    },
+    {
+      "epoch": 0.05588058107654775,
+      "grad_norm": 1.0093368291854858,
+      "learning_rate": 0.00016758184628354642,
+      "loss": 4.8782,
+      "step": 39770
+    },
+    {
+      "epoch": 0.05589463201471133,
+      "grad_norm": 1.0431386232376099,
+      "learning_rate": 0.00016762399887593085,
+      "loss": 4.9162,
+      "step": 39780
+    },
+    {
+      "epoch": 0.05590868295287491,
+      "grad_norm": 1.052478313446045,
+      "learning_rate": 0.0001676661514683153,
+      "loss": 5.017,
+      "step": 39790
+    },
+    {
+      "epoch": 0.05592273389103849,
+      "grad_norm": 1.0899916887283325,
+      "learning_rate": 0.00016770830406069972,
+      "loss": 4.9324,
+      "step": 39800
+    },
+    {
+      "epoch": 0.055936784829202064,
+      "grad_norm": 1.0490303039550781,
+      "learning_rate": 0.00016775045665308415,
+      "loss": 4.9905,
+      "step": 39810
+    },
+    {
+      "epoch": 0.05595083576736564,
+      "grad_norm": 1.129417896270752,
+      "learning_rate": 0.00016779260924546858,
+      "loss": 4.8127,
+      "step": 39820
+    },
+    {
+      "epoch": 0.05596488670552922,
+      "grad_norm": 1.0525208711624146,
+      "learning_rate": 0.000167834761837853,
+      "loss": 4.9774,
+      "step": 39830
+    },
+    {
+      "epoch": 0.0559789376436928,
+      "grad_norm": 1.0649760961532593,
+      "learning_rate": 0.00016787691443023744,
+      "loss": 4.8925,
+      "step": 39840
+    },
+    {
+      "epoch": 0.055992988581856376,
+      "grad_norm": 1.0432103872299194,
+      "learning_rate": 0.0001679190670226219,
+      "loss": 5.1254,
+      "step": 39850
+    },
+    {
+      "epoch": 0.05600703952001995,
+      "grad_norm": 1.046905517578125,
+      "learning_rate": 0.0001679612196150063,
+      "loss": 5.0319,
+      "step": 39860
+    },
+    {
+      "epoch": 0.05602109045818353,
+      "grad_norm": 0.994096040725708,
+      "learning_rate": 0.00016800337220739073,
+      "loss": 5.0441,
+      "step": 39870
+    },
+    {
+      "epoch": 0.056035141396347105,
+      "grad_norm": 1.0326930284500122,
+      "learning_rate": 0.00016804552479977516,
+      "loss": 4.9293,
+      "step": 39880
+    },
+    {
+      "epoch": 0.05604919233451069,
+      "grad_norm": 1.0579875707626343,
+      "learning_rate": 0.0001680876773921596,
+      "loss": 4.9475,
+      "step": 39890
+    },
+    {
+      "epoch": 0.056063243272674264,
+      "grad_norm": 1.1440166234970093,
+      "learning_rate": 0.00016812982998454403,
+      "loss": 4.8159,
+      "step": 39900
+    },
+    {
+      "epoch": 0.05607729421083784,
+      "grad_norm": 1.0132042169570923,
+      "learning_rate": 0.00016817198257692849,
+      "loss": 5.0357,
+      "step": 39910
+    },
+    {
+      "epoch": 0.056091345149001416,
+      "grad_norm": 1.0793256759643555,
+      "learning_rate": 0.00016821413516931292,
+      "loss": 4.8232,
+      "step": 39920
+    },
+    {
+      "epoch": 0.05610539608716499,
+      "grad_norm": 1.0349791049957275,
+      "learning_rate": 0.00016825628776169732,
+      "loss": 5.0093,
+      "step": 39930
+    },
+    {
+      "epoch": 0.056119447025328575,
+      "grad_norm": 1.054980754852295,
+      "learning_rate": 0.00016829844035408175,
+      "loss": 4.905,
+      "step": 39940
+    },
+    {
+      "epoch": 0.05613349796349215,
+      "grad_norm": 1.1767756938934326,
+      "learning_rate": 0.00016834059294646618,
+      "loss": 4.9174,
+      "step": 39950
+    },
+    {
+      "epoch": 0.05614754890165573,
+      "grad_norm": 1.0445729494094849,
+      "learning_rate": 0.00016838274553885064,
+      "loss": 4.9582,
+      "step": 39960
+    },
+    {
+      "epoch": 0.056161599839819304,
+      "grad_norm": 1.0728813409805298,
+      "learning_rate": 0.00016842489813123507,
+      "loss": 4.8951,
+      "step": 39970
+    },
+    {
+      "epoch": 0.05617565077798288,
+      "grad_norm": 1.0347257852554321,
+      "learning_rate": 0.0001684670507236195,
+      "loss": 4.9209,
+      "step": 39980
+    },
+    {
+      "epoch": 0.05618970171614646,
+      "grad_norm": 1.0311264991760254,
+      "learning_rate": 0.0001685092033160039,
+      "loss": 4.9572,
+      "step": 39990
+    },
+    {
+      "epoch": 0.05620375265431004,
+      "grad_norm": 1.029685378074646,
+      "learning_rate": 0.00016855135590838834,
+      "loss": 4.8857,
+      "step": 40000
+    },
+    {
+      "epoch": 0.056217803592473616,
+      "grad_norm": 1.0317046642303467,
+      "learning_rate": 0.00016859350850077277,
+      "loss": 4.9465,
+      "step": 40010
+    },
+    {
+      "epoch": 0.05623185453063719,
+      "grad_norm": 1.0499831438064575,
+      "learning_rate": 0.00016863566109315723,
+      "loss": 4.914,
+      "step": 40020
+    },
+    {
+      "epoch": 0.05624590546880077,
+      "grad_norm": 1.0336376428604126,
+      "learning_rate": 0.00016867781368554166,
+      "loss": 5.013,
+      "step": 40030
+    },
+    {
+      "epoch": 0.05625995640696435,
+      "grad_norm": 1.024095892906189,
+      "learning_rate": 0.0001687199662779261,
+      "loss": 4.8756,
+      "step": 40040
+    },
+    {
+      "epoch": 0.05627400734512793,
+      "grad_norm": 1.0317274332046509,
+      "learning_rate": 0.00016876211887031052,
+      "loss": 4.9187,
+      "step": 40050
+    },
+    {
+      "epoch": 0.0562880582832915,
+      "grad_norm": 1.834438681602478,
+      "learning_rate": 0.00016880427146269493,
+      "loss": 4.8996,
+      "step": 40060
+    },
+    {
+      "epoch": 0.05630210922145508,
+      "grad_norm": 1.0885308980941772,
+      "learning_rate": 0.00016884642405507936,
+      "loss": 5.0043,
+      "step": 40070
+    },
+    {
+      "epoch": 0.056316160159618656,
+      "grad_norm": 1.1309927701950073,
+      "learning_rate": 0.00016888857664746381,
+      "loss": 4.9811,
+      "step": 40080
+    },
+    {
+      "epoch": 0.05633021109778223,
+      "grad_norm": 1.034248948097229,
+      "learning_rate": 0.00016893072923984825,
+      "loss": 4.9089,
+      "step": 40090
+    },
+    {
+      "epoch": 0.056344262035945815,
+      "grad_norm": 1.0103942155838013,
+      "learning_rate": 0.00016897288183223268,
+      "loss": 4.8929,
+      "step": 40100
+    },
+    {
+      "epoch": 0.05635831297410939,
+      "grad_norm": 1.0718656778335571,
+      "learning_rate": 0.0001690150344246171,
+      "loss": 4.9839,
+      "step": 40110
+    },
+    {
+      "epoch": 0.05637236391227297,
+      "grad_norm": 1.044500470161438,
+      "learning_rate": 0.0001690571870170015,
+      "loss": 4.8699,
+      "step": 40120
+    },
+    {
+      "epoch": 0.05638641485043654,
+      "grad_norm": 1.024745225906372,
+      "learning_rate": 0.00016909933960938594,
+      "loss": 5.0166,
+      "step": 40130
+    },
+    {
+      "epoch": 0.05640046578860012,
+      "grad_norm": 1.0210412740707397,
+      "learning_rate": 0.0001691414922017704,
+      "loss": 4.958,
+      "step": 40140
+    },
+    {
+      "epoch": 0.0564145167267637,
+      "grad_norm": 1.0404855012893677,
+      "learning_rate": 0.00016918364479415483,
+      "loss": 4.9962,
+      "step": 40150
+    },
+    {
+      "epoch": 0.05642856766492728,
+      "grad_norm": 1.0807390213012695,
+      "learning_rate": 0.00016922579738653926,
+      "loss": 4.9206,
+      "step": 40160
+    },
+    {
+      "epoch": 0.056442618603090855,
+      "grad_norm": 1.0216072797775269,
+      "learning_rate": 0.0001692679499789237,
+      "loss": 4.9689,
+      "step": 40170
+    },
+    {
+      "epoch": 0.05645666954125443,
+      "grad_norm": 1.0541982650756836,
+      "learning_rate": 0.00016931010257130813,
+      "loss": 4.9347,
+      "step": 40180
+    },
+    {
+      "epoch": 0.05647072047941801,
+      "grad_norm": 1.072575569152832,
+      "learning_rate": 0.00016935225516369253,
+      "loss": 5.0225,
+      "step": 40190
+    },
+    {
+      "epoch": 0.05648477141758159,
+      "grad_norm": 1.0966130495071411,
+      "learning_rate": 0.000169394407756077,
+      "loss": 4.9683,
+      "step": 40200
+    },
+    {
+      "epoch": 0.05649882235574517,
+      "grad_norm": 1.0755553245544434,
+      "learning_rate": 0.00016943656034846142,
+      "loss": 4.8777,
+      "step": 40210
+    },
+    {
+      "epoch": 0.05651287329390874,
+      "grad_norm": 1.0782699584960938,
+      "learning_rate": 0.00016947871294084585,
+      "loss": 4.9303,
+      "step": 40220
+    },
+    {
+      "epoch": 0.05652692423207232,
+      "grad_norm": 1.0849295854568481,
+      "learning_rate": 0.00016952086553323028,
+      "loss": 4.9465,
+      "step": 40230
+    },
+    {
+      "epoch": 0.056540975170235895,
+      "grad_norm": 1.0700656175613403,
+      "learning_rate": 0.0001695630181256147,
+      "loss": 4.9312,
+      "step": 40240
+    },
+    {
+      "epoch": 0.05655502610839948,
+      "grad_norm": 1.053043246269226,
+      "learning_rate": 0.00016960517071799912,
+      "loss": 4.9405,
+      "step": 40250
+    },
+    {
+      "epoch": 0.056569077046563054,
+      "grad_norm": 1.0179575681686401,
+      "learning_rate": 0.00016964732331038358,
+      "loss": 4.91,
+      "step": 40260
+    },
+    {
+      "epoch": 0.05658312798472663,
+      "grad_norm": 1.0473299026489258,
+      "learning_rate": 0.000169689475902768,
+      "loss": 5.0573,
+      "step": 40270
+    },
+    {
+      "epoch": 0.05659717892289021,
+      "grad_norm": 1.0085793733596802,
+      "learning_rate": 0.00016973162849515244,
+      "loss": 4.9596,
+      "step": 40280
+    },
+    {
+      "epoch": 0.05661122986105378,
+      "grad_norm": 1.210898518562317,
+      "learning_rate": 0.00016977378108753687,
+      "loss": 4.9208,
+      "step": 40290
+    },
+    {
+      "epoch": 0.056625280799217366,
+      "grad_norm": 1.0420124530792236,
+      "learning_rate": 0.0001698159336799213,
+      "loss": 4.8712,
+      "step": 40300
+    },
+    {
+      "epoch": 0.05663933173738094,
+      "grad_norm": 1.0410534143447876,
+      "learning_rate": 0.0001698580862723057,
+      "loss": 4.8958,
+      "step": 40310
+    },
+    {
+      "epoch": 0.05665338267554452,
+      "grad_norm": 1.0174304246902466,
+      "learning_rate": 0.00016990023886469016,
+      "loss": 5.0153,
+      "step": 40320
+    },
+    {
+      "epoch": 0.056667433613708094,
+      "grad_norm": 1.1811450719833374,
+      "learning_rate": 0.0001699423914570746,
+      "loss": 4.8823,
+      "step": 40330
+    },
+    {
+      "epoch": 0.05668148455187167,
+      "grad_norm": 1.0314950942993164,
+      "learning_rate": 0.00016998454404945903,
+      "loss": 5.0111,
+      "step": 40340
+    },
+    {
+      "epoch": 0.056695535490035254,
+      "grad_norm": 1.0767461061477661,
+      "learning_rate": 0.00017002669664184346,
+      "loss": 5.0005,
+      "step": 40350
+    },
+    {
+      "epoch": 0.05670958642819883,
+      "grad_norm": 1.0583289861679077,
+      "learning_rate": 0.0001700688492342279,
+      "loss": 4.9518,
+      "step": 40360
+    },
+    {
+      "epoch": 0.056723637366362406,
+      "grad_norm": 1.068487286567688,
+      "learning_rate": 0.00017011100182661235,
+      "loss": 5.0675,
+      "step": 40370
+    },
+    {
+      "epoch": 0.05673768830452598,
+      "grad_norm": 1.0626734495162964,
+      "learning_rate": 0.00017015315441899675,
+      "loss": 4.979,
+      "step": 40380
+    },
+    {
+      "epoch": 0.05675173924268956,
+      "grad_norm": 1.0374014377593994,
+      "learning_rate": 0.00017019530701138118,
+      "loss": 4.7829,
+      "step": 40390
+    },
+    {
+      "epoch": 0.056765790180853135,
+      "grad_norm": 1.0241491794586182,
+      "learning_rate": 0.0001702374596037656,
+      "loss": 4.9447,
+      "step": 40400
+    },
+    {
+      "epoch": 0.05677984111901672,
+      "grad_norm": 1.0468249320983887,
+      "learning_rate": 0.00017027961219615004,
+      "loss": 4.8091,
+      "step": 40410
+    },
+    {
+      "epoch": 0.056793892057180294,
+      "grad_norm": 1.0289359092712402,
+      "learning_rate": 0.00017032176478853447,
+      "loss": 4.9062,
+      "step": 40420
+    },
+    {
+      "epoch": 0.05680794299534387,
+      "grad_norm": 1.02260422706604,
+      "learning_rate": 0.00017036391738091893,
+      "loss": 4.9701,
+      "step": 40430
+    },
+    {
+      "epoch": 0.056821993933507446,
+      "grad_norm": 1.0522793531417847,
+      "learning_rate": 0.00017040606997330334,
+      "loss": 4.9831,
+      "step": 40440
+    },
+    {
+      "epoch": 0.05683604487167102,
+      "grad_norm": 1.0370879173278809,
+      "learning_rate": 0.00017044822256568777,
+      "loss": 4.8978,
+      "step": 40450
+    },
+    {
+      "epoch": 0.056850095809834605,
+      "grad_norm": 1.0785061120986938,
+      "learning_rate": 0.0001704903751580722,
+      "loss": 4.914,
+      "step": 40460
+    },
+    {
+      "epoch": 0.05686414674799818,
+      "grad_norm": 1.0285074710845947,
+      "learning_rate": 0.00017053252775045663,
+      "loss": 5.0287,
+      "step": 40470
+    },
+    {
+      "epoch": 0.05687819768616176,
+      "grad_norm": 1.0587258338928223,
+      "learning_rate": 0.00017057468034284106,
+      "loss": 4.8259,
+      "step": 40480
+    },
+    {
+      "epoch": 0.056892248624325334,
+      "grad_norm": 1.0602926015853882,
+      "learning_rate": 0.00017061683293522552,
+      "loss": 4.9808,
+      "step": 40490
+    },
+    {
+      "epoch": 0.05690629956248891,
+      "grad_norm": 1.0697758197784424,
+      "learning_rate": 0.00017065898552760995,
+      "loss": 5.0533,
+      "step": 40500
+    },
+    {
+      "epoch": 0.05692035050065249,
+      "grad_norm": 1.0304670333862305,
+      "learning_rate": 0.00017070113811999436,
+      "loss": 5.0054,
+      "step": 40510
+    },
+    {
+      "epoch": 0.05693440143881607,
+      "grad_norm": 1.0418838262557983,
+      "learning_rate": 0.00017074329071237879,
+      "loss": 4.9798,
+      "step": 40520
+    },
+    {
+      "epoch": 0.056948452376979646,
+      "grad_norm": 0.9913653135299683,
+      "learning_rate": 0.00017078544330476322,
+      "loss": 4.9925,
+      "step": 40530
+    },
+    {
+      "epoch": 0.05696250331514322,
+      "grad_norm": 0.988510012626648,
+      "learning_rate": 0.00017082759589714768,
+      "loss": 4.9377,
+      "step": 40540
+    },
+    {
+      "epoch": 0.0569765542533068,
+      "grad_norm": 1.019424557685852,
+      "learning_rate": 0.0001708697484895321,
+      "loss": 5.0321,
+      "step": 40550
+    },
+    {
+      "epoch": 0.05699060519147038,
+      "grad_norm": 1.0009490251541138,
+      "learning_rate": 0.00017091190108191654,
+      "loss": 4.9732,
+      "step": 40560
+    },
+    {
+      "epoch": 0.05700465612963396,
+      "grad_norm": 1.026256799697876,
+      "learning_rate": 0.00017095405367430094,
+      "loss": 4.912,
+      "step": 40570
+    },
+    {
+      "epoch": 0.05701870706779753,
+      "grad_norm": 1.0279102325439453,
+      "learning_rate": 0.00017099620626668537,
+      "loss": 4.9669,
+      "step": 40580
+    },
+    {
+      "epoch": 0.05703275800596111,
+      "grad_norm": 1.028504490852356,
+      "learning_rate": 0.0001710383588590698,
+      "loss": 4.9007,
+      "step": 40590
+    },
+    {
+      "epoch": 0.057046808944124686,
+      "grad_norm": 1.0499134063720703,
+      "learning_rate": 0.00017108051145145426,
+      "loss": 4.9829,
+      "step": 40600
+    },
+    {
+      "epoch": 0.05706085988228827,
+      "grad_norm": 1.0060728788375854,
+      "learning_rate": 0.0001711226640438387,
+      "loss": 4.9123,
+      "step": 40610
+    },
+    {
+      "epoch": 0.057074910820451845,
+      "grad_norm": 1.0461713075637817,
+      "learning_rate": 0.00017116481663622312,
+      "loss": 5.0092,
+      "step": 40620
+    },
+    {
+      "epoch": 0.05708896175861542,
+      "grad_norm": 1.0552021265029907,
+      "learning_rate": 0.00017120696922860756,
+      "loss": 4.8915,
+      "step": 40630
+    },
+    {
+      "epoch": 0.057103012696779,
+      "grad_norm": 1.0912543535232544,
+      "learning_rate": 0.00017124912182099196,
+      "loss": 4.9294,
+      "step": 40640
+    },
+    {
+      "epoch": 0.05711706363494257,
+      "grad_norm": 1.0685486793518066,
+      "learning_rate": 0.0001712912744133764,
+      "loss": 4.9328,
+      "step": 40650
+    },
+    {
+      "epoch": 0.057131114573106156,
+      "grad_norm": 1.093159556388855,
+      "learning_rate": 0.00017133342700576085,
+      "loss": 5.0754,
+      "step": 40660
+    },
+    {
+      "epoch": 0.05714516551126973,
+      "grad_norm": 1.0238051414489746,
+      "learning_rate": 0.00017137557959814528,
+      "loss": 4.8441,
+      "step": 40670
+    },
+    {
+      "epoch": 0.05715921644943331,
+      "grad_norm": 1.0513436794281006,
+      "learning_rate": 0.0001714177321905297,
+      "loss": 4.9075,
+      "step": 40680
+    },
+    {
+      "epoch": 0.057173267387596885,
+      "grad_norm": 1.0914078950881958,
+      "learning_rate": 0.00017145988478291414,
+      "loss": 4.9667,
+      "step": 40690
+    },
+    {
+      "epoch": 0.05718731832576046,
+      "grad_norm": 1.0058307647705078,
+      "learning_rate": 0.00017150203737529855,
+      "loss": 5.0005,
+      "step": 40700
+    },
+    {
+      "epoch": 0.05720136926392404,
+      "grad_norm": 0.9954103827476501,
+      "learning_rate": 0.00017154418996768298,
+      "loss": 5.0039,
+      "step": 40710
+    },
+    {
+      "epoch": 0.05721542020208762,
+      "grad_norm": 1.0139034986495972,
+      "learning_rate": 0.00017158634256006744,
+      "loss": 4.9052,
+      "step": 40720
+    },
+    {
+      "epoch": 0.0572294711402512,
+      "grad_norm": 1.0923491716384888,
+      "learning_rate": 0.00017162849515245187,
+      "loss": 4.8337,
+      "step": 40730
+    },
+    {
+      "epoch": 0.05724352207841477,
+      "grad_norm": 1.025458574295044,
+      "learning_rate": 0.0001716706477448363,
+      "loss": 4.941,
+      "step": 40740
+    },
+    {
+      "epoch": 0.05725757301657835,
+      "grad_norm": 1.1214048862457275,
+      "learning_rate": 0.00017171280033722073,
+      "loss": 4.9764,
+      "step": 40750
+    },
+    {
+      "epoch": 0.057271623954741925,
+      "grad_norm": 1.0020577907562256,
+      "learning_rate": 0.00017175495292960516,
+      "loss": 4.9437,
+      "step": 40760
+    },
+    {
+      "epoch": 0.05728567489290551,
+      "grad_norm": 1.0177772045135498,
+      "learning_rate": 0.00017179710552198957,
+      "loss": 4.8961,
+      "step": 40770
+    },
+    {
+      "epoch": 0.057299725831069084,
+      "grad_norm": 1.029328465461731,
+      "learning_rate": 0.00017183925811437402,
+      "loss": 4.8494,
+      "step": 40780
+    },
+    {
+      "epoch": 0.05731377676923266,
+      "grad_norm": 1.0189824104309082,
+      "learning_rate": 0.00017188141070675845,
+      "loss": 4.912,
+      "step": 40790
+    },
+    {
+      "epoch": 0.05732782770739624,
+      "grad_norm": 0.9939380884170532,
+      "learning_rate": 0.00017192356329914289,
+      "loss": 4.8773,
+      "step": 40800
+    },
+    {
+      "epoch": 0.05734187864555981,
+      "grad_norm": 1.076353907585144,
+      "learning_rate": 0.00017196571589152732,
+      "loss": 4.9416,
+      "step": 40810
+    },
+    {
+      "epoch": 0.057355929583723396,
+      "grad_norm": 1.049857258796692,
+      "learning_rate": 0.00017200786848391175,
+      "loss": 4.8408,
+      "step": 40820
+    },
+    {
+      "epoch": 0.05736998052188697,
+      "grad_norm": 1.0945760011672974,
+      "learning_rate": 0.00017205002107629615,
+      "loss": 4.928,
+      "step": 40830
+    },
+    {
+      "epoch": 0.05738403146005055,
+      "grad_norm": 1.0334876775741577,
+      "learning_rate": 0.0001720921736686806,
+      "loss": 5.0276,
+      "step": 40840
+    },
+    {
+      "epoch": 0.057398082398214124,
+      "grad_norm": 1.0239782333374023,
+      "learning_rate": 0.00017213432626106504,
+      "loss": 5.0475,
+      "step": 40850
+    },
+    {
+      "epoch": 0.0574121333363777,
+      "grad_norm": 1.0112613439559937,
+      "learning_rate": 0.00017217647885344947,
+      "loss": 5.0883,
+      "step": 40860
+    },
+    {
+      "epoch": 0.057426184274541284,
+      "grad_norm": 1.0588569641113281,
+      "learning_rate": 0.0001722186314458339,
+      "loss": 4.9827,
+      "step": 40870
+    },
+    {
+      "epoch": 0.05744023521270486,
+      "grad_norm": 1.0560933351516724,
+      "learning_rate": 0.00017226078403821834,
+      "loss": 4.953,
+      "step": 40880
+    },
+    {
+      "epoch": 0.057454286150868436,
+      "grad_norm": 1.08881413936615,
+      "learning_rate": 0.00017230293663060274,
+      "loss": 4.9457,
+      "step": 40890
+    },
+    {
+      "epoch": 0.05746833708903201,
+      "grad_norm": 1.0113881826400757,
+      "learning_rate": 0.0001723450892229872,
+      "loss": 5.0579,
+      "step": 40900
+    },
+    {
+      "epoch": 0.05748238802719559,
+      "grad_norm": 0.9996208548545837,
+      "learning_rate": 0.00017238724181537163,
+      "loss": 4.8815,
+      "step": 40910
+    },
+    {
+      "epoch": 0.05749643896535917,
+      "grad_norm": 1.0179857015609741,
+      "learning_rate": 0.00017242939440775606,
+      "loss": 4.8752,
+      "step": 40920
+    },
+    {
+      "epoch": 0.05751048990352275,
+      "grad_norm": 1.0453020334243774,
+      "learning_rate": 0.0001724715470001405,
+      "loss": 4.8984,
+      "step": 40930
+    },
+    {
+      "epoch": 0.057524540841686324,
+      "grad_norm": 1.0377174615859985,
+      "learning_rate": 0.00017251369959252492,
+      "loss": 4.9946,
+      "step": 40940
+    },
+    {
+      "epoch": 0.0575385917798499,
+      "grad_norm": 1.2559747695922852,
+      "learning_rate": 0.00017255585218490938,
+      "loss": 4.9518,
+      "step": 40950
+    },
+    {
+      "epoch": 0.057552642718013476,
+      "grad_norm": 1.0141798257827759,
+      "learning_rate": 0.00017259800477729378,
+      "loss": 4.9145,
+      "step": 40960
+    },
+    {
+      "epoch": 0.05756669365617706,
+      "grad_norm": 1.0602003335952759,
+      "learning_rate": 0.00017264015736967822,
+      "loss": 4.8659,
+      "step": 40970
+    },
+    {
+      "epoch": 0.057580744594340635,
+      "grad_norm": 1.0627175569534302,
+      "learning_rate": 0.00017268230996206265,
+      "loss": 5.0545,
+      "step": 40980
+    },
+    {
+      "epoch": 0.05759479553250421,
+      "grad_norm": 1.0080777406692505,
+      "learning_rate": 0.00017272446255444708,
+      "loss": 4.936,
+      "step": 40990
+    },
+    {
+      "epoch": 0.05760884647066779,
+      "grad_norm": 1.0756409168243408,
+      "learning_rate": 0.0001727666151468315,
+      "loss": 4.8759,
+      "step": 41000
+    },
+    {
+      "epoch": 0.057622897408831364,
+      "grad_norm": 1.0565085411071777,
+      "learning_rate": 0.00017280876773921597,
+      "loss": 4.9084,
+      "step": 41010
+    },
+    {
+      "epoch": 0.05763694834699494,
+      "grad_norm": 0.9990535378456116,
+      "learning_rate": 0.00017285092033160037,
+      "loss": 4.8277,
+      "step": 41020
+    },
+    {
+      "epoch": 0.05765099928515852,
+      "grad_norm": 1.066185712814331,
+      "learning_rate": 0.0001728930729239848,
+      "loss": 4.906,
+      "step": 41030
+    },
+    {
+      "epoch": 0.0576650502233221,
+      "grad_norm": 0.9987258315086365,
+      "learning_rate": 0.00017293522551636923,
+      "loss": 4.863,
+      "step": 41040
+    },
+    {
+      "epoch": 0.057679101161485676,
+      "grad_norm": 1.011893391609192,
+      "learning_rate": 0.00017297737810875367,
+      "loss": 4.931,
+      "step": 41050
+    },
+    {
+      "epoch": 0.05769315209964925,
+      "grad_norm": 1.0528898239135742,
+      "learning_rate": 0.0001730195307011381,
+      "loss": 5.0409,
+      "step": 41060
+    },
+    {
+      "epoch": 0.05770720303781283,
+      "grad_norm": 1.0291653871536255,
+      "learning_rate": 0.00017306168329352255,
+      "loss": 4.87,
+      "step": 41070
+    },
+    {
+      "epoch": 0.05772125397597641,
+      "grad_norm": 1.029781460762024,
+      "learning_rate": 0.00017310383588590699,
+      "loss": 5.0709,
+      "step": 41080
+    },
+    {
+      "epoch": 0.05773530491413999,
+      "grad_norm": 1.0627702474594116,
+      "learning_rate": 0.0001731459884782914,
+      "loss": 4.8982,
+      "step": 41090
+    },
+    {
+      "epoch": 0.05774935585230356,
+      "grad_norm": 1.027671456336975,
+      "learning_rate": 0.00017318814107067582,
+      "loss": 4.8947,
+      "step": 41100
+    },
+    {
+      "epoch": 0.05776340679046714,
+      "grad_norm": 1.0236762762069702,
+      "learning_rate": 0.00017323029366306025,
+      "loss": 4.9642,
+      "step": 41110
+    },
+    {
+      "epoch": 0.057777457728630716,
+      "grad_norm": 1.0269731283187866,
+      "learning_rate": 0.00017327244625544468,
+      "loss": 4.9839,
+      "step": 41120
+    },
+    {
+      "epoch": 0.0577915086667943,
+      "grad_norm": 1.002811074256897,
+      "learning_rate": 0.00017331459884782914,
+      "loss": 4.96,
+      "step": 41130
+    },
+    {
+      "epoch": 0.057805559604957875,
+      "grad_norm": 1.0268176794052124,
+      "learning_rate": 0.00017335675144021357,
+      "loss": 4.9521,
+      "step": 41140
+    },
+    {
+      "epoch": 0.05781961054312145,
+      "grad_norm": 1.0485950708389282,
+      "learning_rate": 0.00017339890403259798,
+      "loss": 4.9922,
+      "step": 41150
+    },
+    {
+      "epoch": 0.05783366148128503,
+      "grad_norm": 1.0000367164611816,
+      "learning_rate": 0.0001734410566249824,
+      "loss": 4.8602,
+      "step": 41160
+    },
+    {
+      "epoch": 0.0578477124194486,
+      "grad_norm": 1.0004191398620605,
+      "learning_rate": 0.00017348320921736684,
+      "loss": 4.8732,
+      "step": 41170
+    },
+    {
+      "epoch": 0.057861763357612186,
+      "grad_norm": 1.0630460977554321,
+      "learning_rate": 0.0001735253618097513,
+      "loss": 5.0591,
+      "step": 41180
+    },
+    {
+      "epoch": 0.05787581429577576,
+      "grad_norm": 1.069863200187683,
+      "learning_rate": 0.00017356751440213573,
+      "loss": 4.9227,
+      "step": 41190
+    },
+    {
+      "epoch": 0.05788986523393934,
+      "grad_norm": 1.0146735906600952,
+      "learning_rate": 0.00017360966699452016,
+      "loss": 4.9153,
+      "step": 41200
+    },
+    {
+      "epoch": 0.057903916172102915,
+      "grad_norm": 1.0278013944625854,
+      "learning_rate": 0.0001736518195869046,
+      "loss": 4.9406,
+      "step": 41210
+    },
+    {
+      "epoch": 0.05791796711026649,
+      "grad_norm": 1.0474538803100586,
+      "learning_rate": 0.000173693972179289,
+      "loss": 4.8249,
+      "step": 41220
+    },
+    {
+      "epoch": 0.057932018048430074,
+      "grad_norm": 1.0575395822525024,
+      "learning_rate": 0.00017373612477167343,
+      "loss": 4.9606,
+      "step": 41230
+    },
+    {
+      "epoch": 0.05794606898659365,
+      "grad_norm": 1.0785536766052246,
+      "learning_rate": 0.00017377827736405788,
+      "loss": 4.9818,
+      "step": 41240
+    },
+    {
+      "epoch": 0.05796011992475723,
+      "grad_norm": 1.034358024597168,
+      "learning_rate": 0.00017382042995644232,
+      "loss": 5.0435,
+      "step": 41250
+    },
+    {
+      "epoch": 0.0579741708629208,
+      "grad_norm": 1.0407671928405762,
+      "learning_rate": 0.00017386258254882675,
+      "loss": 4.9362,
+      "step": 41260
+    },
+    {
+      "epoch": 0.05798822180108438,
+      "grad_norm": 1.1351573467254639,
+      "learning_rate": 0.00017390473514121118,
+      "loss": 4.9251,
+      "step": 41270
+    },
+    {
+      "epoch": 0.05800227273924796,
+      "grad_norm": 1.0201154947280884,
+      "learning_rate": 0.00017394688773359558,
+      "loss": 4.9085,
+      "step": 41280
+    },
+    {
+      "epoch": 0.05801632367741154,
+      "grad_norm": 1.0584053993225098,
+      "learning_rate": 0.00017398904032598,
+      "loss": 4.9704,
+      "step": 41290
+    },
+    {
+      "epoch": 0.058030374615575114,
+      "grad_norm": 1.0306909084320068,
+      "learning_rate": 0.00017403119291836447,
+      "loss": 4.9058,
+      "step": 41300
+    },
+    {
+      "epoch": 0.05804442555373869,
+      "grad_norm": 1.0459680557250977,
+      "learning_rate": 0.0001740733455107489,
+      "loss": 4.949,
+      "step": 41310
+    },
+    {
+      "epoch": 0.05805847649190227,
+      "grad_norm": 0.9976636171340942,
+      "learning_rate": 0.00017411549810313333,
+      "loss": 5.0384,
+      "step": 41320
+    },
+    {
+      "epoch": 0.05807252743006584,
+      "grad_norm": 1.01653254032135,
+      "learning_rate": 0.00017415765069551776,
+      "loss": 4.9653,
+      "step": 41330
+    },
+    {
+      "epoch": 0.058086578368229426,
+      "grad_norm": 1.0457319021224976,
+      "learning_rate": 0.0001741998032879022,
+      "loss": 4.849,
+      "step": 41340
+    },
+    {
+      "epoch": 0.058100629306393,
+      "grad_norm": 1.0464872121810913,
+      "learning_rate": 0.0001742419558802866,
+      "loss": 4.8891,
+      "step": 41350
+    },
+    {
+      "epoch": 0.05811468024455658,
+      "grad_norm": 1.0535956621170044,
+      "learning_rate": 0.00017428410847267106,
+      "loss": 4.988,
+      "step": 41360
+    },
+    {
+      "epoch": 0.058128731182720154,
+      "grad_norm": 1.0425530672073364,
+      "learning_rate": 0.0001743262610650555,
+      "loss": 4.816,
+      "step": 41370
+    },
+    {
+      "epoch": 0.05814278212088373,
+      "grad_norm": 1.0837699174880981,
+      "learning_rate": 0.00017436841365743992,
+      "loss": 4.9787,
+      "step": 41380
+    },
+    {
+      "epoch": 0.058156833059047314,
+      "grad_norm": 1.025530457496643,
+      "learning_rate": 0.00017441056624982435,
+      "loss": 4.9268,
+      "step": 41390
+    },
+    {
+      "epoch": 0.05817088399721089,
+      "grad_norm": 1.0290592908859253,
+      "learning_rate": 0.00017445271884220878,
+      "loss": 4.9541,
+      "step": 41400
+    },
+    {
+      "epoch": 0.058184934935374466,
+      "grad_norm": 1.0412163734436035,
+      "learning_rate": 0.0001744948714345932,
+      "loss": 4.938,
+      "step": 41410
+    },
+    {
+      "epoch": 0.05819898587353804,
+      "grad_norm": 1.013218879699707,
+      "learning_rate": 0.00017453702402697765,
+      "loss": 5.0052,
+      "step": 41420
+    },
+    {
+      "epoch": 0.05821303681170162,
+      "grad_norm": 1.0787619352340698,
+      "learning_rate": 0.00017457917661936208,
+      "loss": 4.8859,
+      "step": 41430
+    },
+    {
+      "epoch": 0.0582270877498652,
+      "grad_norm": 1.0490294694900513,
+      "learning_rate": 0.0001746213292117465,
+      "loss": 4.8447,
+      "step": 41440
+    },
+    {
+      "epoch": 0.05824113868802878,
+      "grad_norm": 1.0566773414611816,
+      "learning_rate": 0.00017466348180413094,
+      "loss": 5.0328,
+      "step": 41450
+    },
+    {
+      "epoch": 0.058255189626192354,
+      "grad_norm": 1.0242851972579956,
+      "learning_rate": 0.00017470563439651537,
+      "loss": 4.9068,
+      "step": 41460
+    },
+    {
+      "epoch": 0.05826924056435593,
+      "grad_norm": 1.0676864385604858,
+      "learning_rate": 0.00017474778698889983,
+      "loss": 4.9384,
+      "step": 41470
+    },
+    {
+      "epoch": 0.058283291502519506,
+      "grad_norm": 1.0332306623458862,
+      "learning_rate": 0.00017478993958128423,
+      "loss": 5.0788,
+      "step": 41480
+    },
+    {
+      "epoch": 0.05829734244068309,
+      "grad_norm": 0.9824172258377075,
+      "learning_rate": 0.00017483209217366866,
+      "loss": 4.8847,
+      "step": 41490
+    },
+    {
+      "epoch": 0.058311393378846665,
+      "grad_norm": 1.071233868598938,
+      "learning_rate": 0.0001748742447660531,
+      "loss": 4.8921,
+      "step": 41500
+    },
+    {
+      "epoch": 0.05832544431701024,
+      "grad_norm": 1.0408722162246704,
+      "learning_rate": 0.00017491639735843753,
+      "loss": 5.0539,
+      "step": 41510
+    },
+    {
+      "epoch": 0.05833949525517382,
+      "grad_norm": 1.0348032712936401,
+      "learning_rate": 0.00017495854995082196,
+      "loss": 4.8807,
+      "step": 41520
+    },
+    {
+      "epoch": 0.058353546193337394,
+      "grad_norm": 1.0092370510101318,
+      "learning_rate": 0.00017500070254320642,
+      "loss": 4.9697,
+      "step": 41530
+    },
+    {
+      "epoch": 0.05836759713150098,
+      "grad_norm": 1.1517890691757202,
+      "learning_rate": 0.00017504285513559082,
+      "loss": 4.9005,
+      "step": 41540
+    },
+    {
+      "epoch": 0.05838164806966455,
+      "grad_norm": 1.056350827217102,
+      "learning_rate": 0.00017508500772797525,
+      "loss": 4.9908,
+      "step": 41550
+    },
+    {
+      "epoch": 0.05839569900782813,
+      "grad_norm": 1.0199683904647827,
+      "learning_rate": 0.00017512716032035968,
+      "loss": 4.9737,
+      "step": 41560
+    },
+    {
+      "epoch": 0.058409749945991705,
+      "grad_norm": 1.0125006437301636,
+      "learning_rate": 0.0001751693129127441,
+      "loss": 4.989,
+      "step": 41570
+    },
+    {
+      "epoch": 0.05842380088415528,
+      "grad_norm": 1.0040638446807861,
+      "learning_rate": 0.00017521146550512854,
+      "loss": 5.0114,
+      "step": 41580
+    },
+    {
+      "epoch": 0.058437851822318865,
+      "grad_norm": 1.3754082918167114,
+      "learning_rate": 0.000175253618097513,
+      "loss": 5.0118,
+      "step": 41590
+    },
+    {
+      "epoch": 0.05845190276048244,
+      "grad_norm": 0.9892797470092773,
+      "learning_rate": 0.0001752957706898974,
+      "loss": 4.9575,
+      "step": 41600
+    },
+    {
+      "epoch": 0.05846595369864602,
+      "grad_norm": 1.073715090751648,
+      "learning_rate": 0.00017533792328228184,
+      "loss": 4.9162,
+      "step": 41610
+    },
+    {
+      "epoch": 0.05848000463680959,
+      "grad_norm": 1.0172522068023682,
+      "learning_rate": 0.00017538007587466627,
+      "loss": 5.0024,
+      "step": 41620
+    },
+    {
+      "epoch": 0.05849405557497317,
+      "grad_norm": 1.0679402351379395,
+      "learning_rate": 0.0001754222284670507,
+      "loss": 4.9627,
+      "step": 41630
+    },
+    {
+      "epoch": 0.058508106513136746,
+      "grad_norm": 1.0129616260528564,
+      "learning_rate": 0.00017546438105943513,
+      "loss": 4.9732,
+      "step": 41640
+    },
+    {
+      "epoch": 0.05852215745130033,
+      "grad_norm": 1.0495364665985107,
+      "learning_rate": 0.0001755065336518196,
+      "loss": 4.8889,
+      "step": 41650
+    },
+    {
+      "epoch": 0.058536208389463905,
+      "grad_norm": 0.9987136721611023,
+      "learning_rate": 0.00017554868624420402,
+      "loss": 4.9349,
+      "step": 41660
+    },
+    {
+      "epoch": 0.05855025932762748,
+      "grad_norm": 1.0522116422653198,
+      "learning_rate": 0.00017559083883658842,
+      "loss": 4.9909,
+      "step": 41670
+    },
+    {
+      "epoch": 0.05856431026579106,
+      "grad_norm": 1.0374841690063477,
+      "learning_rate": 0.00017563299142897286,
+      "loss": 4.9103,
+      "step": 41680
+    },
+    {
+      "epoch": 0.05857836120395463,
+      "grad_norm": 1.0144661664962769,
+      "learning_rate": 0.0001756751440213573,
+      "loss": 4.9681,
+      "step": 41690
+    },
+    {
+      "epoch": 0.058592412142118216,
+      "grad_norm": 1.0426274538040161,
+      "learning_rate": 0.00017571729661374172,
+      "loss": 5.0318,
+      "step": 41700
+    },
+    {
+      "epoch": 0.05860646308028179,
+      "grad_norm": 1.0269166231155396,
+      "learning_rate": 0.00017575944920612618,
+      "loss": 5.0346,
+      "step": 41710
+    },
+    {
+      "epoch": 0.05862051401844537,
+      "grad_norm": 1.0402437448501587,
+      "learning_rate": 0.0001758016017985106,
+      "loss": 4.8424,
+      "step": 41720
+    },
+    {
+      "epoch": 0.058634564956608945,
+      "grad_norm": 0.9999310970306396,
+      "learning_rate": 0.000175843754390895,
+      "loss": 4.8544,
+      "step": 41730
+    },
+    {
+      "epoch": 0.05864861589477252,
+      "grad_norm": 1.0584388971328735,
+      "learning_rate": 0.00017588590698327944,
+      "loss": 4.976,
+      "step": 41740
+    },
+    {
+      "epoch": 0.058662666832936104,
+      "grad_norm": 1.0145403146743774,
+      "learning_rate": 0.00017592805957566387,
+      "loss": 4.9392,
+      "step": 41750
+    },
+    {
+      "epoch": 0.05867671777109968,
+      "grad_norm": 1.0221564769744873,
+      "learning_rate": 0.00017597021216804833,
+      "loss": 4.9297,
+      "step": 41760
+    },
+    {
+      "epoch": 0.05869076870926326,
+      "grad_norm": 1.0751060247421265,
+      "learning_rate": 0.00017601236476043276,
+      "loss": 4.9083,
+      "step": 41770
+    },
+    {
+      "epoch": 0.05870481964742683,
+      "grad_norm": 1.022549033164978,
+      "learning_rate": 0.0001760545173528172,
+      "loss": 4.9086,
+      "step": 41780
+    },
+    {
+      "epoch": 0.05871887058559041,
+      "grad_norm": 1.0539672374725342,
+      "learning_rate": 0.00017609666994520163,
+      "loss": 4.8246,
+      "step": 41790
+    },
+    {
+      "epoch": 0.05873292152375399,
+      "grad_norm": 1.0099142789840698,
+      "learning_rate": 0.00017613882253758603,
+      "loss": 4.8999,
+      "step": 41800
+    },
+    {
+      "epoch": 0.05874697246191757,
+      "grad_norm": 1.0530145168304443,
+      "learning_rate": 0.00017617675987073204,
+      "loss": 4.9517,
+      "step": 41810
+    },
+    {
+      "epoch": 0.058761023400081144,
+      "grad_norm": 1.0349324941635132,
+      "learning_rate": 0.00017621891246311647,
+      "loss": 4.9144,
+      "step": 41820
+    },
+    {
+      "epoch": 0.05877507433824472,
+      "grad_norm": 0.9777777791023254,
+      "learning_rate": 0.00017626106505550087,
+      "loss": 4.8404,
+      "step": 41830
+    },
+    {
+      "epoch": 0.0587891252764083,
+      "grad_norm": 1.0137258768081665,
+      "learning_rate": 0.00017630321764788533,
+      "loss": 4.7486,
+      "step": 41840
+    },
+    {
+      "epoch": 0.05880317621457188,
+      "grad_norm": 0.9945560097694397,
+      "learning_rate": 0.00017634537024026976,
+      "loss": 4.9688,
+      "step": 41850
+    },
+    {
+      "epoch": 0.058817227152735456,
+      "grad_norm": 1.0335863828659058,
+      "learning_rate": 0.0001763875228326542,
+      "loss": 4.9175,
+      "step": 41860
+    },
+    {
+      "epoch": 0.05883127809089903,
+      "grad_norm": 1.0477795600891113,
+      "learning_rate": 0.00017642967542503863,
+      "loss": 5.0537,
+      "step": 41870
+    },
+    {
+      "epoch": 0.05884532902906261,
+      "grad_norm": 1.0379626750946045,
+      "learning_rate": 0.00017647182801742306,
+      "loss": 4.8798,
+      "step": 41880
+    },
+    {
+      "epoch": 0.058859379967226184,
+      "grad_norm": 1.0096701383590698,
+      "learning_rate": 0.00017651398060980752,
+      "loss": 4.9598,
+      "step": 41890
+    },
+    {
+      "epoch": 0.05887343090538977,
+      "grad_norm": 0.9835436940193176,
+      "learning_rate": 0.00017655613320219192,
+      "loss": 4.8876,
+      "step": 41900
+    },
+    {
+      "epoch": 0.058887481843553344,
+      "grad_norm": 1.0108128786087036,
+      "learning_rate": 0.00017659828579457635,
+      "loss": 4.9465,
+      "step": 41910
+    },
+    {
+      "epoch": 0.05890153278171692,
+      "grad_norm": 0.9913715720176697,
+      "learning_rate": 0.00017664043838696078,
+      "loss": 4.8998,
+      "step": 41920
+    },
+    {
+      "epoch": 0.058915583719880496,
+      "grad_norm": 1.0140135288238525,
+      "learning_rate": 0.0001766825909793452,
+      "loss": 4.9478,
+      "step": 41930
+    },
+    {
+      "epoch": 0.05892963465804407,
+      "grad_norm": 1.0244427919387817,
+      "learning_rate": 0.00017672474357172964,
+      "loss": 4.8803,
+      "step": 41940
+    },
+    {
+      "epoch": 0.05894368559620765,
+      "grad_norm": 1.0371630191802979,
+      "learning_rate": 0.0001767668961641141,
+      "loss": 4.9136,
+      "step": 41950
+    },
+    {
+      "epoch": 0.05895773653437123,
+      "grad_norm": 1.0342624187469482,
+      "learning_rate": 0.0001768090487564985,
+      "loss": 5.0633,
+      "step": 41960
+    },
+    {
+      "epoch": 0.05897178747253481,
+      "grad_norm": 1.0081278085708618,
+      "learning_rate": 0.00017685120134888294,
+      "loss": 4.972,
+      "step": 41970
+    },
+    {
+      "epoch": 0.058985838410698384,
+      "grad_norm": 1.0183650255203247,
+      "learning_rate": 0.00017689335394126737,
+      "loss": 4.8956,
+      "step": 41980
+    },
+    {
+      "epoch": 0.05899988934886196,
+      "grad_norm": 1.0174623727798462,
+      "learning_rate": 0.0001769355065336518,
+      "loss": 4.7951,
+      "step": 41990
+    },
+    {
+      "epoch": 0.059013940287025536,
+      "grad_norm": 1.0399575233459473,
+      "learning_rate": 0.00017697765912603623,
+      "loss": 4.8732,
+      "step": 42000
+    },
+    {
+      "epoch": 0.05902799122518912,
+      "grad_norm": 1.0714610815048218,
+      "learning_rate": 0.0001770198117184207,
+      "loss": 4.8844,
+      "step": 42010
+    },
+    {
+      "epoch": 0.059042042163352695,
+      "grad_norm": 1.07644522190094,
+      "learning_rate": 0.00017706196431080512,
+      "loss": 4.8854,
+      "step": 42020
+    },
+    {
+      "epoch": 0.05905609310151627,
+      "grad_norm": 1.018179178237915,
+      "learning_rate": 0.00017710411690318953,
+      "loss": 4.9124,
+      "step": 42030
+    },
+    {
+      "epoch": 0.05907014403967985,
+      "grad_norm": 1.0028572082519531,
+      "learning_rate": 0.00017714626949557396,
+      "loss": 5.0101,
+      "step": 42040
+    },
+    {
+      "epoch": 0.059084194977843424,
+      "grad_norm": 1.0452194213867188,
+      "learning_rate": 0.0001771884220879584,
+      "loss": 4.9129,
+      "step": 42050
+    },
+    {
+      "epoch": 0.05909824591600701,
+      "grad_norm": 0.9863782525062561,
+      "learning_rate": 0.00017723057468034282,
+      "loss": 4.961,
+      "step": 42060
+    },
+    {
+      "epoch": 0.05911229685417058,
+      "grad_norm": 1.0302046537399292,
+      "learning_rate": 0.00017727272727272728,
+      "loss": 4.896,
+      "step": 42070
+    },
+    {
+      "epoch": 0.05912634779233416,
+      "grad_norm": 0.994310200214386,
+      "learning_rate": 0.0001773148798651117,
+      "loss": 4.9747,
+      "step": 42080
+    },
+    {
+      "epoch": 0.059140398730497735,
+      "grad_norm": 1.021562933921814,
+      "learning_rate": 0.0001773570324574961,
+      "loss": 4.8488,
+      "step": 42090
+    },
+    {
+      "epoch": 0.05915444966866131,
+      "grad_norm": 1.034226655960083,
+      "learning_rate": 0.00017739918504988054,
+      "loss": 4.8405,
+      "step": 42100
+    },
+    {
+      "epoch": 0.059168500606824895,
+      "grad_norm": 1.0121209621429443,
+      "learning_rate": 0.00017744133764226497,
+      "loss": 4.9374,
+      "step": 42110
+    },
+    {
+      "epoch": 0.05918255154498847,
+      "grad_norm": 0.9862031936645508,
+      "learning_rate": 0.0001774834902346494,
+      "loss": 4.8789,
+      "step": 42120
+    },
+    {
+      "epoch": 0.05919660248315205,
+      "grad_norm": 1.0321632623672485,
+      "learning_rate": 0.00017752564282703386,
+      "loss": 4.9114,
+      "step": 42130
+    },
+    {
+      "epoch": 0.05921065342131562,
+      "grad_norm": 1.0482895374298096,
+      "learning_rate": 0.0001775677954194183,
+      "loss": 4.8856,
+      "step": 42140
+    },
+    {
+      "epoch": 0.0592247043594792,
+      "grad_norm": 1.1908639669418335,
+      "learning_rate": 0.0001776099480118027,
+      "loss": 4.8354,
+      "step": 42150
+    },
+    {
+      "epoch": 0.05923875529764278,
+      "grad_norm": 1.1106269359588623,
+      "learning_rate": 0.00017765210060418713,
+      "loss": 4.8703,
+      "step": 42160
+    },
+    {
+      "epoch": 0.05925280623580636,
+      "grad_norm": 1.042508840560913,
+      "learning_rate": 0.00017769425319657156,
+      "loss": 4.8856,
+      "step": 42170
+    },
+    {
+      "epoch": 0.059266857173969935,
+      "grad_norm": 1.398681640625,
+      "learning_rate": 0.000177736405788956,
+      "loss": 4.9736,
+      "step": 42180
+    },
+    {
+      "epoch": 0.05928090811213351,
+      "grad_norm": 1.0562206506729126,
+      "learning_rate": 0.00017777855838134045,
+      "loss": 4.9827,
+      "step": 42190
+    },
+    {
+      "epoch": 0.05929495905029709,
+      "grad_norm": 0.977316677570343,
+      "learning_rate": 0.00017782071097372488,
+      "loss": 4.9283,
+      "step": 42200
+    },
+    {
+      "epoch": 0.05930900998846067,
+      "grad_norm": 1.0145175457000732,
+      "learning_rate": 0.0001778628635661093,
+      "loss": 4.9636,
+      "step": 42210
+    },
+    {
+      "epoch": 0.059323060926624246,
+      "grad_norm": 1.0313082933425903,
+      "learning_rate": 0.00017790501615849372,
+      "loss": 4.7409,
+      "step": 42220
+    },
+    {
+      "epoch": 0.05933711186478782,
+      "grad_norm": 1.0246168375015259,
+      "learning_rate": 0.00017794716875087815,
+      "loss": 4.9663,
+      "step": 42230
+    },
+    {
+      "epoch": 0.0593511628029514,
+      "grad_norm": 1.1417378187179565,
+      "learning_rate": 0.00017798932134326258,
+      "loss": 4.9773,
+      "step": 42240
+    },
+    {
+      "epoch": 0.059365213741114975,
+      "grad_norm": 1.0039445161819458,
+      "learning_rate": 0.00017803147393564704,
+      "loss": 4.9977,
+      "step": 42250
+    },
+    {
+      "epoch": 0.05937926467927855,
+      "grad_norm": 1.012376070022583,
+      "learning_rate": 0.00017807362652803147,
+      "loss": 4.8092,
+      "step": 42260
+    },
+    {
+      "epoch": 0.059393315617442134,
+      "grad_norm": 1.0877492427825928,
+      "learning_rate": 0.0001781157791204159,
+      "loss": 4.9357,
+      "step": 42270
+    },
+    {
+      "epoch": 0.05940736655560571,
+      "grad_norm": 1.0365524291992188,
+      "learning_rate": 0.0001781579317128003,
+      "loss": 4.8717,
+      "step": 42280
+    },
+    {
+      "epoch": 0.05942141749376929,
+      "grad_norm": 0.9986197352409363,
+      "learning_rate": 0.00017820008430518474,
+      "loss": 4.9256,
+      "step": 42290
+    },
+    {
+      "epoch": 0.05943546843193286,
+      "grad_norm": 1.0150564908981323,
+      "learning_rate": 0.00017824223689756917,
+      "loss": 4.8873,
+      "step": 42300
+    },
+    {
+      "epoch": 0.05944951937009644,
+      "grad_norm": 1.0646934509277344,
+      "learning_rate": 0.00017828438948995362,
+      "loss": 4.8436,
+      "step": 42310
+    },
+    {
+      "epoch": 0.05946357030826002,
+      "grad_norm": 1.0046230554580688,
+      "learning_rate": 0.00017832654208233806,
+      "loss": 5.0225,
+      "step": 42320
+    },
+    {
+      "epoch": 0.0594776212464236,
+      "grad_norm": 1.0435608625411987,
+      "learning_rate": 0.0001783686946747225,
+      "loss": 4.9509,
+      "step": 42330
+    },
+    {
+      "epoch": 0.059491672184587174,
+      "grad_norm": 1.004775047302246,
+      "learning_rate": 0.00017841084726710692,
+      "loss": 4.8448,
+      "step": 42340
+    },
+    {
+      "epoch": 0.05950572312275075,
+      "grad_norm": 1.0181119441986084,
+      "learning_rate": 0.00017845299985949132,
+      "loss": 4.9456,
+      "step": 42350
+    },
+    {
+      "epoch": 0.05951977406091433,
+      "grad_norm": 1.0563257932662964,
+      "learning_rate": 0.00017849515245187575,
+      "loss": 4.938,
+      "step": 42360
+    },
+    {
+      "epoch": 0.05953382499907791,
+      "grad_norm": 1.2527403831481934,
+      "learning_rate": 0.0001785373050442602,
+      "loss": 4.8606,
+      "step": 42370
+    },
+    {
+      "epoch": 0.059547875937241486,
+      "grad_norm": 0.9892441630363464,
+      "learning_rate": 0.00017857945763664464,
+      "loss": 4.9537,
+      "step": 42380
+    },
+    {
+      "epoch": 0.05956192687540506,
+      "grad_norm": 0.9748817086219788,
+      "learning_rate": 0.00017862161022902907,
+      "loss": 4.9919,
+      "step": 42390
+    },
+    {
+      "epoch": 0.05957597781356864,
+      "grad_norm": 1.0507982969284058,
+      "learning_rate": 0.0001786637628214135,
+      "loss": 4.9252,
+      "step": 42400
+    },
+    {
+      "epoch": 0.059590028751732214,
+      "grad_norm": 1.108594536781311,
+      "learning_rate": 0.0001787059154137979,
+      "loss": 4.9004,
+      "step": 42410
+    },
+    {
+      "epoch": 0.0596040796898958,
+      "grad_norm": 1.0259724855422974,
+      "learning_rate": 0.00017874806800618237,
+      "loss": 4.7901,
+      "step": 42420
+    },
+    {
+      "epoch": 0.059618130628059374,
+      "grad_norm": 1.0419907569885254,
+      "learning_rate": 0.0001787902205985668,
+      "loss": 4.9264,
+      "step": 42430
+    },
+    {
+      "epoch": 0.05963218156622295,
+      "grad_norm": 1.0113157033920288,
+      "learning_rate": 0.00017883237319095123,
+      "loss": 4.9435,
+      "step": 42440
+    },
+    {
+      "epoch": 0.059646232504386526,
+      "grad_norm": 1.0024772882461548,
+      "learning_rate": 0.00017887452578333566,
+      "loss": 5.0503,
+      "step": 42450
+    },
+    {
+      "epoch": 0.0596602834425501,
+      "grad_norm": 1.024277687072754,
+      "learning_rate": 0.0001789166783757201,
+      "loss": 5.0154,
+      "step": 42460
+    },
+    {
+      "epoch": 0.059674334380713685,
+      "grad_norm": 0.9590503573417664,
+      "learning_rate": 0.00017895883096810455,
+      "loss": 4.9712,
+      "step": 42470
+    },
+    {
+      "epoch": 0.05968838531887726,
+      "grad_norm": 1.0252444744110107,
+      "learning_rate": 0.00017900098356048895,
+      "loss": 4.8889,
+      "step": 42480
+    },
+    {
+      "epoch": 0.05970243625704084,
+      "grad_norm": 0.9953712821006775,
+      "learning_rate": 0.00017904313615287339,
+      "loss": 5.0119,
+      "step": 42490
+    },
+    {
+      "epoch": 0.059716487195204414,
+      "grad_norm": 1.018610954284668,
+      "learning_rate": 0.00017908528874525782,
+      "loss": 4.8525,
+      "step": 42500
+    },
+    {
+      "epoch": 0.05973053813336799,
+      "grad_norm": 0.9951494336128235,
+      "learning_rate": 0.00017912744133764225,
+      "loss": 4.9818,
+      "step": 42510
+    },
+    {
+      "epoch": 0.05974458907153157,
+      "grad_norm": 1.0133929252624512,
+      "learning_rate": 0.00017916959393002668,
+      "loss": 5.0259,
+      "step": 42520
+    },
+    {
+      "epoch": 0.05975864000969515,
+      "grad_norm": 0.9819827079772949,
+      "learning_rate": 0.00017921174652241114,
+      "loss": 4.9843,
+      "step": 42530
+    },
+    {
+      "epoch": 0.059772690947858725,
+      "grad_norm": 0.9785075187683105,
+      "learning_rate": 0.00017925389911479554,
+      "loss": 4.9552,
+      "step": 42540
+    },
+    {
+      "epoch": 0.0597867418860223,
+      "grad_norm": 1.0113465785980225,
+      "learning_rate": 0.00017929605170717997,
+      "loss": 4.9311,
+      "step": 42550
+    },
+    {
+      "epoch": 0.05980079282418588,
+      "grad_norm": 0.9836959838867188,
+      "learning_rate": 0.0001793382042995644,
+      "loss": 4.84,
+      "step": 42560
+    },
+    {
+      "epoch": 0.05981484376234946,
+      "grad_norm": 0.9870015382766724,
+      "learning_rate": 0.00017938035689194884,
+      "loss": 4.9337,
+      "step": 42570
+    },
+    {
+      "epoch": 0.05982889470051304,
+      "grad_norm": 0.9871739745140076,
+      "learning_rate": 0.00017942250948433327,
+      "loss": 4.8171,
+      "step": 42580
+    },
+    {
+      "epoch": 0.05984294563867661,
+      "grad_norm": 0.9959299564361572,
+      "learning_rate": 0.00017946466207671772,
+      "loss": 4.8464,
+      "step": 42590
+    },
+    {
+      "epoch": 0.05985699657684019,
+      "grad_norm": 1.0436452627182007,
+      "learning_rate": 0.00017950681466910216,
+      "loss": 4.941,
+      "step": 42600
+    },
+    {
+      "epoch": 0.059871047515003765,
+      "grad_norm": 1.0525795221328735,
+      "learning_rate": 0.00017954896726148656,
+      "loss": 4.8133,
+      "step": 42610
+    },
+    {
+      "epoch": 0.05988509845316734,
+      "grad_norm": 1.0073230266571045,
+      "learning_rate": 0.000179591119853871,
+      "loss": 4.8356,
+      "step": 42620
+    },
+    {
+      "epoch": 0.059899149391330925,
+      "grad_norm": 0.9857484102249146,
+      "learning_rate": 0.00017963327244625542,
+      "loss": 4.8508,
+      "step": 42630
+    },
+    {
+      "epoch": 0.0599132003294945,
+      "grad_norm": 1.013698935508728,
+      "learning_rate": 0.00017967542503863985,
+      "loss": 4.806,
+      "step": 42640
+    },
+    {
+      "epoch": 0.05992725126765808,
+      "grad_norm": 1.013667106628418,
+      "learning_rate": 0.0001797175776310243,
+      "loss": 4.9422,
+      "step": 42650
+    },
+    {
+      "epoch": 0.05994130220582165,
+      "grad_norm": 1.0081895589828491,
+      "learning_rate": 0.00017975973022340874,
+      "loss": 4.9618,
+      "step": 42660
+    },
+    {
+      "epoch": 0.05995535314398523,
+      "grad_norm": 1.0109422206878662,
+      "learning_rate": 0.00017980188281579315,
+      "loss": 4.8748,
+      "step": 42670
+    },
+    {
+      "epoch": 0.05996940408214881,
+      "grad_norm": 1.035300374031067,
+      "learning_rate": 0.00017984403540817758,
+      "loss": 4.8972,
+      "step": 42680
+    },
+    {
+      "epoch": 0.05998345502031239,
+      "grad_norm": 0.994915783405304,
+      "learning_rate": 0.000179886188000562,
+      "loss": 4.889,
+      "step": 42690
+    },
+    {
+      "epoch": 0.059997505958475965,
+      "grad_norm": 1.0624393224716187,
+      "learning_rate": 0.00017992834059294644,
+      "loss": 4.9058,
+      "step": 42700
+    },
+    {
+      "epoch": 0.06001155689663954,
+      "grad_norm": 1.0263959169387817,
+      "learning_rate": 0.0001799704931853309,
+      "loss": 4.8691,
+      "step": 42710
+    },
+    {
+      "epoch": 0.06002560783480312,
+      "grad_norm": 1.0453038215637207,
+      "learning_rate": 0.00018001264577771533,
+      "loss": 4.6771,
+      "step": 42720
+    },
+    {
+      "epoch": 0.0600396587729667,
+      "grad_norm": 1.070002794265747,
+      "learning_rate": 0.00018005479837009973,
+      "loss": 4.9346,
+      "step": 42730
+    },
+    {
+      "epoch": 0.060053709711130276,
+      "grad_norm": 1.0732343196868896,
+      "learning_rate": 0.00018009695096248417,
+      "loss": 5.021,
+      "step": 42740
+    },
+    {
+      "epoch": 0.06006776064929385,
+      "grad_norm": 1.127801775932312,
+      "learning_rate": 0.0001801391035548686,
+      "loss": 4.9084,
+      "step": 42750
+    },
+    {
+      "epoch": 0.06008181158745743,
+      "grad_norm": 1.0170589685440063,
+      "learning_rate": 0.00018018125614725303,
+      "loss": 4.9463,
+      "step": 42760
+    },
+    {
+      "epoch": 0.060095862525621005,
+      "grad_norm": 1.0990118980407715,
+      "learning_rate": 0.00018022340873963749,
+      "loss": 4.9675,
+      "step": 42770
+    },
+    {
+      "epoch": 0.06010991346378459,
+      "grad_norm": 1.0004825592041016,
+      "learning_rate": 0.00018026556133202192,
+      "loss": 5.0224,
+      "step": 42780
+    },
+    {
+      "epoch": 0.060123964401948164,
+      "grad_norm": 1.0344843864440918,
+      "learning_rate": 0.00018030771392440635,
+      "loss": 4.8195,
+      "step": 42790
+    },
+    {
+      "epoch": 0.06013801534011174,
+      "grad_norm": 1.0126923322677612,
+      "learning_rate": 0.00018034986651679075,
+      "loss": 4.9871,
+      "step": 42800
+    },
+    {
+      "epoch": 0.06015206627827532,
+      "grad_norm": 1.0566537380218506,
+      "learning_rate": 0.00018039201910917518,
+      "loss": 4.959,
+      "step": 42810
+    },
+    {
+      "epoch": 0.06016611721643889,
+      "grad_norm": 1.052660584449768,
+      "learning_rate": 0.00018043417170155961,
+      "loss": 4.995,
+      "step": 42820
+    },
+    {
+      "epoch": 0.060180168154602476,
+      "grad_norm": 1.0248833894729614,
+      "learning_rate": 0.00018047632429394407,
+      "loss": 4.8879,
+      "step": 42830
+    },
+    {
+      "epoch": 0.06019421909276605,
+      "grad_norm": 1.013227939605713,
+      "learning_rate": 0.0001805184768863285,
+      "loss": 4.9213,
+      "step": 42840
+    },
+    {
+      "epoch": 0.06020827003092963,
+      "grad_norm": 1.00045645236969,
+      "learning_rate": 0.00018056062947871294,
+      "loss": 5.0058,
+      "step": 42850
+    },
+    {
+      "epoch": 0.060222320969093204,
+      "grad_norm": 0.959171712398529,
+      "learning_rate": 0.00018060278207109734,
+      "loss": 4.9514,
+      "step": 42860
+    },
+    {
+      "epoch": 0.06023637190725678,
+      "grad_norm": 1.0098928213119507,
+      "learning_rate": 0.00018064493466348177,
+      "loss": 4.9365,
+      "step": 42870
+    },
+    {
+      "epoch": 0.060250422845420364,
+      "grad_norm": 1.0477204322814941,
+      "learning_rate": 0.0001806870872558662,
+      "loss": 4.9944,
+      "step": 42880
+    },
+    {
+      "epoch": 0.06026447378358394,
+      "grad_norm": 1.0132365226745605,
+      "learning_rate": 0.00018072923984825066,
+      "loss": 4.8793,
+      "step": 42890
+    },
+    {
+      "epoch": 0.060278524721747516,
+      "grad_norm": 1.0186618566513062,
+      "learning_rate": 0.0001807713924406351,
+      "loss": 4.9529,
+      "step": 42900
+    },
+    {
+      "epoch": 0.06029257565991109,
+      "grad_norm": 1.0929832458496094,
+      "learning_rate": 0.00018081354503301952,
+      "loss": 4.9407,
+      "step": 42910
+    },
+    {
+      "epoch": 0.06030662659807467,
+      "grad_norm": 1.0314265489578247,
+      "learning_rate": 0.00018085569762540395,
+      "loss": 4.819,
+      "step": 42920
+    },
+    {
+      "epoch": 0.060320677536238244,
+      "grad_norm": 0.9774079918861389,
+      "learning_rate": 0.00018089785021778836,
+      "loss": 4.9518,
+      "step": 42930
+    },
+    {
+      "epoch": 0.06033472847440183,
+      "grad_norm": 0.9911931157112122,
+      "learning_rate": 0.0001809400028101728,
+      "loss": 5.0276,
+      "step": 42940
+    },
+    {
+      "epoch": 0.060348779412565404,
+      "grad_norm": 1.0376858711242676,
+      "learning_rate": 0.00018098215540255725,
+      "loss": 4.8859,
+      "step": 42950
+    },
+    {
+      "epoch": 0.06036283035072898,
+      "grad_norm": 0.9969538450241089,
+      "learning_rate": 0.00018102430799494168,
+      "loss": 4.9157,
+      "step": 42960
+    },
+    {
+      "epoch": 0.060376881288892556,
+      "grad_norm": 1.0148366689682007,
+      "learning_rate": 0.0001810664605873261,
+      "loss": 4.942,
+      "step": 42970
+    },
+    {
+      "epoch": 0.06039093222705613,
+      "grad_norm": 0.9672243595123291,
+      "learning_rate": 0.00018110861317971054,
+      "loss": 4.9629,
+      "step": 42980
+    },
+    {
+      "epoch": 0.060404983165219715,
+      "grad_norm": 1.0580873489379883,
+      "learning_rate": 0.00018115076577209494,
+      "loss": 4.8622,
+      "step": 42990
+    },
+    {
+      "epoch": 0.06041903410338329,
+      "grad_norm": 1.0431230068206787,
+      "learning_rate": 0.0001811929183644794,
+      "loss": 4.8916,
+      "step": 43000
+    },
+    {
+      "epoch": 0.06043308504154687,
+      "grad_norm": 1.0647096633911133,
+      "learning_rate": 0.00018123507095686383,
+      "loss": 4.9537,
+      "step": 43010
+    },
+    {
+      "epoch": 0.060447135979710444,
+      "grad_norm": 1.0209344625473022,
+      "learning_rate": 0.00018127722354924826,
+      "loss": 4.8759,
+      "step": 43020
+    },
+    {
+      "epoch": 0.06046118691787402,
+      "grad_norm": 1.0193747282028198,
+      "learning_rate": 0.0001813193761416327,
+      "loss": 4.8229,
+      "step": 43030
+    },
+    {
+      "epoch": 0.0604752378560376,
+      "grad_norm": 1.0239225625991821,
+      "learning_rate": 0.00018136152873401713,
+      "loss": 4.9135,
+      "step": 43040
+    },
+    {
+      "epoch": 0.06048928879420118,
+      "grad_norm": 1.0313153266906738,
+      "learning_rate": 0.00018140368132640159,
+      "loss": 4.9518,
+      "step": 43050
+    },
+    {
+      "epoch": 0.060503339732364755,
+      "grad_norm": 1.0050175189971924,
+      "learning_rate": 0.000181445833918786,
+      "loss": 4.9435,
+      "step": 43060
+    },
+    {
+      "epoch": 0.06051739067052833,
+      "grad_norm": 0.9589523077011108,
+      "learning_rate": 0.00018148798651117042,
+      "loss": 4.8134,
+      "step": 43070
+    },
+    {
+      "epoch": 0.06053144160869191,
+      "grad_norm": 0.9744222164154053,
+      "learning_rate": 0.00018153013910355485,
+      "loss": 4.9269,
+      "step": 43080
+    },
+    {
+      "epoch": 0.06054549254685549,
+      "grad_norm": 0.9929285049438477,
+      "learning_rate": 0.00018157229169593928,
+      "loss": 4.9625,
+      "step": 43090
+    },
+    {
+      "epoch": 0.06055954348501907,
+      "grad_norm": 1.0369280576705933,
+      "learning_rate": 0.00018161444428832371,
+      "loss": 4.9153,
+      "step": 43100
+    },
+    {
+      "epoch": 0.06057359442318264,
+      "grad_norm": 0.994658350944519,
+      "learning_rate": 0.00018165659688070817,
+      "loss": 4.9614,
+      "step": 43110
+    },
+    {
+      "epoch": 0.06058764536134622,
+      "grad_norm": 1.1135982275009155,
+      "learning_rate": 0.00018169874947309258,
+      "loss": 4.897,
+      "step": 43120
+    },
+    {
+      "epoch": 0.060601696299509795,
+      "grad_norm": 0.9817977547645569,
+      "learning_rate": 0.000181740902065477,
+      "loss": 4.9774,
+      "step": 43130
+    },
+    {
+      "epoch": 0.06061574723767338,
+      "grad_norm": 1.008264422416687,
+      "learning_rate": 0.00018178305465786144,
+      "loss": 4.8703,
+      "step": 43140
+    },
+    {
+      "epoch": 0.060629798175836955,
+      "grad_norm": 1.0090116262435913,
+      "learning_rate": 0.00018182520725024587,
+      "loss": 4.7993,
+      "step": 43150
+    },
+    {
+      "epoch": 0.06064384911400053,
+      "grad_norm": 1.018884301185608,
+      "learning_rate": 0.0001818673598426303,
+      "loss": 5.0122,
+      "step": 43160
+    },
+    {
+      "epoch": 0.06065790005216411,
+      "grad_norm": 1.0300415754318237,
+      "learning_rate": 0.00018190951243501476,
+      "loss": 4.8523,
+      "step": 43170
+    },
+    {
+      "epoch": 0.06067195099032768,
+      "grad_norm": 1.0521844625473022,
+      "learning_rate": 0.0001819516650273992,
+      "loss": 4.8799,
+      "step": 43180
+    },
+    {
+      "epoch": 0.060686001928491266,
+      "grad_norm": 1.026895523071289,
+      "learning_rate": 0.0001819938176197836,
+      "loss": 4.9328,
+      "step": 43190
+    },
+    {
+      "epoch": 0.06070005286665484,
+      "grad_norm": 1.0086530447006226,
+      "learning_rate": 0.00018203597021216803,
+      "loss": 4.8162,
+      "step": 43200
+    },
+    {
+      "epoch": 0.06071410380481842,
+      "grad_norm": 1.0305287837982178,
+      "learning_rate": 0.00018207812280455246,
+      "loss": 4.858,
+      "step": 43210
+    },
+    {
+      "epoch": 0.060728154742981995,
+      "grad_norm": 0.997772216796875,
+      "learning_rate": 0.0001821202753969369,
+      "loss": 4.9303,
+      "step": 43220
+    },
+    {
+      "epoch": 0.06074220568114557,
+      "grad_norm": 1.0172392129898071,
+      "learning_rate": 0.00018216242798932135,
+      "loss": 4.9411,
+      "step": 43230
+    },
+    {
+      "epoch": 0.06075625661930915,
+      "grad_norm": 0.9829245805740356,
+      "learning_rate": 0.00018220458058170578,
+      "loss": 5.0541,
+      "step": 43240
+    },
+    {
+      "epoch": 0.06077030755747273,
+      "grad_norm": 0.9785235524177551,
+      "learning_rate": 0.00018224673317409018,
+      "loss": 4.953,
+      "step": 43250
+    },
+    {
+      "epoch": 0.060784358495636306,
+      "grad_norm": 0.9770336747169495,
+      "learning_rate": 0.0001822888857664746,
+      "loss": 5.0195,
+      "step": 43260
+    },
+    {
+      "epoch": 0.06079840943379988,
+      "grad_norm": 0.994461178779602,
+      "learning_rate": 0.00018233103835885904,
+      "loss": 4.8224,
+      "step": 43270
+    },
+    {
+      "epoch": 0.06081246037196346,
+      "grad_norm": 0.9978111982345581,
+      "learning_rate": 0.00018237319095124348,
+      "loss": 4.8499,
+      "step": 43280
+    },
+    {
+      "epoch": 0.060826511310127035,
+      "grad_norm": 1.0382758378982544,
+      "learning_rate": 0.00018241534354362793,
+      "loss": 4.9601,
+      "step": 43290
+    },
+    {
+      "epoch": 0.06084056224829062,
+      "grad_norm": 1.0007946491241455,
+      "learning_rate": 0.00018245749613601236,
+      "loss": 4.8906,
+      "step": 43300
+    },
+    {
+      "epoch": 0.060854613186454194,
+      "grad_norm": 0.9689368009567261,
+      "learning_rate": 0.0001824996487283968,
+      "loss": 4.9713,
+      "step": 43310
+    },
+    {
+      "epoch": 0.06086866412461777,
+      "grad_norm": 0.9696866273880005,
+      "learning_rate": 0.0001825418013207812,
+      "loss": 4.9007,
+      "step": 43320
+    },
+    {
+      "epoch": 0.06088271506278135,
+      "grad_norm": 0.9870903491973877,
+      "learning_rate": 0.00018258395391316563,
+      "loss": 4.9312,
+      "step": 43330
+    },
+    {
+      "epoch": 0.06089676600094492,
+      "grad_norm": 1.0196123123168945,
+      "learning_rate": 0.00018262610650555006,
+      "loss": 4.7852,
+      "step": 43340
+    },
+    {
+      "epoch": 0.060910816939108506,
+      "grad_norm": 0.9788631200790405,
+      "learning_rate": 0.00018266825909793452,
+      "loss": 4.8479,
+      "step": 43350
+    },
+    {
+      "epoch": 0.06092486787727208,
+      "grad_norm": 0.9982026219367981,
+      "learning_rate": 0.00018271041169031895,
+      "loss": 4.9585,
+      "step": 43360
+    },
+    {
+      "epoch": 0.06093891881543566,
+      "grad_norm": 1.0699291229248047,
+      "learning_rate": 0.00018275256428270338,
+      "loss": 4.9751,
+      "step": 43370
+    },
+    {
+      "epoch": 0.060952969753599234,
+      "grad_norm": 1.0497549772262573,
+      "learning_rate": 0.0001827947168750878,
+      "loss": 4.9279,
+      "step": 43380
+    },
+    {
+      "epoch": 0.06096702069176281,
+      "grad_norm": 0.9782308340072632,
+      "learning_rate": 0.00018283686946747222,
+      "loss": 4.9041,
+      "step": 43390
+    },
+    {
+      "epoch": 0.060981071629926394,
+      "grad_norm": 0.964344322681427,
+      "learning_rate": 0.00018287902205985665,
+      "loss": 4.9652,
+      "step": 43400
+    },
+    {
+      "epoch": 0.06099512256808997,
+      "grad_norm": 0.9865180253982544,
+      "learning_rate": 0.0001829211746522411,
+      "loss": 5.0092,
+      "step": 43410
+    },
+    {
+      "epoch": 0.061009173506253546,
+      "grad_norm": 1.071978211402893,
+      "learning_rate": 0.00018296332724462554,
+      "loss": 4.8976,
+      "step": 43420
+    },
+    {
+      "epoch": 0.06102322444441712,
+      "grad_norm": 0.9803053736686707,
+      "learning_rate": 0.00018300547983700997,
+      "loss": 4.8913,
+      "step": 43430
+    },
+    {
+      "epoch": 0.0610372753825807,
+      "grad_norm": 0.9943988919258118,
+      "learning_rate": 0.00018304763242939437,
+      "loss": 5.0052,
+      "step": 43440
+    },
+    {
+      "epoch": 0.06105132632074428,
+      "grad_norm": 0.9977701306343079,
+      "learning_rate": 0.0001830897850217788,
+      "loss": 4.8374,
+      "step": 43450
+    },
+    {
+      "epoch": 0.06106537725890786,
+      "grad_norm": 1.0006359815597534,
+      "learning_rate": 0.00018313193761416324,
+      "loss": 4.9586,
+      "step": 43460
+    },
+    {
+      "epoch": 0.061079428197071434,
+      "grad_norm": 1.0693806409835815,
+      "learning_rate": 0.0001831740902065477,
+      "loss": 4.932,
+      "step": 43470
+    },
+    {
+      "epoch": 0.06109347913523501,
+      "grad_norm": 0.9895679950714111,
+      "learning_rate": 0.00018321624279893213,
+      "loss": 4.9462,
+      "step": 43480
+    },
+    {
+      "epoch": 0.061107530073398586,
+      "grad_norm": 1.0478029251098633,
+      "learning_rate": 0.00018325839539131656,
+      "loss": 4.9418,
+      "step": 43490
+    },
+    {
+      "epoch": 0.06112158101156217,
+      "grad_norm": 0.9845918416976929,
+      "learning_rate": 0.000183300547983701,
+      "loss": 4.9005,
+      "step": 43500
+    },
+    {
+      "epoch": 0.061135631949725745,
+      "grad_norm": 1.0808178186416626,
+      "learning_rate": 0.0001833427005760854,
+      "loss": 4.8593,
+      "step": 43510
+    },
+    {
+      "epoch": 0.06114968288788932,
+      "grad_norm": 0.9863511919975281,
+      "learning_rate": 0.00018338485316846982,
+      "loss": 4.9801,
+      "step": 43520
+    },
+    {
+      "epoch": 0.0611637338260529,
+      "grad_norm": 1.0400804281234741,
+      "learning_rate": 0.00018342700576085428,
+      "loss": 4.8457,
+      "step": 43530
+    },
+    {
+      "epoch": 0.061177784764216474,
+      "grad_norm": 1.01461660861969,
+      "learning_rate": 0.0001834691583532387,
+      "loss": 4.969,
+      "step": 43540
+    },
+    {
+      "epoch": 0.06119183570238005,
+      "grad_norm": 0.9691588282585144,
+      "learning_rate": 0.00018351131094562314,
+      "loss": 4.8088,
+      "step": 43550
+    },
+    {
+      "epoch": 0.06120588664054363,
+      "grad_norm": 1.038812279701233,
+      "learning_rate": 0.00018355346353800757,
+      "loss": 4.8867,
+      "step": 43560
+    },
+    {
+      "epoch": 0.06121993757870721,
+      "grad_norm": 0.9811944365501404,
+      "learning_rate": 0.00018359561613039198,
+      "loss": 4.9723,
+      "step": 43570
+    },
+    {
+      "epoch": 0.061233988516870785,
+      "grad_norm": 1.0180128812789917,
+      "learning_rate": 0.00018363776872277644,
+      "loss": 4.6997,
+      "step": 43580
+    },
+    {
+      "epoch": 0.06124803945503436,
+      "grad_norm": 1.078423023223877,
+      "learning_rate": 0.00018367992131516087,
+      "loss": 4.8831,
+      "step": 43590
+    },
+    {
+      "epoch": 0.06126209039319794,
+      "grad_norm": 0.9817917346954346,
+      "learning_rate": 0.0001837220739075453,
+      "loss": 4.9102,
+      "step": 43600
+    },
+    {
+      "epoch": 0.06127614133136152,
+      "grad_norm": 0.977243959903717,
+      "learning_rate": 0.00018376422649992973,
+      "loss": 4.883,
+      "step": 43610
+    },
+    {
+      "epoch": 0.0612901922695251,
+      "grad_norm": 1.1097636222839355,
+      "learning_rate": 0.00018380637909231416,
+      "loss": 4.8,
+      "step": 43620
+    },
+    {
+      "epoch": 0.06130424320768867,
+      "grad_norm": 0.9828046560287476,
+      "learning_rate": 0.00018384853168469862,
+      "loss": 4.8724,
+      "step": 43630
+    },
+    {
+      "epoch": 0.06131829414585225,
+      "grad_norm": 0.9683628678321838,
+      "learning_rate": 0.00018389068427708302,
+      "loss": 4.9274,
+      "step": 43640
+    },
+    {
+      "epoch": 0.061332345084015825,
+      "grad_norm": 0.989050030708313,
+      "learning_rate": 0.00018393283686946746,
+      "loss": 4.9037,
+      "step": 43650
+    },
+    {
+      "epoch": 0.06134639602217941,
+      "grad_norm": 1.0087295770645142,
+      "learning_rate": 0.0001839749894618519,
+      "loss": 4.9268,
+      "step": 43660
+    },
+    {
+      "epoch": 0.061360446960342985,
+      "grad_norm": 1.0443772077560425,
+      "learning_rate": 0.00018401714205423632,
+      "loss": 4.9625,
+      "step": 43670
+    },
+    {
+      "epoch": 0.06137449789850656,
+      "grad_norm": 1.0253444910049438,
+      "learning_rate": 0.00018405929464662075,
+      "loss": 4.7028,
+      "step": 43680
+    },
+    {
+      "epoch": 0.06138854883667014,
+      "grad_norm": 1.0143331289291382,
+      "learning_rate": 0.0001841014472390052,
+      "loss": 4.8486,
+      "step": 43690
+    },
+    {
+      "epoch": 0.06140259977483371,
+      "grad_norm": 0.9996601939201355,
+      "learning_rate": 0.0001841435998313896,
+      "loss": 4.9462,
+      "step": 43700
+    },
+    {
+      "epoch": 0.061416650712997296,
+      "grad_norm": 0.9822000861167908,
+      "learning_rate": 0.00018418575242377404,
+      "loss": 4.9844,
+      "step": 43710
+    },
+    {
+      "epoch": 0.06143070165116087,
+      "grad_norm": 1.0107418298721313,
+      "learning_rate": 0.00018422790501615847,
+      "loss": 5.0152,
+      "step": 43720
+    },
+    {
+      "epoch": 0.06144475258932445,
+      "grad_norm": 1.0787968635559082,
+      "learning_rate": 0.0001842700576085429,
+      "loss": 4.9467,
+      "step": 43730
+    },
+    {
+      "epoch": 0.061458803527488025,
+      "grad_norm": 1.0120429992675781,
+      "learning_rate": 0.00018431221020092734,
+      "loss": 4.8599,
+      "step": 43740
+    },
+    {
+      "epoch": 0.0614728544656516,
+      "grad_norm": 1.002541422843933,
+      "learning_rate": 0.0001843543627933118,
+      "loss": 5.0134,
+      "step": 43750
+    },
+    {
+      "epoch": 0.061486905403815184,
+      "grad_norm": 1.0086458921432495,
+      "learning_rate": 0.00018439651538569623,
+      "loss": 4.8738,
+      "step": 43760
+    },
+    {
+      "epoch": 0.06150095634197876,
+      "grad_norm": 1.0043725967407227,
+      "learning_rate": 0.00018443866797808063,
+      "loss": 4.9894,
+      "step": 43770
+    },
+    {
+      "epoch": 0.061515007280142336,
+      "grad_norm": 0.9925393462181091,
+      "learning_rate": 0.00018448082057046506,
+      "loss": 4.9451,
+      "step": 43780
+    },
+    {
+      "epoch": 0.06152905821830591,
+      "grad_norm": 0.9828112721443176,
+      "learning_rate": 0.0001845229731628495,
+      "loss": 4.8894,
+      "step": 43790
+    },
+    {
+      "epoch": 0.06154310915646949,
+      "grad_norm": 1.0312516689300537,
+      "learning_rate": 0.00018456512575523392,
+      "loss": 4.8535,
+      "step": 43800
+    },
+    {
+      "epoch": 0.06155716009463307,
+      "grad_norm": 1.0502887964248657,
+      "learning_rate": 0.00018460727834761838,
+      "loss": 4.8503,
+      "step": 43810
+    },
+    {
+      "epoch": 0.06157121103279665,
+      "grad_norm": 0.9553511738777161,
+      "learning_rate": 0.0001846494309400028,
+      "loss": 4.931,
+      "step": 43820
+    },
+    {
+      "epoch": 0.061585261970960224,
+      "grad_norm": 1.0687919855117798,
+      "learning_rate": 0.00018469158353238722,
+      "loss": 4.858,
+      "step": 43830
+    },
+    {
+      "epoch": 0.0615993129091238,
+      "grad_norm": 0.9931618571281433,
+      "learning_rate": 0.00018473373612477165,
+      "loss": 4.9647,
+      "step": 43840
+    },
+    {
+      "epoch": 0.061613363847287377,
+      "grad_norm": 1.0885789394378662,
+      "learning_rate": 0.00018477588871715608,
+      "loss": 4.8131,
+      "step": 43850
+    },
+    {
+      "epoch": 0.06162741478545095,
+      "grad_norm": 0.9883633255958557,
+      "learning_rate": 0.0001848180413095405,
+      "loss": 4.8692,
+      "step": 43860
+    },
+    {
+      "epoch": 0.061641465723614536,
+      "grad_norm": 0.9546381235122681,
+      "learning_rate": 0.00018486019390192497,
+      "loss": 4.9282,
+      "step": 43870
+    },
+    {
+      "epoch": 0.06165551666177811,
+      "grad_norm": 0.9708349704742432,
+      "learning_rate": 0.0001849023464943094,
+      "loss": 4.9329,
+      "step": 43880
+    },
+    {
+      "epoch": 0.06166956759994169,
+      "grad_norm": 0.985491156578064,
+      "learning_rate": 0.00018494449908669383,
+      "loss": 5.066,
+      "step": 43890
+    },
+    {
+      "epoch": 0.061683618538105264,
+      "grad_norm": 1.0057768821716309,
+      "learning_rate": 0.00018498665167907823,
+      "loss": 5.0658,
+      "step": 43900
+    },
+    {
+      "epoch": 0.06169766947626884,
+      "grad_norm": 1.0529582500457764,
+      "learning_rate": 0.00018502880427146267,
+      "loss": 4.9211,
+      "step": 43910
+    },
+    {
+      "epoch": 0.061711720414432424,
+      "grad_norm": 0.9484959840774536,
+      "learning_rate": 0.0001850709568638471,
+      "loss": 4.8917,
+      "step": 43920
+    },
+    {
+      "epoch": 0.061725771352596,
+      "grad_norm": 0.9802437424659729,
+      "learning_rate": 0.00018511310945623156,
+      "loss": 4.853,
+      "step": 43930
+    },
+    {
+      "epoch": 0.061739822290759576,
+      "grad_norm": 1.045055627822876,
+      "learning_rate": 0.00018515526204861599,
+      "loss": 4.8402,
+      "step": 43940
+    },
+    {
+      "epoch": 0.06175387322892315,
+      "grad_norm": 1.073882818222046,
+      "learning_rate": 0.00018519741464100042,
+      "loss": 4.9227,
+      "step": 43950
+    },
+    {
+      "epoch": 0.06176792416708673,
+      "grad_norm": 0.9608358144760132,
+      "learning_rate": 0.00018523956723338482,
+      "loss": 4.9701,
+      "step": 43960
+    },
+    {
+      "epoch": 0.06178197510525031,
+      "grad_norm": 0.9510095119476318,
+      "learning_rate": 0.00018528171982576925,
+      "loss": 4.9003,
+      "step": 43970
+    },
+    {
+      "epoch": 0.06179602604341389,
+      "grad_norm": 1.0726205110549927,
+      "learning_rate": 0.00018532387241815368,
+      "loss": 4.9263,
+      "step": 43980
+    },
+    {
+      "epoch": 0.061810076981577464,
+      "grad_norm": 0.9834354519844055,
+      "learning_rate": 0.00018536602501053814,
+      "loss": 5.0427,
+      "step": 43990
+    },
+    {
+      "epoch": 0.06182412791974104,
+      "grad_norm": 0.9882546067237854,
+      "learning_rate": 0.00018540817760292257,
+      "loss": 5.026,
+      "step": 44000
+    },
+    {
+      "epoch": 0.061838178857904616,
+      "grad_norm": 0.9732726216316223,
+      "learning_rate": 0.000185450330195307,
+      "loss": 4.9425,
+      "step": 44010
+    },
+    {
+      "epoch": 0.0618522297960682,
+      "grad_norm": 0.9791545271873474,
+      "learning_rate": 0.00018549248278769144,
+      "loss": 4.965,
+      "step": 44020
+    },
+    {
+      "epoch": 0.061866280734231775,
+      "grad_norm": 1.0380457639694214,
+      "learning_rate": 0.00018553463538007584,
+      "loss": 4.9099,
+      "step": 44030
+    },
+    {
+      "epoch": 0.06188033167239535,
+      "grad_norm": 1.010048747062683,
+      "learning_rate": 0.00018557678797246027,
+      "loss": 4.9624,
+      "step": 44040
+    },
+    {
+      "epoch": 0.06189438261055893,
+      "grad_norm": 0.9912928342819214,
+      "learning_rate": 0.00018561894056484473,
+      "loss": 4.9728,
+      "step": 44050
+    },
+    {
+      "epoch": 0.061908433548722504,
+      "grad_norm": 0.9759913682937622,
+      "learning_rate": 0.00018566109315722916,
+      "loss": 4.9517,
+      "step": 44060
+    },
+    {
+      "epoch": 0.06192248448688609,
+      "grad_norm": 0.997549295425415,
+      "learning_rate": 0.0001857032457496136,
+      "loss": 4.8432,
+      "step": 44070
+    },
+    {
+      "epoch": 0.06193653542504966,
+      "grad_norm": 1.0062215328216553,
+      "learning_rate": 0.00018574539834199802,
+      "loss": 4.9373,
+      "step": 44080
+    },
+    {
+      "epoch": 0.06195058636321324,
+      "grad_norm": 1.012075424194336,
+      "learning_rate": 0.00018578755093438243,
+      "loss": 4.8813,
+      "step": 44090
+    },
+    {
+      "epoch": 0.061964637301376815,
+      "grad_norm": 0.95697021484375,
+      "learning_rate": 0.00018582970352676686,
+      "loss": 4.944,
+      "step": 44100
+    },
+    {
+      "epoch": 0.06197868823954039,
+      "grad_norm": 0.9985111951828003,
+      "learning_rate": 0.00018587185611915132,
+      "loss": 4.8227,
+      "step": 44110
+    },
+    {
+      "epoch": 0.061992739177703975,
+      "grad_norm": 0.9926140308380127,
+      "learning_rate": 0.00018591400871153575,
+      "loss": 5.0079,
+      "step": 44120
+    },
+    {
+      "epoch": 0.06200679011586755,
+      "grad_norm": 1.1736414432525635,
+      "learning_rate": 0.00018595616130392018,
+      "loss": 4.7146,
+      "step": 44130
+    },
+    {
+      "epoch": 0.06202084105403113,
+      "grad_norm": 1.0231581926345825,
+      "learning_rate": 0.0001859983138963046,
+      "loss": 5.0206,
+      "step": 44140
+    },
+    {
+      "epoch": 0.0620348919921947,
+      "grad_norm": 1.0800116062164307,
+      "learning_rate": 0.00018604046648868901,
+      "loss": 4.8657,
+      "step": 44150
+    },
+    {
+      "epoch": 0.06204894293035828,
+      "grad_norm": 0.9866999387741089,
+      "learning_rate": 0.00018608261908107347,
+      "loss": 4.9722,
+      "step": 44160
+    },
+    {
+      "epoch": 0.062062993868521855,
+      "grad_norm": 1.0231711864471436,
+      "learning_rate": 0.0001861247716734579,
+      "loss": 4.8777,
+      "step": 44170
+    },
+    {
+      "epoch": 0.06207704480668544,
+      "grad_norm": 0.9434810876846313,
+      "learning_rate": 0.00018616692426584233,
+      "loss": 4.9722,
+      "step": 44180
+    },
+    {
+      "epoch": 0.062091095744849015,
+      "grad_norm": 1.0458160638809204,
+      "learning_rate": 0.00018620907685822677,
+      "loss": 4.8806,
+      "step": 44190
+    },
+    {
+      "epoch": 0.06210514668301259,
+      "grad_norm": 0.9809548258781433,
+      "learning_rate": 0.0001862512294506112,
+      "loss": 4.7962,
+      "step": 44200
+    },
+    {
+      "epoch": 0.06211919762117617,
+      "grad_norm": 1.012259840965271,
+      "learning_rate": 0.00018629338204299565,
+      "loss": 4.9335,
+      "step": 44210
+    },
+    {
+      "epoch": 0.06213324855933974,
+      "grad_norm": 0.9664605855941772,
+      "learning_rate": 0.00018633553463538006,
+      "loss": 4.9491,
+      "step": 44220
+    },
+    {
+      "epoch": 0.062147299497503326,
+      "grad_norm": 0.9853652715682983,
+      "learning_rate": 0.0001863776872277645,
+      "loss": 4.9029,
+      "step": 44230
+    },
+    {
+      "epoch": 0.0621613504356669,
+      "grad_norm": 1.0576822757720947,
+      "learning_rate": 0.00018641983982014892,
+      "loss": 4.8949,
+      "step": 44240
+    },
+    {
+      "epoch": 0.06217540137383048,
+      "grad_norm": 1.0025372505187988,
+      "learning_rate": 0.00018646199241253335,
+      "loss": 4.9004,
+      "step": 44250
+    },
+    {
+      "epoch": 0.062189452311994055,
+      "grad_norm": 1.0175806283950806,
+      "learning_rate": 0.00018650414500491778,
+      "loss": 5.03,
+      "step": 44260
+    },
+    {
+      "epoch": 0.06220350325015763,
+      "grad_norm": 1.0149425268173218,
+      "learning_rate": 0.00018654629759730224,
+      "loss": 4.9339,
+      "step": 44270
+    },
+    {
+      "epoch": 0.062217554188321214,
+      "grad_norm": 1.0029298067092896,
+      "learning_rate": 0.00018658845018968665,
+      "loss": 4.8234,
+      "step": 44280
+    },
+    {
+      "epoch": 0.06223160512648479,
+      "grad_norm": 0.9579691290855408,
+      "learning_rate": 0.00018663060278207108,
+      "loss": 5.1016,
+      "step": 44290
+    },
+    {
+      "epoch": 0.062245656064648366,
+      "grad_norm": 1.061834454536438,
+      "learning_rate": 0.0001866727553744555,
+      "loss": 4.9704,
+      "step": 44300
+    },
+    {
+      "epoch": 0.06225970700281194,
+      "grad_norm": 1.004465103149414,
+      "learning_rate": 0.00018671490796683994,
+      "loss": 4.9067,
+      "step": 44310
+    },
+    {
+      "epoch": 0.06227375794097552,
+      "grad_norm": 0.9837499856948853,
+      "learning_rate": 0.00018675706055922437,
+      "loss": 4.9309,
+      "step": 44320
+    },
+    {
+      "epoch": 0.0622878088791391,
+      "grad_norm": 1.000868558883667,
+      "learning_rate": 0.00018679921315160883,
+      "loss": 4.8866,
+      "step": 44330
+    },
+    {
+      "epoch": 0.06230185981730268,
+      "grad_norm": 0.9788700938224792,
+      "learning_rate": 0.00018684136574399326,
+      "loss": 4.9327,
+      "step": 44340
+    },
+    {
+      "epoch": 0.062315910755466254,
+      "grad_norm": 0.9754438400268555,
+      "learning_rate": 0.00018688351833637766,
+      "loss": 4.8594,
+      "step": 44350
+    },
+    {
+      "epoch": 0.06232996169362983,
+      "grad_norm": 0.951196014881134,
+      "learning_rate": 0.0001869256709287621,
+      "loss": 4.9354,
+      "step": 44360
+    },
+    {
+      "epoch": 0.062344012631793407,
+      "grad_norm": 0.9627026915550232,
+      "learning_rate": 0.00018696782352114653,
+      "loss": 4.9621,
+      "step": 44370
+    },
+    {
+      "epoch": 0.06235806356995699,
+      "grad_norm": 1.063202977180481,
+      "learning_rate": 0.00018700997611353096,
+      "loss": 4.8807,
+      "step": 44380
+    },
+    {
+      "epoch": 0.062372114508120566,
+      "grad_norm": 0.9841533899307251,
+      "learning_rate": 0.00018705212870591542,
+      "loss": 4.9947,
+      "step": 44390
+    },
+    {
+      "epoch": 0.06238616544628414,
+      "grad_norm": 0.9734062552452087,
+      "learning_rate": 0.00018709428129829985,
+      "loss": 4.9473,
+      "step": 44400
+    },
+    {
+      "epoch": 0.06240021638444772,
+      "grad_norm": 0.9729593396186829,
+      "learning_rate": 0.00018713643389068425,
+      "loss": 5.1047,
+      "step": 44410
+    },
+    {
+      "epoch": 0.062414267322611294,
+      "grad_norm": 1.000727891921997,
+      "learning_rate": 0.00018717858648306868,
+      "loss": 4.9258,
+      "step": 44420
+    },
+    {
+      "epoch": 0.06242831826077488,
+      "grad_norm": 0.9832708835601807,
+      "learning_rate": 0.0001872207390754531,
+      "loss": 4.7362,
+      "step": 44430
+    },
+    {
+      "epoch": 0.062442369198938454,
+      "grad_norm": 1.075731635093689,
+      "learning_rate": 0.00018726289166783754,
+      "loss": 5.0231,
+      "step": 44440
+    },
+    {
+      "epoch": 0.06245642013710203,
+      "grad_norm": 0.965521514415741,
+      "learning_rate": 0.000187305044260222,
+      "loss": 4.8227,
+      "step": 44450
+    },
+    {
+      "epoch": 0.062470471075265606,
+      "grad_norm": 1.0088485479354858,
+      "learning_rate": 0.00018734719685260643,
+      "loss": 4.8802,
+      "step": 44460
+    },
+    {
+      "epoch": 0.06248452201342918,
+      "grad_norm": 0.9651148319244385,
+      "learning_rate": 0.00018738934944499087,
+      "loss": 4.8197,
+      "step": 44470
+    },
+    {
+      "epoch": 0.06249857295159276,
+      "grad_norm": 0.9971210360527039,
+      "learning_rate": 0.00018743150203737527,
+      "loss": 4.9242,
+      "step": 44480
+    },
+    {
+      "epoch": 0.06251262388975634,
+      "grad_norm": 0.9815917015075684,
+      "learning_rate": 0.0001874736546297597,
+      "loss": 4.8368,
+      "step": 44490
+    },
+    {
+      "epoch": 0.06252667482791992,
+      "grad_norm": 0.9839165806770325,
+      "learning_rate": 0.00018751580722214413,
+      "loss": 4.8778,
+      "step": 44500
+    },
+    {
+      "epoch": 0.0625407257660835,
+      "grad_norm": 0.990401029586792,
+      "learning_rate": 0.0001875579598145286,
+      "loss": 4.9428,
+      "step": 44510
+    },
+    {
+      "epoch": 0.06255477670424707,
+      "grad_norm": 0.9693237543106079,
+      "learning_rate": 0.00018760011240691302,
+      "loss": 4.8596,
+      "step": 44520
+    },
+    {
+      "epoch": 0.06256882764241065,
+      "grad_norm": 1.0013371706008911,
+      "learning_rate": 0.00018764226499929745,
+      "loss": 4.9533,
+      "step": 44530
+    },
+    {
+      "epoch": 0.06258287858057422,
+      "grad_norm": 0.9999648928642273,
+      "learning_rate": 0.00018768441759168186,
+      "loss": 4.9845,
+      "step": 44540
+    },
+    {
+      "epoch": 0.0625969295187378,
+      "grad_norm": 0.982397198677063,
+      "learning_rate": 0.0001877265701840663,
+      "loss": 4.8334,
+      "step": 44550
+    },
+    {
+      "epoch": 0.06261098045690139,
+      "grad_norm": 0.9798036217689514,
+      "learning_rate": 0.0001877645075172123,
+      "loss": 4.8915,
+      "step": 44560
+    },
+    {
+      "epoch": 0.06262503139506496,
+      "grad_norm": 0.9902676343917847,
+      "learning_rate": 0.0001878066601095967,
+      "loss": 4.9485,
+      "step": 44570
+    },
+    {
+      "epoch": 0.06263908233322854,
+      "grad_norm": 0.9916227459907532,
+      "learning_rate": 0.00018784881270198113,
+      "loss": 4.8654,
+      "step": 44580
+    },
+    {
+      "epoch": 0.06265313327139212,
+      "grad_norm": 1.0029773712158203,
+      "learning_rate": 0.0001878909652943656,
+      "loss": 4.7969,
+      "step": 44590
+    },
+    {
+      "epoch": 0.06266718420955569,
+      "grad_norm": 0.985836386680603,
+      "learning_rate": 0.00018793311788675002,
+      "loss": 4.9815,
+      "step": 44600
+    },
+    {
+      "epoch": 0.06268123514771927,
+      "grad_norm": 1.05781888961792,
+      "learning_rate": 0.00018797527047913445,
+      "loss": 4.9361,
+      "step": 44610
+    },
+    {
+      "epoch": 0.06269528608588285,
+      "grad_norm": 1.0255253314971924,
+      "learning_rate": 0.00018801742307151888,
+      "loss": 4.7969,
+      "step": 44620
+    },
+    {
+      "epoch": 0.06270933702404642,
+      "grad_norm": 0.9807506799697876,
+      "learning_rate": 0.00018805957566390332,
+      "loss": 4.8693,
+      "step": 44630
+    },
+    {
+      "epoch": 0.06272338796221,
+      "grad_norm": 0.9755240082740784,
+      "learning_rate": 0.00018810172825628772,
+      "loss": 4.8811,
+      "step": 44640
+    },
+    {
+      "epoch": 0.06273743890037357,
+      "grad_norm": 0.9982534050941467,
+      "learning_rate": 0.00018814388084867218,
+      "loss": 5.0296,
+      "step": 44650
+    },
+    {
+      "epoch": 0.06275148983853715,
+      "grad_norm": 0.9563882350921631,
+      "learning_rate": 0.0001881860334410566,
+      "loss": 4.9605,
+      "step": 44660
+    },
+    {
+      "epoch": 0.06276554077670074,
+      "grad_norm": 0.9405322074890137,
+      "learning_rate": 0.00018822818603344104,
+      "loss": 4.9641,
+      "step": 44670
+    },
+    {
+      "epoch": 0.06277959171486432,
+      "grad_norm": 0.9542915225028992,
+      "learning_rate": 0.00018827033862582547,
+      "loss": 4.99,
+      "step": 44680
+    },
+    {
+      "epoch": 0.06279364265302789,
+      "grad_norm": 0.9676512479782104,
+      "learning_rate": 0.0001883124912182099,
+      "loss": 4.9995,
+      "step": 44690
+    },
+    {
+      "epoch": 0.06280769359119147,
+      "grad_norm": 0.9878407120704651,
+      "learning_rate": 0.0001883546438105943,
+      "loss": 4.962,
+      "step": 44700
+    },
+    {
+      "epoch": 0.06282174452935504,
+      "grad_norm": 0.9911555647850037,
+      "learning_rate": 0.00018839679640297876,
+      "loss": 4.8163,
+      "step": 44710
+    },
+    {
+      "epoch": 0.06283579546751862,
+      "grad_norm": 1.0555427074432373,
+      "learning_rate": 0.0001884389489953632,
+      "loss": 5.0145,
+      "step": 44720
+    },
+    {
+      "epoch": 0.0628498464056822,
+      "grad_norm": 0.9912372827529907,
+      "learning_rate": 0.00018848110158774763,
+      "loss": 4.8577,
+      "step": 44730
+    },
+    {
+      "epoch": 0.06286389734384577,
+      "grad_norm": 0.9817824959754944,
+      "learning_rate": 0.00018852325418013206,
+      "loss": 4.8828,
+      "step": 44740
+    },
+    {
+      "epoch": 0.06287794828200935,
+      "grad_norm": 0.9970001578330994,
+      "learning_rate": 0.0001885654067725165,
+      "loss": 4.8374,
+      "step": 44750
+    },
+    {
+      "epoch": 0.06289199922017293,
+      "grad_norm": 1.0151931047439575,
+      "learning_rate": 0.00018860755936490095,
+      "loss": 4.9853,
+      "step": 44760
+    },
+    {
+      "epoch": 0.06290605015833652,
+      "grad_norm": 0.9986475110054016,
+      "learning_rate": 0.00018864971195728535,
+      "loss": 4.9118,
+      "step": 44770
+    },
+    {
+      "epoch": 0.06292010109650009,
+      "grad_norm": 1.008439064025879,
+      "learning_rate": 0.00018869186454966978,
+      "loss": 4.9462,
+      "step": 44780
+    },
+    {
+      "epoch": 0.06293415203466367,
+      "grad_norm": 1.03360116481781,
+      "learning_rate": 0.00018873401714205421,
+      "loss": 4.8674,
+      "step": 44790
+    },
+    {
+      "epoch": 0.06294820297282724,
+      "grad_norm": 0.9503140449523926,
+      "learning_rate": 0.00018877616973443865,
+      "loss": 4.9065,
+      "step": 44800
+    },
+    {
+      "epoch": 0.06296225391099082,
+      "grad_norm": 0.9909397959709167,
+      "learning_rate": 0.0001888183223268231,
+      "loss": 5.0051,
+      "step": 44810
+    },
+    {
+      "epoch": 0.0629763048491544,
+      "grad_norm": 0.9907909035682678,
+      "learning_rate": 0.00018886047491920753,
+      "loss": 4.8448,
+      "step": 44820
+    },
+    {
+      "epoch": 0.06299035578731797,
+      "grad_norm": 0.962094783782959,
+      "learning_rate": 0.00018890262751159194,
+      "loss": 4.962,
+      "step": 44830
+    },
+    {
+      "epoch": 0.06300440672548155,
+      "grad_norm": 0.9932377934455872,
+      "learning_rate": 0.00018894478010397637,
+      "loss": 5.0061,
+      "step": 44840
+    },
+    {
+      "epoch": 0.06301845766364512,
+      "grad_norm": 0.9481516480445862,
+      "learning_rate": 0.0001889869326963608,
+      "loss": 4.9518,
+      "step": 44850
+    },
+    {
+      "epoch": 0.0630325086018087,
+      "grad_norm": 0.9638423323631287,
+      "learning_rate": 0.00018902908528874523,
+      "loss": 4.9881,
+      "step": 44860
+    },
+    {
+      "epoch": 0.06304655953997229,
+      "grad_norm": 1.0749502182006836,
+      "learning_rate": 0.0001890712378811297,
+      "loss": 4.915,
+      "step": 44870
+    },
+    {
+      "epoch": 0.06306061047813587,
+      "grad_norm": 0.9810838103294373,
+      "learning_rate": 0.00018911339047351412,
+      "loss": 4.9,
+      "step": 44880
+    },
+    {
+      "epoch": 0.06307466141629944,
+      "grad_norm": 1.0337384939193726,
+      "learning_rate": 0.00018915554306589855,
+      "loss": 4.845,
+      "step": 44890
+    },
+    {
+      "epoch": 0.06308871235446302,
+      "grad_norm": 0.9709575772285461,
+      "learning_rate": 0.00018919769565828296,
+      "loss": 4.8708,
+      "step": 44900
+    },
+    {
+      "epoch": 0.0631027632926266,
+      "grad_norm": 1.0215866565704346,
+      "learning_rate": 0.0001892398482506674,
+      "loss": 4.939,
+      "step": 44910
+    },
+    {
+      "epoch": 0.06311681423079017,
+      "grad_norm": 0.9717178344726562,
+      "learning_rate": 0.00018928200084305182,
+      "loss": 4.9069,
+      "step": 44920
+    },
+    {
+      "epoch": 0.06313086516895375,
+      "grad_norm": 0.9674050807952881,
+      "learning_rate": 0.00018932415343543628,
+      "loss": 4.8942,
+      "step": 44930
+    },
+    {
+      "epoch": 0.06314491610711732,
+      "grad_norm": 1.011043906211853,
+      "learning_rate": 0.0001893663060278207,
+      "loss": 4.9006,
+      "step": 44940
+    },
+    {
+      "epoch": 0.0631589670452809,
+      "grad_norm": 0.9711986184120178,
+      "learning_rate": 0.00018940845862020514,
+      "loss": 4.8699,
+      "step": 44950
+    },
+    {
+      "epoch": 0.06317301798344448,
+      "grad_norm": 1.0319536924362183,
+      "learning_rate": 0.00018945061121258954,
+      "loss": 4.8679,
+      "step": 44960
+    },
+    {
+      "epoch": 0.06318706892160805,
+      "grad_norm": 1.0072388648986816,
+      "learning_rate": 0.00018949276380497398,
+      "loss": 4.9792,
+      "step": 44970
+    },
+    {
+      "epoch": 0.06320111985977164,
+      "grad_norm": 0.9475632905960083,
+      "learning_rate": 0.0001895349163973584,
+      "loss": 4.8833,
+      "step": 44980
+    },
+    {
+      "epoch": 0.06321517079793522,
+      "grad_norm": 1.02812922000885,
+      "learning_rate": 0.00018957706898974286,
+      "loss": 4.9171,
+      "step": 44990
+    },
+    {
+      "epoch": 0.0632292217360988,
+      "grad_norm": 0.9555014967918396,
+      "learning_rate": 0.0001896192215821273,
+      "loss": 4.9232,
+      "step": 45000
+    },
+    {
+      "epoch": 0.06324327267426237,
+      "grad_norm": 0.9676048755645752,
+      "learning_rate": 0.00018966137417451173,
+      "loss": 4.9209,
+      "step": 45010
+    },
+    {
+      "epoch": 0.06325732361242595,
+      "grad_norm": 1.054318904876709,
+      "learning_rate": 0.00018970352676689616,
+      "loss": 4.9587,
+      "step": 45020
+    },
+    {
+      "epoch": 0.06327137455058952,
+      "grad_norm": 0.9832106232643127,
+      "learning_rate": 0.00018974567935928056,
+      "loss": 4.8852,
+      "step": 45030
+    },
+    {
+      "epoch": 0.0632854254887531,
+      "grad_norm": 0.9964709281921387,
+      "learning_rate": 0.000189787831951665,
+      "loss": 4.8634,
+      "step": 45040
+    },
+    {
+      "epoch": 0.06329947642691668,
+      "grad_norm": 0.959078848361969,
+      "learning_rate": 0.00018982998454404945,
+      "loss": 4.9014,
+      "step": 45050
+    },
+    {
+      "epoch": 0.06331352736508025,
+      "grad_norm": 0.993674099445343,
+      "learning_rate": 0.00018987213713643388,
+      "loss": 4.7264,
+      "step": 45060
+    },
+    {
+      "epoch": 0.06332757830324383,
+      "grad_norm": 0.9697414636611938,
+      "learning_rate": 0.00018991428972881831,
+      "loss": 4.8757,
+      "step": 45070
+    },
+    {
+      "epoch": 0.06334162924140742,
+      "grad_norm": 0.9635653495788574,
+      "learning_rate": 0.00018995644232120275,
+      "loss": 4.9426,
+      "step": 45080
+    },
+    {
+      "epoch": 0.063355680179571,
+      "grad_norm": 0.9881270527839661,
+      "learning_rate": 0.00018999859491358715,
+      "loss": 4.9085,
+      "step": 45090
+    },
+    {
+      "epoch": 0.06336973111773457,
+      "grad_norm": 0.9849804639816284,
+      "learning_rate": 0.00019004074750597158,
+      "loss": 4.9223,
+      "step": 45100
+    },
+    {
+      "epoch": 0.06338378205589815,
+      "grad_norm": 0.953628659248352,
+      "learning_rate": 0.00019008290009835604,
+      "loss": 4.9461,
+      "step": 45110
+    },
+    {
+      "epoch": 0.06339783299406172,
+      "grad_norm": 0.9483082890510559,
+      "learning_rate": 0.00019012505269074047,
+      "loss": 4.9306,
+      "step": 45120
+    },
+    {
+      "epoch": 0.0634118839322253,
+      "grad_norm": 1.0407912731170654,
+      "learning_rate": 0.0001901672052831249,
+      "loss": 4.7811,
+      "step": 45130
+    },
+    {
+      "epoch": 0.06342593487038888,
+      "grad_norm": 0.9633585214614868,
+      "learning_rate": 0.00019020935787550933,
+      "loss": 4.8575,
+      "step": 45140
+    },
+    {
+      "epoch": 0.06343998580855245,
+      "grad_norm": 0.9573111534118652,
+      "learning_rate": 0.00019025151046789376,
+      "loss": 4.873,
+      "step": 45150
+    },
+    {
+      "epoch": 0.06345403674671603,
+      "grad_norm": 1.0147194862365723,
+      "learning_rate": 0.00019029366306027817,
+      "loss": 4.7867,
+      "step": 45160
+    },
+    {
+      "epoch": 0.0634680876848796,
+      "grad_norm": 0.9639556407928467,
+      "learning_rate": 0.00019033581565266263,
+      "loss": 4.8389,
+      "step": 45170
+    },
+    {
+      "epoch": 0.0634821386230432,
+      "grad_norm": 0.9909672141075134,
+      "learning_rate": 0.00019037796824504706,
+      "loss": 4.9519,
+      "step": 45180
+    },
+    {
+      "epoch": 0.06349618956120677,
+      "grad_norm": 0.9403594732284546,
+      "learning_rate": 0.0001904201208374315,
+      "loss": 5.012,
+      "step": 45190
+    },
+    {
+      "epoch": 0.06351024049937035,
+      "grad_norm": 1.0044161081314087,
+      "learning_rate": 0.00019046227342981592,
+      "loss": 4.8788,
+      "step": 45200
+    },
+    {
+      "epoch": 0.06352429143753392,
+      "grad_norm": 0.9971259832382202,
+      "learning_rate": 0.00019050442602220035,
+      "loss": 4.9838,
+      "step": 45210
+    },
+    {
+      "epoch": 0.0635383423756975,
+      "grad_norm": 0.9722863435745239,
+      "learning_rate": 0.00019054657861458475,
+      "loss": 4.9108,
+      "step": 45220
+    },
+    {
+      "epoch": 0.06355239331386107,
+      "grad_norm": 0.967170774936676,
+      "learning_rate": 0.0001905887312069692,
+      "loss": 4.8078,
+      "step": 45230
+    },
+    {
+      "epoch": 0.06356644425202465,
+      "grad_norm": 0.9585663676261902,
+      "learning_rate": 0.00019063088379935364,
+      "loss": 4.9564,
+      "step": 45240
+    },
+    {
+      "epoch": 0.06358049519018823,
+      "grad_norm": 1.025834083557129,
+      "learning_rate": 0.00019067303639173807,
+      "loss": 4.8575,
+      "step": 45250
+    },
+    {
+      "epoch": 0.0635945461283518,
+      "grad_norm": 0.9579908847808838,
+      "learning_rate": 0.0001907151889841225,
+      "loss": 4.8684,
+      "step": 45260
+    },
+    {
+      "epoch": 0.06360859706651538,
+      "grad_norm": 1.0075056552886963,
+      "learning_rate": 0.00019075734157650694,
+      "loss": 4.9019,
+      "step": 45270
+    },
+    {
+      "epoch": 0.06362264800467896,
+      "grad_norm": 0.9414964914321899,
+      "learning_rate": 0.00019079949416889134,
+      "loss": 4.8507,
+      "step": 45280
+    },
+    {
+      "epoch": 0.06363669894284255,
+      "grad_norm": 0.947604775428772,
+      "learning_rate": 0.0001908416467612758,
+      "loss": 4.9015,
+      "step": 45290
+    },
+    {
+      "epoch": 0.06365074988100612,
+      "grad_norm": 0.961848795413971,
+      "learning_rate": 0.00019088379935366023,
+      "loss": 4.9397,
+      "step": 45300
+    },
+    {
+      "epoch": 0.0636648008191697,
+      "grad_norm": 0.9270375370979309,
+      "learning_rate": 0.00019092595194604466,
+      "loss": 4.8831,
+      "step": 45310
+    },
+    {
+      "epoch": 0.06367885175733327,
+      "grad_norm": 0.9623255729675293,
+      "learning_rate": 0.0001909681045384291,
+      "loss": 4.9062,
+      "step": 45320
+    },
+    {
+      "epoch": 0.06369290269549685,
+      "grad_norm": 0.9620530605316162,
+      "learning_rate": 0.00019101025713081352,
+      "loss": 4.9139,
+      "step": 45330
+    },
+    {
+      "epoch": 0.06370695363366043,
+      "grad_norm": 0.9391373991966248,
+      "learning_rate": 0.00019105240972319798,
+      "loss": 4.9787,
+      "step": 45340
+    },
+    {
+      "epoch": 0.063721004571824,
+      "grad_norm": 1.0350898504257202,
+      "learning_rate": 0.0001910945623155824,
+      "loss": 4.7686,
+      "step": 45350
+    },
+    {
+      "epoch": 0.06373505550998758,
+      "grad_norm": 0.9722337126731873,
+      "learning_rate": 0.00019113671490796682,
+      "loss": 4.8961,
+      "step": 45360
+    },
+    {
+      "epoch": 0.06374910644815115,
+      "grad_norm": 0.9319054484367371,
+      "learning_rate": 0.00019117886750035125,
+      "loss": 5.0057,
+      "step": 45370
+    },
+    {
+      "epoch": 0.06376315738631473,
+      "grad_norm": 0.9770829081535339,
+      "learning_rate": 0.00019122102009273568,
+      "loss": 4.9013,
+      "step": 45380
+    },
+    {
+      "epoch": 0.06377720832447832,
+      "grad_norm": 0.9620832204818726,
+      "learning_rate": 0.00019126317268512014,
+      "loss": 4.974,
+      "step": 45390
+    },
+    {
+      "epoch": 0.0637912592626419,
+      "grad_norm": 1.0096343755722046,
+      "learning_rate": 0.00019130532527750457,
+      "loss": 4.8912,
+      "step": 45400
+    },
+    {
+      "epoch": 0.06380531020080547,
+      "grad_norm": 0.962091326713562,
+      "learning_rate": 0.00019134747786988897,
+      "loss": 4.9313,
+      "step": 45410
+    },
+    {
+      "epoch": 0.06381936113896905,
+      "grad_norm": 0.9614576697349548,
+      "learning_rate": 0.0001913896304622734,
+      "loss": 4.9008,
+      "step": 45420
+    },
+    {
+      "epoch": 0.06383341207713263,
+      "grad_norm": 0.9405627846717834,
+      "learning_rate": 0.00019143178305465784,
+      "loss": 4.8902,
+      "step": 45430
+    },
+    {
+      "epoch": 0.0638474630152962,
+      "grad_norm": 1.075329065322876,
+      "learning_rate": 0.00019147393564704227,
+      "loss": 4.7951,
+      "step": 45440
+    },
+    {
+      "epoch": 0.06386151395345978,
+      "grad_norm": 0.9246218800544739,
+      "learning_rate": 0.00019151608823942673,
+      "loss": 4.8835,
+      "step": 45450
+    },
+    {
+      "epoch": 0.06387556489162335,
+      "grad_norm": 0.9820383191108704,
+      "learning_rate": 0.00019155824083181116,
+      "loss": 4.9678,
+      "step": 45460
+    },
+    {
+      "epoch": 0.06388961582978693,
+      "grad_norm": 0.9855993390083313,
+      "learning_rate": 0.0001916003934241956,
+      "loss": 4.9499,
+      "step": 45470
+    },
+    {
+      "epoch": 0.0639036667679505,
+      "grad_norm": 0.9557892084121704,
+      "learning_rate": 0.00019164254601658,
+      "loss": 4.9888,
+      "step": 45480
+    },
+    {
+      "epoch": 0.0639177177061141,
+      "grad_norm": 0.9590747356414795,
+      "learning_rate": 0.00019168469860896442,
+      "loss": 4.9599,
+      "step": 45490
+    },
+    {
+      "epoch": 0.06393176864427767,
+      "grad_norm": 0.9951782822608948,
+      "learning_rate": 0.00019172685120134885,
+      "loss": 4.8487,
+      "step": 45500
+    },
+    {
+      "epoch": 0.06394581958244125,
+      "grad_norm": 0.9474301338195801,
+      "learning_rate": 0.0001917690037937333,
+      "loss": 4.9257,
+      "step": 45510
+    },
+    {
+      "epoch": 0.06395987052060483,
+      "grad_norm": 0.9949396252632141,
+      "learning_rate": 0.00019181115638611774,
+      "loss": 4.8466,
+      "step": 45520
+    },
+    {
+      "epoch": 0.0639739214587684,
+      "grad_norm": 0.9865704774856567,
+      "learning_rate": 0.00019185330897850217,
+      "loss": 4.9049,
+      "step": 45530
+    },
+    {
+      "epoch": 0.06398797239693198,
+      "grad_norm": 1.0222458839416504,
+      "learning_rate": 0.00019189546157088658,
+      "loss": 4.8955,
+      "step": 45540
+    },
+    {
+      "epoch": 0.06400202333509555,
+      "grad_norm": 0.9480180144309998,
+      "learning_rate": 0.000191937614163271,
+      "loss": 4.9021,
+      "step": 45550
+    },
+    {
+      "epoch": 0.06401607427325913,
+      "grad_norm": 0.9689090847969055,
+      "learning_rate": 0.00019197976675565544,
+      "loss": 5.0263,
+      "step": 45560
+    },
+    {
+      "epoch": 0.0640301252114227,
+      "grad_norm": 0.9644342064857483,
+      "learning_rate": 0.0001920219193480399,
+      "loss": 4.9652,
+      "step": 45570
+    },
+    {
+      "epoch": 0.06404417614958628,
+      "grad_norm": 0.9961534738540649,
+      "learning_rate": 0.00019206407194042433,
+      "loss": 4.8413,
+      "step": 45580
+    },
+    {
+      "epoch": 0.06405822708774986,
+      "grad_norm": 0.9438686966896057,
+      "learning_rate": 0.00019210622453280876,
+      "loss": 4.9344,
+      "step": 45590
+    },
+    {
+      "epoch": 0.06407227802591345,
+      "grad_norm": 0.9871533513069153,
+      "learning_rate": 0.0001921483771251932,
+      "loss": 4.9071,
+      "step": 45600
+    },
+    {
+      "epoch": 0.06408632896407702,
+      "grad_norm": 0.9871090650558472,
+      "learning_rate": 0.0001921905297175776,
+      "loss": 4.8008,
+      "step": 45610
+    },
+    {
+      "epoch": 0.0641003799022406,
+      "grad_norm": 1.0225098133087158,
+      "learning_rate": 0.00019223268230996203,
+      "loss": 5.0841,
+      "step": 45620
+    },
+    {
+      "epoch": 0.06411443084040418,
+      "grad_norm": 0.9514864087104797,
+      "learning_rate": 0.00019227483490234649,
+      "loss": 4.8433,
+      "step": 45630
+    },
+    {
+      "epoch": 0.06412848177856775,
+      "grad_norm": 0.9759644865989685,
+      "learning_rate": 0.00019231698749473092,
+      "loss": 4.9319,
+      "step": 45640
+    },
+    {
+      "epoch": 0.06414253271673133,
+      "grad_norm": 0.9882025122642517,
+      "learning_rate": 0.00019235914008711535,
+      "loss": 4.9566,
+      "step": 45650
+    },
+    {
+      "epoch": 0.0641565836548949,
+      "grad_norm": 0.9645729660987854,
+      "learning_rate": 0.00019240129267949978,
+      "loss": 4.963,
+      "step": 45660
+    },
+    {
+      "epoch": 0.06417063459305848,
+      "grad_norm": 0.9985837936401367,
+      "learning_rate": 0.00019244344527188418,
+      "loss": 4.968,
+      "step": 45670
+    },
+    {
+      "epoch": 0.06418468553122206,
+      "grad_norm": 0.9520775675773621,
+      "learning_rate": 0.00019248559786426862,
+      "loss": 4.9343,
+      "step": 45680
+    },
+    {
+      "epoch": 0.06419873646938563,
+      "grad_norm": 0.947435736656189,
+      "learning_rate": 0.00019252775045665307,
+      "loss": 4.8316,
+      "step": 45690
+    },
+    {
+      "epoch": 0.06421278740754922,
+      "grad_norm": 1.099055528640747,
+      "learning_rate": 0.0001925699030490375,
+      "loss": 4.7128,
+      "step": 45700
+    },
+    {
+      "epoch": 0.0642268383457128,
+      "grad_norm": 0.9727609753608704,
+      "learning_rate": 0.00019261205564142194,
+      "loss": 4.8179,
+      "step": 45710
+    },
+    {
+      "epoch": 0.06424088928387638,
+      "grad_norm": 0.9793810248374939,
+      "learning_rate": 0.00019265420823380637,
+      "loss": 4.9058,
+      "step": 45720
+    },
+    {
+      "epoch": 0.06425494022203995,
+      "grad_norm": 0.9617207646369934,
+      "learning_rate": 0.0001926963608261908,
+      "loss": 4.8955,
+      "step": 45730
+    },
+    {
+      "epoch": 0.06426899116020353,
+      "grad_norm": 0.9949877262115479,
+      "learning_rate": 0.0001927385134185752,
+      "loss": 4.9032,
+      "step": 45740
+    },
+    {
+      "epoch": 0.0642830420983671,
+      "grad_norm": 0.9807795882225037,
+      "learning_rate": 0.00019278066601095966,
+      "loss": 4.9258,
+      "step": 45750
+    },
+    {
+      "epoch": 0.06429709303653068,
+      "grad_norm": 0.9766243696212769,
+      "learning_rate": 0.0001928228186033441,
+      "loss": 4.8925,
+      "step": 45760
+    },
+    {
+      "epoch": 0.06431114397469426,
+      "grad_norm": 0.971154510974884,
+      "learning_rate": 0.00019286497119572852,
+      "loss": 5.0002,
+      "step": 45770
+    },
+    {
+      "epoch": 0.06432519491285783,
+      "grad_norm": 0.966464102268219,
+      "learning_rate": 0.00019290712378811295,
+      "loss": 4.8888,
+      "step": 45780
+    },
+    {
+      "epoch": 0.06433924585102141,
+      "grad_norm": 0.9556533694267273,
+      "learning_rate": 0.00019294927638049738,
+      "loss": 4.8204,
+      "step": 45790
+    },
+    {
+      "epoch": 0.064353296789185,
+      "grad_norm": 0.9517030715942383,
+      "learning_rate": 0.0001929914289728818,
+      "loss": 4.8391,
+      "step": 45800
+    },
+    {
+      "epoch": 0.06436734772734858,
+      "grad_norm": 1.0709158182144165,
+      "learning_rate": 0.00019303358156526625,
+      "loss": 4.8775,
+      "step": 45810
+    },
+    {
+      "epoch": 0.06438139866551215,
+      "grad_norm": 0.9663071036338806,
+      "learning_rate": 0.00019307573415765068,
+      "loss": 4.7677,
+      "step": 45820
+    },
+    {
+      "epoch": 0.06439544960367573,
+      "grad_norm": 0.9344285726547241,
+      "learning_rate": 0.0001931178867500351,
+      "loss": 4.9828,
+      "step": 45830
+    },
+    {
+      "epoch": 0.0644095005418393,
+      "grad_norm": 0.9900237917900085,
+      "learning_rate": 0.00019316003934241954,
+      "loss": 4.7965,
+      "step": 45840
+    },
+    {
+      "epoch": 0.06442355148000288,
+      "grad_norm": 0.9743140935897827,
+      "learning_rate": 0.00019320219193480397,
+      "loss": 4.963,
+      "step": 45850
+    },
+    {
+      "epoch": 0.06443760241816646,
+      "grad_norm": 0.9134804606437683,
+      "learning_rate": 0.00019324434452718843,
+      "loss": 4.8572,
+      "step": 45860
+    },
+    {
+      "epoch": 0.06445165335633003,
+      "grad_norm": 0.9742643237113953,
+      "learning_rate": 0.00019328649711957283,
+      "loss": 4.8383,
+      "step": 45870
+    },
+    {
+      "epoch": 0.06446570429449361,
+      "grad_norm": 0.9616056680679321,
+      "learning_rate": 0.00019332864971195727,
+      "loss": 4.7778,
+      "step": 45880
+    },
+    {
+      "epoch": 0.06447975523265718,
+      "grad_norm": 0.9398157596588135,
+      "learning_rate": 0.0001933708023043417,
+      "loss": 4.9049,
+      "step": 45890
+    },
+    {
+      "epoch": 0.06449380617082076,
+      "grad_norm": 1.0289732217788696,
+      "learning_rate": 0.00019341295489672613,
+      "loss": 4.9867,
+      "step": 45900
+    },
+    {
+      "epoch": 0.06450785710898435,
+      "grad_norm": 0.9898787140846252,
+      "learning_rate": 0.00019345510748911056,
+      "loss": 4.9202,
+      "step": 45910
+    },
+    {
+      "epoch": 0.06452190804714793,
+      "grad_norm": 0.9590668082237244,
+      "learning_rate": 0.00019349726008149502,
+      "loss": 4.8323,
+      "step": 45920
+    },
+    {
+      "epoch": 0.0645359589853115,
+      "grad_norm": 0.9680622220039368,
+      "learning_rate": 0.00019353941267387942,
+      "loss": 4.9099,
+      "step": 45930
+    },
+    {
+      "epoch": 0.06455000992347508,
+      "grad_norm": 0.9715376496315002,
+      "learning_rate": 0.00019358156526626385,
+      "loss": 4.8441,
+      "step": 45940
+    },
+    {
+      "epoch": 0.06456406086163866,
+      "grad_norm": 1.0132158994674683,
+      "learning_rate": 0.00019362371785864828,
+      "loss": 4.9372,
+      "step": 45950
+    },
+    {
+      "epoch": 0.06457811179980223,
+      "grad_norm": 1.0213428735733032,
+      "learning_rate": 0.00019366587045103271,
+      "loss": 4.9999,
+      "step": 45960
+    },
+    {
+      "epoch": 0.06459216273796581,
+      "grad_norm": 0.9593140482902527,
+      "learning_rate": 0.00019370802304341717,
+      "loss": 4.9006,
+      "step": 45970
+    },
+    {
+      "epoch": 0.06460621367612938,
+      "grad_norm": 0.9583016037940979,
+      "learning_rate": 0.0001937501756358016,
+      "loss": 4.9547,
+      "step": 45980
+    },
+    {
+      "epoch": 0.06462026461429296,
+      "grad_norm": 1.043340802192688,
+      "learning_rate": 0.000193792328228186,
+      "loss": 4.8028,
+      "step": 45990
+    },
+    {
+      "epoch": 0.06463431555245654,
+      "grad_norm": 0.9959349036216736,
+      "learning_rate": 0.00019383448082057044,
+      "loss": 4.8156,
+      "step": 46000
+    },
+    {
+      "epoch": 0.06464836649062013,
+      "grad_norm": 0.9628585577011108,
+      "learning_rate": 0.00019387663341295487,
+      "loss": 4.9479,
+      "step": 46010
+    },
+    {
+      "epoch": 0.0646624174287837,
+      "grad_norm": 0.9335030317306519,
+      "learning_rate": 0.0001939187860053393,
+      "loss": 4.8707,
+      "step": 46020
+    },
+    {
+      "epoch": 0.06467646836694728,
+      "grad_norm": 0.9447938203811646,
+      "learning_rate": 0.00019396093859772376,
+      "loss": 4.8819,
+      "step": 46030
+    },
+    {
+      "epoch": 0.06469051930511086,
+      "grad_norm": 0.9535121917724609,
+      "learning_rate": 0.0001940030911901082,
+      "loss": 4.9952,
+      "step": 46040
+    },
+    {
+      "epoch": 0.06470457024327443,
+      "grad_norm": 0.9345034956932068,
+      "learning_rate": 0.00019404524378249262,
+      "loss": 4.8327,
+      "step": 46050
+    },
+    {
+      "epoch": 0.06471862118143801,
+      "grad_norm": 0.9597700238227844,
+      "learning_rate": 0.00019408739637487703,
+      "loss": 4.9953,
+      "step": 46060
+    },
+    {
+      "epoch": 0.06473267211960158,
+      "grad_norm": 0.9878246188163757,
+      "learning_rate": 0.00019412954896726146,
+      "loss": 4.9171,
+      "step": 46070
+    },
+    {
+      "epoch": 0.06474672305776516,
+      "grad_norm": 0.9477070569992065,
+      "learning_rate": 0.0001941717015596459,
+      "loss": 4.8915,
+      "step": 46080
+    },
+    {
+      "epoch": 0.06476077399592874,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019421385415203035,
+      "loss": 4.9158,
+      "step": 46090
+    },
+    {
+      "epoch": 0.06477482493409231,
+      "grad_norm": 0.9720420837402344,
+      "learning_rate": 0.00019425600674441478,
+      "loss": 5.008,
+      "step": 46100
+    },
+    {
+      "epoch": 0.0647888758722559,
+      "grad_norm": 0.9742789268493652,
+      "learning_rate": 0.0001942981593367992,
+      "loss": 4.8,
+      "step": 46110
+    },
+    {
+      "epoch": 0.06480292681041948,
+      "grad_norm": 0.9673479795455933,
+      "learning_rate": 0.0001943403119291836,
+      "loss": 4.8455,
+      "step": 46120
+    },
+    {
+      "epoch": 0.06481697774858305,
+      "grad_norm": 0.9082801938056946,
+      "learning_rate": 0.00019438246452156804,
+      "loss": 4.8671,
+      "step": 46130
+    },
+    {
+      "epoch": 0.06483102868674663,
+      "grad_norm": 0.9455775022506714,
+      "learning_rate": 0.00019442461711395248,
+      "loss": 4.8722,
+      "step": 46140
+    },
+    {
+      "epoch": 0.0648450796249102,
+      "grad_norm": 0.9338746666908264,
+      "learning_rate": 0.00019446676970633693,
+      "loss": 4.8412,
+      "step": 46150
+    },
+    {
+      "epoch": 0.06485913056307378,
+      "grad_norm": 0.976826012134552,
+      "learning_rate": 0.00019450892229872137,
+      "loss": 4.7848,
+      "step": 46160
+    },
+    {
+      "epoch": 0.06487318150123736,
+      "grad_norm": 0.9593846797943115,
+      "learning_rate": 0.0001945510748911058,
+      "loss": 4.8292,
+      "step": 46170
+    },
+    {
+      "epoch": 0.06488723243940094,
+      "grad_norm": 0.9429964423179626,
+      "learning_rate": 0.00019459322748349023,
+      "loss": 4.8929,
+      "step": 46180
+    },
+    {
+      "epoch": 0.06490128337756451,
+      "grad_norm": 0.9602091312408447,
+      "learning_rate": 0.00019463538007587463,
+      "loss": 4.8737,
+      "step": 46190
+    },
+    {
+      "epoch": 0.06491533431572809,
+      "grad_norm": 0.9551581740379333,
+      "learning_rate": 0.00019467753266825906,
+      "loss": 4.8948,
+      "step": 46200
+    },
+    {
+      "epoch": 0.06492938525389168,
+      "grad_norm": 0.9924933314323425,
+      "learning_rate": 0.00019471968526064352,
+      "loss": 4.9313,
+      "step": 46210
+    },
+    {
+      "epoch": 0.06494343619205525,
+      "grad_norm": 0.9613593220710754,
+      "learning_rate": 0.00019476183785302795,
+      "loss": 4.8588,
+      "step": 46220
+    },
+    {
+      "epoch": 0.06495748713021883,
+      "grad_norm": 0.9572134017944336,
+      "learning_rate": 0.00019480399044541238,
+      "loss": 4.8414,
+      "step": 46230
+    },
+    {
+      "epoch": 0.0649715380683824,
+      "grad_norm": 0.9708147048950195,
+      "learning_rate": 0.00019484614303779681,
+      "loss": 4.8422,
+      "step": 46240
+    },
+    {
+      "epoch": 0.06498558900654598,
+      "grad_norm": 0.9796114563941956,
+      "learning_rate": 0.00019488829563018122,
+      "loss": 4.843,
+      "step": 46250
+    },
+    {
+      "epoch": 0.06499963994470956,
+      "grad_norm": 0.9643402099609375,
+      "learning_rate": 0.00019493044822256565,
+      "loss": 4.9446,
+      "step": 46260
+    },
+    {
+      "epoch": 0.06501369088287313,
+      "grad_norm": 0.9505442976951599,
+      "learning_rate": 0.0001949726008149501,
+      "loss": 5.0311,
+      "step": 46270
+    },
+    {
+      "epoch": 0.06502774182103671,
+      "grad_norm": 0.9346354007720947,
+      "learning_rate": 0.00019501475340733454,
+      "loss": 4.8329,
+      "step": 46280
+    },
+    {
+      "epoch": 0.06504179275920029,
+      "grad_norm": 1.1094235181808472,
+      "learning_rate": 0.00019505690599971897,
+      "loss": 4.9727,
+      "step": 46290
+    },
+    {
+      "epoch": 0.06505584369736386,
+      "grad_norm": 0.9481832385063171,
+      "learning_rate": 0.0001950990585921034,
+      "loss": 4.9542,
+      "step": 46300
+    },
+    {
+      "epoch": 0.06506989463552744,
+      "grad_norm": 0.9947605729103088,
+      "learning_rate": 0.00019514121118448783,
+      "loss": 4.979,
+      "step": 46310
+    },
+    {
+      "epoch": 0.06508394557369103,
+      "grad_norm": 0.9515612125396729,
+      "learning_rate": 0.00019518336377687224,
+      "loss": 4.9198,
+      "step": 46320
+    },
+    {
+      "epoch": 0.0650979965118546,
+      "grad_norm": 0.9424987435340881,
+      "learning_rate": 0.0001952255163692567,
+      "loss": 4.8822,
+      "step": 46330
+    },
+    {
+      "epoch": 0.06511204745001818,
+      "grad_norm": 1.0159428119659424,
+      "learning_rate": 0.00019526766896164113,
+      "loss": 4.7915,
+      "step": 46340
+    },
+    {
+      "epoch": 0.06512609838818176,
+      "grad_norm": 0.993810772895813,
+      "learning_rate": 0.00019530982155402556,
+      "loss": 4.9791,
+      "step": 46350
+    },
+    {
+      "epoch": 0.06514014932634533,
+      "grad_norm": 1.029266357421875,
+      "learning_rate": 0.00019535197414641,
+      "loss": 4.7561,
+      "step": 46360
+    },
+    {
+      "epoch": 0.06515420026450891,
+      "grad_norm": 0.9550025463104248,
+      "learning_rate": 0.00019539412673879442,
+      "loss": 4.9486,
+      "step": 46370
+    },
+    {
+      "epoch": 0.06516825120267249,
+      "grad_norm": 0.9962372183799744,
+      "learning_rate": 0.00019543627933117882,
+      "loss": 4.8646,
+      "step": 46380
+    },
+    {
+      "epoch": 0.06518230214083606,
+      "grad_norm": 0.9664082527160645,
+      "learning_rate": 0.00019547843192356328,
+      "loss": 4.9335,
+      "step": 46390
+    },
+    {
+      "epoch": 0.06519635307899964,
+      "grad_norm": 1.0118764638900757,
+      "learning_rate": 0.0001955205845159477,
+      "loss": 4.8674,
+      "step": 46400
+    },
+    {
+      "epoch": 0.06521040401716321,
+      "grad_norm": 1.034801959991455,
+      "learning_rate": 0.00019556273710833214,
+      "loss": 4.8237,
+      "step": 46410
+    },
+    {
+      "epoch": 0.0652244549553268,
+      "grad_norm": 0.9353872537612915,
+      "learning_rate": 0.00019560488970071658,
+      "loss": 4.9462,
+      "step": 46420
+    },
+    {
+      "epoch": 0.06523850589349038,
+      "grad_norm": 1.0210577249526978,
+      "learning_rate": 0.000195647042293101,
+      "loss": 4.914,
+      "step": 46430
+    },
+    {
+      "epoch": 0.06525255683165396,
+      "grad_norm": 0.9545332789421082,
+      "learning_rate": 0.00019568919488548547,
+      "loss": 4.9091,
+      "step": 46440
+    },
+    {
+      "epoch": 0.06526660776981753,
+      "grad_norm": 0.9154708981513977,
+      "learning_rate": 0.00019573134747786987,
+      "loss": 4.88,
+      "step": 46450
+    },
+    {
+      "epoch": 0.06528065870798111,
+      "grad_norm": 0.9885104894638062,
+      "learning_rate": 0.0001957735000702543,
+      "loss": 4.8147,
+      "step": 46460
+    },
+    {
+      "epoch": 0.06529470964614469,
+      "grad_norm": 1.0183314085006714,
+      "learning_rate": 0.00019581565266263873,
+      "loss": 4.8397,
+      "step": 46470
+    },
+    {
+      "epoch": 0.06530876058430826,
+      "grad_norm": 0.969292163848877,
+      "learning_rate": 0.00019585780525502316,
+      "loss": 4.82,
+      "step": 46480
+    },
+    {
+      "epoch": 0.06532281152247184,
+      "grad_norm": 0.9941924810409546,
+      "learning_rate": 0.0001958999578474076,
+      "loss": 4.8408,
+      "step": 46490
+    },
+    {
+      "epoch": 0.06533686246063541,
+      "grad_norm": 0.9590076804161072,
+      "learning_rate": 0.00019594211043979205,
+      "loss": 4.8889,
+      "step": 46500
+    },
+    {
+      "epoch": 0.06535091339879899,
+      "grad_norm": 0.9533814787864685,
+      "learning_rate": 0.00019598426303217646,
+      "loss": 4.8636,
+      "step": 46510
+    },
+    {
+      "epoch": 0.06536496433696258,
+      "grad_norm": 0.9973676800727844,
+      "learning_rate": 0.0001960264156245609,
+      "loss": 4.8238,
+      "step": 46520
+    },
+    {
+      "epoch": 0.06537901527512616,
+      "grad_norm": 0.9768134951591492,
+      "learning_rate": 0.00019606856821694532,
+      "loss": 4.8206,
+      "step": 46530
+    },
+    {
+      "epoch": 0.06539306621328973,
+      "grad_norm": 0.9430971145629883,
+      "learning_rate": 0.00019611072080932975,
+      "loss": 4.919,
+      "step": 46540
+    },
+    {
+      "epoch": 0.06540711715145331,
+      "grad_norm": 0.9878526329994202,
+      "learning_rate": 0.0001961528734017142,
+      "loss": 4.9512,
+      "step": 46550
+    },
+    {
+      "epoch": 0.06542116808961689,
+      "grad_norm": 0.9809000492095947,
+      "learning_rate": 0.00019619502599409864,
+      "loss": 4.8933,
+      "step": 46560
+    },
+    {
+      "epoch": 0.06543521902778046,
+      "grad_norm": 0.957922637462616,
+      "learning_rate": 0.00019623717858648307,
+      "loss": 4.8238,
+      "step": 46570
+    },
+    {
+      "epoch": 0.06544926996594404,
+      "grad_norm": 0.9714646935462952,
+      "learning_rate": 0.00019627511591962905,
+      "loss": 4.928,
+      "step": 46580
+    },
+    {
+      "epoch": 0.06546332090410761,
+      "grad_norm": 0.9626190662384033,
+      "learning_rate": 0.00019631726851201348,
+      "loss": 4.7571,
+      "step": 46590
+    },
+    {
+      "epoch": 0.06547737184227119,
+      "grad_norm": 0.9654691815376282,
+      "learning_rate": 0.00019635942110439792,
+      "loss": 4.9651,
+      "step": 46600
+    },
+    {
+      "epoch": 0.06549142278043477,
+      "grad_norm": 0.9982614517211914,
+      "learning_rate": 0.00019640157369678232,
+      "loss": 5.0936,
+      "step": 46610
+    },
+    {
+      "epoch": 0.06550547371859834,
+      "grad_norm": 0.9625030159950256,
+      "learning_rate": 0.00019644372628916675,
+      "loss": 4.9313,
+      "step": 46620
+    },
+    {
+      "epoch": 0.06551952465676193,
+      "grad_norm": 0.9751782417297363,
+      "learning_rate": 0.0001964858788815512,
+      "loss": 4.9248,
+      "step": 46630
+    },
+    {
+      "epoch": 0.06553357559492551,
+      "grad_norm": 0.9599964022636414,
+      "learning_rate": 0.00019652803147393564,
+      "loss": 4.9034,
+      "step": 46640
+    },
+    {
+      "epoch": 0.06554762653308908,
+      "grad_norm": 0.9146353602409363,
+      "learning_rate": 0.00019657018406632007,
+      "loss": 4.9831,
+      "step": 46650
+    },
+    {
+      "epoch": 0.06556167747125266,
+      "grad_norm": 0.9709069728851318,
+      "learning_rate": 0.0001966123366587045,
+      "loss": 4.8672,
+      "step": 46660
+    },
+    {
+      "epoch": 0.06557572840941624,
+      "grad_norm": 0.9541670083999634,
+      "learning_rate": 0.0001966544892510889,
+      "loss": 4.9094,
+      "step": 46670
+    },
+    {
+      "epoch": 0.06558977934757981,
+      "grad_norm": 0.9834936857223511,
+      "learning_rate": 0.00019669664184347334,
+      "loss": 4.7472,
+      "step": 46680
+    },
+    {
+      "epoch": 0.06560383028574339,
+      "grad_norm": 0.919543445110321,
+      "learning_rate": 0.0001967387944358578,
+      "loss": 5.0209,
+      "step": 46690
+    },
+    {
+      "epoch": 0.06561788122390697,
+      "grad_norm": 0.9145947694778442,
+      "learning_rate": 0.00019678094702824223,
+      "loss": 4.7879,
+      "step": 46700
+    },
+    {
+      "epoch": 0.06563193216207054,
+      "grad_norm": 0.9651200771331787,
+      "learning_rate": 0.00019682309962062666,
+      "loss": 4.756,
+      "step": 46710
+    },
+    {
+      "epoch": 0.06564598310023412,
+      "grad_norm": 0.9358447790145874,
+      "learning_rate": 0.0001968652522130111,
+      "loss": 4.9653,
+      "step": 46720
+    },
+    {
+      "epoch": 0.06566003403839771,
+      "grad_norm": 0.9291307330131531,
+      "learning_rate": 0.00019690740480539552,
+      "loss": 4.8989,
+      "step": 46730
+    },
+    {
+      "epoch": 0.06567408497656128,
+      "grad_norm": 0.9250601530075073,
+      "learning_rate": 0.00019694955739777992,
+      "loss": 4.7435,
+      "step": 46740
+    },
+    {
+      "epoch": 0.06568813591472486,
+      "grad_norm": 0.9375638961791992,
+      "learning_rate": 0.00019699170999016438,
+      "loss": 4.8347,
+      "step": 46750
+    },
+    {
+      "epoch": 0.06570218685288844,
+      "grad_norm": 0.9673593044281006,
+      "learning_rate": 0.00019703386258254881,
+      "loss": 4.8107,
+      "step": 46760
+    },
+    {
+      "epoch": 0.06571623779105201,
+      "grad_norm": 0.9529427886009216,
+      "learning_rate": 0.00019707601517493324,
+      "loss": 4.8632,
+      "step": 46770
+    },
+    {
+      "epoch": 0.06573028872921559,
+      "grad_norm": 0.9514247179031372,
+      "learning_rate": 0.00019711816776731768,
+      "loss": 4.9264,
+      "step": 46780
+    },
+    {
+      "epoch": 0.06574433966737916,
+      "grad_norm": 1.0021986961364746,
+      "learning_rate": 0.0001971603203597021,
+      "loss": 4.8666,
+      "step": 46790
+    },
+    {
+      "epoch": 0.06575839060554274,
+      "grad_norm": 1.545295000076294,
+      "learning_rate": 0.0001972024729520865,
+      "loss": 4.8099,
+      "step": 46800
+    },
+    {
+      "epoch": 0.06577244154370632,
+      "grad_norm": 0.9640814065933228,
+      "learning_rate": 0.00019724462554447097,
+      "loss": 4.9203,
+      "step": 46810
+    },
+    {
+      "epoch": 0.0657864924818699,
+      "grad_norm": 0.9138042330741882,
+      "learning_rate": 0.0001972867781368554,
+      "loss": 5.0355,
+      "step": 46820
+    },
+    {
+      "epoch": 0.06580054342003348,
+      "grad_norm": 0.9734580516815186,
+      "learning_rate": 0.00019732893072923983,
+      "loss": 4.8711,
+      "step": 46830
+    },
+    {
+      "epoch": 0.06581459435819706,
+      "grad_norm": 1.0923768281936646,
+      "learning_rate": 0.00019737108332162426,
+      "loss": 4.9167,
+      "step": 46840
+    },
+    {
+      "epoch": 0.06582864529636064,
+      "grad_norm": 1.087605595588684,
+      "learning_rate": 0.0001974132359140087,
+      "loss": 4.9004,
+      "step": 46850
+    },
+    {
+      "epoch": 0.06584269623452421,
+      "grad_norm": 0.9761277437210083,
+      "learning_rate": 0.00019745538850639315,
+      "loss": 4.9228,
+      "step": 46860
+    },
+    {
+      "epoch": 0.06585674717268779,
+      "grad_norm": 1.0201919078826904,
+      "learning_rate": 0.00019749754109877756,
+      "loss": 4.9418,
+      "step": 46870
+    },
+    {
+      "epoch": 0.06587079811085136,
+      "grad_norm": 0.9499254822731018,
+      "learning_rate": 0.000197539693691162,
+      "loss": 4.9016,
+      "step": 46880
+    },
+    {
+      "epoch": 0.06588484904901494,
+      "grad_norm": 0.9977640509605408,
+      "learning_rate": 0.00019758184628354642,
+      "loss": 4.8839,
+      "step": 46890
+    },
+    {
+      "epoch": 0.06589889998717852,
+      "grad_norm": 0.9540080428123474,
+      "learning_rate": 0.00019762399887593085,
+      "loss": 4.8135,
+      "step": 46900
+    },
+    {
+      "epoch": 0.06591295092534209,
+      "grad_norm": 0.9515107274055481,
+      "learning_rate": 0.00019766615146831528,
+      "loss": 4.7421,
+      "step": 46910
+    },
+    {
+      "epoch": 0.06592700186350567,
+      "grad_norm": 0.9759949445724487,
+      "learning_rate": 0.00019770830406069974,
+      "loss": 4.9691,
+      "step": 46920
+    },
+    {
+      "epoch": 0.06594105280166924,
+      "grad_norm": 0.9368551969528198,
+      "learning_rate": 0.00019775045665308414,
+      "loss": 4.9767,
+      "step": 46930
+    },
+    {
+      "epoch": 0.06595510373983283,
+      "grad_norm": 0.9469462037086487,
+      "learning_rate": 0.00019779260924546857,
+      "loss": 4.8049,
+      "step": 46940
+    },
+    {
+      "epoch": 0.06596915467799641,
+      "grad_norm": 0.979142427444458,
+      "learning_rate": 0.000197834761837853,
+      "loss": 4.8257,
+      "step": 46950
+    },
+    {
+      "epoch": 0.06598320561615999,
+      "grad_norm": 0.9656456112861633,
+      "learning_rate": 0.00019787691443023744,
+      "loss": 4.9793,
+      "step": 46960
+    },
+    {
+      "epoch": 0.06599725655432356,
+      "grad_norm": 0.9597137570381165,
+      "learning_rate": 0.00019791906702262187,
+      "loss": 4.882,
+      "step": 46970
+    },
+    {
+      "epoch": 0.06601130749248714,
+      "grad_norm": 0.9437071084976196,
+      "learning_rate": 0.00019796121961500633,
+      "loss": 4.8506,
+      "step": 46980
+    },
+    {
+      "epoch": 0.06602535843065072,
+      "grad_norm": 0.9833372235298157,
+      "learning_rate": 0.00019800337220739076,
+      "loss": 4.8654,
+      "step": 46990
+    },
+    {
+      "epoch": 0.06603940936881429,
+      "grad_norm": 0.9324950575828552,
+      "learning_rate": 0.00019804552479977516,
+      "loss": 4.8477,
+      "step": 47000
+    },
+    {
+      "epoch": 0.06605346030697787,
+      "grad_norm": 0.9253446459770203,
+      "learning_rate": 0.0001980876773921596,
+      "loss": 4.831,
+      "step": 47010
+    },
+    {
+      "epoch": 0.06606751124514144,
+      "grad_norm": 0.9337993264198303,
+      "learning_rate": 0.00019812982998454402,
+      "loss": 4.8415,
+      "step": 47020
+    },
+    {
+      "epoch": 0.06608156218330502,
+      "grad_norm": 0.9470283389091492,
+      "learning_rate": 0.00019817198257692846,
+      "loss": 4.9581,
+      "step": 47030
+    },
+    {
+      "epoch": 0.06609561312146861,
+      "grad_norm": 0.9741782546043396,
+      "learning_rate": 0.00019821413516931291,
+      "loss": 4.9539,
+      "step": 47040
+    },
+    {
+      "epoch": 0.06610966405963219,
+      "grad_norm": 0.9608268141746521,
+      "learning_rate": 0.00019825628776169734,
+      "loss": 4.8291,
+      "step": 47050
+    },
+    {
+      "epoch": 0.06612371499779576,
+      "grad_norm": 1.0971790552139282,
+      "learning_rate": 0.00019829844035408175,
+      "loss": 4.8117,
+      "step": 47060
+    },
+    {
+      "epoch": 0.06613776593595934,
+      "grad_norm": 0.9461948871612549,
+      "learning_rate": 0.00019834059294646618,
+      "loss": 4.9276,
+      "step": 47070
+    },
+    {
+      "epoch": 0.06615181687412292,
+      "grad_norm": 0.9624106287956238,
+      "learning_rate": 0.0001983827455388506,
+      "loss": 4.8393,
+      "step": 47080
+    },
+    {
+      "epoch": 0.06616586781228649,
+      "grad_norm": 0.9574129581451416,
+      "learning_rate": 0.00019842489813123504,
+      "loss": 4.7859,
+      "step": 47090
+    },
+    {
+      "epoch": 0.06617991875045007,
+      "grad_norm": 0.9906713962554932,
+      "learning_rate": 0.0001984670507236195,
+      "loss": 4.9673,
+      "step": 47100
+    },
+    {
+      "epoch": 0.06619396968861364,
+      "grad_norm": 0.9565755128860474,
+      "learning_rate": 0.00019850920331600393,
+      "loss": 4.9318,
+      "step": 47110
+    },
+    {
+      "epoch": 0.06620802062677722,
+      "grad_norm": 0.9409468770027161,
+      "learning_rate": 0.00019855135590838834,
+      "loss": 4.7968,
+      "step": 47120
+    },
+    {
+      "epoch": 0.0662220715649408,
+      "grad_norm": 0.9226107001304626,
+      "learning_rate": 0.00019859350850077277,
+      "loss": 4.7978,
+      "step": 47130
+    },
+    {
+      "epoch": 0.06623612250310439,
+      "grad_norm": 0.919445812702179,
+      "learning_rate": 0.0001986356610931572,
+      "loss": 4.9299,
+      "step": 47140
+    },
+    {
+      "epoch": 0.06625017344126796,
+      "grad_norm": 0.9550936818122864,
+      "learning_rate": 0.00019867781368554163,
+      "loss": 4.8339,
+      "step": 47150
+    },
+    {
+      "epoch": 0.06626422437943154,
+      "grad_norm": 0.9679076075553894,
+      "learning_rate": 0.0001987199662779261,
+      "loss": 4.912,
+      "step": 47160
+    },
+    {
+      "epoch": 0.06627827531759511,
+      "grad_norm": 0.9844717979431152,
+      "learning_rate": 0.00019876211887031052,
+      "loss": 4.9034,
+      "step": 47170
+    },
+    {
+      "epoch": 0.06629232625575869,
+      "grad_norm": 0.9533633589744568,
+      "learning_rate": 0.00019880427146269495,
+      "loss": 4.9254,
+      "step": 47180
+    },
+    {
+      "epoch": 0.06630637719392227,
+      "grad_norm": 0.9623854160308838,
+      "learning_rate": 0.00019884642405507935,
+      "loss": 4.9148,
+      "step": 47190
+    },
+    {
+      "epoch": 0.06632042813208584,
+      "grad_norm": 0.9844478964805603,
+      "learning_rate": 0.00019888857664746379,
+      "loss": 4.901,
+      "step": 47200
+    },
+    {
+      "epoch": 0.06633447907024942,
+      "grad_norm": 0.9062860012054443,
+      "learning_rate": 0.00019893072923984824,
+      "loss": 4.8345,
+      "step": 47210
+    },
+    {
+      "epoch": 0.066348530008413,
+      "grad_norm": 0.9490434527397156,
+      "learning_rate": 0.00019897288183223267,
+      "loss": 4.8597,
+      "step": 47220
+    },
+    {
+      "epoch": 0.06636258094657657,
+      "grad_norm": 0.9500983357429504,
+      "learning_rate": 0.0001990150344246171,
+      "loss": 4.8869,
+      "step": 47230
+    },
+    {
+      "epoch": 0.06637663188474015,
+      "grad_norm": 0.9309141039848328,
+      "learning_rate": 0.00019905718701700154,
+      "loss": 4.864,
+      "step": 47240
+    },
+    {
+      "epoch": 0.06639068282290374,
+      "grad_norm": 1.0119329690933228,
+      "learning_rate": 0.00019909933960938594,
+      "loss": 4.8581,
+      "step": 47250
+    },
+    {
+      "epoch": 0.06640473376106731,
+      "grad_norm": 0.9545915126800537,
+      "learning_rate": 0.00019914149220177037,
+      "loss": 5.0142,
+      "step": 47260
+    },
+    {
+      "epoch": 0.06641878469923089,
+      "grad_norm": 0.948043167591095,
+      "learning_rate": 0.00019918364479415483,
+      "loss": 4.9292,
+      "step": 47270
+    },
+    {
+      "epoch": 0.06643283563739447,
+      "grad_norm": 0.9219819903373718,
+      "learning_rate": 0.00019922579738653926,
+      "loss": 4.9284,
+      "step": 47280
+    },
+    {
+      "epoch": 0.06644688657555804,
+      "grad_norm": 0.9398238658905029,
+      "learning_rate": 0.0001992679499789237,
+      "loss": 4.8902,
+      "step": 47290
+    },
+    {
+      "epoch": 0.06646093751372162,
+      "grad_norm": 0.9383578896522522,
+      "learning_rate": 0.00019931010257130812,
+      "loss": 4.827,
+      "step": 47300
+    },
+    {
+      "epoch": 0.0664749884518852,
+      "grad_norm": 0.9550478458404541,
+      "learning_rate": 0.00019935225516369256,
+      "loss": 5.0553,
+      "step": 47310
+    },
+    {
+      "epoch": 0.06648903939004877,
+      "grad_norm": 0.9616899490356445,
+      "learning_rate": 0.00019939440775607696,
+      "loss": 4.8483,
+      "step": 47320
+    },
+    {
+      "epoch": 0.06650309032821235,
+      "grad_norm": 0.9487091898918152,
+      "learning_rate": 0.00019943656034846142,
+      "loss": 4.9295,
+      "step": 47330
+    },
+    {
+      "epoch": 0.06651714126637592,
+      "grad_norm": 0.939850926399231,
+      "learning_rate": 0.00019947871294084585,
+      "loss": 4.7803,
+      "step": 47340
+    },
+    {
+      "epoch": 0.06653119220453951,
+      "grad_norm": 0.9599955081939697,
+      "learning_rate": 0.00019952086553323028,
+      "loss": 4.8776,
+      "step": 47350
+    },
+    {
+      "epoch": 0.06654524314270309,
+      "grad_norm": 0.9502904415130615,
+      "learning_rate": 0.0001995630181256147,
+      "loss": 4.8173,
+      "step": 47360
+    },
+    {
+      "epoch": 0.06655929408086667,
+      "grad_norm": 0.9396896958351135,
+      "learning_rate": 0.00019960517071799914,
+      "loss": 4.9296,
+      "step": 47370
+    },
+    {
+      "epoch": 0.06657334501903024,
+      "grad_norm": 0.9196685552597046,
+      "learning_rate": 0.00019964732331038355,
+      "loss": 5.0329,
+      "step": 47380
+    },
+    {
+      "epoch": 0.06658739595719382,
+      "grad_norm": 0.9280641078948975,
+      "learning_rate": 0.000199689475902768,
+      "loss": 4.8225,
+      "step": 47390
+    },
+    {
+      "epoch": 0.0666014468953574,
+      "grad_norm": 0.9672068357467651,
+      "learning_rate": 0.00019973162849515244,
+      "loss": 4.8844,
+      "step": 47400
+    },
+    {
+      "epoch": 0.06661549783352097,
+      "grad_norm": 0.9293968677520752,
+      "learning_rate": 0.00019977378108753687,
+      "loss": 4.828,
+      "step": 47410
+    },
+    {
+      "epoch": 0.06662954877168455,
+      "grad_norm": 0.9214613437652588,
+      "learning_rate": 0.0001998159336799213,
+      "loss": 4.8509,
+      "step": 47420
+    },
+    {
+      "epoch": 0.06664359970984812,
+      "grad_norm": 0.9712655544281006,
+      "learning_rate": 0.00019985808627230573,
+      "loss": 4.8579,
+      "step": 47430
+    },
+    {
+      "epoch": 0.0666576506480117,
+      "grad_norm": 0.9182402491569519,
+      "learning_rate": 0.0001999002388646902,
+      "loss": 4.789,
+      "step": 47440
+    },
+    {
+      "epoch": 0.06667170158617529,
+      "grad_norm": 0.9156343936920166,
+      "learning_rate": 0.0001999423914570746,
+      "loss": 4.8536,
+      "step": 47450
+    },
+    {
+      "epoch": 0.06668575252433886,
+      "grad_norm": 0.9341609477996826,
+      "learning_rate": 0.00019998454404945902,
+      "loss": 4.989,
+      "step": 47460
+    },
+    {
+      "epoch": 0.06669980346250244,
+      "grad_norm": 0.9106461405754089,
+      "learning_rate": 0.00020002669664184345,
+      "loss": 4.9627,
+      "step": 47470
+    },
+    {
+      "epoch": 0.06671385440066602,
+      "grad_norm": 1.2454720735549927,
+      "learning_rate": 0.00020006884923422788,
+      "loss": 4.8649,
+      "step": 47480
+    },
+    {
+      "epoch": 0.0667279053388296,
+      "grad_norm": 0.9427316784858704,
+      "learning_rate": 0.00020011100182661232,
+      "loss": 4.8691,
+      "step": 47490
+    },
+    {
+      "epoch": 0.06674195627699317,
+      "grad_norm": 0.9439139366149902,
+      "learning_rate": 0.00020015315441899677,
+      "loss": 4.9893,
+      "step": 47500
+    },
+    {
+      "epoch": 0.06675600721515675,
+      "grad_norm": 0.9804021120071411,
+      "learning_rate": 0.00020019530701138118,
+      "loss": 5.011,
+      "step": 47510
+    },
+    {
+      "epoch": 0.06677005815332032,
+      "grad_norm": 0.9366146922111511,
+      "learning_rate": 0.0002002374596037656,
+      "loss": 4.9722,
+      "step": 47520
+    },
+    {
+      "epoch": 0.0667841090914839,
+      "grad_norm": 0.9678098559379578,
+      "learning_rate": 0.00020027961219615004,
+      "loss": 4.91,
+      "step": 47530
+    },
+    {
+      "epoch": 0.06679816002964747,
+      "grad_norm": 0.97898268699646,
+      "learning_rate": 0.00020032176478853447,
+      "loss": 4.8503,
+      "step": 47540
+    },
+    {
+      "epoch": 0.06681221096781105,
+      "grad_norm": 0.9911873936653137,
+      "learning_rate": 0.0002003639173809189,
+      "loss": 4.8557,
+      "step": 47550
+    },
+    {
+      "epoch": 0.06682626190597464,
+      "grad_norm": 0.9423503875732422,
+      "learning_rate": 0.00020040606997330336,
+      "loss": 4.783,
+      "step": 47560
+    },
+    {
+      "epoch": 0.06684031284413822,
+      "grad_norm": 0.956548810005188,
+      "learning_rate": 0.0002004482225656878,
+      "loss": 5.0057,
+      "step": 47570
+    },
+    {
+      "epoch": 0.06685436378230179,
+      "grad_norm": 0.9642254114151001,
+      "learning_rate": 0.0002004903751580722,
+      "loss": 4.9451,
+      "step": 47580
+    },
+    {
+      "epoch": 0.06686841472046537,
+      "grad_norm": 0.9153860211372375,
+      "learning_rate": 0.00020053252775045663,
+      "loss": 4.8504,
+      "step": 47590
+    },
+    {
+      "epoch": 0.06688246565862895,
+      "grad_norm": 0.9987194538116455,
+      "learning_rate": 0.00020057468034284106,
+      "loss": 4.7961,
+      "step": 47600
+    },
+    {
+      "epoch": 0.06689651659679252,
+      "grad_norm": 0.8986924290657043,
+      "learning_rate": 0.0002006168329352255,
+      "loss": 4.8971,
+      "step": 47610
+    },
+    {
+      "epoch": 0.0669105675349561,
+      "grad_norm": 0.9505689740180969,
+      "learning_rate": 0.00020065898552760995,
+      "loss": 4.9268,
+      "step": 47620
+    },
+    {
+      "epoch": 0.06692461847311967,
+      "grad_norm": 0.9821462631225586,
+      "learning_rate": 0.00020070113811999438,
+      "loss": 4.8518,
+      "step": 47630
+    },
+    {
+      "epoch": 0.06693866941128325,
+      "grad_norm": 0.9475671648979187,
+      "learning_rate": 0.00020074329071237878,
+      "loss": 4.9195,
+      "step": 47640
+    },
+    {
+      "epoch": 0.06695272034944683,
+      "grad_norm": 0.9619015455245972,
+      "learning_rate": 0.00020078544330476321,
+      "loss": 4.8156,
+      "step": 47650
+    },
+    {
+      "epoch": 0.06696677128761042,
+      "grad_norm": 0.9441516995429993,
+      "learning_rate": 0.00020082759589714765,
+      "loss": 4.9198,
+      "step": 47660
+    },
+    {
+      "epoch": 0.06698082222577399,
+      "grad_norm": 0.923410177230835,
+      "learning_rate": 0.00020086974848953208,
+      "loss": 4.9305,
+      "step": 47670
+    },
+    {
+      "epoch": 0.06699487316393757,
+      "grad_norm": 0.9332021474838257,
+      "learning_rate": 0.00020091190108191654,
+      "loss": 4.7413,
+      "step": 47680
+    },
+    {
+      "epoch": 0.06700892410210114,
+      "grad_norm": 0.9328610897064209,
+      "learning_rate": 0.00020095405367430097,
+      "loss": 4.847,
+      "step": 47690
+    },
+    {
+      "epoch": 0.06702297504026472,
+      "grad_norm": 0.9511117339134216,
+      "learning_rate": 0.0002009962062666854,
+      "loss": 4.8878,
+      "step": 47700
+    },
+    {
+      "epoch": 0.0670370259784283,
+      "grad_norm": 0.9161847829818726,
+      "learning_rate": 0.0002010383588590698,
+      "loss": 4.9897,
+      "step": 47710
+    },
+    {
+      "epoch": 0.06705107691659187,
+      "grad_norm": 0.9360916614532471,
+      "learning_rate": 0.00020108051145145423,
+      "loss": 4.9512,
+      "step": 47720
+    },
+    {
+      "epoch": 0.06706512785475545,
+      "grad_norm": 1.0616661310195923,
+      "learning_rate": 0.00020112266404383866,
+      "loss": 4.9006,
+      "step": 47730
+    },
+    {
+      "epoch": 0.06707917879291903,
+      "grad_norm": 0.9333809614181519,
+      "learning_rate": 0.00020116481663622312,
+      "loss": 4.7788,
+      "step": 47740
+    },
+    {
+      "epoch": 0.0670932297310826,
+      "grad_norm": 0.9299507737159729,
+      "learning_rate": 0.00020120696922860755,
+      "loss": 4.9081,
+      "step": 47750
+    },
+    {
+      "epoch": 0.06710728066924619,
+      "grad_norm": 0.961584746837616,
+      "learning_rate": 0.00020124912182099198,
+      "loss": 4.8686,
+      "step": 47760
+    },
+    {
+      "epoch": 0.06712133160740977,
+      "grad_norm": 0.9510294198989868,
+      "learning_rate": 0.0002012912744133764,
+      "loss": 4.8925,
+      "step": 47770
+    },
+    {
+      "epoch": 0.06713538254557334,
+      "grad_norm": 0.9562880396842957,
+      "learning_rate": 0.00020133342700576082,
+      "loss": 4.8545,
+      "step": 47780
+    },
+    {
+      "epoch": 0.06714943348373692,
+      "grad_norm": 0.9377868175506592,
+      "learning_rate": 0.00020137557959814528,
+      "loss": 4.8412,
+      "step": 47790
+    },
+    {
+      "epoch": 0.0671634844219005,
+      "grad_norm": 0.9249438643455505,
+      "learning_rate": 0.0002014177321905297,
+      "loss": 4.9385,
+      "step": 47800
+    },
+    {
+      "epoch": 0.06717753536006407,
+      "grad_norm": 0.9281932711601257,
+      "learning_rate": 0.00020145988478291414,
+      "loss": 4.8762,
+      "step": 47810
+    },
+    {
+      "epoch": 0.06719158629822765,
+      "grad_norm": 0.9970051050186157,
+      "learning_rate": 0.00020150203737529857,
+      "loss": 4.7625,
+      "step": 47820
+    },
+    {
+      "epoch": 0.06720563723639122,
+      "grad_norm": 0.9395272731781006,
+      "learning_rate": 0.00020154418996768298,
+      "loss": 4.753,
+      "step": 47830
+    },
+    {
+      "epoch": 0.0672196881745548,
+      "grad_norm": 1.0339237451553345,
+      "learning_rate": 0.0002015863425600674,
+      "loss": 4.8091,
+      "step": 47840
+    },
+    {
+      "epoch": 0.06723373911271838,
+      "grad_norm": 0.9100208282470703,
+      "learning_rate": 0.00020162849515245187,
+      "loss": 5.0206,
+      "step": 47850
+    },
+    {
+      "epoch": 0.06724779005088195,
+      "grad_norm": 0.9560271501541138,
+      "learning_rate": 0.0002016706477448363,
+      "loss": 4.9883,
+      "step": 47860
+    },
+    {
+      "epoch": 0.06726184098904554,
+      "grad_norm": 0.9476458430290222,
+      "learning_rate": 0.00020171280033722073,
+      "loss": 4.7588,
+      "step": 47870
+    },
+    {
+      "epoch": 0.06727589192720912,
+      "grad_norm": 0.9497449398040771,
+      "learning_rate": 0.00020175495292960516,
+      "loss": 4.8974,
+      "step": 47880
+    },
+    {
+      "epoch": 0.0672899428653727,
+      "grad_norm": 0.9444364309310913,
+      "learning_rate": 0.0002017971055219896,
+      "loss": 4.884,
+      "step": 47890
+    },
+    {
+      "epoch": 0.06730399380353627,
+      "grad_norm": 0.9414916634559631,
+      "learning_rate": 0.000201839258114374,
+      "loss": 4.9842,
+      "step": 47900
+    },
+    {
+      "epoch": 0.06731804474169985,
+      "grad_norm": 0.9677934646606445,
+      "learning_rate": 0.00020188141070675845,
+      "loss": 4.8616,
+      "step": 47910
+    },
+    {
+      "epoch": 0.06733209567986342,
+      "grad_norm": 0.9179354906082153,
+      "learning_rate": 0.00020192356329914288,
+      "loss": 4.9397,
+      "step": 47920
+    },
+    {
+      "epoch": 0.067346146618027,
+      "grad_norm": 0.9381024241447449,
+      "learning_rate": 0.00020196571589152731,
+      "loss": 4.957,
+      "step": 47930
+    },
+    {
+      "epoch": 0.06736019755619058,
+      "grad_norm": 0.9283067584037781,
+      "learning_rate": 0.00020200786848391175,
+      "loss": 4.6955,
+      "step": 47940
+    },
+    {
+      "epoch": 0.06737424849435415,
+      "grad_norm": 0.9692529439926147,
+      "learning_rate": 0.00020205002107629618,
+      "loss": 4.8843,
+      "step": 47950
+    },
+    {
+      "epoch": 0.06738829943251773,
+      "grad_norm": 0.9457141160964966,
+      "learning_rate": 0.00020209217366868058,
+      "loss": 4.955,
+      "step": 47960
+    },
+    {
+      "epoch": 0.06740235037068132,
+      "grad_norm": 0.9691393375396729,
+      "learning_rate": 0.00020213432626106504,
+      "loss": 4.8771,
+      "step": 47970
+    },
+    {
+      "epoch": 0.0674164013088449,
+      "grad_norm": 0.9532297253608704,
+      "learning_rate": 0.00020217647885344947,
+      "loss": 4.7826,
+      "step": 47980
+    },
+    {
+      "epoch": 0.06743045224700847,
+      "grad_norm": 0.9230175614356995,
+      "learning_rate": 0.0002022186314458339,
+      "loss": 4.8278,
+      "step": 47990
+    },
+    {
+      "epoch": 0.06744450318517205,
+      "grad_norm": 0.9743686318397522,
+      "learning_rate": 0.00020226078403821833,
+      "loss": 4.9041,
+      "step": 48000
+    },
+    {
+      "epoch": 0.06745855412333562,
+      "grad_norm": 0.942179799079895,
+      "learning_rate": 0.00020230293663060276,
+      "loss": 4.8774,
+      "step": 48010
+    },
+    {
+      "epoch": 0.0674726050614992,
+      "grad_norm": 0.9336681962013245,
+      "learning_rate": 0.00020234508922298722,
+      "loss": 4.9366,
+      "step": 48020
+    },
+    {
+      "epoch": 0.06748665599966278,
+      "grad_norm": 1.0104656219482422,
+      "learning_rate": 0.00020238724181537163,
+      "loss": 4.9661,
+      "step": 48030
+    },
+    {
+      "epoch": 0.06750070693782635,
+      "grad_norm": 0.9617270827293396,
+      "learning_rate": 0.00020242939440775606,
+      "loss": 4.8729,
+      "step": 48040
+    },
+    {
+      "epoch": 0.06751475787598993,
+      "grad_norm": 0.8925577402114868,
+      "learning_rate": 0.0002024715470001405,
+      "loss": 4.9942,
+      "step": 48050
+    },
+    {
+      "epoch": 0.0675288088141535,
+      "grad_norm": 0.9957881569862366,
+      "learning_rate": 0.00020251369959252492,
+      "loss": 4.8822,
+      "step": 48060
+    },
+    {
+      "epoch": 0.0675428597523171,
+      "grad_norm": 0.8969820737838745,
+      "learning_rate": 0.00020255585218490935,
+      "loss": 4.8443,
+      "step": 48070
+    },
+    {
+      "epoch": 0.06755691069048067,
+      "grad_norm": 0.8763751983642578,
+      "learning_rate": 0.0002025980047772938,
+      "loss": 4.8714,
+      "step": 48080
+    },
+    {
+      "epoch": 0.06757096162864425,
+      "grad_norm": 0.9305700659751892,
+      "learning_rate": 0.0002026401573696782,
+      "loss": 4.8867,
+      "step": 48090
+    },
+    {
+      "epoch": 0.06758501256680782,
+      "grad_norm": 0.9393258690834045,
+      "learning_rate": 0.00020268230996206264,
+      "loss": 4.9085,
+      "step": 48100
+    },
+    {
+      "epoch": 0.0675990635049714,
+      "grad_norm": 0.9406274557113647,
+      "learning_rate": 0.00020272446255444708,
+      "loss": 4.9558,
+      "step": 48110
+    },
+    {
+      "epoch": 0.06761311444313498,
+      "grad_norm": 0.9472900032997131,
+      "learning_rate": 0.0002027666151468315,
+      "loss": 4.8328,
+      "step": 48120
+    },
+    {
+      "epoch": 0.06762716538129855,
+      "grad_norm": 0.9359706044197083,
+      "learning_rate": 0.00020280876773921594,
+      "loss": 4.9482,
+      "step": 48130
+    },
+    {
+      "epoch": 0.06764121631946213,
+      "grad_norm": 0.9522212743759155,
+      "learning_rate": 0.0002028509203316004,
+      "loss": 4.8838,
+      "step": 48140
+    },
+    {
+      "epoch": 0.0676552672576257,
+      "grad_norm": 0.9575254917144775,
+      "learning_rate": 0.00020289307292398483,
+      "loss": 4.9737,
+      "step": 48150
+    },
+    {
+      "epoch": 0.06766931819578928,
+      "grad_norm": 0.9331732988357544,
+      "learning_rate": 0.00020293522551636923,
+      "loss": 4.7875,
+      "step": 48160
+    },
+    {
+      "epoch": 0.06768336913395286,
+      "grad_norm": 0.9446621537208557,
+      "learning_rate": 0.00020297737810875366,
+      "loss": 4.8592,
+      "step": 48170
+    },
+    {
+      "epoch": 0.06769742007211645,
+      "grad_norm": 0.9311910271644592,
+      "learning_rate": 0.0002030195307011381,
+      "loss": 4.8407,
+      "step": 48180
+    },
+    {
+      "epoch": 0.06771147101028002,
+      "grad_norm": 0.9321035146713257,
+      "learning_rate": 0.00020306168329352252,
+      "loss": 4.9385,
+      "step": 48190
+    },
+    {
+      "epoch": 0.0677255219484436,
+      "grad_norm": 0.9516061544418335,
+      "learning_rate": 0.00020310383588590698,
+      "loss": 4.9021,
+      "step": 48200
+    },
+    {
+      "epoch": 0.06773957288660717,
+      "grad_norm": 0.9016589522361755,
+      "learning_rate": 0.00020314598847829141,
+      "loss": 5.0427,
+      "step": 48210
+    },
+    {
+      "epoch": 0.06775362382477075,
+      "grad_norm": 0.9039486646652222,
+      "learning_rate": 0.00020318814107067582,
+      "loss": 4.9687,
+      "step": 48220
+    },
+    {
+      "epoch": 0.06776767476293433,
+      "grad_norm": 0.9151666164398193,
+      "learning_rate": 0.00020323029366306025,
+      "loss": 4.9319,
+      "step": 48230
+    },
+    {
+      "epoch": 0.0677817257010979,
+      "grad_norm": 0.9722932577133179,
+      "learning_rate": 0.00020327244625544468,
+      "loss": 4.9052,
+      "step": 48240
+    },
+    {
+      "epoch": 0.06779577663926148,
+      "grad_norm": 0.9459034204483032,
+      "learning_rate": 0.0002033145988478291,
+      "loss": 4.7892,
+      "step": 48250
+    },
+    {
+      "epoch": 0.06780982757742506,
+      "grad_norm": 0.9252526760101318,
+      "learning_rate": 0.00020335675144021357,
+      "loss": 4.8437,
+      "step": 48260
+    },
+    {
+      "epoch": 0.06782387851558863,
+      "grad_norm": 0.9979690909385681,
+      "learning_rate": 0.000203398904032598,
+      "loss": 4.9336,
+      "step": 48270
+    },
+    {
+      "epoch": 0.06783792945375222,
+      "grad_norm": 1.0930742025375366,
+      "learning_rate": 0.00020344105662498243,
+      "loss": 4.8809,
+      "step": 48280
+    },
+    {
+      "epoch": 0.0678519803919158,
+      "grad_norm": 0.9363275766372681,
+      "learning_rate": 0.00020348320921736684,
+      "loss": 4.829,
+      "step": 48290
+    },
+    {
+      "epoch": 0.06786603133007937,
+      "grad_norm": 0.9431924223899841,
+      "learning_rate": 0.00020352536180975127,
+      "loss": 4.8887,
+      "step": 48300
+    },
+    {
+      "epoch": 0.06788008226824295,
+      "grad_norm": 0.949468731880188,
+      "learning_rate": 0.0002035675144021357,
+      "loss": 4.9473,
+      "step": 48310
+    },
+    {
+      "epoch": 0.06789413320640653,
+      "grad_norm": 0.9422009587287903,
+      "learning_rate": 0.00020360966699452016,
+      "loss": 4.8516,
+      "step": 48320
+    },
+    {
+      "epoch": 0.0679081841445701,
+      "grad_norm": 0.9336867928504944,
+      "learning_rate": 0.0002036518195869046,
+      "loss": 4.8706,
+      "step": 48330
+    },
+    {
+      "epoch": 0.06792223508273368,
+      "grad_norm": 0.9072600603103638,
+      "learning_rate": 0.00020369397217928902,
+      "loss": 4.9884,
+      "step": 48340
+    },
+    {
+      "epoch": 0.06793628602089725,
+      "grad_norm": 0.9766586422920227,
+      "learning_rate": 0.00020373612477167342,
+      "loss": 4.8165,
+      "step": 48350
+    },
+    {
+      "epoch": 0.06795033695906083,
+      "grad_norm": 0.9324946403503418,
+      "learning_rate": 0.00020377827736405785,
+      "loss": 4.8802,
+      "step": 48360
+    },
+    {
+      "epoch": 0.06796438789722441,
+      "grad_norm": 0.949245035648346,
+      "learning_rate": 0.00020382042995644229,
+      "loss": 4.8723,
+      "step": 48370
+    },
+    {
+      "epoch": 0.067978438835388,
+      "grad_norm": 0.9230749011039734,
+      "learning_rate": 0.00020386258254882674,
+      "loss": 4.8103,
+      "step": 48380
+    },
+    {
+      "epoch": 0.06799248977355157,
+      "grad_norm": 0.9191190004348755,
+      "learning_rate": 0.00020390473514121118,
+      "loss": 4.8711,
+      "step": 48390
+    },
+    {
+      "epoch": 0.06800654071171515,
+      "grad_norm": 0.9459436535835266,
+      "learning_rate": 0.0002039468877335956,
+      "loss": 4.8038,
+      "step": 48400
+    },
+    {
+      "epoch": 0.06802059164987873,
+      "grad_norm": 0.9071534872055054,
+      "learning_rate": 0.00020398904032598004,
+      "loss": 4.8562,
+      "step": 48410
+    },
+    {
+      "epoch": 0.0680346425880423,
+      "grad_norm": 0.9326480627059937,
+      "learning_rate": 0.00020403119291836444,
+      "loss": 4.9268,
+      "step": 48420
+    },
+    {
+      "epoch": 0.06804869352620588,
+      "grad_norm": 0.9467244744300842,
+      "learning_rate": 0.0002040733455107489,
+      "loss": 4.92,
+      "step": 48430
+    },
+    {
+      "epoch": 0.06806274446436945,
+      "grad_norm": 0.9642537832260132,
+      "learning_rate": 0.00020411549810313333,
+      "loss": 4.7312,
+      "step": 48440
+    },
+    {
+      "epoch": 0.06807679540253303,
+      "grad_norm": 0.9437421560287476,
+      "learning_rate": 0.00020415765069551776,
+      "loss": 4.8087,
+      "step": 48450
+    },
+    {
+      "epoch": 0.0680908463406966,
+      "grad_norm": 0.9311556816101074,
+      "learning_rate": 0.0002041998032879022,
+      "loss": 4.8312,
+      "step": 48460
+    },
+    {
+      "epoch": 0.06810489727886018,
+      "grad_norm": 1.035953164100647,
+      "learning_rate": 0.00020424195588028662,
+      "loss": 4.9495,
+      "step": 48470
+    },
+    {
+      "epoch": 0.06811894821702376,
+      "grad_norm": 0.9514090418815613,
+      "learning_rate": 0.00020428410847267103,
+      "loss": 4.8289,
+      "step": 48480
+    },
+    {
+      "epoch": 0.06813299915518735,
+      "grad_norm": 1.0147558450698853,
+      "learning_rate": 0.0002043262610650555,
+      "loss": 4.793,
+      "step": 48490
+    },
+    {
+      "epoch": 0.06814705009335092,
+      "grad_norm": 0.947508692741394,
+      "learning_rate": 0.00020436841365743992,
+      "loss": 4.836,
+      "step": 48500
+    },
+    {
+      "epoch": 0.0681611010315145,
+      "grad_norm": 0.9145636558532715,
+      "learning_rate": 0.00020441056624982435,
+      "loss": 4.7227,
+      "step": 48510
+    },
+    {
+      "epoch": 0.06817515196967808,
+      "grad_norm": 0.9304710030555725,
+      "learning_rate": 0.00020445271884220878,
+      "loss": 4.9376,
+      "step": 48520
+    },
+    {
+      "epoch": 0.06818920290784165,
+      "grad_norm": 0.9380136728286743,
+      "learning_rate": 0.0002044948714345932,
+      "loss": 4.9731,
+      "step": 48530
+    },
+    {
+      "epoch": 0.06820325384600523,
+      "grad_norm": 0.9320715665817261,
+      "learning_rate": 0.00020453702402697762,
+      "loss": 4.7998,
+      "step": 48540
+    },
+    {
+      "epoch": 0.0682173047841688,
+      "grad_norm": 0.939115047454834,
+      "learning_rate": 0.00020457917661936207,
+      "loss": 4.9202,
+      "step": 48550
+    },
+    {
+      "epoch": 0.06823135572233238,
+      "grad_norm": 0.9177758693695068,
+      "learning_rate": 0.0002046213292117465,
+      "loss": 4.865,
+      "step": 48560
+    },
+    {
+      "epoch": 0.06824540666049596,
+      "grad_norm": 0.9489643573760986,
+      "learning_rate": 0.00020466348180413094,
+      "loss": 4.8896,
+      "step": 48570
+    },
+    {
+      "epoch": 0.06825945759865953,
+      "grad_norm": 0.9503336548805237,
+      "learning_rate": 0.00020470563439651537,
+      "loss": 4.7754,
+      "step": 48580
+    },
+    {
+      "epoch": 0.06827350853682312,
+      "grad_norm": 0.920091986656189,
+      "learning_rate": 0.0002047477869888998,
+      "loss": 4.9028,
+      "step": 48590
+    },
+    {
+      "epoch": 0.0682875594749867,
+      "grad_norm": 0.9541656970977783,
+      "learning_rate": 0.00020478993958128426,
+      "loss": 4.8475,
+      "step": 48600
+    },
+    {
+      "epoch": 0.06830161041315028,
+      "grad_norm": 0.9243221879005432,
+      "learning_rate": 0.00020483209217366866,
+      "loss": 4.8828,
+      "step": 48610
+    },
+    {
+      "epoch": 0.06831566135131385,
+      "grad_norm": 0.9231835603713989,
+      "learning_rate": 0.0002048742447660531,
+      "loss": 4.9098,
+      "step": 48620
+    },
+    {
+      "epoch": 0.06832971228947743,
+      "grad_norm": 0.9024877548217773,
+      "learning_rate": 0.00020491639735843752,
+      "loss": 4.8797,
+      "step": 48630
+    },
+    {
+      "epoch": 0.068343763227641,
+      "grad_norm": 0.9409641623497009,
+      "learning_rate": 0.00020495854995082195,
+      "loss": 4.864,
+      "step": 48640
+    },
+    {
+      "epoch": 0.06835781416580458,
+      "grad_norm": 0.9287962317466736,
+      "learning_rate": 0.00020500070254320639,
+      "loss": 4.8829,
+      "step": 48650
+    },
+    {
+      "epoch": 0.06837186510396816,
+      "grad_norm": 0.9256884455680847,
+      "learning_rate": 0.00020504285513559084,
+      "loss": 4.7937,
+      "step": 48660
+    },
+    {
+      "epoch": 0.06838591604213173,
+      "grad_norm": 0.9251719117164612,
+      "learning_rate": 0.00020508500772797525,
+      "loss": 4.9763,
+      "step": 48670
+    },
+    {
+      "epoch": 0.06839996698029531,
+      "grad_norm": 0.9523289203643799,
+      "learning_rate": 0.00020512716032035968,
+      "loss": 4.9023,
+      "step": 48680
+    },
+    {
+      "epoch": 0.0684140179184589,
+      "grad_norm": 0.9629274606704712,
+      "learning_rate": 0.0002051693129127441,
+      "loss": 4.9088,
+      "step": 48690
+    },
+    {
+      "epoch": 0.06842806885662248,
+      "grad_norm": 0.9680508971214294,
+      "learning_rate": 0.00020521146550512854,
+      "loss": 4.9068,
+      "step": 48700
+    },
+    {
+      "epoch": 0.06844211979478605,
+      "grad_norm": 0.989648699760437,
+      "learning_rate": 0.00020525361809751297,
+      "loss": 4.7496,
+      "step": 48710
+    },
+    {
+      "epoch": 0.06845617073294963,
+      "grad_norm": 0.9049251079559326,
+      "learning_rate": 0.00020529577068989743,
+      "loss": 4.9653,
+      "step": 48720
+    },
+    {
+      "epoch": 0.0684702216711132,
+      "grad_norm": 0.9711537957191467,
+      "learning_rate": 0.00020533792328228186,
+      "loss": 4.9024,
+      "step": 48730
+    },
+    {
+      "epoch": 0.06848427260927678,
+      "grad_norm": 0.987939178943634,
+      "learning_rate": 0.00020538007587466627,
+      "loss": 4.8989,
+      "step": 48740
+    },
+    {
+      "epoch": 0.06849832354744036,
+      "grad_norm": 0.9508666396141052,
+      "learning_rate": 0.0002054222284670507,
+      "loss": 4.8888,
+      "step": 48750
+    },
+    {
+      "epoch": 0.06851237448560393,
+      "grad_norm": 1.0984941720962524,
+      "learning_rate": 0.00020546438105943513,
+      "loss": 4.8828,
+      "step": 48760
+    },
+    {
+      "epoch": 0.06852642542376751,
+      "grad_norm": 0.9066498875617981,
+      "learning_rate": 0.00020550653365181956,
+      "loss": 4.7879,
+      "step": 48770
+    },
+    {
+      "epoch": 0.06854047636193109,
+      "grad_norm": 0.957650899887085,
+      "learning_rate": 0.00020554868624420402,
+      "loss": 4.9034,
+      "step": 48780
+    },
+    {
+      "epoch": 0.06855452730009466,
+      "grad_norm": 0.9184542894363403,
+      "learning_rate": 0.00020559083883658845,
+      "loss": 4.9151,
+      "step": 48790
+    },
+    {
+      "epoch": 0.06856857823825825,
+      "grad_norm": 0.9331673383712769,
+      "learning_rate": 0.00020563299142897285,
+      "loss": 4.764,
+      "step": 48800
+    },
+    {
+      "epoch": 0.06858262917642183,
+      "grad_norm": 0.9273619055747986,
+      "learning_rate": 0.00020567514402135728,
+      "loss": 4.8304,
+      "step": 48810
+    },
+    {
+      "epoch": 0.0685966801145854,
+      "grad_norm": 0.962047815322876,
+      "learning_rate": 0.00020571729661374172,
+      "loss": 4.7868,
+      "step": 48820
+    },
+    {
+      "epoch": 0.06861073105274898,
+      "grad_norm": 0.9674386382102966,
+      "learning_rate": 0.00020575944920612615,
+      "loss": 4.8832,
+      "step": 48830
+    },
+    {
+      "epoch": 0.06862478199091256,
+      "grad_norm": 0.925841212272644,
+      "learning_rate": 0.0002058016017985106,
+      "loss": 4.8394,
+      "step": 48840
+    },
+    {
+      "epoch": 0.06863883292907613,
+      "grad_norm": 0.9452566504478455,
+      "learning_rate": 0.00020584375439089504,
+      "loss": 4.8387,
+      "step": 48850
+    },
+    {
+      "epoch": 0.06865288386723971,
+      "grad_norm": 1.0907858610153198,
+      "learning_rate": 0.00020588590698327947,
+      "loss": 4.873,
+      "step": 48860
+    },
+    {
+      "epoch": 0.06866693480540328,
+      "grad_norm": 0.9650761485099792,
+      "learning_rate": 0.00020592805957566387,
+      "loss": 4.9087,
+      "step": 48870
+    },
+    {
+      "epoch": 0.06868098574356686,
+      "grad_norm": 0.8910475373268127,
+      "learning_rate": 0.0002059702121680483,
+      "loss": 4.8725,
+      "step": 48880
+    },
+    {
+      "epoch": 0.06869503668173044,
+      "grad_norm": 0.9114874005317688,
+      "learning_rate": 0.00020601236476043273,
+      "loss": 4.9084,
+      "step": 48890
+    },
+    {
+      "epoch": 0.06870908761989403,
+      "grad_norm": 0.9370610117912292,
+      "learning_rate": 0.0002060545173528172,
+      "loss": 4.8315,
+      "step": 48900
+    },
+    {
+      "epoch": 0.0687231385580576,
+      "grad_norm": 0.9323891401290894,
+      "learning_rate": 0.00020609666994520162,
+      "loss": 4.7051,
+      "step": 48910
+    },
+    {
+      "epoch": 0.06873718949622118,
+      "grad_norm": 1.0027904510498047,
+      "learning_rate": 0.00020613882253758605,
+      "loss": 4.87,
+      "step": 48920
+    },
+    {
+      "epoch": 0.06875124043438476,
+      "grad_norm": 0.9343486428260803,
+      "learning_rate": 0.00020618097512997046,
+      "loss": 4.8427,
+      "step": 48930
+    },
+    {
+      "epoch": 0.06876529137254833,
+      "grad_norm": 0.9870249032974243,
+      "learning_rate": 0.0002062231277223549,
+      "loss": 4.8214,
+      "step": 48940
+    },
+    {
+      "epoch": 0.06877934231071191,
+      "grad_norm": 0.9833707213401794,
+      "learning_rate": 0.00020626528031473932,
+      "loss": 4.9394,
+      "step": 48950
+    },
+    {
+      "epoch": 0.06879339324887548,
+      "grad_norm": 0.9330481290817261,
+      "learning_rate": 0.00020630743290712378,
+      "loss": 4.8596,
+      "step": 48960
+    },
+    {
+      "epoch": 0.06880744418703906,
+      "grad_norm": 0.9530215859413147,
+      "learning_rate": 0.0002063495854995082,
+      "loss": 4.892,
+      "step": 48970
+    },
+    {
+      "epoch": 0.06882149512520264,
+      "grad_norm": 0.9184706807136536,
+      "learning_rate": 0.00020639173809189264,
+      "loss": 4.9649,
+      "step": 48980
+    },
+    {
+      "epoch": 0.06883554606336621,
+      "grad_norm": 0.9311974048614502,
+      "learning_rate": 0.00020643389068427707,
+      "loss": 4.9251,
+      "step": 48990
+    },
+    {
+      "epoch": 0.0688495970015298,
+      "grad_norm": 0.9431549310684204,
+      "learning_rate": 0.00020647604327666148,
+      "loss": 4.8339,
+      "step": 49000
+    },
+    {
+      "epoch": 0.06886364793969338,
+      "grad_norm": 1.0781751871109009,
+      "learning_rate": 0.00020651819586904593,
+      "loss": 4.915,
+      "step": 49010
+    },
+    {
+      "epoch": 0.06887769887785695,
+      "grad_norm": 0.9329824447631836,
+      "learning_rate": 0.00020656034846143037,
+      "loss": 4.8891,
+      "step": 49020
+    },
+    {
+      "epoch": 0.06889174981602053,
+      "grad_norm": 0.9006100296974182,
+      "learning_rate": 0.0002066025010538148,
+      "loss": 4.9299,
+      "step": 49030
+    },
+    {
+      "epoch": 0.06890580075418411,
+      "grad_norm": 0.8924393057823181,
+      "learning_rate": 0.00020664465364619923,
+      "loss": 4.7221,
+      "step": 49040
+    },
+    {
+      "epoch": 0.06891985169234768,
+      "grad_norm": 0.9394928216934204,
+      "learning_rate": 0.00020668680623858366,
+      "loss": 4.8564,
+      "step": 49050
+    },
+    {
+      "epoch": 0.06893390263051126,
+      "grad_norm": 0.9694632291793823,
+      "learning_rate": 0.00020672895883096806,
+      "loss": 4.8552,
+      "step": 49060
+    },
+    {
+      "epoch": 0.06894795356867484,
+      "grad_norm": 0.9108564853668213,
+      "learning_rate": 0.00020677111142335252,
+      "loss": 4.8766,
+      "step": 49070
+    },
+    {
+      "epoch": 0.06896200450683841,
+      "grad_norm": 0.9487257599830627,
+      "learning_rate": 0.00020681326401573695,
+      "loss": 4.7678,
+      "step": 49080
+    },
+    {
+      "epoch": 0.06897605544500199,
+      "grad_norm": 0.9499151706695557,
+      "learning_rate": 0.00020685541660812138,
+      "loss": 4.822,
+      "step": 49090
+    },
+    {
+      "epoch": 0.06899010638316556,
+      "grad_norm": 1.0061317682266235,
+      "learning_rate": 0.00020689756920050582,
+      "loss": 4.8902,
+      "step": 49100
+    },
+    {
+      "epoch": 0.06900415732132915,
+      "grad_norm": 0.9726933240890503,
+      "learning_rate": 0.00020693972179289025,
+      "loss": 4.8946,
+      "step": 49110
+    },
+    {
+      "epoch": 0.06901820825949273,
+      "grad_norm": 0.8891914486885071,
+      "learning_rate": 0.00020698187438527465,
+      "loss": 4.9583,
+      "step": 49120
+    },
+    {
+      "epoch": 0.0690322591976563,
+      "grad_norm": 0.9320970773696899,
+      "learning_rate": 0.0002070240269776591,
+      "loss": 4.825,
+      "step": 49130
+    },
+    {
+      "epoch": 0.06904631013581988,
+      "grad_norm": 0.8775174021720886,
+      "learning_rate": 0.00020706617957004354,
+      "loss": 4.8636,
+      "step": 49140
+    },
+    {
+      "epoch": 0.06906036107398346,
+      "grad_norm": 0.9681304097175598,
+      "learning_rate": 0.00020710411690318955,
+      "loss": 4.8397,
+      "step": 49150
+    },
+    {
+      "epoch": 0.06907441201214704,
+      "grad_norm": 0.9166766405105591,
+      "learning_rate": 0.00020714626949557395,
+      "loss": 4.8822,
+      "step": 49160
+    },
+    {
+      "epoch": 0.06908846295031061,
+      "grad_norm": 0.9451488852500916,
+      "learning_rate": 0.00020718842208795838,
+      "loss": 4.9481,
+      "step": 49170
+    },
+    {
+      "epoch": 0.06910251388847419,
+      "grad_norm": 1.025514006614685,
+      "learning_rate": 0.00020723057468034282,
+      "loss": 4.7886,
+      "step": 49180
+    },
+    {
+      "epoch": 0.06911656482663776,
+      "grad_norm": 1.0833834409713745,
+      "learning_rate": 0.00020727272727272725,
+      "loss": 4.8875,
+      "step": 49190
+    },
+    {
+      "epoch": 0.06913061576480134,
+      "grad_norm": 0.9185973405838013,
+      "learning_rate": 0.0002073148798651117,
+      "loss": 4.9811,
+      "step": 49200
+    },
+    {
+      "epoch": 0.06914466670296493,
+      "grad_norm": 0.9156882762908936,
+      "learning_rate": 0.00020735703245749614,
+      "loss": 4.8988,
+      "step": 49210
+    },
+    {
+      "epoch": 0.0691587176411285,
+      "grad_norm": 0.9855953454971313,
+      "learning_rate": 0.00020739918504988054,
+      "loss": 4.7664,
+      "step": 49220
+    },
+    {
+      "epoch": 0.06917276857929208,
+      "grad_norm": 0.9379180073738098,
+      "learning_rate": 0.00020744133764226497,
+      "loss": 4.7998,
+      "step": 49230
+    },
+    {
+      "epoch": 0.06918681951745566,
+      "grad_norm": 0.901465892791748,
+      "learning_rate": 0.0002074834902346494,
+      "loss": 4.9366,
+      "step": 49240
+    },
+    {
+      "epoch": 0.06920087045561923,
+      "grad_norm": 0.9280766844749451,
+      "learning_rate": 0.00020752564282703383,
+      "loss": 4.9037,
+      "step": 49250
+    },
+    {
+      "epoch": 0.06921492139378281,
+      "grad_norm": 1.0117437839508057,
+      "learning_rate": 0.0002075677954194183,
+      "loss": 4.8711,
+      "step": 49260
+    },
+    {
+      "epoch": 0.06922897233194639,
+      "grad_norm": 0.936126708984375,
+      "learning_rate": 0.00020760994801180272,
+      "loss": 4.8852,
+      "step": 49270
+    },
+    {
+      "epoch": 0.06924302327010996,
+      "grad_norm": 0.907888650894165,
+      "learning_rate": 0.00020765210060418715,
+      "loss": 4.8348,
+      "step": 49280
+    },
+    {
+      "epoch": 0.06925707420827354,
+      "grad_norm": 0.9258394241333008,
+      "learning_rate": 0.00020769425319657156,
+      "loss": 4.912,
+      "step": 49290
+    },
+    {
+      "epoch": 0.06927112514643712,
+      "grad_norm": 0.9378911852836609,
+      "learning_rate": 0.000207736405788956,
+      "loss": 4.9453,
+      "step": 49300
+    },
+    {
+      "epoch": 0.0692851760846007,
+      "grad_norm": 0.9445339441299438,
+      "learning_rate": 0.00020777855838134042,
+      "loss": 4.7425,
+      "step": 49310
+    },
+    {
+      "epoch": 0.06929922702276428,
+      "grad_norm": 0.9408323168754578,
+      "learning_rate": 0.00020782071097372488,
+      "loss": 4.8162,
+      "step": 49320
+    },
+    {
+      "epoch": 0.06931327796092786,
+      "grad_norm": 0.9732528328895569,
+      "learning_rate": 0.0002078628635661093,
+      "loss": 4.8715,
+      "step": 49330
+    },
+    {
+      "epoch": 0.06932732889909143,
+      "grad_norm": 0.9590655565261841,
+      "learning_rate": 0.00020790501615849374,
+      "loss": 4.7933,
+      "step": 49340
+    },
+    {
+      "epoch": 0.06934137983725501,
+      "grad_norm": 0.9126473069190979,
+      "learning_rate": 0.00020794716875087815,
+      "loss": 4.8097,
+      "step": 49350
+    },
+    {
+      "epoch": 0.06935543077541859,
+      "grad_norm": 0.9477401375770569,
+      "learning_rate": 0.00020798932134326258,
+      "loss": 4.9911,
+      "step": 49360
+    },
+    {
+      "epoch": 0.06936948171358216,
+      "grad_norm": 0.897299587726593,
+      "learning_rate": 0.000208031473935647,
+      "loss": 4.9375,
+      "step": 49370
+    },
+    {
+      "epoch": 0.06938353265174574,
+      "grad_norm": 0.951884388923645,
+      "learning_rate": 0.00020807362652803147,
+      "loss": 4.78,
+      "step": 49380
+    },
+    {
+      "epoch": 0.06939758358990931,
+      "grad_norm": 0.9042193293571472,
+      "learning_rate": 0.0002081157791204159,
+      "loss": 4.8414,
+      "step": 49390
+    },
+    {
+      "epoch": 0.06941163452807289,
+      "grad_norm": 0.9469078183174133,
+      "learning_rate": 0.00020815793171280033,
+      "loss": 4.9658,
+      "step": 49400
+    },
+    {
+      "epoch": 0.06942568546623647,
+      "grad_norm": 0.9061809778213501,
+      "learning_rate": 0.00020820008430518476,
+      "loss": 4.7509,
+      "step": 49410
+    },
+    {
+      "epoch": 0.06943973640440006,
+      "grad_norm": 0.9353413581848145,
+      "learning_rate": 0.00020824223689756916,
+      "loss": 4.8421,
+      "step": 49420
+    },
+    {
+      "epoch": 0.06945378734256363,
+      "grad_norm": 0.9837194085121155,
+      "learning_rate": 0.0002082843894899536,
+      "loss": 4.8281,
+      "step": 49430
+    },
+    {
+      "epoch": 0.06946783828072721,
+      "grad_norm": 0.9294975996017456,
+      "learning_rate": 0.00020832654208233805,
+      "loss": 4.7543,
+      "step": 49440
+    },
+    {
+      "epoch": 0.06948188921889079,
+      "grad_norm": 0.9024710059165955,
+      "learning_rate": 0.00020836869467472248,
+      "loss": 4.8502,
+      "step": 49450
+    },
+    {
+      "epoch": 0.06949594015705436,
+      "grad_norm": 0.9258474707603455,
+      "learning_rate": 0.00020841084726710692,
+      "loss": 4.9781,
+      "step": 49460
+    },
+    {
+      "epoch": 0.06950999109521794,
+      "grad_norm": 1.0224697589874268,
+      "learning_rate": 0.00020845299985949135,
+      "loss": 4.766,
+      "step": 49470
+    },
+    {
+      "epoch": 0.06952404203338151,
+      "grad_norm": 0.9513781070709229,
+      "learning_rate": 0.00020849515245187575,
+      "loss": 4.879,
+      "step": 49480
+    },
+    {
+      "epoch": 0.06953809297154509,
+      "grad_norm": 0.9261512160301208,
+      "learning_rate": 0.00020853730504426018,
+      "loss": 4.7467,
+      "step": 49490
+    },
+    {
+      "epoch": 0.06955214390970867,
+      "grad_norm": 0.9599387049674988,
+      "learning_rate": 0.00020857945763664464,
+      "loss": 4.9011,
+      "step": 49500
+    },
+    {
+      "epoch": 0.06956619484787224,
+      "grad_norm": 0.908785343170166,
+      "learning_rate": 0.00020862161022902907,
+      "loss": 4.8237,
+      "step": 49510
+    },
+    {
+      "epoch": 0.06958024578603583,
+      "grad_norm": 0.8983234167098999,
+      "learning_rate": 0.0002086637628214135,
+      "loss": 4.9282,
+      "step": 49520
+    },
+    {
+      "epoch": 0.06959429672419941,
+      "grad_norm": 0.8948048949241638,
+      "learning_rate": 0.00020870591541379793,
+      "loss": 4.9866,
+      "step": 49530
+    },
+    {
+      "epoch": 0.06960834766236298,
+      "grad_norm": 0.8868510127067566,
+      "learning_rate": 0.00020874806800618237,
+      "loss": 4.8474,
+      "step": 49540
+    },
+    {
+      "epoch": 0.06962239860052656,
+      "grad_norm": 1.2011126279830933,
+      "learning_rate": 0.00020879022059856677,
+      "loss": 4.9062,
+      "step": 49550
+    },
+    {
+      "epoch": 0.06963644953869014,
+      "grad_norm": 0.9646673202514648,
+      "learning_rate": 0.00020883237319095123,
+      "loss": 4.7145,
+      "step": 49560
+    },
+    {
+      "epoch": 0.06965050047685371,
+      "grad_norm": 0.9322654604911804,
+      "learning_rate": 0.00020887452578333566,
+      "loss": 4.9656,
+      "step": 49570
+    },
+    {
+      "epoch": 0.06966455141501729,
+      "grad_norm": 0.9516728520393372,
+      "learning_rate": 0.0002089166783757201,
+      "loss": 4.8092,
+      "step": 49580
+    },
+    {
+      "epoch": 0.06967860235318087,
+      "grad_norm": 0.9191582202911377,
+      "learning_rate": 0.00020895883096810452,
+      "loss": 4.8314,
+      "step": 49590
+    },
+    {
+      "epoch": 0.06969265329134444,
+      "grad_norm": 0.9767720699310303,
+      "learning_rate": 0.00020900098356048898,
+      "loss": 4.9208,
+      "step": 49600
+    },
+    {
+      "epoch": 0.06970670422950802,
+      "grad_norm": 0.9100189208984375,
+      "learning_rate": 0.00020904313615287336,
+      "loss": 4.8623,
+      "step": 49610
+    },
+    {
+      "epoch": 0.06972075516767161,
+      "grad_norm": 0.9272692203521729,
+      "learning_rate": 0.00020908528874525781,
+      "loss": 4.7475,
+      "step": 49620
+    },
+    {
+      "epoch": 0.06973480610583518,
+      "grad_norm": 0.9022219181060791,
+      "learning_rate": 0.00020912744133764225,
+      "loss": 4.8501,
+      "step": 49630
+    },
+    {
+      "epoch": 0.06974885704399876,
+      "grad_norm": 0.9093071222305298,
+      "learning_rate": 0.00020916959393002668,
+      "loss": 4.9728,
+      "step": 49640
+    },
+    {
+      "epoch": 0.06976290798216234,
+      "grad_norm": 0.9411072134971619,
+      "learning_rate": 0.0002092117465224111,
+      "loss": 4.7739,
+      "step": 49650
+    },
+    {
+      "epoch": 0.06977695892032591,
+      "grad_norm": 0.9116350412368774,
+      "learning_rate": 0.00020925389911479557,
+      "loss": 4.8922,
+      "step": 49660
+    },
+    {
+      "epoch": 0.06979100985848949,
+      "grad_norm": 0.9344385266304016,
+      "learning_rate": 0.00020929605170717997,
+      "loss": 4.8928,
+      "step": 49670
+    },
+    {
+      "epoch": 0.06980506079665307,
+      "grad_norm": 0.9087563157081604,
+      "learning_rate": 0.0002093382042995644,
+      "loss": 4.818,
+      "step": 49680
+    },
+    {
+      "epoch": 0.06981911173481664,
+      "grad_norm": 0.8912515044212341,
+      "learning_rate": 0.00020938035689194883,
+      "loss": 4.8745,
+      "step": 49690
+    },
+    {
+      "epoch": 0.06983316267298022,
+      "grad_norm": 0.9308001399040222,
+      "learning_rate": 0.00020942250948433326,
+      "loss": 4.8263,
+      "step": 49700
+    },
+    {
+      "epoch": 0.0698472136111438,
+      "grad_norm": 0.910260021686554,
+      "learning_rate": 0.0002094646620767177,
+      "loss": 4.753,
+      "step": 49710
+    },
+    {
+      "epoch": 0.06986126454930737,
+      "grad_norm": 0.9524611830711365,
+      "learning_rate": 0.00020950681466910215,
+      "loss": 4.8124,
+      "step": 49720
+    },
+    {
+      "epoch": 0.06987531548747096,
+      "grad_norm": 0.8876780271530151,
+      "learning_rate": 0.00020954896726148658,
+      "loss": 4.934,
+      "step": 49730
+    },
+    {
+      "epoch": 0.06988936642563454,
+      "grad_norm": 0.9967767596244812,
+      "learning_rate": 0.000209591119853871,
+      "loss": 4.8565,
+      "step": 49740
+    },
+    {
+      "epoch": 0.06990341736379811,
+      "grad_norm": 0.9247635006904602,
+      "learning_rate": 0.000209629057187017,
+      "loss": 4.8617,
+      "step": 49750
+    },
+    {
+      "epoch": 0.06991746830196169,
+      "grad_norm": 0.886182427406311,
+      "learning_rate": 0.00020967120977940143,
+      "loss": 4.9107,
+      "step": 49760
+    },
+    {
+      "epoch": 0.06993151924012526,
+      "grad_norm": 0.963778555393219,
+      "learning_rate": 0.00020971336237178583,
+      "loss": 4.8586,
+      "step": 49770
+    },
+    {
+      "epoch": 0.06994557017828884,
+      "grad_norm": 0.9098008275032043,
+      "learning_rate": 0.00020975551496417026,
+      "loss": 4.8441,
+      "step": 49780
+    },
+    {
+      "epoch": 0.06995962111645242,
+      "grad_norm": 0.9119469523429871,
+      "learning_rate": 0.0002097976675565547,
+      "loss": 4.8072,
+      "step": 49790
+    },
+    {
+      "epoch": 0.06997367205461599,
+      "grad_norm": 0.920755922794342,
+      "learning_rate": 0.00020983982014893915,
+      "loss": 4.8332,
+      "step": 49800
+    },
+    {
+      "epoch": 0.06998772299277957,
+      "grad_norm": 0.9371573328971863,
+      "learning_rate": 0.00020988197274132359,
+      "loss": 4.7674,
+      "step": 49810
+    },
+    {
+      "epoch": 0.07000177393094315,
+      "grad_norm": 0.926888108253479,
+      "learning_rate": 0.00020992412533370802,
+      "loss": 4.7776,
+      "step": 49820
+    },
+    {
+      "epoch": 0.07001582486910674,
+      "grad_norm": 0.9252434968948364,
+      "learning_rate": 0.00020996627792609245,
+      "loss": 4.8627,
+      "step": 49830
+    },
+    {
+      "epoch": 0.07002987580727031,
+      "grad_norm": 0.8703329563140869,
+      "learning_rate": 0.00021000843051847685,
+      "loss": 5.0105,
+      "step": 49840
+    },
+    {
+      "epoch": 0.07004392674543389,
+      "grad_norm": 0.9439218640327454,
+      "learning_rate": 0.00021005058311086128,
+      "loss": 4.7864,
+      "step": 49850
+    },
+    {
+      "epoch": 0.07005797768359746,
+      "grad_norm": 0.9055263996124268,
+      "learning_rate": 0.00021009273570324574,
+      "loss": 4.8109,
+      "step": 49860
+    },
+    {
+      "epoch": 0.07007202862176104,
+      "grad_norm": 0.9342459440231323,
+      "learning_rate": 0.00021013488829563017,
+      "loss": 4.8661,
+      "step": 49870
+    },
+    {
+      "epoch": 0.07008607955992462,
+      "grad_norm": 0.910137414932251,
+      "learning_rate": 0.0002101770408880146,
+      "loss": 4.8083,
+      "step": 49880
+    },
+    {
+      "epoch": 0.07010013049808819,
+      "grad_norm": 0.9078051447868347,
+      "learning_rate": 0.00021021919348039903,
+      "loss": 5.0483,
+      "step": 49890
+    },
+    {
+      "epoch": 0.07011418143625177,
+      "grad_norm": 0.9174194931983948,
+      "learning_rate": 0.00021026134607278344,
+      "loss": 4.9562,
+      "step": 49900
+    },
+    {
+      "epoch": 0.07012823237441534,
+      "grad_norm": 0.9276567101478577,
+      "learning_rate": 0.00021030349866516787,
+      "loss": 4.8575,
+      "step": 49910
+    },
+    {
+      "epoch": 0.07014228331257892,
+      "grad_norm": 0.9092664122581482,
+      "learning_rate": 0.00021034565125755233,
+      "loss": 4.673,
+      "step": 49920
+    },
+    {
+      "epoch": 0.07015633425074251,
+      "grad_norm": 0.9359709620475769,
+      "learning_rate": 0.00021038780384993676,
+      "loss": 4.8121,
+      "step": 49930
+    },
+    {
+      "epoch": 0.07017038518890609,
+      "grad_norm": 0.9264384508132935,
+      "learning_rate": 0.0002104299564423212,
+      "loss": 4.825,
+      "step": 49940
+    },
+    {
+      "epoch": 0.07018443612706966,
+      "grad_norm": 0.881436824798584,
+      "learning_rate": 0.00021047210903470562,
+      "loss": 4.7949,
+      "step": 49950
+    },
+    {
+      "epoch": 0.07019848706523324,
+      "grad_norm": 0.9295757412910461,
+      "learning_rate": 0.00021051426162709005,
+      "loss": 4.8378,
+      "step": 49960
+    },
+    {
+      "epoch": 0.07021253800339682,
+      "grad_norm": 0.9156924486160278,
+      "learning_rate": 0.00021055641421947446,
+      "loss": 4.9212,
+      "step": 49970
+    },
+    {
+      "epoch": 0.07022658894156039,
+      "grad_norm": 0.9250329732894897,
+      "learning_rate": 0.00021059856681185892,
+      "loss": 4.8159,
+      "step": 49980
+    },
+    {
+      "epoch": 0.07024063987972397,
+      "grad_norm": 0.88548743724823,
+      "learning_rate": 0.00021064071940424335,
+      "loss": 4.858,
+      "step": 49990
+    },
+    {
+      "epoch": 0.07025469081788754,
+      "grad_norm": 0.9146063327789307,
+      "learning_rate": 0.00021068287199662778,
+      "loss": 4.8865,
+      "step": 50000
+    },
+    {
+      "epoch": 0.07026874175605112,
+      "grad_norm": 0.9075859785079956,
+      "learning_rate": 0.0002107250245890122,
+      "loss": 4.8735,
+      "step": 50010
+    },
+    {
+      "epoch": 0.0702827926942147,
+      "grad_norm": 0.9209066033363342,
+      "learning_rate": 0.00021076717718139664,
+      "loss": 4.8269,
+      "step": 50020
+    },
+    {
+      "epoch": 0.07029684363237827,
+      "grad_norm": 0.9381097555160522,
+      "learning_rate": 0.00021080932977378104,
+      "loss": 4.9601,
+      "step": 50030
+    },
+    {
+      "epoch": 0.07031089457054186,
+      "grad_norm": 0.8956311941146851,
+      "learning_rate": 0.0002108514823661655,
+      "loss": 4.9086,
+      "step": 50040
+    },
+    {
+      "epoch": 0.07032494550870544,
+      "grad_norm": 0.9148889780044556,
+      "learning_rate": 0.00021089363495854993,
+      "loss": 4.7701,
+      "step": 50050
+    },
+    {
+      "epoch": 0.07033899644686901,
+      "grad_norm": 0.9117963314056396,
+      "learning_rate": 0.00021093578755093436,
+      "loss": 4.9762,
+      "step": 50060
+    },
+    {
+      "epoch": 0.07035304738503259,
+      "grad_norm": 0.9547762870788574,
+      "learning_rate": 0.0002109779401433188,
+      "loss": 4.8947,
+      "step": 50070
+    },
+    {
+      "epoch": 0.07036709832319617,
+      "grad_norm": 0.9054858088493347,
+      "learning_rate": 0.00021102009273570323,
+      "loss": 4.8054,
+      "step": 50080
+    },
+    {
+      "epoch": 0.07038114926135974,
+      "grad_norm": 0.8751288652420044,
+      "learning_rate": 0.00021106224532808763,
+      "loss": 4.8845,
+      "step": 50090
+    },
+    {
+      "epoch": 0.07039520019952332,
+      "grad_norm": 0.9098647236824036,
+      "learning_rate": 0.0002111043979204721,
+      "loss": 4.7945,
+      "step": 50100
+    },
+    {
+      "epoch": 0.0704092511376869,
+      "grad_norm": 0.9443636536598206,
+      "learning_rate": 0.00021114655051285652,
+      "loss": 4.8985,
+      "step": 50110
+    },
+    {
+      "epoch": 0.07042330207585047,
+      "grad_norm": 0.9238690137863159,
+      "learning_rate": 0.00021118870310524095,
+      "loss": 4.78,
+      "step": 50120
+    },
+    {
+      "epoch": 0.07043735301401405,
+      "grad_norm": 0.9531887173652649,
+      "learning_rate": 0.00021123085569762538,
+      "loss": 4.7975,
+      "step": 50130
+    },
+    {
+      "epoch": 0.07045140395217764,
+      "grad_norm": 0.9312260150909424,
+      "learning_rate": 0.00021127300829000981,
+      "loss": 4.8473,
+      "step": 50140
+    },
+    {
+      "epoch": 0.07046545489034121,
+      "grad_norm": 0.9800652861595154,
+      "learning_rate": 0.00021131516088239427,
+      "loss": 4.8799,
+      "step": 50150
+    },
+    {
+      "epoch": 0.07047950582850479,
+      "grad_norm": 0.8918007016181946,
+      "learning_rate": 0.00021135731347477868,
+      "loss": 4.967,
+      "step": 50160
+    },
+    {
+      "epoch": 0.07049355676666837,
+      "grad_norm": 0.9277655482292175,
+      "learning_rate": 0.0002113994660671631,
+      "loss": 4.7375,
+      "step": 50170
+    },
+    {
+      "epoch": 0.07050760770483194,
+      "grad_norm": 0.9381228089332581,
+      "learning_rate": 0.00021144161865954754,
+      "loss": 4.8449,
+      "step": 50180
+    },
+    {
+      "epoch": 0.07052165864299552,
+      "grad_norm": 0.9662410616874695,
+      "learning_rate": 0.00021148377125193197,
+      "loss": 4.8809,
+      "step": 50190
+    },
+    {
+      "epoch": 0.0705357095811591,
+      "grad_norm": 0.9184268116950989,
+      "learning_rate": 0.0002115259238443164,
+      "loss": 4.9633,
+      "step": 50200
+    },
+    {
+      "epoch": 0.07054976051932267,
+      "grad_norm": 0.9069191217422485,
+      "learning_rate": 0.00021156807643670086,
+      "loss": 4.8526,
+      "step": 50210
+    },
+    {
+      "epoch": 0.07056381145748625,
+      "grad_norm": 0.908988356590271,
+      "learning_rate": 0.00021161022902908526,
+      "loss": 4.7715,
+      "step": 50220
+    },
+    {
+      "epoch": 0.07057786239564982,
+      "grad_norm": 0.9809160828590393,
+      "learning_rate": 0.0002116523816214697,
+      "loss": 4.9594,
+      "step": 50230
+    },
+    {
+      "epoch": 0.07059191333381341,
+      "grad_norm": 0.9043667912483215,
+      "learning_rate": 0.00021169453421385413,
+      "loss": 4.8035,
+      "step": 50240
+    },
+    {
+      "epoch": 0.07060596427197699,
+      "grad_norm": 1.00430166721344,
+      "learning_rate": 0.00021173668680623856,
+      "loss": 4.8976,
+      "step": 50250
+    },
+    {
+      "epoch": 0.07062001521014057,
+      "grad_norm": 0.9381590485572815,
+      "learning_rate": 0.00021177883939862301,
+      "loss": 4.8531,
+      "step": 50260
+    },
+    {
+      "epoch": 0.07063406614830414,
+      "grad_norm": 0.8797604441642761,
+      "learning_rate": 0.00021182099199100745,
+      "loss": 4.8615,
+      "step": 50270
+    },
+    {
+      "epoch": 0.07064811708646772,
+      "grad_norm": 0.8880553245544434,
+      "learning_rate": 0.00021186314458339188,
+      "loss": 4.8398,
+      "step": 50280
+    },
+    {
+      "epoch": 0.0706621680246313,
+      "grad_norm": 0.9292892217636108,
+      "learning_rate": 0.00021190529717577628,
+      "loss": 4.9145,
+      "step": 50290
+    },
+    {
+      "epoch": 0.07067621896279487,
+      "grad_norm": 0.9727832078933716,
+      "learning_rate": 0.0002119474497681607,
+      "loss": 4.7891,
+      "step": 50300
+    },
+    {
+      "epoch": 0.07069026990095845,
+      "grad_norm": 0.9402398467063904,
+      "learning_rate": 0.00021198960236054514,
+      "loss": 4.8532,
+      "step": 50310
+    },
+    {
+      "epoch": 0.07070432083912202,
+      "grad_norm": 0.9445501565933228,
+      "learning_rate": 0.0002120317549529296,
+      "loss": 4.8188,
+      "step": 50320
+    },
+    {
+      "epoch": 0.0707183717772856,
+      "grad_norm": 0.9223546385765076,
+      "learning_rate": 0.00021207390754531403,
+      "loss": 4.9161,
+      "step": 50330
+    },
+    {
+      "epoch": 0.07073242271544918,
+      "grad_norm": 0.9386568069458008,
+      "learning_rate": 0.00021211606013769846,
+      "loss": 4.9415,
+      "step": 50340
+    },
+    {
+      "epoch": 0.07074647365361277,
+      "grad_norm": 0.9313529133796692,
+      "learning_rate": 0.00021215821273008287,
+      "loss": 4.8792,
+      "step": 50350
+    },
+    {
+      "epoch": 0.07076052459177634,
+      "grad_norm": 0.9341327548027039,
+      "learning_rate": 0.0002122003653224673,
+      "loss": 4.8473,
+      "step": 50360
+    },
+    {
+      "epoch": 0.07077457552993992,
+      "grad_norm": 0.8974583745002747,
+      "learning_rate": 0.00021224251791485173,
+      "loss": 4.919,
+      "step": 50370
+    },
+    {
+      "epoch": 0.0707886264681035,
+      "grad_norm": 0.9477220177650452,
+      "learning_rate": 0.0002122846705072362,
+      "loss": 4.9454,
+      "step": 50380
+    },
+    {
+      "epoch": 0.07080267740626707,
+      "grad_norm": 0.9393336176872253,
+      "learning_rate": 0.00021232682309962062,
+      "loss": 5.068,
+      "step": 50390
+    },
+    {
+      "epoch": 0.07081672834443065,
+      "grad_norm": 0.9357321858406067,
+      "learning_rate": 0.00021236897569200505,
+      "loss": 4.794,
+      "step": 50400
+    },
+    {
+      "epoch": 0.07083077928259422,
+      "grad_norm": 0.9102789163589478,
+      "learning_rate": 0.00021241112828438948,
+      "loss": 4.9233,
+      "step": 50410
+    },
+    {
+      "epoch": 0.0708448302207578,
+      "grad_norm": 0.9070646166801453,
+      "learning_rate": 0.00021245328087677389,
+      "loss": 4.9097,
+      "step": 50420
+    },
+    {
+      "epoch": 0.07085888115892137,
+      "grad_norm": 0.8896940350532532,
+      "learning_rate": 0.00021249543346915832,
+      "loss": 4.9443,
+      "step": 50430
+    },
+    {
+      "epoch": 0.07087293209708495,
+      "grad_norm": 0.9510523080825806,
+      "learning_rate": 0.00021253758606154278,
+      "loss": 4.7866,
+      "step": 50440
+    },
+    {
+      "epoch": 0.07088698303524854,
+      "grad_norm": 0.9242532253265381,
+      "learning_rate": 0.0002125797386539272,
+      "loss": 4.8039,
+      "step": 50450
+    },
+    {
+      "epoch": 0.07090103397341212,
+      "grad_norm": 0.9695941209793091,
+      "learning_rate": 0.00021262189124631164,
+      "loss": 4.859,
+      "step": 50460
+    },
+    {
+      "epoch": 0.0709150849115757,
+      "grad_norm": 0.8837727904319763,
+      "learning_rate": 0.00021266404383869607,
+      "loss": 4.9475,
+      "step": 50470
+    },
+    {
+      "epoch": 0.07092913584973927,
+      "grad_norm": 0.9693347811698914,
+      "learning_rate": 0.00021270619643108047,
+      "loss": 4.8531,
+      "step": 50480
+    },
+    {
+      "epoch": 0.07094318678790285,
+      "grad_norm": 0.9365562796592712,
+      "learning_rate": 0.0002127483490234649,
+      "loss": 4.8912,
+      "step": 50490
+    },
+    {
+      "epoch": 0.07095723772606642,
+      "grad_norm": 0.8954295516014099,
+      "learning_rate": 0.00021279050161584936,
+      "loss": 4.8255,
+      "step": 50500
+    },
+    {
+      "epoch": 0.07097128866423,
+      "grad_norm": 0.9566641449928284,
+      "learning_rate": 0.0002128326542082338,
+      "loss": 4.9466,
+      "step": 50510
+    },
+    {
+      "epoch": 0.07098533960239357,
+      "grad_norm": 0.8944385647773743,
+      "learning_rate": 0.00021287480680061823,
+      "loss": 4.9177,
+      "step": 50520
+    },
+    {
+      "epoch": 0.07099939054055715,
+      "grad_norm": 0.8918579816818237,
+      "learning_rate": 0.00021291695939300266,
+      "loss": 4.7941,
+      "step": 50530
+    },
+    {
+      "epoch": 0.07101344147872073,
+      "grad_norm": 0.8937610387802124,
+      "learning_rate": 0.0002129591119853871,
+      "loss": 4.8411,
+      "step": 50540
+    },
+    {
+      "epoch": 0.07102749241688432,
+      "grad_norm": 0.9060569405555725,
+      "learning_rate": 0.0002130012645777715,
+      "loss": 4.7955,
+      "step": 50550
+    },
+    {
+      "epoch": 0.07104154335504789,
+      "grad_norm": 0.9308838248252869,
+      "learning_rate": 0.00021304341717015595,
+      "loss": 4.8368,
+      "step": 50560
+    },
+    {
+      "epoch": 0.07105559429321147,
+      "grad_norm": 0.8938664793968201,
+      "learning_rate": 0.00021308556976254038,
+      "loss": 4.8399,
+      "step": 50570
+    },
+    {
+      "epoch": 0.07106964523137504,
+      "grad_norm": 0.8650916814804077,
+      "learning_rate": 0.0002131277223549248,
+      "loss": 4.9222,
+      "step": 50580
+    },
+    {
+      "epoch": 0.07108369616953862,
+      "grad_norm": 0.9728739857673645,
+      "learning_rate": 0.00021316987494730924,
+      "loss": 4.7345,
+      "step": 50590
+    },
+    {
+      "epoch": 0.0710977471077022,
+      "grad_norm": 0.912001371383667,
+      "learning_rate": 0.00021321202753969367,
+      "loss": 4.88,
+      "step": 50600
+    },
+    {
+      "epoch": 0.07111179804586577,
+      "grad_norm": 0.9506505131721497,
+      "learning_rate": 0.00021325418013207808,
+      "loss": 4.8297,
+      "step": 50610
+    },
+    {
+      "epoch": 0.07112584898402935,
+      "grad_norm": 0.9795990586280823,
+      "learning_rate": 0.00021329633272446254,
+      "loss": 4.8647,
+      "step": 50620
+    },
+    {
+      "epoch": 0.07113989992219293,
+      "grad_norm": 0.9694528579711914,
+      "learning_rate": 0.00021333848531684697,
+      "loss": 4.8235,
+      "step": 50630
+    },
+    {
+      "epoch": 0.0711539508603565,
+      "grad_norm": 0.9510214924812317,
+      "learning_rate": 0.0002133806379092314,
+      "loss": 4.9937,
+      "step": 50640
+    },
+    {
+      "epoch": 0.07116800179852008,
+      "grad_norm": 0.9015481472015381,
+      "learning_rate": 0.00021342279050161583,
+      "loss": 4.8967,
+      "step": 50650
+    },
+    {
+      "epoch": 0.07118205273668367,
+      "grad_norm": 0.8986161947250366,
+      "learning_rate": 0.00021346494309400026,
+      "loss": 4.9995,
+      "step": 50660
+    },
+    {
+      "epoch": 0.07119610367484724,
+      "grad_norm": 0.8803780674934387,
+      "learning_rate": 0.00021350709568638472,
+      "loss": 4.8779,
+      "step": 50670
+    },
+    {
+      "epoch": 0.07121015461301082,
+      "grad_norm": 0.893484354019165,
+      "learning_rate": 0.00021354924827876912,
+      "loss": 5.0183,
+      "step": 50680
+    },
+    {
+      "epoch": 0.0712242055511744,
+      "grad_norm": 0.9707047343254089,
+      "learning_rate": 0.00021359140087115355,
+      "loss": 4.8389,
+      "step": 50690
+    },
+    {
+      "epoch": 0.07123825648933797,
+      "grad_norm": 0.8995065689086914,
+      "learning_rate": 0.00021363355346353799,
+      "loss": 4.7205,
+      "step": 50700
+    },
+    {
+      "epoch": 0.07125230742750155,
+      "grad_norm": 1.0342682600021362,
+      "learning_rate": 0.00021367570605592242,
+      "loss": 4.7475,
+      "step": 50710
+    },
+    {
+      "epoch": 0.07126635836566513,
+      "grad_norm": 0.9065954685211182,
+      "learning_rate": 0.00021371785864830685,
+      "loss": 4.8589,
+      "step": 50720
+    },
+    {
+      "epoch": 0.0712804093038287,
+      "grad_norm": 0.9274217486381531,
+      "learning_rate": 0.0002137600112406913,
+      "loss": 4.9241,
+      "step": 50730
+    },
+    {
+      "epoch": 0.07129446024199228,
+      "grad_norm": 0.9372475743293762,
+      "learning_rate": 0.0002138021638330757,
+      "loss": 4.7475,
+      "step": 50740
+    },
+    {
+      "epoch": 0.07130851118015585,
+      "grad_norm": 0.8814763426780701,
+      "learning_rate": 0.00021384431642546014,
+      "loss": 4.9215,
+      "step": 50750
+    },
+    {
+      "epoch": 0.07132256211831944,
+      "grad_norm": 0.9730595946311951,
+      "learning_rate": 0.00021388646901784457,
+      "loss": 4.9178,
+      "step": 50760
+    },
+    {
+      "epoch": 0.07133661305648302,
+      "grad_norm": 0.9046145081520081,
+      "learning_rate": 0.000213928621610229,
+      "loss": 4.7846,
+      "step": 50770
+    },
+    {
+      "epoch": 0.0713506639946466,
+      "grad_norm": 0.8996490240097046,
+      "learning_rate": 0.00021397077420261344,
+      "loss": 4.8804,
+      "step": 50780
+    },
+    {
+      "epoch": 0.07136471493281017,
+      "grad_norm": 0.9104477763175964,
+      "learning_rate": 0.0002140129267949979,
+      "loss": 5.0551,
+      "step": 50790
+    },
+    {
+      "epoch": 0.07137876587097375,
+      "grad_norm": 0.9244865775108337,
+      "learning_rate": 0.0002140550793873823,
+      "loss": 4.9711,
+      "step": 50800
+    },
+    {
+      "epoch": 0.07139281680913732,
+      "grad_norm": 0.9118911623954773,
+      "learning_rate": 0.00021409723197976673,
+      "loss": 4.9435,
+      "step": 50810
+    },
+    {
+      "epoch": 0.0714068677473009,
+      "grad_norm": 0.8913031816482544,
+      "learning_rate": 0.00021413938457215116,
+      "loss": 4.7714,
+      "step": 50820
+    },
+    {
+      "epoch": 0.07142091868546448,
+      "grad_norm": 0.9230485558509827,
+      "learning_rate": 0.0002141815371645356,
+      "loss": 4.8122,
+      "step": 50830
+    },
+    {
+      "epoch": 0.07143496962362805,
+      "grad_norm": 0.9266636967658997,
+      "learning_rate": 0.00021422368975692005,
+      "loss": 4.8106,
+      "step": 50840
+    },
+    {
+      "epoch": 0.07144902056179163,
+      "grad_norm": 0.9038143157958984,
+      "learning_rate": 0.00021426584234930448,
+      "loss": 4.8459,
+      "step": 50850
+    },
+    {
+      "epoch": 0.07146307149995522,
+      "grad_norm": 0.9343311190605164,
+      "learning_rate": 0.0002143079949416889,
+      "loss": 4.8616,
+      "step": 50860
+    },
+    {
+      "epoch": 0.0714771224381188,
+      "grad_norm": 0.8671455979347229,
+      "learning_rate": 0.00021435014753407332,
+      "loss": 4.8417,
+      "step": 50870
+    },
+    {
+      "epoch": 0.07149117337628237,
+      "grad_norm": 0.9161444306373596,
+      "learning_rate": 0.00021439230012645775,
+      "loss": 4.7486,
+      "step": 50880
+    },
+    {
+      "epoch": 0.07150522431444595,
+      "grad_norm": 0.8902353644371033,
+      "learning_rate": 0.00021443445271884218,
+      "loss": 4.9153,
+      "step": 50890
+    },
+    {
+      "epoch": 0.07151927525260952,
+      "grad_norm": 0.9097145199775696,
+      "learning_rate": 0.00021447660531122664,
+      "loss": 4.8412,
+      "step": 50900
+    },
+    {
+      "epoch": 0.0715333261907731,
+      "grad_norm": 0.9682596921920776,
+      "learning_rate": 0.00021451875790361107,
+      "loss": 4.7094,
+      "step": 50910
+    },
+    {
+      "epoch": 0.07154737712893668,
+      "grad_norm": 0.9312680959701538,
+      "learning_rate": 0.0002145609104959955,
+      "loss": 4.8971,
+      "step": 50920
+    },
+    {
+      "epoch": 0.07156142806710025,
+      "grad_norm": 0.8974460959434509,
+      "learning_rate": 0.0002146030630883799,
+      "loss": 4.7942,
+      "step": 50930
+    },
+    {
+      "epoch": 0.07157547900526383,
+      "grad_norm": 0.9163692593574524,
+      "learning_rate": 0.00021464521568076433,
+      "loss": 4.8219,
+      "step": 50940
+    },
+    {
+      "epoch": 0.0715895299434274,
+      "grad_norm": 0.9357579946517944,
+      "learning_rate": 0.00021468736827314877,
+      "loss": 4.865,
+      "step": 50950
+    },
+    {
+      "epoch": 0.071603580881591,
+      "grad_norm": 0.9122851490974426,
+      "learning_rate": 0.00021472952086553322,
+      "loss": 4.9114,
+      "step": 50960
+    },
+    {
+      "epoch": 0.07161763181975457,
+      "grad_norm": 0.8831797242164612,
+      "learning_rate": 0.00021477167345791765,
+      "loss": 4.8798,
+      "step": 50970
+    },
+    {
+      "epoch": 0.07163168275791815,
+      "grad_norm": 0.9927180409431458,
+      "learning_rate": 0.00021481382605030209,
+      "loss": 4.7515,
+      "step": 50980
+    },
+    {
+      "epoch": 0.07164573369608172,
+      "grad_norm": 0.9619544148445129,
+      "learning_rate": 0.00021485597864268652,
+      "loss": 4.7979,
+      "step": 50990
+    },
+    {
+      "epoch": 0.0716597846342453,
+      "grad_norm": 0.92621248960495,
+      "learning_rate": 0.00021489813123507092,
+      "loss": 4.8098,
+      "step": 51000
+    },
+    {
+      "epoch": 0.07167383557240888,
+      "grad_norm": 0.9142638444900513,
+      "learning_rate": 0.00021494028382745535,
+      "loss": 4.8371,
+      "step": 51010
+    },
+    {
+      "epoch": 0.07168788651057245,
+      "grad_norm": 0.9437655210494995,
+      "learning_rate": 0.0002149824364198398,
+      "loss": 4.8507,
+      "step": 51020
+    },
+    {
+      "epoch": 0.07170193744873603,
+      "grad_norm": 0.9370262026786804,
+      "learning_rate": 0.00021502458901222424,
+      "loss": 4.8096,
+      "step": 51030
+    },
+    {
+      "epoch": 0.0717159883868996,
+      "grad_norm": 0.8940436244010925,
+      "learning_rate": 0.00021506674160460867,
+      "loss": 4.7918,
+      "step": 51040
+    },
+    {
+      "epoch": 0.07173003932506318,
+      "grad_norm": 0.9073984026908875,
+      "learning_rate": 0.0002151088941969931,
+      "loss": 4.8127,
+      "step": 51050
+    },
+    {
+      "epoch": 0.07174409026322676,
+      "grad_norm": 0.9012202024459839,
+      "learning_rate": 0.0002151510467893775,
+      "loss": 4.7126,
+      "step": 51060
+    },
+    {
+      "epoch": 0.07175814120139035,
+      "grad_norm": 0.8825278282165527,
+      "learning_rate": 0.00021519319938176194,
+      "loss": 4.8573,
+      "step": 51070
+    },
+    {
+      "epoch": 0.07177219213955392,
+      "grad_norm": 0.9400650262832642,
+      "learning_rate": 0.0002152353519741464,
+      "loss": 4.8074,
+      "step": 51080
+    },
+    {
+      "epoch": 0.0717862430777175,
+      "grad_norm": 0.8937569260597229,
+      "learning_rate": 0.00021527750456653083,
+      "loss": 4.7746,
+      "step": 51090
+    },
+    {
+      "epoch": 0.07180029401588107,
+      "grad_norm": 0.9248559474945068,
+      "learning_rate": 0.00021531965715891526,
+      "loss": 4.889,
+      "step": 51100
+    },
+    {
+      "epoch": 0.07181434495404465,
+      "grad_norm": 0.8893438577651978,
+      "learning_rate": 0.0002153618097512997,
+      "loss": 4.7216,
+      "step": 51110
+    },
+    {
+      "epoch": 0.07182839589220823,
+      "grad_norm": 0.9111591577529907,
+      "learning_rate": 0.00021540396234368412,
+      "loss": 4.8645,
+      "step": 51120
+    },
+    {
+      "epoch": 0.0718424468303718,
+      "grad_norm": 1.1774877309799194,
+      "learning_rate": 0.00021544611493606853,
+      "loss": 4.8673,
+      "step": 51130
+    },
+    {
+      "epoch": 0.07185649776853538,
+      "grad_norm": 0.8930349946022034,
+      "learning_rate": 0.00021548826752845298,
+      "loss": 4.9113,
+      "step": 51140
+    },
+    {
+      "epoch": 0.07187054870669896,
+      "grad_norm": 0.9083362221717834,
+      "learning_rate": 0.00021553042012083742,
+      "loss": 4.9166,
+      "step": 51150
+    },
+    {
+      "epoch": 0.07188459964486253,
+      "grad_norm": 0.9521656632423401,
+      "learning_rate": 0.00021557257271322185,
+      "loss": 4.8467,
+      "step": 51160
+    },
+    {
+      "epoch": 0.07189865058302612,
+      "grad_norm": 0.8885960578918457,
+      "learning_rate": 0.00021561472530560628,
+      "loss": 4.9143,
+      "step": 51170
+    },
+    {
+      "epoch": 0.0719127015211897,
+      "grad_norm": 0.9528680443763733,
+      "learning_rate": 0.0002156568778979907,
+      "loss": 4.936,
+      "step": 51180
+    },
+    {
+      "epoch": 0.07192675245935327,
+      "grad_norm": 0.9279723763465881,
+      "learning_rate": 0.0002156990304903751,
+      "loss": 4.9369,
+      "step": 51190
+    },
+    {
+      "epoch": 0.07194080339751685,
+      "grad_norm": 0.8902644515037537,
+      "learning_rate": 0.00021574118308275957,
+      "loss": 4.9081,
+      "step": 51200
+    },
+    {
+      "epoch": 0.07195485433568043,
+      "grad_norm": 0.8919149041175842,
+      "learning_rate": 0.000215783335675144,
+      "loss": 4.7189,
+      "step": 51210
+    },
+    {
+      "epoch": 0.071968905273844,
+      "grad_norm": 0.8641528487205505,
+      "learning_rate": 0.00021582548826752843,
+      "loss": 4.9004,
+      "step": 51220
+    },
+    {
+      "epoch": 0.07198295621200758,
+      "grad_norm": 0.9722802639007568,
+      "learning_rate": 0.00021586764085991287,
+      "loss": 4.8593,
+      "step": 51230
+    },
+    {
+      "epoch": 0.07199700715017116,
+      "grad_norm": 0.9114046692848206,
+      "learning_rate": 0.0002159097934522973,
+      "loss": 4.8033,
+      "step": 51240
+    },
+    {
+      "epoch": 0.07201105808833473,
+      "grad_norm": 0.9105363488197327,
+      "learning_rate": 0.00021595194604468175,
+      "loss": 4.929,
+      "step": 51250
+    },
+    {
+      "epoch": 0.07202510902649831,
+      "grad_norm": 0.9299967288970947,
+      "learning_rate": 0.00021599409863706616,
+      "loss": 4.8742,
+      "step": 51260
+    },
+    {
+      "epoch": 0.0720391599646619,
+      "grad_norm": 0.905307948589325,
+      "learning_rate": 0.0002160362512294506,
+      "loss": 4.9712,
+      "step": 51270
+    },
+    {
+      "epoch": 0.07205321090282547,
+      "grad_norm": 0.8946942090988159,
+      "learning_rate": 0.00021607840382183502,
+      "loss": 4.8168,
+      "step": 51280
+    },
+    {
+      "epoch": 0.07206726184098905,
+      "grad_norm": 0.8690630197525024,
+      "learning_rate": 0.00021612055641421945,
+      "loss": 4.9673,
+      "step": 51290
+    },
+    {
+      "epoch": 0.07208131277915263,
+      "grad_norm": 0.9104512333869934,
+      "learning_rate": 0.00021616270900660388,
+      "loss": 4.8752,
+      "step": 51300
+    },
+    {
+      "epoch": 0.0720953637173162,
+      "grad_norm": 0.9049569368362427,
+      "learning_rate": 0.00021620486159898834,
+      "loss": 4.8247,
+      "step": 51310
+    },
+    {
+      "epoch": 0.07210941465547978,
+      "grad_norm": 0.9620559215545654,
+      "learning_rate": 0.00021624701419137275,
+      "loss": 4.8299,
+      "step": 51320
+    },
+    {
+      "epoch": 0.07212346559364335,
+      "grad_norm": 0.925479531288147,
+      "learning_rate": 0.00021628916678375718,
+      "loss": 4.9163,
+      "step": 51330
+    },
+    {
+      "epoch": 0.07213751653180693,
+      "grad_norm": 0.9238383769989014,
+      "learning_rate": 0.0002163313193761416,
+      "loss": 4.7718,
+      "step": 51340
+    },
+    {
+      "epoch": 0.0721515674699705,
+      "grad_norm": 0.9379529356956482,
+      "learning_rate": 0.00021637347196852604,
+      "loss": 4.8578,
+      "step": 51350
+    },
+    {
+      "epoch": 0.07216561840813408,
+      "grad_norm": 0.9757288694381714,
+      "learning_rate": 0.00021641562456091047,
+      "loss": 4.7786,
+      "step": 51360
+    },
+    {
+      "epoch": 0.07217966934629766,
+      "grad_norm": 0.9008384943008423,
+      "learning_rate": 0.00021645777715329493,
+      "loss": 4.7713,
+      "step": 51370
+    },
+    {
+      "epoch": 0.07219372028446125,
+      "grad_norm": 0.9475241899490356,
+      "learning_rate": 0.00021649992974567936,
+      "loss": 4.7283,
+      "step": 51380
+    },
+    {
+      "epoch": 0.07220777122262483,
+      "grad_norm": 0.8943010568618774,
+      "learning_rate": 0.00021654208233806376,
+      "loss": 4.9106,
+      "step": 51390
+    },
+    {
+      "epoch": 0.0722218221607884,
+      "grad_norm": 0.9169595241546631,
+      "learning_rate": 0.0002165842349304482,
+      "loss": 4.8022,
+      "step": 51400
+    },
+    {
+      "epoch": 0.07223587309895198,
+      "grad_norm": 0.9076150059700012,
+      "learning_rate": 0.00021662638752283263,
+      "loss": 4.8884,
+      "step": 51410
+    },
+    {
+      "epoch": 0.07224992403711555,
+      "grad_norm": 0.8879525661468506,
+      "learning_rate": 0.00021666854011521708,
+      "loss": 4.8106,
+      "step": 51420
+    },
+    {
+      "epoch": 0.07226397497527913,
+      "grad_norm": 0.9336953163146973,
+      "learning_rate": 0.00021671069270760152,
+      "loss": 4.8502,
+      "step": 51430
+    },
+    {
+      "epoch": 0.0722780259134427,
+      "grad_norm": 0.9571032524108887,
+      "learning_rate": 0.00021675284529998595,
+      "loss": 4.8924,
+      "step": 51440
+    },
+    {
+      "epoch": 0.07229207685160628,
+      "grad_norm": 0.8778558969497681,
+      "learning_rate": 0.00021679499789237035,
+      "loss": 4.7895,
+      "step": 51450
+    },
+    {
+      "epoch": 0.07230612778976986,
+      "grad_norm": 0.9267377853393555,
+      "learning_rate": 0.00021683715048475478,
+      "loss": 4.8497,
+      "step": 51460
+    },
+    {
+      "epoch": 0.07232017872793343,
+      "grad_norm": 0.8599752187728882,
+      "learning_rate": 0.0002168793030771392,
+      "loss": 4.7963,
+      "step": 51470
+    },
+    {
+      "epoch": 0.07233422966609702,
+      "grad_norm": 0.9141520261764526,
+      "learning_rate": 0.00021692145566952367,
+      "loss": 4.889,
+      "step": 51480
+    },
+    {
+      "epoch": 0.0723482806042606,
+      "grad_norm": 0.9507299065589905,
+      "learning_rate": 0.0002169636082619081,
+      "loss": 4.9028,
+      "step": 51490
+    },
+    {
+      "epoch": 0.07236233154242418,
+      "grad_norm": 0.9014940857887268,
+      "learning_rate": 0.00021700576085429253,
+      "loss": 4.8435,
+      "step": 51500
+    },
+    {
+      "epoch": 0.07237638248058775,
+      "grad_norm": 0.9214157462120056,
+      "learning_rate": 0.00021704791344667694,
+      "loss": 4.8943,
+      "step": 51510
+    },
+    {
+      "epoch": 0.07239043341875133,
+      "grad_norm": 0.8904531598091125,
+      "learning_rate": 0.00021709006603906137,
+      "loss": 4.788,
+      "step": 51520
+    },
+    {
+      "epoch": 0.0724044843569149,
+      "grad_norm": 0.8769239187240601,
+      "learning_rate": 0.0002171322186314458,
+      "loss": 4.8987,
+      "step": 51530
+    },
+    {
+      "epoch": 0.07241853529507848,
+      "grad_norm": 0.8964927196502686,
+      "learning_rate": 0.00021717437122383026,
+      "loss": 4.7834,
+      "step": 51540
+    },
+    {
+      "epoch": 0.07243258623324206,
+      "grad_norm": 0.9342873096466064,
+      "learning_rate": 0.0002172165238162147,
+      "loss": 4.8952,
+      "step": 51550
+    },
+    {
+      "epoch": 0.07244663717140563,
+      "grad_norm": 0.9120386838912964,
+      "learning_rate": 0.00021725867640859912,
+      "loss": 4.927,
+      "step": 51560
+    },
+    {
+      "epoch": 0.07246068810956921,
+      "grad_norm": 0.8827990889549255,
+      "learning_rate": 0.00021730082900098355,
+      "loss": 4.9028,
+      "step": 51570
+    },
+    {
+      "epoch": 0.0724747390477328,
+      "grad_norm": 0.8882848620414734,
+      "learning_rate": 0.00021734298159336796,
+      "loss": 4.8229,
+      "step": 51580
+    },
+    {
+      "epoch": 0.07248878998589638,
+      "grad_norm": 0.9415622353553772,
+      "learning_rate": 0.0002173851341857524,
+      "loss": 4.851,
+      "step": 51590
+    },
+    {
+      "epoch": 0.07250284092405995,
+      "grad_norm": 0.936032772064209,
+      "learning_rate": 0.00021742728677813685,
+      "loss": 4.7324,
+      "step": 51600
+    },
+    {
+      "epoch": 0.07251689186222353,
+      "grad_norm": 0.9080641865730286,
+      "learning_rate": 0.00021746943937052128,
+      "loss": 4.8478,
+      "step": 51610
+    },
+    {
+      "epoch": 0.0725309428003871,
+      "grad_norm": 0.8853969573974609,
+      "learning_rate": 0.0002175115919629057,
+      "loss": 4.8263,
+      "step": 51620
+    },
+    {
+      "epoch": 0.07254499373855068,
+      "grad_norm": 0.9154071807861328,
+      "learning_rate": 0.00021755374455529014,
+      "loss": 4.9162,
+      "step": 51630
+    },
+    {
+      "epoch": 0.07255904467671426,
+      "grad_norm": 0.894340991973877,
+      "learning_rate": 0.00021759589714767454,
+      "loss": 4.8624,
+      "step": 51640
+    },
+    {
+      "epoch": 0.07257309561487783,
+      "grad_norm": 0.9048279523849487,
+      "learning_rate": 0.00021763804974005897,
+      "loss": 4.9428,
+      "step": 51650
+    },
+    {
+      "epoch": 0.07258714655304141,
+      "grad_norm": 0.8889136910438538,
+      "learning_rate": 0.00021768020233244343,
+      "loss": 4.8428,
+      "step": 51660
+    },
+    {
+      "epoch": 0.07260119749120499,
+      "grad_norm": 1.0818347930908203,
+      "learning_rate": 0.00021772235492482786,
+      "loss": 4.9329,
+      "step": 51670
+    },
+    {
+      "epoch": 0.07261524842936856,
+      "grad_norm": 0.9882572889328003,
+      "learning_rate": 0.0002177645075172123,
+      "loss": 4.8468,
+      "step": 51680
+    },
+    {
+      "epoch": 0.07262929936753215,
+      "grad_norm": 0.8729943633079529,
+      "learning_rate": 0.00021780666010959673,
+      "loss": 5.0247,
+      "step": 51690
+    },
+    {
+      "epoch": 0.07264335030569573,
+      "grad_norm": 0.8866937160491943,
+      "learning_rate": 0.00021784881270198116,
+      "loss": 4.7614,
+      "step": 51700
+    },
+    {
+      "epoch": 0.0726574012438593,
+      "grad_norm": 0.9050795435905457,
+      "learning_rate": 0.00021789096529436556,
+      "loss": 4.9629,
+      "step": 51710
+    },
+    {
+      "epoch": 0.07267145218202288,
+      "grad_norm": 0.9173886775970459,
+      "learning_rate": 0.00021793311788675002,
+      "loss": 4.7605,
+      "step": 51720
+    },
+    {
+      "epoch": 0.07268550312018646,
+      "grad_norm": 0.8916087746620178,
+      "learning_rate": 0.00021797527047913445,
+      "loss": 4.8462,
+      "step": 51730
+    },
+    {
+      "epoch": 0.07269955405835003,
+      "grad_norm": 0.8887311220169067,
+      "learning_rate": 0.00021801742307151888,
+      "loss": 4.9066,
+      "step": 51740
+    },
+    {
+      "epoch": 0.07271360499651361,
+      "grad_norm": 0.871232271194458,
+      "learning_rate": 0.0002180595756639033,
+      "loss": 4.9546,
+      "step": 51750
+    },
+    {
+      "epoch": 0.07272765593467719,
+      "grad_norm": 0.8541275858879089,
+      "learning_rate": 0.00021810172825628774,
+      "loss": 4.8691,
+      "step": 51760
+    },
+    {
+      "epoch": 0.07274170687284076,
+      "grad_norm": 0.8941138982772827,
+      "learning_rate": 0.00021814388084867215,
+      "loss": 4.8881,
+      "step": 51770
+    },
+    {
+      "epoch": 0.07275575781100434,
+      "grad_norm": 0.9860665202140808,
+      "learning_rate": 0.0002181860334410566,
+      "loss": 4.7749,
+      "step": 51780
+    },
+    {
+      "epoch": 0.07276980874916793,
+      "grad_norm": 0.8882085680961609,
+      "learning_rate": 0.00021822818603344104,
+      "loss": 4.8333,
+      "step": 51790
+    },
+    {
+      "epoch": 0.0727838596873315,
+      "grad_norm": 0.9299302697181702,
+      "learning_rate": 0.00021827033862582547,
+      "loss": 4.8661,
+      "step": 51800
+    },
+    {
+      "epoch": 0.07279791062549508,
+      "grad_norm": 0.8716930747032166,
+      "learning_rate": 0.0002183124912182099,
+      "loss": 4.8118,
+      "step": 51810
+    },
+    {
+      "epoch": 0.07281196156365866,
+      "grad_norm": 0.9178931713104248,
+      "learning_rate": 0.00021835464381059433,
+      "loss": 4.867,
+      "step": 51820
+    },
+    {
+      "epoch": 0.07282601250182223,
+      "grad_norm": 0.9466065764427185,
+      "learning_rate": 0.0002183967964029788,
+      "loss": 4.8518,
+      "step": 51830
+    },
+    {
+      "epoch": 0.07284006343998581,
+      "grad_norm": 0.9389024972915649,
+      "learning_rate": 0.0002184389489953632,
+      "loss": 4.7342,
+      "step": 51840
+    },
+    {
+      "epoch": 0.07285411437814938,
+      "grad_norm": 0.9677282571792603,
+      "learning_rate": 0.00021848110158774762,
+      "loss": 4.8186,
+      "step": 51850
+    },
+    {
+      "epoch": 0.07286816531631296,
+      "grad_norm": 0.9958227872848511,
+      "learning_rate": 0.00021852325418013206,
+      "loss": 4.9577,
+      "step": 51860
+    },
+    {
+      "epoch": 0.07288221625447654,
+      "grad_norm": 0.8936029076576233,
+      "learning_rate": 0.0002185654067725165,
+      "loss": 4.9129,
+      "step": 51870
+    },
+    {
+      "epoch": 0.07289626719264011,
+      "grad_norm": 0.9088529944419861,
+      "learning_rate": 0.00021860755936490092,
+      "loss": 4.8115,
+      "step": 51880
+    },
+    {
+      "epoch": 0.0729103181308037,
+      "grad_norm": 0.9063525795936584,
+      "learning_rate": 0.00021864971195728538,
+      "loss": 4.9166,
+      "step": 51890
+    },
+    {
+      "epoch": 0.07292436906896728,
+      "grad_norm": 0.9061357378959656,
+      "learning_rate": 0.00021869186454966978,
+      "loss": 4.8306,
+      "step": 51900
+    },
+    {
+      "epoch": 0.07293842000713086,
+      "grad_norm": 0.9639674425125122,
+      "learning_rate": 0.0002187340171420542,
+      "loss": 4.8523,
+      "step": 51910
+    },
+    {
+      "epoch": 0.07295247094529443,
+      "grad_norm": 0.9000791311264038,
+      "learning_rate": 0.00021877616973443864,
+      "loss": 4.7562,
+      "step": 51920
+    },
+    {
+      "epoch": 0.07296652188345801,
+      "grad_norm": 0.9156561493873596,
+      "learning_rate": 0.00021881832232682307,
+      "loss": 5.0079,
+      "step": 51930
+    },
+    {
+      "epoch": 0.07298057282162158,
+      "grad_norm": 0.9364044666290283,
+      "learning_rate": 0.0002188604749192075,
+      "loss": 4.9567,
+      "step": 51940
+    },
+    {
+      "epoch": 0.07299462375978516,
+      "grad_norm": 0.8726660013198853,
+      "learning_rate": 0.00021890262751159196,
+      "loss": 4.9127,
+      "step": 51950
+    },
+    {
+      "epoch": 0.07300867469794874,
+      "grad_norm": 0.9417910575866699,
+      "learning_rate": 0.0002189447801039764,
+      "loss": 4.8765,
+      "step": 51960
+    },
+    {
+      "epoch": 0.07302272563611231,
+      "grad_norm": 0.8728827238082886,
+      "learning_rate": 0.0002189869326963608,
+      "loss": 4.7915,
+      "step": 51970
+    },
+    {
+      "epoch": 0.07303677657427589,
+      "grad_norm": 0.9182146787643433,
+      "learning_rate": 0.00021902908528874523,
+      "loss": 4.9794,
+      "step": 51980
+    },
+    {
+      "epoch": 0.07305082751243946,
+      "grad_norm": 0.8773520588874817,
+      "learning_rate": 0.00021907123788112966,
+      "loss": 4.863,
+      "step": 51990
+    },
+    {
+      "epoch": 0.07306487845060305,
+      "grad_norm": 0.9333748817443848,
+      "learning_rate": 0.0002191133904735141,
+      "loss": 4.7767,
+      "step": 52000
+    },
+    {
+      "epoch": 0.07307892938876663,
+      "grad_norm": 0.9405140280723572,
+      "learning_rate": 0.00021915554306589855,
+      "loss": 4.8579,
+      "step": 52010
+    },
+    {
+      "epoch": 0.07309298032693021,
+      "grad_norm": 0.8939602971076965,
+      "learning_rate": 0.00021919769565828298,
+      "loss": 4.9197,
+      "step": 52020
+    },
+    {
+      "epoch": 0.07310703126509378,
+      "grad_norm": 1.0516551733016968,
+      "learning_rate": 0.00021923984825066739,
+      "loss": 4.7807,
+      "step": 52030
+    },
+    {
+      "epoch": 0.07312108220325736,
+      "grad_norm": 0.9042669534683228,
+      "learning_rate": 0.00021928200084305182,
+      "loss": 4.8805,
+      "step": 52040
+    },
+    {
+      "epoch": 0.07313513314142094,
+      "grad_norm": 0.8705670833587646,
+      "learning_rate": 0.00021932415343543625,
+      "loss": 4.8592,
+      "step": 52050
+    },
+    {
+      "epoch": 0.07314918407958451,
+      "grad_norm": 0.9007607698440552,
+      "learning_rate": 0.0002193663060278207,
+      "loss": 4.9571,
+      "step": 52060
+    },
+    {
+      "epoch": 0.07316323501774809,
+      "grad_norm": 0.9101800322532654,
+      "learning_rate": 0.00021940845862020514,
+      "loss": 4.8341,
+      "step": 52070
+    },
+    {
+      "epoch": 0.07317728595591166,
+      "grad_norm": 0.8472366333007812,
+      "learning_rate": 0.00021945061121258957,
+      "loss": 5.002,
+      "step": 52080
+    },
+    {
+      "epoch": 0.07319133689407524,
+      "grad_norm": 0.8725908994674683,
+      "learning_rate": 0.00021949276380497397,
+      "loss": 4.8479,
+      "step": 52090
+    },
+    {
+      "epoch": 0.07320538783223883,
+      "grad_norm": 0.8859214782714844,
+      "learning_rate": 0.0002195349163973584,
+      "loss": 4.8376,
+      "step": 52100
+    },
+    {
+      "epoch": 0.0732194387704024,
+      "grad_norm": 0.9071083068847656,
+      "learning_rate": 0.00021957706898974283,
+      "loss": 4.7523,
+      "step": 52110
+    },
+    {
+      "epoch": 0.07323348970856598,
+      "grad_norm": 0.8878831267356873,
+      "learning_rate": 0.0002196192215821273,
+      "loss": 4.937,
+      "step": 52120
+    },
+    {
+      "epoch": 0.07324754064672956,
+      "grad_norm": 0.924068033695221,
+      "learning_rate": 0.00021966137417451172,
+      "loss": 4.6908,
+      "step": 52130
+    },
+    {
+      "epoch": 0.07326159158489313,
+      "grad_norm": 0.9516233205795288,
+      "learning_rate": 0.00021970352676689616,
+      "loss": 4.8146,
+      "step": 52140
+    },
+    {
+      "epoch": 0.07327564252305671,
+      "grad_norm": 0.8874931931495667,
+      "learning_rate": 0.0002197456793592806,
+      "loss": 4.799,
+      "step": 52150
+    },
+    {
+      "epoch": 0.07328969346122029,
+      "grad_norm": 0.8723164200782776,
+      "learning_rate": 0.000219787831951665,
+      "loss": 4.6445,
+      "step": 52160
+    },
+    {
+      "epoch": 0.07330374439938386,
+      "grad_norm": 0.8742164373397827,
+      "learning_rate": 0.00021982998454404942,
+      "loss": 5.0286,
+      "step": 52170
+    },
+    {
+      "epoch": 0.07331779533754744,
+      "grad_norm": 1.0268549919128418,
+      "learning_rate": 0.00021987213713643388,
+      "loss": 4.7709,
+      "step": 52180
+    },
+    {
+      "epoch": 0.07333184627571102,
+      "grad_norm": 0.8598684668540955,
+      "learning_rate": 0.0002199142897288183,
+      "loss": 4.9128,
+      "step": 52190
+    },
+    {
+      "epoch": 0.0733458972138746,
+      "grad_norm": 0.9072014689445496,
+      "learning_rate": 0.00021995644232120274,
+      "loss": 4.8774,
+      "step": 52200
+    },
+    {
+      "epoch": 0.07335994815203818,
+      "grad_norm": 0.9579162001609802,
+      "learning_rate": 0.00021999859491358717,
+      "loss": 4.8299,
+      "step": 52210
+    },
+    {
+      "epoch": 0.07337399909020176,
+      "grad_norm": 0.9101182222366333,
+      "learning_rate": 0.00022004074750597158,
+      "loss": 4.7819,
+      "step": 52220
+    },
+    {
+      "epoch": 0.07338805002836533,
+      "grad_norm": 0.8982588052749634,
+      "learning_rate": 0.000220082900098356,
+      "loss": 4.8396,
+      "step": 52230
+    },
+    {
+      "epoch": 0.07340210096652891,
+      "grad_norm": 0.9071044921875,
+      "learning_rate": 0.00022012505269074047,
+      "loss": 4.8595,
+      "step": 52240
+    },
+    {
+      "epoch": 0.07341615190469249,
+      "grad_norm": 0.952265739440918,
+      "learning_rate": 0.0002201672052831249,
+      "loss": 4.8164,
+      "step": 52250
+    },
+    {
+      "epoch": 0.07343020284285606,
+      "grad_norm": 0.9162461161613464,
+      "learning_rate": 0.00022020935787550933,
+      "loss": 4.8448,
+      "step": 52260
+    },
+    {
+      "epoch": 0.07344425378101964,
+      "grad_norm": 0.8878207802772522,
+      "learning_rate": 0.00022025151046789376,
+      "loss": 4.8139,
+      "step": 52270
+    },
+    {
+      "epoch": 0.07345830471918322,
+      "grad_norm": 0.8891446590423584,
+      "learning_rate": 0.0002202936630602782,
+      "loss": 4.9729,
+      "step": 52280
+    },
+    {
+      "epoch": 0.07347235565734679,
+      "grad_norm": 0.9611853957176208,
+      "learning_rate": 0.0002203358156526626,
+      "loss": 4.9082,
+      "step": 52290
+    },
+    {
+      "epoch": 0.07348640659551037,
+      "grad_norm": 0.8403993248939514,
+      "learning_rate": 0.00022037796824504705,
+      "loss": 4.904,
+      "step": 52300
+    },
+    {
+      "epoch": 0.07350045753367396,
+      "grad_norm": 0.873360276222229,
+      "learning_rate": 0.00022042012083743149,
+      "loss": 4.9462,
+      "step": 52310
+    },
+    {
+      "epoch": 0.07351450847183753,
+      "grad_norm": 0.8688433766365051,
+      "learning_rate": 0.00022046227342981592,
+      "loss": 5.0245,
+      "step": 52320
+    },
+    {
+      "epoch": 0.07352855941000111,
+      "grad_norm": 0.899304986000061,
+      "learning_rate": 0.00022050442602220035,
+      "loss": 4.7631,
+      "step": 52330
+    },
+    {
+      "epoch": 0.07354261034816469,
+      "grad_norm": 0.8732203841209412,
+      "learning_rate": 0.00022054657861458478,
+      "loss": 4.9718,
+      "step": 52340
+    },
+    {
+      "epoch": 0.07355666128632826,
+      "grad_norm": 0.8838686943054199,
+      "learning_rate": 0.00022058873120696918,
+      "loss": 4.9258,
+      "step": 52350
+    },
+    {
+      "epoch": 0.07357071222449184,
+      "grad_norm": 0.9056248068809509,
+      "learning_rate": 0.00022063088379935364,
+      "loss": 4.789,
+      "step": 52360
+    },
+    {
+      "epoch": 0.07358476316265541,
+      "grad_norm": 0.8867168426513672,
+      "learning_rate": 0.00022067303639173807,
+      "loss": 4.8421,
+      "step": 52370
+    },
+    {
+      "epoch": 0.07359881410081899,
+      "grad_norm": 1.1361435651779175,
+      "learning_rate": 0.0002207151889841225,
+      "loss": 4.8528,
+      "step": 52380
+    },
+    {
+      "epoch": 0.07361286503898257,
+      "grad_norm": 0.897068977355957,
+      "learning_rate": 0.00022075734157650693,
+      "loss": 4.7161,
+      "step": 52390
+    },
+    {
+      "epoch": 0.07362691597714614,
+      "grad_norm": 0.9281494617462158,
+      "learning_rate": 0.00022079949416889137,
+      "loss": 4.7663,
+      "step": 52400
+    },
+    {
+      "epoch": 0.07364096691530973,
+      "grad_norm": 0.9386702179908752,
+      "learning_rate": 0.00022084164676127582,
+      "loss": 4.8699,
+      "step": 52410
+    },
+    {
+      "epoch": 0.07365501785347331,
+      "grad_norm": 0.8745991587638855,
+      "learning_rate": 0.00022088379935366023,
+      "loss": 4.7858,
+      "step": 52420
+    },
+    {
+      "epoch": 0.07366906879163689,
+      "grad_norm": 0.9027533531188965,
+      "learning_rate": 0.00022092595194604466,
+      "loss": 4.8626,
+      "step": 52430
+    },
+    {
+      "epoch": 0.07368311972980046,
+      "grad_norm": 0.8828858733177185,
+      "learning_rate": 0.0002209681045384291,
+      "loss": 4.6567,
+      "step": 52440
+    },
+    {
+      "epoch": 0.07369717066796404,
+      "grad_norm": 0.8733296990394592,
+      "learning_rate": 0.00022101025713081352,
+      "loss": 4.8987,
+      "step": 52450
+    },
+    {
+      "epoch": 0.07371122160612761,
+      "grad_norm": 0.8760149478912354,
+      "learning_rate": 0.00022105240972319795,
+      "loss": 4.8488,
+      "step": 52460
+    },
+    {
+      "epoch": 0.07372527254429119,
+      "grad_norm": 0.9006288647651672,
+      "learning_rate": 0.0002210945623155824,
+      "loss": 4.8752,
+      "step": 52470
+    },
+    {
+      "epoch": 0.07373932348245477,
+      "grad_norm": 0.8950282335281372,
+      "learning_rate": 0.00022113671490796682,
+      "loss": 4.9373,
+      "step": 52480
+    },
+    {
+      "epoch": 0.07375337442061834,
+      "grad_norm": 0.8998293876647949,
+      "learning_rate": 0.00022117886750035125,
+      "loss": 4.8634,
+      "step": 52490
+    },
+    {
+      "epoch": 0.07376742535878192,
+      "grad_norm": 0.8735572099685669,
+      "learning_rate": 0.00022122102009273568,
+      "loss": 4.869,
+      "step": 52500
+    },
+    {
+      "epoch": 0.07378147629694551,
+      "grad_norm": 0.9324967265129089,
+      "learning_rate": 0.0002212631726851201,
+      "loss": 4.7894,
+      "step": 52510
+    },
+    {
+      "epoch": 0.07379552723510908,
+      "grad_norm": 0.8685128092765808,
+      "learning_rate": 0.00022130532527750454,
+      "loss": 4.8566,
+      "step": 52520
+    },
+    {
+      "epoch": 0.07380957817327266,
+      "grad_norm": 0.916084885597229,
+      "learning_rate": 0.000221347477869889,
+      "loss": 4.6707,
+      "step": 52530
+    },
+    {
+      "epoch": 0.07382362911143624,
+      "grad_norm": 0.8901728987693787,
+      "learning_rate": 0.00022138963046227343,
+      "loss": 4.8058,
+      "step": 52540
+    },
+    {
+      "epoch": 0.07383768004959981,
+      "grad_norm": 0.871108889579773,
+      "learning_rate": 0.00022143178305465783,
+      "loss": 4.9019,
+      "step": 52550
+    },
+    {
+      "epoch": 0.07385173098776339,
+      "grad_norm": 0.8851538300514221,
+      "learning_rate": 0.00022147393564704226,
+      "loss": 4.8358,
+      "step": 52560
+    },
+    {
+      "epoch": 0.07386578192592697,
+      "grad_norm": 0.9221569895744324,
+      "learning_rate": 0.0002215160882394267,
+      "loss": 4.9551,
+      "step": 52570
+    },
+    {
+      "epoch": 0.07387983286409054,
+      "grad_norm": 0.9061564803123474,
+      "learning_rate": 0.00022155824083181113,
+      "loss": 4.6869,
+      "step": 52580
+    },
+    {
+      "epoch": 0.07389388380225412,
+      "grad_norm": 0.8911418914794922,
+      "learning_rate": 0.00022160039342419559,
+      "loss": 4.8726,
+      "step": 52590
+    },
+    {
+      "epoch": 0.0739079347404177,
+      "grad_norm": 0.9063213467597961,
+      "learning_rate": 0.00022164254601658002,
+      "loss": 4.7479,
+      "step": 52600
+    },
+    {
+      "epoch": 0.07392198567858127,
+      "grad_norm": 0.8305318355560303,
+      "learning_rate": 0.00022168469860896442,
+      "loss": 4.9001,
+      "step": 52610
+    },
+    {
+      "epoch": 0.07393603661674486,
+      "grad_norm": 0.871112585067749,
+      "learning_rate": 0.00022172685120134885,
+      "loss": 4.8549,
+      "step": 52620
+    },
+    {
+      "epoch": 0.07395008755490844,
+      "grad_norm": 0.8946985006332397,
+      "learning_rate": 0.00022176900379373328,
+      "loss": 4.8635,
+      "step": 52630
+    },
+    {
+      "epoch": 0.07396413849307201,
+      "grad_norm": 0.9011940360069275,
+      "learning_rate": 0.00022181115638611774,
+      "loss": 4.8687,
+      "step": 52640
+    },
+    {
+      "epoch": 0.07397818943123559,
+      "grad_norm": 0.8744351267814636,
+      "learning_rate": 0.00022185330897850217,
+      "loss": 4.9532,
+      "step": 52650
+    },
+    {
+      "epoch": 0.07399224036939916,
+      "grad_norm": 0.886094331741333,
+      "learning_rate": 0.0002218954615708866,
+      "loss": 4.9025,
+      "step": 52660
+    },
+    {
+      "epoch": 0.07400629130756274,
+      "grad_norm": 0.8723426461219788,
+      "learning_rate": 0.00022193761416327103,
+      "loss": 4.8615,
+      "step": 52670
+    },
+    {
+      "epoch": 0.07402034224572632,
+      "grad_norm": 0.8950220942497253,
+      "learning_rate": 0.00022197976675565544,
+      "loss": 4.825,
+      "step": 52680
+    },
+    {
+      "epoch": 0.0740343931838899,
+      "grad_norm": 0.8471283316612244,
+      "learning_rate": 0.00022202191934803987,
+      "loss": 4.7089,
+      "step": 52690
+    },
+    {
+      "epoch": 0.07404844412205347,
+      "grad_norm": 0.8990341424942017,
+      "learning_rate": 0.00022206407194042433,
+      "loss": 4.9334,
+      "step": 52700
+    },
+    {
+      "epoch": 0.07406249506021705,
+      "grad_norm": 0.9074385762214661,
+      "learning_rate": 0.00022210622453280876,
+      "loss": 4.8409,
+      "step": 52710
+    },
+    {
+      "epoch": 0.07407654599838064,
+      "grad_norm": 0.8630465865135193,
+      "learning_rate": 0.0002221483771251932,
+      "loss": 4.8109,
+      "step": 52720
+    },
+    {
+      "epoch": 0.07409059693654421,
+      "grad_norm": 0.9616298675537109,
+      "learning_rate": 0.00022219052971757762,
+      "loss": 4.8484,
+      "step": 52730
+    },
+    {
+      "epoch": 0.07410464787470779,
+      "grad_norm": 0.9482759237289429,
+      "learning_rate": 0.00022223268230996203,
+      "loss": 4.886,
+      "step": 52740
+    },
+    {
+      "epoch": 0.07411869881287136,
+      "grad_norm": 0.8867536187171936,
+      "learning_rate": 0.00022227483490234646,
+      "loss": 4.8773,
+      "step": 52750
+    },
+    {
+      "epoch": 0.07413274975103494,
+      "grad_norm": 0.9389952421188354,
+      "learning_rate": 0.00022231698749473091,
+      "loss": 4.8021,
+      "step": 52760
+    },
+    {
+      "epoch": 0.07414680068919852,
+      "grad_norm": 0.8756046891212463,
+      "learning_rate": 0.00022235914008711535,
+      "loss": 4.708,
+      "step": 52770
+    },
+    {
+      "epoch": 0.07416085162736209,
+      "grad_norm": 0.8724471926689148,
+      "learning_rate": 0.00022240129267949978,
+      "loss": 4.794,
+      "step": 52780
+    },
+    {
+      "epoch": 0.07417490256552567,
+      "grad_norm": 0.895776629447937,
+      "learning_rate": 0.0002224434452718842,
+      "loss": 4.8679,
+      "step": 52790
+    },
+    {
+      "epoch": 0.07418895350368925,
+      "grad_norm": 0.9033660292625427,
+      "learning_rate": 0.0002224855978642686,
+      "loss": 4.6832,
+      "step": 52800
+    },
+    {
+      "epoch": 0.07420300444185282,
+      "grad_norm": 0.9038524627685547,
+      "learning_rate": 0.00022252775045665304,
+      "loss": 4.7711,
+      "step": 52810
+    },
+    {
+      "epoch": 0.07421705538001641,
+      "grad_norm": 0.8903810381889343,
+      "learning_rate": 0.0002225699030490375,
+      "loss": 4.8756,
+      "step": 52820
+    },
+    {
+      "epoch": 0.07423110631817999,
+      "grad_norm": 0.9286917448043823,
+      "learning_rate": 0.00022261205564142193,
+      "loss": 4.8495,
+      "step": 52830
+    },
+    {
+      "epoch": 0.07424515725634356,
+      "grad_norm": 0.8560378551483154,
+      "learning_rate": 0.00022265420823380636,
+      "loss": 4.8827,
+      "step": 52840
+    },
+    {
+      "epoch": 0.07425920819450714,
+      "grad_norm": 1.5267508029937744,
+      "learning_rate": 0.0002226963608261908,
+      "loss": 4.7531,
+      "step": 52850
+    },
+    {
+      "epoch": 0.07427325913267072,
+      "grad_norm": 0.8825668096542358,
+      "learning_rate": 0.00022273851341857523,
+      "loss": 4.9219,
+      "step": 52860
+    },
+    {
+      "epoch": 0.07428731007083429,
+      "grad_norm": 0.8723922967910767,
+      "learning_rate": 0.00022278066601095963,
+      "loss": 4.7371,
+      "step": 52870
+    },
+    {
+      "epoch": 0.07430136100899787,
+      "grad_norm": 0.8615224361419678,
+      "learning_rate": 0.0002228228186033441,
+      "loss": 4.7961,
+      "step": 52880
+    },
+    {
+      "epoch": 0.07431541194716144,
+      "grad_norm": 0.8599326610565186,
+      "learning_rate": 0.00022286497119572852,
+      "loss": 4.9175,
+      "step": 52890
+    },
+    {
+      "epoch": 0.07432946288532502,
+      "grad_norm": 0.9164808988571167,
+      "learning_rate": 0.00022290712378811295,
+      "loss": 4.7722,
+      "step": 52900
+    },
+    {
+      "epoch": 0.0743435138234886,
+      "grad_norm": 0.8620802760124207,
+      "learning_rate": 0.00022294927638049738,
+      "loss": 4.8847,
+      "step": 52910
+    },
+    {
+      "epoch": 0.07435756476165217,
+      "grad_norm": 0.88958340883255,
+      "learning_rate": 0.0002229914289728818,
+      "loss": 4.8875,
+      "step": 52920
+    },
+    {
+      "epoch": 0.07437161569981576,
+      "grad_norm": 0.894120454788208,
+      "learning_rate": 0.00022303358156526622,
+      "loss": 4.8375,
+      "step": 52930
+    },
+    {
+      "epoch": 0.07438566663797934,
+      "grad_norm": 0.8956342935562134,
+      "learning_rate": 0.00022307573415765068,
+      "loss": 4.9163,
+      "step": 52940
+    },
+    {
+      "epoch": 0.07439971757614292,
+      "grad_norm": 0.8793826699256897,
+      "learning_rate": 0.0002231178867500351,
+      "loss": 4.8802,
+      "step": 52950
+    },
+    {
+      "epoch": 0.07441376851430649,
+      "grad_norm": 0.8765672445297241,
+      "learning_rate": 0.00022316003934241954,
+      "loss": 4.8051,
+      "step": 52960
+    },
+    {
+      "epoch": 0.07442781945247007,
+      "grad_norm": 0.9064963459968567,
+      "learning_rate": 0.00022320219193480397,
+      "loss": 4.804,
+      "step": 52970
+    },
+    {
+      "epoch": 0.07444187039063364,
+      "grad_norm": 0.879390299320221,
+      "learning_rate": 0.0002232443445271884,
+      "loss": 5.0118,
+      "step": 52980
+    },
+    {
+      "epoch": 0.07445592132879722,
+      "grad_norm": 0.8609398603439331,
+      "learning_rate": 0.00022328649711957286,
+      "loss": 4.8256,
+      "step": 52990
+    },
+    {
+      "epoch": 0.0744699722669608,
+      "grad_norm": 0.9296293258666992,
+      "learning_rate": 0.00022332864971195726,
+      "loss": 4.8331,
+      "step": 53000
+    },
+    {
+      "epoch": 0.07448402320512437,
+      "grad_norm": 1.0869863033294678,
+      "learning_rate": 0.0002233708023043417,
+      "loss": 4.7803,
+      "step": 53010
+    },
+    {
+      "epoch": 0.07449807414328795,
+      "grad_norm": 0.9066222906112671,
+      "learning_rate": 0.00022341295489672613,
+      "loss": 4.868,
+      "step": 53020
+    },
+    {
+      "epoch": 0.07451212508145154,
+      "grad_norm": 0.889058530330658,
+      "learning_rate": 0.00022345510748911056,
+      "loss": 4.7838,
+      "step": 53030
+    },
+    {
+      "epoch": 0.07452617601961511,
+      "grad_norm": 1.2822096347808838,
+      "learning_rate": 0.000223497260081495,
+      "loss": 4.7383,
+      "step": 53040
+    },
+    {
+      "epoch": 0.07454022695777869,
+      "grad_norm": 0.8843401074409485,
+      "learning_rate": 0.00022353941267387945,
+      "loss": 4.8443,
+      "step": 53050
+    },
+    {
+      "epoch": 0.07455427789594227,
+      "grad_norm": 0.9470616579055786,
+      "learning_rate": 0.00022358156526626385,
+      "loss": 4.836,
+      "step": 53060
+    },
+    {
+      "epoch": 0.07456832883410584,
+      "grad_norm": 0.878521740436554,
+      "learning_rate": 0.00022362371785864828,
+      "loss": 4.9428,
+      "step": 53070
+    },
+    {
+      "epoch": 0.07458237977226942,
+      "grad_norm": 0.894644558429718,
+      "learning_rate": 0.0002236658704510327,
+      "loss": 4.911,
+      "step": 53080
+    },
+    {
+      "epoch": 0.074596430710433,
+      "grad_norm": 0.8709197044372559,
+      "learning_rate": 0.00022370802304341714,
+      "loss": 4.8799,
+      "step": 53090
+    },
+    {
+      "epoch": 0.07461048164859657,
+      "grad_norm": 0.8657320141792297,
+      "learning_rate": 0.00022375017563580157,
+      "loss": 4.9005,
+      "step": 53100
+    },
+    {
+      "epoch": 0.07462453258676015,
+      "grad_norm": 0.9332450032234192,
+      "learning_rate": 0.00022379232822818603,
+      "loss": 4.994,
+      "step": 53110
+    },
+    {
+      "epoch": 0.07463858352492372,
+      "grad_norm": 0.9028719067573547,
+      "learning_rate": 0.00022383448082057046,
+      "loss": 4.9586,
+      "step": 53120
+    },
+    {
+      "epoch": 0.07465263446308731,
+      "grad_norm": 0.9070411920547485,
+      "learning_rate": 0.00022387663341295487,
+      "loss": 5.001,
+      "step": 53130
+    },
+    {
+      "epoch": 0.07466668540125089,
+      "grad_norm": 0.8787438869476318,
+      "learning_rate": 0.0002239187860053393,
+      "loss": 4.9275,
+      "step": 53140
+    },
+    {
+      "epoch": 0.07468073633941447,
+      "grad_norm": 0.901603102684021,
+      "learning_rate": 0.00022396093859772373,
+      "loss": 4.9226,
+      "step": 53150
+    },
+    {
+      "epoch": 0.07469478727757804,
+      "grad_norm": 0.9032551646232605,
+      "learning_rate": 0.00022400309119010816,
+      "loss": 4.7562,
+      "step": 53160
+    },
+    {
+      "epoch": 0.07470883821574162,
+      "grad_norm": 0.9277701377868652,
+      "learning_rate": 0.00022404524378249262,
+      "loss": 4.9293,
+      "step": 53170
+    },
+    {
+      "epoch": 0.0747228891539052,
+      "grad_norm": 0.9006163477897644,
+      "learning_rate": 0.00022408739637487705,
+      "loss": 4.7839,
+      "step": 53180
+    },
+    {
+      "epoch": 0.07473694009206877,
+      "grad_norm": 0.9267420768737793,
+      "learning_rate": 0.00022412954896726146,
+      "loss": 4.7228,
+      "step": 53190
+    },
+    {
+      "epoch": 0.07475099103023235,
+      "grad_norm": 0.8903605937957764,
+      "learning_rate": 0.00022417170155964589,
+      "loss": 4.7873,
+      "step": 53200
+    },
+    {
+      "epoch": 0.07476504196839592,
+      "grad_norm": 0.8836460113525391,
+      "learning_rate": 0.00022421385415203032,
+      "loss": 4.9616,
+      "step": 53210
+    },
+    {
+      "epoch": 0.0747790929065595,
+      "grad_norm": 0.8987323045730591,
+      "learning_rate": 0.00022425600674441478,
+      "loss": 4.8712,
+      "step": 53220
+    },
+    {
+      "epoch": 0.07479314384472308,
+      "grad_norm": 0.8787208199501038,
+      "learning_rate": 0.0002242981593367992,
+      "loss": 4.8302,
+      "step": 53230
+    },
+    {
+      "epoch": 0.07480719478288667,
+      "grad_norm": 0.8872378468513489,
+      "learning_rate": 0.00022434031192918364,
+      "loss": 4.7994,
+      "step": 53240
+    },
+    {
+      "epoch": 0.07482124572105024,
+      "grad_norm": 0.8948333263397217,
+      "learning_rate": 0.00022438246452156807,
+      "loss": 4.9177,
+      "step": 53250
+    },
+    {
+      "epoch": 0.07483529665921382,
+      "grad_norm": 0.8879920840263367,
+      "learning_rate": 0.00022442461711395247,
+      "loss": 4.8703,
+      "step": 53260
+    },
+    {
+      "epoch": 0.0748493475973774,
+      "grad_norm": 0.897305965423584,
+      "learning_rate": 0.0002244667697063369,
+      "loss": 4.8631,
+      "step": 53270
+    },
+    {
+      "epoch": 0.07486339853554097,
+      "grad_norm": 0.8983936905860901,
+      "learning_rate": 0.00022450892229872136,
+      "loss": 4.7904,
+      "step": 53280
+    },
+    {
+      "epoch": 0.07487744947370455,
+      "grad_norm": 0.9298288822174072,
+      "learning_rate": 0.0002245510748911058,
+      "loss": 4.9048,
+      "step": 53290
+    },
+    {
+      "epoch": 0.07489150041186812,
+      "grad_norm": 0.9018476009368896,
+      "learning_rate": 0.00022459322748349022,
+      "loss": 4.8721,
+      "step": 53300
+    },
+    {
+      "epoch": 0.0749055513500317,
+      "grad_norm": 0.8535480499267578,
+      "learning_rate": 0.00022463538007587466,
+      "loss": 4.9276,
+      "step": 53310
+    },
+    {
+      "epoch": 0.07491960228819528,
+      "grad_norm": 0.8795814514160156,
+      "learning_rate": 0.00022467753266825906,
+      "loss": 4.8589,
+      "step": 53320
+    },
+    {
+      "epoch": 0.07493365322635885,
+      "grad_norm": 0.9002819061279297,
+      "learning_rate": 0.0002247196852606435,
+      "loss": 5.0146,
+      "step": 53330
+    },
+    {
+      "epoch": 0.07494770416452244,
+      "grad_norm": 0.9197555780410767,
+      "learning_rate": 0.00022476183785302795,
+      "loss": 4.9211,
+      "step": 53340
+    },
+    {
+      "epoch": 0.07496175510268602,
+      "grad_norm": 0.9005763530731201,
+      "learning_rate": 0.00022480399044541238,
+      "loss": 4.9638,
+      "step": 53350
+    },
+    {
+      "epoch": 0.0749758060408496,
+      "grad_norm": 0.8709673881530762,
+      "learning_rate": 0.0002248461430377968,
+      "loss": 4.8595,
+      "step": 53360
+    },
+    {
+      "epoch": 0.07498985697901317,
+      "grad_norm": 0.8983860611915588,
+      "learning_rate": 0.00022488829563018124,
+      "loss": 4.7567,
+      "step": 53370
+    },
+    {
+      "epoch": 0.07500390791717675,
+      "grad_norm": 0.8729245066642761,
+      "learning_rate": 0.00022493044822256567,
+      "loss": 4.8404,
+      "step": 53380
+    },
+    {
+      "epoch": 0.07501795885534032,
+      "grad_norm": 0.8825687170028687,
+      "learning_rate": 0.00022497260081495008,
+      "loss": 4.8511,
+      "step": 53390
+    },
+    {
+      "epoch": 0.0750320097935039,
+      "grad_norm": 1.1445651054382324,
+      "learning_rate": 0.00022501475340733454,
+      "loss": 4.8355,
+      "step": 53400
+    },
+    {
+      "epoch": 0.07504606073166747,
+      "grad_norm": 0.8848526477813721,
+      "learning_rate": 0.00022505690599971897,
+      "loss": 4.8923,
+      "step": 53410
+    },
+    {
+      "epoch": 0.07506011166983105,
+      "grad_norm": 0.9172847867012024,
+      "learning_rate": 0.0002250990585921034,
+      "loss": 4.9146,
+      "step": 53420
+    },
+    {
+      "epoch": 0.07507416260799463,
+      "grad_norm": 0.8731652498245239,
+      "learning_rate": 0.00022514121118448783,
+      "loss": 5.0147,
+      "step": 53430
+    },
+    {
+      "epoch": 0.07508821354615822,
+      "grad_norm": 0.8790530562400818,
+      "learning_rate": 0.00022518336377687226,
+      "loss": 4.8628,
+      "step": 53440
+    },
+    {
+      "epoch": 0.07510226448432179,
+      "grad_norm": 0.8736057281494141,
+      "learning_rate": 0.00022522551636925667,
+      "loss": 4.7773,
+      "step": 53450
+    },
+    {
+      "epoch": 0.07511631542248537,
+      "grad_norm": 0.9131947755813599,
+      "learning_rate": 0.00022526766896164112,
+      "loss": 4.7888,
+      "step": 53460
+    },
+    {
+      "epoch": 0.07513036636064895,
+      "grad_norm": 0.8734030723571777,
+      "learning_rate": 0.00022530982155402555,
+      "loss": 4.6335,
+      "step": 53470
+    },
+    {
+      "epoch": 0.07514441729881252,
+      "grad_norm": 0.8606353402137756,
+      "learning_rate": 0.00022535197414640999,
+      "loss": 4.7898,
+      "step": 53480
+    },
+    {
+      "epoch": 0.0751584682369761,
+      "grad_norm": 0.9181352853775024,
+      "learning_rate": 0.00022539412673879442,
+      "loss": 4.7451,
+      "step": 53490
+    },
+    {
+      "epoch": 0.07517251917513967,
+      "grad_norm": 0.8751580119132996,
+      "learning_rate": 0.00022543627933117885,
+      "loss": 4.9376,
+      "step": 53500
+    },
+    {
+      "epoch": 0.07518657011330325,
+      "grad_norm": 0.8582903742790222,
+      "learning_rate": 0.00022547843192356325,
+      "loss": 4.8904,
+      "step": 53510
+    },
+    {
+      "epoch": 0.07520062105146683,
+      "grad_norm": 0.9157120585441589,
+      "learning_rate": 0.0002255205845159477,
+      "loss": 4.8151,
+      "step": 53520
+    },
+    {
+      "epoch": 0.0752146719896304,
+      "grad_norm": 0.866913914680481,
+      "learning_rate": 0.00022556273710833214,
+      "loss": 4.8619,
+      "step": 53530
+    },
+    {
+      "epoch": 0.07522872292779398,
+      "grad_norm": 0.9174304008483887,
+      "learning_rate": 0.00022560488970071657,
+      "loss": 4.8327,
+      "step": 53540
+    },
+    {
+      "epoch": 0.07524277386595757,
+      "grad_norm": 0.8768737316131592,
+      "learning_rate": 0.000225647042293101,
+      "loss": 4.7639,
+      "step": 53550
+    },
+    {
+      "epoch": 0.07525682480412114,
+      "grad_norm": 0.8870129585266113,
+      "learning_rate": 0.00022568919488548544,
+      "loss": 4.7549,
+      "step": 53560
+    },
+    {
+      "epoch": 0.07527087574228472,
+      "grad_norm": 0.8497852087020874,
+      "learning_rate": 0.0002257313474778699,
+      "loss": 4.6357,
+      "step": 53570
+    },
+    {
+      "epoch": 0.0752849266804483,
+      "grad_norm": 0.8594181537628174,
+      "learning_rate": 0.0002257735000702543,
+      "loss": 4.9172,
+      "step": 53580
+    },
+    {
+      "epoch": 0.07529897761861187,
+      "grad_norm": 0.8665555715560913,
+      "learning_rate": 0.00022581565266263873,
+      "loss": 4.8535,
+      "step": 53590
+    },
+    {
+      "epoch": 0.07531302855677545,
+      "grad_norm": 0.8849197030067444,
+      "learning_rate": 0.00022585780525502316,
+      "loss": 4.8296,
+      "step": 53600
+    },
+    {
+      "epoch": 0.07532707949493903,
+      "grad_norm": 0.8826219439506531,
+      "learning_rate": 0.0002258999578474076,
+      "loss": 4.8087,
+      "step": 53610
+    },
+    {
+      "epoch": 0.0753411304331026,
+      "grad_norm": 0.9253659248352051,
+      "learning_rate": 0.00022594211043979202,
+      "loss": 4.8322,
+      "step": 53620
+    },
+    {
+      "epoch": 0.07535518137126618,
+      "grad_norm": 0.9128340482711792,
+      "learning_rate": 0.00022598426303217648,
+      "loss": 4.6608,
+      "step": 53630
+    },
+    {
+      "epoch": 0.07536923230942975,
+      "grad_norm": 0.8583227396011353,
+      "learning_rate": 0.00022602641562456088,
+      "loss": 4.7733,
+      "step": 53640
+    },
+    {
+      "epoch": 0.07538328324759334,
+      "grad_norm": 0.8400426506996155,
+      "learning_rate": 0.00022606856821694532,
+      "loss": 4.799,
+      "step": 53650
+    },
+    {
+      "epoch": 0.07539733418575692,
+      "grad_norm": 0.9156240224838257,
+      "learning_rate": 0.00022611072080932975,
+      "loss": 4.7519,
+      "step": 53660
+    },
+    {
+      "epoch": 0.0754113851239205,
+      "grad_norm": 0.902900218963623,
+      "learning_rate": 0.00022615287340171418,
+      "loss": 4.6597,
+      "step": 53670
+    },
+    {
+      "epoch": 0.07542543606208407,
+      "grad_norm": 0.8428563475608826,
+      "learning_rate": 0.0002261950259940986,
+      "loss": 4.8529,
+      "step": 53680
+    },
+    {
+      "epoch": 0.07543948700024765,
+      "grad_norm": 0.9369030594825745,
+      "learning_rate": 0.00022623717858648307,
+      "loss": 4.6786,
+      "step": 53690
+    },
+    {
+      "epoch": 0.07545353793841122,
+      "grad_norm": 0.8624921441078186,
+      "learning_rate": 0.0002262793311788675,
+      "loss": 4.8759,
+      "step": 53700
+    },
+    {
+      "epoch": 0.0754675888765748,
+      "grad_norm": 0.864963710308075,
+      "learning_rate": 0.0002263214837712519,
+      "loss": 4.8758,
+      "step": 53710
+    },
+    {
+      "epoch": 0.07548163981473838,
+      "grad_norm": 0.8992347121238708,
+      "learning_rate": 0.00022636363636363633,
+      "loss": 4.7599,
+      "step": 53720
+    },
+    {
+      "epoch": 0.07549569075290195,
+      "grad_norm": 0.9037030935287476,
+      "learning_rate": 0.00022640578895602077,
+      "loss": 4.8639,
+      "step": 53730
+    },
+    {
+      "epoch": 0.07550974169106553,
+      "grad_norm": 0.917029857635498,
+      "learning_rate": 0.0002264479415484052,
+      "loss": 4.7925,
+      "step": 53740
+    },
+    {
+      "epoch": 0.07552379262922912,
+      "grad_norm": 0.8903194665908813,
+      "learning_rate": 0.00022649009414078965,
+      "loss": 4.7699,
+      "step": 53750
+    },
+    {
+      "epoch": 0.0755378435673927,
+      "grad_norm": 0.8409320712089539,
+      "learning_rate": 0.00022653224673317409,
+      "loss": 4.7887,
+      "step": 53760
+    },
+    {
+      "epoch": 0.07555189450555627,
+      "grad_norm": 0.9254723787307739,
+      "learning_rate": 0.0002265743993255585,
+      "loss": 4.7081,
+      "step": 53770
+    },
+    {
+      "epoch": 0.07556594544371985,
+      "grad_norm": 0.9033946990966797,
+      "learning_rate": 0.00022661655191794292,
+      "loss": 4.8571,
+      "step": 53780
+    },
+    {
+      "epoch": 0.07557999638188342,
+      "grad_norm": 0.8677847385406494,
+      "learning_rate": 0.00022665870451032735,
+      "loss": 4.8679,
+      "step": 53790
+    },
+    {
+      "epoch": 0.075594047320047,
+      "grad_norm": 0.871347963809967,
+      "learning_rate": 0.0002267008571027118,
+      "loss": 4.7151,
+      "step": 53800
+    },
+    {
+      "epoch": 0.07560809825821058,
+      "grad_norm": 0.8525243997573853,
+      "learning_rate": 0.00022674300969509624,
+      "loss": 4.9235,
+      "step": 53810
+    },
+    {
+      "epoch": 0.07562214919637415,
+      "grad_norm": 0.850013256072998,
+      "learning_rate": 0.00022678516228748067,
+      "loss": 4.9236,
+      "step": 53820
+    },
+    {
+      "epoch": 0.07563620013453773,
+      "grad_norm": 0.8850587606430054,
+      "learning_rate": 0.0002268273148798651,
+      "loss": 4.8146,
+      "step": 53830
+    },
+    {
+      "epoch": 0.0756502510727013,
+      "grad_norm": 0.9525415301322937,
+      "learning_rate": 0.0002268694674722495,
+      "loss": 4.6886,
+      "step": 53840
+    },
+    {
+      "epoch": 0.07566430201086488,
+      "grad_norm": 0.8798964619636536,
+      "learning_rate": 0.00022691162006463394,
+      "loss": 4.7848,
+      "step": 53850
+    },
+    {
+      "epoch": 0.07567835294902847,
+      "grad_norm": 0.8539289236068726,
+      "learning_rate": 0.0002269537726570184,
+      "loss": 4.7355,
+      "step": 53860
+    },
+    {
+      "epoch": 0.07569240388719205,
+      "grad_norm": 0.8648083806037903,
+      "learning_rate": 0.00022699592524940283,
+      "loss": 4.7876,
+      "step": 53870
+    },
+    {
+      "epoch": 0.07570645482535562,
+      "grad_norm": 0.906799852848053,
+      "learning_rate": 0.00022703807784178726,
+      "loss": 4.8246,
+      "step": 53880
+    },
+    {
+      "epoch": 0.0757205057635192,
+      "grad_norm": 0.9276502132415771,
+      "learning_rate": 0.0002270802304341717,
+      "loss": 4.8731,
+      "step": 53890
+    },
+    {
+      "epoch": 0.07573455670168278,
+      "grad_norm": 0.8921993970870972,
+      "learning_rate": 0.0002271223830265561,
+      "loss": 4.8887,
+      "step": 53900
+    },
+    {
+      "epoch": 0.07574860763984635,
+      "grad_norm": 0.8780595064163208,
+      "learning_rate": 0.00022716453561894053,
+      "loss": 4.7841,
+      "step": 53910
+    },
+    {
+      "epoch": 0.07576265857800993,
+      "grad_norm": 0.9392471313476562,
+      "learning_rate": 0.00022720668821132498,
+      "loss": 4.8724,
+      "step": 53920
+    },
+    {
+      "epoch": 0.0757767095161735,
+      "grad_norm": 0.89900803565979,
+      "learning_rate": 0.00022724884080370942,
+      "loss": 4.7855,
+      "step": 53930
+    },
+    {
+      "epoch": 0.07579076045433708,
+      "grad_norm": 0.9714470505714417,
+      "learning_rate": 0.00022729099339609385,
+      "loss": 4.8981,
+      "step": 53940
+    },
+    {
+      "epoch": 0.07580481139250066,
+      "grad_norm": 0.8538759350776672,
+      "learning_rate": 0.00022733314598847828,
+      "loss": 4.9213,
+      "step": 53950
+    },
+    {
+      "epoch": 0.07581886233066425,
+      "grad_norm": 0.879152238368988,
+      "learning_rate": 0.0002273752985808627,
+      "loss": 4.8243,
+      "step": 53960
+    },
+    {
+      "epoch": 0.07583291326882782,
+      "grad_norm": 0.8458477258682251,
+      "learning_rate": 0.0002274174511732471,
+      "loss": 4.8997,
+      "step": 53970
+    },
+    {
+      "epoch": 0.0758469642069914,
+      "grad_norm": 0.9155076146125793,
+      "learning_rate": 0.00022745960376563157,
+      "loss": 4.8447,
+      "step": 53980
+    },
+    {
+      "epoch": 0.07586101514515498,
+      "grad_norm": 0.9187341332435608,
+      "learning_rate": 0.000227501756358016,
+      "loss": 4.9255,
+      "step": 53990
+    },
+    {
+      "epoch": 0.07587506608331855,
+      "grad_norm": 0.8751149773597717,
+      "learning_rate": 0.00022754390895040043,
+      "loss": 4.8512,
+      "step": 54000
+    },
+    {
+      "epoch": 0.07588911702148213,
+      "grad_norm": 0.8937033414840698,
+      "learning_rate": 0.00022758606154278486,
+      "loss": 4.8653,
+      "step": 54010
+    },
+    {
+      "epoch": 0.0759031679596457,
+      "grad_norm": 0.862657904624939,
+      "learning_rate": 0.0002276282141351693,
+      "loss": 4.748,
+      "step": 54020
+    },
+    {
+      "epoch": 0.07591721889780928,
+      "grad_norm": 0.8705819845199585,
+      "learning_rate": 0.0002276703667275537,
+      "loss": 4.8872,
+      "step": 54030
+    },
+    {
+      "epoch": 0.07593126983597286,
+      "grad_norm": 0.8772851824760437,
+      "learning_rate": 0.00022771251931993816,
+      "loss": 4.9373,
+      "step": 54040
+    },
+    {
+      "epoch": 0.07594532077413643,
+      "grad_norm": 0.8797933459281921,
+      "learning_rate": 0.0002277546719123226,
+      "loss": 4.8121,
+      "step": 54050
+    },
+    {
+      "epoch": 0.07595937171230002,
+      "grad_norm": 0.8654122948646545,
+      "learning_rate": 0.00022779682450470702,
+      "loss": 4.76,
+      "step": 54060
+    },
+    {
+      "epoch": 0.0759734226504636,
+      "grad_norm": 0.875860869884491,
+      "learning_rate": 0.00022783897709709145,
+      "loss": 4.7232,
+      "step": 54070
+    },
+    {
+      "epoch": 0.07598747358862717,
+      "grad_norm": 0.9174049496650696,
+      "learning_rate": 0.00022788112968947588,
+      "loss": 4.7754,
+      "step": 54080
+    },
+    {
+      "epoch": 0.07600152452679075,
+      "grad_norm": 0.8807012438774109,
+      "learning_rate": 0.00022792328228186034,
+      "loss": 4.8693,
+      "step": 54090
+    },
+    {
+      "epoch": 0.07601557546495433,
+      "grad_norm": 0.8439174294471741,
+      "learning_rate": 0.00022796543487424475,
+      "loss": 4.8432,
+      "step": 54100
+    },
+    {
+      "epoch": 0.0760296264031179,
+      "grad_norm": 0.9406492710113525,
+      "learning_rate": 0.00022800758746662918,
+      "loss": 4.8385,
+      "step": 54110
+    },
+    {
+      "epoch": 0.07604367734128148,
+      "grad_norm": 0.8603804707527161,
+      "learning_rate": 0.0002280497400590136,
+      "loss": 4.7904,
+      "step": 54120
+    },
+    {
+      "epoch": 0.07605772827944506,
+      "grad_norm": 0.8673272132873535,
+      "learning_rate": 0.0002280876773921596,
+      "loss": 4.8706,
+      "step": 54130
+    },
+    {
+      "epoch": 0.07607177921760863,
+      "grad_norm": 0.8606926202774048,
+      "learning_rate": 0.00022812982998454402,
+      "loss": 4.7967,
+      "step": 54140
+    },
+    {
+      "epoch": 0.07608583015577221,
+      "grad_norm": 0.8870376348495483,
+      "learning_rate": 0.00022817198257692845,
+      "loss": 4.8095,
+      "step": 54150
+    },
+    {
+      "epoch": 0.07609988109393578,
+      "grad_norm": 0.8442344665527344,
+      "learning_rate": 0.00022821413516931288,
+      "loss": 4.8743,
+      "step": 54160
+    },
+    {
+      "epoch": 0.07611393203209937,
+      "grad_norm": 0.8706654906272888,
+      "learning_rate": 0.00022825628776169734,
+      "loss": 4.839,
+      "step": 54170
+    },
+    {
+      "epoch": 0.07612798297026295,
+      "grad_norm": 0.8698878288269043,
+      "learning_rate": 0.00022829844035408177,
+      "loss": 4.8972,
+      "step": 54180
+    },
+    {
+      "epoch": 0.07614203390842653,
+      "grad_norm": 0.8605942726135254,
+      "learning_rate": 0.00022834059294646618,
+      "loss": 4.7414,
+      "step": 54190
+    },
+    {
+      "epoch": 0.0761560848465901,
+      "grad_norm": 0.8853992819786072,
+      "learning_rate": 0.0002283827455388506,
+      "loss": 4.7935,
+      "step": 54200
+    },
+    {
+      "epoch": 0.07617013578475368,
+      "grad_norm": 1.3746801614761353,
+      "learning_rate": 0.00022842489813123504,
+      "loss": 4.8538,
+      "step": 54210
+    },
+    {
+      "epoch": 0.07618418672291725,
+      "grad_norm": 0.8895174264907837,
+      "learning_rate": 0.00022846705072361947,
+      "loss": 4.8864,
+      "step": 54220
+    },
+    {
+      "epoch": 0.07619823766108083,
+      "grad_norm": 0.9199739098548889,
+      "learning_rate": 0.00022850920331600393,
+      "loss": 4.7924,
+      "step": 54230
+    },
+    {
+      "epoch": 0.07621228859924441,
+      "grad_norm": 0.9062716364860535,
+      "learning_rate": 0.00022855135590838836,
+      "loss": 4.8258,
+      "step": 54240
+    },
+    {
+      "epoch": 0.07622633953740798,
+      "grad_norm": 0.8298748731613159,
+      "learning_rate": 0.0002285935085007728,
+      "loss": 4.8469,
+      "step": 54250
+    },
+    {
+      "epoch": 0.07624039047557156,
+      "grad_norm": 0.866485059261322,
+      "learning_rate": 0.0002286356610931572,
+      "loss": 4.8599,
+      "step": 54260
+    },
+    {
+      "epoch": 0.07625444141373515,
+      "grad_norm": 0.8970810174942017,
+      "learning_rate": 0.00022867781368554163,
+      "loss": 4.8573,
+      "step": 54270
+    },
+    {
+      "epoch": 0.07626849235189873,
+      "grad_norm": 0.8525856137275696,
+      "learning_rate": 0.00022871996627792606,
+      "loss": 4.7595,
+      "step": 54280
+    },
+    {
+      "epoch": 0.0762825432900623,
+      "grad_norm": 0.9844761490821838,
+      "learning_rate": 0.00022876211887031052,
+      "loss": 4.829,
+      "step": 54290
+    },
+    {
+      "epoch": 0.07629659422822588,
+      "grad_norm": 0.8804373741149902,
+      "learning_rate": 0.00022880427146269495,
+      "loss": 4.8698,
+      "step": 54300
+    },
+    {
+      "epoch": 0.07631064516638945,
+      "grad_norm": 0.9130948781967163,
+      "learning_rate": 0.00022884642405507938,
+      "loss": 4.8689,
+      "step": 54310
+    },
+    {
+      "epoch": 0.07632469610455303,
+      "grad_norm": 0.8662232160568237,
+      "learning_rate": 0.00022888857664746378,
+      "loss": 4.9434,
+      "step": 54320
+    },
+    {
+      "epoch": 0.0763387470427166,
+      "grad_norm": 0.8689405918121338,
+      "learning_rate": 0.00022893072923984821,
+      "loss": 4.892,
+      "step": 54330
+    },
+    {
+      "epoch": 0.07635279798088018,
+      "grad_norm": 0.8666510581970215,
+      "learning_rate": 0.00022897288183223264,
+      "loss": 4.9568,
+      "step": 54340
+    },
+    {
+      "epoch": 0.07636684891904376,
+      "grad_norm": 0.885850191116333,
+      "learning_rate": 0.0002290150344246171,
+      "loss": 4.8736,
+      "step": 54350
+    },
+    {
+      "epoch": 0.07638089985720733,
+      "grad_norm": 0.8362758159637451,
+      "learning_rate": 0.00022905718701700153,
+      "loss": 4.7962,
+      "step": 54360
+    },
+    {
+      "epoch": 0.07639495079537093,
+      "grad_norm": 0.8462679982185364,
+      "learning_rate": 0.00022909933960938597,
+      "loss": 4.7975,
+      "step": 54370
+    },
+    {
+      "epoch": 0.0764090017335345,
+      "grad_norm": 0.8839713335037231,
+      "learning_rate": 0.0002291414922017704,
+      "loss": 4.7665,
+      "step": 54380
+    },
+    {
+      "epoch": 0.07642305267169808,
+      "grad_norm": 0.8993695974349976,
+      "learning_rate": 0.0002291836447941548,
+      "loss": 4.8324,
+      "step": 54390
+    },
+    {
+      "epoch": 0.07643710360986165,
+      "grad_norm": 0.8608269095420837,
+      "learning_rate": 0.00022922579738653923,
+      "loss": 4.7931,
+      "step": 54400
+    },
+    {
+      "epoch": 0.07645115454802523,
+      "grad_norm": 0.8435550928115845,
+      "learning_rate": 0.0002292679499789237,
+      "loss": 4.7409,
+      "step": 54410
+    },
+    {
+      "epoch": 0.0764652054861888,
+      "grad_norm": 0.9165601134300232,
+      "learning_rate": 0.00022931010257130812,
+      "loss": 4.7406,
+      "step": 54420
+    },
+    {
+      "epoch": 0.07647925642435238,
+      "grad_norm": 0.8588644862174988,
+      "learning_rate": 0.00022935225516369255,
+      "loss": 4.7544,
+      "step": 54430
+    },
+    {
+      "epoch": 0.07649330736251596,
+      "grad_norm": 0.9699477553367615,
+      "learning_rate": 0.00022939440775607698,
+      "loss": 4.8401,
+      "step": 54440
+    },
+    {
+      "epoch": 0.07650735830067953,
+      "grad_norm": 0.822721004486084,
+      "learning_rate": 0.0002294365603484614,
+      "loss": 4.9051,
+      "step": 54450
+    },
+    {
+      "epoch": 0.07652140923884311,
+      "grad_norm": 0.8678251504898071,
+      "learning_rate": 0.00022947871294084585,
+      "loss": 4.9236,
+      "step": 54460
+    },
+    {
+      "epoch": 0.07653546017700669,
+      "grad_norm": 0.8811358213424683,
+      "learning_rate": 0.00022952086553323028,
+      "loss": 4.9256,
+      "step": 54470
+    },
+    {
+      "epoch": 0.07654951111517028,
+      "grad_norm": 0.8711695075035095,
+      "learning_rate": 0.0002295630181256147,
+      "loss": 4.8998,
+      "step": 54480
+    },
+    {
+      "epoch": 0.07656356205333385,
+      "grad_norm": 0.8371874094009399,
+      "learning_rate": 0.00022960517071799914,
+      "loss": 4.7928,
+      "step": 54490
+    },
+    {
+      "epoch": 0.07657761299149743,
+      "grad_norm": 0.8641277551651001,
+      "learning_rate": 0.00022964732331038357,
+      "loss": 4.8052,
+      "step": 54500
+    },
+    {
+      "epoch": 0.076591663929661,
+      "grad_norm": 0.8634895086288452,
+      "learning_rate": 0.00022968947590276803,
+      "loss": 4.9161,
+      "step": 54510
+    },
+    {
+      "epoch": 0.07660571486782458,
+      "grad_norm": 0.9361057281494141,
+      "learning_rate": 0.00022973162849515243,
+      "loss": 4.8,
+      "step": 54520
+    },
+    {
+      "epoch": 0.07661976580598816,
+      "grad_norm": 0.8911561965942383,
+      "learning_rate": 0.00022977378108753686,
+      "loss": 4.8536,
+      "step": 54530
+    },
+    {
+      "epoch": 0.07663381674415173,
+      "grad_norm": 0.8985777497291565,
+      "learning_rate": 0.0002298159336799213,
+      "loss": 4.6861,
+      "step": 54540
+    },
+    {
+      "epoch": 0.07664786768231531,
+      "grad_norm": 0.8521196246147156,
+      "learning_rate": 0.00022985808627230573,
+      "loss": 4.8514,
+      "step": 54550
+    },
+    {
+      "epoch": 0.07666191862047889,
+      "grad_norm": 0.8784716129302979,
+      "learning_rate": 0.00022990023886469016,
+      "loss": 4.8231,
+      "step": 54560
+    },
+    {
+      "epoch": 0.07667596955864246,
+      "grad_norm": 0.8700553774833679,
+      "learning_rate": 0.00022994239145707462,
+      "loss": 4.8225,
+      "step": 54570
+    },
+    {
+      "epoch": 0.07669002049680605,
+      "grad_norm": 0.8594143986701965,
+      "learning_rate": 0.00022998454404945902,
+      "loss": 4.8293,
+      "step": 54580
+    },
+    {
+      "epoch": 0.07670407143496963,
+      "grad_norm": 0.8826280832290649,
+      "learning_rate": 0.00023002669664184345,
+      "loss": 4.8003,
+      "step": 54590
+    },
+    {
+      "epoch": 0.0767181223731332,
+      "grad_norm": 0.9344685673713684,
+      "learning_rate": 0.00023006884923422788,
+      "loss": 4.688,
+      "step": 54600
+    },
+    {
+      "epoch": 0.07673217331129678,
+      "grad_norm": 0.8549621105194092,
+      "learning_rate": 0.00023010678656737386,
+      "loss": 4.8083,
+      "step": 54610
+    },
+    {
+      "epoch": 0.07674622424946036,
+      "grad_norm": 0.8863023519515991,
+      "learning_rate": 0.0002301489391597583,
+      "loss": 4.681,
+      "step": 54620
+    },
+    {
+      "epoch": 0.07676027518762393,
+      "grad_norm": 0.8559615015983582,
+      "learning_rate": 0.00023019109175214273,
+      "loss": 4.8223,
+      "step": 54630
+    },
+    {
+      "epoch": 0.07677432612578751,
+      "grad_norm": 0.8779832720756531,
+      "learning_rate": 0.00023023324434452716,
+      "loss": 4.8405,
+      "step": 54640
+    },
+    {
+      "epoch": 0.07678837706395109,
+      "grad_norm": 0.8753707408905029,
+      "learning_rate": 0.00023027539693691162,
+      "loss": 4.8169,
+      "step": 54650
+    },
+    {
+      "epoch": 0.07680242800211466,
+      "grad_norm": 0.8928582668304443,
+      "learning_rate": 0.00023031754952929605,
+      "loss": 4.7668,
+      "step": 54660
+    },
+    {
+      "epoch": 0.07681647894027824,
+      "grad_norm": 0.8623060584068298,
+      "learning_rate": 0.00023035970212168048,
+      "loss": 4.7727,
+      "step": 54670
+    },
+    {
+      "epoch": 0.07683052987844183,
+      "grad_norm": 0.9036771655082703,
+      "learning_rate": 0.00023040185471406488,
+      "loss": 4.7807,
+      "step": 54680
+    },
+    {
+      "epoch": 0.0768445808166054,
+      "grad_norm": 0.8482112884521484,
+      "learning_rate": 0.00023044400730644931,
+      "loss": 4.7243,
+      "step": 54690
+    },
+    {
+      "epoch": 0.07685863175476898,
+      "grad_norm": 0.8324148058891296,
+      "learning_rate": 0.00023048615989883375,
+      "loss": 4.8233,
+      "step": 54700
+    },
+    {
+      "epoch": 0.07687268269293256,
+      "grad_norm": 0.8993374109268188,
+      "learning_rate": 0.0002305283124912182,
+      "loss": 4.9474,
+      "step": 54710
+    },
+    {
+      "epoch": 0.07688673363109613,
+      "grad_norm": 0.8752866387367249,
+      "learning_rate": 0.00023057046508360263,
+      "loss": 4.8148,
+      "step": 54720
+    },
+    {
+      "epoch": 0.07690078456925971,
+      "grad_norm": 0.884166419506073,
+      "learning_rate": 0.00023061261767598707,
+      "loss": 4.9107,
+      "step": 54730
+    },
+    {
+      "epoch": 0.07691483550742328,
+      "grad_norm": 0.8490076661109924,
+      "learning_rate": 0.00023065477026837147,
+      "loss": 4.7861,
+      "step": 54740
+    },
+    {
+      "epoch": 0.07692888644558686,
+      "grad_norm": 0.8822079300880432,
+      "learning_rate": 0.0002306969228607559,
+      "loss": 4.959,
+      "step": 54750
+    },
+    {
+      "epoch": 0.07694293738375044,
+      "grad_norm": 0.8657336831092834,
+      "learning_rate": 0.00023073907545314033,
+      "loss": 4.7831,
+      "step": 54760
+    },
+    {
+      "epoch": 0.07695698832191401,
+      "grad_norm": 0.8511672019958496,
+      "learning_rate": 0.0002307812280455248,
+      "loss": 4.7675,
+      "step": 54770
+    },
+    {
+      "epoch": 0.07697103926007759,
+      "grad_norm": 0.8640127778053284,
+      "learning_rate": 0.00023082338063790922,
+      "loss": 4.8313,
+      "step": 54780
+    },
+    {
+      "epoch": 0.07698509019824118,
+      "grad_norm": 0.8668428659439087,
+      "learning_rate": 0.00023086553323029365,
+      "loss": 4.9068,
+      "step": 54790
+    },
+    {
+      "epoch": 0.07699914113640476,
+      "grad_norm": 0.8341074585914612,
+      "learning_rate": 0.00023090768582267808,
+      "loss": 4.8259,
+      "step": 54800
+    },
+    {
+      "epoch": 0.07701319207456833,
+      "grad_norm": 0.8660095930099487,
+      "learning_rate": 0.0002309498384150625,
+      "loss": 4.7503,
+      "step": 54810
+    },
+    {
+      "epoch": 0.07702724301273191,
+      "grad_norm": 0.8701623678207397,
+      "learning_rate": 0.00023099199100744692,
+      "loss": 4.7477,
+      "step": 54820
+    },
+    {
+      "epoch": 0.07704129395089548,
+      "grad_norm": 0.8704190850257874,
+      "learning_rate": 0.00023103414359983138,
+      "loss": 4.7662,
+      "step": 54830
+    },
+    {
+      "epoch": 0.07705534488905906,
+      "grad_norm": 0.8704859614372253,
+      "learning_rate": 0.0002310762961922158,
+      "loss": 4.8061,
+      "step": 54840
+    },
+    {
+      "epoch": 0.07706939582722264,
+      "grad_norm": 0.8688639998435974,
+      "learning_rate": 0.00023111844878460024,
+      "loss": 4.6981,
+      "step": 54850
+    },
+    {
+      "epoch": 0.07708344676538621,
+      "grad_norm": 0.8302347660064697,
+      "learning_rate": 0.00023116060137698467,
+      "loss": 4.7741,
+      "step": 54860
+    },
+    {
+      "epoch": 0.07709749770354979,
+      "grad_norm": 0.8849836587905884,
+      "learning_rate": 0.00023120275396936908,
+      "loss": 4.8139,
+      "step": 54870
+    },
+    {
+      "epoch": 0.07711154864171336,
+      "grad_norm": 0.8696654438972473,
+      "learning_rate": 0.0002312449065617535,
+      "loss": 4.8226,
+      "step": 54880
+    },
+    {
+      "epoch": 0.07712559957987696,
+      "grad_norm": 0.8935073018074036,
+      "learning_rate": 0.00023128705915413796,
+      "loss": 4.7954,
+      "step": 54890
+    },
+    {
+      "epoch": 0.07713965051804053,
+      "grad_norm": 0.8748064637184143,
+      "learning_rate": 0.0002313292117465224,
+      "loss": 4.8052,
+      "step": 54900
+    },
+    {
+      "epoch": 0.07715370145620411,
+      "grad_norm": 0.8696599006652832,
+      "learning_rate": 0.00023137136433890683,
+      "loss": 4.8104,
+      "step": 54910
+    },
+    {
+      "epoch": 0.07716775239436768,
+      "grad_norm": 0.852148175239563,
+      "learning_rate": 0.00023141351693129126,
+      "loss": 4.7914,
+      "step": 54920
+    },
+    {
+      "epoch": 0.07718180333253126,
+      "grad_norm": 0.849692702293396,
+      "learning_rate": 0.0002314556695236757,
+      "loss": 4.8389,
+      "step": 54930
+    },
+    {
+      "epoch": 0.07719585427069484,
+      "grad_norm": 0.8747760057449341,
+      "learning_rate": 0.0002314978221160601,
+      "loss": 4.8114,
+      "step": 54940
+    },
+    {
+      "epoch": 0.07720990520885841,
+      "grad_norm": 0.972870409488678,
+      "learning_rate": 0.00023153997470844455,
+      "loss": 4.9658,
+      "step": 54950
+    },
+    {
+      "epoch": 0.07722395614702199,
+      "grad_norm": 0.883500337600708,
+      "learning_rate": 0.00023158212730082898,
+      "loss": 4.915,
+      "step": 54960
+    },
+    {
+      "epoch": 0.07723800708518556,
+      "grad_norm": 0.860719621181488,
+      "learning_rate": 0.00023162427989321341,
+      "loss": 4.7428,
+      "step": 54970
+    },
+    {
+      "epoch": 0.07725205802334914,
+      "grad_norm": 0.8673244118690491,
+      "learning_rate": 0.00023166643248559785,
+      "loss": 4.9087,
+      "step": 54980
+    },
+    {
+      "epoch": 0.07726610896151273,
+      "grad_norm": 0.862847626209259,
+      "learning_rate": 0.00023170858507798228,
+      "loss": 4.7927,
+      "step": 54990
+    },
+    {
+      "epoch": 0.0772801598996763,
+      "grad_norm": 0.873631477355957,
+      "learning_rate": 0.00023175073767036668,
+      "loss": 4.7431,
+      "step": 55000
+    },
+    {
+      "epoch": 0.07729421083783988,
+      "grad_norm": 0.8591784834861755,
+      "learning_rate": 0.00023179289026275114,
+      "loss": 4.8536,
+      "step": 55010
+    },
+    {
+      "epoch": 0.07730826177600346,
+      "grad_norm": 0.8631607890129089,
+      "learning_rate": 0.00023183504285513557,
+      "loss": 4.7317,
+      "step": 55020
+    },
+    {
+      "epoch": 0.07732231271416704,
+      "grad_norm": 0.8369123339653015,
+      "learning_rate": 0.00023187719544752,
+      "loss": 4.7036,
+      "step": 55030
+    },
+    {
+      "epoch": 0.07733636365233061,
+      "grad_norm": 0.8408652544021606,
+      "learning_rate": 0.00023191934803990443,
+      "loss": 4.9028,
+      "step": 55040
+    },
+    {
+      "epoch": 0.07735041459049419,
+      "grad_norm": 0.8339114189147949,
+      "learning_rate": 0.0002319615006322889,
+      "loss": 4.7352,
+      "step": 55050
+    },
+    {
+      "epoch": 0.07736446552865776,
+      "grad_norm": 0.8363640308380127,
+      "learning_rate": 0.00023200365322467327,
+      "loss": 4.7126,
+      "step": 55060
+    },
+    {
+      "epoch": 0.07737851646682134,
+      "grad_norm": 0.8503546714782715,
+      "learning_rate": 0.00023204580581705773,
+      "loss": 4.8552,
+      "step": 55070
+    },
+    {
+      "epoch": 0.07739256740498492,
+      "grad_norm": 0.8623889088630676,
+      "learning_rate": 0.00023208795840944216,
+      "loss": 4.7625,
+      "step": 55080
+    },
+    {
+      "epoch": 0.07740661834314849,
+      "grad_norm": 0.8672446608543396,
+      "learning_rate": 0.0002321301110018266,
+      "loss": 4.703,
+      "step": 55090
+    },
+    {
+      "epoch": 0.07742066928131208,
+      "grad_norm": 0.83352130651474,
+      "learning_rate": 0.00023217226359421102,
+      "loss": 4.9044,
+      "step": 55100
+    },
+    {
+      "epoch": 0.07743472021947566,
+      "grad_norm": 0.8357213139533997,
+      "learning_rate": 0.00023221441618659548,
+      "loss": 4.7765,
+      "step": 55110
+    },
+    {
+      "epoch": 0.07744877115763923,
+      "grad_norm": 0.8617483973503113,
+      "learning_rate": 0.0002322565687789799,
+      "loss": 4.8928,
+      "step": 55120
+    },
+    {
+      "epoch": 0.07746282209580281,
+      "grad_norm": 0.8757510185241699,
+      "learning_rate": 0.0002322987213713643,
+      "loss": 4.9618,
+      "step": 55130
+    },
+    {
+      "epoch": 0.07747687303396639,
+      "grad_norm": 0.8584406971931458,
+      "learning_rate": 0.00023234087396374874,
+      "loss": 4.7733,
+      "step": 55140
+    },
+    {
+      "epoch": 0.07749092397212996,
+      "grad_norm": 0.8585185408592224,
+      "learning_rate": 0.00023238302655613318,
+      "loss": 4.8477,
+      "step": 55150
+    },
+    {
+      "epoch": 0.07750497491029354,
+      "grad_norm": 0.9038747549057007,
+      "learning_rate": 0.0002324251791485176,
+      "loss": 4.9543,
+      "step": 55160
+    },
+    {
+      "epoch": 0.07751902584845712,
+      "grad_norm": 0.8479253649711609,
+      "learning_rate": 0.00023246733174090206,
+      "loss": 4.7465,
+      "step": 55170
+    },
+    {
+      "epoch": 0.07753307678662069,
+      "grad_norm": 0.8460027575492859,
+      "learning_rate": 0.0002325094843332865,
+      "loss": 4.8782,
+      "step": 55180
+    },
+    {
+      "epoch": 0.07754712772478427,
+      "grad_norm": 0.8895059823989868,
+      "learning_rate": 0.0002325516369256709,
+      "loss": 4.9051,
+      "step": 55190
+    },
+    {
+      "epoch": 0.07756117866294786,
+      "grad_norm": 0.8584068417549133,
+      "learning_rate": 0.00023259378951805533,
+      "loss": 4.8017,
+      "step": 55200
+    },
+    {
+      "epoch": 0.07757522960111143,
+      "grad_norm": 0.8654706478118896,
+      "learning_rate": 0.00023263594211043976,
+      "loss": 4.8685,
+      "step": 55210
+    },
+    {
+      "epoch": 0.07758928053927501,
+      "grad_norm": 0.8444214463233948,
+      "learning_rate": 0.0002326780947028242,
+      "loss": 4.7856,
+      "step": 55220
+    },
+    {
+      "epoch": 0.07760333147743859,
+      "grad_norm": 0.9347013235092163,
+      "learning_rate": 0.00023272024729520865,
+      "loss": 4.8631,
+      "step": 55230
+    },
+    {
+      "epoch": 0.07761738241560216,
+      "grad_norm": 0.8600069880485535,
+      "learning_rate": 0.00023276239988759308,
+      "loss": 4.953,
+      "step": 55240
+    },
+    {
+      "epoch": 0.07763143335376574,
+      "grad_norm": 0.8897565007209778,
+      "learning_rate": 0.00023280455247997751,
+      "loss": 4.8058,
+      "step": 55250
+    },
+    {
+      "epoch": 0.07764548429192931,
+      "grad_norm": 0.868504524230957,
+      "learning_rate": 0.00023284670507236192,
+      "loss": 4.7797,
+      "step": 55260
+    },
+    {
+      "epoch": 0.07765953523009289,
+      "grad_norm": 0.8592216372489929,
+      "learning_rate": 0.00023288885766474635,
+      "loss": 4.7845,
+      "step": 55270
+    },
+    {
+      "epoch": 0.07767358616825647,
+      "grad_norm": 0.8826098442077637,
+      "learning_rate": 0.00023293101025713078,
+      "loss": 4.7159,
+      "step": 55280
+    },
+    {
+      "epoch": 0.07768763710642004,
+      "grad_norm": 0.8514786958694458,
+      "learning_rate": 0.00023297316284951524,
+      "loss": 4.7917,
+      "step": 55290
+    },
+    {
+      "epoch": 0.07770168804458363,
+      "grad_norm": 0.9170621037483215,
+      "learning_rate": 0.00023301531544189967,
+      "loss": 4.7785,
+      "step": 55300
+    },
+    {
+      "epoch": 0.07771573898274721,
+      "grad_norm": 0.8372135162353516,
+      "learning_rate": 0.0002330574680342841,
+      "loss": 4.8629,
+      "step": 55310
+    },
+    {
+      "epoch": 0.07772978992091079,
+      "grad_norm": 0.867605447769165,
+      "learning_rate": 0.0002330996206266685,
+      "loss": 4.8731,
+      "step": 55320
+    },
+    {
+      "epoch": 0.07774384085907436,
+      "grad_norm": 0.8326741456985474,
+      "learning_rate": 0.00023314177321905294,
+      "loss": 4.9606,
+      "step": 55330
+    },
+    {
+      "epoch": 0.07775789179723794,
+      "grad_norm": 0.8772097229957581,
+      "learning_rate": 0.00023318392581143737,
+      "loss": 4.7784,
+      "step": 55340
+    },
+    {
+      "epoch": 0.07777194273540151,
+      "grad_norm": 0.8476175665855408,
+      "learning_rate": 0.00023322607840382183,
+      "loss": 4.9128,
+      "step": 55350
+    },
+    {
+      "epoch": 0.07778599367356509,
+      "grad_norm": 0.8802474737167358,
+      "learning_rate": 0.00023326823099620626,
+      "loss": 4.7922,
+      "step": 55360
+    },
+    {
+      "epoch": 0.07780004461172867,
+      "grad_norm": 0.8601409792900085,
+      "learning_rate": 0.0002333103835885907,
+      "loss": 4.8804,
+      "step": 55370
+    },
+    {
+      "epoch": 0.07781409554989224,
+      "grad_norm": 0.8463119864463806,
+      "learning_rate": 0.00023335253618097512,
+      "loss": 4.8471,
+      "step": 55380
+    },
+    {
+      "epoch": 0.07782814648805582,
+      "grad_norm": 0.8972960114479065,
+      "learning_rate": 0.00023339468877335952,
+      "loss": 4.8792,
+      "step": 55390
+    },
+    {
+      "epoch": 0.0778421974262194,
+      "grad_norm": 0.9296767711639404,
+      "learning_rate": 0.00023343684136574395,
+      "loss": 4.7318,
+      "step": 55400
+    },
+    {
+      "epoch": 0.07785624836438299,
+      "grad_norm": 0.9014337062835693,
+      "learning_rate": 0.0002334789939581284,
+      "loss": 4.8008,
+      "step": 55410
+    },
+    {
+      "epoch": 0.07787029930254656,
+      "grad_norm": 0.8471907377243042,
+      "learning_rate": 0.00023352114655051284,
+      "loss": 4.9053,
+      "step": 55420
+    },
+    {
+      "epoch": 0.07788435024071014,
+      "grad_norm": 0.8656927347183228,
+      "learning_rate": 0.00023356329914289727,
+      "loss": 4.8065,
+      "step": 55430
+    },
+    {
+      "epoch": 0.07789840117887371,
+      "grad_norm": 0.8613098859786987,
+      "learning_rate": 0.0002336054517352817,
+      "loss": 5.0078,
+      "step": 55440
+    },
+    {
+      "epoch": 0.07791245211703729,
+      "grad_norm": 0.8597536683082581,
+      "learning_rate": 0.0002336476043276661,
+      "loss": 4.8672,
+      "step": 55450
+    },
+    {
+      "epoch": 0.07792650305520087,
+      "grad_norm": 0.8532637357711792,
+      "learning_rate": 0.00023368975692005054,
+      "loss": 4.7714,
+      "step": 55460
+    },
+    {
+      "epoch": 0.07794055399336444,
+      "grad_norm": 0.9325604438781738,
+      "learning_rate": 0.000233731909512435,
+      "loss": 4.6782,
+      "step": 55470
+    },
+    {
+      "epoch": 0.07795460493152802,
+      "grad_norm": 0.8740974068641663,
+      "learning_rate": 0.00023377406210481943,
+      "loss": 4.7074,
+      "step": 55480
+    },
+    {
+      "epoch": 0.0779686558696916,
+      "grad_norm": 0.8521115183830261,
+      "learning_rate": 0.00023381621469720386,
+      "loss": 4.8376,
+      "step": 55490
+    },
+    {
+      "epoch": 0.07798270680785517,
+      "grad_norm": 0.8538252711296082,
+      "learning_rate": 0.0002338583672895883,
+      "loss": 4.8655,
+      "step": 55500
+    },
+    {
+      "epoch": 0.07799675774601876,
+      "grad_norm": 0.8827087879180908,
+      "learning_rate": 0.00023390051988197272,
+      "loss": 4.9127,
+      "step": 55510
+    },
+    {
+      "epoch": 0.07801080868418234,
+      "grad_norm": 0.8596060872077942,
+      "learning_rate": 0.00023394267247435713,
+      "loss": 4.7259,
+      "step": 55520
+    },
+    {
+      "epoch": 0.07802485962234591,
+      "grad_norm": 0.8459311723709106,
+      "learning_rate": 0.0002339848250667416,
+      "loss": 4.9033,
+      "step": 55530
+    },
+    {
+      "epoch": 0.07803891056050949,
+      "grad_norm": 0.8380910754203796,
+      "learning_rate": 0.00023402697765912602,
+      "loss": 4.7834,
+      "step": 55540
+    },
+    {
+      "epoch": 0.07805296149867307,
+      "grad_norm": 0.8815401792526245,
+      "learning_rate": 0.00023406913025151045,
+      "loss": 4.8435,
+      "step": 55550
+    },
+    {
+      "epoch": 0.07806701243683664,
+      "grad_norm": 0.8380628228187561,
+      "learning_rate": 0.00023411128284389488,
+      "loss": 4.8127,
+      "step": 55560
+    },
+    {
+      "epoch": 0.07808106337500022,
+      "grad_norm": 0.8409835696220398,
+      "learning_rate": 0.0002341534354362793,
+      "loss": 4.8411,
+      "step": 55570
+    },
+    {
+      "epoch": 0.0780951143131638,
+      "grad_norm": 0.8814390897750854,
+      "learning_rate": 0.00023419558802866372,
+      "loss": 4.8431,
+      "step": 55580
+    },
+    {
+      "epoch": 0.07810916525132737,
+      "grad_norm": 0.8419836163520813,
+      "learning_rate": 0.00023423774062104817,
+      "loss": 4.8255,
+      "step": 55590
+    },
+    {
+      "epoch": 0.07812321618949095,
+      "grad_norm": 0.8572084903717041,
+      "learning_rate": 0.0002342798932134326,
+      "loss": 4.727,
+      "step": 55600
+    },
+    {
+      "epoch": 0.07813726712765454,
+      "grad_norm": 0.8860167860984802,
+      "learning_rate": 0.00023432204580581704,
+      "loss": 4.9694,
+      "step": 55610
+    },
+    {
+      "epoch": 0.07815131806581811,
+      "grad_norm": 0.8654281497001648,
+      "learning_rate": 0.00023436419839820147,
+      "loss": 4.8643,
+      "step": 55620
+    },
+    {
+      "epoch": 0.07816536900398169,
+      "grad_norm": 0.848604142665863,
+      "learning_rate": 0.0002344063509905859,
+      "loss": 4.9082,
+      "step": 55630
+    },
+    {
+      "epoch": 0.07817941994214526,
+      "grad_norm": 0.8291561007499695,
+      "learning_rate": 0.00023444850358297036,
+      "loss": 4.9041,
+      "step": 55640
+    },
+    {
+      "epoch": 0.07819347088030884,
+      "grad_norm": 0.8431739807128906,
+      "learning_rate": 0.00023449065617535476,
+      "loss": 4.8418,
+      "step": 55650
+    },
+    {
+      "epoch": 0.07820752181847242,
+      "grad_norm": 0.8272885084152222,
+      "learning_rate": 0.0002345328087677392,
+      "loss": 4.7878,
+      "step": 55660
+    },
+    {
+      "epoch": 0.078221572756636,
+      "grad_norm": 0.8414096236228943,
+      "learning_rate": 0.00023457496136012362,
+      "loss": 4.9341,
+      "step": 55670
+    },
+    {
+      "epoch": 0.07823562369479957,
+      "grad_norm": 0.8919362425804138,
+      "learning_rate": 0.00023461711395250805,
+      "loss": 4.9505,
+      "step": 55680
+    },
+    {
+      "epoch": 0.07824967463296315,
+      "grad_norm": 0.8313416242599487,
+      "learning_rate": 0.0002346592665448925,
+      "loss": 4.9542,
+      "step": 55690
+    },
+    {
+      "epoch": 0.07826372557112672,
+      "grad_norm": 0.902398943901062,
+      "learning_rate": 0.00023470141913727694,
+      "loss": 4.8456,
+      "step": 55700
+    },
+    {
+      "epoch": 0.07827777650929031,
+      "grad_norm": 0.8654336333274841,
+      "learning_rate": 0.00023474357172966135,
+      "loss": 4.7756,
+      "step": 55710
+    },
+    {
+      "epoch": 0.07829182744745389,
+      "grad_norm": 0.848102867603302,
+      "learning_rate": 0.00023478572432204578,
+      "loss": 4.9353,
+      "step": 55720
+    },
+    {
+      "epoch": 0.07830587838561746,
+      "grad_norm": 0.8511942028999329,
+      "learning_rate": 0.0002348278769144302,
+      "loss": 4.7061,
+      "step": 55730
+    },
+    {
+      "epoch": 0.07831992932378104,
+      "grad_norm": 0.8495092988014221,
+      "learning_rate": 0.00023487002950681464,
+      "loss": 4.8499,
+      "step": 55740
+    },
+    {
+      "epoch": 0.07833398026194462,
+      "grad_norm": 0.8534985780715942,
+      "learning_rate": 0.0002349121820991991,
+      "loss": 4.7557,
+      "step": 55750
+    },
+    {
+      "epoch": 0.07834803120010819,
+      "grad_norm": 0.8438852429389954,
+      "learning_rate": 0.00023495433469158353,
+      "loss": 4.7647,
+      "step": 55760
+    },
+    {
+      "epoch": 0.07836208213827177,
+      "grad_norm": 0.8534382581710815,
+      "learning_rate": 0.00023499648728396793,
+      "loss": 4.7887,
+      "step": 55770
+    },
+    {
+      "epoch": 0.07837613307643534,
+      "grad_norm": 0.8537501096725464,
+      "learning_rate": 0.00023503863987635237,
+      "loss": 4.8719,
+      "step": 55780
+    },
+    {
+      "epoch": 0.07839018401459892,
+      "grad_norm": 0.862980306148529,
+      "learning_rate": 0.0002350807924687368,
+      "loss": 4.8047,
+      "step": 55790
+    },
+    {
+      "epoch": 0.0784042349527625,
+      "grad_norm": 0.8612253069877625,
+      "learning_rate": 0.00023512294506112123,
+      "loss": 4.8798,
+      "step": 55800
+    },
+    {
+      "epoch": 0.07841828589092607,
+      "grad_norm": 0.8723065257072449,
+      "learning_rate": 0.00023516509765350569,
+      "loss": 4.8036,
+      "step": 55810
+    },
+    {
+      "epoch": 0.07843233682908966,
+      "grad_norm": 0.8549647331237793,
+      "learning_rate": 0.00023520725024589012,
+      "loss": 4.9236,
+      "step": 55820
+    },
+    {
+      "epoch": 0.07844638776725324,
+      "grad_norm": 0.8528120517730713,
+      "learning_rate": 0.00023524940283827455,
+      "loss": 4.8222,
+      "step": 55830
+    },
+    {
+      "epoch": 0.07846043870541682,
+      "grad_norm": 0.8351045250892639,
+      "learning_rate": 0.00023529155543065895,
+      "loss": 4.8075,
+      "step": 55840
+    },
+    {
+      "epoch": 0.07847448964358039,
+      "grad_norm": 0.8468853831291199,
+      "learning_rate": 0.00023533370802304338,
+      "loss": 4.7578,
+      "step": 55850
+    },
+    {
+      "epoch": 0.07848854058174397,
+      "grad_norm": 0.9013736844062805,
+      "learning_rate": 0.00023537586061542781,
+      "loss": 4.7043,
+      "step": 55860
+    },
+    {
+      "epoch": 0.07850259151990754,
+      "grad_norm": 0.8511090874671936,
+      "learning_rate": 0.00023541801320781227,
+      "loss": 4.8785,
+      "step": 55870
+    },
+    {
+      "epoch": 0.07851664245807112,
+      "grad_norm": 0.8878459930419922,
+      "learning_rate": 0.0002354601658001967,
+      "loss": 4.8643,
+      "step": 55880
+    },
+    {
+      "epoch": 0.0785306933962347,
+      "grad_norm": 0.873618483543396,
+      "learning_rate": 0.00023550231839258114,
+      "loss": 4.7879,
+      "step": 55890
+    },
+    {
+      "epoch": 0.07854474433439827,
+      "grad_norm": 0.829479992389679,
+      "learning_rate": 0.00023554447098496554,
+      "loss": 4.6622,
+      "step": 55900
+    },
+    {
+      "epoch": 0.07855879527256185,
+      "grad_norm": 0.8406625986099243,
+      "learning_rate": 0.00023558662357734997,
+      "loss": 4.972,
+      "step": 55910
+    },
+    {
+      "epoch": 0.07857284621072544,
+      "grad_norm": 0.8553104400634766,
+      "learning_rate": 0.0002356287761697344,
+      "loss": 4.7771,
+      "step": 55920
+    },
+    {
+      "epoch": 0.07858689714888902,
+      "grad_norm": 0.8392889499664307,
+      "learning_rate": 0.00023567092876211886,
+      "loss": 4.7373,
+      "step": 55930
+    },
+    {
+      "epoch": 0.07860094808705259,
+      "grad_norm": 0.8430394530296326,
+      "learning_rate": 0.0002357130813545033,
+      "loss": 4.9143,
+      "step": 55940
+    },
+    {
+      "epoch": 0.07861499902521617,
+      "grad_norm": 0.8386268615722656,
+      "learning_rate": 0.00023575523394688772,
+      "loss": 4.8825,
+      "step": 55950
+    },
+    {
+      "epoch": 0.07862904996337974,
+      "grad_norm": 0.8521925210952759,
+      "learning_rate": 0.00023579738653927215,
+      "loss": 4.7222,
+      "step": 55960
+    },
+    {
+      "epoch": 0.07864310090154332,
+      "grad_norm": 0.8255799412727356,
+      "learning_rate": 0.00023583953913165656,
+      "loss": 4.8464,
+      "step": 55970
+    },
+    {
+      "epoch": 0.0786571518397069,
+      "grad_norm": 0.9146137237548828,
+      "learning_rate": 0.000235881691724041,
+      "loss": 4.7756,
+      "step": 55980
+    },
+    {
+      "epoch": 0.07867120277787047,
+      "grad_norm": 0.9210695624351501,
+      "learning_rate": 0.00023592384431642545,
+      "loss": 4.7982,
+      "step": 55990
+    },
+    {
+      "epoch": 0.07868525371603405,
+      "grad_norm": 0.8495543003082275,
+      "learning_rate": 0.00023596599690880988,
+      "loss": 4.6518,
+      "step": 56000
+    },
+    {
+      "epoch": 0.07869930465419762,
+      "grad_norm": 0.8346798419952393,
+      "learning_rate": 0.0002360081495011943,
+      "loss": 4.8465,
+      "step": 56010
+    },
+    {
+      "epoch": 0.07871335559236121,
+      "grad_norm": 0.8985382914543152,
+      "learning_rate": 0.00023605030209357874,
+      "loss": 4.7423,
+      "step": 56020
+    },
+    {
+      "epoch": 0.07872740653052479,
+      "grad_norm": 0.8527060747146606,
+      "learning_rate": 0.00023609245468596314,
+      "loss": 4.9118,
+      "step": 56030
+    },
+    {
+      "epoch": 0.07874145746868837,
+      "grad_norm": 0.8444516658782959,
+      "learning_rate": 0.00023613460727834758,
+      "loss": 4.9117,
+      "step": 56040
+    },
+    {
+      "epoch": 0.07875550840685194,
+      "grad_norm": 0.879189133644104,
+      "learning_rate": 0.00023617675987073203,
+      "loss": 4.6753,
+      "step": 56050
+    },
+    {
+      "epoch": 0.07876955934501552,
+      "grad_norm": 0.8668951392173767,
+      "learning_rate": 0.00023621891246311647,
+      "loss": 4.8826,
+      "step": 56060
+    },
+    {
+      "epoch": 0.0787836102831791,
+      "grad_norm": 0.8374396562576294,
+      "learning_rate": 0.0002362610650555009,
+      "loss": 4.8564,
+      "step": 56070
+    },
+    {
+      "epoch": 0.07879766122134267,
+      "grad_norm": 0.8434193730354309,
+      "learning_rate": 0.00023630321764788533,
+      "loss": 4.85,
+      "step": 56080
+    },
+    {
+      "epoch": 0.07881171215950625,
+      "grad_norm": 0.8440457582473755,
+      "learning_rate": 0.00023634537024026976,
+      "loss": 4.8728,
+      "step": 56090
+    },
+    {
+      "epoch": 0.07882576309766982,
+      "grad_norm": 0.8732414245605469,
+      "learning_rate": 0.00023638752283265416,
+      "loss": 4.9149,
+      "step": 56100
+    },
+    {
+      "epoch": 0.0788398140358334,
+      "grad_norm": 0.8345431685447693,
+      "learning_rate": 0.00023642967542503862,
+      "loss": 4.9031,
+      "step": 56110
+    },
+    {
+      "epoch": 0.07885386497399698,
+      "grad_norm": 0.8340375423431396,
+      "learning_rate": 0.00023647182801742305,
+      "loss": 4.8705,
+      "step": 56120
+    },
+    {
+      "epoch": 0.07886791591216057,
+      "grad_norm": 0.8727152943611145,
+      "learning_rate": 0.00023651398060980748,
+      "loss": 4.8331,
+      "step": 56130
+    },
+    {
+      "epoch": 0.07888196685032414,
+      "grad_norm": 0.8571959733963013,
+      "learning_rate": 0.00023655613320219191,
+      "loss": 4.8701,
+      "step": 56140
+    },
+    {
+      "epoch": 0.07889601778848772,
+      "grad_norm": 0.8859211802482605,
+      "learning_rate": 0.00023659828579457635,
+      "loss": 4.7979,
+      "step": 56150
+    },
+    {
+      "epoch": 0.0789100687266513,
+      "grad_norm": 0.8285291790962219,
+      "learning_rate": 0.00023664043838696075,
+      "loss": 4.862,
+      "step": 56160
+    },
+    {
+      "epoch": 0.07892411966481487,
+      "grad_norm": 0.8508478403091431,
+      "learning_rate": 0.0002366825909793452,
+      "loss": 4.8707,
+      "step": 56170
+    },
+    {
+      "epoch": 0.07893817060297845,
+      "grad_norm": 0.8419012427330017,
+      "learning_rate": 0.00023672474357172964,
+      "loss": 4.8833,
+      "step": 56180
+    },
+    {
+      "epoch": 0.07895222154114202,
+      "grad_norm": 0.9277992844581604,
+      "learning_rate": 0.00023676689616411407,
+      "loss": 4.8056,
+      "step": 56190
+    },
+    {
+      "epoch": 0.0789662724793056,
+      "grad_norm": 0.879784882068634,
+      "learning_rate": 0.0002368090487564985,
+      "loss": 4.9411,
+      "step": 56200
+    },
+    {
+      "epoch": 0.07898032341746918,
+      "grad_norm": 0.8691972494125366,
+      "learning_rate": 0.00023685120134888293,
+      "loss": 4.7991,
+      "step": 56210
+    },
+    {
+      "epoch": 0.07899437435563275,
+      "grad_norm": 0.8471894860267639,
+      "learning_rate": 0.0002368933539412674,
+      "loss": 4.6985,
+      "step": 56220
+    },
+    {
+      "epoch": 0.07900842529379634,
+      "grad_norm": 0.8567607402801514,
+      "learning_rate": 0.0002369355065336518,
+      "loss": 4.8068,
+      "step": 56230
+    },
+    {
+      "epoch": 0.07902247623195992,
+      "grad_norm": 0.8540003895759583,
+      "learning_rate": 0.00023697765912603623,
+      "loss": 4.6384,
+      "step": 56240
+    },
+    {
+      "epoch": 0.0790365271701235,
+      "grad_norm": 0.886030912399292,
+      "learning_rate": 0.00023701981171842066,
+      "loss": 4.8265,
+      "step": 56250
+    },
+    {
+      "epoch": 0.07905057810828707,
+      "grad_norm": 0.9057062268257141,
+      "learning_rate": 0.0002370619643108051,
+      "loss": 4.881,
+      "step": 56260
+    },
+    {
+      "epoch": 0.07906462904645065,
+      "grad_norm": 0.8884521126747131,
+      "learning_rate": 0.00023710411690318955,
+      "loss": 4.7785,
+      "step": 56270
+    },
+    {
+      "epoch": 0.07907867998461422,
+      "grad_norm": 0.8333665132522583,
+      "learning_rate": 0.00023714626949557398,
+      "loss": 4.83,
+      "step": 56280
+    },
+    {
+      "epoch": 0.0790927309227778,
+      "grad_norm": 0.8531104922294617,
+      "learning_rate": 0.00023718842208795838,
+      "loss": 4.7392,
+      "step": 56290
+    },
+    {
+      "epoch": 0.07910678186094137,
+      "grad_norm": 0.853764533996582,
+      "learning_rate": 0.0002372305746803428,
+      "loss": 4.8426,
+      "step": 56300
+    },
+    {
+      "epoch": 0.07912083279910495,
+      "grad_norm": 0.8597157597541809,
+      "learning_rate": 0.00023727272727272724,
+      "loss": 4.9028,
+      "step": 56310
+    },
+    {
+      "epoch": 0.07913488373726853,
+      "grad_norm": 0.8901702165603638,
+      "learning_rate": 0.00023731487986511168,
+      "loss": 4.7871,
+      "step": 56320
+    },
+    {
+      "epoch": 0.07914893467543212,
+      "grad_norm": 0.8476951122283936,
+      "learning_rate": 0.00023735703245749613,
+      "loss": 4.7935,
+      "step": 56330
+    },
+    {
+      "epoch": 0.0791629856135957,
+      "grad_norm": 0.8251456022262573,
+      "learning_rate": 0.00023739918504988057,
+      "loss": 4.8088,
+      "step": 56340
+    },
+    {
+      "epoch": 0.07917703655175927,
+      "grad_norm": 0.8737718462944031,
+      "learning_rate": 0.000237441337642265,
+      "loss": 4.8629,
+      "step": 56350
+    },
+    {
+      "epoch": 0.07919108748992285,
+      "grad_norm": 0.847468912601471,
+      "learning_rate": 0.0002374834902346494,
+      "loss": 4.8208,
+      "step": 56360
+    },
+    {
+      "epoch": 0.07920513842808642,
+      "grad_norm": 0.8241137266159058,
+      "learning_rate": 0.00023752564282703383,
+      "loss": 4.7132,
+      "step": 56370
+    },
+    {
+      "epoch": 0.07921918936625,
+      "grad_norm": 0.8659803867340088,
+      "learning_rate": 0.00023756779541941826,
+      "loss": 4.7646,
+      "step": 56380
+    },
+    {
+      "epoch": 0.07923324030441357,
+      "grad_norm": 0.876525342464447,
+      "learning_rate": 0.00023760994801180272,
+      "loss": 4.8045,
+      "step": 56390
+    },
+    {
+      "epoch": 0.07924729124257715,
+      "grad_norm": 0.8721202611923218,
+      "learning_rate": 0.00023765210060418715,
+      "loss": 4.8596,
+      "step": 56400
+    },
+    {
+      "epoch": 0.07926134218074073,
+      "grad_norm": 0.8766302466392517,
+      "learning_rate": 0.00023769425319657158,
+      "loss": 4.8604,
+      "step": 56410
+    },
+    {
+      "epoch": 0.0792753931189043,
+      "grad_norm": 0.8864064812660217,
+      "learning_rate": 0.000237736405788956,
+      "loss": 4.7736,
+      "step": 56420
+    },
+    {
+      "epoch": 0.07928944405706788,
+      "grad_norm": 0.8010143041610718,
+      "learning_rate": 0.00023777855838134042,
+      "loss": 4.8483,
+      "step": 56430
+    },
+    {
+      "epoch": 0.07930349499523147,
+      "grad_norm": 0.8585301637649536,
+      "learning_rate": 0.00023782071097372485,
+      "loss": 4.8176,
+      "step": 56440
+    },
+    {
+      "epoch": 0.07931754593339504,
+      "grad_norm": 0.8217198848724365,
+      "learning_rate": 0.0002378628635661093,
+      "loss": 4.8178,
+      "step": 56450
+    },
+    {
+      "epoch": 0.07933159687155862,
+      "grad_norm": 0.8594200015068054,
+      "learning_rate": 0.00023790501615849374,
+      "loss": 4.7902,
+      "step": 56460
+    },
+    {
+      "epoch": 0.0793456478097222,
+      "grad_norm": 0.8369775414466858,
+      "learning_rate": 0.00023794716875087817,
+      "loss": 4.6128,
+      "step": 56470
+    },
+    {
+      "epoch": 0.07935969874788577,
+      "grad_norm": 0.8443807363510132,
+      "learning_rate": 0.00023798932134326257,
+      "loss": 4.9548,
+      "step": 56480
+    },
+    {
+      "epoch": 0.07937374968604935,
+      "grad_norm": 0.921536386013031,
+      "learning_rate": 0.000238031473935647,
+      "loss": 4.9101,
+      "step": 56490
+    },
+    {
+      "epoch": 0.07938780062421293,
+      "grad_norm": 0.842056155204773,
+      "learning_rate": 0.00023807362652803144,
+      "loss": 4.825,
+      "step": 56500
+    },
+    {
+      "epoch": 0.0794018515623765,
+      "grad_norm": 0.8614702224731445,
+      "learning_rate": 0.0002381157791204159,
+      "loss": 4.7713,
+      "step": 56510
+    },
+    {
+      "epoch": 0.07941590250054008,
+      "grad_norm": 0.8810839056968689,
+      "learning_rate": 0.00023815793171280033,
+      "loss": 4.9506,
+      "step": 56520
+    },
+    {
+      "epoch": 0.07942995343870365,
+      "grad_norm": 0.8495205640792847,
+      "learning_rate": 0.00023820008430518476,
+      "loss": 4.8596,
+      "step": 56530
+    },
+    {
+      "epoch": 0.07944400437686724,
+      "grad_norm": 0.829397439956665,
+      "learning_rate": 0.0002382422368975692,
+      "loss": 4.7906,
+      "step": 56540
+    },
+    {
+      "epoch": 0.07945805531503082,
+      "grad_norm": 0.8445591926574707,
+      "learning_rate": 0.0002382843894899536,
+      "loss": 4.8183,
+      "step": 56550
+    },
+    {
+      "epoch": 0.0794721062531944,
+      "grad_norm": 0.8689804673194885,
+      "learning_rate": 0.00023832654208233802,
+      "loss": 4.9633,
+      "step": 56560
+    },
+    {
+      "epoch": 0.07948615719135797,
+      "grad_norm": 0.8556300401687622,
+      "learning_rate": 0.00023836869467472248,
+      "loss": 4.8468,
+      "step": 56570
+    },
+    {
+      "epoch": 0.07950020812952155,
+      "grad_norm": 0.9189597964286804,
+      "learning_rate": 0.0002384108472671069,
+      "loss": 4.7865,
+      "step": 56580
+    },
+    {
+      "epoch": 0.07951425906768513,
+      "grad_norm": 0.8929227590560913,
+      "learning_rate": 0.00023845299985949134,
+      "loss": 4.9381,
+      "step": 56590
+    },
+    {
+      "epoch": 0.0795283100058487,
+      "grad_norm": 0.8549538254737854,
+      "learning_rate": 0.00023849515245187578,
+      "loss": 4.8449,
+      "step": 56600
+    },
+    {
+      "epoch": 0.07954236094401228,
+      "grad_norm": 0.8399584293365479,
+      "learning_rate": 0.00023853730504426018,
+      "loss": 4.8057,
+      "step": 56610
+    },
+    {
+      "epoch": 0.07955641188217585,
+      "grad_norm": 0.8562041521072388,
+      "learning_rate": 0.0002385794576366446,
+      "loss": 4.7622,
+      "step": 56620
+    },
+    {
+      "epoch": 0.07957046282033943,
+      "grad_norm": 0.8549632430076599,
+      "learning_rate": 0.00023862161022902907,
+      "loss": 4.7727,
+      "step": 56630
+    },
+    {
+      "epoch": 0.07958451375850302,
+      "grad_norm": 0.8222755789756775,
+      "learning_rate": 0.0002386637628214135,
+      "loss": 4.8657,
+      "step": 56640
+    },
+    {
+      "epoch": 0.0795985646966666,
+      "grad_norm": 0.8658560514450073,
+      "learning_rate": 0.00023870591541379793,
+      "loss": 4.8183,
+      "step": 56650
+    },
+    {
+      "epoch": 0.07961261563483017,
+      "grad_norm": 0.8258546590805054,
+      "learning_rate": 0.00023874806800618236,
+      "loss": 4.8242,
+      "step": 56660
+    },
+    {
+      "epoch": 0.07962666657299375,
+      "grad_norm": 0.8417488932609558,
+      "learning_rate": 0.0002387902205985668,
+      "loss": 4.7378,
+      "step": 56670
+    },
+    {
+      "epoch": 0.07964071751115732,
+      "grad_norm": 0.8733723759651184,
+      "learning_rate": 0.0002388323731909512,
+      "loss": 4.7247,
+      "step": 56680
+    },
+    {
+      "epoch": 0.0796547684493209,
+      "grad_norm": 0.8495579361915588,
+      "learning_rate": 0.00023887452578333566,
+      "loss": 4.8123,
+      "step": 56690
+    },
+    {
+      "epoch": 0.07966881938748448,
+      "grad_norm": 0.8252209424972534,
+      "learning_rate": 0.0002389166783757201,
+      "loss": 4.791,
+      "step": 56700
+    },
+    {
+      "epoch": 0.07968287032564805,
+      "grad_norm": 0.8873644471168518,
+      "learning_rate": 0.00023895883096810452,
+      "loss": 4.7452,
+      "step": 56710
+    },
+    {
+      "epoch": 0.07969692126381163,
+      "grad_norm": 0.8369097113609314,
+      "learning_rate": 0.00023900098356048895,
+      "loss": 4.6592,
+      "step": 56720
+    },
+    {
+      "epoch": 0.0797109722019752,
+      "grad_norm": 0.9035923480987549,
+      "learning_rate": 0.00023904313615287338,
+      "loss": 4.646,
+      "step": 56730
+    },
+    {
+      "epoch": 0.07972502314013878,
+      "grad_norm": 0.87725430727005,
+      "learning_rate": 0.00023908528874525778,
+      "loss": 4.6869,
+      "step": 56740
+    },
+    {
+      "epoch": 0.07973907407830237,
+      "grad_norm": 0.8985244631767273,
+      "learning_rate": 0.00023912744133764224,
+      "loss": 4.7142,
+      "step": 56750
+    },
+    {
+      "epoch": 0.07975312501646595,
+      "grad_norm": 0.8276535868644714,
+      "learning_rate": 0.00023916959393002667,
+      "loss": 4.8682,
+      "step": 56760
+    },
+    {
+      "epoch": 0.07976717595462952,
+      "grad_norm": 0.8182491660118103,
+      "learning_rate": 0.0002392117465224111,
+      "loss": 4.9084,
+      "step": 56770
+    },
+    {
+      "epoch": 0.0797812268927931,
+      "grad_norm": 0.8240747451782227,
+      "learning_rate": 0.00023925389911479554,
+      "loss": 4.957,
+      "step": 56780
+    },
+    {
+      "epoch": 0.07979527783095668,
+      "grad_norm": 0.8549389243125916,
+      "learning_rate": 0.00023929605170717997,
+      "loss": 4.8275,
+      "step": 56790
+    },
+    {
+      "epoch": 0.07980932876912025,
+      "grad_norm": 0.8908678293228149,
+      "learning_rate": 0.00023933820429956443,
+      "loss": 4.9102,
+      "step": 56800
+    },
+    {
+      "epoch": 0.07982337970728383,
+      "grad_norm": 0.8527509570121765,
+      "learning_rate": 0.00023938035689194883,
+      "loss": 4.9285,
+      "step": 56810
+    },
+    {
+      "epoch": 0.0798374306454474,
+      "grad_norm": 0.8254045248031616,
+      "learning_rate": 0.00023942250948433326,
+      "loss": 4.8112,
+      "step": 56820
+    },
+    {
+      "epoch": 0.07985148158361098,
+      "grad_norm": 0.850629448890686,
+      "learning_rate": 0.0002394646620767177,
+      "loss": 4.8258,
+      "step": 56830
+    },
+    {
+      "epoch": 0.07986553252177456,
+      "grad_norm": 0.8825765252113342,
+      "learning_rate": 0.00023950681466910212,
+      "loss": 4.8338,
+      "step": 56840
+    },
+    {
+      "epoch": 0.07987958345993815,
+      "grad_norm": 0.8650710582733154,
+      "learning_rate": 0.00023954896726148658,
+      "loss": 4.8267,
+      "step": 56850
+    },
+    {
+      "epoch": 0.07989363439810172,
+      "grad_norm": 0.8440834283828735,
+      "learning_rate": 0.000239591119853871,
+      "loss": 4.9396,
+      "step": 56860
+    },
+    {
+      "epoch": 0.0799076853362653,
+      "grad_norm": 0.8128529191017151,
+      "learning_rate": 0.00023963327244625542,
+      "loss": 4.9086,
+      "step": 56870
+    },
+    {
+      "epoch": 0.07992173627442888,
+      "grad_norm": 0.9077975749969482,
+      "learning_rate": 0.00023967542503863985,
+      "loss": 4.7803,
+      "step": 56880
+    },
+    {
+      "epoch": 0.07993578721259245,
+      "grad_norm": 0.9005282521247864,
+      "learning_rate": 0.00023971757763102428,
+      "loss": 4.8147,
+      "step": 56890
+    },
+    {
+      "epoch": 0.07994983815075603,
+      "grad_norm": 0.8272696137428284,
+      "learning_rate": 0.0002397597302234087,
+      "loss": 4.7913,
+      "step": 56900
+    },
+    {
+      "epoch": 0.0799638890889196,
+      "grad_norm": 0.8787935972213745,
+      "learning_rate": 0.00023980188281579317,
+      "loss": 4.817,
+      "step": 56910
+    },
+    {
+      "epoch": 0.07997794002708318,
+      "grad_norm": 0.8999579548835754,
+      "learning_rate": 0.0002398440354081776,
+      "loss": 4.8105,
+      "step": 56920
+    },
+    {
+      "epoch": 0.07999199096524676,
+      "grad_norm": 0.8166282773017883,
+      "learning_rate": 0.00023988618800056203,
+      "loss": 4.8415,
+      "step": 56930
+    },
+    {
+      "epoch": 0.08000604190341033,
+      "grad_norm": 0.8887068033218384,
+      "learning_rate": 0.00023992834059294644,
+      "loss": 4.8513,
+      "step": 56940
+    },
+    {
+      "epoch": 0.08002009284157392,
+      "grad_norm": 0.8588322997093201,
+      "learning_rate": 0.00023997049318533087,
+      "loss": 4.7216,
+      "step": 56950
+    },
+    {
+      "epoch": 0.0800341437797375,
+      "grad_norm": 0.8577817678451538,
+      "learning_rate": 0.0002400126457777153,
+      "loss": 4.8628,
+      "step": 56960
+    },
+    {
+      "epoch": 0.08004819471790107,
+      "grad_norm": 0.8334136605262756,
+      "learning_rate": 0.00024005479837009976,
+      "loss": 4.8069,
+      "step": 56970
+    },
+    {
+      "epoch": 0.08006224565606465,
+      "grad_norm": 0.8446940779685974,
+      "learning_rate": 0.0002400969509624842,
+      "loss": 4.771,
+      "step": 56980
+    },
+    {
+      "epoch": 0.08007629659422823,
+      "grad_norm": 0.817052960395813,
+      "learning_rate": 0.00024013910355486862,
+      "loss": 4.8963,
+      "step": 56990
+    },
+    {
+      "epoch": 0.0800903475323918,
+      "grad_norm": 0.8921146392822266,
+      "learning_rate": 0.00024018125614725302,
+      "loss": 4.8283,
+      "step": 57000
+    },
+    {
+      "epoch": 0.08010439847055538,
+      "grad_norm": 0.8406715989112854,
+      "learning_rate": 0.00024022340873963745,
+      "loss": 4.759,
+      "step": 57010
+    },
+    {
+      "epoch": 0.08011844940871896,
+      "grad_norm": 0.8852546811103821,
+      "learning_rate": 0.00024026556133202188,
+      "loss": 4.8704,
+      "step": 57020
+    },
+    {
+      "epoch": 0.08013250034688253,
+      "grad_norm": 0.8223075866699219,
+      "learning_rate": 0.00024030771392440634,
+      "loss": 4.7321,
+      "step": 57030
+    },
+    {
+      "epoch": 0.08014655128504611,
+      "grad_norm": 0.8500157594680786,
+      "learning_rate": 0.00024034986651679077,
+      "loss": 4.9298,
+      "step": 57040
+    },
+    {
+      "epoch": 0.08016060222320968,
+      "grad_norm": 0.8550832867622375,
+      "learning_rate": 0.0002403920191091752,
+      "loss": 4.8613,
+      "step": 57050
+    },
+    {
+      "epoch": 0.08017465316137327,
+      "grad_norm": 0.9191502928733826,
+      "learning_rate": 0.00024043417170155964,
+      "loss": 4.6645,
+      "step": 57060
+    },
+    {
+      "epoch": 0.08018870409953685,
+      "grad_norm": 0.8604241013526917,
+      "learning_rate": 0.00024047632429394404,
+      "loss": 4.7982,
+      "step": 57070
+    },
+    {
+      "epoch": 0.08020275503770043,
+      "grad_norm": 0.8823716640472412,
+      "learning_rate": 0.00024051847688632847,
+      "loss": 4.7106,
+      "step": 57080
+    },
+    {
+      "epoch": 0.080216805975864,
+      "grad_norm": 0.8595419526100159,
+      "learning_rate": 0.00024056062947871293,
+      "loss": 4.835,
+      "step": 57090
+    },
+    {
+      "epoch": 0.08023085691402758,
+      "grad_norm": 0.8353159427642822,
+      "learning_rate": 0.00024060278207109736,
+      "loss": 4.7358,
+      "step": 57100
+    },
+    {
+      "epoch": 0.08024490785219116,
+      "grad_norm": 0.8417918682098389,
+      "learning_rate": 0.0002406449346634818,
+      "loss": 4.9533,
+      "step": 57110
+    },
+    {
+      "epoch": 0.08025895879035473,
+      "grad_norm": 0.8293863534927368,
+      "learning_rate": 0.00024068708725586622,
+      "loss": 4.9028,
+      "step": 57120
+    },
+    {
+      "epoch": 0.08027300972851831,
+      "grad_norm": 0.8590785264968872,
+      "learning_rate": 0.00024072923984825063,
+      "loss": 4.8322,
+      "step": 57130
+    },
+    {
+      "epoch": 0.08028706066668188,
+      "grad_norm": 0.8248816728591919,
+      "learning_rate": 0.00024077139244063506,
+      "loss": 4.7921,
+      "step": 57140
+    },
+    {
+      "epoch": 0.08030111160484546,
+      "grad_norm": 0.8333460688591003,
+      "learning_rate": 0.00024081354503301952,
+      "loss": 4.9145,
+      "step": 57150
+    },
+    {
+      "epoch": 0.08031516254300905,
+      "grad_norm": 0.8591793179512024,
+      "learning_rate": 0.00024085569762540395,
+      "loss": 4.8857,
+      "step": 57160
+    },
+    {
+      "epoch": 0.08032921348117263,
+      "grad_norm": 0.8233816623687744,
+      "learning_rate": 0.00024089785021778838,
+      "loss": 4.8737,
+      "step": 57170
+    },
+    {
+      "epoch": 0.0803432644193362,
+      "grad_norm": 0.8355622887611389,
+      "learning_rate": 0.0002409400028101728,
+      "loss": 4.7534,
+      "step": 57180
+    },
+    {
+      "epoch": 0.08035731535749978,
+      "grad_norm": 0.8862990140914917,
+      "learning_rate": 0.00024098215540255721,
+      "loss": 4.952,
+      "step": 57190
+    },
+    {
+      "epoch": 0.08037136629566335,
+      "grad_norm": 0.8540744781494141,
+      "learning_rate": 0.00024102430799494165,
+      "loss": 4.8611,
+      "step": 57200
+    },
+    {
+      "epoch": 0.08038541723382693,
+      "grad_norm": 0.8337244987487793,
+      "learning_rate": 0.0002410664605873261,
+      "loss": 4.8597,
+      "step": 57210
+    },
+    {
+      "epoch": 0.0803994681719905,
+      "grad_norm": 0.859093189239502,
+      "learning_rate": 0.00024110861317971053,
+      "loss": 4.7506,
+      "step": 57220
+    },
+    {
+      "epoch": 0.08041351911015408,
+      "grad_norm": 0.8373620510101318,
+      "learning_rate": 0.00024115076577209497,
+      "loss": 4.7512,
+      "step": 57230
+    },
+    {
+      "epoch": 0.08042757004831766,
+      "grad_norm": 0.9458513855934143,
+      "learning_rate": 0.0002411929183644794,
+      "loss": 4.849,
+      "step": 57240
+    },
+    {
+      "epoch": 0.08044162098648124,
+      "grad_norm": 0.830441415309906,
+      "learning_rate": 0.00024123507095686383,
+      "loss": 4.8426,
+      "step": 57250
+    },
+    {
+      "epoch": 0.08045567192464483,
+      "grad_norm": 0.8454274535179138,
+      "learning_rate": 0.00024127722354924823,
+      "loss": 4.931,
+      "step": 57260
+    },
+    {
+      "epoch": 0.0804697228628084,
+      "grad_norm": 0.8949636816978455,
+      "learning_rate": 0.0002413193761416327,
+      "loss": 4.8015,
+      "step": 57270
+    },
+    {
+      "epoch": 0.08048377380097198,
+      "grad_norm": 0.8301366567611694,
+      "learning_rate": 0.00024136152873401712,
+      "loss": 4.7237,
+      "step": 57280
+    },
+    {
+      "epoch": 0.08049782473913555,
+      "grad_norm": 0.8346551060676575,
+      "learning_rate": 0.00024140368132640155,
+      "loss": 4.8849,
+      "step": 57290
+    },
+    {
+      "epoch": 0.08051187567729913,
+      "grad_norm": 0.8517528772354126,
+      "learning_rate": 0.00024144583391878598,
+      "loss": 4.9193,
+      "step": 57300
+    },
+    {
+      "epoch": 0.0805259266154627,
+      "grad_norm": 0.8681390285491943,
+      "learning_rate": 0.00024148798651117042,
+      "loss": 4.8459,
+      "step": 57310
+    },
+    {
+      "epoch": 0.08053997755362628,
+      "grad_norm": 0.8334530591964722,
+      "learning_rate": 0.00024153013910355482,
+      "loss": 4.7559,
+      "step": 57320
+    },
+    {
+      "epoch": 0.08055402849178986,
+      "grad_norm": 0.844562828540802,
+      "learning_rate": 0.00024157229169593928,
+      "loss": 4.7291,
+      "step": 57330
+    },
+    {
+      "epoch": 0.08056807942995343,
+      "grad_norm": 0.8150333166122437,
+      "learning_rate": 0.0002416144442883237,
+      "loss": 4.8139,
+      "step": 57340
+    },
+    {
+      "epoch": 0.08058213036811701,
+      "grad_norm": 0.8444922566413879,
+      "learning_rate": 0.00024165659688070814,
+      "loss": 4.8298,
+      "step": 57350
+    },
+    {
+      "epoch": 0.08059618130628059,
+      "grad_norm": 0.832202672958374,
+      "learning_rate": 0.00024169874947309257,
+      "loss": 4.6788,
+      "step": 57360
+    },
+    {
+      "epoch": 0.08061023224444418,
+      "grad_norm": 0.819246232509613,
+      "learning_rate": 0.000241740902065477,
+      "loss": 4.8065,
+      "step": 57370
+    },
+    {
+      "epoch": 0.08062428318260775,
+      "grad_norm": 0.8348022103309631,
+      "learning_rate": 0.00024178305465786146,
+      "loss": 4.8509,
+      "step": 57380
+    },
+    {
+      "epoch": 0.08063833412077133,
+      "grad_norm": 0.85079425573349,
+      "learning_rate": 0.00024182520725024586,
+      "loss": 4.7028,
+      "step": 57390
+    },
+    {
+      "epoch": 0.0806523850589349,
+      "grad_norm": 0.8373984098434448,
+      "learning_rate": 0.0002418673598426303,
+      "loss": 4.7505,
+      "step": 57400
+    },
+    {
+      "epoch": 0.08066643599709848,
+      "grad_norm": 0.8485515713691711,
+      "learning_rate": 0.00024190951243501473,
+      "loss": 4.8502,
+      "step": 57410
+    },
+    {
+      "epoch": 0.08068048693526206,
+      "grad_norm": 0.8618393540382385,
+      "learning_rate": 0.00024195166502739916,
+      "loss": 4.8078,
+      "step": 57420
+    },
+    {
+      "epoch": 0.08069453787342563,
+      "grad_norm": 0.8273009657859802,
+      "learning_rate": 0.00024199381761978362,
+      "loss": 4.8212,
+      "step": 57430
+    },
+    {
+      "epoch": 0.08070858881158921,
+      "grad_norm": 0.8302605748176575,
+      "learning_rate": 0.00024203597021216805,
+      "loss": 4.6817,
+      "step": 57440
+    },
+    {
+      "epoch": 0.08072263974975279,
+      "grad_norm": 0.8489684462547302,
+      "learning_rate": 0.00024207812280455245,
+      "loss": 4.79,
+      "step": 57450
+    },
+    {
+      "epoch": 0.08073669068791636,
+      "grad_norm": 0.842799186706543,
+      "learning_rate": 0.00024212027539693688,
+      "loss": 4.8192,
+      "step": 57460
+    },
+    {
+      "epoch": 0.08075074162607995,
+      "grad_norm": 0.8465150594711304,
+      "learning_rate": 0.00024216242798932131,
+      "loss": 4.8832,
+      "step": 57470
+    },
+    {
+      "epoch": 0.08076479256424353,
+      "grad_norm": 0.8409611582756042,
+      "learning_rate": 0.00024220458058170575,
+      "loss": 4.7202,
+      "step": 57480
+    },
+    {
+      "epoch": 0.0807788435024071,
+      "grad_norm": 0.8487250804901123,
+      "learning_rate": 0.0002422467331740902,
+      "loss": 4.748,
+      "step": 57490
+    },
+    {
+      "epoch": 0.08079289444057068,
+      "grad_norm": 0.845812976360321,
+      "learning_rate": 0.00024228888576647463,
+      "loss": 4.7638,
+      "step": 57500
+    },
+    {
+      "epoch": 0.08080694537873426,
+      "grad_norm": 0.8647302985191345,
+      "learning_rate": 0.00024233103835885907,
+      "loss": 4.8452,
+      "step": 57510
+    },
+    {
+      "epoch": 0.08082099631689783,
+      "grad_norm": 0.8373607993125916,
+      "learning_rate": 0.00024237319095124347,
+      "loss": 4.8014,
+      "step": 57520
+    },
+    {
+      "epoch": 0.08083504725506141,
+      "grad_norm": 0.8152607679367065,
+      "learning_rate": 0.0002424153435436279,
+      "loss": 4.8544,
+      "step": 57530
+    },
+    {
+      "epoch": 0.08084909819322499,
+      "grad_norm": 0.8385706543922424,
+      "learning_rate": 0.00024245749613601233,
+      "loss": 4.8766,
+      "step": 57540
+    },
+    {
+      "epoch": 0.08086314913138856,
+      "grad_norm": 0.8651167154312134,
+      "learning_rate": 0.0002424996487283968,
+      "loss": 4.8027,
+      "step": 57550
+    },
+    {
+      "epoch": 0.08087720006955214,
+      "grad_norm": 0.837224543094635,
+      "learning_rate": 0.00024254180132078122,
+      "loss": 4.7422,
+      "step": 57560
+    },
+    {
+      "epoch": 0.08089125100771573,
+      "grad_norm": 0.8273807168006897,
+      "learning_rate": 0.00024258395391316565,
+      "loss": 4.7841,
+      "step": 57570
+    },
+    {
+      "epoch": 0.0809053019458793,
+      "grad_norm": 0.8563796281814575,
+      "learning_rate": 0.00024262610650555006,
+      "loss": 4.825,
+      "step": 57580
+    },
+    {
+      "epoch": 0.08091935288404288,
+      "grad_norm": 0.8224353790283203,
+      "learning_rate": 0.0002426682590979345,
+      "loss": 4.8982,
+      "step": 57590
+    },
+    {
+      "epoch": 0.08093340382220646,
+      "grad_norm": 0.8756199479103088,
+      "learning_rate": 0.00024271041169031892,
+      "loss": 4.7504,
+      "step": 57600
+    },
+    {
+      "epoch": 0.08094745476037003,
+      "grad_norm": 0.8563529253005981,
+      "learning_rate": 0.00024275256428270338,
+      "loss": 4.8682,
+      "step": 57610
+    },
+    {
+      "epoch": 0.08096150569853361,
+      "grad_norm": 0.8547507524490356,
+      "learning_rate": 0.0002427947168750878,
+      "loss": 4.8426,
+      "step": 57620
+    },
+    {
+      "epoch": 0.08097555663669719,
+      "grad_norm": 0.8033429980278015,
+      "learning_rate": 0.00024283686946747224,
+      "loss": 4.8511,
+      "step": 57630
+    },
+    {
+      "epoch": 0.08098960757486076,
+      "grad_norm": 0.8401547074317932,
+      "learning_rate": 0.00024287902205985667,
+      "loss": 4.7927,
+      "step": 57640
+    },
+    {
+      "epoch": 0.08100365851302434,
+      "grad_norm": 0.8384739756584167,
+      "learning_rate": 0.00024292117465224108,
+      "loss": 4.8009,
+      "step": 57650
+    },
+    {
+      "epoch": 0.08101770945118791,
+      "grad_norm": 0.840840220451355,
+      "learning_rate": 0.0002429633272446255,
+      "loss": 4.7733,
+      "step": 57660
+    },
+    {
+      "epoch": 0.08103176038935149,
+      "grad_norm": 0.8522785902023315,
+      "learning_rate": 0.00024300547983700996,
+      "loss": 4.7992,
+      "step": 57670
+    },
+    {
+      "epoch": 0.08104581132751508,
+      "grad_norm": 0.8566286563873291,
+      "learning_rate": 0.0002430476324293944,
+      "loss": 4.8891,
+      "step": 57680
+    },
+    {
+      "epoch": 0.08105986226567866,
+      "grad_norm": 0.8864015340805054,
+      "learning_rate": 0.00024308978502177883,
+      "loss": 4.8248,
+      "step": 57690
+    },
+    {
+      "epoch": 0.08107391320384223,
+      "grad_norm": 0.8461414575576782,
+      "learning_rate": 0.00024313193761416326,
+      "loss": 4.8703,
+      "step": 57700
+    },
+    {
+      "epoch": 0.08108796414200581,
+      "grad_norm": 0.8322480916976929,
+      "learning_rate": 0.00024317409020654766,
+      "loss": 4.872,
+      "step": 57710
+    },
+    {
+      "epoch": 0.08110201508016938,
+      "grad_norm": 0.8064298033714294,
+      "learning_rate": 0.0002432162427989321,
+      "loss": 4.9469,
+      "step": 57720
+    },
+    {
+      "epoch": 0.08111606601833296,
+      "grad_norm": 0.8545445799827576,
+      "learning_rate": 0.00024325839539131655,
+      "loss": 4.8237,
+      "step": 57730
+    },
+    {
+      "epoch": 0.08113011695649654,
+      "grad_norm": 0.8561162948608398,
+      "learning_rate": 0.00024330054798370098,
+      "loss": 4.8889,
+      "step": 57740
+    },
+    {
+      "epoch": 0.08114416789466011,
+      "grad_norm": 0.862308919429779,
+      "learning_rate": 0.00024334270057608541,
+      "loss": 4.6817,
+      "step": 57750
+    },
+    {
+      "epoch": 0.08115821883282369,
+      "grad_norm": 0.832358717918396,
+      "learning_rate": 0.00024338485316846985,
+      "loss": 4.819,
+      "step": 57760
+    },
+    {
+      "epoch": 0.08117226977098727,
+      "grad_norm": 0.8506022095680237,
+      "learning_rate": 0.00024342700576085428,
+      "loss": 4.7971,
+      "step": 57770
+    },
+    {
+      "epoch": 0.08118632070915086,
+      "grad_norm": 0.8205908536911011,
+      "learning_rate": 0.00024346915835323868,
+      "loss": 4.7174,
+      "step": 57780
+    },
+    {
+      "epoch": 0.08120037164731443,
+      "grad_norm": 0.843183159828186,
+      "learning_rate": 0.00024351131094562314,
+      "loss": 4.8403,
+      "step": 57790
+    },
+    {
+      "epoch": 0.08121442258547801,
+      "grad_norm": 0.8414803147315979,
+      "learning_rate": 0.00024355346353800757,
+      "loss": 4.7169,
+      "step": 57800
+    },
+    {
+      "epoch": 0.08122847352364158,
+      "grad_norm": 0.8241091370582581,
+      "learning_rate": 0.000243595616130392,
+      "loss": 4.7469,
+      "step": 57810
+    },
+    {
+      "epoch": 0.08124252446180516,
+      "grad_norm": 0.8284342288970947,
+      "learning_rate": 0.00024363776872277643,
+      "loss": 4.8104,
+      "step": 57820
+    },
+    {
+      "epoch": 0.08125657539996874,
+      "grad_norm": 0.8515377044677734,
+      "learning_rate": 0.00024367992131516086,
+      "loss": 4.7973,
+      "step": 57830
+    },
+    {
+      "epoch": 0.08127062633813231,
+      "grad_norm": 0.8532301187515259,
+      "learning_rate": 0.00024372207390754527,
+      "loss": 4.7666,
+      "step": 57840
+    },
+    {
+      "epoch": 0.08128467727629589,
+      "grad_norm": 0.8152916431427002,
+      "learning_rate": 0.00024376422649992973,
+      "loss": 4.797,
+      "step": 57850
+    },
+    {
+      "epoch": 0.08129872821445946,
+      "grad_norm": 0.834984540939331,
+      "learning_rate": 0.00024380637909231416,
+      "loss": 4.7428,
+      "step": 57860
+    },
+    {
+      "epoch": 0.08131277915262304,
+      "grad_norm": 0.8661616444587708,
+      "learning_rate": 0.0002438485316846986,
+      "loss": 4.8754,
+      "step": 57870
+    },
+    {
+      "epoch": 0.08132683009078663,
+      "grad_norm": 0.843447744846344,
+      "learning_rate": 0.00024389068427708302,
+      "loss": 5.0731,
+      "step": 57880
+    },
+    {
+      "epoch": 0.08134088102895021,
+      "grad_norm": 0.8291600346565247,
+      "learning_rate": 0.00024393283686946745,
+      "loss": 4.8219,
+      "step": 57890
+    },
+    {
+      "epoch": 0.08135493196711378,
+      "grad_norm": 0.8383052349090576,
+      "learning_rate": 0.00024397077420261343,
+      "loss": 4.7993,
+      "step": 57900
+    },
+    {
+      "epoch": 0.08136898290527736,
+      "grad_norm": 0.8022859692573547,
+      "learning_rate": 0.00024401292679499786,
+      "loss": 4.8967,
+      "step": 57910
+    },
+    {
+      "epoch": 0.08138303384344094,
+      "grad_norm": 0.8225985169410706,
+      "learning_rate": 0.00024405507938738232,
+      "loss": 4.7647,
+      "step": 57920
+    },
+    {
+      "epoch": 0.08139708478160451,
+      "grad_norm": 0.8517218828201294,
+      "learning_rate": 0.00024409723197976675,
+      "loss": 4.7826,
+      "step": 57930
+    },
+    {
+      "epoch": 0.08141113571976809,
+      "grad_norm": 0.8240352869033813,
+      "learning_rate": 0.00024413938457215116,
+      "loss": 4.7381,
+      "step": 57940
+    },
+    {
+      "epoch": 0.08142518665793166,
+      "grad_norm": 0.8103241324424744,
+      "learning_rate": 0.0002441815371645356,
+      "loss": 4.7835,
+      "step": 57950
+    },
+    {
+      "epoch": 0.08143923759609524,
+      "grad_norm": 0.8353750705718994,
+      "learning_rate": 0.00024422368975692005,
+      "loss": 4.8407,
+      "step": 57960
+    },
+    {
+      "epoch": 0.08145328853425882,
+      "grad_norm": 0.8388403058052063,
+      "learning_rate": 0.0002442658423493045,
+      "loss": 4.8119,
+      "step": 57970
+    },
+    {
+      "epoch": 0.08146733947242239,
+      "grad_norm": 0.8061336278915405,
+      "learning_rate": 0.0002443079949416889,
+      "loss": 4.7948,
+      "step": 57980
+    },
+    {
+      "epoch": 0.08148139041058598,
+      "grad_norm": 0.9352821707725525,
+      "learning_rate": 0.00024435014753407334,
+      "loss": 4.8472,
+      "step": 57990
+    },
+    {
+      "epoch": 0.08149544134874956,
+      "grad_norm": 0.8587878346443176,
+      "learning_rate": 0.0002443923001264577,
+      "loss": 4.8216,
+      "step": 58000
+    },
+    {
+      "epoch": 0.08150949228691313,
+      "grad_norm": 0.8197136521339417,
+      "learning_rate": 0.0002444344527188422,
+      "loss": 4.7632,
+      "step": 58010
+    },
+    {
+      "epoch": 0.08152354322507671,
+      "grad_norm": 0.8324862718582153,
+      "learning_rate": 0.00024447660531122663,
+      "loss": 4.8624,
+      "step": 58020
+    },
+    {
+      "epoch": 0.08153759416324029,
+      "grad_norm": 0.8072025179862976,
+      "learning_rate": 0.00024451875790361107,
+      "loss": 4.8684,
+      "step": 58030
+    },
+    {
+      "epoch": 0.08155164510140386,
+      "grad_norm": 0.8620959520339966,
+      "learning_rate": 0.0002445609104959955,
+      "loss": 4.8949,
+      "step": 58040
+    },
+    {
+      "epoch": 0.08156569603956744,
+      "grad_norm": 0.8675084114074707,
+      "learning_rate": 0.00024460306308837993,
+      "loss": 4.7576,
+      "step": 58050
+    },
+    {
+      "epoch": 0.08157974697773102,
+      "grad_norm": 0.8540964126586914,
+      "learning_rate": 0.00024464521568076436,
+      "loss": 4.7649,
+      "step": 58060
+    },
+    {
+      "epoch": 0.08159379791589459,
+      "grad_norm": 0.8450266122817993,
+      "learning_rate": 0.0002446873682731488,
+      "loss": 4.8923,
+      "step": 58070
+    },
+    {
+      "epoch": 0.08160784885405817,
+      "grad_norm": 0.8445958495140076,
+      "learning_rate": 0.0002447295208655332,
+      "loss": 4.7243,
+      "step": 58080
+    },
+    {
+      "epoch": 0.08162189979222176,
+      "grad_norm": 0.823774516582489,
+      "learning_rate": 0.00024477167345791765,
+      "loss": 4.7552,
+      "step": 58090
+    },
+    {
+      "epoch": 0.08163595073038533,
+      "grad_norm": 0.8373647928237915,
+      "learning_rate": 0.0002448138260503021,
+      "loss": 4.7746,
+      "step": 58100
+    },
+    {
+      "epoch": 0.08165000166854891,
+      "grad_norm": 0.8526290655136108,
+      "learning_rate": 0.0002448559786426865,
+      "loss": 4.7328,
+      "step": 58110
+    },
+    {
+      "epoch": 0.08166405260671249,
+      "grad_norm": 0.8044569492340088,
+      "learning_rate": 0.00024489813123507095,
+      "loss": 4.8296,
+      "step": 58120
+    },
+    {
+      "epoch": 0.08167810354487606,
+      "grad_norm": 0.8177915811538696,
+      "learning_rate": 0.0002449402838274554,
+      "loss": 4.8325,
+      "step": 58130
+    },
+    {
+      "epoch": 0.08169215448303964,
+      "grad_norm": 0.8297937512397766,
+      "learning_rate": 0.0002449824364198398,
+      "loss": 4.8649,
+      "step": 58140
+    },
+    {
+      "epoch": 0.08170620542120322,
+      "grad_norm": 0.8474819660186768,
+      "learning_rate": 0.00024502458901222424,
+      "loss": 4.7662,
+      "step": 58150
+    },
+    {
+      "epoch": 0.08172025635936679,
+      "grad_norm": 0.876619815826416,
+      "learning_rate": 0.00024506674160460867,
+      "loss": 4.8893,
+      "step": 58160
+    },
+    {
+      "epoch": 0.08173430729753037,
+      "grad_norm": 0.8410592675209045,
+      "learning_rate": 0.0002451088941969931,
+      "loss": 4.7708,
+      "step": 58170
+    },
+    {
+      "epoch": 0.08174835823569394,
+      "grad_norm": 0.8169021010398865,
+      "learning_rate": 0.00024515104678937753,
+      "loss": 4.892,
+      "step": 58180
+    },
+    {
+      "epoch": 0.08176240917385753,
+      "grad_norm": 0.8390880823135376,
+      "learning_rate": 0.00024519319938176196,
+      "loss": 4.8694,
+      "step": 58190
+    },
+    {
+      "epoch": 0.08177646011202111,
+      "grad_norm": 0.8502917885780334,
+      "learning_rate": 0.0002452353519741464,
+      "loss": 4.7237,
+      "step": 58200
+    },
+    {
+      "epoch": 0.08179051105018469,
+      "grad_norm": 0.8233117461204529,
+      "learning_rate": 0.0002452775045665308,
+      "loss": 4.7619,
+      "step": 58210
+    },
+    {
+      "epoch": 0.08180456198834826,
+      "grad_norm": 0.8351155519485474,
+      "learning_rate": 0.00024531965715891526,
+      "loss": 4.8656,
+      "step": 58220
+    },
+    {
+      "epoch": 0.08181861292651184,
+      "grad_norm": 0.8331232666969299,
+      "learning_rate": 0.0002453618097512997,
+      "loss": 4.7766,
+      "step": 58230
+    },
+    {
+      "epoch": 0.08183266386467541,
+      "grad_norm": 0.8282060623168945,
+      "learning_rate": 0.0002454039623436841,
+      "loss": 4.7699,
+      "step": 58240
+    },
+    {
+      "epoch": 0.08184671480283899,
+      "grad_norm": 0.8700007200241089,
+      "learning_rate": 0.00024544611493606855,
+      "loss": 4.7489,
+      "step": 58250
+    },
+    {
+      "epoch": 0.08186076574100257,
+      "grad_norm": 0.8399953842163086,
+      "learning_rate": 0.000245488267528453,
+      "loss": 4.7302,
+      "step": 58260
+    },
+    {
+      "epoch": 0.08187481667916614,
+      "grad_norm": 0.8028297424316406,
+      "learning_rate": 0.0002455304201208374,
+      "loss": 4.9267,
+      "step": 58270
+    },
+    {
+      "epoch": 0.08188886761732972,
+      "grad_norm": 0.8101634979248047,
+      "learning_rate": 0.00024557257271322184,
+      "loss": 4.8224,
+      "step": 58280
+    },
+    {
+      "epoch": 0.0819029185554933,
+      "grad_norm": 0.8356708884239197,
+      "learning_rate": 0.0002456147253056063,
+      "loss": 4.7565,
+      "step": 58290
+    },
+    {
+      "epoch": 0.08191696949365689,
+      "grad_norm": 0.8344118595123291,
+      "learning_rate": 0.0002456568778979907,
+      "loss": 4.812,
+      "step": 58300
+    },
+    {
+      "epoch": 0.08193102043182046,
+      "grad_norm": 0.8281126022338867,
+      "learning_rate": 0.00024569903049037514,
+      "loss": 4.8933,
+      "step": 58310
+    },
+    {
+      "epoch": 0.08194507136998404,
+      "grad_norm": 0.8315061926841736,
+      "learning_rate": 0.00024574118308275957,
+      "loss": 4.6364,
+      "step": 58320
+    },
+    {
+      "epoch": 0.08195912230814761,
+      "grad_norm": 0.8570355772972107,
+      "learning_rate": 0.000245783335675144,
+      "loss": 4.8018,
+      "step": 58330
+    },
+    {
+      "epoch": 0.08197317324631119,
+      "grad_norm": 0.8422298431396484,
+      "learning_rate": 0.00024582548826752843,
+      "loss": 4.7245,
+      "step": 58340
+    },
+    {
+      "epoch": 0.08198722418447477,
+      "grad_norm": 0.8204920887947083,
+      "learning_rate": 0.00024586764085991286,
+      "loss": 4.7966,
+      "step": 58350
+    },
+    {
+      "epoch": 0.08200127512263834,
+      "grad_norm": 0.8395097851753235,
+      "learning_rate": 0.0002459097934522973,
+      "loss": 4.7934,
+      "step": 58360
+    },
+    {
+      "epoch": 0.08201532606080192,
+      "grad_norm": 0.8092932105064392,
+      "learning_rate": 0.0002459519460446817,
+      "loss": 4.812,
+      "step": 58370
+    },
+    {
+      "epoch": 0.0820293769989655,
+      "grad_norm": 0.818016529083252,
+      "learning_rate": 0.00024599409863706616,
+      "loss": 4.8381,
+      "step": 58380
+    },
+    {
+      "epoch": 0.08204342793712907,
+      "grad_norm": 0.866174042224884,
+      "learning_rate": 0.0002460362512294506,
+      "loss": 4.8195,
+      "step": 58390
+    },
+    {
+      "epoch": 0.08205747887529266,
+      "grad_norm": 0.8038418292999268,
+      "learning_rate": 0.000246078403821835,
+      "loss": 4.8522,
+      "step": 58400
+    },
+    {
+      "epoch": 0.08207152981345624,
+      "grad_norm": 0.8310818672180176,
+      "learning_rate": 0.00024612055641421945,
+      "loss": 4.8415,
+      "step": 58410
+    },
+    {
+      "epoch": 0.08208558075161981,
+      "grad_norm": 0.8277154564857483,
+      "learning_rate": 0.0002461627090066039,
+      "loss": 4.738,
+      "step": 58420
+    },
+    {
+      "epoch": 0.08209963168978339,
+      "grad_norm": 0.8364769220352173,
+      "learning_rate": 0.0002462048615989883,
+      "loss": 4.8246,
+      "step": 58430
+    },
+    {
+      "epoch": 0.08211368262794697,
+      "grad_norm": 0.8868595361709595,
+      "learning_rate": 0.00024624701419137274,
+      "loss": 4.8204,
+      "step": 58440
+    },
+    {
+      "epoch": 0.08212773356611054,
+      "grad_norm": 0.8325658440589905,
+      "learning_rate": 0.0002462891667837572,
+      "loss": 4.7931,
+      "step": 58450
+    },
+    {
+      "epoch": 0.08214178450427412,
+      "grad_norm": 0.8351956605911255,
+      "learning_rate": 0.0002463313193761416,
+      "loss": 4.8991,
+      "step": 58460
+    },
+    {
+      "epoch": 0.0821558354424377,
+      "grad_norm": 0.8221379518508911,
+      "learning_rate": 0.00024637347196852604,
+      "loss": 4.867,
+      "step": 58470
+    },
+    {
+      "epoch": 0.08216988638060127,
+      "grad_norm": 0.827820897102356,
+      "learning_rate": 0.00024641562456091047,
+      "loss": 4.793,
+      "step": 58480
+    },
+    {
+      "epoch": 0.08218393731876485,
+      "grad_norm": 0.902683436870575,
+      "learning_rate": 0.0002464577771532949,
+      "loss": 4.7763,
+      "step": 58490
+    },
+    {
+      "epoch": 0.08219798825692844,
+      "grad_norm": 0.8596385717391968,
+      "learning_rate": 0.00024649992974567933,
+      "loss": 4.7973,
+      "step": 58500
+    },
+    {
+      "epoch": 0.08221203919509201,
+      "grad_norm": 0.8017410635948181,
+      "learning_rate": 0.00024654208233806376,
+      "loss": 4.8217,
+      "step": 58510
+    },
+    {
+      "epoch": 0.08222609013325559,
+      "grad_norm": 0.8284733891487122,
+      "learning_rate": 0.0002465842349304482,
+      "loss": 4.8301,
+      "step": 58520
+    },
+    {
+      "epoch": 0.08224014107141916,
+      "grad_norm": 0.8326720595359802,
+      "learning_rate": 0.0002466263875228326,
+      "loss": 4.7582,
+      "step": 58530
+    },
+    {
+      "epoch": 0.08225419200958274,
+      "grad_norm": 0.8211853504180908,
+      "learning_rate": 0.00024666854011521705,
+      "loss": 4.8276,
+      "step": 58540
+    },
+    {
+      "epoch": 0.08226824294774632,
+      "grad_norm": 0.8203056454658508,
+      "learning_rate": 0.0002467106927076015,
+      "loss": 4.8199,
+      "step": 58550
+    },
+    {
+      "epoch": 0.0822822938859099,
+      "grad_norm": 0.8489205837249756,
+      "learning_rate": 0.0002467528452999859,
+      "loss": 4.8442,
+      "step": 58560
+    },
+    {
+      "epoch": 0.08229634482407347,
+      "grad_norm": 0.8325367569923401,
+      "learning_rate": 0.00024679499789237035,
+      "loss": 4.8388,
+      "step": 58570
+    },
+    {
+      "epoch": 0.08231039576223705,
+      "grad_norm": 0.8521776795387268,
+      "learning_rate": 0.0002468371504847548,
+      "loss": 4.9109,
+      "step": 58580
+    },
+    {
+      "epoch": 0.08232444670040062,
+      "grad_norm": 0.8650569915771484,
+      "learning_rate": 0.0002468793030771392,
+      "loss": 4.8815,
+      "step": 58590
+    },
+    {
+      "epoch": 0.0823384976385642,
+      "grad_norm": 0.8365474343299866,
+      "learning_rate": 0.00024692145566952364,
+      "loss": 4.8285,
+      "step": 58600
+    },
+    {
+      "epoch": 0.08235254857672779,
+      "grad_norm": 0.8498002886772156,
+      "learning_rate": 0.00024696360826190807,
+      "loss": 4.8788,
+      "step": 58610
+    },
+    {
+      "epoch": 0.08236659951489136,
+      "grad_norm": 0.8518571257591248,
+      "learning_rate": 0.0002470057608542925,
+      "loss": 4.7712,
+      "step": 58620
+    },
+    {
+      "epoch": 0.08238065045305494,
+      "grad_norm": 0.8263545632362366,
+      "learning_rate": 0.00024704791344667694,
+      "loss": 4.9409,
+      "step": 58630
+    },
+    {
+      "epoch": 0.08239470139121852,
+      "grad_norm": 0.8218898177146912,
+      "learning_rate": 0.0002470900660390614,
+      "loss": 4.6849,
+      "step": 58640
+    },
+    {
+      "epoch": 0.08240875232938209,
+      "grad_norm": 0.8200307488441467,
+      "learning_rate": 0.0002471322186314458,
+      "loss": 4.7979,
+      "step": 58650
+    },
+    {
+      "epoch": 0.08242280326754567,
+      "grad_norm": 0.8574158549308777,
+      "learning_rate": 0.00024717437122383023,
+      "loss": 4.8704,
+      "step": 58660
+    },
+    {
+      "epoch": 0.08243685420570925,
+      "grad_norm": 0.8805510401725769,
+      "learning_rate": 0.00024721652381621466,
+      "loss": 4.8024,
+      "step": 58670
+    },
+    {
+      "epoch": 0.08245090514387282,
+      "grad_norm": 0.8428569436073303,
+      "learning_rate": 0.0002472586764085991,
+      "loss": 4.7948,
+      "step": 58680
+    },
+    {
+      "epoch": 0.0824649560820364,
+      "grad_norm": 0.8235595226287842,
+      "learning_rate": 0.0002473008290009835,
+      "loss": 4.709,
+      "step": 58690
+    },
+    {
+      "epoch": 0.08247900702019997,
+      "grad_norm": 0.8152335286140442,
+      "learning_rate": 0.000247342981593368,
+      "loss": 4.8758,
+      "step": 58700
+    },
+    {
+      "epoch": 0.08249305795836356,
+      "grad_norm": 0.8103764653205872,
+      "learning_rate": 0.0002473851341857524,
+      "loss": 4.7552,
+      "step": 58710
+    },
+    {
+      "epoch": 0.08250710889652714,
+      "grad_norm": 0.8165539503097534,
+      "learning_rate": 0.0002474272867781368,
+      "loss": 4.8179,
+      "step": 58720
+    },
+    {
+      "epoch": 0.08252115983469072,
+      "grad_norm": 0.851290225982666,
+      "learning_rate": 0.00024746943937052125,
+      "loss": 4.9018,
+      "step": 58730
+    },
+    {
+      "epoch": 0.08253521077285429,
+      "grad_norm": 0.8036884665489197,
+      "learning_rate": 0.0002475115919629057,
+      "loss": 4.8221,
+      "step": 58740
+    },
+    {
+      "epoch": 0.08254926171101787,
+      "grad_norm": 0.8429344296455383,
+      "learning_rate": 0.0002475537445552901,
+      "loss": 4.8122,
+      "step": 58750
+    },
+    {
+      "epoch": 0.08256331264918144,
+      "grad_norm": 0.8305610418319702,
+      "learning_rate": 0.0002475958971476746,
+      "loss": 4.7718,
+      "step": 58760
+    },
+    {
+      "epoch": 0.08257736358734502,
+      "grad_norm": 0.8201099038124084,
+      "learning_rate": 0.000247638049740059,
+      "loss": 4.9241,
+      "step": 58770
+    },
+    {
+      "epoch": 0.0825914145255086,
+      "grad_norm": 0.8135986328125,
+      "learning_rate": 0.0002476802023324434,
+      "loss": 4.8023,
+      "step": 58780
+    },
+    {
+      "epoch": 0.08260546546367217,
+      "grad_norm": 0.8189361095428467,
+      "learning_rate": 0.00024772235492482783,
+      "loss": 4.8384,
+      "step": 58790
+    },
+    {
+      "epoch": 0.08261951640183575,
+      "grad_norm": 0.8032187223434448,
+      "learning_rate": 0.00024776450751721226,
+      "loss": 4.8052,
+      "step": 58800
+    },
+    {
+      "epoch": 0.08263356733999934,
+      "grad_norm": 0.8336382508277893,
+      "learning_rate": 0.0002478066601095967,
+      "loss": 4.755,
+      "step": 58810
+    },
+    {
+      "epoch": 0.08264761827816292,
+      "grad_norm": 0.8608211278915405,
+      "learning_rate": 0.0002478488127019812,
+      "loss": 4.9017,
+      "step": 58820
+    },
+    {
+      "epoch": 0.08266166921632649,
+      "grad_norm": 0.8575011491775513,
+      "learning_rate": 0.0002478909652943656,
+      "loss": 4.7826,
+      "step": 58830
+    },
+    {
+      "epoch": 0.08267572015449007,
+      "grad_norm": 0.8528128266334534,
+      "learning_rate": 0.00024793311788675,
+      "loss": 4.8539,
+      "step": 58840
+    },
+    {
+      "epoch": 0.08268977109265364,
+      "grad_norm": 0.8594315052032471,
+      "learning_rate": 0.0002479752704791344,
+      "loss": 4.7169,
+      "step": 58850
+    },
+    {
+      "epoch": 0.08270382203081722,
+      "grad_norm": 0.8435611128807068,
+      "learning_rate": 0.00024801742307151885,
+      "loss": 4.7116,
+      "step": 58860
+    },
+    {
+      "epoch": 0.0827178729689808,
+      "grad_norm": 0.8331279158592224,
+      "learning_rate": 0.0002480595756639033,
+      "loss": 4.933,
+      "step": 58870
+    },
+    {
+      "epoch": 0.08273192390714437,
+      "grad_norm": 0.8534746766090393,
+      "learning_rate": 0.00024810172825628777,
+      "loss": 4.8714,
+      "step": 58880
+    },
+    {
+      "epoch": 0.08274597484530795,
+      "grad_norm": 0.8149212002754211,
+      "learning_rate": 0.0002481438808486722,
+      "loss": 4.917,
+      "step": 58890
+    },
+    {
+      "epoch": 0.08276002578347152,
+      "grad_norm": 0.8224017024040222,
+      "learning_rate": 0.00024818603344105663,
+      "loss": 4.7634,
+      "step": 58900
+    },
+    {
+      "epoch": 0.0827740767216351,
+      "grad_norm": 0.8118835687637329,
+      "learning_rate": 0.000248228186033441,
+      "loss": 4.782,
+      "step": 58910
+    },
+    {
+      "epoch": 0.08278812765979869,
+      "grad_norm": 0.8306914567947388,
+      "learning_rate": 0.00024827033862582544,
+      "loss": 4.8674,
+      "step": 58920
+    },
+    {
+      "epoch": 0.08280217859796227,
+      "grad_norm": 0.8186638355255127,
+      "learning_rate": 0.00024831249121820987,
+      "loss": 4.8705,
+      "step": 58930
+    },
+    {
+      "epoch": 0.08281622953612584,
+      "grad_norm": 0.8312268257141113,
+      "learning_rate": 0.00024835464381059436,
+      "loss": 4.8295,
+      "step": 58940
+    },
+    {
+      "epoch": 0.08283028047428942,
+      "grad_norm": 0.8370024561882019,
+      "learning_rate": 0.0002483967964029788,
+      "loss": 4.8434,
+      "step": 58950
+    },
+    {
+      "epoch": 0.082844331412453,
+      "grad_norm": 0.8207914233207703,
+      "learning_rate": 0.0002484389489953632,
+      "loss": 4.7712,
+      "step": 58960
+    },
+    {
+      "epoch": 0.08285838235061657,
+      "grad_norm": 0.8279985785484314,
+      "learning_rate": 0.0002484811015877476,
+      "loss": 4.8821,
+      "step": 58970
+    },
+    {
+      "epoch": 0.08287243328878015,
+      "grad_norm": 0.8251729011535645,
+      "learning_rate": 0.000248523254180132,
+      "loss": 4.6615,
+      "step": 58980
+    },
+    {
+      "epoch": 0.08288648422694372,
+      "grad_norm": 0.8180922269821167,
+      "learning_rate": 0.0002485654067725165,
+      "loss": 4.778,
+      "step": 58990
+    },
+    {
+      "epoch": 0.0829005351651073,
+      "grad_norm": 0.833755612373352,
+      "learning_rate": 0.00024860755936490094,
+      "loss": 4.8855,
+      "step": 59000
+    },
+    {
+      "epoch": 0.08291458610327088,
+      "grad_norm": 0.8239709734916687,
+      "learning_rate": 0.0002486497119572854,
+      "loss": 4.7605,
+      "step": 59010
+    },
+    {
+      "epoch": 0.08292863704143447,
+      "grad_norm": 0.8077985644340515,
+      "learning_rate": 0.0002486918645496698,
+      "loss": 4.8525,
+      "step": 59020
+    },
+    {
+      "epoch": 0.08294268797959804,
+      "grad_norm": 0.795622706413269,
+      "learning_rate": 0.0002487340171420542,
+      "loss": 4.7966,
+      "step": 59030
+    },
+    {
+      "epoch": 0.08295673891776162,
+      "grad_norm": 0.8494805693626404,
+      "learning_rate": 0.0002487761697344386,
+      "loss": 4.8679,
+      "step": 59040
+    },
+    {
+      "epoch": 0.0829707898559252,
+      "grad_norm": 0.8523160219192505,
+      "learning_rate": 0.0002488183223268231,
+      "loss": 4.824,
+      "step": 59050
+    },
+    {
+      "epoch": 0.08298484079408877,
+      "grad_norm": 0.8548047542572021,
+      "learning_rate": 0.00024886047491920753,
+      "loss": 4.7349,
+      "step": 59060
+    },
+    {
+      "epoch": 0.08299889173225235,
+      "grad_norm": 0.8346095681190491,
+      "learning_rate": 0.00024890262751159196,
+      "loss": 4.858,
+      "step": 59070
+    },
+    {
+      "epoch": 0.08301294267041592,
+      "grad_norm": 0.8644514679908752,
+      "learning_rate": 0.0002489447801039764,
+      "loss": 4.897,
+      "step": 59080
+    },
+    {
+      "epoch": 0.0830269936085795,
+      "grad_norm": 0.8482093811035156,
+      "learning_rate": 0.0002489869326963608,
+      "loss": 4.8537,
+      "step": 59090
+    },
+    {
+      "epoch": 0.08304104454674308,
+      "grad_norm": 0.8546515703201294,
+      "learning_rate": 0.0002490290852887452,
+      "loss": 4.7443,
+      "step": 59100
+    },
+    {
+      "epoch": 0.08305509548490665,
+      "grad_norm": 0.8156818151473999,
+      "learning_rate": 0.0002490712378811297,
+      "loss": 4.7512,
+      "step": 59110
+    },
+    {
+      "epoch": 0.08306914642307024,
+      "grad_norm": 0.8068788647651672,
+      "learning_rate": 0.0002491133904735141,
+      "loss": 4.8865,
+      "step": 59120
+    },
+    {
+      "epoch": 0.08308319736123382,
+      "grad_norm": 0.8569709658622742,
+      "learning_rate": 0.00024915554306589855,
+      "loss": 4.8543,
+      "step": 59130
+    },
+    {
+      "epoch": 0.0830972482993974,
+      "grad_norm": 0.8595927357673645,
+      "learning_rate": 0.000249197695658283,
+      "loss": 4.9016,
+      "step": 59140
+    },
+    {
+      "epoch": 0.08311129923756097,
+      "grad_norm": 0.8086974024772644,
+      "learning_rate": 0.0002492398482506674,
+      "loss": 4.8776,
+      "step": 59150
+    },
+    {
+      "epoch": 0.08312535017572455,
+      "grad_norm": 0.8013572096824646,
+      "learning_rate": 0.0002492820008430518,
+      "loss": 4.7825,
+      "step": 59160
+    },
+    {
+      "epoch": 0.08313940111388812,
+      "grad_norm": 0.8331977725028992,
+      "learning_rate": 0.00024932415343543627,
+      "loss": 4.8848,
+      "step": 59170
+    },
+    {
+      "epoch": 0.0831534520520517,
+      "grad_norm": 0.8577580451965332,
+      "learning_rate": 0.0002493663060278207,
+      "loss": 4.8004,
+      "step": 59180
+    },
+    {
+      "epoch": 0.08316750299021528,
+      "grad_norm": 0.8222034573554993,
+      "learning_rate": 0.00024940845862020513,
+      "loss": 4.7706,
+      "step": 59190
+    },
+    {
+      "epoch": 0.08318155392837885,
+      "grad_norm": 0.8116173148155212,
+      "learning_rate": 0.00024945061121258957,
+      "loss": 4.925,
+      "step": 59200
+    },
+    {
+      "epoch": 0.08319560486654243,
+      "grad_norm": 0.8104360699653625,
+      "learning_rate": 0.000249492763804974,
+      "loss": 4.8693,
+      "step": 59210
+    },
+    {
+      "epoch": 0.083209655804706,
+      "grad_norm": 0.8252411484718323,
+      "learning_rate": 0.00024953491639735843,
+      "loss": 4.7542,
+      "step": 59220
+    },
+    {
+      "epoch": 0.0832237067428696,
+      "grad_norm": 0.8130877614021301,
+      "learning_rate": 0.00024957706898974286,
+      "loss": 4.8668,
+      "step": 59230
+    },
+    {
+      "epoch": 0.08323775768103317,
+      "grad_norm": 0.8746052980422974,
+      "learning_rate": 0.0002496192215821273,
+      "loss": 4.732,
+      "step": 59240
+    },
+    {
+      "epoch": 0.08325180861919675,
+      "grad_norm": 0.8369506001472473,
+      "learning_rate": 0.0002496613741745117,
+      "loss": 4.6947,
+      "step": 59250
+    },
+    {
+      "epoch": 0.08326585955736032,
+      "grad_norm": 0.8340523838996887,
+      "learning_rate": 0.00024970352676689615,
+      "loss": 4.8168,
+      "step": 59260
+    },
+    {
+      "epoch": 0.0832799104955239,
+      "grad_norm": 0.8198563456535339,
+      "learning_rate": 0.0002497456793592806,
+      "loss": 4.8669,
+      "step": 59270
+    },
+    {
+      "epoch": 0.08329396143368747,
+      "grad_norm": 0.7938926219940186,
+      "learning_rate": 0.000249787831951665,
+      "loss": 4.9119,
+      "step": 59280
+    },
+    {
+      "epoch": 0.08330801237185105,
+      "grad_norm": 0.8195042610168457,
+      "learning_rate": 0.00024982998454404945,
+      "loss": 4.7753,
+      "step": 59290
+    },
+    {
+      "epoch": 0.08332206331001463,
+      "grad_norm": 0.786556601524353,
+      "learning_rate": 0.0002498721371364339,
+      "loss": 4.8474,
+      "step": 59300
+    },
+    {
+      "epoch": 0.0833361142481782,
+      "grad_norm": 0.7999930381774902,
+      "learning_rate": 0.0002499142897288183,
+      "loss": 4.8823,
+      "step": 59310
+    },
+    {
+      "epoch": 0.08335016518634178,
+      "grad_norm": 0.8049028515815735,
+      "learning_rate": 0.00024995644232120274,
+      "loss": 4.758,
+      "step": 59320
+    },
+    {
+      "epoch": 0.08336421612450537,
+      "grad_norm": 0.8140138983726501,
+      "learning_rate": 0.00024999859491358717,
+      "loss": 4.8163,
+      "step": 59330
+    },
+    {
+      "epoch": 0.08337826706266895,
+      "grad_norm": 0.7999250292778015,
+      "learning_rate": 0.0002500407475059716,
+      "loss": 4.8743,
+      "step": 59340
+    },
+    {
+      "epoch": 0.08339231800083252,
+      "grad_norm": 0.8304258584976196,
+      "learning_rate": 0.00025008290009835603,
+      "loss": 4.8675,
+      "step": 59350
+    },
+    {
+      "epoch": 0.0834063689389961,
+      "grad_norm": 0.8012201189994812,
+      "learning_rate": 0.00025012505269074046,
+      "loss": 4.8525,
+      "step": 59360
+    },
+    {
+      "epoch": 0.08342041987715967,
+      "grad_norm": 0.7982044219970703,
+      "learning_rate": 0.0002501672052831249,
+      "loss": 4.8047,
+      "step": 59370
+    },
+    {
+      "epoch": 0.08343447081532325,
+      "grad_norm": 0.8110559582710266,
+      "learning_rate": 0.0002502093578755093,
+      "loss": 4.83,
+      "step": 59380
+    },
+    {
+      "epoch": 0.08344852175348683,
+      "grad_norm": 0.8042994141578674,
+      "learning_rate": 0.00025025151046789376,
+      "loss": 4.6822,
+      "step": 59390
+    },
+    {
+      "epoch": 0.0834625726916504,
+      "grad_norm": 0.9221730828285217,
+      "learning_rate": 0.0002502936630602782,
+      "loss": 4.8834,
+      "step": 59400
+    },
+    {
+      "epoch": 0.08347662362981398,
+      "grad_norm": 0.8320198059082031,
+      "learning_rate": 0.0002503358156526626,
+      "loss": 4.7958,
+      "step": 59410
+    },
+    {
+      "epoch": 0.08349067456797755,
+      "grad_norm": 0.8428821563720703,
+      "learning_rate": 0.00025037796824504705,
+      "loss": 4.7707,
+      "step": 59420
+    },
+    {
+      "epoch": 0.08350472550614114,
+      "grad_norm": 0.7867874503135681,
+      "learning_rate": 0.0002504201208374315,
+      "loss": 4.8291,
+      "step": 59430
+    },
+    {
+      "epoch": 0.08351877644430472,
+      "grad_norm": 0.8074051737785339,
+      "learning_rate": 0.0002504622734298159,
+      "loss": 4.8084,
+      "step": 59440
+    },
+    {
+      "epoch": 0.0835328273824683,
+      "grad_norm": 0.8142467141151428,
+      "learning_rate": 0.00025050442602220034,
+      "loss": 4.8289,
+      "step": 59450
+    },
+    {
+      "epoch": 0.08354687832063187,
+      "grad_norm": 0.8267741203308105,
+      "learning_rate": 0.0002505465786145848,
+      "loss": 4.734,
+      "step": 59460
+    },
+    {
+      "epoch": 0.08356092925879545,
+      "grad_norm": 0.786417543888092,
+      "learning_rate": 0.0002505887312069692,
+      "loss": 4.7357,
+      "step": 59470
+    },
+    {
+      "epoch": 0.08357498019695903,
+      "grad_norm": 0.8515885472297668,
+      "learning_rate": 0.00025063088379935364,
+      "loss": 4.7698,
+      "step": 59480
+    },
+    {
+      "epoch": 0.0835890311351226,
+      "grad_norm": 0.8577207922935486,
+      "learning_rate": 0.00025067303639173807,
+      "loss": 4.7342,
+      "step": 59490
+    },
+    {
+      "epoch": 0.08360308207328618,
+      "grad_norm": 0.8472533226013184,
+      "learning_rate": 0.0002507151889841225,
+      "loss": 4.7212,
+      "step": 59500
+    },
+    {
+      "epoch": 0.08361713301144975,
+      "grad_norm": 0.8497911691665649,
+      "learning_rate": 0.00025075734157650693,
+      "loss": 4.6883,
+      "step": 59510
+    },
+    {
+      "epoch": 0.08363118394961333,
+      "grad_norm": 0.7981038689613342,
+      "learning_rate": 0.00025079949416889136,
+      "loss": 4.7996,
+      "step": 59520
+    },
+    {
+      "epoch": 0.0836452348877769,
+      "grad_norm": 0.8073769807815552,
+      "learning_rate": 0.0002508416467612758,
+      "loss": 4.8362,
+      "step": 59530
+    },
+    {
+      "epoch": 0.0836592858259405,
+      "grad_norm": 0.8505818247795105,
+      "learning_rate": 0.0002508837993536602,
+      "loss": 4.7964,
+      "step": 59540
+    },
+    {
+      "epoch": 0.08367333676410407,
+      "grad_norm": 0.822891891002655,
+      "learning_rate": 0.00025092595194604466,
+      "loss": 4.9098,
+      "step": 59550
+    },
+    {
+      "epoch": 0.08368738770226765,
+      "grad_norm": 0.8498794436454773,
+      "learning_rate": 0.0002509681045384291,
+      "loss": 4.8038,
+      "step": 59560
+    },
+    {
+      "epoch": 0.08370143864043122,
+      "grad_norm": 0.8308042883872986,
+      "learning_rate": 0.0002510102571308135,
+      "loss": 4.8735,
+      "step": 59570
+    },
+    {
+      "epoch": 0.0837154895785948,
+      "grad_norm": 0.8018273711204529,
+      "learning_rate": 0.00025105240972319795,
+      "loss": 4.908,
+      "step": 59580
+    },
+    {
+      "epoch": 0.08372954051675838,
+      "grad_norm": 0.8206508159637451,
+      "learning_rate": 0.0002510945623155824,
+      "loss": 4.7587,
+      "step": 59590
+    },
+    {
+      "epoch": 0.08374359145492195,
+      "grad_norm": 0.841090977191925,
+      "learning_rate": 0.0002511367149079668,
+      "loss": 4.7037,
+      "step": 59600
+    },
+    {
+      "epoch": 0.08375764239308553,
+      "grad_norm": 0.7994279265403748,
+      "learning_rate": 0.00025117886750035124,
+      "loss": 4.7894,
+      "step": 59610
+    },
+    {
+      "epoch": 0.0837716933312491,
+      "grad_norm": 0.7760536074638367,
+      "learning_rate": 0.0002512210200927357,
+      "loss": 4.8517,
+      "step": 59620
+    },
+    {
+      "epoch": 0.08378574426941268,
+      "grad_norm": 0.8113006949424744,
+      "learning_rate": 0.0002512631726851201,
+      "loss": 4.9551,
+      "step": 59630
+    },
+    {
+      "epoch": 0.08379979520757627,
+      "grad_norm": 0.8211884498596191,
+      "learning_rate": 0.00025130532527750454,
+      "loss": 4.9197,
+      "step": 59640
+    },
+    {
+      "epoch": 0.08381384614573985,
+      "grad_norm": 0.8168864846229553,
+      "learning_rate": 0.00025134747786988897,
+      "loss": 4.8356,
+      "step": 59650
+    },
+    {
+      "epoch": 0.08382789708390342,
+      "grad_norm": 0.8315194845199585,
+      "learning_rate": 0.0002513896304622734,
+      "loss": 4.7761,
+      "step": 59660
+    },
+    {
+      "epoch": 0.083841948022067,
+      "grad_norm": 0.8301543593406677,
+      "learning_rate": 0.00025143178305465783,
+      "loss": 4.8829,
+      "step": 59670
+    },
+    {
+      "epoch": 0.08385599896023058,
+      "grad_norm": 0.8475062251091003,
+      "learning_rate": 0.00025147393564704226,
+      "loss": 4.7885,
+      "step": 59680
+    },
+    {
+      "epoch": 0.08387004989839415,
+      "grad_norm": 0.8310939073562622,
+      "learning_rate": 0.0002515160882394267,
+      "loss": 4.7185,
+      "step": 59690
+    },
+    {
+      "epoch": 0.08388410083655773,
+      "grad_norm": 0.8392112851142883,
+      "learning_rate": 0.0002515582408318111,
+      "loss": 4.7873,
+      "step": 59700
+    },
+    {
+      "epoch": 0.0838981517747213,
+      "grad_norm": 0.8230659365653992,
+      "learning_rate": 0.00025160039342419556,
+      "loss": 4.7909,
+      "step": 59710
+    },
+    {
+      "epoch": 0.08391220271288488,
+      "grad_norm": 0.834247350692749,
+      "learning_rate": 0.00025164254601658,
+      "loss": 4.7701,
+      "step": 59720
+    },
+    {
+      "epoch": 0.08392625365104846,
+      "grad_norm": 0.8113618493080139,
+      "learning_rate": 0.0002516846986089644,
+      "loss": 4.8853,
+      "step": 59730
+    },
+    {
+      "epoch": 0.08394030458921205,
+      "grad_norm": 0.8198868632316589,
+      "learning_rate": 0.00025172685120134885,
+      "loss": 4.8926,
+      "step": 59740
+    },
+    {
+      "epoch": 0.08395435552737562,
+      "grad_norm": 0.8316317200660706,
+      "learning_rate": 0.0002517690037937333,
+      "loss": 4.8605,
+      "step": 59750
+    },
+    {
+      "epoch": 0.0839684064655392,
+      "grad_norm": 0.9213061332702637,
+      "learning_rate": 0.0002518111563861177,
+      "loss": 4.8161,
+      "step": 59760
+    },
+    {
+      "epoch": 0.08398245740370278,
+      "grad_norm": 0.8067527413368225,
+      "learning_rate": 0.00025185330897850214,
+      "loss": 4.8305,
+      "step": 59770
+    },
+    {
+      "epoch": 0.08399650834186635,
+      "grad_norm": 0.8328394293785095,
+      "learning_rate": 0.0002518954615708866,
+      "loss": 4.8528,
+      "step": 59780
+    },
+    {
+      "epoch": 0.08401055928002993,
+      "grad_norm": 0.8063852190971375,
+      "learning_rate": 0.000251937614163271,
+      "loss": 4.7661,
+      "step": 59790
+    },
+    {
+      "epoch": 0.0840246102181935,
+      "grad_norm": 0.8150149583816528,
+      "learning_rate": 0.0002519797667556555,
+      "loss": 4.6904,
+      "step": 59800
+    },
+    {
+      "epoch": 0.08403866115635708,
+      "grad_norm": 0.8086451292037964,
+      "learning_rate": 0.00025202191934803987,
+      "loss": 4.8242,
+      "step": 59810
+    },
+    {
+      "epoch": 0.08405271209452066,
+      "grad_norm": 0.8651347160339355,
+      "learning_rate": 0.0002520640719404243,
+      "loss": 4.7535,
+      "step": 59820
+    },
+    {
+      "epoch": 0.08406676303268423,
+      "grad_norm": 0.8142220377922058,
+      "learning_rate": 0.00025210622453280873,
+      "loss": 4.8362,
+      "step": 59830
+    },
+    {
+      "epoch": 0.08408081397084781,
+      "grad_norm": 0.8657679557800293,
+      "learning_rate": 0.00025214837712519316,
+      "loss": 4.7861,
+      "step": 59840
+    },
+    {
+      "epoch": 0.0840948649090114,
+      "grad_norm": 0.87697833776474,
+      "learning_rate": 0.0002521905297175776,
+      "loss": 4.739,
+      "step": 59850
+    },
+    {
+      "epoch": 0.08410891584717498,
+      "grad_norm": 0.870245099067688,
+      "learning_rate": 0.0002522326823099621,
+      "loss": 4.7173,
+      "step": 59860
+    },
+    {
+      "epoch": 0.08412296678533855,
+      "grad_norm": 0.8389632701873779,
+      "learning_rate": 0.00025227483490234645,
+      "loss": 4.7693,
+      "step": 59870
+    },
+    {
+      "epoch": 0.08413701772350213,
+      "grad_norm": 0.8237364888191223,
+      "learning_rate": 0.0002523169874947309,
+      "loss": 4.8892,
+      "step": 59880
+    },
+    {
+      "epoch": 0.0841510686616657,
+      "grad_norm": 0.8023500442504883,
+      "learning_rate": 0.0002523591400871153,
+      "loss": 4.7923,
+      "step": 59890
+    },
+    {
+      "epoch": 0.08416511959982928,
+      "grad_norm": 0.8338358402252197,
+      "learning_rate": 0.00025240129267949975,
+      "loss": 4.8156,
+      "step": 59900
+    },
+    {
+      "epoch": 0.08417917053799286,
+      "grad_norm": 0.7917497158050537,
+      "learning_rate": 0.00025243923001264573,
+      "loss": 4.7169,
+      "step": 59910
+    },
+    {
+      "epoch": 0.08419322147615643,
+      "grad_norm": 0.8231458067893982,
+      "learning_rate": 0.00025248138260503016,
+      "loss": 4.8951,
+      "step": 59920
+    },
+    {
+      "epoch": 0.08420727241432001,
+      "grad_norm": 0.8315156102180481,
+      "learning_rate": 0.0002525235351974146,
+      "loss": 4.8144,
+      "step": 59930
+    },
+    {
+      "epoch": 0.08422132335248358,
+      "grad_norm": 0.8496831059455872,
+      "learning_rate": 0.0002525656877897991,
+      "loss": 4.9156,
+      "step": 59940
+    },
+    {
+      "epoch": 0.08423537429064717,
+      "grad_norm": 0.8502749800682068,
+      "learning_rate": 0.0002526078403821835,
+      "loss": 4.9349,
+      "step": 59950
+    },
+    {
+      "epoch": 0.08424942522881075,
+      "grad_norm": 0.8212767243385315,
+      "learning_rate": 0.00025264999297456794,
+      "loss": 4.801,
+      "step": 59960
+    },
+    {
+      "epoch": 0.08426347616697433,
+      "grad_norm": 0.8079051971435547,
+      "learning_rate": 0.0002526921455669523,
+      "loss": 4.821,
+      "step": 59970
+    },
+    {
+      "epoch": 0.0842775271051379,
+      "grad_norm": 0.7926346063613892,
+      "learning_rate": 0.00025273429815933675,
+      "loss": 4.7747,
+      "step": 59980
+    },
+    {
+      "epoch": 0.08429157804330148,
+      "grad_norm": 0.8248520493507385,
+      "learning_rate": 0.0002527764507517212,
+      "loss": 4.8311,
+      "step": 59990
+    },
+    {
+      "epoch": 0.08430562898146506,
+      "grad_norm": 0.8126881122589111,
+      "learning_rate": 0.00025281860334410566,
+      "loss": 4.8472,
+      "step": 60000
+    },
+    {
+      "epoch": 0.08431967991962863,
+      "grad_norm": 0.8101698160171509,
+      "learning_rate": 0.0002528607559364901,
+      "loss": 4.7401,
+      "step": 60010
+    },
+    {
+      "epoch": 0.08433373085779221,
+      "grad_norm": 0.8176490664482117,
+      "learning_rate": 0.00025290290852887453,
+      "loss": 4.7447,
+      "step": 60020
+    },
+    {
+      "epoch": 0.08434778179595578,
+      "grad_norm": 0.8406776189804077,
+      "learning_rate": 0.00025294506112125896,
+      "loss": 4.7898,
+      "step": 60030
+    },
+    {
+      "epoch": 0.08436183273411936,
+      "grad_norm": 0.8245927691459656,
+      "learning_rate": 0.00025298721371364334,
+      "loss": 4.6811,
+      "step": 60040
+    },
+    {
+      "epoch": 0.08437588367228295,
+      "grad_norm": 0.8441291451454163,
+      "learning_rate": 0.00025302936630602777,
+      "loss": 4.7632,
+      "step": 60050
+    },
+    {
+      "epoch": 0.08438993461044653,
+      "grad_norm": 0.8327770829200745,
+      "learning_rate": 0.00025307151889841225,
+      "loss": 4.8536,
+      "step": 60060
+    },
+    {
+      "epoch": 0.0844039855486101,
+      "grad_norm": 0.850288987159729,
+      "learning_rate": 0.0002531136714907967,
+      "loss": 4.8284,
+      "step": 60070
+    },
+    {
+      "epoch": 0.08441803648677368,
+      "grad_norm": 0.7964315414428711,
+      "learning_rate": 0.0002531558240831811,
+      "loss": 4.9407,
+      "step": 60080
+    },
+    {
+      "epoch": 0.08443208742493725,
+      "grad_norm": 0.8162838220596313,
+      "learning_rate": 0.00025319797667556555,
+      "loss": 4.8588,
+      "step": 60090
+    },
+    {
+      "epoch": 0.08444613836310083,
+      "grad_norm": 0.8507556319236755,
+      "learning_rate": 0.0002532401292679499,
+      "loss": 4.7351,
+      "step": 60100
+    },
+    {
+      "epoch": 0.08446018930126441,
+      "grad_norm": 0.7655229568481445,
+      "learning_rate": 0.00025328228186033435,
+      "loss": 4.9912,
+      "step": 60110
+    },
+    {
+      "epoch": 0.08447424023942798,
+      "grad_norm": 0.8107126951217651,
+      "learning_rate": 0.00025332443445271884,
+      "loss": 4.7462,
+      "step": 60120
+    },
+    {
+      "epoch": 0.08448829117759156,
+      "grad_norm": 0.8109584450721741,
+      "learning_rate": 0.00025336658704510327,
+      "loss": 4.8539,
+      "step": 60130
+    },
+    {
+      "epoch": 0.08450234211575514,
+      "grad_norm": 0.8400088548660278,
+      "learning_rate": 0.0002534087396374877,
+      "loss": 4.9024,
+      "step": 60140
+    },
+    {
+      "epoch": 0.08451639305391871,
+      "grad_norm": 0.8335424065589905,
+      "learning_rate": 0.00025345089222987213,
+      "loss": 4.8311,
+      "step": 60150
+    },
+    {
+      "epoch": 0.0845304439920823,
+      "grad_norm": 0.8169661164283752,
+      "learning_rate": 0.0002534930448222565,
+      "loss": 4.8582,
+      "step": 60160
+    },
+    {
+      "epoch": 0.08454449493024588,
+      "grad_norm": 0.7944623827934265,
+      "learning_rate": 0.00025353519741464094,
+      "loss": 4.7601,
+      "step": 60170
+    },
+    {
+      "epoch": 0.08455854586840945,
+      "grad_norm": 0.8325213193893433,
+      "learning_rate": 0.0002535773500070254,
+      "loss": 4.819,
+      "step": 60180
+    },
+    {
+      "epoch": 0.08457259680657303,
+      "grad_norm": 0.8035892844200134,
+      "learning_rate": 0.00025361950259940986,
+      "loss": 4.8387,
+      "step": 60190
+    },
+    {
+      "epoch": 0.0845866477447366,
+      "grad_norm": 0.8531663417816162,
+      "learning_rate": 0.0002536616551917943,
+      "loss": 4.9727,
+      "step": 60200
+    },
+    {
+      "epoch": 0.08460069868290018,
+      "grad_norm": 0.8064233660697937,
+      "learning_rate": 0.0002537038077841787,
+      "loss": 4.739,
+      "step": 60210
+    },
+    {
+      "epoch": 0.08461474962106376,
+      "grad_norm": 0.8612556457519531,
+      "learning_rate": 0.00025374596037656315,
+      "loss": 4.763,
+      "step": 60220
+    },
+    {
+      "epoch": 0.08462880055922734,
+      "grad_norm": 0.8533815145492554,
+      "learning_rate": 0.0002537881129689476,
+      "loss": 4.7813,
+      "step": 60230
+    },
+    {
+      "epoch": 0.08464285149739091,
+      "grad_norm": 0.7971927523612976,
+      "learning_rate": 0.000253830265561332,
+      "loss": 4.7823,
+      "step": 60240
+    },
+    {
+      "epoch": 0.08465690243555449,
+      "grad_norm": 0.8081861734390259,
+      "learning_rate": 0.00025387241815371644,
+      "loss": 4.7512,
+      "step": 60250
+    },
+    {
+      "epoch": 0.08467095337371808,
+      "grad_norm": 0.8018068075180054,
+      "learning_rate": 0.0002539145707461009,
+      "loss": 4.7706,
+      "step": 60260
+    },
+    {
+      "epoch": 0.08468500431188165,
+      "grad_norm": 0.8136337995529175,
+      "learning_rate": 0.0002539567233384853,
+      "loss": 4.7486,
+      "step": 60270
+    },
+    {
+      "epoch": 0.08469905525004523,
+      "grad_norm": 0.837989091873169,
+      "learning_rate": 0.00025399887593086974,
+      "loss": 4.7813,
+      "step": 60280
+    },
+    {
+      "epoch": 0.0847131061882088,
+      "grad_norm": 0.8240730166435242,
+      "learning_rate": 0.00025404102852325417,
+      "loss": 4.8329,
+      "step": 60290
+    },
+    {
+      "epoch": 0.08472715712637238,
+      "grad_norm": 0.8388275504112244,
+      "learning_rate": 0.0002540831811156386,
+      "loss": 4.8039,
+      "step": 60300
+    },
+    {
+      "epoch": 0.08474120806453596,
+      "grad_norm": 0.8347736597061157,
+      "learning_rate": 0.00025412533370802303,
+      "loss": 4.6845,
+      "step": 60310
+    },
+    {
+      "epoch": 0.08475525900269953,
+      "grad_norm": 0.8188750147819519,
+      "learning_rate": 0.00025416748630040746,
+      "loss": 4.7612,
+      "step": 60320
+    },
+    {
+      "epoch": 0.08476930994086311,
+      "grad_norm": 0.8329417109489441,
+      "learning_rate": 0.0002542096388927919,
+      "loss": 4.8708,
+      "step": 60330
+    },
+    {
+      "epoch": 0.08478336087902669,
+      "grad_norm": 0.8301446437835693,
+      "learning_rate": 0.0002542517914851763,
+      "loss": 4.8245,
+      "step": 60340
+    },
+    {
+      "epoch": 0.08479741181719026,
+      "grad_norm": 0.8130600452423096,
+      "learning_rate": 0.00025429394407756076,
+      "loss": 4.8549,
+      "step": 60350
+    },
+    {
+      "epoch": 0.08481146275535385,
+      "grad_norm": 0.8038048148155212,
+      "learning_rate": 0.0002543360966699452,
+      "loss": 4.8175,
+      "step": 60360
+    },
+    {
+      "epoch": 0.08482551369351743,
+      "grad_norm": 0.7973541021347046,
+      "learning_rate": 0.0002543782492623296,
+      "loss": 4.852,
+      "step": 60370
+    },
+    {
+      "epoch": 0.084839564631681,
+      "grad_norm": 0.8406749367713928,
+      "learning_rate": 0.00025442040185471405,
+      "loss": 4.8618,
+      "step": 60380
+    },
+    {
+      "epoch": 0.08485361556984458,
+      "grad_norm": 0.8773977160453796,
+      "learning_rate": 0.0002544625544470985,
+      "loss": 4.7474,
+      "step": 60390
+    },
+    {
+      "epoch": 0.08486766650800816,
+      "grad_norm": 0.7995550632476807,
+      "learning_rate": 0.0002545047070394829,
+      "loss": 4.7697,
+      "step": 60400
+    },
+    {
+      "epoch": 0.08488171744617173,
+      "grad_norm": 0.8282455801963806,
+      "learning_rate": 0.00025454685963186734,
+      "loss": 4.8029,
+      "step": 60410
+    },
+    {
+      "epoch": 0.08489576838433531,
+      "grad_norm": 0.8432449102401733,
+      "learning_rate": 0.0002545890122242518,
+      "loss": 4.872,
+      "step": 60420
+    },
+    {
+      "epoch": 0.08490981932249889,
+      "grad_norm": 0.8184887170791626,
+      "learning_rate": 0.0002546311648166362,
+      "loss": 4.7083,
+      "step": 60430
+    },
+    {
+      "epoch": 0.08492387026066246,
+      "grad_norm": 0.7973591685295105,
+      "learning_rate": 0.00025467331740902064,
+      "loss": 4.8724,
+      "step": 60440
+    },
+    {
+      "epoch": 0.08493792119882604,
+      "grad_norm": 0.8287274837493896,
+      "learning_rate": 0.00025471547000140507,
+      "loss": 4.9017,
+      "step": 60450
+    },
+    {
+      "epoch": 0.08495197213698961,
+      "grad_norm": 0.8042613863945007,
+      "learning_rate": 0.0002547576225937895,
+      "loss": 4.7804,
+      "step": 60460
+    },
+    {
+      "epoch": 0.0849660230751532,
+      "grad_norm": 0.8201180100440979,
+      "learning_rate": 0.00025479977518617393,
+      "loss": 4.8187,
+      "step": 60470
+    },
+    {
+      "epoch": 0.08498007401331678,
+      "grad_norm": 0.8212829232215881,
+      "learning_rate": 0.00025484192777855836,
+      "loss": 4.8745,
+      "step": 60480
+    },
+    {
+      "epoch": 0.08499412495148036,
+      "grad_norm": 0.8027827143669128,
+      "learning_rate": 0.0002548840803709428,
+      "loss": 4.7172,
+      "step": 60490
+    },
+    {
+      "epoch": 0.08500817588964393,
+      "grad_norm": 1.7494938373565674,
+      "learning_rate": 0.0002549262329633272,
+      "loss": 4.827,
+      "step": 60500
+    },
+    {
+      "epoch": 0.08502222682780751,
+      "grad_norm": 0.8321921825408936,
+      "learning_rate": 0.00025496838555571165,
+      "loss": 4.7602,
+      "step": 60510
+    },
+    {
+      "epoch": 0.08503627776597109,
+      "grad_norm": 0.8253518342971802,
+      "learning_rate": 0.0002550105381480961,
+      "loss": 4.7594,
+      "step": 60520
+    },
+    {
+      "epoch": 0.08505032870413466,
+      "grad_norm": 0.8793212175369263,
+      "learning_rate": 0.0002550526907404805,
+      "loss": 4.7981,
+      "step": 60530
+    },
+    {
+      "epoch": 0.08506437964229824,
+      "grad_norm": 0.8195023536682129,
+      "learning_rate": 0.00025509484333286495,
+      "loss": 4.7146,
+      "step": 60540
+    },
+    {
+      "epoch": 0.08507843058046181,
+      "grad_norm": 0.8089900016784668,
+      "learning_rate": 0.0002551369959252494,
+      "loss": 4.835,
+      "step": 60550
+    },
+    {
+      "epoch": 0.08509248151862539,
+      "grad_norm": 0.798321008682251,
+      "learning_rate": 0.0002551791485176338,
+      "loss": 4.9411,
+      "step": 60560
+    },
+    {
+      "epoch": 0.08510653245678898,
+      "grad_norm": 0.7749428153038025,
+      "learning_rate": 0.00025522130111001824,
+      "loss": 4.7919,
+      "step": 60570
+    },
+    {
+      "epoch": 0.08512058339495256,
+      "grad_norm": 0.832476019859314,
+      "learning_rate": 0.00025526345370240267,
+      "loss": 4.8104,
+      "step": 60580
+    },
+    {
+      "epoch": 0.08513463433311613,
+      "grad_norm": 0.8051169514656067,
+      "learning_rate": 0.0002553056062947871,
+      "loss": 4.7459,
+      "step": 60590
+    },
+    {
+      "epoch": 0.08514868527127971,
+      "grad_norm": 0.8199285268783569,
+      "learning_rate": 0.00025534775888717153,
+      "loss": 4.7434,
+      "step": 60600
+    },
+    {
+      "epoch": 0.08516273620944328,
+      "grad_norm": 0.820178210735321,
+      "learning_rate": 0.00025538991147955597,
+      "loss": 4.8498,
+      "step": 60610
+    },
+    {
+      "epoch": 0.08517678714760686,
+      "grad_norm": 0.8192396759986877,
+      "learning_rate": 0.0002554320640719404,
+      "loss": 4.8455,
+      "step": 60620
+    },
+    {
+      "epoch": 0.08519083808577044,
+      "grad_norm": 0.8456838130950928,
+      "learning_rate": 0.00025547421666432483,
+      "loss": 4.9442,
+      "step": 60630
+    },
+    {
+      "epoch": 0.08520488902393401,
+      "grad_norm": 0.7862314581871033,
+      "learning_rate": 0.00025551636925670926,
+      "loss": 4.8109,
+      "step": 60640
+    },
+    {
+      "epoch": 0.08521893996209759,
+      "grad_norm": 0.781730592250824,
+      "learning_rate": 0.0002555585218490937,
+      "loss": 4.8463,
+      "step": 60650
+    },
+    {
+      "epoch": 0.08523299090026117,
+      "grad_norm": 0.8605624437332153,
+      "learning_rate": 0.0002556006744414781,
+      "loss": 4.7406,
+      "step": 60660
+    },
+    {
+      "epoch": 0.08524704183842476,
+      "grad_norm": 0.798784613609314,
+      "learning_rate": 0.00025564282703386255,
+      "loss": 4.9813,
+      "step": 60670
+    },
+    {
+      "epoch": 0.08526109277658833,
+      "grad_norm": 0.8223904967308044,
+      "learning_rate": 0.000255684979626247,
+      "loss": 4.9145,
+      "step": 60680
+    },
+    {
+      "epoch": 0.08527514371475191,
+      "grad_norm": 0.8261276483535767,
+      "learning_rate": 0.0002557271322186314,
+      "loss": 4.815,
+      "step": 60690
+    },
+    {
+      "epoch": 0.08528919465291548,
+      "grad_norm": 0.7942939400672913,
+      "learning_rate": 0.00025576928481101585,
+      "loss": 4.7418,
+      "step": 60700
+    },
+    {
+      "epoch": 0.08530324559107906,
+      "grad_norm": 0.7955240607261658,
+      "learning_rate": 0.0002558114374034003,
+      "loss": 4.7581,
+      "step": 60710
+    },
+    {
+      "epoch": 0.08531729652924264,
+      "grad_norm": 0.8077800273895264,
+      "learning_rate": 0.0002558535899957847,
+      "loss": 4.8254,
+      "step": 60720
+    },
+    {
+      "epoch": 0.08533134746740621,
+      "grad_norm": 0.8321365714073181,
+      "learning_rate": 0.00025589574258816914,
+      "loss": 4.8636,
+      "step": 60730
+    },
+    {
+      "epoch": 0.08534539840556979,
+      "grad_norm": 0.8244563341140747,
+      "learning_rate": 0.0002559378951805536,
+      "loss": 4.8733,
+      "step": 60740
+    },
+    {
+      "epoch": 0.08535944934373337,
+      "grad_norm": 0.8096931576728821,
+      "learning_rate": 0.000255980047772938,
+      "loss": 4.8622,
+      "step": 60750
+    },
+    {
+      "epoch": 0.08537350028189694,
+      "grad_norm": 0.8346720337867737,
+      "learning_rate": 0.00025602220036532243,
+      "loss": 4.8147,
+      "step": 60760
+    },
+    {
+      "epoch": 0.08538755122006053,
+      "grad_norm": 0.7969753742218018,
+      "learning_rate": 0.00025606435295770686,
+      "loss": 4.803,
+      "step": 60770
+    },
+    {
+      "epoch": 0.08540160215822411,
+      "grad_norm": 0.7898983359336853,
+      "learning_rate": 0.0002561065055500913,
+      "loss": 4.7337,
+      "step": 60780
+    },
+    {
+      "epoch": 0.08541565309638768,
+      "grad_norm": 0.815297544002533,
+      "learning_rate": 0.0002561486581424757,
+      "loss": 4.8547,
+      "step": 60790
+    },
+    {
+      "epoch": 0.08542970403455126,
+      "grad_norm": 0.8263185620307922,
+      "learning_rate": 0.0002561908107348602,
+      "loss": 4.7106,
+      "step": 60800
+    },
+    {
+      "epoch": 0.08544375497271484,
+      "grad_norm": 0.7924615740776062,
+      "learning_rate": 0.0002562329633272446,
+      "loss": 4.628,
+      "step": 60810
+    },
+    {
+      "epoch": 0.08545780591087841,
+      "grad_norm": 0.8533799052238464,
+      "learning_rate": 0.000256275115919629,
+      "loss": 4.8976,
+      "step": 60820
+    },
+    {
+      "epoch": 0.08547185684904199,
+      "grad_norm": 0.7679029703140259,
+      "learning_rate": 0.00025631726851201345,
+      "loss": 4.8269,
+      "step": 60830
+    },
+    {
+      "epoch": 0.08548590778720556,
+      "grad_norm": 0.9749653935432434,
+      "learning_rate": 0.0002563594211043979,
+      "loss": 4.8443,
+      "step": 60840
+    },
+    {
+      "epoch": 0.08549995872536914,
+      "grad_norm": 0.8148539066314697,
+      "learning_rate": 0.0002564015736967823,
+      "loss": 4.8212,
+      "step": 60850
+    },
+    {
+      "epoch": 0.08551400966353272,
+      "grad_norm": 0.8042306900024414,
+      "learning_rate": 0.0002564437262891668,
+      "loss": 4.9523,
+      "step": 60860
+    },
+    {
+      "epoch": 0.08552806060169629,
+      "grad_norm": 0.8610725998878479,
+      "learning_rate": 0.0002564858788815512,
+      "loss": 4.822,
+      "step": 60870
+    },
+    {
+      "epoch": 0.08554211153985988,
+      "grad_norm": 0.8803948760032654,
+      "learning_rate": 0.0002565280314739356,
+      "loss": 4.7024,
+      "step": 60880
+    },
+    {
+      "epoch": 0.08555616247802346,
+      "grad_norm": 0.805877149105072,
+      "learning_rate": 0.00025657018406632004,
+      "loss": 4.7908,
+      "step": 60890
+    },
+    {
+      "epoch": 0.08557021341618704,
+      "grad_norm": 0.8154386878013611,
+      "learning_rate": 0.00025661233665870447,
+      "loss": 4.9642,
+      "step": 60900
+    },
+    {
+      "epoch": 0.08558426435435061,
+      "grad_norm": 0.8559610843658447,
+      "learning_rate": 0.0002566544892510889,
+      "loss": 4.7765,
+      "step": 60910
+    },
+    {
+      "epoch": 0.08559831529251419,
+      "grad_norm": 0.8591551184654236,
+      "learning_rate": 0.0002566966418434734,
+      "loss": 4.7464,
+      "step": 60920
+    },
+    {
+      "epoch": 0.08561236623067776,
+      "grad_norm": 0.8162509202957153,
+      "learning_rate": 0.0002567387944358578,
+      "loss": 4.7748,
+      "step": 60930
+    },
+    {
+      "epoch": 0.08562641716884134,
+      "grad_norm": 0.8140776753425598,
+      "learning_rate": 0.0002567809470282422,
+      "loss": 4.888,
+      "step": 60940
+    },
+    {
+      "epoch": 0.08564046810700492,
+      "grad_norm": 0.8785784244537354,
+      "learning_rate": 0.0002568230996206266,
+      "loss": 4.9443,
+      "step": 60950
+    },
+    {
+      "epoch": 0.08565451904516849,
+      "grad_norm": 0.7831446528434753,
+      "learning_rate": 0.00025686525221301106,
+      "loss": 4.9672,
+      "step": 60960
+    },
+    {
+      "epoch": 0.08566856998333207,
+      "grad_norm": 0.8083533048629761,
+      "learning_rate": 0.0002569074048053955,
+      "loss": 4.7751,
+      "step": 60970
+    },
+    {
+      "epoch": 0.08568262092149566,
+      "grad_norm": 0.8388922810554504,
+      "learning_rate": 0.00025694955739778,
+      "loss": 4.8792,
+      "step": 60980
+    },
+    {
+      "epoch": 0.08569667185965923,
+      "grad_norm": 0.8060362339019775,
+      "learning_rate": 0.0002569917099901644,
+      "loss": 4.796,
+      "step": 60990
+    },
+    {
+      "epoch": 0.08571072279782281,
+      "grad_norm": 0.8279473185539246,
+      "learning_rate": 0.0002570338625825488,
+      "loss": 4.7601,
+      "step": 61000
+    },
+    {
+      "epoch": 0.08572477373598639,
+      "grad_norm": 0.8248935341835022,
+      "learning_rate": 0.0002570760151749332,
+      "loss": 4.8548,
+      "step": 61010
+    },
+    {
+      "epoch": 0.08573882467414996,
+      "grad_norm": 0.8512712717056274,
+      "learning_rate": 0.00025711816776731764,
+      "loss": 4.7709,
+      "step": 61020
+    },
+    {
+      "epoch": 0.08575287561231354,
+      "grad_norm": 0.7810263633728027,
+      "learning_rate": 0.0002571603203597021,
+      "loss": 4.8958,
+      "step": 61030
+    },
+    {
+      "epoch": 0.08576692655047712,
+      "grad_norm": 0.8753952980041504,
+      "learning_rate": 0.00025720247295208656,
+      "loss": 4.6734,
+      "step": 61040
+    },
+    {
+      "epoch": 0.08578097748864069,
+      "grad_norm": 0.8158033490180969,
+      "learning_rate": 0.000257244625544471,
+      "loss": 4.8609,
+      "step": 61050
+    },
+    {
+      "epoch": 0.08579502842680427,
+      "grad_norm": 0.8370834589004517,
+      "learning_rate": 0.0002572867781368554,
+      "loss": 4.8195,
+      "step": 61060
+    },
+    {
+      "epoch": 0.08580907936496784,
+      "grad_norm": 0.792068362236023,
+      "learning_rate": 0.0002573289307292398,
+      "loss": 4.7818,
+      "step": 61070
+    },
+    {
+      "epoch": 0.08582313030313143,
+      "grad_norm": 0.8027522563934326,
+      "learning_rate": 0.00025737108332162423,
+      "loss": 4.7769,
+      "step": 61080
+    },
+    {
+      "epoch": 0.08583718124129501,
+      "grad_norm": 0.8090192079544067,
+      "learning_rate": 0.00025741323591400866,
+      "loss": 4.6341,
+      "step": 61090
+    },
+    {
+      "epoch": 0.08585123217945859,
+      "grad_norm": 0.7874997854232788,
+      "learning_rate": 0.00025745538850639315,
+      "loss": 4.9727,
+      "step": 61100
+    },
+    {
+      "epoch": 0.08586528311762216,
+      "grad_norm": 0.7994272708892822,
+      "learning_rate": 0.0002574975410987776,
+      "loss": 4.7984,
+      "step": 61110
+    },
+    {
+      "epoch": 0.08587933405578574,
+      "grad_norm": 0.8127740025520325,
+      "learning_rate": 0.000257539693691162,
+      "loss": 4.8371,
+      "step": 61120
+    },
+    {
+      "epoch": 0.08589338499394931,
+      "grad_norm": 0.7839046120643616,
+      "learning_rate": 0.0002575818462835464,
+      "loss": 4.9259,
+      "step": 61130
+    },
+    {
+      "epoch": 0.08590743593211289,
+      "grad_norm": 0.812605619430542,
+      "learning_rate": 0.0002576239988759308,
+      "loss": 4.7804,
+      "step": 61140
+    },
+    {
+      "epoch": 0.08592148687027647,
+      "grad_norm": 0.8251674175262451,
+      "learning_rate": 0.00025766615146831525,
+      "loss": 4.7948,
+      "step": 61150
+    },
+    {
+      "epoch": 0.08593553780844004,
+      "grad_norm": 0.7972586750984192,
+      "learning_rate": 0.00025770830406069973,
+      "loss": 4.7815,
+      "step": 61160
+    },
+    {
+      "epoch": 0.08594958874660362,
+      "grad_norm": 0.8652247190475464,
+      "learning_rate": 0.00025775045665308417,
+      "loss": 4.9054,
+      "step": 61170
+    },
+    {
+      "epoch": 0.0859636396847672,
+      "grad_norm": 0.7779868245124817,
+      "learning_rate": 0.00025778839398623015,
+      "loss": 4.8678,
+      "step": 61180
+    },
+    {
+      "epoch": 0.08597769062293079,
+      "grad_norm": 0.8390598893165588,
+      "learning_rate": 0.0002578305465786146,
+      "loss": 4.796,
+      "step": 61190
+    },
+    {
+      "epoch": 0.08599174156109436,
+      "grad_norm": 0.7976048588752747,
+      "learning_rate": 0.000257872699170999,
+      "loss": 4.8985,
+      "step": 61200
+    },
+    {
+      "epoch": 0.08600579249925794,
+      "grad_norm": 0.8002855777740479,
+      "learning_rate": 0.00025791485176338344,
+      "loss": 4.8251,
+      "step": 61210
+    },
+    {
+      "epoch": 0.08601984343742151,
+      "grad_norm": 0.787406861782074,
+      "learning_rate": 0.00025795700435576787,
+      "loss": 4.8762,
+      "step": 61220
+    },
+    {
+      "epoch": 0.08603389437558509,
+      "grad_norm": 0.8275022506713867,
+      "learning_rate": 0.00025799915694815225,
+      "loss": 4.8124,
+      "step": 61230
+    },
+    {
+      "epoch": 0.08604794531374867,
+      "grad_norm": 0.8134986162185669,
+      "learning_rate": 0.00025804130954053674,
+      "loss": 4.6739,
+      "step": 61240
+    },
+    {
+      "epoch": 0.08606199625191224,
+      "grad_norm": 0.8022597432136536,
+      "learning_rate": 0.00025808346213292117,
+      "loss": 4.8794,
+      "step": 61250
+    },
+    {
+      "epoch": 0.08607604719007582,
+      "grad_norm": 0.789439857006073,
+      "learning_rate": 0.0002581256147253056,
+      "loss": 4.8165,
+      "step": 61260
+    },
+    {
+      "epoch": 0.0860900981282394,
+      "grad_norm": 0.7730726003646851,
+      "learning_rate": 0.00025816776731769003,
+      "loss": 4.8367,
+      "step": 61270
+    },
+    {
+      "epoch": 0.08610414906640297,
+      "grad_norm": 0.8278162479400635,
+      "learning_rate": 0.00025820991991007446,
+      "loss": 4.6802,
+      "step": 61280
+    },
+    {
+      "epoch": 0.08611820000456656,
+      "grad_norm": 0.8260532021522522,
+      "learning_rate": 0.00025825207250245884,
+      "loss": 4.7331,
+      "step": 61290
+    },
+    {
+      "epoch": 0.08613225094273014,
+      "grad_norm": 0.8389633297920227,
+      "learning_rate": 0.0002582942250948433,
+      "loss": 4.8219,
+      "step": 61300
+    },
+    {
+      "epoch": 0.08614630188089371,
+      "grad_norm": 0.7943553328514099,
+      "learning_rate": 0.00025833637768722775,
+      "loss": 4.9142,
+      "step": 61310
+    },
+    {
+      "epoch": 0.08616035281905729,
+      "grad_norm": 0.815380334854126,
+      "learning_rate": 0.0002583785302796122,
+      "loss": 4.7978,
+      "step": 61320
+    },
+    {
+      "epoch": 0.08617440375722087,
+      "grad_norm": 0.7899944186210632,
+      "learning_rate": 0.0002584206828719966,
+      "loss": 4.7829,
+      "step": 61330
+    },
+    {
+      "epoch": 0.08618845469538444,
+      "grad_norm": 0.8026753664016724,
+      "learning_rate": 0.00025846283546438105,
+      "loss": 4.7086,
+      "step": 61340
+    },
+    {
+      "epoch": 0.08620250563354802,
+      "grad_norm": 0.8403603434562683,
+      "learning_rate": 0.0002585049880567655,
+      "loss": 4.8447,
+      "step": 61350
+    },
+    {
+      "epoch": 0.0862165565717116,
+      "grad_norm": 0.7885310649871826,
+      "learning_rate": 0.0002585471406491499,
+      "loss": 4.818,
+      "step": 61360
+    },
+    {
+      "epoch": 0.08623060750987517,
+      "grad_norm": 0.7821835279464722,
+      "learning_rate": 0.00025858929324153434,
+      "loss": 4.7951,
+      "step": 61370
+    },
+    {
+      "epoch": 0.08624465844803875,
+      "grad_norm": 0.8090482354164124,
+      "learning_rate": 0.00025863144583391877,
+      "loss": 4.7686,
+      "step": 61380
+    },
+    {
+      "epoch": 0.08625870938620234,
+      "grad_norm": 0.8011608719825745,
+      "learning_rate": 0.0002586735984263032,
+      "loss": 4.8789,
+      "step": 61390
+    },
+    {
+      "epoch": 0.08627276032436591,
+      "grad_norm": 0.7893279194831848,
+      "learning_rate": 0.00025871575101868763,
+      "loss": 4.8815,
+      "step": 61400
+    },
+    {
+      "epoch": 0.08628681126252949,
+      "grad_norm": 0.8958577513694763,
+      "learning_rate": 0.00025875790361107206,
+      "loss": 4.6991,
+      "step": 61410
+    },
+    {
+      "epoch": 0.08630086220069307,
+      "grad_norm": 0.8177340030670166,
+      "learning_rate": 0.0002588000562034565,
+      "loss": 4.8222,
+      "step": 61420
+    },
+    {
+      "epoch": 0.08631491313885664,
+      "grad_norm": 0.8018973469734192,
+      "learning_rate": 0.00025884220879584093,
+      "loss": 4.9089,
+      "step": 61430
+    },
+    {
+      "epoch": 0.08632896407702022,
+      "grad_norm": 0.7807876467704773,
+      "learning_rate": 0.00025888436138822536,
+      "loss": 4.826,
+      "step": 61440
+    },
+    {
+      "epoch": 0.0863430150151838,
+      "grad_norm": 0.8257559537887573,
+      "learning_rate": 0.0002589265139806098,
+      "loss": 4.8224,
+      "step": 61450
+    },
+    {
+      "epoch": 0.08635706595334737,
+      "grad_norm": 0.8308781385421753,
+      "learning_rate": 0.0002589686665729942,
+      "loss": 4.6974,
+      "step": 61460
+    },
+    {
+      "epoch": 0.08637111689151095,
+      "grad_norm": 0.7814127206802368,
+      "learning_rate": 0.00025901081916537865,
+      "loss": 4.8122,
+      "step": 61470
+    },
+    {
+      "epoch": 0.08638516782967452,
+      "grad_norm": 0.7803528904914856,
+      "learning_rate": 0.0002590529717577631,
+      "loss": 4.8447,
+      "step": 61480
+    },
+    {
+      "epoch": 0.0863992187678381,
+      "grad_norm": 0.8149372935295105,
+      "learning_rate": 0.0002590951243501475,
+      "loss": 4.7555,
+      "step": 61490
+    },
+    {
+      "epoch": 0.08641326970600169,
+      "grad_norm": 0.8497858047485352,
+      "learning_rate": 0.00025913727694253195,
+      "loss": 4.7493,
+      "step": 61500
+    },
+    {
+      "epoch": 0.08642732064416526,
+      "grad_norm": 0.8034449219703674,
+      "learning_rate": 0.0002591794295349164,
+      "loss": 4.8812,
+      "step": 61510
+    },
+    {
+      "epoch": 0.08644137158232884,
+      "grad_norm": 0.8024821877479553,
+      "learning_rate": 0.0002592215821273008,
+      "loss": 4.928,
+      "step": 61520
+    },
+    {
+      "epoch": 0.08645542252049242,
+      "grad_norm": 0.7921434640884399,
+      "learning_rate": 0.00025926373471968524,
+      "loss": 4.751,
+      "step": 61530
+    },
+    {
+      "epoch": 0.086469473458656,
+      "grad_norm": 0.8643134832382202,
+      "learning_rate": 0.00025930588731206967,
+      "loss": 4.8568,
+      "step": 61540
+    },
+    {
+      "epoch": 0.08648352439681957,
+      "grad_norm": 0.7830716967582703,
+      "learning_rate": 0.0002593480399044541,
+      "loss": 4.853,
+      "step": 61550
+    },
+    {
+      "epoch": 0.08649757533498315,
+      "grad_norm": 0.7742080092430115,
+      "learning_rate": 0.00025939019249683853,
+      "loss": 4.8559,
+      "step": 61560
+    },
+    {
+      "epoch": 0.08651162627314672,
+      "grad_norm": 0.8120469450950623,
+      "learning_rate": 0.00025943234508922296,
+      "loss": 4.856,
+      "step": 61570
+    },
+    {
+      "epoch": 0.0865256772113103,
+      "grad_norm": 0.827467679977417,
+      "learning_rate": 0.0002594744976816074,
+      "loss": 4.7491,
+      "step": 61580
+    },
+    {
+      "epoch": 0.08653972814947387,
+      "grad_norm": 0.7826569080352783,
+      "learning_rate": 0.0002595166502739918,
+      "loss": 4.7901,
+      "step": 61590
+    },
+    {
+      "epoch": 0.08655377908763746,
+      "grad_norm": 0.7950142621994019,
+      "learning_rate": 0.00025955880286637626,
+      "loss": 4.7976,
+      "step": 61600
+    },
+    {
+      "epoch": 0.08656783002580104,
+      "grad_norm": 0.821763277053833,
+      "learning_rate": 0.0002596009554587607,
+      "loss": 4.821,
+      "step": 61610
+    },
+    {
+      "epoch": 0.08658188096396462,
+      "grad_norm": 0.8673090934753418,
+      "learning_rate": 0.0002596431080511451,
+      "loss": 4.7369,
+      "step": 61620
+    },
+    {
+      "epoch": 0.08659593190212819,
+      "grad_norm": 0.7934688925743103,
+      "learning_rate": 0.00025968526064352955,
+      "loss": 4.6503,
+      "step": 61630
+    },
+    {
+      "epoch": 0.08660998284029177,
+      "grad_norm": 0.7949123978614807,
+      "learning_rate": 0.000259727413235914,
+      "loss": 4.8444,
+      "step": 61640
+    },
+    {
+      "epoch": 0.08662403377845534,
+      "grad_norm": 0.7871426939964294,
+      "learning_rate": 0.0002597695658282984,
+      "loss": 4.8069,
+      "step": 61650
+    },
+    {
+      "epoch": 0.08663808471661892,
+      "grad_norm": 0.8564624786376953,
+      "learning_rate": 0.00025981171842068284,
+      "loss": 4.863,
+      "step": 61660
+    },
+    {
+      "epoch": 0.0866521356547825,
+      "grad_norm": 0.8065398931503296,
+      "learning_rate": 0.0002598538710130673,
+      "loss": 4.9085,
+      "step": 61670
+    },
+    {
+      "epoch": 0.08666618659294607,
+      "grad_norm": 0.79127436876297,
+      "learning_rate": 0.0002598960236054517,
+      "loss": 4.7051,
+      "step": 61680
+    },
+    {
+      "epoch": 0.08668023753110965,
+      "grad_norm": 0.7989571690559387,
+      "learning_rate": 0.00025993817619783614,
+      "loss": 4.6685,
+      "step": 61690
+    },
+    {
+      "epoch": 0.08669428846927324,
+      "grad_norm": 0.794864296913147,
+      "learning_rate": 0.00025998032879022057,
+      "loss": 4.8343,
+      "step": 61700
+    },
+    {
+      "epoch": 0.08670833940743682,
+      "grad_norm": 0.803548276424408,
+      "learning_rate": 0.000260022481382605,
+      "loss": 4.7731,
+      "step": 61710
+    },
+    {
+      "epoch": 0.08672239034560039,
+      "grad_norm": 0.8012824654579163,
+      "learning_rate": 0.00026006463397498943,
+      "loss": 4.8466,
+      "step": 61720
+    },
+    {
+      "epoch": 0.08673644128376397,
+      "grad_norm": 0.792219340801239,
+      "learning_rate": 0.00026010678656737386,
+      "loss": 4.7721,
+      "step": 61730
+    },
+    {
+      "epoch": 0.08675049222192754,
+      "grad_norm": 0.7680630087852478,
+      "learning_rate": 0.0002601489391597583,
+      "loss": 4.7905,
+      "step": 61740
+    },
+    {
+      "epoch": 0.08676454316009112,
+      "grad_norm": 0.7854152917861938,
+      "learning_rate": 0.0002601910917521427,
+      "loss": 4.7413,
+      "step": 61750
+    },
+    {
+      "epoch": 0.0867785940982547,
+      "grad_norm": 0.8030341267585754,
+      "learning_rate": 0.00026023324434452716,
+      "loss": 4.8224,
+      "step": 61760
+    },
+    {
+      "epoch": 0.08679264503641827,
+      "grad_norm": 0.807295560836792,
+      "learning_rate": 0.0002602753969369116,
+      "loss": 4.8628,
+      "step": 61770
+    },
+    {
+      "epoch": 0.08680669597458185,
+      "grad_norm": 0.7911593317985535,
+      "learning_rate": 0.000260317549529296,
+      "loss": 4.6736,
+      "step": 61780
+    },
+    {
+      "epoch": 0.08682074691274543,
+      "grad_norm": 0.8324936032295227,
+      "learning_rate": 0.00026035970212168045,
+      "loss": 4.8149,
+      "step": 61790
+    },
+    {
+      "epoch": 0.086834797850909,
+      "grad_norm": 0.8072502017021179,
+      "learning_rate": 0.0002604018547140649,
+      "loss": 4.8918,
+      "step": 61800
+    },
+    {
+      "epoch": 0.08684884878907259,
+      "grad_norm": 0.8104373216629028,
+      "learning_rate": 0.0002604440073064493,
+      "loss": 4.8896,
+      "step": 61810
+    },
+    {
+      "epoch": 0.08686289972723617,
+      "grad_norm": 1.0080177783966064,
+      "learning_rate": 0.00026048615989883374,
+      "loss": 4.8292,
+      "step": 61820
+    },
+    {
+      "epoch": 0.08687695066539974,
+      "grad_norm": 0.8252009749412537,
+      "learning_rate": 0.0002605283124912182,
+      "loss": 4.7111,
+      "step": 61830
+    },
+    {
+      "epoch": 0.08689100160356332,
+      "grad_norm": 0.8061153888702393,
+      "learning_rate": 0.0002605704650836026,
+      "loss": 4.781,
+      "step": 61840
+    },
+    {
+      "epoch": 0.0869050525417269,
+      "grad_norm": 0.8102174401283264,
+      "learning_rate": 0.00026061261767598704,
+      "loss": 4.8626,
+      "step": 61850
+    },
+    {
+      "epoch": 0.08691910347989047,
+      "grad_norm": 0.8249738812446594,
+      "learning_rate": 0.00026065477026837147,
+      "loss": 4.9965,
+      "step": 61860
+    },
+    {
+      "epoch": 0.08693315441805405,
+      "grad_norm": 0.8087854385375977,
+      "learning_rate": 0.00026069692286075595,
+      "loss": 4.8119,
+      "step": 61870
+    },
+    {
+      "epoch": 0.08694720535621762,
+      "grad_norm": 0.7888401746749878,
+      "learning_rate": 0.00026073907545314033,
+      "loss": 4.7734,
+      "step": 61880
+    },
+    {
+      "epoch": 0.0869612562943812,
+      "grad_norm": 0.7908345460891724,
+      "learning_rate": 0.00026078122804552476,
+      "loss": 4.792,
+      "step": 61890
+    },
+    {
+      "epoch": 0.08697530723254478,
+      "grad_norm": 0.8059920072555542,
+      "learning_rate": 0.0002608233806379092,
+      "loss": 4.8439,
+      "step": 61900
+    },
+    {
+      "epoch": 0.08698935817070837,
+      "grad_norm": 0.8107774257659912,
+      "learning_rate": 0.0002608655332302936,
+      "loss": 4.8511,
+      "step": 61910
+    },
+    {
+      "epoch": 0.08700340910887194,
+      "grad_norm": 0.7921923995018005,
+      "learning_rate": 0.00026090768582267805,
+      "loss": 4.6943,
+      "step": 61920
+    },
+    {
+      "epoch": 0.08701746004703552,
+      "grad_norm": 0.8148934245109558,
+      "learning_rate": 0.00026094983841506254,
+      "loss": 4.7229,
+      "step": 61930
+    },
+    {
+      "epoch": 0.0870315109851991,
+      "grad_norm": 0.7990559935569763,
+      "learning_rate": 0.0002609919910074469,
+      "loss": 4.7258,
+      "step": 61940
+    },
+    {
+      "epoch": 0.08704556192336267,
+      "grad_norm": 0.7900142073631287,
+      "learning_rate": 0.00026103414359983135,
+      "loss": 4.7876,
+      "step": 61950
+    },
+    {
+      "epoch": 0.08705961286152625,
+      "grad_norm": 0.7979053258895874,
+      "learning_rate": 0.0002610762961922158,
+      "loss": 4.7788,
+      "step": 61960
+    },
+    {
+      "epoch": 0.08707366379968982,
+      "grad_norm": 0.8106703162193298,
+      "learning_rate": 0.0002611184487846002,
+      "loss": 4.7476,
+      "step": 61970
+    },
+    {
+      "epoch": 0.0870877147378534,
+      "grad_norm": 0.8180921673774719,
+      "learning_rate": 0.00026116060137698464,
+      "loss": 4.6959,
+      "step": 61980
+    },
+    {
+      "epoch": 0.08710176567601698,
+      "grad_norm": 0.7843952178955078,
+      "learning_rate": 0.0002612027539693691,
+      "loss": 4.7037,
+      "step": 61990
+    },
+    {
+      "epoch": 0.08711581661418055,
+      "grad_norm": 0.7980206608772278,
+      "learning_rate": 0.0002612449065617535,
+      "loss": 4.8514,
+      "step": 62000
+    },
+    {
+      "epoch": 0.08712986755234414,
+      "grad_norm": 0.8227792382240295,
+      "learning_rate": 0.00026128705915413793,
+      "loss": 4.7442,
+      "step": 62010
+    },
+    {
+      "epoch": 0.08714391849050772,
+      "grad_norm": 0.8404500484466553,
+      "learning_rate": 0.00026132921174652237,
+      "loss": 4.89,
+      "step": 62020
+    },
+    {
+      "epoch": 0.0871579694286713,
+      "grad_norm": 0.8126966953277588,
+      "learning_rate": 0.0002613713643389068,
+      "loss": 4.8632,
+      "step": 62030
+    },
+    {
+      "epoch": 0.08717202036683487,
+      "grad_norm": 0.817398726940155,
+      "learning_rate": 0.0002614135169312913,
+      "loss": 4.8315,
+      "step": 62040
+    },
+    {
+      "epoch": 0.08718607130499845,
+      "grad_norm": 0.81786048412323,
+      "learning_rate": 0.0002614556695236757,
+      "loss": 4.7713,
+      "step": 62050
+    },
+    {
+      "epoch": 0.08720012224316202,
+      "grad_norm": 0.8178216814994812,
+      "learning_rate": 0.00026149782211606014,
+      "loss": 4.7856,
+      "step": 62060
+    },
+    {
+      "epoch": 0.0872141731813256,
+      "grad_norm": 0.8115580081939697,
+      "learning_rate": 0.0002615399747084445,
+      "loss": 4.7314,
+      "step": 62070
+    },
+    {
+      "epoch": 0.08722822411948918,
+      "grad_norm": 0.7866354584693909,
+      "learning_rate": 0.00026158212730082895,
+      "loss": 4.839,
+      "step": 62080
+    },
+    {
+      "epoch": 0.08724227505765275,
+      "grad_norm": 0.808581531047821,
+      "learning_rate": 0.0002616242798932134,
+      "loss": 4.8308,
+      "step": 62090
+    },
+    {
+      "epoch": 0.08725632599581633,
+      "grad_norm": 0.7981185913085938,
+      "learning_rate": 0.00026166643248559787,
+      "loss": 4.7796,
+      "step": 62100
+    },
+    {
+      "epoch": 0.0872703769339799,
+      "grad_norm": 0.8379929065704346,
+      "learning_rate": 0.0002617085850779823,
+      "loss": 4.737,
+      "step": 62110
+    },
+    {
+      "epoch": 0.0872844278721435,
+      "grad_norm": 0.8175838589668274,
+      "learning_rate": 0.00026175073767036673,
+      "loss": 4.8416,
+      "step": 62120
+    },
+    {
+      "epoch": 0.08729847881030707,
+      "grad_norm": 0.8132398724555969,
+      "learning_rate": 0.0002617928902627511,
+      "loss": 4.8088,
+      "step": 62130
+    },
+    {
+      "epoch": 0.08731252974847065,
+      "grad_norm": 0.8097195625305176,
+      "learning_rate": 0.00026183504285513554,
+      "loss": 4.8075,
+      "step": 62140
+    },
+    {
+      "epoch": 0.08732658068663422,
+      "grad_norm": 0.8951746225357056,
+      "learning_rate": 0.00026187719544751997,
+      "loss": 4.8201,
+      "step": 62150
+    },
+    {
+      "epoch": 0.0873406316247978,
+      "grad_norm": 0.8372669219970703,
+      "learning_rate": 0.00026191934803990446,
+      "loss": 4.8856,
+      "step": 62160
+    },
+    {
+      "epoch": 0.08735468256296137,
+      "grad_norm": 0.7923863530158997,
+      "learning_rate": 0.0002619615006322889,
+      "loss": 4.8744,
+      "step": 62170
+    },
+    {
+      "epoch": 0.08736873350112495,
+      "grad_norm": 0.7726845741271973,
+      "learning_rate": 0.0002620036532246733,
+      "loss": 4.8476,
+      "step": 62180
+    },
+    {
+      "epoch": 0.08738278443928853,
+      "grad_norm": 0.7641924619674683,
+      "learning_rate": 0.00026204580581705775,
+      "loss": 4.8731,
+      "step": 62190
+    },
+    {
+      "epoch": 0.0873968353774521,
+      "grad_norm": 0.8116229772567749,
+      "learning_rate": 0.00026208795840944213,
+      "loss": 4.8149,
+      "step": 62200
+    },
+    {
+      "epoch": 0.08741088631561568,
+      "grad_norm": 0.7808107137680054,
+      "learning_rate": 0.00026213011100182656,
+      "loss": 4.7596,
+      "step": 62210
+    },
+    {
+      "epoch": 0.08742493725377927,
+      "grad_norm": 0.7787986397743225,
+      "learning_rate": 0.00026217226359421104,
+      "loss": 4.7875,
+      "step": 62220
+    },
+    {
+      "epoch": 0.08743898819194285,
+      "grad_norm": 0.7925445437431335,
+      "learning_rate": 0.0002622144161865955,
+      "loss": 4.7452,
+      "step": 62230
+    },
+    {
+      "epoch": 0.08745303913010642,
+      "grad_norm": 0.8389007449150085,
+      "learning_rate": 0.0002622565687789799,
+      "loss": 4.7547,
+      "step": 62240
+    },
+    {
+      "epoch": 0.08746709006827,
+      "grad_norm": 0.7930507659912109,
+      "learning_rate": 0.00026229872137136434,
+      "loss": 4.7961,
+      "step": 62250
+    },
+    {
+      "epoch": 0.08748114100643357,
+      "grad_norm": 0.7937973737716675,
+      "learning_rate": 0.0002623408739637487,
+      "loss": 4.7547,
+      "step": 62260
+    },
+    {
+      "epoch": 0.08749519194459715,
+      "grad_norm": 0.797887921333313,
+      "learning_rate": 0.00026238302655613315,
+      "loss": 4.9243,
+      "step": 62270
+    },
+    {
+      "epoch": 0.08750924288276073,
+      "grad_norm": 0.7737236618995667,
+      "learning_rate": 0.00026242517914851763,
+      "loss": 4.8146,
+      "step": 62280
+    },
+    {
+      "epoch": 0.0875232938209243,
+      "grad_norm": 0.8101477026939392,
+      "learning_rate": 0.00026246733174090206,
+      "loss": 4.7729,
+      "step": 62290
+    },
+    {
+      "epoch": 0.08753734475908788,
+      "grad_norm": 0.8078815340995789,
+      "learning_rate": 0.0002625094843332865,
+      "loss": 4.6863,
+      "step": 62300
+    },
+    {
+      "epoch": 0.08755139569725146,
+      "grad_norm": 0.7906801700592041,
+      "learning_rate": 0.0002625516369256709,
+      "loss": 4.7801,
+      "step": 62310
+    },
+    {
+      "epoch": 0.08756544663541505,
+      "grad_norm": 0.8077771663665771,
+      "learning_rate": 0.00026259378951805536,
+      "loss": 4.7214,
+      "step": 62320
+    },
+    {
+      "epoch": 0.08757949757357862,
+      "grad_norm": 0.8051345348358154,
+      "learning_rate": 0.00026263594211043973,
+      "loss": 4.8615,
+      "step": 62330
+    },
+    {
+      "epoch": 0.0875935485117422,
+      "grad_norm": 0.798250675201416,
+      "learning_rate": 0.0002626780947028242,
+      "loss": 4.7257,
+      "step": 62340
+    },
+    {
+      "epoch": 0.08760759944990577,
+      "grad_norm": 0.7771165370941162,
+      "learning_rate": 0.00026272024729520865,
+      "loss": 4.7493,
+      "step": 62350
+    },
+    {
+      "epoch": 0.08762165038806935,
+      "grad_norm": 0.7755945920944214,
+      "learning_rate": 0.0002627623998875931,
+      "loss": 4.9485,
+      "step": 62360
+    },
+    {
+      "epoch": 0.08763570132623293,
+      "grad_norm": 0.7809829115867615,
+      "learning_rate": 0.0002628045524799775,
+      "loss": 4.7587,
+      "step": 62370
+    },
+    {
+      "epoch": 0.0876497522643965,
+      "grad_norm": 0.80907142162323,
+      "learning_rate": 0.00026284670507236194,
+      "loss": 4.7375,
+      "step": 62380
+    },
+    {
+      "epoch": 0.08766380320256008,
+      "grad_norm": 0.8076225519180298,
+      "learning_rate": 0.0002628888576647463,
+      "loss": 4.898,
+      "step": 62390
+    },
+    {
+      "epoch": 0.08767785414072365,
+      "grad_norm": 0.7816614508628845,
+      "learning_rate": 0.0002629310102571308,
+      "loss": 4.689,
+      "step": 62400
+    },
+    {
+      "epoch": 0.08769190507888723,
+      "grad_norm": 0.8814823031425476,
+      "learning_rate": 0.00026297316284951524,
+      "loss": 4.7298,
+      "step": 62410
+    },
+    {
+      "epoch": 0.0877059560170508,
+      "grad_norm": 0.8320509791374207,
+      "learning_rate": 0.00026301531544189967,
+      "loss": 4.7035,
+      "step": 62420
+    },
+    {
+      "epoch": 0.0877200069552144,
+      "grad_norm": 0.8074789643287659,
+      "learning_rate": 0.0002630574680342841,
+      "loss": 4.8431,
+      "step": 62430
+    },
+    {
+      "epoch": 0.08773405789337797,
+      "grad_norm": 0.7870790362358093,
+      "learning_rate": 0.00026309962062666853,
+      "loss": 4.8224,
+      "step": 62440
+    },
+    {
+      "epoch": 0.08774810883154155,
+      "grad_norm": 0.8190731406211853,
+      "learning_rate": 0.00026314177321905296,
+      "loss": 4.7692,
+      "step": 62450
+    },
+    {
+      "epoch": 0.08776215976970513,
+      "grad_norm": 0.8550110459327698,
+      "learning_rate": 0.0002631839258114374,
+      "loss": 4.8563,
+      "step": 62460
+    },
+    {
+      "epoch": 0.0877762107078687,
+      "grad_norm": 0.7676360607147217,
+      "learning_rate": 0.0002632260784038218,
+      "loss": 4.7988,
+      "step": 62470
+    },
+    {
+      "epoch": 0.08779026164603228,
+      "grad_norm": 0.8471920490264893,
+      "learning_rate": 0.00026326823099620625,
+      "loss": 4.7623,
+      "step": 62480
+    },
+    {
+      "epoch": 0.08780431258419585,
+      "grad_norm": 0.7891804575920105,
+      "learning_rate": 0.0002633103835885907,
+      "loss": 4.8314,
+      "step": 62490
+    },
+    {
+      "epoch": 0.08781836352235943,
+      "grad_norm": 0.8159024119377136,
+      "learning_rate": 0.0002633525361809751,
+      "loss": 4.8448,
+      "step": 62500
+    },
+    {
+      "epoch": 0.087832414460523,
+      "grad_norm": 0.7949129939079285,
+      "learning_rate": 0.00026339468877335955,
+      "loss": 4.7038,
+      "step": 62510
+    },
+    {
+      "epoch": 0.08784646539868658,
+      "grad_norm": 0.8131535649299622,
+      "learning_rate": 0.000263436841365744,
+      "loss": 4.8487,
+      "step": 62520
+    },
+    {
+      "epoch": 0.08786051633685017,
+      "grad_norm": 0.856124222278595,
+      "learning_rate": 0.0002634789939581284,
+      "loss": 4.7969,
+      "step": 62530
+    },
+    {
+      "epoch": 0.08787456727501375,
+      "grad_norm": 0.7902341485023499,
+      "learning_rate": 0.00026352114655051284,
+      "loss": 4.7348,
+      "step": 62540
+    },
+    {
+      "epoch": 0.08788861821317732,
+      "grad_norm": 0.7894250154495239,
+      "learning_rate": 0.00026356329914289727,
+      "loss": 4.7934,
+      "step": 62550
+    },
+    {
+      "epoch": 0.0879026691513409,
+      "grad_norm": 0.7873814702033997,
+      "learning_rate": 0.0002636054517352817,
+      "loss": 4.8012,
+      "step": 62560
+    },
+    {
+      "epoch": 0.08791672008950448,
+      "grad_norm": 0.8066341876983643,
+      "learning_rate": 0.00026364760432766613,
+      "loss": 4.8533,
+      "step": 62570
+    },
+    {
+      "epoch": 0.08793077102766805,
+      "grad_norm": 0.7610378861427307,
+      "learning_rate": 0.00026368975692005057,
+      "loss": 4.7305,
+      "step": 62580
+    },
+    {
+      "epoch": 0.08794482196583163,
+      "grad_norm": 0.8013112545013428,
+      "learning_rate": 0.000263731909512435,
+      "loss": 4.7925,
+      "step": 62590
+    },
+    {
+      "epoch": 0.0879588729039952,
+      "grad_norm": 0.7849205732345581,
+      "learning_rate": 0.00026377406210481943,
+      "loss": 4.793,
+      "step": 62600
+    },
+    {
+      "epoch": 0.08797292384215878,
+      "grad_norm": 0.8190938234329224,
+      "learning_rate": 0.00026381621469720386,
+      "loss": 4.7288,
+      "step": 62610
+    },
+    {
+      "epoch": 0.08798697478032236,
+      "grad_norm": 0.8454349040985107,
+      "learning_rate": 0.0002638583672895883,
+      "loss": 4.8616,
+      "step": 62620
+    },
+    {
+      "epoch": 0.08800102571848595,
+      "grad_norm": 0.7780125737190247,
+      "learning_rate": 0.0002639005198819727,
+      "loss": 4.7913,
+      "step": 62630
+    },
+    {
+      "epoch": 0.08801507665664952,
+      "grad_norm": 0.8072920441627502,
+      "learning_rate": 0.00026394267247435715,
+      "loss": 4.8829,
+      "step": 62640
+    },
+    {
+      "epoch": 0.0880291275948131,
+      "grad_norm": 0.8076897263526917,
+      "learning_rate": 0.0002639848250667416,
+      "loss": 4.7353,
+      "step": 62650
+    },
+    {
+      "epoch": 0.08804317853297668,
+      "grad_norm": 0.7699461579322815,
+      "learning_rate": 0.000264026977659126,
+      "loss": 4.8364,
+      "step": 62660
+    },
+    {
+      "epoch": 0.08805722947114025,
+      "grad_norm": 0.8059861063957214,
+      "learning_rate": 0.00026406913025151045,
+      "loss": 4.779,
+      "step": 62670
+    },
+    {
+      "epoch": 0.08807128040930383,
+      "grad_norm": 0.7961968779563904,
+      "learning_rate": 0.0002641112828438949,
+      "loss": 4.8785,
+      "step": 62680
+    },
+    {
+      "epoch": 0.0880853313474674,
+      "grad_norm": 0.8171684741973877,
+      "learning_rate": 0.0002641534354362793,
+      "loss": 4.7068,
+      "step": 62690
+    },
+    {
+      "epoch": 0.08809938228563098,
+      "grad_norm": 0.7811867594718933,
+      "learning_rate": 0.00026419558802866374,
+      "loss": 4.971,
+      "step": 62700
+    },
+    {
+      "epoch": 0.08811343322379456,
+      "grad_norm": 0.7805197834968567,
+      "learning_rate": 0.00026423774062104817,
+      "loss": 4.9694,
+      "step": 62710
+    },
+    {
+      "epoch": 0.08812748416195813,
+      "grad_norm": 0.8082363605499268,
+      "learning_rate": 0.0002642798932134326,
+      "loss": 4.7084,
+      "step": 62720
+    },
+    {
+      "epoch": 0.08814153510012171,
+      "grad_norm": 0.8304884433746338,
+      "learning_rate": 0.00026432204580581703,
+      "loss": 4.7155,
+      "step": 62730
+    },
+    {
+      "epoch": 0.0881555860382853,
+      "grad_norm": 0.8257015943527222,
+      "learning_rate": 0.00026436419839820146,
+      "loss": 4.8118,
+      "step": 62740
+    },
+    {
+      "epoch": 0.08816963697644888,
+      "grad_norm": 0.7988607883453369,
+      "learning_rate": 0.0002644063509905859,
+      "loss": 4.7102,
+      "step": 62750
+    },
+    {
+      "epoch": 0.08818368791461245,
+      "grad_norm": 0.7881811857223511,
+      "learning_rate": 0.0002644485035829703,
+      "loss": 4.8997,
+      "step": 62760
+    },
+    {
+      "epoch": 0.08819773885277603,
+      "grad_norm": 0.859298825263977,
+      "learning_rate": 0.00026449065617535476,
+      "loss": 4.8189,
+      "step": 62770
+    },
+    {
+      "epoch": 0.0882117897909396,
+      "grad_norm": 0.781113862991333,
+      "learning_rate": 0.0002645328087677392,
+      "loss": 4.7552,
+      "step": 62780
+    },
+    {
+      "epoch": 0.08822584072910318,
+      "grad_norm": 0.7915051579475403,
+      "learning_rate": 0.0002645749613601236,
+      "loss": 4.7368,
+      "step": 62790
+    },
+    {
+      "epoch": 0.08823989166726676,
+      "grad_norm": 0.8124852180480957,
+      "learning_rate": 0.00026461711395250805,
+      "loss": 4.8497,
+      "step": 62800
+    },
+    {
+      "epoch": 0.08825394260543033,
+      "grad_norm": 0.8054794073104858,
+      "learning_rate": 0.0002646592665448925,
+      "loss": 4.7871,
+      "step": 62810
+    },
+    {
+      "epoch": 0.08826799354359391,
+      "grad_norm": 0.8467390537261963,
+      "learning_rate": 0.0002647014191372769,
+      "loss": 4.6806,
+      "step": 62820
+    },
+    {
+      "epoch": 0.08828204448175749,
+      "grad_norm": 0.8043726682662964,
+      "learning_rate": 0.00026474357172966134,
+      "loss": 4.8889,
+      "step": 62830
+    },
+    {
+      "epoch": 0.08829609541992108,
+      "grad_norm": 0.7874152660369873,
+      "learning_rate": 0.0002647857243220458,
+      "loss": 4.8625,
+      "step": 62840
+    },
+    {
+      "epoch": 0.08831014635808465,
+      "grad_norm": 0.8490023016929626,
+      "learning_rate": 0.0002648278769144302,
+      "loss": 4.7956,
+      "step": 62850
+    },
+    {
+      "epoch": 0.08832419729624823,
+      "grad_norm": 0.8019155263900757,
+      "learning_rate": 0.00026487002950681464,
+      "loss": 4.7038,
+      "step": 62860
+    },
+    {
+      "epoch": 0.0883382482344118,
+      "grad_norm": 0.7915338277816772,
+      "learning_rate": 0.00026491218209919907,
+      "loss": 4.8408,
+      "step": 62870
+    },
+    {
+      "epoch": 0.08835229917257538,
+      "grad_norm": 0.7774483561515808,
+      "learning_rate": 0.0002649543346915835,
+      "loss": 4.8746,
+      "step": 62880
+    },
+    {
+      "epoch": 0.08836635011073896,
+      "grad_norm": 0.7984682321548462,
+      "learning_rate": 0.00026499648728396793,
+      "loss": 4.7794,
+      "step": 62890
+    },
+    {
+      "epoch": 0.08838040104890253,
+      "grad_norm": 0.803041398525238,
+      "learning_rate": 0.00026503863987635236,
+      "loss": 4.7,
+      "step": 62900
+    },
+    {
+      "epoch": 0.08839445198706611,
+      "grad_norm": 0.7951576113700867,
+      "learning_rate": 0.0002650807924687368,
+      "loss": 4.7453,
+      "step": 62910
+    },
+    {
+      "epoch": 0.08840850292522968,
+      "grad_norm": 0.7813413143157959,
+      "learning_rate": 0.0002651229450611212,
+      "loss": 4.7845,
+      "step": 62920
+    },
+    {
+      "epoch": 0.08842255386339326,
+      "grad_norm": 0.7739083170890808,
+      "learning_rate": 0.00026516509765350566,
+      "loss": 4.7668,
+      "step": 62930
+    },
+    {
+      "epoch": 0.08843660480155685,
+      "grad_norm": 0.8090055584907532,
+      "learning_rate": 0.0002652072502458901,
+      "loss": 4.8,
+      "step": 62940
+    },
+    {
+      "epoch": 0.08845065573972043,
+      "grad_norm": 0.7967948317527771,
+      "learning_rate": 0.0002652494028382745,
+      "loss": 4.9125,
+      "step": 62950
+    },
+    {
+      "epoch": 0.088464706677884,
+      "grad_norm": 0.7919458150863647,
+      "learning_rate": 0.00026529155543065895,
+      "loss": 4.8298,
+      "step": 62960
+    },
+    {
+      "epoch": 0.08847875761604758,
+      "grad_norm": 0.7871581315994263,
+      "learning_rate": 0.0002653337080230434,
+      "loss": 4.8399,
+      "step": 62970
+    },
+    {
+      "epoch": 0.08849280855421116,
+      "grad_norm": 0.7785304188728333,
+      "learning_rate": 0.0002653758606154278,
+      "loss": 4.856,
+      "step": 62980
+    },
+    {
+      "epoch": 0.08850685949237473,
+      "grad_norm": 0.7989861369132996,
+      "learning_rate": 0.00026541801320781224,
+      "loss": 4.7617,
+      "step": 62990
+    },
+    {
+      "epoch": 0.08852091043053831,
+      "grad_norm": 0.7797976732254028,
+      "learning_rate": 0.0002654601658001967,
+      "loss": 4.9164,
+      "step": 63000
+    },
+    {
+      "epoch": 0.08853496136870188,
+      "grad_norm": 0.766815721988678,
+      "learning_rate": 0.0002655023183925811,
+      "loss": 4.809,
+      "step": 63010
+    },
+    {
+      "epoch": 0.08854901230686546,
+      "grad_norm": 0.8306536078453064,
+      "learning_rate": 0.00026554447098496554,
+      "loss": 4.8476,
+      "step": 63020
+    },
+    {
+      "epoch": 0.08856306324502904,
+      "grad_norm": 0.8039864897727966,
+      "learning_rate": 0.00026558662357735,
+      "loss": 4.7684,
+      "step": 63030
+    },
+    {
+      "epoch": 0.08857711418319261,
+      "grad_norm": 0.8326762914657593,
+      "learning_rate": 0.0002656287761697344,
+      "loss": 4.7807,
+      "step": 63040
+    },
+    {
+      "epoch": 0.0885911651213562,
+      "grad_norm": 0.7910907864570618,
+      "learning_rate": 0.00026567092876211883,
+      "loss": 4.7499,
+      "step": 63050
+    },
+    {
+      "epoch": 0.08860521605951978,
+      "grad_norm": 0.7897930145263672,
+      "learning_rate": 0.00026571308135450326,
+      "loss": 4.8467,
+      "step": 63060
+    },
+    {
+      "epoch": 0.08861926699768335,
+      "grad_norm": 0.7836660146713257,
+      "learning_rate": 0.0002657552339468877,
+      "loss": 4.7958,
+      "step": 63070
+    },
+    {
+      "epoch": 0.08863331793584693,
+      "grad_norm": 0.7851306796073914,
+      "learning_rate": 0.0002657973865392721,
+      "loss": 4.7329,
+      "step": 63080
+    },
+    {
+      "epoch": 0.08864736887401051,
+      "grad_norm": 0.7947384715080261,
+      "learning_rate": 0.0002658395391316566,
+      "loss": 4.8723,
+      "step": 63090
+    },
+    {
+      "epoch": 0.08866141981217408,
+      "grad_norm": 0.8200597167015076,
+      "learning_rate": 0.000265881691724041,
+      "loss": 4.8698,
+      "step": 63100
+    },
+    {
+      "epoch": 0.08867547075033766,
+      "grad_norm": 0.7728275060653687,
+      "learning_rate": 0.0002659238443164254,
+      "loss": 4.7902,
+      "step": 63110
+    },
+    {
+      "epoch": 0.08868952168850124,
+      "grad_norm": 0.7961565852165222,
+      "learning_rate": 0.00026596599690880985,
+      "loss": 4.736,
+      "step": 63120
+    },
+    {
+      "epoch": 0.08870357262666481,
+      "grad_norm": 0.7639411091804504,
+      "learning_rate": 0.0002660081495011943,
+      "loss": 4.9944,
+      "step": 63130
+    },
+    {
+      "epoch": 0.08871762356482839,
+      "grad_norm": 0.8078753352165222,
+      "learning_rate": 0.0002660503020935787,
+      "loss": 4.7914,
+      "step": 63140
+    },
+    {
+      "epoch": 0.08873167450299198,
+      "grad_norm": 0.8076030611991882,
+      "learning_rate": 0.0002660924546859632,
+      "loss": 4.9144,
+      "step": 63150
+    },
+    {
+      "epoch": 0.08874572544115555,
+      "grad_norm": 0.8135642409324646,
+      "learning_rate": 0.00026613460727834763,
+      "loss": 4.7739,
+      "step": 63160
+    },
+    {
+      "epoch": 0.08875977637931913,
+      "grad_norm": 0.81586754322052,
+      "learning_rate": 0.000266176759870732,
+      "loss": 4.7433,
+      "step": 63170
+    },
+    {
+      "epoch": 0.0887738273174827,
+      "grad_norm": 0.8089296817779541,
+      "learning_rate": 0.00026621891246311644,
+      "loss": 4.8924,
+      "step": 63180
+    },
+    {
+      "epoch": 0.08878787825564628,
+      "grad_norm": 0.7918955683708191,
+      "learning_rate": 0.00026626106505550087,
+      "loss": 4.7475,
+      "step": 63190
+    },
+    {
+      "epoch": 0.08880192919380986,
+      "grad_norm": 0.7879622578620911,
+      "learning_rate": 0.00026630321764788535,
+      "loss": 4.8674,
+      "step": 63200
+    },
+    {
+      "epoch": 0.08881598013197343,
+      "grad_norm": 0.7970558404922485,
+      "learning_rate": 0.0002663453702402698,
+      "loss": 4.8004,
+      "step": 63210
+    },
+    {
+      "epoch": 0.08883003107013701,
+      "grad_norm": 0.7629973292350769,
+      "learning_rate": 0.0002663875228326542,
+      "loss": 4.715,
+      "step": 63220
+    },
+    {
+      "epoch": 0.08884408200830059,
+      "grad_norm": 0.8390489816665649,
+      "learning_rate": 0.0002664296754250386,
+      "loss": 4.7549,
+      "step": 63230
+    },
+    {
+      "epoch": 0.08885813294646416,
+      "grad_norm": 0.7946515679359436,
+      "learning_rate": 0.000266471828017423,
+      "loss": 4.846,
+      "step": 63240
+    },
+    {
+      "epoch": 0.08887218388462775,
+      "grad_norm": 0.790104866027832,
+      "learning_rate": 0.00026651398060980745,
+      "loss": 5.0164,
+      "step": 63250
+    },
+    {
+      "epoch": 0.08888623482279133,
+      "grad_norm": 0.7630163431167603,
+      "learning_rate": 0.00026655613320219194,
+      "loss": 4.895,
+      "step": 63260
+    },
+    {
+      "epoch": 0.0889002857609549,
+      "grad_norm": 0.8079057931900024,
+      "learning_rate": 0.00026659828579457637,
+      "loss": 4.8883,
+      "step": 63270
+    },
+    {
+      "epoch": 0.08891433669911848,
+      "grad_norm": 0.7824141383171082,
+      "learning_rate": 0.0002666404383869608,
+      "loss": 4.7367,
+      "step": 63280
+    },
+    {
+      "epoch": 0.08892838763728206,
+      "grad_norm": 1.028652310371399,
+      "learning_rate": 0.0002666825909793452,
+      "loss": 4.8906,
+      "step": 63290
+    },
+    {
+      "epoch": 0.08894243857544563,
+      "grad_norm": 0.8163354396820068,
+      "learning_rate": 0.0002667247435717296,
+      "loss": 4.8534,
+      "step": 63300
+    },
+    {
+      "epoch": 0.08895648951360921,
+      "grad_norm": 0.7730701565742493,
+      "learning_rate": 0.00026676689616411404,
+      "loss": 4.7245,
+      "step": 63310
+    },
+    {
+      "epoch": 0.08897054045177279,
+      "grad_norm": 0.7872650623321533,
+      "learning_rate": 0.0002668090487564985,
+      "loss": 4.799,
+      "step": 63320
+    },
+    {
+      "epoch": 0.08898459138993636,
+      "grad_norm": 0.7956480979919434,
+      "learning_rate": 0.00026685120134888296,
+      "loss": 4.8097,
+      "step": 63330
+    },
+    {
+      "epoch": 0.08899864232809994,
+      "grad_norm": 0.7865951061248779,
+      "learning_rate": 0.0002668933539412674,
+      "loss": 4.9028,
+      "step": 63340
+    },
+    {
+      "epoch": 0.08901269326626352,
+      "grad_norm": 0.7755113244056702,
+      "learning_rate": 0.0002669355065336518,
+      "loss": 4.8193,
+      "step": 63350
+    },
+    {
+      "epoch": 0.0890267442044271,
+      "grad_norm": 0.7839175462722778,
+      "learning_rate": 0.0002669776591260362,
+      "loss": 4.7572,
+      "step": 63360
+    },
+    {
+      "epoch": 0.08904079514259068,
+      "grad_norm": 0.801013708114624,
+      "learning_rate": 0.00026701981171842063,
+      "loss": 4.7782,
+      "step": 63370
+    },
+    {
+      "epoch": 0.08905484608075426,
+      "grad_norm": 0.8248993754386902,
+      "learning_rate": 0.0002670619643108051,
+      "loss": 4.8563,
+      "step": 63380
+    },
+    {
+      "epoch": 0.08906889701891783,
+      "grad_norm": 0.8062582612037659,
+      "learning_rate": 0.00026710411690318954,
+      "loss": 4.8467,
+      "step": 63390
+    },
+    {
+      "epoch": 0.08908294795708141,
+      "grad_norm": 0.7893478274345398,
+      "learning_rate": 0.000267146269495574,
+      "loss": 4.8678,
+      "step": 63400
+    },
+    {
+      "epoch": 0.08909699889524499,
+      "grad_norm": 0.8045288920402527,
+      "learning_rate": 0.0002671884220879584,
+      "loss": 4.7405,
+      "step": 63410
+    },
+    {
+      "epoch": 0.08911104983340856,
+      "grad_norm": 0.7915014028549194,
+      "learning_rate": 0.0002672305746803428,
+      "loss": 4.7565,
+      "step": 63420
+    },
+    {
+      "epoch": 0.08912510077157214,
+      "grad_norm": 0.8055747747421265,
+      "learning_rate": 0.0002672727272727272,
+      "loss": 4.8634,
+      "step": 63430
+    },
+    {
+      "epoch": 0.08913915170973571,
+      "grad_norm": 0.8131148815155029,
+      "learning_rate": 0.0002673148798651117,
+      "loss": 4.7419,
+      "step": 63440
+    },
+    {
+      "epoch": 0.08915320264789929,
+      "grad_norm": 0.7881912589073181,
+      "learning_rate": 0.00026735703245749613,
+      "loss": 4.8812,
+      "step": 63450
+    },
+    {
+      "epoch": 0.08916725358606288,
+      "grad_norm": 0.7696599960327148,
+      "learning_rate": 0.00026739918504988056,
+      "loss": 4.7284,
+      "step": 63460
+    },
+    {
+      "epoch": 0.08918130452422646,
+      "grad_norm": 0.7833225727081299,
+      "learning_rate": 0.000267441337642265,
+      "loss": 4.8861,
+      "step": 63470
+    },
+    {
+      "epoch": 0.08919535546239003,
+      "grad_norm": 0.7945125102996826,
+      "learning_rate": 0.0002674834902346494,
+      "loss": 4.8175,
+      "step": 63480
+    },
+    {
+      "epoch": 0.08920940640055361,
+      "grad_norm": 0.8102738857269287,
+      "learning_rate": 0.0002675256428270338,
+      "loss": 4.8052,
+      "step": 63490
+    },
+    {
+      "epoch": 0.08922345733871719,
+      "grad_norm": 0.7833125591278076,
+      "learning_rate": 0.0002675677954194183,
+      "loss": 4.8178,
+      "step": 63500
+    },
+    {
+      "epoch": 0.08923750827688076,
+      "grad_norm": 0.7774162888526917,
+      "learning_rate": 0.0002676099480118027,
+      "loss": 4.9522,
+      "step": 63510
+    },
+    {
+      "epoch": 0.08925155921504434,
+      "grad_norm": 0.7914357781410217,
+      "learning_rate": 0.00026765210060418715,
+      "loss": 4.7583,
+      "step": 63520
+    },
+    {
+      "epoch": 0.08926561015320791,
+      "grad_norm": 0.780647337436676,
+      "learning_rate": 0.0002676942531965716,
+      "loss": 4.7975,
+      "step": 63530
+    },
+    {
+      "epoch": 0.08927966109137149,
+      "grad_norm": 0.7892720699310303,
+      "learning_rate": 0.000267736405788956,
+      "loss": 4.8141,
+      "step": 63540
+    },
+    {
+      "epoch": 0.08929371202953507,
+      "grad_norm": 0.7910761833190918,
+      "learning_rate": 0.0002677785583813404,
+      "loss": 4.7316,
+      "step": 63550
+    },
+    {
+      "epoch": 0.08930776296769866,
+      "grad_norm": 0.7864344716072083,
+      "learning_rate": 0.0002678207109737249,
+      "loss": 4.6778,
+      "step": 63560
+    },
+    {
+      "epoch": 0.08932181390586223,
+      "grad_norm": 0.8073357343673706,
+      "learning_rate": 0.0002678628635661093,
+      "loss": 4.7731,
+      "step": 63570
+    },
+    {
+      "epoch": 0.08933586484402581,
+      "grad_norm": 0.784953236579895,
+      "learning_rate": 0.00026790501615849374,
+      "loss": 4.7903,
+      "step": 63580
+    },
+    {
+      "epoch": 0.08934991578218938,
+      "grad_norm": 0.7666168808937073,
+      "learning_rate": 0.00026794716875087817,
+      "loss": 4.8284,
+      "step": 63590
+    },
+    {
+      "epoch": 0.08936396672035296,
+      "grad_norm": 0.8397361636161804,
+      "learning_rate": 0.0002679893213432626,
+      "loss": 4.872,
+      "step": 63600
+    },
+    {
+      "epoch": 0.08937801765851654,
+      "grad_norm": 0.801026463508606,
+      "learning_rate": 0.00026803147393564703,
+      "loss": 4.9165,
+      "step": 63610
+    },
+    {
+      "epoch": 0.08939206859668011,
+      "grad_norm": 0.7671746015548706,
+      "learning_rate": 0.00026807362652803146,
+      "loss": 4.9064,
+      "step": 63620
+    },
+    {
+      "epoch": 0.08940611953484369,
+      "grad_norm": 0.791368305683136,
+      "learning_rate": 0.0002681157791204159,
+      "loss": 4.8467,
+      "step": 63630
+    },
+    {
+      "epoch": 0.08942017047300727,
+      "grad_norm": 0.7889533638954163,
+      "learning_rate": 0.0002681579317128003,
+      "loss": 4.7742,
+      "step": 63640
+    },
+    {
+      "epoch": 0.08943422141117084,
+      "grad_norm": 0.8022910356521606,
+      "learning_rate": 0.00026820008430518475,
+      "loss": 4.7715,
+      "step": 63650
+    },
+    {
+      "epoch": 0.08944827234933442,
+      "grad_norm": 0.8055729866027832,
+      "learning_rate": 0.0002682422368975692,
+      "loss": 4.8834,
+      "step": 63660
+    },
+    {
+      "epoch": 0.08946232328749801,
+      "grad_norm": 0.8192464709281921,
+      "learning_rate": 0.0002682843894899536,
+      "loss": 4.8719,
+      "step": 63670
+    },
+    {
+      "epoch": 0.08947637422566158,
+      "grad_norm": 0.7914919257164001,
+      "learning_rate": 0.00026832654208233805,
+      "loss": 4.8239,
+      "step": 63680
+    },
+    {
+      "epoch": 0.08949042516382516,
+      "grad_norm": 0.7857105135917664,
+      "learning_rate": 0.0002683686946747225,
+      "loss": 4.798,
+      "step": 63690
+    },
+    {
+      "epoch": 0.08950447610198874,
+      "grad_norm": 0.7843837738037109,
+      "learning_rate": 0.0002684108472671069,
+      "loss": 4.8482,
+      "step": 63700
+    },
+    {
+      "epoch": 0.08951852704015231,
+      "grad_norm": 0.7536832690238953,
+      "learning_rate": 0.00026845299985949134,
+      "loss": 4.9153,
+      "step": 63710
+    },
+    {
+      "epoch": 0.08953257797831589,
+      "grad_norm": 0.8098011016845703,
+      "learning_rate": 0.00026849515245187577,
+      "loss": 4.7243,
+      "step": 63720
+    },
+    {
+      "epoch": 0.08954662891647946,
+      "grad_norm": 0.7772601842880249,
+      "learning_rate": 0.0002685373050442602,
+      "loss": 4.7708,
+      "step": 63730
+    },
+    {
+      "epoch": 0.08956067985464304,
+      "grad_norm": 0.7769831418991089,
+      "learning_rate": 0.00026857945763664464,
+      "loss": 4.8495,
+      "step": 63740
+    },
+    {
+      "epoch": 0.08957473079280662,
+      "grad_norm": 0.7833921909332275,
+      "learning_rate": 0.00026862161022902907,
+      "loss": 4.7753,
+      "step": 63750
+    },
+    {
+      "epoch": 0.0895887817309702,
+      "grad_norm": 0.7920560836791992,
+      "learning_rate": 0.0002686637628214135,
+      "loss": 4.892,
+      "step": 63760
+    },
+    {
+      "epoch": 0.08960283266913378,
+      "grad_norm": 0.7888614535331726,
+      "learning_rate": 0.00026870591541379793,
+      "loss": 4.8269,
+      "step": 63770
+    },
+    {
+      "epoch": 0.08961688360729736,
+      "grad_norm": 0.7918998599052429,
+      "learning_rate": 0.00026874806800618236,
+      "loss": 4.7976,
+      "step": 63780
+    },
+    {
+      "epoch": 0.08963093454546094,
+      "grad_norm": 0.8044359683990479,
+      "learning_rate": 0.0002687902205985668,
+      "loss": 4.6607,
+      "step": 63790
+    },
+    {
+      "epoch": 0.08964498548362451,
+      "grad_norm": 0.8149663209915161,
+      "learning_rate": 0.0002688323731909512,
+      "loss": 4.9424,
+      "step": 63800
+    },
+    {
+      "epoch": 0.08965903642178809,
+      "grad_norm": 0.8094214797019958,
+      "learning_rate": 0.00026887452578333565,
+      "loss": 4.7464,
+      "step": 63810
+    },
+    {
+      "epoch": 0.08967308735995166,
+      "grad_norm": 0.8028861284255981,
+      "learning_rate": 0.0002689166783757201,
+      "loss": 4.7546,
+      "step": 63820
+    },
+    {
+      "epoch": 0.08968713829811524,
+      "grad_norm": 0.7879270315170288,
+      "learning_rate": 0.0002689588309681045,
+      "loss": 4.7557,
+      "step": 63830
+    },
+    {
+      "epoch": 0.08970118923627882,
+      "grad_norm": 0.802869439125061,
+      "learning_rate": 0.00026900098356048895,
+      "loss": 4.7707,
+      "step": 63840
+    },
+    {
+      "epoch": 0.08971524017444239,
+      "grad_norm": 0.825872004032135,
+      "learning_rate": 0.0002690431361528734,
+      "loss": 4.7376,
+      "step": 63850
+    },
+    {
+      "epoch": 0.08972929111260597,
+      "grad_norm": 0.7914631366729736,
+      "learning_rate": 0.0002690852887452578,
+      "loss": 4.7492,
+      "step": 63860
+    },
+    {
+      "epoch": 0.08974334205076956,
+      "grad_norm": 0.8055576682090759,
+      "learning_rate": 0.00026912744133764224,
+      "loss": 4.8859,
+      "step": 63870
+    },
+    {
+      "epoch": 0.08975739298893314,
+      "grad_norm": 0.7824854254722595,
+      "learning_rate": 0.00026916959393002667,
+      "loss": 4.6911,
+      "step": 63880
+    },
+    {
+      "epoch": 0.08977144392709671,
+      "grad_norm": 0.7842464447021484,
+      "learning_rate": 0.0002692117465224111,
+      "loss": 4.8765,
+      "step": 63890
+    },
+    {
+      "epoch": 0.08978549486526029,
+      "grad_norm": 0.7833605408668518,
+      "learning_rate": 0.00026925389911479553,
+      "loss": 4.9597,
+      "step": 63900
+    },
+    {
+      "epoch": 0.08979954580342386,
+      "grad_norm": 0.8475800156593323,
+      "learning_rate": 0.00026929605170717997,
+      "loss": 4.8712,
+      "step": 63910
+    },
+    {
+      "epoch": 0.08981359674158744,
+      "grad_norm": 0.7852799296379089,
+      "learning_rate": 0.0002693382042995644,
+      "loss": 4.7098,
+      "step": 63920
+    },
+    {
+      "epoch": 0.08982764767975102,
+      "grad_norm": 0.7918947339057922,
+      "learning_rate": 0.00026938035689194883,
+      "loss": 4.7586,
+      "step": 63930
+    },
+    {
+      "epoch": 0.08984169861791459,
+      "grad_norm": 0.7760900855064392,
+      "learning_rate": 0.00026942250948433326,
+      "loss": 4.6896,
+      "step": 63940
+    },
+    {
+      "epoch": 0.08985574955607817,
+      "grad_norm": 0.7669195532798767,
+      "learning_rate": 0.0002694646620767177,
+      "loss": 4.7557,
+      "step": 63950
+    },
+    {
+      "epoch": 0.08986980049424174,
+      "grad_norm": 0.7910531163215637,
+      "learning_rate": 0.0002695068146691021,
+      "loss": 4.7617,
+      "step": 63960
+    },
+    {
+      "epoch": 0.08988385143240532,
+      "grad_norm": 0.7762371301651001,
+      "learning_rate": 0.00026954896726148655,
+      "loss": 4.8493,
+      "step": 63970
+    },
+    {
+      "epoch": 0.08989790237056891,
+      "grad_norm": 0.7670621275901794,
+      "learning_rate": 0.000269591119853871,
+      "loss": 4.8319,
+      "step": 63980
+    },
+    {
+      "epoch": 0.08991195330873249,
+      "grad_norm": 0.7959233522415161,
+      "learning_rate": 0.0002696332724462554,
+      "loss": 4.8357,
+      "step": 63990
+    },
+    {
+      "epoch": 0.08992600424689606,
+      "grad_norm": 0.7903863191604614,
+      "learning_rate": 0.00026967542503863985,
+      "loss": 4.9134,
+      "step": 64000
+    },
+    {
+      "epoch": 0.08994005518505964,
+      "grad_norm": 0.7668089866638184,
+      "learning_rate": 0.0002697175776310243,
+      "loss": 4.8917,
+      "step": 64010
+    },
+    {
+      "epoch": 0.08995410612322322,
+      "grad_norm": 0.7648235559463501,
+      "learning_rate": 0.0002697597302234087,
+      "loss": 4.7716,
+      "step": 64020
+    },
+    {
+      "epoch": 0.08996815706138679,
+      "grad_norm": 0.7938455939292908,
+      "learning_rate": 0.00026980188281579314,
+      "loss": 4.8386,
+      "step": 64030
+    },
+    {
+      "epoch": 0.08998220799955037,
+      "grad_norm": 0.7752029895782471,
+      "learning_rate": 0.00026984403540817757,
+      "loss": 4.7078,
+      "step": 64040
+    },
+    {
+      "epoch": 0.08999625893771394,
+      "grad_norm": 0.7659651041030884,
+      "learning_rate": 0.000269886188000562,
+      "loss": 4.83,
+      "step": 64050
+    },
+    {
+      "epoch": 0.09001030987587752,
+      "grad_norm": 0.7914566993713379,
+      "learning_rate": 0.00026992834059294643,
+      "loss": 4.8551,
+      "step": 64060
+    },
+    {
+      "epoch": 0.0900243608140411,
+      "grad_norm": 0.7842085957527161,
+      "learning_rate": 0.00026997049318533086,
+      "loss": 4.7521,
+      "step": 64070
+    },
+    {
+      "epoch": 0.09003841175220469,
+      "grad_norm": 0.8064524531364441,
+      "learning_rate": 0.0002700126457777153,
+      "loss": 4.6835,
+      "step": 64080
+    },
+    {
+      "epoch": 0.09005246269036826,
+      "grad_norm": 0.7767420411109924,
+      "learning_rate": 0.0002700547983700997,
+      "loss": 4.8526,
+      "step": 64090
+    },
+    {
+      "epoch": 0.09006651362853184,
+      "grad_norm": 0.8287174105644226,
+      "learning_rate": 0.00027009695096248416,
+      "loss": 4.8894,
+      "step": 64100
+    },
+    {
+      "epoch": 0.09008056456669541,
+      "grad_norm": 0.7760927677154541,
+      "learning_rate": 0.0002701391035548686,
+      "loss": 4.8314,
+      "step": 64110
+    },
+    {
+      "epoch": 0.09009461550485899,
+      "grad_norm": 0.7992680072784424,
+      "learning_rate": 0.000270181256147253,
+      "loss": 4.6976,
+      "step": 64120
+    },
+    {
+      "epoch": 0.09010866644302257,
+      "grad_norm": 0.7589333057403564,
+      "learning_rate": 0.00027022340873963745,
+      "loss": 4.7346,
+      "step": 64130
+    },
+    {
+      "epoch": 0.09012271738118614,
+      "grad_norm": 0.8577151298522949,
+      "learning_rate": 0.0002702655613320219,
+      "loss": 4.8193,
+      "step": 64140
+    },
+    {
+      "epoch": 0.09013676831934972,
+      "grad_norm": 0.8156781792640686,
+      "learning_rate": 0.0002703077139244063,
+      "loss": 4.8104,
+      "step": 64150
+    },
+    {
+      "epoch": 0.0901508192575133,
+      "grad_norm": 0.780096709728241,
+      "learning_rate": 0.00027034986651679074,
+      "loss": 4.8663,
+      "step": 64160
+    },
+    {
+      "epoch": 0.09016487019567687,
+      "grad_norm": 0.798434853553772,
+      "learning_rate": 0.0002703920191091752,
+      "loss": 4.971,
+      "step": 64170
+    },
+    {
+      "epoch": 0.09017892113384046,
+      "grad_norm": 0.7590034008026123,
+      "learning_rate": 0.0002704341717015596,
+      "loss": 4.7796,
+      "step": 64180
+    },
+    {
+      "epoch": 0.09019297207200404,
+      "grad_norm": 0.7929971814155579,
+      "learning_rate": 0.0002704763242939441,
+      "loss": 4.8087,
+      "step": 64190
+    },
+    {
+      "epoch": 0.09020702301016761,
+      "grad_norm": 0.8228213787078857,
+      "learning_rate": 0.00027051847688632847,
+      "loss": 4.7715,
+      "step": 64200
+    },
+    {
+      "epoch": 0.09022107394833119,
+      "grad_norm": 0.7692803144454956,
+      "learning_rate": 0.0002705606294787129,
+      "loss": 4.7081,
+      "step": 64210
+    },
+    {
+      "epoch": 0.09023512488649477,
+      "grad_norm": 0.7841417789459229,
+      "learning_rate": 0.00027060278207109733,
+      "loss": 4.8686,
+      "step": 64220
+    },
+    {
+      "epoch": 0.09024917582465834,
+      "grad_norm": 0.7873497009277344,
+      "learning_rate": 0.00027064493466348176,
+      "loss": 4.8794,
+      "step": 64230
+    },
+    {
+      "epoch": 0.09026322676282192,
+      "grad_norm": 0.7846882343292236,
+      "learning_rate": 0.0002706870872558662,
+      "loss": 4.671,
+      "step": 64240
+    },
+    {
+      "epoch": 0.0902772777009855,
+      "grad_norm": 0.7908580303192139,
+      "learning_rate": 0.0002707292398482507,
+      "loss": 4.7099,
+      "step": 64250
+    },
+    {
+      "epoch": 0.09029132863914907,
+      "grad_norm": 0.8587797284126282,
+      "learning_rate": 0.00027077139244063506,
+      "loss": 4.7138,
+      "step": 64260
+    },
+    {
+      "epoch": 0.09030537957731265,
+      "grad_norm": 0.8411611914634705,
+      "learning_rate": 0.0002708135450330195,
+      "loss": 4.8517,
+      "step": 64270
+    },
+    {
+      "epoch": 0.09031943051547622,
+      "grad_norm": 0.8374800086021423,
+      "learning_rate": 0.0002708556976254039,
+      "loss": 4.7404,
+      "step": 64280
+    },
+    {
+      "epoch": 0.09033348145363981,
+      "grad_norm": 0.74876469373703,
+      "learning_rate": 0.00027089785021778835,
+      "loss": 4.8666,
+      "step": 64290
+    },
+    {
+      "epoch": 0.09034753239180339,
+      "grad_norm": 0.7894213199615479,
+      "learning_rate": 0.0002709400028101728,
+      "loss": 4.6872,
+      "step": 64300
+    },
+    {
+      "epoch": 0.09036158332996697,
+      "grad_norm": 0.7926090955734253,
+      "learning_rate": 0.00027098215540255727,
+      "loss": 4.7816,
+      "step": 64310
+    },
+    {
+      "epoch": 0.09037563426813054,
+      "grad_norm": 0.7763543725013733,
+      "learning_rate": 0.0002710243079949417,
+      "loss": 4.8655,
+      "step": 64320
+    },
+    {
+      "epoch": 0.09038968520629412,
+      "grad_norm": 0.7839246988296509,
+      "learning_rate": 0.0002710664605873261,
+      "loss": 4.9012,
+      "step": 64330
+    },
+    {
+      "epoch": 0.0904037361444577,
+      "grad_norm": 0.8784212470054626,
+      "learning_rate": 0.0002711086131797105,
+      "loss": 4.8065,
+      "step": 64340
+    },
+    {
+      "epoch": 0.09041778708262127,
+      "grad_norm": 0.7844157218933105,
+      "learning_rate": 0.00027115076577209494,
+      "loss": 4.7653,
+      "step": 64350
+    },
+    {
+      "epoch": 0.09043183802078485,
+      "grad_norm": 0.7894060015678406,
+      "learning_rate": 0.00027119291836447937,
+      "loss": 4.7516,
+      "step": 64360
+    },
+    {
+      "epoch": 0.09044588895894842,
+      "grad_norm": 0.7836933732032776,
+      "learning_rate": 0.00027123085569762535,
+      "loss": 4.7689,
+      "step": 64370
+    },
+    {
+      "epoch": 0.090459939897112,
+      "grad_norm": 0.7689288854598999,
+      "learning_rate": 0.0002712730082900098,
+      "loss": 4.8232,
+      "step": 64380
+    },
+    {
+      "epoch": 0.09047399083527559,
+      "grad_norm": 0.83738112449646,
+      "learning_rate": 0.00027131516088239427,
+      "loss": 4.8224,
+      "step": 64390
+    },
+    {
+      "epoch": 0.09048804177343917,
+      "grad_norm": 0.7634382843971252,
+      "learning_rate": 0.0002713573134747787,
+      "loss": 4.9363,
+      "step": 64400
+    },
+    {
+      "epoch": 0.09050209271160274,
+      "grad_norm": 0.7892853617668152,
+      "learning_rate": 0.00027139946606716313,
+      "loss": 4.8402,
+      "step": 64410
+    },
+    {
+      "epoch": 0.09051614364976632,
+      "grad_norm": 0.7666093111038208,
+      "learning_rate": 0.0002714416186595475,
+      "loss": 4.8348,
+      "step": 64420
+    },
+    {
+      "epoch": 0.0905301945879299,
+      "grad_norm": 0.780320405960083,
+      "learning_rate": 0.00027148377125193194,
+      "loss": 4.695,
+      "step": 64430
+    },
+    {
+      "epoch": 0.09054424552609347,
+      "grad_norm": 0.7658125758171082,
+      "learning_rate": 0.00027152592384431637,
+      "loss": 4.8615,
+      "step": 64440
+    },
+    {
+      "epoch": 0.09055829646425705,
+      "grad_norm": 0.7964998483657837,
+      "learning_rate": 0.00027156807643670085,
+      "loss": 4.7311,
+      "step": 64450
+    },
+    {
+      "epoch": 0.09057234740242062,
+      "grad_norm": 0.7794137001037598,
+      "learning_rate": 0.0002716102290290853,
+      "loss": 4.8909,
+      "step": 64460
+    },
+    {
+      "epoch": 0.0905863983405842,
+      "grad_norm": 0.763118326663971,
+      "learning_rate": 0.0002716523816214697,
+      "loss": 4.6842,
+      "step": 64470
+    },
+    {
+      "epoch": 0.09060044927874777,
+      "grad_norm": 0.758752167224884,
+      "learning_rate": 0.00027169453421385415,
+      "loss": 4.7289,
+      "step": 64480
+    },
+    {
+      "epoch": 0.09061450021691136,
+      "grad_norm": 0.8015297055244446,
+      "learning_rate": 0.0002717366868062385,
+      "loss": 4.8431,
+      "step": 64490
+    },
+    {
+      "epoch": 0.09062855115507494,
+      "grad_norm": 0.7830663323402405,
+      "learning_rate": 0.000271778839398623,
+      "loss": 4.8457,
+      "step": 64500
+    },
+    {
+      "epoch": 0.09064260209323852,
+      "grad_norm": 0.7865242958068848,
+      "learning_rate": 0.00027182099199100744,
+      "loss": 4.7674,
+      "step": 64510
+    },
+    {
+      "epoch": 0.09065665303140209,
+      "grad_norm": 0.829883337020874,
+      "learning_rate": 0.00027186314458339187,
+      "loss": 4.7804,
+      "step": 64520
+    },
+    {
+      "epoch": 0.09067070396956567,
+      "grad_norm": 0.7968289852142334,
+      "learning_rate": 0.0002719052971757763,
+      "loss": 4.8857,
+      "step": 64530
+    },
+    {
+      "epoch": 0.09068475490772925,
+      "grad_norm": 0.7666770815849304,
+      "learning_rate": 0.00027194744976816073,
+      "loss": 4.7485,
+      "step": 64540
+    },
+    {
+      "epoch": 0.09069880584589282,
+      "grad_norm": 0.7687329053878784,
+      "learning_rate": 0.0002719896023605451,
+      "loss": 4.8631,
+      "step": 64550
+    },
+    {
+      "epoch": 0.0907128567840564,
+      "grad_norm": 0.8181332945823669,
+      "learning_rate": 0.0002720317549529296,
+      "loss": 4.8318,
+      "step": 64560
+    },
+    {
+      "epoch": 0.09072690772221997,
+      "grad_norm": 0.7772801518440247,
+      "learning_rate": 0.00027207390754531403,
+      "loss": 4.8091,
+      "step": 64570
+    },
+    {
+      "epoch": 0.09074095866038355,
+      "grad_norm": 0.7849490642547607,
+      "learning_rate": 0.00027211606013769846,
+      "loss": 4.8445,
+      "step": 64580
+    },
+    {
+      "epoch": 0.09075500959854713,
+      "grad_norm": 0.7753223180770874,
+      "learning_rate": 0.0002721582127300829,
+      "loss": 4.9714,
+      "step": 64590
+    },
+    {
+      "epoch": 0.09076906053671072,
+      "grad_norm": 0.7721008062362671,
+      "learning_rate": 0.0002722003653224673,
+      "loss": 4.8823,
+      "step": 64600
+    },
+    {
+      "epoch": 0.09078311147487429,
+      "grad_norm": 0.7952477335929871,
+      "learning_rate": 0.00027224251791485175,
+      "loss": 4.8502,
+      "step": 64610
+    },
+    {
+      "epoch": 0.09079716241303787,
+      "grad_norm": 0.7919043302536011,
+      "learning_rate": 0.0002722846705072362,
+      "loss": 4.757,
+      "step": 64620
+    },
+    {
+      "epoch": 0.09081121335120144,
+      "grad_norm": 0.7945954203605652,
+      "learning_rate": 0.0002723268230996206,
+      "loss": 4.8094,
+      "step": 64630
+    },
+    {
+      "epoch": 0.09082526428936502,
+      "grad_norm": 0.768685519695282,
+      "learning_rate": 0.00027236897569200505,
+      "loss": 4.796,
+      "step": 64640
+    },
+    {
+      "epoch": 0.0908393152275286,
+      "grad_norm": 0.7738785743713379,
+      "learning_rate": 0.0002724111282843895,
+      "loss": 4.8588,
+      "step": 64650
+    },
+    {
+      "epoch": 0.09085336616569217,
+      "grad_norm": 0.76154625415802,
+      "learning_rate": 0.0002724532808767739,
+      "loss": 4.8627,
+      "step": 64660
+    },
+    {
+      "epoch": 0.09086741710385575,
+      "grad_norm": 0.7614012956619263,
+      "learning_rate": 0.00027249543346915834,
+      "loss": 4.9136,
+      "step": 64670
+    },
+    {
+      "epoch": 0.09088146804201933,
+      "grad_norm": 0.7839372754096985,
+      "learning_rate": 0.00027253758606154277,
+      "loss": 4.8921,
+      "step": 64680
+    },
+    {
+      "epoch": 0.0908955189801829,
+      "grad_norm": 0.8074344396591187,
+      "learning_rate": 0.0002725797386539272,
+      "loss": 4.8253,
+      "step": 64690
+    },
+    {
+      "epoch": 0.09090956991834649,
+      "grad_norm": 0.7729949951171875,
+      "learning_rate": 0.00027262189124631163,
+      "loss": 4.6527,
+      "step": 64700
+    },
+    {
+      "epoch": 0.09092362085651007,
+      "grad_norm": 0.7652543187141418,
+      "learning_rate": 0.00027266404383869606,
+      "loss": 4.7971,
+      "step": 64710
+    },
+    {
+      "epoch": 0.09093767179467364,
+      "grad_norm": 0.7983514070510864,
+      "learning_rate": 0.0002727061964310805,
+      "loss": 4.7203,
+      "step": 64720
+    },
+    {
+      "epoch": 0.09095172273283722,
+      "grad_norm": 0.7539568543434143,
+      "learning_rate": 0.0002727483490234649,
+      "loss": 4.8282,
+      "step": 64730
+    },
+    {
+      "epoch": 0.0909657736710008,
+      "grad_norm": 0.7831068634986877,
+      "learning_rate": 0.00027279050161584936,
+      "loss": 4.7831,
+      "step": 64740
+    },
+    {
+      "epoch": 0.09097982460916437,
+      "grad_norm": 0.7707744836807251,
+      "learning_rate": 0.0002728326542082338,
+      "loss": 4.8327,
+      "step": 64750
+    },
+    {
+      "epoch": 0.09099387554732795,
+      "grad_norm": 0.7709292769432068,
+      "learning_rate": 0.0002728748068006182,
+      "loss": 4.7731,
+      "step": 64760
+    },
+    {
+      "epoch": 0.09100792648549152,
+      "grad_norm": 0.8126934766769409,
+      "learning_rate": 0.00027291695939300265,
+      "loss": 4.7199,
+      "step": 64770
+    },
+    {
+      "epoch": 0.0910219774236551,
+      "grad_norm": 0.7724310755729675,
+      "learning_rate": 0.0002729591119853871,
+      "loss": 4.9449,
+      "step": 64780
+    },
+    {
+      "epoch": 0.09103602836181868,
+      "grad_norm": 0.787074089050293,
+      "learning_rate": 0.0002730012645777715,
+      "loss": 4.8887,
+      "step": 64790
+    },
+    {
+      "epoch": 0.09105007929998227,
+      "grad_norm": 0.8085893988609314,
+      "learning_rate": 0.00027304341717015594,
+      "loss": 4.6988,
+      "step": 64800
+    },
+    {
+      "epoch": 0.09106413023814584,
+      "grad_norm": 0.7805407047271729,
+      "learning_rate": 0.0002730855697625404,
+      "loss": 4.752,
+      "step": 64810
+    },
+    {
+      "epoch": 0.09107818117630942,
+      "grad_norm": 0.8128443360328674,
+      "learning_rate": 0.0002731277223549248,
+      "loss": 4.8374,
+      "step": 64820
+    },
+    {
+      "epoch": 0.091092232114473,
+      "grad_norm": 0.745376467704773,
+      "learning_rate": 0.00027316987494730924,
+      "loss": 4.684,
+      "step": 64830
+    },
+    {
+      "epoch": 0.09110628305263657,
+      "grad_norm": 0.7663626670837402,
+      "learning_rate": 0.00027321202753969367,
+      "loss": 4.897,
+      "step": 64840
+    },
+    {
+      "epoch": 0.09112033399080015,
+      "grad_norm": 0.7975637316703796,
+      "learning_rate": 0.0002732541801320781,
+      "loss": 4.8247,
+      "step": 64850
+    },
+    {
+      "epoch": 0.09113438492896372,
+      "grad_norm": 0.7657324075698853,
+      "learning_rate": 0.00027329633272446253,
+      "loss": 4.7546,
+      "step": 64860
+    },
+    {
+      "epoch": 0.0911484358671273,
+      "grad_norm": 0.7604333162307739,
+      "learning_rate": 0.00027333848531684696,
+      "loss": 4.6362,
+      "step": 64870
+    },
+    {
+      "epoch": 0.09116248680529088,
+      "grad_norm": 0.8322310447692871,
+      "learning_rate": 0.0002733806379092314,
+      "loss": 4.7038,
+      "step": 64880
+    },
+    {
+      "epoch": 0.09117653774345445,
+      "grad_norm": 0.7785294055938721,
+      "learning_rate": 0.0002734227905016158,
+      "loss": 4.7769,
+      "step": 64890
+    },
+    {
+      "epoch": 0.09119058868161803,
+      "grad_norm": 0.763119637966156,
+      "learning_rate": 0.00027346494309400026,
+      "loss": 4.8997,
+      "step": 64900
+    },
+    {
+      "epoch": 0.09120463961978162,
+      "grad_norm": 0.7625258564949036,
+      "learning_rate": 0.0002735070956863847,
+      "loss": 4.8272,
+      "step": 64910
+    },
+    {
+      "epoch": 0.0912186905579452,
+      "grad_norm": 0.777725100517273,
+      "learning_rate": 0.0002735492482787691,
+      "loss": 4.7371,
+      "step": 64920
+    },
+    {
+      "epoch": 0.09123274149610877,
+      "grad_norm": 0.8185275793075562,
+      "learning_rate": 0.00027359140087115355,
+      "loss": 4.7327,
+      "step": 64930
+    },
+    {
+      "epoch": 0.09124679243427235,
+      "grad_norm": 0.7645296454429626,
+      "learning_rate": 0.000273633553463538,
+      "loss": 4.7193,
+      "step": 64940
+    },
+    {
+      "epoch": 0.09126084337243592,
+      "grad_norm": 0.7671161890029907,
+      "learning_rate": 0.0002736757060559224,
+      "loss": 4.8717,
+      "step": 64950
+    },
+    {
+      "epoch": 0.0912748943105995,
+      "grad_norm": 0.7883368730545044,
+      "learning_rate": 0.00027371785864830684,
+      "loss": 4.8019,
+      "step": 64960
+    },
+    {
+      "epoch": 0.09128894524876308,
+      "grad_norm": 0.7796012759208679,
+      "learning_rate": 0.0002737600112406913,
+      "loss": 4.8924,
+      "step": 64970
+    },
+    {
+      "epoch": 0.09130299618692665,
+      "grad_norm": 0.7437483072280884,
+      "learning_rate": 0.0002738021638330757,
+      "loss": 4.7397,
+      "step": 64980
+    },
+    {
+      "epoch": 0.09131704712509023,
+      "grad_norm": 0.7617444396018982,
+      "learning_rate": 0.00027384431642546014,
+      "loss": 4.7948,
+      "step": 64990
+    },
+    {
+      "epoch": 0.0913310980632538,
+      "grad_norm": 0.7927044034004211,
+      "learning_rate": 0.00027388646901784457,
+      "loss": 4.7253,
+      "step": 65000
+    },
+    {
+      "epoch": 0.0913451490014174,
+      "grad_norm": 0.8112998604774475,
+      "learning_rate": 0.000273928621610229,
+      "loss": 4.7289,
+      "step": 65010
+    },
+    {
+      "epoch": 0.09135919993958097,
+      "grad_norm": 0.7460951805114746,
+      "learning_rate": 0.00027397077420261343,
+      "loss": 4.6933,
+      "step": 65020
+    },
+    {
+      "epoch": 0.09137325087774455,
+      "grad_norm": 0.7493196129798889,
+      "learning_rate": 0.00027401292679499786,
+      "loss": 4.8195,
+      "step": 65030
+    },
+    {
+      "epoch": 0.09138730181590812,
+      "grad_norm": 0.7835866808891296,
+      "learning_rate": 0.0002740550793873823,
+      "loss": 4.7021,
+      "step": 65040
+    },
+    {
+      "epoch": 0.0914013527540717,
+      "grad_norm": 0.7718499302864075,
+      "learning_rate": 0.0002740972319797667,
+      "loss": 4.8433,
+      "step": 65050
+    },
+    {
+      "epoch": 0.09141540369223528,
+      "grad_norm": 0.7729384899139404,
+      "learning_rate": 0.00027413938457215115,
+      "loss": 4.6984,
+      "step": 65060
+    },
+    {
+      "epoch": 0.09142945463039885,
+      "grad_norm": 0.7586498260498047,
+      "learning_rate": 0.0002741815371645356,
+      "loss": 4.836,
+      "step": 65070
+    },
+    {
+      "epoch": 0.09144350556856243,
+      "grad_norm": 0.8164992332458496,
+      "learning_rate": 0.00027422368975692,
+      "loss": 4.7512,
+      "step": 65080
+    },
+    {
+      "epoch": 0.091457556506726,
+      "grad_norm": 0.7848567962646484,
+      "learning_rate": 0.00027426584234930445,
+      "loss": 4.8578,
+      "step": 65090
+    },
+    {
+      "epoch": 0.09147160744488958,
+      "grad_norm": 0.7660877704620361,
+      "learning_rate": 0.0002743079949416889,
+      "loss": 4.8616,
+      "step": 65100
+    },
+    {
+      "epoch": 0.09148565838305317,
+      "grad_norm": 0.7506645321846008,
+      "learning_rate": 0.0002743501475340733,
+      "loss": 4.6539,
+      "step": 65110
+    },
+    {
+      "epoch": 0.09149970932121675,
+      "grad_norm": 0.7743892669677734,
+      "learning_rate": 0.00027439230012645774,
+      "loss": 4.7529,
+      "step": 65120
+    },
+    {
+      "epoch": 0.09151376025938032,
+      "grad_norm": 0.7679518461227417,
+      "learning_rate": 0.0002744344527188422,
+      "loss": 4.768,
+      "step": 65130
+    },
+    {
+      "epoch": 0.0915278111975439,
+      "grad_norm": 0.7692804336547852,
+      "learning_rate": 0.0002744766053112266,
+      "loss": 4.8296,
+      "step": 65140
+    },
+    {
+      "epoch": 0.09154186213570747,
+      "grad_norm": 0.7586258053779602,
+      "learning_rate": 0.00027451875790361104,
+      "loss": 4.7978,
+      "step": 65150
+    },
+    {
+      "epoch": 0.09155591307387105,
+      "grad_norm": 0.799343466758728,
+      "learning_rate": 0.00027456091049599547,
+      "loss": 4.6653,
+      "step": 65160
+    },
+    {
+      "epoch": 0.09156996401203463,
+      "grad_norm": 0.8051402568817139,
+      "learning_rate": 0.0002746030630883799,
+      "loss": 4.8429,
+      "step": 65170
+    },
+    {
+      "epoch": 0.0915840149501982,
+      "grad_norm": 0.7682332396507263,
+      "learning_rate": 0.00027464521568076433,
+      "loss": 4.7701,
+      "step": 65180
+    },
+    {
+      "epoch": 0.09159806588836178,
+      "grad_norm": 0.7532578110694885,
+      "learning_rate": 0.0002746873682731488,
+      "loss": 4.9643,
+      "step": 65190
+    },
+    {
+      "epoch": 0.09161211682652536,
+      "grad_norm": 0.7631083130836487,
+      "learning_rate": 0.0002747295208655332,
+      "loss": 4.8076,
+      "step": 65200
+    },
+    {
+      "epoch": 0.09162616776468893,
+      "grad_norm": 0.7771109342575073,
+      "learning_rate": 0.0002747716734579176,
+      "loss": 4.894,
+      "step": 65210
+    },
+    {
+      "epoch": 0.09164021870285252,
+      "grad_norm": 0.8161442279815674,
+      "learning_rate": 0.00027481382605030205,
+      "loss": 4.7906,
+      "step": 65220
+    },
+    {
+      "epoch": 0.0916542696410161,
+      "grad_norm": 0.7733320593833923,
+      "learning_rate": 0.0002748559786426865,
+      "loss": 4.7649,
+      "step": 65230
+    },
+    {
+      "epoch": 0.09166832057917967,
+      "grad_norm": 0.7682567238807678,
+      "learning_rate": 0.0002748981312350709,
+      "loss": 4.6973,
+      "step": 65240
+    },
+    {
+      "epoch": 0.09168237151734325,
+      "grad_norm": 0.7778883576393127,
+      "learning_rate": 0.0002749402838274554,
+      "loss": 4.7464,
+      "step": 65250
+    },
+    {
+      "epoch": 0.09169642245550683,
+      "grad_norm": 0.7723760008811951,
+      "learning_rate": 0.0002749824364198398,
+      "loss": 4.8092,
+      "step": 65260
+    },
+    {
+      "epoch": 0.0917104733936704,
+      "grad_norm": 0.7590900659561157,
+      "learning_rate": 0.0002750245890122242,
+      "loss": 4.7138,
+      "step": 65270
+    },
+    {
+      "epoch": 0.09172452433183398,
+      "grad_norm": 0.7957313656806946,
+      "learning_rate": 0.00027506674160460864,
+      "loss": 4.7174,
+      "step": 65280
+    },
+    {
+      "epoch": 0.09173857526999755,
+      "grad_norm": 0.7801516056060791,
+      "learning_rate": 0.00027510889419699307,
+      "loss": 4.8837,
+      "step": 65290
+    },
+    {
+      "epoch": 0.09175262620816113,
+      "grad_norm": 0.7635098099708557,
+      "learning_rate": 0.0002751510467893775,
+      "loss": 4.7774,
+      "step": 65300
+    },
+    {
+      "epoch": 0.09176667714632471,
+      "grad_norm": 0.8163349032402039,
+      "learning_rate": 0.000275193199381762,
+      "loss": 4.902,
+      "step": 65310
+    },
+    {
+      "epoch": 0.0917807280844883,
+      "grad_norm": 0.7679600715637207,
+      "learning_rate": 0.0002752353519741464,
+      "loss": 4.8026,
+      "step": 65320
+    },
+    {
+      "epoch": 0.09179477902265187,
+      "grad_norm": 0.794219970703125,
+      "learning_rate": 0.0002752775045665308,
+      "loss": 4.8016,
+      "step": 65330
+    },
+    {
+      "epoch": 0.09180882996081545,
+      "grad_norm": 0.8230299353599548,
+      "learning_rate": 0.00027531965715891523,
+      "loss": 4.7612,
+      "step": 65340
+    },
+    {
+      "epoch": 0.09182288089897903,
+      "grad_norm": 0.7828049659729004,
+      "learning_rate": 0.00027536180975129966,
+      "loss": 4.8781,
+      "step": 65350
+    },
+    {
+      "epoch": 0.0918369318371426,
+      "grad_norm": 0.7481945753097534,
+      "learning_rate": 0.0002754039623436841,
+      "loss": 4.818,
+      "step": 65360
+    },
+    {
+      "epoch": 0.09185098277530618,
+      "grad_norm": 0.8005225658416748,
+      "learning_rate": 0.0002754461149360686,
+      "loss": 4.7186,
+      "step": 65370
+    },
+    {
+      "epoch": 0.09186503371346975,
+      "grad_norm": 0.8023760318756104,
+      "learning_rate": 0.000275488267528453,
+      "loss": 4.7263,
+      "step": 65380
+    },
+    {
+      "epoch": 0.09187908465163333,
+      "grad_norm": 0.8292908072471619,
+      "learning_rate": 0.0002755304201208374,
+      "loss": 4.9251,
+      "step": 65390
+    },
+    {
+      "epoch": 0.0918931355897969,
+      "grad_norm": 0.7676294445991516,
+      "learning_rate": 0.0002755725727132218,
+      "loss": 4.7774,
+      "step": 65400
+    },
+    {
+      "epoch": 0.09190718652796048,
+      "grad_norm": 0.7877136468887329,
+      "learning_rate": 0.00027561472530560625,
+      "loss": 4.8057,
+      "step": 65410
+    },
+    {
+      "epoch": 0.09192123746612407,
+      "grad_norm": 0.7636573314666748,
+      "learning_rate": 0.0002756568778979907,
+      "loss": 4.9751,
+      "step": 65420
+    },
+    {
+      "epoch": 0.09193528840428765,
+      "grad_norm": 0.7771406173706055,
+      "learning_rate": 0.00027569903049037516,
+      "loss": 4.785,
+      "step": 65430
+    },
+    {
+      "epoch": 0.09194933934245123,
+      "grad_norm": 0.7473503351211548,
+      "learning_rate": 0.0002757411830827596,
+      "loss": 4.7884,
+      "step": 65440
+    },
+    {
+      "epoch": 0.0919633902806148,
+      "grad_norm": 0.7925262451171875,
+      "learning_rate": 0.000275783335675144,
+      "loss": 4.7761,
+      "step": 65450
+    },
+    {
+      "epoch": 0.09197744121877838,
+      "grad_norm": 0.7829357385635376,
+      "learning_rate": 0.0002758254882675284,
+      "loss": 4.7649,
+      "step": 65460
+    },
+    {
+      "epoch": 0.09199149215694195,
+      "grad_norm": 0.8279735445976257,
+      "learning_rate": 0.00027586764085991283,
+      "loss": 4.7296,
+      "step": 65470
+    },
+    {
+      "epoch": 0.09200554309510553,
+      "grad_norm": 0.7664202451705933,
+      "learning_rate": 0.00027590979345229726,
+      "loss": 4.8287,
+      "step": 65480
+    },
+    {
+      "epoch": 0.0920195940332691,
+      "grad_norm": 0.7685601711273193,
+      "learning_rate": 0.00027595194604468175,
+      "loss": 4.7864,
+      "step": 65490
+    },
+    {
+      "epoch": 0.09203364497143268,
+      "grad_norm": 0.784343421459198,
+      "learning_rate": 0.0002759940986370662,
+      "loss": 4.817,
+      "step": 65500
+    },
+    {
+      "epoch": 0.09204769590959626,
+      "grad_norm": 0.8135119080543518,
+      "learning_rate": 0.0002760362512294506,
+      "loss": 4.8879,
+      "step": 65510
+    },
+    {
+      "epoch": 0.09206174684775985,
+      "grad_norm": 0.7757298350334167,
+      "learning_rate": 0.000276078403821835,
+      "loss": 4.793,
+      "step": 65520
+    },
+    {
+      "epoch": 0.09207579778592342,
+      "grad_norm": 0.8407139778137207,
+      "learning_rate": 0.0002761205564142194,
+      "loss": 4.8615,
+      "step": 65530
+    },
+    {
+      "epoch": 0.092089848724087,
+      "grad_norm": 0.7666981220245361,
+      "learning_rate": 0.00027616270900660385,
+      "loss": 4.721,
+      "step": 65540
+    },
+    {
+      "epoch": 0.09210389966225058,
+      "grad_norm": 0.781695544719696,
+      "learning_rate": 0.00027620486159898834,
+      "loss": 4.7766,
+      "step": 65550
+    },
+    {
+      "epoch": 0.09211795060041415,
+      "grad_norm": 0.7683060169219971,
+      "learning_rate": 0.00027624701419137277,
+      "loss": 4.8538,
+      "step": 65560
+    },
+    {
+      "epoch": 0.09213200153857773,
+      "grad_norm": 0.7697662115097046,
+      "learning_rate": 0.0002762891667837572,
+      "loss": 4.8836,
+      "step": 65570
+    },
+    {
+      "epoch": 0.0921460524767413,
+      "grad_norm": 0.7881425023078918,
+      "learning_rate": 0.00027633131937614163,
+      "loss": 4.7934,
+      "step": 65580
+    },
+    {
+      "epoch": 0.09216010341490488,
+      "grad_norm": 0.7787525653839111,
+      "learning_rate": 0.000276373471968526,
+      "loss": 4.8257,
+      "step": 65590
+    },
+    {
+      "epoch": 0.09217415435306846,
+      "grad_norm": 0.757805585861206,
+      "learning_rate": 0.00027641562456091044,
+      "loss": 4.8827,
+      "step": 65600
+    },
+    {
+      "epoch": 0.09218820529123203,
+      "grad_norm": 0.7534720301628113,
+      "learning_rate": 0.0002764577771532949,
+      "loss": 4.7524,
+      "step": 65610
+    },
+    {
+      "epoch": 0.09220225622939561,
+      "grad_norm": 0.7800748348236084,
+      "learning_rate": 0.00027649992974567935,
+      "loss": 4.8452,
+      "step": 65620
+    },
+    {
+      "epoch": 0.0922163071675592,
+      "grad_norm": 0.7849797010421753,
+      "learning_rate": 0.0002765420823380638,
+      "loss": 4.7981,
+      "step": 65630
+    },
+    {
+      "epoch": 0.09223035810572278,
+      "grad_norm": 0.7599306106567383,
+      "learning_rate": 0.0002765842349304482,
+      "loss": 4.7777,
+      "step": 65640
+    },
+    {
+      "epoch": 0.09224440904388635,
+      "grad_norm": 0.789528489112854,
+      "learning_rate": 0.0002766263875228326,
+      "loss": 4.7558,
+      "step": 65650
+    },
+    {
+      "epoch": 0.09225845998204993,
+      "grad_norm": 0.7765876650810242,
+      "learning_rate": 0.0002766685401152171,
+      "loss": 4.8782,
+      "step": 65660
+    },
+    {
+      "epoch": 0.0922725109202135,
+      "grad_norm": 0.7817826271057129,
+      "learning_rate": 0.0002767106927076015,
+      "loss": 4.8516,
+      "step": 65670
+    },
+    {
+      "epoch": 0.09228656185837708,
+      "grad_norm": 0.7477931380271912,
+      "learning_rate": 0.00027675284529998594,
+      "loss": 4.9344,
+      "step": 65680
+    },
+    {
+      "epoch": 0.09230061279654066,
+      "grad_norm": 0.7491680979728699,
+      "learning_rate": 0.00027679499789237037,
+      "loss": 4.8218,
+      "step": 65690
+    },
+    {
+      "epoch": 0.09231466373470423,
+      "grad_norm": 0.7636551260948181,
+      "learning_rate": 0.0002768371504847548,
+      "loss": 4.8828,
+      "step": 65700
+    },
+    {
+      "epoch": 0.09232871467286781,
+      "grad_norm": 0.7482210993766785,
+      "learning_rate": 0.00027687930307713923,
+      "loss": 4.7725,
+      "step": 65710
+    },
+    {
+      "epoch": 0.09234276561103139,
+      "grad_norm": 0.7747960090637207,
+      "learning_rate": 0.00027692145566952367,
+      "loss": 4.7205,
+      "step": 65720
+    },
+    {
+      "epoch": 0.09235681654919498,
+      "grad_norm": 0.7718808650970459,
+      "learning_rate": 0.0002769636082619081,
+      "loss": 4.7258,
+      "step": 65730
+    },
+    {
+      "epoch": 0.09237086748735855,
+      "grad_norm": 0.7397331595420837,
+      "learning_rate": 0.00027700576085429253,
+      "loss": 4.7103,
+      "step": 65740
+    },
+    {
+      "epoch": 0.09238491842552213,
+      "grad_norm": 0.7860981822013855,
+      "learning_rate": 0.00027704791344667696,
+      "loss": 4.752,
+      "step": 65750
+    },
+    {
+      "epoch": 0.0923989693636857,
+      "grad_norm": 0.7629745006561279,
+      "learning_rate": 0.0002770900660390614,
+      "loss": 4.8679,
+      "step": 65760
+    },
+    {
+      "epoch": 0.09241302030184928,
+      "grad_norm": 0.8002822995185852,
+      "learning_rate": 0.0002771322186314458,
+      "loss": 4.6234,
+      "step": 65770
+    },
+    {
+      "epoch": 0.09242707124001286,
+      "grad_norm": 0.7651919722557068,
+      "learning_rate": 0.00027717437122383025,
+      "loss": 4.8368,
+      "step": 65780
+    },
+    {
+      "epoch": 0.09244112217817643,
+      "grad_norm": 0.7951807379722595,
+      "learning_rate": 0.0002772165238162147,
+      "loss": 4.7385,
+      "step": 65790
+    },
+    {
+      "epoch": 0.09245517311634001,
+      "grad_norm": 0.7759065628051758,
+      "learning_rate": 0.0002772586764085991,
+      "loss": 4.8701,
+      "step": 65800
+    },
+    {
+      "epoch": 0.09246922405450358,
+      "grad_norm": 0.753464937210083,
+      "learning_rate": 0.00027730082900098355,
+      "loss": 4.7545,
+      "step": 65810
+    },
+    {
+      "epoch": 0.09248327499266716,
+      "grad_norm": 0.7959068417549133,
+      "learning_rate": 0.000277342981593368,
+      "loss": 4.8609,
+      "step": 65820
+    },
+    {
+      "epoch": 0.09249732593083075,
+      "grad_norm": 0.7836405038833618,
+      "learning_rate": 0.0002773851341857524,
+      "loss": 4.768,
+      "step": 65830
+    },
+    {
+      "epoch": 0.09251137686899433,
+      "grad_norm": 0.7547979950904846,
+      "learning_rate": 0.00027742728677813684,
+      "loss": 4.7868,
+      "step": 65840
+    },
+    {
+      "epoch": 0.0925254278071579,
+      "grad_norm": 0.793201744556427,
+      "learning_rate": 0.00027746943937052127,
+      "loss": 4.7997,
+      "step": 65850
+    },
+    {
+      "epoch": 0.09253947874532148,
+      "grad_norm": 0.7729663252830505,
+      "learning_rate": 0.0002775115919629057,
+      "loss": 4.7825,
+      "step": 65860
+    },
+    {
+      "epoch": 0.09255352968348506,
+      "grad_norm": 0.8203871846199036,
+      "learning_rate": 0.00027755374455529013,
+      "loss": 4.7756,
+      "step": 65870
+    },
+    {
+      "epoch": 0.09256758062164863,
+      "grad_norm": 0.8373938798904419,
+      "learning_rate": 0.00027759589714767456,
+      "loss": 4.7466,
+      "step": 65880
+    },
+    {
+      "epoch": 0.09258163155981221,
+      "grad_norm": 0.77327960729599,
+      "learning_rate": 0.000277638049740059,
+      "loss": 4.812,
+      "step": 65890
+    },
+    {
+      "epoch": 0.09259568249797578,
+      "grad_norm": 0.7683082818984985,
+      "learning_rate": 0.0002776802023324434,
+      "loss": 4.8128,
+      "step": 65900
+    },
+    {
+      "epoch": 0.09260973343613936,
+      "grad_norm": 0.7626813650131226,
+      "learning_rate": 0.00027772235492482786,
+      "loss": 4.8501,
+      "step": 65910
+    },
+    {
+      "epoch": 0.09262378437430294,
+      "grad_norm": 0.7827718257904053,
+      "learning_rate": 0.0002777645075172123,
+      "loss": 4.7699,
+      "step": 65920
+    },
+    {
+      "epoch": 0.09263783531246651,
+      "grad_norm": 0.7622679471969604,
+      "learning_rate": 0.0002778066601095967,
+      "loss": 4.6728,
+      "step": 65930
+    },
+    {
+      "epoch": 0.0926518862506301,
+      "grad_norm": 0.7683987021446228,
+      "learning_rate": 0.00027784881270198115,
+      "loss": 4.9069,
+      "step": 65940
+    },
+    {
+      "epoch": 0.09266593718879368,
+      "grad_norm": 0.7734614610671997,
+      "learning_rate": 0.0002778909652943656,
+      "loss": 4.8696,
+      "step": 65950
+    },
+    {
+      "epoch": 0.09267998812695726,
+      "grad_norm": 1.3228286504745483,
+      "learning_rate": 0.00027793311788675,
+      "loss": 4.7319,
+      "step": 65960
+    },
+    {
+      "epoch": 0.09269403906512083,
+      "grad_norm": 0.7824644446372986,
+      "learning_rate": 0.00027797527047913445,
+      "loss": 4.7125,
+      "step": 65970
+    },
+    {
+      "epoch": 0.09270809000328441,
+      "grad_norm": 0.7849623560905457,
+      "learning_rate": 0.0002780174230715189,
+      "loss": 4.8618,
+      "step": 65980
+    },
+    {
+      "epoch": 0.09272214094144798,
+      "grad_norm": 0.7580630779266357,
+      "learning_rate": 0.0002780595756639033,
+      "loss": 4.7867,
+      "step": 65990
+    },
+    {
+      "epoch": 0.09273619187961156,
+      "grad_norm": 0.7988112568855286,
+      "learning_rate": 0.00027810172825628774,
+      "loss": 4.7798,
+      "step": 66000
+    },
+    {
+      "epoch": 0.09275024281777514,
+      "grad_norm": 0.7714882493019104,
+      "learning_rate": 0.00027814388084867217,
+      "loss": 4.8462,
+      "step": 66010
+    },
+    {
+      "epoch": 0.09276429375593871,
+      "grad_norm": 0.7701422572135925,
+      "learning_rate": 0.0002781860334410566,
+      "loss": 4.7374,
+      "step": 66020
+    },
+    {
+      "epoch": 0.09277834469410229,
+      "grad_norm": 0.7310877442359924,
+      "learning_rate": 0.00027822818603344103,
+      "loss": 4.8362,
+      "step": 66030
+    },
+    {
+      "epoch": 0.09279239563226588,
+      "grad_norm": 0.7519302368164062,
+      "learning_rate": 0.00027827033862582546,
+      "loss": 4.8217,
+      "step": 66040
+    },
+    {
+      "epoch": 0.09280644657042945,
+      "grad_norm": 0.7439327239990234,
+      "learning_rate": 0.0002783124912182099,
+      "loss": 4.899,
+      "step": 66050
+    },
+    {
+      "epoch": 0.09282049750859303,
+      "grad_norm": 0.7650001645088196,
+      "learning_rate": 0.0002783546438105943,
+      "loss": 4.8668,
+      "step": 66060
+    },
+    {
+      "epoch": 0.0928345484467566,
+      "grad_norm": 0.7906643152236938,
+      "learning_rate": 0.00027839679640297876,
+      "loss": 4.85,
+      "step": 66070
+    },
+    {
+      "epoch": 0.09284859938492018,
+      "grad_norm": 0.772290825843811,
+      "learning_rate": 0.0002784389489953632,
+      "loss": 4.8879,
+      "step": 66080
+    },
+    {
+      "epoch": 0.09286265032308376,
+      "grad_norm": 0.7907188534736633,
+      "learning_rate": 0.0002784811015877476,
+      "loss": 4.7498,
+      "step": 66090
+    },
+    {
+      "epoch": 0.09287670126124734,
+      "grad_norm": 0.7830239534378052,
+      "learning_rate": 0.00027852325418013205,
+      "loss": 4.8478,
+      "step": 66100
+    },
+    {
+      "epoch": 0.09289075219941091,
+      "grad_norm": 0.7706902027130127,
+      "learning_rate": 0.0002785654067725165,
+      "loss": 4.8905,
+      "step": 66110
+    },
+    {
+      "epoch": 0.09290480313757449,
+      "grad_norm": 0.7300134301185608,
+      "learning_rate": 0.0002786075593649009,
+      "loss": 4.7552,
+      "step": 66120
+    },
+    {
+      "epoch": 0.09291885407573806,
+      "grad_norm": 0.7618225812911987,
+      "learning_rate": 0.00027864971195728534,
+      "loss": 4.7571,
+      "step": 66130
+    },
+    {
+      "epoch": 0.09293290501390165,
+      "grad_norm": 0.8037118911743164,
+      "learning_rate": 0.0002786918645496698,
+      "loss": 4.749,
+      "step": 66140
+    },
+    {
+      "epoch": 0.09294695595206523,
+      "grad_norm": 0.7755313515663147,
+      "learning_rate": 0.0002787340171420542,
+      "loss": 4.8419,
+      "step": 66150
+    },
+    {
+      "epoch": 0.0929610068902288,
+      "grad_norm": 0.753610372543335,
+      "learning_rate": 0.00027877616973443864,
+      "loss": 4.8936,
+      "step": 66160
+    },
+    {
+      "epoch": 0.09297505782839238,
+      "grad_norm": 0.8028730750083923,
+      "learning_rate": 0.00027881832232682307,
+      "loss": 4.8275,
+      "step": 66170
+    },
+    {
+      "epoch": 0.09298910876655596,
+      "grad_norm": 0.7614244818687439,
+      "learning_rate": 0.0002788604749192075,
+      "loss": 4.8163,
+      "step": 66180
+    },
+    {
+      "epoch": 0.09300315970471953,
+      "grad_norm": 0.7953831553459167,
+      "learning_rate": 0.00027890262751159193,
+      "loss": 4.9137,
+      "step": 66190
+    },
+    {
+      "epoch": 0.09301721064288311,
+      "grad_norm": 0.7856250405311584,
+      "learning_rate": 0.00027894478010397636,
+      "loss": 4.8262,
+      "step": 66200
+    },
+    {
+      "epoch": 0.09303126158104669,
+      "grad_norm": 0.749713659286499,
+      "learning_rate": 0.0002789869326963608,
+      "loss": 4.8025,
+      "step": 66210
+    },
+    {
+      "epoch": 0.09304531251921026,
+      "grad_norm": 0.7905395030975342,
+      "learning_rate": 0.0002790290852887452,
+      "loss": 4.8134,
+      "step": 66220
+    },
+    {
+      "epoch": 0.09305936345737384,
+      "grad_norm": 0.7537453174591064,
+      "learning_rate": 0.00027907123788112966,
+      "loss": 4.6562,
+      "step": 66230
+    },
+    {
+      "epoch": 0.09307341439553742,
+      "grad_norm": 0.7715054154396057,
+      "learning_rate": 0.0002791133904735141,
+      "loss": 4.8059,
+      "step": 66240
+    },
+    {
+      "epoch": 0.093087465333701,
+      "grad_norm": 0.7600757479667664,
+      "learning_rate": 0.0002791555430658985,
+      "loss": 4.7877,
+      "step": 66250
+    },
+    {
+      "epoch": 0.09310151627186458,
+      "grad_norm": 0.7605622410774231,
+      "learning_rate": 0.00027919769565828295,
+      "loss": 4.8044,
+      "step": 66260
+    },
+    {
+      "epoch": 0.09311556721002816,
+      "grad_norm": 0.7521841526031494,
+      "learning_rate": 0.0002792398482506674,
+      "loss": 4.771,
+      "step": 66270
+    },
+    {
+      "epoch": 0.09312961814819173,
+      "grad_norm": 0.751024067401886,
+      "learning_rate": 0.0002792820008430518,
+      "loss": 4.7984,
+      "step": 66280
+    },
+    {
+      "epoch": 0.09314366908635531,
+      "grad_norm": 0.7688322067260742,
+      "learning_rate": 0.0002793241534354363,
+      "loss": 4.8356,
+      "step": 66290
+    },
+    {
+      "epoch": 0.09315772002451889,
+      "grad_norm": 0.7705650329589844,
+      "learning_rate": 0.0002793663060278207,
+      "loss": 4.7877,
+      "step": 66300
+    },
+    {
+      "epoch": 0.09317177096268246,
+      "grad_norm": 0.769523024559021,
+      "learning_rate": 0.0002794084586202051,
+      "loss": 4.8034,
+      "step": 66310
+    },
+    {
+      "epoch": 0.09318582190084604,
+      "grad_norm": 0.7682741284370422,
+      "learning_rate": 0.00027945061121258954,
+      "loss": 4.6817,
+      "step": 66320
+    },
+    {
+      "epoch": 0.09319987283900961,
+      "grad_norm": 0.765863835811615,
+      "learning_rate": 0.00027949276380497397,
+      "loss": 4.7184,
+      "step": 66330
+    },
+    {
+      "epoch": 0.09321392377717319,
+      "grad_norm": 0.7482745051383972,
+      "learning_rate": 0.0002795349163973584,
+      "loss": 4.766,
+      "step": 66340
+    },
+    {
+      "epoch": 0.09322797471533678,
+      "grad_norm": 0.7587314248085022,
+      "learning_rate": 0.0002795770689897429,
+      "loss": 4.8509,
+      "step": 66350
+    },
+    {
+      "epoch": 0.09324202565350036,
+      "grad_norm": 0.7830142378807068,
+      "learning_rate": 0.00027961922158212726,
+      "loss": 4.6339,
+      "step": 66360
+    },
+    {
+      "epoch": 0.09325607659166393,
+      "grad_norm": 0.7715080380439758,
+      "learning_rate": 0.0002796613741745117,
+      "loss": 4.7213,
+      "step": 66370
+    },
+    {
+      "epoch": 0.09327012752982751,
+      "grad_norm": 0.7741284966468811,
+      "learning_rate": 0.0002797035267668961,
+      "loss": 4.7524,
+      "step": 66380
+    },
+    {
+      "epoch": 0.09328417846799109,
+      "grad_norm": 0.7751002907752991,
+      "learning_rate": 0.00027974567935928055,
+      "loss": 4.8013,
+      "step": 66390
+    },
+    {
+      "epoch": 0.09329822940615466,
+      "grad_norm": 0.8361513614654541,
+      "learning_rate": 0.000279787831951665,
+      "loss": 4.8237,
+      "step": 66400
+    },
+    {
+      "epoch": 0.09331228034431824,
+      "grad_norm": 0.7796192169189453,
+      "learning_rate": 0.00027982998454404947,
+      "loss": 4.8101,
+      "step": 66410
+    },
+    {
+      "epoch": 0.09332633128248181,
+      "grad_norm": 0.8053950071334839,
+      "learning_rate": 0.0002798721371364339,
+      "loss": 4.6938,
+      "step": 66420
+    },
+    {
+      "epoch": 0.09334038222064539,
+      "grad_norm": 0.8083071112632751,
+      "learning_rate": 0.0002799142897288183,
+      "loss": 4.8032,
+      "step": 66430
+    },
+    {
+      "epoch": 0.09335443315880897,
+      "grad_norm": 0.8004527688026428,
+      "learning_rate": 0.0002799564423212027,
+      "loss": 4.7322,
+      "step": 66440
+    },
+    {
+      "epoch": 0.09336848409697256,
+      "grad_norm": 0.7978747487068176,
+      "learning_rate": 0.00027999859491358714,
+      "loss": 4.8168,
+      "step": 66450
+    },
+    {
+      "epoch": 0.09338253503513613,
+      "grad_norm": 0.8221896290779114,
+      "learning_rate": 0.00028004074750597157,
+      "loss": 4.5916,
+      "step": 66460
+    },
+    {
+      "epoch": 0.09339658597329971,
+      "grad_norm": 0.8084535002708435,
+      "learning_rate": 0.00028008290009835606,
+      "loss": 4.8748,
+      "step": 66470
+    },
+    {
+      "epoch": 0.09341063691146329,
+      "grad_norm": 0.8202959895133972,
+      "learning_rate": 0.0002801250526907405,
+      "loss": 4.7996,
+      "step": 66480
+    },
+    {
+      "epoch": 0.09342468784962686,
+      "grad_norm": 0.7502172589302063,
+      "learning_rate": 0.00028016720528312487,
+      "loss": 4.9115,
+      "step": 66490
+    },
+    {
+      "epoch": 0.09343873878779044,
+      "grad_norm": 0.7529955506324768,
+      "learning_rate": 0.0002802093578755093,
+      "loss": 4.8029,
+      "step": 66500
+    },
+    {
+      "epoch": 0.09345278972595401,
+      "grad_norm": 0.7608353495597839,
+      "learning_rate": 0.00028025151046789373,
+      "loss": 4.7967,
+      "step": 66510
+    },
+    {
+      "epoch": 0.09346684066411759,
+      "grad_norm": 0.8318527936935425,
+      "learning_rate": 0.00028029366306027816,
+      "loss": 4.7791,
+      "step": 66520
+    },
+    {
+      "epoch": 0.09348089160228117,
+      "grad_norm": 0.7857316732406616,
+      "learning_rate": 0.00028033581565266264,
+      "loss": 4.7374,
+      "step": 66530
+    },
+    {
+      "epoch": 0.09349494254044474,
+      "grad_norm": 0.9135415554046631,
+      "learning_rate": 0.0002803779682450471,
+      "loss": 4.761,
+      "step": 66540
+    },
+    {
+      "epoch": 0.09350899347860832,
+      "grad_norm": 0.7974305748939514,
+      "learning_rate": 0.00028042012083743145,
+      "loss": 4.6771,
+      "step": 66550
+    },
+    {
+      "epoch": 0.09352304441677191,
+      "grad_norm": 0.7421619892120361,
+      "learning_rate": 0.0002804622734298159,
+      "loss": 4.7538,
+      "step": 66560
+    },
+    {
+      "epoch": 0.09353709535493548,
+      "grad_norm": 0.796992838382721,
+      "learning_rate": 0.0002805044260222003,
+      "loss": 4.8697,
+      "step": 66570
+    },
+    {
+      "epoch": 0.09355114629309906,
+      "grad_norm": 1.0140798091888428,
+      "learning_rate": 0.00028054657861458475,
+      "loss": 4.8219,
+      "step": 66580
+    },
+    {
+      "epoch": 0.09356519723126264,
+      "grad_norm": 0.8030684590339661,
+      "learning_rate": 0.00028058873120696923,
+      "loss": 4.7397,
+      "step": 66590
+    },
+    {
+      "epoch": 0.09357924816942621,
+      "grad_norm": 0.8484904766082764,
+      "learning_rate": 0.00028063088379935366,
+      "loss": 4.7481,
+      "step": 66600
+    },
+    {
+      "epoch": 0.09359329910758979,
+      "grad_norm": 0.7419153451919556,
+      "learning_rate": 0.0002806730363917381,
+      "loss": 4.8027,
+      "step": 66610
+    },
+    {
+      "epoch": 0.09360735004575337,
+      "grad_norm": 0.7505769729614258,
+      "learning_rate": 0.00028071518898412247,
+      "loss": 4.7095,
+      "step": 66620
+    },
+    {
+      "epoch": 0.09362140098391694,
+      "grad_norm": 0.7895868420600891,
+      "learning_rate": 0.0002807573415765069,
+      "loss": 4.7332,
+      "step": 66630
+    },
+    {
+      "epoch": 0.09363545192208052,
+      "grad_norm": 0.7834476232528687,
+      "learning_rate": 0.00028079949416889133,
+      "loss": 4.6707,
+      "step": 66640
+    },
+    {
+      "epoch": 0.0936495028602441,
+      "grad_norm": 0.7624843716621399,
+      "learning_rate": 0.0002808416467612758,
+      "loss": 4.7925,
+      "step": 66650
+    },
+    {
+      "epoch": 0.09366355379840768,
+      "grad_norm": 0.8170449733734131,
+      "learning_rate": 0.00028088379935366025,
+      "loss": 4.7729,
+      "step": 66660
+    },
+    {
+      "epoch": 0.09367760473657126,
+      "grad_norm": 0.8319101929664612,
+      "learning_rate": 0.0002809259519460447,
+      "loss": 4.933,
+      "step": 66670
+    },
+    {
+      "epoch": 0.09369165567473484,
+      "grad_norm": 0.7560040354728699,
+      "learning_rate": 0.00028096810453842906,
+      "loss": 4.7946,
+      "step": 66680
+    },
+    {
+      "epoch": 0.09370570661289841,
+      "grad_norm": 0.7837918996810913,
+      "learning_rate": 0.0002810102571308135,
+      "loss": 4.7895,
+      "step": 66690
+    },
+    {
+      "epoch": 0.09371975755106199,
+      "grad_norm": 0.7507804036140442,
+      "learning_rate": 0.0002810524097231979,
+      "loss": 4.7504,
+      "step": 66700
+    },
+    {
+      "epoch": 0.09373380848922556,
+      "grad_norm": 0.771360456943512,
+      "learning_rate": 0.0002810945623155824,
+      "loss": 4.8561,
+      "step": 66710
+    },
+    {
+      "epoch": 0.09374785942738914,
+      "grad_norm": 0.7771738171577454,
+      "learning_rate": 0.00028113671490796684,
+      "loss": 4.874,
+      "step": 66720
+    },
+    {
+      "epoch": 0.09376191036555272,
+      "grad_norm": 0.7817769646644592,
+      "learning_rate": 0.00028117886750035127,
+      "loss": 4.8408,
+      "step": 66730
+    },
+    {
+      "epoch": 0.0937759613037163,
+      "grad_norm": 0.9140200018882751,
+      "learning_rate": 0.0002812210200927357,
+      "loss": 4.8285,
+      "step": 66740
+    },
+    {
+      "epoch": 0.09379001224187987,
+      "grad_norm": 0.7840200662612915,
+      "learning_rate": 0.0002812631726851201,
+      "loss": 4.8193,
+      "step": 66750
+    },
+    {
+      "epoch": 0.09380406318004346,
+      "grad_norm": 0.7683082818984985,
+      "learning_rate": 0.0002813053252775045,
+      "loss": 4.5404,
+      "step": 66760
+    },
+    {
+      "epoch": 0.09381811411820704,
+      "grad_norm": 0.7743772268295288,
+      "learning_rate": 0.000281347477869889,
+      "loss": 4.9332,
+      "step": 66770
+    },
+    {
+      "epoch": 0.09383216505637061,
+      "grad_norm": 0.7601823210716248,
+      "learning_rate": 0.0002813896304622734,
+      "loss": 4.8454,
+      "step": 66780
+    },
+    {
+      "epoch": 0.09384621599453419,
+      "grad_norm": 0.7685533761978149,
+      "learning_rate": 0.00028143178305465786,
+      "loss": 4.8704,
+      "step": 66790
+    },
+    {
+      "epoch": 0.09386026693269776,
+      "grad_norm": 0.7852855324745178,
+      "learning_rate": 0.0002814739356470423,
+      "loss": 4.7243,
+      "step": 66800
+    },
+    {
+      "epoch": 0.09387431787086134,
+      "grad_norm": 0.7344205379486084,
+      "learning_rate": 0.00028151608823942666,
+      "loss": 4.6807,
+      "step": 66810
+    },
+    {
+      "epoch": 0.09388836880902492,
+      "grad_norm": 0.7859446406364441,
+      "learning_rate": 0.00028155824083181115,
+      "loss": 4.6914,
+      "step": 66820
+    },
+    {
+      "epoch": 0.09390241974718849,
+      "grad_norm": 0.769825279712677,
+      "learning_rate": 0.0002816003934241956,
+      "loss": 4.795,
+      "step": 66830
+    },
+    {
+      "epoch": 0.09391647068535207,
+      "grad_norm": 0.7718509435653687,
+      "learning_rate": 0.00028164254601658,
+      "loss": 4.7383,
+      "step": 66840
+    },
+    {
+      "epoch": 0.09393052162351564,
+      "grad_norm": 0.7945657968521118,
+      "learning_rate": 0.00028168469860896444,
+      "loss": 4.8164,
+      "step": 66850
+    },
+    {
+      "epoch": 0.09394457256167922,
+      "grad_norm": 0.9426447749137878,
+      "learning_rate": 0.0002817268512013489,
+      "loss": 4.674,
+      "step": 66860
+    },
+    {
+      "epoch": 0.09395862349984281,
+      "grad_norm": 0.7499502301216125,
+      "learning_rate": 0.0002817690037937333,
+      "loss": 4.7949,
+      "step": 66870
+    },
+    {
+      "epoch": 0.09397267443800639,
+      "grad_norm": 0.7673668265342712,
+      "learning_rate": 0.00028181115638611774,
+      "loss": 4.8771,
+      "step": 66880
+    },
+    {
+      "epoch": 0.09398672537616996,
+      "grad_norm": 0.7585043907165527,
+      "learning_rate": 0.00028185330897850217,
+      "loss": 4.8944,
+      "step": 66890
+    },
+    {
+      "epoch": 0.09400077631433354,
+      "grad_norm": 0.7782924175262451,
+      "learning_rate": 0.0002818954615708866,
+      "loss": 4.7822,
+      "step": 66900
+    },
+    {
+      "epoch": 0.09401482725249712,
+      "grad_norm": 0.7514545917510986,
+      "learning_rate": 0.00028193761416327103,
+      "loss": 4.6599,
+      "step": 66910
+    },
+    {
+      "epoch": 0.09402887819066069,
+      "grad_norm": 0.7437326312065125,
+      "learning_rate": 0.00028197976675565546,
+      "loss": 4.8388,
+      "step": 66920
+    },
+    {
+      "epoch": 0.09404292912882427,
+      "grad_norm": 0.7777213454246521,
+      "learning_rate": 0.0002820219193480399,
+      "loss": 4.7727,
+      "step": 66930
+    },
+    {
+      "epoch": 0.09405698006698784,
+      "grad_norm": 0.7418848276138306,
+      "learning_rate": 0.0002820640719404243,
+      "loss": 4.7165,
+      "step": 66940
+    },
+    {
+      "epoch": 0.09407103100515142,
+      "grad_norm": 0.7748672962188721,
+      "learning_rate": 0.00028210622453280875,
+      "loss": 4.7824,
+      "step": 66950
+    },
+    {
+      "epoch": 0.094085081943315,
+      "grad_norm": 0.7600992321968079,
+      "learning_rate": 0.0002821483771251932,
+      "loss": 4.8642,
+      "step": 66960
+    },
+    {
+      "epoch": 0.09409913288147859,
+      "grad_norm": 0.7563418745994568,
+      "learning_rate": 0.0002821905297175776,
+      "loss": 4.8016,
+      "step": 66970
+    },
+    {
+      "epoch": 0.09411318381964216,
+      "grad_norm": 0.7561931014060974,
+      "learning_rate": 0.00028223268230996205,
+      "loss": 4.8931,
+      "step": 66980
+    },
+    {
+      "epoch": 0.09412723475780574,
+      "grad_norm": 0.8095070719718933,
+      "learning_rate": 0.0002822748349023465,
+      "loss": 4.6954,
+      "step": 66990
+    },
+    {
+      "epoch": 0.09414128569596932,
+      "grad_norm": 0.7557608485221863,
+      "learning_rate": 0.0002823169874947309,
+      "loss": 4.9589,
+      "step": 67000
+    },
+    {
+      "epoch": 0.09415533663413289,
+      "grad_norm": 0.7877722978591919,
+      "learning_rate": 0.00028235914008711534,
+      "loss": 4.8427,
+      "step": 67010
+    },
+    {
+      "epoch": 0.09416938757229647,
+      "grad_norm": 0.7584842443466187,
+      "learning_rate": 0.00028240129267949977,
+      "loss": 4.8917,
+      "step": 67020
+    },
+    {
+      "epoch": 0.09418343851046004,
+      "grad_norm": 0.7679328322410583,
+      "learning_rate": 0.0002824434452718842,
+      "loss": 4.684,
+      "step": 67030
+    },
+    {
+      "epoch": 0.09419748944862362,
+      "grad_norm": 0.7546048760414124,
+      "learning_rate": 0.00028248559786426863,
+      "loss": 4.6522,
+      "step": 67040
+    },
+    {
+      "epoch": 0.0942115403867872,
+      "grad_norm": 0.7862738966941833,
+      "learning_rate": 0.00028252775045665307,
+      "loss": 4.7412,
+      "step": 67050
+    },
+    {
+      "epoch": 0.09422559132495077,
+      "grad_norm": 0.7737675905227661,
+      "learning_rate": 0.0002825699030490375,
+      "loss": 4.7039,
+      "step": 67060
+    },
+    {
+      "epoch": 0.09423964226311436,
+      "grad_norm": 0.7403519153594971,
+      "learning_rate": 0.00028261205564142193,
+      "loss": 4.8726,
+      "step": 67070
+    },
+    {
+      "epoch": 0.09425369320127794,
+      "grad_norm": 0.7542514204978943,
+      "learning_rate": 0.00028265420823380636,
+      "loss": 4.8402,
+      "step": 67080
+    },
+    {
+      "epoch": 0.09426774413944151,
+      "grad_norm": 0.7455792427062988,
+      "learning_rate": 0.0002826963608261908,
+      "loss": 4.8167,
+      "step": 67090
+    },
+    {
+      "epoch": 0.09428179507760509,
+      "grad_norm": 0.7721514701843262,
+      "learning_rate": 0.0002827385134185752,
+      "loss": 4.8949,
+      "step": 67100
+    },
+    {
+      "epoch": 0.09429584601576867,
+      "grad_norm": 0.7985861301422119,
+      "learning_rate": 0.00028278066601095965,
+      "loss": 4.8274,
+      "step": 67110
+    },
+    {
+      "epoch": 0.09430989695393224,
+      "grad_norm": 0.7794066071510315,
+      "learning_rate": 0.0002828228186033441,
+      "loss": 4.7993,
+      "step": 67120
+    },
+    {
+      "epoch": 0.09432394789209582,
+      "grad_norm": 0.7830890417098999,
+      "learning_rate": 0.0002828649711957285,
+      "loss": 4.7902,
+      "step": 67130
+    },
+    {
+      "epoch": 0.0943379988302594,
+      "grad_norm": 0.7577866911888123,
+      "learning_rate": 0.00028290712378811295,
+      "loss": 4.8907,
+      "step": 67140
+    },
+    {
+      "epoch": 0.09435204976842297,
+      "grad_norm": 0.8127627968788147,
+      "learning_rate": 0.0002829492763804974,
+      "loss": 4.7554,
+      "step": 67150
+    },
+    {
+      "epoch": 0.09436610070658655,
+      "grad_norm": 0.7625774145126343,
+      "learning_rate": 0.0002829914289728818,
+      "loss": 4.7444,
+      "step": 67160
+    },
+    {
+      "epoch": 0.09438015164475012,
+      "grad_norm": 0.7992131114006042,
+      "learning_rate": 0.00028303358156526624,
+      "loss": 4.8071,
+      "step": 67170
+    },
+    {
+      "epoch": 0.09439420258291371,
+      "grad_norm": 0.7639414668083191,
+      "learning_rate": 0.00028307573415765067,
+      "loss": 4.7122,
+      "step": 67180
+    },
+    {
+      "epoch": 0.09440825352107729,
+      "grad_norm": 0.7498297691345215,
+      "learning_rate": 0.0002831178867500351,
+      "loss": 4.7293,
+      "step": 67190
+    },
+    {
+      "epoch": 0.09442230445924087,
+      "grad_norm": 0.7553703188896179,
+      "learning_rate": 0.00028316003934241953,
+      "loss": 4.9202,
+      "step": 67200
+    },
+    {
+      "epoch": 0.09443635539740444,
+      "grad_norm": 0.7734369039535522,
+      "learning_rate": 0.00028320219193480396,
+      "loss": 4.7912,
+      "step": 67210
+    },
+    {
+      "epoch": 0.09445040633556802,
+      "grad_norm": 0.7736976146697998,
+      "learning_rate": 0.0002832443445271884,
+      "loss": 4.8836,
+      "step": 67220
+    },
+    {
+      "epoch": 0.0944644572737316,
+      "grad_norm": 0.7649384140968323,
+      "learning_rate": 0.0002832864971195728,
+      "loss": 4.8357,
+      "step": 67230
+    },
+    {
+      "epoch": 0.09447850821189517,
+      "grad_norm": 0.7400626540184021,
+      "learning_rate": 0.00028332864971195726,
+      "loss": 4.8263,
+      "step": 67240
+    },
+    {
+      "epoch": 0.09449255915005875,
+      "grad_norm": 0.8864115476608276,
+      "learning_rate": 0.0002833708023043417,
+      "loss": 4.7784,
+      "step": 67250
+    },
+    {
+      "epoch": 0.09450661008822232,
+      "grad_norm": 0.7719464302062988,
+      "learning_rate": 0.0002834129548967261,
+      "loss": 4.927,
+      "step": 67260
+    },
+    {
+      "epoch": 0.0945206610263859,
+      "grad_norm": 0.7647191286087036,
+      "learning_rate": 0.00028345510748911055,
+      "loss": 4.776,
+      "step": 67270
+    },
+    {
+      "epoch": 0.09453471196454949,
+      "grad_norm": 0.7322827577590942,
+      "learning_rate": 0.000283497260081495,
+      "loss": 4.8069,
+      "step": 67280
+    },
+    {
+      "epoch": 0.09454876290271307,
+      "grad_norm": 0.7592594623565674,
+      "learning_rate": 0.0002835394126738794,
+      "loss": 4.7898,
+      "step": 67290
+    },
+    {
+      "epoch": 0.09456281384087664,
+      "grad_norm": 0.7448276281356812,
+      "learning_rate": 0.00028358156526626384,
+      "loss": 4.8365,
+      "step": 67300
+    },
+    {
+      "epoch": 0.09457686477904022,
+      "grad_norm": 0.786535382270813,
+      "learning_rate": 0.0002836237178586483,
+      "loss": 4.7254,
+      "step": 67310
+    },
+    {
+      "epoch": 0.0945909157172038,
+      "grad_norm": 0.7626109719276428,
+      "learning_rate": 0.0002836658704510327,
+      "loss": 4.9295,
+      "step": 67320
+    },
+    {
+      "epoch": 0.09460496665536737,
+      "grad_norm": 0.8014910817146301,
+      "learning_rate": 0.00028370802304341714,
+      "loss": 4.8296,
+      "step": 67330
+    },
+    {
+      "epoch": 0.09461901759353095,
+      "grad_norm": 0.7696790099143982,
+      "learning_rate": 0.00028375017563580157,
+      "loss": 4.7296,
+      "step": 67340
+    },
+    {
+      "epoch": 0.09463306853169452,
+      "grad_norm": 0.770531952381134,
+      "learning_rate": 0.000283792328228186,
+      "loss": 4.8663,
+      "step": 67350
+    },
+    {
+      "epoch": 0.0946471194698581,
+      "grad_norm": 0.7845657467842102,
+      "learning_rate": 0.00028383448082057043,
+      "loss": 4.7573,
+      "step": 67360
+    },
+    {
+      "epoch": 0.09466117040802167,
+      "grad_norm": 0.798072099685669,
+      "learning_rate": 0.00028387663341295486,
+      "loss": 4.7633,
+      "step": 67370
+    },
+    {
+      "epoch": 0.09467522134618526,
+      "grad_norm": 0.7655891180038452,
+      "learning_rate": 0.0002839187860053393,
+      "loss": 4.729,
+      "step": 67380
+    },
+    {
+      "epoch": 0.09468927228434884,
+      "grad_norm": 0.7528164386749268,
+      "learning_rate": 0.0002839609385977237,
+      "loss": 4.7863,
+      "step": 67390
+    },
+    {
+      "epoch": 0.09470332322251242,
+      "grad_norm": 0.7557967305183411,
+      "learning_rate": 0.00028400309119010816,
+      "loss": 4.8202,
+      "step": 67400
+    },
+    {
+      "epoch": 0.094717374160676,
+      "grad_norm": 0.7755447626113892,
+      "learning_rate": 0.0002840452437824926,
+      "loss": 4.7798,
+      "step": 67410
+    },
+    {
+      "epoch": 0.09473142509883957,
+      "grad_norm": 0.7739848494529724,
+      "learning_rate": 0.000284087396374877,
+      "loss": 4.8155,
+      "step": 67420
+    },
+    {
+      "epoch": 0.09474547603700315,
+      "grad_norm": 0.7496004700660706,
+      "learning_rate": 0.00028412954896726145,
+      "loss": 4.579,
+      "step": 67430
+    },
+    {
+      "epoch": 0.09475952697516672,
+      "grad_norm": 0.7430173754692078,
+      "learning_rate": 0.0002841717015596459,
+      "loss": 4.7696,
+      "step": 67440
+    },
+    {
+      "epoch": 0.0947735779133303,
+      "grad_norm": 0.7696523666381836,
+      "learning_rate": 0.00028421385415203037,
+      "loss": 4.9943,
+      "step": 67450
+    },
+    {
+      "epoch": 0.09478762885149387,
+      "grad_norm": 0.8024923801422119,
+      "learning_rate": 0.00028425600674441474,
+      "loss": 4.7865,
+      "step": 67460
+    },
+    {
+      "epoch": 0.09480167978965745,
+      "grad_norm": 0.7850974202156067,
+      "learning_rate": 0.0002842981593367992,
+      "loss": 4.813,
+      "step": 67470
+    },
+    {
+      "epoch": 0.09481573072782103,
+      "grad_norm": 0.7672629952430725,
+      "learning_rate": 0.0002843403119291836,
+      "loss": 4.8848,
+      "step": 67480
+    },
+    {
+      "epoch": 0.09482978166598462,
+      "grad_norm": 0.7520509958267212,
+      "learning_rate": 0.00028438246452156804,
+      "loss": 4.821,
+      "step": 67490
+    },
+    {
+      "epoch": 0.09484383260414819,
+      "grad_norm": 0.7907251715660095,
+      "learning_rate": 0.00028442461711395247,
+      "loss": 4.7222,
+      "step": 67500
+    },
+    {
+      "epoch": 0.09485788354231177,
+      "grad_norm": 0.7532843947410583,
+      "learning_rate": 0.00028446676970633695,
+      "loss": 4.7545,
+      "step": 67510
+    },
+    {
+      "epoch": 0.09487193448047535,
+      "grad_norm": 0.7584605813026428,
+      "learning_rate": 0.00028450892229872133,
+      "loss": 4.7687,
+      "step": 67520
+    },
+    {
+      "epoch": 0.09488598541863892,
+      "grad_norm": 0.7673341035842896,
+      "learning_rate": 0.00028455107489110576,
+      "loss": 4.7871,
+      "step": 67530
+    },
+    {
+      "epoch": 0.0949000363568025,
+      "grad_norm": 0.7831926345825195,
+      "learning_rate": 0.0002845932274834902,
+      "loss": 4.7859,
+      "step": 67540
+    },
+    {
+      "epoch": 0.09491408729496607,
+      "grad_norm": 0.7618937492370605,
+      "learning_rate": 0.0002846353800758746,
+      "loss": 4.8277,
+      "step": 67550
+    },
+    {
+      "epoch": 0.09492813823312965,
+      "grad_norm": 0.813814640045166,
+      "learning_rate": 0.00028467753266825905,
+      "loss": 4.7525,
+      "step": 67560
+    },
+    {
+      "epoch": 0.09494218917129323,
+      "grad_norm": 0.7494709491729736,
+      "learning_rate": 0.00028471968526064354,
+      "loss": 4.8525,
+      "step": 67570
+    },
+    {
+      "epoch": 0.0949562401094568,
+      "grad_norm": 0.7520213723182678,
+      "learning_rate": 0.00028476183785302797,
+      "loss": 4.7942,
+      "step": 67580
+    },
+    {
+      "epoch": 0.09497029104762039,
+      "grad_norm": 0.7849497199058533,
+      "learning_rate": 0.00028480399044541235,
+      "loss": 4.7749,
+      "step": 67590
+    },
+    {
+      "epoch": 0.09498434198578397,
+      "grad_norm": 0.7550547122955322,
+      "learning_rate": 0.0002848461430377968,
+      "loss": 4.7389,
+      "step": 67600
+    },
+    {
+      "epoch": 0.09499839292394754,
+      "grad_norm": 0.7957075834274292,
+      "learning_rate": 0.0002848882956301812,
+      "loss": 4.7788,
+      "step": 67610
+    },
+    {
+      "epoch": 0.09501244386211112,
+      "grad_norm": 0.7405850291252136,
+      "learning_rate": 0.00028493044822256564,
+      "loss": 4.7592,
+      "step": 67620
+    },
+    {
+      "epoch": 0.0950264948002747,
+      "grad_norm": 0.7549129128456116,
+      "learning_rate": 0.00028497260081495013,
+      "loss": 4.8282,
+      "step": 67630
+    },
+    {
+      "epoch": 0.09504054573843827,
+      "grad_norm": 0.7701460123062134,
+      "learning_rate": 0.00028501475340733456,
+      "loss": 4.8901,
+      "step": 67640
+    },
+    {
+      "epoch": 0.09505459667660185,
+      "grad_norm": 0.7383482456207275,
+      "learning_rate": 0.00028505690599971894,
+      "loss": 4.7539,
+      "step": 67650
+    },
+    {
+      "epoch": 0.09506864761476543,
+      "grad_norm": 0.7618653774261475,
+      "learning_rate": 0.00028509905859210337,
+      "loss": 4.812,
+      "step": 67660
+    },
+    {
+      "epoch": 0.095082698552929,
+      "grad_norm": 0.7711530327796936,
+      "learning_rate": 0.0002851412111844878,
+      "loss": 4.7898,
+      "step": 67670
+    },
+    {
+      "epoch": 0.09509674949109258,
+      "grad_norm": 0.7828205227851868,
+      "learning_rate": 0.00028518336377687223,
+      "loss": 4.8524,
+      "step": 67680
+    },
+    {
+      "epoch": 0.09511080042925617,
+      "grad_norm": 0.7714737057685852,
+      "learning_rate": 0.0002852255163692567,
+      "loss": 4.731,
+      "step": 67690
+    },
+    {
+      "epoch": 0.09512485136741974,
+      "grad_norm": 0.7553906440734863,
+      "learning_rate": 0.00028526766896164115,
+      "loss": 4.7764,
+      "step": 67700
+    },
+    {
+      "epoch": 0.09513890230558332,
+      "grad_norm": 0.7760165333747864,
+      "learning_rate": 0.0002853098215540256,
+      "loss": 4.7859,
+      "step": 67710
+    },
+    {
+      "epoch": 0.0951529532437469,
+      "grad_norm": 0.7785241603851318,
+      "learning_rate": 0.00028535197414640995,
+      "loss": 4.7499,
+      "step": 67720
+    },
+    {
+      "epoch": 0.09516700418191047,
+      "grad_norm": 0.7613420486450195,
+      "learning_rate": 0.0002853941267387944,
+      "loss": 4.806,
+      "step": 67730
+    },
+    {
+      "epoch": 0.09518105512007405,
+      "grad_norm": 0.7879531979560852,
+      "learning_rate": 0.0002854362793311788,
+      "loss": 4.8668,
+      "step": 67740
+    },
+    {
+      "epoch": 0.09519510605823762,
+      "grad_norm": 0.7439324855804443,
+      "learning_rate": 0.0002854784319235633,
+      "loss": 4.8048,
+      "step": 67750
+    },
+    {
+      "epoch": 0.0952091569964012,
+      "grad_norm": 0.8519731163978577,
+      "learning_rate": 0.00028552058451594773,
+      "loss": 4.6315,
+      "step": 67760
+    },
+    {
+      "epoch": 0.09522320793456478,
+      "grad_norm": 0.7999274134635925,
+      "learning_rate": 0.00028556273710833216,
+      "loss": 4.8338,
+      "step": 67770
+    },
+    {
+      "epoch": 0.09523725887272835,
+      "grad_norm": 0.7844228148460388,
+      "learning_rate": 0.00028560488970071654,
+      "loss": 4.7579,
+      "step": 67780
+    },
+    {
+      "epoch": 0.09525130981089193,
+      "grad_norm": 0.7809820175170898,
+      "learning_rate": 0.00028564704229310097,
+      "loss": 4.7089,
+      "step": 67790
+    },
+    {
+      "epoch": 0.09526536074905552,
+      "grad_norm": 0.759044885635376,
+      "learning_rate": 0.0002856891948854854,
+      "loss": 4.7492,
+      "step": 67800
+    },
+    {
+      "epoch": 0.0952794116872191,
+      "grad_norm": 0.755091667175293,
+      "learning_rate": 0.0002857313474778699,
+      "loss": 4.6888,
+      "step": 67810
+    },
+    {
+      "epoch": 0.09529346262538267,
+      "grad_norm": 0.7296857237815857,
+      "learning_rate": 0.0002857735000702543,
+      "loss": 4.7357,
+      "step": 67820
+    },
+    {
+      "epoch": 0.09530751356354625,
+      "grad_norm": 0.7496383190155029,
+      "learning_rate": 0.00028581565266263875,
+      "loss": 4.7345,
+      "step": 67830
+    },
+    {
+      "epoch": 0.09532156450170982,
+      "grad_norm": 0.7716061472892761,
+      "learning_rate": 0.00028585780525502313,
+      "loss": 4.8233,
+      "step": 67840
+    },
+    {
+      "epoch": 0.0953356154398734,
+      "grad_norm": 0.7615830302238464,
+      "learning_rate": 0.00028589995784740756,
+      "loss": 4.7433,
+      "step": 67850
+    },
+    {
+      "epoch": 0.09534966637803698,
+      "grad_norm": 0.7765911817550659,
+      "learning_rate": 0.000285942110439792,
+      "loss": 4.7496,
+      "step": 67860
+    },
+    {
+      "epoch": 0.09536371731620055,
+      "grad_norm": 0.7490736842155457,
+      "learning_rate": 0.0002859842630321765,
+      "loss": 4.8747,
+      "step": 67870
+    },
+    {
+      "epoch": 0.09537776825436413,
+      "grad_norm": 0.7704018354415894,
+      "learning_rate": 0.0002860264156245609,
+      "loss": 4.8234,
+      "step": 67880
+    },
+    {
+      "epoch": 0.0953918191925277,
+      "grad_norm": 0.7510921359062195,
+      "learning_rate": 0.00028606856821694534,
+      "loss": 4.7961,
+      "step": 67890
+    },
+    {
+      "epoch": 0.0954058701306913,
+      "grad_norm": 0.7316580414772034,
+      "learning_rate": 0.00028611072080932977,
+      "loss": 4.7473,
+      "step": 67900
+    },
+    {
+      "epoch": 0.09541992106885487,
+      "grad_norm": 0.7586633563041687,
+      "learning_rate": 0.00028615287340171415,
+      "loss": 4.8093,
+      "step": 67910
+    },
+    {
+      "epoch": 0.09543397200701845,
+      "grad_norm": 0.7565239071846008,
+      "learning_rate": 0.0002861950259940986,
+      "loss": 4.7894,
+      "step": 67920
+    },
+    {
+      "epoch": 0.09544802294518202,
+      "grad_norm": 0.7565003037452698,
+      "learning_rate": 0.00028623717858648306,
+      "loss": 4.8842,
+      "step": 67930
+    },
+    {
+      "epoch": 0.0954620738833456,
+      "grad_norm": 0.7603064179420471,
+      "learning_rate": 0.0002862793311788675,
+      "loss": 4.7224,
+      "step": 67940
+    },
+    {
+      "epoch": 0.09547612482150918,
+      "grad_norm": 0.7563630938529968,
+      "learning_rate": 0.0002863214837712519,
+      "loss": 4.7289,
+      "step": 67950
+    },
+    {
+      "epoch": 0.09549017575967275,
+      "grad_norm": 0.7232897281646729,
+      "learning_rate": 0.00028636363636363636,
+      "loss": 4.718,
+      "step": 67960
+    },
+    {
+      "epoch": 0.09550422669783633,
+      "grad_norm": 0.747141420841217,
+      "learning_rate": 0.00028640578895602073,
+      "loss": 4.7984,
+      "step": 67970
+    },
+    {
+      "epoch": 0.0955182776359999,
+      "grad_norm": 0.7786388397216797,
+      "learning_rate": 0.00028644794154840516,
+      "loss": 4.6962,
+      "step": 67980
+    },
+    {
+      "epoch": 0.09553232857416348,
+      "grad_norm": 0.7422512769699097,
+      "learning_rate": 0.00028649009414078965,
+      "loss": 4.8789,
+      "step": 67990
+    },
+    {
+      "epoch": 0.09554637951232707,
+      "grad_norm": 0.8502141833305359,
+      "learning_rate": 0.0002865322467331741,
+      "loss": 4.8269,
+      "step": 68000
+    },
+    {
+      "epoch": 0.09556043045049065,
+      "grad_norm": 0.7409288287162781,
+      "learning_rate": 0.0002865743993255585,
+      "loss": 4.7592,
+      "step": 68010
+    },
+    {
+      "epoch": 0.09557448138865422,
+      "grad_norm": 0.7689024209976196,
+      "learning_rate": 0.00028661655191794294,
+      "loss": 4.7512,
+      "step": 68020
+    },
+    {
+      "epoch": 0.0955885323268178,
+      "grad_norm": 0.7674072980880737,
+      "learning_rate": 0.0002866587045103274,
+      "loss": 4.9744,
+      "step": 68030
+    },
+    {
+      "epoch": 0.09560258326498138,
+      "grad_norm": 0.7291126251220703,
+      "learning_rate": 0.0002867008571027118,
+      "loss": 4.7998,
+      "step": 68040
+    },
+    {
+      "epoch": 0.09561663420314495,
+      "grad_norm": 0.7502152323722839,
+      "learning_rate": 0.00028674300969509624,
+      "loss": 4.7427,
+      "step": 68050
+    },
+    {
+      "epoch": 0.09563068514130853,
+      "grad_norm": 0.744105875492096,
+      "learning_rate": 0.00028678516228748067,
+      "loss": 4.8043,
+      "step": 68060
+    },
+    {
+      "epoch": 0.0956447360794721,
+      "grad_norm": 0.7957160472869873,
+      "learning_rate": 0.0002868273148798651,
+      "loss": 4.8657,
+      "step": 68070
+    },
+    {
+      "epoch": 0.09565878701763568,
+      "grad_norm": 0.857458233833313,
+      "learning_rate": 0.00028686946747224953,
+      "loss": 4.685,
+      "step": 68080
+    },
+    {
+      "epoch": 0.09567283795579926,
+      "grad_norm": 0.7583727240562439,
+      "learning_rate": 0.00028691162006463396,
+      "loss": 4.8969,
+      "step": 68090
+    },
+    {
+      "epoch": 0.09568688889396283,
+      "grad_norm": 0.7256913781166077,
+      "learning_rate": 0.0002869537726570184,
+      "loss": 4.8367,
+      "step": 68100
+    },
+    {
+      "epoch": 0.09570093983212642,
+      "grad_norm": 0.7758738994598389,
+      "learning_rate": 0.0002869959252494028,
+      "loss": 4.8103,
+      "step": 68110
+    },
+    {
+      "epoch": 0.09571499077029,
+      "grad_norm": 0.7613289952278137,
+      "learning_rate": 0.00028703807784178725,
+      "loss": 4.8833,
+      "step": 68120
+    },
+    {
+      "epoch": 0.09572904170845357,
+      "grad_norm": 0.7221300601959229,
+      "learning_rate": 0.0002870802304341717,
+      "loss": 4.7747,
+      "step": 68130
+    },
+    {
+      "epoch": 0.09574309264661715,
+      "grad_norm": 0.7547207474708557,
+      "learning_rate": 0.0002871223830265561,
+      "loss": 4.8013,
+      "step": 68140
+    },
+    {
+      "epoch": 0.09575714358478073,
+      "grad_norm": 0.7506118416786194,
+      "learning_rate": 0.00028716453561894055,
+      "loss": 4.8435,
+      "step": 68150
+    },
+    {
+      "epoch": 0.0957711945229443,
+      "grad_norm": 0.7563244700431824,
+      "learning_rate": 0.000287206688211325,
+      "loss": 4.7599,
+      "step": 68160
+    },
+    {
+      "epoch": 0.09578524546110788,
+      "grad_norm": 0.7807770371437073,
+      "learning_rate": 0.0002872488408037094,
+      "loss": 4.8858,
+      "step": 68170
+    },
+    {
+      "epoch": 0.09579929639927146,
+      "grad_norm": 0.7905330657958984,
+      "learning_rate": 0.00028729099339609384,
+      "loss": 4.8305,
+      "step": 68180
+    },
+    {
+      "epoch": 0.09581334733743503,
+      "grad_norm": 0.7471740245819092,
+      "learning_rate": 0.00028733314598847827,
+      "loss": 4.7711,
+      "step": 68190
+    },
+    {
+      "epoch": 0.09582739827559861,
+      "grad_norm": 0.7244470119476318,
+      "learning_rate": 0.0002873752985808627,
+      "loss": 4.9656,
+      "step": 68200
+    },
+    {
+      "epoch": 0.0958414492137622,
+      "grad_norm": 0.7297165989875793,
+      "learning_rate": 0.00028741745117324713,
+      "loss": 4.783,
+      "step": 68210
+    },
+    {
+      "epoch": 0.09585550015192577,
+      "grad_norm": 0.7688151597976685,
+      "learning_rate": 0.00028745960376563157,
+      "loss": 4.7866,
+      "step": 68220
+    },
+    {
+      "epoch": 0.09586955109008935,
+      "grad_norm": 0.7747588157653809,
+      "learning_rate": 0.000287501756358016,
+      "loss": 4.7084,
+      "step": 68230
+    },
+    {
+      "epoch": 0.09588360202825293,
+      "grad_norm": 0.766656219959259,
+      "learning_rate": 0.00028754390895040043,
+      "loss": 4.7271,
+      "step": 68240
+    },
+    {
+      "epoch": 0.0958976529664165,
+      "grad_norm": 0.7433441281318665,
+      "learning_rate": 0.00028758606154278486,
+      "loss": 4.8076,
+      "step": 68250
+    },
+    {
+      "epoch": 0.09591170390458008,
+      "grad_norm": 0.7474486231803894,
+      "learning_rate": 0.0002876282141351693,
+      "loss": 4.7575,
+      "step": 68260
+    },
+    {
+      "epoch": 0.09592575484274365,
+      "grad_norm": 0.7397088408470154,
+      "learning_rate": 0.0002876703667275537,
+      "loss": 4.8102,
+      "step": 68270
+    },
+    {
+      "epoch": 0.09593980578090723,
+      "grad_norm": 0.7497064471244812,
+      "learning_rate": 0.00028771251931993815,
+      "loss": 4.8034,
+      "step": 68280
+    },
+    {
+      "epoch": 0.0959538567190708,
+      "grad_norm": 0.7426905632019043,
+      "learning_rate": 0.0002877546719123226,
+      "loss": 4.7192,
+      "step": 68290
+    },
+    {
+      "epoch": 0.09596790765723438,
+      "grad_norm": 0.7634294629096985,
+      "learning_rate": 0.000287796824504707,
+      "loss": 4.7343,
+      "step": 68300
+    },
+    {
+      "epoch": 0.09598195859539797,
+      "grad_norm": 0.7681379318237305,
+      "learning_rate": 0.00028783897709709145,
+      "loss": 4.7821,
+      "step": 68310
+    },
+    {
+      "epoch": 0.09599600953356155,
+      "grad_norm": 0.7707437872886658,
+      "learning_rate": 0.0002878811296894759,
+      "loss": 4.839,
+      "step": 68320
+    },
+    {
+      "epoch": 0.09601006047172513,
+      "grad_norm": 0.7358832359313965,
+      "learning_rate": 0.0002879232822818603,
+      "loss": 4.7412,
+      "step": 68330
+    },
+    {
+      "epoch": 0.0960241114098887,
+      "grad_norm": 0.7570725083351135,
+      "learning_rate": 0.00028796543487424474,
+      "loss": 4.7562,
+      "step": 68340
+    },
+    {
+      "epoch": 0.09603816234805228,
+      "grad_norm": 0.7852720022201538,
+      "learning_rate": 0.00028800758746662917,
+      "loss": 4.6697,
+      "step": 68350
+    },
+    {
+      "epoch": 0.09605221328621585,
+      "grad_norm": 0.7732431888580322,
+      "learning_rate": 0.0002880497400590136,
+      "loss": 4.7449,
+      "step": 68360
+    },
+    {
+      "epoch": 0.09606626422437943,
+      "grad_norm": 0.7763086557388306,
+      "learning_rate": 0.00028809189265139803,
+      "loss": 4.7639,
+      "step": 68370
+    },
+    {
+      "epoch": 0.096080315162543,
+      "grad_norm": 0.7549687623977661,
+      "learning_rate": 0.00028813404524378246,
+      "loss": 4.7652,
+      "step": 68380
+    },
+    {
+      "epoch": 0.09609436610070658,
+      "grad_norm": 0.7978414297103882,
+      "learning_rate": 0.0002881761978361669,
+      "loss": 4.8397,
+      "step": 68390
+    },
+    {
+      "epoch": 0.09610841703887016,
+      "grad_norm": 0.7623574137687683,
+      "learning_rate": 0.00028821835042855133,
+      "loss": 4.7205,
+      "step": 68400
+    },
+    {
+      "epoch": 0.09612246797703373,
+      "grad_norm": 0.7838755249977112,
+      "learning_rate": 0.00028826050302093576,
+      "loss": 4.8583,
+      "step": 68410
+    },
+    {
+      "epoch": 0.09613651891519732,
+      "grad_norm": 0.7429123520851135,
+      "learning_rate": 0.0002883026556133202,
+      "loss": 4.9437,
+      "step": 68420
+    },
+    {
+      "epoch": 0.0961505698533609,
+      "grad_norm": 0.7363834381103516,
+      "learning_rate": 0.0002883448082057046,
+      "loss": 4.8511,
+      "step": 68430
+    },
+    {
+      "epoch": 0.09616462079152448,
+      "grad_norm": 0.7411858439445496,
+      "learning_rate": 0.00028838696079808905,
+      "loss": 4.8572,
+      "step": 68440
+    },
+    {
+      "epoch": 0.09617867172968805,
+      "grad_norm": 0.8027516007423401,
+      "learning_rate": 0.0002884291133904735,
+      "loss": 4.7389,
+      "step": 68450
+    },
+    {
+      "epoch": 0.09619272266785163,
+      "grad_norm": 0.7942211031913757,
+      "learning_rate": 0.0002884712659828579,
+      "loss": 4.8992,
+      "step": 68460
+    },
+    {
+      "epoch": 0.0962067736060152,
+      "grad_norm": 0.7656008005142212,
+      "learning_rate": 0.00028851341857524235,
+      "loss": 4.9126,
+      "step": 68470
+    },
+    {
+      "epoch": 0.09622082454417878,
+      "grad_norm": 0.7987953424453735,
+      "learning_rate": 0.0002885555711676268,
+      "loss": 4.7477,
+      "step": 68480
+    },
+    {
+      "epoch": 0.09623487548234236,
+      "grad_norm": 0.7650423645973206,
+      "learning_rate": 0.0002885977237600112,
+      "loss": 4.8695,
+      "step": 68490
+    },
+    {
+      "epoch": 0.09624892642050593,
+      "grad_norm": 0.7581927180290222,
+      "learning_rate": 0.00028863987635239564,
+      "loss": 4.9698,
+      "step": 68500
+    },
+    {
+      "epoch": 0.09626297735866951,
+      "grad_norm": 0.7571653127670288,
+      "learning_rate": 0.00028868202894478007,
+      "loss": 4.7919,
+      "step": 68510
+    },
+    {
+      "epoch": 0.0962770282968331,
+      "grad_norm": 0.752564549446106,
+      "learning_rate": 0.0002887241815371645,
+      "loss": 4.7388,
+      "step": 68520
+    },
+    {
+      "epoch": 0.09629107923499668,
+      "grad_norm": 0.7321598529815674,
+      "learning_rate": 0.00028876633412954893,
+      "loss": 4.7807,
+      "step": 68530
+    },
+    {
+      "epoch": 0.09630513017316025,
+      "grad_norm": 0.7438421845436096,
+      "learning_rate": 0.00028880848672193336,
+      "loss": 4.8229,
+      "step": 68540
+    },
+    {
+      "epoch": 0.09631918111132383,
+      "grad_norm": 0.7429989576339722,
+      "learning_rate": 0.0002888506393143178,
+      "loss": 4.7067,
+      "step": 68550
+    },
+    {
+      "epoch": 0.0963332320494874,
+      "grad_norm": 0.749879777431488,
+      "learning_rate": 0.0002888927919067022,
+      "loss": 4.8433,
+      "step": 68560
+    },
+    {
+      "epoch": 0.09634728298765098,
+      "grad_norm": 0.8272457718849182,
+      "learning_rate": 0.00028893494449908666,
+      "loss": 4.6546,
+      "step": 68570
+    },
+    {
+      "epoch": 0.09636133392581456,
+      "grad_norm": 0.7289469242095947,
+      "learning_rate": 0.0002889770970914711,
+      "loss": 4.8699,
+      "step": 68580
+    },
+    {
+      "epoch": 0.09637538486397813,
+      "grad_norm": 0.7849155068397522,
+      "learning_rate": 0.0002890192496838555,
+      "loss": 4.5681,
+      "step": 68590
+    },
+    {
+      "epoch": 0.09638943580214171,
+      "grad_norm": 0.7392459511756897,
+      "learning_rate": 0.00028906140227623995,
+      "loss": 4.8062,
+      "step": 68600
+    },
+    {
+      "epoch": 0.09640348674030529,
+      "grad_norm": 0.771606981754303,
+      "learning_rate": 0.00028910355486862444,
+      "loss": 4.7789,
+      "step": 68610
+    },
+    {
+      "epoch": 0.09641753767846888,
+      "grad_norm": 0.7495245337486267,
+      "learning_rate": 0.0002891457074610088,
+      "loss": 4.725,
+      "step": 68620
+    },
+    {
+      "epoch": 0.09643158861663245,
+      "grad_norm": 0.758076012134552,
+      "learning_rate": 0.00028918786005339324,
+      "loss": 4.8515,
+      "step": 68630
+    },
+    {
+      "epoch": 0.09644563955479603,
+      "grad_norm": 0.7334091663360596,
+      "learning_rate": 0.0002892300126457777,
+      "loss": 5.0253,
+      "step": 68640
+    },
+    {
+      "epoch": 0.0964596904929596,
+      "grad_norm": 0.7746126651763916,
+      "learning_rate": 0.0002892721652381621,
+      "loss": 4.7502,
+      "step": 68650
+    },
+    {
+      "epoch": 0.09647374143112318,
+      "grad_norm": 0.7248702645301819,
+      "learning_rate": 0.00028931431783054654,
+      "loss": 4.9443,
+      "step": 68660
+    },
+    {
+      "epoch": 0.09648779236928676,
+      "grad_norm": 0.8012595772743225,
+      "learning_rate": 0.000289356470422931,
+      "loss": 4.8209,
+      "step": 68670
+    },
+    {
+      "epoch": 0.09650184330745033,
+      "grad_norm": 0.7670885324478149,
+      "learning_rate": 0.0002893986230153154,
+      "loss": 4.7838,
+      "step": 68680
+    },
+    {
+      "epoch": 0.09651589424561391,
+      "grad_norm": 0.7456730008125305,
+      "learning_rate": 0.00028944077560769983,
+      "loss": 4.7761,
+      "step": 68690
+    },
+    {
+      "epoch": 0.09652994518377749,
+      "grad_norm": 0.727116584777832,
+      "learning_rate": 0.00028948292820008426,
+      "loss": 4.7292,
+      "step": 68700
+    },
+    {
+      "epoch": 0.09654399612194106,
+      "grad_norm": 0.758237898349762,
+      "learning_rate": 0.0002895250807924687,
+      "loss": 4.7278,
+      "step": 68710
+    },
+    {
+      "epoch": 0.09655804706010464,
+      "grad_norm": 0.7621706128120422,
+      "learning_rate": 0.0002895672333848531,
+      "loss": 4.7813,
+      "step": 68720
+    },
+    {
+      "epoch": 0.09657209799826823,
+      "grad_norm": 0.7740551829338074,
+      "learning_rate": 0.0002896093859772376,
+      "loss": 4.8008,
+      "step": 68730
+    },
+    {
+      "epoch": 0.0965861489364318,
+      "grad_norm": 0.7670646905899048,
+      "learning_rate": 0.00028965153856962204,
+      "loss": 4.762,
+      "step": 68740
+    },
+    {
+      "epoch": 0.09660019987459538,
+      "grad_norm": 0.76221764087677,
+      "learning_rate": 0.0002896936911620064,
+      "loss": 4.6957,
+      "step": 68750
+    },
+    {
+      "epoch": 0.09661425081275896,
+      "grad_norm": 0.7280772924423218,
+      "learning_rate": 0.00028973584375439085,
+      "loss": 4.8566,
+      "step": 68760
+    },
+    {
+      "epoch": 0.09662830175092253,
+      "grad_norm": 0.7589055299758911,
+      "learning_rate": 0.0002897779963467753,
+      "loss": 4.7532,
+      "step": 68770
+    },
+    {
+      "epoch": 0.09664235268908611,
+      "grad_norm": 0.8092201948165894,
+      "learning_rate": 0.0002898201489391597,
+      "loss": 4.7692,
+      "step": 68780
+    },
+    {
+      "epoch": 0.09665640362724968,
+      "grad_norm": 0.7353342175483704,
+      "learning_rate": 0.0002898623015315442,
+      "loss": 4.8688,
+      "step": 68790
+    },
+    {
+      "epoch": 0.09667045456541326,
+      "grad_norm": 0.7660097479820251,
+      "learning_rate": 0.00028990445412392863,
+      "loss": 4.809,
+      "step": 68800
+    },
+    {
+      "epoch": 0.09668450550357684,
+      "grad_norm": 0.7710687518119812,
+      "learning_rate": 0.000289946606716313,
+      "loss": 4.7622,
+      "step": 68810
+    },
+    {
+      "epoch": 0.09669855644174041,
+      "grad_norm": 0.7906742095947266,
+      "learning_rate": 0.00028998875930869744,
+      "loss": 4.8203,
+      "step": 68820
+    },
+    {
+      "epoch": 0.096712607379904,
+      "grad_norm": 0.7781643271446228,
+      "learning_rate": 0.00029003091190108187,
+      "loss": 4.8479,
+      "step": 68830
+    },
+    {
+      "epoch": 0.09672665831806758,
+      "grad_norm": 0.7606348991394043,
+      "learning_rate": 0.0002900730644934663,
+      "loss": 4.8669,
+      "step": 68840
+    },
+    {
+      "epoch": 0.09674070925623116,
+      "grad_norm": 0.7359606623649597,
+      "learning_rate": 0.0002901152170858508,
+      "loss": 4.861,
+      "step": 68850
+    },
+    {
+      "epoch": 0.09675476019439473,
+      "grad_norm": 0.7380146384239197,
+      "learning_rate": 0.0002901573696782352,
+      "loss": 4.7876,
+      "step": 68860
+    },
+    {
+      "epoch": 0.09676881113255831,
+      "grad_norm": 0.7908768057823181,
+      "learning_rate": 0.00029019952227061965,
+      "loss": 4.765,
+      "step": 68870
+    },
+    {
+      "epoch": 0.09678286207072188,
+      "grad_norm": 0.7726009488105774,
+      "learning_rate": 0.000290241674863004,
+      "loss": 4.6684,
+      "step": 68880
+    },
+    {
+      "epoch": 0.09679691300888546,
+      "grad_norm": 0.7417333126068115,
+      "learning_rate": 0.00029028382745538845,
+      "loss": 4.8782,
+      "step": 68890
+    },
+    {
+      "epoch": 0.09681096394704904,
+      "grad_norm": 0.7703368067741394,
+      "learning_rate": 0.0002903259800477729,
+      "loss": 4.8342,
+      "step": 68900
+    },
+    {
+      "epoch": 0.09682501488521261,
+      "grad_norm": 0.774889349937439,
+      "learning_rate": 0.00029036813264015737,
+      "loss": 4.8245,
+      "step": 68910
+    },
+    {
+      "epoch": 0.09683906582337619,
+      "grad_norm": 0.7258095145225525,
+      "learning_rate": 0.0002904102852325418,
+      "loss": 4.7255,
+      "step": 68920
+    },
+    {
+      "epoch": 0.09685311676153978,
+      "grad_norm": 0.8192217946052551,
+      "learning_rate": 0.00029045243782492623,
+      "loss": 4.7965,
+      "step": 68930
+    },
+    {
+      "epoch": 0.09686716769970335,
+      "grad_norm": 0.7315029501914978,
+      "learning_rate": 0.0002904945904173106,
+      "loss": 4.6835,
+      "step": 68940
+    },
+    {
+      "epoch": 0.09688121863786693,
+      "grad_norm": 0.7798453569412231,
+      "learning_rate": 0.00029053674300969504,
+      "loss": 4.8065,
+      "step": 68950
+    },
+    {
+      "epoch": 0.09689526957603051,
+      "grad_norm": 0.7461909651756287,
+      "learning_rate": 0.00029057889560207947,
+      "loss": 4.8489,
+      "step": 68960
+    },
+    {
+      "epoch": 0.09690932051419408,
+      "grad_norm": 0.7602362036705017,
+      "learning_rate": 0.00029062104819446396,
+      "loss": 4.7368,
+      "step": 68970
+    },
+    {
+      "epoch": 0.09692337145235766,
+      "grad_norm": 0.7881399989128113,
+      "learning_rate": 0.0002906632007868484,
+      "loss": 4.8113,
+      "step": 68980
+    },
+    {
+      "epoch": 0.09693742239052124,
+      "grad_norm": 0.7365151643753052,
+      "learning_rate": 0.0002907053533792328,
+      "loss": 4.873,
+      "step": 68990
+    },
+    {
+      "epoch": 0.09695147332868481,
+      "grad_norm": 0.7622225284576416,
+      "learning_rate": 0.00029074750597161725,
+      "loss": 4.7345,
+      "step": 69000
+    },
+    {
+      "epoch": 0.09696552426684839,
+      "grad_norm": 0.7960168123245239,
+      "learning_rate": 0.00029078965856400163,
+      "loss": 4.6768,
+      "step": 69010
+    },
+    {
+      "epoch": 0.09697957520501196,
+      "grad_norm": 0.7706853747367859,
+      "learning_rate": 0.00029083181115638606,
+      "loss": 4.8081,
+      "step": 69020
+    },
+    {
+      "epoch": 0.09699362614317554,
+      "grad_norm": 0.7176666259765625,
+      "learning_rate": 0.00029087396374877054,
+      "loss": 4.8542,
+      "step": 69030
+    },
+    {
+      "epoch": 0.09700767708133913,
+      "grad_norm": 0.7501884698867798,
+      "learning_rate": 0.000290916116341155,
+      "loss": 4.7623,
+      "step": 69040
+    },
+    {
+      "epoch": 0.0970217280195027,
+      "grad_norm": 0.7589862942695618,
+      "learning_rate": 0.0002909582689335394,
+      "loss": 4.8451,
+      "step": 69050
+    },
+    {
+      "epoch": 0.09703577895766628,
+      "grad_norm": 0.7779807448387146,
+      "learning_rate": 0.00029100042152592384,
+      "loss": 4.7506,
+      "step": 69060
+    },
+    {
+      "epoch": 0.09704982989582986,
+      "grad_norm": 0.745506227016449,
+      "learning_rate": 0.0002910425741183082,
+      "loss": 4.7947,
+      "step": 69070
+    },
+    {
+      "epoch": 0.09706388083399344,
+      "grad_norm": 0.7517091035842896,
+      "learning_rate": 0.00029108472671069265,
+      "loss": 4.7302,
+      "step": 69080
+    },
+    {
+      "epoch": 0.09707793177215701,
+      "grad_norm": 0.7594832181930542,
+      "learning_rate": 0.00029112687930307713,
+      "loss": 4.7494,
+      "step": 69090
+    },
+    {
+      "epoch": 0.09709198271032059,
+      "grad_norm": 0.7366914749145508,
+      "learning_rate": 0.00029116903189546156,
+      "loss": 4.7735,
+      "step": 69100
+    },
+    {
+      "epoch": 0.09710603364848416,
+      "grad_norm": 0.7551494836807251,
+      "learning_rate": 0.000291211184487846,
+      "loss": 4.8079,
+      "step": 69110
+    },
+    {
+      "epoch": 0.09712008458664774,
+      "grad_norm": 0.7677338123321533,
+      "learning_rate": 0.0002912533370802304,
+      "loss": 4.7275,
+      "step": 69120
+    },
+    {
+      "epoch": 0.09713413552481132,
+      "grad_norm": 0.7898170948028564,
+      "learning_rate": 0.00029129548967261486,
+      "loss": 4.7946,
+      "step": 69130
+    },
+    {
+      "epoch": 0.0971481864629749,
+      "grad_norm": 0.7417213320732117,
+      "learning_rate": 0.00029133764226499923,
+      "loss": 4.7658,
+      "step": 69140
+    },
+    {
+      "epoch": 0.09716223740113848,
+      "grad_norm": 0.7414938807487488,
+      "learning_rate": 0.0002913797948573837,
+      "loss": 4.8031,
+      "step": 69150
+    },
+    {
+      "epoch": 0.09717628833930206,
+      "grad_norm": 0.7222784161567688,
+      "learning_rate": 0.00029142194744976815,
+      "loss": 4.7848,
+      "step": 69160
+    },
+    {
+      "epoch": 0.09719033927746563,
+      "grad_norm": 0.7901402115821838,
+      "learning_rate": 0.0002914641000421526,
+      "loss": 4.7291,
+      "step": 69170
+    },
+    {
+      "epoch": 0.09720439021562921,
+      "grad_norm": 0.7619618773460388,
+      "learning_rate": 0.000291506252634537,
+      "loss": 4.772,
+      "step": 69180
+    },
+    {
+      "epoch": 0.09721844115379279,
+      "grad_norm": 0.7558301687240601,
+      "learning_rate": 0.00029154840522692144,
+      "loss": 4.6799,
+      "step": 69190
+    },
+    {
+      "epoch": 0.09723249209195636,
+      "grad_norm": 0.7339478731155396,
+      "learning_rate": 0.0002915905578193059,
+      "loss": 4.699,
+      "step": 69200
+    },
+    {
+      "epoch": 0.09724654303011994,
+      "grad_norm": 0.7412484288215637,
+      "learning_rate": 0.0002916327104116903,
+      "loss": 4.7946,
+      "step": 69210
+    },
+    {
+      "epoch": 0.09726059396828352,
+      "grad_norm": 0.7897161245346069,
+      "learning_rate": 0.00029167486300407474,
+      "loss": 4.7929,
+      "step": 69220
+    },
+    {
+      "epoch": 0.09727464490644709,
+      "grad_norm": 0.7667302489280701,
+      "learning_rate": 0.00029171701559645917,
+      "loss": 4.7734,
+      "step": 69230
+    },
+    {
+      "epoch": 0.09728869584461068,
+      "grad_norm": 0.7448514699935913,
+      "learning_rate": 0.0002917591681888436,
+      "loss": 4.7505,
+      "step": 69240
+    },
+    {
+      "epoch": 0.09730274678277426,
+      "grad_norm": 0.7377836108207703,
+      "learning_rate": 0.00029180132078122803,
+      "loss": 4.6303,
+      "step": 69250
+    },
+    {
+      "epoch": 0.09731679772093783,
+      "grad_norm": 0.742834746837616,
+      "learning_rate": 0.00029184347337361246,
+      "loss": 4.8095,
+      "step": 69260
+    },
+    {
+      "epoch": 0.09733084865910141,
+      "grad_norm": 0.7830805778503418,
+      "learning_rate": 0.0002918856259659969,
+      "loss": 4.7535,
+      "step": 69270
+    },
+    {
+      "epoch": 0.09734489959726499,
+      "grad_norm": 0.7334023118019104,
+      "learning_rate": 0.0002919277785583813,
+      "loss": 4.8267,
+      "step": 69280
+    },
+    {
+      "epoch": 0.09735895053542856,
+      "grad_norm": 0.7822585105895996,
+      "learning_rate": 0.00029196993115076576,
+      "loss": 4.743,
+      "step": 69290
+    },
+    {
+      "epoch": 0.09737300147359214,
+      "grad_norm": 0.7647944092750549,
+      "learning_rate": 0.0002920120837431502,
+      "loss": 4.7674,
+      "step": 69300
+    },
+    {
+      "epoch": 0.09738705241175571,
+      "grad_norm": 0.7947744727134705,
+      "learning_rate": 0.0002920542363355346,
+      "loss": 4.7308,
+      "step": 69310
+    },
+    {
+      "epoch": 0.09740110334991929,
+      "grad_norm": 0.7814948558807373,
+      "learning_rate": 0.0002920921736686806,
+      "loss": 4.8563,
+      "step": 69320
+    },
+    {
+      "epoch": 0.09741515428808287,
+      "grad_norm": 0.7444278001785278,
+      "learning_rate": 0.00029213432626106503,
+      "loss": 4.7429,
+      "step": 69330
+    },
+    {
+      "epoch": 0.09742920522624644,
+      "grad_norm": 0.7391352653503418,
+      "learning_rate": 0.00029217647885344946,
+      "loss": 4.6809,
+      "step": 69340
+    },
+    {
+      "epoch": 0.09744325616441003,
+      "grad_norm": 0.7300538420677185,
+      "learning_rate": 0.0002922186314458339,
+      "loss": 4.8616,
+      "step": 69350
+    },
+    {
+      "epoch": 0.09745730710257361,
+      "grad_norm": 0.7259089350700378,
+      "learning_rate": 0.0002922607840382183,
+      "loss": 4.7652,
+      "step": 69360
+    },
+    {
+      "epoch": 0.09747135804073719,
+      "grad_norm": 0.7582126259803772,
+      "learning_rate": 0.00029230293663060276,
+      "loss": 4.7776,
+      "step": 69370
+    },
+    {
+      "epoch": 0.09748540897890076,
+      "grad_norm": 0.7297702431678772,
+      "learning_rate": 0.0002923450892229872,
+      "loss": 4.7676,
+      "step": 69380
+    },
+    {
+      "epoch": 0.09749945991706434,
+      "grad_norm": 0.753105878829956,
+      "learning_rate": 0.0002923872418153716,
+      "loss": 4.7513,
+      "step": 69390
+    },
+    {
+      "epoch": 0.09751351085522791,
+      "grad_norm": 0.7406522631645203,
+      "learning_rate": 0.00029242939440775605,
+      "loss": 4.8253,
+      "step": 69400
+    },
+    {
+      "epoch": 0.09752756179339149,
+      "grad_norm": 0.8035889863967896,
+      "learning_rate": 0.0002924715470001405,
+      "loss": 4.7889,
+      "step": 69410
+    },
+    {
+      "epoch": 0.09754161273155507,
+      "grad_norm": 0.7537881731987,
+      "learning_rate": 0.0002925136995925249,
+      "loss": 4.8387,
+      "step": 69420
+    },
+    {
+      "epoch": 0.09755566366971864,
+      "grad_norm": 0.7590875625610352,
+      "learning_rate": 0.00029255585218490934,
+      "loss": 4.8179,
+      "step": 69430
+    },
+    {
+      "epoch": 0.09756971460788222,
+      "grad_norm": 0.7489945292472839,
+      "learning_rate": 0.0002925980047772938,
+      "loss": 4.8359,
+      "step": 69440
+    },
+    {
+      "epoch": 0.09758376554604581,
+      "grad_norm": 0.7501441240310669,
+      "learning_rate": 0.0002926401573696782,
+      "loss": 4.7967,
+      "step": 69450
+    },
+    {
+      "epoch": 0.09759781648420938,
+      "grad_norm": 0.7316177487373352,
+      "learning_rate": 0.00029268230996206264,
+      "loss": 4.772,
+      "step": 69460
+    },
+    {
+      "epoch": 0.09761186742237296,
+      "grad_norm": 0.7441616654396057,
+      "learning_rate": 0.00029272446255444707,
+      "loss": 4.7145,
+      "step": 69470
+    },
+    {
+      "epoch": 0.09762591836053654,
+      "grad_norm": 0.733923077583313,
+      "learning_rate": 0.0002927666151468315,
+      "loss": 4.8215,
+      "step": 69480
+    },
+    {
+      "epoch": 0.09763996929870011,
+      "grad_norm": 0.7444280385971069,
+      "learning_rate": 0.00029280876773921593,
+      "loss": 4.8405,
+      "step": 69490
+    },
+    {
+      "epoch": 0.09765402023686369,
+      "grad_norm": 0.7458165287971497,
+      "learning_rate": 0.00029285092033160036,
+      "loss": 4.7594,
+      "step": 69500
+    },
+    {
+      "epoch": 0.09766807117502727,
+      "grad_norm": 0.7323271632194519,
+      "learning_rate": 0.0002928930729239848,
+      "loss": 4.6669,
+      "step": 69510
+    },
+    {
+      "epoch": 0.09768212211319084,
+      "grad_norm": 0.7334503531455994,
+      "learning_rate": 0.0002929352255163692,
+      "loss": 4.8641,
+      "step": 69520
+    },
+    {
+      "epoch": 0.09769617305135442,
+      "grad_norm": 0.747018039226532,
+      "learning_rate": 0.00029297737810875365,
+      "loss": 4.8795,
+      "step": 69530
+    },
+    {
+      "epoch": 0.097710223989518,
+      "grad_norm": 0.7300179600715637,
+      "learning_rate": 0.0002930195307011381,
+      "loss": 4.6616,
+      "step": 69540
+    },
+    {
+      "epoch": 0.09772427492768158,
+      "grad_norm": 0.7466294765472412,
+      "learning_rate": 0.0002930616832935225,
+      "loss": 4.8153,
+      "step": 69550
+    },
+    {
+      "epoch": 0.09773832586584516,
+      "grad_norm": 0.7623385787010193,
+      "learning_rate": 0.00029310383588590695,
+      "loss": 4.7444,
+      "step": 69560
+    },
+    {
+      "epoch": 0.09775237680400874,
+      "grad_norm": 0.7668961882591248,
+      "learning_rate": 0.0002931459884782914,
+      "loss": 4.8579,
+      "step": 69570
+    },
+    {
+      "epoch": 0.09776642774217231,
+      "grad_norm": 0.7659717202186584,
+      "learning_rate": 0.0002931881410706758,
+      "loss": 4.6188,
+      "step": 69580
+    },
+    {
+      "epoch": 0.09778047868033589,
+      "grad_norm": 0.7393763065338135,
+      "learning_rate": 0.00029323029366306024,
+      "loss": 4.8071,
+      "step": 69590
+    },
+    {
+      "epoch": 0.09779452961849947,
+      "grad_norm": 0.7585105299949646,
+      "learning_rate": 0.00029327244625544467,
+      "loss": 4.7279,
+      "step": 69600
+    },
+    {
+      "epoch": 0.09780858055666304,
+      "grad_norm": 0.9059628844261169,
+      "learning_rate": 0.0002933145988478291,
+      "loss": 4.931,
+      "step": 69610
+    },
+    {
+      "epoch": 0.09782263149482662,
+      "grad_norm": 0.7281200289726257,
+      "learning_rate": 0.00029335675144021354,
+      "loss": 4.8205,
+      "step": 69620
+    },
+    {
+      "epoch": 0.0978366824329902,
+      "grad_norm": 0.7609696984291077,
+      "learning_rate": 0.00029339890403259797,
+      "loss": 4.8716,
+      "step": 69630
+    },
+    {
+      "epoch": 0.09785073337115377,
+      "grad_norm": 0.7602007389068604,
+      "learning_rate": 0.0002934410566249824,
+      "loss": 4.8642,
+      "step": 69640
+    },
+    {
+      "epoch": 0.09786478430931735,
+      "grad_norm": 0.7322178483009338,
+      "learning_rate": 0.00029348320921736683,
+      "loss": 4.8271,
+      "step": 69650
+    },
+    {
+      "epoch": 0.09787883524748094,
+      "grad_norm": 0.757675290107727,
+      "learning_rate": 0.00029352536180975126,
+      "loss": 4.8059,
+      "step": 69660
+    },
+    {
+      "epoch": 0.09789288618564451,
+      "grad_norm": 0.770977258682251,
+      "learning_rate": 0.0002935675144021357,
+      "loss": 4.7335,
+      "step": 69670
+    },
+    {
+      "epoch": 0.09790693712380809,
+      "grad_norm": 0.7639873623847961,
+      "learning_rate": 0.0002936096669945201,
+      "loss": 4.7993,
+      "step": 69680
+    },
+    {
+      "epoch": 0.09792098806197166,
+      "grad_norm": 0.7296125888824463,
+      "learning_rate": 0.00029365181958690455,
+      "loss": 4.9388,
+      "step": 69690
+    },
+    {
+      "epoch": 0.09793503900013524,
+      "grad_norm": 0.7537563443183899,
+      "learning_rate": 0.000293693972179289,
+      "loss": 4.8099,
+      "step": 69700
+    },
+    {
+      "epoch": 0.09794908993829882,
+      "grad_norm": 0.7601600289344788,
+      "learning_rate": 0.0002937361247716734,
+      "loss": 4.7219,
+      "step": 69710
+    },
+    {
+      "epoch": 0.09796314087646239,
+      "grad_norm": 0.7119511365890503,
+      "learning_rate": 0.00029377827736405785,
+      "loss": 4.7607,
+      "step": 69720
+    },
+    {
+      "epoch": 0.09797719181462597,
+      "grad_norm": 0.7661904096603394,
+      "learning_rate": 0.0002938204299564423,
+      "loss": 4.7596,
+      "step": 69730
+    },
+    {
+      "epoch": 0.09799124275278955,
+      "grad_norm": 0.7884121537208557,
+      "learning_rate": 0.00029386258254882676,
+      "loss": 4.7573,
+      "step": 69740
+    },
+    {
+      "epoch": 0.09800529369095312,
+      "grad_norm": 0.7419071793556213,
+      "learning_rate": 0.00029390473514121114,
+      "loss": 4.852,
+      "step": 69750
+    },
+    {
+      "epoch": 0.09801934462911671,
+      "grad_norm": 0.8159481287002563,
+      "learning_rate": 0.00029394688773359557,
+      "loss": 4.6828,
+      "step": 69760
+    },
+    {
+      "epoch": 0.09803339556728029,
+      "grad_norm": 0.755358874797821,
+      "learning_rate": 0.00029398904032598,
+      "loss": 4.7223,
+      "step": 69770
+    },
+    {
+      "epoch": 0.09804744650544386,
+      "grad_norm": 0.7439599633216858,
+      "learning_rate": 0.00029403119291836443,
+      "loss": 4.7543,
+      "step": 69780
+    },
+    {
+      "epoch": 0.09806149744360744,
+      "grad_norm": 0.7651283740997314,
+      "learning_rate": 0.0002940733455107489,
+      "loss": 4.8595,
+      "step": 69790
+    },
+    {
+      "epoch": 0.09807554838177102,
+      "grad_norm": 0.7742921113967896,
+      "learning_rate": 0.00029411549810313335,
+      "loss": 4.6861,
+      "step": 69800
+    },
+    {
+      "epoch": 0.09808959931993459,
+      "grad_norm": 0.7510424256324768,
+      "learning_rate": 0.00029415765069551773,
+      "loss": 4.7856,
+      "step": 69810
+    },
+    {
+      "epoch": 0.09810365025809817,
+      "grad_norm": 0.7651505470275879,
+      "learning_rate": 0.00029419980328790216,
+      "loss": 4.7194,
+      "step": 69820
+    },
+    {
+      "epoch": 0.09811770119626174,
+      "grad_norm": 0.7609485387802124,
+      "learning_rate": 0.0002942419558802866,
+      "loss": 4.7036,
+      "step": 69830
+    },
+    {
+      "epoch": 0.09813175213442532,
+      "grad_norm": 0.7432272434234619,
+      "learning_rate": 0.000294284108472671,
+      "loss": 4.7723,
+      "step": 69840
+    },
+    {
+      "epoch": 0.0981458030725889,
+      "grad_norm": 0.7429347038269043,
+      "learning_rate": 0.0002943262610650555,
+      "loss": 4.7675,
+      "step": 69850
+    },
+    {
+      "epoch": 0.09815985401075249,
+      "grad_norm": 0.7958447933197021,
+      "learning_rate": 0.00029436841365743994,
+      "loss": 4.7454,
+      "step": 69860
+    },
+    {
+      "epoch": 0.09817390494891606,
+      "grad_norm": 0.7396420836448669,
+      "learning_rate": 0.00029441056624982437,
+      "loss": 4.6856,
+      "step": 69870
+    },
+    {
+      "epoch": 0.09818795588707964,
+      "grad_norm": 0.7574613690376282,
+      "learning_rate": 0.00029445271884220875,
+      "loss": 4.8366,
+      "step": 69880
+    },
+    {
+      "epoch": 0.09820200682524322,
+      "grad_norm": 0.7506481409072876,
+      "learning_rate": 0.0002944948714345932,
+      "loss": 4.6661,
+      "step": 69890
+    },
+    {
+      "epoch": 0.09821605776340679,
+      "grad_norm": 0.7857638001441956,
+      "learning_rate": 0.0002945370240269776,
+      "loss": 4.7884,
+      "step": 69900
+    },
+    {
+      "epoch": 0.09823010870157037,
+      "grad_norm": 0.7581502199172974,
+      "learning_rate": 0.0002945791766193621,
+      "loss": 4.7728,
+      "step": 69910
+    },
+    {
+      "epoch": 0.09824415963973394,
+      "grad_norm": 0.7380070090293884,
+      "learning_rate": 0.0002946213292117465,
+      "loss": 4.7034,
+      "step": 69920
+    },
+    {
+      "epoch": 0.09825821057789752,
+      "grad_norm": 0.719098687171936,
+      "learning_rate": 0.00029466348180413096,
+      "loss": 4.8379,
+      "step": 69930
+    },
+    {
+      "epoch": 0.0982722615160611,
+      "grad_norm": 0.7294281125068665,
+      "learning_rate": 0.00029470563439651533,
+      "loss": 4.8387,
+      "step": 69940
+    },
+    {
+      "epoch": 0.09828631245422467,
+      "grad_norm": 0.7245777249336243,
+      "learning_rate": 0.00029474778698889976,
+      "loss": 4.9148,
+      "step": 69950
+    },
+    {
+      "epoch": 0.09830036339238825,
+      "grad_norm": 0.7560135126113892,
+      "learning_rate": 0.0002947899395812842,
+      "loss": 4.8186,
+      "step": 69960
+    },
+    {
+      "epoch": 0.09831441433055184,
+      "grad_norm": 0.7155381441116333,
+      "learning_rate": 0.0002948320921736687,
+      "loss": 4.8529,
+      "step": 69970
+    },
+    {
+      "epoch": 0.09832846526871541,
+      "grad_norm": 0.751598060131073,
+      "learning_rate": 0.0002948742447660531,
+      "loss": 4.6971,
+      "step": 69980
+    },
+    {
+      "epoch": 0.09834251620687899,
+      "grad_norm": 0.7454250454902649,
+      "learning_rate": 0.00029491639735843754,
+      "loss": 4.8887,
+      "step": 69990
+    },
+    {
+      "epoch": 0.09835656714504257,
+      "grad_norm": 0.745444118976593,
+      "learning_rate": 0.000294958549950822,
+      "loss": 4.8025,
+      "step": 70000
+    },
+    {
+      "epoch": 0.09837061808320614,
+      "grad_norm": 0.7454139590263367,
+      "learning_rate": 0.00029500070254320635,
+      "loss": 4.8269,
+      "step": 70010
+    },
+    {
+      "epoch": 0.09838466902136972,
+      "grad_norm": 0.7288814187049866,
+      "learning_rate": 0.0002950428551355908,
+      "loss": 4.8821,
+      "step": 70020
+    },
+    {
+      "epoch": 0.0983987199595333,
+      "grad_norm": 0.7203482985496521,
+      "learning_rate": 0.00029508500772797527,
+      "loss": 4.7518,
+      "step": 70030
+    },
+    {
+      "epoch": 0.09841277089769687,
+      "grad_norm": 0.7820789813995361,
+      "learning_rate": 0.0002951271603203597,
+      "loss": 4.7813,
+      "step": 70040
+    },
+    {
+      "epoch": 0.09842682183586045,
+      "grad_norm": 0.7446337938308716,
+      "learning_rate": 0.00029516931291274413,
+      "loss": 4.7992,
+      "step": 70050
+    },
+    {
+      "epoch": 0.09844087277402402,
+      "grad_norm": 0.7281479835510254,
+      "learning_rate": 0.00029521146550512856,
+      "loss": 4.8715,
+      "step": 70060
+    },
+    {
+      "epoch": 0.09845492371218761,
+      "grad_norm": 0.7503718137741089,
+      "learning_rate": 0.00029525361809751294,
+      "loss": 4.7856,
+      "step": 70070
+    },
+    {
+      "epoch": 0.09846897465035119,
+      "grad_norm": 0.7504997849464417,
+      "learning_rate": 0.00029529577068989737,
+      "loss": 4.7455,
+      "step": 70080
+    },
+    {
+      "epoch": 0.09848302558851477,
+      "grad_norm": 0.755368709564209,
+      "learning_rate": 0.00029533792328228185,
+      "loss": 4.7809,
+      "step": 70090
+    },
+    {
+      "epoch": 0.09849707652667834,
+      "grad_norm": 0.7570355534553528,
+      "learning_rate": 0.0002953800758746663,
+      "loss": 4.7163,
+      "step": 70100
+    },
+    {
+      "epoch": 0.09851112746484192,
+      "grad_norm": 0.7263005971908569,
+      "learning_rate": 0.0002954222284670507,
+      "loss": 4.8162,
+      "step": 70110
+    },
+    {
+      "epoch": 0.0985251784030055,
+      "grad_norm": 0.7369113564491272,
+      "learning_rate": 0.00029546438105943515,
+      "loss": 4.7484,
+      "step": 70120
+    },
+    {
+      "epoch": 0.09853922934116907,
+      "grad_norm": 0.7835990190505981,
+      "learning_rate": 0.0002955065336518196,
+      "loss": 4.7219,
+      "step": 70130
+    },
+    {
+      "epoch": 0.09855328027933265,
+      "grad_norm": 0.7315928936004639,
+      "learning_rate": 0.00029554868624420396,
+      "loss": 4.7446,
+      "step": 70140
+    },
+    {
+      "epoch": 0.09856733121749622,
+      "grad_norm": 0.7778264284133911,
+      "learning_rate": 0.00029559083883658844,
+      "loss": 4.7838,
+      "step": 70150
+    },
+    {
+      "epoch": 0.0985813821556598,
+      "grad_norm": 0.7490993142127991,
+      "learning_rate": 0.00029563299142897287,
+      "loss": 4.7523,
+      "step": 70160
+    },
+    {
+      "epoch": 0.09859543309382339,
+      "grad_norm": 0.728198766708374,
+      "learning_rate": 0.0002956751440213573,
+      "loss": 4.7791,
+      "step": 70170
+    },
+    {
+      "epoch": 0.09860948403198697,
+      "grad_norm": 0.7419266700744629,
+      "learning_rate": 0.00029571729661374173,
+      "loss": 4.7178,
+      "step": 70180
+    },
+    {
+      "epoch": 0.09862353497015054,
+      "grad_norm": 0.7586491703987122,
+      "learning_rate": 0.00029575944920612617,
+      "loss": 4.7541,
+      "step": 70190
+    },
+    {
+      "epoch": 0.09863758590831412,
+      "grad_norm": 0.7712236046791077,
+      "learning_rate": 0.00029580160179851054,
+      "loss": 4.7473,
+      "step": 70200
+    },
+    {
+      "epoch": 0.0986516368464777,
+      "grad_norm": 0.7614474892616272,
+      "learning_rate": 0.00029584375439089503,
+      "loss": 4.7263,
+      "step": 70210
+    },
+    {
+      "epoch": 0.09866568778464127,
+      "grad_norm": 0.7652673125267029,
+      "learning_rate": 0.00029588590698327946,
+      "loss": 4.7051,
+      "step": 70220
+    },
+    {
+      "epoch": 0.09867973872280485,
+      "grad_norm": 0.7385839223861694,
+      "learning_rate": 0.0002959280595756639,
+      "loss": 4.7849,
+      "step": 70230
+    },
+    {
+      "epoch": 0.09869378966096842,
+      "grad_norm": 0.8170595765113831,
+      "learning_rate": 0.0002959702121680483,
+      "loss": 4.8655,
+      "step": 70240
+    },
+    {
+      "epoch": 0.098707840599132,
+      "grad_norm": 0.7230207324028015,
+      "learning_rate": 0.00029601236476043275,
+      "loss": 4.8548,
+      "step": 70250
+    },
+    {
+      "epoch": 0.09872189153729558,
+      "grad_norm": 0.7606765031814575,
+      "learning_rate": 0.0002960545173528172,
+      "loss": 4.7995,
+      "step": 70260
+    },
+    {
+      "epoch": 0.09873594247545917,
+      "grad_norm": 0.7488109469413757,
+      "learning_rate": 0.0002960966699452016,
+      "loss": 4.827,
+      "step": 70270
+    },
+    {
+      "epoch": 0.09874999341362274,
+      "grad_norm": 0.7458518743515015,
+      "learning_rate": 0.00029613882253758605,
+      "loss": 4.8889,
+      "step": 70280
+    },
+    {
+      "epoch": 0.09876404435178632,
+      "grad_norm": 0.7448275089263916,
+      "learning_rate": 0.0002961809751299705,
+      "loss": 4.7753,
+      "step": 70290
+    },
+    {
+      "epoch": 0.0987780952899499,
+      "grad_norm": 0.7387959957122803,
+      "learning_rate": 0.0002962231277223549,
+      "loss": 4.738,
+      "step": 70300
+    },
+    {
+      "epoch": 0.09879214622811347,
+      "grad_norm": 0.7459929585456848,
+      "learning_rate": 0.00029626528031473934,
+      "loss": 4.6784,
+      "step": 70310
+    },
+    {
+      "epoch": 0.09880619716627705,
+      "grad_norm": 0.7338137030601501,
+      "learning_rate": 0.00029630743290712377,
+      "loss": 4.8043,
+      "step": 70320
+    },
+    {
+      "epoch": 0.09882024810444062,
+      "grad_norm": 0.7475922107696533,
+      "learning_rate": 0.0002963495854995082,
+      "loss": 4.8699,
+      "step": 70330
+    },
+    {
+      "epoch": 0.0988342990426042,
+      "grad_norm": 0.7375343441963196,
+      "learning_rate": 0.00029639173809189263,
+      "loss": 4.7544,
+      "step": 70340
+    },
+    {
+      "epoch": 0.09884834998076777,
+      "grad_norm": 0.7328280210494995,
+      "learning_rate": 0.00029643389068427706,
+      "loss": 4.7604,
+      "step": 70350
+    },
+    {
+      "epoch": 0.09886240091893135,
+      "grad_norm": 0.723840594291687,
+      "learning_rate": 0.0002964760432766615,
+      "loss": 4.888,
+      "step": 70360
+    },
+    {
+      "epoch": 0.09887645185709493,
+      "grad_norm": 0.7332075238227844,
+      "learning_rate": 0.0002965181958690459,
+      "loss": 4.7549,
+      "step": 70370
+    },
+    {
+      "epoch": 0.09889050279525852,
+      "grad_norm": 0.7340726852416992,
+      "learning_rate": 0.00029656034846143036,
+      "loss": 4.7467,
+      "step": 70380
+    },
+    {
+      "epoch": 0.0989045537334221,
+      "grad_norm": 0.7343766093254089,
+      "learning_rate": 0.0002966025010538148,
+      "loss": 4.7236,
+      "step": 70390
+    },
+    {
+      "epoch": 0.09891860467158567,
+      "grad_norm": 0.7564096450805664,
+      "learning_rate": 0.0002966446536461992,
+      "loss": 4.6312,
+      "step": 70400
+    },
+    {
+      "epoch": 0.09893265560974925,
+      "grad_norm": 0.7509664297103882,
+      "learning_rate": 0.00029668680623858365,
+      "loss": 4.7879,
+      "step": 70410
+    },
+    {
+      "epoch": 0.09894670654791282,
+      "grad_norm": 0.7478966116905212,
+      "learning_rate": 0.0002967289588309681,
+      "loss": 4.8726,
+      "step": 70420
+    },
+    {
+      "epoch": 0.0989607574860764,
+      "grad_norm": 0.7956492304801941,
+      "learning_rate": 0.0002967711114233525,
+      "loss": 4.8496,
+      "step": 70430
+    },
+    {
+      "epoch": 0.09897480842423997,
+      "grad_norm": 0.7533037662506104,
+      "learning_rate": 0.00029681326401573694,
+      "loss": 4.8447,
+      "step": 70440
+    },
+    {
+      "epoch": 0.09898885936240355,
+      "grad_norm": 0.7333579063415527,
+      "learning_rate": 0.0002968554166081214,
+      "loss": 4.7172,
+      "step": 70450
+    },
+    {
+      "epoch": 0.09900291030056713,
+      "grad_norm": 0.745343029499054,
+      "learning_rate": 0.0002968975692005058,
+      "loss": 4.7113,
+      "step": 70460
+    },
+    {
+      "epoch": 0.0990169612387307,
+      "grad_norm": 0.7377544045448303,
+      "learning_rate": 0.00029693972179289024,
+      "loss": 5.0384,
+      "step": 70470
+    },
+    {
+      "epoch": 0.09903101217689429,
+      "grad_norm": 0.756903886795044,
+      "learning_rate": 0.00029698187438527467,
+      "loss": 4.6786,
+      "step": 70480
+    },
+    {
+      "epoch": 0.09904506311505787,
+      "grad_norm": 0.7782666087150574,
+      "learning_rate": 0.0002970240269776591,
+      "loss": 4.8214,
+      "step": 70490
+    },
+    {
+      "epoch": 0.09905911405322144,
+      "grad_norm": 0.7528187036514282,
+      "learning_rate": 0.00029706617957004353,
+      "loss": 4.7897,
+      "step": 70500
+    },
+    {
+      "epoch": 0.09907316499138502,
+      "grad_norm": 0.7484343647956848,
+      "learning_rate": 0.00029710833216242796,
+      "loss": 4.7155,
+      "step": 70510
+    },
+    {
+      "epoch": 0.0990872159295486,
+      "grad_norm": 0.7250393629074097,
+      "learning_rate": 0.0002971504847548124,
+      "loss": 4.8713,
+      "step": 70520
+    },
+    {
+      "epoch": 0.09910126686771217,
+      "grad_norm": 0.7773247957229614,
+      "learning_rate": 0.0002971926373471968,
+      "loss": 4.7164,
+      "step": 70530
+    },
+    {
+      "epoch": 0.09911531780587575,
+      "grad_norm": 0.729132890701294,
+      "learning_rate": 0.00029723478993958126,
+      "loss": 4.7736,
+      "step": 70540
+    },
+    {
+      "epoch": 0.09912936874403933,
+      "grad_norm": 0.7351367473602295,
+      "learning_rate": 0.0002972769425319657,
+      "loss": 4.9012,
+      "step": 70550
+    },
+    {
+      "epoch": 0.0991434196822029,
+      "grad_norm": 0.7388243675231934,
+      "learning_rate": 0.0002973190951243501,
+      "loss": 4.6448,
+      "step": 70560
+    },
+    {
+      "epoch": 0.09915747062036648,
+      "grad_norm": 0.7806434631347656,
+      "learning_rate": 0.00029736124771673455,
+      "loss": 4.7922,
+      "step": 70570
+    },
+    {
+      "epoch": 0.09917152155853007,
+      "grad_norm": 0.788178026676178,
+      "learning_rate": 0.000297403400309119,
+      "loss": 4.7924,
+      "step": 70580
+    },
+    {
+      "epoch": 0.09918557249669364,
+      "grad_norm": 0.7238157391548157,
+      "learning_rate": 0.0002974455529015034,
+      "loss": 4.8356,
+      "step": 70590
+    },
+    {
+      "epoch": 0.09919962343485722,
+      "grad_norm": 0.7454667687416077,
+      "learning_rate": 0.00029748770549388784,
+      "loss": 4.7332,
+      "step": 70600
+    },
+    {
+      "epoch": 0.0992136743730208,
+      "grad_norm": 0.7305548191070557,
+      "learning_rate": 0.0002975298580862723,
+      "loss": 4.8754,
+      "step": 70610
+    },
+    {
+      "epoch": 0.09922772531118437,
+      "grad_norm": 0.7459691762924194,
+      "learning_rate": 0.0002975720106786567,
+      "loss": 4.7751,
+      "step": 70620
+    },
+    {
+      "epoch": 0.09924177624934795,
+      "grad_norm": 0.7162081599235535,
+      "learning_rate": 0.00029761416327104114,
+      "loss": 4.8116,
+      "step": 70630
+    },
+    {
+      "epoch": 0.09925582718751153,
+      "grad_norm": 0.7404140830039978,
+      "learning_rate": 0.00029765631586342557,
+      "loss": 4.8285,
+      "step": 70640
+    },
+    {
+      "epoch": 0.0992698781256751,
+      "grad_norm": 0.7442372441291809,
+      "learning_rate": 0.00029769846845581,
+      "loss": 4.7367,
+      "step": 70650
+    },
+    {
+      "epoch": 0.09928392906383868,
+      "grad_norm": 0.7542151212692261,
+      "learning_rate": 0.00029774062104819443,
+      "loss": 4.7881,
+      "step": 70660
+    },
+    {
+      "epoch": 0.09929798000200225,
+      "grad_norm": 0.7247451543807983,
+      "learning_rate": 0.00029778277364057886,
+      "loss": 4.7453,
+      "step": 70670
+    },
+    {
+      "epoch": 0.09931203094016583,
+      "grad_norm": 0.7268667221069336,
+      "learning_rate": 0.0002978249262329633,
+      "loss": 4.7807,
+      "step": 70680
+    },
+    {
+      "epoch": 0.09932608187832942,
+      "grad_norm": 0.7007578611373901,
+      "learning_rate": 0.0002978670788253477,
+      "loss": 4.8148,
+      "step": 70690
+    },
+    {
+      "epoch": 0.099340132816493,
+      "grad_norm": 0.7512852549552917,
+      "learning_rate": 0.00029790923141773216,
+      "loss": 4.7867,
+      "step": 70700
+    },
+    {
+      "epoch": 0.09935418375465657,
+      "grad_norm": 0.7661580443382263,
+      "learning_rate": 0.0002979513840101166,
+      "loss": 4.721,
+      "step": 70710
+    },
+    {
+      "epoch": 0.09936823469282015,
+      "grad_norm": 0.7639215588569641,
+      "learning_rate": 0.000297993536602501,
+      "loss": 4.7987,
+      "step": 70720
+    },
+    {
+      "epoch": 0.09938228563098372,
+      "grad_norm": 0.7315471172332764,
+      "learning_rate": 0.00029803568919488545,
+      "loss": 4.9177,
+      "step": 70730
+    },
+    {
+      "epoch": 0.0993963365691473,
+      "grad_norm": 0.7605272531509399,
+      "learning_rate": 0.0002980778417872699,
+      "loss": 4.8585,
+      "step": 70740
+    },
+    {
+      "epoch": 0.09941038750731088,
+      "grad_norm": 0.71612948179245,
+      "learning_rate": 0.0002981199943796543,
+      "loss": 4.7394,
+      "step": 70750
+    },
+    {
+      "epoch": 0.09942443844547445,
+      "grad_norm": 0.7437877058982849,
+      "learning_rate": 0.00029816214697203874,
+      "loss": 4.7909,
+      "step": 70760
+    },
+    {
+      "epoch": 0.09943848938363803,
+      "grad_norm": 0.7186173796653748,
+      "learning_rate": 0.0002982042995644232,
+      "loss": 4.7169,
+      "step": 70770
+    },
+    {
+      "epoch": 0.0994525403218016,
+      "grad_norm": 0.7542218565940857,
+      "learning_rate": 0.0002982464521568076,
+      "loss": 4.7704,
+      "step": 70780
+    },
+    {
+      "epoch": 0.0994665912599652,
+      "grad_norm": 0.7415812611579895,
+      "learning_rate": 0.00029828860474919204,
+      "loss": 4.8173,
+      "step": 70790
+    },
+    {
+      "epoch": 0.09948064219812877,
+      "grad_norm": 0.7335118651390076,
+      "learning_rate": 0.00029833075734157647,
+      "loss": 4.8941,
+      "step": 70800
+    },
+    {
+      "epoch": 0.09949469313629235,
+      "grad_norm": 0.7257421612739563,
+      "learning_rate": 0.0002983729099339609,
+      "loss": 4.8061,
+      "step": 70810
+    },
+    {
+      "epoch": 0.09950874407445592,
+      "grad_norm": 0.7506051659584045,
+      "learning_rate": 0.00029841506252634533,
+      "loss": 4.7745,
+      "step": 70820
+    },
+    {
+      "epoch": 0.0995227950126195,
+      "grad_norm": 0.765600323677063,
+      "learning_rate": 0.00029845721511872976,
+      "loss": 4.7226,
+      "step": 70830
+    },
+    {
+      "epoch": 0.09953684595078308,
+      "grad_norm": 0.7521647214889526,
+      "learning_rate": 0.00029849936771111425,
+      "loss": 4.8212,
+      "step": 70840
+    },
+    {
+      "epoch": 0.09955089688894665,
+      "grad_norm": 0.7769646644592285,
+      "learning_rate": 0.0002985415203034986,
+      "loss": 4.7042,
+      "step": 70850
+    },
+    {
+      "epoch": 0.09956494782711023,
+      "grad_norm": 0.751598060131073,
+      "learning_rate": 0.00029858367289588305,
+      "loss": 4.7392,
+      "step": 70860
+    },
+    {
+      "epoch": 0.0995789987652738,
+      "grad_norm": 0.7508914470672607,
+      "learning_rate": 0.0002986258254882675,
+      "loss": 4.8002,
+      "step": 70870
+    },
+    {
+      "epoch": 0.09959304970343738,
+      "grad_norm": 0.728714644908905,
+      "learning_rate": 0.0002986679780806519,
+      "loss": 4.7734,
+      "step": 70880
+    },
+    {
+      "epoch": 0.09960710064160097,
+      "grad_norm": 0.7441499829292297,
+      "learning_rate": 0.00029871013067303635,
+      "loss": 4.8044,
+      "step": 70890
+    },
+    {
+      "epoch": 0.09962115157976455,
+      "grad_norm": 0.74562007188797,
+      "learning_rate": 0.00029875228326542083,
+      "loss": 4.8104,
+      "step": 70900
+    },
+    {
+      "epoch": 0.09963520251792812,
+      "grad_norm": 0.7518746256828308,
+      "learning_rate": 0.0002987944358578052,
+      "loss": 4.7614,
+      "step": 70910
+    },
+    {
+      "epoch": 0.0996492534560917,
+      "grad_norm": 0.7399574518203735,
+      "learning_rate": 0.00029883658845018964,
+      "loss": 4.7511,
+      "step": 70920
+    },
+    {
+      "epoch": 0.09966330439425528,
+      "grad_norm": 0.7430614233016968,
+      "learning_rate": 0.00029887874104257407,
+      "loss": 4.7823,
+      "step": 70930
+    },
+    {
+      "epoch": 0.09967735533241885,
+      "grad_norm": 0.7251525521278381,
+      "learning_rate": 0.0002989208936349585,
+      "loss": 4.8084,
+      "step": 70940
+    },
+    {
+      "epoch": 0.09969140627058243,
+      "grad_norm": 0.8306521773338318,
+      "learning_rate": 0.000298963046227343,
+      "loss": 4.7226,
+      "step": 70950
+    },
+    {
+      "epoch": 0.099705457208746,
+      "grad_norm": 0.7492359280586243,
+      "learning_rate": 0.0002990051988197274,
+      "loss": 4.7497,
+      "step": 70960
+    },
+    {
+      "epoch": 0.09971950814690958,
+      "grad_norm": 0.736562967300415,
+      "learning_rate": 0.00029904735141211185,
+      "loss": 4.6997,
+      "step": 70970
+    },
+    {
+      "epoch": 0.09973355908507316,
+      "grad_norm": 0.740244448184967,
+      "learning_rate": 0.00029908950400449623,
+      "loss": 4.7356,
+      "step": 70980
+    },
+    {
+      "epoch": 0.09974761002323673,
+      "grad_norm": 0.7707760334014893,
+      "learning_rate": 0.00029913165659688066,
+      "loss": 4.6801,
+      "step": 70990
+    },
+    {
+      "epoch": 0.09976166096140032,
+      "grad_norm": 0.7498148679733276,
+      "learning_rate": 0.0002991738091892651,
+      "loss": 4.6808,
+      "step": 71000
+    },
+    {
+      "epoch": 0.0997757118995639,
+      "grad_norm": 0.7346001863479614,
+      "learning_rate": 0.0002992159617816496,
+      "loss": 4.7796,
+      "step": 71010
+    },
+    {
+      "epoch": 0.09978976283772747,
+      "grad_norm": 0.7446746826171875,
+      "learning_rate": 0.000299258114374034,
+      "loss": 4.7906,
+      "step": 71020
+    },
+    {
+      "epoch": 0.09980381377589105,
+      "grad_norm": 0.7236793637275696,
+      "learning_rate": 0.00029930026696641844,
+      "loss": 4.7742,
+      "step": 71030
+    },
+    {
+      "epoch": 0.09981786471405463,
+      "grad_norm": 0.7329582571983337,
+      "learning_rate": 0.0002993424195588028,
+      "loss": 4.6221,
+      "step": 71040
+    },
+    {
+      "epoch": 0.0998319156522182,
+      "grad_norm": 0.7298672199249268,
+      "learning_rate": 0.00029938457215118725,
+      "loss": 4.8417,
+      "step": 71050
+    },
+    {
+      "epoch": 0.09984596659038178,
+      "grad_norm": 0.7621816396713257,
+      "learning_rate": 0.0002994267247435717,
+      "loss": 4.8465,
+      "step": 71060
+    },
+    {
+      "epoch": 0.09986001752854536,
+      "grad_norm": 0.7588155269622803,
+      "learning_rate": 0.00029946887733595616,
+      "loss": 4.7579,
+      "step": 71070
+    },
+    {
+      "epoch": 0.09987406846670893,
+      "grad_norm": 0.7053050398826599,
+      "learning_rate": 0.0002995110299283406,
+      "loss": 4.7375,
+      "step": 71080
+    },
+    {
+      "epoch": 0.09988811940487251,
+      "grad_norm": 0.7316890954971313,
+      "learning_rate": 0.000299553182520725,
+      "loss": 4.8224,
+      "step": 71090
+    },
+    {
+      "epoch": 0.0999021703430361,
+      "grad_norm": 0.7458178400993347,
+      "learning_rate": 0.0002995953351131094,
+      "loss": 4.9259,
+      "step": 71100
+    },
+    {
+      "epoch": 0.09991622128119967,
+      "grad_norm": 0.7310707569122314,
+      "learning_rate": 0.00029963748770549383,
+      "loss": 4.7937,
+      "step": 71110
+    },
+    {
+      "epoch": 0.09993027221936325,
+      "grad_norm": 0.7549211978912354,
+      "learning_rate": 0.00029967964029787826,
+      "loss": 4.7397,
+      "step": 71120
+    },
+    {
+      "epoch": 0.09994432315752683,
+      "grad_norm": 0.743523120880127,
+      "learning_rate": 0.00029972179289026275,
+      "loss": 4.7553,
+      "step": 71130
+    },
+    {
+      "epoch": 0.0999583740956904,
+      "grad_norm": 0.7451346516609192,
+      "learning_rate": 0.0002997639454826472,
+      "loss": 4.7489,
+      "step": 71140
+    },
+    {
+      "epoch": 0.09997242503385398,
+      "grad_norm": 0.7571463584899902,
+      "learning_rate": 0.0002998060980750316,
+      "loss": 4.7289,
+      "step": 71150
+    },
+    {
+      "epoch": 0.09998647597201755,
+      "grad_norm": 0.7282206416130066,
+      "learning_rate": 0.00029984825066741604,
+      "loss": 4.8128,
+      "step": 71160
+    },
+    {
+      "epoch": 0.10000052691018113,
+      "grad_norm": 0.7786749601364136,
+      "learning_rate": 0.0002998904032598004,
+      "loss": 4.928,
+      "step": 71170
+    },
+    {
+      "epoch": 0.10001457784834471,
+      "grad_norm": 0.7948459982872009,
+      "learning_rate": 0.00029993255585218485,
+      "loss": 4.7095,
+      "step": 71180
+    },
+    {
+      "epoch": 0.10002862878650828,
+      "grad_norm": 0.7715887427330017,
+      "learning_rate": 0.00029997470844456934,
+      "loss": 4.8289,
+      "step": 71190
+    },
+    {
+      "epoch": 0.10004267972467187,
+      "grad_norm": 0.7413100004196167,
+      "learning_rate": 0.0002999999999711326,
+      "loss": 4.6136,
+      "step": 71200
+    },
+    {
+      "epoch": 0.10005673066283545,
+      "grad_norm": 0.8110563158988953,
+      "learning_rate": 0.00029999999964637453,
+      "loss": 4.7762,
+      "step": 71210
+    },
+    {
+      "epoch": 0.10007078160099903,
+      "grad_norm": 0.7453218698501587,
+      "learning_rate": 0.0002999999989607741,
+      "loss": 4.7854,
+      "step": 71220
+    },
+    {
+      "epoch": 0.1000848325391626,
+      "grad_norm": 0.787434995174408,
+      "learning_rate": 0.0002999999979143313,
+      "loss": 4.7152,
+      "step": 71230
+    },
+    {
+      "epoch": 0.10009888347732618,
+      "grad_norm": 0.7521796226501465,
+      "learning_rate": 0.00029999999650704627,
+      "loss": 4.863,
+      "step": 71240
+    },
+    {
+      "epoch": 0.10011293441548975,
+      "grad_norm": 0.7553433179855347,
+      "learning_rate": 0.00029999999473891885,
+      "loss": 4.7434,
+      "step": 71250
+    },
+    {
+      "epoch": 0.10012698535365333,
+      "grad_norm": 0.7574853301048279,
+      "learning_rate": 0.00029999999260994916,
+      "loss": 4.8025,
+      "step": 71260
+    },
+    {
+      "epoch": 0.1001410362918169,
+      "grad_norm": 0.7284955978393555,
+      "learning_rate": 0.0002999999901201371,
+      "loss": 4.7283,
+      "step": 71270
+    },
+    {
+      "epoch": 0.10015508722998048,
+      "grad_norm": 0.7206137776374817,
+      "learning_rate": 0.00029999998726948276,
+      "loss": 4.6774,
+      "step": 71280
+    },
+    {
+      "epoch": 0.10016913816814406,
+      "grad_norm": 0.7361604571342468,
+      "learning_rate": 0.00029999998405798615,
+      "loss": 4.8914,
+      "step": 71290
+    },
+    {
+      "epoch": 0.10018318910630764,
+      "grad_norm": 0.7506524920463562,
+      "learning_rate": 0.0002999999804856472,
+      "loss": 4.7574,
+      "step": 71300
+    },
+    {
+      "epoch": 0.10019724004447123,
+      "grad_norm": 0.7192894220352173,
+      "learning_rate": 0.00029999997655246597,
+      "loss": 4.7978,
+      "step": 71310
+    },
+    {
+      "epoch": 0.1002112909826348,
+      "grad_norm": 0.6896945238113403,
+      "learning_rate": 0.0002999999722584425,
+      "loss": 4.7519,
+      "step": 71320
+    },
+    {
+      "epoch": 0.10022534192079838,
+      "grad_norm": 0.7413153648376465,
+      "learning_rate": 0.00029999996760357676,
+      "loss": 4.878,
+      "step": 71330
+    },
+    {
+      "epoch": 0.10023939285896195,
+      "grad_norm": 0.7378984689712524,
+      "learning_rate": 0.0002999999625878688,
+      "loss": 4.7571,
+      "step": 71340
+    },
+    {
+      "epoch": 0.10025344379712553,
+      "grad_norm": 0.7499397397041321,
+      "learning_rate": 0.0002999999572113185,
+      "loss": 4.7636,
+      "step": 71350
+    },
+    {
+      "epoch": 0.1002674947352891,
+      "grad_norm": 0.725183367729187,
+      "learning_rate": 0.0002999999514739261,
+      "loss": 4.8262,
+      "step": 71360
+    },
+    {
+      "epoch": 0.10028154567345268,
+      "grad_norm": 0.783112645149231,
+      "learning_rate": 0.0002999999453756914,
+      "loss": 4.7547,
+      "step": 71370
+    },
+    {
+      "epoch": 0.10029559661161626,
+      "grad_norm": 0.7308584451675415,
+      "learning_rate": 0.00029999993891661455,
+      "loss": 4.7689,
+      "step": 71380
+    },
+    {
+      "epoch": 0.10030964754977983,
+      "grad_norm": 0.7266840934753418,
+      "learning_rate": 0.0002999999320966955,
+      "loss": 4.7863,
+      "step": 71390
+    },
+    {
+      "epoch": 0.10032369848794341,
+      "grad_norm": 0.7103331089019775,
+      "learning_rate": 0.00029999992491593427,
+      "loss": 4.6985,
+      "step": 71400
+    },
+    {
+      "epoch": 0.100337749426107,
+      "grad_norm": 0.7648615837097168,
+      "learning_rate": 0.0002999999173743309,
+      "loss": 4.8235,
+      "step": 71410
+    },
+    {
+      "epoch": 0.10035180036427058,
+      "grad_norm": 0.7262908220291138,
+      "learning_rate": 0.0002999999094718854,
+      "loss": 4.8409,
+      "step": 71420
+    },
+    {
+      "epoch": 0.10036585130243415,
+      "grad_norm": 0.7537401914596558,
+      "learning_rate": 0.0002999999012085978,
+      "loss": 4.7755,
+      "step": 71430
+    },
+    {
+      "epoch": 0.10037990224059773,
+      "grad_norm": 0.7262635827064514,
+      "learning_rate": 0.0002999998925844681,
+      "loss": 4.7393,
+      "step": 71440
+    },
+    {
+      "epoch": 0.1003939531787613,
+      "grad_norm": 0.7246768474578857,
+      "learning_rate": 0.0002999998835994963,
+      "loss": 4.747,
+      "step": 71450
+    },
+    {
+      "epoch": 0.10040800411692488,
+      "grad_norm": 0.7420998811721802,
+      "learning_rate": 0.0002999998742536825,
+      "loss": 4.7245,
+      "step": 71460
+    },
+    {
+      "epoch": 0.10042205505508846,
+      "grad_norm": 0.7534216046333313,
+      "learning_rate": 0.0002999998645470266,
+      "loss": 4.7235,
+      "step": 71470
+    },
+    {
+      "epoch": 0.10043610599325203,
+      "grad_norm": 0.7178711891174316,
+      "learning_rate": 0.00029999985447952876,
+      "loss": 4.6908,
+      "step": 71480
+    },
+    {
+      "epoch": 0.10045015693141561,
+      "grad_norm": 0.7422319054603577,
+      "learning_rate": 0.0002999998440511889,
+      "loss": 4.8748,
+      "step": 71490
+    },
+    {
+      "epoch": 0.10046420786957919,
+      "grad_norm": 0.7381262183189392,
+      "learning_rate": 0.00029999983326200715,
+      "loss": 4.7584,
+      "step": 71500
+    },
+    {
+      "epoch": 0.10047825880774278,
+      "grad_norm": 0.726177453994751,
+      "learning_rate": 0.0002999998232432237,
+      "loss": 4.7548,
+      "step": 71510
+    },
+    {
+      "epoch": 0.10049230974590635,
+      "grad_norm": 0.732631266117096,
+      "learning_rate": 0.00029999981176844225,
+      "loss": 4.8067,
+      "step": 71520
+    },
+    {
+      "epoch": 0.10050636068406993,
+      "grad_norm": 0.7357652187347412,
+      "learning_rate": 0.00029999979993281894,
+      "loss": 4.8273,
+      "step": 71530
+    },
+    {
+      "epoch": 0.1005204116222335,
+      "grad_norm": 0.7101837396621704,
+      "learning_rate": 0.0002999997877363538,
+      "loss": 4.6746,
+      "step": 71540
+    },
+    {
+      "epoch": 0.10053446256039708,
+      "grad_norm": 0.7089740633964539,
+      "learning_rate": 0.0002999997751790468,
+      "loss": 4.8879,
+      "step": 71550
+    },
+    {
+      "epoch": 0.10054851349856066,
+      "grad_norm": 0.7296656966209412,
+      "learning_rate": 0.000299999762260898,
+      "loss": 4.7599,
+      "step": 71560
+    },
+    {
+      "epoch": 0.10056256443672423,
+      "grad_norm": 0.7465136051177979,
+      "learning_rate": 0.00029999974898190754,
+      "loss": 4.8796,
+      "step": 71570
+    },
+    {
+      "epoch": 0.10057661537488781,
+      "grad_norm": 0.7571253776550293,
+      "learning_rate": 0.0002999997353420753,
+      "loss": 4.7633,
+      "step": 71580
+    },
+    {
+      "epoch": 0.10059066631305139,
+      "grad_norm": 0.7460293173789978,
+      "learning_rate": 0.00029999972134140134,
+      "loss": 4.9302,
+      "step": 71590
+    },
+    {
+      "epoch": 0.10060471725121496,
+      "grad_norm": 0.7485179305076599,
+      "learning_rate": 0.0002999997069798857,
+      "loss": 4.8744,
+      "step": 71600
+    },
+    {
+      "epoch": 0.10061876818937854,
+      "grad_norm": 0.7496930956840515,
+      "learning_rate": 0.0002999996922575285,
+      "loss": 4.7929,
+      "step": 71610
+    },
+    {
+      "epoch": 0.10063281912754213,
+      "grad_norm": 0.7394983768463135,
+      "learning_rate": 0.0002999996771743297,
+      "loss": 4.7198,
+      "step": 71620
+    },
+    {
+      "epoch": 0.1006468700657057,
+      "grad_norm": 0.7388569712638855,
+      "learning_rate": 0.00029999966173028934,
+      "loss": 4.7331,
+      "step": 71630
+    },
+    {
+      "epoch": 0.10066092100386928,
+      "grad_norm": 0.7156569361686707,
+      "learning_rate": 0.00029999964592540747,
+      "loss": 4.8351,
+      "step": 71640
+    },
+    {
+      "epoch": 0.10067497194203286,
+      "grad_norm": 0.7694973349571228,
+      "learning_rate": 0.0002999996297596841,
+      "loss": 4.8795,
+      "step": 71650
+    },
+    {
+      "epoch": 0.10068902288019643,
+      "grad_norm": 0.8083022236824036,
+      "learning_rate": 0.0002999996132331193,
+      "loss": 4.7547,
+      "step": 71660
+    },
+    {
+      "epoch": 0.10070307381836001,
+      "grad_norm": 0.7387674450874329,
+      "learning_rate": 0.00029999959634571306,
+      "loss": 4.6957,
+      "step": 71670
+    },
+    {
+      "epoch": 0.10071712475652358,
+      "grad_norm": 0.7180886268615723,
+      "learning_rate": 0.0002999995790974655,
+      "loss": 4.9078,
+      "step": 71680
+    },
+    {
+      "epoch": 0.10073117569468716,
+      "grad_norm": 0.7135482430458069,
+      "learning_rate": 0.00029999956148837664,
+      "loss": 4.8903,
+      "step": 71690
+    },
+    {
+      "epoch": 0.10074522663285074,
+      "grad_norm": 0.7912961840629578,
+      "learning_rate": 0.00029999954351844646,
+      "loss": 4.7291,
+      "step": 71700
+    },
+    {
+      "epoch": 0.10075927757101431,
+      "grad_norm": 0.7425175905227661,
+      "learning_rate": 0.0002999995251876751,
+      "loss": 4.8633,
+      "step": 71710
+    },
+    {
+      "epoch": 0.1007733285091779,
+      "grad_norm": 0.741531491279602,
+      "learning_rate": 0.00029999950649606255,
+      "loss": 4.8959,
+      "step": 71720
+    },
+    {
+      "epoch": 0.10078737944734148,
+      "grad_norm": 0.7444692254066467,
+      "learning_rate": 0.0002999994874436088,
+      "loss": 4.7001,
+      "step": 71730
+    },
+    {
+      "epoch": 0.10080143038550506,
+      "grad_norm": 0.7535449862480164,
+      "learning_rate": 0.000299999468030314,
+      "loss": 4.7645,
+      "step": 71740
+    },
+    {
+      "epoch": 0.10081548132366863,
+      "grad_norm": 0.7291529178619385,
+      "learning_rate": 0.0002999994482561781,
+      "loss": 4.8059,
+      "step": 71750
+    },
+    {
+      "epoch": 0.10082953226183221,
+      "grad_norm": 0.7530328035354614,
+      "learning_rate": 0.0002999994281212013,
+      "loss": 4.7958,
+      "step": 71760
+    },
+    {
+      "epoch": 0.10084358319999578,
+      "grad_norm": 0.7359017133712769,
+      "learning_rate": 0.0002999994076253835,
+      "loss": 4.7982,
+      "step": 71770
+    },
+    {
+      "epoch": 0.10085763413815936,
+      "grad_norm": 0.7342458367347717,
+      "learning_rate": 0.00029999938676872475,
+      "loss": 4.7397,
+      "step": 71780
+    },
+    {
+      "epoch": 0.10087168507632294,
+      "grad_norm": 0.7454365491867065,
+      "learning_rate": 0.0002999993655512252,
+      "loss": 4.7391,
+      "step": 71790
+    },
+    {
+      "epoch": 0.10088573601448651,
+      "grad_norm": 0.771918773651123,
+      "learning_rate": 0.0002999993439728848,
+      "loss": 4.8666,
+      "step": 71800
+    },
+    {
+      "epoch": 0.10089978695265009,
+      "grad_norm": 0.7269641160964966,
+      "learning_rate": 0.00029999932203370364,
+      "loss": 4.8573,
+      "step": 71810
+    },
+    {
+      "epoch": 0.10091383789081368,
+      "grad_norm": 0.7324143648147583,
+      "learning_rate": 0.00029999929973368184,
+      "loss": 4.6735,
+      "step": 71820
+    },
+    {
+      "epoch": 0.10092788882897726,
+      "grad_norm": 0.7486067414283752,
+      "learning_rate": 0.0002999992770728194,
+      "loss": 4.6773,
+      "step": 71830
+    },
+    {
+      "epoch": 0.10094193976714083,
+      "grad_norm": 0.7198307514190674,
+      "learning_rate": 0.0002999992540511163,
+      "loss": 4.7123,
+      "step": 71840
+    },
+    {
+      "epoch": 0.10095599070530441,
+      "grad_norm": 0.7431245446205139,
+      "learning_rate": 0.00029999923066857274,
+      "loss": 4.7798,
+      "step": 71850
+    },
+    {
+      "epoch": 0.10097004164346798,
+      "grad_norm": 0.7345250844955444,
+      "learning_rate": 0.00029999920692518864,
+      "loss": 4.8772,
+      "step": 71860
+    },
+    {
+      "epoch": 0.10098409258163156,
+      "grad_norm": 0.7660800814628601,
+      "learning_rate": 0.0002999991828209641,
+      "loss": 4.8468,
+      "step": 71870
+    },
+    {
+      "epoch": 0.10099814351979514,
+      "grad_norm": 0.7611793279647827,
+      "learning_rate": 0.0002999991583558993,
+      "loss": 4.8363,
+      "step": 71880
+    },
+    {
+      "epoch": 0.10101219445795871,
+      "grad_norm": 0.7366256713867188,
+      "learning_rate": 0.0002999991335299941,
+      "loss": 4.7184,
+      "step": 71890
+    },
+    {
+      "epoch": 0.10102624539612229,
+      "grad_norm": 0.7112072706222534,
+      "learning_rate": 0.0002999991083432487,
+      "loss": 4.724,
+      "step": 71900
+    },
+    {
+      "epoch": 0.10104029633428586,
+      "grad_norm": 0.7414090633392334,
+      "learning_rate": 0.0002999990827956631,
+      "loss": 4.8933,
+      "step": 71910
+    },
+    {
+      "epoch": 0.10105434727244944,
+      "grad_norm": 0.7189602851867676,
+      "learning_rate": 0.0002999990568872374,
+      "loss": 4.7637,
+      "step": 71920
+    },
+    {
+      "epoch": 0.10106839821061303,
+      "grad_norm": 0.7293269038200378,
+      "learning_rate": 0.0002999990306179716,
+      "loss": 4.8199,
+      "step": 71930
+    },
+    {
+      "epoch": 0.1010824491487766,
+      "grad_norm": 0.7632038593292236,
+      "learning_rate": 0.0002999990039878658,
+      "loss": 4.6565,
+      "step": 71940
+    },
+    {
+      "epoch": 0.10109650008694018,
+      "grad_norm": 0.7255117297172546,
+      "learning_rate": 0.0002999989769969201,
+      "loss": 4.6914,
+      "step": 71950
+    },
+    {
+      "epoch": 0.10111055102510376,
+      "grad_norm": 0.7196149230003357,
+      "learning_rate": 0.0002999989496451345,
+      "loss": 4.7191,
+      "step": 71960
+    },
+    {
+      "epoch": 0.10112460196326734,
+      "grad_norm": 0.7434830665588379,
+      "learning_rate": 0.00029999892193250917,
+      "loss": 4.7101,
+      "step": 71970
+    },
+    {
+      "epoch": 0.10113865290143091,
+      "grad_norm": 0.7188177108764648,
+      "learning_rate": 0.000299998893859044,
+      "loss": 4.8254,
+      "step": 71980
+    },
+    {
+      "epoch": 0.10115270383959449,
+      "grad_norm": 0.7340826988220215,
+      "learning_rate": 0.00029999886542473925,
+      "loss": 4.7947,
+      "step": 71990
+    },
+    {
+      "epoch": 0.10116675477775806,
+      "grad_norm": 0.7305575609207153,
+      "learning_rate": 0.0002999988366295948,
+      "loss": 4.5447,
+      "step": 72000
+    },
+    {
+      "epoch": 0.10118080571592164,
+      "grad_norm": 0.7468332052230835,
+      "learning_rate": 0.0002999988074736109,
+      "loss": 4.7055,
+      "step": 72010
+    },
+    {
+      "epoch": 0.10119485665408522,
+      "grad_norm": 0.7357817888259888,
+      "learning_rate": 0.00029999877795678753,
+      "loss": 4.8088,
+      "step": 72020
+    },
+    {
+      "epoch": 0.1012089075922488,
+      "grad_norm": 0.7364190816879272,
+      "learning_rate": 0.0002999987480791248,
+      "loss": 4.6873,
+      "step": 72030
+    },
+    {
+      "epoch": 0.10122295853041238,
+      "grad_norm": 0.7412238717079163,
+      "learning_rate": 0.00029999871784062266,
+      "loss": 4.7859,
+      "step": 72040
+    },
+    {
+      "epoch": 0.10123700946857596,
+      "grad_norm": 0.7454258799552917,
+      "learning_rate": 0.00029999868724128135,
+      "loss": 4.7342,
+      "step": 72050
+    },
+    {
+      "epoch": 0.10125106040673953,
+      "grad_norm": 0.7320020794868469,
+      "learning_rate": 0.0002999986562811008,
+      "loss": 4.7076,
+      "step": 72060
+    },
+    {
+      "epoch": 0.10126511134490311,
+      "grad_norm": 0.7472354769706726,
+      "learning_rate": 0.00029999862496008124,
+      "loss": 4.7221,
+      "step": 72070
+    },
+    {
+      "epoch": 0.10127916228306669,
+      "grad_norm": 0.7363531589508057,
+      "learning_rate": 0.0002999985932782226,
+      "loss": 4.681,
+      "step": 72080
+    },
+    {
+      "epoch": 0.10129321322123026,
+      "grad_norm": 0.7143122553825378,
+      "learning_rate": 0.000299998561235525,
+      "loss": 4.7911,
+      "step": 72090
+    },
+    {
+      "epoch": 0.10130726415939384,
+      "grad_norm": 0.7466495037078857,
+      "learning_rate": 0.0002999985288319886,
+      "loss": 4.9298,
+      "step": 72100
+    },
+    {
+      "epoch": 0.10132131509755742,
+      "grad_norm": 0.7105987668037415,
+      "learning_rate": 0.00029999849606761336,
+      "loss": 4.8656,
+      "step": 72110
+    },
+    {
+      "epoch": 0.10133536603572099,
+      "grad_norm": 0.7323604822158813,
+      "learning_rate": 0.0002999984629423994,
+      "loss": 4.8375,
+      "step": 72120
+    },
+    {
+      "epoch": 0.10134941697388458,
+      "grad_norm": 0.7545437812805176,
+      "learning_rate": 0.00029999842945634684,
+      "loss": 4.7479,
+      "step": 72130
+    },
+    {
+      "epoch": 0.10136346791204816,
+      "grad_norm": 0.732039213180542,
+      "learning_rate": 0.0002999983956094557,
+      "loss": 4.8971,
+      "step": 72140
+    },
+    {
+      "epoch": 0.10137751885021173,
+      "grad_norm": 0.7388688325881958,
+      "learning_rate": 0.0002999983614017261,
+      "loss": 4.7152,
+      "step": 72150
+    },
+    {
+      "epoch": 0.10139156978837531,
+      "grad_norm": 0.7269540429115295,
+      "learning_rate": 0.0002999983268331582,
+      "loss": 4.8155,
+      "step": 72160
+    },
+    {
+      "epoch": 0.10140562072653889,
+      "grad_norm": 0.7393106818199158,
+      "learning_rate": 0.00029999829190375187,
+      "loss": 4.7825,
+      "step": 72170
+    },
+    {
+      "epoch": 0.10141967166470246,
+      "grad_norm": 0.7458699941635132,
+      "learning_rate": 0.00029999825661350735,
+      "loss": 4.696,
+      "step": 72180
+    },
+    {
+      "epoch": 0.10143372260286604,
+      "grad_norm": 0.7346116900444031,
+      "learning_rate": 0.00029999822096242473,
+      "loss": 4.7549,
+      "step": 72190
+    },
+    {
+      "epoch": 0.10144777354102961,
+      "grad_norm": 0.7251469492912292,
+      "learning_rate": 0.000299998184950504,
+      "loss": 4.7406,
+      "step": 72200
+    },
+    {
+      "epoch": 0.10146182447919319,
+      "grad_norm": 0.7549038529396057,
+      "learning_rate": 0.0002999981485777453,
+      "loss": 4.7226,
+      "step": 72210
+    },
+    {
+      "epoch": 0.10147587541735677,
+      "grad_norm": 0.7311290502548218,
+      "learning_rate": 0.0002999981118441488,
+      "loss": 4.7761,
+      "step": 72220
+    },
+    {
+      "epoch": 0.10148992635552034,
+      "grad_norm": 0.7395356297492981,
+      "learning_rate": 0.0002999980747497145,
+      "loss": 4.8128,
+      "step": 72230
+    },
+    {
+      "epoch": 0.10150397729368393,
+      "grad_norm": 0.7352813482284546,
+      "learning_rate": 0.0002999980372944425,
+      "loss": 4.7336,
+      "step": 72240
+    },
+    {
+      "epoch": 0.10151802823184751,
+      "grad_norm": 0.7374472618103027,
+      "learning_rate": 0.00029999799947833287,
+      "loss": 4.6786,
+      "step": 72250
+    },
+    {
+      "epoch": 0.10153207917001109,
+      "grad_norm": 0.7531014680862427,
+      "learning_rate": 0.0002999979613013857,
+      "loss": 4.7935,
+      "step": 72260
+    },
+    {
+      "epoch": 0.10154613010817466,
+      "grad_norm": 0.7118738889694214,
+      "learning_rate": 0.00029999792276360113,
+      "loss": 4.7567,
+      "step": 72270
+    },
+    {
+      "epoch": 0.10156018104633824,
+      "grad_norm": 0.7456701397895813,
+      "learning_rate": 0.00029999788386497923,
+      "loss": 4.7232,
+      "step": 72280
+    },
+    {
+      "epoch": 0.10157423198450181,
+      "grad_norm": 0.7232758402824402,
+      "learning_rate": 0.00029999784460552005,
+      "loss": 4.7032,
+      "step": 72290
+    },
+    {
+      "epoch": 0.10158828292266539,
+      "grad_norm": 0.7644693851470947,
+      "learning_rate": 0.0002999978049852238,
+      "loss": 4.8459,
+      "step": 72300
+    },
+    {
+      "epoch": 0.10160233386082897,
+      "grad_norm": 0.7518143057823181,
+      "learning_rate": 0.00029999776500409044,
+      "loss": 4.7958,
+      "step": 72310
+    },
+    {
+      "epoch": 0.10161638479899254,
+      "grad_norm": 0.7484323382377625,
+      "learning_rate": 0.00029999772466212017,
+      "loss": 4.7845,
+      "step": 72320
+    },
+    {
+      "epoch": 0.10163043573715612,
+      "grad_norm": 0.7363878488540649,
+      "learning_rate": 0.000299997683959313,
+      "loss": 4.8805,
+      "step": 72330
+    },
+    {
+      "epoch": 0.10164448667531971,
+      "grad_norm": 0.771102786064148,
+      "learning_rate": 0.00029999764289566914,
+      "loss": 4.766,
+      "step": 72340
+    },
+    {
+      "epoch": 0.10165853761348329,
+      "grad_norm": 0.7849813103675842,
+      "learning_rate": 0.00029999760147118854,
+      "loss": 4.7616,
+      "step": 72350
+    },
+    {
+      "epoch": 0.10167258855164686,
+      "grad_norm": 0.7605536580085754,
+      "learning_rate": 0.0002999975596858714,
+      "loss": 4.8137,
+      "step": 72360
+    },
+    {
+      "epoch": 0.10168663948981044,
+      "grad_norm": 0.7546150088310242,
+      "learning_rate": 0.0002999975175397178,
+      "loss": 4.7202,
+      "step": 72370
+    },
+    {
+      "epoch": 0.10170069042797401,
+      "grad_norm": 0.7308611869812012,
+      "learning_rate": 0.0002999974750327279,
+      "loss": 4.7466,
+      "step": 72380
+    },
+    {
+      "epoch": 0.10171474136613759,
+      "grad_norm": 0.7356964349746704,
+      "learning_rate": 0.0002999974321649017,
+      "loss": 4.8318,
+      "step": 72390
+    },
+    {
+      "epoch": 0.10172879230430117,
+      "grad_norm": 0.73519366979599,
+      "learning_rate": 0.00029999738893623937,
+      "loss": 4.7815,
+      "step": 72400
+    },
+    {
+      "epoch": 0.10174284324246474,
+      "grad_norm": 0.6988383531570435,
+      "learning_rate": 0.000299997345346741,
+      "loss": 4.7789,
+      "step": 72410
+    },
+    {
+      "epoch": 0.10175689418062832,
+      "grad_norm": 0.745654284954071,
+      "learning_rate": 0.0002999973013964067,
+      "loss": 4.8238,
+      "step": 72420
+    },
+    {
+      "epoch": 0.1017709451187919,
+      "grad_norm": 0.7207648158073425,
+      "learning_rate": 0.00029999725708523643,
+      "loss": 4.7722,
+      "step": 72430
+    },
+    {
+      "epoch": 0.10178499605695548,
+      "grad_norm": 0.8427481651306152,
+      "learning_rate": 0.00029999721241323054,
+      "loss": 4.7605,
+      "step": 72440
+    },
+    {
+      "epoch": 0.10179904699511906,
+      "grad_norm": 0.7269243597984314,
+      "learning_rate": 0.00029999716738038904,
+      "loss": 4.8267,
+      "step": 72450
+    },
+    {
+      "epoch": 0.10181309793328264,
+      "grad_norm": 0.7501411437988281,
+      "learning_rate": 0.00029999712198671195,
+      "loss": 4.7538,
+      "step": 72460
+    },
+    {
+      "epoch": 0.10182714887144621,
+      "grad_norm": 0.7097275853157043,
+      "learning_rate": 0.0002999970762321995,
+      "loss": 4.7286,
+      "step": 72470
+    },
+    {
+      "epoch": 0.10184119980960979,
+      "grad_norm": 0.7278454303741455,
+      "learning_rate": 0.00029999703011685176,
+      "loss": 4.8126,
+      "step": 72480
+    },
+    {
+      "epoch": 0.10185525074777337,
+      "grad_norm": 0.7060046195983887,
+      "learning_rate": 0.00029999698364066883,
+      "loss": 4.715,
+      "step": 72490
+    },
+    {
+      "epoch": 0.10186930168593694,
+      "grad_norm": 0.7554420828819275,
+      "learning_rate": 0.0002999969368036508,
+      "loss": 4.9029,
+      "step": 72500
+    },
+    {
+      "epoch": 0.10188335262410052,
+      "grad_norm": 0.733871340751648,
+      "learning_rate": 0.00029999688960579785,
+      "loss": 4.808,
+      "step": 72510
+    },
+    {
+      "epoch": 0.1018974035622641,
+      "grad_norm": 0.7300422191619873,
+      "learning_rate": 0.00029999684204711006,
+      "loss": 4.8701,
+      "step": 72520
+    },
+    {
+      "epoch": 0.10191145450042767,
+      "grad_norm": 0.7577276825904846,
+      "learning_rate": 0.0002999967941275875,
+      "loss": 4.6814,
+      "step": 72530
+    },
+    {
+      "epoch": 0.10192550543859125,
+      "grad_norm": 0.7170461416244507,
+      "learning_rate": 0.00029999674584723035,
+      "loss": 4.7992,
+      "step": 72540
+    },
+    {
+      "epoch": 0.10193955637675484,
+      "grad_norm": 0.7378881573677063,
+      "learning_rate": 0.00029999669720603864,
+      "loss": 4.7144,
+      "step": 72550
+    },
+    {
+      "epoch": 0.10195360731491841,
+      "grad_norm": 0.7653374671936035,
+      "learning_rate": 0.0002999966482040126,
+      "loss": 4.8943,
+      "step": 72560
+    },
+    {
+      "epoch": 0.10196765825308199,
+      "grad_norm": 0.7405737042427063,
+      "learning_rate": 0.0002999965988411523,
+      "loss": 4.7858,
+      "step": 72570
+    },
+    {
+      "epoch": 0.10198170919124556,
+      "grad_norm": 0.712135910987854,
+      "learning_rate": 0.0002999965491174578,
+      "loss": 4.9162,
+      "step": 72580
+    },
+    {
+      "epoch": 0.10199576012940914,
+      "grad_norm": 0.742899477481842,
+      "learning_rate": 0.0002999964990329293,
+      "loss": 4.7942,
+      "step": 72590
+    },
+    {
+      "epoch": 0.10200981106757272,
+      "grad_norm": 0.7559585571289062,
+      "learning_rate": 0.0002999964485875669,
+      "loss": 4.6854,
+      "step": 72600
+    },
+    {
+      "epoch": 0.1020238620057363,
+      "grad_norm": 0.7186324000358582,
+      "learning_rate": 0.00029999639778137074,
+      "loss": 4.7303,
+      "step": 72610
+    },
+    {
+      "epoch": 0.10203791294389987,
+      "grad_norm": 0.7546454071998596,
+      "learning_rate": 0.0002999963466143408,
+      "loss": 4.6544,
+      "step": 72620
+    },
+    {
+      "epoch": 0.10205196388206345,
+      "grad_norm": 0.6856681108474731,
+      "learning_rate": 0.00029999629508647746,
+      "loss": 4.7369,
+      "step": 72630
+    },
+    {
+      "epoch": 0.10206601482022702,
+      "grad_norm": 0.7506888508796692,
+      "learning_rate": 0.00029999624319778067,
+      "loss": 4.7411,
+      "step": 72640
+    },
+    {
+      "epoch": 0.10208006575839061,
+      "grad_norm": 0.7311187982559204,
+      "learning_rate": 0.00029999619094825056,
+      "loss": 4.6478,
+      "step": 72650
+    },
+    {
+      "epoch": 0.10209411669655419,
+      "grad_norm": 0.7421939969062805,
+      "learning_rate": 0.0002999961383378873,
+      "loss": 4.7328,
+      "step": 72660
+    },
+    {
+      "epoch": 0.10210816763471776,
+      "grad_norm": 0.7084130644798279,
+      "learning_rate": 0.00029999608536669095,
+      "loss": 4.6734,
+      "step": 72670
+    },
+    {
+      "epoch": 0.10212221857288134,
+      "grad_norm": 0.7495949864387512,
+      "learning_rate": 0.0002999960320346617,
+      "loss": 4.7531,
+      "step": 72680
+    },
+    {
+      "epoch": 0.10213626951104492,
+      "grad_norm": 0.7394850254058838,
+      "learning_rate": 0.0002999959783417997,
+      "loss": 4.7793,
+      "step": 72690
+    },
+    {
+      "epoch": 0.10215032044920849,
+      "grad_norm": 0.7716978788375854,
+      "learning_rate": 0.00029999592428810504,
+      "loss": 4.7758,
+      "step": 72700
+    },
+    {
+      "epoch": 0.10216437138737207,
+      "grad_norm": 0.7250087857246399,
+      "learning_rate": 0.0002999958698735779,
+      "loss": 4.8447,
+      "step": 72710
+    },
+    {
+      "epoch": 0.10217842232553564,
+      "grad_norm": 0.7172320485115051,
+      "learning_rate": 0.0002999958150982183,
+      "loss": 4.9669,
+      "step": 72720
+    },
+    {
+      "epoch": 0.10219247326369922,
+      "grad_norm": 0.7347210049629211,
+      "learning_rate": 0.00029999575996202647,
+      "loss": 4.7817,
+      "step": 72730
+    },
+    {
+      "epoch": 0.1022065242018628,
+      "grad_norm": 0.7101386785507202,
+      "learning_rate": 0.00029999570446500245,
+      "loss": 4.7991,
+      "step": 72740
+    },
+    {
+      "epoch": 0.10222057514002639,
+      "grad_norm": 0.7327938079833984,
+      "learning_rate": 0.0002999956486071465,
+      "loss": 4.707,
+      "step": 72750
+    },
+    {
+      "epoch": 0.10223462607818996,
+      "grad_norm": 0.7212610244750977,
+      "learning_rate": 0.0002999955923884587,
+      "loss": 4.7766,
+      "step": 72760
+    },
+    {
+      "epoch": 0.10224867701635354,
+      "grad_norm": 0.720736563205719,
+      "learning_rate": 0.0002999955358089391,
+      "loss": 4.7198,
+      "step": 72770
+    },
+    {
+      "epoch": 0.10226272795451712,
+      "grad_norm": 0.7179093360900879,
+      "learning_rate": 0.000299995478868588,
+      "loss": 4.7682,
+      "step": 72780
+    },
+    {
+      "epoch": 0.10227677889268069,
+      "grad_norm": 0.7311738133430481,
+      "learning_rate": 0.0002999954215674054,
+      "loss": 4.8445,
+      "step": 72790
+    },
+    {
+      "epoch": 0.10229082983084427,
+      "grad_norm": 0.7069105505943298,
+      "learning_rate": 0.00029999536390539145,
+      "loss": 4.7873,
+      "step": 72800
+    },
+    {
+      "epoch": 0.10230488076900784,
+      "grad_norm": 0.7458084225654602,
+      "learning_rate": 0.00029999530588254635,
+      "loss": 4.6617,
+      "step": 72810
+    },
+    {
+      "epoch": 0.10231893170717142,
+      "grad_norm": 0.7396774291992188,
+      "learning_rate": 0.00029999524749887025,
+      "loss": 4.8425,
+      "step": 72820
+    },
+    {
+      "epoch": 0.102332982645335,
+      "grad_norm": 0.7057560682296753,
+      "learning_rate": 0.0002999951887543632,
+      "loss": 4.8336,
+      "step": 72830
+    },
+    {
+      "epoch": 0.10234703358349857,
+      "grad_norm": 0.716718852519989,
+      "learning_rate": 0.0002999951296490254,
+      "loss": 4.6897,
+      "step": 72840
+    },
+    {
+      "epoch": 0.10236108452166215,
+      "grad_norm": 0.7290700674057007,
+      "learning_rate": 0.000299995070182857,
+      "loss": 4.8415,
+      "step": 72850
+    },
+    {
+      "epoch": 0.10237513545982574,
+      "grad_norm": 0.7181606292724609,
+      "learning_rate": 0.00029999501035585816,
+      "loss": 4.8283,
+      "step": 72860
+    },
+    {
+      "epoch": 0.10238918639798932,
+      "grad_norm": 0.7337809801101685,
+      "learning_rate": 0.00029999495016802893,
+      "loss": 4.8772,
+      "step": 72870
+    },
+    {
+      "epoch": 0.10240323733615289,
+      "grad_norm": 0.7575526833534241,
+      "learning_rate": 0.0002999948896193696,
+      "loss": 4.787,
+      "step": 72880
+    },
+    {
+      "epoch": 0.10241728827431647,
+      "grad_norm": Infinity,
+      "learning_rate": 0.00029999482870988017,
+      "loss": 4.7359,
+      "step": 72890
+    },
+    {
+      "epoch": 0.10243133921248004,
+      "grad_norm": 0.7764653563499451,
+      "learning_rate": 0.0002999947735828301,
+      "loss": 4.7426,
+      "step": 72900
+    },
+    {
+      "epoch": 0.10244539015064362,
+      "grad_norm": 0.7185347676277161,
+      "learning_rate": 0.00029999471198776404,
+      "loss": 4.867,
+      "step": 72910
+    },
+    {
+      "epoch": 0.1024594410888072,
+      "grad_norm": 0.755716860294342,
+      "learning_rate": 0.0002999946500318684,
+      "loss": 4.7958,
+      "step": 72920
+    },
+    {
+      "epoch": 0.10247349202697077,
+      "grad_norm": 0.7544811367988586,
+      "learning_rate": 0.0002999945877151432,
+      "loss": 4.9392,
+      "step": 72930
+    },
+    {
+      "epoch": 0.10248754296513435,
+      "grad_norm": 0.7307287454605103,
+      "learning_rate": 0.00029999452503758875,
+      "loss": 4.8479,
+      "step": 72940
+    },
+    {
+      "epoch": 0.10250159390329792,
+      "grad_norm": 0.7607468366622925,
+      "learning_rate": 0.00029999446199920516,
+      "loss": 4.6962,
+      "step": 72950
+    },
+    {
+      "epoch": 0.10251564484146151,
+      "grad_norm": 0.721833348274231,
+      "learning_rate": 0.00029999439859999257,
+      "loss": 4.7497,
+      "step": 72960
+    },
+    {
+      "epoch": 0.10252969577962509,
+      "grad_norm": 0.7424848079681396,
+      "learning_rate": 0.00029999433483995106,
+      "loss": 4.7419,
+      "step": 72970
+    },
+    {
+      "epoch": 0.10254374671778867,
+      "grad_norm": 0.721015453338623,
+      "learning_rate": 0.0002999942707190809,
+      "loss": 4.5932,
+      "step": 72980
+    },
+    {
+      "epoch": 0.10255779765595224,
+      "grad_norm": 0.7474051713943481,
+      "learning_rate": 0.00029999420623738214,
+      "loss": 4.6855,
+      "step": 72990
+    },
+    {
+      "epoch": 0.10257184859411582,
+      "grad_norm": 0.7426432967185974,
+      "learning_rate": 0.000299994141394855,
+      "loss": 4.7178,
+      "step": 73000
+    },
+    {
+      "epoch": 0.1025858995322794,
+      "grad_norm": 0.7148788571357727,
+      "learning_rate": 0.00029999407619149966,
+      "loss": 4.8307,
+      "step": 73010
+    },
+    {
+      "epoch": 0.10259995047044297,
+      "grad_norm": 0.7621682286262512,
+      "learning_rate": 0.00029999401062731623,
+      "loss": 4.7439,
+      "step": 73020
+    },
+    {
+      "epoch": 0.10261400140860655,
+      "grad_norm": 0.7121155858039856,
+      "learning_rate": 0.00029999394470230485,
+      "loss": 4.6666,
+      "step": 73030
+    },
+    {
+      "epoch": 0.10262805234677012,
+      "grad_norm": 0.7072491645812988,
+      "learning_rate": 0.0002999938784164658,
+      "loss": 4.6901,
+      "step": 73040
+    },
+    {
+      "epoch": 0.1026421032849337,
+      "grad_norm": 0.7037854790687561,
+      "learning_rate": 0.00029999381176979903,
+      "loss": 4.7861,
+      "step": 73050
+    },
+    {
+      "epoch": 0.10265615422309729,
+      "grad_norm": 0.7043249011039734,
+      "learning_rate": 0.0002999937447623049,
+      "loss": 4.8758,
+      "step": 73060
+    },
+    {
+      "epoch": 0.10267020516126087,
+      "grad_norm": 0.7356353998184204,
+      "learning_rate": 0.00029999367739398343,
+      "loss": 4.7525,
+      "step": 73070
+    },
+    {
+      "epoch": 0.10268425609942444,
+      "grad_norm": 0.7560482025146484,
+      "learning_rate": 0.00029999360966483486,
+      "loss": 4.642,
+      "step": 73080
+    },
+    {
+      "epoch": 0.10269830703758802,
+      "grad_norm": 0.7081610560417175,
+      "learning_rate": 0.00029999354157485937,
+      "loss": 4.6725,
+      "step": 73090
+    },
+    {
+      "epoch": 0.1027123579757516,
+      "grad_norm": 0.7189093828201294,
+      "learning_rate": 0.00029999347312405705,
+      "loss": 4.852,
+      "step": 73100
+    },
+    {
+      "epoch": 0.10272640891391517,
+      "grad_norm": 0.7496188879013062,
+      "learning_rate": 0.0002999934043124281,
+      "loss": 4.7643,
+      "step": 73110
+    },
+    {
+      "epoch": 0.10274045985207875,
+      "grad_norm": 0.8219110369682312,
+      "learning_rate": 0.0002999933351399727,
+      "loss": 4.8383,
+      "step": 73120
+    },
+    {
+      "epoch": 0.10275451079024232,
+      "grad_norm": 0.714171826839447,
+      "learning_rate": 0.00029999326560669096,
+      "loss": 4.6523,
+      "step": 73130
+    },
+    {
+      "epoch": 0.1027685617284059,
+      "grad_norm": 0.7769915461540222,
+      "learning_rate": 0.00029999319571258316,
+      "loss": 4.7178,
+      "step": 73140
+    },
+    {
+      "epoch": 0.10278261266656948,
+      "grad_norm": 0.7325097918510437,
+      "learning_rate": 0.00029999312545764934,
+      "loss": 4.7205,
+      "step": 73150
+    },
+    {
+      "epoch": 0.10279666360473305,
+      "grad_norm": 0.7231411933898926,
+      "learning_rate": 0.00029999305484188974,
+      "loss": 4.8554,
+      "step": 73160
+    },
+    {
+      "epoch": 0.10281071454289664,
+      "grad_norm": 0.7304064035415649,
+      "learning_rate": 0.00029999298386530456,
+      "loss": 4.7141,
+      "step": 73170
+    },
+    {
+      "epoch": 0.10282476548106022,
+      "grad_norm": 0.7381361722946167,
+      "learning_rate": 0.0002999929125278939,
+      "loss": 4.8076,
+      "step": 73180
+    },
+    {
+      "epoch": 0.1028388164192238,
+      "grad_norm": 0.7440902590751648,
+      "learning_rate": 0.00029999284082965797,
+      "loss": 4.7352,
+      "step": 73190
+    },
+    {
+      "epoch": 0.10285286735738737,
+      "grad_norm": 0.7400135397911072,
+      "learning_rate": 0.0002999927687705969,
+      "loss": 4.8078,
+      "step": 73200
+    },
+    {
+      "epoch": 0.10286691829555095,
+      "grad_norm": 0.6989032626152039,
+      "learning_rate": 0.00029999269635071094,
+      "loss": 4.8402,
+      "step": 73210
+    },
+    {
+      "epoch": 0.10288096923371452,
+      "grad_norm": 0.7391857504844666,
+      "learning_rate": 0.00029999262357000017,
+      "loss": 4.8794,
+      "step": 73220
+    },
+    {
+      "epoch": 0.1028950201718781,
+      "grad_norm": 0.7222636342048645,
+      "learning_rate": 0.00029999255042846486,
+      "loss": 4.6988,
+      "step": 73230
+    },
+    {
+      "epoch": 0.10290907111004167,
+      "grad_norm": 0.7194198369979858,
+      "learning_rate": 0.0002999924769261051,
+      "loss": 4.649,
+      "step": 73240
+    },
+    {
+      "epoch": 0.10292312204820525,
+      "grad_norm": 0.7942532300949097,
+      "learning_rate": 0.00029999240306292116,
+      "loss": 4.7988,
+      "step": 73250
+    },
+    {
+      "epoch": 0.10293717298636883,
+      "grad_norm": 0.7117690443992615,
+      "learning_rate": 0.00029999232883891314,
+      "loss": 4.6962,
+      "step": 73260
+    },
+    {
+      "epoch": 0.10295122392453242,
+      "grad_norm": 0.7358763813972473,
+      "learning_rate": 0.00029999225425408123,
+      "loss": 4.7445,
+      "step": 73270
+    },
+    {
+      "epoch": 0.102965274862696,
+      "grad_norm": 0.7358459234237671,
+      "learning_rate": 0.00029999217930842565,
+      "loss": 4.7767,
+      "step": 73280
+    },
+    {
+      "epoch": 0.10297932580085957,
+      "grad_norm": 0.7460939884185791,
+      "learning_rate": 0.0002999921040019465,
+      "loss": 4.7906,
+      "step": 73290
+    },
+    {
+      "epoch": 0.10299337673902315,
+      "grad_norm": 0.7220245599746704,
+      "learning_rate": 0.0002999920283346441,
+      "loss": 4.818,
+      "step": 73300
+    },
+    {
+      "epoch": 0.10300742767718672,
+      "grad_norm": 0.7411149740219116,
+      "learning_rate": 0.00029999195230651847,
+      "loss": 4.7419,
+      "step": 73310
+    },
+    {
+      "epoch": 0.1030214786153503,
+      "grad_norm": 0.7259693145751953,
+      "learning_rate": 0.0002999918759175699,
+      "loss": 4.7345,
+      "step": 73320
+    },
+    {
+      "epoch": 0.10303552955351387,
+      "grad_norm": 0.7284397482872009,
+      "learning_rate": 0.00029999179916779856,
+      "loss": 4.873,
+      "step": 73330
+    },
+    {
+      "epoch": 0.10304958049167745,
+      "grad_norm": 0.7188897132873535,
+      "learning_rate": 0.0002999917220572046,
+      "loss": 4.8309,
+      "step": 73340
+    },
+    {
+      "epoch": 0.10306363142984103,
+      "grad_norm": 0.7184074521064758,
+      "learning_rate": 0.0002999916445857882,
+      "loss": 4.7298,
+      "step": 73350
+    },
+    {
+      "epoch": 0.1030776823680046,
+      "grad_norm": 0.7120581269264221,
+      "learning_rate": 0.00029999156675354963,
+      "loss": 4.8255,
+      "step": 73360
+    },
+    {
+      "epoch": 0.10309173330616819,
+      "grad_norm": 0.733837902545929,
+      "learning_rate": 0.000299991488560489,
+      "loss": 4.7718,
+      "step": 73370
+    },
+    {
+      "epoch": 0.10310578424433177,
+      "grad_norm": 0.7166063785552979,
+      "learning_rate": 0.00029999141000660646,
+      "loss": 4.801,
+      "step": 73380
+    },
+    {
+      "epoch": 0.10311983518249535,
+      "grad_norm": 0.7153290510177612,
+      "learning_rate": 0.0002999913310919023,
+      "loss": 4.7146,
+      "step": 73390
+    },
+    {
+      "epoch": 0.10313388612065892,
+      "grad_norm": 0.7479586005210876,
+      "learning_rate": 0.00029999125181637666,
+      "loss": 4.8948,
+      "step": 73400
+    },
+    {
+      "epoch": 0.1031479370588225,
+      "grad_norm": 0.7473878860473633,
+      "learning_rate": 0.0002999911721800297,
+      "loss": 4.7084,
+      "step": 73410
+    },
+    {
+      "epoch": 0.10316198799698607,
+      "grad_norm": 0.7283495664596558,
+      "learning_rate": 0.0002999910921828617,
+      "loss": 4.7356,
+      "step": 73420
+    },
+    {
+      "epoch": 0.10317603893514965,
+      "grad_norm": 0.7387956976890564,
+      "learning_rate": 0.0002999910118248728,
+      "loss": 4.8818,
+      "step": 73430
+    },
+    {
+      "epoch": 0.10319008987331323,
+      "grad_norm": 0.7675484418869019,
+      "learning_rate": 0.0002999909311060632,
+      "loss": 4.8588,
+      "step": 73440
+    },
+    {
+      "epoch": 0.1032041408114768,
+      "grad_norm": 0.7122921943664551,
+      "learning_rate": 0.00029999085002643306,
+      "loss": 4.7725,
+      "step": 73450
+    },
+    {
+      "epoch": 0.10321819174964038,
+      "grad_norm": 0.7225001454353333,
+      "learning_rate": 0.0002999907685859826,
+      "loss": 4.6845,
+      "step": 73460
+    },
+    {
+      "epoch": 0.10323224268780395,
+      "grad_norm": 0.7381240725517273,
+      "learning_rate": 0.00029999068678471203,
+      "loss": 4.7242,
+      "step": 73470
+    },
+    {
+      "epoch": 0.10324629362596754,
+      "grad_norm": 0.7431597709655762,
+      "learning_rate": 0.00029999060462262153,
+      "loss": 4.8332,
+      "step": 73480
+    },
+    {
+      "epoch": 0.10326034456413112,
+      "grad_norm": 0.7094483971595764,
+      "learning_rate": 0.00029999052209971126,
+      "loss": 4.812,
+      "step": 73490
+    },
+    {
+      "epoch": 0.1032743955022947,
+      "grad_norm": 0.7248532176017761,
+      "learning_rate": 0.0002999904392159815,
+      "loss": 4.734,
+      "step": 73500
+    },
+    {
+      "epoch": 0.10328844644045827,
+      "grad_norm": 0.7661448121070862,
+      "learning_rate": 0.00029999035597143243,
+      "loss": 4.8685,
+      "step": 73510
+    },
+    {
+      "epoch": 0.10330249737862185,
+      "grad_norm": 0.7161885499954224,
+      "learning_rate": 0.0002999902723660642,
+      "loss": 4.7242,
+      "step": 73520
+    },
+    {
+      "epoch": 0.10331654831678543,
+      "grad_norm": 0.8056687116622925,
+      "learning_rate": 0.00029999018839987706,
+      "loss": 4.7779,
+      "step": 73530
+    },
+    {
+      "epoch": 0.103330599254949,
+      "grad_norm": 0.7297375798225403,
+      "learning_rate": 0.0002999901040728712,
+      "loss": 4.699,
+      "step": 73540
+    },
+    {
+      "epoch": 0.10334465019311258,
+      "grad_norm": 0.7313802242279053,
+      "learning_rate": 0.00029999001938504677,
+      "loss": 4.7828,
+      "step": 73550
+    },
+    {
+      "epoch": 0.10335870113127615,
+      "grad_norm": 0.7233887314796448,
+      "learning_rate": 0.000299989934336404,
+      "loss": 4.7871,
+      "step": 73560
+    },
+    {
+      "epoch": 0.10337275206943973,
+      "grad_norm": 0.7726687788963318,
+      "learning_rate": 0.0002999898489269432,
+      "loss": 4.7328,
+      "step": 73570
+    },
+    {
+      "epoch": 0.10338680300760332,
+      "grad_norm": 0.7139991521835327,
+      "learning_rate": 0.00029998976315666445,
+      "loss": 4.6724,
+      "step": 73580
+    },
+    {
+      "epoch": 0.1034008539457669,
+      "grad_norm": 0.7393600940704346,
+      "learning_rate": 0.00029998967702556795,
+      "loss": 4.7427,
+      "step": 73590
+    },
+    {
+      "epoch": 0.10341490488393047,
+      "grad_norm": 0.7399834394454956,
+      "learning_rate": 0.00029998959053365406,
+      "loss": 4.7802,
+      "step": 73600
+    },
+    {
+      "epoch": 0.10342895582209405,
+      "grad_norm": 0.7233583331108093,
+      "learning_rate": 0.0002999895036809228,
+      "loss": 4.8864,
+      "step": 73610
+    },
+    {
+      "epoch": 0.10344300676025762,
+      "grad_norm": 0.7222065925598145,
+      "learning_rate": 0.0002999894164673745,
+      "loss": 4.771,
+      "step": 73620
+    },
+    {
+      "epoch": 0.1034570576984212,
+      "grad_norm": 0.7875381112098694,
+      "learning_rate": 0.0002999893288930093,
+      "loss": 4.6923,
+      "step": 73630
+    },
+    {
+      "epoch": 0.10347110863658478,
+      "grad_norm": 0.7450695037841797,
+      "learning_rate": 0.00029998924095782744,
+      "loss": 4.7587,
+      "step": 73640
+    },
+    {
+      "epoch": 0.10348515957474835,
+      "grad_norm": 0.7188162803649902,
+      "learning_rate": 0.00029998915266182916,
+      "loss": 4.8141,
+      "step": 73650
+    },
+    {
+      "epoch": 0.10349921051291193,
+      "grad_norm": 0.7098068594932556,
+      "learning_rate": 0.00029998906400501463,
+      "loss": 4.7548,
+      "step": 73660
+    },
+    {
+      "epoch": 0.1035132614510755,
+      "grad_norm": 0.730060875415802,
+      "learning_rate": 0.00029998897498738407,
+      "loss": 4.788,
+      "step": 73670
+    },
+    {
+      "epoch": 0.1035273123892391,
+      "grad_norm": 0.7018395662307739,
+      "learning_rate": 0.00029998888560893775,
+      "loss": 4.8347,
+      "step": 73680
+    },
+    {
+      "epoch": 0.10354136332740267,
+      "grad_norm": 0.7120745778083801,
+      "learning_rate": 0.0002999887958696758,
+      "loss": 4.8095,
+      "step": 73690
+    },
+    {
+      "epoch": 0.10355541426556625,
+      "grad_norm": 0.7242351174354553,
+      "learning_rate": 0.00029998870576959847,
+      "loss": 4.6808,
+      "step": 73700
+    },
+    {
+      "epoch": 0.10356946520372982,
+      "grad_norm": 0.7491177320480347,
+      "learning_rate": 0.00029998861530870595,
+      "loss": 4.7441,
+      "step": 73710
+    },
+    {
+      "epoch": 0.1035835161418934,
+      "grad_norm": 0.7404350638389587,
+      "learning_rate": 0.00029998852448699854,
+      "loss": 4.7472,
+      "step": 73720
+    },
+    {
+      "epoch": 0.10359756708005698,
+      "grad_norm": 0.7084073424339294,
+      "learning_rate": 0.0002999884333044764,
+      "loss": 4.7107,
+      "step": 73730
+    },
+    {
+      "epoch": 0.10361161801822055,
+      "grad_norm": 0.7354574799537659,
+      "learning_rate": 0.00029998834176113973,
+      "loss": 4.7334,
+      "step": 73740
+    },
+    {
+      "epoch": 0.10362566895638413,
+      "grad_norm": 0.7333629131317139,
+      "learning_rate": 0.0002999882498569888,
+      "loss": 4.7645,
+      "step": 73750
+    },
+    {
+      "epoch": 0.1036397198945477,
+      "grad_norm": 0.7341249585151672,
+      "learning_rate": 0.0002999881575920238,
+      "loss": 4.8465,
+      "step": 73760
+    },
+    {
+      "epoch": 0.10365377083271128,
+      "grad_norm": 0.7545302510261536,
+      "learning_rate": 0.00029998806496624495,
+      "loss": 4.7715,
+      "step": 73770
+    },
+    {
+      "epoch": 0.10366782177087486,
+      "grad_norm": 0.7346596121788025,
+      "learning_rate": 0.00029998797197965246,
+      "loss": 4.7478,
+      "step": 73780
+    },
+    {
+      "epoch": 0.10368187270903845,
+      "grad_norm": 0.7493293285369873,
+      "learning_rate": 0.0002999878786322466,
+      "loss": 4.6977,
+      "step": 73790
+    },
+    {
+      "epoch": 0.10369592364720202,
+      "grad_norm": 0.713477611541748,
+      "learning_rate": 0.00029998778492402756,
+      "loss": 4.7183,
+      "step": 73800
+    },
+    {
+      "epoch": 0.1037099745853656,
+      "grad_norm": 0.7505142688751221,
+      "learning_rate": 0.0002999876908549956,
+      "loss": 4.7706,
+      "step": 73810
+    },
+    {
+      "epoch": 0.10372402552352918,
+      "grad_norm": 0.7234914898872375,
+      "learning_rate": 0.0002999875964251509,
+      "loss": 4.7092,
+      "step": 73820
+    },
+    {
+      "epoch": 0.10373807646169275,
+      "grad_norm": 0.7134677171707153,
+      "learning_rate": 0.0002999875016344937,
+      "loss": 4.8585,
+      "step": 73830
+    },
+    {
+      "epoch": 0.10375212739985633,
+      "grad_norm": 0.7248352766036987,
+      "learning_rate": 0.00029998740648302425,
+      "loss": 4.7321,
+      "step": 73840
+    },
+    {
+      "epoch": 0.1037661783380199,
+      "grad_norm": 0.729803204536438,
+      "learning_rate": 0.0002999873109707428,
+      "loss": 4.8516,
+      "step": 73850
+    },
+    {
+      "epoch": 0.10378022927618348,
+      "grad_norm": 0.7194018363952637,
+      "learning_rate": 0.0002999872150976495,
+      "loss": 4.7418,
+      "step": 73860
+    },
+    {
+      "epoch": 0.10379428021434706,
+      "grad_norm": 0.7106111645698547,
+      "learning_rate": 0.0002999871188637446,
+      "loss": 4.7227,
+      "step": 73870
+    },
+    {
+      "epoch": 0.10380833115251063,
+      "grad_norm": 0.724184513092041,
+      "learning_rate": 0.0002999870222690284,
+      "loss": 4.7338,
+      "step": 73880
+    },
+    {
+      "epoch": 0.10382238209067422,
+      "grad_norm": 0.7321856617927551,
+      "learning_rate": 0.0002999869253135011,
+      "loss": 4.7236,
+      "step": 73890
+    },
+    {
+      "epoch": 0.1038364330288378,
+      "grad_norm": 0.7175441384315491,
+      "learning_rate": 0.00029998682799716294,
+      "loss": 4.8629,
+      "step": 73900
+    },
+    {
+      "epoch": 0.10385048396700138,
+      "grad_norm": 0.7437195181846619,
+      "learning_rate": 0.00029998673032001407,
+      "loss": 4.7403,
+      "step": 73910
+    },
+    {
+      "epoch": 0.10386453490516495,
+      "grad_norm": 0.7287278771400452,
+      "learning_rate": 0.0002999866322820548,
+      "loss": 4.7742,
+      "step": 73920
+    },
+    {
+      "epoch": 0.10387858584332853,
+      "grad_norm": 0.7289449572563171,
+      "learning_rate": 0.00029998653388328537,
+      "loss": 4.7351,
+      "step": 73930
+    },
+    {
+      "epoch": 0.1038926367814921,
+      "grad_norm": 0.7480982542037964,
+      "learning_rate": 0.0002999864351237061,
+      "loss": 4.7372,
+      "step": 73940
+    },
+    {
+      "epoch": 0.10390668771965568,
+      "grad_norm": 0.7479860186576843,
+      "learning_rate": 0.00029998633600331705,
+      "loss": 4.7979,
+      "step": 73950
+    },
+    {
+      "epoch": 0.10392073865781926,
+      "grad_norm": 0.756632387638092,
+      "learning_rate": 0.0002999862365221185,
+      "loss": 4.7318,
+      "step": 73960
+    },
+    {
+      "epoch": 0.10393478959598283,
+      "grad_norm": 0.722648561000824,
+      "learning_rate": 0.0002999861366801108,
+      "loss": 4.7916,
+      "step": 73970
+    },
+    {
+      "epoch": 0.10394884053414641,
+      "grad_norm": 0.7514876127243042,
+      "learning_rate": 0.0002999860364772941,
+      "loss": 4.8139,
+      "step": 73980
+    },
+    {
+      "epoch": 0.10396289147231,
+      "grad_norm": 0.7162560224533081,
+      "learning_rate": 0.0002999859359136686,
+      "loss": 4.7387,
+      "step": 73990
+    },
+    {
+      "epoch": 0.10397694241047357,
+      "grad_norm": 0.7170782089233398,
+      "learning_rate": 0.0002999858349892347,
+      "loss": 4.8234,
+      "step": 74000
+    },
+    {
+      "epoch": 0.10399099334863715,
+      "grad_norm": 0.7494903206825256,
+      "learning_rate": 0.0002999857337039925,
+      "loss": 4.7603,
+      "step": 74010
+    },
+    {
+      "epoch": 0.10400504428680073,
+      "grad_norm": 0.7144988775253296,
+      "learning_rate": 0.00029998563205794223,
+      "loss": 4.7661,
+      "step": 74020
+    },
+    {
+      "epoch": 0.1040190952249643,
+      "grad_norm": 0.7347742319107056,
+      "learning_rate": 0.0002999855300510843,
+      "loss": 4.8141,
+      "step": 74030
+    },
+    {
+      "epoch": 0.10403314616312788,
+      "grad_norm": 0.7606138586997986,
+      "learning_rate": 0.0002999854276834188,
+      "loss": 4.7885,
+      "step": 74040
+    },
+    {
+      "epoch": 0.10404719710129146,
+      "grad_norm": 0.7385061979293823,
+      "learning_rate": 0.00029998532495494604,
+      "loss": 4.8078,
+      "step": 74050
+    },
+    {
+      "epoch": 0.10406124803945503,
+      "grad_norm": 0.7367041110992432,
+      "learning_rate": 0.00029998522186566626,
+      "loss": 4.802,
+      "step": 74060
+    },
+    {
+      "epoch": 0.10407529897761861,
+      "grad_norm": 0.7521949410438538,
+      "learning_rate": 0.0002999851184155797,
+      "loss": 4.7509,
+      "step": 74070
+    },
+    {
+      "epoch": 0.10408934991578218,
+      "grad_norm": 0.746960461139679,
+      "learning_rate": 0.00029998501460468656,
+      "loss": 4.7074,
+      "step": 74080
+    },
+    {
+      "epoch": 0.10410340085394576,
+      "grad_norm": 0.7234981060028076,
+      "learning_rate": 0.0002999849104329872,
+      "loss": 4.8579,
+      "step": 74090
+    },
+    {
+      "epoch": 0.10411745179210935,
+      "grad_norm": 0.6993427872657776,
+      "learning_rate": 0.00029998480590048176,
+      "loss": 4.7195,
+      "step": 74100
+    },
+    {
+      "epoch": 0.10413150273027293,
+      "grad_norm": 0.7664044499397278,
+      "learning_rate": 0.0002999847010071706,
+      "loss": 4.8389,
+      "step": 74110
+    },
+    {
+      "epoch": 0.1041455536684365,
+      "grad_norm": 0.7423094511032104,
+      "learning_rate": 0.0002999845957530538,
+      "loss": 4.7013,
+      "step": 74120
+    },
+    {
+      "epoch": 0.10415960460660008,
+      "grad_norm": 0.7351399660110474,
+      "learning_rate": 0.00029998449013813184,
+      "loss": 4.8115,
+      "step": 74130
+    },
+    {
+      "epoch": 0.10417365554476365,
+      "grad_norm": 0.7567272186279297,
+      "learning_rate": 0.00029998438416240484,
+      "loss": 4.7673,
+      "step": 74140
+    },
+    {
+      "epoch": 0.10418770648292723,
+      "grad_norm": 0.7193549275398254,
+      "learning_rate": 0.00029998427782587303,
+      "loss": 4.7971,
+      "step": 74150
+    },
+    {
+      "epoch": 0.10420175742109081,
+      "grad_norm": 0.7518367767333984,
+      "learning_rate": 0.0002999841711285368,
+      "loss": 4.8342,
+      "step": 74160
+    },
+    {
+      "epoch": 0.10421580835925438,
+      "grad_norm": 0.7178200483322144,
+      "learning_rate": 0.0002999840640703963,
+      "loss": 4.7677,
+      "step": 74170
+    },
+    {
+      "epoch": 0.10422985929741796,
+      "grad_norm": 0.8699411749839783,
+      "learning_rate": 0.0002999839566514517,
+      "loss": 4.6964,
+      "step": 74180
+    },
+    {
+      "epoch": 0.10424391023558154,
+      "grad_norm": 0.7391669750213623,
+      "learning_rate": 0.0002999838488717035,
+      "loss": 4.7222,
+      "step": 74190
+    },
+    {
+      "epoch": 0.10425796117374513,
+      "grad_norm": 0.7199887633323669,
+      "learning_rate": 0.0002999837407311517,
+      "loss": 4.7128,
+      "step": 74200
+    },
+    {
+      "epoch": 0.1042720121119087,
+      "grad_norm": 0.7740340232849121,
+      "learning_rate": 0.00029998363222979676,
+      "loss": 4.6918,
+      "step": 74210
+    },
+    {
+      "epoch": 0.10428606305007228,
+      "grad_norm": 0.7272385954856873,
+      "learning_rate": 0.0002999835233676388,
+      "loss": 4.6641,
+      "step": 74220
+    },
+    {
+      "epoch": 0.10430011398823585,
+      "grad_norm": 0.7720179557800293,
+      "learning_rate": 0.0002999834141446782,
+      "loss": 4.8111,
+      "step": 74230
+    },
+    {
+      "epoch": 0.10431416492639943,
+      "grad_norm": 0.7326408624649048,
+      "learning_rate": 0.0002999833045609152,
+      "loss": 4.7686,
+      "step": 74240
+    },
+    {
+      "epoch": 0.104328215864563,
+      "grad_norm": 0.708678126335144,
+      "learning_rate": 0.00029998319461635,
+      "loss": 4.7862,
+      "step": 74250
+    },
+    {
+      "epoch": 0.10434226680272658,
+      "grad_norm": 0.7066006660461426,
+      "learning_rate": 0.00029998308431098287,
+      "loss": 4.7004,
+      "step": 74260
+    },
+    {
+      "epoch": 0.10435631774089016,
+      "grad_norm": 0.7312318682670593,
+      "learning_rate": 0.00029998297364481415,
+      "loss": 4.8703,
+      "step": 74270
+    },
+    {
+      "epoch": 0.10437036867905373,
+      "grad_norm": 0.7320531606674194,
+      "learning_rate": 0.00029998286261784403,
+      "loss": 4.694,
+      "step": 74280
+    },
+    {
+      "epoch": 0.10438441961721731,
+      "grad_norm": 0.743951141834259,
+      "learning_rate": 0.0002999827512300728,
+      "loss": 4.8147,
+      "step": 74290
+    },
+    {
+      "epoch": 0.1043984705553809,
+      "grad_norm": 0.7393697500228882,
+      "learning_rate": 0.00029998263948150075,
+      "loss": 4.8336,
+      "step": 74300
+    },
+    {
+      "epoch": 0.10441252149354448,
+      "grad_norm": 0.7255837917327881,
+      "learning_rate": 0.00029998252737212813,
+      "loss": 4.7326,
+      "step": 74310
+    },
+    {
+      "epoch": 0.10442657243170805,
+      "grad_norm": 0.7354263067245483,
+      "learning_rate": 0.0002999824149019552,
+      "loss": 4.7204,
+      "step": 74320
+    },
+    {
+      "epoch": 0.10444062336987163,
+      "grad_norm": 0.7385523915290833,
+      "learning_rate": 0.0002999823020709823,
+      "loss": 4.7987,
+      "step": 74330
+    },
+    {
+      "epoch": 0.1044546743080352,
+      "grad_norm": 0.7048619985580444,
+      "learning_rate": 0.0002999821888792096,
+      "loss": 4.7279,
+      "step": 74340
+    },
+    {
+      "epoch": 0.10446872524619878,
+      "grad_norm": 0.7030783295631409,
+      "learning_rate": 0.00029998207532663736,
+      "loss": 4.8038,
+      "step": 74350
+    },
+    {
+      "epoch": 0.10448277618436236,
+      "grad_norm": 0.7411651611328125,
+      "learning_rate": 0.000299981961413266,
+      "loss": 4.9361,
+      "step": 74360
+    },
+    {
+      "epoch": 0.10449682712252593,
+      "grad_norm": 0.7514051198959351,
+      "learning_rate": 0.00029998184713909563,
+      "loss": 4.7624,
+      "step": 74370
+    },
+    {
+      "epoch": 0.10451087806068951,
+      "grad_norm": 0.7228174209594727,
+      "learning_rate": 0.0002999817325041267,
+      "loss": 4.8236,
+      "step": 74380
+    },
+    {
+      "epoch": 0.10452492899885309,
+      "grad_norm": 0.7124325633049011,
+      "learning_rate": 0.0002999816175083593,
+      "loss": 4.8186,
+      "step": 74390
+    },
+    {
+      "epoch": 0.10453897993701666,
+      "grad_norm": 0.7320361137390137,
+      "learning_rate": 0.00029998150215179383,
+      "loss": 4.8853,
+      "step": 74400
+    },
+    {
+      "epoch": 0.10455303087518025,
+      "grad_norm": 0.7228125333786011,
+      "learning_rate": 0.0002999813864344305,
+      "loss": 4.6899,
+      "step": 74410
+    },
+    {
+      "epoch": 0.10456708181334383,
+      "grad_norm": 0.7609881162643433,
+      "learning_rate": 0.00029998127035626965,
+      "loss": 4.6897,
+      "step": 74420
+    },
+    {
+      "epoch": 0.1045811327515074,
+      "grad_norm": 0.7024797797203064,
+      "learning_rate": 0.0002999811539173115,
+      "loss": 4.756,
+      "step": 74430
+    },
+    {
+      "epoch": 0.10459518368967098,
+      "grad_norm": 0.7212551832199097,
+      "learning_rate": 0.00029998103711755637,
+      "loss": 4.8677,
+      "step": 74440
+    },
+    {
+      "epoch": 0.10460923462783456,
+      "grad_norm": 0.7069899439811707,
+      "learning_rate": 0.00029998091995700455,
+      "loss": 4.8609,
+      "step": 74450
+    },
+    {
+      "epoch": 0.10462328556599813,
+      "grad_norm": 0.7125039100646973,
+      "learning_rate": 0.00029998080243565626,
+      "loss": 4.6833,
+      "step": 74460
+    },
+    {
+      "epoch": 0.10463733650416171,
+      "grad_norm": 0.7615266442298889,
+      "learning_rate": 0.00029998068455351184,
+      "loss": 4.6864,
+      "step": 74470
+    },
+    {
+      "epoch": 0.10465138744232529,
+      "grad_norm": 0.7260366082191467,
+      "learning_rate": 0.00029998056631057155,
+      "loss": 4.7931,
+      "step": 74480
+    },
+    {
+      "epoch": 0.10466543838048886,
+      "grad_norm": 0.7347187995910645,
+      "learning_rate": 0.00029998044770683567,
+      "loss": 4.797,
+      "step": 74490
+    },
+    {
+      "epoch": 0.10467948931865244,
+      "grad_norm": 0.7333105206489563,
+      "learning_rate": 0.0002999803287423045,
+      "loss": 4.7057,
+      "step": 74500
+    },
+    {
+      "epoch": 0.10469354025681603,
+      "grad_norm": 0.7438333630561829,
+      "learning_rate": 0.0002999802094169783,
+      "loss": 4.8193,
+      "step": 74510
+    },
+    {
+      "epoch": 0.1047075911949796,
+      "grad_norm": 0.836247980594635,
+      "learning_rate": 0.00029998008973085744,
+      "loss": 4.6534,
+      "step": 74520
+    },
+    {
+      "epoch": 0.10472164213314318,
+      "grad_norm": 0.7281263470649719,
+      "learning_rate": 0.00029997996968394216,
+      "loss": 4.7365,
+      "step": 74530
+    },
+    {
+      "epoch": 0.10473569307130676,
+      "grad_norm": 0.7797707319259644,
+      "learning_rate": 0.0002999798492762327,
+      "loss": 4.6875,
+      "step": 74540
+    },
+    {
+      "epoch": 0.10474974400947033,
+      "grad_norm": 0.7314170598983765,
+      "learning_rate": 0.00029997972850772937,
+      "loss": 4.7991,
+      "step": 74550
+    },
+    {
+      "epoch": 0.10476379494763391,
+      "grad_norm": 0.8902124762535095,
+      "learning_rate": 0.00029997960737843246,
+      "loss": 4.7621,
+      "step": 74560
+    },
+    {
+      "epoch": 0.10477784588579749,
+      "grad_norm": 0.7190029621124268,
+      "learning_rate": 0.00029997948588834233,
+      "loss": 4.8784,
+      "step": 74570
+    },
+    {
+      "epoch": 0.10479189682396106,
+      "grad_norm": 0.7397128343582153,
+      "learning_rate": 0.00029997936403745916,
+      "loss": 4.8362,
+      "step": 74580
+    },
+    {
+      "epoch": 0.10480594776212464,
+      "grad_norm": 0.7313927412033081,
+      "learning_rate": 0.0002999792418257833,
+      "loss": 4.788,
+      "step": 74590
+    },
+    {
+      "epoch": 0.10481999870028821,
+      "grad_norm": 0.7316232919692993,
+      "learning_rate": 0.00029997911925331514,
+      "loss": 4.6484,
+      "step": 74600
+    },
+    {
+      "epoch": 0.1048340496384518,
+      "grad_norm": 0.7240854501724243,
+      "learning_rate": 0.00029997899632005484,
+      "loss": 4.6801,
+      "step": 74610
+    },
+    {
+      "epoch": 0.10484810057661538,
+      "grad_norm": 0.7367461323738098,
+      "learning_rate": 0.0002999788730260027,
+      "loss": 4.7329,
+      "step": 74620
+    },
+    {
+      "epoch": 0.10486215151477896,
+      "grad_norm": 0.7352172136306763,
+      "learning_rate": 0.00029997874937115905,
+      "loss": 4.6665,
+      "step": 74630
+    },
+    {
+      "epoch": 0.10487620245294253,
+      "grad_norm": 0.7014811635017395,
+      "learning_rate": 0.0002999786253555242,
+      "loss": 4.7753,
+      "step": 74640
+    },
+    {
+      "epoch": 0.10489025339110611,
+      "grad_norm": 0.7095927596092224,
+      "learning_rate": 0.0002999785009790985,
+      "loss": 4.7157,
+      "step": 74650
+    },
+    {
+      "epoch": 0.10490430432926968,
+      "grad_norm": 0.7442272305488586,
+      "learning_rate": 0.00029997837624188206,
+      "loss": 4.7303,
+      "step": 74660
+    },
+    {
+      "epoch": 0.10491835526743326,
+      "grad_norm": 0.7289188504219055,
+      "learning_rate": 0.00029997825114387545,
+      "loss": 4.8395,
+      "step": 74670
+    },
+    {
+      "epoch": 0.10493240620559684,
+      "grad_norm": 0.7389072179794312,
+      "learning_rate": 0.0002999781256850788,
+      "loss": 4.7305,
+      "step": 74680
+    },
+    {
+      "epoch": 0.10494645714376041,
+      "grad_norm": 0.7263525128364563,
+      "learning_rate": 0.00029997799986549237,
+      "loss": 4.8211,
+      "step": 74690
+    },
+    {
+      "epoch": 0.10496050808192399,
+      "grad_norm": 0.7205453515052795,
+      "learning_rate": 0.00029997787368511654,
+      "loss": 4.8692,
+      "step": 74700
+    },
+    {
+      "epoch": 0.10497455902008757,
+      "grad_norm": 0.7371070384979248,
+      "learning_rate": 0.0002999777471439517,
+      "loss": 4.6703,
+      "step": 74710
+    },
+    {
+      "epoch": 0.10498860995825116,
+      "grad_norm": 0.7681359648704529,
+      "learning_rate": 0.000299977620241998,
+      "loss": 4.6302,
+      "step": 74720
+    },
+    {
+      "epoch": 0.10500266089641473,
+      "grad_norm": 0.7186245322227478,
+      "learning_rate": 0.00029997749297925577,
+      "loss": 4.8669,
+      "step": 74730
+    },
+    {
+      "epoch": 0.10501671183457831,
+      "grad_norm": 0.7410948872566223,
+      "learning_rate": 0.00029997736535572543,
+      "loss": 4.8782,
+      "step": 74740
+    },
+    {
+      "epoch": 0.10503076277274188,
+      "grad_norm": 0.7192717790603638,
+      "learning_rate": 0.00029997723737140713,
+      "loss": 4.6931,
+      "step": 74750
+    },
+    {
+      "epoch": 0.10504481371090546,
+      "grad_norm": 0.6977980732917786,
+      "learning_rate": 0.0002999771090263013,
+      "loss": 4.8123,
+      "step": 74760
+    },
+    {
+      "epoch": 0.10505886464906904,
+      "grad_norm": 0.7145591378211975,
+      "learning_rate": 0.00029997698032040827,
+      "loss": 4.7737,
+      "step": 74770
+    },
+    {
+      "epoch": 0.10507291558723261,
+      "grad_norm": 0.7509664297103882,
+      "learning_rate": 0.0002999768512537282,
+      "loss": 4.6107,
+      "step": 74780
+    },
+    {
+      "epoch": 0.10508696652539619,
+      "grad_norm": 0.7300460338592529,
+      "learning_rate": 0.0002999767218262615,
+      "loss": 4.6742,
+      "step": 74790
+    },
+    {
+      "epoch": 0.10510101746355976,
+      "grad_norm": 0.711122453212738,
+      "learning_rate": 0.0002999765920380085,
+      "loss": 4.8037,
+      "step": 74800
+    },
+    {
+      "epoch": 0.10511506840172334,
+      "grad_norm": 0.6955252289772034,
+      "learning_rate": 0.0002999764618889695,
+      "loss": 4.9123,
+      "step": 74810
+    },
+    {
+      "epoch": 0.10512911933988693,
+      "grad_norm": 0.7196218967437744,
+      "learning_rate": 0.00029997633137914475,
+      "loss": 4.7226,
+      "step": 74820
+    },
+    {
+      "epoch": 0.10514317027805051,
+      "grad_norm": 0.7192176580429077,
+      "learning_rate": 0.0002999762005085346,
+      "loss": 4.5709,
+      "step": 74830
+    },
+    {
+      "epoch": 0.10515722121621408,
+      "grad_norm": 0.7230987548828125,
+      "learning_rate": 0.0002999760692771394,
+      "loss": 4.7651,
+      "step": 74840
+    },
+    {
+      "epoch": 0.10517127215437766,
+      "grad_norm": 0.705202579498291,
+      "learning_rate": 0.0002999759376849594,
+      "loss": 4.6931,
+      "step": 74850
+    },
+    {
+      "epoch": 0.10518532309254124,
+      "grad_norm": 0.7315492630004883,
+      "learning_rate": 0.000299975805731995,
+      "loss": 4.7949,
+      "step": 74860
+    },
+    {
+      "epoch": 0.10519937403070481,
+      "grad_norm": 0.7159463763237,
+      "learning_rate": 0.0002999756734182464,
+      "loss": 4.72,
+      "step": 74870
+    },
+    {
+      "epoch": 0.10521342496886839,
+      "grad_norm": 0.7130573987960815,
+      "learning_rate": 0.00029997554074371414,
+      "loss": 4.7058,
+      "step": 74880
+    },
+    {
+      "epoch": 0.10522747590703196,
+      "grad_norm": 0.714362621307373,
+      "learning_rate": 0.00029997540770839825,
+      "loss": 4.7595,
+      "step": 74890
+    },
+    {
+      "epoch": 0.10524152684519554,
+      "grad_norm": 0.7358441352844238,
+      "learning_rate": 0.00029997527431229923,
+      "loss": 4.69,
+      "step": 74900
+    },
+    {
+      "epoch": 0.10525557778335912,
+      "grad_norm": 0.7290864586830139,
+      "learning_rate": 0.00029997514055541736,
+      "loss": 4.7994,
+      "step": 74910
+    },
+    {
+      "epoch": 0.1052696287215227,
+      "grad_norm": 0.7372201085090637,
+      "learning_rate": 0.000299975006437753,
+      "loss": 4.8169,
+      "step": 74920
+    },
+    {
+      "epoch": 0.10528367965968628,
+      "grad_norm": 0.7162683010101318,
+      "learning_rate": 0.0002999748719593064,
+      "loss": 4.7578,
+      "step": 74930
+    },
+    {
+      "epoch": 0.10529773059784986,
+      "grad_norm": 0.7158368825912476,
+      "learning_rate": 0.0002999747371200779,
+      "loss": 4.8043,
+      "step": 74940
+    },
+    {
+      "epoch": 0.10531178153601344,
+      "grad_norm": 0.7241293787956238,
+      "learning_rate": 0.00029997460192006793,
+      "loss": 4.7606,
+      "step": 74950
+    },
+    {
+      "epoch": 0.10532583247417701,
+      "grad_norm": 0.7109899520874023,
+      "learning_rate": 0.00029997446635927666,
+      "loss": 4.6894,
+      "step": 74960
+    },
+    {
+      "epoch": 0.10533988341234059,
+      "grad_norm": 0.7288115620613098,
+      "learning_rate": 0.0002999743304377045,
+      "loss": 4.7761,
+      "step": 74970
+    },
+    {
+      "epoch": 0.10535393435050416,
+      "grad_norm": 0.7241887450218201,
+      "learning_rate": 0.00029997419415535173,
+      "loss": 4.726,
+      "step": 74980
+    },
+    {
+      "epoch": 0.10536798528866774,
+      "grad_norm": 0.7299656271934509,
+      "learning_rate": 0.0002999740575122187,
+      "loss": 4.766,
+      "step": 74990
+    },
+    {
+      "epoch": 0.10538203622683132,
+      "grad_norm": 0.9255989789962769,
+      "learning_rate": 0.0002999739205083058,
+      "loss": 4.792,
+      "step": 75000
+    },
+    {
+      "epoch": 0.10539608716499489,
+      "grad_norm": 0.7332300543785095,
+      "learning_rate": 0.0002999737831436133,
+      "loss": 4.7162,
+      "step": 75010
+    },
+    {
+      "epoch": 0.10541013810315847,
+      "grad_norm": 0.7358729839324951,
+      "learning_rate": 0.00029997364541814155,
+      "loss": 4.8229,
+      "step": 75020
+    },
+    {
+      "epoch": 0.10542418904132206,
+      "grad_norm": 0.7899588346481323,
+      "learning_rate": 0.00029997350733189084,
+      "loss": 4.6358,
+      "step": 75030
+    },
+    {
+      "epoch": 0.10543823997948563,
+      "grad_norm": 0.7667545676231384,
+      "learning_rate": 0.0002999733688848615,
+      "loss": 4.8529,
+      "step": 75040
+    },
+    {
+      "epoch": 0.10545229091764921,
+      "grad_norm": 0.7302919626235962,
+      "learning_rate": 0.00029997323007705396,
+      "loss": 4.6845,
+      "step": 75050
+    },
+    {
+      "epoch": 0.10546634185581279,
+      "grad_norm": 0.7035412788391113,
+      "learning_rate": 0.0002999730909084684,
+      "loss": 4.7805,
+      "step": 75060
+    },
+    {
+      "epoch": 0.10548039279397636,
+      "grad_norm": 0.7011492848396301,
+      "learning_rate": 0.00029997295137910535,
+      "loss": 4.6031,
+      "step": 75070
+    },
+    {
+      "epoch": 0.10549444373213994,
+      "grad_norm": 0.7378577589988708,
+      "learning_rate": 0.00029997281148896497,
+      "loss": 4.8141,
+      "step": 75080
+    },
+    {
+      "epoch": 0.10550849467030352,
+      "grad_norm": 0.7121955156326294,
+      "learning_rate": 0.0002999726712380477,
+      "loss": 4.7603,
+      "step": 75090
+    },
+    {
+      "epoch": 0.10552254560846709,
+      "grad_norm": 0.7434091567993164,
+      "learning_rate": 0.00029997253062635386,
+      "loss": 4.9379,
+      "step": 75100
+    },
+    {
+      "epoch": 0.10553659654663067,
+      "grad_norm": 0.7290818095207214,
+      "learning_rate": 0.0002999723896538837,
+      "loss": 4.7528,
+      "step": 75110
+    },
+    {
+      "epoch": 0.10555064748479424,
+      "grad_norm": 0.7576467394828796,
+      "learning_rate": 0.0002999722483206377,
+      "loss": 4.8125,
+      "step": 75120
+    },
+    {
+      "epoch": 0.10556469842295783,
+      "grad_norm": 0.7280282974243164,
+      "learning_rate": 0.0002999721066266161,
+      "loss": 4.7116,
+      "step": 75130
+    },
+    {
+      "epoch": 0.10557874936112141,
+      "grad_norm": 0.7176198959350586,
+      "learning_rate": 0.0002999719645718193,
+      "loss": 4.8024,
+      "step": 75140
+    },
+    {
+      "epoch": 0.10559280029928499,
+      "grad_norm": 0.7369096279144287,
+      "learning_rate": 0.00029997182215624755,
+      "loss": 4.7124,
+      "step": 75150
+    },
+    {
+      "epoch": 0.10560685123744856,
+      "grad_norm": 0.7322725653648376,
+      "learning_rate": 0.00029997167937990125,
+      "loss": 4.7161,
+      "step": 75160
+    },
+    {
+      "epoch": 0.10562090217561214,
+      "grad_norm": 0.7204955816268921,
+      "learning_rate": 0.0002999715362427808,
+      "loss": 4.7834,
+      "step": 75170
+    },
+    {
+      "epoch": 0.10563495311377571,
+      "grad_norm": 0.7370813488960266,
+      "learning_rate": 0.0002999713927448865,
+      "loss": 4.7807,
+      "step": 75180
+    },
+    {
+      "epoch": 0.10564900405193929,
+      "grad_norm": 0.737962543964386,
+      "learning_rate": 0.00029997124888621866,
+      "loss": 4.7968,
+      "step": 75190
+    },
+    {
+      "epoch": 0.10566305499010287,
+      "grad_norm": 0.7260681986808777,
+      "learning_rate": 0.0002999711046667777,
+      "loss": 4.6648,
+      "step": 75200
+    },
+    {
+      "epoch": 0.10567710592826644,
+      "grad_norm": 0.7455945014953613,
+      "learning_rate": 0.00029997096008656384,
+      "loss": 4.7407,
+      "step": 75210
+    },
+    {
+      "epoch": 0.10569115686643002,
+      "grad_norm": 0.710437536239624,
+      "learning_rate": 0.0002999708151455776,
+      "loss": 4.7474,
+      "step": 75220
+    },
+    {
+      "epoch": 0.10570520780459361,
+      "grad_norm": 0.7501201033592224,
+      "learning_rate": 0.0002999706698438192,
+      "loss": 4.6896,
+      "step": 75230
+    },
+    {
+      "epoch": 0.10571925874275719,
+      "grad_norm": 0.7423140406608582,
+      "learning_rate": 0.00029997052418128897,
+      "loss": 4.8219,
+      "step": 75240
+    },
+    {
+      "epoch": 0.10573330968092076,
+      "grad_norm": 0.8542389869689941,
+      "learning_rate": 0.0002999703781579874,
+      "loss": 4.7801,
+      "step": 75250
+    },
+    {
+      "epoch": 0.10574736061908434,
+      "grad_norm": 0.736545979976654,
+      "learning_rate": 0.0002999702317739147,
+      "loss": 4.8571,
+      "step": 75260
+    },
+    {
+      "epoch": 0.10576141155724791,
+      "grad_norm": 0.7311723828315735,
+      "learning_rate": 0.00029997008502907134,
+      "loss": 4.8457,
+      "step": 75270
+    },
+    {
+      "epoch": 0.10577546249541149,
+      "grad_norm": 0.7157546877861023,
+      "learning_rate": 0.00029996993792345757,
+      "loss": 4.827,
+      "step": 75280
+    },
+    {
+      "epoch": 0.10578951343357507,
+      "grad_norm": 0.7155654430389404,
+      "learning_rate": 0.0002999697904570738,
+      "loss": 4.9051,
+      "step": 75290
+    },
+    {
+      "epoch": 0.10580356437173864,
+      "grad_norm": 0.7166204452514648,
+      "learning_rate": 0.0002999696426299204,
+      "loss": 4.7967,
+      "step": 75300
+    },
+    {
+      "epoch": 0.10581761530990222,
+      "grad_norm": 0.7281050086021423,
+      "learning_rate": 0.0002999694944419976,
+      "loss": 4.7054,
+      "step": 75310
+    },
+    {
+      "epoch": 0.1058316662480658,
+      "grad_norm": 0.7242557406425476,
+      "learning_rate": 0.00029996934589330594,
+      "loss": 4.7892,
+      "step": 75320
+    },
+    {
+      "epoch": 0.10584571718622938,
+      "grad_norm": 0.7574252486228943,
+      "learning_rate": 0.0002999691969838457,
+      "loss": 4.7271,
+      "step": 75330
+    },
+    {
+      "epoch": 0.10585976812439296,
+      "grad_norm": 0.7095407843589783,
+      "learning_rate": 0.0002999690477136172,
+      "loss": 4.7575,
+      "step": 75340
+    },
+    {
+      "epoch": 0.10587381906255654,
+      "grad_norm": 0.7685112953186035,
+      "learning_rate": 0.00029996889808262087,
+      "loss": 4.8053,
+      "step": 75350
+    },
+    {
+      "epoch": 0.10588787000072011,
+      "grad_norm": 0.7352884411811829,
+      "learning_rate": 0.000299968748090857,
+      "loss": 4.7534,
+      "step": 75360
+    },
+    {
+      "epoch": 0.10590192093888369,
+      "grad_norm": 0.7439702153205872,
+      "learning_rate": 0.000299968597738326,
+      "loss": 4.7675,
+      "step": 75370
+    },
+    {
+      "epoch": 0.10591597187704727,
+      "grad_norm": 0.7124572396278381,
+      "learning_rate": 0.00029996844702502817,
+      "loss": 4.8264,
+      "step": 75380
+    },
+    {
+      "epoch": 0.10593002281521084,
+      "grad_norm": 0.7124819159507751,
+      "learning_rate": 0.00029996829595096394,
+      "loss": 4.764,
+      "step": 75390
+    },
+    {
+      "epoch": 0.10594407375337442,
+      "grad_norm": 0.7227981686592102,
+      "learning_rate": 0.00029996814451613366,
+      "loss": 4.672,
+      "step": 75400
+    },
+    {
+      "epoch": 0.105958124691538,
+      "grad_norm": 0.7123106718063354,
+      "learning_rate": 0.00029996799272053766,
+      "loss": 4.8143,
+      "step": 75410
+    },
+    {
+      "epoch": 0.10597217562970157,
+      "grad_norm": 0.7216246724128723,
+      "learning_rate": 0.00029996784056417636,
+      "loss": 4.8079,
+      "step": 75420
+    },
+    {
+      "epoch": 0.10598622656786515,
+      "grad_norm": 0.7304632067680359,
+      "learning_rate": 0.00029996768804705005,
+      "loss": 4.7621,
+      "step": 75430
+    },
+    {
+      "epoch": 0.10600027750602874,
+      "grad_norm": 0.7460058927536011,
+      "learning_rate": 0.00029996753516915915,
+      "loss": 4.7263,
+      "step": 75440
+    },
+    {
+      "epoch": 0.10601432844419231,
+      "grad_norm": 0.7226387858390808,
+      "learning_rate": 0.00029996738193050404,
+      "loss": 4.8106,
+      "step": 75450
+    },
+    {
+      "epoch": 0.10602837938235589,
+      "grad_norm": 0.7301837801933289,
+      "learning_rate": 0.00029996722833108505,
+      "loss": 4.8293,
+      "step": 75460
+    },
+    {
+      "epoch": 0.10604243032051947,
+      "grad_norm": 0.7361141443252563,
+      "learning_rate": 0.00029996707437090255,
+      "loss": 4.8255,
+      "step": 75470
+    },
+    {
+      "epoch": 0.10605648125868304,
+      "grad_norm": 0.726216733455658,
+      "learning_rate": 0.000299966920049957,
+      "loss": 4.7971,
+      "step": 75480
+    },
+    {
+      "epoch": 0.10607053219684662,
+      "grad_norm": 0.7054420113563538,
+      "learning_rate": 0.0002999667653682486,
+      "loss": 4.8014,
+      "step": 75490
+    },
+    {
+      "epoch": 0.1060845831350102,
+      "grad_norm": 0.7391892075538635,
+      "learning_rate": 0.00029996661032577785,
+      "loss": 4.8289,
+      "step": 75500
+    },
+    {
+      "epoch": 0.10609863407317377,
+      "grad_norm": 0.7168707251548767,
+      "learning_rate": 0.00029996645492254514,
+      "loss": 4.6734,
+      "step": 75510
+    },
+    {
+      "epoch": 0.10611268501133735,
+      "grad_norm": 0.7321151494979858,
+      "learning_rate": 0.0002999662991585507,
+      "loss": 4.7675,
+      "step": 75520
+    },
+    {
+      "epoch": 0.10612673594950092,
+      "grad_norm": 0.7170401215553284,
+      "learning_rate": 0.00029996614303379513,
+      "loss": 4.6391,
+      "step": 75530
+    },
+    {
+      "epoch": 0.10614078688766451,
+      "grad_norm": 0.6838593482971191,
+      "learning_rate": 0.00029996598654827864,
+      "loss": 4.7761,
+      "step": 75540
+    },
+    {
+      "epoch": 0.10615483782582809,
+      "grad_norm": 0.7038040161132812,
+      "learning_rate": 0.00029996582970200157,
+      "loss": 4.8264,
+      "step": 75550
+    },
+    {
+      "epoch": 0.10616888876399166,
+      "grad_norm": 0.7399590015411377,
+      "learning_rate": 0.00029996567249496447,
+      "loss": 4.7957,
+      "step": 75560
+    },
+    {
+      "epoch": 0.10618293970215524,
+      "grad_norm": 0.7103729248046875,
+      "learning_rate": 0.0002999655149271675,
+      "loss": 4.8062,
+      "step": 75570
+    },
+    {
+      "epoch": 0.10619699064031882,
+      "grad_norm": 0.7498480081558228,
+      "learning_rate": 0.0002999653569986113,
+      "loss": 4.7943,
+      "step": 75580
+    },
+    {
+      "epoch": 0.10621104157848239,
+      "grad_norm": 0.730226993560791,
+      "learning_rate": 0.000299965198709296,
+      "loss": 4.7162,
+      "step": 75590
+    },
+    {
+      "epoch": 0.10622509251664597,
+      "grad_norm": 0.7429940700531006,
+      "learning_rate": 0.0002999650400592221,
+      "loss": 4.7841,
+      "step": 75600
+    },
+    {
+      "epoch": 0.10623914345480955,
+      "grad_norm": 0.7538554668426514,
+      "learning_rate": 0.00029996488104839,
+      "loss": 4.7471,
+      "step": 75610
+    },
+    {
+      "epoch": 0.10625319439297312,
+      "grad_norm": 0.7402827739715576,
+      "learning_rate": 0.0002999647216768001,
+      "loss": 4.6182,
+      "step": 75620
+    },
+    {
+      "epoch": 0.1062672453311367,
+      "grad_norm": 0.7203502058982849,
+      "learning_rate": 0.0002999645619444526,
+      "loss": 4.7311,
+      "step": 75630
+    },
+    {
+      "epoch": 0.10628129626930029,
+      "grad_norm": 0.729570209980011,
+      "learning_rate": 0.00029996440185134815,
+      "loss": 4.7514,
+      "step": 75640
+    },
+    {
+      "epoch": 0.10629534720746386,
+      "grad_norm": 0.8635588884353638,
+      "learning_rate": 0.00029996424139748696,
+      "loss": 4.7753,
+      "step": 75650
+    },
+    {
+      "epoch": 0.10630939814562744,
+      "grad_norm": 0.7464445233345032,
+      "learning_rate": 0.00029996408058286943,
+      "loss": 4.6894,
+      "step": 75660
+    },
+    {
+      "epoch": 0.10632344908379102,
+      "grad_norm": 0.7225930094718933,
+      "learning_rate": 0.0002999639194074961,
+      "loss": 4.7286,
+      "step": 75670
+    },
+    {
+      "epoch": 0.10633750002195459,
+      "grad_norm": 0.7263953685760498,
+      "learning_rate": 0.0002999637578713671,
+      "loss": 4.7748,
+      "step": 75680
+    },
+    {
+      "epoch": 0.10635155096011817,
+      "grad_norm": 0.7385531663894653,
+      "learning_rate": 0.000299963595974483,
+      "loss": 4.7724,
+      "step": 75690
+    },
+    {
+      "epoch": 0.10636560189828174,
+      "grad_norm": 0.6999334096908569,
+      "learning_rate": 0.0002999634337168441,
+      "loss": 4.8438,
+      "step": 75700
+    },
+    {
+      "epoch": 0.10637965283644532,
+      "grad_norm": 0.7192821502685547,
+      "learning_rate": 0.0002999632710984509,
+      "loss": 4.7399,
+      "step": 75710
+    },
+    {
+      "epoch": 0.1063937037746089,
+      "grad_norm": 0.7049062252044678,
+      "learning_rate": 0.0002999631081193037,
+      "loss": 4.7818,
+      "step": 75720
+    },
+    {
+      "epoch": 0.10640775471277247,
+      "grad_norm": 0.7248371839523315,
+      "learning_rate": 0.00029996294477940296,
+      "loss": 4.624,
+      "step": 75730
+    },
+    {
+      "epoch": 0.10642180565093605,
+      "grad_norm": 0.6994637250900269,
+      "learning_rate": 0.000299962781078749,
+      "loss": 4.7474,
+      "step": 75740
+    },
+    {
+      "epoch": 0.10643585658909964,
+      "grad_norm": 0.744132936000824,
+      "learning_rate": 0.0002999626170173422,
+      "loss": 4.7988,
+      "step": 75750
+    },
+    {
+      "epoch": 0.10644990752726322,
+      "grad_norm": 0.7178496718406677,
+      "learning_rate": 0.00029996245259518304,
+      "loss": 4.7616,
+      "step": 75760
+    },
+    {
+      "epoch": 0.10646395846542679,
+      "grad_norm": 0.7215403318405151,
+      "learning_rate": 0.0002999622878122719,
+      "loss": 4.7007,
+      "step": 75770
+    },
+    {
+      "epoch": 0.10647800940359037,
+      "grad_norm": 0.7298187613487244,
+      "learning_rate": 0.00029996212266860913,
+      "loss": 4.7248,
+      "step": 75780
+    },
+    {
+      "epoch": 0.10649206034175394,
+      "grad_norm": 0.7294317483901978,
+      "learning_rate": 0.00029996195716419514,
+      "loss": 4.6084,
+      "step": 75790
+    },
+    {
+      "epoch": 0.10650611127991752,
+      "grad_norm": 0.7263827919960022,
+      "learning_rate": 0.0002999617912990303,
+      "loss": 4.7366,
+      "step": 75800
+    },
+    {
+      "epoch": 0.1065201622180811,
+      "grad_norm": 0.768453061580658,
+      "learning_rate": 0.00029996162507311513,
+      "loss": 4.8619,
+      "step": 75810
+    },
+    {
+      "epoch": 0.10653421315624467,
+      "grad_norm": 0.7260246276855469,
+      "learning_rate": 0.00029996145848644993,
+      "loss": 4.7207,
+      "step": 75820
+    },
+    {
+      "epoch": 0.10654826409440825,
+      "grad_norm": 0.7110903859138489,
+      "learning_rate": 0.00029996129153903506,
+      "loss": 4.7166,
+      "step": 75830
+    },
+    {
+      "epoch": 0.10656231503257182,
+      "grad_norm": 0.7710065245628357,
+      "learning_rate": 0.00029996112423087104,
+      "loss": 4.7668,
+      "step": 75840
+    },
+    {
+      "epoch": 0.10657636597073541,
+      "grad_norm": 0.7256232500076294,
+      "learning_rate": 0.00029996095656195816,
+      "loss": 4.7379,
+      "step": 75850
+    },
+    {
+      "epoch": 0.10659041690889899,
+      "grad_norm": 0.7331588864326477,
+      "learning_rate": 0.0002999607885322969,
+      "loss": 4.7213,
+      "step": 75860
+    },
+    {
+      "epoch": 0.10660446784706257,
+      "grad_norm": 0.7234667539596558,
+      "learning_rate": 0.0002999606201418876,
+      "loss": 4.9289,
+      "step": 75870
+    },
+    {
+      "epoch": 0.10661851878522614,
+      "grad_norm": 0.7169747352600098,
+      "learning_rate": 0.0002999604513907308,
+      "loss": 4.804,
+      "step": 75880
+    },
+    {
+      "epoch": 0.10663256972338972,
+      "grad_norm": 0.7151200771331787,
+      "learning_rate": 0.0002999602822788267,
+      "loss": 4.7486,
+      "step": 75890
+    },
+    {
+      "epoch": 0.1066466206615533,
+      "grad_norm": 0.7405186295509338,
+      "learning_rate": 0.00029996011280617585,
+      "loss": 4.7887,
+      "step": 75900
+    },
+    {
+      "epoch": 0.10666067159971687,
+      "grad_norm": 0.712820291519165,
+      "learning_rate": 0.0002999599429727786,
+      "loss": 4.7063,
+      "step": 75910
+    },
+    {
+      "epoch": 0.10667472253788045,
+      "grad_norm": 0.733707845211029,
+      "learning_rate": 0.00029995977277863544,
+      "loss": 4.7957,
+      "step": 75920
+    },
+    {
+      "epoch": 0.10668877347604402,
+      "grad_norm": 0.7222506403923035,
+      "learning_rate": 0.0002999596022237467,
+      "loss": 4.6025,
+      "step": 75930
+    },
+    {
+      "epoch": 0.1067028244142076,
+      "grad_norm": 0.7260400056838989,
+      "learning_rate": 0.0002999594313081128,
+      "loss": 4.7407,
+      "step": 75940
+    },
+    {
+      "epoch": 0.10671687535237119,
+      "grad_norm": 0.7056910395622253,
+      "learning_rate": 0.0002999592600317342,
+      "loss": 4.8546,
+      "step": 75950
+    },
+    {
+      "epoch": 0.10673092629053477,
+      "grad_norm": 0.7579191327095032,
+      "learning_rate": 0.0002999590883946112,
+      "loss": 4.6756,
+      "step": 75960
+    },
+    {
+      "epoch": 0.10674497722869834,
+      "grad_norm": 0.7058764100074768,
+      "learning_rate": 0.00029995891639674437,
+      "loss": 4.7785,
+      "step": 75970
+    },
+    {
+      "epoch": 0.10675902816686192,
+      "grad_norm": 0.7291330099105835,
+      "learning_rate": 0.000299958744038134,
+      "loss": 4.7122,
+      "step": 75980
+    },
+    {
+      "epoch": 0.1067730791050255,
+      "grad_norm": 0.7083228230476379,
+      "learning_rate": 0.00029995857131878055,
+      "loss": 4.8688,
+      "step": 75990
+    },
+    {
+      "epoch": 0.10678713004318907,
+      "grad_norm": 0.7070737481117249,
+      "learning_rate": 0.00029995839823868445,
+      "loss": 4.7306,
+      "step": 76000
+    },
+    {
+      "epoch": 0.10680118098135265,
+      "grad_norm": 0.7116759419441223,
+      "learning_rate": 0.00029995822479784605,
+      "loss": 4.6925,
+      "step": 76010
+    },
+    {
+      "epoch": 0.10681523191951622,
+      "grad_norm": 0.7482094764709473,
+      "learning_rate": 0.0002999580509962658,
+      "loss": 4.6478,
+      "step": 76020
+    },
+    {
+      "epoch": 0.1068292828576798,
+      "grad_norm": 0.7246078252792358,
+      "learning_rate": 0.00029995787683394427,
+      "loss": 4.7802,
+      "step": 76030
+    },
+    {
+      "epoch": 0.10684333379584338,
+      "grad_norm": 0.7256546020507812,
+      "learning_rate": 0.0002999577023108816,
+      "loss": 4.7196,
+      "step": 76040
+    },
+    {
+      "epoch": 0.10685738473400695,
+      "grad_norm": 0.7039734721183777,
+      "learning_rate": 0.0002999575274270784,
+      "loss": 4.7635,
+      "step": 76050
+    },
+    {
+      "epoch": 0.10687143567217054,
+      "grad_norm": 0.6815078258514404,
+      "learning_rate": 0.0002999573521825351,
+      "loss": 4.8917,
+      "step": 76060
+    },
+    {
+      "epoch": 0.10688548661033412,
+      "grad_norm": 0.7078894972801208,
+      "learning_rate": 0.00029995717657725195,
+      "loss": 4.7096,
+      "step": 76070
+    },
+    {
+      "epoch": 0.1068995375484977,
+      "grad_norm": 0.7273938655853271,
+      "learning_rate": 0.0002999570006112296,
+      "loss": 4.644,
+      "step": 76080
+    },
+    {
+      "epoch": 0.10691358848666127,
+      "grad_norm": 0.7119658589363098,
+      "learning_rate": 0.0002999568242844683,
+      "loss": 4.788,
+      "step": 76090
+    },
+    {
+      "epoch": 0.10692763942482485,
+      "grad_norm": 0.7011669874191284,
+      "learning_rate": 0.00029995664759696855,
+      "loss": 4.5298,
+      "step": 76100
+    },
+    {
+      "epoch": 0.10694169036298842,
+      "grad_norm": 0.7396077513694763,
+      "learning_rate": 0.0002999564705487307,
+      "loss": 4.6668,
+      "step": 76110
+    },
+    {
+      "epoch": 0.106955741301152,
+      "grad_norm": 0.7068682312965393,
+      "learning_rate": 0.00029995629313975535,
+      "loss": 4.7548,
+      "step": 76120
+    },
+    {
+      "epoch": 0.10696979223931558,
+      "grad_norm": 0.7131781578063965,
+      "learning_rate": 0.0002999561153700427,
+      "loss": 4.8285,
+      "step": 76130
+    },
+    {
+      "epoch": 0.10698384317747915,
+      "grad_norm": 0.7171995639801025,
+      "learning_rate": 0.00029995593723959336,
+      "loss": 4.6529,
+      "step": 76140
+    },
+    {
+      "epoch": 0.10699789411564273,
+      "grad_norm": 0.6993269920349121,
+      "learning_rate": 0.0002999557587484077,
+      "loss": 4.6662,
+      "step": 76150
+    },
+    {
+      "epoch": 0.10701194505380632,
+      "grad_norm": 0.7163041234016418,
+      "learning_rate": 0.0002999555798964861,
+      "loss": 4.8221,
+      "step": 76160
+    },
+    {
+      "epoch": 0.1070259959919699,
+      "grad_norm": 0.6855682134628296,
+      "learning_rate": 0.0002999554006838291,
+      "loss": 4.7832,
+      "step": 76170
+    },
+    {
+      "epoch": 0.10704004693013347,
+      "grad_norm": 0.7540141940116882,
+      "learning_rate": 0.00029995522111043695,
+      "loss": 4.7666,
+      "step": 76180
+    },
+    {
+      "epoch": 0.10705409786829705,
+      "grad_norm": 0.7331625819206238,
+      "learning_rate": 0.0002999550411763102,
+      "loss": 4.7275,
+      "step": 76190
+    },
+    {
+      "epoch": 0.10706814880646062,
+      "grad_norm": 0.7579591870307922,
+      "learning_rate": 0.0002999548608814493,
+      "loss": 4.6774,
+      "step": 76200
+    },
+    {
+      "epoch": 0.1070821997446242,
+      "grad_norm": 0.7343604564666748,
+      "learning_rate": 0.0002999546802258547,
+      "loss": 4.716,
+      "step": 76210
+    },
+    {
+      "epoch": 0.10709625068278777,
+      "grad_norm": 0.7140637040138245,
+      "learning_rate": 0.00029995449920952675,
+      "loss": 4.6961,
+      "step": 76220
+    },
+    {
+      "epoch": 0.10711030162095135,
+      "grad_norm": 0.7120891213417053,
+      "learning_rate": 0.000299954317832466,
+      "loss": 4.8529,
+      "step": 76230
+    },
+    {
+      "epoch": 0.10712435255911493,
+      "grad_norm": 0.7201858758926392,
+      "learning_rate": 0.0002999541360946727,
+      "loss": 4.7678,
+      "step": 76240
+    },
+    {
+      "epoch": 0.1071384034972785,
+      "grad_norm": 0.7303602695465088,
+      "learning_rate": 0.0002999539539961475,
+      "loss": 4.8727,
+      "step": 76250
+    },
+    {
+      "epoch": 0.1071524544354421,
+      "grad_norm": 0.7096266150474548,
+      "learning_rate": 0.00029995377153689063,
+      "loss": 4.7656,
+      "step": 76260
+    },
+    {
+      "epoch": 0.10716650537360567,
+      "grad_norm": 0.6913262009620667,
+      "learning_rate": 0.00029995358871690275,
+      "loss": 4.7481,
+      "step": 76270
+    },
+    {
+      "epoch": 0.10718055631176925,
+      "grad_norm": 0.7402442693710327,
+      "learning_rate": 0.0002999534055361841,
+      "loss": 4.7493,
+      "step": 76280
+    },
+    {
+      "epoch": 0.10719460724993282,
+      "grad_norm": 0.7259013652801514,
+      "learning_rate": 0.00029995322199473525,
+      "loss": 4.7733,
+      "step": 76290
+    },
+    {
+      "epoch": 0.1072086581880964,
+      "grad_norm": 0.7574719786643982,
+      "learning_rate": 0.0002999530380925566,
+      "loss": 4.65,
+      "step": 76300
+    },
+    {
+      "epoch": 0.10722270912625997,
+      "grad_norm": 0.7451621294021606,
+      "learning_rate": 0.00029995285382964856,
+      "loss": 4.6954,
+      "step": 76310
+    },
+    {
+      "epoch": 0.10723676006442355,
+      "grad_norm": 0.704237163066864,
+      "learning_rate": 0.00029995266920601165,
+      "loss": 4.7913,
+      "step": 76320
+    },
+    {
+      "epoch": 0.10725081100258713,
+      "grad_norm": 0.7237526774406433,
+      "learning_rate": 0.0002999524842216463,
+      "loss": 4.7553,
+      "step": 76330
+    },
+    {
+      "epoch": 0.1072648619407507,
+      "grad_norm": 0.724061131477356,
+      "learning_rate": 0.0002999522988765529,
+      "loss": 4.7573,
+      "step": 76340
+    },
+    {
+      "epoch": 0.10727891287891428,
+      "grad_norm": 0.6993366479873657,
+      "learning_rate": 0.0002999521131707318,
+      "loss": 4.6088,
+      "step": 76350
+    },
+    {
+      "epoch": 0.10729296381707785,
+      "grad_norm": 0.6915683746337891,
+      "learning_rate": 0.00029995192710418375,
+      "loss": 4.8146,
+      "step": 76360
+    },
+    {
+      "epoch": 0.10730701475524144,
+      "grad_norm": 0.7251079678535461,
+      "learning_rate": 0.0002999517406769089,
+      "loss": 4.6804,
+      "step": 76370
+    },
+    {
+      "epoch": 0.10732106569340502,
+      "grad_norm": 0.7160983681678772,
+      "learning_rate": 0.00029995155388890785,
+      "loss": 4.8018,
+      "step": 76380
+    },
+    {
+      "epoch": 0.1073351166315686,
+      "grad_norm": 0.6960489153862,
+      "learning_rate": 0.00029995136674018105,
+      "loss": 4.6292,
+      "step": 76390
+    },
+    {
+      "epoch": 0.10734916756973217,
+      "grad_norm": 0.7182583808898926,
+      "learning_rate": 0.00029995117923072885,
+      "loss": 4.7744,
+      "step": 76400
+    },
+    {
+      "epoch": 0.10736321850789575,
+      "grad_norm": 0.7371787428855896,
+      "learning_rate": 0.0002999509913605518,
+      "loss": 4.7254,
+      "step": 76410
+    },
+    {
+      "epoch": 0.10737726944605933,
+      "grad_norm": 0.7134677171707153,
+      "learning_rate": 0.0002999508031296503,
+      "loss": 4.8213,
+      "step": 76420
+    },
+    {
+      "epoch": 0.1073913203842229,
+      "grad_norm": 0.7464840412139893,
+      "learning_rate": 0.00029995061453802485,
+      "loss": 4.722,
+      "step": 76430
+    },
+    {
+      "epoch": 0.10740537132238648,
+      "grad_norm": 0.7147547602653503,
+      "learning_rate": 0.0002999504255856758,
+      "loss": 4.8029,
+      "step": 76440
+    },
+    {
+      "epoch": 0.10741942226055005,
+      "grad_norm": 0.7210128307342529,
+      "learning_rate": 0.00029995023627260375,
+      "loss": 4.7076,
+      "step": 76450
+    },
+    {
+      "epoch": 0.10743347319871363,
+      "grad_norm": 0.6943991184234619,
+      "learning_rate": 0.00029995004659880907,
+      "loss": 4.7665,
+      "step": 76460
+    },
+    {
+      "epoch": 0.10744752413687722,
+      "grad_norm": 0.7254976034164429,
+      "learning_rate": 0.0002999498565642922,
+      "loss": 4.6883,
+      "step": 76470
+    },
+    {
+      "epoch": 0.1074615750750408,
+      "grad_norm": 0.733951210975647,
+      "learning_rate": 0.00029994966616905366,
+      "loss": 4.6997,
+      "step": 76480
+    },
+    {
+      "epoch": 0.10747562601320437,
+      "grad_norm": 0.7102512121200562,
+      "learning_rate": 0.00029994947541309386,
+      "loss": 4.6974,
+      "step": 76490
+    },
+    {
+      "epoch": 0.10748967695136795,
+      "grad_norm": 0.7103749513626099,
+      "learning_rate": 0.00029994928429641327,
+      "loss": 4.677,
+      "step": 76500
+    },
+    {
+      "epoch": 0.10750372788953153,
+      "grad_norm": 0.7255563735961914,
+      "learning_rate": 0.00029994909281901233,
+      "loss": 4.7501,
+      "step": 76510
+    },
+    {
+      "epoch": 0.1075177788276951,
+      "grad_norm": 0.7186533808708191,
+      "learning_rate": 0.00029994890098089153,
+      "loss": 4.6903,
+      "step": 76520
+    },
+    {
+      "epoch": 0.10753182976585868,
+      "grad_norm": 0.7188341617584229,
+      "learning_rate": 0.00029994870878205136,
+      "loss": 4.7485,
+      "step": 76530
+    },
+    {
+      "epoch": 0.10754588070402225,
+      "grad_norm": 0.7477817535400391,
+      "learning_rate": 0.00029994851622249224,
+      "loss": 4.7999,
+      "step": 76540
+    },
+    {
+      "epoch": 0.10755993164218583,
+      "grad_norm": 0.7210610508918762,
+      "learning_rate": 0.0002999483233022146,
+      "loss": 4.7705,
+      "step": 76550
+    },
+    {
+      "epoch": 0.1075739825803494,
+      "grad_norm": 0.7165090441703796,
+      "learning_rate": 0.000299948130021219,
+      "loss": 4.7389,
+      "step": 76560
+    },
+    {
+      "epoch": 0.107588033518513,
+      "grad_norm": 0.6880157589912415,
+      "learning_rate": 0.00029994793637950584,
+      "loss": 4.7931,
+      "step": 76570
+    },
+    {
+      "epoch": 0.10760208445667657,
+      "grad_norm": 0.719469428062439,
+      "learning_rate": 0.00029994774237707556,
+      "loss": 4.751,
+      "step": 76580
+    },
+    {
+      "epoch": 0.10761613539484015,
+      "grad_norm": 0.7153851389884949,
+      "learning_rate": 0.0002999475480139287,
+      "loss": 4.6964,
+      "step": 76590
+    },
+    {
+      "epoch": 0.10763018633300372,
+      "grad_norm": 0.7175468802452087,
+      "learning_rate": 0.00029994735329006567,
+      "loss": 4.6719,
+      "step": 76600
+    },
+    {
+      "epoch": 0.1076442372711673,
+      "grad_norm": 0.7065703868865967,
+      "learning_rate": 0.0002999471582054869,
+      "loss": 4.7861,
+      "step": 76610
+    },
+    {
+      "epoch": 0.10765828820933088,
+      "grad_norm": 0.7555844187736511,
+      "learning_rate": 0.00029994696276019307,
+      "loss": 4.7068,
+      "step": 76620
+    },
+    {
+      "epoch": 0.10767233914749445,
+      "grad_norm": 0.7342779040336609,
+      "learning_rate": 0.0002999467669541844,
+      "loss": 4.7435,
+      "step": 76630
+    },
+    {
+      "epoch": 0.10768639008565803,
+      "grad_norm": 0.7292844653129578,
+      "learning_rate": 0.00029994657078746145,
+      "loss": 4.6379,
+      "step": 76640
+    },
+    {
+      "epoch": 0.1077004410238216,
+      "grad_norm": 0.7343292832374573,
+      "learning_rate": 0.0002999463742600247,
+      "loss": 4.6573,
+      "step": 76650
+    },
+    {
+      "epoch": 0.10771449196198518,
+      "grad_norm": 0.7275290489196777,
+      "learning_rate": 0.0002999461773718746,
+      "loss": 4.7651,
+      "step": 76660
+    },
+    {
+      "epoch": 0.10772854290014876,
+      "grad_norm": 0.7052959203720093,
+      "learning_rate": 0.0002999459801230117,
+      "loss": 4.806,
+      "step": 76670
+    },
+    {
+      "epoch": 0.10774259383831235,
+      "grad_norm": 0.7254387140274048,
+      "learning_rate": 0.0002999457825134364,
+      "loss": 4.5679,
+      "step": 76680
+    },
+    {
+      "epoch": 0.10775664477647592,
+      "grad_norm": 0.7066879272460938,
+      "learning_rate": 0.0002999455845431492,
+      "loss": 4.6913,
+      "step": 76690
+    },
+    {
+      "epoch": 0.1077706957146395,
+      "grad_norm": 0.716200590133667,
+      "learning_rate": 0.0002999453862121506,
+      "loss": 4.567,
+      "step": 76700
+    },
+    {
+      "epoch": 0.10778474665280308,
+      "grad_norm": 0.7112560868263245,
+      "learning_rate": 0.00029994518752044104,
+      "loss": 4.7576,
+      "step": 76710
+    },
+    {
+      "epoch": 0.10779879759096665,
+      "grad_norm": 0.7039072513580322,
+      "learning_rate": 0.000299944988468021,
+      "loss": 4.6923,
+      "step": 76720
+    },
+    {
+      "epoch": 0.10781284852913023,
+      "grad_norm": 0.7240103483200073,
+      "learning_rate": 0.00029994478905489094,
+      "loss": 4.8887,
+      "step": 76730
+    },
+    {
+      "epoch": 0.1078268994672938,
+      "grad_norm": 0.6706738471984863,
+      "learning_rate": 0.0002999445892810514,
+      "loss": 4.848,
+      "step": 76740
+    },
+    {
+      "epoch": 0.10784095040545738,
+      "grad_norm": 0.7257563471794128,
+      "learning_rate": 0.0002999443891465028,
+      "loss": 4.7367,
+      "step": 76750
+    },
+    {
+      "epoch": 0.10785500134362096,
+      "grad_norm": 0.7115849256515503,
+      "learning_rate": 0.0002999441886512456,
+      "loss": 4.7701,
+      "step": 76760
+    },
+    {
+      "epoch": 0.10786905228178453,
+      "grad_norm": 0.7439100742340088,
+      "learning_rate": 0.00029994398779528047,
+      "loss": 4.7311,
+      "step": 76770
+    },
+    {
+      "epoch": 0.10788310321994812,
+      "grad_norm": 0.7145756483078003,
+      "learning_rate": 0.00029994378657860764,
+      "loss": 4.7226,
+      "step": 76780
+    },
+    {
+      "epoch": 0.1078971541581117,
+      "grad_norm": 0.7000148892402649,
+      "learning_rate": 0.00029994358500122774,
+      "loss": 4.8696,
+      "step": 76790
+    },
+    {
+      "epoch": 0.10791120509627528,
+      "grad_norm": 0.7268564105033875,
+      "learning_rate": 0.00029994338306314126,
+      "loss": 4.7751,
+      "step": 76800
+    },
+    {
+      "epoch": 0.10792525603443885,
+      "grad_norm": 0.6927092671394348,
+      "learning_rate": 0.00029994318076434864,
+      "loss": 4.6979,
+      "step": 76810
+    },
+    {
+      "epoch": 0.10793930697260243,
+      "grad_norm": 0.7400745749473572,
+      "learning_rate": 0.0002999429781048503,
+      "loss": 4.7819,
+      "step": 76820
+    },
+    {
+      "epoch": 0.107953357910766,
+      "grad_norm": 0.7177280783653259,
+      "learning_rate": 0.00029994277508464684,
+      "loss": 4.7256,
+      "step": 76830
+    },
+    {
+      "epoch": 0.10796740884892958,
+      "grad_norm": 0.7067402601242065,
+      "learning_rate": 0.00029994257170373875,
+      "loss": 4.8645,
+      "step": 76840
+    },
+    {
+      "epoch": 0.10798145978709316,
+      "grad_norm": 0.7211511731147766,
+      "learning_rate": 0.0002999423679621265,
+      "loss": 4.768,
+      "step": 76850
+    },
+    {
+      "epoch": 0.10799551072525673,
+      "grad_norm": 0.717394232749939,
+      "learning_rate": 0.0002999421638598105,
+      "loss": 4.7647,
+      "step": 76860
+    },
+    {
+      "epoch": 0.10800956166342031,
+      "grad_norm": 0.741280734539032,
+      "learning_rate": 0.0002999419593967913,
+      "loss": 4.6856,
+      "step": 76870
+    },
+    {
+      "epoch": 0.1080236126015839,
+      "grad_norm": 0.7437717318534851,
+      "learning_rate": 0.0002999417545730694,
+      "loss": 4.7239,
+      "step": 76880
+    },
+    {
+      "epoch": 0.10803766353974747,
+      "grad_norm": 0.7169762849807739,
+      "learning_rate": 0.00029994154938864533,
+      "loss": 4.7881,
+      "step": 76890
+    },
+    {
+      "epoch": 0.10805171447791105,
+      "grad_norm": 0.7102547287940979,
+      "learning_rate": 0.0002999413438435195,
+      "loss": 4.6919,
+      "step": 76900
+    },
+    {
+      "epoch": 0.10806576541607463,
+      "grad_norm": 0.7318692803382874,
+      "learning_rate": 0.00029994113793769243,
+      "loss": 4.8037,
+      "step": 76910
+    },
+    {
+      "epoch": 0.1080798163542382,
+      "grad_norm": 0.7535057663917542,
+      "learning_rate": 0.0002999409316711647,
+      "loss": 4.7232,
+      "step": 76920
+    },
+    {
+      "epoch": 0.10809386729240178,
+      "grad_norm": 0.7213916182518005,
+      "learning_rate": 0.0002999407250439367,
+      "loss": 4.6257,
+      "step": 76930
+    },
+    {
+      "epoch": 0.10810791823056536,
+      "grad_norm": 0.7251282930374146,
+      "learning_rate": 0.00029994051805600896,
+      "loss": 4.6825,
+      "step": 76940
+    },
+    {
+      "epoch": 0.10812196916872893,
+      "grad_norm": 0.8058227300643921,
+      "learning_rate": 0.0002999403107073819,
+      "loss": 4.6347,
+      "step": 76950
+    },
+    {
+      "epoch": 0.10813602010689251,
+      "grad_norm": 0.7237033843994141,
+      "learning_rate": 0.0002999401029980562,
+      "loss": 4.7505,
+      "step": 76960
+    },
+    {
+      "epoch": 0.10815007104505608,
+      "grad_norm": 0.7178425192832947,
+      "learning_rate": 0.0002999398949280323,
+      "loss": 4.7016,
+      "step": 76970
+    },
+    {
+      "epoch": 0.10816412198321966,
+      "grad_norm": 0.7172357439994812,
+      "learning_rate": 0.0002999396864973106,
+      "loss": 4.8938,
+      "step": 76980
+    },
+    {
+      "epoch": 0.10817817292138325,
+      "grad_norm": 0.7308405637741089,
+      "learning_rate": 0.0002999394777058916,
+      "loss": 4.7532,
+      "step": 76990
+    },
+    {
+      "epoch": 0.10819222385954683,
+      "grad_norm": 0.715327799320221,
+      "learning_rate": 0.00029993926855377594,
+      "loss": 4.7328,
+      "step": 77000
+    },
+    {
+      "epoch": 0.1082062747977104,
+      "grad_norm": 0.6993787288665771,
+      "learning_rate": 0.000299939059040964,
+      "loss": 4.7433,
+      "step": 77010
+    },
+    {
+      "epoch": 0.10822032573587398,
+      "grad_norm": 0.7134409546852112,
+      "learning_rate": 0.0002999388491674564,
+      "loss": 4.6656,
+      "step": 77020
+    },
+    {
+      "epoch": 0.10823437667403756,
+      "grad_norm": 0.7323780655860901,
+      "learning_rate": 0.00029993863893325355,
+      "loss": 4.7395,
+      "step": 77030
+    },
+    {
+      "epoch": 0.10824842761220113,
+      "grad_norm": 0.7384430170059204,
+      "learning_rate": 0.00029993844941407694,
+      "loss": 4.701,
+      "step": 77040
+    },
+    {
+      "epoch": 0.10826247855036471,
+      "grad_norm": 0.7190428972244263,
+      "learning_rate": 0.00029993823849455454,
+      "loss": 4.8667,
+      "step": 77050
+    },
+    {
+      "epoch": 0.10827652948852828,
+      "grad_norm": 0.7207445502281189,
+      "learning_rate": 0.0002999380272143384,
+      "loss": 4.8387,
+      "step": 77060
+    },
+    {
+      "epoch": 0.10829058042669186,
+      "grad_norm": 0.7112932205200195,
+      "learning_rate": 0.000299937815573429,
+      "loss": 4.7327,
+      "step": 77070
+    },
+    {
+      "epoch": 0.10830463136485544,
+      "grad_norm": 0.7273050546646118,
+      "learning_rate": 0.00029993760357182686,
+      "loss": 4.7299,
+      "step": 77080
+    },
+    {
+      "epoch": 0.10831868230301903,
+      "grad_norm": 0.7051961421966553,
+      "learning_rate": 0.0002999373912095325,
+      "loss": 4.7734,
+      "step": 77090
+    },
+    {
+      "epoch": 0.1083327332411826,
+      "grad_norm": 0.7308077216148376,
+      "learning_rate": 0.0002999371784865464,
+      "loss": 4.859,
+      "step": 77100
+    },
+    {
+      "epoch": 0.10834678417934618,
+      "grad_norm": 0.7368705868721008,
+      "learning_rate": 0.00029993696540286913,
+      "loss": 4.7365,
+      "step": 77110
+    },
+    {
+      "epoch": 0.10836083511750975,
+      "grad_norm": 0.7260712385177612,
+      "learning_rate": 0.00029993675195850116,
+      "loss": 4.822,
+      "step": 77120
+    },
+    {
+      "epoch": 0.10837488605567333,
+      "grad_norm": 0.7681806087493896,
+      "learning_rate": 0.000299936538153443,
+      "loss": 4.7356,
+      "step": 77130
+    },
+    {
+      "epoch": 0.1083889369938369,
+      "grad_norm": 0.7338926196098328,
+      "learning_rate": 0.0002999363239876952,
+      "loss": 4.8259,
+      "step": 77140
+    },
+    {
+      "epoch": 0.10840298793200048,
+      "grad_norm": 0.6925027370452881,
+      "learning_rate": 0.0002999361094612582,
+      "loss": 4.723,
+      "step": 77150
+    },
+    {
+      "epoch": 0.10841703887016406,
+      "grad_norm": 0.7058929800987244,
+      "learning_rate": 0.0002999358945741326,
+      "loss": 4.7304,
+      "step": 77160
+    },
+    {
+      "epoch": 0.10843108980832764,
+      "grad_norm": 0.7068637013435364,
+      "learning_rate": 0.00029993567932631886,
+      "loss": 4.9076,
+      "step": 77170
+    },
+    {
+      "epoch": 0.10844514074649121,
+      "grad_norm": 0.7397764921188354,
+      "learning_rate": 0.00029993546371781754,
+      "loss": 4.7557,
+      "step": 77180
+    },
+    {
+      "epoch": 0.1084591916846548,
+      "grad_norm": 0.7128493189811707,
+      "learning_rate": 0.0002999352477486291,
+      "loss": 4.7219,
+      "step": 77190
+    },
+    {
+      "epoch": 0.10847324262281838,
+      "grad_norm": 0.7162886261940002,
+      "learning_rate": 0.00029993503141875413,
+      "loss": 4.7888,
+      "step": 77200
+    },
+    {
+      "epoch": 0.10848729356098195,
+      "grad_norm": 0.7667913436889648,
+      "learning_rate": 0.00029993481472819313,
+      "loss": 4.8192,
+      "step": 77210
+    },
+    {
+      "epoch": 0.10850134449914553,
+      "grad_norm": 0.707908034324646,
+      "learning_rate": 0.0002999345976769466,
+      "loss": 4.7827,
+      "step": 77220
+    },
+    {
+      "epoch": 0.1085153954373091,
+      "grad_norm": 0.7176185250282288,
+      "learning_rate": 0.00029993438026501513,
+      "loss": 4.6943,
+      "step": 77230
+    },
+    {
+      "epoch": 0.10852944637547268,
+      "grad_norm": 0.7139223217964172,
+      "learning_rate": 0.0002999341624923991,
+      "loss": 4.7028,
+      "step": 77240
+    },
+    {
+      "epoch": 0.10854349731363626,
+      "grad_norm": 0.7057112455368042,
+      "learning_rate": 0.00029993394435909917,
+      "loss": 4.7016,
+      "step": 77250
+    },
+    {
+      "epoch": 0.10855754825179983,
+      "grad_norm": 0.7334666848182678,
+      "learning_rate": 0.0002999337258651158,
+      "loss": 4.7411,
+      "step": 77260
+    },
+    {
+      "epoch": 0.10857159918996341,
+      "grad_norm": 0.7039465308189392,
+      "learning_rate": 0.0002999335070104496,
+      "loss": 4.8021,
+      "step": 77270
+    },
+    {
+      "epoch": 0.10858565012812699,
+      "grad_norm": 0.716081976890564,
+      "learning_rate": 0.0002999332877951009,
+      "loss": 4.7026,
+      "step": 77280
+    },
+    {
+      "epoch": 0.10859970106629056,
+      "grad_norm": 0.725543737411499,
+      "learning_rate": 0.00029993306821907037,
+      "loss": 4.7908,
+      "step": 77290
+    },
+    {
+      "epoch": 0.10861375200445415,
+      "grad_norm": 0.7131525874137878,
+      "learning_rate": 0.0002999328482823586,
+      "loss": 4.7378,
+      "step": 77300
+    },
+    {
+      "epoch": 0.10862780294261773,
+      "grad_norm": 0.7272675037384033,
+      "learning_rate": 0.00029993262798496604,
+      "loss": 4.6323,
+      "step": 77310
+    },
+    {
+      "epoch": 0.1086418538807813,
+      "grad_norm": 0.8117364645004272,
+      "learning_rate": 0.0002999324073268931,
+      "loss": 4.7449,
+      "step": 77320
+    },
+    {
+      "epoch": 0.10865590481894488,
+      "grad_norm": 0.7216888666152954,
+      "learning_rate": 0.00029993218630814056,
+      "loss": 4.6925,
+      "step": 77330
+    },
+    {
+      "epoch": 0.10866995575710846,
+      "grad_norm": 0.7425556778907776,
+      "learning_rate": 0.0002999319649287088,
+      "loss": 4.8035,
+      "step": 77340
+    },
+    {
+      "epoch": 0.10868400669527203,
+      "grad_norm": 0.7141205668449402,
+      "learning_rate": 0.00029993174318859836,
+      "loss": 4.7453,
+      "step": 77350
+    },
+    {
+      "epoch": 0.10869805763343561,
+      "grad_norm": 0.7087002992630005,
+      "learning_rate": 0.0002999315210878098,
+      "loss": 4.6944,
+      "step": 77360
+    },
+    {
+      "epoch": 0.10871210857159919,
+      "grad_norm": 0.7184770107269287,
+      "learning_rate": 0.0002999312986263436,
+      "loss": 4.8394,
+      "step": 77370
+    },
+    {
+      "epoch": 0.10872615950976276,
+      "grad_norm": 0.7024494409561157,
+      "learning_rate": 0.0002999310758042004,
+      "loss": 4.7444,
+      "step": 77380
+    },
+    {
+      "epoch": 0.10874021044792634,
+      "grad_norm": 0.723019540309906,
+      "learning_rate": 0.0002999308526213806,
+      "loss": 4.7096,
+      "step": 77390
+    },
+    {
+      "epoch": 0.10875426138608993,
+      "grad_norm": 0.731758713722229,
+      "learning_rate": 0.0002999306290778849,
+      "loss": 4.7041,
+      "step": 77400
+    },
+    {
+      "epoch": 0.1087683123242535,
+      "grad_norm": 0.7360407710075378,
+      "learning_rate": 0.0002999304051737137,
+      "loss": 4.7408,
+      "step": 77410
+    },
+    {
+      "epoch": 0.10878236326241708,
+      "grad_norm": 0.9013959169387817,
+      "learning_rate": 0.00029993018090886765,
+      "loss": 4.8065,
+      "step": 77420
+    },
+    {
+      "epoch": 0.10879641420058066,
+      "grad_norm": 0.716956377029419,
+      "learning_rate": 0.00029992995628334715,
+      "loss": 4.7094,
+      "step": 77430
+    },
+    {
+      "epoch": 0.10881046513874423,
+      "grad_norm": 0.7028824090957642,
+      "learning_rate": 0.0002999297312971529,
+      "loss": 4.7542,
+      "step": 77440
+    },
+    {
+      "epoch": 0.10882451607690781,
+      "grad_norm": 0.7272639274597168,
+      "learning_rate": 0.00029992950595028524,
+      "loss": 4.7021,
+      "step": 77450
+    },
+    {
+      "epoch": 0.10883856701507139,
+      "grad_norm": 0.769668459892273,
+      "learning_rate": 0.0002999292802427449,
+      "loss": 4.7857,
+      "step": 77460
+    },
+    {
+      "epoch": 0.10885261795323496,
+      "grad_norm": 0.7413691878318787,
+      "learning_rate": 0.00029992905417453237,
+      "loss": 4.7665,
+      "step": 77470
+    },
+    {
+      "epoch": 0.10886666889139854,
+      "grad_norm": 0.7249032258987427,
+      "learning_rate": 0.0002999288277456482,
+      "loss": 4.6758,
+      "step": 77480
+    },
+    {
+      "epoch": 0.10888071982956211,
+      "grad_norm": 0.7224403023719788,
+      "learning_rate": 0.00029992860095609286,
+      "loss": 4.696,
+      "step": 77490
+    },
+    {
+      "epoch": 0.1088947707677257,
+      "grad_norm": 0.708949863910675,
+      "learning_rate": 0.00029992837380586694,
+      "loss": 4.7162,
+      "step": 77500
+    },
+    {
+      "epoch": 0.10890882170588928,
+      "grad_norm": 0.7245976328849792,
+      "learning_rate": 0.000299928146294971,
+      "loss": 4.6874,
+      "step": 77510
+    },
+    {
+      "epoch": 0.10892287264405286,
+      "grad_norm": 1.1022865772247314,
+      "learning_rate": 0.00029992791842340565,
+      "loss": 4.7513,
+      "step": 77520
+    },
+    {
+      "epoch": 0.10893692358221643,
+      "grad_norm": 0.7656620740890503,
+      "learning_rate": 0.0002999276901911713,
+      "loss": 4.6487,
+      "step": 77530
+    },
+    {
+      "epoch": 0.10895097452038001,
+      "grad_norm": 0.698890209197998,
+      "learning_rate": 0.00029992746159826863,
+      "loss": 4.7322,
+      "step": 77540
+    },
+    {
+      "epoch": 0.10896502545854359,
+      "grad_norm": 0.7267652153968811,
+      "learning_rate": 0.00029992723264469807,
+      "loss": 4.7124,
+      "step": 77550
+    },
+    {
+      "epoch": 0.10897907639670716,
+      "grad_norm": 0.7437712550163269,
+      "learning_rate": 0.0002999270033304603,
+      "loss": 4.7506,
+      "step": 77560
+    },
+    {
+      "epoch": 0.10899312733487074,
+      "grad_norm": 0.7425222396850586,
+      "learning_rate": 0.0002999267736555557,
+      "loss": 4.6336,
+      "step": 77570
+    },
+    {
+      "epoch": 0.10900717827303431,
+      "grad_norm": 0.735388994216919,
+      "learning_rate": 0.00029992654361998497,
+      "loss": 4.7389,
+      "step": 77580
+    },
+    {
+      "epoch": 0.10902122921119789,
+      "grad_norm": 0.7018815279006958,
+      "learning_rate": 0.00029992631322374864,
+      "loss": 4.7882,
+      "step": 77590
+    },
+    {
+      "epoch": 0.10903528014936147,
+      "grad_norm": 0.7219749689102173,
+      "learning_rate": 0.00029992608246684726,
+      "loss": 4.773,
+      "step": 77600
+    },
+    {
+      "epoch": 0.10904933108752506,
+      "grad_norm": 0.7270717024803162,
+      "learning_rate": 0.00029992585134928137,
+      "loss": 4.7005,
+      "step": 77610
+    },
+    {
+      "epoch": 0.10906338202568863,
+      "grad_norm": 0.6950241327285767,
+      "learning_rate": 0.00029992561987105145,
+      "loss": 4.7569,
+      "step": 77620
+    },
+    {
+      "epoch": 0.10907743296385221,
+      "grad_norm": 0.711341142654419,
+      "learning_rate": 0.0002999253880321582,
+      "loss": 4.7915,
+      "step": 77630
+    },
+    {
+      "epoch": 0.10909148390201578,
+      "grad_norm": 0.7145726084709167,
+      "learning_rate": 0.00029992515583260207,
+      "loss": 4.6679,
+      "step": 77640
+    },
+    {
+      "epoch": 0.10910553484017936,
+      "grad_norm": 0.7017688155174255,
+      "learning_rate": 0.0002999249232723837,
+      "loss": 4.897,
+      "step": 77650
+    },
+    {
+      "epoch": 0.10911958577834294,
+      "grad_norm": 0.7016028165817261,
+      "learning_rate": 0.0002999246903515035,
+      "loss": 4.7972,
+      "step": 77660
+    },
+    {
+      "epoch": 0.10913363671650651,
+      "grad_norm": 0.739503026008606,
+      "learning_rate": 0.00029992445706996225,
+      "loss": 4.7323,
+      "step": 77670
+    },
+    {
+      "epoch": 0.10914768765467009,
+      "grad_norm": 0.7204867601394653,
+      "learning_rate": 0.00029992422342776037,
+      "loss": 4.7506,
+      "step": 77680
+    },
+    {
+      "epoch": 0.10916173859283367,
+      "grad_norm": 0.713930070400238,
+      "learning_rate": 0.0002999239894248984,
+      "loss": 4.8078,
+      "step": 77690
+    },
+    {
+      "epoch": 0.10917578953099724,
+      "grad_norm": 0.7046233415603638,
+      "learning_rate": 0.00029992375506137704,
+      "loss": 4.6674,
+      "step": 77700
+    },
+    {
+      "epoch": 0.10918984046916083,
+      "grad_norm": 0.731946587562561,
+      "learning_rate": 0.0002999235203371967,
+      "loss": 4.7776,
+      "step": 77710
+    },
+    {
+      "epoch": 0.10920389140732441,
+      "grad_norm": 0.7064023017883301,
+      "learning_rate": 0.00029992328525235804,
+      "loss": 4.749,
+      "step": 77720
+    },
+    {
+      "epoch": 0.10921794234548798,
+      "grad_norm": 0.693246066570282,
+      "learning_rate": 0.00029992304980686155,
+      "loss": 4.7499,
+      "step": 77730
+    },
+    {
+      "epoch": 0.10923199328365156,
+      "grad_norm": 0.7170222997665405,
+      "learning_rate": 0.0002999228140007079,
+      "loss": 4.6803,
+      "step": 77740
+    },
+    {
+      "epoch": 0.10924604422181514,
+      "grad_norm": 0.7952947616577148,
+      "learning_rate": 0.0002999225778338976,
+      "loss": 4.6546,
+      "step": 77750
+    },
+    {
+      "epoch": 0.10926009515997871,
+      "grad_norm": 0.7248690128326416,
+      "learning_rate": 0.0002999223413064312,
+      "loss": 4.6298,
+      "step": 77760
+    },
+    {
+      "epoch": 0.10927414609814229,
+      "grad_norm": 0.7593909502029419,
+      "learning_rate": 0.0002999221044183093,
+      "loss": 4.7174,
+      "step": 77770
+    },
+    {
+      "epoch": 0.10928819703630586,
+      "grad_norm": 0.7026767730712891,
+      "learning_rate": 0.0002999218671695324,
+      "loss": 4.5835,
+      "step": 77780
+    },
+    {
+      "epoch": 0.10930224797446944,
+      "grad_norm": 0.7110097408294678,
+      "learning_rate": 0.0002999216295601012,
+      "loss": 4.8037,
+      "step": 77790
+    },
+    {
+      "epoch": 0.10931629891263302,
+      "grad_norm": 0.7158304452896118,
+      "learning_rate": 0.00029992139159001614,
+      "loss": 4.7756,
+      "step": 77800
+    },
+    {
+      "epoch": 0.10933034985079661,
+      "grad_norm": 0.7022632360458374,
+      "learning_rate": 0.0002999211532592779,
+      "loss": 4.7069,
+      "step": 77810
+    },
+    {
+      "epoch": 0.10934440078896018,
+      "grad_norm": 0.7358798384666443,
+      "learning_rate": 0.000299920914567887,
+      "loss": 4.8487,
+      "step": 77820
+    },
+    {
+      "epoch": 0.10935845172712376,
+      "grad_norm": 0.721831738948822,
+      "learning_rate": 0.00029992067551584396,
+      "loss": 4.7852,
+      "step": 77830
+    },
+    {
+      "epoch": 0.10937250266528734,
+      "grad_norm": 0.7131444811820984,
+      "learning_rate": 0.0002999204361031495,
+      "loss": 4.8763,
+      "step": 77840
+    },
+    {
+      "epoch": 0.10938655360345091,
+      "grad_norm": 0.7156862616539001,
+      "learning_rate": 0.00029992019632980405,
+      "loss": 4.7805,
+      "step": 77850
+    },
+    {
+      "epoch": 0.10940060454161449,
+      "grad_norm": 0.7103162407875061,
+      "learning_rate": 0.00029991995619580824,
+      "loss": 4.7374,
+      "step": 77860
+    },
+    {
+      "epoch": 0.10941465547977806,
+      "grad_norm": 0.6913536190986633,
+      "learning_rate": 0.0002999197157011627,
+      "loss": 4.6408,
+      "step": 77870
+    },
+    {
+      "epoch": 0.10942870641794164,
+      "grad_norm": 0.7195391058921814,
+      "learning_rate": 0.0002999194748458679,
+      "loss": 4.8017,
+      "step": 77880
+    },
+    {
+      "epoch": 0.10944275735610522,
+      "grad_norm": 0.7326173186302185,
+      "learning_rate": 0.0002999192336299246,
+      "loss": 4.729,
+      "step": 77890
+    },
+    {
+      "epoch": 0.10945680829426879,
+      "grad_norm": 0.721824049949646,
+      "learning_rate": 0.0002999189920533332,
+      "loss": 4.7534,
+      "step": 77900
+    },
+    {
+      "epoch": 0.10947085923243237,
+      "grad_norm": 0.719916582107544,
+      "learning_rate": 0.00029991875011609433,
+      "loss": 4.7614,
+      "step": 77910
+    },
+    {
+      "epoch": 0.10948491017059596,
+      "grad_norm": 0.7365716099739075,
+      "learning_rate": 0.0002999185078182086,
+      "loss": 4.6417,
+      "step": 77920
+    },
+    {
+      "epoch": 0.10949896110875953,
+      "grad_norm": 0.7198102474212646,
+      "learning_rate": 0.00029991826515967653,
+      "loss": 4.6395,
+      "step": 77930
+    },
+    {
+      "epoch": 0.10951301204692311,
+      "grad_norm": 0.7074626088142395,
+      "learning_rate": 0.0002999180221404988,
+      "loss": 4.8106,
+      "step": 77940
+    },
+    {
+      "epoch": 0.10952706298508669,
+      "grad_norm": 0.7036644220352173,
+      "learning_rate": 0.000299917778760676,
+      "loss": 4.6494,
+      "step": 77950
+    },
+    {
+      "epoch": 0.10954111392325026,
+      "grad_norm": 0.7054417729377747,
+      "learning_rate": 0.00029991753502020855,
+      "loss": 4.8741,
+      "step": 77960
+    },
+    {
+      "epoch": 0.10955516486141384,
+      "grad_norm": 0.7201747298240662,
+      "learning_rate": 0.0002999172909190972,
+      "loss": 4.8208,
+      "step": 77970
+    },
+    {
+      "epoch": 0.10956921579957742,
+      "grad_norm": 0.7417471408843994,
+      "learning_rate": 0.0002999170464573425,
+      "loss": 4.6939,
+      "step": 77980
+    },
+    {
+      "epoch": 0.10958326673774099,
+      "grad_norm": 0.7059158086776733,
+      "learning_rate": 0.000299916801634945,
+      "loss": 4.6488,
+      "step": 77990
+    },
+    {
+      "epoch": 0.10959731767590457,
+      "grad_norm": 0.7165995836257935,
+      "learning_rate": 0.00029991655645190534,
+      "loss": 4.6077,
+      "step": 78000
+    },
+    {
+      "epoch": 0.10961136861406814,
+      "grad_norm": 0.6998836398124695,
+      "learning_rate": 0.0002999163109082241,
+      "loss": 4.7637,
+      "step": 78010
+    },
+    {
+      "epoch": 0.10962541955223173,
+      "grad_norm": 0.7285990715026855,
+      "learning_rate": 0.00029991606500390184,
+      "loss": 4.7262,
+      "step": 78020
+    },
+    {
+      "epoch": 0.10963947049039531,
+      "grad_norm": 0.7066509127616882,
+      "learning_rate": 0.0002999158187389391,
+      "loss": 4.8026,
+      "step": 78030
+    },
+    {
+      "epoch": 0.10965352142855889,
+      "grad_norm": 0.7119672894477844,
+      "learning_rate": 0.0002999155721133366,
+      "loss": 4.7852,
+      "step": 78040
+    },
+    {
+      "epoch": 0.10966757236672246,
+      "grad_norm": 0.7337003946304321,
+      "learning_rate": 0.0002999153251270949,
+      "loss": 4.7541,
+      "step": 78050
+    },
+    {
+      "epoch": 0.10968162330488604,
+      "grad_norm": 0.7333500981330872,
+      "learning_rate": 0.0002999150777802145,
+      "loss": 4.8055,
+      "step": 78060
+    },
+    {
+      "epoch": 0.10969567424304962,
+      "grad_norm": 0.7334858775138855,
+      "learning_rate": 0.0002999148300726961,
+      "loss": 4.719,
+      "step": 78070
+    },
+    {
+      "epoch": 0.10970972518121319,
+      "grad_norm": 0.7448481917381287,
+      "learning_rate": 0.00029991458200454024,
+      "loss": 4.591,
+      "step": 78080
+    },
+    {
+      "epoch": 0.10972377611937677,
+      "grad_norm": 0.7082262635231018,
+      "learning_rate": 0.0002999143335757475,
+      "loss": 4.6573,
+      "step": 78090
+    },
+    {
+      "epoch": 0.10973782705754034,
+      "grad_norm": 0.752067506313324,
+      "learning_rate": 0.00029991408478631857,
+      "loss": 4.801,
+      "step": 78100
+    },
+    {
+      "epoch": 0.10975187799570392,
+      "grad_norm": 0.7344422936439514,
+      "learning_rate": 0.000299913835636254,
+      "loss": 4.6786,
+      "step": 78110
+    },
+    {
+      "epoch": 0.10976592893386751,
+      "grad_norm": 0.7434372901916504,
+      "learning_rate": 0.00029991358612555435,
+      "loss": 4.7506,
+      "step": 78120
+    },
+    {
+      "epoch": 0.10977997987203109,
+      "grad_norm": 0.7248328924179077,
+      "learning_rate": 0.00029991333625422023,
+      "loss": 4.766,
+      "step": 78130
+    },
+    {
+      "epoch": 0.10979403081019466,
+      "grad_norm": 0.7229834794998169,
+      "learning_rate": 0.00029991308602225227,
+      "loss": 4.7095,
+      "step": 78140
+    },
+    {
+      "epoch": 0.10980808174835824,
+      "grad_norm": 0.718464732170105,
+      "learning_rate": 0.0002999128354296511,
+      "loss": 4.731,
+      "step": 78150
+    },
+    {
+      "epoch": 0.10982213268652181,
+      "grad_norm": 0.7103954553604126,
+      "learning_rate": 0.00029991258447641724,
+      "loss": 4.7028,
+      "step": 78160
+    },
+    {
+      "epoch": 0.10983618362468539,
+      "grad_norm": 0.7132537961006165,
+      "learning_rate": 0.00029991233316255136,
+      "loss": 4.5993,
+      "step": 78170
+    },
+    {
+      "epoch": 0.10985023456284897,
+      "grad_norm": 0.7273306846618652,
+      "learning_rate": 0.00029991208148805404,
+      "loss": 4.7597,
+      "step": 78180
+    },
+    {
+      "epoch": 0.10986428550101254,
+      "grad_norm": 0.7169373035430908,
+      "learning_rate": 0.00029991182945292587,
+      "loss": 4.821,
+      "step": 78190
+    },
+    {
+      "epoch": 0.10987833643917612,
+      "grad_norm": 0.7016663551330566,
+      "learning_rate": 0.0002999115770571675,
+      "loss": 4.7387,
+      "step": 78200
+    },
+    {
+      "epoch": 0.1098923873773397,
+      "grad_norm": 0.7153110504150391,
+      "learning_rate": 0.00029991132430077946,
+      "loss": 4.6402,
+      "step": 78210
+    },
+    {
+      "epoch": 0.10990643831550327,
+      "grad_norm": 0.7164067625999451,
+      "learning_rate": 0.0002999110711837625,
+      "loss": 4.7666,
+      "step": 78220
+    },
+    {
+      "epoch": 0.10992048925366686,
+      "grad_norm": 0.7064258456230164,
+      "learning_rate": 0.0002999108177061171,
+      "loss": 4.7787,
+      "step": 78230
+    },
+    {
+      "epoch": 0.10993454019183044,
+      "grad_norm": 0.7285356521606445,
+      "learning_rate": 0.0002999105638678439,
+      "loss": 4.8512,
+      "step": 78240
+    },
+    {
+      "epoch": 0.10994859112999401,
+      "grad_norm": 0.7336903810501099,
+      "learning_rate": 0.00029991030966894345,
+      "loss": 4.7763,
+      "step": 78250
+    },
+    {
+      "epoch": 0.10996264206815759,
+      "grad_norm": 0.7162684798240662,
+      "learning_rate": 0.0002999100551094165,
+      "loss": 4.755,
+      "step": 78260
+    },
+    {
+      "epoch": 0.10997669300632117,
+      "grad_norm": 0.7184701561927795,
+      "learning_rate": 0.0002999098001892636,
+      "loss": 4.7169,
+      "step": 78270
+    },
+    {
+      "epoch": 0.10999074394448474,
+      "grad_norm": 0.7117341756820679,
+      "learning_rate": 0.00029990954490848536,
+      "loss": 4.7958,
+      "step": 78280
+    },
+    {
+      "epoch": 0.11000479488264832,
+      "grad_norm": 0.7427325248718262,
+      "learning_rate": 0.00029990928926708235,
+      "loss": 4.7711,
+      "step": 78290
+    },
+    {
+      "epoch": 0.1100188458208119,
+      "grad_norm": 0.7277485728263855,
+      "learning_rate": 0.00029990903326505527,
+      "loss": 4.7021,
+      "step": 78300
+    },
+    {
+      "epoch": 0.11003289675897547,
+      "grad_norm": 0.7296305894851685,
+      "learning_rate": 0.0002999087769024047,
+      "loss": 4.73,
+      "step": 78310
+    },
+    {
+      "epoch": 0.11004694769713905,
+      "grad_norm": 0.7299262285232544,
+      "learning_rate": 0.0002999085201791312,
+      "loss": 4.8023,
+      "step": 78320
+    },
+    {
+      "epoch": 0.11006099863530264,
+      "grad_norm": 0.7146306037902832,
+      "learning_rate": 0.0002999082630952354,
+      "loss": 4.6505,
+      "step": 78330
+    },
+    {
+      "epoch": 0.11007504957346621,
+      "grad_norm": 0.7327890396118164,
+      "learning_rate": 0.00029990800565071806,
+      "loss": 4.7601,
+      "step": 78340
+    },
+    {
+      "epoch": 0.11008910051162979,
+      "grad_norm": 0.716513454914093,
+      "learning_rate": 0.00029990774784557965,
+      "loss": 4.7456,
+      "step": 78350
+    },
+    {
+      "epoch": 0.11010315144979337,
+      "grad_norm": 0.7176522612571716,
+      "learning_rate": 0.0002999074896798208,
+      "loss": 4.856,
+      "step": 78360
+    },
+    {
+      "epoch": 0.11011720238795694,
+      "grad_norm": 0.7281580567359924,
+      "learning_rate": 0.0002999072311534422,
+      "loss": 4.7127,
+      "step": 78370
+    },
+    {
+      "epoch": 0.11013125332612052,
+      "grad_norm": 0.7641294002532959,
+      "learning_rate": 0.00029990697226644445,
+      "loss": 4.6719,
+      "step": 78380
+    },
+    {
+      "epoch": 0.1101453042642841,
+      "grad_norm": 0.7364638447761536,
+      "learning_rate": 0.00029990671301882813,
+      "loss": 4.672,
+      "step": 78390
+    },
+    {
+      "epoch": 0.11015935520244767,
+      "grad_norm": 0.7401412725448608,
+      "learning_rate": 0.0002999064534105939,
+      "loss": 4.8599,
+      "step": 78400
+    },
+    {
+      "epoch": 0.11017340614061125,
+      "grad_norm": 0.7288405895233154,
+      "learning_rate": 0.0002999061934417424,
+      "loss": 4.7307,
+      "step": 78410
+    },
+    {
+      "epoch": 0.11018745707877482,
+      "grad_norm": 0.7121359705924988,
+      "learning_rate": 0.0002999059331122742,
+      "loss": 4.7776,
+      "step": 78420
+    },
+    {
+      "epoch": 0.11020150801693841,
+      "grad_norm": 0.6956414580345154,
+      "learning_rate": 0.00029990567242219,
+      "loss": 4.6886,
+      "step": 78430
+    },
+    {
+      "epoch": 0.11021555895510199,
+      "grad_norm": 0.7270907163619995,
+      "learning_rate": 0.00029990541137149036,
+      "loss": 4.6126,
+      "step": 78440
+    },
+    {
+      "epoch": 0.11022960989326556,
+      "grad_norm": 0.731920063495636,
+      "learning_rate": 0.0002999051499601759,
+      "loss": 4.7614,
+      "step": 78450
+    },
+    {
+      "epoch": 0.11024366083142914,
+      "grad_norm": 0.7504376769065857,
+      "learning_rate": 0.0002999048881882473,
+      "loss": 4.6018,
+      "step": 78460
+    },
+    {
+      "epoch": 0.11025771176959272,
+      "grad_norm": 0.7335795164108276,
+      "learning_rate": 0.0002999046260557052,
+      "loss": 4.6259,
+      "step": 78470
+    },
+    {
+      "epoch": 0.1102717627077563,
+      "grad_norm": 0.7405069470405579,
+      "learning_rate": 0.0002999043635625502,
+      "loss": 4.7949,
+      "step": 78480
+    },
+    {
+      "epoch": 0.11028581364591987,
+      "grad_norm": 0.7165108323097229,
+      "learning_rate": 0.00029990410070878293,
+      "loss": 4.8312,
+      "step": 78490
+    },
+    {
+      "epoch": 0.11029986458408345,
+      "grad_norm": 0.7114459872245789,
+      "learning_rate": 0.00029990383749440403,
+      "loss": 4.6969,
+      "step": 78500
+    },
+    {
+      "epoch": 0.11031391552224702,
+      "grad_norm": 0.7313840389251709,
+      "learning_rate": 0.00029990357391941413,
+      "loss": 4.7718,
+      "step": 78510
+    },
+    {
+      "epoch": 0.1103279664604106,
+      "grad_norm": 0.7081531882286072,
+      "learning_rate": 0.0002999033099838138,
+      "loss": 4.7343,
+      "step": 78520
+    },
+    {
+      "epoch": 0.11034201739857417,
+      "grad_norm": 0.7249537706375122,
+      "learning_rate": 0.00029990304568760384,
+      "loss": 4.7134,
+      "step": 78530
+    },
+    {
+      "epoch": 0.11035606833673776,
+      "grad_norm": 0.7393092513084412,
+      "learning_rate": 0.00029990278103078475,
+      "loss": 4.7877,
+      "step": 78540
+    },
+    {
+      "epoch": 0.11037011927490134,
+      "grad_norm": 0.7348365783691406,
+      "learning_rate": 0.00029990251601335715,
+      "loss": 4.693,
+      "step": 78550
+    },
+    {
+      "epoch": 0.11038417021306492,
+      "grad_norm": 0.7228670120239258,
+      "learning_rate": 0.00029990225063532185,
+      "loss": 4.7112,
+      "step": 78560
+    },
+    {
+      "epoch": 0.11039822115122849,
+      "grad_norm": 0.7742883563041687,
+      "learning_rate": 0.0002999019848966793,
+      "loss": 4.7404,
+      "step": 78570
+    },
+    {
+      "epoch": 0.11041227208939207,
+      "grad_norm": 0.7462913393974304,
+      "learning_rate": 0.00029990171879743015,
+      "loss": 4.699,
+      "step": 78580
+    },
+    {
+      "epoch": 0.11042632302755565,
+      "grad_norm": 0.7195731997489929,
+      "learning_rate": 0.00029990145233757515,
+      "loss": 4.745,
+      "step": 78590
+    },
+    {
+      "epoch": 0.11044037396571922,
+      "grad_norm": 0.7176849246025085,
+      "learning_rate": 0.00029990118551711484,
+      "loss": 4.7519,
+      "step": 78600
+    },
+    {
+      "epoch": 0.1104544249038828,
+      "grad_norm": 0.7326333522796631,
+      "learning_rate": 0.00029990091833604997,
+      "loss": 4.6879,
+      "step": 78610
+    },
+    {
+      "epoch": 0.11046847584204637,
+      "grad_norm": 0.7096735239028931,
+      "learning_rate": 0.00029990065079438114,
+      "loss": 4.7295,
+      "step": 78620
+    },
+    {
+      "epoch": 0.11048252678020995,
+      "grad_norm": 0.7012742757797241,
+      "learning_rate": 0.00029990038289210894,
+      "loss": 4.5788,
+      "step": 78630
+    },
+    {
+      "epoch": 0.11049657771837354,
+      "grad_norm": 0.7373076677322388,
+      "learning_rate": 0.00029990011462923403,
+      "loss": 4.7283,
+      "step": 78640
+    },
+    {
+      "epoch": 0.11051062865653712,
+      "grad_norm": 0.7383387684822083,
+      "learning_rate": 0.0002998998460057571,
+      "loss": 4.718,
+      "step": 78650
+    },
+    {
+      "epoch": 0.11052467959470069,
+      "grad_norm": 0.7086196541786194,
+      "learning_rate": 0.00029989957702167877,
+      "loss": 4.7736,
+      "step": 78660
+    },
+    {
+      "epoch": 0.11053873053286427,
+      "grad_norm": 0.7198120951652527,
+      "learning_rate": 0.0002998993076769997,
+      "loss": 4.7757,
+      "step": 78670
+    },
+    {
+      "epoch": 0.11055278147102784,
+      "grad_norm": 0.7587843537330627,
+      "learning_rate": 0.00029989903797172056,
+      "loss": 4.801,
+      "step": 78680
+    },
+    {
+      "epoch": 0.11056683240919142,
+      "grad_norm": 0.7173702716827393,
+      "learning_rate": 0.0002998987679058419,
+      "loss": 4.7577,
+      "step": 78690
+    },
+    {
+      "epoch": 0.110580883347355,
+      "grad_norm": 0.7113364934921265,
+      "learning_rate": 0.0002998984974793645,
+      "loss": 4.7938,
+      "step": 78700
+    },
+    {
+      "epoch": 0.11059493428551857,
+      "grad_norm": 0.7075303196907043,
+      "learning_rate": 0.00029989822669228885,
+      "loss": 4.8154,
+      "step": 78710
+    },
+    {
+      "epoch": 0.11060898522368215,
+      "grad_norm": 0.749019205570221,
+      "learning_rate": 0.00029989795554461574,
+      "loss": 4.6428,
+      "step": 78720
+    },
+    {
+      "epoch": 0.11062303616184573,
+      "grad_norm": 0.7221829295158386,
+      "learning_rate": 0.00029989768403634583,
+      "loss": 4.7743,
+      "step": 78730
+    },
+    {
+      "epoch": 0.11063708710000932,
+      "grad_norm": 0.7469913959503174,
+      "learning_rate": 0.0002998974121674797,
+      "loss": 4.7479,
+      "step": 78740
+    },
+    {
+      "epoch": 0.11065113803817289,
+      "grad_norm": 0.7294034361839294,
+      "learning_rate": 0.000299897139938018,
+      "loss": 4.6667,
+      "step": 78750
+    },
+    {
+      "epoch": 0.11066518897633647,
+      "grad_norm": 0.7280092239379883,
+      "learning_rate": 0.00029989686734796143,
+      "loss": 4.8348,
+      "step": 78760
+    },
+    {
+      "epoch": 0.11067923991450004,
+      "grad_norm": 0.7117518782615662,
+      "learning_rate": 0.0002998965943973106,
+      "loss": 4.7427,
+      "step": 78770
+    },
+    {
+      "epoch": 0.11069329085266362,
+      "grad_norm": 0.7216711640357971,
+      "learning_rate": 0.0002998963210860662,
+      "loss": 4.6675,
+      "step": 78780
+    },
+    {
+      "epoch": 0.1107073417908272,
+      "grad_norm": 0.7047522664070129,
+      "learning_rate": 0.00029989604741422897,
+      "loss": 4.7768,
+      "step": 78790
+    },
+    {
+      "epoch": 0.11072139272899077,
+      "grad_norm": 0.7834486961364746,
+      "learning_rate": 0.0002998957733817994,
+      "loss": 4.5577,
+      "step": 78800
+    },
+    {
+      "epoch": 0.11073544366715435,
+      "grad_norm": 0.6964229345321655,
+      "learning_rate": 0.00029989549898877825,
+      "loss": 4.7269,
+      "step": 78810
+    },
+    {
+      "epoch": 0.11074949460531792,
+      "grad_norm": 0.7209952473640442,
+      "learning_rate": 0.00029989522423516614,
+      "loss": 4.6683,
+      "step": 78820
+    },
+    {
+      "epoch": 0.1107635455434815,
+      "grad_norm": 0.7614384293556213,
+      "learning_rate": 0.0002998949491209638,
+      "loss": 4.6271,
+      "step": 78830
+    },
+    {
+      "epoch": 0.11077759648164508,
+      "grad_norm": 0.706362247467041,
+      "learning_rate": 0.00029989467364617176,
+      "loss": 4.7318,
+      "step": 78840
+    },
+    {
+      "epoch": 0.11079164741980867,
+      "grad_norm": 0.7691898941993713,
+      "learning_rate": 0.0002998943978107908,
+      "loss": 4.7161,
+      "step": 78850
+    },
+    {
+      "epoch": 0.11080569835797224,
+      "grad_norm": 0.6833986043930054,
+      "learning_rate": 0.00029989412161482155,
+      "loss": 4.7538,
+      "step": 78860
+    },
+    {
+      "epoch": 0.11081974929613582,
+      "grad_norm": 0.7353855967521667,
+      "learning_rate": 0.0002998938450582647,
+      "loss": 4.8051,
+      "step": 78870
+    },
+    {
+      "epoch": 0.1108338002342994,
+      "grad_norm": 0.7171785831451416,
+      "learning_rate": 0.00029989356814112087,
+      "loss": 4.7027,
+      "step": 78880
+    },
+    {
+      "epoch": 0.11084785117246297,
+      "grad_norm": 0.6899160146713257,
+      "learning_rate": 0.00029989329086339074,
+      "loss": 4.7299,
+      "step": 78890
+    },
+    {
+      "epoch": 0.11086190211062655,
+      "grad_norm": 0.6934748888015747,
+      "learning_rate": 0.000299893013225075,
+      "loss": 4.7226,
+      "step": 78900
+    },
+    {
+      "epoch": 0.11087595304879012,
+      "grad_norm": 0.7488577961921692,
+      "learning_rate": 0.0002998927352261743,
+      "loss": 4.7633,
+      "step": 78910
+    },
+    {
+      "epoch": 0.1108900039869537,
+      "grad_norm": 0.6993347406387329,
+      "learning_rate": 0.0002998924568666893,
+      "loss": 4.6352,
+      "step": 78920
+    },
+    {
+      "epoch": 0.11090405492511728,
+      "grad_norm": 0.7227916717529297,
+      "learning_rate": 0.00029989217814662067,
+      "loss": 4.7705,
+      "step": 78930
+    },
+    {
+      "epoch": 0.11091810586328085,
+      "grad_norm": 0.7275491952896118,
+      "learning_rate": 0.00029989189906596907,
+      "loss": 4.7141,
+      "step": 78940
+    },
+    {
+      "epoch": 0.11093215680144444,
+      "grad_norm": 0.7532659769058228,
+      "learning_rate": 0.0002998916196247352,
+      "loss": 4.6849,
+      "step": 78950
+    },
+    {
+      "epoch": 0.11094620773960802,
+      "grad_norm": 0.7284137606620789,
+      "learning_rate": 0.00029989133982291974,
+      "loss": 4.7539,
+      "step": 78960
+    },
+    {
+      "epoch": 0.1109602586777716,
+      "grad_norm": 0.7159876823425293,
+      "learning_rate": 0.0002998910596605233,
+      "loss": 4.7526,
+      "step": 78970
+    },
+    {
+      "epoch": 0.11097430961593517,
+      "grad_norm": 0.7106112837791443,
+      "learning_rate": 0.00029989077913754667,
+      "loss": 4.7353,
+      "step": 78980
+    },
+    {
+      "epoch": 0.11098836055409875,
+      "grad_norm": 0.7276374697685242,
+      "learning_rate": 0.0002998904982539904,
+      "loss": 4.7902,
+      "step": 78990
+    },
+    {
+      "epoch": 0.11100241149226232,
+      "grad_norm": 0.7346372604370117,
+      "learning_rate": 0.0002998902170098552,
+      "loss": 4.8002,
+      "step": 79000
+    },
+    {
+      "epoch": 0.1110164624304259,
+      "grad_norm": 0.7175330519676208,
+      "learning_rate": 0.00029988993540514187,
+      "loss": 4.7478,
+      "step": 79010
+    },
+    {
+      "epoch": 0.11103051336858948,
+      "grad_norm": 0.7902714014053345,
+      "learning_rate": 0.00029988965343985085,
+      "loss": 4.6789,
+      "step": 79020
+    },
+    {
+      "epoch": 0.11104456430675305,
+      "grad_norm": 0.7199444770812988,
+      "learning_rate": 0.00029988937111398304,
+      "loss": 4.7025,
+      "step": 79030
+    },
+    {
+      "epoch": 0.11105861524491663,
+      "grad_norm": 0.7127149105072021,
+      "learning_rate": 0.000299889088427539,
+      "loss": 4.6838,
+      "step": 79040
+    },
+    {
+      "epoch": 0.11107266618308022,
+      "grad_norm": 0.700092613697052,
+      "learning_rate": 0.0002998888053805195,
+      "loss": 4.7037,
+      "step": 79050
+    },
+    {
+      "epoch": 0.1110867171212438,
+      "grad_norm": 0.704186737537384,
+      "learning_rate": 0.0002998885219729251,
+      "loss": 4.9441,
+      "step": 79060
+    },
+    {
+      "epoch": 0.11110076805940737,
+      "grad_norm": 0.714832603931427,
+      "learning_rate": 0.00029988823820475654,
+      "loss": 4.8514,
+      "step": 79070
+    },
+    {
+      "epoch": 0.11111481899757095,
+      "grad_norm": 0.7268180251121521,
+      "learning_rate": 0.0002998879540760145,
+      "loss": 4.708,
+      "step": 79080
+    },
+    {
+      "epoch": 0.11112886993573452,
+      "grad_norm": 0.7234380841255188,
+      "learning_rate": 0.0002998876695866997,
+      "loss": 4.7114,
+      "step": 79090
+    },
+    {
+      "epoch": 0.1111429208738981,
+      "grad_norm": 0.7124701738357544,
+      "learning_rate": 0.0002998873847368128,
+      "loss": 4.7032,
+      "step": 79100
+    },
+    {
+      "epoch": 0.11115697181206168,
+      "grad_norm": 0.7144205570220947,
+      "learning_rate": 0.0002998870995263545,
+      "loss": 4.7765,
+      "step": 79110
+    },
+    {
+      "epoch": 0.11117102275022525,
+      "grad_norm": 0.7110472917556763,
+      "learning_rate": 0.00029988684252865396,
+      "loss": 4.7136,
+      "step": 79120
+    },
+    {
+      "epoch": 0.11118507368838883,
+      "grad_norm": 0.7378934025764465,
+      "learning_rate": 0.0002998865566331118,
+      "loss": 4.8435,
+      "step": 79130
+    },
+    {
+      "epoch": 0.1111991246265524,
+      "grad_norm": 0.7095884084701538,
+      "learning_rate": 0.00029988627037700027,
+      "loss": 4.8469,
+      "step": 79140
+    },
+    {
+      "epoch": 0.11121317556471598,
+      "grad_norm": 0.7001661062240601,
+      "learning_rate": 0.00029988598376031996,
+      "loss": 4.8903,
+      "step": 79150
+    },
+    {
+      "epoch": 0.11122722650287957,
+      "grad_norm": 0.7070997357368469,
+      "learning_rate": 0.00029988569678307157,
+      "loss": 4.8263,
+      "step": 79160
+    },
+    {
+      "epoch": 0.11124127744104315,
+      "grad_norm": 0.7273538112640381,
+      "learning_rate": 0.00029988540944525586,
+      "loss": 4.758,
+      "step": 79170
+    },
+    {
+      "epoch": 0.11125532837920672,
+      "grad_norm": 0.7323654890060425,
+      "learning_rate": 0.0002998851217468735,
+      "loss": 4.645,
+      "step": 79180
+    },
+    {
+      "epoch": 0.1112693793173703,
+      "grad_norm": 0.6991866230964661,
+      "learning_rate": 0.0002998848336879251,
+      "loss": 4.687,
+      "step": 79190
+    },
+    {
+      "epoch": 0.11128343025553387,
+      "grad_norm": 0.7104476094245911,
+      "learning_rate": 0.0002998845452684114,
+      "loss": 4.7952,
+      "step": 79200
+    },
+    {
+      "epoch": 0.11129748119369745,
+      "grad_norm": 0.7157678604125977,
+      "learning_rate": 0.0002998842564883332,
+      "loss": 4.7413,
+      "step": 79210
+    },
+    {
+      "epoch": 0.11131153213186103,
+      "grad_norm": 0.7549179196357727,
+      "learning_rate": 0.0002998839673476911,
+      "loss": 4.6288,
+      "step": 79220
+    },
+    {
+      "epoch": 0.1113255830700246,
+      "grad_norm": 0.7367541193962097,
+      "learning_rate": 0.00029988367784648573,
+      "loss": 4.8033,
+      "step": 79230
+    },
+    {
+      "epoch": 0.11133963400818818,
+      "grad_norm": 0.7407969832420349,
+      "learning_rate": 0.0002998833879847179,
+      "loss": 4.7142,
+      "step": 79240
+    },
+    {
+      "epoch": 0.11135368494635176,
+      "grad_norm": 0.7445090413093567,
+      "learning_rate": 0.0002998830977623883,
+      "loss": 4.6799,
+      "step": 79250
+    },
+    {
+      "epoch": 0.11136773588451535,
+      "grad_norm": 0.7357890605926514,
+      "learning_rate": 0.0002998828071794975,
+      "loss": 4.6795,
+      "step": 79260
+    },
+    {
+      "epoch": 0.11138178682267892,
+      "grad_norm": 0.7026675343513489,
+      "learning_rate": 0.00029988251623604633,
+      "loss": 4.7248,
+      "step": 79270
+    },
+    {
+      "epoch": 0.1113958377608425,
+      "grad_norm": 0.7261384129524231,
+      "learning_rate": 0.00029988222493203547,
+      "loss": 4.8116,
+      "step": 79280
+    },
+    {
+      "epoch": 0.11140988869900607,
+      "grad_norm": 0.7369160056114197,
+      "learning_rate": 0.0002998819332674656,
+      "loss": 4.6674,
+      "step": 79290
+    },
+    {
+      "epoch": 0.11142393963716965,
+      "grad_norm": 0.7252683043479919,
+      "learning_rate": 0.00029988164124233744,
+      "loss": 4.7722,
+      "step": 79300
+    },
+    {
+      "epoch": 0.11143799057533323,
+      "grad_norm": 0.694120466709137,
+      "learning_rate": 0.0002998813488566516,
+      "loss": 4.7246,
+      "step": 79310
+    },
+    {
+      "epoch": 0.1114520415134968,
+      "grad_norm": 0.7220812439918518,
+      "learning_rate": 0.0002998810561104089,
+      "loss": 4.7342,
+      "step": 79320
+    },
+    {
+      "epoch": 0.11146609245166038,
+      "grad_norm": 0.7195317149162292,
+      "learning_rate": 0.00029988076300361004,
+      "loss": 4.5972,
+      "step": 79330
+    },
+    {
+      "epoch": 0.11148014338982395,
+      "grad_norm": 0.7166000008583069,
+      "learning_rate": 0.00029988046953625567,
+      "loss": 4.5616,
+      "step": 79340
+    },
+    {
+      "epoch": 0.11149419432798753,
+      "grad_norm": 0.7502967715263367,
+      "learning_rate": 0.0002998801757083465,
+      "loss": 4.8734,
+      "step": 79350
+    },
+    {
+      "epoch": 0.11150824526615112,
+      "grad_norm": 0.7110496163368225,
+      "learning_rate": 0.00029987988151988326,
+      "loss": 4.7498,
+      "step": 79360
+    },
+    {
+      "epoch": 0.1115222962043147,
+      "grad_norm": 0.7079729437828064,
+      "learning_rate": 0.0002998795869708666,
+      "loss": 4.7823,
+      "step": 79370
+    },
+    {
+      "epoch": 0.11153634714247827,
+      "grad_norm": 0.7025529146194458,
+      "learning_rate": 0.00029987929206129737,
+      "loss": 4.7594,
+      "step": 79380
+    },
+    {
+      "epoch": 0.11155039808064185,
+      "grad_norm": 0.6984058022499084,
+      "learning_rate": 0.00029987899679117613,
+      "loss": 4.8166,
+      "step": 79390
+    },
+    {
+      "epoch": 0.11156444901880543,
+      "grad_norm": 0.7246695160865784,
+      "learning_rate": 0.0002998787011605037,
+      "loss": 4.7495,
+      "step": 79400
+    },
+    {
+      "epoch": 0.111578499956969,
+      "grad_norm": 0.7176758050918579,
+      "learning_rate": 0.0002998784051692807,
+      "loss": 4.661,
+      "step": 79410
+    },
+    {
+      "epoch": 0.11159255089513258,
+      "grad_norm": 0.719412624835968,
+      "learning_rate": 0.00029987810881750785,
+      "loss": 4.7986,
+      "step": 79420
+    },
+    {
+      "epoch": 0.11160660183329615,
+      "grad_norm": 0.7081352472305298,
+      "learning_rate": 0.00029987781210518593,
+      "loss": 4.6177,
+      "step": 79430
+    },
+    {
+      "epoch": 0.11162065277145973,
+      "grad_norm": 0.7022585272789001,
+      "learning_rate": 0.0002998775150323156,
+      "loss": 4.7622,
+      "step": 79440
+    },
+    {
+      "epoch": 0.1116347037096233,
+      "grad_norm": 0.7345691919326782,
+      "learning_rate": 0.0002998772175988976,
+      "loss": 4.7247,
+      "step": 79450
+    },
+    {
+      "epoch": 0.11164875464778688,
+      "grad_norm": 0.7096884250640869,
+      "learning_rate": 0.00029987691980493266,
+      "loss": 4.6959,
+      "step": 79460
+    },
+    {
+      "epoch": 0.11166280558595047,
+      "grad_norm": 0.7393691539764404,
+      "learning_rate": 0.00029987662165042143,
+      "loss": 4.6315,
+      "step": 79470
+    },
+    {
+      "epoch": 0.11167685652411405,
+      "grad_norm": 0.7320327758789062,
+      "learning_rate": 0.0002998763231353647,
+      "loss": 4.7236,
+      "step": 79480
+    },
+    {
+      "epoch": 0.11169090746227762,
+      "grad_norm": 0.730957567691803,
+      "learning_rate": 0.00029987602425976317,
+      "loss": 4.5586,
+      "step": 79490
+    },
+    {
+      "epoch": 0.1117049584004412,
+      "grad_norm": 0.7240452766418457,
+      "learning_rate": 0.00029987572502361754,
+      "loss": 4.7775,
+      "step": 79500
+    },
+    {
+      "epoch": 0.11171900933860478,
+      "grad_norm": 0.7169708013534546,
+      "learning_rate": 0.0002998754254269285,
+      "loss": 4.8488,
+      "step": 79510
+    },
+    {
+      "epoch": 0.11173306027676835,
+      "grad_norm": 0.7280612587928772,
+      "learning_rate": 0.00029987512546969687,
+      "loss": 4.6919,
+      "step": 79520
+    },
+    {
+      "epoch": 0.11174711121493193,
+      "grad_norm": 0.7089032530784607,
+      "learning_rate": 0.0002998748251519233,
+      "loss": 4.7049,
+      "step": 79530
+    },
+    {
+      "epoch": 0.1117611621530955,
+      "grad_norm": 0.7215970158576965,
+      "learning_rate": 0.0002998745244736085,
+      "loss": 4.8457,
+      "step": 79540
+    },
+    {
+      "epoch": 0.11177521309125908,
+      "grad_norm": 0.7137922048568726,
+      "learning_rate": 0.00029987422343475317,
+      "loss": 4.6256,
+      "step": 79550
+    },
+    {
+      "epoch": 0.11178926402942266,
+      "grad_norm": 0.768875241279602,
+      "learning_rate": 0.00029987392203535815,
+      "loss": 4.7273,
+      "step": 79560
+    },
+    {
+      "epoch": 0.11180331496758625,
+      "grad_norm": 0.7097663879394531,
+      "learning_rate": 0.0002998736202754241,
+      "loss": 4.7569,
+      "step": 79570
+    },
+    {
+      "epoch": 0.11181736590574982,
+      "grad_norm": 0.7040906548500061,
+      "learning_rate": 0.00029987331815495175,
+      "loss": 4.8235,
+      "step": 79580
+    },
+    {
+      "epoch": 0.1118314168439134,
+      "grad_norm": 0.734401285648346,
+      "learning_rate": 0.0002998730156739418,
+      "loss": 4.7113,
+      "step": 79590
+    },
+    {
+      "epoch": 0.11184546778207698,
+      "grad_norm": 0.7665032148361206,
+      "learning_rate": 0.0002998727128323949,
+      "loss": 4.7411,
+      "step": 79600
+    },
+    {
+      "epoch": 0.11185951872024055,
+      "grad_norm": 0.7201694846153259,
+      "learning_rate": 0.000299872409630312,
+      "loss": 4.5992,
+      "step": 79610
+    },
+    {
+      "epoch": 0.11187356965840413,
+      "grad_norm": 0.7197522521018982,
+      "learning_rate": 0.00029987210606769365,
+      "loss": 4.7049,
+      "step": 79620
+    },
+    {
+      "epoch": 0.1118876205965677,
+      "grad_norm": 0.734275221824646,
+      "learning_rate": 0.0002998718021445406,
+      "loss": 4.761,
+      "step": 79630
+    },
+    {
+      "epoch": 0.11190167153473128,
+      "grad_norm": 0.7246469259262085,
+      "learning_rate": 0.0002998714978608537,
+      "loss": 4.6607,
+      "step": 79640
+    },
+    {
+      "epoch": 0.11191572247289486,
+      "grad_norm": 0.7006386518478394,
+      "learning_rate": 0.00029987119321663356,
+      "loss": 4.6202,
+      "step": 79650
+    },
+    {
+      "epoch": 0.11192977341105843,
+      "grad_norm": 0.7070479989051819,
+      "learning_rate": 0.00029987088821188093,
+      "loss": 4.817,
+      "step": 79660
+    },
+    {
+      "epoch": 0.11194382434922202,
+      "grad_norm": 0.7431702613830566,
+      "learning_rate": 0.00029987058284659655,
+      "loss": 4.9736,
+      "step": 79670
+    },
+    {
+      "epoch": 0.1119578752873856,
+      "grad_norm": 0.7081708908081055,
+      "learning_rate": 0.0002998702771207812,
+      "loss": 4.7651,
+      "step": 79680
+    },
+    {
+      "epoch": 0.11197192622554918,
+      "grad_norm": 0.7065561413764954,
+      "learning_rate": 0.0002998699710344356,
+      "loss": 4.8145,
+      "step": 79690
+    },
+    {
+      "epoch": 0.11198597716371275,
+      "grad_norm": 0.7250800728797913,
+      "learning_rate": 0.00029986966458756046,
+      "loss": 4.7197,
+      "step": 79700
+    },
+    {
+      "epoch": 0.11200002810187633,
+      "grad_norm": 0.7182665467262268,
+      "learning_rate": 0.0002998693577801565,
+      "loss": 4.6853,
+      "step": 79710
+    },
+    {
+      "epoch": 0.1120140790400399,
+      "grad_norm": 0.721691906452179,
+      "learning_rate": 0.0002998690506122245,
+      "loss": 4.7432,
+      "step": 79720
+    },
+    {
+      "epoch": 0.11202812997820348,
+      "grad_norm": 0.7138781547546387,
+      "learning_rate": 0.0002998687430837652,
+      "loss": 4.8657,
+      "step": 79730
+    },
+    {
+      "epoch": 0.11204218091636706,
+      "grad_norm": 0.7386598587036133,
+      "learning_rate": 0.0002998684351947793,
+      "loss": 4.5605,
+      "step": 79740
+    },
+    {
+      "epoch": 0.11205623185453063,
+      "grad_norm": 0.72069251537323,
+      "learning_rate": 0.0002998681269452676,
+      "loss": 4.7569,
+      "step": 79750
+    },
+    {
+      "epoch": 0.11207028279269421,
+      "grad_norm": 0.7185590863227844,
+      "learning_rate": 0.0002998678183352308,
+      "loss": 4.8125,
+      "step": 79760
+    },
+    {
+      "epoch": 0.11208433373085779,
+      "grad_norm": 0.71347576379776,
+      "learning_rate": 0.00029986750936466963,
+      "loss": 4.77,
+      "step": 79770
+    },
+    {
+      "epoch": 0.11209838466902138,
+      "grad_norm": 0.7057868242263794,
+      "learning_rate": 0.00029986720003358485,
+      "loss": 4.7322,
+      "step": 79780
+    },
+    {
+      "epoch": 0.11211243560718495,
+      "grad_norm": 0.7145351767539978,
+      "learning_rate": 0.0002998668903419772,
+      "loss": 4.7988,
+      "step": 79790
+    },
+    {
+      "epoch": 0.11212648654534853,
+      "grad_norm": 0.7262371778488159,
+      "learning_rate": 0.00029986658028984746,
+      "loss": 4.7664,
+      "step": 79800
+    },
+    {
+      "epoch": 0.1121405374835121,
+      "grad_norm": 0.702280580997467,
+      "learning_rate": 0.0002998662698771963,
+      "loss": 4.8152,
+      "step": 79810
+    },
+    {
+      "epoch": 0.11215458842167568,
+      "grad_norm": 0.7286804914474487,
+      "learning_rate": 0.00029986595910402457,
+      "loss": 4.7686,
+      "step": 79820
+    },
+    {
+      "epoch": 0.11216863935983926,
+      "grad_norm": 0.7360864281654358,
+      "learning_rate": 0.0002998656479703329,
+      "loss": 4.7821,
+      "step": 79830
+    },
+    {
+      "epoch": 0.11218269029800283,
+      "grad_norm": 0.7273938059806824,
+      "learning_rate": 0.0002998653364761221,
+      "loss": 4.8064,
+      "step": 79840
+    },
+    {
+      "epoch": 0.11219674123616641,
+      "grad_norm": 0.7493408918380737,
+      "learning_rate": 0.000299865024621393,
+      "loss": 4.7691,
+      "step": 79850
+    },
+    {
+      "epoch": 0.11221079217432998,
+      "grad_norm": 0.7027277946472168,
+      "learning_rate": 0.00029986471240614623,
+      "loss": 4.5972,
+      "step": 79860
+    },
+    {
+      "epoch": 0.11222484311249356,
+      "grad_norm": 0.7099219560623169,
+      "learning_rate": 0.00029986439983038256,
+      "loss": 4.8091,
+      "step": 79870
+    },
+    {
+      "epoch": 0.11223889405065715,
+      "grad_norm": 0.726952314376831,
+      "learning_rate": 0.00029986408689410275,
+      "loss": 4.6214,
+      "step": 79880
+    },
+    {
+      "epoch": 0.11225294498882073,
+      "grad_norm": 0.7291332483291626,
+      "learning_rate": 0.0002998637735973076,
+      "loss": 4.7047,
+      "step": 79890
+    },
+    {
+      "epoch": 0.1122669959269843,
+      "grad_norm": 0.7252445220947266,
+      "learning_rate": 0.00029986345993999786,
+      "loss": 4.7274,
+      "step": 79900
+    },
+    {
+      "epoch": 0.11228104686514788,
+      "grad_norm": 0.7106611132621765,
+      "learning_rate": 0.0002998631459221742,
+      "loss": 4.7364,
+      "step": 79910
+    },
+    {
+      "epoch": 0.11229509780331146,
+      "grad_norm": 0.7336950898170471,
+      "learning_rate": 0.0002998628315438374,
+      "loss": 4.8461,
+      "step": 79920
+    },
+    {
+      "epoch": 0.11230914874147503,
+      "grad_norm": 0.7143675088882446,
+      "learning_rate": 0.00029986251680498826,
+      "loss": 4.6871,
+      "step": 79930
+    },
+    {
+      "epoch": 0.11232319967963861,
+      "grad_norm": 0.7315747737884521,
+      "learning_rate": 0.0002998622017056276,
+      "loss": 4.7463,
+      "step": 79940
+    },
+    {
+      "epoch": 0.11233725061780218,
+      "grad_norm": 0.7288627028465271,
+      "learning_rate": 0.00029986188624575604,
+      "loss": 4.7344,
+      "step": 79950
+    },
+    {
+      "epoch": 0.11235130155596576,
+      "grad_norm": 0.7390293478965759,
+      "learning_rate": 0.0002998615704253744,
+      "loss": 4.7487,
+      "step": 79960
+    },
+    {
+      "epoch": 0.11236535249412934,
+      "grad_norm": 0.7393094301223755,
+      "learning_rate": 0.00029986125424448344,
+      "loss": 4.6628,
+      "step": 79970
+    },
+    {
+      "epoch": 0.11237940343229293,
+      "grad_norm": 0.7108803391456604,
+      "learning_rate": 0.0002998609377030839,
+      "loss": 4.6948,
+      "step": 79980
+    },
+    {
+      "epoch": 0.1123934543704565,
+      "grad_norm": 0.7048096656799316,
+      "learning_rate": 0.00029986062080117657,
+      "loss": 4.7815,
+      "step": 79990
+    },
+    {
+      "epoch": 0.11240750530862008,
+      "grad_norm": 0.722067654132843,
+      "learning_rate": 0.00029986030353876226,
+      "loss": 4.6608,
+      "step": 80000
+    },
+    {
+      "epoch": 0.11242155624678365,
+      "grad_norm": 0.6960885524749756,
+      "learning_rate": 0.0002998599859158416,
+      "loss": 4.7304,
+      "step": 80010
+    },
+    {
+      "epoch": 0.11243560718494723,
+      "grad_norm": 0.7393103241920471,
+      "learning_rate": 0.0002998596679324155,
+      "loss": 4.6923,
+      "step": 80020
+    },
+    {
+      "epoch": 0.11244965812311081,
+      "grad_norm": 0.7443263530731201,
+      "learning_rate": 0.0002998593495884846,
+      "loss": 4.7715,
+      "step": 80030
+    },
+    {
+      "epoch": 0.11246370906127438,
+      "grad_norm": 0.7111883163452148,
+      "learning_rate": 0.00029985903088404974,
+      "loss": 4.6946,
+      "step": 80040
+    },
+    {
+      "epoch": 0.11247775999943796,
+      "grad_norm": 0.7264695763587952,
+      "learning_rate": 0.00029985871181911164,
+      "loss": 4.6478,
+      "step": 80050
+    },
+    {
+      "epoch": 0.11249181093760154,
+      "grad_norm": 0.7133517265319824,
+      "learning_rate": 0.0002998583923936711,
+      "loss": 4.7459,
+      "step": 80060
+    },
+    {
+      "epoch": 0.11250586187576511,
+      "grad_norm": 0.7024343609809875,
+      "learning_rate": 0.0002998580726077289,
+      "loss": 4.7401,
+      "step": 80070
+    },
+    {
+      "epoch": 0.1125199128139287,
+      "grad_norm": 0.7126244306564331,
+      "learning_rate": 0.0002998577524612858,
+      "loss": 4.8324,
+      "step": 80080
+    },
+    {
+      "epoch": 0.11253396375209228,
+      "grad_norm": 0.750248908996582,
+      "learning_rate": 0.0002998574319543426,
+      "loss": 4.6754,
+      "step": 80090
+    },
+    {
+      "epoch": 0.11254801469025585,
+      "grad_norm": 0.6958588361740112,
+      "learning_rate": 0.0002998571110869,
+      "loss": 4.7199,
+      "step": 80100
+    },
+    {
+      "epoch": 0.11256206562841943,
+      "grad_norm": 0.7012098431587219,
+      "learning_rate": 0.0002998567898589588,
+      "loss": 4.5606,
+      "step": 80110
+    },
+    {
+      "epoch": 0.112576116566583,
+      "grad_norm": 0.739078938961029,
+      "learning_rate": 0.0002998564682705198,
+      "loss": 4.8491,
+      "step": 80120
+    },
+    {
+      "epoch": 0.11259016750474658,
+      "grad_norm": 0.745468258857727,
+      "learning_rate": 0.00029985614632158374,
+      "loss": 4.7074,
+      "step": 80130
+    },
+    {
+      "epoch": 0.11260421844291016,
+      "grad_norm": 0.7302658557891846,
+      "learning_rate": 0.0002998558240121514,
+      "loss": 4.8606,
+      "step": 80140
+    },
+    {
+      "epoch": 0.11261826938107374,
+      "grad_norm": 0.7549187541007996,
+      "learning_rate": 0.00029985550134222356,
+      "loss": 4.7794,
+      "step": 80150
+    },
+    {
+      "epoch": 0.11263232031923731,
+      "grad_norm": 0.7186284065246582,
+      "learning_rate": 0.000299855178311801,
+      "loss": 4.6571,
+      "step": 80160
+    },
+    {
+      "epoch": 0.11264637125740089,
+      "grad_norm": 0.7297990918159485,
+      "learning_rate": 0.0002998548549208845,
+      "loss": 4.691,
+      "step": 80170
+    },
+    {
+      "epoch": 0.11266042219556446,
+      "grad_norm": 0.6997066140174866,
+      "learning_rate": 0.00029985453116947484,
+      "loss": 4.7066,
+      "step": 80180
+    },
+    {
+      "epoch": 0.11267447313372805,
+      "grad_norm": 0.7195040583610535,
+      "learning_rate": 0.00029985420705757275,
+      "loss": 4.7252,
+      "step": 80190
+    },
+    {
+      "epoch": 0.11268852407189163,
+      "grad_norm": 0.699199914932251,
+      "learning_rate": 0.0002998538825851791,
+      "loss": 4.7883,
+      "step": 80200
+    },
+    {
+      "epoch": 0.1127025750100552,
+      "grad_norm": 0.6951486468315125,
+      "learning_rate": 0.00029985355775229464,
+      "loss": 4.8286,
+      "step": 80210
+    },
+    {
+      "epoch": 0.11271662594821878,
+      "grad_norm": 0.7475593686103821,
+      "learning_rate": 0.0002998532325589201,
+      "loss": 4.7873,
+      "step": 80220
+    },
+    {
+      "epoch": 0.11273067688638236,
+      "grad_norm": 0.7170931100845337,
+      "learning_rate": 0.00029985290700505627,
+      "loss": 4.6795,
+      "step": 80230
+    },
+    {
+      "epoch": 0.11274472782454593,
+      "grad_norm": 0.7242776155471802,
+      "learning_rate": 0.000299852581090704,
+      "loss": 4.7406,
+      "step": 80240
+    },
+    {
+      "epoch": 0.11275877876270951,
+      "grad_norm": 0.6967573165893555,
+      "learning_rate": 0.000299852254815864,
+      "loss": 4.8199,
+      "step": 80250
+    },
+    {
+      "epoch": 0.11277282970087309,
+      "grad_norm": 0.7147936820983887,
+      "learning_rate": 0.0002998519281805371,
+      "loss": 4.7201,
+      "step": 80260
+    },
+    {
+      "epoch": 0.11278688063903666,
+      "grad_norm": 0.6991884708404541,
+      "learning_rate": 0.0002998516011847241,
+      "loss": 4.6309,
+      "step": 80270
+    },
+    {
+      "epoch": 0.11280093157720024,
+      "grad_norm": 0.7079862356185913,
+      "learning_rate": 0.0002998512738284257,
+      "loss": 4.7821,
+      "step": 80280
+    },
+    {
+      "epoch": 0.11281498251536383,
+      "grad_norm": 0.7197107076644897,
+      "learning_rate": 0.00029985094611164286,
+      "loss": 4.8384,
+      "step": 80290
+    },
+    {
+      "epoch": 0.1128290334535274,
+      "grad_norm": 0.7062892317771912,
+      "learning_rate": 0.0002998506180343762,
+      "loss": 4.6175,
+      "step": 80300
+    },
+    {
+      "epoch": 0.11284308439169098,
+      "grad_norm": 0.7397512793540955,
+      "learning_rate": 0.0002998502895966265,
+      "loss": 4.7381,
+      "step": 80310
+    },
+    {
+      "epoch": 0.11285713532985456,
+      "grad_norm": 0.7265103459358215,
+      "learning_rate": 0.0002998499607983947,
+      "loss": 4.7411,
+      "step": 80320
+    },
+    {
+      "epoch": 0.11287118626801813,
+      "grad_norm": 0.7303242087364197,
+      "learning_rate": 0.00029984963163968146,
+      "loss": 4.6545,
+      "step": 80330
+    },
+    {
+      "epoch": 0.11288523720618171,
+      "grad_norm": 0.6990224719047546,
+      "learning_rate": 0.0002998493021204877,
+      "loss": 4.715,
+      "step": 80340
+    },
+    {
+      "epoch": 0.11289928814434529,
+      "grad_norm": 0.7091715931892395,
+      "learning_rate": 0.000299848972240814,
+      "loss": 4.7124,
+      "step": 80350
+    },
+    {
+      "epoch": 0.11291333908250886,
+      "grad_norm": 0.718712329864502,
+      "learning_rate": 0.0002998486420006614,
+      "loss": 4.6979,
+      "step": 80360
+    },
+    {
+      "epoch": 0.11292739002067244,
+      "grad_norm": 0.7297937870025635,
+      "learning_rate": 0.00029984831140003056,
+      "loss": 4.7455,
+      "step": 80370
+    },
+    {
+      "epoch": 0.11294144095883601,
+      "grad_norm": 0.7224186062812805,
+      "learning_rate": 0.0002998479804389222,
+      "loss": 4.6676,
+      "step": 80380
+    },
+    {
+      "epoch": 0.1129554918969996,
+      "grad_norm": 0.72234046459198,
+      "learning_rate": 0.0002998476491173373,
+      "loss": 4.7358,
+      "step": 80390
+    },
+    {
+      "epoch": 0.11296954283516318,
+      "grad_norm": 0.7080051898956299,
+      "learning_rate": 0.0002998473174352766,
+      "loss": 4.6412,
+      "step": 80400
+    },
+    {
+      "epoch": 0.11298359377332676,
+      "grad_norm": 0.7055094242095947,
+      "learning_rate": 0.00029984698539274075,
+      "loss": 4.7382,
+      "step": 80410
+    },
+    {
+      "epoch": 0.11299764471149033,
+      "grad_norm": 0.7096608281135559,
+      "learning_rate": 0.00029984665298973074,
+      "loss": 4.6255,
+      "step": 80420
+    },
+    {
+      "epoch": 0.11301169564965391,
+      "grad_norm": 0.7352139949798584,
+      "learning_rate": 0.0002998463202262473,
+      "loss": 4.6637,
+      "step": 80430
+    },
+    {
+      "epoch": 0.11302574658781749,
+      "grad_norm": 0.692638635635376,
+      "learning_rate": 0.0002998459871022912,
+      "loss": 4.7391,
+      "step": 80440
+    },
+    {
+      "epoch": 0.11303979752598106,
+      "grad_norm": 0.7054898738861084,
+      "learning_rate": 0.00029984565361786326,
+      "loss": 4.7407,
+      "step": 80450
+    },
+    {
+      "epoch": 0.11305384846414464,
+      "grad_norm": 0.7301256656646729,
+      "learning_rate": 0.0002998453197729643,
+      "loss": 4.8302,
+      "step": 80460
+    },
+    {
+      "epoch": 0.11306789940230821,
+      "grad_norm": 0.7004954814910889,
+      "learning_rate": 0.00029984498556759516,
+      "loss": 4.7458,
+      "step": 80470
+    },
+    {
+      "epoch": 0.11308195034047179,
+      "grad_norm": 0.7302805781364441,
+      "learning_rate": 0.00029984465100175653,
+      "loss": 4.6252,
+      "step": 80480
+    },
+    {
+      "epoch": 0.11309600127863537,
+      "grad_norm": 0.7323416471481323,
+      "learning_rate": 0.0002998443160754493,
+      "loss": 4.7379,
+      "step": 80490
+    },
+    {
+      "epoch": 0.11311005221679896,
+      "grad_norm": 0.7046477198600769,
+      "learning_rate": 0.00029984398078867423,
+      "loss": 4.7576,
+      "step": 80500
+    },
+    {
+      "epoch": 0.11312410315496253,
+      "grad_norm": 0.719993531703949,
+      "learning_rate": 0.00029984364514143224,
+      "loss": 4.7161,
+      "step": 80510
+    },
+    {
+      "epoch": 0.11313815409312611,
+      "grad_norm": 0.7029352784156799,
+      "learning_rate": 0.000299843309133724,
+      "loss": 4.7524,
+      "step": 80520
+    },
+    {
+      "epoch": 0.11315220503128968,
+      "grad_norm": 0.7699235677719116,
+      "learning_rate": 0.00029984297276555033,
+      "loss": 4.7803,
+      "step": 80530
+    },
+    {
+      "epoch": 0.11316625596945326,
+      "grad_norm": 0.7614515423774719,
+      "learning_rate": 0.0002998426360369121,
+      "loss": 4.6891,
+      "step": 80540
+    },
+    {
+      "epoch": 0.11318030690761684,
+      "grad_norm": 0.7042279243469238,
+      "learning_rate": 0.00029984229894781006,
+      "loss": 4.7154,
+      "step": 80550
+    },
+    {
+      "epoch": 0.11319435784578041,
+      "grad_norm": 0.7094720602035522,
+      "learning_rate": 0.00029984196149824514,
+      "loss": 4.7491,
+      "step": 80560
+    },
+    {
+      "epoch": 0.11320840878394399,
+      "grad_norm": 0.722616970539093,
+      "learning_rate": 0.000299841623688218,
+      "loss": 4.6867,
+      "step": 80570
+    },
+    {
+      "epoch": 0.11322245972210757,
+      "grad_norm": 0.719673216342926,
+      "learning_rate": 0.0002998412855177296,
+      "loss": 4.6543,
+      "step": 80580
+    },
+    {
+      "epoch": 0.11323651066027114,
+      "grad_norm": 0.7150025963783264,
+      "learning_rate": 0.0002998409469867806,
+      "loss": 4.7076,
+      "step": 80590
+    },
+    {
+      "epoch": 0.11325056159843473,
+      "grad_norm": 0.7299203276634216,
+      "learning_rate": 0.0002998406080953719,
+      "loss": 4.7068,
+      "step": 80600
+    },
+    {
+      "epoch": 0.11326461253659831,
+      "grad_norm": 0.7360802292823792,
+      "learning_rate": 0.00029984026884350437,
+      "loss": 4.6402,
+      "step": 80610
+    },
+    {
+      "epoch": 0.11327866347476188,
+      "grad_norm": 0.7183154225349426,
+      "learning_rate": 0.0002998399292311787,
+      "loss": 4.6114,
+      "step": 80620
+    },
+    {
+      "epoch": 0.11329271441292546,
+      "grad_norm": 0.6950372457504272,
+      "learning_rate": 0.0002998395892583958,
+      "loss": 4.7284,
+      "step": 80630
+    },
+    {
+      "epoch": 0.11330676535108904,
+      "grad_norm": 0.7128282785415649,
+      "learning_rate": 0.0002998392489251564,
+      "loss": 4.8155,
+      "step": 80640
+    },
+    {
+      "epoch": 0.11332081628925261,
+      "grad_norm": 0.7069471478462219,
+      "learning_rate": 0.00029983890823146147,
+      "loss": 4.7038,
+      "step": 80650
+    },
+    {
+      "epoch": 0.11333486722741619,
+      "grad_norm": 0.7087200880050659,
+      "learning_rate": 0.00029983856717731165,
+      "loss": 4.5934,
+      "step": 80660
+    },
+    {
+      "epoch": 0.11334891816557977,
+      "grad_norm": 0.7316790819168091,
+      "learning_rate": 0.0002998382257627079,
+      "loss": 4.6918,
+      "step": 80670
+    },
+    {
+      "epoch": 0.11336296910374334,
+      "grad_norm": 0.7532024383544922,
+      "learning_rate": 0.00029983788398765094,
+      "loss": 4.667,
+      "step": 80680
+    },
+    {
+      "epoch": 0.11337702004190692,
+      "grad_norm": 0.7689909338951111,
+      "learning_rate": 0.00029983754185214167,
+      "loss": 4.6237,
+      "step": 80690
+    },
+    {
+      "epoch": 0.11339107098007051,
+      "grad_norm": 0.7248286604881287,
+      "learning_rate": 0.0002998371993561809,
+      "loss": 4.7349,
+      "step": 80700
+    },
+    {
+      "epoch": 0.11340512191823408,
+      "grad_norm": 0.7164115309715271,
+      "learning_rate": 0.00029983685649976943,
+      "loss": 4.7656,
+      "step": 80710
+    },
+    {
+      "epoch": 0.11341917285639766,
+      "grad_norm": 0.6937947273254395,
+      "learning_rate": 0.0002998365132829081,
+      "loss": 4.7744,
+      "step": 80720
+    },
+    {
+      "epoch": 0.11343322379456124,
+      "grad_norm": 0.700322687625885,
+      "learning_rate": 0.00029983616970559766,
+      "loss": 4.7036,
+      "step": 80730
+    },
+    {
+      "epoch": 0.11344727473272481,
+      "grad_norm": 0.7567490935325623,
+      "learning_rate": 0.00029983582576783907,
+      "loss": 4.7032,
+      "step": 80740
+    },
+    {
+      "epoch": 0.11346132567088839,
+      "grad_norm": 0.7181823253631592,
+      "learning_rate": 0.0002998354814696331,
+      "loss": 4.8266,
+      "step": 80750
+    },
+    {
+      "epoch": 0.11347537660905196,
+      "grad_norm": 0.7397735714912415,
+      "learning_rate": 0.0002998351368109805,
+      "loss": 4.6778,
+      "step": 80760
+    },
+    {
+      "epoch": 0.11348942754721554,
+      "grad_norm": 0.698970377445221,
+      "learning_rate": 0.0002998347917918822,
+      "loss": 4.6136,
+      "step": 80770
+    },
+    {
+      "epoch": 0.11350347848537912,
+      "grad_norm": 0.6962791085243225,
+      "learning_rate": 0.00029983444641233897,
+      "loss": 4.687,
+      "step": 80780
+    },
+    {
+      "epoch": 0.11351752942354269,
+      "grad_norm": 0.7418192028999329,
+      "learning_rate": 0.0002998341006723517,
+      "loss": 4.7033,
+      "step": 80790
+    },
+    {
+      "epoch": 0.11353158036170627,
+      "grad_norm": 0.7161971926689148,
+      "learning_rate": 0.00029983375457192117,
+      "loss": 4.8678,
+      "step": 80800
+    },
+    {
+      "epoch": 0.11354563129986986,
+      "grad_norm": 0.711134672164917,
+      "learning_rate": 0.0002998334081110483,
+      "loss": 4.7424,
+      "step": 80810
+    },
+    {
+      "epoch": 0.11355968223803344,
+      "grad_norm": 0.7014956474304199,
+      "learning_rate": 0.00029983306128973376,
+      "loss": 4.7054,
+      "step": 80820
+    },
+    {
+      "epoch": 0.11357373317619701,
+      "grad_norm": 0.7056500911712646,
+      "learning_rate": 0.00029983271410797857,
+      "loss": 4.7097,
+      "step": 80830
+    },
+    {
+      "epoch": 0.11358778411436059,
+      "grad_norm": 0.7953014373779297,
+      "learning_rate": 0.0002998323665657834,
+      "loss": 4.7563,
+      "step": 80840
+    },
+    {
+      "epoch": 0.11360183505252416,
+      "grad_norm": 0.7236987948417664,
+      "learning_rate": 0.00029983201866314924,
+      "loss": 4.6917,
+      "step": 80850
+    },
+    {
+      "epoch": 0.11361588599068774,
+      "grad_norm": 0.7303849458694458,
+      "learning_rate": 0.0002998316704000768,
+      "loss": 4.6084,
+      "step": 80860
+    },
+    {
+      "epoch": 0.11362993692885132,
+      "grad_norm": 0.6903831958770752,
+      "learning_rate": 0.00029983132177656695,
+      "loss": 4.768,
+      "step": 80870
+    },
+    {
+      "epoch": 0.11364398786701489,
+      "grad_norm": 0.744113564491272,
+      "learning_rate": 0.0002998309727926206,
+      "loss": 4.6886,
+      "step": 80880
+    },
+    {
+      "epoch": 0.11365803880517847,
+      "grad_norm": 0.7207052707672119,
+      "learning_rate": 0.0002998306234482385,
+      "loss": 4.7185,
+      "step": 80890
+    },
+    {
+      "epoch": 0.11367208974334204,
+      "grad_norm": 0.7463589906692505,
+      "learning_rate": 0.00029983027374342155,
+      "loss": 4.7078,
+      "step": 80900
+    },
+    {
+      "epoch": 0.11368614068150563,
+      "grad_norm": 1.0752589702606201,
+      "learning_rate": 0.0002998299236781705,
+      "loss": 4.6764,
+      "step": 80910
+    },
+    {
+      "epoch": 0.11370019161966921,
+      "grad_norm": 0.7272537350654602,
+      "learning_rate": 0.0002998295732524863,
+      "loss": 4.8209,
+      "step": 80920
+    },
+    {
+      "epoch": 0.11371424255783279,
+      "grad_norm": 0.7111218571662903,
+      "learning_rate": 0.00029982922246636975,
+      "loss": 4.71,
+      "step": 80930
+    },
+    {
+      "epoch": 0.11372829349599636,
+      "grad_norm": 0.7262582182884216,
+      "learning_rate": 0.0002998288713198217,
+      "loss": 4.7863,
+      "step": 80940
+    },
+    {
+      "epoch": 0.11374234443415994,
+      "grad_norm": 0.7486706972122192,
+      "learning_rate": 0.00029982851981284303,
+      "loss": 4.7294,
+      "step": 80950
+    },
+    {
+      "epoch": 0.11375639537232352,
+      "grad_norm": 0.764664351940155,
+      "learning_rate": 0.00029982816794543445,
+      "loss": 4.714,
+      "step": 80960
+    },
+    {
+      "epoch": 0.11377044631048709,
+      "grad_norm": 0.701084554195404,
+      "learning_rate": 0.000299827815717597,
+      "loss": 4.8014,
+      "step": 80970
+    },
+    {
+      "epoch": 0.11378449724865067,
+      "grad_norm": 1.806844711303711,
+      "learning_rate": 0.0002998274631293314,
+      "loss": 4.7518,
+      "step": 80980
+    },
+    {
+      "epoch": 0.11379854818681424,
+      "grad_norm": 0.7623975276947021,
+      "learning_rate": 0.0002998271101806385,
+      "loss": 4.8233,
+      "step": 80990
+    },
+    {
+      "epoch": 0.11381259912497782,
+      "grad_norm": 0.7140116095542908,
+      "learning_rate": 0.0002998267568715192,
+      "loss": 4.7421,
+      "step": 81000
+    },
+    {
+      "epoch": 0.11382665006314141,
+      "grad_norm": 0.7124093174934387,
+      "learning_rate": 0.0002998264032019743,
+      "loss": 4.7544,
+      "step": 81010
+    },
+    {
+      "epoch": 0.11384070100130499,
+      "grad_norm": 0.7354893088340759,
+      "learning_rate": 0.0002998260491720047,
+      "loss": 4.8484,
+      "step": 81020
+    },
+    {
+      "epoch": 0.11385475193946856,
+      "grad_norm": 0.7834903001785278,
+      "learning_rate": 0.0002998256947816112,
+      "loss": 4.7304,
+      "step": 81030
+    },
+    {
+      "epoch": 0.11386880287763214,
+      "grad_norm": 0.7008675336837769,
+      "learning_rate": 0.00029982534003079474,
+      "loss": 4.7707,
+      "step": 81040
+    },
+    {
+      "epoch": 0.11388285381579571,
+      "grad_norm": 0.7236028909683228,
+      "learning_rate": 0.0002998249849195561,
+      "loss": 4.7268,
+      "step": 81050
+    },
+    {
+      "epoch": 0.11389690475395929,
+      "grad_norm": 0.7143322825431824,
+      "learning_rate": 0.0002998246294478961,
+      "loss": 4.7511,
+      "step": 81060
+    },
+    {
+      "epoch": 0.11391095569212287,
+      "grad_norm": 0.7715073227882385,
+      "learning_rate": 0.0002998242736158157,
+      "loss": 4.7602,
+      "step": 81070
+    },
+    {
+      "epoch": 0.11392500663028644,
+      "grad_norm": 0.7184918522834778,
+      "learning_rate": 0.00029982391742331565,
+      "loss": 4.7352,
+      "step": 81080
+    },
+    {
+      "epoch": 0.11393905756845002,
+      "grad_norm": 0.6928876638412476,
+      "learning_rate": 0.0002998235608703969,
+      "loss": 4.5818,
+      "step": 81090
+    },
+    {
+      "epoch": 0.1139531085066136,
+      "grad_norm": 0.7220043540000916,
+      "learning_rate": 0.00029982320395706023,
+      "loss": 4.7761,
+      "step": 81100
+    },
+    {
+      "epoch": 0.11396715944477717,
+      "grad_norm": 0.6904178261756897,
+      "learning_rate": 0.00029982284668330657,
+      "loss": 4.8029,
+      "step": 81110
+    },
+    {
+      "epoch": 0.11398121038294076,
+      "grad_norm": 0.7251768708229065,
+      "learning_rate": 0.00029982248904913673,
+      "loss": 4.7282,
+      "step": 81120
+    },
+    {
+      "epoch": 0.11399526132110434,
+      "grad_norm": 0.6947658658027649,
+      "learning_rate": 0.00029982213105455153,
+      "loss": 4.7695,
+      "step": 81130
+    },
+    {
+      "epoch": 0.11400931225926791,
+      "grad_norm": 0.7100722193717957,
+      "learning_rate": 0.000299821772699552,
+      "loss": 4.8002,
+      "step": 81140
+    },
+    {
+      "epoch": 0.11402336319743149,
+      "grad_norm": 0.7027919292449951,
+      "learning_rate": 0.0002998214139841388,
+      "loss": 4.598,
+      "step": 81150
+    },
+    {
+      "epoch": 0.11403741413559507,
+      "grad_norm": 0.693475604057312,
+      "learning_rate": 0.0002998210549083129,
+      "loss": 4.7001,
+      "step": 81160
+    },
+    {
+      "epoch": 0.11405146507375864,
+      "grad_norm": 0.720072329044342,
+      "learning_rate": 0.0002998206954720751,
+      "loss": 4.7215,
+      "step": 81170
+    },
+    {
+      "epoch": 0.11406551601192222,
+      "grad_norm": 0.6978077292442322,
+      "learning_rate": 0.0002998203356754264,
+      "loss": 4.7458,
+      "step": 81180
+    },
+    {
+      "epoch": 0.1140795669500858,
+      "grad_norm": 0.721059262752533,
+      "learning_rate": 0.0002998199755183675,
+      "loss": 4.6789,
+      "step": 81190
+    },
+    {
+      "epoch": 0.11409361788824937,
+      "grad_norm": 0.7521209716796875,
+      "learning_rate": 0.0002998196150008994,
+      "loss": 4.5958,
+      "step": 81200
+    },
+    {
+      "epoch": 0.11410766882641295,
+      "grad_norm": 0.7243664860725403,
+      "learning_rate": 0.00029981925412302286,
+      "loss": 4.7493,
+      "step": 81210
+    },
+    {
+      "epoch": 0.11412171976457654,
+      "grad_norm": 0.6969353556632996,
+      "learning_rate": 0.0002998189290247855,
+      "loss": 4.8096,
+      "step": 81220
+    },
+    {
+      "epoch": 0.11413577070274011,
+      "grad_norm": 0.7307901978492737,
+      "learning_rate": 0.0002998185674621355,
+      "loss": 4.582,
+      "step": 81230
+    },
+    {
+      "epoch": 0.11414982164090369,
+      "grad_norm": 0.7377418875694275,
+      "learning_rate": 0.00029981820553907944,
+      "loss": 4.7947,
+      "step": 81240
+    },
+    {
+      "epoch": 0.11416387257906727,
+      "grad_norm": 0.7029874324798584,
+      "learning_rate": 0.00029981784325561856,
+      "loss": 4.6826,
+      "step": 81250
+    },
+    {
+      "epoch": 0.11417792351723084,
+      "grad_norm": 0.7139830589294434,
+      "learning_rate": 0.00029981748061175343,
+      "loss": 4.7922,
+      "step": 81260
+    },
+    {
+      "epoch": 0.11419197445539442,
+      "grad_norm": 0.7318224906921387,
+      "learning_rate": 0.00029981711760748513,
+      "loss": 4.6779,
+      "step": 81270
+    },
+    {
+      "epoch": 0.114206025393558,
+      "grad_norm": 0.7297080159187317,
+      "learning_rate": 0.00029981675424281437,
+      "loss": 4.7369,
+      "step": 81280
+    },
+    {
+      "epoch": 0.11422007633172157,
+      "grad_norm": 0.7171475887298584,
+      "learning_rate": 0.0002998163905177421,
+      "loss": 4.8174,
+      "step": 81290
+    },
+    {
+      "epoch": 0.11423412726988515,
+      "grad_norm": 0.710830569267273,
+      "learning_rate": 0.0002998160264322693,
+      "loss": 4.7841,
+      "step": 81300
+    },
+    {
+      "epoch": 0.11424817820804872,
+      "grad_norm": 0.7462754845619202,
+      "learning_rate": 0.0002998156619863966,
+      "loss": 4.7658,
+      "step": 81310
+    },
+    {
+      "epoch": 0.11426222914621231,
+      "grad_norm": 0.7007805705070496,
+      "learning_rate": 0.0002998152971801251,
+      "loss": 4.8825,
+      "step": 81320
+    },
+    {
+      "epoch": 0.11427628008437589,
+      "grad_norm": 0.6981909275054932,
+      "learning_rate": 0.0002998149320134556,
+      "loss": 4.7546,
+      "step": 81330
+    },
+    {
+      "epoch": 0.11429033102253947,
+      "grad_norm": 0.7107502818107605,
+      "learning_rate": 0.0002998145664863889,
+      "loss": 4.7203,
+      "step": 81340
+    },
+    {
+      "epoch": 0.11430438196070304,
+      "grad_norm": 0.7196722030639648,
+      "learning_rate": 0.000299814200598926,
+      "loss": 4.7632,
+      "step": 81350
+    },
+    {
+      "epoch": 0.11431843289886662,
+      "grad_norm": 0.7067022919654846,
+      "learning_rate": 0.00029981383435106774,
+      "loss": 4.7363,
+      "step": 81360
+    },
+    {
+      "epoch": 0.1143324838370302,
+      "grad_norm": 0.7224134802818298,
+      "learning_rate": 0.000299813467742815,
+      "loss": 4.6321,
+      "step": 81370
+    },
+    {
+      "epoch": 0.11434653477519377,
+      "grad_norm": 0.722435712814331,
+      "learning_rate": 0.00029981310077416864,
+      "loss": 4.7949,
+      "step": 81380
+    },
+    {
+      "epoch": 0.11436058571335735,
+      "grad_norm": 0.7146986722946167,
+      "learning_rate": 0.00029981273344512955,
+      "loss": 4.7668,
+      "step": 81390
+    },
+    {
+      "epoch": 0.11437463665152092,
+      "grad_norm": 0.7286592125892639,
+      "learning_rate": 0.0002998123657556986,
+      "loss": 4.7215,
+      "step": 81400
+    },
+    {
+      "epoch": 0.1143886875896845,
+      "grad_norm": 0.7215551137924194,
+      "learning_rate": 0.00029981199770587677,
+      "loss": 4.7592,
+      "step": 81410
+    },
+    {
+      "epoch": 0.11440273852784807,
+      "grad_norm": 0.737603485584259,
+      "learning_rate": 0.0002998116292956648,
+      "loss": 4.651,
+      "step": 81420
+    },
+    {
+      "epoch": 0.11441678946601166,
+      "grad_norm": 0.7101167440414429,
+      "learning_rate": 0.00029981126052506374,
+      "loss": 4.8422,
+      "step": 81430
+    },
+    {
+      "epoch": 0.11443084040417524,
+      "grad_norm": 0.7299492955207825,
+      "learning_rate": 0.0002998108913940743,
+      "loss": 4.885,
+      "step": 81440
+    },
+    {
+      "epoch": 0.11444489134233882,
+      "grad_norm": 0.6973890662193298,
+      "learning_rate": 0.00029981052190269745,
+      "loss": 4.8612,
+      "step": 81450
+    },
+    {
+      "epoch": 0.1144589422805024,
+      "grad_norm": 0.7072581648826599,
+      "learning_rate": 0.0002998101520509341,
+      "loss": 4.683,
+      "step": 81460
+    },
+    {
+      "epoch": 0.11447299321866597,
+      "grad_norm": 0.7035022974014282,
+      "learning_rate": 0.0002998097818387851,
+      "loss": 4.6592,
+      "step": 81470
+    },
+    {
+      "epoch": 0.11448704415682955,
+      "grad_norm": 0.7320752143859863,
+      "learning_rate": 0.0002998094112662514,
+      "loss": 4.6978,
+      "step": 81480
+    },
+    {
+      "epoch": 0.11450109509499312,
+      "grad_norm": 0.6897929906845093,
+      "learning_rate": 0.00029980904033333384,
+      "loss": 4.9475,
+      "step": 81490
+    },
+    {
+      "epoch": 0.1145151460331567,
+      "grad_norm": 0.6888123750686646,
+      "learning_rate": 0.0002998086690400334,
+      "loss": 4.6484,
+      "step": 81500
+    },
+    {
+      "epoch": 0.11452919697132027,
+      "grad_norm": 0.7391230463981628,
+      "learning_rate": 0.0002998082973863508,
+      "loss": 4.7175,
+      "step": 81510
+    },
+    {
+      "epoch": 0.11454324790948385,
+      "grad_norm": 0.7109723091125488,
+      "learning_rate": 0.000299807925372287,
+      "loss": 4.7042,
+      "step": 81520
+    },
+    {
+      "epoch": 0.11455729884764744,
+      "grad_norm": 0.6957715749740601,
+      "learning_rate": 0.000299807552997843,
+      "loss": 4.7355,
+      "step": 81530
+    },
+    {
+      "epoch": 0.11457134978581102,
+      "grad_norm": 0.7242121696472168,
+      "learning_rate": 0.00029980718026301965,
+      "loss": 4.8314,
+      "step": 81540
+    },
+    {
+      "epoch": 0.11458540072397459,
+      "grad_norm": 0.7289703488349915,
+      "learning_rate": 0.0002998068071678178,
+      "loss": 4.709,
+      "step": 81550
+    },
+    {
+      "epoch": 0.11459945166213817,
+      "grad_norm": 0.7176446914672852,
+      "learning_rate": 0.00029980643371223833,
+      "loss": 4.7353,
+      "step": 81560
+    },
+    {
+      "epoch": 0.11461350260030174,
+      "grad_norm": 0.7373208403587341,
+      "learning_rate": 0.0002998060598962822,
+      "loss": 4.754,
+      "step": 81570
+    },
+    {
+      "epoch": 0.11462755353846532,
+      "grad_norm": 0.7444453835487366,
+      "learning_rate": 0.00029980568571995027,
+      "loss": 4.6487,
+      "step": 81580
+    },
+    {
+      "epoch": 0.1146416044766289,
+      "grad_norm": 0.7543371915817261,
+      "learning_rate": 0.00029980531118324345,
+      "loss": 4.8064,
+      "step": 81590
+    },
+    {
+      "epoch": 0.11465565541479247,
+      "grad_norm": 2.069936513900757,
+      "learning_rate": 0.0002998049362861627,
+      "loss": 4.6117,
+      "step": 81600
+    },
+    {
+      "epoch": 0.11466970635295605,
+      "grad_norm": 0.7130931615829468,
+      "learning_rate": 0.0002998045610287088,
+      "loss": 4.7338,
+      "step": 81610
+    },
+    {
+      "epoch": 0.11468375729111963,
+      "grad_norm": 0.7223381400108337,
+      "learning_rate": 0.0002998041854108828,
+      "loss": 4.8274,
+      "step": 81620
+    },
+    {
+      "epoch": 0.11469780822928322,
+      "grad_norm": 0.7344392538070679,
+      "learning_rate": 0.0002998038094326855,
+      "loss": 4.7107,
+      "step": 81630
+    },
+    {
+      "epoch": 0.11471185916744679,
+      "grad_norm": 0.735009491443634,
+      "learning_rate": 0.00029980343309411777,
+      "loss": 4.7712,
+      "step": 81640
+    },
+    {
+      "epoch": 0.11472591010561037,
+      "grad_norm": 0.7493471503257751,
+      "learning_rate": 0.0002998030563951806,
+      "loss": 4.692,
+      "step": 81650
+    },
+    {
+      "epoch": 0.11473996104377394,
+      "grad_norm": 0.7441632747650146,
+      "learning_rate": 0.0002998026793358749,
+      "loss": 4.7691,
+      "step": 81660
+    },
+    {
+      "epoch": 0.11475401198193752,
+      "grad_norm": 0.7013207674026489,
+      "learning_rate": 0.00029980230191620157,
+      "loss": 4.6536,
+      "step": 81670
+    },
+    {
+      "epoch": 0.1147680629201011,
+      "grad_norm": 0.697334349155426,
+      "learning_rate": 0.00029980192413616146,
+      "loss": 4.6575,
+      "step": 81680
+    },
+    {
+      "epoch": 0.11478211385826467,
+      "grad_norm": 0.6946876645088196,
+      "learning_rate": 0.0002998015459957555,
+      "loss": 4.7414,
+      "step": 81690
+    },
+    {
+      "epoch": 0.11479616479642825,
+      "grad_norm": 0.7325066924095154,
+      "learning_rate": 0.0002998011674949846,
+      "loss": 4.7229,
+      "step": 81700
+    },
+    {
+      "epoch": 0.11481021573459183,
+      "grad_norm": 0.6877426505088806,
+      "learning_rate": 0.00029980078863384976,
+      "loss": 4.7216,
+      "step": 81710
+    },
+    {
+      "epoch": 0.1148242666727554,
+      "grad_norm": 0.7010763883590698,
+      "learning_rate": 0.00029980040941235175,
+      "loss": 4.77,
+      "step": 81720
+    },
+    {
+      "epoch": 0.11483831761091898,
+      "grad_norm": 0.7128811478614807,
+      "learning_rate": 0.00029980002983049156,
+      "loss": 4.555,
+      "step": 81730
+    },
+    {
+      "epoch": 0.11485236854908257,
+      "grad_norm": 0.6962786316871643,
+      "learning_rate": 0.0002997996498882701,
+      "loss": 4.699,
+      "step": 81740
+    },
+    {
+      "epoch": 0.11486641948724614,
+      "grad_norm": 0.7425155639648438,
+      "learning_rate": 0.0002997992695856883,
+      "loss": 4.7556,
+      "step": 81750
+    },
+    {
+      "epoch": 0.11488047042540972,
+      "grad_norm": 0.7460845112800598,
+      "learning_rate": 0.000299798888922747,
+      "loss": 4.7343,
+      "step": 81760
+    },
+    {
+      "epoch": 0.1148945213635733,
+      "grad_norm": 0.7301716804504395,
+      "learning_rate": 0.0002997985078994472,
+      "loss": 4.8511,
+      "step": 81770
+    },
+    {
+      "epoch": 0.11490857230173687,
+      "grad_norm": 0.7091270089149475,
+      "learning_rate": 0.00029979812651578975,
+      "loss": 4.6521,
+      "step": 81780
+    },
+    {
+      "epoch": 0.11492262323990045,
+      "grad_norm": 0.7062439918518066,
+      "learning_rate": 0.00029979774477177564,
+      "loss": 4.7847,
+      "step": 81790
+    },
+    {
+      "epoch": 0.11493667417806402,
+      "grad_norm": 0.6802224516868591,
+      "learning_rate": 0.00029979736266740574,
+      "loss": 4.7821,
+      "step": 81800
+    },
+    {
+      "epoch": 0.1149507251162276,
+      "grad_norm": 0.7181935906410217,
+      "learning_rate": 0.00029979698020268097,
+      "loss": 4.7611,
+      "step": 81810
+    },
+    {
+      "epoch": 0.11496477605439118,
+      "grad_norm": 0.6938358545303345,
+      "learning_rate": 0.0002997965973776023,
+      "loss": 4.7246,
+      "step": 81820
+    },
+    {
+      "epoch": 0.11497882699255475,
+      "grad_norm": 0.6967886686325073,
+      "learning_rate": 0.0002997962141921706,
+      "loss": 4.7603,
+      "step": 81830
+    },
+    {
+      "epoch": 0.11499287793071834,
+      "grad_norm": 0.7130746841430664,
+      "learning_rate": 0.0002997958306463867,
+      "loss": 4.7692,
+      "step": 81840
+    },
+    {
+      "epoch": 0.11500692886888192,
+      "grad_norm": 0.7128005027770996,
+      "learning_rate": 0.0002997954467402517,
+      "loss": 4.7132,
+      "step": 81850
+    },
+    {
+      "epoch": 0.1150209798070455,
+      "grad_norm": 0.7023301124572754,
+      "learning_rate": 0.00029979506247376645,
+      "loss": 4.7823,
+      "step": 81860
+    },
+    {
+      "epoch": 0.11503503074520907,
+      "grad_norm": 0.7159565091133118,
+      "learning_rate": 0.0002997946778469319,
+      "loss": 4.7408,
+      "step": 81870
+    },
+    {
+      "epoch": 0.11504908168337265,
+      "grad_norm": 0.6996265649795532,
+      "learning_rate": 0.00029979429285974886,
+      "loss": 4.681,
+      "step": 81880
+    },
+    {
+      "epoch": 0.11506313262153622,
+      "grad_norm": 0.6762425899505615,
+      "learning_rate": 0.0002997939075122184,
+      "loss": 4.6677,
+      "step": 81890
+    },
+    {
+      "epoch": 0.1150771835596998,
+      "grad_norm": 0.6953459978103638,
+      "learning_rate": 0.0002997935218043414,
+      "loss": 4.7948,
+      "step": 81900
+    },
+    {
+      "epoch": 0.11509123449786338,
+      "grad_norm": 0.7166847586631775,
+      "learning_rate": 0.00029979313573611876,
+      "loss": 4.7171,
+      "step": 81910
+    },
+    {
+      "epoch": 0.11510528543602695,
+      "grad_norm": 0.7089852094650269,
+      "learning_rate": 0.0002997927493075514,
+      "loss": 4.6225,
+      "step": 81920
+    },
+    {
+      "epoch": 0.11511933637419053,
+      "grad_norm": 0.7163013815879822,
+      "learning_rate": 0.00029979236251864033,
+      "loss": 4.8264,
+      "step": 81930
+    },
+    {
+      "epoch": 0.11513338731235412,
+      "grad_norm": 0.7131022810935974,
+      "learning_rate": 0.0002997919753693864,
+      "loss": 4.7408,
+      "step": 81940
+    },
+    {
+      "epoch": 0.1151474382505177,
+      "grad_norm": 0.7089284658432007,
+      "learning_rate": 0.0002997915878597905,
+      "loss": 4.7225,
+      "step": 81950
+    },
+    {
+      "epoch": 0.11516148918868127,
+      "grad_norm": 0.770504355430603,
+      "learning_rate": 0.00029979119998985377,
+      "loss": 4.5838,
+      "step": 81960
+    },
+    {
+      "epoch": 0.11517554012684485,
+      "grad_norm": 0.7346895933151245,
+      "learning_rate": 0.0002997908117595769,
+      "loss": 4.6957,
+      "step": 81970
+    },
+    {
+      "epoch": 0.11518959106500842,
+      "grad_norm": 0.7275788187980652,
+      "learning_rate": 0.00029979042316896095,
+      "loss": 4.7569,
+      "step": 81980
+    },
+    {
+      "epoch": 0.115203642003172,
+      "grad_norm": 0.7080996036529541,
+      "learning_rate": 0.0002997900342180068,
+      "loss": 4.6806,
+      "step": 81990
+    },
+    {
+      "epoch": 0.11521769294133558,
+      "grad_norm": 0.6918423175811768,
+      "learning_rate": 0.0002997896449067155,
+      "loss": 4.6623,
+      "step": 82000
+    },
+    {
+      "epoch": 0.11523174387949915,
+      "grad_norm": 0.7253255844116211,
+      "learning_rate": 0.00029978925523508785,
+      "loss": 4.7897,
+      "step": 82010
+    },
+    {
+      "epoch": 0.11524579481766273,
+      "grad_norm": 0.750601053237915,
+      "learning_rate": 0.00029978886520312485,
+      "loss": 4.7144,
+      "step": 82020
+    },
+    {
+      "epoch": 0.1152598457558263,
+      "grad_norm": 0.7183839678764343,
+      "learning_rate": 0.0002997884748108274,
+      "loss": 4.8058,
+      "step": 82030
+    },
+    {
+      "epoch": 0.11527389669398988,
+      "grad_norm": 0.7069101333618164,
+      "learning_rate": 0.00029978808405819654,
+      "loss": 4.6957,
+      "step": 82040
+    },
+    {
+      "epoch": 0.11528794763215347,
+      "grad_norm": 0.7078718543052673,
+      "learning_rate": 0.000299787692945233,
+      "loss": 4.6645,
+      "step": 82050
+    },
+    {
+      "epoch": 0.11530199857031705,
+      "grad_norm": 0.7066739797592163,
+      "learning_rate": 0.000299787301471938,
+      "loss": 4.7598,
+      "step": 82060
+    },
+    {
+      "epoch": 0.11531604950848062,
+      "grad_norm": 0.7062867283821106,
+      "learning_rate": 0.0002997869096383123,
+      "loss": 4.6227,
+      "step": 82070
+    },
+    {
+      "epoch": 0.1153301004466442,
+      "grad_norm": 0.686747670173645,
+      "learning_rate": 0.00029978651744435686,
+      "loss": 4.6724,
+      "step": 82080
+    },
+    {
+      "epoch": 0.11534415138480777,
+      "grad_norm": 0.729067862033844,
+      "learning_rate": 0.0002997861248900727,
+      "loss": 4.7635,
+      "step": 82090
+    },
+    {
+      "epoch": 0.11535820232297135,
+      "grad_norm": 0.7246768474578857,
+      "learning_rate": 0.00029978573197546066,
+      "loss": 4.662,
+      "step": 82100
+    },
+    {
+      "epoch": 0.11537225326113493,
+      "grad_norm": 0.743782103061676,
+      "learning_rate": 0.0002997853387005218,
+      "loss": 4.8008,
+      "step": 82110
+    },
+    {
+      "epoch": 0.1153863041992985,
+      "grad_norm": 0.6988118290901184,
+      "learning_rate": 0.00029978494506525694,
+      "loss": 4.8016,
+      "step": 82120
+    },
+    {
+      "epoch": 0.11540035513746208,
+      "grad_norm": 0.7115128040313721,
+      "learning_rate": 0.0002997845510696671,
+      "loss": 4.6794,
+      "step": 82130
+    },
+    {
+      "epoch": 0.11541440607562566,
+      "grad_norm": 0.7527329921722412,
+      "learning_rate": 0.0002997841567137532,
+      "loss": 4.7338,
+      "step": 82140
+    },
+    {
+      "epoch": 0.11542845701378925,
+      "grad_norm": 0.738523006439209,
+      "learning_rate": 0.0002997837619975162,
+      "loss": 4.6762,
+      "step": 82150
+    },
+    {
+      "epoch": 0.11544250795195282,
+      "grad_norm": 0.7189664244651794,
+      "learning_rate": 0.00029978336692095713,
+      "loss": 4.5602,
+      "step": 82160
+    },
+    {
+      "epoch": 0.1154565588901164,
+      "grad_norm": 0.7112995386123657,
+      "learning_rate": 0.00029978297148407684,
+      "loss": 4.7785,
+      "step": 82170
+    },
+    {
+      "epoch": 0.11547060982827997,
+      "grad_norm": 0.6832894682884216,
+      "learning_rate": 0.00029978257568687627,
+      "loss": 4.6804,
+      "step": 82180
+    },
+    {
+      "epoch": 0.11548466076644355,
+      "grad_norm": 0.6907115578651428,
+      "learning_rate": 0.00029978217952935645,
+      "loss": 4.6355,
+      "step": 82190
+    },
+    {
+      "epoch": 0.11549871170460713,
+      "grad_norm": 0.6926548480987549,
+      "learning_rate": 0.00029978178301151825,
+      "loss": 4.6094,
+      "step": 82200
+    },
+    {
+      "epoch": 0.1155127626427707,
+      "grad_norm": 0.7217580676078796,
+      "learning_rate": 0.0002997813861333627,
+      "loss": 4.6527,
+      "step": 82210
+    },
+    {
+      "epoch": 0.11552681358093428,
+      "grad_norm": 0.7224411368370056,
+      "learning_rate": 0.0002997809888948907,
+      "loss": 4.7685,
+      "step": 82220
+    },
+    {
+      "epoch": 0.11554086451909786,
+      "grad_norm": 0.7286561131477356,
+      "learning_rate": 0.00029978059129610326,
+      "loss": 4.5463,
+      "step": 82230
+    },
+    {
+      "epoch": 0.11555491545726143,
+      "grad_norm": 0.7249482870101929,
+      "learning_rate": 0.0002997801933370013,
+      "loss": 4.8454,
+      "step": 82240
+    },
+    {
+      "epoch": 0.11556896639542502,
+      "grad_norm": 0.7178618907928467,
+      "learning_rate": 0.00029977979501758574,
+      "loss": 4.7182,
+      "step": 82250
+    },
+    {
+      "epoch": 0.1155830173335886,
+      "grad_norm": 0.7384633421897888,
+      "learning_rate": 0.00029977939633785757,
+      "loss": 4.7321,
+      "step": 82260
+    },
+    {
+      "epoch": 0.11559706827175217,
+      "grad_norm": 0.7196593284606934,
+      "learning_rate": 0.0002997789972978178,
+      "loss": 4.7556,
+      "step": 82270
+    },
+    {
+      "epoch": 0.11561111920991575,
+      "grad_norm": 0.7066304087638855,
+      "learning_rate": 0.0002997785978974674,
+      "loss": 4.7843,
+      "step": 82280
+    },
+    {
+      "epoch": 0.11562517014807933,
+      "grad_norm": 0.736116886138916,
+      "learning_rate": 0.00029977819813680715,
+      "loss": 4.8727,
+      "step": 82290
+    },
+    {
+      "epoch": 0.1156392210862429,
+      "grad_norm": 0.708854079246521,
+      "learning_rate": 0.00029977779801583824,
+      "loss": 4.7171,
+      "step": 82300
+    },
+    {
+      "epoch": 0.11565327202440648,
+      "grad_norm": 0.7159163355827332,
+      "learning_rate": 0.00029977739753456146,
+      "loss": 4.7331,
+      "step": 82310
+    },
+    {
+      "epoch": 0.11566732296257005,
+      "grad_norm": 0.7191676497459412,
+      "learning_rate": 0.0002997769966929779,
+      "loss": 4.7406,
+      "step": 82320
+    },
+    {
+      "epoch": 0.11568137390073363,
+      "grad_norm": 0.7123490571975708,
+      "learning_rate": 0.0002997765954910885,
+      "loss": 4.7577,
+      "step": 82330
+    },
+    {
+      "epoch": 0.1156954248388972,
+      "grad_norm": 0.8625441193580627,
+      "learning_rate": 0.0002997761939288941,
+      "loss": 4.661,
+      "step": 82340
+    },
+    {
+      "epoch": 0.11570947577706078,
+      "grad_norm": 0.6980680227279663,
+      "learning_rate": 0.0002997757920063958,
+      "loss": 4.7982,
+      "step": 82350
+    },
+    {
+      "epoch": 0.11572352671522437,
+      "grad_norm": 0.7026267647743225,
+      "learning_rate": 0.00029977538972359455,
+      "loss": 4.6444,
+      "step": 82360
+    },
+    {
+      "epoch": 0.11573757765338795,
+      "grad_norm": 0.7010255455970764,
+      "learning_rate": 0.0002997749870804913,
+      "loss": 4.7762,
+      "step": 82370
+    },
+    {
+      "epoch": 0.11575162859155153,
+      "grad_norm": 0.7031188011169434,
+      "learning_rate": 0.000299774584077087,
+      "loss": 4.7665,
+      "step": 82380
+    },
+    {
+      "epoch": 0.1157656795297151,
+      "grad_norm": 0.6966001391410828,
+      "learning_rate": 0.0002997741807133827,
+      "loss": 4.6379,
+      "step": 82390
+    },
+    {
+      "epoch": 0.11577973046787868,
+      "grad_norm": 0.7564621567726135,
+      "learning_rate": 0.00029977381737799296,
+      "loss": 4.7131,
+      "step": 82400
+    },
+    {
+      "epoch": 0.11579378140604225,
+      "grad_norm": 0.7084454298019409,
+      "learning_rate": 0.0002997734133297211,
+      "loss": 4.6699,
+      "step": 82410
+    },
+    {
+      "epoch": 0.11580783234420583,
+      "grad_norm": 0.6875171661376953,
+      "learning_rate": 0.0002997730089211521,
+      "loss": 4.7788,
+      "step": 82420
+    },
+    {
+      "epoch": 0.1158218832823694,
+      "grad_norm": 0.706990122795105,
+      "learning_rate": 0.00029977260415228675,
+      "loss": 4.6794,
+      "step": 82430
+    },
+    {
+      "epoch": 0.11583593422053298,
+      "grad_norm": 0.6801334023475647,
+      "learning_rate": 0.0002997721990231261,
+      "loss": 4.6598,
+      "step": 82440
+    },
+    {
+      "epoch": 0.11584998515869656,
+      "grad_norm": 0.7689122557640076,
+      "learning_rate": 0.0002997717935336711,
+      "loss": 4.5744,
+      "step": 82450
+    },
+    {
+      "epoch": 0.11586403609686015,
+      "grad_norm": 0.7127279043197632,
+      "learning_rate": 0.00029977138768392283,
+      "loss": 4.6187,
+      "step": 82460
+    },
+    {
+      "epoch": 0.11587808703502372,
+      "grad_norm": 0.7756818532943726,
+      "learning_rate": 0.0002997709814738821,
+      "loss": 4.7612,
+      "step": 82470
+    },
+    {
+      "epoch": 0.1158921379731873,
+      "grad_norm": 0.7288581132888794,
+      "learning_rate": 0.00029977057490355004,
+      "loss": 4.7149,
+      "step": 82480
+    },
+    {
+      "epoch": 0.11590618891135088,
+      "grad_norm": 0.7076622843742371,
+      "learning_rate": 0.0002997701679729275,
+      "loss": 4.6853,
+      "step": 82490
+    },
+    {
+      "epoch": 0.11592023984951445,
+      "grad_norm": 0.7319123148918152,
+      "learning_rate": 0.00029976976068201556,
+      "loss": 4.5778,
+      "step": 82500
+    },
+    {
+      "epoch": 0.11593429078767803,
+      "grad_norm": 0.719047486782074,
+      "learning_rate": 0.0002997693530308151,
+      "loss": 4.6536,
+      "step": 82510
+    },
+    {
+      "epoch": 0.1159483417258416,
+      "grad_norm": 0.7204477190971375,
+      "learning_rate": 0.00029976894501932723,
+      "loss": 4.7233,
+      "step": 82520
+    },
+    {
+      "epoch": 0.11596239266400518,
+      "grad_norm": 0.7214565277099609,
+      "learning_rate": 0.0002997685366475528,
+      "loss": 4.7507,
+      "step": 82530
+    },
+    {
+      "epoch": 0.11597644360216876,
+      "grad_norm": 0.7362378239631653,
+      "learning_rate": 0.0002997681279154929,
+      "loss": 4.8398,
+      "step": 82540
+    },
+    {
+      "epoch": 0.11599049454033233,
+      "grad_norm": 0.7328546643257141,
+      "learning_rate": 0.0002997677188231484,
+      "loss": 4.7098,
+      "step": 82550
+    },
+    {
+      "epoch": 0.11600454547849592,
+      "grad_norm": 0.7075091004371643,
+      "learning_rate": 0.0002997673093705204,
+      "loss": 4.7569,
+      "step": 82560
+    },
+    {
+      "epoch": 0.1160185964166595,
+      "grad_norm": 0.7006434202194214,
+      "learning_rate": 0.00029976689955760985,
+      "loss": 4.725,
+      "step": 82570
+    },
+    {
+      "epoch": 0.11603264735482308,
+      "grad_norm": 0.70038241147995,
+      "learning_rate": 0.0002997664893844177,
+      "loss": 4.6548,
+      "step": 82580
+    },
+    {
+      "epoch": 0.11604669829298665,
+      "grad_norm": 0.7156614065170288,
+      "learning_rate": 0.00029976607885094495,
+      "loss": 4.7975,
+      "step": 82590
+    },
+    {
+      "epoch": 0.11606074923115023,
+      "grad_norm": 0.7034133076667786,
+      "learning_rate": 0.0002997656679571926,
+      "loss": 4.7715,
+      "step": 82600
+    },
+    {
+      "epoch": 0.1160748001693138,
+      "grad_norm": 0.7613126635551453,
+      "learning_rate": 0.0002997652567031616,
+      "loss": 4.8339,
+      "step": 82610
+    },
+    {
+      "epoch": 0.11608885110747738,
+      "grad_norm": 0.7240241765975952,
+      "learning_rate": 0.000299764845088853,
+      "loss": 4.6923,
+      "step": 82620
+    },
+    {
+      "epoch": 0.11610290204564096,
+      "grad_norm": 0.6997239589691162,
+      "learning_rate": 0.00029976443311426776,
+      "loss": 4.7306,
+      "step": 82630
+    },
+    {
+      "epoch": 0.11611695298380453,
+      "grad_norm": 0.7083427309989929,
+      "learning_rate": 0.0002997640207794069,
+      "loss": 4.6962,
+      "step": 82640
+    },
+    {
+      "epoch": 0.11613100392196811,
+      "grad_norm": 0.7667652368545532,
+      "learning_rate": 0.00029976360808427134,
+      "loss": 4.6436,
+      "step": 82650
+    },
+    {
+      "epoch": 0.11614505486013169,
+      "grad_norm": 0.6891388297080994,
+      "learning_rate": 0.0002997631950288622,
+      "loss": 4.6205,
+      "step": 82660
+    },
+    {
+      "epoch": 0.11615910579829528,
+      "grad_norm": 0.7161304354667664,
+      "learning_rate": 0.00029976278161318034,
+      "loss": 4.7486,
+      "step": 82670
+    },
+    {
+      "epoch": 0.11617315673645885,
+      "grad_norm": 0.715480625629425,
+      "learning_rate": 0.00029976236783722674,
+      "loss": 4.6952,
+      "step": 82680
+    },
+    {
+      "epoch": 0.11618720767462243,
+      "grad_norm": 0.7926939725875854,
+      "learning_rate": 0.00029976195370100256,
+      "loss": 4.7267,
+      "step": 82690
+    },
+    {
+      "epoch": 0.116201258612786,
+      "grad_norm": 0.7394651770591736,
+      "learning_rate": 0.00029976153920450863,
+      "loss": 4.6537,
+      "step": 82700
+    },
+    {
+      "epoch": 0.11621530955094958,
+      "grad_norm": 0.7032037973403931,
+      "learning_rate": 0.0002997611243477461,
+      "loss": 4.8165,
+      "step": 82710
+    },
+    {
+      "epoch": 0.11622936048911316,
+      "grad_norm": 0.7072632908821106,
+      "learning_rate": 0.00029976070913071583,
+      "loss": 4.6947,
+      "step": 82720
+    },
+    {
+      "epoch": 0.11624341142727673,
+      "grad_norm": 0.7013282179832458,
+      "learning_rate": 0.0002997602935534188,
+      "loss": 4.6732,
+      "step": 82730
+    },
+    {
+      "epoch": 0.11625746236544031,
+      "grad_norm": 0.7069587111473083,
+      "learning_rate": 0.0002997598776158562,
+      "loss": 4.6504,
+      "step": 82740
+    },
+    {
+      "epoch": 0.11627151330360389,
+      "grad_norm": 0.7177491188049316,
+      "learning_rate": 0.0002997594613180289,
+      "loss": 4.5903,
+      "step": 82750
+    },
+    {
+      "epoch": 0.11628556424176746,
+      "grad_norm": 0.7011220455169678,
+      "learning_rate": 0.0002997590446599379,
+      "loss": 4.8592,
+      "step": 82760
+    },
+    {
+      "epoch": 0.11629961517993105,
+      "grad_norm": 0.7004333138465881,
+      "learning_rate": 0.00029975862764158417,
+      "loss": 4.8258,
+      "step": 82770
+    },
+    {
+      "epoch": 0.11631366611809463,
+      "grad_norm": 0.7188377976417542,
+      "learning_rate": 0.0002997582102629688,
+      "loss": 4.7412,
+      "step": 82780
+    },
+    {
+      "epoch": 0.1163277170562582,
+      "grad_norm": 0.7212134003639221,
+      "learning_rate": 0.00029975779252409274,
+      "loss": 4.7791,
+      "step": 82790
+    },
+    {
+      "epoch": 0.11634176799442178,
+      "grad_norm": 0.7132065892219543,
+      "learning_rate": 0.00029975737442495696,
+      "loss": 4.8067,
+      "step": 82800
+    },
+    {
+      "epoch": 0.11635581893258536,
+      "grad_norm": 0.7281680107116699,
+      "learning_rate": 0.0002997569559655626,
+      "loss": 4.7152,
+      "step": 82810
+    },
+    {
+      "epoch": 0.11636986987074893,
+      "grad_norm": 0.6972888708114624,
+      "learning_rate": 0.0002997565371459105,
+      "loss": 4.7821,
+      "step": 82820
+    },
+    {
+      "epoch": 0.11638392080891251,
+      "grad_norm": 1.2375811338424683,
+      "learning_rate": 0.0002997561179660018,
+      "loss": 4.7507,
+      "step": 82830
+    },
+    {
+      "epoch": 0.11639797174707608,
+      "grad_norm": 0.6999583840370178,
+      "learning_rate": 0.0002997556984258375,
+      "loss": 4.7406,
+      "step": 82840
+    },
+    {
+      "epoch": 0.11641202268523966,
+      "grad_norm": 0.7113127708435059,
+      "learning_rate": 0.0002997552785254185,
+      "loss": 4.7202,
+      "step": 82850
+    },
+    {
+      "epoch": 0.11642607362340324,
+      "grad_norm": 0.6991801857948303,
+      "learning_rate": 0.0002997548582647459,
+      "loss": 4.6655,
+      "step": 82860
+    },
+    {
+      "epoch": 0.11644012456156683,
+      "grad_norm": 0.7328364849090576,
+      "learning_rate": 0.00029975443764382066,
+      "loss": 4.7223,
+      "step": 82870
+    },
+    {
+      "epoch": 0.1164541754997304,
+      "grad_norm": 0.7123426795005798,
+      "learning_rate": 0.0002997540166626438,
+      "loss": 4.6433,
+      "step": 82880
+    },
+    {
+      "epoch": 0.11646822643789398,
+      "grad_norm": 0.7378365397453308,
+      "learning_rate": 0.0002997535953212164,
+      "loss": 4.8043,
+      "step": 82890
+    },
+    {
+      "epoch": 0.11648227737605756,
+      "grad_norm": 0.7234169840812683,
+      "learning_rate": 0.0002997531736195394,
+      "loss": 4.726,
+      "step": 82900
+    },
+    {
+      "epoch": 0.11649632831422113,
+      "grad_norm": 0.709543764591217,
+      "learning_rate": 0.0002997527515576139,
+      "loss": 4.6436,
+      "step": 82910
+    },
+    {
+      "epoch": 0.11651037925238471,
+      "grad_norm": 0.7029217481613159,
+      "learning_rate": 0.0002997523291354408,
+      "loss": 4.7416,
+      "step": 82920
+    },
+    {
+      "epoch": 0.11652443019054828,
+      "grad_norm": 0.7221389412879944,
+      "learning_rate": 0.0002997519063530212,
+      "loss": 4.8052,
+      "step": 82930
+    },
+    {
+      "epoch": 0.11653848112871186,
+      "grad_norm": 0.7333707809448242,
+      "learning_rate": 0.00029975148321035607,
+      "loss": 4.6743,
+      "step": 82940
+    },
+    {
+      "epoch": 0.11655253206687544,
+      "grad_norm": 0.712275505065918,
+      "learning_rate": 0.00029975105970744646,
+      "loss": 4.7026,
+      "step": 82950
+    },
+    {
+      "epoch": 0.11656658300503901,
+      "grad_norm": 0.7303036451339722,
+      "learning_rate": 0.00029975063584429334,
+      "loss": 4.7575,
+      "step": 82960
+    },
+    {
+      "epoch": 0.11658063394320259,
+      "grad_norm": 0.7000608444213867,
+      "learning_rate": 0.00029975021162089783,
+      "loss": 4.8079,
+      "step": 82970
+    },
+    {
+      "epoch": 0.11659468488136618,
+      "grad_norm": 0.7333692312240601,
+      "learning_rate": 0.0002997497870372608,
+      "loss": 4.5493,
+      "step": 82980
+    },
+    {
+      "epoch": 0.11660873581952975,
+      "grad_norm": 0.7260574102401733,
+      "learning_rate": 0.0002997493620933834,
+      "loss": 4.6226,
+      "step": 82990
+    },
+    {
+      "epoch": 0.11662278675769333,
+      "grad_norm": 0.712442934513092,
+      "learning_rate": 0.0002997489367892666,
+      "loss": 4.8145,
+      "step": 83000
+    },
+    {
+      "epoch": 0.11663683769585691,
+      "grad_norm": 0.7070791125297546,
+      "learning_rate": 0.00029974851112491146,
+      "loss": 4.6327,
+      "step": 83010
+    },
+    {
+      "epoch": 0.11665088863402048,
+      "grad_norm": 0.7452596426010132,
+      "learning_rate": 0.0002997480851003189,
+      "loss": 4.6218,
+      "step": 83020
+    },
+    {
+      "epoch": 0.11666493957218406,
+      "grad_norm": 0.7247107625007629,
+      "learning_rate": 0.0002997476587154901,
+      "loss": 4.7157,
+      "step": 83030
+    },
+    {
+      "epoch": 0.11667899051034764,
+      "grad_norm": 0.7191839814186096,
+      "learning_rate": 0.000299747231970426,
+      "loss": 4.5815,
+      "step": 83040
+    },
+    {
+      "epoch": 0.11669304144851121,
+      "grad_norm": 0.7884148955345154,
+      "learning_rate": 0.0002997468048651276,
+      "loss": 4.5333,
+      "step": 83050
+    },
+    {
+      "epoch": 0.11670709238667479,
+      "grad_norm": 0.7189491987228394,
+      "learning_rate": 0.00029974637739959594,
+      "loss": 4.7521,
+      "step": 83060
+    },
+    {
+      "epoch": 0.11672114332483836,
+      "grad_norm": 0.7118692994117737,
+      "learning_rate": 0.00029974594957383213,
+      "loss": 4.7522,
+      "step": 83070
+    },
+    {
+      "epoch": 0.11673519426300195,
+      "grad_norm": 0.7117485404014587,
+      "learning_rate": 0.0002997455213878371,
+      "loss": 4.76,
+      "step": 83080
+    },
+    {
+      "epoch": 0.11674924520116553,
+      "grad_norm": 0.7011753916740417,
+      "learning_rate": 0.00029974509284161197,
+      "loss": 4.6493,
+      "step": 83090
+    },
+    {
+      "epoch": 0.1167632961393291,
+      "grad_norm": 0.7218371033668518,
+      "learning_rate": 0.0002997446639351576,
+      "loss": 4.6864,
+      "step": 83100
+    },
+    {
+      "epoch": 0.11677734707749268,
+      "grad_norm": 0.7155550718307495,
+      "learning_rate": 0.0002997442346684753,
+      "loss": 4.7834,
+      "step": 83110
+    },
+    {
+      "epoch": 0.11679139801565626,
+      "grad_norm": 0.7008214592933655,
+      "learning_rate": 0.00029974380504156585,
+      "loss": 4.6983,
+      "step": 83120
+    },
+    {
+      "epoch": 0.11680544895381983,
+      "grad_norm": 0.697800874710083,
+      "learning_rate": 0.0002997433750544304,
+      "loss": 4.7975,
+      "step": 83130
+    },
+    {
+      "epoch": 0.11681949989198341,
+      "grad_norm": 0.7439045310020447,
+      "learning_rate": 0.00029974294470706993,
+      "loss": 4.7438,
+      "step": 83140
+    },
+    {
+      "epoch": 0.11683355083014699,
+      "grad_norm": 0.7132053971290588,
+      "learning_rate": 0.00029974251399948553,
+      "loss": 4.7585,
+      "step": 83150
+    },
+    {
+      "epoch": 0.11684760176831056,
+      "grad_norm": 0.6998278498649597,
+      "learning_rate": 0.0002997420829316782,
+      "loss": 4.7226,
+      "step": 83160
+    },
+    {
+      "epoch": 0.11686165270647414,
+      "grad_norm": 0.6966537833213806,
+      "learning_rate": 0.000299741651503649,
+      "loss": 4.6966,
+      "step": 83170
+    },
+    {
+      "epoch": 0.11687570364463773,
+      "grad_norm": 0.6955662965774536,
+      "learning_rate": 0.00029974121971539894,
+      "loss": 4.6016,
+      "step": 83180
+    },
+    {
+      "epoch": 0.1168897545828013,
+      "grad_norm": 0.6958626508712769,
+      "learning_rate": 0.0002997407875669291,
+      "loss": 4.6607,
+      "step": 83190
+    },
+    {
+      "epoch": 0.11690380552096488,
+      "grad_norm": 0.7100691199302673,
+      "learning_rate": 0.00029974035505824046,
+      "loss": 4.7568,
+      "step": 83200
+    },
+    {
+      "epoch": 0.11691785645912846,
+      "grad_norm": 0.7105290293693542,
+      "learning_rate": 0.00029973992218933416,
+      "loss": 4.8082,
+      "step": 83210
+    },
+    {
+      "epoch": 0.11693190739729203,
+      "grad_norm": 0.7148688435554504,
+      "learning_rate": 0.00029973948896021117,
+      "loss": 4.6506,
+      "step": 83220
+    },
+    {
+      "epoch": 0.11694595833545561,
+      "grad_norm": 0.7135083079338074,
+      "learning_rate": 0.0002997390553708725,
+      "loss": 4.7098,
+      "step": 83230
+    },
+    {
+      "epoch": 0.11696000927361919,
+      "grad_norm": 0.6871196627616882,
+      "learning_rate": 0.0002997386214213192,
+      "loss": 4.7369,
+      "step": 83240
+    },
+    {
+      "epoch": 0.11697406021178276,
+      "grad_norm": 0.7220321297645569,
+      "learning_rate": 0.00029973818711155246,
+      "loss": 4.7927,
+      "step": 83250
+    },
+    {
+      "epoch": 0.11698811114994634,
+      "grad_norm": 0.6984358429908752,
+      "learning_rate": 0.0002997377524415731,
+      "loss": 4.7761,
+      "step": 83260
+    },
+    {
+      "epoch": 0.11700216208810992,
+      "grad_norm": 0.7106263637542725,
+      "learning_rate": 0.00029973731741138234,
+      "loss": 4.7305,
+      "step": 83270
+    },
+    {
+      "epoch": 0.11701621302627349,
+      "grad_norm": 0.715644121170044,
+      "learning_rate": 0.00029973688202098115,
+      "loss": 4.781,
+      "step": 83280
+    },
+    {
+      "epoch": 0.11703026396443708,
+      "grad_norm": 0.7058255076408386,
+      "learning_rate": 0.0002997364462703706,
+      "loss": 4.6378,
+      "step": 83290
+    },
+    {
+      "epoch": 0.11704431490260066,
+      "grad_norm": 0.6990760564804077,
+      "learning_rate": 0.0002997360101595517,
+      "loss": 4.7514,
+      "step": 83300
+    },
+    {
+      "epoch": 0.11705836584076423,
+      "grad_norm": 0.7147160172462463,
+      "learning_rate": 0.00029973557368852553,
+      "loss": 4.6166,
+      "step": 83310
+    },
+    {
+      "epoch": 0.11707241677892781,
+      "grad_norm": 0.712631344795227,
+      "learning_rate": 0.00029973513685729316,
+      "loss": 4.6634,
+      "step": 83320
+    },
+    {
+      "epoch": 0.11708646771709139,
+      "grad_norm": 0.7080478668212891,
+      "learning_rate": 0.0002997346996658556,
+      "loss": 4.6785,
+      "step": 83330
+    },
+    {
+      "epoch": 0.11710051865525496,
+      "grad_norm": 0.6984758973121643,
+      "learning_rate": 0.0002997342621142139,
+      "loss": 4.7836,
+      "step": 83340
+    },
+    {
+      "epoch": 0.11711456959341854,
+      "grad_norm": 0.7084048986434937,
+      "learning_rate": 0.00029973382420236917,
+      "loss": 4.8846,
+      "step": 83350
+    },
+    {
+      "epoch": 0.11712862053158211,
+      "grad_norm": 0.7941184043884277,
+      "learning_rate": 0.0002997333859303225,
+      "loss": 4.5988,
+      "step": 83360
+    },
+    {
+      "epoch": 0.11714267146974569,
+      "grad_norm": 0.7254170775413513,
+      "learning_rate": 0.0002997329472980747,
+      "loss": 4.7335,
+      "step": 83370
+    },
+    {
+      "epoch": 0.11715672240790927,
+      "grad_norm": 0.7091358304023743,
+      "learning_rate": 0.0002997325083056271,
+      "loss": 4.7572,
+      "step": 83380
+    },
+    {
+      "epoch": 0.11717077334607286,
+      "grad_norm": 0.7274218797683716,
+      "learning_rate": 0.0002997320689529807,
+      "loss": 4.6113,
+      "step": 83390
+    },
+    {
+      "epoch": 0.11718482428423643,
+      "grad_norm": 0.699985146522522,
+      "learning_rate": 0.0002997316292401364,
+      "loss": 4.7306,
+      "step": 83400
+    },
+    {
+      "epoch": 0.11719887522240001,
+      "grad_norm": 0.7258591055870056,
+      "learning_rate": 0.0002997311891670954,
+      "loss": 4.7404,
+      "step": 83410
+    },
+    {
+      "epoch": 0.11721292616056359,
+      "grad_norm": 0.7040428519248962,
+      "learning_rate": 0.00029973074873385877,
+      "loss": 4.8231,
+      "step": 83420
+    },
+    {
+      "epoch": 0.11722697709872716,
+      "grad_norm": 0.7231871485710144,
+      "learning_rate": 0.0002997303079404275,
+      "loss": 4.7028,
+      "step": 83430
+    },
+    {
+      "epoch": 0.11724102803689074,
+      "grad_norm": 0.701480507850647,
+      "learning_rate": 0.0002997298667868027,
+      "loss": 4.9066,
+      "step": 83440
+    },
+    {
+      "epoch": 0.11725507897505431,
+      "grad_norm": 0.739446759223938,
+      "learning_rate": 0.0002997294252729854,
+      "loss": 4.8486,
+      "step": 83450
+    },
+    {
+      "epoch": 0.11726912991321789,
+      "grad_norm": 0.7062489986419678,
+      "learning_rate": 0.00029972898339897664,
+      "loss": 4.7646,
+      "step": 83460
+    },
+    {
+      "epoch": 0.11728318085138147,
+      "grad_norm": 0.7041696906089783,
+      "learning_rate": 0.0002997285411647775,
+      "loss": 4.7471,
+      "step": 83470
+    },
+    {
+      "epoch": 0.11729723178954504,
+      "grad_norm": 0.7124684453010559,
+      "learning_rate": 0.00029972809857038913,
+      "loss": 4.719,
+      "step": 83480
+    },
+    {
+      "epoch": 0.11731128272770863,
+      "grad_norm": 0.7017519474029541,
+      "learning_rate": 0.0002997276556158124,
+      "loss": 4.6908,
+      "step": 83490
+    },
+    {
+      "epoch": 0.11732533366587221,
+      "grad_norm": 0.7156013250350952,
+      "learning_rate": 0.0002997272123010486,
+      "loss": 4.7906,
+      "step": 83500
+    },
+    {
+      "epoch": 0.11733938460403578,
+      "grad_norm": 0.715263307094574,
+      "learning_rate": 0.0002997267686260987,
+      "loss": 4.6222,
+      "step": 83510
+    },
+    {
+      "epoch": 0.11735343554219936,
+      "grad_norm": 0.7254456877708435,
+      "learning_rate": 0.0002997263245909637,
+      "loss": 4.585,
+      "step": 83520
+    },
+    {
+      "epoch": 0.11736748648036294,
+      "grad_norm": 0.7293530106544495,
+      "learning_rate": 0.00029972588019564476,
+      "loss": 4.6986,
+      "step": 83530
+    },
+    {
+      "epoch": 0.11738153741852651,
+      "grad_norm": 0.7089353799819946,
+      "learning_rate": 0.00029972543544014285,
+      "loss": 4.6565,
+      "step": 83540
+    },
+    {
+      "epoch": 0.11739558835669009,
+      "grad_norm": 0.7219105958938599,
+      "learning_rate": 0.0002997249903244592,
+      "loss": 4.6934,
+      "step": 83550
+    },
+    {
+      "epoch": 0.11740963929485367,
+      "grad_norm": 0.7276551723480225,
+      "learning_rate": 0.00029972454484859475,
+      "loss": 4.6889,
+      "step": 83560
+    },
+    {
+      "epoch": 0.11742369023301724,
+      "grad_norm": 0.6939069628715515,
+      "learning_rate": 0.00029972409901255056,
+      "loss": 4.7927,
+      "step": 83570
+    },
+    {
+      "epoch": 0.11743774117118082,
+      "grad_norm": 0.7165120244026184,
+      "learning_rate": 0.0002997236528163278,
+      "loss": 4.757,
+      "step": 83580
+    },
+    {
+      "epoch": 0.1174517921093444,
+      "grad_norm": 0.774430513381958,
+      "learning_rate": 0.00029972320625992754,
+      "loss": 4.657,
+      "step": 83590
+    },
+    {
+      "epoch": 0.11746584304750798,
+      "grad_norm": 0.7265297770500183,
+      "learning_rate": 0.00029972275934335077,
+      "loss": 4.6596,
+      "step": 83600
+    },
+    {
+      "epoch": 0.11747989398567156,
+      "grad_norm": 0.7017237544059753,
+      "learning_rate": 0.00029972231206659857,
+      "loss": 4.7227,
+      "step": 83610
+    },
+    {
+      "epoch": 0.11749394492383514,
+      "grad_norm": 0.703671395778656,
+      "learning_rate": 0.00029972186442967213,
+      "loss": 4.6661,
+      "step": 83620
+    },
+    {
+      "epoch": 0.11750799586199871,
+      "grad_norm": 0.7623944878578186,
+      "learning_rate": 0.00029972141643257237,
+      "loss": 4.6292,
+      "step": 83630
+    },
+    {
+      "epoch": 0.11752204680016229,
+      "grad_norm": 0.73209547996521,
+      "learning_rate": 0.00029972096807530043,
+      "loss": 4.6943,
+      "step": 83640
+    },
+    {
+      "epoch": 0.11753609773832586,
+      "grad_norm": 0.7071471810340881,
+      "learning_rate": 0.0002997205193578575,
+      "loss": 4.6732,
+      "step": 83650
+    },
+    {
+      "epoch": 0.11755014867648944,
+      "grad_norm": 0.7387067675590515,
+      "learning_rate": 0.00029972007028024447,
+      "loss": 4.6494,
+      "step": 83660
+    },
+    {
+      "epoch": 0.11756419961465302,
+      "grad_norm": 0.7155556678771973,
+      "learning_rate": 0.00029971962084246256,
+      "loss": 4.6376,
+      "step": 83670
+    },
+    {
+      "epoch": 0.1175782505528166,
+      "grad_norm": 0.7402138113975525,
+      "learning_rate": 0.0002997191710445128,
+      "loss": 4.6761,
+      "step": 83680
+    },
+    {
+      "epoch": 0.11759230149098017,
+      "grad_norm": 0.7761551141738892,
+      "learning_rate": 0.0002997187208863963,
+      "loss": 4.6724,
+      "step": 83690
+    },
+    {
+      "epoch": 0.11760635242914376,
+      "grad_norm": 0.7482668161392212,
+      "learning_rate": 0.0002997182703681141,
+      "loss": 4.7362,
+      "step": 83700
+    },
+    {
+      "epoch": 0.11762040336730734,
+      "grad_norm": 0.7007404565811157,
+      "learning_rate": 0.00029971781948966733,
+      "loss": 4.8267,
+      "step": 83710
+    },
+    {
+      "epoch": 0.11763445430547091,
+      "grad_norm": 0.6935082077980042,
+      "learning_rate": 0.000299717368251057,
+      "loss": 4.6847,
+      "step": 83720
+    },
+    {
+      "epoch": 0.11764850524363449,
+      "grad_norm": 0.7338686585426331,
+      "learning_rate": 0.00029971691665228424,
+      "loss": 4.8848,
+      "step": 83730
+    },
+    {
+      "epoch": 0.11766255618179806,
+      "grad_norm": 0.7095941305160522,
+      "learning_rate": 0.00029971646469335015,
+      "loss": 4.7018,
+      "step": 83740
+    },
+    {
+      "epoch": 0.11767660711996164,
+      "grad_norm": 0.7719713449478149,
+      "learning_rate": 0.00029971601237425586,
+      "loss": 4.6187,
+      "step": 83750
+    },
+    {
+      "epoch": 0.11769065805812522,
+      "grad_norm": 0.7228941321372986,
+      "learning_rate": 0.00029971555969500235,
+      "loss": 4.6907,
+      "step": 83760
+    },
+    {
+      "epoch": 0.11770470899628879,
+      "grad_norm": 0.7366059422492981,
+      "learning_rate": 0.0002997151066555908,
+      "loss": 4.7233,
+      "step": 83770
+    },
+    {
+      "epoch": 0.11771875993445237,
+      "grad_norm": 0.8631296157836914,
+      "learning_rate": 0.0002997146532560222,
+      "loss": 4.7148,
+      "step": 83780
+    },
+    {
+      "epoch": 0.11773281087261595,
+      "grad_norm": 0.7031307816505432,
+      "learning_rate": 0.00029971419949629776,
+      "loss": 4.7728,
+      "step": 83790
+    },
+    {
+      "epoch": 0.11774686181077954,
+      "grad_norm": 0.7433533668518066,
+      "learning_rate": 0.00029971374537641853,
+      "loss": 4.6781,
+      "step": 83800
+    },
+    {
+      "epoch": 0.11776091274894311,
+      "grad_norm": 0.760563850402832,
+      "learning_rate": 0.00029971329089638555,
+      "loss": 4.7035,
+      "step": 83810
+    },
+    {
+      "epoch": 0.11777496368710669,
+      "grad_norm": 0.7019788026809692,
+      "learning_rate": 0.00029971283605619996,
+      "loss": 4.6762,
+      "step": 83820
+    },
+    {
+      "epoch": 0.11778901462527026,
+      "grad_norm": 0.6909261345863342,
+      "learning_rate": 0.00029971238085586284,
+      "loss": 4.6912,
+      "step": 83830
+    },
+    {
+      "epoch": 0.11780306556343384,
+      "grad_norm": 0.7148104310035706,
+      "learning_rate": 0.00029971192529537534,
+      "loss": 4.7071,
+      "step": 83840
+    },
+    {
+      "epoch": 0.11781711650159742,
+      "grad_norm": 0.7196040749549866,
+      "learning_rate": 0.00029971146937473844,
+      "loss": 4.7632,
+      "step": 83850
+    },
+    {
+      "epoch": 0.11783116743976099,
+      "grad_norm": 0.7324225902557373,
+      "learning_rate": 0.0002997110130939533,
+      "loss": 4.5944,
+      "step": 83860
+    },
+    {
+      "epoch": 0.11784521837792457,
+      "grad_norm": 0.7406251430511475,
+      "learning_rate": 0.00029971055645302115,
+      "loss": 4.7681,
+      "step": 83870
+    },
+    {
+      "epoch": 0.11785926931608814,
+      "grad_norm": 0.6971359252929688,
+      "learning_rate": 0.00029971009945194284,
+      "loss": 4.5862,
+      "step": 83880
+    },
+    {
+      "epoch": 0.11787332025425172,
+      "grad_norm": 0.7004740834236145,
+      "learning_rate": 0.00029970964209071957,
+      "loss": 4.6547,
+      "step": 83890
+    },
+    {
+      "epoch": 0.1178873711924153,
+      "grad_norm": 0.736107349395752,
+      "learning_rate": 0.0002997091843693525,
+      "loss": 4.5973,
+      "step": 83900
+    },
+    {
+      "epoch": 0.11790142213057889,
+      "grad_norm": 0.7069373726844788,
+      "learning_rate": 0.00029970872628784273,
+      "loss": 4.6779,
+      "step": 83910
+    },
+    {
+      "epoch": 0.11791547306874246,
+      "grad_norm": 0.7382087707519531,
+      "learning_rate": 0.0002997082678461913,
+      "loss": 4.5928,
+      "step": 83920
+    },
+    {
+      "epoch": 0.11792952400690604,
+      "grad_norm": 0.7047857642173767,
+      "learning_rate": 0.0002997078090443993,
+      "loss": 4.7603,
+      "step": 83930
+    },
+    {
+      "epoch": 0.11794357494506962,
+      "grad_norm": 0.7913258075714111,
+      "learning_rate": 0.0002997073498824679,
+      "loss": 4.5967,
+      "step": 83940
+    },
+    {
+      "epoch": 0.11795762588323319,
+      "grad_norm": 0.7597509026527405,
+      "learning_rate": 0.0002997068903603982,
+      "loss": 4.6874,
+      "step": 83950
+    },
+    {
+      "epoch": 0.11797167682139677,
+      "grad_norm": 0.7429311871528625,
+      "learning_rate": 0.00029970643047819124,
+      "loss": 4.6269,
+      "step": 83960
+    },
+    {
+      "epoch": 0.11798572775956034,
+      "grad_norm": 0.7132638692855835,
+      "learning_rate": 0.00029970597023584814,
+      "loss": 4.6915,
+      "step": 83970
+    },
+    {
+      "epoch": 0.11799977869772392,
+      "grad_norm": 0.7404999136924744,
+      "learning_rate": 0.0002997055096333701,
+      "loss": 4.7002,
+      "step": 83980
+    },
+    {
+      "epoch": 0.1180138296358875,
+      "grad_norm": 0.7566960453987122,
+      "learning_rate": 0.0002997050486707581,
+      "loss": 4.6367,
+      "step": 83990
+    },
+    {
+      "epoch": 0.11802788057405107,
+      "grad_norm": 0.7191115617752075,
+      "learning_rate": 0.0002997045873480134,
+      "loss": 4.6247,
+      "step": 84000
+    },
+    {
+      "epoch": 0.11804193151221466,
+      "grad_norm": 0.7127372026443481,
+      "learning_rate": 0.0002997041256651369,
+      "loss": 4.5988,
+      "step": 84010
+    },
+    {
+      "epoch": 0.11805598245037824,
+      "grad_norm": 0.7138316035270691,
+      "learning_rate": 0.0002997036636221299,
+      "loss": 4.694,
+      "step": 84020
+    },
+    {
+      "epoch": 0.11807003338854181,
+      "grad_norm": 0.7099950313568115,
+      "learning_rate": 0.0002997032012189934,
+      "loss": 4.7058,
+      "step": 84030
+    },
+    {
+      "epoch": 0.11808408432670539,
+      "grad_norm": 0.7554578185081482,
+      "learning_rate": 0.00029970273845572864,
+      "loss": 4.6664,
+      "step": 84040
+    },
+    {
+      "epoch": 0.11809813526486897,
+      "grad_norm": 0.7338352799415588,
+      "learning_rate": 0.00029970227533233657,
+      "loss": 4.6666,
+      "step": 84050
+    },
+    {
+      "epoch": 0.11811218620303254,
+      "grad_norm": 0.7148082852363586,
+      "learning_rate": 0.0002997018118488184,
+      "loss": 4.6799,
+      "step": 84060
+    },
+    {
+      "epoch": 0.11812623714119612,
+      "grad_norm": 0.7274507284164429,
+      "learning_rate": 0.0002997013480051753,
+      "loss": 4.6302,
+      "step": 84070
+    },
+    {
+      "epoch": 0.1181402880793597,
+      "grad_norm": 0.7034341096878052,
+      "learning_rate": 0.0002997008838014082,
+      "loss": 4.743,
+      "step": 84080
+    },
+    {
+      "epoch": 0.11815433901752327,
+      "grad_norm": 0.7157617807388306,
+      "learning_rate": 0.00029970041923751837,
+      "loss": 4.6052,
+      "step": 84090
+    },
+    {
+      "epoch": 0.11816838995568685,
+      "grad_norm": 0.8156205415725708,
+      "learning_rate": 0.00029969995431350686,
+      "loss": 4.7544,
+      "step": 84100
+    },
+    {
+      "epoch": 0.11818244089385044,
+      "grad_norm": 0.7375916838645935,
+      "learning_rate": 0.0002996994890293749,
+      "loss": 4.7907,
+      "step": 84110
+    },
+    {
+      "epoch": 0.11819649183201401,
+      "grad_norm": 0.6925612092018127,
+      "learning_rate": 0.00029969902338512343,
+      "loss": 4.7709,
+      "step": 84120
+    },
+    {
+      "epoch": 0.11821054277017759,
+      "grad_norm": 0.7296881079673767,
+      "learning_rate": 0.0002996985573807537,
+      "loss": 4.7123,
+      "step": 84130
+    },
+    {
+      "epoch": 0.11822459370834117,
+      "grad_norm": 0.717957079410553,
+      "learning_rate": 0.0002996980910162668,
+      "loss": 4.6673,
+      "step": 84140
+    },
+    {
+      "epoch": 0.11823864464650474,
+      "grad_norm": 0.6915928721427917,
+      "learning_rate": 0.0002996976242916638,
+      "loss": 4.6409,
+      "step": 84150
+    },
+    {
+      "epoch": 0.11825269558466832,
+      "grad_norm": 0.69838547706604,
+      "learning_rate": 0.00029969715720694595,
+      "loss": 4.8507,
+      "step": 84160
+    },
+    {
+      "epoch": 0.1182667465228319,
+      "grad_norm": 0.701752245426178,
+      "learning_rate": 0.0002996966897621143,
+      "loss": 4.5703,
+      "step": 84170
+    },
+    {
+      "epoch": 0.11828079746099547,
+      "grad_norm": 0.7115376591682434,
+      "learning_rate": 0.0002996962219571699,
+      "loss": 4.7654,
+      "step": 84180
+    },
+    {
+      "epoch": 0.11829484839915905,
+      "grad_norm": 0.7013298869132996,
+      "learning_rate": 0.00029969575379211396,
+      "loss": 4.6597,
+      "step": 84190
+    },
+    {
+      "epoch": 0.11830889933732262,
+      "grad_norm": 0.7234477996826172,
+      "learning_rate": 0.0002996952852669476,
+      "loss": 4.6087,
+      "step": 84200
+    },
+    {
+      "epoch": 0.1183229502754862,
+      "grad_norm": 0.6927001476287842,
+      "learning_rate": 0.00029969481638167193,
+      "loss": 4.622,
+      "step": 84210
+    },
+    {
+      "epoch": 0.11833700121364979,
+      "grad_norm": 0.7288459539413452,
+      "learning_rate": 0.00029969434713628807,
+      "loss": 4.6142,
+      "step": 84220
+    },
+    {
+      "epoch": 0.11835105215181337,
+      "grad_norm": 0.7043154239654541,
+      "learning_rate": 0.0002996938775307972,
+      "loss": 4.6994,
+      "step": 84230
+    },
+    {
+      "epoch": 0.11836510308997694,
+      "grad_norm": 0.7041778564453125,
+      "learning_rate": 0.0002996934075652004,
+      "loss": 4.6914,
+      "step": 84240
+    },
+    {
+      "epoch": 0.11837915402814052,
+      "grad_norm": 0.7131189703941345,
+      "learning_rate": 0.00029969293723949885,
+      "loss": 4.6957,
+      "step": 84250
+    },
+    {
+      "epoch": 0.1183932049663041,
+      "grad_norm": 0.7089633345603943,
+      "learning_rate": 0.0002996924665536936,
+      "loss": 4.7275,
+      "step": 84260
+    },
+    {
+      "epoch": 0.11840725590446767,
+      "grad_norm": 0.7021405696868896,
+      "learning_rate": 0.00029969199550778584,
+      "loss": 4.7756,
+      "step": 84270
+    },
+    {
+      "epoch": 0.11842130684263125,
+      "grad_norm": 0.706847071647644,
+      "learning_rate": 0.00029969152410177667,
+      "loss": 4.792,
+      "step": 84280
+    },
+    {
+      "epoch": 0.11843535778079482,
+      "grad_norm": 0.7018424272537231,
+      "learning_rate": 0.0002996910523356672,
+      "loss": 4.6697,
+      "step": 84290
+    },
+    {
+      "epoch": 0.1184494087189584,
+      "grad_norm": 0.7254888415336609,
+      "learning_rate": 0.0002996905802094587,
+      "loss": 4.81,
+      "step": 84300
+    },
+    {
+      "epoch": 0.11846345965712198,
+      "grad_norm": 0.7102382779121399,
+      "learning_rate": 0.0002996901077231522,
+      "loss": 4.6677,
+      "step": 84310
+    },
+    {
+      "epoch": 0.11847751059528557,
+      "grad_norm": 0.6817678213119507,
+      "learning_rate": 0.00029968963487674884,
+      "loss": 4.6627,
+      "step": 84320
+    },
+    {
+      "epoch": 0.11849156153344914,
+      "grad_norm": 0.7572276592254639,
+      "learning_rate": 0.00029968916167024973,
+      "loss": 4.8695,
+      "step": 84330
+    },
+    {
+      "epoch": 0.11850561247161272,
+      "grad_norm": 0.7169764041900635,
+      "learning_rate": 0.00029968868810365613,
+      "loss": 4.6564,
+      "step": 84340
+    },
+    {
+      "epoch": 0.1185196634097763,
+      "grad_norm": 0.708909273147583,
+      "learning_rate": 0.000299688214176969,
+      "loss": 4.5852,
+      "step": 84350
+    },
+    {
+      "epoch": 0.11853371434793987,
+      "grad_norm": 0.7388038039207458,
+      "learning_rate": 0.0002996877398901896,
+      "loss": 4.7354,
+      "step": 84360
+    },
+    {
+      "epoch": 0.11854776528610345,
+      "grad_norm": 0.7097702622413635,
+      "learning_rate": 0.00029968726524331915,
+      "loss": 4.7295,
+      "step": 84370
+    },
+    {
+      "epoch": 0.11856181622426702,
+      "grad_norm": 0.7070819735527039,
+      "learning_rate": 0.00029968679023635857,
+      "loss": 4.6223,
+      "step": 84380
+    },
+    {
+      "epoch": 0.1185758671624306,
+      "grad_norm": 0.7111035585403442,
+      "learning_rate": 0.0002996863148693092,
+      "loss": 4.5126,
+      "step": 84390
+    },
+    {
+      "epoch": 0.11858991810059417,
+      "grad_norm": 0.7140163779258728,
+      "learning_rate": 0.00029968583914217204,
+      "loss": 4.7595,
+      "step": 84400
+    },
+    {
+      "epoch": 0.11860396903875775,
+      "grad_norm": 0.7231415510177612,
+      "learning_rate": 0.00029968536305494836,
+      "loss": 4.7312,
+      "step": 84410
+    },
+    {
+      "epoch": 0.11861801997692134,
+      "grad_norm": 0.74226975440979,
+      "learning_rate": 0.00029968488660763923,
+      "loss": 4.6488,
+      "step": 84420
+    },
+    {
+      "epoch": 0.11863207091508492,
+      "grad_norm": 0.7267938852310181,
+      "learning_rate": 0.0002996844098002458,
+      "loss": 4.5626,
+      "step": 84430
+    },
+    {
+      "epoch": 0.11864612185324849,
+      "grad_norm": 0.7126814126968384,
+      "learning_rate": 0.00029968393263276927,
+      "loss": 4.6538,
+      "step": 84440
+    },
+    {
+      "epoch": 0.11866017279141207,
+      "grad_norm": 0.7250803709030151,
+      "learning_rate": 0.0002996834551052107,
+      "loss": 4.6957,
+      "step": 84450
+    },
+    {
+      "epoch": 0.11867422372957565,
+      "grad_norm": 0.7476853728294373,
+      "learning_rate": 0.0002996829772175713,
+      "loss": 4.6687,
+      "step": 84460
+    },
+    {
+      "epoch": 0.11868827466773922,
+      "grad_norm": 0.7315259575843811,
+      "learning_rate": 0.0002996824989698522,
+      "loss": 4.7679,
+      "step": 84470
+    },
+    {
+      "epoch": 0.1187023256059028,
+      "grad_norm": 0.7134382128715515,
+      "learning_rate": 0.00029968202036205455,
+      "loss": 4.7786,
+      "step": 84480
+    },
+    {
+      "epoch": 0.11871637654406637,
+      "grad_norm": 0.7032891511917114,
+      "learning_rate": 0.0002996815413941795,
+      "loss": 4.6831,
+      "step": 84490
+    },
+    {
+      "epoch": 0.11873042748222995,
+      "grad_norm": 0.708054780960083,
+      "learning_rate": 0.0002996810620662282,
+      "loss": 4.7639,
+      "step": 84500
+    },
+    {
+      "epoch": 0.11874447842039353,
+      "grad_norm": 0.7284858822822571,
+      "learning_rate": 0.00029968058237820185,
+      "loss": 4.7167,
+      "step": 84510
+    },
+    {
+      "epoch": 0.1187585293585571,
+      "grad_norm": 0.7195180058479309,
+      "learning_rate": 0.0002996801023301015,
+      "loss": 4.7123,
+      "step": 84520
+    },
+    {
+      "epoch": 0.11877258029672069,
+      "grad_norm": 0.7547293305397034,
+      "learning_rate": 0.00029967962192192847,
+      "loss": 4.7145,
+      "step": 84530
+    },
+    {
+      "epoch": 0.11878663123488427,
+      "grad_norm": 0.7729194760322571,
+      "learning_rate": 0.00029967914115368375,
+      "loss": 4.7879,
+      "step": 84540
+    },
+    {
+      "epoch": 0.11880068217304784,
+      "grad_norm": 0.7287982702255249,
+      "learning_rate": 0.0002996786600253686,
+      "loss": 4.7128,
+      "step": 84550
+    },
+    {
+      "epoch": 0.11881473311121142,
+      "grad_norm": 0.7168266177177429,
+      "learning_rate": 0.00029967817853698413,
+      "loss": 4.6013,
+      "step": 84560
+    },
+    {
+      "epoch": 0.118828784049375,
+      "grad_norm": 0.6961681246757507,
+      "learning_rate": 0.0002996776966885315,
+      "loss": 4.7291,
+      "step": 84570
+    },
+    {
+      "epoch": 0.11884283498753857,
+      "grad_norm": 0.7002440690994263,
+      "learning_rate": 0.00029967721448001186,
+      "loss": 4.6898,
+      "step": 84580
+    },
+    {
+      "epoch": 0.11885688592570215,
+      "grad_norm": 0.7198019027709961,
+      "learning_rate": 0.00029967673191142645,
+      "loss": 4.6502,
+      "step": 84590
+    },
+    {
+      "epoch": 0.11887093686386573,
+      "grad_norm": 0.7356411814689636,
+      "learning_rate": 0.0002996762489827763,
+      "loss": 4.7287,
+      "step": 84600
+    },
+    {
+      "epoch": 0.1188849878020293,
+      "grad_norm": 0.7238462567329407,
+      "learning_rate": 0.0002996757656940627,
+      "loss": 4.6479,
+      "step": 84610
+    },
+    {
+      "epoch": 0.11889903874019288,
+      "grad_norm": 0.7430039644241333,
+      "learning_rate": 0.0002996752820452867,
+      "loss": 4.7358,
+      "step": 84620
+    },
+    {
+      "epoch": 0.11891308967835647,
+      "grad_norm": 0.6926167011260986,
+      "learning_rate": 0.0002996747980364496,
+      "loss": 4.6281,
+      "step": 84630
+    },
+    {
+      "epoch": 0.11892714061652004,
+      "grad_norm": 0.7018167972564697,
+      "learning_rate": 0.00029967431366755237,
+      "loss": 4.7712,
+      "step": 84640
+    },
+    {
+      "epoch": 0.11894119155468362,
+      "grad_norm": 0.6989160180091858,
+      "learning_rate": 0.00029967382893859634,
+      "loss": 4.7982,
+      "step": 84650
+    },
+    {
+      "epoch": 0.1189552424928472,
+      "grad_norm": 0.7366704940795898,
+      "learning_rate": 0.0002996733438495826,
+      "loss": 4.7934,
+      "step": 84660
+    },
+    {
+      "epoch": 0.11896929343101077,
+      "grad_norm": 0.698725163936615,
+      "learning_rate": 0.00029967285840051236,
+      "loss": 4.7432,
+      "step": 84670
+    },
+    {
+      "epoch": 0.11898334436917435,
+      "grad_norm": 0.7053683400154114,
+      "learning_rate": 0.0002996723725913868,
+      "loss": 4.7379,
+      "step": 84680
+    },
+    {
+      "epoch": 0.11899739530733792,
+      "grad_norm": 0.7268297076225281,
+      "learning_rate": 0.000299671886422207,
+      "loss": 4.6208,
+      "step": 84690
+    },
+    {
+      "epoch": 0.1190114462455015,
+      "grad_norm": 0.7235490679740906,
+      "learning_rate": 0.0002996713998929742,
+      "loss": 4.6095,
+      "step": 84700
+    },
+    {
+      "epoch": 0.11902549718366508,
+      "grad_norm": 0.706933319568634,
+      "learning_rate": 0.00029967091300368956,
+      "loss": 4.6821,
+      "step": 84710
+    },
+    {
+      "epoch": 0.11903954812182865,
+      "grad_norm": 0.7260755300521851,
+      "learning_rate": 0.0002996704257543542,
+      "loss": 4.6393,
+      "step": 84720
+    },
+    {
+      "epoch": 0.11905359905999224,
+      "grad_norm": 0.7088666558265686,
+      "learning_rate": 0.00029966993814496944,
+      "loss": 4.7713,
+      "step": 84730
+    },
+    {
+      "epoch": 0.11906764999815582,
+      "grad_norm": 0.6888231635093689,
+      "learning_rate": 0.0002996694501755363,
+      "loss": 4.6284,
+      "step": 84740
+    },
+    {
+      "epoch": 0.1190817009363194,
+      "grad_norm": 0.7280553579330444,
+      "learning_rate": 0.000299668961846056,
+      "loss": 4.6492,
+      "step": 84750
+    },
+    {
+      "epoch": 0.11909575187448297,
+      "grad_norm": 0.7076670527458191,
+      "learning_rate": 0.00029966847315652975,
+      "loss": 4.7118,
+      "step": 84760
+    },
+    {
+      "epoch": 0.11910980281264655,
+      "grad_norm": 0.7475929856300354,
+      "learning_rate": 0.0002996679841069587,
+      "loss": 4.6426,
+      "step": 84770
+    },
+    {
+      "epoch": 0.11912385375081012,
+      "grad_norm": 0.6810241937637329,
+      "learning_rate": 0.00029966749469734395,
+      "loss": 4.7599,
+      "step": 84780
+    },
+    {
+      "epoch": 0.1191379046889737,
+      "grad_norm": 0.7051819562911987,
+      "learning_rate": 0.00029966700492768683,
+      "loss": 4.7097,
+      "step": 84790
+    },
+    {
+      "epoch": 0.11915195562713728,
+      "grad_norm": 0.7109501361846924,
+      "learning_rate": 0.0002996665147979884,
+      "loss": 4.625,
+      "step": 84800
+    },
+    {
+      "epoch": 0.11916600656530085,
+      "grad_norm": 0.6987961530685425,
+      "learning_rate": 0.0002996660243082499,
+      "loss": 4.6492,
+      "step": 84810
+    },
+    {
+      "epoch": 0.11918005750346443,
+      "grad_norm": 0.7196264863014221,
+      "learning_rate": 0.00029966553345847245,
+      "loss": 4.5642,
+      "step": 84820
+    },
+    {
+      "epoch": 0.11919410844162802,
+      "grad_norm": 0.6904036998748779,
+      "learning_rate": 0.0002996650422486573,
+      "loss": 4.7637,
+      "step": 84830
+    },
+    {
+      "epoch": 0.1192081593797916,
+      "grad_norm": 0.7117659449577332,
+      "learning_rate": 0.0002996645506788056,
+      "loss": 4.7709,
+      "step": 84840
+    },
+    {
+      "epoch": 0.11922221031795517,
+      "grad_norm": 0.7033788561820984,
+      "learning_rate": 0.0002996640587489185,
+      "loss": 4.654,
+      "step": 84850
+    },
+    {
+      "epoch": 0.11923626125611875,
+      "grad_norm": 0.710970938205719,
+      "learning_rate": 0.00029966356645899726,
+      "loss": 4.6868,
+      "step": 84860
+    },
+    {
+      "epoch": 0.11925031219428232,
+      "grad_norm": 0.7019822597503662,
+      "learning_rate": 0.00029966307380904295,
+      "loss": 4.6512,
+      "step": 84870
+    },
+    {
+      "epoch": 0.1192643631324459,
+      "grad_norm": 0.7262449860572815,
+      "learning_rate": 0.00029966258079905687,
+      "loss": 4.6789,
+      "step": 84880
+    },
+    {
+      "epoch": 0.11927841407060948,
+      "grad_norm": 0.7337659001350403,
+      "learning_rate": 0.0002996620874290402,
+      "loss": 4.7618,
+      "step": 84890
+    },
+    {
+      "epoch": 0.11929246500877305,
+      "grad_norm": 0.7070848941802979,
+      "learning_rate": 0.00029966159369899405,
+      "loss": 4.7406,
+      "step": 84900
+    },
+    {
+      "epoch": 0.11930651594693663,
+      "grad_norm": 0.7408710718154907,
+      "learning_rate": 0.00029966109960891964,
+      "loss": 4.8364,
+      "step": 84910
+    },
+    {
+      "epoch": 0.1193205668851002,
+      "grad_norm": 0.6921396851539612,
+      "learning_rate": 0.0002996606051588182,
+      "loss": 4.767,
+      "step": 84920
+    },
+    {
+      "epoch": 0.11933461782326378,
+      "grad_norm": 0.7204297780990601,
+      "learning_rate": 0.00029966011034869084,
+      "loss": 4.6362,
+      "step": 84930
+    },
+    {
+      "epoch": 0.11934866876142737,
+      "grad_norm": 0.7283856868743896,
+      "learning_rate": 0.00029965961517853883,
+      "loss": 4.6364,
+      "step": 84940
+    },
+    {
+      "epoch": 0.11936271969959095,
+      "grad_norm": 0.7073463201522827,
+      "learning_rate": 0.0002996591196483633,
+      "loss": 4.7698,
+      "step": 84950
+    },
+    {
+      "epoch": 0.11937677063775452,
+      "grad_norm": 0.7231873869895935,
+      "learning_rate": 0.00029965862375816546,
+      "loss": 4.7165,
+      "step": 84960
+    },
+    {
+      "epoch": 0.1193908215759181,
+      "grad_norm": 0.7474141716957092,
+      "learning_rate": 0.00029965812750794654,
+      "loss": 4.7594,
+      "step": 84970
+    },
+    {
+      "epoch": 0.11940487251408168,
+      "grad_norm": 0.752204179763794,
+      "learning_rate": 0.00029965763089770766,
+      "loss": 4.7098,
+      "step": 84980
+    },
+    {
+      "epoch": 0.11941892345224525,
+      "grad_norm": 0.6901721954345703,
+      "learning_rate": 0.0002996571339274501,
+      "loss": 4.5758,
+      "step": 84990
+    },
+    {
+      "epoch": 0.11943297439040883,
+      "grad_norm": 0.711621105670929,
+      "learning_rate": 0.000299656636597175,
+      "loss": 4.7077,
+      "step": 85000
+    },
+    {
+      "epoch": 0.1194470253285724,
+      "grad_norm": 0.7142300605773926,
+      "learning_rate": 0.0002996561389068836,
+      "loss": 4.7274,
+      "step": 85010
+    },
+    {
+      "epoch": 0.11946107626673598,
+      "grad_norm": 0.7306178212165833,
+      "learning_rate": 0.000299655640856577,
+      "loss": 4.7378,
+      "step": 85020
+    },
+    {
+      "epoch": 0.11947512720489956,
+      "grad_norm": 0.7129437923431396,
+      "learning_rate": 0.0002996551424462565,
+      "loss": 4.679,
+      "step": 85030
+    },
+    {
+      "epoch": 0.11948917814306315,
+      "grad_norm": 0.7050901055335999,
+      "learning_rate": 0.00029965464367592327,
+      "loss": 4.7632,
+      "step": 85040
+    },
+    {
+      "epoch": 0.11950322908122672,
+      "grad_norm": 0.6919633150100708,
+      "learning_rate": 0.0002996541445455785,
+      "loss": 4.629,
+      "step": 85050
+    },
+    {
+      "epoch": 0.1195172800193903,
+      "grad_norm": 0.705011785030365,
+      "learning_rate": 0.0002996536450552234,
+      "loss": 4.639,
+      "step": 85060
+    },
+    {
+      "epoch": 0.11953133095755387,
+      "grad_norm": 0.7281025648117065,
+      "learning_rate": 0.0002996531452048592,
+      "loss": 4.7754,
+      "step": 85070
+    },
+    {
+      "epoch": 0.11954538189571745,
+      "grad_norm": 0.7474626898765564,
+      "learning_rate": 0.00029965264499448704,
+      "loss": 4.6266,
+      "step": 85080
+    },
+    {
+      "epoch": 0.11955943283388103,
+      "grad_norm": 0.7219833731651306,
+      "learning_rate": 0.00029965214442410817,
+      "loss": 4.6425,
+      "step": 85090
+    },
+    {
+      "epoch": 0.1195734837720446,
+      "grad_norm": 0.7083914875984192,
+      "learning_rate": 0.0002996516434937237,
+      "loss": 4.7466,
+      "step": 85100
+    },
+    {
+      "epoch": 0.11958753471020818,
+      "grad_norm": 0.7148147225379944,
+      "learning_rate": 0.000299651142203335,
+      "loss": 4.5847,
+      "step": 85110
+    },
+    {
+      "epoch": 0.11960158564837176,
+      "grad_norm": 0.7254843711853027,
+      "learning_rate": 0.00029965064055294315,
+      "loss": 4.6418,
+      "step": 85120
+    },
+    {
+      "epoch": 0.11961563658653533,
+      "grad_norm": 0.7078420519828796,
+      "learning_rate": 0.0002996501385425494,
+      "loss": 4.6431,
+      "step": 85130
+    },
+    {
+      "epoch": 0.11962968752469892,
+      "grad_norm": 0.7228267192840576,
+      "learning_rate": 0.00029964963617215497,
+      "loss": 4.6071,
+      "step": 85140
+    },
+    {
+      "epoch": 0.1196437384628625,
+      "grad_norm": 0.700187623500824,
+      "learning_rate": 0.00029964913344176104,
+      "loss": 4.765,
+      "step": 85150
+    },
+    {
+      "epoch": 0.11965778940102607,
+      "grad_norm": 0.723201334476471,
+      "learning_rate": 0.00029964863035136883,
+      "loss": 4.7408,
+      "step": 85160
+    },
+    {
+      "epoch": 0.11967184033918965,
+      "grad_norm": 0.7586082220077515,
+      "learning_rate": 0.00029964812690097954,
+      "loss": 4.7016,
+      "step": 85170
+    },
+    {
+      "epoch": 0.11968589127735323,
+      "grad_norm": 0.7144497632980347,
+      "learning_rate": 0.0002996476230905944,
+      "loss": 4.6846,
+      "step": 85180
+    },
+    {
+      "epoch": 0.1196999422155168,
+      "grad_norm": 0.7355106472969055,
+      "learning_rate": 0.0002996471189202146,
+      "loss": 4.7114,
+      "step": 85190
+    },
+    {
+      "epoch": 0.11971399315368038,
+      "grad_norm": 0.7208760380744934,
+      "learning_rate": 0.0002996466143898414,
+      "loss": 4.6776,
+      "step": 85200
+    },
+    {
+      "epoch": 0.11972804409184395,
+      "grad_norm": 0.724746823310852,
+      "learning_rate": 0.00029964610949947593,
+      "loss": 4.8292,
+      "step": 85210
+    },
+    {
+      "epoch": 0.11974209503000753,
+      "grad_norm": 0.6953820586204529,
+      "learning_rate": 0.0002996456042491195,
+      "loss": 4.7256,
+      "step": 85220
+    },
+    {
+      "epoch": 0.11975614596817111,
+      "grad_norm": 0.7167063355445862,
+      "learning_rate": 0.0002996450986387733,
+      "loss": 4.6268,
+      "step": 85230
+    },
+    {
+      "epoch": 0.11977019690633468,
+      "grad_norm": 0.7081119418144226,
+      "learning_rate": 0.0002996445926684385,
+      "loss": 4.644,
+      "step": 85240
+    },
+    {
+      "epoch": 0.11978424784449827,
+      "grad_norm": 0.7091720700263977,
+      "learning_rate": 0.0002996440863381163,
+      "loss": 4.7266,
+      "step": 85250
+    },
+    {
+      "epoch": 0.11979829878266185,
+      "grad_norm": 0.7209839820861816,
+      "learning_rate": 0.000299643579647808,
+      "loss": 4.7454,
+      "step": 85260
+    },
+    {
+      "epoch": 0.11981234972082543,
+      "grad_norm": 0.6966519951820374,
+      "learning_rate": 0.00029964307259751474,
+      "loss": 4.6377,
+      "step": 85270
+    },
+    {
+      "epoch": 0.119826400658989,
+      "grad_norm": 0.701129674911499,
+      "learning_rate": 0.00029964256518723786,
+      "loss": 4.6073,
+      "step": 85280
+    },
+    {
+      "epoch": 0.11984045159715258,
+      "grad_norm": 0.6923808455467224,
+      "learning_rate": 0.0002996420574169785,
+      "loss": 4.6931,
+      "step": 85290
+    },
+    {
+      "epoch": 0.11985450253531615,
+      "grad_norm": 0.7123744487762451,
+      "learning_rate": 0.0002996415492867378,
+      "loss": 4.6669,
+      "step": 85300
+    },
+    {
+      "epoch": 0.11986855347347973,
+      "grad_norm": 0.7286718487739563,
+      "learning_rate": 0.0002996410407965171,
+      "loss": 4.6736,
+      "step": 85310
+    },
+    {
+      "epoch": 0.1198826044116433,
+      "grad_norm": 0.7242770791053772,
+      "learning_rate": 0.0002996405319463176,
+      "loss": 4.6789,
+      "step": 85320
+    },
+    {
+      "epoch": 0.11989665534980688,
+      "grad_norm": 0.7120243310928345,
+      "learning_rate": 0.0002996400227361405,
+      "loss": 4.6072,
+      "step": 85330
+    },
+    {
+      "epoch": 0.11991070628797046,
+      "grad_norm": 0.7126664519309998,
+      "learning_rate": 0.00029963951316598706,
+      "loss": 4.7034,
+      "step": 85340
+    },
+    {
+      "epoch": 0.11992475722613405,
+      "grad_norm": 0.7311067581176758,
+      "learning_rate": 0.00029963900323585845,
+      "loss": 4.7734,
+      "step": 85350
+    },
+    {
+      "epoch": 0.11993880816429763,
+      "grad_norm": 0.736724317073822,
+      "learning_rate": 0.00029963849294575596,
+      "loss": 4.6918,
+      "step": 85360
+    },
+    {
+      "epoch": 0.1199528591024612,
+      "grad_norm": 0.7146928310394287,
+      "learning_rate": 0.0002996379822956808,
+      "loss": 4.5878,
+      "step": 85370
+    },
+    {
+      "epoch": 0.11996691004062478,
+      "grad_norm": 0.6940208673477173,
+      "learning_rate": 0.00029963747128563417,
+      "loss": 4.7649,
+      "step": 85380
+    },
+    {
+      "epoch": 0.11998096097878835,
+      "grad_norm": 0.7303279638290405,
+      "learning_rate": 0.0002996369599156173,
+      "loss": 4.6806,
+      "step": 85390
+    },
+    {
+      "epoch": 0.11999501191695193,
+      "grad_norm": 0.7200376391410828,
+      "learning_rate": 0.00029963644818563145,
+      "loss": 4.7124,
+      "step": 85400
+    },
+    {
+      "epoch": 0.1200090628551155,
+      "grad_norm": 0.6997661590576172,
+      "learning_rate": 0.00029963593609567783,
+      "loss": 4.7171,
+      "step": 85410
+    },
+    {
+      "epoch": 0.12002311379327908,
+      "grad_norm": 0.7147961854934692,
+      "learning_rate": 0.00029963542364575766,
+      "loss": 4.5216,
+      "step": 85420
+    },
+    {
+      "epoch": 0.12003716473144266,
+      "grad_norm": 0.7196703553199768,
+      "learning_rate": 0.0002996349108358722,
+      "loss": 4.7912,
+      "step": 85430
+    },
+    {
+      "epoch": 0.12005121566960623,
+      "grad_norm": 0.7229997515678406,
+      "learning_rate": 0.0002996343976660227,
+      "loss": 4.6595,
+      "step": 85440
+    },
+    {
+      "epoch": 0.12006526660776982,
+      "grad_norm": 0.7053361535072327,
+      "learning_rate": 0.00029963388413621037,
+      "loss": 4.6721,
+      "step": 85450
+    },
+    {
+      "epoch": 0.1200793175459334,
+      "grad_norm": 0.7070173621177673,
+      "learning_rate": 0.00029963337024643637,
+      "loss": 4.6046,
+      "step": 85460
+    },
+    {
+      "epoch": 0.12009336848409698,
+      "grad_norm": 0.6948744654655457,
+      "learning_rate": 0.0002996328559967021,
+      "loss": 4.7345,
+      "step": 85470
+    },
+    {
+      "epoch": 0.12010741942226055,
+      "grad_norm": 0.7141450643539429,
+      "learning_rate": 0.00029963234138700865,
+      "loss": 4.6867,
+      "step": 85480
+    },
+    {
+      "epoch": 0.12012147036042413,
+      "grad_norm": 0.7317810654640198,
+      "learning_rate": 0.0002996318264173573,
+      "loss": 4.5998,
+      "step": 85490
+    },
+    {
+      "epoch": 0.1201355212985877,
+      "grad_norm": 0.7293758392333984,
+      "learning_rate": 0.0002996313110877494,
+      "loss": 4.7669,
+      "step": 85500
+    },
+    {
+      "epoch": 0.12014957223675128,
+      "grad_norm": 0.7065700888633728,
+      "learning_rate": 0.000299630795398186,
+      "loss": 4.6155,
+      "step": 85510
+    },
+    {
+      "epoch": 0.12016362317491486,
+      "grad_norm": 0.7263256907463074,
+      "learning_rate": 0.0002996302793486685,
+      "loss": 4.6669,
+      "step": 85520
+    },
+    {
+      "epoch": 0.12017767411307843,
+      "grad_norm": 0.7085826396942139,
+      "learning_rate": 0.000299629762939198,
+      "loss": 4.6789,
+      "step": 85530
+    },
+    {
+      "epoch": 0.12019172505124201,
+      "grad_norm": 0.7257848381996155,
+      "learning_rate": 0.0002996292461697759,
+      "loss": 4.7586,
+      "step": 85540
+    },
+    {
+      "epoch": 0.12020577598940559,
+      "grad_norm": 0.7008543610572815,
+      "learning_rate": 0.0002996287290404033,
+      "loss": 4.7479,
+      "step": 85550
+    },
+    {
+      "epoch": 0.12021982692756918,
+      "grad_norm": 0.6934198141098022,
+      "learning_rate": 0.00029962821155108153,
+      "loss": 4.7428,
+      "step": 85560
+    },
+    {
+      "epoch": 0.12023387786573275,
+      "grad_norm": 0.6932348012924194,
+      "learning_rate": 0.0002996276937018119,
+      "loss": 4.701,
+      "step": 85570
+    },
+    {
+      "epoch": 0.12024792880389633,
+      "grad_norm": 0.6966341733932495,
+      "learning_rate": 0.0002996271754925954,
+      "loss": 4.7036,
+      "step": 85580
+    },
+    {
+      "epoch": 0.1202619797420599,
+      "grad_norm": 0.7221991419792175,
+      "learning_rate": 0.00029962665692343356,
+      "loss": 4.6009,
+      "step": 85590
+    },
+    {
+      "epoch": 0.12027603068022348,
+      "grad_norm": 0.7265183925628662,
+      "learning_rate": 0.00029962613799432744,
+      "loss": 4.6958,
+      "step": 85600
+    },
+    {
+      "epoch": 0.12029008161838706,
+      "grad_norm": 0.7104468941688538,
+      "learning_rate": 0.0002996256187052784,
+      "loss": 4.6442,
+      "step": 85610
+    },
+    {
+      "epoch": 0.12030413255655063,
+      "grad_norm": 0.725537896156311,
+      "learning_rate": 0.0002996250990562877,
+      "loss": 4.7405,
+      "step": 85620
+    },
+    {
+      "epoch": 0.12031818349471421,
+      "grad_norm": 0.7155194282531738,
+      "learning_rate": 0.0002996245790473564,
+      "loss": 4.7107,
+      "step": 85630
+    },
+    {
+      "epoch": 0.12033223443287779,
+      "grad_norm": 0.7233233451843262,
+      "learning_rate": 0.00029962405867848597,
+      "loss": 4.6326,
+      "step": 85640
+    },
+    {
+      "epoch": 0.12034628537104136,
+      "grad_norm": 0.7190825939178467,
+      "learning_rate": 0.0002996235379496776,
+      "loss": 4.7711,
+      "step": 85650
+    },
+    {
+      "epoch": 0.12036033630920495,
+      "grad_norm": 0.7331857681274414,
+      "learning_rate": 0.00029962301686093246,
+      "loss": 4.7669,
+      "step": 85660
+    },
+    {
+      "epoch": 0.12037438724736853,
+      "grad_norm": 0.7344217896461487,
+      "learning_rate": 0.00029962249541225195,
+      "loss": 4.7563,
+      "step": 85670
+    },
+    {
+      "epoch": 0.1203884381855321,
+      "grad_norm": 0.706200122833252,
+      "learning_rate": 0.0002996219736036372,
+      "loss": 4.7661,
+      "step": 85680
+    },
+    {
+      "epoch": 0.12040248912369568,
+      "grad_norm": 0.7147502899169922,
+      "learning_rate": 0.0002996214514350895,
+      "loss": 4.5101,
+      "step": 85690
+    },
+    {
+      "epoch": 0.12041654006185926,
+      "grad_norm": 0.7082406878471375,
+      "learning_rate": 0.0002996209289066101,
+      "loss": 4.6323,
+      "step": 85700
+    },
+    {
+      "epoch": 0.12043059100002283,
+      "grad_norm": 0.7128416299819946,
+      "learning_rate": 0.0002996204060182003,
+      "loss": 4.6708,
+      "step": 85710
+    },
+    {
+      "epoch": 0.12044464193818641,
+      "grad_norm": 0.7045454382896423,
+      "learning_rate": 0.0002996198827698613,
+      "loss": 4.6902,
+      "step": 85720
+    },
+    {
+      "epoch": 0.12045869287634998,
+      "grad_norm": 0.7640816569328308,
+      "learning_rate": 0.0002996193591615944,
+      "loss": 4.7689,
+      "step": 85730
+    },
+    {
+      "epoch": 0.12047274381451356,
+      "grad_norm": 0.7236823439598083,
+      "learning_rate": 0.00029961883519340085,
+      "loss": 4.7993,
+      "step": 85740
+    },
+    {
+      "epoch": 0.12048679475267714,
+      "grad_norm": 0.7274718880653381,
+      "learning_rate": 0.0002996183108652819,
+      "loss": 4.7037,
+      "step": 85750
+    },
+    {
+      "epoch": 0.12050084569084073,
+      "grad_norm": 0.7061973810195923,
+      "learning_rate": 0.00029961778617723883,
+      "loss": 4.6896,
+      "step": 85760
+    },
+    {
+      "epoch": 0.1205148966290043,
+      "grad_norm": 0.7118090391159058,
+      "learning_rate": 0.00029961726112927284,
+      "loss": 4.6545,
+      "step": 85770
+    },
+    {
+      "epoch": 0.12052894756716788,
+      "grad_norm": 0.6875846982002258,
+      "learning_rate": 0.00029961673572138534,
+      "loss": 4.6925,
+      "step": 85780
+    },
+    {
+      "epoch": 0.12054299850533146,
+      "grad_norm": 0.7023685574531555,
+      "learning_rate": 0.00029961620995357747,
+      "loss": 4.6063,
+      "step": 85790
+    },
+    {
+      "epoch": 0.12055704944349503,
+      "grad_norm": 0.7327830195426941,
+      "learning_rate": 0.0002996156838258505,
+      "loss": 4.7115,
+      "step": 85800
+    },
+    {
+      "epoch": 0.12057110038165861,
+      "grad_norm": 0.7080143690109253,
+      "learning_rate": 0.00029961515733820565,
+      "loss": 4.7993,
+      "step": 85810
+    },
+    {
+      "epoch": 0.12058515131982218,
+      "grad_norm": 0.7189124822616577,
+      "learning_rate": 0.00029961463049064436,
+      "loss": 4.6521,
+      "step": 85820
+    },
+    {
+      "epoch": 0.12059920225798576,
+      "grad_norm": 0.7159058451652527,
+      "learning_rate": 0.0002996141032831677,
+      "loss": 4.798,
+      "step": 85830
+    },
+    {
+      "epoch": 0.12061325319614934,
+      "grad_norm": 0.7728943228721619,
+      "learning_rate": 0.0002996135757157771,
+      "loss": 4.6605,
+      "step": 85840
+    },
+    {
+      "epoch": 0.12062730413431291,
+      "grad_norm": 0.9222893118858337,
+      "learning_rate": 0.0002996130477884737,
+      "loss": 4.6658,
+      "step": 85850
+    },
+    {
+      "epoch": 0.12064135507247649,
+      "grad_norm": 0.7004067301750183,
+      "learning_rate": 0.00029961251950125884,
+      "loss": 4.7422,
+      "step": 85860
+    },
+    {
+      "epoch": 0.12065540601064008,
+      "grad_norm": 0.7140782475471497,
+      "learning_rate": 0.00029961199085413385,
+      "loss": 4.6833,
+      "step": 85870
+    },
+    {
+      "epoch": 0.12066945694880366,
+      "grad_norm": 0.7055903077125549,
+      "learning_rate": 0.00029961146184709987,
+      "loss": 4.7995,
+      "step": 85880
+    },
+    {
+      "epoch": 0.12068350788696723,
+      "grad_norm": 0.7646137475967407,
+      "learning_rate": 0.00029961093248015826,
+      "loss": 4.7584,
+      "step": 85890
+    },
+    {
+      "epoch": 0.12069755882513081,
+      "grad_norm": 0.710187554359436,
+      "learning_rate": 0.0002996104027533102,
+      "loss": 4.5475,
+      "step": 85900
+    },
+    {
+      "epoch": 0.12071160976329438,
+      "grad_norm": 0.7333691120147705,
+      "learning_rate": 0.00029960987266655707,
+      "loss": 4.7007,
+      "step": 85910
+    },
+    {
+      "epoch": 0.12072566070145796,
+      "grad_norm": 0.7043735384941101,
+      "learning_rate": 0.00029960934221990015,
+      "loss": 4.7965,
+      "step": 85920
+    },
+    {
+      "epoch": 0.12073971163962154,
+      "grad_norm": 0.7109283804893494,
+      "learning_rate": 0.00029960881141334063,
+      "loss": 4.6432,
+      "step": 85930
+    },
+    {
+      "epoch": 0.12075376257778511,
+      "grad_norm": 0.6858183145523071,
+      "learning_rate": 0.00029960828024687983,
+      "loss": 4.7364,
+      "step": 85940
+    },
+    {
+      "epoch": 0.12076781351594869,
+      "grad_norm": 0.7360689043998718,
+      "learning_rate": 0.0002996077487205191,
+      "loss": 4.5901,
+      "step": 85950
+    },
+    {
+      "epoch": 0.12078186445411226,
+      "grad_norm": 0.6898038387298584,
+      "learning_rate": 0.0002996072700390809,
+      "loss": 4.6589,
+      "step": 85960
+    },
+    {
+      "epoch": 0.12079591539227585,
+      "grad_norm": 1.0662989616394043,
+      "learning_rate": 0.00029960673782891364,
+      "loss": 4.5829,
+      "step": 85970
+    },
+    {
+      "epoch": 0.12080996633043943,
+      "grad_norm": 0.7156269550323486,
+      "learning_rate": 0.0002996062052588501,
+      "loss": 4.722,
+      "step": 85980
+    },
+    {
+      "epoch": 0.120824017268603,
+      "grad_norm": 0.7196338176727295,
+      "learning_rate": 0.0002996056723288915,
+      "loss": 4.655,
+      "step": 85990
+    },
+    {
+      "epoch": 0.12083806820676658,
+      "grad_norm": 0.6985352039337158,
+      "learning_rate": 0.0002996051390390392,
+      "loss": 4.7494,
+      "step": 86000
+    },
+    {
+      "epoch": 0.12085211914493016,
+      "grad_norm": 0.7045134902000427,
+      "learning_rate": 0.0002996046053892945,
+      "loss": 4.7251,
+      "step": 86010
+    },
+    {
+      "epoch": 0.12086617008309374,
+      "grad_norm": 0.7500072121620178,
+      "learning_rate": 0.00029960407137965856,
+      "loss": 4.732,
+      "step": 86020
+    },
+    {
+      "epoch": 0.12088022102125731,
+      "grad_norm": 0.7060306668281555,
+      "learning_rate": 0.0002996035370101327,
+      "loss": 4.7745,
+      "step": 86030
+    },
+    {
+      "epoch": 0.12089427195942089,
+      "grad_norm": 0.7595226764678955,
+      "learning_rate": 0.0002996030022807184,
+      "loss": 4.6405,
+      "step": 86040
+    },
+    {
+      "epoch": 0.12090832289758446,
+      "grad_norm": 0.696519136428833,
+      "learning_rate": 0.0002996024671914167,
+      "loss": 4.724,
+      "step": 86050
+    },
+    {
+      "epoch": 0.12092237383574804,
+      "grad_norm": 0.7316704988479614,
+      "learning_rate": 0.000299601931742229,
+      "loss": 4.5628,
+      "step": 86060
+    },
+    {
+      "epoch": 0.12093642477391163,
+      "grad_norm": 0.6876169443130493,
+      "learning_rate": 0.0002996013959331566,
+      "loss": 4.7434,
+      "step": 86070
+    },
+    {
+      "epoch": 0.1209504757120752,
+      "grad_norm": 0.7389705777168274,
+      "learning_rate": 0.00029960085976420066,
+      "loss": 4.7681,
+      "step": 86080
+    },
+    {
+      "epoch": 0.12096452665023878,
+      "grad_norm": 0.702412486076355,
+      "learning_rate": 0.00029960032323536265,
+      "loss": 4.7209,
+      "step": 86090
+    },
+    {
+      "epoch": 0.12097857758840236,
+      "grad_norm": 0.6958481669425964,
+      "learning_rate": 0.0002995997863466437,
+      "loss": 4.8133,
+      "step": 86100
+    },
+    {
+      "epoch": 0.12099262852656593,
+      "grad_norm": 0.717162013053894,
+      "learning_rate": 0.00029959924909804525,
+      "loss": 4.7312,
+      "step": 86110
+    },
+    {
+      "epoch": 0.12100667946472951,
+      "grad_norm": 0.7261740565299988,
+      "learning_rate": 0.00029959871148956853,
+      "loss": 4.6485,
+      "step": 86120
+    },
+    {
+      "epoch": 0.12102073040289309,
+      "grad_norm": 0.7102240324020386,
+      "learning_rate": 0.00029959817352121485,
+      "loss": 4.8627,
+      "step": 86130
+    },
+    {
+      "epoch": 0.12103478134105666,
+      "grad_norm": 0.7106133699417114,
+      "learning_rate": 0.0002995976351929854,
+      "loss": 4.7682,
+      "step": 86140
+    },
+    {
+      "epoch": 0.12104883227922024,
+      "grad_norm": 0.6954272389411926,
+      "learning_rate": 0.00029959709650488164,
+      "loss": 4.6691,
+      "step": 86150
+    },
+    {
+      "epoch": 0.12106288321738382,
+      "grad_norm": 0.7176921367645264,
+      "learning_rate": 0.0002995965574569047,
+      "loss": 4.6487,
+      "step": 86160
+    },
+    {
+      "epoch": 0.12107693415554739,
+      "grad_norm": 0.8005315661430359,
+      "learning_rate": 0.000299596018049056,
+      "loss": 4.735,
+      "step": 86170
+    },
+    {
+      "epoch": 0.12109098509371098,
+      "grad_norm": 0.7331059575080872,
+      "learning_rate": 0.0002995954782813368,
+      "loss": 4.7956,
+      "step": 86180
+    },
+    {
+      "epoch": 0.12110503603187456,
+      "grad_norm": 0.7373118996620178,
+      "learning_rate": 0.0002995949381537484,
+      "loss": 4.584,
+      "step": 86190
+    },
+    {
+      "epoch": 0.12111908697003813,
+      "grad_norm": 0.7049098610877991,
+      "learning_rate": 0.00029959439766629207,
+      "loss": 4.6658,
+      "step": 86200
+    },
+    {
+      "epoch": 0.12113313790820171,
+      "grad_norm": 0.7197480797767639,
+      "learning_rate": 0.00029959385681896914,
+      "loss": 4.8936,
+      "step": 86210
+    },
+    {
+      "epoch": 0.12114718884636529,
+      "grad_norm": 0.7270209193229675,
+      "learning_rate": 0.00029959331561178087,
+      "loss": 4.6427,
+      "step": 86220
+    },
+    {
+      "epoch": 0.12116123978452886,
+      "grad_norm": 0.7451220750808716,
+      "learning_rate": 0.0002995927740447287,
+      "loss": 4.7832,
+      "step": 86230
+    },
+    {
+      "epoch": 0.12117529072269244,
+      "grad_norm": 0.7173677682876587,
+      "learning_rate": 0.00029959223211781374,
+      "loss": 4.6468,
+      "step": 86240
+    },
+    {
+      "epoch": 0.12118934166085601,
+      "grad_norm": 0.7185055613517761,
+      "learning_rate": 0.0002995916898310374,
+      "loss": 4.7216,
+      "step": 86250
+    },
+    {
+      "epoch": 0.12120339259901959,
+      "grad_norm": 0.729455828666687,
+      "learning_rate": 0.00029959114718440094,
+      "loss": 4.7083,
+      "step": 86260
+    },
+    {
+      "epoch": 0.12121744353718317,
+      "grad_norm": 0.70658940076828,
+      "learning_rate": 0.0002995906041779057,
+      "loss": 4.6727,
+      "step": 86270
+    },
+    {
+      "epoch": 0.12123149447534676,
+      "grad_norm": 0.7007445693016052,
+      "learning_rate": 0.00029959006081155303,
+      "loss": 4.7122,
+      "step": 86280
+    },
+    {
+      "epoch": 0.12124554541351033,
+      "grad_norm": 0.7127397656440735,
+      "learning_rate": 0.00029958951708534417,
+      "loss": 4.6473,
+      "step": 86290
+    },
+    {
+      "epoch": 0.12125959635167391,
+      "grad_norm": 0.7260735630989075,
+      "learning_rate": 0.0002995889729992805,
+      "loss": 4.6155,
+      "step": 86300
+    },
+    {
+      "epoch": 0.12127364728983749,
+      "grad_norm": 0.7390902638435364,
+      "learning_rate": 0.00029958842855336314,
+      "loss": 4.6914,
+      "step": 86310
+    },
+    {
+      "epoch": 0.12128769822800106,
+      "grad_norm": 0.7290669083595276,
+      "learning_rate": 0.00029958788374759364,
+      "loss": 4.6071,
+      "step": 86320
+    },
+    {
+      "epoch": 0.12130174916616464,
+      "grad_norm": 0.7232752442359924,
+      "learning_rate": 0.00029958733858197314,
+      "loss": 4.5553,
+      "step": 86330
+    },
+    {
+      "epoch": 0.12131580010432821,
+      "grad_norm": 0.7251487970352173,
+      "learning_rate": 0.0002995867930565031,
+      "loss": 4.6294,
+      "step": 86340
+    },
+    {
+      "epoch": 0.12132985104249179,
+      "grad_norm": 0.7375115156173706,
+      "learning_rate": 0.00029958624717118463,
+      "loss": 4.7586,
+      "step": 86350
+    },
+    {
+      "epoch": 0.12134390198065537,
+      "grad_norm": 0.7383313775062561,
+      "learning_rate": 0.0002995857009260192,
+      "loss": 4.7191,
+      "step": 86360
+    },
+    {
+      "epoch": 0.12135795291881894,
+      "grad_norm": 0.730225145816803,
+      "learning_rate": 0.0002995851543210081,
+      "loss": 4.6134,
+      "step": 86370
+    },
+    {
+      "epoch": 0.12137200385698253,
+      "grad_norm": 1.1462152004241943,
+      "learning_rate": 0.00029958460735615266,
+      "loss": 4.6346,
+      "step": 86380
+    },
+    {
+      "epoch": 0.12138605479514611,
+      "grad_norm": 0.7488716840744019,
+      "learning_rate": 0.00029958406003145413,
+      "loss": 4.7212,
+      "step": 86390
+    },
+    {
+      "epoch": 0.12140010573330969,
+      "grad_norm": 0.7324610352516174,
+      "learning_rate": 0.0002995835123469139,
+      "loss": 4.6655,
+      "step": 86400
+    },
+    {
+      "epoch": 0.12141415667147326,
+      "grad_norm": 0.7316756248474121,
+      "learning_rate": 0.0002995829643025332,
+      "loss": 4.5674,
+      "step": 86410
+    },
+    {
+      "epoch": 0.12142820760963684,
+      "grad_norm": 0.7131539583206177,
+      "learning_rate": 0.0002995824158983134,
+      "loss": 4.7464,
+      "step": 86420
+    },
+    {
+      "epoch": 0.12144225854780041,
+      "grad_norm": 0.7157332897186279,
+      "learning_rate": 0.0002995818671342559,
+      "loss": 4.7049,
+      "step": 86430
+    },
+    {
+      "epoch": 0.12145630948596399,
+      "grad_norm": 0.7550666332244873,
+      "learning_rate": 0.0002995813180103618,
+      "loss": 4.5843,
+      "step": 86440
+    },
+    {
+      "epoch": 0.12147036042412757,
+      "grad_norm": 0.7143164277076721,
+      "learning_rate": 0.00029958076852663266,
+      "loss": 4.6484,
+      "step": 86450
+    },
+    {
+      "epoch": 0.12148441136229114,
+      "grad_norm": 0.7037560343742371,
+      "learning_rate": 0.0002995802186830697,
+      "loss": 4.806,
+      "step": 86460
+    },
+    {
+      "epoch": 0.12149846230045472,
+      "grad_norm": 0.710610032081604,
+      "learning_rate": 0.0002995796684796742,
+      "loss": 4.7,
+      "step": 86470
+    },
+    {
+      "epoch": 0.1215125132386183,
+      "grad_norm": 0.7024252414703369,
+      "learning_rate": 0.0002995791179164475,
+      "loss": 4.6566,
+      "step": 86480
+    },
+    {
+      "epoch": 0.12152656417678188,
+      "grad_norm": 0.7062261700630188,
+      "learning_rate": 0.00029957856699339104,
+      "loss": 4.69,
+      "step": 86490
+    },
+    {
+      "epoch": 0.12154061511494546,
+      "grad_norm": 0.7243465185165405,
+      "learning_rate": 0.00029957801571050597,
+      "loss": 4.7999,
+      "step": 86500
+    },
+    {
+      "epoch": 0.12155466605310904,
+      "grad_norm": 0.7086811661720276,
+      "learning_rate": 0.0002995774640677937,
+      "loss": 4.6722,
+      "step": 86510
+    },
+    {
+      "epoch": 0.12156871699127261,
+      "grad_norm": 0.7045823931694031,
+      "learning_rate": 0.0002995769672817016,
+      "loss": 4.8183,
+      "step": 86520
+    },
+    {
+      "epoch": 0.12158276792943619,
+      "grad_norm": 0.7064098119735718,
+      "learning_rate": 0.0002995764149553213,
+      "loss": 4.7609,
+      "step": 86530
+    },
+    {
+      "epoch": 0.12159681886759977,
+      "grad_norm": 0.7484224438667297,
+      "learning_rate": 0.0002995758622691177,
+      "loss": 4.681,
+      "step": 86540
+    },
+    {
+      "epoch": 0.12161086980576334,
+      "grad_norm": 0.6937485933303833,
+      "learning_rate": 0.0002995753092230921,
+      "loss": 4.7115,
+      "step": 86550
+    },
+    {
+      "epoch": 0.12162492074392692,
+      "grad_norm": 0.7272822856903076,
+      "learning_rate": 0.00029957475581724586,
+      "loss": 4.5993,
+      "step": 86560
+    },
+    {
+      "epoch": 0.1216389716820905,
+      "grad_norm": 0.696310818195343,
+      "learning_rate": 0.0002995742020515802,
+      "loss": 4.7079,
+      "step": 86570
+    },
+    {
+      "epoch": 0.12165302262025407,
+      "grad_norm": 0.7192110419273376,
+      "learning_rate": 0.0002995736479260965,
+      "loss": 4.7158,
+      "step": 86580
+    },
+    {
+      "epoch": 0.12166707355841766,
+      "grad_norm": 0.7095620632171631,
+      "learning_rate": 0.0002995730934407961,
+      "loss": 4.6992,
+      "step": 86590
+    },
+    {
+      "epoch": 0.12168112449658124,
+      "grad_norm": 0.6891617178916931,
+      "learning_rate": 0.0002995725385956804,
+      "loss": 4.7227,
+      "step": 86600
+    },
+    {
+      "epoch": 0.12169517543474481,
+      "grad_norm": 0.7383580207824707,
+      "learning_rate": 0.0002995719833907506,
+      "loss": 4.7221,
+      "step": 86610
+    },
+    {
+      "epoch": 0.12170922637290839,
+      "grad_norm": 0.7053825259208679,
+      "learning_rate": 0.0002995714278260082,
+      "loss": 4.6793,
+      "step": 86620
+    },
+    {
+      "epoch": 0.12172327731107196,
+      "grad_norm": 0.7410792708396912,
+      "learning_rate": 0.0002995708719014544,
+      "loss": 4.769,
+      "step": 86630
+    },
+    {
+      "epoch": 0.12173732824923554,
+      "grad_norm": 0.711789071559906,
+      "learning_rate": 0.0002995703156170906,
+      "loss": 4.7196,
+      "step": 86640
+    },
+    {
+      "epoch": 0.12175137918739912,
+      "grad_norm": 0.6985942125320435,
+      "learning_rate": 0.0002995697589729181,
+      "loss": 4.6477,
+      "step": 86650
+    },
+    {
+      "epoch": 0.1217654301255627,
+      "grad_norm": 0.6995700001716614,
+      "learning_rate": 0.00029956920196893834,
+      "loss": 4.5672,
+      "step": 86660
+    },
+    {
+      "epoch": 0.12177948106372627,
+      "grad_norm": 0.7413903474807739,
+      "learning_rate": 0.0002995686446051525,
+      "loss": 4.7654,
+      "step": 86670
+    },
+    {
+      "epoch": 0.12179353200188985,
+      "grad_norm": 0.7320995926856995,
+      "learning_rate": 0.00029956808688156206,
+      "loss": 4.7567,
+      "step": 86680
+    },
+    {
+      "epoch": 0.12180758294005344,
+      "grad_norm": 0.7222046852111816,
+      "learning_rate": 0.0002995675287981683,
+      "loss": 4.6457,
+      "step": 86690
+    },
+    {
+      "epoch": 0.12182163387821701,
+      "grad_norm": 0.711199939250946,
+      "learning_rate": 0.00029956697035497256,
+      "loss": 4.6399,
+      "step": 86700
+    },
+    {
+      "epoch": 0.12183568481638059,
+      "grad_norm": 0.7224550247192383,
+      "learning_rate": 0.00029956641155197617,
+      "loss": 4.6731,
+      "step": 86710
+    },
+    {
+      "epoch": 0.12184973575454416,
+      "grad_norm": 0.6902201175689697,
+      "learning_rate": 0.00029956585238918057,
+      "loss": 4.6802,
+      "step": 86720
+    },
+    {
+      "epoch": 0.12186378669270774,
+      "grad_norm": 0.7383785247802734,
+      "learning_rate": 0.00029956529286658695,
+      "loss": 4.7232,
+      "step": 86730
+    },
+    {
+      "epoch": 0.12187783763087132,
+      "grad_norm": 0.7981114387512207,
+      "learning_rate": 0.00029956473298419677,
+      "loss": 4.7792,
+      "step": 86740
+    },
+    {
+      "epoch": 0.12189188856903489,
+      "grad_norm": 0.7090650200843811,
+      "learning_rate": 0.00029956417274201134,
+      "loss": 4.7199,
+      "step": 86750
+    },
+    {
+      "epoch": 0.12190593950719847,
+      "grad_norm": 0.7152296900749207,
+      "learning_rate": 0.00029956361214003205,
+      "loss": 4.6483,
+      "step": 86760
+    },
+    {
+      "epoch": 0.12191999044536204,
+      "grad_norm": 0.7130936980247498,
+      "learning_rate": 0.0002995630511782602,
+      "loss": 4.6469,
+      "step": 86770
+    },
+    {
+      "epoch": 0.12193404138352562,
+      "grad_norm": 0.7285060286521912,
+      "learning_rate": 0.0002995624898566971,
+      "loss": 4.7693,
+      "step": 86780
+    },
+    {
+      "epoch": 0.1219480923216892,
+      "grad_norm": 0.7463589906692505,
+      "learning_rate": 0.0002995619281753442,
+      "loss": 4.8208,
+      "step": 86790
+    },
+    {
+      "epoch": 0.12196214325985279,
+      "grad_norm": 0.7308253049850464,
+      "learning_rate": 0.00029956136613420276,
+      "loss": 4.7281,
+      "step": 86800
+    },
+    {
+      "epoch": 0.12197619419801636,
+      "grad_norm": 0.7179209589958191,
+      "learning_rate": 0.0002995608037332742,
+      "loss": 4.672,
+      "step": 86810
+    },
+    {
+      "epoch": 0.12199024513617994,
+      "grad_norm": 0.731372594833374,
+      "learning_rate": 0.00029956024097255987,
+      "loss": 4.8004,
+      "step": 86820
+    },
+    {
+      "epoch": 0.12200429607434352,
+      "grad_norm": 0.7444266080856323,
+      "learning_rate": 0.00029955967785206107,
+      "loss": 4.5729,
+      "step": 86830
+    },
+    {
+      "epoch": 0.12201834701250709,
+      "grad_norm": 0.7302293181419373,
+      "learning_rate": 0.0002995591143717792,
+      "loss": 4.662,
+      "step": 86840
+    },
+    {
+      "epoch": 0.12203239795067067,
+      "grad_norm": 0.7154730558395386,
+      "learning_rate": 0.0002995585505317156,
+      "loss": 4.5486,
+      "step": 86850
+    },
+    {
+      "epoch": 0.12204644888883424,
+      "grad_norm": 0.701775074005127,
+      "learning_rate": 0.00029955798633187164,
+      "loss": 4.5835,
+      "step": 86860
+    },
+    {
+      "epoch": 0.12206049982699782,
+      "grad_norm": 0.6990428566932678,
+      "learning_rate": 0.0002995574217722487,
+      "loss": 4.6875,
+      "step": 86870
+    },
+    {
+      "epoch": 0.1220745507651614,
+      "grad_norm": 0.7048460245132446,
+      "learning_rate": 0.000299556856852848,
+      "loss": 4.6615,
+      "step": 86880
+    },
+    {
+      "epoch": 0.12208860170332497,
+      "grad_norm": 0.7063455581665039,
+      "learning_rate": 0.00029955629157367105,
+      "loss": 4.8373,
+      "step": 86890
+    },
+    {
+      "epoch": 0.12210265264148856,
+      "grad_norm": 0.7515528202056885,
+      "learning_rate": 0.0002995557259347192,
+      "loss": 4.7006,
+      "step": 86900
+    },
+    {
+      "epoch": 0.12211670357965214,
+      "grad_norm": 0.7269037365913391,
+      "learning_rate": 0.0002995551599359937,
+      "loss": 4.7502,
+      "step": 86910
+    },
+    {
+      "epoch": 0.12213075451781572,
+      "grad_norm": 0.7405303716659546,
+      "learning_rate": 0.0002995545935774961,
+      "loss": 4.6159,
+      "step": 86920
+    },
+    {
+      "epoch": 0.12214480545597929,
+      "grad_norm": 0.7235696315765381,
+      "learning_rate": 0.00029955402685922757,
+      "loss": 4.7358,
+      "step": 86930
+    },
+    {
+      "epoch": 0.12215885639414287,
+      "grad_norm": 0.711112916469574,
+      "learning_rate": 0.0002995534597811895,
+      "loss": 4.7783,
+      "step": 86940
+    },
+    {
+      "epoch": 0.12217290733230644,
+      "grad_norm": 0.7069167494773865,
+      "learning_rate": 0.0002995528923433834,
+      "loss": 4.4935,
+      "step": 86950
+    },
+    {
+      "epoch": 0.12218695827047002,
+      "grad_norm": 0.7162838578224182,
+      "learning_rate": 0.0002995523245458105,
+      "loss": 4.5865,
+      "step": 86960
+    },
+    {
+      "epoch": 0.1222010092086336,
+      "grad_norm": 0.7079212069511414,
+      "learning_rate": 0.0002995517563884722,
+      "loss": 4.6902,
+      "step": 86970
+    },
+    {
+      "epoch": 0.12221506014679717,
+      "grad_norm": 0.7136911749839783,
+      "learning_rate": 0.0002995511878713699,
+      "loss": 4.7081,
+      "step": 86980
+    },
+    {
+      "epoch": 0.12222911108496075,
+      "grad_norm": 0.7026124000549316,
+      "learning_rate": 0.0002995506189945049,
+      "loss": 4.7789,
+      "step": 86990
+    },
+    {
+      "epoch": 0.12224316202312434,
+      "grad_norm": 0.8226521611213684,
+      "learning_rate": 0.00029955004975787863,
+      "loss": 4.6564,
+      "step": 87000
+    },
+    {
+      "epoch": 0.12225721296128791,
+      "grad_norm": 0.7236669659614563,
+      "learning_rate": 0.0002995494801614924,
+      "loss": 4.785,
+      "step": 87010
+    },
+    {
+      "epoch": 0.12227126389945149,
+      "grad_norm": 0.7194319367408752,
+      "learning_rate": 0.0002995489102053477,
+      "loss": 4.7189,
+      "step": 87020
+    },
+    {
+      "epoch": 0.12228531483761507,
+      "grad_norm": 0.710686445236206,
+      "learning_rate": 0.00029954833988944576,
+      "loss": 4.5753,
+      "step": 87030
+    },
+    {
+      "epoch": 0.12229936577577864,
+      "grad_norm": 0.7050729393959045,
+      "learning_rate": 0.000299547769213788,
+      "loss": 4.4799,
+      "step": 87040
+    },
+    {
+      "epoch": 0.12231341671394222,
+      "grad_norm": 0.7316957712173462,
+      "learning_rate": 0.00029954719817837583,
+      "loss": 4.772,
+      "step": 87050
+    },
+    {
+      "epoch": 0.1223274676521058,
+      "grad_norm": 0.6963942050933838,
+      "learning_rate": 0.0002995466267832106,
+      "loss": 4.8665,
+      "step": 87060
+    },
+    {
+      "epoch": 0.12234151859026937,
+      "grad_norm": 0.7359170913696289,
+      "learning_rate": 0.00029954605502829364,
+      "loss": 4.6619,
+      "step": 87070
+    },
+    {
+      "epoch": 0.12235556952843295,
+      "grad_norm": 0.6984687447547913,
+      "learning_rate": 0.0002995454829136264,
+      "loss": 4.7002,
+      "step": 87080
+    },
+    {
+      "epoch": 0.12236962046659652,
+      "grad_norm": 0.7410992383956909,
+      "learning_rate": 0.00029954491043921025,
+      "loss": 4.7953,
+      "step": 87090
+    },
+    {
+      "epoch": 0.1223836714047601,
+      "grad_norm": 0.7628210186958313,
+      "learning_rate": 0.0002995443376050465,
+      "loss": 4.5949,
+      "step": 87100
+    },
+    {
+      "epoch": 0.12239772234292369,
+      "grad_norm": 0.7284057140350342,
+      "learning_rate": 0.0002995437644111366,
+      "loss": 4.6736,
+      "step": 87110
+    },
+    {
+      "epoch": 0.12241177328108727,
+      "grad_norm": 0.706147313117981,
+      "learning_rate": 0.0002995431908574819,
+      "loss": 4.7252,
+      "step": 87120
+    },
+    {
+      "epoch": 0.12242582421925084,
+      "grad_norm": 0.7244818806648254,
+      "learning_rate": 0.00029954261694408374,
+      "loss": 4.6662,
+      "step": 87130
+    },
+    {
+      "epoch": 0.12243987515741442,
+      "grad_norm": 0.7074465155601501,
+      "learning_rate": 0.0002995420426709436,
+      "loss": 4.7113,
+      "step": 87140
+    },
+    {
+      "epoch": 0.122453926095578,
+      "grad_norm": 0.7343944311141968,
+      "learning_rate": 0.0002995414680380627,
+      "loss": 4.6968,
+      "step": 87150
+    },
+    {
+      "epoch": 0.12246797703374157,
+      "grad_norm": 0.7110148668289185,
+      "learning_rate": 0.0002995408930454426,
+      "loss": 4.7478,
+      "step": 87160
+    },
+    {
+      "epoch": 0.12248202797190515,
+      "grad_norm": 0.6801456809043884,
+      "learning_rate": 0.0002995403176930846,
+      "loss": 4.6153,
+      "step": 87170
+    },
+    {
+      "epoch": 0.12249607891006872,
+      "grad_norm": 0.7084956169128418,
+      "learning_rate": 0.0002995397419809901,
+      "loss": 4.7813,
+      "step": 87180
+    },
+    {
+      "epoch": 0.1225101298482323,
+      "grad_norm": 0.7134016156196594,
+      "learning_rate": 0.0002995391659091604,
+      "loss": 4.7072,
+      "step": 87190
+    },
+    {
+      "epoch": 0.12252418078639588,
+      "grad_norm": 0.7109416723251343,
+      "learning_rate": 0.000299538589477597,
+      "loss": 4.6814,
+      "step": 87200
+    },
+    {
+      "epoch": 0.12253823172455947,
+      "grad_norm": 0.7339860200881958,
+      "learning_rate": 0.00029953801268630125,
+      "loss": 4.7426,
+      "step": 87210
+    },
+    {
+      "epoch": 0.12255228266272304,
+      "grad_norm": 0.7281706929206848,
+      "learning_rate": 0.00029953743553527457,
+      "loss": 4.6708,
+      "step": 87220
+    },
+    {
+      "epoch": 0.12256633360088662,
+      "grad_norm": 0.7278706431388855,
+      "learning_rate": 0.0002995368580245183,
+      "loss": 4.6032,
+      "step": 87230
+    },
+    {
+      "epoch": 0.1225803845390502,
+      "grad_norm": 0.6926674842834473,
+      "learning_rate": 0.00029953628015403375,
+      "loss": 4.7332,
+      "step": 87240
+    },
+    {
+      "epoch": 0.12259443547721377,
+      "grad_norm": 0.6893281936645508,
+      "learning_rate": 0.00029953570192382253,
+      "loss": 4.6901,
+      "step": 87250
+    },
+    {
+      "epoch": 0.12260848641537735,
+      "grad_norm": 0.7695411443710327,
+      "learning_rate": 0.0002995351233338858,
+      "loss": 4.6268,
+      "step": 87260
+    },
+    {
+      "epoch": 0.12262253735354092,
+      "grad_norm": 0.7140504121780396,
+      "learning_rate": 0.00029953454438422515,
+      "loss": 4.6299,
+      "step": 87270
+    },
+    {
+      "epoch": 0.1226365882917045,
+      "grad_norm": 0.7018819451332092,
+      "learning_rate": 0.0002995339650748418,
+      "loss": 4.7092,
+      "step": 87280
+    },
+    {
+      "epoch": 0.12265063922986807,
+      "grad_norm": 0.7678191065788269,
+      "learning_rate": 0.00029953338540573727,
+      "loss": 4.7711,
+      "step": 87290
+    },
+    {
+      "epoch": 0.12266469016803165,
+      "grad_norm": 0.7389617562294006,
+      "learning_rate": 0.00029953280537691287,
+      "loss": 4.7441,
+      "step": 87300
+    },
+    {
+      "epoch": 0.12267874110619524,
+      "grad_norm": 0.7068891525268555,
+      "learning_rate": 0.00029953222498837003,
+      "loss": 4.6663,
+      "step": 87310
+    },
+    {
+      "epoch": 0.12269279204435882,
+      "grad_norm": 0.7148218750953674,
+      "learning_rate": 0.00029953164424011017,
+      "loss": 4.7392,
+      "step": 87320
+    },
+    {
+      "epoch": 0.1227068429825224,
+      "grad_norm": 0.68925541639328,
+      "learning_rate": 0.0002995310631321347,
+      "loss": 4.5429,
+      "step": 87330
+    },
+    {
+      "epoch": 0.12272089392068597,
+      "grad_norm": 0.7171045541763306,
+      "learning_rate": 0.00029953048166444493,
+      "loss": 4.7119,
+      "step": 87340
+    },
+    {
+      "epoch": 0.12273494485884955,
+      "grad_norm": 0.7084298133850098,
+      "learning_rate": 0.0002995298998370423,
+      "loss": 4.7834,
+      "step": 87350
+    },
+    {
+      "epoch": 0.12274899579701312,
+      "grad_norm": 0.7399872541427612,
+      "learning_rate": 0.00029952931764992826,
+      "loss": 4.6349,
+      "step": 87360
+    },
+    {
+      "epoch": 0.1227630467351767,
+      "grad_norm": 0.7395424246788025,
+      "learning_rate": 0.00029952873510310417,
+      "loss": 4.6944,
+      "step": 87370
+    },
+    {
+      "epoch": 0.12277709767334027,
+      "grad_norm": 0.7185593843460083,
+      "learning_rate": 0.00029952815219657145,
+      "loss": 4.6208,
+      "step": 87380
+    },
+    {
+      "epoch": 0.12279114861150385,
+      "grad_norm": 0.7182762622833252,
+      "learning_rate": 0.0002995275689303314,
+      "loss": 4.536,
+      "step": 87390
+    },
+    {
+      "epoch": 0.12280519954966743,
+      "grad_norm": 0.7268804311752319,
+      "learning_rate": 0.0002995269853043856,
+      "loss": 4.6515,
+      "step": 87400
+    },
+    {
+      "epoch": 0.122819250487831,
+      "grad_norm": 0.7242562174797058,
+      "learning_rate": 0.0002995264013187353,
+      "loss": 4.7664,
+      "step": 87410
+    },
+    {
+      "epoch": 0.12283330142599459,
+      "grad_norm": 0.7390630841255188,
+      "learning_rate": 0.000299525816973382,
+      "loss": 4.7246,
+      "step": 87420
+    },
+    {
+      "epoch": 0.12284735236415817,
+      "grad_norm": 0.7184797525405884,
+      "learning_rate": 0.00029952523226832705,
+      "loss": 4.6634,
+      "step": 87430
+    },
+    {
+      "epoch": 0.12286140330232175,
+      "grad_norm": 0.7052783370018005,
+      "learning_rate": 0.0002995246472035719,
+      "loss": 4.6488,
+      "step": 87440
+    },
+    {
+      "epoch": 0.12287545424048532,
+      "grad_norm": 0.7244933247566223,
+      "learning_rate": 0.0002995240617791179,
+      "loss": 4.655,
+      "step": 87450
+    },
+    {
+      "epoch": 0.1228895051786489,
+      "grad_norm": 0.6974210739135742,
+      "learning_rate": 0.00029952347599496653,
+      "loss": 4.6001,
+      "step": 87460
+    },
+    {
+      "epoch": 0.12290355611681247,
+      "grad_norm": 0.7236213088035583,
+      "learning_rate": 0.00029952288985111915,
+      "loss": 4.7134,
+      "step": 87470
+    },
+    {
+      "epoch": 0.12291760705497605,
+      "grad_norm": 0.7080331444740295,
+      "learning_rate": 0.0002995223033475772,
+      "loss": 4.6978,
+      "step": 87480
+    },
+    {
+      "epoch": 0.12293165799313963,
+      "grad_norm": 0.7316338419914246,
+      "learning_rate": 0.00029952171648434204,
+      "loss": 4.601,
+      "step": 87490
+    },
+    {
+      "epoch": 0.1229457089313032,
+      "grad_norm": 0.687455415725708,
+      "learning_rate": 0.0002995211292614151,
+      "loss": 4.5823,
+      "step": 87500
+    },
+    {
+      "epoch": 0.12295975986946678,
+      "grad_norm": 0.7364456653594971,
+      "learning_rate": 0.0002995205416787979,
+      "loss": 4.6627,
+      "step": 87510
+    },
+    {
+      "epoch": 0.12297381080763037,
+      "grad_norm": 0.7389670014381409,
+      "learning_rate": 0.00029951995373649174,
+      "loss": 4.6651,
+      "step": 87520
+    },
+    {
+      "epoch": 0.12298786174579394,
+      "grad_norm": 0.6907218098640442,
+      "learning_rate": 0.000299519365434498,
+      "loss": 4.6806,
+      "step": 87530
+    },
+    {
+      "epoch": 0.12300191268395752,
+      "grad_norm": 0.7170994281768799,
+      "learning_rate": 0.0002995187767728182,
+      "loss": 4.6415,
+      "step": 87540
+    },
+    {
+      "epoch": 0.1230159636221211,
+      "grad_norm": 0.7091425061225891,
+      "learning_rate": 0.0002995181877514537,
+      "loss": 4.669,
+      "step": 87550
+    },
+    {
+      "epoch": 0.12303001456028467,
+      "grad_norm": 0.7479362487792969,
+      "learning_rate": 0.0002995175983704059,
+      "loss": 4.6114,
+      "step": 87560
+    },
+    {
+      "epoch": 0.12304406549844825,
+      "grad_norm": 0.7046723961830139,
+      "learning_rate": 0.00029951700862967625,
+      "loss": 4.6076,
+      "step": 87570
+    },
+    {
+      "epoch": 0.12305811643661183,
+      "grad_norm": 0.7411142587661743,
+      "learning_rate": 0.0002995164185292662,
+      "loss": 4.6606,
+      "step": 87580
+    },
+    {
+      "epoch": 0.1230721673747754,
+      "grad_norm": 0.7122248411178589,
+      "learning_rate": 0.00029951582806917707,
+      "loss": 4.8045,
+      "step": 87590
+    },
+    {
+      "epoch": 0.12308621831293898,
+      "grad_norm": 0.7476603984832764,
+      "learning_rate": 0.00029951523724941037,
+      "loss": 4.6634,
+      "step": 87600
+    },
+    {
+      "epoch": 0.12310026925110255,
+      "grad_norm": 0.7404193878173828,
+      "learning_rate": 0.0002995146460699675,
+      "loss": 4.6486,
+      "step": 87610
+    },
+    {
+      "epoch": 0.12311432018926614,
+      "grad_norm": 0.707002580165863,
+      "learning_rate": 0.0002995140545308499,
+      "loss": 4.7272,
+      "step": 87620
+    },
+    {
+      "epoch": 0.12312837112742972,
+      "grad_norm": 0.762656569480896,
+      "learning_rate": 0.0002995134626320589,
+      "loss": 4.6558,
+      "step": 87630
+    },
+    {
+      "epoch": 0.1231424220655933,
+      "grad_norm": 0.7162299752235413,
+      "learning_rate": 0.00029951287037359605,
+      "loss": 4.6148,
+      "step": 87640
+    },
+    {
+      "epoch": 0.12315647300375687,
+      "grad_norm": 0.7232511043548584,
+      "learning_rate": 0.0002995122777554627,
+      "loss": 4.6304,
+      "step": 87650
+    },
+    {
+      "epoch": 0.12317052394192045,
+      "grad_norm": 0.7200028896331787,
+      "learning_rate": 0.00029951168477766033,
+      "loss": 4.7234,
+      "step": 87660
+    },
+    {
+      "epoch": 0.12318457488008402,
+      "grad_norm": 0.705317497253418,
+      "learning_rate": 0.00029951109144019024,
+      "loss": 4.6568,
+      "step": 87670
+    },
+    {
+      "epoch": 0.1231986258182476,
+      "grad_norm": 0.7253167033195496,
+      "learning_rate": 0.000299510497743054,
+      "loss": 4.7056,
+      "step": 87680
+    },
+    {
+      "epoch": 0.12321267675641118,
+      "grad_norm": 0.7323938012123108,
+      "learning_rate": 0.000299509903686253,
+      "loss": 4.6481,
+      "step": 87690
+    },
+    {
+      "epoch": 0.12322672769457475,
+      "grad_norm": 0.6936154961585999,
+      "learning_rate": 0.00029950930926978865,
+      "loss": 4.7263,
+      "step": 87700
+    },
+    {
+      "epoch": 0.12324077863273833,
+      "grad_norm": 0.7268804311752319,
+      "learning_rate": 0.0002995087144936624,
+      "loss": 4.6468,
+      "step": 87710
+    },
+    {
+      "epoch": 0.1232548295709019,
+      "grad_norm": 0.7117732167243958,
+      "learning_rate": 0.0002995081193578756,
+      "loss": 4.7268,
+      "step": 87720
+    },
+    {
+      "epoch": 0.1232688805090655,
+      "grad_norm": 0.7191809415817261,
+      "learning_rate": 0.0002995075238624298,
+      "loss": 4.7574,
+      "step": 87730
+    },
+    {
+      "epoch": 0.12328293144722907,
+      "grad_norm": 0.7095280885696411,
+      "learning_rate": 0.0002995069280073263,
+      "loss": 4.689,
+      "step": 87740
+    },
+    {
+      "epoch": 0.12329698238539265,
+      "grad_norm": 0.7109695076942444,
+      "learning_rate": 0.0002995063317925667,
+      "loss": 4.6289,
+      "step": 87750
+    },
+    {
+      "epoch": 0.12331103332355622,
+      "grad_norm": 0.7158311605453491,
+      "learning_rate": 0.0002995057352181523,
+      "loss": 4.6884,
+      "step": 87760
+    },
+    {
+      "epoch": 0.1233250842617198,
+      "grad_norm": 0.7213734984397888,
+      "learning_rate": 0.00029950513828408464,
+      "loss": 4.8162,
+      "step": 87770
+    },
+    {
+      "epoch": 0.12333913519988338,
+      "grad_norm": 0.7377204895019531,
+      "learning_rate": 0.00029950454099036505,
+      "loss": 4.6634,
+      "step": 87780
+    },
+    {
+      "epoch": 0.12335318613804695,
+      "grad_norm": 0.7066074013710022,
+      "learning_rate": 0.000299503943336995,
+      "loss": 4.6969,
+      "step": 87790
+    },
+    {
+      "epoch": 0.12336723707621053,
+      "grad_norm": 0.7790802121162415,
+      "learning_rate": 0.00029950334532397597,
+      "loss": 4.7552,
+      "step": 87800
+    },
+    {
+      "epoch": 0.1233812880143741,
+      "grad_norm": 0.7286075949668884,
+      "learning_rate": 0.0002995027469513094,
+      "loss": 4.6512,
+      "step": 87810
+    },
+    {
+      "epoch": 0.12339533895253768,
+      "grad_norm": 0.7531540989875793,
+      "learning_rate": 0.00029950214821899664,
+      "loss": 4.7612,
+      "step": 87820
+    },
+    {
+      "epoch": 0.12340938989070127,
+      "grad_norm": 0.7376166582107544,
+      "learning_rate": 0.00029950154912703923,
+      "loss": 4.6591,
+      "step": 87830
+    },
+    {
+      "epoch": 0.12342344082886485,
+      "grad_norm": 0.7482818365097046,
+      "learning_rate": 0.0002995009496754385,
+      "loss": 4.5681,
+      "step": 87840
+    },
+    {
+      "epoch": 0.12343749176702842,
+      "grad_norm": 0.7085266709327698,
+      "learning_rate": 0.00029950034986419606,
+      "loss": 4.7676,
+      "step": 87850
+    },
+    {
+      "epoch": 0.123451542705192,
+      "grad_norm": 0.7395931482315063,
+      "learning_rate": 0.0002994997496933132,
+      "loss": 4.6275,
+      "step": 87860
+    },
+    {
+      "epoch": 0.12346559364335558,
+      "grad_norm": 0.7464209198951721,
+      "learning_rate": 0.0002994991491627914,
+      "loss": 4.746,
+      "step": 87870
+    },
+    {
+      "epoch": 0.12347964458151915,
+      "grad_norm": 0.7421188950538635,
+      "learning_rate": 0.0002994985482726322,
+      "loss": 4.6564,
+      "step": 87880
+    },
+    {
+      "epoch": 0.12349369551968273,
+      "grad_norm": 0.7147485017776489,
+      "learning_rate": 0.0002994979470228369,
+      "loss": 4.6841,
+      "step": 87890
+    },
+    {
+      "epoch": 0.1235077464578463,
+      "grad_norm": 0.7064347267150879,
+      "learning_rate": 0.0002994973454134071,
+      "loss": 4.6705,
+      "step": 87900
+    },
+    {
+      "epoch": 0.12352179739600988,
+      "grad_norm": 0.6962974071502686,
+      "learning_rate": 0.00029949674344434407,
+      "loss": 4.5298,
+      "step": 87910
+    },
+    {
+      "epoch": 0.12353584833417346,
+      "grad_norm": 0.7260435223579407,
+      "learning_rate": 0.0002994961411156494,
+      "loss": 4.8508,
+      "step": 87920
+    },
+    {
+      "epoch": 0.12354989927233705,
+      "grad_norm": 0.7157175540924072,
+      "learning_rate": 0.00029949553842732447,
+      "loss": 4.5779,
+      "step": 87930
+    },
+    {
+      "epoch": 0.12356395021050062,
+      "grad_norm": 0.7105233669281006,
+      "learning_rate": 0.0002994949353793708,
+      "loss": 4.6836,
+      "step": 87940
+    },
+    {
+      "epoch": 0.1235780011486642,
+      "grad_norm": 0.7243243455886841,
+      "learning_rate": 0.0002994943319717897,
+      "loss": 4.6184,
+      "step": 87950
+    },
+    {
+      "epoch": 0.12359205208682777,
+      "grad_norm": 0.7264546751976013,
+      "learning_rate": 0.0002994937282045828,
+      "loss": 4.6754,
+      "step": 87960
+    },
+    {
+      "epoch": 0.12360610302499135,
+      "grad_norm": 0.7493976354598999,
+      "learning_rate": 0.00029949312407775145,
+      "loss": 4.709,
+      "step": 87970
+    },
+    {
+      "epoch": 0.12362015396315493,
+      "grad_norm": 0.7562281489372253,
+      "learning_rate": 0.0002994925195912971,
+      "loss": 4.6625,
+      "step": 87980
+    },
+    {
+      "epoch": 0.1236342049013185,
+      "grad_norm": 0.7009782195091248,
+      "learning_rate": 0.00029949191474522123,
+      "loss": 4.6659,
+      "step": 87990
+    },
+    {
+      "epoch": 0.12364825583948208,
+      "grad_norm": 0.7240691781044006,
+      "learning_rate": 0.0002994913095395253,
+      "loss": 4.6298,
+      "step": 88000
+    },
+    {
+      "epoch": 0.12366230677764566,
+      "grad_norm": 0.7109857797622681,
+      "learning_rate": 0.0002994907039742107,
+      "loss": 4.829,
+      "step": 88010
+    },
+    {
+      "epoch": 0.12367635771580923,
+      "grad_norm": 0.734104335308075,
+      "learning_rate": 0.000299490098049279,
+      "loss": 4.7854,
+      "step": 88020
+    },
+    {
+      "epoch": 0.12369040865397281,
+      "grad_norm": 0.707338273525238,
+      "learning_rate": 0.00029948949176473153,
+      "loss": 4.4639,
+      "step": 88030
+    },
+    {
+      "epoch": 0.1237044595921364,
+      "grad_norm": 0.7034321427345276,
+      "learning_rate": 0.0002994888851205699,
+      "loss": 4.5313,
+      "step": 88040
+    },
+    {
+      "epoch": 0.12371851053029997,
+      "grad_norm": 0.7377339005470276,
+      "learning_rate": 0.0002994882781167955,
+      "loss": 4.6773,
+      "step": 88050
+    },
+    {
+      "epoch": 0.12373256146846355,
+      "grad_norm": 0.7189347147941589,
+      "learning_rate": 0.0002994876707534097,
+      "loss": 4.7458,
+      "step": 88060
+    },
+    {
+      "epoch": 0.12374661240662713,
+      "grad_norm": 0.7128537893295288,
+      "learning_rate": 0.00029948706303041403,
+      "loss": 4.7733,
+      "step": 88070
+    },
+    {
+      "epoch": 0.1237606633447907,
+      "grad_norm": 0.7069502472877502,
+      "learning_rate": 0.00029948645494781,
+      "loss": 4.6594,
+      "step": 88080
+    },
+    {
+      "epoch": 0.12377471428295428,
+      "grad_norm": 0.7175171375274658,
+      "learning_rate": 0.00029948584650559907,
+      "loss": 4.8268,
+      "step": 88090
+    },
+    {
+      "epoch": 0.12378876522111786,
+      "grad_norm": 0.7087175846099854,
+      "learning_rate": 0.0002994852377037826,
+      "loss": 4.7501,
+      "step": 88100
+    },
+    {
+      "epoch": 0.12380281615928143,
+      "grad_norm": 0.7500482201576233,
+      "learning_rate": 0.00029948462854236213,
+      "loss": 4.6788,
+      "step": 88110
+    },
+    {
+      "epoch": 0.12381686709744501,
+      "grad_norm": 0.7136823534965515,
+      "learning_rate": 0.00029948401902133917,
+      "loss": 4.6901,
+      "step": 88120
+    },
+    {
+      "epoch": 0.12383091803560858,
+      "grad_norm": 0.7292098999023438,
+      "learning_rate": 0.0002994834091407151,
+      "loss": 4.7064,
+      "step": 88130
+    },
+    {
+      "epoch": 0.12384496897377217,
+      "grad_norm": 0.698124349117279,
+      "learning_rate": 0.0002994827989004914,
+      "loss": 4.8037,
+      "step": 88140
+    },
+    {
+      "epoch": 0.12385901991193575,
+      "grad_norm": 0.7261769771575928,
+      "learning_rate": 0.00029948218830066955,
+      "loss": 4.5345,
+      "step": 88150
+    },
+    {
+      "epoch": 0.12387307085009933,
+      "grad_norm": 0.706026554107666,
+      "learning_rate": 0.000299481577341251,
+      "loss": 4.6582,
+      "step": 88160
+    },
+    {
+      "epoch": 0.1238871217882629,
+      "grad_norm": 0.7198874354362488,
+      "learning_rate": 0.0002994809660222373,
+      "loss": 4.7463,
+      "step": 88170
+    },
+    {
+      "epoch": 0.12390117272642648,
+      "grad_norm": 0.7321640849113464,
+      "learning_rate": 0.0002994803543436299,
+      "loss": 4.6111,
+      "step": 88180
+    },
+    {
+      "epoch": 0.12391522366459005,
+      "grad_norm": 0.7120465040206909,
+      "learning_rate": 0.0002994797423054302,
+      "loss": 4.7361,
+      "step": 88190
+    },
+    {
+      "epoch": 0.12392927460275363,
+      "grad_norm": 0.718183159828186,
+      "learning_rate": 0.0002994791299076397,
+      "loss": 4.6667,
+      "step": 88200
+    },
+    {
+      "epoch": 0.1239433255409172,
+      "grad_norm": 0.7051864266395569,
+      "learning_rate": 0.0002994785171502599,
+      "loss": 4.7235,
+      "step": 88210
+    },
+    {
+      "epoch": 0.12395737647908078,
+      "grad_norm": 0.71971595287323,
+      "learning_rate": 0.0002994779040332922,
+      "loss": 4.7439,
+      "step": 88220
+    },
+    {
+      "epoch": 0.12397142741724436,
+      "grad_norm": 0.6985116004943848,
+      "learning_rate": 0.0002994772905567382,
+      "loss": 4.6536,
+      "step": 88230
+    },
+    {
+      "epoch": 0.12398547835540795,
+      "grad_norm": 0.7011168003082275,
+      "learning_rate": 0.0002994766767205993,
+      "loss": 4.7449,
+      "step": 88240
+    },
+    {
+      "epoch": 0.12399952929357153,
+      "grad_norm": 0.7223268747329712,
+      "learning_rate": 0.00029947606252487695,
+      "loss": 4.7109,
+      "step": 88250
+    },
+    {
+      "epoch": 0.1240135802317351,
+      "grad_norm": 0.7336310744285583,
+      "learning_rate": 0.0002994754479695727,
+      "loss": 4.5966,
+      "step": 88260
+    },
+    {
+      "epoch": 0.12402763116989868,
+      "grad_norm": 0.7206735014915466,
+      "learning_rate": 0.000299474833054688,
+      "loss": 4.5686,
+      "step": 88270
+    },
+    {
+      "epoch": 0.12404168210806225,
+      "grad_norm": 0.7099548578262329,
+      "learning_rate": 0.0002994742177802243,
+      "loss": 4.7035,
+      "step": 88280
+    },
+    {
+      "epoch": 0.12405573304622583,
+      "grad_norm": 0.7359626293182373,
+      "learning_rate": 0.0002994736021461831,
+      "loss": 4.6886,
+      "step": 88290
+    },
+    {
+      "epoch": 0.1240697839843894,
+      "grad_norm": 0.7080744504928589,
+      "learning_rate": 0.00029947298615256587,
+      "loss": 4.7505,
+      "step": 88300
+    },
+    {
+      "epoch": 0.12408383492255298,
+      "grad_norm": 0.7350786924362183,
+      "learning_rate": 0.00029947236979937413,
+      "loss": 4.5652,
+      "step": 88310
+    },
+    {
+      "epoch": 0.12409788586071656,
+      "grad_norm": 1.0457093715667725,
+      "learning_rate": 0.0002994717530866093,
+      "loss": 4.7684,
+      "step": 88320
+    },
+    {
+      "epoch": 0.12411193679888013,
+      "grad_norm": 0.7474514245986938,
+      "learning_rate": 0.00029947113601427286,
+      "loss": 4.7522,
+      "step": 88330
+    },
+    {
+      "epoch": 0.12412598773704371,
+      "grad_norm": 0.7411333322525024,
+      "learning_rate": 0.0002994705185823664,
+      "loss": 4.7131,
+      "step": 88340
+    },
+    {
+      "epoch": 0.1241400386752073,
+      "grad_norm": 0.7307577133178711,
+      "learning_rate": 0.0002994699007908914,
+      "loss": 4.5768,
+      "step": 88350
+    },
+    {
+      "epoch": 0.12415408961337088,
+      "grad_norm": 0.7115901708602905,
+      "learning_rate": 0.00029946928263984915,
+      "loss": 4.6929,
+      "step": 88360
+    },
+    {
+      "epoch": 0.12416814055153445,
+      "grad_norm": 0.7410994172096252,
+      "learning_rate": 0.00029946866412924135,
+      "loss": 4.7275,
+      "step": 88370
+    },
+    {
+      "epoch": 0.12418219148969803,
+      "grad_norm": 0.7032224535942078,
+      "learning_rate": 0.0002994680452590694,
+      "loss": 4.7629,
+      "step": 88380
+    },
+    {
+      "epoch": 0.1241962424278616,
+      "grad_norm": 0.6944802403450012,
+      "learning_rate": 0.00029946742602933484,
+      "loss": 4.7391,
+      "step": 88390
+    },
+    {
+      "epoch": 0.12421029336602518,
+      "grad_norm": 0.7015830874443054,
+      "learning_rate": 0.000299466806440039,
+      "loss": 4.5912,
+      "step": 88400
+    },
+    {
+      "epoch": 0.12422434430418876,
+      "grad_norm": 0.7076965570449829,
+      "learning_rate": 0.0002994661864911836,
+      "loss": 4.7955,
+      "step": 88410
+    },
+    {
+      "epoch": 0.12423839524235233,
+      "grad_norm": 0.7263289093971252,
+      "learning_rate": 0.00029946556618277,
+      "loss": 4.7882,
+      "step": 88420
+    },
+    {
+      "epoch": 0.12425244618051591,
+      "grad_norm": 0.7141406536102295,
+      "learning_rate": 0.0002994649455147997,
+      "loss": 4.7778,
+      "step": 88430
+    },
+    {
+      "epoch": 0.12426649711867949,
+      "grad_norm": 0.7139621376991272,
+      "learning_rate": 0.00029946432448727417,
+      "loss": 4.7299,
+      "step": 88440
+    },
+    {
+      "epoch": 0.12428054805684308,
+      "grad_norm": 0.7191937565803528,
+      "learning_rate": 0.000299463703100195,
+      "loss": 4.6505,
+      "step": 88450
+    },
+    {
+      "epoch": 0.12429459899500665,
+      "grad_norm": 0.7122807502746582,
+      "learning_rate": 0.00029946308135356356,
+      "loss": 4.7084,
+      "step": 88460
+    },
+    {
+      "epoch": 0.12430864993317023,
+      "grad_norm": 0.707273006439209,
+      "learning_rate": 0.00029946245924738146,
+      "loss": 4.6373,
+      "step": 88470
+    },
+    {
+      "epoch": 0.1243227008713338,
+      "grad_norm": 0.6925734281539917,
+      "learning_rate": 0.00029946183678165015,
+      "loss": 4.7202,
+      "step": 88480
+    },
+    {
+      "epoch": 0.12433675180949738,
+      "grad_norm": 0.6993895769119263,
+      "learning_rate": 0.0002994612139563711,
+      "loss": 4.7537,
+      "step": 88490
+    },
+    {
+      "epoch": 0.12435080274766096,
+      "grad_norm": 0.7005565166473389,
+      "learning_rate": 0.00029946059077154587,
+      "loss": 4.7321,
+      "step": 88500
+    },
+    {
+      "epoch": 0.12436485368582453,
+      "grad_norm": 0.7049751281738281,
+      "learning_rate": 0.00029945996722717587,
+      "loss": 4.7177,
+      "step": 88510
+    },
+    {
+      "epoch": 0.12437890462398811,
+      "grad_norm": 0.7042818069458008,
+      "learning_rate": 0.00029945934332326264,
+      "loss": 4.5704,
+      "step": 88520
+    },
+    {
+      "epoch": 0.12439295556215169,
+      "grad_norm": 0.6895237565040588,
+      "learning_rate": 0.00029945871905980774,
+      "loss": 4.6741,
+      "step": 88530
+    },
+    {
+      "epoch": 0.12440700650031526,
+      "grad_norm": 0.7173465490341187,
+      "learning_rate": 0.0002994580944368126,
+      "loss": 4.7058,
+      "step": 88540
+    },
+    {
+      "epoch": 0.12442105743847885,
+      "grad_norm": 0.7089383006095886,
+      "learning_rate": 0.00029945746945427877,
+      "loss": 4.7225,
+      "step": 88550
+    },
+    {
+      "epoch": 0.12443510837664243,
+      "grad_norm": 0.7033636569976807,
+      "learning_rate": 0.00029945684411220773,
+      "loss": 4.7401,
+      "step": 88560
+    },
+    {
+      "epoch": 0.124449159314806,
+      "grad_norm": 0.7240288257598877,
+      "learning_rate": 0.00029945621841060094,
+      "loss": 4.7217,
+      "step": 88570
+    },
+    {
+      "epoch": 0.12446321025296958,
+      "grad_norm": 0.7067498564720154,
+      "learning_rate": 0.00029945559234945997,
+      "loss": 4.6413,
+      "step": 88580
+    },
+    {
+      "epoch": 0.12447726119113316,
+      "grad_norm": 0.7195616364479065,
+      "learning_rate": 0.00029945496592878633,
+      "loss": 4.7053,
+      "step": 88590
+    },
+    {
+      "epoch": 0.12449131212929673,
+      "grad_norm": 0.7112637162208557,
+      "learning_rate": 0.0002994543391485815,
+      "loss": 4.6635,
+      "step": 88600
+    },
+    {
+      "epoch": 0.12450536306746031,
+      "grad_norm": 0.6900596618652344,
+      "learning_rate": 0.000299453712008847,
+      "loss": 4.7249,
+      "step": 88610
+    },
+    {
+      "epoch": 0.12451941400562389,
+      "grad_norm": 0.6978746652603149,
+      "learning_rate": 0.00029945308450958426,
+      "loss": 4.6273,
+      "step": 88620
+    },
+    {
+      "epoch": 0.12453346494378746,
+      "grad_norm": 0.707542359828949,
+      "learning_rate": 0.00029945245665079494,
+      "loss": 4.648,
+      "step": 88630
+    },
+    {
+      "epoch": 0.12454751588195104,
+      "grad_norm": 0.7132956981658936,
+      "learning_rate": 0.00029945182843248044,
+      "loss": 4.6548,
+      "step": 88640
+    },
+    {
+      "epoch": 0.12456156682011461,
+      "grad_norm": 0.7130081057548523,
+      "learning_rate": 0.0002994511998546423,
+      "loss": 4.666,
+      "step": 88650
+    },
+    {
+      "epoch": 0.1245756177582782,
+      "grad_norm": 0.7389416098594666,
+      "learning_rate": 0.000299450570917282,
+      "loss": 4.6872,
+      "step": 88660
+    },
+    {
+      "epoch": 0.12458966869644178,
+      "grad_norm": 0.7738654613494873,
+      "learning_rate": 0.00029944994162040114,
+      "loss": 4.6537,
+      "step": 88670
+    },
+    {
+      "epoch": 0.12460371963460536,
+      "grad_norm": 0.7259090542793274,
+      "learning_rate": 0.00029944931196400115,
+      "loss": 4.6315,
+      "step": 88680
+    },
+    {
+      "epoch": 0.12461777057276893,
+      "grad_norm": 0.7047815322875977,
+      "learning_rate": 0.00029944868194808357,
+      "loss": 4.7483,
+      "step": 88690
+    },
+    {
+      "epoch": 0.12463182151093251,
+      "grad_norm": 0.722084641456604,
+      "learning_rate": 0.0002994480515726499,
+      "loss": 4.7039,
+      "step": 88700
+    },
+    {
+      "epoch": 0.12464587244909608,
+      "grad_norm": 0.717406153678894,
+      "learning_rate": 0.00029944742083770176,
+      "loss": 4.622,
+      "step": 88710
+    },
+    {
+      "epoch": 0.12465992338725966,
+      "grad_norm": 0.7036492228507996,
+      "learning_rate": 0.00029944678974324045,
+      "loss": 4.7124,
+      "step": 88720
+    },
+    {
+      "epoch": 0.12467397432542324,
+      "grad_norm": 0.7213016748428345,
+      "learning_rate": 0.0002994461582892678,
+      "loss": 4.6973,
+      "step": 88730
+    },
+    {
+      "epoch": 0.12468802526358681,
+      "grad_norm": 0.775976300239563,
+      "learning_rate": 0.000299445526475785,
+      "loss": 4.7035,
+      "step": 88740
+    },
+    {
+      "epoch": 0.12470207620175039,
+      "grad_norm": 0.7207806706428528,
+      "learning_rate": 0.00029944489430279377,
+      "loss": 4.7464,
+      "step": 88750
+    },
+    {
+      "epoch": 0.12471612713991398,
+      "grad_norm": 0.6966215968132019,
+      "learning_rate": 0.00029944426177029555,
+      "loss": 4.7001,
+      "step": 88760
+    },
+    {
+      "epoch": 0.12473017807807756,
+      "grad_norm": 0.7105135321617126,
+      "learning_rate": 0.000299443628878292,
+      "loss": 4.778,
+      "step": 88770
+    },
+    {
+      "epoch": 0.12474422901624113,
+      "grad_norm": 0.7118408679962158,
+      "learning_rate": 0.0002994429956267844,
+      "loss": 4.7942,
+      "step": 88780
+    },
+    {
+      "epoch": 0.12475827995440471,
+      "grad_norm": 0.7115956544876099,
+      "learning_rate": 0.0002994423620157745,
+      "loss": 4.7289,
+      "step": 88790
+    },
+    {
+      "epoch": 0.12477233089256828,
+      "grad_norm": 0.8141958117485046,
+      "learning_rate": 0.0002994417280452637,
+      "loss": 4.7197,
+      "step": 88800
+    },
+    {
+      "epoch": 0.12478638183073186,
+      "grad_norm": 0.6964655518531799,
+      "learning_rate": 0.00029944109371525356,
+      "loss": 4.6832,
+      "step": 88810
+    },
+    {
+      "epoch": 0.12480043276889544,
+      "grad_norm": 0.6905677318572998,
+      "learning_rate": 0.0002994404590257456,
+      "loss": 4.6532,
+      "step": 88820
+    },
+    {
+      "epoch": 0.12481448370705901,
+      "grad_norm": 0.7176385521888733,
+      "learning_rate": 0.0002994398239767413,
+      "loss": 4.6367,
+      "step": 88830
+    },
+    {
+      "epoch": 0.12482853464522259,
+      "grad_norm": 0.723534107208252,
+      "learning_rate": 0.0002994391885682423,
+      "loss": 4.6079,
+      "step": 88840
+    },
+    {
+      "epoch": 0.12484258558338616,
+      "grad_norm": 0.7028889060020447,
+      "learning_rate": 0.0002994385528002501,
+      "loss": 4.7336,
+      "step": 88850
+    },
+    {
+      "epoch": 0.12485663652154975,
+      "grad_norm": 0.7260593771934509,
+      "learning_rate": 0.0002994379166727661,
+      "loss": 4.5638,
+      "step": 88860
+    },
+    {
+      "epoch": 0.12487068745971333,
+      "grad_norm": 0.7041234374046326,
+      "learning_rate": 0.00029943728018579205,
+      "loss": 4.7209,
+      "step": 88870
+    },
+    {
+      "epoch": 0.12488473839787691,
+      "grad_norm": 0.6976323127746582,
+      "learning_rate": 0.0002994366433393293,
+      "loss": 4.6149,
+      "step": 88880
+    },
+    {
+      "epoch": 0.12489878933604048,
+      "grad_norm": 0.7066828012466431,
+      "learning_rate": 0.00029943600613337943,
+      "loss": 4.6027,
+      "step": 88890
+    },
+    {
+      "epoch": 0.12491284027420406,
+      "grad_norm": 0.7120361328125,
+      "learning_rate": 0.000299435368567944,
+      "loss": 4.5695,
+      "step": 88900
+    },
+    {
+      "epoch": 0.12492689121236764,
+      "grad_norm": 0.7178036570549011,
+      "learning_rate": 0.0002994347306430245,
+      "loss": 4.7551,
+      "step": 88910
+    },
+    {
+      "epoch": 0.12494094215053121,
+      "grad_norm": 1.0115084648132324,
+      "learning_rate": 0.0002994340923586225,
+      "loss": 4.5328,
+      "step": 88920
+    },
+    {
+      "epoch": 0.12495499308869479,
+      "grad_norm": 0.7211881279945374,
+      "learning_rate": 0.0002994334537147395,
+      "loss": 4.753,
+      "step": 88930
+    },
+    {
+      "epoch": 0.12496904402685836,
+      "grad_norm": 0.72914057970047,
+      "learning_rate": 0.00029943281471137706,
+      "loss": 4.6328,
+      "step": 88940
+    },
+    {
+      "epoch": 0.12498309496502194,
+      "grad_norm": 0.7203587889671326,
+      "learning_rate": 0.00029943217534853677,
+      "loss": 4.7458,
+      "step": 88950
+    },
+    {
+      "epoch": 0.12499714590318552,
+      "grad_norm": 0.7104395031929016,
+      "learning_rate": 0.0002994315356262201,
+      "loss": 4.6597,
+      "step": 88960
+    },
+    {
+      "epoch": 0.1250111968413491,
+      "grad_norm": 0.7179931998252869,
+      "learning_rate": 0.0002994308955444286,
+      "loss": 4.6605,
+      "step": 88970
+    },
+    {
+      "epoch": 0.12502524777951268,
+      "grad_norm": 0.7127233743667603,
+      "learning_rate": 0.00029943025510316385,
+      "loss": 4.7703,
+      "step": 88980
+    },
+    {
+      "epoch": 0.12503929871767624,
+      "grad_norm": 0.7128562331199646,
+      "learning_rate": 0.00029942961430242727,
+      "loss": 4.8032,
+      "step": 88990
+    },
+    {
+      "epoch": 0.12505334965583983,
+      "grad_norm": 0.7086876630783081,
+      "learning_rate": 0.0002994289731422206,
+      "loss": 4.6504,
+      "step": 89000
+    },
+    {
+      "epoch": 0.1250674005940034,
+      "grad_norm": 0.7091412544250488,
+      "learning_rate": 0.0002994283316225452,
+      "loss": 4.6319,
+      "step": 89010
+    },
+    {
+      "epoch": 0.125081451532167,
+      "grad_norm": 0.7031800746917725,
+      "learning_rate": 0.0002994276897434027,
+      "loss": 4.6304,
+      "step": 89020
+    },
+    {
+      "epoch": 0.12509550247033058,
+      "grad_norm": 0.7421759963035583,
+      "learning_rate": 0.00029942704750479464,
+      "loss": 4.7184,
+      "step": 89030
+    },
+    {
+      "epoch": 0.12510955340849414,
+      "grad_norm": 0.7047469019889832,
+      "learning_rate": 0.00029942640490672256,
+      "loss": 4.6966,
+      "step": 89040
+    },
+    {
+      "epoch": 0.12512360434665773,
+      "grad_norm": 0.717984139919281,
+      "learning_rate": 0.00029942576194918796,
+      "loss": 4.6959,
+      "step": 89050
+    },
+    {
+      "epoch": 0.1251376552848213,
+      "grad_norm": 0.674823522567749,
+      "learning_rate": 0.0002994251186321925,
+      "loss": 4.7184,
+      "step": 89060
+    },
+    {
+      "epoch": 0.12515170622298488,
+      "grad_norm": 0.725084662437439,
+      "learning_rate": 0.0002994244749557376,
+      "loss": 4.5896,
+      "step": 89070
+    },
+    {
+      "epoch": 0.12516575716114844,
+      "grad_norm": 0.7343477010726929,
+      "learning_rate": 0.00029942383091982486,
+      "loss": 4.6372,
+      "step": 89080
+    },
+    {
+      "epoch": 0.12517980809931203,
+      "grad_norm": 0.6985862255096436,
+      "learning_rate": 0.0002994231865244558,
+      "loss": 4.7568,
+      "step": 89090
+    },
+    {
+      "epoch": 0.1251938590374756,
+      "grad_norm": 0.7344580292701721,
+      "learning_rate": 0.00029942254176963207,
+      "loss": 4.774,
+      "step": 89100
+    },
+    {
+      "epoch": 0.1252079099756392,
+      "grad_norm": 0.714970588684082,
+      "learning_rate": 0.0002994218966553551,
+      "loss": 4.7269,
+      "step": 89110
+    },
+    {
+      "epoch": 0.12522196091380278,
+      "grad_norm": 0.7142235636711121,
+      "learning_rate": 0.00029942125118162657,
+      "loss": 4.7118,
+      "step": 89120
+    },
+    {
+      "epoch": 0.12523601185196634,
+      "grad_norm": 0.7200468182563782,
+      "learning_rate": 0.00029942060534844786,
+      "loss": 4.5759,
+      "step": 89130
+    },
+    {
+      "epoch": 0.12525006279012993,
+      "grad_norm": 0.7273971438407898,
+      "learning_rate": 0.0002994199591558207,
+      "loss": 4.596,
+      "step": 89140
+    },
+    {
+      "epoch": 0.1252641137282935,
+      "grad_norm": 0.708501398563385,
+      "learning_rate": 0.00029941931260374653,
+      "loss": 4.7371,
+      "step": 89150
+    },
+    {
+      "epoch": 0.12527816466645708,
+      "grad_norm": 0.6819218397140503,
+      "learning_rate": 0.00029941866569222695,
+      "loss": 4.6924,
+      "step": 89160
+    },
+    {
+      "epoch": 0.12529221560462064,
+      "grad_norm": 0.7276637554168701,
+      "learning_rate": 0.0002994180184212635,
+      "loss": 4.5807,
+      "step": 89170
+    },
+    {
+      "epoch": 0.12530626654278423,
+      "grad_norm": 0.8149365186691284,
+      "learning_rate": 0.00029941737079085777,
+      "loss": 4.7547,
+      "step": 89180
+    },
+    {
+      "epoch": 0.1253203174809478,
+      "grad_norm": 0.7019021511077881,
+      "learning_rate": 0.00029941672280101126,
+      "loss": 4.7022,
+      "step": 89190
+    },
+    {
+      "epoch": 0.12533436841911139,
+      "grad_norm": 0.7072144150733948,
+      "learning_rate": 0.00029941607445172554,
+      "loss": 4.6415,
+      "step": 89200
+    },
+    {
+      "epoch": 0.12534841935727495,
+      "grad_norm": 0.6896094083786011,
+      "learning_rate": 0.00029941542574300225,
+      "loss": 4.5062,
+      "step": 89210
+    },
+    {
+      "epoch": 0.12536247029543854,
+      "grad_norm": 0.7020169496536255,
+      "learning_rate": 0.00029941477667484284,
+      "loss": 4.8094,
+      "step": 89220
+    },
+    {
+      "epoch": 0.12537652123360213,
+      "grad_norm": 0.6777108311653137,
+      "learning_rate": 0.000299414127247249,
+      "loss": 4.6478,
+      "step": 89230
+    },
+    {
+      "epoch": 0.1253905721717657,
+      "grad_norm": 0.7213195562362671,
+      "learning_rate": 0.00029941347746022213,
+      "loss": 4.7103,
+      "step": 89240
+    },
+    {
+      "epoch": 0.12540462310992928,
+      "grad_norm": 0.6935251951217651,
+      "learning_rate": 0.00029941282731376387,
+      "loss": 4.3878,
+      "step": 89250
+    },
+    {
+      "epoch": 0.12541867404809284,
+      "grad_norm": 0.8068158626556396,
+      "learning_rate": 0.00029941217680787583,
+      "loss": 4.6871,
+      "step": 89260
+    },
+    {
+      "epoch": 0.12543272498625643,
+      "grad_norm": 0.7309907078742981,
+      "learning_rate": 0.00029941152594255955,
+      "loss": 4.7632,
+      "step": 89270
+    },
+    {
+      "epoch": 0.12544677592442,
+      "grad_norm": 0.7052793502807617,
+      "learning_rate": 0.0002994108747178166,
+      "loss": 4.8095,
+      "step": 89280
+    },
+    {
+      "epoch": 0.12546082686258359,
+      "grad_norm": 0.7147111892700195,
+      "learning_rate": 0.00029941022313364845,
+      "loss": 4.6167,
+      "step": 89290
+    },
+    {
+      "epoch": 0.12547487780074715,
+      "grad_norm": 0.7007347345352173,
+      "learning_rate": 0.00029940957119005683,
+      "loss": 4.7428,
+      "step": 89300
+    },
+    {
+      "epoch": 0.12548892873891074,
+      "grad_norm": 0.6927651166915894,
+      "learning_rate": 0.0002994089188870432,
+      "loss": 4.7348,
+      "step": 89310
+    },
+    {
+      "epoch": 0.1255029796770743,
+      "grad_norm": 0.7285720109939575,
+      "learning_rate": 0.0002994082662246091,
+      "loss": 4.6294,
+      "step": 89320
+    },
+    {
+      "epoch": 0.1255170306152379,
+      "grad_norm": 0.7167104482650757,
+      "learning_rate": 0.0002994076132027563,
+      "loss": 4.6708,
+      "step": 89330
+    },
+    {
+      "epoch": 0.12553108155340148,
+      "grad_norm": 0.7202353477478027,
+      "learning_rate": 0.00029940695982148607,
+      "loss": 4.7455,
+      "step": 89340
+    },
+    {
+      "epoch": 0.12554513249156504,
+      "grad_norm": 0.7299073934555054,
+      "learning_rate": 0.0002994063060808002,
+      "loss": 4.6263,
+      "step": 89350
+    },
+    {
+      "epoch": 0.12555918342972863,
+      "grad_norm": 0.7049494981765747,
+      "learning_rate": 0.00029940565198070015,
+      "loss": 4.6685,
+      "step": 89360
+    },
+    {
+      "epoch": 0.1255732343678922,
+      "grad_norm": 0.7378440499305725,
+      "learning_rate": 0.00029940499752118764,
+      "loss": 4.628,
+      "step": 89370
+    },
+    {
+      "epoch": 0.12558728530605578,
+      "grad_norm": 0.705937922000885,
+      "learning_rate": 0.00029940434270226405,
+      "loss": 4.6245,
+      "step": 89380
+    },
+    {
+      "epoch": 0.12560133624421935,
+      "grad_norm": 0.7085596323013306,
+      "learning_rate": 0.00029940368752393113,
+      "loss": 4.607,
+      "step": 89390
+    },
+    {
+      "epoch": 0.12561538718238294,
+      "grad_norm": 0.7117762565612793,
+      "learning_rate": 0.00029940303198619034,
+      "loss": 4.5603,
+      "step": 89400
+    },
+    {
+      "epoch": 0.1256294381205465,
+      "grad_norm": 0.7034866809844971,
+      "learning_rate": 0.0002994023760890433,
+      "loss": 4.6884,
+      "step": 89410
+    },
+    {
+      "epoch": 0.1256434890587101,
+      "grad_norm": 0.6858923435211182,
+      "learning_rate": 0.00029940171983249155,
+      "loss": 4.6888,
+      "step": 89420
+    },
+    {
+      "epoch": 0.12565753999687368,
+      "grad_norm": 0.7072145938873291,
+      "learning_rate": 0.00029940106321653674,
+      "loss": 4.6249,
+      "step": 89430
+    },
+    {
+      "epoch": 0.12567159093503724,
+      "grad_norm": 0.730768620967865,
+      "learning_rate": 0.0002994004062411804,
+      "loss": 4.5767,
+      "step": 89440
+    },
+    {
+      "epoch": 0.12568564187320083,
+      "grad_norm": 0.7527211308479309,
+      "learning_rate": 0.0002993997489064242,
+      "loss": 4.6003,
+      "step": 89450
+    },
+    {
+      "epoch": 0.1256996928113644,
+      "grad_norm": 0.7296200394630432,
+      "learning_rate": 0.0002993990912122695,
+      "loss": 4.7804,
+      "step": 89460
+    },
+    {
+      "epoch": 0.12571374374952798,
+      "grad_norm": 0.7212024927139282,
+      "learning_rate": 0.00029939843315871813,
+      "loss": 4.6649,
+      "step": 89470
+    },
+    {
+      "epoch": 0.12572779468769155,
+      "grad_norm": 0.7408531308174133,
+      "learning_rate": 0.00029939777474577157,
+      "loss": 4.7058,
+      "step": 89480
+    },
+    {
+      "epoch": 0.12574184562585514,
+      "grad_norm": 0.7241307497024536,
+      "learning_rate": 0.00029939711597343135,
+      "loss": 4.7249,
+      "step": 89490
+    },
+    {
+      "epoch": 0.1257558965640187,
+      "grad_norm": 0.7100182175636292,
+      "learning_rate": 0.00029939645684169913,
+      "loss": 4.5997,
+      "step": 89500
+    },
+    {
+      "epoch": 0.1257699475021823,
+      "grad_norm": 0.7237468957901001,
+      "learning_rate": 0.00029939579735057645,
+      "loss": 4.698,
+      "step": 89510
+    },
+    {
+      "epoch": 0.12578399844034585,
+      "grad_norm": 0.6810537576675415,
+      "learning_rate": 0.0002993951375000649,
+      "loss": 4.6484,
+      "step": 89520
+    },
+    {
+      "epoch": 0.12579804937850944,
+      "grad_norm": 0.7313446402549744,
+      "learning_rate": 0.0002993944772901662,
+      "loss": 4.5818,
+      "step": 89530
+    },
+    {
+      "epoch": 0.12581210031667303,
+      "grad_norm": 0.7293576598167419,
+      "learning_rate": 0.00029939381672088175,
+      "loss": 4.6622,
+      "step": 89540
+    },
+    {
+      "epoch": 0.1258261512548366,
+      "grad_norm": 0.7267341613769531,
+      "learning_rate": 0.0002993931557922132,
+      "loss": 4.5724,
+      "step": 89550
+    },
+    {
+      "epoch": 0.12584020219300018,
+      "grad_norm": 0.722690999507904,
+      "learning_rate": 0.00029939249450416217,
+      "loss": 4.7577,
+      "step": 89560
+    },
+    {
+      "epoch": 0.12585425313116375,
+      "grad_norm": 0.7288978695869446,
+      "learning_rate": 0.00029939183285673023,
+      "loss": 4.6926,
+      "step": 89570
+    },
+    {
+      "epoch": 0.12586830406932734,
+      "grad_norm": 0.7535728216171265,
+      "learning_rate": 0.000299391170849919,
+      "loss": 4.6465,
+      "step": 89580
+    },
+    {
+      "epoch": 0.1258823550074909,
+      "grad_norm": 0.7304617166519165,
+      "learning_rate": 0.00029939050848373,
+      "loss": 4.6726,
+      "step": 89590
+    },
+    {
+      "epoch": 0.1258964059456545,
+      "grad_norm": 0.7804543972015381,
+      "learning_rate": 0.00029938984575816495,
+      "loss": 4.7101,
+      "step": 89600
+    },
+    {
+      "epoch": 0.12591045688381805,
+      "grad_norm": 0.7636072635650635,
+      "learning_rate": 0.0002993891826732253,
+      "loss": 4.7559,
+      "step": 89610
+    },
+    {
+      "epoch": 0.12592450782198164,
+      "grad_norm": 0.8159419298171997,
+      "learning_rate": 0.0002993885192289128,
+      "loss": 4.5094,
+      "step": 89620
+    },
+    {
+      "epoch": 0.1259385587601452,
+      "grad_norm": 0.7560939192771912,
+      "learning_rate": 0.00029938785542522883,
+      "loss": 4.7041,
+      "step": 89630
+    },
+    {
+      "epoch": 0.1259526096983088,
+      "grad_norm": 0.7355748414993286,
+      "learning_rate": 0.0002993871912621752,
+      "loss": 4.7708,
+      "step": 89640
+    },
+    {
+      "epoch": 0.12596666063647238,
+      "grad_norm": 0.7174279689788818,
+      "learning_rate": 0.0002993865267397534,
+      "loss": 4.6629,
+      "step": 89650
+    },
+    {
+      "epoch": 0.12598071157463595,
+      "grad_norm": 0.7072028517723083,
+      "learning_rate": 0.00029938586185796504,
+      "loss": 4.708,
+      "step": 89660
+    },
+    {
+      "epoch": 0.12599476251279954,
+      "grad_norm": 0.7467581033706665,
+      "learning_rate": 0.00029938519661681175,
+      "loss": 4.6189,
+      "step": 89670
+    },
+    {
+      "epoch": 0.1260088134509631,
+      "grad_norm": 0.724288821220398,
+      "learning_rate": 0.0002993845310162951,
+      "loss": 4.6856,
+      "step": 89680
+    },
+    {
+      "epoch": 0.1260228643891267,
+      "grad_norm": 0.7614923715591431,
+      "learning_rate": 0.0002993838650564167,
+      "loss": 4.6472,
+      "step": 89690
+    },
+    {
+      "epoch": 0.12603691532729025,
+      "grad_norm": 0.7412780523300171,
+      "learning_rate": 0.00029938319873717815,
+      "loss": 4.6539,
+      "step": 89700
+    },
+    {
+      "epoch": 0.12605096626545384,
+      "grad_norm": 0.7168459892272949,
+      "learning_rate": 0.00029938253205858106,
+      "loss": 4.775,
+      "step": 89710
+    },
+    {
+      "epoch": 0.1260650172036174,
+      "grad_norm": 0.7272344827651978,
+      "learning_rate": 0.00029938186502062705,
+      "loss": 4.698,
+      "step": 89720
+    },
+    {
+      "epoch": 0.126079068141781,
+      "grad_norm": 0.6984730958938599,
+      "learning_rate": 0.0002993811976233177,
+      "loss": 4.5482,
+      "step": 89730
+    },
+    {
+      "epoch": 0.12609311907994458,
+      "grad_norm": 0.698017418384552,
+      "learning_rate": 0.0002993805298666546,
+      "loss": 4.6122,
+      "step": 89740
+    },
+    {
+      "epoch": 0.12610717001810814,
+      "grad_norm": 0.7286360263824463,
+      "learning_rate": 0.0002993798617506394,
+      "loss": 4.7628,
+      "step": 89750
+    },
+    {
+      "epoch": 0.12612122095627173,
+      "grad_norm": 0.6975075006484985,
+      "learning_rate": 0.00029937919327527367,
+      "loss": 4.7699,
+      "step": 89760
+    },
+    {
+      "epoch": 0.1261352718944353,
+      "grad_norm": 0.7229320406913757,
+      "learning_rate": 0.000299378524440559,
+      "loss": 4.7071,
+      "step": 89770
+    },
+    {
+      "epoch": 0.1261493228325989,
+      "grad_norm": 0.7221950888633728,
+      "learning_rate": 0.00029937785524649713,
+      "loss": 4.6989,
+      "step": 89780
+    },
+    {
+      "epoch": 0.12616337377076245,
+      "grad_norm": 0.6790987849235535,
+      "learning_rate": 0.00029937718569308946,
+      "loss": 4.7405,
+      "step": 89790
+    },
+    {
+      "epoch": 0.12617742470892604,
+      "grad_norm": 0.7140920162200928,
+      "learning_rate": 0.0002993765157803378,
+      "loss": 4.7422,
+      "step": 89800
+    },
+    {
+      "epoch": 0.1261914756470896,
+      "grad_norm": 0.7047672867774963,
+      "learning_rate": 0.0002993758455082436,
+      "loss": 4.6582,
+      "step": 89810
+    },
+    {
+      "epoch": 0.1262055265852532,
+      "grad_norm": 0.7307078242301941,
+      "learning_rate": 0.0002993751748768086,
+      "loss": 4.6722,
+      "step": 89820
+    },
+    {
+      "epoch": 0.12621957752341675,
+      "grad_norm": 0.7230780124664307,
+      "learning_rate": 0.0002993745038860344,
+      "loss": 4.7186,
+      "step": 89830
+    },
+    {
+      "epoch": 0.12623362846158034,
+      "grad_norm": 0.7278953790664673,
+      "learning_rate": 0.00029937383253592244,
+      "loss": 4.7258,
+      "step": 89840
+    },
+    {
+      "epoch": 0.12624767939974393,
+      "grad_norm": 0.7198377251625061,
+      "learning_rate": 0.00029937316082647456,
+      "loss": 4.7229,
+      "step": 89850
+    },
+    {
+      "epoch": 0.1262617303379075,
+      "grad_norm": 0.6939035058021545,
+      "learning_rate": 0.00029937248875769226,
+      "loss": 4.6455,
+      "step": 89860
+    },
+    {
+      "epoch": 0.1262757812760711,
+      "grad_norm": 0.7266935706138611,
+      "learning_rate": 0.00029937181632957717,
+      "loss": 4.5993,
+      "step": 89870
+    },
+    {
+      "epoch": 0.12628983221423465,
+      "grad_norm": 0.7101845145225525,
+      "learning_rate": 0.00029937114354213097,
+      "loss": 4.6855,
+      "step": 89880
+    },
+    {
+      "epoch": 0.12630388315239824,
+      "grad_norm": 0.7384141087532043,
+      "learning_rate": 0.0002993704703953552,
+      "loss": 4.7441,
+      "step": 89890
+    },
+    {
+      "epoch": 0.1263179340905618,
+      "grad_norm": 0.6860256791114807,
+      "learning_rate": 0.0002993697968892515,
+      "loss": 4.6038,
+      "step": 89900
+    },
+    {
+      "epoch": 0.1263319850287254,
+      "grad_norm": 0.7191735506057739,
+      "learning_rate": 0.0002993691230238215,
+      "loss": 4.6831,
+      "step": 89910
+    },
+    {
+      "epoch": 0.12634603596688895,
+      "grad_norm": 0.7158761620521545,
+      "learning_rate": 0.0002993684487990668,
+      "loss": 4.5944,
+      "step": 89920
+    },
+    {
+      "epoch": 0.12636008690505254,
+      "grad_norm": 0.7162553071975708,
+      "learning_rate": 0.0002993677742149891,
+      "loss": 4.6579,
+      "step": 89930
+    },
+    {
+      "epoch": 0.1263741378432161,
+      "grad_norm": 0.7150312662124634,
+      "learning_rate": 0.0002993670992715899,
+      "loss": 4.7404,
+      "step": 89940
+    },
+    {
+      "epoch": 0.1263881887813797,
+      "grad_norm": 0.7158754467964172,
+      "learning_rate": 0.0002993664239688709,
+      "loss": 4.6764,
+      "step": 89950
+    },
+    {
+      "epoch": 0.12640223971954329,
+      "grad_norm": 0.7216680645942688,
+      "learning_rate": 0.0002993657483068337,
+      "loss": 4.6328,
+      "step": 89960
+    },
+    {
+      "epoch": 0.12641629065770685,
+      "grad_norm": 0.7339308261871338,
+      "learning_rate": 0.00029936507228547996,
+      "loss": 4.6651,
+      "step": 89970
+    },
+    {
+      "epoch": 0.12643034159587044,
+      "grad_norm": 0.700171709060669,
+      "learning_rate": 0.00029936439590481126,
+      "loss": 4.6854,
+      "step": 89980
+    },
+    {
+      "epoch": 0.126444392534034,
+      "grad_norm": 0.7425507307052612,
+      "learning_rate": 0.0002993637191648293,
+      "loss": 4.5799,
+      "step": 89990
+    },
+    {
+      "epoch": 0.1264584434721976,
+      "grad_norm": 0.7096763849258423,
+      "learning_rate": 0.0002993630420655356,
+      "loss": 4.5681,
+      "step": 90000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 711696,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.7405109256192e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}