diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21701 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 100.0,
+  "eval_steps": 20000,
+  "global_step": 309400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03232062055591468,
+      "grad_norm": 13.562658309936523,
+      "learning_rate": 9.900000000000002e-06,
+      "loss": 4.2119,
+      "step": 100
+    },
+    {
+      "epoch": 0.06464124111182935,
+      "grad_norm": 17.95138931274414,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 4.0516,
+      "step": 200
+    },
+    {
+      "epoch": 0.09696186166774402,
+      "grad_norm": 11.31137752532959,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 3.9654,
+      "step": 300
+    },
+    {
+      "epoch": 0.1292824822236587,
+      "grad_norm": 17.110918045043945,
+      "learning_rate": 3.99e-05,
+      "loss": 3.7782,
+      "step": 400
+    },
+    {
+      "epoch": 0.16160310277957338,
+      "grad_norm": 11.72215461730957,
+      "learning_rate": 4.99e-05,
+      "loss": 3.4622,
+      "step": 500
+    },
+    {
+      "epoch": 0.19392372333548805,
+      "grad_norm": 7.6839447021484375,
+      "learning_rate": 5.9900000000000006e-05,
+      "loss": 3.0781,
+      "step": 600
+    },
+    {
+      "epoch": 0.22624434389140272,
+      "grad_norm": 9.97439193725586,
+      "learning_rate": 6.99e-05,
+      "loss": 2.8028,
+      "step": 700
+    },
+    {
+      "epoch": 0.2585649644473174,
+      "grad_norm": 7.437990665435791,
+      "learning_rate": 7.99e-05,
+      "loss": 2.6209,
+      "step": 800
+    },
+    {
+      "epoch": 0.2908855850032321,
+      "grad_norm": 5.37725830078125,
+      "learning_rate": 8.989999999999999e-05,
+      "loss": 2.4793,
+      "step": 900
+    },
+    {
+      "epoch": 0.32320620555914675,
+      "grad_norm": 3.4234330654144287,
+      "learning_rate": 9.99e-05,
+      "loss": 2.4139,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3555268261150614,
+      "grad_norm": 3.771662473678589,
+      "learning_rate": 0.0001099,
+      "loss": 2.3273,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3878474466709761,
+      "grad_norm": 3.9465994834899902,
+      "learning_rate": 0.00011990000000000001,
+      "loss": 2.3011,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 3.0348823070526123,
+      "learning_rate": 0.00012989999999999999,
+      "loss": 2.2512,
+      "step": 1300
+    },
+    {
+      "epoch": 0.45248868778280543,
+      "grad_norm": 2.580744504928589,
+      "learning_rate": 0.0001399,
+      "loss": 2.2338,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4848093083387201,
+      "grad_norm": 2.7396256923675537,
+      "learning_rate": 0.0001499,
+      "loss": 2.2165,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5171299288946348,
+      "grad_norm": 2.427551507949829,
+      "learning_rate": 0.00015989999999999998,
+      "loss": 2.1936,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 4.367044448852539,
+      "learning_rate": 0.0001699,
+      "loss": 2.1448,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5817711700064642,
+      "grad_norm": 3.5235722064971924,
+      "learning_rate": 0.0001799,
+      "loss": 2.1785,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6140917905623788,
+      "grad_norm": 1.780901551246643,
+      "learning_rate": 0.0001899,
+      "loss": 2.1839,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6464124111182935,
+      "grad_norm": 1.8098434209823608,
+      "learning_rate": 0.0001999,
+      "loss": 2.1043,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6787330316742082,
+      "grad_norm": 2.296581268310547,
+      "learning_rate": 0.0002099,
+      "loss": 2.1415,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7110536522301228,
+      "grad_norm": 2.007957696914673,
+      "learning_rate": 0.0002199,
+      "loss": 2.1096,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7433742727860375,
+      "grad_norm": 3.1119742393493652,
+      "learning_rate": 0.0002299,
+      "loss": 2.1146,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7756948933419522,
+      "grad_norm": 1.6894687414169312,
+      "learning_rate": 0.0002399,
+      "loss": 2.0824,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8080155138978669,
+      "grad_norm": 1.2437957525253296,
+      "learning_rate": 0.0002499,
+      "loss": 2.0671,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 2.4846837520599365,
+      "learning_rate": 0.00025990000000000003,
+      "loss": 2.1032,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8726567550096962,
+      "grad_norm": 1.7763832807540894,
+      "learning_rate": 0.0002699,
+      "loss": 2.1034,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9049773755656109,
+      "grad_norm": 2.0607247352600098,
+      "learning_rate": 0.0002799,
+      "loss": 2.0898,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9372979961215255,
+      "grad_norm": 1.1315553188323975,
+      "learning_rate": 0.0002899,
+      "loss": 2.0906,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9696186166774402,
+      "grad_norm": 1.3324776887893677,
+      "learning_rate": 0.0002999,
+      "loss": 2.065,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0019392372333549,
+      "grad_norm": 1.5390121936798096,
+      "learning_rate": 0.0003099,
+      "loss": 2.0899,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0342598577892697,
+      "grad_norm": 3.3471102714538574,
+      "learning_rate": 0.0003199,
+      "loss": 1.986,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0665804783451842,
+      "grad_norm": 2.176851749420166,
+      "learning_rate": 0.00032990000000000005,
+      "loss": 2.0288,
+      "step": 3300
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 1.6547337770462036,
+      "learning_rate": 0.00033989999999999997,
+      "loss": 2.0137,
+      "step": 3400
+    },
+    {
+      "epoch": 1.1312217194570136,
+      "grad_norm": 2.7004611492156982,
+      "learning_rate": 0.0003499,
+      "loss": 2.0361,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1635423400129283,
+      "grad_norm": 9.764275550842285,
+      "learning_rate": 0.0003599,
+      "loss": 2.0461,
+      "step": 3600
+    },
+    {
+      "epoch": 1.195862960568843,
+      "grad_norm": 2.408719062805176,
+      "learning_rate": 0.0003699,
+      "loss": 2.013,
+      "step": 3700
+    },
+    {
+      "epoch": 1.2281835811247577,
+      "grad_norm": 1.4798204898834229,
+      "learning_rate": 0.0003799,
+      "loss": 2.0373,
+      "step": 3800
+    },
+    {
+      "epoch": 1.2605042016806722,
+      "grad_norm": 1.6942540407180786,
+      "learning_rate": 0.00038990000000000004,
+      "loss": 2.0252,
+      "step": 3900
+    },
+    {
+      "epoch": 1.292824822236587,
+      "grad_norm": 1.9354304075241089,
+      "learning_rate": 0.00039989999999999996,
+      "loss": 1.968,
+      "step": 4000
+    },
+    {
+      "epoch": 1.3251454427925016,
+      "grad_norm": 1.3031123876571655,
+      "learning_rate": 0.0004099,
+      "loss": 1.9987,
+      "step": 4100
+    },
+    {
+      "epoch": 1.3574660633484164,
+      "grad_norm": 2.0432260036468506,
+      "learning_rate": 0.0004199,
+      "loss": 2.0022,
+      "step": 4200
+    },
+    {
+      "epoch": 1.389786683904331,
+      "grad_norm": 1.3270591497421265,
+      "learning_rate": 0.0004299,
+      "loss": 1.9947,
+      "step": 4300
+    },
+    {
+      "epoch": 1.4221073044602457,
+      "grad_norm": 2.2205493450164795,
+      "learning_rate": 0.0004399,
+      "loss": 2.0379,
+      "step": 4400
+    },
+    {
+      "epoch": 1.4544279250161603,
+      "grad_norm": 1.547240138053894,
+      "learning_rate": 0.00044990000000000004,
+      "loss": 1.9993,
+      "step": 4500
+    },
+    {
+      "epoch": 1.486748545572075,
+      "grad_norm": 1.6051744222640991,
+      "learning_rate": 0.0004599,
+      "loss": 1.9934,
+      "step": 4600
+    },
+    {
+      "epoch": 1.5190691661279896,
+      "grad_norm": 1.6245908737182617,
+      "learning_rate": 0.0004699,
+      "loss": 2.0052,
+      "step": 4700
+    },
+    {
+      "epoch": 1.5513897866839044,
+      "grad_norm": 1.2813587188720703,
+      "learning_rate": 0.0004799,
+      "loss": 2.0224,
+      "step": 4800
+    },
+    {
+      "epoch": 1.5837104072398192,
+      "grad_norm": 1.2587623596191406,
+      "learning_rate": 0.0004899,
+      "loss": 2.0377,
+      "step": 4900
+    },
+    {
+      "epoch": 1.6160310277957337,
+      "grad_norm": 1.2050402164459229,
+      "learning_rate": 0.0004999000000000001,
+      "loss": 2.0159,
+      "step": 5000
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 1.2096720933914185,
+      "learning_rate": 0.0005099,
+      "loss": 1.9913,
+      "step": 5100
+    },
+    {
+      "epoch": 1.680672268907563,
+      "grad_norm": 1.371106743812561,
+      "learning_rate": 0.0005199,
+      "loss": 1.9823,
+      "step": 5200
+    },
+    {
+      "epoch": 1.7129928894634778,
+      "grad_norm": 1.6344568729400635,
+      "learning_rate": 0.0005299,
+      "loss": 1.9601,
+      "step": 5300
+    },
+    {
+      "epoch": 1.7453135100193924,
+      "grad_norm": 1.475213646888733,
+      "learning_rate": 0.0005399000000000001,
+      "loss": 2.0069,
+      "step": 5400
+    },
+    {
+      "epoch": 1.777634130575307,
+      "grad_norm": 1.147307276725769,
+      "learning_rate": 0.0005499000000000001,
+      "loss": 1.9979,
+      "step": 5500
+    },
+    {
+      "epoch": 1.8099547511312217,
+      "grad_norm": 1.5867249965667725,
+      "learning_rate": 0.0005599,
+      "loss": 1.9819,
+      "step": 5600
+    },
+    {
+      "epoch": 1.8422753716871365,
+      "grad_norm": 1.2033599615097046,
+      "learning_rate": 0.0005698999999999999,
+      "loss": 1.9972,
+      "step": 5700
+    },
+    {
+      "epoch": 1.874595992243051,
+      "grad_norm": 1.5934315919876099,
+      "learning_rate": 0.0005799,
+      "loss": 1.9636,
+      "step": 5800
+    },
+    {
+      "epoch": 1.9069166127989656,
+      "grad_norm": 1.7510161399841309,
+      "learning_rate": 0.0005899,
+      "loss": 1.9905,
+      "step": 5900
+    },
+    {
+      "epoch": 1.9392372333548804,
+      "grad_norm": 1.3442281484603882,
+      "learning_rate": 0.0005999,
+      "loss": 2.0058,
+      "step": 6000
+    },
+    {
+      "epoch": 1.9715578539107952,
+      "grad_norm": 1.8174034357070923,
+      "learning_rate": 0.0006099,
+      "loss": 1.9725,
+      "step": 6100
+    },
+    {
+      "epoch": 2.0038784744667097,
+      "grad_norm": 1.3350085020065308,
+      "learning_rate": 0.0006199,
+      "loss": 2.0277,
+      "step": 6200
+    },
+    {
+      "epoch": 2.0361990950226243,
+      "grad_norm": 1.1999424695968628,
+      "learning_rate": 0.0006299000000000001,
+      "loss": 1.9098,
+      "step": 6300
+    },
+    {
+      "epoch": 2.0685197155785393,
+      "grad_norm": 1.2381353378295898,
+      "learning_rate": 0.0006399,
+      "loss": 1.9531,
+      "step": 6400
+    },
+    {
+      "epoch": 2.100840336134454,
+      "grad_norm": 1.414468765258789,
+      "learning_rate": 0.0006499,
+      "loss": 1.9276,
+      "step": 6500
+    },
+    {
+      "epoch": 2.1331609566903684,
+      "grad_norm": 1.050279140472412,
+      "learning_rate": 0.0006599,
+      "loss": 1.9068,
+      "step": 6600
+    },
+    {
+      "epoch": 2.165481577246283,
+      "grad_norm": 1.3866891860961914,
+      "learning_rate": 0.0006699000000000001,
+      "loss": 1.9283,
+      "step": 6700
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 1.0788445472717285,
+      "learning_rate": 0.0006799,
+      "loss": 1.9478,
+      "step": 6800
+    },
+    {
+      "epoch": 2.2301228183581125,
+      "grad_norm": 1.032123327255249,
+      "learning_rate": 0.0006899,
+      "loss": 1.9247,
+      "step": 6900
+    },
+    {
+      "epoch": 2.262443438914027,
+      "grad_norm": 1.3964852094650269,
+      "learning_rate": 0.0006998999999999999,
+      "loss": 1.9256,
+      "step": 7000
+    },
+    {
+      "epoch": 2.2947640594699417,
+      "grad_norm": 1.1780961751937866,
+      "learning_rate": 0.0007099,
+      "loss": 1.9308,
+      "step": 7100
+    },
+    {
+      "epoch": 2.3270846800258567,
+      "grad_norm": 1.3616302013397217,
+      "learning_rate": 0.0007199,
+      "loss": 1.9456,
+      "step": 7200
+    },
+    {
+      "epoch": 2.3594053005817712,
+      "grad_norm": 1.266806960105896,
+      "learning_rate": 0.0007299,
+      "loss": 1.9409,
+      "step": 7300
+    },
+    {
+      "epoch": 2.391725921137686,
+      "grad_norm": 1.1740697622299194,
+      "learning_rate": 0.0007399,
+      "loss": 1.9325,
+      "step": 7400
+    },
+    {
+      "epoch": 2.4240465416936003,
+      "grad_norm": 1.1045465469360352,
+      "learning_rate": 0.0007499000000000001,
+      "loss": 1.9336,
+      "step": 7500
+    },
+    {
+      "epoch": 2.4563671622495153,
+      "grad_norm": 1.4932281970977783,
+      "learning_rate": 0.0007599,
+      "loss": 1.9374,
+      "step": 7600
+    },
+    {
+      "epoch": 2.48868778280543,
+      "grad_norm": 1.0262266397476196,
+      "learning_rate": 0.0007699,
+      "loss": 1.9295,
+      "step": 7700
+    },
+    {
+      "epoch": 2.5210084033613445,
+      "grad_norm": 1.2064828872680664,
+      "learning_rate": 0.0007799,
+      "loss": 1.9178,
+      "step": 7800
+    },
+    {
+      "epoch": 2.553329023917259,
+      "grad_norm": 1.2155147790908813,
+      "learning_rate": 0.0007899000000000001,
+      "loss": 1.9659,
+      "step": 7900
+    },
+    {
+      "epoch": 2.585649644473174,
+      "grad_norm": 1.2359004020690918,
+      "learning_rate": 0.0007999000000000001,
+      "loss": 1.943,
+      "step": 8000
+    },
+    {
+      "epoch": 2.6179702650290886,
+      "grad_norm": 1.1473153829574585,
+      "learning_rate": 0.0008099,
+      "loss": 1.9309,
+      "step": 8100
+    },
+    {
+      "epoch": 2.650290885585003,
+      "grad_norm": 1.13937246799469,
+      "learning_rate": 0.0008198999999999999,
+      "loss": 1.9133,
+      "step": 8200
+    },
+    {
+      "epoch": 2.682611506140918,
+      "grad_norm": 1.46646249294281,
+      "learning_rate": 0.0008299,
+      "loss": 1.9746,
+      "step": 8300
+    },
+    {
+      "epoch": 2.7149321266968327,
+      "grad_norm": 1.3891866207122803,
+      "learning_rate": 0.0008399,
+      "loss": 1.9353,
+      "step": 8400
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 1.0530701875686646,
+      "learning_rate": 0.0008499,
+      "loss": 1.9152,
+      "step": 8500
+    },
+    {
+      "epoch": 2.779573367808662,
+      "grad_norm": 1.3477870225906372,
+      "learning_rate": 0.0008599,
+      "loss": 1.9334,
+      "step": 8600
+    },
+    {
+      "epoch": 2.8118939883645764,
+      "grad_norm": 1.0107545852661133,
+      "learning_rate": 0.0008699000000000001,
+      "loss": 1.9586,
+      "step": 8700
+    },
+    {
+      "epoch": 2.8442146089204914,
+      "grad_norm": 1.2573806047439575,
+      "learning_rate": 0.0008799000000000001,
+      "loss": 1.9526,
+      "step": 8800
+    },
+    {
+      "epoch": 2.876535229476406,
+      "grad_norm": 1.7003774642944336,
+      "learning_rate": 0.0008899,
+      "loss": 1.933,
+      "step": 8900
+    },
+    {
+      "epoch": 2.9088558500323205,
+      "grad_norm": 0.9627268314361572,
+      "learning_rate": 0.0008999,
+      "loss": 1.9588,
+      "step": 9000
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 1.1297580003738403,
+      "learning_rate": 0.0009099,
+      "loss": 1.961,
+      "step": 9100
+    },
+    {
+      "epoch": 2.97349709114415,
+      "grad_norm": 1.230385184288025,
+      "learning_rate": 0.0009199000000000001,
+      "loss": 1.9194,
+      "step": 9200
+    },
+    {
+      "epoch": 3.0058177117000646,
+      "grad_norm": 0.9041159152984619,
+      "learning_rate": 0.0009299,
+      "loss": 1.9607,
+      "step": 9300
+    },
+    {
+      "epoch": 3.038138332255979,
+      "grad_norm": 1.1520296335220337,
+      "learning_rate": 0.0009399,
+      "loss": 1.8419,
+      "step": 9400
+    },
+    {
+      "epoch": 3.070458952811894,
+      "grad_norm": 0.8856098651885986,
+      "learning_rate": 0.0009498999999999999,
+      "loss": 1.865,
+      "step": 9500
+    },
+    {
+      "epoch": 3.1027795733678087,
+      "grad_norm": 1.082818627357483,
+      "learning_rate": 0.0009599,
+      "loss": 1.8975,
+      "step": 9600
+    },
+    {
+      "epoch": 3.1351001939237233,
+      "grad_norm": 1.1733167171478271,
+      "learning_rate": 0.0009699,
+      "loss": 1.8623,
+      "step": 9700
+    },
+    {
+      "epoch": 3.167420814479638,
+      "grad_norm": 1.0028611421585083,
+      "learning_rate": 0.0009799,
+      "loss": 1.8665,
+      "step": 9800
+    },
+    {
+      "epoch": 3.199741435035553,
+      "grad_norm": 1.318609356880188,
+      "learning_rate": 0.0009899,
+      "loss": 1.8917,
+      "step": 9900
+    },
+    {
+      "epoch": 3.2320620555914674,
+      "grad_norm": 1.452332854270935,
+      "learning_rate": 0.0009999,
+      "loss": 1.8757,
+      "step": 10000
+    },
+    {
+      "epoch": 3.264382676147382,
+      "grad_norm": 1.2408781051635742,
+      "learning_rate": 0.001,
+      "loss": 1.825,
+      "step": 10100
+    },
+    {
+      "epoch": 3.2967032967032965,
+      "grad_norm": 0.9759476780891418,
+      "learning_rate": 0.001,
+      "loss": 1.8435,
+      "step": 10200
+    },
+    {
+      "epoch": 3.3290239172592115,
+      "grad_norm": 1.2983310222625732,
+      "learning_rate": 0.001,
+      "loss": 1.8448,
+      "step": 10300
+    },
+    {
+      "epoch": 3.361344537815126,
+      "grad_norm": 1.7093514204025269,
+      "learning_rate": 0.001,
+      "loss": 1.8535,
+      "step": 10400
+    },
+    {
+      "epoch": 3.3936651583710407,
+      "grad_norm": 0.990372359752655,
+      "learning_rate": 0.001,
+      "loss": 1.8681,
+      "step": 10500
+    },
+    {
+      "epoch": 3.425985778926955,
+      "grad_norm": 30.755449295043945,
+      "learning_rate": 0.001,
+      "loss": 1.8494,
+      "step": 10600
+    },
+    {
+      "epoch": 3.45830639948287,
+      "grad_norm": 1.0039843320846558,
+      "learning_rate": 0.001,
+      "loss": 1.8672,
+      "step": 10700
+    },
+    {
+      "epoch": 3.490627020038785,
+      "grad_norm": 1.0978604555130005,
+      "learning_rate": 0.001,
+      "loss": 1.8695,
+      "step": 10800
+    },
+    {
+      "epoch": 3.5229476405946993,
+      "grad_norm": 1.1592726707458496,
+      "learning_rate": 0.001,
+      "loss": 1.8769,
+      "step": 10900
+    },
+    {
+      "epoch": 3.555268261150614,
+      "grad_norm": 1.8122249841690063,
+      "learning_rate": 0.001,
+      "loss": 1.8361,
+      "step": 11000
+    },
+    {
+      "epoch": 3.587588881706529,
+      "grad_norm": 1.182767391204834,
+      "learning_rate": 0.001,
+      "loss": 1.866,
+      "step": 11100
+    },
+    {
+      "epoch": 3.6199095022624435,
+      "grad_norm": 1.7454653978347778,
+      "learning_rate": 0.001,
+      "loss": 1.8712,
+      "step": 11200
+    },
+    {
+      "epoch": 3.652230122818358,
+      "grad_norm": 1.3222342729568481,
+      "learning_rate": 0.001,
+      "loss": 1.8495,
+      "step": 11300
+    },
+    {
+      "epoch": 3.684550743374273,
+      "grad_norm": 1.1197530031204224,
+      "learning_rate": 0.001,
+      "loss": 1.8636,
+      "step": 11400
+    },
+    {
+      "epoch": 3.7168713639301876,
+      "grad_norm": 1.0089489221572876,
+      "learning_rate": 0.001,
+      "loss": 1.8698,
+      "step": 11500
+    },
+    {
+      "epoch": 3.749191984486102,
+      "grad_norm": 1.1160805225372314,
+      "learning_rate": 0.001,
+      "loss": 1.8626,
+      "step": 11600
+    },
+    {
+      "epoch": 3.7815126050420167,
+      "grad_norm": 1.6491235494613647,
+      "learning_rate": 0.001,
+      "loss": 1.8567,
+      "step": 11700
+    },
+    {
+      "epoch": 3.8138332255979313,
+      "grad_norm": 1.2563141584396362,
+      "learning_rate": 0.001,
+      "loss": 1.8366,
+      "step": 11800
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 1.25520658493042,
+      "learning_rate": 0.001,
+      "loss": 1.8873,
+      "step": 11900
+    },
+    {
+      "epoch": 3.878474466709761,
+      "grad_norm": 0.7658872604370117,
+      "learning_rate": 0.001,
+      "loss": 1.8845,
+      "step": 12000
+    },
+    {
+      "epoch": 3.9107950872656754,
+      "grad_norm": 1.1348446607589722,
+      "learning_rate": 0.001,
+      "loss": 1.8875,
+      "step": 12100
+    },
+    {
+      "epoch": 3.9431157078215904,
+      "grad_norm": 1.0871976613998413,
+      "learning_rate": 0.001,
+      "loss": 1.8727,
+      "step": 12200
+    },
+    {
+      "epoch": 3.975436328377505,
+      "grad_norm": 1.2896612882614136,
+      "learning_rate": 0.001,
+      "loss": 1.865,
+      "step": 12300
+    },
+    {
+      "epoch": 4.0077569489334195,
+      "grad_norm": 1.233090877532959,
+      "learning_rate": 0.001,
+      "loss": 1.8577,
+      "step": 12400
+    },
+    {
+      "epoch": 4.040077569489334,
+      "grad_norm": 1.2701627016067505,
+      "learning_rate": 0.001,
+      "loss": 1.7695,
+      "step": 12500
+    },
+    {
+      "epoch": 4.072398190045249,
+      "grad_norm": 0.9071168899536133,
+      "learning_rate": 0.001,
+      "loss": 1.7788,
+      "step": 12600
+    },
+    {
+      "epoch": 4.104718810601163,
+      "grad_norm": 0.851148247718811,
+      "learning_rate": 0.001,
+      "loss": 1.8106,
+      "step": 12700
+    },
+    {
+      "epoch": 4.137039431157079,
+      "grad_norm": 1.1509605646133423,
+      "learning_rate": 0.001,
+      "loss": 1.7861,
+      "step": 12800
+    },
+    {
+      "epoch": 4.169360051712993,
+      "grad_norm": 1.0977354049682617,
+      "learning_rate": 0.001,
+      "loss": 1.8142,
+      "step": 12900
+    },
+    {
+      "epoch": 4.201680672268908,
+      "grad_norm": 1.4111992120742798,
+      "learning_rate": 0.001,
+      "loss": 1.8264,
+      "step": 13000
+    },
+    {
+      "epoch": 4.234001292824822,
+      "grad_norm": 0.900395929813385,
+      "learning_rate": 0.001,
+      "loss": 1.8613,
+      "step": 13100
+    },
+    {
+      "epoch": 4.266321913380737,
+      "grad_norm": 1.2737363576889038,
+      "learning_rate": 0.001,
+      "loss": 1.7886,
+      "step": 13200
+    },
+    {
+      "epoch": 4.298642533936651,
+      "grad_norm": 1.435176968574524,
+      "learning_rate": 0.001,
+      "loss": 1.8192,
+      "step": 13300
+    },
+    {
+      "epoch": 4.330963154492566,
+      "grad_norm": 0.7658246755599976,
+      "learning_rate": 0.001,
+      "loss": 1.8316,
+      "step": 13400
+    },
+    {
+      "epoch": 4.3632837750484805,
+      "grad_norm": 1.1032602787017822,
+      "learning_rate": 0.001,
+      "loss": 1.8161,
+      "step": 13500
+    },
+    {
+      "epoch": 4.395604395604396,
+      "grad_norm": 1.1794319152832031,
+      "learning_rate": 0.001,
+      "loss": 1.7846,
+      "step": 13600
+    },
+    {
+      "epoch": 4.4279250161603105,
+      "grad_norm": 0.8004180192947388,
+      "learning_rate": 0.001,
+      "loss": 1.7987,
+      "step": 13700
+    },
+    {
+      "epoch": 4.460245636716225,
+      "grad_norm": 1.0232980251312256,
+      "learning_rate": 0.001,
+      "loss": 1.7836,
+      "step": 13800
+    },
+    {
+      "epoch": 4.49256625727214,
+      "grad_norm": 1.3699309825897217,
+      "learning_rate": 0.001,
+      "loss": 1.8018,
+      "step": 13900
+    },
+    {
+      "epoch": 4.524886877828054,
+      "grad_norm": 1.0403003692626953,
+      "learning_rate": 0.001,
+      "loss": 1.7962,
+      "step": 14000
+    },
+    {
+      "epoch": 4.557207498383969,
+      "grad_norm": 1.176257610321045,
+      "learning_rate": 0.001,
+      "loss": 1.8054,
+      "step": 14100
+    },
+    {
+      "epoch": 4.589528118939883,
+      "grad_norm": 0.7099631428718567,
+      "learning_rate": 0.001,
+      "loss": 1.8299,
+      "step": 14200
+    },
+    {
+      "epoch": 4.621848739495798,
+      "grad_norm": 0.7507422566413879,
+      "learning_rate": 0.001,
+      "loss": 1.8118,
+      "step": 14300
+    },
+    {
+      "epoch": 4.654169360051713,
+      "grad_norm": 0.9232106804847717,
+      "learning_rate": 0.001,
+      "loss": 1.8049,
+      "step": 14400
+    },
+    {
+      "epoch": 4.686489980607628,
+      "grad_norm": 1.074141263961792,
+      "learning_rate": 0.001,
+      "loss": 1.8488,
+      "step": 14500
+    },
+    {
+      "epoch": 4.7188106011635425,
+      "grad_norm": 2.9028780460357666,
+      "learning_rate": 0.001,
+      "loss": 1.8251,
+      "step": 14600
+    },
+    {
+      "epoch": 4.751131221719457,
+      "grad_norm": 0.8118910789489746,
+      "learning_rate": 0.001,
+      "loss": 1.8049,
+      "step": 14700
+    },
+    {
+      "epoch": 4.783451842275372,
+      "grad_norm": 0.8545799255371094,
+      "learning_rate": 0.001,
+      "loss": 1.8078,
+      "step": 14800
+    },
+    {
+      "epoch": 4.815772462831286,
+      "grad_norm": 1.1466169357299805,
+      "learning_rate": 0.001,
+      "loss": 1.7987,
+      "step": 14900
+    },
+    {
+      "epoch": 4.848093083387201,
+      "grad_norm": 1.126753568649292,
+      "learning_rate": 0.001,
+      "loss": 1.7949,
+      "step": 15000
+    },
+    {
+      "epoch": 4.880413703943116,
+      "grad_norm": 0.7958624362945557,
+      "learning_rate": 0.001,
+      "loss": 1.7877,
+      "step": 15100
+    },
+    {
+      "epoch": 4.912734324499031,
+      "grad_norm": 1.2399948835372925,
+      "learning_rate": 0.001,
+      "loss": 1.8136,
+      "step": 15200
+    },
+    {
+      "epoch": 4.945054945054945,
+      "grad_norm": 1.478546380996704,
+      "learning_rate": 0.001,
+      "loss": 1.8386,
+      "step": 15300
+    },
+    {
+      "epoch": 4.97737556561086,
+      "grad_norm": 1.4831817150115967,
+      "learning_rate": 0.001,
+      "loss": 1.828,
+      "step": 15400
+    },
+    {
+      "epoch": 5.009696186166774,
+      "grad_norm": 1.126725196838379,
+      "learning_rate": 0.001,
+      "loss": 1.7936,
+      "step": 15500
+    },
+    {
+      "epoch": 5.042016806722689,
+      "grad_norm": 0.9983710646629333,
+      "learning_rate": 0.001,
+      "loss": 1.6942,
+      "step": 15600
+    },
+    {
+      "epoch": 5.0743374272786035,
+      "grad_norm": 1.1261931657791138,
+      "learning_rate": 0.001,
+      "loss": 1.7046,
+      "step": 15700
+    },
+    {
+      "epoch": 5.106658047834518,
+      "grad_norm": 0.8670969605445862,
+      "learning_rate": 0.001,
+      "loss": 1.7256,
+      "step": 15800
+    },
+    {
+      "epoch": 5.1389786683904335,
+      "grad_norm": 1.0237650871276855,
+      "learning_rate": 0.001,
+      "loss": 1.7319,
+      "step": 15900
+    },
+    {
+      "epoch": 5.171299288946348,
+      "grad_norm": 0.8006519079208374,
+      "learning_rate": 0.001,
+      "loss": 1.7529,
+      "step": 16000
+    },
+    {
+      "epoch": 5.203619909502263,
+      "grad_norm": 0.9771319627761841,
+      "learning_rate": 0.001,
+      "loss": 1.7032,
+      "step": 16100
+    },
+    {
+      "epoch": 5.235940530058177,
+      "grad_norm": 0.9730778336524963,
+      "learning_rate": 0.001,
+      "loss": 1.7376,
+      "step": 16200
+    },
+    {
+      "epoch": 5.268261150614092,
+      "grad_norm": 0.8748031854629517,
+      "learning_rate": 0.001,
+      "loss": 1.7531,
+      "step": 16300
+    },
+    {
+      "epoch": 5.300581771170006,
+      "grad_norm": 0.7881172895431519,
+      "learning_rate": 0.001,
+      "loss": 1.7506,
+      "step": 16400
+    },
+    {
+      "epoch": 5.332902391725921,
+      "grad_norm": 2.841874599456787,
+      "learning_rate": 0.001,
+      "loss": 1.7433,
+      "step": 16500
+    },
+    {
+      "epoch": 5.365223012281835,
+      "grad_norm": 0.8375452756881714,
+      "learning_rate": 0.001,
+      "loss": 1.7232,
+      "step": 16600
+    },
+    {
+      "epoch": 5.397543632837751,
+      "grad_norm": 1.30898916721344,
+      "learning_rate": 0.001,
+      "loss": 1.7317,
+      "step": 16700
+    },
+    {
+      "epoch": 5.429864253393665,
+      "grad_norm": 0.9335517883300781,
+      "learning_rate": 0.001,
+      "loss": 1.7807,
+      "step": 16800
+    },
+    {
+      "epoch": 5.46218487394958,
+      "grad_norm": 1.097780704498291,
+      "learning_rate": 0.001,
+      "loss": 1.7398,
+      "step": 16900
+    },
+    {
+      "epoch": 5.4945054945054945,
+      "grad_norm": 0.9672789573669434,
+      "learning_rate": 0.001,
+      "loss": 1.7514,
+      "step": 17000
+    },
+    {
+      "epoch": 5.526826115061409,
+      "grad_norm": 0.6645662784576416,
+      "learning_rate": 0.001,
+      "loss": 1.7464,
+      "step": 17100
+    },
+    {
+      "epoch": 5.559146735617324,
+      "grad_norm": 0.8699595928192139,
+      "learning_rate": 0.001,
+      "loss": 1.7568,
+      "step": 17200
+    },
+    {
+      "epoch": 5.591467356173238,
+      "grad_norm": 0.897167980670929,
+      "learning_rate": 0.001,
+      "loss": 1.7206,
+      "step": 17300
+    },
+    {
+      "epoch": 5.623787976729153,
+      "grad_norm": 0.8125122785568237,
+      "learning_rate": 0.001,
+      "loss": 1.731,
+      "step": 17400
+    },
+    {
+      "epoch": 5.656108597285068,
+      "grad_norm": 1.1313683986663818,
+      "learning_rate": 0.001,
+      "loss": 1.7657,
+      "step": 17500
+    },
+    {
+      "epoch": 5.688429217840983,
+      "grad_norm": 0.751796543598175,
+      "learning_rate": 0.001,
+      "loss": 1.7696,
+      "step": 17600
+    },
+    {
+      "epoch": 5.720749838396897,
+      "grad_norm": 0.7507619857788086,
+      "learning_rate": 0.001,
+      "loss": 1.7562,
+      "step": 17700
+    },
+    {
+      "epoch": 5.753070458952812,
+      "grad_norm": 0.7618012428283691,
+      "learning_rate": 0.001,
+      "loss": 1.7589,
+      "step": 17800
+    },
+    {
+      "epoch": 5.785391079508726,
+      "grad_norm": 1.1747350692749023,
+      "learning_rate": 0.001,
+      "loss": 1.7558,
+      "step": 17900
+    },
+    {
+      "epoch": 5.817711700064641,
+      "grad_norm": 0.8985292911529541,
+      "learning_rate": 0.001,
+      "loss": 1.7461,
+      "step": 18000
+    },
+    {
+      "epoch": 5.850032320620556,
+      "grad_norm": 0.7552205324172974,
+      "learning_rate": 0.001,
+      "loss": 1.7702,
+      "step": 18100
+    },
+    {
+      "epoch": 5.882352941176471,
+      "grad_norm": 0.8019037246704102,
+      "learning_rate": 0.001,
+      "loss": 1.7779,
+      "step": 18200
+    },
+    {
+      "epoch": 5.914673561732386,
+      "grad_norm": 0.9410333037376404,
+      "learning_rate": 0.001,
+      "loss": 1.7969,
+      "step": 18300
+    },
+    {
+      "epoch": 5.9469941822883,
+      "grad_norm": 1.1927149295806885,
+      "learning_rate": 0.001,
+      "loss": 1.7573,
+      "step": 18400
+    },
+    {
+      "epoch": 5.979314802844215,
+      "grad_norm": 1.060541033744812,
+      "learning_rate": 0.001,
+      "loss": 1.7674,
+      "step": 18500
+    },
+    {
+      "epoch": 6.011635423400129,
+      "grad_norm": 1.0835535526275635,
+      "learning_rate": 0.001,
+      "loss": 1.6796,
+      "step": 18600
+    },
+    {
+      "epoch": 6.043956043956044,
+      "grad_norm": 0.9343781471252441,
+      "learning_rate": 0.001,
+      "loss": 1.6642,
+      "step": 18700
+    },
+    {
+      "epoch": 6.076276664511958,
+      "grad_norm": 0.9074199795722961,
+      "learning_rate": 0.001,
+      "loss": 1.6493,
+      "step": 18800
+    },
+    {
+      "epoch": 6.108597285067873,
+      "grad_norm": 0.9927520155906677,
+      "learning_rate": 0.001,
+      "loss": 1.6346,
+      "step": 18900
+    },
+    {
+      "epoch": 6.140917905623788,
+      "grad_norm": 0.7810798287391663,
+      "learning_rate": 0.001,
+      "loss": 1.6908,
+      "step": 19000
+    },
+    {
+      "epoch": 6.173238526179703,
+      "grad_norm": 0.6857879757881165,
+      "learning_rate": 0.001,
+      "loss": 1.6729,
+      "step": 19100
+    },
+    {
+      "epoch": 6.2055591467356175,
+      "grad_norm": 0.8595390915870667,
+      "learning_rate": 0.001,
+      "loss": 1.6679,
+      "step": 19200
+    },
+    {
+      "epoch": 6.237879767291532,
+      "grad_norm": 0.9147341847419739,
+      "learning_rate": 0.001,
+      "loss": 1.7043,
+      "step": 19300
+    },
+    {
+      "epoch": 6.270200387847447,
+      "grad_norm": 0.9340370893478394,
+      "learning_rate": 0.001,
+      "loss": 1.7241,
+      "step": 19400
+    },
+    {
+      "epoch": 6.302521008403361,
+      "grad_norm": 1.0489479303359985,
+      "learning_rate": 0.001,
+      "loss": 1.6671,
+      "step": 19500
+    },
+    {
+      "epoch": 6.334841628959276,
+      "grad_norm": 0.7707657814025879,
+      "learning_rate": 0.001,
+      "loss": 1.6772,
+      "step": 19600
+    },
+    {
+      "epoch": 6.36716224951519,
+      "grad_norm": 0.748698353767395,
+      "learning_rate": 0.001,
+      "loss": 1.7033,
+      "step": 19700
+    },
+    {
+      "epoch": 6.399482870071106,
+      "grad_norm": 0.7804653644561768,
+      "learning_rate": 0.001,
+      "loss": 1.7091,
+      "step": 19800
+    },
+    {
+      "epoch": 6.43180349062702,
+      "grad_norm": 1.2131924629211426,
+      "learning_rate": 0.001,
+      "loss": 1.6997,
+      "step": 19900
+    },
+    {
+      "epoch": 6.464124111182935,
+      "grad_norm": 0.9598875641822815,
+      "learning_rate": 0.001,
+      "loss": 1.676,
+      "step": 20000
+    },
+    {
+      "epoch": 6.496444731738849,
+      "grad_norm": 1.1061313152313232,
+      "learning_rate": 0.001,
+      "loss": 1.7048,
+      "step": 20100
+    },
+    {
+      "epoch": 6.528765352294764,
+      "grad_norm": 0.8944084644317627,
+      "learning_rate": 0.001,
+      "loss": 1.6876,
+      "step": 20200
+    },
+    {
+      "epoch": 6.5610859728506785,
+      "grad_norm": 1.101098895072937,
+      "learning_rate": 0.001,
+      "loss": 1.6805,
+      "step": 20300
+    },
+    {
+      "epoch": 6.593406593406593,
+      "grad_norm": 0.9273419380187988,
+      "learning_rate": 0.001,
+      "loss": 1.6776,
+      "step": 20400
+    },
+    {
+      "epoch": 6.625727213962508,
+      "grad_norm": 1.4434058666229248,
+      "learning_rate": 0.001,
+      "loss": 1.6771,
+      "step": 20500
+    },
+    {
+      "epoch": 6.658047834518423,
+      "grad_norm": 0.9152409434318542,
+      "learning_rate": 0.001,
+      "loss": 1.685,
+      "step": 20600
+    },
+    {
+      "epoch": 6.690368455074338,
+      "grad_norm": 1.1112756729125977,
+      "learning_rate": 0.001,
+      "loss": 1.6881,
+      "step": 20700
+    },
+    {
+      "epoch": 6.722689075630252,
+      "grad_norm": 0.9489529132843018,
+      "learning_rate": 0.001,
+      "loss": 1.6868,
+      "step": 20800
+    },
+    {
+      "epoch": 6.755009696186167,
+      "grad_norm": 0.7932117581367493,
+      "learning_rate": 0.001,
+      "loss": 1.6925,
+      "step": 20900
+    },
+    {
+      "epoch": 6.787330316742081,
+      "grad_norm": 0.893706202507019,
+      "learning_rate": 0.001,
+      "loss": 1.7308,
+      "step": 21000
+    },
+    {
+      "epoch": 6.819650937297996,
+      "grad_norm": 1.0104913711547852,
+      "learning_rate": 0.001,
+      "loss": 1.7075,
+      "step": 21100
+    },
+    {
+      "epoch": 6.85197155785391,
+      "grad_norm": 0.8653020858764648,
+      "learning_rate": 0.001,
+      "loss": 1.7352,
+      "step": 21200
+    },
+    {
+      "epoch": 6.884292178409826,
+      "grad_norm": 0.8224719166755676,
+      "learning_rate": 0.001,
+      "loss": 1.7129,
+      "step": 21300
+    },
+    {
+      "epoch": 6.91661279896574,
+      "grad_norm": 1.0622234344482422,
+      "learning_rate": 0.001,
+      "loss": 1.7153,
+      "step": 21400
+    },
+    {
+      "epoch": 6.948933419521655,
+      "grad_norm": 1.1205211877822876,
+      "learning_rate": 0.001,
+      "loss": 1.741,
+      "step": 21500
+    },
+    {
+      "epoch": 6.98125404007757,
+      "grad_norm": 1.1996233463287354,
+      "learning_rate": 0.001,
+      "loss": 1.7005,
+      "step": 21600
+    },
+    {
+      "epoch": 7.013574660633484,
+      "grad_norm": 1.043961763381958,
+      "learning_rate": 0.001,
+      "loss": 1.6296,
+      "step": 21700
+    },
+    {
+      "epoch": 7.045895281189399,
+      "grad_norm": 0.9840327501296997,
+      "learning_rate": 0.001,
+      "loss": 1.5904,
+      "step": 21800
+    },
+    {
+      "epoch": 7.078215901745313,
+      "grad_norm": 1.1553939580917358,
+      "learning_rate": 0.001,
+      "loss": 1.5837,
+      "step": 21900
+    },
+    {
+      "epoch": 7.110536522301228,
+      "grad_norm": 0.9404619336128235,
+      "learning_rate": 0.001,
+      "loss": 1.6182,
+      "step": 22000
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.7575173377990723,
+      "learning_rate": 0.001,
+      "loss": 1.6091,
+      "step": 22100
+    },
+    {
+      "epoch": 7.175177763413058,
+      "grad_norm": 0.8651929497718811,
+      "learning_rate": 0.001,
+      "loss": 1.6097,
+      "step": 22200
+    },
+    {
+      "epoch": 7.207498383968972,
+      "grad_norm": 0.7588925361633301,
+      "learning_rate": 0.001,
+      "loss": 1.6596,
+      "step": 22300
+    },
+    {
+      "epoch": 7.239819004524887,
+      "grad_norm": 0.9948194026947021,
+      "learning_rate": 0.001,
+      "loss": 1.6307,
+      "step": 22400
+    },
+    {
+      "epoch": 7.2721396250808015,
+      "grad_norm": 0.7452443242073059,
+      "learning_rate": 0.001,
+      "loss": 1.6545,
+      "step": 22500
+    },
+    {
+      "epoch": 7.304460245636716,
+      "grad_norm": 0.7591370940208435,
+      "learning_rate": 0.001,
+      "loss": 1.6257,
+      "step": 22600
+    },
+    {
+      "epoch": 7.336780866192631,
+      "grad_norm": 1.1810401678085327,
+      "learning_rate": 0.001,
+      "loss": 1.6381,
+      "step": 22700
+    },
+    {
+      "epoch": 7.369101486748546,
+      "grad_norm": 0.785140335559845,
+      "learning_rate": 0.001,
+      "loss": 1.6453,
+      "step": 22800
+    },
+    {
+      "epoch": 7.401422107304461,
+      "grad_norm": 1.0665206909179688,
+      "learning_rate": 0.001,
+      "loss": 1.656,
+      "step": 22900
+    },
+    {
+      "epoch": 7.433742727860375,
+      "grad_norm": 1.0718743801116943,
+      "learning_rate": 0.001,
+      "loss": 1.6676,
+      "step": 23000
+    },
+    {
+      "epoch": 7.46606334841629,
+      "grad_norm": 0.8834295868873596,
+      "learning_rate": 0.001,
+      "loss": 1.6464,
+      "step": 23100
+    },
+    {
+      "epoch": 7.498383968972204,
+      "grad_norm": 0.9542046189308167,
+      "learning_rate": 0.001,
+      "loss": 1.6544,
+      "step": 23200
+    },
+    {
+      "epoch": 7.530704589528119,
+      "grad_norm": 0.7912190556526184,
+      "learning_rate": 0.001,
+      "loss": 1.6348,
+      "step": 23300
+    },
+    {
+      "epoch": 7.563025210084033,
+      "grad_norm": 46.410430908203125,
+      "learning_rate": 0.001,
+      "loss": 1.6686,
+      "step": 23400
+    },
+    {
+      "epoch": 7.595345830639948,
+      "grad_norm": 0.779313325881958,
+      "learning_rate": 0.001,
+      "loss": 1.6581,
+      "step": 23500
+    },
+    {
+      "epoch": 7.6276664511958625,
+      "grad_norm": 1.3436274528503418,
+      "learning_rate": 0.001,
+      "loss": 1.6553,
+      "step": 23600
+    },
+    {
+      "epoch": 7.659987071751778,
+      "grad_norm": 0.8226367831230164,
+      "learning_rate": 0.001,
+      "loss": 1.6682,
+      "step": 23700
+    },
+    {
+      "epoch": 7.6923076923076925,
+      "grad_norm": 0.8935831189155579,
+      "learning_rate": 0.001,
+      "loss": 1.6315,
+      "step": 23800
+    },
+    {
+      "epoch": 7.724628312863607,
+      "grad_norm": 1.181753158569336,
+      "learning_rate": 0.001,
+      "loss": 1.6274,
+      "step": 23900
+    },
+    {
+      "epoch": 7.756948933419522,
+      "grad_norm": 0.674148440361023,
+      "learning_rate": 0.001,
+      "loss": 1.6386,
+      "step": 24000
+    },
+    {
+      "epoch": 7.789269553975436,
+      "grad_norm": 1.0500868558883667,
+      "learning_rate": 0.001,
+      "loss": 1.6534,
+      "step": 24100
+    },
+    {
+      "epoch": 7.821590174531351,
+      "grad_norm": 0.9021979570388794,
+      "learning_rate": 0.001,
+      "loss": 1.6714,
+      "step": 24200
+    },
+    {
+      "epoch": 7.853910795087265,
+      "grad_norm": 1.0396559238433838,
+      "learning_rate": 0.001,
+      "loss": 1.6539,
+      "step": 24300
+    },
+    {
+      "epoch": 7.886231415643181,
+      "grad_norm": 0.9892787337303162,
+      "learning_rate": 0.001,
+      "loss": 1.6734,
+      "step": 24400
+    },
+    {
+      "epoch": 7.918552036199095,
+      "grad_norm": 0.8489758968353271,
+      "learning_rate": 0.001,
+      "loss": 1.6485,
+      "step": 24500
+    },
+    {
+      "epoch": 7.95087265675501,
+      "grad_norm": 1.0946519374847412,
+      "learning_rate": 0.001,
+      "loss": 1.6992,
+      "step": 24600
+    },
+    {
+      "epoch": 7.983193277310924,
+      "grad_norm": 0.9034314155578613,
+      "learning_rate": 0.001,
+      "loss": 1.6708,
+      "step": 24700
+    },
+    {
+      "epoch": 8.015513897866839,
+      "grad_norm": 0.7086787819862366,
+      "learning_rate": 0.001,
+      "loss": 1.548,
+      "step": 24800
+    },
+    {
+      "epoch": 8.047834518422754,
+      "grad_norm": 0.42088550329208374,
+      "learning_rate": 0.001,
+      "loss": 1.5713,
+      "step": 24900
+    },
+    {
+      "epoch": 8.080155138978668,
+      "grad_norm": 0.5306487083435059,
+      "learning_rate": 0.001,
+      "loss": 1.568,
+      "step": 25000
+    },
+    {
+      "epoch": 8.112475759534583,
+      "grad_norm": 0.7567152976989746,
+      "learning_rate": 0.001,
+      "loss": 1.5902,
+      "step": 25100
+    },
+    {
+      "epoch": 8.144796380090497,
+      "grad_norm": 0.8452807664871216,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 25200
+    },
+    {
+      "epoch": 8.177117000646412,
+      "grad_norm": 0.7313001751899719,
+      "learning_rate": 0.001,
+      "loss": 1.5539,
+      "step": 25300
+    },
+    {
+      "epoch": 8.209437621202326,
+      "grad_norm": 0.8463912606239319,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 25400
+    },
+    {
+      "epoch": 8.241758241758241,
+      "grad_norm": 0.7595904469490051,
+      "learning_rate": 0.001,
+      "loss": 1.5831,
+      "step": 25500
+    },
+    {
+      "epoch": 8.274078862314157,
+      "grad_norm": 0.8072571158409119,
+      "learning_rate": 0.001,
+      "loss": 1.5837,
+      "step": 25600
+    },
+    {
+      "epoch": 8.306399482870072,
+      "grad_norm": 0.5768963694572449,
+      "learning_rate": 0.001,
+      "loss": 1.589,
+      "step": 25700
+    },
+    {
+      "epoch": 8.338720103425986,
+      "grad_norm": 0.7995852828025818,
+      "learning_rate": 0.001,
+      "loss": 1.5792,
+      "step": 25800
+    },
+    {
+      "epoch": 8.371040723981901,
+      "grad_norm": 0.8774004578590393,
+      "learning_rate": 0.001,
+      "loss": 1.5867,
+      "step": 25900
+    },
+    {
+      "epoch": 8.403361344537815,
+      "grad_norm": 1.035139560699463,
+      "learning_rate": 0.001,
+      "loss": 1.5771,
+      "step": 26000
+    },
+    {
+      "epoch": 8.43568196509373,
+      "grad_norm": 0.6128062605857849,
+      "learning_rate": 0.001,
+      "loss": 1.573,
+      "step": 26100
+    },
+    {
+      "epoch": 8.468002585649645,
+      "grad_norm": 0.6861528158187866,
+      "learning_rate": 0.001,
+      "loss": 1.5973,
+      "step": 26200
+    },
+    {
+      "epoch": 8.50032320620556,
+      "grad_norm": 0.6741127371788025,
+      "learning_rate": 0.001,
+      "loss": 1.6055,
+      "step": 26300
+    },
+    {
+      "epoch": 8.532643826761474,
+      "grad_norm": 0.7249501943588257,
+      "learning_rate": 0.001,
+      "loss": 1.6032,
+      "step": 26400
+    },
+    {
+      "epoch": 8.564964447317388,
+      "grad_norm": 0.5071559548377991,
+      "learning_rate": 0.001,
+      "loss": 1.6135,
+      "step": 26500
+    },
+    {
+      "epoch": 8.597285067873303,
+      "grad_norm": 0.6695942282676697,
+      "learning_rate": 0.001,
+      "loss": 1.6317,
+      "step": 26600
+    },
+    {
+      "epoch": 8.629605688429217,
+      "grad_norm": 0.8934532999992371,
+      "learning_rate": 0.001,
+      "loss": 1.6195,
+      "step": 26700
+    },
+    {
+      "epoch": 8.661926308985132,
+      "grad_norm": 0.6500742435455322,
+      "learning_rate": 0.001,
+      "loss": 1.5722,
+      "step": 26800
+    },
+    {
+      "epoch": 8.694246929541046,
+      "grad_norm": 0.6032618880271912,
+      "learning_rate": 0.001,
+      "loss": 1.5891,
+      "step": 26900
+    },
+    {
+      "epoch": 8.726567550096961,
+      "grad_norm": 0.8172997236251831,
+      "learning_rate": 0.001,
+      "loss": 1.6224,
+      "step": 27000
+    },
+    {
+      "epoch": 8.758888170652877,
+      "grad_norm": 0.7582260370254517,
+      "learning_rate": 0.001,
+      "loss": 1.6274,
+      "step": 27100
+    },
+    {
+      "epoch": 8.791208791208792,
+      "grad_norm": 0.7059329152107239,
+      "learning_rate": 0.001,
+      "loss": 1.6109,
+      "step": 27200
+    },
+    {
+      "epoch": 8.823529411764707,
+      "grad_norm": 0.8909659385681152,
+      "learning_rate": 0.001,
+      "loss": 1.6163,
+      "step": 27300
+    },
+    {
+      "epoch": 8.855850032320621,
+      "grad_norm": 0.6082348227500916,
+      "learning_rate": 0.001,
+      "loss": 1.6392,
+      "step": 27400
+    },
+    {
+      "epoch": 8.888170652876536,
+      "grad_norm": 0.6296877861022949,
+      "learning_rate": 0.001,
+      "loss": 1.6235,
+      "step": 27500
+    },
+    {
+      "epoch": 8.92049127343245,
+      "grad_norm": 0.5239854454994202,
+      "learning_rate": 0.001,
+      "loss": 1.5809,
+      "step": 27600
+    },
+    {
+      "epoch": 8.952811893988365,
+      "grad_norm": 0.5151014924049377,
+      "learning_rate": 0.001,
+      "loss": 1.6231,
+      "step": 27700
+    },
+    {
+      "epoch": 8.98513251454428,
+      "grad_norm": 0.6107050180435181,
+      "learning_rate": 0.001,
+      "loss": 1.6294,
+      "step": 27800
+    },
+    {
+      "epoch": 9.017453135100194,
+      "grad_norm": 1.571207880973816,
+      "learning_rate": 0.001,
+      "loss": 1.5311,
+      "step": 27900
+    },
+    {
+      "epoch": 9.049773755656108,
+      "grad_norm": 1.3317564725875854,
+      "learning_rate": 0.001,
+      "loss": 1.5232,
+      "step": 28000
+    },
+    {
+      "epoch": 9.082094376212023,
+      "grad_norm": 0.9770411252975464,
+      "learning_rate": 0.001,
+      "loss": 1.5323,
+      "step": 28100
+    },
+    {
+      "epoch": 9.114414996767938,
+      "grad_norm": 1.5269267559051514,
+      "learning_rate": 0.001,
+      "loss": 1.538,
+      "step": 28200
+    },
+    {
+      "epoch": 9.146735617323852,
+      "grad_norm": 1.4465818405151367,
+      "learning_rate": 0.001,
+      "loss": 1.5356,
+      "step": 28300
+    },
+    {
+      "epoch": 9.179056237879767,
+      "grad_norm": 1.4123685359954834,
+      "learning_rate": 0.001,
+      "loss": 1.5354,
+      "step": 28400
+    },
+    {
+      "epoch": 9.211376858435681,
+      "grad_norm": 1.1594544649124146,
+      "learning_rate": 0.001,
+      "loss": 1.4929,
+      "step": 28500
+    },
+    {
+      "epoch": 9.243697478991596,
+      "grad_norm": 1.5278046131134033,
+      "learning_rate": 0.001,
+      "loss": 1.5353,
+      "step": 28600
+    },
+    {
+      "epoch": 9.276018099547512,
+      "grad_norm": 1.3489573001861572,
+      "learning_rate": 0.001,
+      "loss": 1.5223,
+      "step": 28700
+    },
+    {
+      "epoch": 9.308338720103427,
+      "grad_norm": 0.9945598244667053,
+      "learning_rate": 0.001,
+      "loss": 1.565,
+      "step": 28800
+    },
+    {
+      "epoch": 9.340659340659341,
+      "grad_norm": 1.3939306735992432,
+      "learning_rate": 0.001,
+      "loss": 1.5747,
+      "step": 28900
+    },
+    {
+      "epoch": 9.372979961215256,
+      "grad_norm": 1.0500473976135254,
+      "learning_rate": 0.001,
+      "loss": 1.5677,
+      "step": 29000
+    },
+    {
+      "epoch": 9.40530058177117,
+      "grad_norm": 1.1538209915161133,
+      "learning_rate": 0.001,
+      "loss": 1.5602,
+      "step": 29100
+    },
+    {
+      "epoch": 9.437621202327085,
+      "grad_norm": 1.2732844352722168,
+      "learning_rate": 0.001,
+      "loss": 1.5582,
+      "step": 29200
+    },
+    {
+      "epoch": 9.469941822883,
+      "grad_norm": 1.4475446939468384,
+      "learning_rate": 0.001,
+      "loss": 1.5542,
+      "step": 29300
+    },
+    {
+      "epoch": 9.502262443438914,
+      "grad_norm": 1.5488345623016357,
+      "learning_rate": 0.001,
+      "loss": 1.5589,
+      "step": 29400
+    },
+    {
+      "epoch": 9.534583063994829,
+      "grad_norm": 0.9935588836669922,
+      "learning_rate": 0.001,
+      "loss": 1.5645,
+      "step": 29500
+    },
+    {
+      "epoch": 9.566903684550743,
+      "grad_norm": 1.3134706020355225,
+      "learning_rate": 0.001,
+      "loss": 1.5479,
+      "step": 29600
+    },
+    {
+      "epoch": 9.599224305106658,
+      "grad_norm": 1.4117892980575562,
+      "learning_rate": 0.001,
+      "loss": 1.5534,
+      "step": 29700
+    },
+    {
+      "epoch": 9.631544925662572,
+      "grad_norm": 1.0989189147949219,
+      "learning_rate": 0.001,
+      "loss": 1.5924,
+      "step": 29800
+    },
+    {
+      "epoch": 9.663865546218487,
+      "grad_norm": 0.9962360858917236,
+      "learning_rate": 0.001,
+      "loss": 1.5521,
+      "step": 29900
+    },
+    {
+      "epoch": 9.696186166774401,
+      "grad_norm": 1.0200791358947754,
+      "learning_rate": 0.001,
+      "loss": 1.5421,
+      "step": 30000
+    },
+    {
+      "epoch": 9.728506787330316,
+      "grad_norm": 1.2421667575836182,
+      "learning_rate": 0.001,
+      "loss": 1.5971,
+      "step": 30100
+    },
+    {
+      "epoch": 9.760827407886232,
+      "grad_norm": 1.2540074586868286,
+      "learning_rate": 0.001,
+      "loss": 1.562,
+      "step": 30200
+    },
+    {
+      "epoch": 9.793148028442147,
+      "grad_norm": 1.1804842948913574,
+      "learning_rate": 0.001,
+      "loss": 1.5673,
+      "step": 30300
+    },
+    {
+      "epoch": 9.825468648998061,
+      "grad_norm": 1.2499170303344727,
+      "learning_rate": 0.001,
+      "loss": 1.5783,
+      "step": 30400
+    },
+    {
+      "epoch": 9.857789269553976,
+      "grad_norm": 1.1028128862380981,
+      "learning_rate": 0.001,
+      "loss": 1.5888,
+      "step": 30500
+    },
+    {
+      "epoch": 9.89010989010989,
+      "grad_norm": 1.251753330230713,
+      "learning_rate": 0.001,
+      "loss": 1.5749,
+      "step": 30600
+    },
+    {
+      "epoch": 9.922430510665805,
+      "grad_norm": 1.2316412925720215,
+      "learning_rate": 0.001,
+      "loss": 1.5755,
+      "step": 30700
+    },
+    {
+      "epoch": 9.95475113122172,
+      "grad_norm": 0.9406437873840332,
+      "learning_rate": 0.001,
+      "loss": 1.5741,
+      "step": 30800
+    },
+    {
+      "epoch": 9.987071751777634,
+      "grad_norm": 1.3652178049087524,
+      "learning_rate": 0.001,
+      "loss": 1.5867,
+      "step": 30900
+    },
+    {
+      "epoch": 10.019392372333549,
+      "grad_norm": 1.1299371719360352,
+      "learning_rate": 0.001,
+      "loss": 1.5492,
+      "step": 31000
+    },
+    {
+      "epoch": 10.051712992889463,
+      "grad_norm": 1.1812490224838257,
+      "learning_rate": 0.001,
+      "loss": 1.4886,
+      "step": 31100
+    },
+    {
+      "epoch": 10.084033613445378,
+      "grad_norm": 1.179519534111023,
+      "learning_rate": 0.001,
+      "loss": 1.4934,
+      "step": 31200
+    },
+    {
+      "epoch": 10.116354234001292,
+      "grad_norm": 1.120004653930664,
+      "learning_rate": 0.001,
+      "loss": 1.4879,
+      "step": 31300
+    },
+    {
+      "epoch": 10.148674854557207,
+      "grad_norm": 1.0279499292373657,
+      "learning_rate": 0.001,
+      "loss": 1.4963,
+      "step": 31400
+    },
+    {
+      "epoch": 10.180995475113122,
+      "grad_norm": 1.1619842052459717,
+      "learning_rate": 0.001,
+      "loss": 1.5114,
+      "step": 31500
+    },
+    {
+      "epoch": 10.213316095669036,
+      "grad_norm": 1.1536353826522827,
+      "learning_rate": 0.001,
+      "loss": 1.4935,
+      "step": 31600
+    },
+    {
+      "epoch": 10.24563671622495,
+      "grad_norm": 1.0345031023025513,
+      "learning_rate": 0.001,
+      "loss": 1.506,
+      "step": 31700
+    },
+    {
+      "epoch": 10.277957336780867,
+      "grad_norm": 1.0397800207138062,
+      "learning_rate": 0.001,
+      "loss": 1.5026,
+      "step": 31800
+    },
+    {
+      "epoch": 10.310277957336782,
+      "grad_norm": 0.8596133589744568,
+      "learning_rate": 0.001,
+      "loss": 1.5256,
+      "step": 31900
+    },
+    {
+      "epoch": 10.342598577892696,
+      "grad_norm": 1.1593185663223267,
+      "learning_rate": 0.001,
+      "loss": 1.5055,
+      "step": 32000
+    },
+    {
+      "epoch": 10.37491919844861,
+      "grad_norm": 1.0919194221496582,
+      "learning_rate": 0.001,
+      "loss": 1.4956,
+      "step": 32100
+    },
+    {
+      "epoch": 10.407239819004525,
+      "grad_norm": 1.1217265129089355,
+      "learning_rate": 0.001,
+      "loss": 1.5337,
+      "step": 32200
+    },
+    {
+      "epoch": 10.43956043956044,
+      "grad_norm": 0.9850316643714905,
+      "learning_rate": 0.001,
+      "loss": 1.516,
+      "step": 32300
+    },
+    {
+      "epoch": 10.471881060116354,
+      "grad_norm": 1.0271371603012085,
+      "learning_rate": 0.001,
+      "loss": 1.5346,
+      "step": 32400
+    },
+    {
+      "epoch": 10.504201680672269,
+      "grad_norm": 1.0458983182907104,
+      "learning_rate": 0.001,
+      "loss": 1.5167,
+      "step": 32500
+    },
+    {
+      "epoch": 10.536522301228183,
+      "grad_norm": 1.0349985361099243,
+      "learning_rate": 0.001,
+      "loss": 1.5203,
+      "step": 32600
+    },
+    {
+      "epoch": 10.568842921784098,
+      "grad_norm": 1.091556191444397,
+      "learning_rate": 0.001,
+      "loss": 1.5097,
+      "step": 32700
+    },
+    {
+      "epoch": 10.601163542340013,
+      "grad_norm": 1.1958649158477783,
+      "learning_rate": 0.001,
+      "loss": 1.5038,
+      "step": 32800
+    },
+    {
+      "epoch": 10.633484162895927,
+      "grad_norm": 1.1409047842025757,
+      "learning_rate": 0.001,
+      "loss": 1.4975,
+      "step": 32900
+    },
+    {
+      "epoch": 10.665804783451842,
+      "grad_norm": 1.205556035041809,
+      "learning_rate": 0.001,
+      "loss": 1.5596,
+      "step": 33000
+    },
+    {
+      "epoch": 10.698125404007756,
+      "grad_norm": 1.0502017736434937,
+      "learning_rate": 0.001,
+      "loss": 1.5111,
+      "step": 33100
+    },
+    {
+      "epoch": 10.73044602456367,
+      "grad_norm": 1.3859450817108154,
+      "learning_rate": 0.001,
+      "loss": 1.508,
+      "step": 33200
+    },
+    {
+      "epoch": 10.762766645119587,
+      "grad_norm": 1.0951238870620728,
+      "learning_rate": 0.001,
+      "loss": 1.5417,
+      "step": 33300
+    },
+    {
+      "epoch": 10.795087265675502,
+      "grad_norm": 2.344174385070801,
+      "learning_rate": 0.001,
+      "loss": 1.5599,
+      "step": 33400
+    },
+    {
+      "epoch": 10.827407886231416,
+      "grad_norm": 0.9748075604438782,
+      "learning_rate": 0.001,
+      "loss": 1.5558,
+      "step": 33500
+    },
+    {
+      "epoch": 10.85972850678733,
+      "grad_norm": 0.9509823322296143,
+      "learning_rate": 0.001,
+      "loss": 1.5253,
+      "step": 33600
+    },
+    {
+      "epoch": 10.892049127343245,
+      "grad_norm": 0.9078757166862488,
+      "learning_rate": 0.001,
+      "loss": 1.543,
+      "step": 33700
+    },
+    {
+      "epoch": 10.92436974789916,
+      "grad_norm": 0.9902774691581726,
+      "learning_rate": 0.001,
+      "loss": 1.5438,
+      "step": 33800
+    },
+    {
+      "epoch": 10.956690368455074,
+      "grad_norm": 1.2797644138336182,
+      "learning_rate": 0.001,
+      "loss": 1.5345,
+      "step": 33900
+    },
+    {
+      "epoch": 10.989010989010989,
+      "grad_norm": 1.1902170181274414,
+      "learning_rate": 0.001,
+      "loss": 1.5337,
+      "step": 34000
+    },
+    {
+      "epoch": 11.021331609566904,
+      "grad_norm": 1.0825999975204468,
+      "learning_rate": 0.001,
+      "loss": 1.4888,
+      "step": 34100
+    },
+    {
+      "epoch": 11.053652230122818,
+      "grad_norm": 1.1632471084594727,
+      "learning_rate": 0.001,
+      "loss": 1.4294,
+      "step": 34200
+    },
+    {
+      "epoch": 11.085972850678733,
+      "grad_norm": 1.212774634361267,
+      "learning_rate": 0.001,
+      "loss": 1.4312,
+      "step": 34300
+    },
+    {
+      "epoch": 11.118293471234647,
+      "grad_norm": 1.018050193786621,
+      "learning_rate": 0.001,
+      "loss": 1.4696,
+      "step": 34400
+    },
+    {
+      "epoch": 11.150614091790562,
+      "grad_norm": 1.0512523651123047,
+      "learning_rate": 0.001,
+      "loss": 1.4331,
+      "step": 34500
+    },
+    {
+      "epoch": 11.182934712346476,
+      "grad_norm": 0.9138154983520508,
+      "learning_rate": 0.001,
+      "loss": 1.4497,
+      "step": 34600
+    },
+    {
+      "epoch": 11.215255332902391,
+      "grad_norm": 1.0956708192825317,
+      "learning_rate": 0.001,
+      "loss": 1.4666,
+      "step": 34700
+    },
+    {
+      "epoch": 11.247575953458306,
+      "grad_norm": 1.0326133966445923,
+      "learning_rate": 0.001,
+      "loss": 1.4538,
+      "step": 34800
+    },
+    {
+      "epoch": 11.279896574014222,
+      "grad_norm": 0.9127147793769836,
+      "learning_rate": 0.001,
+      "loss": 1.4677,
+      "step": 34900
+    },
+    {
+      "epoch": 11.312217194570136,
+      "grad_norm": 0.9023076295852661,
+      "learning_rate": 0.001,
+      "loss": 1.4532,
+      "step": 35000
+    },
+    {
+      "epoch": 11.344537815126051,
+      "grad_norm": 1.0806233882904053,
+      "learning_rate": 0.001,
+      "loss": 1.464,
+      "step": 35100
+    },
+    {
+      "epoch": 11.376858435681966,
+      "grad_norm": 1.0325735807418823,
+      "learning_rate": 0.001,
+      "loss": 1.4889,
+      "step": 35200
+    },
+    {
+      "epoch": 11.40917905623788,
+      "grad_norm": 0.9904654026031494,
+      "learning_rate": 0.001,
+      "loss": 1.4935,
+      "step": 35300
+    },
+    {
+      "epoch": 11.441499676793795,
+      "grad_norm": 1.2254970073699951,
+      "learning_rate": 0.001,
+      "loss": 1.491,
+      "step": 35400
+    },
+    {
+      "epoch": 11.47382029734971,
+      "grad_norm": 1.5757197141647339,
+      "learning_rate": 0.001,
+      "loss": 1.4954,
+      "step": 35500
+    },
+    {
+      "epoch": 11.506140917905624,
+      "grad_norm": 0.9046617746353149,
+      "learning_rate": 0.001,
+      "loss": 1.4765,
+      "step": 35600
+    },
+    {
+      "epoch": 11.538461538461538,
+      "grad_norm": 0.987343430519104,
+      "learning_rate": 0.001,
+      "loss": 1.5013,
+      "step": 35700
+    },
+    {
+      "epoch": 11.570782159017453,
+      "grad_norm": 1.0150471925735474,
+      "learning_rate": 0.001,
+      "loss": 1.4783,
+      "step": 35800
+    },
+    {
+      "epoch": 11.603102779573367,
+      "grad_norm": 1.2784874439239502,
+      "learning_rate": 0.001,
+      "loss": 1.4966,
+      "step": 35900
+    },
+    {
+      "epoch": 11.635423400129282,
+      "grad_norm": 1.0940210819244385,
+      "learning_rate": 0.001,
+      "loss": 1.5001,
+      "step": 36000
+    },
+    {
+      "epoch": 11.667744020685197,
+      "grad_norm": 1.2780746221542358,
+      "learning_rate": 0.001,
+      "loss": 1.4962,
+      "step": 36100
+    },
+    {
+      "epoch": 11.700064641241111,
+      "grad_norm": 0.9342361688613892,
+      "learning_rate": 0.001,
+      "loss": 1.5119,
+      "step": 36200
+    },
+    {
+      "epoch": 11.732385261797026,
+      "grad_norm": 1.034030556678772,
+      "learning_rate": 0.001,
+      "loss": 1.4994,
+      "step": 36300
+    },
+    {
+      "epoch": 11.764705882352942,
+      "grad_norm": 1.0301884412765503,
+      "learning_rate": 0.001,
+      "loss": 1.5109,
+      "step": 36400
+    },
+    {
+      "epoch": 11.797026502908857,
+      "grad_norm": 1.0798345804214478,
+      "learning_rate": 0.001,
+      "loss": 1.4931,
+      "step": 36500
+    },
+    {
+      "epoch": 11.829347123464771,
+      "grad_norm": 0.9824477434158325,
+      "learning_rate": 0.001,
+      "loss": 1.4757,
+      "step": 36600
+    },
+    {
+      "epoch": 11.861667744020686,
+      "grad_norm": 0.970503568649292,
+      "learning_rate": 0.001,
+      "loss": 1.5167,
+      "step": 36700
+    },
+    {
+      "epoch": 11.8939883645766,
+      "grad_norm": 1.0813010931015015,
+      "learning_rate": 0.001,
+      "loss": 1.497,
+      "step": 36800
+    },
+    {
+      "epoch": 11.926308985132515,
+      "grad_norm": 1.0717248916625977,
+      "learning_rate": 0.001,
+      "loss": 1.5072,
+      "step": 36900
+    },
+    {
+      "epoch": 11.95862960568843,
+      "grad_norm": 0.9872753024101257,
+      "learning_rate": 0.001,
+      "loss": 1.5011,
+      "step": 37000
+    },
+    {
+      "epoch": 11.990950226244344,
+      "grad_norm": 0.9820966124534607,
+      "learning_rate": 0.001,
+      "loss": 1.5042,
+      "step": 37100
+    },
+    {
+      "epoch": 12.023270846800258,
+      "grad_norm": 0.968085527420044,
+      "learning_rate": 0.001,
+      "loss": 1.4501,
+      "step": 37200
+    },
+    {
+      "epoch": 12.055591467356173,
+      "grad_norm": 0.980858325958252,
+      "learning_rate": 0.001,
+      "loss": 1.4073,
+      "step": 37300
+    },
+    {
+      "epoch": 12.087912087912088,
+      "grad_norm": 1.4215143918991089,
+      "learning_rate": 0.001,
+      "loss": 1.4116,
+      "step": 37400
+    },
+    {
+      "epoch": 12.120232708468002,
+      "grad_norm": 0.9262951612472534,
+      "learning_rate": 0.001,
+      "loss": 1.4064,
+      "step": 37500
+    },
+    {
+      "epoch": 12.152553329023917,
+      "grad_norm": 1.0522440671920776,
+      "learning_rate": 0.001,
+      "loss": 1.4034,
+      "step": 37600
+    },
+    {
+      "epoch": 12.184873949579831,
+      "grad_norm": 1.0187525749206543,
+      "learning_rate": 0.001,
+      "loss": 1.4318,
+      "step": 37700
+    },
+    {
+      "epoch": 12.217194570135746,
+      "grad_norm": 0.9597002863883972,
+      "learning_rate": 0.001,
+      "loss": 1.4324,
+      "step": 37800
+    },
+    {
+      "epoch": 12.24951519069166,
+      "grad_norm": 1.0441052913665771,
+      "learning_rate": 0.001,
+      "loss": 1.4142,
+      "step": 37900
+    },
+    {
+      "epoch": 12.281835811247577,
+      "grad_norm": 0.8995744585990906,
+      "learning_rate": 0.001,
+      "loss": 1.432,
+      "step": 38000
+    },
+    {
+      "epoch": 12.314156431803491,
+      "grad_norm": 1.096145510673523,
+      "learning_rate": 0.001,
+      "loss": 1.4187,
+      "step": 38100
+    },
+    {
+      "epoch": 12.346477052359406,
+      "grad_norm": 0.9527760148048401,
+      "learning_rate": 0.001,
+      "loss": 1.4671,
+      "step": 38200
+    },
+    {
+      "epoch": 12.37879767291532,
+      "grad_norm": 1.1196210384368896,
+      "learning_rate": 0.001,
+      "loss": 1.4443,
+      "step": 38300
+    },
+    {
+      "epoch": 12.411118293471235,
+      "grad_norm": 0.9554662108421326,
+      "learning_rate": 0.001,
+      "loss": 1.4614,
+      "step": 38400
+    },
+    {
+      "epoch": 12.44343891402715,
+      "grad_norm": 0.9521270394325256,
+      "learning_rate": 0.001,
+      "loss": 1.4679,
+      "step": 38500
+    },
+    {
+      "epoch": 12.475759534583064,
+      "grad_norm": 1.0394660234451294,
+      "learning_rate": 0.001,
+      "loss": 1.4449,
+      "step": 38600
+    },
+    {
+      "epoch": 12.508080155138979,
+      "grad_norm": 1.0146692991256714,
+      "learning_rate": 0.001,
+      "loss": 1.4368,
+      "step": 38700
+    },
+    {
+      "epoch": 12.540400775694893,
+      "grad_norm": 1.3197181224822998,
+      "learning_rate": 0.001,
+      "loss": 1.4654,
+      "step": 38800
+    },
+    {
+      "epoch": 12.572721396250808,
+      "grad_norm": 1.0358250141143799,
+      "learning_rate": 0.001,
+      "loss": 1.4495,
+      "step": 38900
+    },
+    {
+      "epoch": 12.605042016806722,
+      "grad_norm": 1.08975088596344,
+      "learning_rate": 0.001,
+      "loss": 1.4504,
+      "step": 39000
+    },
+    {
+      "epoch": 12.637362637362637,
+      "grad_norm": 0.9866904020309448,
+      "learning_rate": 0.001,
+      "loss": 1.435,
+      "step": 39100
+    },
+    {
+      "epoch": 12.669683257918551,
+      "grad_norm": 1.085909366607666,
+      "learning_rate": 0.001,
+      "loss": 1.4448,
+      "step": 39200
+    },
+    {
+      "epoch": 12.702003878474466,
+      "grad_norm": 0.9458845257759094,
+      "learning_rate": 0.001,
+      "loss": 1.4559,
+      "step": 39300
+    },
+    {
+      "epoch": 12.73432449903038,
+      "grad_norm": 1.0192725658416748,
+      "learning_rate": 0.001,
+      "loss": 1.458,
+      "step": 39400
+    },
+    {
+      "epoch": 12.766645119586297,
+      "grad_norm": 1.2316535711288452,
+      "learning_rate": 0.001,
+      "loss": 1.4697,
+      "step": 39500
+    },
+    {
+      "epoch": 12.798965740142211,
+      "grad_norm": 0.9104325771331787,
+      "learning_rate": 0.001,
+      "loss": 1.5081,
+      "step": 39600
+    },
+    {
+      "epoch": 12.831286360698126,
+      "grad_norm": 1.11668860912323,
+      "learning_rate": 0.001,
+      "loss": 1.4984,
+      "step": 39700
+    },
+    {
+      "epoch": 12.86360698125404,
+      "grad_norm": 0.9999051690101624,
+      "learning_rate": 0.001,
+      "loss": 1.4485,
+      "step": 39800
+    },
+    {
+      "epoch": 12.895927601809955,
+      "grad_norm": 1.0887517929077148,
+      "learning_rate": 0.001,
+      "loss": 1.4528,
+      "step": 39900
+    },
+    {
+      "epoch": 12.92824822236587,
+      "grad_norm": 1.1782604455947876,
+      "learning_rate": 0.001,
+      "loss": 1.4756,
+      "step": 40000
+    },
+    {
+      "epoch": 12.960568842921784,
+      "grad_norm": 0.8981242775917053,
+      "learning_rate": 0.001,
+      "loss": 1.4514,
+      "step": 40100
+    },
+    {
+      "epoch": 12.992889463477699,
+      "grad_norm": 0.875098705291748,
+      "learning_rate": 0.001,
+      "loss": 1.4544,
+      "step": 40200
+    },
+    {
+      "epoch": 13.025210084033613,
+      "grad_norm": 1.2988208532333374,
+      "learning_rate": 0.001,
+      "loss": 1.3966,
+      "step": 40300
+    },
+    {
+      "epoch": 13.057530704589528,
+      "grad_norm": 1.8159008026123047,
+      "learning_rate": 0.001,
+      "loss": 1.3734,
+      "step": 40400
+    },
+    {
+      "epoch": 13.089851325145442,
+      "grad_norm": 0.9084588289260864,
+      "learning_rate": 0.001,
+      "loss": 1.369,
+      "step": 40500
+    },
+    {
+      "epoch": 13.122171945701357,
+      "grad_norm": 0.8925555944442749,
+      "learning_rate": 0.001,
+      "loss": 1.3684,
+      "step": 40600
+    },
+    {
+      "epoch": 13.154492566257272,
+      "grad_norm": 0.8304650187492371,
+      "learning_rate": 0.001,
+      "loss": 1.3887,
+      "step": 40700
+    },
+    {
+      "epoch": 13.186813186813186,
+      "grad_norm": 1.0589717626571655,
+      "learning_rate": 0.001,
+      "loss": 1.3834,
+      "step": 40800
+    },
+    {
+      "epoch": 13.2191338073691,
+      "grad_norm": 0.9229005575180054,
+      "learning_rate": 0.001,
+      "loss": 1.3972,
+      "step": 40900
+    },
+    {
+      "epoch": 13.251454427925015,
+      "grad_norm": 0.9296209216117859,
+      "learning_rate": 0.001,
+      "loss": 1.3909,
+      "step": 41000
+    },
+    {
+      "epoch": 13.283775048480932,
+      "grad_norm": 0.8965741991996765,
+      "learning_rate": 0.001,
+      "loss": 1.3834,
+      "step": 41100
+    },
+    {
+      "epoch": 13.316095669036846,
+      "grad_norm": 0.9377841353416443,
+      "learning_rate": 0.001,
+      "loss": 1.4088,
+      "step": 41200
+    },
+    {
+      "epoch": 13.34841628959276,
+      "grad_norm": 1.0339752435684204,
+      "learning_rate": 0.001,
+      "loss": 1.4098,
+      "step": 41300
+    },
+    {
+      "epoch": 13.380736910148675,
+      "grad_norm": 0.8679028749465942,
+      "learning_rate": 0.001,
+      "loss": 1.3976,
+      "step": 41400
+    },
+    {
+      "epoch": 13.41305753070459,
+      "grad_norm": 0.9596776366233826,
+      "learning_rate": 0.001,
+      "loss": 1.3852,
+      "step": 41500
+    },
+    {
+      "epoch": 13.445378151260504,
+      "grad_norm": 1.042893886566162,
+      "learning_rate": 0.001,
+      "loss": 1.4066,
+      "step": 41600
+    },
+    {
+      "epoch": 13.477698771816419,
+      "grad_norm": 1.0267068147659302,
+      "learning_rate": 0.001,
+      "loss": 1.4032,
+      "step": 41700
+    },
+    {
+      "epoch": 13.510019392372334,
+      "grad_norm": 1.0795466899871826,
+      "learning_rate": 0.001,
+      "loss": 1.4233,
+      "step": 41800
+    },
+    {
+      "epoch": 13.542340012928248,
+      "grad_norm": 0.9310310482978821,
+      "learning_rate": 0.001,
+      "loss": 1.4176,
+      "step": 41900
+    },
+    {
+      "epoch": 13.574660633484163,
+      "grad_norm": 0.7695964574813843,
+      "learning_rate": 0.001,
+      "loss": 1.4068,
+      "step": 42000
+    },
+    {
+      "epoch": 13.606981254040077,
+      "grad_norm": 0.9554638266563416,
+      "learning_rate": 0.001,
+      "loss": 1.4462,
+      "step": 42100
+    },
+    {
+      "epoch": 13.639301874595992,
+      "grad_norm": 0.963115394115448,
+      "learning_rate": 0.001,
+      "loss": 1.4159,
+      "step": 42200
+    },
+    {
+      "epoch": 13.671622495151906,
+      "grad_norm": 0.9717909693717957,
+      "learning_rate": 0.001,
+      "loss": 1.4081,
+      "step": 42300
+    },
+    {
+      "epoch": 13.70394311570782,
+      "grad_norm": 1.1150710582733154,
+      "learning_rate": 0.001,
+      "loss": 1.4377,
+      "step": 42400
+    },
+    {
+      "epoch": 13.736263736263737,
+      "grad_norm": 1.0072553157806396,
+      "learning_rate": 0.001,
+      "loss": 1.4576,
+      "step": 42500
+    },
+    {
+      "epoch": 13.768584356819652,
+      "grad_norm": 0.9450471997261047,
+      "learning_rate": 0.001,
+      "loss": 1.4349,
+      "step": 42600
+    },
+    {
+      "epoch": 13.800904977375566,
+      "grad_norm": 0.8816408514976501,
+      "learning_rate": 0.001,
+      "loss": 1.4237,
+      "step": 42700
+    },
+    {
+      "epoch": 13.83322559793148,
+      "grad_norm": 0.9339333772659302,
+      "learning_rate": 0.001,
+      "loss": 1.4434,
+      "step": 42800
+    },
+    {
+      "epoch": 13.865546218487395,
+      "grad_norm": 1.1679960489273071,
+      "learning_rate": 0.001,
+      "loss": 1.4531,
+      "step": 42900
+    },
+    {
+      "epoch": 13.89786683904331,
+      "grad_norm": 1.0511928796768188,
+      "learning_rate": 0.001,
+      "loss": 1.4101,
+      "step": 43000
+    },
+    {
+      "epoch": 13.930187459599225,
+      "grad_norm": 0.8638760447502136,
+      "learning_rate": 0.001,
+      "loss": 1.4366,
+      "step": 43100
+    },
+    {
+      "epoch": 13.96250808015514,
+      "grad_norm": 1.1316864490509033,
+      "learning_rate": 0.001,
+      "loss": 1.434,
+      "step": 43200
+    },
+    {
+      "epoch": 13.994828700711054,
+      "grad_norm": 1.0850309133529663,
+      "learning_rate": 0.001,
+      "loss": 1.4175,
+      "step": 43300
+    },
+    {
+      "epoch": 14.027149321266968,
+      "grad_norm": 1.1530749797821045,
+      "learning_rate": 0.001,
+      "loss": 1.3574,
+      "step": 43400
+    },
+    {
+      "epoch": 14.059469941822883,
+      "grad_norm": 1.2573904991149902,
+      "learning_rate": 0.001,
+      "loss": 1.3626,
+      "step": 43500
+    },
+    {
+      "epoch": 14.091790562378797,
+      "grad_norm": 1.0096023082733154,
+      "learning_rate": 0.001,
+      "loss": 1.3593,
+      "step": 43600
+    },
+    {
+      "epoch": 14.124111182934712,
+      "grad_norm": 1.0445042848587036,
+      "learning_rate": 0.001,
+      "loss": 1.337,
+      "step": 43700
+    },
+    {
+      "epoch": 14.156431803490626,
+      "grad_norm": 0.976283609867096,
+      "learning_rate": 0.001,
+      "loss": 1.3559,
+      "step": 43800
+    },
+    {
+      "epoch": 14.188752424046541,
+      "grad_norm": 0.9800626635551453,
+      "learning_rate": 0.001,
+      "loss": 1.3801,
+      "step": 43900
+    },
+    {
+      "epoch": 14.221073044602456,
+      "grad_norm": 1.07820725440979,
+      "learning_rate": 0.001,
+      "loss": 1.3517,
+      "step": 44000
+    },
+    {
+      "epoch": 14.25339366515837,
+      "grad_norm": 1.0982788801193237,
+      "learning_rate": 0.001,
+      "loss": 1.3222,
+      "step": 44100
+    },
+    {
+      "epoch": 14.285714285714286,
+      "grad_norm": 1.013915777206421,
+      "learning_rate": 0.001,
+      "loss": 1.3454,
+      "step": 44200
+    },
+    {
+      "epoch": 14.318034906270201,
+      "grad_norm": 0.9462135434150696,
+      "learning_rate": 0.001,
+      "loss": 1.3472,
+      "step": 44300
+    },
+    {
+      "epoch": 14.350355526826116,
+      "grad_norm": 1.1129679679870605,
+      "learning_rate": 0.001,
+      "loss": 1.3625,
+      "step": 44400
+    },
+    {
+      "epoch": 14.38267614738203,
+      "grad_norm": 1.1635687351226807,
+      "learning_rate": 0.001,
+      "loss": 1.3752,
+      "step": 44500
+    },
+    {
+      "epoch": 14.414996767937945,
+      "grad_norm": 0.9367783665657043,
+      "learning_rate": 0.001,
+      "loss": 1.398,
+      "step": 44600
+    },
+    {
+      "epoch": 14.44731738849386,
+      "grad_norm": 4.812443256378174,
+      "learning_rate": 0.001,
+      "loss": 1.3708,
+      "step": 44700
+    },
+    {
+      "epoch": 14.479638009049774,
+      "grad_norm": 1.1130398511886597,
+      "learning_rate": 0.001,
+      "loss": 1.3928,
+      "step": 44800
+    },
+    {
+      "epoch": 14.511958629605688,
+      "grad_norm": 1.0559415817260742,
+      "learning_rate": 0.001,
+      "loss": 1.367,
+      "step": 44900
+    },
+    {
+      "epoch": 14.544279250161603,
+      "grad_norm": 1.052778959274292,
+      "learning_rate": 0.001,
+      "loss": 1.3692,
+      "step": 45000
+    },
+    {
+      "epoch": 14.576599870717518,
+      "grad_norm": 1.0826404094696045,
+      "learning_rate": 0.001,
+      "loss": 1.3857,
+      "step": 45100
+    },
+    {
+      "epoch": 14.608920491273432,
+      "grad_norm": 0.9068247675895691,
+      "learning_rate": 0.001,
+      "loss": 1.3805,
+      "step": 45200
+    },
+    {
+      "epoch": 14.641241111829347,
+      "grad_norm": 1.2357306480407715,
+      "learning_rate": 0.001,
+      "loss": 1.4152,
+      "step": 45300
+    },
+    {
+      "epoch": 14.673561732385261,
+      "grad_norm": 0.9457151293754578,
+      "learning_rate": 0.001,
+      "loss": 1.4111,
+      "step": 45400
+    },
+    {
+      "epoch": 14.705882352941176,
+      "grad_norm": 0.8796570301055908,
+      "learning_rate": 0.001,
+      "loss": 1.3951,
+      "step": 45500
+    },
+    {
+      "epoch": 14.738202973497092,
+      "grad_norm": 0.9057885408401489,
+      "learning_rate": 0.001,
+      "loss": 1.4123,
+      "step": 45600
+    },
+    {
+      "epoch": 14.770523594053007,
+      "grad_norm": 1.1413110494613647,
+      "learning_rate": 0.001,
+      "loss": 1.3942,
+      "step": 45700
+    },
+    {
+      "epoch": 14.802844214608921,
+      "grad_norm": 0.8065590262413025,
+      "learning_rate": 0.001,
+      "loss": 1.4036,
+      "step": 45800
+    },
+    {
+      "epoch": 14.835164835164836,
+      "grad_norm": 0.9088504314422607,
+      "learning_rate": 0.001,
+      "loss": 1.3785,
+      "step": 45900
+    },
+    {
+      "epoch": 14.86748545572075,
+      "grad_norm": 1.1512057781219482,
+      "learning_rate": 0.001,
+      "loss": 1.411,
+      "step": 46000
+    },
+    {
+      "epoch": 14.899806076276665,
+      "grad_norm": 0.9676141142845154,
+      "learning_rate": 0.001,
+      "loss": 1.3749,
+      "step": 46100
+    },
+    {
+      "epoch": 14.93212669683258,
+      "grad_norm": 1.2688740491867065,
+      "learning_rate": 0.001,
+      "loss": 1.4197,
+      "step": 46200
+    },
+    {
+      "epoch": 14.964447317388494,
+      "grad_norm": 0.9541943669319153,
+      "learning_rate": 0.001,
+      "loss": 1.4337,
+      "step": 46300
+    },
+    {
+      "epoch": 14.996767937944409,
+      "grad_norm": 1.5543314218521118,
+      "learning_rate": 0.001,
+      "loss": 1.3808,
+      "step": 46400
+    },
+    {
+      "epoch": 15.029088558500323,
+      "grad_norm": 1.0888712406158447,
+      "learning_rate": 0.001,
+      "loss": 1.3064,
+      "step": 46500
+    },
+    {
+      "epoch": 15.061409179056238,
+      "grad_norm": 1.151442527770996,
+      "learning_rate": 0.001,
+      "loss": 1.3064,
+      "step": 46600
+    },
+    {
+      "epoch": 15.093729799612152,
+      "grad_norm": 1.339379072189331,
+      "learning_rate": 0.001,
+      "loss": 1.3393,
+      "step": 46700
+    },
+    {
+      "epoch": 15.126050420168067,
+      "grad_norm": 1.1313862800598145,
+      "learning_rate": 0.001,
+      "loss": 1.313,
+      "step": 46800
+    },
+    {
+      "epoch": 15.158371040723981,
+      "grad_norm": 0.9059498906135559,
+      "learning_rate": 0.001,
+      "loss": 1.3322,
+      "step": 46900
+    },
+    {
+      "epoch": 15.190691661279896,
+      "grad_norm": 0.9920981526374817,
+      "learning_rate": 0.001,
+      "loss": 1.325,
+      "step": 47000
+    },
+    {
+      "epoch": 15.22301228183581,
+      "grad_norm": 0.7398461103439331,
+      "learning_rate": 0.001,
+      "loss": 1.3382,
+      "step": 47100
+    },
+    {
+      "epoch": 15.255332902391725,
+      "grad_norm": 1.0884451866149902,
+      "learning_rate": 0.001,
+      "loss": 1.3054,
+      "step": 47200
+    },
+    {
+      "epoch": 15.287653522947641,
+      "grad_norm": 1.2823668718338013,
+      "learning_rate": 0.001,
+      "loss": 1.3264,
+      "step": 47300
+    },
+    {
+      "epoch": 15.319974143503556,
+      "grad_norm": 1.3318175077438354,
+      "learning_rate": 0.001,
+      "loss": 1.3511,
+      "step": 47400
+    },
+    {
+      "epoch": 15.35229476405947,
+      "grad_norm": 1.1119569540023804,
+      "learning_rate": 0.001,
+      "loss": 1.3449,
+      "step": 47500
+    },
+    {
+      "epoch": 15.384615384615385,
+      "grad_norm": 1.1583483219146729,
+      "learning_rate": 0.001,
+      "loss": 1.3145,
+      "step": 47600
+    },
+    {
+      "epoch": 15.4169360051713,
+      "grad_norm": 1.323159098625183,
+      "learning_rate": 0.001,
+      "loss": 1.3609,
+      "step": 47700
+    },
+    {
+      "epoch": 15.449256625727214,
+      "grad_norm": 1.1489546298980713,
+      "learning_rate": 0.001,
+      "loss": 1.3374,
+      "step": 47800
+    },
+    {
+      "epoch": 15.481577246283129,
+      "grad_norm": 1.1882226467132568,
+      "learning_rate": 0.001,
+      "loss": 1.353,
+      "step": 47900
+    },
+    {
+      "epoch": 15.513897866839043,
+      "grad_norm": 1.6996192932128906,
+      "learning_rate": 0.001,
+      "loss": 1.3764,
+      "step": 48000
+    },
+    {
+      "epoch": 15.546218487394958,
+      "grad_norm": 1.079559564590454,
+      "learning_rate": 0.001,
+      "loss": 1.3654,
+      "step": 48100
+    },
+    {
+      "epoch": 15.578539107950872,
+      "grad_norm": 0.9709523916244507,
+      "learning_rate": 0.001,
+      "loss": 1.361,
+      "step": 48200
+    },
+    {
+      "epoch": 15.610859728506787,
+      "grad_norm": 1.1908799409866333,
+      "learning_rate": 0.001,
+      "loss": 1.3668,
+      "step": 48300
+    },
+    {
+      "epoch": 15.643180349062701,
+      "grad_norm": 0.8918905854225159,
+      "learning_rate": 0.001,
+      "loss": 1.3524,
+      "step": 48400
+    },
+    {
+      "epoch": 15.675500969618616,
+      "grad_norm": 0.880649983882904,
+      "learning_rate": 0.001,
+      "loss": 1.3517,
+      "step": 48500
+    },
+    {
+      "epoch": 15.70782159017453,
+      "grad_norm": 1.1815990209579468,
+      "learning_rate": 0.001,
+      "loss": 1.349,
+      "step": 48600
+    },
+    {
+      "epoch": 15.740142210730447,
+      "grad_norm": 1.071019172668457,
+      "learning_rate": 0.001,
+      "loss": 1.3596,
+      "step": 48700
+    },
+    {
+      "epoch": 15.772462831286362,
+      "grad_norm": 1.1914271116256714,
+      "learning_rate": 0.001,
+      "loss": 1.3663,
+      "step": 48800
+    },
+    {
+      "epoch": 15.804783451842276,
+      "grad_norm": 1.4547832012176514,
+      "learning_rate": 0.001,
+      "loss": 1.3611,
+      "step": 48900
+    },
+    {
+      "epoch": 15.83710407239819,
+      "grad_norm": 1.1460895538330078,
+      "learning_rate": 0.001,
+      "loss": 1.3605,
+      "step": 49000
+    },
+    {
+      "epoch": 15.869424692954105,
+      "grad_norm": 1.155902624130249,
+      "learning_rate": 0.001,
+      "loss": 1.3833,
+      "step": 49100
+    },
+    {
+      "epoch": 15.90174531351002,
+      "grad_norm": 1.1332170963287354,
+      "learning_rate": 0.001,
+      "loss": 1.366,
+      "step": 49200
+    },
+    {
+      "epoch": 15.934065934065934,
+      "grad_norm": 1.009746789932251,
+      "learning_rate": 0.001,
+      "loss": 1.3741,
+      "step": 49300
+    },
+    {
+      "epoch": 15.966386554621849,
+      "grad_norm": 1.1160213947296143,
+      "learning_rate": 0.001,
+      "loss": 1.3727,
+      "step": 49400
+    },
+    {
+      "epoch": 15.998707175177763,
+      "grad_norm": 0.9431664347648621,
+      "learning_rate": 0.001,
+      "loss": 1.3603,
+      "step": 49500
+    },
+    {
+      "epoch": 16.031027795733678,
+      "grad_norm": 1.071413516998291,
+      "learning_rate": 0.001,
+      "loss": 1.2621,
+      "step": 49600
+    },
+    {
+      "epoch": 16.063348416289593,
+      "grad_norm": 1.035800576210022,
+      "learning_rate": 0.001,
+      "loss": 1.2914,
+      "step": 49700
+    },
+    {
+      "epoch": 16.095669036845507,
+      "grad_norm": 1.001190423965454,
+      "learning_rate": 0.001,
+      "loss": 1.2734,
+      "step": 49800
+    },
+    {
+      "epoch": 16.12798965740142,
+      "grad_norm": 0.9272027611732483,
+      "learning_rate": 0.001,
+      "loss": 1.2658,
+      "step": 49900
+    },
+    {
+      "epoch": 16.160310277957336,
+      "grad_norm": 0.8124894499778748,
+      "learning_rate": 0.001,
+      "loss": 1.2945,
+      "step": 50000
+    },
+    {
+      "epoch": 16.19263089851325,
+      "grad_norm": 1.0019245147705078,
+      "learning_rate": 0.001,
+      "loss": 1.3106,
+      "step": 50100
+    },
+    {
+      "epoch": 16.224951519069165,
+      "grad_norm": 1.071997880935669,
+      "learning_rate": 0.001,
+      "loss": 1.3253,
+      "step": 50200
+    },
+    {
+      "epoch": 16.25727213962508,
+      "grad_norm": 0.8945184946060181,
+      "learning_rate": 0.001,
+      "loss": 1.2997,
+      "step": 50300
+    },
+    {
+      "epoch": 16.289592760180994,
+      "grad_norm": 0.8286392092704773,
+      "learning_rate": 0.001,
+      "loss": 1.3153,
+      "step": 50400
+    },
+    {
+      "epoch": 16.32191338073691,
+      "grad_norm": 1.137093424797058,
+      "learning_rate": 0.001,
+      "loss": 1.3035,
+      "step": 50500
+    },
+    {
+      "epoch": 16.354234001292824,
+      "grad_norm": 1.379814863204956,
+      "learning_rate": 0.001,
+      "loss": 1.2933,
+      "step": 50600
+    },
+    {
+      "epoch": 16.386554621848738,
+      "grad_norm": 0.8075309991836548,
+      "learning_rate": 0.001,
+      "loss": 1.3175,
+      "step": 50700
+    },
+    {
+      "epoch": 16.418875242404653,
+      "grad_norm": 1.3296177387237549,
+      "learning_rate": 0.001,
+      "loss": 1.3143,
+      "step": 50800
+    },
+    {
+      "epoch": 16.451195862960567,
+      "grad_norm": 1.5358154773712158,
+      "learning_rate": 0.001,
+      "loss": 1.3137,
+      "step": 50900
+    },
+    {
+      "epoch": 16.483516483516482,
+      "grad_norm": 1.2911955118179321,
+      "learning_rate": 0.001,
+      "loss": 1.3108,
+      "step": 51000
+    },
+    {
+      "epoch": 16.5158371040724,
+      "grad_norm": 0.6920326352119446,
+      "learning_rate": 0.001,
+      "loss": 1.3177,
+      "step": 51100
+    },
+    {
+      "epoch": 16.548157724628314,
+      "grad_norm": 1.075358510017395,
+      "learning_rate": 0.001,
+      "loss": 1.3267,
+      "step": 51200
+    },
+    {
+      "epoch": 16.58047834518423,
+      "grad_norm": 0.9144567251205444,
+      "learning_rate": 0.001,
+      "loss": 1.3318,
+      "step": 51300
+    },
+    {
+      "epoch": 16.612798965740144,
+      "grad_norm": 1.1483515501022339,
+      "learning_rate": 0.001,
+      "loss": 1.3312,
+      "step": 51400
+    },
+    {
+      "epoch": 16.645119586296058,
+      "grad_norm": 1.050889253616333,
+      "learning_rate": 0.001,
+      "loss": 1.3428,
+      "step": 51500
+    },
+    {
+      "epoch": 16.677440206851973,
+      "grad_norm": 1.0265402793884277,
+      "learning_rate": 0.001,
+      "loss": 1.3533,
+      "step": 51600
+    },
+    {
+      "epoch": 16.709760827407887,
+      "grad_norm": 0.9257956743240356,
+      "learning_rate": 0.001,
+      "loss": 1.3497,
+      "step": 51700
+    },
+    {
+      "epoch": 16.742081447963802,
+      "grad_norm": 1.9063913822174072,
+      "learning_rate": 0.001,
+      "loss": 1.3191,
+      "step": 51800
+    },
+    {
+      "epoch": 16.774402068519716,
+      "grad_norm": 1.0440740585327148,
+      "learning_rate": 0.001,
+      "loss": 1.3378,
+      "step": 51900
+    },
+    {
+      "epoch": 16.80672268907563,
+      "grad_norm": 1.0758187770843506,
+      "learning_rate": 0.001,
+      "loss": 1.3278,
+      "step": 52000
+    },
+    {
+      "epoch": 16.839043309631545,
+      "grad_norm": 1.1654984951019287,
+      "learning_rate": 0.001,
+      "loss": 1.3627,
+      "step": 52100
+    },
+    {
+      "epoch": 16.87136393018746,
+      "grad_norm": 0.9538053870201111,
+      "learning_rate": 0.001,
+      "loss": 1.3415,
+      "step": 52200
+    },
+    {
+      "epoch": 16.903684550743375,
+      "grad_norm": 1.0862343311309814,
+      "learning_rate": 0.001,
+      "loss": 1.351,
+      "step": 52300
+    },
+    {
+      "epoch": 16.93600517129929,
+      "grad_norm": 1.0913732051849365,
+      "learning_rate": 0.001,
+      "loss": 1.3614,
+      "step": 52400
+    },
+    {
+      "epoch": 16.968325791855204,
+      "grad_norm": 1.0344431400299072,
+      "learning_rate": 0.001,
+      "loss": 1.3385,
+      "step": 52500
+    },
+    {
+      "epoch": 17.00064641241112,
+      "grad_norm": 1.0967899560928345,
+      "learning_rate": 0.001,
+      "loss": 1.3154,
+      "step": 52600
+    },
+    {
+      "epoch": 17.032967032967033,
+      "grad_norm": 1.7565858364105225,
+      "learning_rate": 0.001,
+      "loss": 1.2238,
+      "step": 52700
+    },
+    {
+      "epoch": 17.065287653522947,
+      "grad_norm": 1.8335599899291992,
+      "learning_rate": 0.001,
+      "loss": 1.2362,
+      "step": 52800
+    },
+    {
+      "epoch": 17.097608274078862,
+      "grad_norm": 1.951370358467102,
+      "learning_rate": 0.001,
+      "loss": 1.2447,
+      "step": 52900
+    },
+    {
+      "epoch": 17.129928894634777,
+      "grad_norm": 1.7428096532821655,
+      "learning_rate": 0.001,
+      "loss": 1.273,
+      "step": 53000
+    },
+    {
+      "epoch": 17.16224951519069,
+      "grad_norm": 1.5402082204818726,
+      "learning_rate": 0.001,
+      "loss": 1.2413,
+      "step": 53100
+    },
+    {
+      "epoch": 17.194570135746606,
+      "grad_norm": 1.3417208194732666,
+      "learning_rate": 0.001,
+      "loss": 1.2456,
+      "step": 53200
+    },
+    {
+      "epoch": 17.22689075630252,
+      "grad_norm": 1.6393389701843262,
+      "learning_rate": 0.001,
+      "loss": 1.2635,
+      "step": 53300
+    },
+    {
+      "epoch": 17.259211376858435,
+      "grad_norm": 1.5430525541305542,
+      "learning_rate": 0.001,
+      "loss": 1.2979,
+      "step": 53400
+    },
+    {
+      "epoch": 17.29153199741435,
+      "grad_norm": 1.640655517578125,
+      "learning_rate": 0.001,
+      "loss": 1.2704,
+      "step": 53500
+    },
+    {
+      "epoch": 17.323852617970264,
+      "grad_norm": 1.7041141986846924,
+      "learning_rate": 0.001,
+      "loss": 1.3015,
+      "step": 53600
+    },
+    {
+      "epoch": 17.35617323852618,
+      "grad_norm": 1.5950192213058472,
+      "learning_rate": 0.001,
+      "loss": 1.2574,
+      "step": 53700
+    },
+    {
+      "epoch": 17.388493859082093,
+      "grad_norm": 1.8514941930770874,
+      "learning_rate": 0.001,
+      "loss": 1.2826,
+      "step": 53800
+    },
+    {
+      "epoch": 17.420814479638008,
+      "grad_norm": 1.5289716720581055,
+      "learning_rate": 0.001,
+      "loss": 1.2847,
+      "step": 53900
+    },
+    {
+      "epoch": 17.453135100193922,
+      "grad_norm": 1.5072901248931885,
+      "learning_rate": 0.001,
+      "loss": 1.3026,
+      "step": 54000
+    },
+    {
+      "epoch": 17.485455720749837,
+      "grad_norm": 1.9386134147644043,
+      "learning_rate": 0.001,
+      "loss": 1.3189,
+      "step": 54100
+    },
+    {
+      "epoch": 17.517776341305755,
+      "grad_norm": 1.3212685585021973,
+      "learning_rate": 0.001,
+      "loss": 1.2961,
+      "step": 54200
+    },
+    {
+      "epoch": 17.55009696186167,
+      "grad_norm": 1.8733723163604736,
+      "learning_rate": 0.001,
+      "loss": 1.2841,
+      "step": 54300
+    },
+    {
+      "epoch": 17.582417582417584,
+      "grad_norm": 1.5845545530319214,
+      "learning_rate": 0.001,
+      "loss": 1.2776,
+      "step": 54400
+    },
+    {
+      "epoch": 17.6147382029735,
+      "grad_norm": 1.2121365070343018,
+      "learning_rate": 0.001,
+      "loss": 1.3145,
+      "step": 54500
+    },
+    {
+      "epoch": 17.647058823529413,
+      "grad_norm": 1.1913347244262695,
+      "learning_rate": 0.001,
+      "loss": 1.3027,
+      "step": 54600
+    },
+    {
+      "epoch": 17.679379444085328,
+      "grad_norm": 1.4041755199432373,
+      "learning_rate": 0.001,
+      "loss": 1.3157,
+      "step": 54700
+    },
+    {
+      "epoch": 17.711700064641242,
+      "grad_norm": 1.5565708875656128,
+      "learning_rate": 0.001,
+      "loss": 1.3094,
+      "step": 54800
+    },
+    {
+      "epoch": 17.744020685197157,
+      "grad_norm": 1.6354928016662598,
+      "learning_rate": 0.001,
+      "loss": 1.3378,
+      "step": 54900
+    },
+    {
+      "epoch": 17.77634130575307,
+      "grad_norm": 1.461625099182129,
+      "learning_rate": 0.001,
+      "loss": 1.3012,
+      "step": 55000
+    },
+    {
+      "epoch": 17.808661926308986,
+      "grad_norm": 1.9901479482650757,
+      "learning_rate": 0.001,
+      "loss": 1.3125,
+      "step": 55100
+    },
+    {
+      "epoch": 17.8409825468649,
+      "grad_norm": 1.8469276428222656,
+      "learning_rate": 0.001,
+      "loss": 1.3086,
+      "step": 55200
+    },
+    {
+      "epoch": 17.873303167420815,
+      "grad_norm": 1.725543737411499,
+      "learning_rate": 0.001,
+      "loss": 1.351,
+      "step": 55300
+    },
+    {
+      "epoch": 17.90562378797673,
+      "grad_norm": 1.799232006072998,
+      "learning_rate": 0.001,
+      "loss": 1.3276,
+      "step": 55400
+    },
+    {
+      "epoch": 17.937944408532644,
+      "grad_norm": 1.447696566581726,
+      "learning_rate": 0.001,
+      "loss": 1.3347,
+      "step": 55500
+    },
+    {
+      "epoch": 17.97026502908856,
+      "grad_norm": 1.706591010093689,
+      "learning_rate": 0.001,
+      "loss": 1.3436,
+      "step": 55600
+    },
+    {
+      "epoch": 18.002585649644473,
+      "grad_norm": 1.0814664363861084,
+      "learning_rate": 0.001,
+      "loss": 1.3459,
+      "step": 55700
+    },
+    {
+      "epoch": 18.034906270200388,
+      "grad_norm": 1.1114346981048584,
+      "learning_rate": 0.001,
+      "loss": 1.1934,
+      "step": 55800
+    },
+    {
+      "epoch": 18.067226890756302,
+      "grad_norm": 1.211247205734253,
+      "learning_rate": 0.001,
+      "loss": 1.2291,
+      "step": 55900
+    },
+    {
+      "epoch": 18.099547511312217,
+      "grad_norm": 1.1112585067749023,
+      "learning_rate": 0.001,
+      "loss": 1.2196,
+      "step": 56000
+    },
+    {
+      "epoch": 18.13186813186813,
+      "grad_norm": 1.2427830696105957,
+      "learning_rate": 0.001,
+      "loss": 1.2401,
+      "step": 56100
+    },
+    {
+      "epoch": 18.164188752424046,
+      "grad_norm": 1.1477326154708862,
+      "learning_rate": 0.001,
+      "loss": 1.2263,
+      "step": 56200
+    },
+    {
+      "epoch": 18.19650937297996,
+      "grad_norm": 1.2850887775421143,
+      "learning_rate": 0.001,
+      "loss": 1.2521,
+      "step": 56300
+    },
+    {
+      "epoch": 18.228829993535875,
+      "grad_norm": 1.1225168704986572,
+      "learning_rate": 0.001,
+      "loss": 1.2548,
+      "step": 56400
+    },
+    {
+      "epoch": 18.26115061409179,
+      "grad_norm": 1.1634271144866943,
+      "learning_rate": 0.001,
+      "loss": 1.2274,
+      "step": 56500
+    },
+    {
+      "epoch": 18.293471234647704,
+      "grad_norm": 1.162834644317627,
+      "learning_rate": 0.001,
+      "loss": 1.2562,
+      "step": 56600
+    },
+    {
+      "epoch": 18.32579185520362,
+      "grad_norm": 1.3091973066329956,
+      "learning_rate": 0.001,
+      "loss": 1.2844,
+      "step": 56700
+    },
+    {
+      "epoch": 18.358112475759533,
+      "grad_norm": 1.1031546592712402,
+      "learning_rate": 0.001,
+      "loss": 1.2595,
+      "step": 56800
+    },
+    {
+      "epoch": 18.390433096315448,
+      "grad_norm": 1.2217282056808472,
+      "learning_rate": 0.001,
+      "loss": 1.2533,
+      "step": 56900
+    },
+    {
+      "epoch": 18.422753716871362,
+      "grad_norm": 1.2840886116027832,
+      "learning_rate": 0.001,
+      "loss": 1.2639,
+      "step": 57000
+    },
+    {
+      "epoch": 18.455074337427277,
+      "grad_norm": 0.935218870639801,
+      "learning_rate": 0.001,
+      "loss": 1.2536,
+      "step": 57100
+    },
+    {
+      "epoch": 18.48739495798319,
+      "grad_norm": 1.423843502998352,
+      "learning_rate": 0.001,
+      "loss": 1.2741,
+      "step": 57200
+    },
+    {
+      "epoch": 18.51971557853911,
+      "grad_norm": 1.232347011566162,
+      "learning_rate": 0.001,
+      "loss": 1.2671,
+      "step": 57300
+    },
+    {
+      "epoch": 18.552036199095024,
+      "grad_norm": 1.2617510557174683,
+      "learning_rate": 0.001,
+      "loss": 1.2726,
+      "step": 57400
+    },
+    {
+      "epoch": 18.58435681965094,
+      "grad_norm": 1.3618507385253906,
+      "learning_rate": 0.001,
+      "loss": 1.2602,
+      "step": 57500
+    },
+    {
+      "epoch": 18.616677440206853,
+      "grad_norm": 1.2484835386276245,
+      "learning_rate": 0.001,
+      "loss": 1.2583,
+      "step": 57600
+    },
+    {
+      "epoch": 18.648998060762768,
+      "grad_norm": 1.2037937641143799,
+      "learning_rate": 0.001,
+      "loss": 1.2757,
+      "step": 57700
+    },
+    {
+      "epoch": 18.681318681318682,
+      "grad_norm": 1.1551131010055542,
+      "learning_rate": 0.001,
+      "loss": 1.2876,
+      "step": 57800
+    },
+    {
+      "epoch": 18.713639301874597,
+      "grad_norm": 1.0815056562423706,
+      "learning_rate": 0.001,
+      "loss": 1.3025,
+      "step": 57900
+    },
+    {
+      "epoch": 18.74595992243051,
+      "grad_norm": 1.0868639945983887,
+      "learning_rate": 0.001,
+      "loss": 1.3064,
+      "step": 58000
+    },
+    {
+      "epoch": 18.778280542986426,
+      "grad_norm": 1.2222222089767456,
+      "learning_rate": 0.001,
+      "loss": 1.2787,
+      "step": 58100
+    },
+    {
+      "epoch": 18.81060116354234,
+      "grad_norm": 1.2130047082901,
+      "learning_rate": 0.001,
+      "loss": 1.2716,
+      "step": 58200
+    },
+    {
+      "epoch": 18.842921784098255,
+      "grad_norm": 1.2509404420852661,
+      "learning_rate": 0.001,
+      "loss": 1.2875,
+      "step": 58300
+    },
+    {
+      "epoch": 18.87524240465417,
+      "grad_norm": 1.3974509239196777,
+      "learning_rate": 0.001,
+      "loss": 1.3006,
+      "step": 58400
+    },
+    {
+      "epoch": 18.907563025210084,
+      "grad_norm": 1.0603617429733276,
+      "learning_rate": 0.001,
+      "loss": 1.3321,
+      "step": 58500
+    },
+    {
+      "epoch": 18.939883645766,
+      "grad_norm": 1.0816590785980225,
+      "learning_rate": 0.001,
+      "loss": 1.2933,
+      "step": 58600
+    },
+    {
+      "epoch": 18.972204266321913,
+      "grad_norm": 1.3309355974197388,
+      "learning_rate": 0.001,
+      "loss": 1.2893,
+      "step": 58700
+    },
+    {
+      "epoch": 19.004524886877828,
+      "grad_norm": 1.2768296003341675,
+      "learning_rate": 0.001,
+      "loss": 1.3172,
+      "step": 58800
+    },
+    {
+      "epoch": 19.036845507433743,
+      "grad_norm": 1.4616053104400635,
+      "learning_rate": 0.001,
+      "loss": 1.1704,
+      "step": 58900
+    },
+    {
+      "epoch": 19.069166127989657,
+      "grad_norm": 1.417331337928772,
+      "learning_rate": 0.001,
+      "loss": 1.1887,
+      "step": 59000
+    },
+    {
+      "epoch": 19.10148674854557,
+      "grad_norm": 1.2152198553085327,
+      "learning_rate": 0.001,
+      "loss": 1.1897,
+      "step": 59100
+    },
+    {
+      "epoch": 19.133807369101486,
+      "grad_norm": 1.1913012266159058,
+      "learning_rate": 0.001,
+      "loss": 1.2236,
+      "step": 59200
+    },
+    {
+      "epoch": 19.1661279896574,
+      "grad_norm": 0.9392966628074646,
+      "learning_rate": 0.001,
+      "loss": 1.226,
+      "step": 59300
+    },
+    {
+      "epoch": 19.198448610213315,
+      "grad_norm": 1.0119884014129639,
+      "learning_rate": 0.001,
+      "loss": 1.2228,
+      "step": 59400
+    },
+    {
+      "epoch": 19.23076923076923,
+      "grad_norm": 1.276890754699707,
+      "learning_rate": 0.001,
+      "loss": 1.2287,
+      "step": 59500
+    },
+    {
+      "epoch": 19.263089851325145,
+      "grad_norm": 1.1096450090408325,
+      "learning_rate": 0.001,
+      "loss": 1.2371,
+      "step": 59600
+    },
+    {
+      "epoch": 19.29541047188106,
+      "grad_norm": 1.1702587604522705,
+      "learning_rate": 0.001,
+      "loss": 1.2268,
+      "step": 59700
+    },
+    {
+      "epoch": 19.327731092436974,
+      "grad_norm": 1.0416254997253418,
+      "learning_rate": 0.001,
+      "loss": 1.2284,
+      "step": 59800
+    },
+    {
+      "epoch": 19.360051712992888,
+      "grad_norm": 1.093361735343933,
+      "learning_rate": 0.001,
+      "loss": 1.2461,
+      "step": 59900
+    },
+    {
+      "epoch": 19.392372333548803,
+      "grad_norm": 1.0348377227783203,
+      "learning_rate": 0.001,
+      "loss": 1.2525,
+      "step": 60000
+    },
+    {
+      "epoch": 19.424692954104717,
+      "grad_norm": 1.9038110971450806,
+      "learning_rate": 0.001,
+      "loss": 1.2481,
+      "step": 60100
+    },
+    {
+      "epoch": 19.457013574660632,
+      "grad_norm": 1.2229783535003662,
+      "learning_rate": 0.001,
+      "loss": 1.2186,
+      "step": 60200
+    },
+    {
+      "epoch": 19.489334195216546,
+      "grad_norm": 1.1477720737457275,
+      "learning_rate": 0.001,
+      "loss": 1.2482,
+      "step": 60300
+    },
+    {
+      "epoch": 19.521654815772465,
+      "grad_norm": 1.255200743675232,
+      "learning_rate": 0.001,
+      "loss": 1.2552,
+      "step": 60400
+    },
+    {
+      "epoch": 19.55397543632838,
+      "grad_norm": 1.0543900728225708,
+      "learning_rate": 0.001,
+      "loss": 1.2373,
+      "step": 60500
+    },
+    {
+      "epoch": 19.586296056884294,
+      "grad_norm": 1.2552344799041748,
+      "learning_rate": 0.001,
+      "loss": 1.2511,
+      "step": 60600
+    },
+    {
+      "epoch": 19.618616677440208,
+      "grad_norm": 1.2354882955551147,
+      "learning_rate": 0.001,
+      "loss": 1.2509,
+      "step": 60700
+    },
+    {
+      "epoch": 19.650937297996123,
+      "grad_norm": 1.1915310621261597,
+      "learning_rate": 0.001,
+      "loss": 1.2541,
+      "step": 60800
+    },
+    {
+      "epoch": 19.683257918552037,
+      "grad_norm": 1.3169018030166626,
+      "learning_rate": 0.001,
+      "loss": 1.246,
+      "step": 60900
+    },
+    {
+      "epoch": 19.715578539107952,
+      "grad_norm": 1.34446120262146,
+      "learning_rate": 0.001,
+      "loss": 1.2651,
+      "step": 61000
+    },
+    {
+      "epoch": 19.747899159663866,
+      "grad_norm": 1.141524076461792,
+      "learning_rate": 0.001,
+      "loss": 1.2607,
+      "step": 61100
+    },
+    {
+      "epoch": 19.78021978021978,
+      "grad_norm": 1.2178977727890015,
+      "learning_rate": 0.001,
+      "loss": 1.2887,
+      "step": 61200
+    },
+    {
+      "epoch": 19.812540400775696,
+      "grad_norm": 0.9515364766120911,
+      "learning_rate": 0.001,
+      "loss": 1.2445,
+      "step": 61300
+    },
+    {
+      "epoch": 19.84486102133161,
+      "grad_norm": 1.2744394540786743,
+      "learning_rate": 0.001,
+      "loss": 1.2612,
+      "step": 61400
+    },
+    {
+      "epoch": 19.877181641887525,
+      "grad_norm": 1.0756638050079346,
+      "learning_rate": 0.001,
+      "loss": 1.2727,
+      "step": 61500
+    },
+    {
+      "epoch": 19.90950226244344,
+      "grad_norm": 1.0071052312850952,
+      "learning_rate": 0.001,
+      "loss": 1.2548,
+      "step": 61600
+    },
+    {
+      "epoch": 19.941822882999354,
+      "grad_norm": 1.349400520324707,
+      "learning_rate": 0.001,
+      "loss": 1.2679,
+      "step": 61700
+    },
+    {
+      "epoch": 19.97414350355527,
+      "grad_norm": 3.5351197719573975,
+      "learning_rate": 0.001,
+      "loss": 1.2702,
+      "step": 61800
+    },
+    {
+      "epoch": 20.006464124111183,
+      "grad_norm": 1.235960602760315,
+      "learning_rate": 0.001,
+      "loss": 1.2695,
+      "step": 61900
+    },
+    {
+      "epoch": 20.038784744667097,
+      "grad_norm": 1.0976554155349731,
+      "learning_rate": 0.001,
+      "loss": 1.1755,
+      "step": 62000
+    },
+    {
+      "epoch": 20.071105365223012,
+      "grad_norm": 1.267794132232666,
+      "learning_rate": 0.001,
+      "loss": 1.1877,
+      "step": 62100
+    },
+    {
+      "epoch": 20.103425985778927,
+      "grad_norm": 1.089699625968933,
+      "learning_rate": 0.001,
+      "loss": 1.1617,
+      "step": 62200
+    },
+    {
+      "epoch": 20.13574660633484,
+      "grad_norm": 1.0829448699951172,
+      "learning_rate": 0.001,
+      "loss": 1.1832,
+      "step": 62300
+    },
+    {
+      "epoch": 20.168067226890756,
+      "grad_norm": 1.2309054136276245,
+      "learning_rate": 0.001,
+      "loss": 1.2009,
+      "step": 62400
+    },
+    {
+      "epoch": 20.20038784744667,
+      "grad_norm": 1.1939396858215332,
+      "learning_rate": 0.001,
+      "loss": 1.1831,
+      "step": 62500
+    },
+    {
+      "epoch": 20.232708468002585,
+      "grad_norm": 1.3018850088119507,
+      "learning_rate": 0.001,
+      "loss": 1.192,
+      "step": 62600
+    },
+    {
+      "epoch": 20.2650290885585,
+      "grad_norm": 0.9886336922645569,
+      "learning_rate": 0.001,
+      "loss": 1.2002,
+      "step": 62700
+    },
+    {
+      "epoch": 20.297349709114414,
+      "grad_norm": 1.4016402959823608,
+      "learning_rate": 0.001,
+      "loss": 1.2176,
+      "step": 62800
+    },
+    {
+      "epoch": 20.32967032967033,
+      "grad_norm": 1.0506421327590942,
+      "learning_rate": 0.001,
+      "loss": 1.2063,
+      "step": 62900
+    },
+    {
+      "epoch": 20.361990950226243,
+      "grad_norm": 1.0052201747894287,
+      "learning_rate": 0.001,
+      "loss": 1.2195,
+      "step": 63000
+    },
+    {
+      "epoch": 20.394311570782158,
+      "grad_norm": 1.0880728960037231,
+      "learning_rate": 0.001,
+      "loss": 1.2034,
+      "step": 63100
+    },
+    {
+      "epoch": 20.426632191338072,
+      "grad_norm": 1.157288670539856,
+      "learning_rate": 0.001,
+      "loss": 1.2042,
+      "step": 63200
+    },
+    {
+      "epoch": 20.458952811893987,
+      "grad_norm": 0.956786036491394,
+      "learning_rate": 0.001,
+      "loss": 1.1999,
+      "step": 63300
+    },
+    {
+      "epoch": 20.4912734324499,
+      "grad_norm": 1.3962632417678833,
+      "learning_rate": 0.001,
+      "loss": 1.2276,
+      "step": 63400
+    },
+    {
+      "epoch": 20.52359405300582,
+      "grad_norm": 1.192475438117981,
+      "learning_rate": 0.001,
+      "loss": 1.196,
+      "step": 63500
+    },
+    {
+      "epoch": 20.555914673561734,
+      "grad_norm": 1.085652470588684,
+      "learning_rate": 0.001,
+      "loss": 1.2409,
+      "step": 63600
+    },
+    {
+      "epoch": 20.58823529411765,
+      "grad_norm": 1.0996172428131104,
+      "learning_rate": 0.001,
+      "loss": 1.2265,
+      "step": 63700
+    },
+    {
+      "epoch": 20.620555914673563,
+      "grad_norm": 1.1674541234970093,
+      "learning_rate": 0.001,
+      "loss": 1.2195,
+      "step": 63800
+    },
+    {
+      "epoch": 20.652876535229478,
+      "grad_norm": 1.3376907110214233,
+      "learning_rate": 0.001,
+      "loss": 1.2255,
+      "step": 63900
+    },
+    {
+      "epoch": 20.685197155785392,
+      "grad_norm": 1.0617411136627197,
+      "learning_rate": 0.001,
+      "loss": 1.2072,
+      "step": 64000
+    },
+    {
+      "epoch": 20.717517776341307,
+      "grad_norm": 0.935806930065155,
+      "learning_rate": 0.001,
+      "loss": 1.2313,
+      "step": 64100
+    },
+    {
+      "epoch": 20.74983839689722,
+      "grad_norm": 1.2648018598556519,
+      "learning_rate": 0.001,
+      "loss": 1.2438,
+      "step": 64200
+    },
+    {
+      "epoch": 20.782159017453136,
+      "grad_norm": 1.2684612274169922,
+      "learning_rate": 0.001,
+      "loss": 1.2323,
+      "step": 64300
+    },
+    {
+      "epoch": 20.81447963800905,
+      "grad_norm": 1.0561871528625488,
+      "learning_rate": 0.001,
+      "loss": 1.2478,
+      "step": 64400
+    },
+    {
+      "epoch": 20.846800258564965,
+      "grad_norm": 1.1016511917114258,
+      "learning_rate": 0.001,
+      "loss": 1.2459,
+      "step": 64500
+    },
+    {
+      "epoch": 20.87912087912088,
+      "grad_norm": 1.1570508480072021,
+      "learning_rate": 0.001,
+      "loss": 1.272,
+      "step": 64600
+    },
+    {
+      "epoch": 20.911441499676794,
+      "grad_norm": 1.2147791385650635,
+      "learning_rate": 0.001,
+      "loss": 1.2439,
+      "step": 64700
+    },
+    {
+      "epoch": 20.94376212023271,
+      "grad_norm": 1.2510056495666504,
+      "learning_rate": 0.001,
+      "loss": 1.2652,
+      "step": 64800
+    },
+    {
+      "epoch": 20.976082740788623,
+      "grad_norm": 0.9977579116821289,
+      "learning_rate": 0.001,
+      "loss": 1.2608,
+      "step": 64900
+    },
+    {
+      "epoch": 21.008403361344538,
+      "grad_norm": 1.4413050413131714,
+      "learning_rate": 0.001,
+      "loss": 1.2153,
+      "step": 65000
+    },
+    {
+      "epoch": 21.040723981900452,
+      "grad_norm": 1.1104098558425903,
+      "learning_rate": 0.001,
+      "loss": 1.1585,
+      "step": 65100
+    },
+    {
+      "epoch": 21.073044602456367,
+      "grad_norm": 1.1735037565231323,
+      "learning_rate": 0.001,
+      "loss": 1.1601,
+      "step": 65200
+    },
+    {
+      "epoch": 21.10536522301228,
+      "grad_norm": 1.2895740270614624,
+      "learning_rate": 0.001,
+      "loss": 1.1557,
+      "step": 65300
+    },
+    {
+      "epoch": 21.137685843568196,
+      "grad_norm": 1.7517699003219604,
+      "learning_rate": 0.001,
+      "loss": 1.1728,
+      "step": 65400
+    },
+    {
+      "epoch": 21.17000646412411,
+      "grad_norm": 1.2219549417495728,
+      "learning_rate": 0.001,
+      "loss": 1.1527,
+      "step": 65500
+    },
+    {
+      "epoch": 21.202327084680025,
+      "grad_norm": 1.2364577054977417,
+      "learning_rate": 0.001,
+      "loss": 1.1771,
+      "step": 65600
+    },
+    {
+      "epoch": 21.23464770523594,
+      "grad_norm": 1.2639278173446655,
+      "learning_rate": 0.001,
+      "loss": 1.1716,
+      "step": 65700
+    },
+    {
+      "epoch": 21.266968325791854,
+      "grad_norm": 1.2081694602966309,
+      "learning_rate": 0.001,
+      "loss": 1.1708,
+      "step": 65800
+    },
+    {
+      "epoch": 21.29928894634777,
+      "grad_norm": 1.219407558441162,
+      "learning_rate": 0.001,
+      "loss": 1.1861,
+      "step": 65900
+    },
+    {
+      "epoch": 21.331609566903683,
+      "grad_norm": 1.067844271659851,
+      "learning_rate": 0.001,
+      "loss": 1.1979,
+      "step": 66000
+    },
+    {
+      "epoch": 21.363930187459598,
+      "grad_norm": 1.2782212495803833,
+      "learning_rate": 0.001,
+      "loss": 1.1796,
+      "step": 66100
+    },
+    {
+      "epoch": 21.396250808015512,
+      "grad_norm": 1.3282757997512817,
+      "learning_rate": 0.001,
+      "loss": 1.186,
+      "step": 66200
+    },
+    {
+      "epoch": 21.428571428571427,
+      "grad_norm": 1.018190860748291,
+      "learning_rate": 0.001,
+      "loss": 1.242,
+      "step": 66300
+    },
+    {
+      "epoch": 21.46089204912734,
+      "grad_norm": 1.1868315935134888,
+      "learning_rate": 0.001,
+      "loss": 1.2631,
+      "step": 66400
+    },
+    {
+      "epoch": 21.49321266968326,
+      "grad_norm": 1.3459444046020508,
+      "learning_rate": 0.001,
+      "loss": 1.2075,
+      "step": 66500
+    },
+    {
+      "epoch": 21.525533290239174,
+      "grad_norm": 0.9880927801132202,
+      "learning_rate": 0.001,
+      "loss": 1.1989,
+      "step": 66600
+    },
+    {
+      "epoch": 21.55785391079509,
+      "grad_norm": 1.5468779802322388,
+      "learning_rate": 0.001,
+      "loss": 1.2444,
+      "step": 66700
+    },
+    {
+      "epoch": 21.590174531351003,
+      "grad_norm": 0.9835416674613953,
+      "learning_rate": 0.001,
+      "loss": 1.2407,
+      "step": 66800
+    },
+    {
+      "epoch": 21.622495151906918,
+      "grad_norm": 1.1654820442199707,
+      "learning_rate": 0.001,
+      "loss": 1.2372,
+      "step": 66900
+    },
+    {
+      "epoch": 21.654815772462833,
+      "grad_norm": 1.2552804946899414,
+      "learning_rate": 0.001,
+      "loss": 1.2291,
+      "step": 67000
+    },
+    {
+      "epoch": 21.687136393018747,
+      "grad_norm": 3.475341796875,
+      "learning_rate": 0.001,
+      "loss": 1.2355,
+      "step": 67100
+    },
+    {
+      "epoch": 21.71945701357466,
+      "grad_norm": 1.0410614013671875,
+      "learning_rate": 0.001,
+      "loss": 1.234,
+      "step": 67200
+    },
+    {
+      "epoch": 21.751777634130576,
+      "grad_norm": 1.168955683708191,
+      "learning_rate": 0.001,
+      "loss": 1.2406,
+      "step": 67300
+    },
+    {
+      "epoch": 21.78409825468649,
+      "grad_norm": 1.7832200527191162,
+      "learning_rate": 0.001,
+      "loss": 1.2689,
+      "step": 67400
+    },
+    {
+      "epoch": 21.816418875242405,
+      "grad_norm": 1.3156075477600098,
+      "learning_rate": 0.001,
+      "loss": 1.2598,
+      "step": 67500
+    },
+    {
+      "epoch": 21.84873949579832,
+      "grad_norm": 1.005856990814209,
+      "learning_rate": 0.001,
+      "loss": 1.2477,
+      "step": 67600
+    },
+    {
+      "epoch": 21.881060116354234,
+      "grad_norm": 0.9797137379646301,
+      "learning_rate": 0.001,
+      "loss": 1.2534,
+      "step": 67700
+    },
+    {
+      "epoch": 21.91338073691015,
+      "grad_norm": 0.9590547680854797,
+      "learning_rate": 0.001,
+      "loss": 1.2328,
+      "step": 67800
+    },
+    {
+      "epoch": 21.945701357466064,
+      "grad_norm": 1.1797035932540894,
+      "learning_rate": 0.001,
+      "loss": 1.2347,
+      "step": 67900
+    },
+    {
+      "epoch": 21.978021978021978,
+      "grad_norm": 1.065169334411621,
+      "learning_rate": 0.001,
+      "loss": 1.2465,
+      "step": 68000
+    },
+    {
+      "epoch": 22.010342598577893,
+      "grad_norm": 1.0149396657943726,
+      "learning_rate": 0.001,
+      "loss": 1.1883,
+      "step": 68100
+    },
+    {
+      "epoch": 22.042663219133807,
+      "grad_norm": 1.1508020162582397,
+      "learning_rate": 0.001,
+      "loss": 1.156,
+      "step": 68200
+    },
+    {
+      "epoch": 22.07498383968972,
+      "grad_norm": 3.199270725250244,
+      "learning_rate": 0.001,
+      "loss": 1.1385,
+      "step": 68300
+    },
+    {
+      "epoch": 22.107304460245636,
+      "grad_norm": 1.0568350553512573,
+      "learning_rate": 0.001,
+      "loss": 1.1544,
+      "step": 68400
+    },
+    {
+      "epoch": 22.13962508080155,
+      "grad_norm": 5.397552013397217,
+      "learning_rate": 0.001,
+      "loss": 1.149,
+      "step": 68500
+    },
+    {
+      "epoch": 22.171945701357465,
+      "grad_norm": 1.1484830379486084,
+      "learning_rate": 0.001,
+      "loss": 1.1669,
+      "step": 68600
+    },
+    {
+      "epoch": 22.20426632191338,
+      "grad_norm": 1.3436305522918701,
+      "learning_rate": 0.001,
+      "loss": 1.1669,
+      "step": 68700
+    },
+    {
+      "epoch": 22.236586942469295,
+      "grad_norm": 1.3443758487701416,
+      "learning_rate": 0.001,
+      "loss": 1.1538,
+      "step": 68800
+    },
+    {
+      "epoch": 22.26890756302521,
+      "grad_norm": 1.4839012622833252,
+      "learning_rate": 0.001,
+      "loss": 1.1557,
+      "step": 68900
+    },
+    {
+      "epoch": 22.301228183581124,
+      "grad_norm": 1.2252734899520874,
+      "learning_rate": 0.001,
+      "loss": 1.1756,
+      "step": 69000
+    },
+    {
+      "epoch": 22.33354880413704,
+      "grad_norm": 1.0860698223114014,
+      "learning_rate": 0.001,
+      "loss": 1.161,
+      "step": 69100
+    },
+    {
+      "epoch": 22.365869424692953,
+      "grad_norm": 0.9651658535003662,
+      "learning_rate": 0.001,
+      "loss": 1.2006,
+      "step": 69200
+    },
+    {
+      "epoch": 22.398190045248867,
+      "grad_norm": 1.0372250080108643,
+      "learning_rate": 0.001,
+      "loss": 1.1829,
+      "step": 69300
+    },
+    {
+      "epoch": 22.430510665804782,
+      "grad_norm": 1.0921285152435303,
+      "learning_rate": 0.001,
+      "loss": 1.1988,
+      "step": 69400
+    },
+    {
+      "epoch": 22.462831286360696,
+      "grad_norm": 1.0957082509994507,
+      "learning_rate": 0.001,
+      "loss": 1.1702,
+      "step": 69500
+    },
+    {
+      "epoch": 22.49515190691661,
+      "grad_norm": 0.9881604909896851,
+      "learning_rate": 0.001,
+      "loss": 1.1969,
+      "step": 69600
+    },
+    {
+      "epoch": 22.52747252747253,
+      "grad_norm": 1.0762852430343628,
+      "learning_rate": 0.001,
+      "loss": 1.1866,
+      "step": 69700
+    },
+    {
+      "epoch": 22.559793148028444,
+      "grad_norm": 1.1281818151474,
+      "learning_rate": 0.001,
+      "loss": 1.1812,
+      "step": 69800
+    },
+    {
+      "epoch": 22.59211376858436,
+      "grad_norm": 1.1947038173675537,
+      "learning_rate": 0.001,
+      "loss": 1.1999,
+      "step": 69900
+    },
+    {
+      "epoch": 22.624434389140273,
+      "grad_norm": 1.0397061109542847,
+      "learning_rate": 0.001,
+      "loss": 1.1993,
+      "step": 70000
+    },
+    {
+      "epoch": 22.656755009696187,
+      "grad_norm": 0.9127522110939026,
+      "learning_rate": 0.001,
+      "loss": 1.1945,
+      "step": 70100
+    },
+    {
+      "epoch": 22.689075630252102,
+      "grad_norm": 1.0395044088363647,
+      "learning_rate": 0.001,
+      "loss": 1.2217,
+      "step": 70200
+    },
+    {
+      "epoch": 22.721396250808017,
+      "grad_norm": 1.1878328323364258,
+      "learning_rate": 0.001,
+      "loss": 1.211,
+      "step": 70300
+    },
+    {
+      "epoch": 22.75371687136393,
+      "grad_norm": 1.0908920764923096,
+      "learning_rate": 0.001,
+      "loss": 1.1992,
+      "step": 70400
+    },
+    {
+      "epoch": 22.786037491919846,
+      "grad_norm": 1.1181707382202148,
+      "learning_rate": 0.001,
+      "loss": 1.1994,
+      "step": 70500
+    },
+    {
+      "epoch": 22.81835811247576,
+      "grad_norm": 0.9277530312538147,
+      "learning_rate": 0.001,
+      "loss": 1.2147,
+      "step": 70600
+    },
+    {
+      "epoch": 22.850678733031675,
+      "grad_norm": 1.4166144132614136,
+      "learning_rate": 0.001,
+      "loss": 1.1931,
+      "step": 70700
+    },
+    {
+      "epoch": 22.88299935358759,
+      "grad_norm": 1.0426658391952515,
+      "learning_rate": 0.001,
+      "loss": 1.2126,
+      "step": 70800
+    },
+    {
+      "epoch": 22.915319974143504,
+      "grad_norm": 1.1179332733154297,
+      "learning_rate": 0.001,
+      "loss": 1.2002,
+      "step": 70900
+    },
+    {
+      "epoch": 22.94764059469942,
+      "grad_norm": 1.1418015956878662,
+      "learning_rate": 0.001,
+      "loss": 1.206,
+      "step": 71000
+    },
+    {
+      "epoch": 22.979961215255333,
+      "grad_norm": 1.1119362115859985,
+      "learning_rate": 0.001,
+      "loss": 1.2136,
+      "step": 71100
+    },
+    {
+      "epoch": 23.012281835811248,
+      "grad_norm": 1.1023640632629395,
+      "learning_rate": 0.001,
+      "loss": 1.133,
+      "step": 71200
+    },
+    {
+      "epoch": 23.044602456367162,
+      "grad_norm": 1.1139122247695923,
+      "learning_rate": 0.001,
+      "loss": 1.1339,
+      "step": 71300
+    },
+    {
+      "epoch": 23.076923076923077,
+      "grad_norm": 1.0510680675506592,
+      "learning_rate": 0.001,
+      "loss": 1.0907,
+      "step": 71400
+    },
+    {
+      "epoch": 23.10924369747899,
+      "grad_norm": 1.1501266956329346,
+      "learning_rate": 0.001,
+      "loss": 1.1127,
+      "step": 71500
+    },
+    {
+      "epoch": 23.141564318034906,
+      "grad_norm": 1.1204190254211426,
+      "learning_rate": 0.001,
+      "loss": 1.1262,
+      "step": 71600
+    },
+    {
+      "epoch": 23.17388493859082,
+      "grad_norm": 1.0314933061599731,
+      "learning_rate": 0.001,
+      "loss": 1.1214,
+      "step": 71700
+    },
+    {
+      "epoch": 23.206205559146735,
+      "grad_norm": 1.4279398918151855,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 71800
+    },
+    {
+      "epoch": 23.23852617970265,
+      "grad_norm": 0.9070366621017456,
+      "learning_rate": 0.001,
+      "loss": 1.1344,
+      "step": 71900
+    },
+    {
+      "epoch": 23.270846800258564,
+      "grad_norm": 1.1681766510009766,
+      "learning_rate": 0.001,
+      "loss": 1.1446,
+      "step": 72000
+    },
+    {
+      "epoch": 23.30316742081448,
+      "grad_norm": 1.0091142654418945,
+      "learning_rate": 0.001,
+      "loss": 1.1534,
+      "step": 72100
+    },
+    {
+      "epoch": 23.335488041370393,
+      "grad_norm": 1.163992166519165,
+      "learning_rate": 0.001,
+      "loss": 1.1348,
+      "step": 72200
+    },
+    {
+      "epoch": 23.367808661926308,
+      "grad_norm": 1.4471473693847656,
+      "learning_rate": 0.001,
+      "loss": 1.1372,
+      "step": 72300
+    },
+    {
+      "epoch": 23.400129282482222,
+      "grad_norm": 1.072084665298462,
+      "learning_rate": 0.001,
+      "loss": 1.1513,
+      "step": 72400
+    },
+    {
+      "epoch": 23.432449903038137,
+      "grad_norm": 1.074723720550537,
+      "learning_rate": 0.001,
+      "loss": 1.1345,
+      "step": 72500
+    },
+    {
+      "epoch": 23.46477052359405,
+      "grad_norm": 1.3358465433120728,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 72600
+    },
+    {
+      "epoch": 23.49709114414997,
+      "grad_norm": 1.1754719018936157,
+      "learning_rate": 0.001,
+      "loss": 1.1531,
+      "step": 72700
+    },
+    {
+      "epoch": 23.529411764705884,
+      "grad_norm": 1.1433322429656982,
+      "learning_rate": 0.001,
+      "loss": 1.1684,
+      "step": 72800
+    },
+    {
+      "epoch": 23.5617323852618,
+      "grad_norm": 1.201206922531128,
+      "learning_rate": 0.001,
+      "loss": 1.1648,
+      "step": 72900
+    },
+    {
+      "epoch": 23.594053005817713,
+      "grad_norm": 1.3061528205871582,
+      "learning_rate": 0.001,
+      "loss": 1.177,
+      "step": 73000
+    },
+    {
+      "epoch": 23.626373626373628,
+      "grad_norm": 1.055266261100769,
+      "learning_rate": 0.001,
+      "loss": 1.1715,
+      "step": 73100
+    },
+    {
+      "epoch": 23.658694246929542,
+      "grad_norm": 1.279536247253418,
+      "learning_rate": 0.001,
+      "loss": 1.1822,
+      "step": 73200
+    },
+    {
+      "epoch": 23.691014867485457,
+      "grad_norm": 1.0013713836669922,
+      "learning_rate": 0.001,
+      "loss": 1.182,
+      "step": 73300
+    },
+    {
+      "epoch": 23.72333548804137,
+      "grad_norm": 1.2121005058288574,
+      "learning_rate": 0.001,
+      "loss": 1.1937,
+      "step": 73400
+    },
+    {
+      "epoch": 23.755656108597286,
+      "grad_norm": 1.1226108074188232,
+      "learning_rate": 0.001,
+      "loss": 1.1872,
+      "step": 73500
+    },
+    {
+      "epoch": 23.7879767291532,
+      "grad_norm": 1.16231369972229,
+      "learning_rate": 0.001,
+      "loss": 1.2158,
+      "step": 73600
+    },
+    {
+      "epoch": 23.820297349709115,
+      "grad_norm": 0.9862212538719177,
+      "learning_rate": 0.001,
+      "loss": 1.1814,
+      "step": 73700
+    },
+    {
+      "epoch": 23.85261797026503,
+      "grad_norm": 1.0801526308059692,
+      "learning_rate": 0.001,
+      "loss": 1.1786,
+      "step": 73800
+    },
+    {
+      "epoch": 23.884938590820944,
+      "grad_norm": 1.2229857444763184,
+      "learning_rate": 0.001,
+      "loss": 1.1822,
+      "step": 73900
+    },
+    {
+      "epoch": 23.91725921137686,
+      "grad_norm": 1.4451088905334473,
+      "learning_rate": 0.001,
+      "loss": 1.1959,
+      "step": 74000
+    },
+    {
+      "epoch": 23.949579831932773,
+      "grad_norm": 0.986847460269928,
+      "learning_rate": 0.001,
+      "loss": 1.1885,
+      "step": 74100
+    },
+    {
+      "epoch": 23.981900452488688,
+      "grad_norm": 1.1042050123214722,
+      "learning_rate": 0.001,
+      "loss": 1.1995,
+      "step": 74200
+    },
+    {
+      "epoch": 24.014221073044602,
+      "grad_norm": 0.8930547833442688,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 74300
+    },
+    {
+      "epoch": 24.046541693600517,
+      "grad_norm": 1.0590393543243408,
+      "learning_rate": 0.001,
+      "loss": 1.0869,
+      "step": 74400
+    },
+    {
+      "epoch": 24.07886231415643,
+      "grad_norm": 1.775429368019104,
+      "learning_rate": 0.001,
+      "loss": 1.0912,
+      "step": 74500
+    },
+    {
+      "epoch": 24.111182934712346,
+      "grad_norm": 1.147539734840393,
+      "learning_rate": 0.001,
+      "loss": 1.088,
+      "step": 74600
+    },
+    {
+      "epoch": 24.14350355526826,
+      "grad_norm": 0.8840130567550659,
+      "learning_rate": 0.001,
+      "loss": 1.1231,
+      "step": 74700
+    },
+    {
+      "epoch": 24.175824175824175,
+      "grad_norm": 1.4045928716659546,
+      "learning_rate": 0.001,
+      "loss": 1.1347,
+      "step": 74800
+    },
+    {
+      "epoch": 24.20814479638009,
+      "grad_norm": 1.090484857559204,
+      "learning_rate": 0.001,
+      "loss": 1.0942,
+      "step": 74900
+    },
+    {
+      "epoch": 24.240465416936004,
+      "grad_norm": 1.1950263977050781,
+      "learning_rate": 0.001,
+      "loss": 1.1123,
+      "step": 75000
+    },
+    {
+      "epoch": 24.27278603749192,
+      "grad_norm": 2.7916829586029053,
+      "learning_rate": 0.001,
+      "loss": 1.113,
+      "step": 75100
+    },
+    {
+      "epoch": 24.305106658047833,
+      "grad_norm": 1.1619802713394165,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 75200
+    },
+    {
+      "epoch": 24.337427278603748,
+      "grad_norm": 0.9073593616485596,
+      "learning_rate": 0.001,
+      "loss": 1.1345,
+      "step": 75300
+    },
+    {
+      "epoch": 24.369747899159663,
+      "grad_norm": 0.8575296401977539,
+      "learning_rate": 0.001,
+      "loss": 1.1108,
+      "step": 75400
+    },
+    {
+      "epoch": 24.402068519715577,
+      "grad_norm": 1.386518955230713,
+      "learning_rate": 0.001,
+      "loss": 1.1076,
+      "step": 75500
+    },
+    {
+      "epoch": 24.43438914027149,
+      "grad_norm": 76.49897003173828,
+      "learning_rate": 0.001,
+      "loss": 1.1144,
+      "step": 75600
+    },
+    {
+      "epoch": 24.466709760827406,
+      "grad_norm": 0.9959169626235962,
+      "learning_rate": 0.001,
+      "loss": 1.1121,
+      "step": 75700
+    },
+    {
+      "epoch": 24.49903038138332,
+      "grad_norm": 0.896315336227417,
+      "learning_rate": 0.001,
+      "loss": 1.1449,
+      "step": 75800
+    },
+    {
+      "epoch": 24.53135100193924,
+      "grad_norm": 0.8647759556770325,
+      "learning_rate": 0.001,
+      "loss": 1.1313,
+      "step": 75900
+    },
+    {
+      "epoch": 24.563671622495153,
+      "grad_norm": 1.080623984336853,
+      "learning_rate": 0.001,
+      "loss": 1.1644,
+      "step": 76000
+    },
+    {
+      "epoch": 24.595992243051068,
+      "grad_norm": 1.1319677829742432,
+      "learning_rate": 0.001,
+      "loss": 1.1349,
+      "step": 76100
+    },
+    {
+      "epoch": 24.628312863606983,
+      "grad_norm": 1.0553691387176514,
+      "learning_rate": 0.001,
+      "loss": 1.1447,
+      "step": 76200
+    },
+    {
+      "epoch": 24.660633484162897,
+      "grad_norm": 1.051260232925415,
+      "learning_rate": 0.001,
+      "loss": 1.1257,
+      "step": 76300
+    },
+    {
+      "epoch": 24.69295410471881,
+      "grad_norm": 1.0142831802368164,
+      "learning_rate": 0.001,
+      "loss": 1.161,
+      "step": 76400
+    },
+    {
+      "epoch": 24.725274725274726,
+      "grad_norm": 1.0020849704742432,
+      "learning_rate": 0.001,
+      "loss": 1.1579,
+      "step": 76500
+    },
+    {
+      "epoch": 24.75759534583064,
+      "grad_norm": 0.996508002281189,
+      "learning_rate": 0.001,
+      "loss": 1.1621,
+      "step": 76600
+    },
+    {
+      "epoch": 24.789915966386555,
+      "grad_norm": 1.1945658922195435,
+      "learning_rate": 0.001,
+      "loss": 1.1346,
+      "step": 76700
+    },
+    {
+      "epoch": 24.82223658694247,
+      "grad_norm": 0.9318341612815857,
+      "learning_rate": 0.001,
+      "loss": 1.1534,
+      "step": 76800
+    },
+    {
+      "epoch": 24.854557207498384,
+      "grad_norm": 1.0282987356185913,
+      "learning_rate": 0.001,
+      "loss": 1.1679,
+      "step": 76900
+    },
+    {
+      "epoch": 24.8868778280543,
+      "grad_norm": 0.9928200840950012,
+      "learning_rate": 0.001,
+      "loss": 1.1851,
+      "step": 77000
+    },
+    {
+      "epoch": 24.919198448610214,
+      "grad_norm": 1.1055632829666138,
+      "learning_rate": 0.001,
+      "loss": 1.1779,
+      "step": 77100
+    },
+    {
+      "epoch": 24.951519069166128,
+      "grad_norm": 0.9744527339935303,
+      "learning_rate": 0.001,
+      "loss": 1.1775,
+      "step": 77200
+    },
+    {
+      "epoch": 24.983839689722043,
+      "grad_norm": 1.0894049406051636,
+      "learning_rate": 0.001,
+      "loss": 1.1786,
+      "step": 77300
+    },
+    {
+      "epoch": 25.016160310277957,
+      "grad_norm": 0.2896379232406616,
+      "learning_rate": 0.001,
+      "loss": 1.0473,
+      "step": 77400
+    },
+    {
+      "epoch": 25.048480930833872,
+      "grad_norm": 0.6147540211677551,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 77500
+    },
+    {
+      "epoch": 25.080801551389786,
+      "grad_norm": 0.5519897937774658,
+      "learning_rate": 0.001,
+      "loss": 1.0645,
+      "step": 77600
+    },
+    {
+      "epoch": 25.1131221719457,
+      "grad_norm": 0.30935660004615784,
+      "learning_rate": 0.001,
+      "loss": 1.0736,
+      "step": 77700
+    },
+    {
+      "epoch": 25.145442792501616,
+      "grad_norm": 1.2920175790786743,
+      "learning_rate": 0.001,
+      "loss": 1.0879,
+      "step": 77800
+    },
+    {
+      "epoch": 25.17776341305753,
+      "grad_norm": 0.775704562664032,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 77900
+    },
+    {
+      "epoch": 25.210084033613445,
+      "grad_norm": 0.5069411993026733,
+      "learning_rate": 0.001,
+      "loss": 1.096,
+      "step": 78000
+    },
+    {
+      "epoch": 25.24240465416936,
+      "grad_norm": 1.0527359247207642,
+      "learning_rate": 0.001,
+      "loss": 1.0937,
+      "step": 78100
+    },
+    {
+      "epoch": 25.274725274725274,
+      "grad_norm": 1.1386781930923462,
+      "learning_rate": 0.001,
+      "loss": 1.0713,
+      "step": 78200
+    },
+    {
+      "epoch": 25.30704589528119,
+      "grad_norm": 0.48567479848861694,
+      "learning_rate": 0.001,
+      "loss": 1.1076,
+      "step": 78300
+    },
+    {
+      "epoch": 25.339366515837103,
+      "grad_norm": 0.3815343379974365,
+      "learning_rate": 0.001,
+      "loss": 1.1034,
+      "step": 78400
+    },
+    {
+      "epoch": 25.371687136393017,
+      "grad_norm": 0.11455032974481583,
+      "learning_rate": 0.001,
+      "loss": 1.1048,
+      "step": 78500
+    },
+    {
+      "epoch": 25.404007756948932,
+      "grad_norm": 0.6132304072380066,
+      "learning_rate": 0.001,
+      "loss": 1.1235,
+      "step": 78600
+    },
+    {
+      "epoch": 25.436328377504847,
+      "grad_norm": 0.564409613609314,
+      "learning_rate": 0.001,
+      "loss": 1.1046,
+      "step": 78700
+    },
+    {
+      "epoch": 25.46864899806076,
+      "grad_norm": 0.3146139085292816,
+      "learning_rate": 0.001,
+      "loss": 1.1204,
+      "step": 78800
+    },
+    {
+      "epoch": 25.50096961861668,
+      "grad_norm": 0.21083228290081024,
+      "learning_rate": 0.001,
+      "loss": 1.0924,
+      "step": 78900
+    },
+    {
+      "epoch": 25.533290239172594,
+      "grad_norm": 0.9692633152008057,
+      "learning_rate": 0.001,
+      "loss": 1.1529,
+      "step": 79000
+    },
+    {
+      "epoch": 25.56561085972851,
+      "grad_norm": 0.813957929611206,
+      "learning_rate": 0.001,
+      "loss": 1.1256,
+      "step": 79100
+    },
+    {
+      "epoch": 25.597931480284423,
+      "grad_norm": 1.2155555486679077,
+      "learning_rate": 0.001,
+      "loss": 1.1369,
+      "step": 79200
+    },
+    {
+      "epoch": 25.630252100840337,
+      "grad_norm": 0.21164648234844208,
+      "learning_rate": 0.001,
+      "loss": 1.1224,
+      "step": 79300
+    },
+    {
+      "epoch": 25.662572721396252,
+      "grad_norm": 1.0323749780654907,
+      "learning_rate": 0.001,
+      "loss": 1.1336,
+      "step": 79400
+    },
+    {
+      "epoch": 25.694893341952167,
+      "grad_norm": 0.2119108885526657,
+      "learning_rate": 0.001,
+      "loss": 1.1362,
+      "step": 79500
+    },
+    {
+      "epoch": 25.72721396250808,
+      "grad_norm": 0.8170724511146545,
+      "learning_rate": 0.001,
+      "loss": 1.125,
+      "step": 79600
+    },
+    {
+      "epoch": 25.759534583063996,
+      "grad_norm": 0.5977357625961304,
+      "learning_rate": 0.001,
+      "loss": 1.1174,
+      "step": 79700
+    },
+    {
+      "epoch": 25.79185520361991,
+      "grad_norm": 0.6432121396064758,
+      "learning_rate": 0.001,
+      "loss": 1.1451,
+      "step": 79800
+    },
+    {
+      "epoch": 25.824175824175825,
+      "grad_norm": 0.4329550862312317,
+      "learning_rate": 0.001,
+      "loss": 1.1333,
+      "step": 79900
+    },
+    {
+      "epoch": 25.85649644473174,
+      "grad_norm": 0.66637122631073,
+      "learning_rate": 0.001,
+      "loss": 1.1565,
+      "step": 80000
+    },
+    {
+      "epoch": 25.888817065287654,
+      "grad_norm": 0.4904533922672272,
+      "learning_rate": 0.001,
+      "loss": 1.1281,
+      "step": 80100
+    },
+    {
+      "epoch": 25.92113768584357,
+      "grad_norm": 0.5748351812362671,
+      "learning_rate": 0.001,
+      "loss": 1.1416,
+      "step": 80200
+    },
+    {
+      "epoch": 25.953458306399483,
+      "grad_norm": 0.5132641792297363,
+      "learning_rate": 0.001,
+      "loss": 1.1453,
+      "step": 80300
+    },
+    {
+      "epoch": 25.985778926955398,
+      "grad_norm": 0.8233006596565247,
+      "learning_rate": 0.001,
+      "loss": 1.142,
+      "step": 80400
+    },
+    {
+      "epoch": 26.018099547511312,
+      "grad_norm": 1.450419306755066,
+      "learning_rate": 0.001,
+      "loss": 1.1022,
+      "step": 80500
+    },
+    {
+      "epoch": 26.050420168067227,
+      "grad_norm": 0.9929932951927185,
+      "learning_rate": 0.001,
+      "loss": 1.0498,
+      "step": 80600
+    },
+    {
+      "epoch": 26.08274078862314,
+      "grad_norm": 1.1767561435699463,
+      "learning_rate": 0.001,
+      "loss": 1.0544,
+      "step": 80700
+    },
+    {
+      "epoch": 26.115061409179056,
+      "grad_norm": 1.4456899166107178,
+      "learning_rate": 0.001,
+      "loss": 1.0704,
+      "step": 80800
+    },
+    {
+      "epoch": 26.14738202973497,
+      "grad_norm": 1.3941863775253296,
+      "learning_rate": 0.001,
+      "loss": 1.0555,
+      "step": 80900
+    },
+    {
+      "epoch": 26.179702650290885,
+      "grad_norm": 1.3099164962768555,
+      "learning_rate": 0.001,
+      "loss": 1.0503,
+      "step": 81000
+    },
+    {
+      "epoch": 26.2120232708468,
+      "grad_norm": 1.0749870538711548,
+      "learning_rate": 0.001,
+      "loss": 1.0674,
+      "step": 81100
+    },
+    {
+      "epoch": 26.244343891402714,
+      "grad_norm": 1.2630252838134766,
+      "learning_rate": 0.001,
+      "loss": 1.069,
+      "step": 81200
+    },
+    {
+      "epoch": 26.27666451195863,
+      "grad_norm": 1.1559174060821533,
+      "learning_rate": 0.001,
+      "loss": 1.0838,
+      "step": 81300
+    },
+    {
+      "epoch": 26.308985132514543,
+      "grad_norm": 1.3933382034301758,
+      "learning_rate": 0.001,
+      "loss": 1.0832,
+      "step": 81400
+    },
+    {
+      "epoch": 26.341305753070458,
+      "grad_norm": 1.12375807762146,
+      "learning_rate": 0.001,
+      "loss": 1.1035,
+      "step": 81500
+    },
+    {
+      "epoch": 26.373626373626372,
+      "grad_norm": 1.1551575660705566,
+      "learning_rate": 0.001,
+      "loss": 1.0778,
+      "step": 81600
+    },
+    {
+      "epoch": 26.405946994182287,
+      "grad_norm": 1.2731724977493286,
+      "learning_rate": 0.001,
+      "loss": 1.0918,
+      "step": 81700
+    },
+    {
+      "epoch": 26.4382676147382,
+      "grad_norm": 1.1205143928527832,
+      "learning_rate": 0.001,
+      "loss": 1.0749,
+      "step": 81800
+    },
+    {
+      "epoch": 26.470588235294116,
+      "grad_norm": 1.176575779914856,
+      "learning_rate": 0.001,
+      "loss": 1.079,
+      "step": 81900
+    },
+    {
+      "epoch": 26.50290885585003,
+      "grad_norm": 1.095289707183838,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 82000
+    },
+    {
+      "epoch": 26.53522947640595,
+      "grad_norm": 1.2680126428604126,
+      "learning_rate": 0.001,
+      "loss": 1.0829,
+      "step": 82100
+    },
+    {
+      "epoch": 26.567550096961863,
+      "grad_norm": 1.43263578414917,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 82200
+    },
+    {
+      "epoch": 26.599870717517778,
+      "grad_norm": 1.2441238164901733,
+      "learning_rate": 0.001,
+      "loss": 1.107,
+      "step": 82300
+    },
+    {
+      "epoch": 26.632191338073692,
+      "grad_norm": 1.180927038192749,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 82400
+    },
+    {
+      "epoch": 26.664511958629607,
+      "grad_norm": 1.0788034200668335,
+      "learning_rate": 0.001,
+      "loss": 1.1133,
+      "step": 82500
+    },
+    {
+      "epoch": 26.69683257918552,
+      "grad_norm": 1.4446241855621338,
+      "learning_rate": 0.001,
+      "loss": 1.1139,
+      "step": 82600
+    },
+    {
+      "epoch": 26.729153199741436,
+      "grad_norm": 1.2069514989852905,
+      "learning_rate": 0.001,
+      "loss": 1.1228,
+      "step": 82700
+    },
+    {
+      "epoch": 26.76147382029735,
+      "grad_norm": 1.1585434675216675,
+      "learning_rate": 0.001,
+      "loss": 1.0998,
+      "step": 82800
+    },
+    {
+      "epoch": 26.793794440853265,
+      "grad_norm": 1.341698169708252,
+      "learning_rate": 0.001,
+      "loss": 1.1213,
+      "step": 82900
+    },
+    {
+      "epoch": 26.82611506140918,
+      "grad_norm": 1.0288625955581665,
+      "learning_rate": 0.001,
+      "loss": 1.1245,
+      "step": 83000
+    },
+    {
+      "epoch": 26.858435681965094,
+      "grad_norm": 1.1505529880523682,
+      "learning_rate": 0.001,
+      "loss": 1.1668,
+      "step": 83100
+    },
+    {
+      "epoch": 26.89075630252101,
+      "grad_norm": 1.243490219116211,
+      "learning_rate": 0.001,
+      "loss": 1.1389,
+      "step": 83200
+    },
+    {
+      "epoch": 26.923076923076923,
+      "grad_norm": 1.4656898975372314,
+      "learning_rate": 0.001,
+      "loss": 1.1215,
+      "step": 83300
+    },
+    {
+      "epoch": 26.955397543632838,
+      "grad_norm": 1.286278247833252,
+      "learning_rate": 0.001,
+      "loss": 1.1255,
+      "step": 83400
+    },
+    {
+      "epoch": 26.987718164188752,
+      "grad_norm": 1.291344404220581,
+      "learning_rate": 0.001,
+      "loss": 1.1213,
+      "step": 83500
+    },
+    {
+      "epoch": 27.020038784744667,
+      "grad_norm": 1.1009141206741333,
+      "learning_rate": 0.001,
+      "loss": 1.0931,
+      "step": 83600
+    },
+    {
+      "epoch": 27.05235940530058,
+      "grad_norm": 1.21869695186615,
+      "learning_rate": 0.001,
+      "loss": 1.0301,
+      "step": 83700
+    },
+    {
+      "epoch": 27.084680025856496,
+      "grad_norm": 1.084806203842163,
+      "learning_rate": 0.001,
+      "loss": 1.0152,
+      "step": 83800
+    },
+    {
+      "epoch": 27.11700064641241,
+      "grad_norm": 1.2538260221481323,
+      "learning_rate": 0.001,
+      "loss": 1.0652,
+      "step": 83900
+    },
+    {
+      "epoch": 27.149321266968325,
+      "grad_norm": 1.1370879411697388,
+      "learning_rate": 0.001,
+      "loss": 1.0567,
+      "step": 84000
+    },
+    {
+      "epoch": 27.18164188752424,
+      "grad_norm": 1.0101512670516968,
+      "learning_rate": 0.001,
+      "loss": 1.0426,
+      "step": 84100
+    },
+    {
+      "epoch": 27.213962508080154,
+      "grad_norm": 1.0626932382583618,
+      "learning_rate": 0.001,
+      "loss": 1.0708,
+      "step": 84200
+    },
+    {
+      "epoch": 27.24628312863607,
+      "grad_norm": 1.3760011196136475,
+      "learning_rate": 0.001,
+      "loss": 1.0445,
+      "step": 84300
+    },
+    {
+      "epoch": 27.278603749191983,
+      "grad_norm": 6.08766508102417,
+      "learning_rate": 0.001,
+      "loss": 1.0801,
+      "step": 84400
+    },
+    {
+      "epoch": 27.310924369747898,
+      "grad_norm": 1.370469570159912,
+      "learning_rate": 0.001,
+      "loss": 1.0628,
+      "step": 84500
+    },
+    {
+      "epoch": 27.343244990303813,
+      "grad_norm": 1.2071462869644165,
+      "learning_rate": 0.001,
+      "loss": 1.091,
+      "step": 84600
+    },
+    {
+      "epoch": 27.375565610859727,
+      "grad_norm": 1.372673511505127,
+      "learning_rate": 0.001,
+      "loss": 1.0651,
+      "step": 84700
+    },
+    {
+      "epoch": 27.40788623141564,
+      "grad_norm": 1.2071818113327026,
+      "learning_rate": 0.001,
+      "loss": 1.0622,
+      "step": 84800
+    },
+    {
+      "epoch": 27.440206851971556,
+      "grad_norm": 1.0134035348892212,
+      "learning_rate": 0.001,
+      "loss": 1.0684,
+      "step": 84900
+    },
+    {
+      "epoch": 27.47252747252747,
+      "grad_norm": 1.3614178895950317,
+      "learning_rate": 0.001,
+      "loss": 1.0721,
+      "step": 85000
+    },
+    {
+      "epoch": 27.50484809308339,
+      "grad_norm": 1.2039440870285034,
+      "learning_rate": 0.001,
+      "loss": 1.0645,
+      "step": 85100
+    },
+    {
+      "epoch": 27.537168713639304,
+      "grad_norm": 1.2189968824386597,
+      "learning_rate": 0.001,
+      "loss": 1.0833,
+      "step": 85200
+    },
+    {
+      "epoch": 27.569489334195218,
+      "grad_norm": 1.1108815670013428,
+      "learning_rate": 0.001,
+      "loss": 1.0945,
+      "step": 85300
+    },
+    {
+      "epoch": 27.601809954751133,
+      "grad_norm": 1.3825979232788086,
+      "learning_rate": 0.001,
+      "loss": 1.0647,
+      "step": 85400
+    },
+    {
+      "epoch": 27.634130575307047,
+      "grad_norm": 1.2646015882492065,
+      "learning_rate": 0.001,
+      "loss": 1.0858,
+      "step": 85500
+    },
+    {
+      "epoch": 27.66645119586296,
+      "grad_norm": 1.1898399591445923,
+      "learning_rate": 0.001,
+      "loss": 1.081,
+      "step": 85600
+    },
+    {
+      "epoch": 27.698771816418876,
+      "grad_norm": 1.1825064420700073,
+      "learning_rate": 0.001,
+      "loss": 1.0948,
+      "step": 85700
+    },
+    {
+      "epoch": 27.73109243697479,
+      "grad_norm": 1.337872862815857,
+      "learning_rate": 0.001,
+      "loss": 1.0779,
+      "step": 85800
+    },
+    {
+      "epoch": 27.763413057530705,
+      "grad_norm": 1.146846890449524,
+      "learning_rate": 0.001,
+      "loss": 1.0819,
+      "step": 85900
+    },
+    {
+      "epoch": 27.79573367808662,
+      "grad_norm": 1.468003511428833,
+      "learning_rate": 0.001,
+      "loss": 1.0798,
+      "step": 86000
+    },
+    {
+      "epoch": 27.828054298642535,
+      "grad_norm": 1.125246286392212,
+      "learning_rate": 0.001,
+      "loss": 1.0924,
+      "step": 86100
+    },
+    {
+      "epoch": 27.86037491919845,
+      "grad_norm": 1.1755690574645996,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 86200
+    },
+    {
+      "epoch": 27.892695539754364,
+      "grad_norm": 1.1506797075271606,
+      "learning_rate": 0.001,
+      "loss": 1.1024,
+      "step": 86300
+    },
+    {
+      "epoch": 27.92501616031028,
+      "grad_norm": 1.084913730621338,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 86400
+    },
+    {
+      "epoch": 27.957336780866193,
+      "grad_norm": 1.0553350448608398,
+      "learning_rate": 0.001,
+      "loss": 1.1164,
+      "step": 86500
+    },
+    {
+      "epoch": 27.989657401422107,
+      "grad_norm": 1.1251089572906494,
+      "learning_rate": 0.001,
+      "loss": 1.1268,
+      "step": 86600
+    },
+    {
+      "epoch": 28.021978021978022,
+      "grad_norm": 1.3656169176101685,
+      "learning_rate": 0.001,
+      "loss": 1.0753,
+      "step": 86700
+    },
+    {
+      "epoch": 28.054298642533936,
+      "grad_norm": 1.0576167106628418,
+      "learning_rate": 0.001,
+      "loss": 0.9991,
+      "step": 86800
+    },
+    {
+      "epoch": 28.08661926308985,
+      "grad_norm": 1.0733096599578857,
+      "learning_rate": 0.001,
+      "loss": 1.0239,
+      "step": 86900
+    },
+    {
+      "epoch": 28.118939883645766,
+      "grad_norm": 1.265825629234314,
+      "learning_rate": 0.001,
+      "loss": 1.0298,
+      "step": 87000
+    },
+    {
+      "epoch": 28.15126050420168,
+      "grad_norm": 1.655713438987732,
+      "learning_rate": 0.001,
+      "loss": 1.016,
+      "step": 87100
+    },
+    {
+      "epoch": 28.183581124757595,
+      "grad_norm": 1.3085395097732544,
+      "learning_rate": 0.001,
+      "loss": 1.0343,
+      "step": 87200
+    },
+    {
+      "epoch": 28.21590174531351,
+      "grad_norm": 1.0143513679504395,
+      "learning_rate": 0.001,
+      "loss": 1.0293,
+      "step": 87300
+    },
+    {
+      "epoch": 28.248222365869424,
+      "grad_norm": 1.4002749919891357,
+      "learning_rate": 0.001,
+      "loss": 1.0416,
+      "step": 87400
+    },
+    {
+      "epoch": 28.28054298642534,
+      "grad_norm": 1.082223892211914,
+      "learning_rate": 0.001,
+      "loss": 1.0368,
+      "step": 87500
+    },
+    {
+      "epoch": 28.312863606981253,
+      "grad_norm": 1.3155796527862549,
+      "learning_rate": 0.001,
+      "loss": 1.051,
+      "step": 87600
+    },
+    {
+      "epoch": 28.345184227537167,
+      "grad_norm": 1.3327791690826416,
+      "learning_rate": 0.001,
+      "loss": 1.0128,
+      "step": 87700
+    },
+    {
+      "epoch": 28.377504848093082,
+      "grad_norm": 0.9752927422523499,
+      "learning_rate": 0.001,
+      "loss": 1.0389,
+      "step": 87800
+    },
+    {
+      "epoch": 28.409825468648997,
+      "grad_norm": 1.3138093948364258,
+      "learning_rate": 0.001,
+      "loss": 1.0696,
+      "step": 87900
+    },
+    {
+      "epoch": 28.44214608920491,
+      "grad_norm": 0.9144354462623596,
+      "learning_rate": 0.001,
+      "loss": 1.047,
+      "step": 88000
+    },
+    {
+      "epoch": 28.474466709760826,
+      "grad_norm": 1.2346453666687012,
+      "learning_rate": 0.001,
+      "loss": 1.0398,
+      "step": 88100
+    },
+    {
+      "epoch": 28.50678733031674,
+      "grad_norm": 1.703855037689209,
+      "learning_rate": 0.001,
+      "loss": 1.0524,
+      "step": 88200
+    },
+    {
+      "epoch": 28.53910795087266,
+      "grad_norm": 1.0008935928344727,
+      "learning_rate": 0.001,
+      "loss": 1.0512,
+      "step": 88300
+    },
+    {
+      "epoch": 28.571428571428573,
+      "grad_norm": 1.1595375537872314,
+      "learning_rate": 0.001,
+      "loss": 1.0562,
+      "step": 88400
+    },
+    {
+      "epoch": 28.603749191984488,
+      "grad_norm": 0.9839758276939392,
+      "learning_rate": 0.001,
+      "loss": 1.0767,
+      "step": 88500
+    },
+    {
+      "epoch": 28.636069812540402,
+      "grad_norm": 0.9803606271743774,
+      "learning_rate": 0.001,
+      "loss": 1.0523,
+      "step": 88600
+    },
+    {
+      "epoch": 28.668390433096317,
+      "grad_norm": 1.0053520202636719,
+      "learning_rate": 0.001,
+      "loss": 1.069,
+      "step": 88700
+    },
+    {
+      "epoch": 28.70071105365223,
+      "grad_norm": 1.0004795789718628,
+      "learning_rate": 0.001,
+      "loss": 1.0402,
+      "step": 88800
+    },
+    {
+      "epoch": 28.733031674208146,
+      "grad_norm": 1.1498825550079346,
+      "learning_rate": 0.001,
+      "loss": 1.0616,
+      "step": 88900
+    },
+    {
+      "epoch": 28.76535229476406,
+      "grad_norm": 1.0905274152755737,
+      "learning_rate": 0.001,
+      "loss": 1.0813,
+      "step": 89000
+    },
+    {
+      "epoch": 28.797672915319975,
+      "grad_norm": 0.9474394917488098,
+      "learning_rate": 0.001,
+      "loss": 1.0742,
+      "step": 89100
+    },
+    {
+      "epoch": 28.82999353587589,
+      "grad_norm": 1.319617509841919,
+      "learning_rate": 0.001,
+      "loss": 1.0922,
+      "step": 89200
+    },
+    {
+      "epoch": 28.862314156431804,
+      "grad_norm": 1.1449226140975952,
+      "learning_rate": 0.001,
+      "loss": 1.1161,
+      "step": 89300
+    },
+    {
+      "epoch": 28.89463477698772,
+      "grad_norm": 1.1758530139923096,
+      "learning_rate": 0.001,
+      "loss": 1.0891,
+      "step": 89400
+    },
+    {
+      "epoch": 28.926955397543633,
+      "grad_norm": 1.2047131061553955,
+      "learning_rate": 0.001,
+      "loss": 1.0833,
+      "step": 89500
+    },
+    {
+      "epoch": 28.959276018099548,
+      "grad_norm": 1.0821508169174194,
+      "learning_rate": 0.001,
+      "loss": 1.0829,
+      "step": 89600
+    },
+    {
+      "epoch": 28.991596638655462,
+      "grad_norm": 1.1592580080032349,
+      "learning_rate": 0.001,
+      "loss": 1.0958,
+      "step": 89700
+    },
+    {
+      "epoch": 29.023917259211377,
+      "grad_norm": 1.1290960311889648,
+      "learning_rate": 0.001,
+      "loss": 1.0154,
+      "step": 89800
+    },
+    {
+      "epoch": 29.05623787976729,
+      "grad_norm": 1.4548909664154053,
+      "learning_rate": 0.001,
+      "loss": 0.9883,
+      "step": 89900
+    },
+    {
+      "epoch": 29.088558500323206,
+      "grad_norm": 1.4414570331573486,
+      "learning_rate": 0.001,
+      "loss": 0.9885,
+      "step": 90000
+    },
+    {
+      "epoch": 29.12087912087912,
+      "grad_norm": 1.0668739080429077,
+      "learning_rate": 0.001,
+      "loss": 1.0079,
+      "step": 90100
+    },
+    {
+      "epoch": 29.153199741435035,
+      "grad_norm": 1.2436254024505615,
+      "learning_rate": 0.001,
+      "loss": 1.0177,
+      "step": 90200
+    },
+    {
+      "epoch": 29.18552036199095,
+      "grad_norm": 1.1421066522598267,
+      "learning_rate": 0.001,
+      "loss": 1.0225,
+      "step": 90300
+    },
+    {
+      "epoch": 29.217840982546864,
+      "grad_norm": 0.9146871566772461,
+      "learning_rate": 0.001,
+      "loss": 1.0021,
+      "step": 90400
+    },
+    {
+      "epoch": 29.25016160310278,
+      "grad_norm": 1.0086992979049683,
+      "learning_rate": 0.001,
+      "loss": 1.0366,
+      "step": 90500
+    },
+    {
+      "epoch": 29.282482223658693,
+      "grad_norm": 0.9744951128959656,
+      "learning_rate": 0.001,
+      "loss": 1.0344,
+      "step": 90600
+    },
+    {
+      "epoch": 29.314802844214608,
+      "grad_norm": 1.1832053661346436,
+      "learning_rate": 0.001,
+      "loss": 1.0138,
+      "step": 90700
+    },
+    {
+      "epoch": 29.347123464770522,
+      "grad_norm": 1.3362852334976196,
+      "learning_rate": 0.001,
+      "loss": 1.0176,
+      "step": 90800
+    },
+    {
+      "epoch": 29.379444085326437,
+      "grad_norm": 1.1680186986923218,
+      "learning_rate": 0.001,
+      "loss": 1.0305,
+      "step": 90900
+    },
+    {
+      "epoch": 29.41176470588235,
+      "grad_norm": 1.1634215116500854,
+      "learning_rate": 0.001,
+      "loss": 1.0087,
+      "step": 91000
+    },
+    {
+      "epoch": 29.444085326438266,
+      "grad_norm": 1.0064040422439575,
+      "learning_rate": 0.001,
+      "loss": 1.0162,
+      "step": 91100
+    },
+    {
+      "epoch": 29.47640594699418,
+      "grad_norm": 1.3981666564941406,
+      "learning_rate": 0.001,
+      "loss": 1.0381,
+      "step": 91200
+    },
+    {
+      "epoch": 29.5087265675501,
+      "grad_norm": 1.1178312301635742,
+      "learning_rate": 0.001,
+      "loss": 1.0377,
+      "step": 91300
+    },
+    {
+      "epoch": 29.541047188106013,
+      "grad_norm": 1.013622522354126,
+      "learning_rate": 0.001,
+      "loss": 1.0503,
+      "step": 91400
+    },
+    {
+      "epoch": 29.573367808661928,
+      "grad_norm": 1.1112215518951416,
+      "learning_rate": 0.001,
+      "loss": 1.0493,
+      "step": 91500
+    },
+    {
+      "epoch": 29.605688429217842,
+      "grad_norm": 1.1293482780456543,
+      "learning_rate": 0.001,
+      "loss": 1.043,
+      "step": 91600
+    },
+    {
+      "epoch": 29.638009049773757,
+      "grad_norm": 1.1543068885803223,
+      "learning_rate": 0.001,
+      "loss": 1.0458,
+      "step": 91700
+    },
+    {
+      "epoch": 29.67032967032967,
+      "grad_norm": 0.9590059518814087,
+      "learning_rate": 0.001,
+      "loss": 1.0692,
+      "step": 91800
+    },
+    {
+      "epoch": 29.702650290885586,
+      "grad_norm": 0.9611120820045471,
+      "learning_rate": 0.001,
+      "loss": 1.0507,
+      "step": 91900
+    },
+    {
+      "epoch": 29.7349709114415,
+      "grad_norm": 1.105459451675415,
+      "learning_rate": 0.001,
+      "loss": 1.0477,
+      "step": 92000
+    },
+    {
+      "epoch": 29.767291531997415,
+      "grad_norm": 1.2844051122665405,
+      "learning_rate": 0.001,
+      "loss": 1.0547,
+      "step": 92100
+    },
+    {
+      "epoch": 29.79961215255333,
+      "grad_norm": 1.194689154624939,
+      "learning_rate": 0.001,
+      "loss": 1.0415,
+      "step": 92200
+    },
+    {
+      "epoch": 29.831932773109244,
+      "grad_norm": 1.2303766012191772,
+      "learning_rate": 0.001,
+      "loss": 1.0577,
+      "step": 92300
+    },
+    {
+      "epoch": 29.86425339366516,
+      "grad_norm": 1.0361992120742798,
+      "learning_rate": 0.001,
+      "loss": 1.0653,
+      "step": 92400
+    },
+    {
+      "epoch": 29.896574014221073,
+      "grad_norm": 1.1771466732025146,
+      "learning_rate": 0.001,
+      "loss": 1.0595,
+      "step": 92500
+    },
+    {
+      "epoch": 29.928894634776988,
+      "grad_norm": 1.1949156522750854,
+      "learning_rate": 0.001,
+      "loss": 1.061,
+      "step": 92600
+    },
+    {
+      "epoch": 29.961215255332903,
+      "grad_norm": 1.147527813911438,
+      "learning_rate": 0.001,
+      "loss": 1.0707,
+      "step": 92700
+    },
+    {
+      "epoch": 29.993535875888817,
+      "grad_norm": 1.1559338569641113,
+      "learning_rate": 0.001,
+      "loss": 1.0782,
+      "step": 92800
+    },
+    {
+      "epoch": 30.02585649644473,
+      "grad_norm": 1.3547965288162231,
+      "learning_rate": 0.001,
+      "loss": 1.0003,
+      "step": 92900
+    },
+    {
+      "epoch": 30.058177117000646,
+      "grad_norm": 1.4437899589538574,
+      "learning_rate": 0.001,
+      "loss": 0.966,
+      "step": 93000
+    },
+    {
+      "epoch": 30.09049773755656,
+      "grad_norm": 1.0247992277145386,
+      "learning_rate": 0.001,
+      "loss": 0.9779,
+      "step": 93100
+    },
+    {
+      "epoch": 30.122818358112475,
+      "grad_norm": 1.0951383113861084,
+      "learning_rate": 0.001,
+      "loss": 0.9782,
+      "step": 93200
+    },
+    {
+      "epoch": 30.15513897866839,
+      "grad_norm": 1.2668837308883667,
+      "learning_rate": 0.001,
+      "loss": 0.977,
+      "step": 93300
+    },
+    {
+      "epoch": 30.187459599224304,
+      "grad_norm": 1.2851572036743164,
+      "learning_rate": 0.001,
+      "loss": 0.9819,
+      "step": 93400
+    },
+    {
+      "epoch": 30.21978021978022,
+      "grad_norm": 1.1531295776367188,
+      "learning_rate": 0.001,
+      "loss": 0.9862,
+      "step": 93500
+    },
+    {
+      "epoch": 30.252100840336134,
+      "grad_norm": 1.1562260389328003,
+      "learning_rate": 0.001,
+      "loss": 0.9874,
+      "step": 93600
+    },
+    {
+      "epoch": 30.284421460892048,
+      "grad_norm": 1.5142194032669067,
+      "learning_rate": 0.001,
+      "loss": 1.0211,
+      "step": 93700
+    },
+    {
+      "epoch": 30.316742081447963,
+      "grad_norm": 1.2942471504211426,
+      "learning_rate": 0.001,
+      "loss": 0.9893,
+      "step": 93800
+    },
+    {
+      "epoch": 30.349062702003877,
+      "grad_norm": 1.273345708847046,
+      "learning_rate": 0.001,
+      "loss": 1.0102,
+      "step": 93900
+    },
+    {
+      "epoch": 30.381383322559792,
+      "grad_norm": 1.251236915588379,
+      "learning_rate": 0.001,
+      "loss": 1.0103,
+      "step": 94000
+    },
+    {
+      "epoch": 30.413703943115706,
+      "grad_norm": 1.4119136333465576,
+      "learning_rate": 0.001,
+      "loss": 0.9897,
+      "step": 94100
+    },
+    {
+      "epoch": 30.44602456367162,
+      "grad_norm": 1.2509639263153076,
+      "learning_rate": 0.001,
+      "loss": 1.0158,
+      "step": 94200
+    },
+    {
+      "epoch": 30.478345184227535,
+      "grad_norm": 1.3147063255310059,
+      "learning_rate": 0.001,
+      "loss": 1.0271,
+      "step": 94300
+    },
+    {
+      "epoch": 30.51066580478345,
+      "grad_norm": 1.2182319164276123,
+      "learning_rate": 0.001,
+      "loss": 1.0446,
+      "step": 94400
+    },
+    {
+      "epoch": 30.542986425339368,
+      "grad_norm": 3.014864206314087,
+      "learning_rate": 0.001,
+      "loss": 1.0268,
+      "step": 94500
+    },
+    {
+      "epoch": 30.575307045895283,
+      "grad_norm": 1.1657381057739258,
+      "learning_rate": 0.001,
+      "loss": 1.0381,
+      "step": 94600
+    },
+    {
+      "epoch": 30.607627666451197,
+      "grad_norm": 1.2702171802520752,
+      "learning_rate": 0.001,
+      "loss": 1.022,
+      "step": 94700
+    },
+    {
+      "epoch": 30.639948287007112,
+      "grad_norm": 1.3262053728103638,
+      "learning_rate": 0.001,
+      "loss": 1.0328,
+      "step": 94800
+    },
+    {
+      "epoch": 30.672268907563026,
+      "grad_norm": 1.1520822048187256,
+      "learning_rate": 0.001,
+      "loss": 1.0414,
+      "step": 94900
+    },
+    {
+      "epoch": 30.70458952811894,
+      "grad_norm": 1.637629747390747,
+      "learning_rate": 0.001,
+      "loss": 1.0357,
+      "step": 95000
+    },
+    {
+      "epoch": 30.736910148674855,
+      "grad_norm": 1.3296183347702026,
+      "learning_rate": 0.001,
+      "loss": 1.0383,
+      "step": 95100
+    },
+    {
+      "epoch": 30.76923076923077,
+      "grad_norm": 1.197227120399475,
+      "learning_rate": 0.001,
+      "loss": 1.0412,
+      "step": 95200
+    },
+    {
+      "epoch": 30.801551389786685,
+      "grad_norm": 1.2573148012161255,
+      "learning_rate": 0.001,
+      "loss": 1.0402,
+      "step": 95300
+    },
+    {
+      "epoch": 30.8338720103426,
+      "grad_norm": 1.1908835172653198,
+      "learning_rate": 0.001,
+      "loss": 1.0467,
+      "step": 95400
+    },
+    {
+      "epoch": 30.866192630898514,
+      "grad_norm": 1.181947946548462,
+      "learning_rate": 0.001,
+      "loss": 1.0488,
+      "step": 95500
+    },
+    {
+      "epoch": 30.89851325145443,
+      "grad_norm": 1.127081036567688,
+      "learning_rate": 0.001,
+      "loss": 1.028,
+      "step": 95600
+    },
+    {
+      "epoch": 30.930833872010343,
+      "grad_norm": 1.8955832719802856,
+      "learning_rate": 0.001,
+      "loss": 1.0532,
+      "step": 95700
+    },
+    {
+      "epoch": 30.963154492566257,
+      "grad_norm": 1.178209900856018,
+      "learning_rate": 0.001,
+      "loss": 1.0188,
+      "step": 95800
+    },
+    {
+      "epoch": 30.995475113122172,
+      "grad_norm": 1.3171616792678833,
+      "learning_rate": 0.001,
+      "loss": 1.019,
+      "step": 95900
+    },
+    {
+      "epoch": 31.027795733678087,
+      "grad_norm": 0.9879239201545715,
+      "learning_rate": 0.001,
+      "loss": 0.9763,
+      "step": 96000
+    },
+    {
+      "epoch": 31.060116354234,
+      "grad_norm": 0.9748243093490601,
+      "learning_rate": 0.001,
+      "loss": 0.9399,
+      "step": 96100
+    },
+    {
+      "epoch": 31.092436974789916,
+      "grad_norm": 1.102300763130188,
+      "learning_rate": 0.001,
+      "loss": 0.9548,
+      "step": 96200
+    },
+    {
+      "epoch": 31.12475759534583,
+      "grad_norm": 1.0693408250808716,
+      "learning_rate": 0.001,
+      "loss": 0.964,
+      "step": 96300
+    },
+    {
+      "epoch": 31.157078215901745,
+      "grad_norm": 1.0911110639572144,
+      "learning_rate": 0.001,
+      "loss": 0.9647,
+      "step": 96400
+    },
+    {
+      "epoch": 31.18939883645766,
+      "grad_norm": 1.2917101383209229,
+      "learning_rate": 0.001,
+      "loss": 0.9822,
+      "step": 96500
+    },
+    {
+      "epoch": 31.221719457013574,
+      "grad_norm": 1.0852818489074707,
+      "learning_rate": 0.001,
+      "loss": 0.9762,
+      "step": 96600
+    },
+    {
+      "epoch": 31.25404007756949,
+      "grad_norm": 1.5321288108825684,
+      "learning_rate": 0.001,
+      "loss": 0.9643,
+      "step": 96700
+    },
+    {
+      "epoch": 31.286360698125403,
+      "grad_norm": 1.1907495260238647,
+      "learning_rate": 0.001,
+      "loss": 0.9786,
+      "step": 96800
+    },
+    {
+      "epoch": 31.318681318681318,
+      "grad_norm": 1.236419439315796,
+      "learning_rate": 0.001,
+      "loss": 1.0093,
+      "step": 96900
+    },
+    {
+      "epoch": 31.351001939237232,
+      "grad_norm": 1.4393523931503296,
+      "learning_rate": 0.001,
+      "loss": 0.9868,
+      "step": 97000
+    },
+    {
+      "epoch": 31.383322559793147,
+      "grad_norm": 1.1615034341812134,
+      "learning_rate": 0.001,
+      "loss": 0.9932,
+      "step": 97100
+    },
+    {
+      "epoch": 31.41564318034906,
+      "grad_norm": 1.4453788995742798,
+      "learning_rate": 0.001,
+      "loss": 0.9853,
+      "step": 97200
+    },
+    {
+      "epoch": 31.447963800904976,
+      "grad_norm": 1.1026290655136108,
+      "learning_rate": 0.001,
+      "loss": 0.9995,
+      "step": 97300
+    },
+    {
+      "epoch": 31.48028442146089,
+      "grad_norm": 1.1974660158157349,
+      "learning_rate": 0.001,
+      "loss": 0.9895,
+      "step": 97400
+    },
+    {
+      "epoch": 31.51260504201681,
+      "grad_norm": 1.2898255586624146,
+      "learning_rate": 0.001,
+      "loss": 1.0166,
+      "step": 97500
+    },
+    {
+      "epoch": 31.544925662572723,
+      "grad_norm": 1.1666028499603271,
+      "learning_rate": 0.001,
+      "loss": 1.0055,
+      "step": 97600
+    },
+    {
+      "epoch": 31.577246283128638,
+      "grad_norm": 1.4688661098480225,
+      "learning_rate": 0.001,
+      "loss": 0.9944,
+      "step": 97700
+    },
+    {
+      "epoch": 31.609566903684552,
+      "grad_norm": 1.2777717113494873,
+      "learning_rate": 0.001,
+      "loss": 1.0156,
+      "step": 97800
+    },
+    {
+      "epoch": 31.641887524240467,
+      "grad_norm": 1.1962950229644775,
+      "learning_rate": 0.001,
+      "loss": 1.0111,
+      "step": 97900
+    },
+    {
+      "epoch": 31.67420814479638,
+      "grad_norm": 1.0366114377975464,
+      "learning_rate": 0.001,
+      "loss": 1.012,
+      "step": 98000
+    },
+    {
+      "epoch": 31.706528765352296,
+      "grad_norm": 1.0451583862304688,
+      "learning_rate": 0.001,
+      "loss": 1.0146,
+      "step": 98100
+    },
+    {
+      "epoch": 31.73884938590821,
+      "grad_norm": 1.020317554473877,
+      "learning_rate": 0.001,
+      "loss": 1.0263,
+      "step": 98200
+    },
+    {
+      "epoch": 31.771170006464125,
+      "grad_norm": 1.1409897804260254,
+      "learning_rate": 0.001,
+      "loss": 1.0167,
+      "step": 98300
+    },
+    {
+      "epoch": 31.80349062702004,
+      "grad_norm": 1.5148199796676636,
+      "learning_rate": 0.001,
+      "loss": 1.0359,
+      "step": 98400
+    },
+    {
+      "epoch": 31.835811247575954,
+      "grad_norm": 1.0931096076965332,
+      "learning_rate": 0.001,
+      "loss": 1.0176,
+      "step": 98500
+    },
+    {
+      "epoch": 31.86813186813187,
+      "grad_norm": 1.3822332620620728,
+      "learning_rate": 0.001,
+      "loss": 1.0394,
+      "step": 98600
+    },
+    {
+      "epoch": 31.900452488687783,
+      "grad_norm": 1.0372686386108398,
+      "learning_rate": 0.001,
+      "loss": 1.0125,
+      "step": 98700
+    },
+    {
+      "epoch": 31.932773109243698,
+      "grad_norm": 1.1177948713302612,
+      "learning_rate": 0.001,
+      "loss": 1.0355,
+      "step": 98800
+    },
+    {
+      "epoch": 31.965093729799612,
+      "grad_norm": 22.84385108947754,
+      "learning_rate": 0.001,
+      "loss": 1.0112,
+      "step": 98900
+    },
+    {
+      "epoch": 31.997414350355527,
+      "grad_norm": 1.431148648262024,
+      "learning_rate": 0.001,
+      "loss": 1.0157,
+      "step": 99000
+    },
+    {
+      "epoch": 32.02973497091144,
+      "grad_norm": 1.1543240547180176,
+      "learning_rate": 0.001,
+      "loss": 0.9505,
+      "step": 99100
+    },
+    {
+      "epoch": 32.062055591467356,
+      "grad_norm": 1.2091968059539795,
+      "learning_rate": 0.001,
+      "loss": 0.9292,
+      "step": 99200
+    },
+    {
+      "epoch": 32.09437621202327,
+      "grad_norm": 1.3215097188949585,
+      "learning_rate": 0.001,
+      "loss": 0.9619,
+      "step": 99300
+    },
+    {
+      "epoch": 32.126696832579185,
+      "grad_norm": 1.0096403360366821,
+      "learning_rate": 0.001,
+      "loss": 0.9312,
+      "step": 99400
+    },
+    {
+      "epoch": 32.1590174531351,
+      "grad_norm": 1.2567592859268188,
+      "learning_rate": 0.001,
+      "loss": 0.9507,
+      "step": 99500
+    },
+    {
+      "epoch": 32.191338073691014,
+      "grad_norm": 1.2970302104949951,
+      "learning_rate": 0.001,
+      "loss": 0.9646,
+      "step": 99600
+    },
+    {
+      "epoch": 32.22365869424693,
+      "grad_norm": 1.2226184606552124,
+      "learning_rate": 0.001,
+      "loss": 0.9644,
+      "step": 99700
+    },
+    {
+      "epoch": 32.25597931480284,
+      "grad_norm": 1.4690126180648804,
+      "learning_rate": 0.001,
+      "loss": 0.9591,
+      "step": 99800
+    },
+    {
+      "epoch": 32.28829993535876,
+      "grad_norm": 5.503747463226318,
+      "learning_rate": 0.001,
+      "loss": 0.9596,
+      "step": 99900
+    },
+    {
+      "epoch": 32.32062055591467,
+      "grad_norm": 1.5522727966308594,
+      "learning_rate": 0.001,
+      "loss": 0.9552,
+      "step": 100000
+    },
+    {
+      "epoch": 32.35294117647059,
+      "grad_norm": 1.183016061782837,
+      "learning_rate": 0.001,
+      "loss": 0.9799,
+      "step": 100100
+    },
+    {
+      "epoch": 32.3852617970265,
+      "grad_norm": 1.1440593004226685,
+      "learning_rate": 0.001,
+      "loss": 0.9745,
+      "step": 100200
+    },
+    {
+      "epoch": 32.417582417582416,
+      "grad_norm": 1.1371045112609863,
+      "learning_rate": 0.001,
+      "loss": 0.9723,
+      "step": 100300
+    },
+    {
+      "epoch": 32.44990303813833,
+      "grad_norm": 1.2485049962997437,
+      "learning_rate": 0.001,
+      "loss": 0.9701,
+      "step": 100400
+    },
+    {
+      "epoch": 32.482223658694245,
+      "grad_norm": 1.0270694494247437,
+      "learning_rate": 0.001,
+      "loss": 0.9945,
+      "step": 100500
+    },
+    {
+      "epoch": 32.51454427925016,
+      "grad_norm": 1.3093137741088867,
+      "learning_rate": 0.001,
+      "loss": 0.9781,
+      "step": 100600
+    },
+    {
+      "epoch": 32.546864899806074,
+      "grad_norm": 0.982565701007843,
+      "learning_rate": 0.001,
+      "loss": 0.9586,
+      "step": 100700
+    },
+    {
+      "epoch": 32.57918552036199,
+      "grad_norm": 1.180829644203186,
+      "learning_rate": 0.001,
+      "loss": 1.0096,
+      "step": 100800
+    },
+    {
+      "epoch": 32.6115061409179,
+      "grad_norm": 1.8596967458724976,
+      "learning_rate": 0.001,
+      "loss": 1.0101,
+      "step": 100900
+    },
+    {
+      "epoch": 32.64382676147382,
+      "grad_norm": 1.1227686405181885,
+      "learning_rate": 0.001,
+      "loss": 0.9968,
+      "step": 101000
+    },
+    {
+      "epoch": 32.67614738202973,
+      "grad_norm": 1.3988664150238037,
+      "learning_rate": 0.001,
+      "loss": 0.9934,
+      "step": 101100
+    },
+    {
+      "epoch": 32.70846800258565,
+      "grad_norm": 1.0867431163787842,
+      "learning_rate": 0.001,
+      "loss": 1.0071,
+      "step": 101200
+    },
+    {
+      "epoch": 32.74078862314156,
+      "grad_norm": 1.2575784921646118,
+      "learning_rate": 0.001,
+      "loss": 1.0145,
+      "step": 101300
+    },
+    {
+      "epoch": 32.773109243697476,
+      "grad_norm": 1.1998168230056763,
+      "learning_rate": 0.001,
+      "loss": 0.9904,
+      "step": 101400
+    },
+    {
+      "epoch": 32.80542986425339,
+      "grad_norm": 0.9470672011375427,
+      "learning_rate": 0.001,
+      "loss": 1.0114,
+      "step": 101500
+    },
+    {
+      "epoch": 32.837750484809305,
+      "grad_norm": 1.0151934623718262,
+      "learning_rate": 0.001,
+      "loss": 1.0152,
+      "step": 101600
+    },
+    {
+      "epoch": 32.87007110536522,
+      "grad_norm": 1.1765100955963135,
+      "learning_rate": 0.001,
+      "loss": 1.023,
+      "step": 101700
+    },
+    {
+      "epoch": 32.902391725921134,
+      "grad_norm": 1.2009962797164917,
+      "learning_rate": 0.001,
+      "loss": 0.9987,
+      "step": 101800
+    },
+    {
+      "epoch": 32.93471234647705,
+      "grad_norm": 1.1376994848251343,
+      "learning_rate": 0.001,
+      "loss": 0.9883,
+      "step": 101900
+    },
+    {
+      "epoch": 32.967032967032964,
+      "grad_norm": 0.9074110984802246,
+      "learning_rate": 0.001,
+      "loss": 1.0078,
+      "step": 102000
+    },
+    {
+      "epoch": 32.999353587588885,
+      "grad_norm": 0.8598896861076355,
+      "learning_rate": 0.001,
+      "loss": 0.9777,
+      "step": 102100
+    },
+    {
+      "epoch": 33.0316742081448,
+      "grad_norm": 0.7127558588981628,
+      "learning_rate": 0.001,
+      "loss": 0.9143,
+      "step": 102200
+    },
+    {
+      "epoch": 33.063994828700714,
+      "grad_norm": 0.7595661878585815,
+      "learning_rate": 0.001,
+      "loss": 0.9234,
+      "step": 102300
+    },
+    {
+      "epoch": 33.09631544925663,
+      "grad_norm": 0.8373937010765076,
+      "learning_rate": 0.001,
+      "loss": 0.9316,
+      "step": 102400
+    },
+    {
+      "epoch": 33.12863606981254,
+      "grad_norm": 1.197033166885376,
+      "learning_rate": 0.001,
+      "loss": 0.949,
+      "step": 102500
+    },
+    {
+      "epoch": 33.16095669036846,
+      "grad_norm": 0.6342483758926392,
+      "learning_rate": 0.001,
+      "loss": 0.9397,
+      "step": 102600
+    },
+    {
+      "epoch": 33.19327731092437,
+      "grad_norm": 1.198174238204956,
+      "learning_rate": 0.001,
+      "loss": 0.9458,
+      "step": 102700
+    },
+    {
+      "epoch": 33.22559793148029,
+      "grad_norm": 0.8614729642868042,
+      "learning_rate": 0.001,
+      "loss": 0.9565,
+      "step": 102800
+    },
+    {
+      "epoch": 33.2579185520362,
+      "grad_norm": 0.9009416699409485,
+      "learning_rate": 0.001,
+      "loss": 0.9174,
+      "step": 102900
+    },
+    {
+      "epoch": 33.290239172592116,
+      "grad_norm": 1.1099083423614502,
+      "learning_rate": 0.001,
+      "loss": 0.9521,
+      "step": 103000
+    },
+    {
+      "epoch": 33.32255979314803,
+      "grad_norm": 0.8563976883888245,
+      "learning_rate": 0.001,
+      "loss": 0.9295,
+      "step": 103100
+    },
+    {
+      "epoch": 33.354880413703945,
+      "grad_norm": 0.8638460040092468,
+      "learning_rate": 0.001,
+      "loss": 0.9511,
+      "step": 103200
+    },
+    {
+      "epoch": 33.38720103425986,
+      "grad_norm": 0.7926732301712036,
+      "learning_rate": 0.001,
+      "loss": 0.9582,
+      "step": 103300
+    },
+    {
+      "epoch": 33.419521654815775,
+      "grad_norm": 0.6091955900192261,
+      "learning_rate": 0.001,
+      "loss": 0.948,
+      "step": 103400
+    },
+    {
+      "epoch": 33.45184227537169,
+      "grad_norm": 1.9351041316986084,
+      "learning_rate": 0.001,
+      "loss": 0.9555,
+      "step": 103500
+    },
+    {
+      "epoch": 33.484162895927604,
+      "grad_norm": 0.9850326776504517,
+      "learning_rate": 0.001,
+      "loss": 0.945,
+      "step": 103600
+    },
+    {
+      "epoch": 33.51648351648352,
+      "grad_norm": 1.0357599258422852,
+      "learning_rate": 0.001,
+      "loss": 0.975,
+      "step": 103700
+    },
+    {
+      "epoch": 33.54880413703943,
+      "grad_norm": 0.6513636112213135,
+      "learning_rate": 0.001,
+      "loss": 0.9802,
+      "step": 103800
+    },
+    {
+      "epoch": 33.58112475759535,
+      "grad_norm": 1.0478070974349976,
+      "learning_rate": 0.001,
+      "loss": 0.9791,
+      "step": 103900
+    },
+    {
+      "epoch": 33.61344537815126,
+      "grad_norm": 0.8873677253723145,
+      "learning_rate": 0.001,
+      "loss": 0.975,
+      "step": 104000
+    },
+    {
+      "epoch": 33.645765998707176,
+      "grad_norm": 0.8857697248458862,
+      "learning_rate": 0.001,
+      "loss": 0.9868,
+      "step": 104100
+    },
+    {
+      "epoch": 33.67808661926309,
+      "grad_norm": 0.8290280699729919,
+      "learning_rate": 0.001,
+      "loss": 0.9874,
+      "step": 104200
+    },
+    {
+      "epoch": 33.710407239819006,
+      "grad_norm": 1.063020944595337,
+      "learning_rate": 0.001,
+      "loss": 0.9847,
+      "step": 104300
+    },
+    {
+      "epoch": 33.74272786037492,
+      "grad_norm": 0.8536366820335388,
+      "learning_rate": 0.001,
+      "loss": 0.9826,
+      "step": 104400
+    },
+    {
+      "epoch": 33.775048480930835,
+      "grad_norm": 1.7910808324813843,
+      "learning_rate": 0.001,
+      "loss": 1.0049,
+      "step": 104500
+    },
+    {
+      "epoch": 33.80736910148675,
+      "grad_norm": 0.9663560390472412,
+      "learning_rate": 0.001,
+      "loss": 0.9915,
+      "step": 104600
+    },
+    {
+      "epoch": 33.839689722042664,
+      "grad_norm": 0.8727264404296875,
+      "learning_rate": 0.001,
+      "loss": 1.0047,
+      "step": 104700
+    },
+    {
+      "epoch": 33.87201034259858,
+      "grad_norm": 1.2965424060821533,
+      "learning_rate": 0.001,
+      "loss": 0.9907,
+      "step": 104800
+    },
+    {
+      "epoch": 33.90433096315449,
+      "grad_norm": 1.0414011478424072,
+      "learning_rate": 0.001,
+      "loss": 1.0034,
+      "step": 104900
+    },
+    {
+      "epoch": 33.93665158371041,
+      "grad_norm": 0.7418168783187866,
+      "learning_rate": 0.001,
+      "loss": 0.9997,
+      "step": 105000
+    },
+    {
+      "epoch": 33.96897220426632,
+      "grad_norm": 0.7445783615112305,
+      "learning_rate": 0.001,
+      "loss": 0.9919,
+      "step": 105100
+    },
+    {
+      "epoch": 34.00129282482224,
+      "grad_norm": 1.2291535139083862,
+      "learning_rate": 0.001,
+      "loss": 0.9792,
+      "step": 105200
+    },
+    {
+      "epoch": 34.03361344537815,
+      "grad_norm": 1.6661193370819092,
+      "learning_rate": 0.001,
+      "loss": 0.9099,
+      "step": 105300
+    },
+    {
+      "epoch": 34.065934065934066,
+      "grad_norm": 1.3818449974060059,
+      "learning_rate": 0.001,
+      "loss": 0.9103,
+      "step": 105400
+    },
+    {
+      "epoch": 34.09825468648998,
+      "grad_norm": 1.2478028535842896,
+      "learning_rate": 0.001,
+      "loss": 0.9196,
+      "step": 105500
+    },
+    {
+      "epoch": 34.130575307045895,
+      "grad_norm": 1.4901819229125977,
+      "learning_rate": 0.001,
+      "loss": 0.9162,
+      "step": 105600
+    },
+    {
+      "epoch": 34.16289592760181,
+      "grad_norm": 1.3518427610397339,
+      "learning_rate": 0.001,
+      "loss": 0.9214,
+      "step": 105700
+    },
+    {
+      "epoch": 34.195216548157724,
+      "grad_norm": 1.1966758966445923,
+      "learning_rate": 0.001,
+      "loss": 0.921,
+      "step": 105800
+    },
+    {
+      "epoch": 34.22753716871364,
+      "grad_norm": 1.0546095371246338,
+      "learning_rate": 0.001,
+      "loss": 0.9203,
+      "step": 105900
+    },
+    {
+      "epoch": 34.25985778926955,
+      "grad_norm": 1.6791573762893677,
+      "learning_rate": 0.001,
+      "loss": 0.9263,
+      "step": 106000
+    },
+    {
+      "epoch": 34.29217840982547,
+      "grad_norm": 1.7650243043899536,
+      "learning_rate": 0.001,
+      "loss": 0.9263,
+      "step": 106100
+    },
+    {
+      "epoch": 34.32449903038138,
+      "grad_norm": 1.4087214469909668,
+      "learning_rate": 0.001,
+      "loss": 0.9372,
+      "step": 106200
+    },
+    {
+      "epoch": 34.3568196509373,
+      "grad_norm": 1.335076928138733,
+      "learning_rate": 0.001,
+      "loss": 0.9456,
+      "step": 106300
+    },
+    {
+      "epoch": 34.38914027149321,
+      "grad_norm": 1.5549242496490479,
+      "learning_rate": 0.001,
+      "loss": 0.9418,
+      "step": 106400
+    },
+    {
+      "epoch": 34.421460892049126,
+      "grad_norm": 2.1881766319274902,
+      "learning_rate": 0.001,
+      "loss": 0.9424,
+      "step": 106500
+    },
+    {
+      "epoch": 34.45378151260504,
+      "grad_norm": 1.419062614440918,
+      "learning_rate": 0.001,
+      "loss": 0.9557,
+      "step": 106600
+    },
+    {
+      "epoch": 34.486102133160955,
+      "grad_norm": 1.5608348846435547,
+      "learning_rate": 0.001,
+      "loss": 0.9554,
+      "step": 106700
+    },
+    {
+      "epoch": 34.51842275371687,
+      "grad_norm": 1.1787161827087402,
+      "learning_rate": 0.001,
+      "loss": 0.934,
+      "step": 106800
+    },
+    {
+      "epoch": 34.550743374272784,
+      "grad_norm": 1.2317980527877808,
+      "learning_rate": 0.001,
+      "loss": 0.9724,
+      "step": 106900
+    },
+    {
+      "epoch": 34.5830639948287,
+      "grad_norm": 1.6141093969345093,
+      "learning_rate": 0.001,
+      "loss": 0.9421,
+      "step": 107000
+    },
+    {
+      "epoch": 34.61538461538461,
+      "grad_norm": 1.3813108205795288,
+      "learning_rate": 0.001,
+      "loss": 0.9603,
+      "step": 107100
+    },
+    {
+      "epoch": 34.64770523594053,
+      "grad_norm": 1.9711265563964844,
+      "learning_rate": 0.001,
+      "loss": 0.9639,
+      "step": 107200
+    },
+    {
+      "epoch": 34.68002585649644,
+      "grad_norm": 1.3804035186767578,
+      "learning_rate": 0.001,
+      "loss": 0.9641,
+      "step": 107300
+    },
+    {
+      "epoch": 34.71234647705236,
+      "grad_norm": 1.8484543561935425,
+      "learning_rate": 0.001,
+      "loss": 0.9766,
+      "step": 107400
+    },
+    {
+      "epoch": 34.74466709760827,
+      "grad_norm": 1.755317211151123,
+      "learning_rate": 0.001,
+      "loss": 0.9875,
+      "step": 107500
+    },
+    {
+      "epoch": 34.776987718164186,
+      "grad_norm": 1.6666924953460693,
+      "learning_rate": 0.001,
+      "loss": 0.9852,
+      "step": 107600
+    },
+    {
+      "epoch": 34.8093083387201,
+      "grad_norm": 1.394085168838501,
+      "learning_rate": 0.001,
+      "loss": 0.9903,
+      "step": 107700
+    },
+    {
+      "epoch": 34.841628959276015,
+      "grad_norm": 1.8693289756774902,
+      "learning_rate": 0.001,
+      "loss": 0.9803,
+      "step": 107800
+    },
+    {
+      "epoch": 34.87394957983193,
+      "grad_norm": 1.5679796934127808,
+      "learning_rate": 0.001,
+      "loss": 0.989,
+      "step": 107900
+    },
+    {
+      "epoch": 34.906270200387844,
+      "grad_norm": 1.6920912265777588,
+      "learning_rate": 0.001,
+      "loss": 0.9789,
+      "step": 108000
+    },
+    {
+      "epoch": 34.93859082094376,
+      "grad_norm": 1.1828668117523193,
+      "learning_rate": 0.001,
+      "loss": 0.9783,
+      "step": 108100
+    },
+    {
+      "epoch": 34.97091144149967,
+      "grad_norm": 1.4724658727645874,
+      "learning_rate": 0.001,
+      "loss": 0.9847,
+      "step": 108200
+    },
+    {
+      "epoch": 35.003232062055595,
+      "grad_norm": 1.3865439891815186,
+      "learning_rate": 0.001,
+      "loss": 0.9699,
+      "step": 108300
+    },
+    {
+      "epoch": 35.03555268261151,
+      "grad_norm": 1.377656102180481,
+      "learning_rate": 0.001,
+      "loss": 0.8899,
+      "step": 108400
+    },
+    {
+      "epoch": 35.067873303167424,
+      "grad_norm": 1.2798742055892944,
+      "learning_rate": 0.001,
+      "loss": 0.9105,
+      "step": 108500
+    },
+    {
+      "epoch": 35.10019392372334,
+      "grad_norm": 1.2324934005737305,
+      "learning_rate": 0.001,
+      "loss": 0.8821,
+      "step": 108600
+    },
+    {
+      "epoch": 35.13251454427925,
+      "grad_norm": 1.3357598781585693,
+      "learning_rate": 0.001,
+      "loss": 0.8884,
+      "step": 108700
+    },
+    {
+      "epoch": 35.16483516483517,
+      "grad_norm": 1.5128265619277954,
+      "learning_rate": 0.001,
+      "loss": 0.9217,
+      "step": 108800
+    },
+    {
+      "epoch": 35.19715578539108,
+      "grad_norm": 1.328216314315796,
+      "learning_rate": 0.001,
+      "loss": 0.9182,
+      "step": 108900
+    },
+    {
+      "epoch": 35.229476405947,
+      "grad_norm": 1.1615173816680908,
+      "learning_rate": 0.001,
+      "loss": 0.8841,
+      "step": 109000
+    },
+    {
+      "epoch": 35.26179702650291,
+      "grad_norm": 1.154482364654541,
+      "learning_rate": 0.001,
+      "loss": 0.9154,
+      "step": 109100
+    },
+    {
+      "epoch": 35.294117647058826,
+      "grad_norm": 1.2462170124053955,
+      "learning_rate": 0.001,
+      "loss": 0.9134,
+      "step": 109200
+    },
+    {
+      "epoch": 35.32643826761474,
+      "grad_norm": 1.3430410623550415,
+      "learning_rate": 0.001,
+      "loss": 0.9442,
+      "step": 109300
+    },
+    {
+      "epoch": 35.358758888170655,
+      "grad_norm": 1.0239200592041016,
+      "learning_rate": 0.001,
+      "loss": 0.9379,
+      "step": 109400
+    },
+    {
+      "epoch": 35.39107950872657,
+      "grad_norm": 1.3149348497390747,
+      "learning_rate": 0.001,
+      "loss": 0.9362,
+      "step": 109500
+    },
+    {
+      "epoch": 35.423400129282484,
+      "grad_norm": 1.2154160737991333,
+      "learning_rate": 0.001,
+      "loss": 0.9421,
+      "step": 109600
+    },
+    {
+      "epoch": 35.4557207498384,
+      "grad_norm": 1.1552187204360962,
+      "learning_rate": 0.001,
+      "loss": 0.9314,
+      "step": 109700
+    },
+    {
+      "epoch": 35.48804137039431,
+      "grad_norm": 1.5669306516647339,
+      "learning_rate": 0.001,
+      "loss": 0.9438,
+      "step": 109800
+    },
+    {
+      "epoch": 35.52036199095023,
+      "grad_norm": 1.1329694986343384,
+      "learning_rate": 0.001,
+      "loss": 0.9255,
+      "step": 109900
+    },
+    {
+      "epoch": 35.55268261150614,
+      "grad_norm": 1.1270161867141724,
+      "learning_rate": 0.001,
+      "loss": 0.9267,
+      "step": 110000
+    },
+    {
+      "epoch": 35.58500323206206,
+      "grad_norm": 1.4306011199951172,
+      "learning_rate": 0.001,
+      "loss": 0.9465,
+      "step": 110100
+    },
+    {
+      "epoch": 35.61732385261797,
+      "grad_norm": 1.260940670967102,
+      "learning_rate": 0.001,
+      "loss": 0.9495,
+      "step": 110200
+    },
+    {
+      "epoch": 35.649644473173886,
+      "grad_norm": 1.591579794883728,
+      "learning_rate": 0.001,
+      "loss": 0.9466,
+      "step": 110300
+    },
+    {
+      "epoch": 35.6819650937298,
+      "grad_norm": 1.207124948501587,
+      "learning_rate": 0.001,
+      "loss": 0.9448,
+      "step": 110400
+    },
+    {
+      "epoch": 35.714285714285715,
+      "grad_norm": 1.2340648174285889,
+      "learning_rate": 0.001,
+      "loss": 0.9448,
+      "step": 110500
+    },
+    {
+      "epoch": 35.74660633484163,
+      "grad_norm": 1.2188169956207275,
+      "learning_rate": 0.001,
+      "loss": 0.962,
+      "step": 110600
+    },
+    {
+      "epoch": 35.778926955397544,
+      "grad_norm": 1.1369333267211914,
+      "learning_rate": 0.001,
+      "loss": 0.9493,
+      "step": 110700
+    },
+    {
+      "epoch": 35.81124757595346,
+      "grad_norm": 1.1521050930023193,
+      "learning_rate": 0.001,
+      "loss": 0.9591,
+      "step": 110800
+    },
+    {
+      "epoch": 35.84356819650937,
+      "grad_norm": 1.3265663385391235,
+      "learning_rate": 0.001,
+      "loss": 0.962,
+      "step": 110900
+    },
+    {
+      "epoch": 35.87588881706529,
+      "grad_norm": 1.1715490818023682,
+      "learning_rate": 0.001,
+      "loss": 0.9482,
+      "step": 111000
+    },
+    {
+      "epoch": 35.9082094376212,
+      "grad_norm": 1.5694321393966675,
+      "learning_rate": 0.001,
+      "loss": 0.9593,
+      "step": 111100
+    },
+    {
+      "epoch": 35.94053005817712,
+      "grad_norm": 1.4879381656646729,
+      "learning_rate": 0.001,
+      "loss": 0.9633,
+      "step": 111200
+    },
+    {
+      "epoch": 35.97285067873303,
+      "grad_norm": 1.109749674797058,
+      "learning_rate": 0.001,
+      "loss": 0.9738,
+      "step": 111300
+    },
+    {
+      "epoch": 36.005171299288946,
+      "grad_norm": 1.32253098487854,
+      "learning_rate": 0.001,
+      "loss": 0.974,
+      "step": 111400
+    },
+    {
+      "epoch": 36.03749191984486,
+      "grad_norm": 1.4396809339523315,
+      "learning_rate": 0.001,
+      "loss": 0.8579,
+      "step": 111500
+    },
+    {
+      "epoch": 36.069812540400775,
+      "grad_norm": 1.4428057670593262,
+      "learning_rate": 0.001,
+      "loss": 0.8854,
+      "step": 111600
+    },
+    {
+      "epoch": 36.10213316095669,
+      "grad_norm": 1.3725905418395996,
+      "learning_rate": 0.001,
+      "loss": 0.9104,
+      "step": 111700
+    },
+    {
+      "epoch": 36.134453781512605,
+      "grad_norm": 1.1569488048553467,
+      "learning_rate": 0.001,
+      "loss": 0.8795,
+      "step": 111800
+    },
+    {
+      "epoch": 36.16677440206852,
+      "grad_norm": 1.3119192123413086,
+      "learning_rate": 0.001,
+      "loss": 0.8881,
+      "step": 111900
+    },
+    {
+      "epoch": 36.199095022624434,
+      "grad_norm": 1.108713984489441,
+      "learning_rate": 0.001,
+      "loss": 0.8895,
+      "step": 112000
+    },
+    {
+      "epoch": 36.23141564318035,
+      "grad_norm": 1.3103444576263428,
+      "learning_rate": 0.001,
+      "loss": 0.884,
+      "step": 112100
+    },
+    {
+      "epoch": 36.26373626373626,
+      "grad_norm": 1.0280040502548218,
+      "learning_rate": 0.001,
+      "loss": 0.8962,
+      "step": 112200
+    },
+    {
+      "epoch": 36.29605688429218,
+      "grad_norm": 1.3514018058776855,
+      "learning_rate": 0.001,
+      "loss": 0.8936,
+      "step": 112300
+    },
+    {
+      "epoch": 36.32837750484809,
+      "grad_norm": 1.2298413515090942,
+      "learning_rate": 0.001,
+      "loss": 0.8838,
+      "step": 112400
+    },
+    {
+      "epoch": 36.36069812540401,
+      "grad_norm": 1.0672255754470825,
+      "learning_rate": 0.001,
+      "loss": 0.9085,
+      "step": 112500
+    },
+    {
+      "epoch": 36.39301874595992,
+      "grad_norm": 1.4790087938308716,
+      "learning_rate": 0.001,
+      "loss": 0.9262,
+      "step": 112600
+    },
+    {
+      "epoch": 36.425339366515836,
+      "grad_norm": 1.1427301168441772,
+      "learning_rate": 0.001,
+      "loss": 0.9084,
+      "step": 112700
+    },
+    {
+      "epoch": 36.45765998707175,
+      "grad_norm": 1.0541187524795532,
+      "learning_rate": 0.001,
+      "loss": 0.9191,
+      "step": 112800
+    },
+    {
+      "epoch": 36.489980607627665,
+      "grad_norm": 1.4663690328598022,
+      "learning_rate": 0.001,
+      "loss": 0.9199,
+      "step": 112900
+    },
+    {
+      "epoch": 36.52230122818358,
+      "grad_norm": 20.26365089416504,
+      "learning_rate": 0.001,
+      "loss": 0.943,
+      "step": 113000
+    },
+    {
+      "epoch": 36.554621848739494,
+      "grad_norm": 1.370599389076233,
+      "learning_rate": 0.001,
+      "loss": 0.9374,
+      "step": 113100
+    },
+    {
+      "epoch": 36.58694246929541,
+      "grad_norm": 1.2954767942428589,
+      "learning_rate": 0.001,
+      "loss": 0.9299,
+      "step": 113200
+    },
+    {
+      "epoch": 36.61926308985132,
+      "grad_norm": 1.1597362756729126,
+      "learning_rate": 0.001,
+      "loss": 0.9112,
+      "step": 113300
+    },
+    {
+      "epoch": 36.65158371040724,
+      "grad_norm": 1.1877658367156982,
+      "learning_rate": 0.001,
+      "loss": 0.9379,
+      "step": 113400
+    },
+    {
+      "epoch": 36.68390433096315,
+      "grad_norm": 1.0057965517044067,
+      "learning_rate": 0.001,
+      "loss": 0.9497,
+      "step": 113500
+    },
+    {
+      "epoch": 36.71622495151907,
+      "grad_norm": 1.238929033279419,
+      "learning_rate": 0.001,
+      "loss": 0.9472,
+      "step": 113600
+    },
+    {
+      "epoch": 36.74854557207498,
+      "grad_norm": 1.2580220699310303,
+      "learning_rate": 0.001,
+      "loss": 0.9394,
+      "step": 113700
+    },
+    {
+      "epoch": 36.780866192630896,
+      "grad_norm": 1.4125406742095947,
+      "learning_rate": 0.001,
+      "loss": 0.9507,
+      "step": 113800
+    },
+    {
+      "epoch": 36.81318681318681,
+      "grad_norm": 1.0683133602142334,
+      "learning_rate": 0.001,
+      "loss": 0.9365,
+      "step": 113900
+    },
+    {
+      "epoch": 36.845507433742725,
+      "grad_norm": 1.269522786140442,
+      "learning_rate": 0.001,
+      "loss": 0.9508,
+      "step": 114000
+    },
+    {
+      "epoch": 36.87782805429864,
+      "grad_norm": 1.5544761419296265,
+      "learning_rate": 0.001,
+      "loss": 0.9311,
+      "step": 114100
+    },
+    {
+      "epoch": 36.910148674854554,
+      "grad_norm": 1.7500112056732178,
+      "learning_rate": 0.001,
+      "loss": 0.9612,
+      "step": 114200
+    },
+    {
+      "epoch": 36.94246929541047,
+      "grad_norm": 1.395186185836792,
+      "learning_rate": 0.001,
+      "loss": 0.9531,
+      "step": 114300
+    },
+    {
+      "epoch": 36.97478991596638,
+      "grad_norm": 1.5248706340789795,
+      "learning_rate": 0.001,
+      "loss": 0.9398,
+      "step": 114400
+    },
+    {
+      "epoch": 37.007110536522305,
+      "grad_norm": 1.1058337688446045,
+      "learning_rate": 0.001,
+      "loss": 0.9588,
+      "step": 114500
+    },
+    {
+      "epoch": 37.03943115707822,
+      "grad_norm": 1.2561908960342407,
+      "learning_rate": 0.001,
+      "loss": 0.8687,
+      "step": 114600
+    },
+    {
+      "epoch": 37.071751777634134,
+      "grad_norm": 1.2578023672103882,
+      "learning_rate": 0.001,
+      "loss": 0.8513,
+      "step": 114700
+    },
+    {
+      "epoch": 37.10407239819005,
+      "grad_norm": 1.7073860168457031,
+      "learning_rate": 0.001,
+      "loss": 0.8689,
+      "step": 114800
+    },
+    {
+      "epoch": 37.13639301874596,
+      "grad_norm": 1.2335535287857056,
+      "learning_rate": 0.001,
+      "loss": 0.8638,
+      "step": 114900
+    },
+    {
+      "epoch": 37.16871363930188,
+      "grad_norm": 1.389021396636963,
+      "learning_rate": 0.001,
+      "loss": 0.8734,
+      "step": 115000
+    },
+    {
+      "epoch": 37.20103425985779,
+      "grad_norm": 1.0842416286468506,
+      "learning_rate": 0.001,
+      "loss": 0.8801,
+      "step": 115100
+    },
+    {
+      "epoch": 37.23335488041371,
+      "grad_norm": 1.200750708580017,
+      "learning_rate": 0.001,
+      "loss": 0.8774,
+      "step": 115200
+    },
+    {
+      "epoch": 37.26567550096962,
+      "grad_norm": 1.2029190063476562,
+      "learning_rate": 0.001,
+      "loss": 0.893,
+      "step": 115300
+    },
+    {
+      "epoch": 37.297996121525536,
+      "grad_norm": 1.0285815000534058,
+      "learning_rate": 0.001,
+      "loss": 0.9137,
+      "step": 115400
+    },
+    {
+      "epoch": 37.33031674208145,
+      "grad_norm": 1.4431654214859009,
+      "learning_rate": 0.001,
+      "loss": 0.9109,
+      "step": 115500
+    },
+    {
+      "epoch": 37.362637362637365,
+      "grad_norm": 1.1851094961166382,
+      "learning_rate": 0.001,
+      "loss": 0.8865,
+      "step": 115600
+    },
+    {
+      "epoch": 37.39495798319328,
+      "grad_norm": 1.4423298835754395,
+      "learning_rate": 0.001,
+      "loss": 0.9152,
+      "step": 115700
+    },
+    {
+      "epoch": 37.427278603749194,
+      "grad_norm": 1.3550646305084229,
+      "learning_rate": 0.001,
+      "loss": 0.9074,
+      "step": 115800
+    },
+    {
+      "epoch": 37.45959922430511,
+      "grad_norm": 1.1404973268508911,
+      "learning_rate": 0.001,
+      "loss": 0.9146,
+      "step": 115900
+    },
+    {
+      "epoch": 37.49191984486102,
+      "grad_norm": 1.348008632659912,
+      "learning_rate": 0.001,
+      "loss": 0.9181,
+      "step": 116000
+    },
+    {
+      "epoch": 37.52424046541694,
+      "grad_norm": 1.3652067184448242,
+      "learning_rate": 0.001,
+      "loss": 0.9212,
+      "step": 116100
+    },
+    {
+      "epoch": 37.55656108597285,
+      "grad_norm": 1.0638148784637451,
+      "learning_rate": 0.001,
+      "loss": 0.9181,
+      "step": 116200
+    },
+    {
+      "epoch": 37.58888170652877,
+      "grad_norm": 1.0108212232589722,
+      "learning_rate": 0.001,
+      "loss": 0.9068,
+      "step": 116300
+    },
+    {
+      "epoch": 37.62120232708468,
+      "grad_norm": 1.4408948421478271,
+      "learning_rate": 0.001,
+      "loss": 0.9113,
+      "step": 116400
+    },
+    {
+      "epoch": 37.653522947640596,
+      "grad_norm": 1.2225804328918457,
+      "learning_rate": 0.001,
+      "loss": 0.9123,
+      "step": 116500
+    },
+    {
+      "epoch": 37.68584356819651,
+      "grad_norm": 2.1255993843078613,
+      "learning_rate": 0.001,
+      "loss": 0.9122,
+      "step": 116600
+    },
+    {
+      "epoch": 37.718164188752425,
+      "grad_norm": 1.3163739442825317,
+      "learning_rate": 0.001,
+      "loss": 0.9171,
+      "step": 116700
+    },
+    {
+      "epoch": 37.75048480930834,
+      "grad_norm": 1.3287923336029053,
+      "learning_rate": 0.001,
+      "loss": 0.9176,
+      "step": 116800
+    },
+    {
+      "epoch": 37.782805429864254,
+      "grad_norm": 1.0648704767227173,
+      "learning_rate": 0.001,
+      "loss": 0.9277,
+      "step": 116900
+    },
+    {
+      "epoch": 37.81512605042017,
+      "grad_norm": 1.2207541465759277,
+      "learning_rate": 0.001,
+      "loss": 0.9486,
+      "step": 117000
+    },
+    {
+      "epoch": 37.84744667097608,
+      "grad_norm": 1.2167179584503174,
+      "learning_rate": 0.001,
+      "loss": 0.9341,
+      "step": 117100
+    },
+    {
+      "epoch": 37.879767291532,
+      "grad_norm": 1.1062008142471313,
+      "learning_rate": 0.001,
+      "loss": 0.9155,
+      "step": 117200
+    },
+    {
+      "epoch": 37.91208791208791,
+      "grad_norm": 1.135310173034668,
+      "learning_rate": 0.001,
+      "loss": 0.9465,
+      "step": 117300
+    },
+    {
+      "epoch": 37.94440853264383,
+      "grad_norm": 1.182563304901123,
+      "learning_rate": 0.001,
+      "loss": 0.9311,
+      "step": 117400
+    },
+    {
+      "epoch": 37.97672915319974,
+      "grad_norm": 1.289273977279663,
+      "learning_rate": 0.001,
+      "loss": 0.9395,
+      "step": 117500
+    },
+    {
+      "epoch": 38.009049773755656,
+      "grad_norm": 2.1775407791137695,
+      "learning_rate": 0.001,
+      "loss": 0.912,
+      "step": 117600
+    },
+    {
+      "epoch": 38.04137039431157,
+      "grad_norm": 1.0216083526611328,
+      "learning_rate": 0.001,
+      "loss": 0.8395,
+      "step": 117700
+    },
+    {
+      "epoch": 38.073691014867485,
+      "grad_norm": 1.5370399951934814,
+      "learning_rate": 0.001,
+      "loss": 0.8599,
+      "step": 117800
+    },
+    {
+      "epoch": 38.1060116354234,
+      "grad_norm": 1.2524052858352661,
+      "learning_rate": 0.001,
+      "loss": 0.8611,
+      "step": 117900
+    },
+    {
+      "epoch": 38.138332255979314,
+      "grad_norm": 1.2202684879302979,
+      "learning_rate": 0.001,
+      "loss": 0.8669,
+      "step": 118000
+    },
+    {
+      "epoch": 38.17065287653523,
+      "grad_norm": 1.1983212232589722,
+      "learning_rate": 0.001,
+      "loss": 0.859,
+      "step": 118100
+    },
+    {
+      "epoch": 38.20297349709114,
+      "grad_norm": 3.6025729179382324,
+      "learning_rate": 0.001,
+      "loss": 0.8675,
+      "step": 118200
+    },
+    {
+      "epoch": 38.23529411764706,
+      "grad_norm": 1.2796443700790405,
+      "learning_rate": 0.001,
+      "loss": 0.8688,
+      "step": 118300
+    },
+    {
+      "epoch": 38.26761473820297,
+      "grad_norm": 1.3514477014541626,
+      "learning_rate": 0.001,
+      "loss": 0.8642,
+      "step": 118400
+    },
+    {
+      "epoch": 38.29993535875889,
+      "grad_norm": 1.2543727159500122,
+      "learning_rate": 0.001,
+      "loss": 0.892,
+      "step": 118500
+    },
+    {
+      "epoch": 38.3322559793148,
+      "grad_norm": 2.0149078369140625,
+      "learning_rate": 0.001,
+      "loss": 0.875,
+      "step": 118600
+    },
+    {
+      "epoch": 38.364576599870716,
+      "grad_norm": 1.5488308668136597,
+      "learning_rate": 0.001,
+      "loss": 0.9035,
+      "step": 118700
+    },
+    {
+      "epoch": 38.39689722042663,
+      "grad_norm": 1.1685779094696045,
+      "learning_rate": 0.001,
+      "loss": 0.8959,
+      "step": 118800
+    },
+    {
+      "epoch": 38.429217840982545,
+      "grad_norm": 1.3017526865005493,
+      "learning_rate": 0.001,
+      "loss": 0.8843,
+      "step": 118900
+    },
+    {
+      "epoch": 38.46153846153846,
+      "grad_norm": 1.4337364435195923,
+      "learning_rate": 0.001,
+      "loss": 0.8842,
+      "step": 119000
+    },
+    {
+      "epoch": 38.493859082094374,
+      "grad_norm": 1.4739630222320557,
+      "learning_rate": 0.001,
+      "loss": 0.8781,
+      "step": 119100
+    },
+    {
+      "epoch": 38.52617970265029,
+      "grad_norm": 1.5166560411453247,
+      "learning_rate": 0.001,
+      "loss": 0.9092,
+      "step": 119200
+    },
+    {
+      "epoch": 38.558500323206204,
+      "grad_norm": 1.242997407913208,
+      "learning_rate": 0.001,
+      "loss": 0.8854,
+      "step": 119300
+    },
+    {
+      "epoch": 38.59082094376212,
+      "grad_norm": 1.5462573766708374,
+      "learning_rate": 0.001,
+      "loss": 0.9008,
+      "step": 119400
+    },
+    {
+      "epoch": 38.62314156431803,
+      "grad_norm": 1.318841576576233,
+      "learning_rate": 0.001,
+      "loss": 0.9186,
+      "step": 119500
+    },
+    {
+      "epoch": 38.65546218487395,
+      "grad_norm": 1.2831882238388062,
+      "learning_rate": 0.001,
+      "loss": 0.9134,
+      "step": 119600
+    },
+    {
+      "epoch": 38.68778280542986,
+      "grad_norm": 1.7237813472747803,
+      "learning_rate": 0.001,
+      "loss": 0.9139,
+      "step": 119700
+    },
+    {
+      "epoch": 38.720103425985776,
+      "grad_norm": 1.2951987981796265,
+      "learning_rate": 0.001,
+      "loss": 0.9191,
+      "step": 119800
+    },
+    {
+      "epoch": 38.75242404654169,
+      "grad_norm": 1.2112561464309692,
+      "learning_rate": 0.001,
+      "loss": 0.8893,
+      "step": 119900
+    },
+    {
+      "epoch": 38.784744667097605,
+      "grad_norm": 1.510880708694458,
+      "learning_rate": 0.001,
+      "loss": 0.9243,
+      "step": 120000
+    },
+    {
+      "epoch": 38.81706528765352,
+      "grad_norm": 1.351163387298584,
+      "learning_rate": 0.001,
+      "loss": 0.9163,
+      "step": 120100
+    },
+    {
+      "epoch": 38.849385908209435,
+      "grad_norm": 1.3247283697128296,
+      "learning_rate": 0.001,
+      "loss": 0.9262,
+      "step": 120200
+    },
+    {
+      "epoch": 38.88170652876535,
+      "grad_norm": 1.0785566568374634,
+      "learning_rate": 0.001,
+      "loss": 0.9076,
+      "step": 120300
+    },
+    {
+      "epoch": 38.914027149321264,
+      "grad_norm": 1.0955142974853516,
+      "learning_rate": 0.001,
+      "loss": 0.9009,
+      "step": 120400
+    },
+    {
+      "epoch": 38.94634776987718,
+      "grad_norm": 1.0944831371307373,
+      "learning_rate": 0.001,
+      "loss": 0.9169,
+      "step": 120500
+    },
+    {
+      "epoch": 38.97866839043309,
+      "grad_norm": 2.2278664112091064,
+      "learning_rate": 0.001,
+      "loss": 0.9378,
+      "step": 120600
+    },
+    {
+      "epoch": 39.010989010989015,
+      "grad_norm": 1.430474042892456,
+      "learning_rate": 0.001,
+      "loss": 0.8951,
+      "step": 120700
+    },
+    {
+      "epoch": 39.04330963154493,
+      "grad_norm": 1.299006462097168,
+      "learning_rate": 0.001,
+      "loss": 0.82,
+      "step": 120800
+    },
+    {
+      "epoch": 39.075630252100844,
+      "grad_norm": 1.3967899084091187,
+      "learning_rate": 0.001,
+      "loss": 0.8329,
+      "step": 120900
+    },
+    {
+      "epoch": 39.10795087265676,
+      "grad_norm": 1.8231886625289917,
+      "learning_rate": 0.001,
+      "loss": 0.8339,
+      "step": 121000
+    },
+    {
+      "epoch": 39.14027149321267,
+      "grad_norm": 2.1227855682373047,
+      "learning_rate": 0.001,
+      "loss": 0.8417,
+      "step": 121100
+    },
+    {
+      "epoch": 39.17259211376859,
+      "grad_norm": 1.2603408098220825,
+      "learning_rate": 0.001,
+      "loss": 0.8658,
+      "step": 121200
+    },
+    {
+      "epoch": 39.2049127343245,
+      "grad_norm": 1.3786275386810303,
+      "learning_rate": 0.001,
+      "loss": 0.8835,
+      "step": 121300
+    },
+    {
+      "epoch": 39.237233354880416,
+      "grad_norm": 1.3917183876037598,
+      "learning_rate": 0.001,
+      "loss": 0.8726,
+      "step": 121400
+    },
+    {
+      "epoch": 39.26955397543633,
+      "grad_norm": 1.4549587965011597,
+      "learning_rate": 0.001,
+      "loss": 0.8656,
+      "step": 121500
+    },
+    {
+      "epoch": 39.301874595992246,
+      "grad_norm": 1.4152559041976929,
+      "learning_rate": 0.001,
+      "loss": 0.8748,
+      "step": 121600
+    },
+    {
+      "epoch": 39.33419521654816,
+      "grad_norm": 2.8188393115997314,
+      "learning_rate": 0.001,
+      "loss": 0.8734,
+      "step": 121700
+    },
+    {
+      "epoch": 39.366515837104075,
+      "grad_norm": 1.3342585563659668,
+      "learning_rate": 0.001,
+      "loss": 0.847,
+      "step": 121800
+    },
+    {
+      "epoch": 39.39883645765999,
+      "grad_norm": 1.2291045188903809,
+      "learning_rate": 0.001,
+      "loss": 0.8779,
+      "step": 121900
+    },
+    {
+      "epoch": 39.431157078215904,
+      "grad_norm": 1.130303978919983,
+      "learning_rate": 0.001,
+      "loss": 0.8561,
+      "step": 122000
+    },
+    {
+      "epoch": 39.46347769877182,
+      "grad_norm": 1.1709375381469727,
+      "learning_rate": 0.001,
+      "loss": 0.8792,
+      "step": 122100
+    },
+    {
+      "epoch": 39.49579831932773,
+      "grad_norm": 1.5080801248550415,
+      "learning_rate": 0.001,
+      "loss": 0.8826,
+      "step": 122200
+    },
+    {
+      "epoch": 39.52811893988365,
+      "grad_norm": 1.1423321962356567,
+      "learning_rate": 0.001,
+      "loss": 0.9035,
+      "step": 122300
+    },
+    {
+      "epoch": 39.56043956043956,
+      "grad_norm": 1.049439549446106,
+      "learning_rate": 0.001,
+      "loss": 0.8805,
+      "step": 122400
+    },
+    {
+      "epoch": 39.59276018099548,
+      "grad_norm": 1.460243821144104,
+      "learning_rate": 0.001,
+      "loss": 0.9,
+      "step": 122500
+    },
+    {
+      "epoch": 39.62508080155139,
+      "grad_norm": 1.9269883632659912,
+      "learning_rate": 0.001,
+      "loss": 0.9058,
+      "step": 122600
+    },
+    {
+      "epoch": 39.657401422107306,
+      "grad_norm": 1.2040832042694092,
+      "learning_rate": 0.001,
+      "loss": 0.8709,
+      "step": 122700
+    },
+    {
+      "epoch": 39.68972204266322,
+      "grad_norm": 1.3963993787765503,
+      "learning_rate": 0.001,
+      "loss": 0.9073,
+      "step": 122800
+    },
+    {
+      "epoch": 39.722042663219135,
+      "grad_norm": 1.2941703796386719,
+      "learning_rate": 0.001,
+      "loss": 0.8873,
+      "step": 122900
+    },
+    {
+      "epoch": 39.75436328377505,
+      "grad_norm": 1.2239911556243896,
+      "learning_rate": 0.001,
+      "loss": 0.8998,
+      "step": 123000
+    },
+    {
+      "epoch": 39.786683904330964,
+      "grad_norm": 1.0870931148529053,
+      "learning_rate": 0.001,
+      "loss": 0.8863,
+      "step": 123100
+    },
+    {
+      "epoch": 39.81900452488688,
+      "grad_norm": 1.0407912731170654,
+      "learning_rate": 0.001,
+      "loss": 0.8919,
+      "step": 123200
+    },
+    {
+      "epoch": 39.85132514544279,
+      "grad_norm": 1.204813838005066,
+      "learning_rate": 0.001,
+      "loss": 0.8963,
+      "step": 123300
+    },
+    {
+      "epoch": 39.88364576599871,
+      "grad_norm": 1.4278241395950317,
+      "learning_rate": 0.001,
+      "loss": 0.9084,
+      "step": 123400
+    },
+    {
+      "epoch": 39.91596638655462,
+      "grad_norm": 1.1825990676879883,
+      "learning_rate": 0.001,
+      "loss": 0.9032,
+      "step": 123500
+    },
+    {
+      "epoch": 39.94828700711054,
+      "grad_norm": 1.0813344717025757,
+      "learning_rate": 0.001,
+      "loss": 0.9231,
+      "step": 123600
+    },
+    {
+      "epoch": 39.98060762766645,
+      "grad_norm": 1.409714937210083,
+      "learning_rate": 0.001,
+      "loss": 0.9177,
+      "step": 123700
+    },
+    {
+      "epoch": 40.012928248222366,
+      "grad_norm": 1.474061131477356,
+      "learning_rate": 0.001,
+      "loss": 0.8616,
+      "step": 123800
+    },
+    {
+      "epoch": 40.04524886877828,
+      "grad_norm": 1.4782670736312866,
+      "learning_rate": 0.001,
+      "loss": 0.8286,
+      "step": 123900
+    },
+    {
+      "epoch": 40.077569489334195,
+      "grad_norm": 1.4716808795928955,
+      "learning_rate": 0.001,
+      "loss": 0.8311,
+      "step": 124000
+    },
+    {
+      "epoch": 40.10989010989011,
+      "grad_norm": 1.3071645498275757,
+      "learning_rate": 0.001,
+      "loss": 0.8232,
+      "step": 124100
+    },
+    {
+      "epoch": 40.142210730446024,
+      "grad_norm": 1.4670727252960205,
+      "learning_rate": 0.001,
+      "loss": 0.8244,
+      "step": 124200
+    },
+    {
+      "epoch": 40.17453135100194,
+      "grad_norm": 1.10783851146698,
+      "learning_rate": 0.001,
+      "loss": 0.8399,
+      "step": 124300
+    },
+    {
+      "epoch": 40.20685197155785,
+      "grad_norm": 1.0578640699386597,
+      "learning_rate": 0.001,
+      "loss": 0.8455,
+      "step": 124400
+    },
+    {
+      "epoch": 40.23917259211377,
+      "grad_norm": 1.236038327217102,
+      "learning_rate": 0.001,
+      "loss": 0.842,
+      "step": 124500
+    },
+    {
+      "epoch": 40.27149321266968,
+      "grad_norm": 1.1498534679412842,
+      "learning_rate": 0.001,
+      "loss": 0.8523,
+      "step": 124600
+    },
+    {
+      "epoch": 40.3038138332256,
+      "grad_norm": 1.199738621711731,
+      "learning_rate": 0.001,
+      "loss": 0.8457,
+      "step": 124700
+    },
+    {
+      "epoch": 40.33613445378151,
+      "grad_norm": 1.2088783979415894,
+      "learning_rate": 0.001,
+      "loss": 0.8572,
+      "step": 124800
+    },
+    {
+      "epoch": 40.368455074337426,
+      "grad_norm": 1.3873686790466309,
+      "learning_rate": 0.001,
+      "loss": 0.8561,
+      "step": 124900
+    },
+    {
+      "epoch": 40.40077569489334,
+      "grad_norm": 1.2533921003341675,
+      "learning_rate": 0.001,
+      "loss": 0.8655,
+      "step": 125000
+    },
+    {
+      "epoch": 40.433096315449255,
+      "grad_norm": 1.8301284313201904,
+      "learning_rate": 0.001,
+      "loss": 0.8748,
+      "step": 125100
+    },
+    {
+      "epoch": 40.46541693600517,
+      "grad_norm": 1.1949044466018677,
+      "learning_rate": 0.001,
+      "loss": 0.859,
+      "step": 125200
+    },
+    {
+      "epoch": 40.497737556561084,
+      "grad_norm": 2.156513214111328,
+      "learning_rate": 0.001,
+      "loss": 0.8567,
+      "step": 125300
+    },
+    {
+      "epoch": 40.530058177117,
+      "grad_norm": 1.2791069746017456,
+      "learning_rate": 0.001,
+      "loss": 0.8792,
+      "step": 125400
+    },
+    {
+      "epoch": 40.56237879767291,
+      "grad_norm": 1.3833438158035278,
+      "learning_rate": 0.001,
+      "loss": 0.8753,
+      "step": 125500
+    },
+    {
+      "epoch": 40.59469941822883,
+      "grad_norm": 1.301124095916748,
+      "learning_rate": 0.001,
+      "loss": 0.8718,
+      "step": 125600
+    },
+    {
+      "epoch": 40.62702003878474,
+      "grad_norm": 1.559802770614624,
+      "learning_rate": 0.001,
+      "loss": 0.8723,
+      "step": 125700
+    },
+    {
+      "epoch": 40.65934065934066,
+      "grad_norm": 1.1438794136047363,
+      "learning_rate": 0.001,
+      "loss": 0.8669,
+      "step": 125800
+    },
+    {
+      "epoch": 40.69166127989657,
+      "grad_norm": 1.6619253158569336,
+      "learning_rate": 0.001,
+      "loss": 0.8853,
+      "step": 125900
+    },
+    {
+      "epoch": 40.723981900452486,
+      "grad_norm": 1.3953354358673096,
+      "learning_rate": 0.001,
+      "loss": 0.8997,
+      "step": 126000
+    },
+    {
+      "epoch": 40.7563025210084,
+      "grad_norm": 1.3383327722549438,
+      "learning_rate": 0.001,
+      "loss": 0.8962,
+      "step": 126100
+    },
+    {
+      "epoch": 40.788623141564315,
+      "grad_norm": 1.0945528745651245,
+      "learning_rate": 0.001,
+      "loss": 0.8861,
+      "step": 126200
+    },
+    {
+      "epoch": 40.82094376212023,
+      "grad_norm": 1.4700912237167358,
+      "learning_rate": 0.001,
+      "loss": 0.899,
+      "step": 126300
+    },
+    {
+      "epoch": 40.853264382676144,
+      "grad_norm": 1.1272400617599487,
+      "learning_rate": 0.001,
+      "loss": 0.9029,
+      "step": 126400
+    },
+    {
+      "epoch": 40.88558500323206,
+      "grad_norm": 1.222775936126709,
+      "learning_rate": 0.001,
+      "loss": 0.8966,
+      "step": 126500
+    },
+    {
+      "epoch": 40.91790562378797,
+      "grad_norm": 1.9689314365386963,
+      "learning_rate": 0.001,
+      "loss": 0.894,
+      "step": 126600
+    },
+    {
+      "epoch": 40.95022624434389,
+      "grad_norm": 1.0816248655319214,
+      "learning_rate": 0.001,
+      "loss": 0.8955,
+      "step": 126700
+    },
+    {
+      "epoch": 40.9825468648998,
+      "grad_norm": 7.644120216369629,
+      "learning_rate": 0.001,
+      "loss": 0.8911,
+      "step": 126800
+    },
+    {
+      "epoch": 41.014867485455724,
+      "grad_norm": 1.3644909858703613,
+      "learning_rate": 0.001,
+      "loss": 0.8324,
+      "step": 126900
+    },
+    {
+      "epoch": 41.04718810601164,
+      "grad_norm": 1.2451497316360474,
+      "learning_rate": 0.001,
+      "loss": 0.8106,
+      "step": 127000
+    },
+    {
+      "epoch": 41.07950872656755,
+      "grad_norm": 2.127196788787842,
+      "learning_rate": 0.001,
+      "loss": 0.8169,
+      "step": 127100
+    },
+    {
+      "epoch": 41.11182934712347,
+      "grad_norm": 1.1432043313980103,
+      "learning_rate": 0.001,
+      "loss": 0.8184,
+      "step": 127200
+    },
+    {
+      "epoch": 41.14414996767938,
+      "grad_norm": 1.2829303741455078,
+      "learning_rate": 0.001,
+      "loss": 0.8124,
+      "step": 127300
+    },
+    {
+      "epoch": 41.1764705882353,
+      "grad_norm": 0.8562968373298645,
+      "learning_rate": 0.001,
+      "loss": 0.8184,
+      "step": 127400
+    },
+    {
+      "epoch": 41.20879120879121,
+      "grad_norm": 0.9684380888938904,
+      "learning_rate": 0.001,
+      "loss": 0.828,
+      "step": 127500
+    },
+    {
+      "epoch": 41.241111829347126,
+      "grad_norm": 1.0784997940063477,
+      "learning_rate": 0.001,
+      "loss": 0.836,
+      "step": 127600
+    },
+    {
+      "epoch": 41.27343244990304,
+      "grad_norm": 1.046213984489441,
+      "learning_rate": 0.001,
+      "loss": 0.844,
+      "step": 127700
+    },
+    {
+      "epoch": 41.305753070458955,
+      "grad_norm": 0.8843748569488525,
+      "learning_rate": 0.001,
+      "loss": 0.84,
+      "step": 127800
+    },
+    {
+      "epoch": 41.33807369101487,
+      "grad_norm": 0.7278721332550049,
+      "learning_rate": 0.001,
+      "loss": 0.8598,
+      "step": 127900
+    },
+    {
+      "epoch": 41.370394311570784,
+      "grad_norm": 0.9991313815116882,
+      "learning_rate": 0.001,
+      "loss": 0.8365,
+      "step": 128000
+    },
+    {
+      "epoch": 41.4027149321267,
+      "grad_norm": 0.8424844145774841,
+      "learning_rate": 0.001,
+      "loss": 0.8411,
+      "step": 128100
+    },
+    {
+      "epoch": 41.43503555268261,
+      "grad_norm": 1.4285770654678345,
+      "learning_rate": 0.001,
+      "loss": 0.8432,
+      "step": 128200
+    },
+    {
+      "epoch": 41.46735617323853,
+      "grad_norm": 1.2143181562423706,
+      "learning_rate": 0.001,
+      "loss": 0.8543,
+      "step": 128300
+    },
+    {
+      "epoch": 41.49967679379444,
+      "grad_norm": 1.2977038621902466,
+      "learning_rate": 0.001,
+      "loss": 0.8526,
+      "step": 128400
+    },
+    {
+      "epoch": 41.53199741435036,
+      "grad_norm": 1.0228110551834106,
+      "learning_rate": 0.001,
+      "loss": 0.8542,
+      "step": 128500
+    },
+    {
+      "epoch": 41.56431803490627,
+      "grad_norm": 1.2492506504058838,
+      "learning_rate": 0.001,
+      "loss": 0.8704,
+      "step": 128600
+    },
+    {
+      "epoch": 41.596638655462186,
+      "grad_norm": 1.5715211629867554,
+      "learning_rate": 0.001,
+      "loss": 0.8777,
+      "step": 128700
+    },
+    {
+      "epoch": 41.6289592760181,
+      "grad_norm": 1.181208610534668,
+      "learning_rate": 0.001,
+      "loss": 0.8637,
+      "step": 128800
+    },
+    {
+      "epoch": 41.661279896574015,
+      "grad_norm": 0.986804187297821,
+      "learning_rate": 0.001,
+      "loss": 0.8951,
+      "step": 128900
+    },
+    {
+      "epoch": 41.69360051712993,
+      "grad_norm": 1.0209100246429443,
+      "learning_rate": 0.001,
+      "loss": 0.8649,
+      "step": 129000
+    },
+    {
+      "epoch": 41.725921137685845,
+      "grad_norm": 1.1363537311553955,
+      "learning_rate": 0.001,
+      "loss": 0.8669,
+      "step": 129100
+    },
+    {
+      "epoch": 41.75824175824176,
+      "grad_norm": 1.1821092367172241,
+      "learning_rate": 0.001,
+      "loss": 0.8749,
+      "step": 129200
+    },
+    {
+      "epoch": 41.790562378797674,
+      "grad_norm": 0.9846389293670654,
+      "learning_rate": 0.001,
+      "loss": 0.8755,
+      "step": 129300
+    },
+    {
+      "epoch": 41.82288299935359,
+      "grad_norm": 1.3120102882385254,
+      "learning_rate": 0.001,
+      "loss": 0.8847,
+      "step": 129400
+    },
+    {
+      "epoch": 41.8552036199095,
+      "grad_norm": 1.2054803371429443,
+      "learning_rate": 0.001,
+      "loss": 0.8704,
+      "step": 129500
+    },
+    {
+      "epoch": 41.88752424046542,
+      "grad_norm": 1.1038309335708618,
+      "learning_rate": 0.001,
+      "loss": 0.8747,
+      "step": 129600
+    },
+    {
+      "epoch": 41.91984486102133,
+      "grad_norm": 3.2323157787323,
+      "learning_rate": 0.001,
+      "loss": 0.8803,
+      "step": 129700
+    },
+    {
+      "epoch": 41.95216548157725,
+      "grad_norm": 1.0762603282928467,
+      "learning_rate": 0.001,
+      "loss": 0.8818,
+      "step": 129800
+    },
+    {
+      "epoch": 41.98448610213316,
+      "grad_norm": 1.0352551937103271,
+      "learning_rate": 0.001,
+      "loss": 0.8977,
+      "step": 129900
+    },
+    {
+      "epoch": 42.016806722689076,
+      "grad_norm": 1.9383422136306763,
+      "learning_rate": 0.001,
+      "loss": 0.8207,
+      "step": 130000
+    },
+    {
+      "epoch": 42.04912734324499,
+      "grad_norm": 1.456878900527954,
+      "learning_rate": 0.001,
+      "loss": 0.7909,
+      "step": 130100
+    },
+    {
+      "epoch": 42.081447963800905,
+      "grad_norm": 2.3285322189331055,
+      "learning_rate": 0.001,
+      "loss": 0.8194,
+      "step": 130200
+    },
+    {
+      "epoch": 42.11376858435682,
+      "grad_norm": 1.8079019784927368,
+      "learning_rate": 0.001,
+      "loss": 0.8073,
+      "step": 130300
+    },
+    {
+      "epoch": 42.146089204912734,
+      "grad_norm": 1.3452057838439941,
+      "learning_rate": 0.001,
+      "loss": 0.7993,
+      "step": 130400
+    },
+    {
+      "epoch": 42.17840982546865,
+      "grad_norm": 1.915332555770874,
+      "learning_rate": 0.001,
+      "loss": 0.8246,
+      "step": 130500
+    },
+    {
+      "epoch": 42.21073044602456,
+      "grad_norm": 1.9111815690994263,
+      "learning_rate": 0.001,
+      "loss": 0.8088,
+      "step": 130600
+    },
+    {
+      "epoch": 42.24305106658048,
+      "grad_norm": 1.762703537940979,
+      "learning_rate": 0.001,
+      "loss": 0.8243,
+      "step": 130700
+    },
+    {
+      "epoch": 42.27537168713639,
+      "grad_norm": 1.3808751106262207,
+      "learning_rate": 0.001,
+      "loss": 0.8224,
+      "step": 130800
+    },
+    {
+      "epoch": 42.30769230769231,
+      "grad_norm": 1.7224359512329102,
+      "learning_rate": 0.001,
+      "loss": 0.8173,
+      "step": 130900
+    },
+    {
+      "epoch": 42.34001292824822,
+      "grad_norm": 2.0048210620880127,
+      "learning_rate": 0.001,
+      "loss": 0.833,
+      "step": 131000
+    },
+    {
+      "epoch": 42.372333548804136,
+      "grad_norm": 1.6992894411087036,
+      "learning_rate": 0.001,
+      "loss": 0.8386,
+      "step": 131100
+    },
+    {
+      "epoch": 42.40465416936005,
+      "grad_norm": 1.8275269269943237,
+      "learning_rate": 0.001,
+      "loss": 0.8428,
+      "step": 131200
+    },
+    {
+      "epoch": 42.436974789915965,
+      "grad_norm": 2.066981554031372,
+      "learning_rate": 0.001,
+      "loss": 0.818,
+      "step": 131300
+    },
+    {
+      "epoch": 42.46929541047188,
+      "grad_norm": 1.7945083379745483,
+      "learning_rate": 0.001,
+      "loss": 0.8342,
+      "step": 131400
+    },
+    {
+      "epoch": 42.501616031027794,
+      "grad_norm": 2.0542707443237305,
+      "learning_rate": 0.001,
+      "loss": 0.8374,
+      "step": 131500
+    },
+    {
+      "epoch": 42.53393665158371,
+      "grad_norm": 1.5740739107131958,
+      "learning_rate": 0.001,
+      "loss": 0.8487,
+      "step": 131600
+    },
+    {
+      "epoch": 42.56625727213962,
+      "grad_norm": 2.1072022914886475,
+      "learning_rate": 0.001,
+      "loss": 0.8606,
+      "step": 131700
+    },
+    {
+      "epoch": 42.59857789269554,
+      "grad_norm": 2.105353593826294,
+      "learning_rate": 0.001,
+      "loss": 0.8614,
+      "step": 131800
+    },
+    {
+      "epoch": 42.63089851325145,
+      "grad_norm": 1.8877348899841309,
+      "learning_rate": 0.001,
+      "loss": 0.8343,
+      "step": 131900
+    },
+    {
+      "epoch": 42.66321913380737,
+      "grad_norm": 1.740787148475647,
+      "learning_rate": 0.001,
+      "loss": 0.8556,
+      "step": 132000
+    },
+    {
+      "epoch": 42.69553975436328,
+      "grad_norm": 1.4763180017471313,
+      "learning_rate": 0.001,
+      "loss": 0.8485,
+      "step": 132100
+    },
+    {
+      "epoch": 42.727860374919196,
+      "grad_norm": 1.9027022123336792,
+      "learning_rate": 0.001,
+      "loss": 0.8436,
+      "step": 132200
+    },
+    {
+      "epoch": 42.76018099547511,
+      "grad_norm": 1.3983405828475952,
+      "learning_rate": 0.001,
+      "loss": 0.8714,
+      "step": 132300
+    },
+    {
+      "epoch": 42.792501616031025,
+      "grad_norm": 2.2473886013031006,
+      "learning_rate": 0.001,
+      "loss": 0.8717,
+      "step": 132400
+    },
+    {
+      "epoch": 42.82482223658694,
+      "grad_norm": 1.3791003227233887,
+      "learning_rate": 0.001,
+      "loss": 0.8692,
+      "step": 132500
+    },
+    {
+      "epoch": 42.857142857142854,
+      "grad_norm": 1.7160332202911377,
+      "learning_rate": 0.001,
+      "loss": 0.8712,
+      "step": 132600
+    },
+    {
+      "epoch": 42.88946347769877,
+      "grad_norm": 1.4543373584747314,
+      "learning_rate": 0.001,
+      "loss": 0.8719,
+      "step": 132700
+    },
+    {
+      "epoch": 42.92178409825468,
+      "grad_norm": 1.5968679189682007,
+      "learning_rate": 0.001,
+      "loss": 0.8879,
+      "step": 132800
+    },
+    {
+      "epoch": 42.9541047188106,
+      "grad_norm": 1.9997758865356445,
+      "learning_rate": 0.001,
+      "loss": 0.8759,
+      "step": 132900
+    },
+    {
+      "epoch": 42.98642533936652,
+      "grad_norm": 1.7506520748138428,
+      "learning_rate": 0.001,
+      "loss": 0.8855,
+      "step": 133000
+    },
+    {
+      "epoch": 43.018745959922434,
+      "grad_norm": 1.527773380279541,
+      "learning_rate": 0.001,
+      "loss": 0.8467,
+      "step": 133100
+    },
+    {
+      "epoch": 43.05106658047835,
+      "grad_norm": 1.4628387689590454,
+      "learning_rate": 0.001,
+      "loss": 0.7821,
+      "step": 133200
+    },
+    {
+      "epoch": 43.08338720103426,
+      "grad_norm": 1.3667227029800415,
+      "learning_rate": 0.001,
+      "loss": 0.7919,
+      "step": 133300
+    },
+    {
+      "epoch": 43.11570782159018,
+      "grad_norm": 1.2891846895217896,
+      "learning_rate": 0.001,
+      "loss": 0.8055,
+      "step": 133400
+    },
+    {
+      "epoch": 43.14802844214609,
+      "grad_norm": 1.349412441253662,
+      "learning_rate": 0.001,
+      "loss": 0.7955,
+      "step": 133500
+    },
+    {
+      "epoch": 43.18034906270201,
+      "grad_norm": 1.5655378103256226,
+      "learning_rate": 0.001,
+      "loss": 0.7913,
+      "step": 133600
+    },
+    {
+      "epoch": 43.21266968325792,
+      "grad_norm": 1.305897831916809,
+      "learning_rate": 0.001,
+      "loss": 0.8004,
+      "step": 133700
+    },
+    {
+      "epoch": 43.244990303813836,
+      "grad_norm": 1.294546365737915,
+      "learning_rate": 0.001,
+      "loss": 0.8053,
+      "step": 133800
+    },
+    {
+      "epoch": 43.27731092436975,
+      "grad_norm": 1.4070929288864136,
+      "learning_rate": 0.001,
+      "loss": 0.8101,
+      "step": 133900
+    },
+    {
+      "epoch": 43.309631544925665,
+      "grad_norm": 1.0786323547363281,
+      "learning_rate": 0.001,
+      "loss": 0.805,
+      "step": 134000
+    },
+    {
+      "epoch": 43.34195216548158,
+      "grad_norm": 1.484237551689148,
+      "learning_rate": 0.001,
+      "loss": 0.8018,
+      "step": 134100
+    },
+    {
+      "epoch": 43.374272786037494,
+      "grad_norm": 1.4975357055664062,
+      "learning_rate": 0.001,
+      "loss": 0.8563,
+      "step": 134200
+    },
+    {
+      "epoch": 43.40659340659341,
+      "grad_norm": 1.1751554012298584,
+      "learning_rate": 0.001,
+      "loss": 0.8259,
+      "step": 134300
+    },
+    {
+      "epoch": 43.43891402714932,
+      "grad_norm": 1.3102048635482788,
+      "learning_rate": 0.001,
+      "loss": 0.796,
+      "step": 134400
+    },
+    {
+      "epoch": 43.47123464770524,
+      "grad_norm": 1.3700244426727295,
+      "learning_rate": 0.001,
+      "loss": 0.8382,
+      "step": 134500
+    },
+    {
+      "epoch": 43.50355526826115,
+      "grad_norm": 1.8419667482376099,
+      "learning_rate": 0.001,
+      "loss": 0.8244,
+      "step": 134600
+    },
+    {
+      "epoch": 43.53587588881707,
+      "grad_norm": 1.279415249824524,
+      "learning_rate": 0.001,
+      "loss": 0.8414,
+      "step": 134700
+    },
+    {
+      "epoch": 43.56819650937298,
+      "grad_norm": 1.1568162441253662,
+      "learning_rate": 0.001,
+      "loss": 0.8424,
+      "step": 134800
+    },
+    {
+      "epoch": 43.600517129928896,
+      "grad_norm": 1.2953382730484009,
+      "learning_rate": 0.001,
+      "loss": 0.8303,
+      "step": 134900
+    },
+    {
+      "epoch": 43.63283775048481,
+      "grad_norm": 1.2967185974121094,
+      "learning_rate": 0.001,
+      "loss": 0.8394,
+      "step": 135000
+    },
+    {
+      "epoch": 43.665158371040725,
+      "grad_norm": 1.3641395568847656,
+      "learning_rate": 0.001,
+      "loss": 0.8408,
+      "step": 135100
+    },
+    {
+      "epoch": 43.69747899159664,
+      "grad_norm": 1.2159711122512817,
+      "learning_rate": 0.001,
+      "loss": 0.8462,
+      "step": 135200
+    },
+    {
+      "epoch": 43.729799612152554,
+      "grad_norm": 1.5351017713546753,
+      "learning_rate": 0.001,
+      "loss": 0.8459,
+      "step": 135300
+    },
+    {
+      "epoch": 43.76212023270847,
+      "grad_norm": 1.5198405981063843,
+      "learning_rate": 0.001,
+      "loss": 0.8499,
+      "step": 135400
+    },
+    {
+      "epoch": 43.79444085326438,
+      "grad_norm": 1.2137715816497803,
+      "learning_rate": 0.001,
+      "loss": 0.8526,
+      "step": 135500
+    },
+    {
+      "epoch": 43.8267614738203,
+      "grad_norm": 1.2001396417617798,
+      "learning_rate": 0.001,
+      "loss": 0.8715,
+      "step": 135600
+    },
+    {
+      "epoch": 43.85908209437621,
+      "grad_norm": 1.7482579946517944,
+      "learning_rate": 0.001,
+      "loss": 0.8561,
+      "step": 135700
+    },
+    {
+      "epoch": 43.89140271493213,
+      "grad_norm": 1.476863980293274,
+      "learning_rate": 0.001,
+      "loss": 0.86,
+      "step": 135800
+    },
+    {
+      "epoch": 43.92372333548804,
+      "grad_norm": 1.6021820306777954,
+      "learning_rate": 0.001,
+      "loss": 0.881,
+      "step": 135900
+    },
+    {
+      "epoch": 43.956043956043956,
+      "grad_norm": 1.5128792524337769,
+      "learning_rate": 0.001,
+      "loss": 0.873,
+      "step": 136000
+    },
+    {
+      "epoch": 43.98836457659987,
+      "grad_norm": 1.4070549011230469,
+      "learning_rate": 0.001,
+      "loss": 0.8732,
+      "step": 136100
+    },
+    {
+      "epoch": 44.020685197155785,
+      "grad_norm": 1.2981127500534058,
+      "learning_rate": 0.001,
+      "loss": 0.8268,
+      "step": 136200
+    },
+    {
+      "epoch": 44.0530058177117,
+      "grad_norm": 1.4790092706680298,
+      "learning_rate": 0.001,
+      "loss": 0.7825,
+      "step": 136300
+    },
+    {
+      "epoch": 44.085326438267614,
+      "grad_norm": 1.3770111799240112,
+      "learning_rate": 0.001,
+      "loss": 0.7824,
+      "step": 136400
+    },
+    {
+      "epoch": 44.11764705882353,
+      "grad_norm": 9.711091041564941,
+      "learning_rate": 0.001,
+      "loss": 0.7852,
+      "step": 136500
+    },
+    {
+      "epoch": 44.14996767937944,
+      "grad_norm": 1.4254145622253418,
+      "learning_rate": 0.001,
+      "loss": 0.7864,
+      "step": 136600
+    },
+    {
+      "epoch": 44.18228829993536,
+      "grad_norm": 1.2787805795669556,
+      "learning_rate": 0.001,
+      "loss": 0.79,
+      "step": 136700
+    },
+    {
+      "epoch": 44.21460892049127,
+      "grad_norm": 1.3041934967041016,
+      "learning_rate": 0.001,
+      "loss": 0.8038,
+      "step": 136800
+    },
+    {
+      "epoch": 44.24692954104719,
+      "grad_norm": 1.2634992599487305,
+      "learning_rate": 0.001,
+      "loss": 0.8016,
+      "step": 136900
+    },
+    {
+      "epoch": 44.2792501616031,
+      "grad_norm": 1.237308144569397,
+      "learning_rate": 0.001,
+      "loss": 0.8099,
+      "step": 137000
+    },
+    {
+      "epoch": 44.311570782159016,
+      "grad_norm": 1.2396942377090454,
+      "learning_rate": 0.001,
+      "loss": 0.8169,
+      "step": 137100
+    },
+    {
+      "epoch": 44.34389140271493,
+      "grad_norm": 1.2848567962646484,
+      "learning_rate": 0.001,
+      "loss": 0.8059,
+      "step": 137200
+    },
+    {
+      "epoch": 44.376212023270845,
+      "grad_norm": 1.5477381944656372,
+      "learning_rate": 0.001,
+      "loss": 0.8163,
+      "step": 137300
+    },
+    {
+      "epoch": 44.40853264382676,
+      "grad_norm": 1.2960271835327148,
+      "learning_rate": 0.001,
+      "loss": 0.8032,
+      "step": 137400
+    },
+    {
+      "epoch": 44.440853264382675,
+      "grad_norm": 1.273160696029663,
+      "learning_rate": 0.001,
+      "loss": 0.8081,
+      "step": 137500
+    },
+    {
+      "epoch": 44.47317388493859,
+      "grad_norm": 1.191059947013855,
+      "learning_rate": 0.001,
+      "loss": 0.8189,
+      "step": 137600
+    },
+    {
+      "epoch": 44.505494505494504,
+      "grad_norm": 1.1510519981384277,
+      "learning_rate": 0.001,
+      "loss": 0.807,
+      "step": 137700
+    },
+    {
+      "epoch": 44.53781512605042,
+      "grad_norm": 1.358547329902649,
+      "learning_rate": 0.001,
+      "loss": 0.8193,
+      "step": 137800
+    },
+    {
+      "epoch": 44.57013574660633,
+      "grad_norm": 1.4065147638320923,
+      "learning_rate": 0.001,
+      "loss": 0.8166,
+      "step": 137900
+    },
+    {
+      "epoch": 44.60245636716225,
+      "grad_norm": 1.5476964712142944,
+      "learning_rate": 0.001,
+      "loss": 0.8333,
+      "step": 138000
+    },
+    {
+      "epoch": 44.63477698771816,
+      "grad_norm": 1.3824409246444702,
+      "learning_rate": 0.001,
+      "loss": 0.8273,
+      "step": 138100
+    },
+    {
+      "epoch": 44.66709760827408,
+      "grad_norm": 1.1956334114074707,
+      "learning_rate": 0.001,
+      "loss": 0.8241,
+      "step": 138200
+    },
+    {
+      "epoch": 44.69941822882999,
+      "grad_norm": 1.1970595121383667,
+      "learning_rate": 0.001,
+      "loss": 0.8477,
+      "step": 138300
+    },
+    {
+      "epoch": 44.731738849385906,
+      "grad_norm": 1.162197470664978,
+      "learning_rate": 0.001,
+      "loss": 0.8377,
+      "step": 138400
+    },
+    {
+      "epoch": 44.76405946994182,
+      "grad_norm": 1.2741224765777588,
+      "learning_rate": 0.001,
+      "loss": 0.8509,
+      "step": 138500
+    },
+    {
+      "epoch": 44.796380090497735,
+      "grad_norm": 1.1326532363891602,
+      "learning_rate": 0.001,
+      "loss": 0.8413,
+      "step": 138600
+    },
+    {
+      "epoch": 44.82870071105365,
+      "grad_norm": 1.5357139110565186,
+      "learning_rate": 0.001,
+      "loss": 0.8304,
+      "step": 138700
+    },
+    {
+      "epoch": 44.861021331609564,
+      "grad_norm": 1.355280876159668,
+      "learning_rate": 0.001,
+      "loss": 0.8572,
+      "step": 138800
+    },
+    {
+      "epoch": 44.89334195216548,
+      "grad_norm": 1.6483911275863647,
+      "learning_rate": 0.001,
+      "loss": 0.8489,
+      "step": 138900
+    },
+    {
+      "epoch": 44.92566257272139,
+      "grad_norm": 1.3355098962783813,
+      "learning_rate": 0.001,
+      "loss": 0.8559,
+      "step": 139000
+    },
+    {
+      "epoch": 44.95798319327731,
+      "grad_norm": 1.3055047988891602,
+      "learning_rate": 0.001,
+      "loss": 0.8465,
+      "step": 139100
+    },
+    {
+      "epoch": 44.99030381383322,
+      "grad_norm": 1.4414533376693726,
+      "learning_rate": 0.001,
+      "loss": 0.8137,
+      "step": 139200
+    },
+    {
+      "epoch": 45.022624434389144,
+      "grad_norm": 1.346257209777832,
+      "learning_rate": 0.001,
+      "loss": 0.8097,
+      "step": 139300
+    },
+    {
+      "epoch": 45.05494505494506,
+      "grad_norm": 1.8192956447601318,
+      "learning_rate": 0.001,
+      "loss": 0.7631,
+      "step": 139400
+    },
+    {
+      "epoch": 45.08726567550097,
+      "grad_norm": 1.2960363626480103,
+      "learning_rate": 0.001,
+      "loss": 0.7741,
+      "step": 139500
+    },
+    {
+      "epoch": 45.11958629605689,
+      "grad_norm": 2.104506731033325,
+      "learning_rate": 0.001,
+      "loss": 0.7705,
+      "step": 139600
+    },
+    {
+      "epoch": 45.1519069166128,
+      "grad_norm": 1.1508853435516357,
+      "learning_rate": 0.001,
+      "loss": 0.7728,
+      "step": 139700
+    },
+    {
+      "epoch": 45.18422753716872,
+      "grad_norm": 1.65923011302948,
+      "learning_rate": 0.001,
+      "loss": 0.7694,
+      "step": 139800
+    },
+    {
+      "epoch": 45.21654815772463,
+      "grad_norm": 1.5621492862701416,
+      "learning_rate": 0.001,
+      "loss": 0.7782,
+      "step": 139900
+    },
+    {
+      "epoch": 45.248868778280546,
+      "grad_norm": 1.388525366783142,
+      "learning_rate": 0.001,
+      "loss": 0.7728,
+      "step": 140000
+    },
+    {
+      "epoch": 45.28118939883646,
+      "grad_norm": 1.4508206844329834,
+      "learning_rate": 0.001,
+      "loss": 0.7853,
+      "step": 140100
+    },
+    {
+      "epoch": 45.313510019392375,
+      "grad_norm": 1.1052316427230835,
+      "learning_rate": 0.001,
+      "loss": 0.7995,
+      "step": 140200
+    },
+    {
+      "epoch": 45.34583063994829,
+      "grad_norm": 1.0586109161376953,
+      "learning_rate": 0.001,
+      "loss": 0.7858,
+      "step": 140300
+    },
+    {
+      "epoch": 45.378151260504204,
+      "grad_norm": 1.1809642314910889,
+      "learning_rate": 0.001,
+      "loss": 0.7982,
+      "step": 140400
+    },
+    {
+      "epoch": 45.41047188106012,
+      "grad_norm": 1.301119327545166,
+      "learning_rate": 0.001,
+      "loss": 0.7995,
+      "step": 140500
+    },
+    {
+      "epoch": 45.44279250161603,
+      "grad_norm": 1.2782262563705444,
+      "learning_rate": 0.001,
+      "loss": 0.8058,
+      "step": 140600
+    },
+    {
+      "epoch": 45.47511312217195,
+      "grad_norm": 1.2796794176101685,
+      "learning_rate": 0.001,
+      "loss": 0.7985,
+      "step": 140700
+    },
+    {
+      "epoch": 45.50743374272786,
+      "grad_norm": 1.1781928539276123,
+      "learning_rate": 0.001,
+      "loss": 0.7911,
+      "step": 140800
+    },
+    {
+      "epoch": 45.53975436328378,
+      "grad_norm": 1.1269559860229492,
+      "learning_rate": 0.001,
+      "loss": 0.8117,
+      "step": 140900
+    },
+    {
+      "epoch": 45.57207498383969,
+      "grad_norm": 1.3490080833435059,
+      "learning_rate": 0.001,
+      "loss": 0.793,
+      "step": 141000
+    },
+    {
+      "epoch": 45.604395604395606,
+      "grad_norm": 1.1441093683242798,
+      "learning_rate": 0.001,
+      "loss": 0.8239,
+      "step": 141100
+    },
+    {
+      "epoch": 45.63671622495152,
+      "grad_norm": 1.4738757610321045,
+      "learning_rate": 0.001,
+      "loss": 0.8282,
+      "step": 141200
+    },
+    {
+      "epoch": 45.669036845507435,
+      "grad_norm": 1.5204062461853027,
+      "learning_rate": 0.001,
+      "loss": 0.8204,
+      "step": 141300
+    },
+    {
+      "epoch": 45.70135746606335,
+      "grad_norm": 1.3744752407073975,
+      "learning_rate": 0.001,
+      "loss": 0.8146,
+      "step": 141400
+    },
+    {
+      "epoch": 45.733678086619264,
+      "grad_norm": 1.4345048666000366,
+      "learning_rate": 0.001,
+      "loss": 0.8251,
+      "step": 141500
+    },
+    {
+      "epoch": 45.76599870717518,
+      "grad_norm": 1.3285446166992188,
+      "learning_rate": 0.001,
+      "loss": 0.8376,
+      "step": 141600
+    },
+    {
+      "epoch": 45.79831932773109,
+      "grad_norm": 1.2864034175872803,
+      "learning_rate": 0.001,
+      "loss": 0.84,
+      "step": 141700
+    },
+    {
+      "epoch": 45.83063994828701,
+      "grad_norm": 1.5362489223480225,
+      "learning_rate": 0.001,
+      "loss": 0.8346,
+      "step": 141800
+    },
+    {
+      "epoch": 45.86296056884292,
+      "grad_norm": 1.177847981452942,
+      "learning_rate": 0.001,
+      "loss": 0.8402,
+      "step": 141900
+    },
+    {
+      "epoch": 45.89528118939884,
+      "grad_norm": 1.168405294418335,
+      "learning_rate": 0.001,
+      "loss": 0.8333,
+      "step": 142000
+    },
+    {
+      "epoch": 45.92760180995475,
+      "grad_norm": 1.4306167364120483,
+      "learning_rate": 0.001,
+      "loss": 0.8347,
+      "step": 142100
+    },
+    {
+      "epoch": 45.959922430510666,
+      "grad_norm": 1.3686728477478027,
+      "learning_rate": 0.001,
+      "loss": 0.8238,
+      "step": 142200
+    },
+    {
+      "epoch": 45.99224305106658,
+      "grad_norm": 1.2857394218444824,
+      "learning_rate": 0.001,
+      "loss": 0.8655,
+      "step": 142300
+    },
+    {
+      "epoch": 46.024563671622495,
+      "grad_norm": 1.2607849836349487,
+      "learning_rate": 0.001,
+      "loss": 0.7783,
+      "step": 142400
+    },
+    {
+      "epoch": 46.05688429217841,
+      "grad_norm": 1.5633095502853394,
+      "learning_rate": 0.001,
+      "loss": 0.7446,
+      "step": 142500
+    },
+    {
+      "epoch": 46.089204912734324,
+      "grad_norm": 4.660933494567871,
+      "learning_rate": 0.001,
+      "loss": 0.7425,
+      "step": 142600
+    },
+    {
+      "epoch": 46.12152553329024,
+      "grad_norm": 1.7380452156066895,
+      "learning_rate": 0.001,
+      "loss": 0.7674,
+      "step": 142700
+    },
+    {
+      "epoch": 46.15384615384615,
+      "grad_norm": 1.2123302221298218,
+      "learning_rate": 0.001,
+      "loss": 0.775,
+      "step": 142800
+    },
+    {
+      "epoch": 46.18616677440207,
+      "grad_norm": 1.2319421768188477,
+      "learning_rate": 0.001,
+      "loss": 0.7753,
+      "step": 142900
+    },
+    {
+      "epoch": 46.21848739495798,
+      "grad_norm": 1.5815069675445557,
+      "learning_rate": 0.001,
+      "loss": 0.7775,
+      "step": 143000
+    },
+    {
+      "epoch": 46.2508080155139,
+      "grad_norm": 1.0354673862457275,
+      "learning_rate": 0.001,
+      "loss": 0.7743,
+      "step": 143100
+    },
+    {
+      "epoch": 46.28312863606981,
+      "grad_norm": 1.3230652809143066,
+      "learning_rate": 0.001,
+      "loss": 0.7778,
+      "step": 143200
+    },
+    {
+      "epoch": 46.315449256625726,
+      "grad_norm": 1.2702628374099731,
+      "learning_rate": 0.001,
+      "loss": 0.7708,
+      "step": 143300
+    },
+    {
+      "epoch": 46.34776987718164,
+      "grad_norm": 1.2347975969314575,
+      "learning_rate": 0.001,
+      "loss": 0.7792,
+      "step": 143400
+    },
+    {
+      "epoch": 46.380090497737555,
+      "grad_norm": 1.1771498918533325,
+      "learning_rate": 0.001,
+      "loss": 0.7826,
+      "step": 143500
+    },
+    {
+      "epoch": 46.41241111829347,
+      "grad_norm": 1.1414570808410645,
+      "learning_rate": 0.001,
+      "loss": 0.7848,
+      "step": 143600
+    },
+    {
+      "epoch": 46.444731738849384,
+      "grad_norm": 1.1657198667526245,
+      "learning_rate": 0.001,
+      "loss": 0.7975,
+      "step": 143700
+    },
+    {
+      "epoch": 46.4770523594053,
+      "grad_norm": 1.267530083656311,
+      "learning_rate": 0.001,
+      "loss": 0.8029,
+      "step": 143800
+    },
+    {
+      "epoch": 46.50937297996121,
+      "grad_norm": 1.1892708539962769,
+      "learning_rate": 0.001,
+      "loss": 0.798,
+      "step": 143900
+    },
+    {
+      "epoch": 46.54169360051713,
+      "grad_norm": 1.89950692653656,
+      "learning_rate": 0.001,
+      "loss": 0.7923,
+      "step": 144000
+    },
+    {
+      "epoch": 46.57401422107304,
+      "grad_norm": 1.305152177810669,
+      "learning_rate": 0.001,
+      "loss": 0.8014,
+      "step": 144100
+    },
+    {
+      "epoch": 46.60633484162896,
+      "grad_norm": 1.5587385892868042,
+      "learning_rate": 0.001,
+      "loss": 0.8099,
+      "step": 144200
+    },
+    {
+      "epoch": 46.63865546218487,
+      "grad_norm": 1.152372121810913,
+      "learning_rate": 0.001,
+      "loss": 0.8062,
+      "step": 144300
+    },
+    {
+      "epoch": 46.670976082740786,
+      "grad_norm": 1.1674656867980957,
+      "learning_rate": 0.001,
+      "loss": 0.8167,
+      "step": 144400
+    },
+    {
+      "epoch": 46.7032967032967,
+      "grad_norm": 1.259466290473938,
+      "learning_rate": 0.001,
+      "loss": 0.8222,
+      "step": 144500
+    },
+    {
+      "epoch": 46.735617323852615,
+      "grad_norm": 1.4729558229446411,
+      "learning_rate": 0.001,
+      "loss": 0.8043,
+      "step": 144600
+    },
+    {
+      "epoch": 46.76793794440853,
+      "grad_norm": 1.0603488683700562,
+      "learning_rate": 0.001,
+      "loss": 0.8125,
+      "step": 144700
+    },
+    {
+      "epoch": 46.800258564964444,
+      "grad_norm": 3.8644630908966064,
+      "learning_rate": 0.001,
+      "loss": 0.8291,
+      "step": 144800
+    },
+    {
+      "epoch": 46.83257918552036,
+      "grad_norm": 1.261315107345581,
+      "learning_rate": 0.001,
+      "loss": 0.8183,
+      "step": 144900
+    },
+    {
+      "epoch": 46.864899806076274,
+      "grad_norm": 1.1630548238754272,
+      "learning_rate": 0.001,
+      "loss": 0.8281,
+      "step": 145000
+    },
+    {
+      "epoch": 46.89722042663219,
+      "grad_norm": 1.2682583332061768,
+      "learning_rate": 0.001,
+      "loss": 0.8263,
+      "step": 145100
+    },
+    {
+      "epoch": 46.9295410471881,
+      "grad_norm": 1.2569491863250732,
+      "learning_rate": 0.001,
+      "loss": 0.8012,
+      "step": 145200
+    },
+    {
+      "epoch": 46.96186166774402,
+      "grad_norm": 1.174126148223877,
+      "learning_rate": 0.001,
+      "loss": 0.8269,
+      "step": 145300
+    },
+    {
+      "epoch": 46.99418228829994,
+      "grad_norm": 1.253170371055603,
+      "learning_rate": 0.001,
+      "loss": 0.826,
+      "step": 145400
+    },
+    {
+      "epoch": 47.02650290885585,
+      "grad_norm": 1.2278573513031006,
+      "learning_rate": 0.001,
+      "loss": 0.7577,
+      "step": 145500
+    },
+    {
+      "epoch": 47.05882352941177,
+      "grad_norm": 1.2207344770431519,
+      "learning_rate": 0.001,
+      "loss": 0.7376,
+      "step": 145600
+    },
+    {
+      "epoch": 47.09114414996768,
+      "grad_norm": 1.3616275787353516,
+      "learning_rate": 0.001,
+      "loss": 0.745,
+      "step": 145700
+    },
+    {
+      "epoch": 47.1234647705236,
+      "grad_norm": 1.287418007850647,
+      "learning_rate": 0.001,
+      "loss": 0.7458,
+      "step": 145800
+    },
+    {
+      "epoch": 47.15578539107951,
+      "grad_norm": 1.1178083419799805,
+      "learning_rate": 0.001,
+      "loss": 0.7562,
+      "step": 145900
+    },
+    {
+      "epoch": 47.188106011635426,
+      "grad_norm": 1.5619690418243408,
+      "learning_rate": 0.001,
+      "loss": 0.7389,
+      "step": 146000
+    },
+    {
+      "epoch": 47.22042663219134,
+      "grad_norm": 1.4271554946899414,
+      "learning_rate": 0.001,
+      "loss": 0.7723,
+      "step": 146100
+    },
+    {
+      "epoch": 47.252747252747255,
+      "grad_norm": 1.7732999324798584,
+      "learning_rate": 0.001,
+      "loss": 0.7538,
+      "step": 146200
+    },
+    {
+      "epoch": 47.28506787330317,
+      "grad_norm": 1.4282053709030151,
+      "learning_rate": 0.001,
+      "loss": 0.7821,
+      "step": 146300
+    },
+    {
+      "epoch": 47.317388493859085,
+      "grad_norm": 1.228203535079956,
+      "learning_rate": 0.001,
+      "loss": 0.7722,
+      "step": 146400
+    },
+    {
+      "epoch": 47.349709114415,
+      "grad_norm": 1.3013367652893066,
+      "learning_rate": 0.001,
+      "loss": 0.7652,
+      "step": 146500
+    },
+    {
+      "epoch": 47.382029734970914,
+      "grad_norm": 1.2661182880401611,
+      "learning_rate": 0.001,
+      "loss": 0.7777,
+      "step": 146600
+    },
+    {
+      "epoch": 47.41435035552683,
+      "grad_norm": 1.3260958194732666,
+      "learning_rate": 0.001,
+      "loss": 0.7804,
+      "step": 146700
+    },
+    {
+      "epoch": 47.44667097608274,
+      "grad_norm": 1.212944507598877,
+      "learning_rate": 0.001,
+      "loss": 0.7679,
+      "step": 146800
+    },
+    {
+      "epoch": 47.47899159663866,
+      "grad_norm": 1.4435935020446777,
+      "learning_rate": 0.001,
+      "loss": 0.765,
+      "step": 146900
+    },
+    {
+      "epoch": 47.51131221719457,
+      "grad_norm": 1.4578791856765747,
+      "learning_rate": 0.001,
+      "loss": 0.7912,
+      "step": 147000
+    },
+    {
+      "epoch": 47.543632837750486,
+      "grad_norm": 1.462074637413025,
+      "learning_rate": 0.001,
+      "loss": 0.7974,
+      "step": 147100
+    },
+    {
+      "epoch": 47.5759534583064,
+      "grad_norm": 1.403319001197815,
+      "learning_rate": 0.001,
+      "loss": 0.7922,
+      "step": 147200
+    },
+    {
+      "epoch": 47.608274078862316,
+      "grad_norm": 1.2772393226623535,
+      "learning_rate": 0.001,
+      "loss": 0.8032,
+      "step": 147300
+    },
+    {
+      "epoch": 47.64059469941823,
+      "grad_norm": 1.217108964920044,
+      "learning_rate": 0.001,
+      "loss": 0.8016,
+      "step": 147400
+    },
+    {
+      "epoch": 47.672915319974145,
+      "grad_norm": 1.2560069561004639,
+      "learning_rate": 0.001,
+      "loss": 0.802,
+      "step": 147500
+    },
+    {
+      "epoch": 47.70523594053006,
+      "grad_norm": 1.161948800086975,
+      "learning_rate": 0.001,
+      "loss": 0.7941,
+      "step": 147600
+    },
+    {
+      "epoch": 47.737556561085974,
+      "grad_norm": 1.2663328647613525,
+      "learning_rate": 0.001,
+      "loss": 0.8108,
+      "step": 147700
+    },
+    {
+      "epoch": 47.76987718164189,
+      "grad_norm": 1.628860592842102,
+      "learning_rate": 0.001,
+      "loss": 0.8044,
+      "step": 147800
+    },
+    {
+      "epoch": 47.8021978021978,
+      "grad_norm": 1.2811692953109741,
+      "learning_rate": 0.001,
+      "loss": 0.8118,
+      "step": 147900
+    },
+    {
+      "epoch": 47.83451842275372,
+      "grad_norm": 1.2758411169052124,
+      "learning_rate": 0.001,
+      "loss": 0.8147,
+      "step": 148000
+    },
+    {
+      "epoch": 47.86683904330963,
+      "grad_norm": 1.3008402585983276,
+      "learning_rate": 0.001,
+      "loss": 0.8128,
+      "step": 148100
+    },
+    {
+      "epoch": 47.89915966386555,
+      "grad_norm": 1.4310789108276367,
+      "learning_rate": 0.001,
+      "loss": 0.8126,
+      "step": 148200
+    },
+    {
+      "epoch": 47.93148028442146,
+      "grad_norm": 1.4643484354019165,
+      "learning_rate": 0.001,
+      "loss": 0.8037,
+      "step": 148300
+    },
+    {
+      "epoch": 47.963800904977376,
+      "grad_norm": 1.2317537069320679,
+      "learning_rate": 0.001,
+      "loss": 0.8193,
+      "step": 148400
+    },
+    {
+      "epoch": 47.99612152553329,
+      "grad_norm": 1.2515267133712769,
+      "learning_rate": 0.001,
+      "loss": 0.8045,
+      "step": 148500
+    },
+    {
+      "epoch": 48.028442146089205,
+      "grad_norm": 1.475560188293457,
+      "learning_rate": 0.001,
+      "loss": 0.7236,
+      "step": 148600
+    },
+    {
+      "epoch": 48.06076276664512,
+      "grad_norm": 1.2725452184677124,
+      "learning_rate": 0.001,
+      "loss": 0.7288,
+      "step": 148700
+    },
+    {
+      "epoch": 48.093083387201034,
+      "grad_norm": 1.295411467552185,
+      "learning_rate": 0.001,
+      "loss": 0.7375,
+      "step": 148800
+    },
+    {
+      "epoch": 48.12540400775695,
+      "grad_norm": 1.436627984046936,
+      "learning_rate": 0.001,
+      "loss": 0.7454,
+      "step": 148900
+    },
+    {
+      "epoch": 48.15772462831286,
+      "grad_norm": 1.3891310691833496,
+      "learning_rate": 0.001,
+      "loss": 0.744,
+      "step": 149000
+    },
+    {
+      "epoch": 48.19004524886878,
+      "grad_norm": 1.15779447555542,
+      "learning_rate": 0.001,
+      "loss": 0.7465,
+      "step": 149100
+    },
+    {
+      "epoch": 48.22236586942469,
+      "grad_norm": 1.1223961114883423,
+      "learning_rate": 0.001,
+      "loss": 0.7363,
+      "step": 149200
+    },
+    {
+      "epoch": 48.25468648998061,
+      "grad_norm": 1.1627105474472046,
+      "learning_rate": 0.001,
+      "loss": 0.7456,
+      "step": 149300
+    },
+    {
+      "epoch": 48.28700711053652,
+      "grad_norm": 1.5019142627716064,
+      "learning_rate": 0.001,
+      "loss": 0.7612,
+      "step": 149400
+    },
+    {
+      "epoch": 48.319327731092436,
+      "grad_norm": 1.1844847202301025,
+      "learning_rate": 0.001,
+      "loss": 0.7567,
+      "step": 149500
+    },
+    {
+      "epoch": 48.35164835164835,
+      "grad_norm": 1.4008020162582397,
+      "learning_rate": 0.001,
+      "loss": 0.7587,
+      "step": 149600
+    },
+    {
+      "epoch": 48.383968972204265,
+      "grad_norm": 1.299863338470459,
+      "learning_rate": 0.001,
+      "loss": 0.7637,
+      "step": 149700
+    },
+    {
+      "epoch": 48.41628959276018,
+      "grad_norm": 1.1990106105804443,
+      "learning_rate": 0.001,
+      "loss": 0.7528,
+      "step": 149800
+    },
+    {
+      "epoch": 48.448610213316094,
+      "grad_norm": 1.5277669429779053,
+      "learning_rate": 0.001,
+      "loss": 0.7751,
+      "step": 149900
+    },
+    {
+      "epoch": 48.48093083387201,
+      "grad_norm": 1.361127495765686,
+      "learning_rate": 0.001,
+      "loss": 0.7523,
+      "step": 150000
+    },
+    {
+      "epoch": 48.51325145442792,
+      "grad_norm": 1.2485038042068481,
+      "learning_rate": 0.001,
+      "loss": 0.7826,
+      "step": 150100
+    },
+    {
+      "epoch": 48.54557207498384,
+      "grad_norm": 1.1917405128479004,
+      "learning_rate": 0.001,
+      "loss": 0.7906,
+      "step": 150200
+    },
+    {
+      "epoch": 48.57789269553975,
+      "grad_norm": 1.3229522705078125,
+      "learning_rate": 0.001,
+      "loss": 0.7801,
+      "step": 150300
+    },
+    {
+      "epoch": 48.61021331609567,
+      "grad_norm": 1.2076053619384766,
+      "learning_rate": 0.001,
+      "loss": 0.7907,
+      "step": 150400
+    },
+    {
+      "epoch": 48.64253393665158,
+      "grad_norm": 1.409424066543579,
+      "learning_rate": 0.001,
+      "loss": 0.7948,
+      "step": 150500
+    },
+    {
+      "epoch": 48.674854557207496,
+      "grad_norm": 1.1984336376190186,
+      "learning_rate": 0.001,
+      "loss": 0.7876,
+      "step": 150600
+    },
+    {
+      "epoch": 48.70717517776341,
+      "grad_norm": 14.626358985900879,
+      "learning_rate": 0.001,
+      "loss": 0.7982,
+      "step": 150700
+    },
+    {
+      "epoch": 48.739495798319325,
+      "grad_norm": 1.6344068050384521,
+      "learning_rate": 0.001,
+      "loss": 0.7986,
+      "step": 150800
+    },
+    {
+      "epoch": 48.77181641887524,
+      "grad_norm": 1.367846965789795,
+      "learning_rate": 0.001,
+      "loss": 0.791,
+      "step": 150900
+    },
+    {
+      "epoch": 48.804137039431154,
+      "grad_norm": 1.1952319145202637,
+      "learning_rate": 0.001,
+      "loss": 0.7965,
+      "step": 151000
+    },
+    {
+      "epoch": 48.83645765998707,
+      "grad_norm": 1.4131128787994385,
+      "learning_rate": 0.001,
+      "loss": 0.8009,
+      "step": 151100
+    },
+    {
+      "epoch": 48.86877828054298,
+      "grad_norm": 1.2616978883743286,
+      "learning_rate": 0.001,
+      "loss": 0.8031,
+      "step": 151200
+    },
+    {
+      "epoch": 48.9010989010989,
+      "grad_norm": 1.5032106637954712,
+      "learning_rate": 0.001,
+      "loss": 0.8055,
+      "step": 151300
+    },
+    {
+      "epoch": 48.93341952165481,
+      "grad_norm": 1.5496946573257446,
+      "learning_rate": 0.001,
+      "loss": 0.801,
+      "step": 151400
+    },
+    {
+      "epoch": 48.96574014221073,
+      "grad_norm": 1.0915486812591553,
+      "learning_rate": 0.001,
+      "loss": 0.8008,
+      "step": 151500
+    },
+    {
+      "epoch": 48.99806076276664,
+      "grad_norm": 1.169753074645996,
+      "learning_rate": 0.001,
+      "loss": 0.7731,
+      "step": 151600
+    },
+    {
+      "epoch": 49.03038138332256,
+      "grad_norm": 1.0841970443725586,
+      "learning_rate": 0.001,
+      "loss": 0.7138,
+      "step": 151700
+    },
+    {
+      "epoch": 49.06270200387848,
+      "grad_norm": 1.2232255935668945,
+      "learning_rate": 0.001,
+      "loss": 0.7295,
+      "step": 151800
+    },
+    {
+      "epoch": 49.09502262443439,
+      "grad_norm": 1.101244330406189,
+      "learning_rate": 0.001,
+      "loss": 0.7295,
+      "step": 151900
+    },
+    {
+      "epoch": 49.12734324499031,
+      "grad_norm": 1.1868951320648193,
+      "learning_rate": 0.001,
+      "loss": 0.7322,
+      "step": 152000
+    },
+    {
+      "epoch": 49.15966386554622,
+      "grad_norm": 1.4374502897262573,
+      "learning_rate": 0.001,
+      "loss": 0.7394,
+      "step": 152100
+    },
+    {
+      "epoch": 49.191984486102136,
+      "grad_norm": 1.1606899499893188,
+      "learning_rate": 0.001,
+      "loss": 0.7298,
+      "step": 152200
+    },
+    {
+      "epoch": 49.22430510665805,
+      "grad_norm": 1.3472959995269775,
+      "learning_rate": 0.001,
+      "loss": 0.7425,
+      "step": 152300
+    },
+    {
+      "epoch": 49.256625727213965,
+      "grad_norm": 1.0685003995895386,
+      "learning_rate": 0.001,
+      "loss": 0.7441,
+      "step": 152400
+    },
+    {
+      "epoch": 49.28894634776988,
+      "grad_norm": 0.9984011650085449,
+      "learning_rate": 0.001,
+      "loss": 0.7484,
+      "step": 152500
+    },
+    {
+      "epoch": 49.321266968325794,
+      "grad_norm": 1.1586939096450806,
+      "learning_rate": 0.001,
+      "loss": 0.7544,
+      "step": 152600
+    },
+    {
+      "epoch": 49.35358758888171,
+      "grad_norm": 1.0123095512390137,
+      "learning_rate": 0.001,
+      "loss": 0.771,
+      "step": 152700
+    },
+    {
+      "epoch": 49.38590820943762,
+      "grad_norm": 1.0273991823196411,
+      "learning_rate": 0.001,
+      "loss": 0.7457,
+      "step": 152800
+    },
+    {
+      "epoch": 49.41822882999354,
+      "grad_norm": 1.0201725959777832,
+      "learning_rate": 0.001,
+      "loss": 0.7645,
+      "step": 152900
+    },
+    {
+      "epoch": 49.45054945054945,
+      "grad_norm": 1.0155104398727417,
+      "learning_rate": 0.001,
+      "loss": 0.7649,
+      "step": 153000
+    },
+    {
+      "epoch": 49.48287007110537,
+      "grad_norm": 1.6812835931777954,
+      "learning_rate": 0.001,
+      "loss": 0.7636,
+      "step": 153100
+    },
+    {
+      "epoch": 49.51519069166128,
+      "grad_norm": 1.271406888961792,
+      "learning_rate": 0.001,
+      "loss": 0.7741,
+      "step": 153200
+    },
+    {
+      "epoch": 49.547511312217196,
+      "grad_norm": 1.2045003175735474,
+      "learning_rate": 0.001,
+      "loss": 0.769,
+      "step": 153300
+    },
+    {
+      "epoch": 49.57983193277311,
+      "grad_norm": 1.3944015502929688,
+      "learning_rate": 0.001,
+      "loss": 0.7857,
+      "step": 153400
+    },
+    {
+      "epoch": 49.612152553329025,
+      "grad_norm": 1.1044825315475464,
+      "learning_rate": 0.001,
+      "loss": 0.7744,
+      "step": 153500
+    },
+    {
+      "epoch": 49.64447317388494,
+      "grad_norm": 1.2409112453460693,
+      "learning_rate": 0.001,
+      "loss": 0.7744,
+      "step": 153600
+    },
+    {
+      "epoch": 49.676793794440854,
+      "grad_norm": 1.0952699184417725,
+      "learning_rate": 0.001,
+      "loss": 0.7621,
+      "step": 153700
+    },
+    {
+      "epoch": 49.70911441499677,
+      "grad_norm": 1.224043607711792,
+      "learning_rate": 0.001,
+      "loss": 0.7822,
+      "step": 153800
+    },
+    {
+      "epoch": 49.74143503555268,
+      "grad_norm": 1.085472822189331,
+      "learning_rate": 0.001,
+      "loss": 0.7815,
+      "step": 153900
+    },
+    {
+      "epoch": 49.7737556561086,
+      "grad_norm": 1.2062139511108398,
+      "learning_rate": 0.001,
+      "loss": 0.7739,
+      "step": 154000
+    },
+    {
+      "epoch": 49.80607627666451,
+      "grad_norm": 1.1115362644195557,
+      "learning_rate": 0.001,
+      "loss": 0.775,
+      "step": 154100
+    },
+    {
+      "epoch": 49.83839689722043,
+      "grad_norm": 1.2838987112045288,
+      "learning_rate": 0.001,
+      "loss": 0.7823,
+      "step": 154200
+    },
+    {
+      "epoch": 49.87071751777634,
+      "grad_norm": 1.238246202468872,
+      "learning_rate": 0.001,
+      "loss": 0.7866,
+      "step": 154300
+    },
+    {
+      "epoch": 49.903038138332256,
+      "grad_norm": 1.235521912574768,
+      "learning_rate": 0.001,
+      "loss": 0.7761,
+      "step": 154400
+    },
+    {
+      "epoch": 49.93535875888817,
+      "grad_norm": 1.0825852155685425,
+      "learning_rate": 0.001,
+      "loss": 0.7827,
+      "step": 154500
+    },
+    {
+      "epoch": 49.967679379444085,
+      "grad_norm": 1.4009476900100708,
+      "learning_rate": 0.001,
+      "loss": 0.7995,
+      "step": 154600
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 0.06359585374593735,
+      "learning_rate": 0.001,
+      "loss": 0.7531,
+      "step": 154700
+    },
+    {
+      "epoch": 50.032320620555915,
+      "grad_norm": 0.6125797033309937,
+      "learning_rate": 0.001,
+      "loss": 0.6978,
+      "step": 154800
+    },
+    {
+      "epoch": 50.06464124111183,
+      "grad_norm": 0.09692316502332687,
+      "learning_rate": 0.001,
+      "loss": 0.7195,
+      "step": 154900
+    },
+    {
+      "epoch": 50.096961861667744,
+      "grad_norm": 0.6651096343994141,
+      "learning_rate": 0.001,
+      "loss": 0.7354,
+      "step": 155000
+    },
+    {
+      "epoch": 50.12928248222366,
+      "grad_norm": 0.3355295658111572,
+      "learning_rate": 0.001,
+      "loss": 0.7262,
+      "step": 155100
+    },
+    {
+      "epoch": 50.16160310277957,
+      "grad_norm": 0.36281466484069824,
+      "learning_rate": 0.001,
+      "loss": 0.7209,
+      "step": 155200
+    },
+    {
+      "epoch": 50.19392372333549,
+      "grad_norm": 0.5302878022193909,
+      "learning_rate": 0.001,
+      "loss": 0.7309,
+      "step": 155300
+    },
+    {
+      "epoch": 50.2262443438914,
+      "grad_norm": 0.20198680460453033,
+      "learning_rate": 0.001,
+      "loss": 0.7168,
+      "step": 155400
+    },
+    {
+      "epoch": 50.25856496444732,
+      "grad_norm": 0.5641872882843018,
+      "learning_rate": 0.001,
+      "loss": 0.7314,
+      "step": 155500
+    },
+    {
+      "epoch": 50.29088558500323,
+      "grad_norm": 0.1572069376707077,
+      "learning_rate": 0.001,
+      "loss": 0.7431,
+      "step": 155600
+    },
+    {
+      "epoch": 50.323206205559146,
+      "grad_norm": 1.641300082206726,
+      "learning_rate": 0.001,
+      "loss": 0.7391,
+      "step": 155700
+    },
+    {
+      "epoch": 50.35552682611506,
+      "grad_norm": 0.5732482671737671,
+      "learning_rate": 0.001,
+      "loss": 0.7487,
+      "step": 155800
+    },
+    {
+      "epoch": 50.387847446670975,
+      "grad_norm": 0.4709911644458771,
+      "learning_rate": 0.001,
+      "loss": 0.7426,
+      "step": 155900
+    },
+    {
+      "epoch": 50.42016806722689,
+      "grad_norm": 0.3790436387062073,
+      "learning_rate": 0.001,
+      "loss": 0.7581,
+      "step": 156000
+    },
+    {
+      "epoch": 50.452488687782804,
+      "grad_norm": 0.8265621662139893,
+      "learning_rate": 0.001,
+      "loss": 0.7376,
+      "step": 156100
+    },
+    {
+      "epoch": 50.48480930833872,
+      "grad_norm": 0.5220683813095093,
+      "learning_rate": 0.001,
+      "loss": 0.7474,
+      "step": 156200
+    },
+    {
+      "epoch": 50.51712992889463,
+      "grad_norm": 0.22513864934444427,
+      "learning_rate": 0.001,
+      "loss": 0.7492,
+      "step": 156300
+    },
+    {
+      "epoch": 50.54945054945055,
+      "grad_norm": 0.5701922178268433,
+      "learning_rate": 0.001,
+      "loss": 0.7517,
+      "step": 156400
+    },
+    {
+      "epoch": 50.58177117000646,
+      "grad_norm": 0.5959166288375854,
+      "learning_rate": 0.001,
+      "loss": 0.7521,
+      "step": 156500
+    },
+    {
+      "epoch": 50.61409179056238,
+      "grad_norm": 0.5421650409698486,
+      "learning_rate": 0.001,
+      "loss": 0.7633,
+      "step": 156600
+    },
+    {
+      "epoch": 50.64641241111829,
+      "grad_norm": 0.604699969291687,
+      "learning_rate": 0.001,
+      "loss": 0.7506,
+      "step": 156700
+    },
+    {
+      "epoch": 50.678733031674206,
+      "grad_norm": 0.7056938409805298,
+      "learning_rate": 0.001,
+      "loss": 0.7583,
+      "step": 156800
+    },
+    {
+      "epoch": 50.71105365223012,
+      "grad_norm": 0.36048054695129395,
+      "learning_rate": 0.001,
+      "loss": 0.7446,
+      "step": 156900
+    },
+    {
+      "epoch": 50.743374272786035,
+      "grad_norm": 1.0495373010635376,
+      "learning_rate": 0.001,
+      "loss": 0.791,
+      "step": 157000
+    },
+    {
+      "epoch": 50.77569489334195,
+      "grad_norm": 0.38334110379219055,
+      "learning_rate": 0.001,
+      "loss": 0.7661,
+      "step": 157100
+    },
+    {
+      "epoch": 50.808015513897864,
+      "grad_norm": 0.34685152769088745,
+      "learning_rate": 0.001,
+      "loss": 0.7642,
+      "step": 157200
+    },
+    {
+      "epoch": 50.84033613445378,
+      "grad_norm": 6.124505043029785,
+      "learning_rate": 0.001,
+      "loss": 0.7831,
+      "step": 157300
+    },
+    {
+      "epoch": 50.87265675500969,
+      "grad_norm": 0.3870513439178467,
+      "learning_rate": 0.001,
+      "loss": 0.7879,
+      "step": 157400
+    },
+    {
+      "epoch": 50.90497737556561,
+      "grad_norm": 0.3915799558162689,
+      "learning_rate": 0.001,
+      "loss": 0.7751,
+      "step": 157500
+    },
+    {
+      "epoch": 50.93729799612152,
+      "grad_norm": 0.4591212868690491,
+      "learning_rate": 0.001,
+      "loss": 0.7986,
+      "step": 157600
+    },
+    {
+      "epoch": 50.96961861667744,
+      "grad_norm": 0.3180168867111206,
+      "learning_rate": 0.001,
+      "loss": 0.7874,
+      "step": 157700
+    },
+    {
+      "epoch": 51.00193923723336,
+      "grad_norm": 1.6160231828689575,
+      "learning_rate": 0.001,
+      "loss": 0.804,
+      "step": 157800
+    },
+    {
+      "epoch": 51.03425985778927,
+      "grad_norm": 1.5989055633544922,
+      "learning_rate": 0.001,
+      "loss": 0.6946,
+      "step": 157900
+    },
+    {
+      "epoch": 51.06658047834519,
+      "grad_norm": 1.248809576034546,
+      "learning_rate": 0.001,
+      "loss": 0.6941,
+      "step": 158000
+    },
+    {
+      "epoch": 51.0989010989011,
+      "grad_norm": 1.428173303604126,
+      "learning_rate": 0.001,
+      "loss": 0.717,
+      "step": 158100
+    },
+    {
+      "epoch": 51.13122171945702,
+      "grad_norm": 1.5812784433364868,
+      "learning_rate": 0.001,
+      "loss": 0.7142,
+      "step": 158200
+    },
+    {
+      "epoch": 51.16354234001293,
+      "grad_norm": 1.3164476156234741,
+      "learning_rate": 0.001,
+      "loss": 0.6995,
+      "step": 158300
+    },
+    {
+      "epoch": 51.195862960568846,
+      "grad_norm": 1.545201301574707,
+      "learning_rate": 0.001,
+      "loss": 0.7093,
+      "step": 158400
+    },
+    {
+      "epoch": 51.22818358112476,
+      "grad_norm": 1.6235681772232056,
+      "learning_rate": 0.001,
+      "loss": 0.728,
+      "step": 158500
+    },
+    {
+      "epoch": 51.260504201680675,
+      "grad_norm": 1.6064990758895874,
+      "learning_rate": 0.001,
+      "loss": 0.7269,
+      "step": 158600
+    },
+    {
+      "epoch": 51.29282482223659,
+      "grad_norm": 1.4662895202636719,
+      "learning_rate": 0.001,
+      "loss": 0.741,
+      "step": 158700
+    },
+    {
+      "epoch": 51.325145442792504,
+      "grad_norm": 1.4283791780471802,
+      "learning_rate": 0.001,
+      "loss": 0.7171,
+      "step": 158800
+    },
+    {
+      "epoch": 51.35746606334842,
+      "grad_norm": 1.3249914646148682,
+      "learning_rate": 0.001,
+      "loss": 0.7335,
+      "step": 158900
+    },
+    {
+      "epoch": 51.38978668390433,
+      "grad_norm": 1.3508305549621582,
+      "learning_rate": 0.001,
+      "loss": 0.7484,
+      "step": 159000
+    },
+    {
+      "epoch": 51.42210730446025,
+      "grad_norm": 1.402336597442627,
+      "learning_rate": 0.001,
+      "loss": 0.7242,
+      "step": 159100
+    },
+    {
+      "epoch": 51.45442792501616,
+      "grad_norm": 1.5387367010116577,
+      "learning_rate": 0.001,
+      "loss": 0.7565,
+      "step": 159200
+    },
+    {
+      "epoch": 51.48674854557208,
+      "grad_norm": 1.280126690864563,
+      "learning_rate": 0.001,
+      "loss": 0.7443,
+      "step": 159300
+    },
+    {
+      "epoch": 51.51906916612799,
+      "grad_norm": 1.4007776975631714,
+      "learning_rate": 0.001,
+      "loss": 0.73,
+      "step": 159400
+    },
+    {
+      "epoch": 51.551389786683906,
+      "grad_norm": 1.4511324167251587,
+      "learning_rate": 0.001,
+      "loss": 0.7601,
+      "step": 159500
+    },
+    {
+      "epoch": 51.58371040723982,
+      "grad_norm": 1.6074110269546509,
+      "learning_rate": 0.001,
+      "loss": 0.7468,
+      "step": 159600
+    },
+    {
+      "epoch": 51.616031027795735,
+      "grad_norm": 1.6182364225387573,
+      "learning_rate": 0.001,
+      "loss": 0.7409,
+      "step": 159700
+    },
+    {
+      "epoch": 51.64835164835165,
+      "grad_norm": 1.3085405826568604,
+      "learning_rate": 0.001,
+      "loss": 0.7545,
+      "step": 159800
+    },
+    {
+      "epoch": 51.680672268907564,
+      "grad_norm": 1.2141468524932861,
+      "learning_rate": 0.001,
+      "loss": 0.7507,
+      "step": 159900
+    },
+    {
+      "epoch": 51.71299288946348,
+      "grad_norm": 1.3837734460830688,
+      "learning_rate": 0.001,
+      "loss": 0.7613,
+      "step": 160000
+    },
+    {
+      "epoch": 51.74531351001939,
+      "grad_norm": 1.4727295637130737,
+      "learning_rate": 0.001,
+      "loss": 0.7631,
+      "step": 160100
+    },
+    {
+      "epoch": 51.77763413057531,
+      "grad_norm": 1.4606034755706787,
+      "learning_rate": 0.001,
+      "loss": 0.7705,
+      "step": 160200
+    },
+    {
+      "epoch": 51.80995475113122,
+      "grad_norm": 1.729146957397461,
+      "learning_rate": 0.001,
+      "loss": 0.7593,
+      "step": 160300
+    },
+    {
+      "epoch": 51.84227537168714,
+      "grad_norm": 1.6483460664749146,
+      "learning_rate": 0.001,
+      "loss": 0.7622,
+      "step": 160400
+    },
+    {
+      "epoch": 51.87459599224305,
+      "grad_norm": 1.164381980895996,
+      "learning_rate": 0.001,
+      "loss": 0.7678,
+      "step": 160500
+    },
+    {
+      "epoch": 51.906916612798966,
+      "grad_norm": 1.6500674486160278,
+      "learning_rate": 0.001,
+      "loss": 0.7659,
+      "step": 160600
+    },
+    {
+      "epoch": 51.93923723335488,
+      "grad_norm": 1.311309814453125,
+      "learning_rate": 0.001,
+      "loss": 0.7597,
+      "step": 160700
+    },
+    {
+      "epoch": 51.971557853910795,
+      "grad_norm": 1.169773817062378,
+      "learning_rate": 0.001,
+      "loss": 0.7661,
+      "step": 160800
+    },
+    {
+      "epoch": 52.00387847446671,
+      "grad_norm": 1.237779140472412,
+      "learning_rate": 0.001,
+      "loss": 0.7649,
+      "step": 160900
+    },
+    {
+      "epoch": 52.036199095022624,
+      "grad_norm": 1.669989824295044,
+      "learning_rate": 0.001,
+      "loss": 0.697,
+      "step": 161000
+    },
+    {
+      "epoch": 52.06851971557854,
+      "grad_norm": 1.4660130739212036,
+      "learning_rate": 0.001,
+      "loss": 0.7028,
+      "step": 161100
+    },
+    {
+      "epoch": 52.10084033613445,
+      "grad_norm": 1.304290533065796,
+      "learning_rate": 0.001,
+      "loss": 0.7115,
+      "step": 161200
+    },
+    {
+      "epoch": 52.13316095669037,
+      "grad_norm": 1.3688267469406128,
+      "learning_rate": 0.001,
+      "loss": 0.7015,
+      "step": 161300
+    },
+    {
+      "epoch": 52.16548157724628,
+      "grad_norm": 1.2158695459365845,
+      "learning_rate": 0.001,
+      "loss": 0.701,
+      "step": 161400
+    },
+    {
+      "epoch": 52.1978021978022,
+      "grad_norm": 1.2809381484985352,
+      "learning_rate": 0.001,
+      "loss": 0.703,
+      "step": 161500
+    },
+    {
+      "epoch": 52.23012281835811,
+      "grad_norm": 1.359277367591858,
+      "learning_rate": 0.001,
+      "loss": 0.7179,
+      "step": 161600
+    },
+    {
+      "epoch": 52.262443438914026,
+      "grad_norm": 1.4113330841064453,
+      "learning_rate": 0.001,
+      "loss": 0.7142,
+      "step": 161700
+    },
+    {
+      "epoch": 52.29476405946994,
+      "grad_norm": 1.4368807077407837,
+      "learning_rate": 0.001,
+      "loss": 0.7208,
+      "step": 161800
+    },
+    {
+      "epoch": 52.327084680025855,
+      "grad_norm": 1.1833715438842773,
+      "learning_rate": 0.001,
+      "loss": 0.715,
+      "step": 161900
+    },
+    {
+      "epoch": 52.35940530058177,
+      "grad_norm": 1.5551220178604126,
+      "learning_rate": 0.001,
+      "loss": 0.7246,
+      "step": 162000
+    },
+    {
+      "epoch": 52.391725921137684,
+      "grad_norm": 1.8916435241699219,
+      "learning_rate": 0.001,
+      "loss": 0.7346,
+      "step": 162100
+    },
+    {
+      "epoch": 52.4240465416936,
+      "grad_norm": 1.30422842502594,
+      "learning_rate": 0.001,
+      "loss": 0.7308,
+      "step": 162200
+    },
+    {
+      "epoch": 52.456367162249514,
+      "grad_norm": 1.31717050075531,
+      "learning_rate": 0.001,
+      "loss": 0.7259,
+      "step": 162300
+    },
+    {
+      "epoch": 52.48868778280543,
+      "grad_norm": 1.4147814512252808,
+      "learning_rate": 0.001,
+      "loss": 0.742,
+      "step": 162400
+    },
+    {
+      "epoch": 52.52100840336134,
+      "grad_norm": 1.0690675973892212,
+      "learning_rate": 0.001,
+      "loss": 0.7292,
+      "step": 162500
+    },
+    {
+      "epoch": 52.55332902391726,
+      "grad_norm": 1.655393123626709,
+      "learning_rate": 0.001,
+      "loss": 0.7393,
+      "step": 162600
+    },
+    {
+      "epoch": 52.58564964447317,
+      "grad_norm": 1.5793589353561401,
+      "learning_rate": 0.001,
+      "loss": 0.7345,
+      "step": 162700
+    },
+    {
+      "epoch": 52.617970265029086,
+      "grad_norm": 1.1870840787887573,
+      "learning_rate": 0.001,
+      "loss": 0.7449,
+      "step": 162800
+    },
+    {
+      "epoch": 52.650290885585,
+      "grad_norm": 1.2543737888336182,
+      "learning_rate": 0.001,
+      "loss": 0.7454,
+      "step": 162900
+    },
+    {
+      "epoch": 52.682611506140915,
+      "grad_norm": 1.7071529626846313,
+      "learning_rate": 0.001,
+      "loss": 0.7386,
+      "step": 163000
+    },
+    {
+      "epoch": 52.71493212669683,
+      "grad_norm": 1.2244914770126343,
+      "learning_rate": 0.001,
+      "loss": 0.7432,
+      "step": 163100
+    },
+    {
+      "epoch": 52.747252747252745,
+      "grad_norm": 1.2786270380020142,
+      "learning_rate": 0.001,
+      "loss": 0.7358,
+      "step": 163200
+    },
+    {
+      "epoch": 52.77957336780866,
+      "grad_norm": 1.573927640914917,
+      "learning_rate": 0.001,
+      "loss": 0.7321,
+      "step": 163300
+    },
+    {
+      "epoch": 52.811893988364574,
+      "grad_norm": 1.7538331747055054,
+      "learning_rate": 0.001,
+      "loss": 0.7337,
+      "step": 163400
+    },
+    {
+      "epoch": 52.84421460892049,
+      "grad_norm": 1.616270661354065,
+      "learning_rate": 0.001,
+      "loss": 0.7482,
+      "step": 163500
+    },
+    {
+      "epoch": 52.8765352294764,
+      "grad_norm": 1.253475546836853,
+      "learning_rate": 0.001,
+      "loss": 0.7635,
+      "step": 163600
+    },
+    {
+      "epoch": 52.90885585003232,
+      "grad_norm": 1.7537295818328857,
+      "learning_rate": 0.001,
+      "loss": 0.751,
+      "step": 163700
+    },
+    {
+      "epoch": 52.94117647058823,
+      "grad_norm": 1.3528518676757812,
+      "learning_rate": 0.001,
+      "loss": 0.7697,
+      "step": 163800
+    },
+    {
+      "epoch": 52.97349709114415,
+      "grad_norm": 1.401299238204956,
+      "learning_rate": 0.001,
+      "loss": 0.7577,
+      "step": 163900
+    },
+    {
+      "epoch": 53.00581771170007,
+      "grad_norm": 1.3519210815429688,
+      "learning_rate": 0.001,
+      "loss": 0.7499,
+      "step": 164000
+    },
+    {
+      "epoch": 53.03813833225598,
+      "grad_norm": 1.4237704277038574,
+      "learning_rate": 0.001,
+      "loss": 0.6907,
+      "step": 164100
+    },
+    {
+      "epoch": 53.0704589528119,
+      "grad_norm": 1.6554957628250122,
+      "learning_rate": 0.001,
+      "loss": 0.6931,
+      "step": 164200
+    },
+    {
+      "epoch": 53.10277957336781,
+      "grad_norm": 1.322135090827942,
+      "learning_rate": 0.001,
+      "loss": 0.6639,
+      "step": 164300
+    },
+    {
+      "epoch": 53.135100193923726,
+      "grad_norm": 1.5357197523117065,
+      "learning_rate": 0.001,
+      "loss": 0.6789,
+      "step": 164400
+    },
+    {
+      "epoch": 53.16742081447964,
+      "grad_norm": 1.1936781406402588,
+      "learning_rate": 0.001,
+      "loss": 0.69,
+      "step": 164500
+    },
+    {
+      "epoch": 53.199741435035556,
+      "grad_norm": 1.3940258026123047,
+      "learning_rate": 0.001,
+      "loss": 0.6974,
+      "step": 164600
+    },
+    {
+      "epoch": 53.23206205559147,
+      "grad_norm": 1.251088261604309,
+      "learning_rate": 0.001,
+      "loss": 0.7037,
+      "step": 164700
+    },
+    {
+      "epoch": 53.264382676147385,
+      "grad_norm": 1.4187980890274048,
+      "learning_rate": 0.001,
+      "loss": 0.6951,
+      "step": 164800
+    },
+    {
+      "epoch": 53.2967032967033,
+      "grad_norm": 1.410966396331787,
+      "learning_rate": 0.001,
+      "loss": 0.7132,
+      "step": 164900
+    },
+    {
+      "epoch": 53.329023917259214,
+      "grad_norm": 1.2439744472503662,
+      "learning_rate": 0.001,
+      "loss": 0.6969,
+      "step": 165000
+    },
+    {
+      "epoch": 53.36134453781513,
+      "grad_norm": 1.6391518115997314,
+      "learning_rate": 0.001,
+      "loss": 0.6999,
+      "step": 165100
+    },
+    {
+      "epoch": 53.39366515837104,
+      "grad_norm": 1.2765934467315674,
+      "learning_rate": 0.001,
+      "loss": 0.7341,
+      "step": 165200
+    },
+    {
+      "epoch": 53.42598577892696,
+      "grad_norm": 1.314164400100708,
+      "learning_rate": 0.001,
+      "loss": 0.7119,
+      "step": 165300
+    },
+    {
+      "epoch": 53.45830639948287,
+      "grad_norm": 1.3228839635849,
+      "learning_rate": 0.001,
+      "loss": 0.7288,
+      "step": 165400
+    },
+    {
+      "epoch": 53.49062702003879,
+      "grad_norm": 1.2543978691101074,
+      "learning_rate": 0.001,
+      "loss": 0.7222,
+      "step": 165500
+    },
+    {
+      "epoch": 53.5229476405947,
+      "grad_norm": 1.3691924810409546,
+      "learning_rate": 0.001,
+      "loss": 0.7289,
+      "step": 165600
+    },
+    {
+      "epoch": 53.555268261150616,
+      "grad_norm": 1.47661292552948,
+      "learning_rate": 0.001,
+      "loss": 0.7187,
+      "step": 165700
+    },
+    {
+      "epoch": 53.58758888170653,
+      "grad_norm": 1.5954946279525757,
+      "learning_rate": 0.001,
+      "loss": 0.7064,
+      "step": 165800
+    },
+    {
+      "epoch": 53.619909502262445,
+      "grad_norm": 1.302594780921936,
+      "learning_rate": 0.001,
+      "loss": 0.7357,
+      "step": 165900
+    },
+    {
+      "epoch": 53.65223012281836,
+      "grad_norm": 1.3874233961105347,
+      "learning_rate": 0.001,
+      "loss": 0.7342,
+      "step": 166000
+    },
+    {
+      "epoch": 53.684550743374274,
+      "grad_norm": 1.2953951358795166,
+      "learning_rate": 0.001,
+      "loss": 0.7399,
+      "step": 166100
+    },
+    {
+      "epoch": 53.71687136393019,
+      "grad_norm": 1.5059915781021118,
+      "learning_rate": 0.001,
+      "loss": 0.7226,
+      "step": 166200
+    },
+    {
+      "epoch": 53.7491919844861,
+      "grad_norm": 1.6534998416900635,
+      "learning_rate": 0.001,
+      "loss": 0.7478,
+      "step": 166300
+    },
+    {
+      "epoch": 53.78151260504202,
+      "grad_norm": 1.4183323383331299,
+      "learning_rate": 0.001,
+      "loss": 0.744,
+      "step": 166400
+    },
+    {
+      "epoch": 53.81383322559793,
+      "grad_norm": 1.4309442043304443,
+      "learning_rate": 0.001,
+      "loss": 0.7421,
+      "step": 166500
+    },
+    {
+      "epoch": 53.84615384615385,
+      "grad_norm": 1.4040199518203735,
+      "learning_rate": 0.001,
+      "loss": 0.7511,
+      "step": 166600
+    },
+    {
+      "epoch": 53.87847446670976,
+      "grad_norm": 1.5059336423873901,
+      "learning_rate": 0.001,
+      "loss": 0.763,
+      "step": 166700
+    },
+    {
+      "epoch": 53.910795087265676,
+      "grad_norm": 1.2663861513137817,
+      "learning_rate": 0.001,
+      "loss": 0.7299,
+      "step": 166800
+    },
+    {
+      "epoch": 53.94311570782159,
+      "grad_norm": 1.3244277238845825,
+      "learning_rate": 0.001,
+      "loss": 0.7438,
+      "step": 166900
+    },
+    {
+      "epoch": 53.975436328377505,
+      "grad_norm": 1.1952475309371948,
+      "learning_rate": 0.001,
+      "loss": 0.7579,
+      "step": 167000
+    },
+    {
+      "epoch": 54.00775694893342,
+      "grad_norm": 1.4403356313705444,
+      "learning_rate": 0.001,
+      "loss": 0.7466,
+      "step": 167100
+    },
+    {
+      "epoch": 54.040077569489334,
+      "grad_norm": 1.3646568059921265,
+      "learning_rate": 0.001,
+      "loss": 0.665,
+      "step": 167200
+    },
+    {
+      "epoch": 54.07239819004525,
+      "grad_norm": 1.3861289024353027,
+      "learning_rate": 0.001,
+      "loss": 0.6708,
+      "step": 167300
+    },
+    {
+      "epoch": 54.10471881060116,
+      "grad_norm": 1.2403484582901,
+      "learning_rate": 0.001,
+      "loss": 0.6773,
+      "step": 167400
+    },
+    {
+      "epoch": 54.13703943115708,
+      "grad_norm": 1.4412761926651,
+      "learning_rate": 0.001,
+      "loss": 0.6714,
+      "step": 167500
+    },
+    {
+      "epoch": 54.16936005171299,
+      "grad_norm": 1.5241587162017822,
+      "learning_rate": 0.001,
+      "loss": 0.6872,
+      "step": 167600
+    },
+    {
+      "epoch": 54.20168067226891,
+      "grad_norm": 1.2935175895690918,
+      "learning_rate": 0.001,
+      "loss": 0.6931,
+      "step": 167700
+    },
+    {
+      "epoch": 54.23400129282482,
+      "grad_norm": 1.2425872087478638,
+      "learning_rate": 0.001,
+      "loss": 0.6821,
+      "step": 167800
+    },
+    {
+      "epoch": 54.266321913380736,
+      "grad_norm": 1.1844414472579956,
+      "learning_rate": 0.001,
+      "loss": 0.6979,
+      "step": 167900
+    },
+    {
+      "epoch": 54.29864253393665,
+      "grad_norm": 1.5796948671340942,
+      "learning_rate": 0.001,
+      "loss": 0.7203,
+      "step": 168000
+    },
+    {
+      "epoch": 54.330963154492565,
+      "grad_norm": 1.386918306350708,
+      "learning_rate": 0.001,
+      "loss": 0.7002,
+      "step": 168100
+    },
+    {
+      "epoch": 54.36328377504848,
+      "grad_norm": 1.1819859743118286,
+      "learning_rate": 0.001,
+      "loss": 0.7158,
+      "step": 168200
+    },
+    {
+      "epoch": 54.395604395604394,
+      "grad_norm": 1.2906848192214966,
+      "learning_rate": 0.001,
+      "loss": 0.7115,
+      "step": 168300
+    },
+    {
+      "epoch": 54.42792501616031,
+      "grad_norm": 1.1121623516082764,
+      "learning_rate": 0.001,
+      "loss": 0.6974,
+      "step": 168400
+    },
+    {
+      "epoch": 54.46024563671622,
+      "grad_norm": 1.2486132383346558,
+      "learning_rate": 0.001,
+      "loss": 0.7067,
+      "step": 168500
+    },
+    {
+      "epoch": 54.49256625727214,
+      "grad_norm": 1.1886978149414062,
+      "learning_rate": 0.001,
+      "loss": 0.7072,
+      "step": 168600
+    },
+    {
+      "epoch": 54.52488687782805,
+      "grad_norm": 1.0904563665390015,
+      "learning_rate": 0.001,
+      "loss": 0.7235,
+      "step": 168700
+    },
+    {
+      "epoch": 54.55720749838397,
+      "grad_norm": 1.4579226970672607,
+      "learning_rate": 0.001,
+      "loss": 0.7086,
+      "step": 168800
+    },
+    {
+      "epoch": 54.58952811893988,
+      "grad_norm": 1.0935107469558716,
+      "learning_rate": 0.001,
+      "loss": 0.7086,
+      "step": 168900
+    },
+    {
+      "epoch": 54.621848739495796,
+      "grad_norm": 1.4180753231048584,
+      "learning_rate": 0.001,
+      "loss": 0.7231,
+      "step": 169000
+    },
+    {
+      "epoch": 54.65416936005171,
+      "grad_norm": 1.4720892906188965,
+      "learning_rate": 0.001,
+      "loss": 0.7279,
+      "step": 169100
+    },
+    {
+      "epoch": 54.686489980607625,
+      "grad_norm": 1.350489616394043,
+      "learning_rate": 0.001,
+      "loss": 0.7294,
+      "step": 169200
+    },
+    {
+      "epoch": 54.71881060116354,
+      "grad_norm": 1.259083867073059,
+      "learning_rate": 0.001,
+      "loss": 0.7361,
+      "step": 169300
+    },
+    {
+      "epoch": 54.751131221719454,
+      "grad_norm": 1.6136354207992554,
+      "learning_rate": 0.001,
+      "loss": 0.7294,
+      "step": 169400
+    },
+    {
+      "epoch": 54.78345184227537,
+      "grad_norm": 1.4642295837402344,
+      "learning_rate": 0.001,
+      "loss": 0.7428,
+      "step": 169500
+    },
+    {
+      "epoch": 54.81577246283128,
+      "grad_norm": 1.4074591398239136,
+      "learning_rate": 0.001,
+      "loss": 0.7217,
+      "step": 169600
+    },
+    {
+      "epoch": 54.8480930833872,
+      "grad_norm": 1.2914643287658691,
+      "learning_rate": 0.001,
+      "loss": 0.7551,
+      "step": 169700
+    },
+    {
+      "epoch": 54.88041370394311,
+      "grad_norm": 1.2137229442596436,
+      "learning_rate": 0.001,
+      "loss": 0.7368,
+      "step": 169800
+    },
+    {
+      "epoch": 54.91273432449903,
+      "grad_norm": 1.177645206451416,
+      "learning_rate": 0.001,
+      "loss": 0.7293,
+      "step": 169900
+    },
+    {
+      "epoch": 54.94505494505494,
+      "grad_norm": 1.1865633726119995,
+      "learning_rate": 0.001,
+      "loss": 0.7241,
+      "step": 170000
+    },
+    {
+      "epoch": 54.977375565610856,
+      "grad_norm": 1.1472890377044678,
+      "learning_rate": 0.001,
+      "loss": 0.7466,
+      "step": 170100
+    },
+    {
+      "epoch": 55.00969618616678,
+      "grad_norm": 1.3177831172943115,
+      "learning_rate": 0.001,
+      "loss": 0.7035,
+      "step": 170200
+    },
+    {
+      "epoch": 55.04201680672269,
+      "grad_norm": 2.511430263519287,
+      "learning_rate": 0.001,
+      "loss": 0.6627,
+      "step": 170300
+    },
+    {
+      "epoch": 55.07433742727861,
+      "grad_norm": 1.3882372379302979,
+      "learning_rate": 0.001,
+      "loss": 0.6713,
+      "step": 170400
+    },
+    {
+      "epoch": 55.10665804783452,
+      "grad_norm": 1.1807273626327515,
+      "learning_rate": 0.001,
+      "loss": 0.6763,
+      "step": 170500
+    },
+    {
+      "epoch": 55.138978668390436,
+      "grad_norm": 1.6859544515609741,
+      "learning_rate": 0.001,
+      "loss": 0.6848,
+      "step": 170600
+    },
+    {
+      "epoch": 55.17129928894635,
+      "grad_norm": 1.858846664428711,
+      "learning_rate": 0.001,
+      "loss": 0.6797,
+      "step": 170700
+    },
+    {
+      "epoch": 55.203619909502265,
+      "grad_norm": 1.37615168094635,
+      "learning_rate": 0.001,
+      "loss": 0.6874,
+      "step": 170800
+    },
+    {
+      "epoch": 55.23594053005818,
+      "grad_norm": 1.270747184753418,
+      "learning_rate": 0.001,
+      "loss": 0.6824,
+      "step": 170900
+    },
+    {
+      "epoch": 55.268261150614094,
+      "grad_norm": 1.7557049989700317,
+      "learning_rate": 0.001,
+      "loss": 0.6876,
+      "step": 171000
+    },
+    {
+      "epoch": 55.30058177117001,
+      "grad_norm": 1.3425112962722778,
+      "learning_rate": 0.001,
+      "loss": 0.674,
+      "step": 171100
+    },
+    {
+      "epoch": 55.33290239172592,
+      "grad_norm": 1.377942681312561,
+      "learning_rate": 0.001,
+      "loss": 0.691,
+      "step": 171200
+    },
+    {
+      "epoch": 55.36522301228184,
+      "grad_norm": 1.1815950870513916,
+      "learning_rate": 0.001,
+      "loss": 0.6899,
+      "step": 171300
+    },
+    {
+      "epoch": 55.39754363283775,
+      "grad_norm": 1.3967550992965698,
+      "learning_rate": 0.001,
+      "loss": 0.6935,
+      "step": 171400
+    },
+    {
+      "epoch": 55.42986425339367,
+      "grad_norm": 1.5351048707962036,
+      "learning_rate": 0.001,
+      "loss": 0.6875,
+      "step": 171500
+    },
+    {
+      "epoch": 55.46218487394958,
+      "grad_norm": 1.2867332696914673,
+      "learning_rate": 0.001,
+      "loss": 0.6963,
+      "step": 171600
+    },
+    {
+      "epoch": 55.494505494505496,
+      "grad_norm": 1.6633144617080688,
+      "learning_rate": 0.001,
+      "loss": 0.707,
+      "step": 171700
+    },
+    {
+      "epoch": 55.52682611506141,
+      "grad_norm": 1.3871257305145264,
+      "learning_rate": 0.001,
+      "loss": 0.7089,
+      "step": 171800
+    },
+    {
+      "epoch": 55.559146735617325,
+      "grad_norm": 1.155198097229004,
+      "learning_rate": 0.001,
+      "loss": 0.6981,
+      "step": 171900
+    },
+    {
+      "epoch": 55.59146735617324,
+      "grad_norm": 1.0028072595596313,
+      "learning_rate": 0.001,
+      "loss": 0.7082,
+      "step": 172000
+    },
+    {
+      "epoch": 55.623787976729155,
+      "grad_norm": 1.5060408115386963,
+      "learning_rate": 0.001,
+      "loss": 0.7144,
+      "step": 172100
+    },
+    {
+      "epoch": 55.65610859728507,
+      "grad_norm": 1.0690950155258179,
+      "learning_rate": 0.001,
+      "loss": 0.697,
+      "step": 172200
+    },
+    {
+      "epoch": 55.688429217840984,
+      "grad_norm": 1.4234610795974731,
+      "learning_rate": 0.001,
+      "loss": 0.7196,
+      "step": 172300
+    },
+    {
+      "epoch": 55.7207498383969,
+      "grad_norm": 1.2625986337661743,
+      "learning_rate": 0.001,
+      "loss": 0.7008,
+      "step": 172400
+    },
+    {
+      "epoch": 55.75307045895281,
+      "grad_norm": 1.2988468408584595,
+      "learning_rate": 0.001,
+      "loss": 0.7296,
+      "step": 172500
+    },
+    {
+      "epoch": 55.78539107950873,
+      "grad_norm": 1.5457489490509033,
+      "learning_rate": 0.001,
+      "loss": 0.7056,
+      "step": 172600
+    },
+    {
+      "epoch": 55.81771170006464,
+      "grad_norm": 1.2999827861785889,
+      "learning_rate": 0.001,
+      "loss": 0.7393,
+      "step": 172700
+    },
+    {
+      "epoch": 55.85003232062056,
+      "grad_norm": 1.3677887916564941,
+      "learning_rate": 0.001,
+      "loss": 0.7106,
+      "step": 172800
+    },
+    {
+      "epoch": 55.88235294117647,
+      "grad_norm": 1.438559889793396,
+      "learning_rate": 0.001,
+      "loss": 0.7275,
+      "step": 172900
+    },
+    {
+      "epoch": 55.914673561732386,
+      "grad_norm": 1.15455162525177,
+      "learning_rate": 0.001,
+      "loss": 0.7265,
+      "step": 173000
+    },
+    {
+      "epoch": 55.9469941822883,
+      "grad_norm": 1.3578943014144897,
+      "learning_rate": 0.001,
+      "loss": 0.734,
+      "step": 173100
+    },
+    {
+      "epoch": 55.979314802844215,
+      "grad_norm": 1.8947703838348389,
+      "learning_rate": 0.001,
+      "loss": 0.7342,
+      "step": 173200
+    },
+    {
+      "epoch": 56.01163542340013,
+      "grad_norm": 1.7013978958129883,
+      "learning_rate": 0.001,
+      "loss": 0.6952,
+      "step": 173300
+    },
+    {
+      "epoch": 56.043956043956044,
+      "grad_norm": 1.2173410654067993,
+      "learning_rate": 0.001,
+      "loss": 0.6489,
+      "step": 173400
+    },
+    {
+      "epoch": 56.07627666451196,
+      "grad_norm": 1.3756160736083984,
+      "learning_rate": 0.001,
+      "loss": 0.6525,
+      "step": 173500
+    },
+    {
+      "epoch": 56.10859728506787,
+      "grad_norm": 1.1060224771499634,
+      "learning_rate": 0.001,
+      "loss": 0.6633,
+      "step": 173600
+    },
+    {
+      "epoch": 56.14091790562379,
+      "grad_norm": 1.318628191947937,
+      "learning_rate": 0.001,
+      "loss": 0.6619,
+      "step": 173700
+    },
+    {
+      "epoch": 56.1732385261797,
+      "grad_norm": 1.0686438083648682,
+      "learning_rate": 0.001,
+      "loss": 0.6631,
+      "step": 173800
+    },
+    {
+      "epoch": 56.20555914673562,
+      "grad_norm": 1.1766424179077148,
+      "learning_rate": 0.001,
+      "loss": 0.6583,
+      "step": 173900
+    },
+    {
+      "epoch": 56.23787976729153,
+      "grad_norm": 1.3714375495910645,
+      "learning_rate": 0.001,
+      "loss": 0.7061,
+      "step": 174000
+    },
+    {
+      "epoch": 56.270200387847446,
+      "grad_norm": 1.0865048170089722,
+      "learning_rate": 0.001,
+      "loss": 0.6906,
+      "step": 174100
+    },
+    {
+      "epoch": 56.30252100840336,
+      "grad_norm": 1.3435384035110474,
+      "learning_rate": 0.001,
+      "loss": 0.6932,
+      "step": 174200
+    },
+    {
+      "epoch": 56.334841628959275,
+      "grad_norm": 1.2520074844360352,
+      "learning_rate": 0.001,
+      "loss": 0.6673,
+      "step": 174300
+    },
+    {
+      "epoch": 56.36716224951519,
+      "grad_norm": 1.3908581733703613,
+      "learning_rate": 0.001,
+      "loss": 0.6668,
+      "step": 174400
+    },
+    {
+      "epoch": 56.399482870071104,
+      "grad_norm": 1.0985134840011597,
+      "learning_rate": 0.001,
+      "loss": 0.6863,
+      "step": 174500
+    },
+    {
+      "epoch": 56.43180349062702,
+      "grad_norm": 1.246180772781372,
+      "learning_rate": 0.001,
+      "loss": 0.6887,
+      "step": 174600
+    },
+    {
+      "epoch": 56.46412411118293,
+      "grad_norm": 1.1640903949737549,
+      "learning_rate": 0.001,
+      "loss": 0.6972,
+      "step": 174700
+    },
+    {
+      "epoch": 56.49644473173885,
+      "grad_norm": 1.4162952899932861,
+      "learning_rate": 0.001,
+      "loss": 0.7061,
+      "step": 174800
+    },
+    {
+      "epoch": 56.52876535229476,
+      "grad_norm": 1.4577606916427612,
+      "learning_rate": 0.001,
+      "loss": 0.6915,
+      "step": 174900
+    },
+    {
+      "epoch": 56.56108597285068,
+      "grad_norm": 1.6326485872268677,
+      "learning_rate": 0.001,
+      "loss": 0.69,
+      "step": 175000
+    },
+    {
+      "epoch": 56.59340659340659,
+      "grad_norm": 1.6330126523971558,
+      "learning_rate": 0.001,
+      "loss": 0.698,
+      "step": 175100
+    },
+    {
+      "epoch": 56.625727213962506,
+      "grad_norm": 1.5364458560943604,
+      "learning_rate": 0.001,
+      "loss": 0.6766,
+      "step": 175200
+    },
+    {
+      "epoch": 56.65804783451842,
+      "grad_norm": 1.5034723281860352,
+      "learning_rate": 0.001,
+      "loss": 0.702,
+      "step": 175300
+    },
+    {
+      "epoch": 56.690368455074335,
+      "grad_norm": 1.1095960140228271,
+      "learning_rate": 0.001,
+      "loss": 0.7051,
+      "step": 175400
+    },
+    {
+      "epoch": 56.72268907563025,
+      "grad_norm": 1.692766785621643,
+      "learning_rate": 0.001,
+      "loss": 0.711,
+      "step": 175500
+    },
+    {
+      "epoch": 56.755009696186164,
+      "grad_norm": 1.2210887670516968,
+      "learning_rate": 0.001,
+      "loss": 0.6972,
+      "step": 175600
+    },
+    {
+      "epoch": 56.78733031674208,
+      "grad_norm": 1.6064265966415405,
+      "learning_rate": 0.001,
+      "loss": 0.7076,
+      "step": 175700
+    },
+    {
+      "epoch": 56.81965093729799,
+      "grad_norm": 1.4132839441299438,
+      "learning_rate": 0.001,
+      "loss": 0.7206,
+      "step": 175800
+    },
+    {
+      "epoch": 56.85197155785391,
+      "grad_norm": 1.3864809274673462,
+      "learning_rate": 0.001,
+      "loss": 0.7339,
+      "step": 175900
+    },
+    {
+      "epoch": 56.88429217840982,
+      "grad_norm": 1.4417805671691895,
+      "learning_rate": 0.001,
+      "loss": 0.7262,
+      "step": 176000
+    },
+    {
+      "epoch": 56.91661279896574,
+      "grad_norm": 1.3630573749542236,
+      "learning_rate": 0.001,
+      "loss": 0.7211,
+      "step": 176100
+    },
+    {
+      "epoch": 56.94893341952165,
+      "grad_norm": 1.3025026321411133,
+      "learning_rate": 0.001,
+      "loss": 0.7153,
+      "step": 176200
+    },
+    {
+      "epoch": 56.981254040077566,
+      "grad_norm": 1.1699614524841309,
+      "learning_rate": 0.001,
+      "loss": 0.7156,
+      "step": 176300
+    },
+    {
+      "epoch": 57.01357466063349,
+      "grad_norm": 1.3091782331466675,
+      "learning_rate": 0.001,
+      "loss": 0.6699,
+      "step": 176400
+    },
+    {
+      "epoch": 57.0458952811894,
+      "grad_norm": 1.0730043649673462,
+      "learning_rate": 0.001,
+      "loss": 0.6543,
+      "step": 176500
+    },
+    {
+      "epoch": 57.07821590174532,
+      "grad_norm": 1.6480448246002197,
+      "learning_rate": 0.001,
+      "loss": 0.6499,
+      "step": 176600
+    },
+    {
+      "epoch": 57.11053652230123,
+      "grad_norm": 1.1549612283706665,
+      "learning_rate": 0.001,
+      "loss": 0.6457,
+      "step": 176700
+    },
+    {
+      "epoch": 57.142857142857146,
+      "grad_norm": 0.9740864634513855,
+      "learning_rate": 0.001,
+      "loss": 0.6583,
+      "step": 176800
+    },
+    {
+      "epoch": 57.17517776341306,
+      "grad_norm": 1.0623377561569214,
+      "learning_rate": 0.001,
+      "loss": 0.6563,
+      "step": 176900
+    },
+    {
+      "epoch": 57.207498383968975,
+      "grad_norm": 1.1579992771148682,
+      "learning_rate": 0.001,
+      "loss": 0.6723,
+      "step": 177000
+    },
+    {
+      "epoch": 57.23981900452489,
+      "grad_norm": 1.4379750490188599,
+      "learning_rate": 0.001,
+      "loss": 0.6672,
+      "step": 177100
+    },
+    {
+      "epoch": 57.272139625080804,
+      "grad_norm": 1.191053032875061,
+      "learning_rate": 0.001,
+      "loss": 0.6594,
+      "step": 177200
+    },
+    {
+      "epoch": 57.30446024563672,
+      "grad_norm": 1.3507729768753052,
+      "learning_rate": 0.001,
+      "loss": 0.6797,
+      "step": 177300
+    },
+    {
+      "epoch": 57.33678086619263,
+      "grad_norm": 1.240713119506836,
+      "learning_rate": 0.001,
+      "loss": 0.6605,
+      "step": 177400
+    },
+    {
+      "epoch": 57.36910148674855,
+      "grad_norm": 1.1125223636627197,
+      "learning_rate": 0.001,
+      "loss": 0.6766,
+      "step": 177500
+    },
+    {
+      "epoch": 57.40142210730446,
+      "grad_norm": 1.2132062911987305,
+      "learning_rate": 0.001,
+      "loss": 0.6815,
+      "step": 177600
+    },
+    {
+      "epoch": 57.43374272786038,
+      "grad_norm": 1.2277776002883911,
+      "learning_rate": 0.001,
+      "loss": 0.684,
+      "step": 177700
+    },
+    {
+      "epoch": 57.46606334841629,
+      "grad_norm": 1.0552458763122559,
+      "learning_rate": 0.001,
+      "loss": 0.6758,
+      "step": 177800
+    },
+    {
+      "epoch": 57.498383968972206,
+      "grad_norm": 1.1401628255844116,
+      "learning_rate": 0.001,
+      "loss": 0.685,
+      "step": 177900
+    },
+    {
+      "epoch": 57.53070458952812,
+      "grad_norm": 1.225094199180603,
+      "learning_rate": 0.001,
+      "loss": 0.6842,
+      "step": 178000
+    },
+    {
+      "epoch": 57.563025210084035,
+      "grad_norm": 1.2406818866729736,
+      "learning_rate": 0.001,
+      "loss": 0.695,
+      "step": 178100
+    },
+    {
+      "epoch": 57.59534583063995,
+      "grad_norm": 0.9722842574119568,
+      "learning_rate": 0.001,
+      "loss": 0.6877,
+      "step": 178200
+    },
+    {
+      "epoch": 57.627666451195864,
+      "grad_norm": 1.3157597780227661,
+      "learning_rate": 0.001,
+      "loss": 0.7033,
+      "step": 178300
+    },
+    {
+      "epoch": 57.65998707175178,
+      "grad_norm": 1.02400803565979,
+      "learning_rate": 0.001,
+      "loss": 0.6946,
+      "step": 178400
+    },
+    {
+      "epoch": 57.69230769230769,
+      "grad_norm": 1.0910509824752808,
+      "learning_rate": 0.001,
+      "loss": 0.7019,
+      "step": 178500
+    },
+    {
+      "epoch": 57.72462831286361,
+      "grad_norm": 1.3276338577270508,
+      "learning_rate": 0.001,
+      "loss": 0.7062,
+      "step": 178600
+    },
+    {
+      "epoch": 57.75694893341952,
+      "grad_norm": 1.2191648483276367,
+      "learning_rate": 0.001,
+      "loss": 0.6972,
+      "step": 178700
+    },
+    {
+      "epoch": 57.78926955397544,
+      "grad_norm": 1.43450129032135,
+      "learning_rate": 0.001,
+      "loss": 0.706,
+      "step": 178800
+    },
+    {
+      "epoch": 57.82159017453135,
+      "grad_norm": 1.2280073165893555,
+      "learning_rate": 0.001,
+      "loss": 0.6939,
+      "step": 178900
+    },
+    {
+      "epoch": 57.853910795087266,
+      "grad_norm": 1.187094807624817,
+      "learning_rate": 0.001,
+      "loss": 0.6895,
+      "step": 179000
+    },
+    {
+      "epoch": 57.88623141564318,
+      "grad_norm": 1.1127578020095825,
+      "learning_rate": 0.001,
+      "loss": 0.709,
+      "step": 179100
+    },
+    {
+      "epoch": 57.918552036199095,
+      "grad_norm": 1.1391856670379639,
+      "learning_rate": 0.001,
+      "loss": 0.7105,
+      "step": 179200
+    },
+    {
+      "epoch": 57.95087265675501,
+      "grad_norm": 1.3863335847854614,
+      "learning_rate": 0.001,
+      "loss": 0.721,
+      "step": 179300
+    },
+    {
+      "epoch": 57.983193277310924,
+      "grad_norm": 0.9819924235343933,
+      "learning_rate": 0.001,
+      "loss": 0.7063,
+      "step": 179400
+    },
+    {
+      "epoch": 58.01551389786684,
+      "grad_norm": 0.818321168422699,
+      "learning_rate": 0.001,
+      "loss": 0.6418,
+      "step": 179500
+    },
+    {
+      "epoch": 58.04783451842275,
+      "grad_norm": 0.7625473737716675,
+      "learning_rate": 0.001,
+      "loss": 0.6364,
+      "step": 179600
+    },
+    {
+      "epoch": 58.08015513897867,
+      "grad_norm": 0.5951278209686279,
+      "learning_rate": 0.001,
+      "loss": 0.6252,
+      "step": 179700
+    },
+    {
+      "epoch": 58.11247575953458,
+      "grad_norm": 0.5998912453651428,
+      "learning_rate": 0.001,
+      "loss": 0.6378,
+      "step": 179800
+    },
+    {
+      "epoch": 58.1447963800905,
+      "grad_norm": 1.1031502485275269,
+      "learning_rate": 0.001,
+      "loss": 0.6487,
+      "step": 179900
+    },
+    {
+      "epoch": 58.17711700064641,
+      "grad_norm": 0.686004102230072,
+      "learning_rate": 0.001,
+      "loss": 0.6581,
+      "step": 180000
+    },
+    {
+      "epoch": 58.209437621202326,
+      "grad_norm": 0.9831950068473816,
+      "learning_rate": 0.001,
+      "loss": 0.6453,
+      "step": 180100
+    },
+    {
+      "epoch": 58.24175824175824,
+      "grad_norm": 0.9751262664794922,
+      "learning_rate": 0.001,
+      "loss": 0.6528,
+      "step": 180200
+    },
+    {
+      "epoch": 58.274078862314155,
+      "grad_norm": 0.9873791337013245,
+      "learning_rate": 0.001,
+      "loss": 0.6597,
+      "step": 180300
+    },
+    {
+      "epoch": 58.30639948287007,
+      "grad_norm": 0.6696386933326721,
+      "learning_rate": 0.001,
+      "loss": 0.6569,
+      "step": 180400
+    },
+    {
+      "epoch": 58.338720103425985,
+      "grad_norm": 0.8646842837333679,
+      "learning_rate": 0.001,
+      "loss": 0.6658,
+      "step": 180500
+    },
+    {
+      "epoch": 58.3710407239819,
+      "grad_norm": 0.4439694881439209,
+      "learning_rate": 0.001,
+      "loss": 0.6511,
+      "step": 180600
+    },
+    {
+      "epoch": 58.403361344537814,
+      "grad_norm": 0.6436874270439148,
+      "learning_rate": 0.001,
+      "loss": 0.6722,
+      "step": 180700
+    },
+    {
+      "epoch": 58.43568196509373,
+      "grad_norm": 1.4539772272109985,
+      "learning_rate": 0.001,
+      "loss": 0.6739,
+      "step": 180800
+    },
+    {
+      "epoch": 58.46800258564964,
+      "grad_norm": 1.248949408531189,
+      "learning_rate": 0.001,
+      "loss": 0.6787,
+      "step": 180900
+    },
+    {
+      "epoch": 58.50032320620556,
+      "grad_norm": 0.9955918788909912,
+      "learning_rate": 0.001,
+      "loss": 0.6826,
+      "step": 181000
+    },
+    {
+      "epoch": 58.53264382676147,
+      "grad_norm": 1.4825260639190674,
+      "learning_rate": 0.001,
+      "loss": 0.68,
+      "step": 181100
+    },
+    {
+      "epoch": 58.56496444731739,
+      "grad_norm": 1.0058106184005737,
+      "learning_rate": 0.001,
+      "loss": 0.6706,
+      "step": 181200
+    },
+    {
+      "epoch": 58.5972850678733,
+      "grad_norm": 1.0586559772491455,
+      "learning_rate": 0.001,
+      "loss": 0.6747,
+      "step": 181300
+    },
+    {
+      "epoch": 58.629605688429216,
+      "grad_norm": 0.8397805094718933,
+      "learning_rate": 0.001,
+      "loss": 0.6897,
+      "step": 181400
+    },
+    {
+      "epoch": 58.66192630898513,
+      "grad_norm": 0.8589861392974854,
+      "learning_rate": 0.001,
+      "loss": 0.6802,
+      "step": 181500
+    },
+    {
+      "epoch": 58.694246929541045,
+      "grad_norm": 0.6333179473876953,
+      "learning_rate": 0.001,
+      "loss": 0.7055,
+      "step": 181600
+    },
+    {
+      "epoch": 58.72656755009696,
+      "grad_norm": 0.7798647284507751,
+      "learning_rate": 0.001,
+      "loss": 0.6974,
+      "step": 181700
+    },
+    {
+      "epoch": 58.758888170652874,
+      "grad_norm": 1.4674975872039795,
+      "learning_rate": 0.001,
+      "loss": 0.6912,
+      "step": 181800
+    },
+    {
+      "epoch": 58.79120879120879,
+      "grad_norm": 0.7736033201217651,
+      "learning_rate": 0.001,
+      "loss": 0.6948,
+      "step": 181900
+    },
+    {
+      "epoch": 58.8235294117647,
+      "grad_norm": 0.7464256882667542,
+      "learning_rate": 0.001,
+      "loss": 0.6924,
+      "step": 182000
+    },
+    {
+      "epoch": 58.85585003232062,
+      "grad_norm": 0.839506983757019,
+      "learning_rate": 0.001,
+      "loss": 0.6994,
+      "step": 182100
+    },
+    {
+      "epoch": 58.88817065287653,
+      "grad_norm": 0.6937113404273987,
+      "learning_rate": 0.001,
+      "loss": 0.6914,
+      "step": 182200
+    },
+    {
+      "epoch": 58.92049127343245,
+      "grad_norm": 0.8933976888656616,
+      "learning_rate": 0.001,
+      "loss": 0.715,
+      "step": 182300
+    },
+    {
+      "epoch": 58.95281189398836,
+      "grad_norm": 0.8206430673599243,
+      "learning_rate": 0.001,
+      "loss": 0.707,
+      "step": 182400
+    },
+    {
+      "epoch": 58.985132514544276,
+      "grad_norm": 0.863739013671875,
+      "learning_rate": 0.001,
+      "loss": 0.7061,
+      "step": 182500
+    },
+    {
+      "epoch": 59.0174531351002,
+      "grad_norm": 1.6361074447631836,
+      "learning_rate": 0.001,
+      "loss": 0.6643,
+      "step": 182600
+    },
+    {
+      "epoch": 59.04977375565611,
+      "grad_norm": 1.581034779548645,
+      "learning_rate": 0.001,
+      "loss": 0.6275,
+      "step": 182700
+    },
+    {
+      "epoch": 59.08209437621203,
+      "grad_norm": 1.680957317352295,
+      "learning_rate": 0.001,
+      "loss": 0.6349,
+      "step": 182800
+    },
+    {
+      "epoch": 59.11441499676794,
+      "grad_norm": 1.7037603855133057,
+      "learning_rate": 0.001,
+      "loss": 0.634,
+      "step": 182900
+    },
+    {
+      "epoch": 59.146735617323856,
+      "grad_norm": 1.5957320928573608,
+      "learning_rate": 0.001,
+      "loss": 0.6395,
+      "step": 183000
+    },
+    {
+      "epoch": 59.17905623787977,
+      "grad_norm": 1.8327863216400146,
+      "learning_rate": 0.001,
+      "loss": 0.6531,
+      "step": 183100
+    },
+    {
+      "epoch": 59.211376858435685,
+      "grad_norm": 1.845932960510254,
+      "learning_rate": 0.001,
+      "loss": 0.6402,
+      "step": 183200
+    },
+    {
+      "epoch": 59.2436974789916,
+      "grad_norm": 1.7632895708084106,
+      "learning_rate": 0.001,
+      "loss": 0.6557,
+      "step": 183300
+    },
+    {
+      "epoch": 59.276018099547514,
+      "grad_norm": 1.6277376413345337,
+      "learning_rate": 0.001,
+      "loss": 0.6491,
+      "step": 183400
+    },
+    {
+      "epoch": 59.30833872010343,
+      "grad_norm": 1.7220267057418823,
+      "learning_rate": 0.001,
+      "loss": 0.6503,
+      "step": 183500
+    },
+    {
+      "epoch": 59.34065934065934,
+      "grad_norm": 1.6623505353927612,
+      "learning_rate": 0.001,
+      "loss": 0.6621,
+      "step": 183600
+    },
+    {
+      "epoch": 59.37297996121526,
+      "grad_norm": 1.4514002799987793,
+      "learning_rate": 0.001,
+      "loss": 0.6541,
+      "step": 183700
+    },
+    {
+      "epoch": 59.40530058177117,
+      "grad_norm": 1.77423095703125,
+      "learning_rate": 0.001,
+      "loss": 0.6583,
+      "step": 183800
+    },
+    {
+      "epoch": 59.43762120232709,
+      "grad_norm": 1.649661898612976,
+      "learning_rate": 0.001,
+      "loss": 0.6752,
+      "step": 183900
+    },
+    {
+      "epoch": 59.469941822883,
+      "grad_norm": 1.8120254278182983,
+      "learning_rate": 0.001,
+      "loss": 0.6613,
+      "step": 184000
+    },
+    {
+      "epoch": 59.502262443438916,
+      "grad_norm": 1.68980872631073,
+      "learning_rate": 0.001,
+      "loss": 0.6742,
+      "step": 184100
+    },
+    {
+      "epoch": 59.53458306399483,
+      "grad_norm": 1.6394644975662231,
+      "learning_rate": 0.001,
+      "loss": 0.6718,
+      "step": 184200
+    },
+    {
+      "epoch": 59.566903684550745,
+      "grad_norm": 1.7204967737197876,
+      "learning_rate": 0.001,
+      "loss": 0.6731,
+      "step": 184300
+    },
+    {
+      "epoch": 59.59922430510666,
+      "grad_norm": 1.4989378452301025,
+      "learning_rate": 0.001,
+      "loss": 0.6842,
+      "step": 184400
+    },
+    {
+      "epoch": 59.631544925662574,
+      "grad_norm": 1.6579501628875732,
+      "learning_rate": 0.001,
+      "loss": 0.6813,
+      "step": 184500
+    },
+    {
+      "epoch": 59.66386554621849,
+      "grad_norm": 1.2888191938400269,
+      "learning_rate": 0.001,
+      "loss": 0.6843,
+      "step": 184600
+    },
+    {
+      "epoch": 59.6961861667744,
+      "grad_norm": 1.838328242301941,
+      "learning_rate": 0.001,
+      "loss": 0.6749,
+      "step": 184700
+    },
+    {
+      "epoch": 59.72850678733032,
+      "grad_norm": 1.8516156673431396,
+      "learning_rate": 0.001,
+      "loss": 0.6793,
+      "step": 184800
+    },
+    {
+      "epoch": 59.76082740788623,
+      "grad_norm": 1.3773984909057617,
+      "learning_rate": 0.001,
+      "loss": 0.6932,
+      "step": 184900
+    },
+    {
+      "epoch": 59.79314802844215,
+      "grad_norm": 1.4964762926101685,
+      "learning_rate": 0.001,
+      "loss": 0.6755,
+      "step": 185000
+    },
+    {
+      "epoch": 59.82546864899806,
+      "grad_norm": 1.436951994895935,
+      "learning_rate": 0.001,
+      "loss": 0.6887,
+      "step": 185100
+    },
+    {
+      "epoch": 59.857789269553976,
+      "grad_norm": 1.999688982963562,
+      "learning_rate": 0.001,
+      "loss": 0.6755,
+      "step": 185200
+    },
+    {
+      "epoch": 59.89010989010989,
+      "grad_norm": 2.138962507247925,
+      "learning_rate": 0.001,
+      "loss": 0.6728,
+      "step": 185300
+    },
+    {
+      "epoch": 59.922430510665805,
+      "grad_norm": 1.625344157218933,
+      "learning_rate": 0.001,
+      "loss": 0.6871,
+      "step": 185400
+    },
+    {
+      "epoch": 59.95475113122172,
+      "grad_norm": 1.556227445602417,
+      "learning_rate": 0.001,
+      "loss": 0.692,
+      "step": 185500
+    },
+    {
+      "epoch": 59.987071751777634,
+      "grad_norm": 1.5213489532470703,
+      "learning_rate": 0.001,
+      "loss": 0.6942,
+      "step": 185600
+    },
+    {
+      "epoch": 60.01939237233355,
+      "grad_norm": 1.5821621417999268,
+      "learning_rate": 0.001,
+      "loss": 0.6547,
+      "step": 185700
+    },
+    {
+      "epoch": 60.05171299288946,
+      "grad_norm": 1.2159498929977417,
+      "learning_rate": 0.001,
+      "loss": 0.6171,
+      "step": 185800
+    },
+    {
+      "epoch": 60.08403361344538,
+      "grad_norm": 1.3063063621520996,
+      "learning_rate": 0.001,
+      "loss": 0.6401,
+      "step": 185900
+    },
+    {
+      "epoch": 60.11635423400129,
+      "grad_norm": 1.501042127609253,
+      "learning_rate": 0.001,
+      "loss": 0.6274,
+      "step": 186000
+    },
+    {
+      "epoch": 60.14867485455721,
+      "grad_norm": 1.4862663745880127,
+      "learning_rate": 0.001,
+      "loss": 0.6215,
+      "step": 186100
+    },
+    {
+      "epoch": 60.18099547511312,
+      "grad_norm": 1.3393819332122803,
+      "learning_rate": 0.001,
+      "loss": 0.6366,
+      "step": 186200
+    },
+    {
+      "epoch": 60.213316095669036,
+      "grad_norm": 1.5676498413085938,
+      "learning_rate": 0.001,
+      "loss": 0.6364,
+      "step": 186300
+    },
+    {
+      "epoch": 60.24563671622495,
+      "grad_norm": 1.6434075832366943,
+      "learning_rate": 0.001,
+      "loss": 0.6319,
+      "step": 186400
+    },
+    {
+      "epoch": 60.277957336780865,
+      "grad_norm": 1.2878286838531494,
+      "learning_rate": 0.001,
+      "loss": 0.6592,
+      "step": 186500
+    },
+    {
+      "epoch": 60.31027795733678,
+      "grad_norm": 1.3613319396972656,
+      "learning_rate": 0.001,
+      "loss": 0.647,
+      "step": 186600
+    },
+    {
+      "epoch": 60.342598577892694,
+      "grad_norm": 1.654203176498413,
+      "learning_rate": 0.001,
+      "loss": 0.6508,
+      "step": 186700
+    },
+    {
+      "epoch": 60.37491919844861,
+      "grad_norm": 1.6589421033859253,
+      "learning_rate": 0.001,
+      "loss": 0.6401,
+      "step": 186800
+    },
+    {
+      "epoch": 60.40723981900452,
+      "grad_norm": 1.7555818557739258,
+      "learning_rate": 0.001,
+      "loss": 0.6528,
+      "step": 186900
+    },
+    {
+      "epoch": 60.43956043956044,
+      "grad_norm": 1.5702903270721436,
+      "learning_rate": 0.001,
+      "loss": 0.6553,
+      "step": 187000
+    },
+    {
+      "epoch": 60.47188106011635,
+      "grad_norm": 1.3548332452774048,
+      "learning_rate": 0.001,
+      "loss": 0.6411,
+      "step": 187100
+    },
+    {
+      "epoch": 60.50420168067227,
+      "grad_norm": 1.6495980024337769,
+      "learning_rate": 0.001,
+      "loss": 0.6673,
+      "step": 187200
+    },
+    {
+      "epoch": 60.53652230122818,
+      "grad_norm": 1.319981336593628,
+      "learning_rate": 0.001,
+      "loss": 0.6493,
+      "step": 187300
+    },
+    {
+      "epoch": 60.568842921784096,
+      "grad_norm": 1.2345103025436401,
+      "learning_rate": 0.001,
+      "loss": 0.6848,
+      "step": 187400
+    },
+    {
+      "epoch": 60.60116354234001,
+      "grad_norm": 1.2018239498138428,
+      "learning_rate": 0.001,
+      "loss": 0.6668,
+      "step": 187500
+    },
+    {
+      "epoch": 60.633484162895925,
+      "grad_norm": 1.442687749862671,
+      "learning_rate": 0.001,
+      "loss": 0.6462,
+      "step": 187600
+    },
+    {
+      "epoch": 60.66580478345184,
+      "grad_norm": 1.4780453443527222,
+      "learning_rate": 0.001,
+      "loss": 0.6695,
+      "step": 187700
+    },
+    {
+      "epoch": 60.698125404007754,
+      "grad_norm": 1.3161441087722778,
+      "learning_rate": 0.001,
+      "loss": 0.6759,
+      "step": 187800
+    },
+    {
+      "epoch": 60.73044602456367,
+      "grad_norm": 1.5663788318634033,
+      "learning_rate": 0.001,
+      "loss": 0.6783,
+      "step": 187900
+    },
+    {
+      "epoch": 60.762766645119584,
+      "grad_norm": 1.2903250455856323,
+      "learning_rate": 0.001,
+      "loss": 0.6802,
+      "step": 188000
+    },
+    {
+      "epoch": 60.7950872656755,
+      "grad_norm": 1.2310444116592407,
+      "learning_rate": 0.001,
+      "loss": 0.6798,
+      "step": 188100
+    },
+    {
+      "epoch": 60.82740788623141,
+      "grad_norm": 1.6267993450164795,
+      "learning_rate": 0.001,
+      "loss": 0.6747,
+      "step": 188200
+    },
+    {
+      "epoch": 60.85972850678733,
+      "grad_norm": 1.3514257669448853,
+      "learning_rate": 0.001,
+      "loss": 0.6954,
+      "step": 188300
+    },
+    {
+      "epoch": 60.89204912734324,
+      "grad_norm": 1.3700529336929321,
+      "learning_rate": 0.001,
+      "loss": 0.6921,
+      "step": 188400
+    },
+    {
+      "epoch": 60.924369747899156,
+      "grad_norm": 1.3592313528060913,
+      "learning_rate": 0.001,
+      "loss": 0.6949,
+      "step": 188500
+    },
+    {
+      "epoch": 60.95669036845507,
+      "grad_norm": 1.6864216327667236,
+      "learning_rate": 0.001,
+      "loss": 0.693,
+      "step": 188600
+    },
+    {
+      "epoch": 60.98901098901099,
+      "grad_norm": 1.476345181465149,
+      "learning_rate": 0.001,
+      "loss": 0.679,
+      "step": 188700
+    },
+    {
+      "epoch": 61.02133160956691,
+      "grad_norm": 1.2383800745010376,
+      "learning_rate": 0.001,
+      "loss": 0.6467,
+      "step": 188800
+    },
+    {
+      "epoch": 61.05365223012282,
+      "grad_norm": 1.6404695510864258,
+      "learning_rate": 0.001,
+      "loss": 0.617,
+      "step": 188900
+    },
+    {
+      "epoch": 61.085972850678736,
+      "grad_norm": 1.4813312292099,
+      "learning_rate": 0.001,
+      "loss": 0.612,
+      "step": 189000
+    },
+    {
+      "epoch": 61.11829347123465,
+      "grad_norm": 1.823527455329895,
+      "learning_rate": 0.001,
+      "loss": 0.6325,
+      "step": 189100
+    },
+    {
+      "epoch": 61.150614091790565,
+      "grad_norm": 1.2506108283996582,
+      "learning_rate": 0.001,
+      "loss": 0.6095,
+      "step": 189200
+    },
+    {
+      "epoch": 61.18293471234648,
+      "grad_norm": 1.376138687133789,
+      "learning_rate": 0.001,
+      "loss": 0.6224,
+      "step": 189300
+    },
+    {
+      "epoch": 61.215255332902395,
+      "grad_norm": 1.1488109827041626,
+      "learning_rate": 0.001,
+      "loss": 0.6249,
+      "step": 189400
+    },
+    {
+      "epoch": 61.24757595345831,
+      "grad_norm": 1.6473678350448608,
+      "learning_rate": 0.001,
+      "loss": 0.6276,
+      "step": 189500
+    },
+    {
+      "epoch": 61.279896574014224,
+      "grad_norm": 1.2410812377929688,
+      "learning_rate": 0.001,
+      "loss": 0.6226,
+      "step": 189600
+    },
+    {
+      "epoch": 61.31221719457014,
+      "grad_norm": 10.854134559631348,
+      "learning_rate": 0.001,
+      "loss": 0.652,
+      "step": 189700
+    },
+    {
+      "epoch": 61.34453781512605,
+      "grad_norm": 2.2806782722473145,
+      "learning_rate": 0.001,
+      "loss": 0.6421,
+      "step": 189800
+    },
+    {
+      "epoch": 61.37685843568197,
+      "grad_norm": 1.4156832695007324,
+      "learning_rate": 0.001,
+      "loss": 0.6338,
+      "step": 189900
+    },
+    {
+      "epoch": 61.40917905623788,
+      "grad_norm": 1.337144374847412,
+      "learning_rate": 0.001,
+      "loss": 0.65,
+      "step": 190000
+    },
+    {
+      "epoch": 61.441499676793796,
+      "grad_norm": 1.3378078937530518,
+      "learning_rate": 0.001,
+      "loss": 0.6391,
+      "step": 190100
+    },
+    {
+      "epoch": 61.47382029734971,
+      "grad_norm": 1.5442296266555786,
+      "learning_rate": 0.001,
+      "loss": 0.6588,
+      "step": 190200
+    },
+    {
+      "epoch": 61.506140917905626,
+      "grad_norm": 1.5696725845336914,
+      "learning_rate": 0.001,
+      "loss": 0.6621,
+      "step": 190300
+    },
+    {
+      "epoch": 61.53846153846154,
+      "grad_norm": 1.3986225128173828,
+      "learning_rate": 0.001,
+      "loss": 0.6463,
+      "step": 190400
+    },
+    {
+      "epoch": 61.570782159017455,
+      "grad_norm": 1.4197642803192139,
+      "learning_rate": 0.001,
+      "loss": 0.6643,
+      "step": 190500
+    },
+    {
+      "epoch": 61.60310277957337,
+      "grad_norm": 1.4512290954589844,
+      "learning_rate": 0.001,
+      "loss": 0.6533,
+      "step": 190600
+    },
+    {
+      "epoch": 61.635423400129284,
+      "grad_norm": 1.717221975326538,
+      "learning_rate": 0.001,
+      "loss": 0.6638,
+      "step": 190700
+    },
+    {
+      "epoch": 61.6677440206852,
+      "grad_norm": 1.4332921504974365,
+      "learning_rate": 0.001,
+      "loss": 0.6626,
+      "step": 190800
+    },
+    {
+      "epoch": 61.70006464124111,
+      "grad_norm": 1.781604290008545,
+      "learning_rate": 0.001,
+      "loss": 0.6637,
+      "step": 190900
+    },
+    {
+      "epoch": 61.73238526179703,
+      "grad_norm": 1.6162261962890625,
+      "learning_rate": 0.001,
+      "loss": 0.6542,
+      "step": 191000
+    },
+    {
+      "epoch": 61.76470588235294,
+      "grad_norm": 1.3693976402282715,
+      "learning_rate": 0.001,
+      "loss": 0.6713,
+      "step": 191100
+    },
+    {
+      "epoch": 61.79702650290886,
+      "grad_norm": 1.339870810508728,
+      "learning_rate": 0.001,
+      "loss": 0.667,
+      "step": 191200
+    },
+    {
+      "epoch": 61.82934712346477,
+      "grad_norm": 1.6282362937927246,
+      "learning_rate": 0.001,
+      "loss": 0.677,
+      "step": 191300
+    },
+    {
+      "epoch": 61.861667744020686,
+      "grad_norm": 1.3749769926071167,
+      "learning_rate": 0.001,
+      "loss": 0.6788,
+      "step": 191400
+    },
+    {
+      "epoch": 61.8939883645766,
+      "grad_norm": 1.6163561344146729,
+      "learning_rate": 0.001,
+      "loss": 0.6856,
+      "step": 191500
+    },
+    {
+      "epoch": 61.926308985132515,
+      "grad_norm": 1.7528187036514282,
+      "learning_rate": 0.001,
+      "loss": 0.6811,
+      "step": 191600
+    },
+    {
+      "epoch": 61.95862960568843,
+      "grad_norm": 1.3337574005126953,
+      "learning_rate": 0.001,
+      "loss": 0.6813,
+      "step": 191700
+    },
+    {
+      "epoch": 61.990950226244344,
+      "grad_norm": 1.4761725664138794,
+      "learning_rate": 0.001,
+      "loss": 0.688,
+      "step": 191800
+    },
+    {
+      "epoch": 62.02327084680026,
+      "grad_norm": 1.1570292711257935,
+      "learning_rate": 0.001,
+      "loss": 0.634,
+      "step": 191900
+    },
+    {
+      "epoch": 62.05559146735617,
+      "grad_norm": 1.2674442529678345,
+      "learning_rate": 0.001,
+      "loss": 0.6063,
+      "step": 192000
+    },
+    {
+      "epoch": 62.08791208791209,
+      "grad_norm": 1.3792743682861328,
+      "learning_rate": 0.001,
+      "loss": 0.6035,
+      "step": 192100
+    },
+    {
+      "epoch": 62.120232708468,
+      "grad_norm": 1.5314244031906128,
+      "learning_rate": 0.001,
+      "loss": 0.6136,
+      "step": 192200
+    },
+    {
+      "epoch": 62.15255332902392,
+      "grad_norm": 1.6034588813781738,
+      "learning_rate": 0.001,
+      "loss": 0.6097,
+      "step": 192300
+    },
+    {
+      "epoch": 62.18487394957983,
+      "grad_norm": 1.53525972366333,
+      "learning_rate": 0.001,
+      "loss": 0.6319,
+      "step": 192400
+    },
+    {
+      "epoch": 62.217194570135746,
+      "grad_norm": 1.1694210767745972,
+      "learning_rate": 0.001,
+      "loss": 0.6177,
+      "step": 192500
+    },
+    {
+      "epoch": 62.24951519069166,
+      "grad_norm": 1.1147392988204956,
+      "learning_rate": 0.001,
+      "loss": 0.6296,
+      "step": 192600
+    },
+    {
+      "epoch": 62.281835811247575,
+      "grad_norm": 2.554081916809082,
+      "learning_rate": 0.001,
+      "loss": 0.6303,
+      "step": 192700
+    },
+    {
+      "epoch": 62.31415643180349,
+      "grad_norm": 1.8635549545288086,
+      "learning_rate": 0.001,
+      "loss": 0.6389,
+      "step": 192800
+    },
+    {
+      "epoch": 62.346477052359404,
+      "grad_norm": 1.2993345260620117,
+      "learning_rate": 0.001,
+      "loss": 0.6266,
+      "step": 192900
+    },
+    {
+      "epoch": 62.37879767291532,
+      "grad_norm": 1.4036643505096436,
+      "learning_rate": 0.001,
+      "loss": 0.6352,
+      "step": 193000
+    },
+    {
+      "epoch": 62.41111829347123,
+      "grad_norm": 1.7309150695800781,
+      "learning_rate": 0.001,
+      "loss": 0.6386,
+      "step": 193100
+    },
+    {
+      "epoch": 62.44343891402715,
+      "grad_norm": 1.4345393180847168,
+      "learning_rate": 0.001,
+      "loss": 0.6329,
+      "step": 193200
+    },
+    {
+      "epoch": 62.47575953458306,
+      "grad_norm": 1.1931675672531128,
+      "learning_rate": 0.001,
+      "loss": 0.6399,
+      "step": 193300
+    },
+    {
+      "epoch": 62.50808015513898,
+      "grad_norm": 1.2835602760314941,
+      "learning_rate": 0.001,
+      "loss": 0.636,
+      "step": 193400
+    },
+    {
+      "epoch": 62.54040077569489,
+      "grad_norm": 1.2109442949295044,
+      "learning_rate": 0.001,
+      "loss": 0.6593,
+      "step": 193500
+    },
+    {
+      "epoch": 62.572721396250806,
+      "grad_norm": 1.3101767301559448,
+      "learning_rate": 0.001,
+      "loss": 0.6553,
+      "step": 193600
+    },
+    {
+      "epoch": 62.60504201680672,
+      "grad_norm": 1.3763705492019653,
+      "learning_rate": 0.001,
+      "loss": 0.6493,
+      "step": 193700
+    },
+    {
+      "epoch": 62.637362637362635,
+      "grad_norm": 1.2936313152313232,
+      "learning_rate": 0.001,
+      "loss": 0.6668,
+      "step": 193800
+    },
+    {
+      "epoch": 62.66968325791855,
+      "grad_norm": 1.3182551860809326,
+      "learning_rate": 0.001,
+      "loss": 0.6578,
+      "step": 193900
+    },
+    {
+      "epoch": 62.702003878474464,
+      "grad_norm": 1.171347737312317,
+      "learning_rate": 0.001,
+      "loss": 0.6648,
+      "step": 194000
+    },
+    {
+      "epoch": 62.73432449903038,
+      "grad_norm": 1.108046293258667,
+      "learning_rate": 0.001,
+      "loss": 0.6601,
+      "step": 194100
+    },
+    {
+      "epoch": 62.76664511958629,
+      "grad_norm": 1.2180066108703613,
+      "learning_rate": 0.001,
+      "loss": 0.662,
+      "step": 194200
+    },
+    {
+      "epoch": 62.79896574014221,
+      "grad_norm": 1.196394443511963,
+      "learning_rate": 0.001,
+      "loss": 0.6833,
+      "step": 194300
+    },
+    {
+      "epoch": 62.83128636069812,
+      "grad_norm": 1.340371012687683,
+      "learning_rate": 0.001,
+      "loss": 0.6584,
+      "step": 194400
+    },
+    {
+      "epoch": 62.86360698125404,
+      "grad_norm": 1.432305932044983,
+      "learning_rate": 0.001,
+      "loss": 0.6577,
+      "step": 194500
+    },
+    {
+      "epoch": 62.89592760180995,
+      "grad_norm": 1.041042447090149,
+      "learning_rate": 0.001,
+      "loss": 0.6583,
+      "step": 194600
+    },
+    {
+      "epoch": 62.928248222365866,
+      "grad_norm": 1.2910453081130981,
+      "learning_rate": 0.001,
+      "loss": 0.6918,
+      "step": 194700
+    },
+    {
+      "epoch": 62.96056884292178,
+      "grad_norm": 1.4799001216888428,
+      "learning_rate": 0.001,
+      "loss": 0.6755,
+      "step": 194800
+    },
+    {
+      "epoch": 62.992889463477695,
+      "grad_norm": 1.722066879272461,
+      "learning_rate": 0.001,
+      "loss": 0.6661,
+      "step": 194900
+    },
+    {
+      "epoch": 63.02521008403362,
+      "grad_norm": 1.3788399696350098,
+      "learning_rate": 0.001,
+      "loss": 0.6136,
+      "step": 195000
+    },
+    {
+      "epoch": 63.05753070458953,
+      "grad_norm": 1.6480493545532227,
+      "learning_rate": 0.001,
+      "loss": 0.608,
+      "step": 195100
+    },
+    {
+      "epoch": 63.089851325145446,
+      "grad_norm": 1.5709948539733887,
+      "learning_rate": 0.001,
+      "loss": 0.6036,
+      "step": 195200
+    },
+    {
+      "epoch": 63.12217194570136,
+      "grad_norm": 1.4029523134231567,
+      "learning_rate": 0.001,
+      "loss": 0.6171,
+      "step": 195300
+    },
+    {
+      "epoch": 63.154492566257275,
+      "grad_norm": 1.2338638305664062,
+      "learning_rate": 0.001,
+      "loss": 0.6006,
+      "step": 195400
+    },
+    {
+      "epoch": 63.18681318681319,
+      "grad_norm": 1.1037108898162842,
+      "learning_rate": 0.001,
+      "loss": 0.6153,
+      "step": 195500
+    },
+    {
+      "epoch": 63.219133807369104,
+      "grad_norm": 1.4087430238723755,
+      "learning_rate": 0.001,
+      "loss": 0.6155,
+      "step": 195600
+    },
+    {
+      "epoch": 63.25145442792502,
+      "grad_norm": 1.2912625074386597,
+      "learning_rate": 0.001,
+      "loss": 0.6161,
+      "step": 195700
+    },
+    {
+      "epoch": 63.28377504848093,
+      "grad_norm": 1.2002849578857422,
+      "learning_rate": 0.001,
+      "loss": 0.611,
+      "step": 195800
+    },
+    {
+      "epoch": 63.31609566903685,
+      "grad_norm": 1.1420046091079712,
+      "learning_rate": 0.001,
+      "loss": 0.6373,
+      "step": 195900
+    },
+    {
+      "epoch": 63.34841628959276,
+      "grad_norm": 1.29826819896698,
+      "learning_rate": 0.001,
+      "loss": 0.6353,
+      "step": 196000
+    },
+    {
+      "epoch": 63.38073691014868,
+      "grad_norm": 1.15346360206604,
+      "learning_rate": 0.001,
+      "loss": 0.6186,
+      "step": 196100
+    },
+    {
+      "epoch": 63.41305753070459,
+      "grad_norm": 1.4631860256195068,
+      "learning_rate": 0.001,
+      "loss": 0.633,
+      "step": 196200
+    },
+    {
+      "epoch": 63.445378151260506,
+      "grad_norm": 1.5273650884628296,
+      "learning_rate": 0.001,
+      "loss": 0.6317,
+      "step": 196300
+    },
+    {
+      "epoch": 63.47769877181642,
+      "grad_norm": 1.2980990409851074,
+      "learning_rate": 0.001,
+      "loss": 0.6256,
+      "step": 196400
+    },
+    {
+      "epoch": 63.510019392372335,
+      "grad_norm": 4.956090927124023,
+      "learning_rate": 0.001,
+      "loss": 0.6367,
+      "step": 196500
+    },
+    {
+      "epoch": 63.54234001292825,
+      "grad_norm": 1.180713176727295,
+      "learning_rate": 0.001,
+      "loss": 0.6315,
+      "step": 196600
+    },
+    {
+      "epoch": 63.574660633484164,
+      "grad_norm": 1.9830008745193481,
+      "learning_rate": 0.001,
+      "loss": 0.6323,
+      "step": 196700
+    },
+    {
+      "epoch": 63.60698125404008,
+      "grad_norm": 1.653559923171997,
+      "learning_rate": 0.001,
+      "loss": 0.6339,
+      "step": 196800
+    },
+    {
+      "epoch": 63.63930187459599,
+      "grad_norm": 1.3870145082473755,
+      "learning_rate": 0.001,
+      "loss": 0.6497,
+      "step": 196900
+    },
+    {
+      "epoch": 63.67162249515191,
+      "grad_norm": 1.1964478492736816,
+      "learning_rate": 0.001,
+      "loss": 0.6457,
+      "step": 197000
+    },
+    {
+      "epoch": 63.70394311570782,
+      "grad_norm": 1.1363627910614014,
+      "learning_rate": 0.001,
+      "loss": 0.6566,
+      "step": 197100
+    },
+    {
+      "epoch": 63.73626373626374,
+      "grad_norm": 1.4076826572418213,
+      "learning_rate": 0.001,
+      "loss": 0.6512,
+      "step": 197200
+    },
+    {
+      "epoch": 63.76858435681965,
+      "grad_norm": 1.313567876815796,
+      "learning_rate": 0.001,
+      "loss": 0.6445,
+      "step": 197300
+    },
+    {
+      "epoch": 63.800904977375566,
+      "grad_norm": 1.2769737243652344,
+      "learning_rate": 0.001,
+      "loss": 0.651,
+      "step": 197400
+    },
+    {
+      "epoch": 63.83322559793148,
+      "grad_norm": 1.3601458072662354,
+      "learning_rate": 0.001,
+      "loss": 0.6534,
+      "step": 197500
+    },
+    {
+      "epoch": 63.865546218487395,
+      "grad_norm": 1.2652206420898438,
+      "learning_rate": 0.001,
+      "loss": 0.6746,
+      "step": 197600
+    },
+    {
+      "epoch": 63.89786683904331,
+      "grad_norm": 1.2474119663238525,
+      "learning_rate": 0.001,
+      "loss": 0.6498,
+      "step": 197700
+    },
+    {
+      "epoch": 63.930187459599225,
+      "grad_norm": 1.1776601076126099,
+      "learning_rate": 0.001,
+      "loss": 0.6568,
+      "step": 197800
+    },
+    {
+      "epoch": 63.96250808015514,
+      "grad_norm": 1.3926881551742554,
+      "learning_rate": 0.001,
+      "loss": 0.6834,
+      "step": 197900
+    },
+    {
+      "epoch": 63.994828700711054,
+      "grad_norm": 1.4210342168807983,
+      "learning_rate": 0.001,
+      "loss": 0.6644,
+      "step": 198000
+    },
+    {
+      "epoch": 64.02714932126698,
+      "grad_norm": 1.1131622791290283,
+      "learning_rate": 0.001,
+      "loss": 0.5911,
+      "step": 198100
+    },
+    {
+      "epoch": 64.05946994182288,
+      "grad_norm": 1.1499837636947632,
+      "learning_rate": 0.001,
+      "loss": 0.5899,
+      "step": 198200
+    },
+    {
+      "epoch": 64.0917905623788,
+      "grad_norm": 1.542824625968933,
+      "learning_rate": 0.001,
+      "loss": 0.5949,
+      "step": 198300
+    },
+    {
+      "epoch": 64.12411118293471,
+      "grad_norm": 0.9843227863311768,
+      "learning_rate": 0.001,
+      "loss": 0.6054,
+      "step": 198400
+    },
+    {
+      "epoch": 64.15643180349063,
+      "grad_norm": 1.3814574480056763,
+      "learning_rate": 0.001,
+      "loss": 0.6014,
+      "step": 198500
+    },
+    {
+      "epoch": 64.18875242404654,
+      "grad_norm": 1.4337519407272339,
+      "learning_rate": 0.001,
+      "loss": 0.6009,
+      "step": 198600
+    },
+    {
+      "epoch": 64.22107304460246,
+      "grad_norm": 1.3197928667068481,
+      "learning_rate": 0.001,
+      "loss": 0.6177,
+      "step": 198700
+    },
+    {
+      "epoch": 64.25339366515837,
+      "grad_norm": 1.289289116859436,
+      "learning_rate": 0.001,
+      "loss": 0.6107,
+      "step": 198800
+    },
+    {
+      "epoch": 64.28571428571429,
+      "grad_norm": 1.8015815019607544,
+      "learning_rate": 0.001,
+      "loss": 0.6158,
+      "step": 198900
+    },
+    {
+      "epoch": 64.3180349062702,
+      "grad_norm": 1.9143645763397217,
+      "learning_rate": 0.001,
+      "loss": 0.6172,
+      "step": 199000
+    },
+    {
+      "epoch": 64.35035552682612,
+      "grad_norm": 1.310492753982544,
+      "learning_rate": 0.001,
+      "loss": 0.6177,
+      "step": 199100
+    },
+    {
+      "epoch": 64.38267614738203,
+      "grad_norm": 1.1836750507354736,
+      "learning_rate": 0.001,
+      "loss": 0.6297,
+      "step": 199200
+    },
+    {
+      "epoch": 64.41499676793795,
+      "grad_norm": 1.273499846458435,
+      "learning_rate": 0.001,
+      "loss": 0.6168,
+      "step": 199300
+    },
+    {
+      "epoch": 64.44731738849386,
+      "grad_norm": 1.3184856176376343,
+      "learning_rate": 0.001,
+      "loss": 0.6205,
+      "step": 199400
+    },
+    {
+      "epoch": 64.47963800904978,
+      "grad_norm": 1.370647668838501,
+      "learning_rate": 0.001,
+      "loss": 0.6336,
+      "step": 199500
+    },
+    {
+      "epoch": 64.51195862960569,
+      "grad_norm": 1.4278833866119385,
+      "learning_rate": 0.001,
+      "loss": 0.6315,
+      "step": 199600
+    },
+    {
+      "epoch": 64.54427925016161,
+      "grad_norm": 1.2103545665740967,
+      "learning_rate": 0.001,
+      "loss": 0.6377,
+      "step": 199700
+    },
+    {
+      "epoch": 64.57659987071752,
+      "grad_norm": 1.6035828590393066,
+      "learning_rate": 0.001,
+      "loss": 0.6336,
+      "step": 199800
+    },
+    {
+      "epoch": 64.60892049127344,
+      "grad_norm": 1.029595971107483,
+      "learning_rate": 0.001,
+      "loss": 0.6382,
+      "step": 199900
+    },
+    {
+      "epoch": 64.64124111182934,
+      "grad_norm": 1.3785440921783447,
+      "learning_rate": 0.001,
+      "loss": 0.6392,
+      "step": 200000
+    },
+    {
+      "epoch": 64.67356173238527,
+      "grad_norm": 1.218138337135315,
+      "learning_rate": 0.001,
+      "loss": 0.6394,
+      "step": 200100
+    },
+    {
+      "epoch": 64.70588235294117,
+      "grad_norm": 1.326785922050476,
+      "learning_rate": 0.001,
+      "loss": 0.6552,
+      "step": 200200
+    },
+    {
+      "epoch": 64.7382029734971,
+      "grad_norm": 1.4117070436477661,
+      "learning_rate": 0.001,
+      "loss": 0.6476,
+      "step": 200300
+    },
+    {
+      "epoch": 64.770523594053,
+      "grad_norm": 1.4691327810287476,
+      "learning_rate": 0.001,
+      "loss": 0.6415,
+      "step": 200400
+    },
+    {
+      "epoch": 64.80284421460892,
+      "grad_norm": 1.4258376359939575,
+      "learning_rate": 0.001,
+      "loss": 0.6622,
+      "step": 200500
+    },
+    {
+      "epoch": 64.83516483516483,
+      "grad_norm": 1.3137234449386597,
+      "learning_rate": 0.001,
+      "loss": 0.6443,
+      "step": 200600
+    },
+    {
+      "epoch": 64.86748545572075,
+      "grad_norm": 1.0861976146697998,
+      "learning_rate": 0.001,
+      "loss": 0.6563,
+      "step": 200700
+    },
+    {
+      "epoch": 64.89980607627666,
+      "grad_norm": 1.148963212966919,
+      "learning_rate": 0.001,
+      "loss": 0.6483,
+      "step": 200800
+    },
+    {
+      "epoch": 64.93212669683258,
+      "grad_norm": 1.3883782625198364,
+      "learning_rate": 0.001,
+      "loss": 0.648,
+      "step": 200900
+    },
+    {
+      "epoch": 64.96444731738849,
+      "grad_norm": 1.6844929456710815,
+      "learning_rate": 0.001,
+      "loss": 0.6661,
+      "step": 201000
+    },
+    {
+      "epoch": 64.99676793794441,
+      "grad_norm": 1.6211166381835938,
+      "learning_rate": 0.001,
+      "loss": 0.6372,
+      "step": 201100
+    },
+    {
+      "epoch": 65.02908855850032,
+      "grad_norm": 1.3715167045593262,
+      "learning_rate": 0.001,
+      "loss": 0.5833,
+      "step": 201200
+    },
+    {
+      "epoch": 65.06140917905624,
+      "grad_norm": 1.0815142393112183,
+      "learning_rate": 0.001,
+      "loss": 0.5804,
+      "step": 201300
+    },
+    {
+      "epoch": 65.09372979961215,
+      "grad_norm": 1.040615439414978,
+      "learning_rate": 0.001,
+      "loss": 0.5859,
+      "step": 201400
+    },
+    {
+      "epoch": 65.12605042016807,
+      "grad_norm": 1.36244797706604,
+      "learning_rate": 0.001,
+      "loss": 0.5935,
+      "step": 201500
+    },
+    {
+      "epoch": 65.15837104072398,
+      "grad_norm": 1.1301552057266235,
+      "learning_rate": 0.001,
+      "loss": 0.5946,
+      "step": 201600
+    },
+    {
+      "epoch": 65.1906916612799,
+      "grad_norm": 1.00641930103302,
+      "learning_rate": 0.001,
+      "loss": 0.5962,
+      "step": 201700
+    },
+    {
+      "epoch": 65.2230122818358,
+      "grad_norm": 1.5438146591186523,
+      "learning_rate": 0.001,
+      "loss": 0.6009,
+      "step": 201800
+    },
+    {
+      "epoch": 65.25533290239173,
+      "grad_norm": 1.7968852519989014,
+      "learning_rate": 0.001,
+      "loss": 0.61,
+      "step": 201900
+    },
+    {
+      "epoch": 65.28765352294764,
+      "grad_norm": 1.2997229099273682,
+      "learning_rate": 0.001,
+      "loss": 0.6055,
+      "step": 202000
+    },
+    {
+      "epoch": 65.31997414350356,
+      "grad_norm": 1.2967760562896729,
+      "learning_rate": 0.001,
+      "loss": 0.5919,
+      "step": 202100
+    },
+    {
+      "epoch": 65.35229476405947,
+      "grad_norm": 1.2664779424667358,
+      "learning_rate": 0.001,
+      "loss": 0.6116,
+      "step": 202200
+    },
+    {
+      "epoch": 65.38461538461539,
+      "grad_norm": 1.5638916492462158,
+      "learning_rate": 0.001,
+      "loss": 0.6073,
+      "step": 202300
+    },
+    {
+      "epoch": 65.4169360051713,
+      "grad_norm": 1.400978922843933,
+      "learning_rate": 0.001,
+      "loss": 0.6127,
+      "step": 202400
+    },
+    {
+      "epoch": 65.44925662572722,
+      "grad_norm": 1.990863561630249,
+      "learning_rate": 0.001,
+      "loss": 0.6285,
+      "step": 202500
+    },
+    {
+      "epoch": 65.48157724628312,
+      "grad_norm": 1.062779188156128,
+      "learning_rate": 0.001,
+      "loss": 0.6176,
+      "step": 202600
+    },
+    {
+      "epoch": 65.51389786683905,
+      "grad_norm": 1.3369722366333008,
+      "learning_rate": 0.001,
+      "loss": 0.6216,
+      "step": 202700
+    },
+    {
+      "epoch": 65.54621848739495,
+      "grad_norm": 1.756347417831421,
+      "learning_rate": 0.001,
+      "loss": 0.6291,
+      "step": 202800
+    },
+    {
+      "epoch": 65.57853910795087,
+      "grad_norm": 1.065873384475708,
+      "learning_rate": 0.001,
+      "loss": 0.6181,
+      "step": 202900
+    },
+    {
+      "epoch": 65.61085972850678,
+      "grad_norm": 1.2089930772781372,
+      "learning_rate": 0.001,
+      "loss": 0.6189,
+      "step": 203000
+    },
+    {
+      "epoch": 65.6431803490627,
+      "grad_norm": 1.3150242567062378,
+      "learning_rate": 0.001,
+      "loss": 0.6453,
+      "step": 203100
+    },
+    {
+      "epoch": 65.67550096961861,
+      "grad_norm": 1.3111896514892578,
+      "learning_rate": 0.001,
+      "loss": 0.6317,
+      "step": 203200
+    },
+    {
+      "epoch": 65.70782159017453,
+      "grad_norm": 1.1152726411819458,
+      "learning_rate": 0.001,
+      "loss": 0.6405,
+      "step": 203300
+    },
+    {
+      "epoch": 65.74014221073044,
+      "grad_norm": 1.3868318796157837,
+      "learning_rate": 0.001,
+      "loss": 0.6381,
+      "step": 203400
+    },
+    {
+      "epoch": 65.77246283128636,
+      "grad_norm": 1.0794872045516968,
+      "learning_rate": 0.001,
+      "loss": 0.6367,
+      "step": 203500
+    },
+    {
+      "epoch": 65.80478345184227,
+      "grad_norm": 0.9957221746444702,
+      "learning_rate": 0.001,
+      "loss": 0.6376,
+      "step": 203600
+    },
+    {
+      "epoch": 65.83710407239819,
+      "grad_norm": 1.7803281545639038,
+      "learning_rate": 0.001,
+      "loss": 0.6441,
+      "step": 203700
+    },
+    {
+      "epoch": 65.8694246929541,
+      "grad_norm": 1.3243619203567505,
+      "learning_rate": 0.001,
+      "loss": 0.6486,
+      "step": 203800
+    },
+    {
+      "epoch": 65.90174531351002,
+      "grad_norm": 1.8025758266448975,
+      "learning_rate": 0.001,
+      "loss": 0.6266,
+      "step": 203900
+    },
+    {
+      "epoch": 65.93406593406593,
+      "grad_norm": 1.5865435600280762,
+      "learning_rate": 0.001,
+      "loss": 0.6534,
+      "step": 204000
+    },
+    {
+      "epoch": 65.96638655462185,
+      "grad_norm": 1.2580561637878418,
+      "learning_rate": 0.001,
+      "loss": 0.6571,
+      "step": 204100
+    },
+    {
+      "epoch": 65.99870717517777,
+      "grad_norm": 1.0204856395721436,
+      "learning_rate": 0.001,
+      "loss": 0.6378,
+      "step": 204200
+    },
+    {
+      "epoch": 66.03102779573368,
+      "grad_norm": 0.7897757887840271,
+      "learning_rate": 0.001,
+      "loss": 0.5688,
+      "step": 204300
+    },
+    {
+      "epoch": 66.0633484162896,
+      "grad_norm": 0.9957269430160522,
+      "learning_rate": 0.001,
+      "loss": 0.5762,
+      "step": 204400
+    },
+    {
+      "epoch": 66.0956690368455,
+      "grad_norm": 0.931441605091095,
+      "learning_rate": 0.001,
+      "loss": 0.5709,
+      "step": 204500
+    },
+    {
+      "epoch": 66.12798965740143,
+      "grad_norm": 1.0603188276290894,
+      "learning_rate": 0.001,
+      "loss": 0.6035,
+      "step": 204600
+    },
+    {
+      "epoch": 66.16031027795734,
+      "grad_norm": 0.9763708114624023,
+      "learning_rate": 0.001,
+      "loss": 0.5772,
+      "step": 204700
+    },
+    {
+      "epoch": 66.19263089851326,
+      "grad_norm": 0.8747771382331848,
+      "learning_rate": 0.001,
+      "loss": 0.5889,
+      "step": 204800
+    },
+    {
+      "epoch": 66.22495151906917,
+      "grad_norm": 0.968600869178772,
+      "learning_rate": 0.001,
+      "loss": 0.6063,
+      "step": 204900
+    },
+    {
+      "epoch": 66.25727213962509,
+      "grad_norm": 0.8659926652908325,
+      "learning_rate": 0.001,
+      "loss": 0.5979,
+      "step": 205000
+    },
+    {
+      "epoch": 66.289592760181,
+      "grad_norm": 0.9334861040115356,
+      "learning_rate": 0.001,
+      "loss": 0.5813,
+      "step": 205100
+    },
+    {
+      "epoch": 66.32191338073692,
+      "grad_norm": 1.1221340894699097,
+      "learning_rate": 0.001,
+      "loss": 0.6035,
+      "step": 205200
+    },
+    {
+      "epoch": 66.35423400129282,
+      "grad_norm": 1.2959271669387817,
+      "learning_rate": 0.001,
+      "loss": 0.5931,
+      "step": 205300
+    },
+    {
+      "epoch": 66.38655462184875,
+      "grad_norm": 0.9807857275009155,
+      "learning_rate": 0.001,
+      "loss": 0.6045,
+      "step": 205400
+    },
+    {
+      "epoch": 66.41887524240465,
+      "grad_norm": 1.0263525247573853,
+      "learning_rate": 0.001,
+      "loss": 0.604,
+      "step": 205500
+    },
+    {
+      "epoch": 66.45119586296057,
+      "grad_norm": 0.9026124477386475,
+      "learning_rate": 0.001,
+      "loss": 0.6266,
+      "step": 205600
+    },
+    {
+      "epoch": 66.48351648351648,
+      "grad_norm": 1.1074974536895752,
+      "learning_rate": 0.001,
+      "loss": 0.6292,
+      "step": 205700
+    },
+    {
+      "epoch": 66.5158371040724,
+      "grad_norm": 0.889342725276947,
+      "learning_rate": 0.001,
+      "loss": 0.6134,
+      "step": 205800
+    },
+    {
+      "epoch": 66.54815772462831,
+      "grad_norm": 1.0654977560043335,
+      "learning_rate": 0.001,
+      "loss": 0.6055,
+      "step": 205900
+    },
+    {
+      "epoch": 66.58047834518423,
+      "grad_norm": 1.0482105016708374,
+      "learning_rate": 0.001,
+      "loss": 0.6176,
+      "step": 206000
+    },
+    {
+      "epoch": 66.61279896574014,
+      "grad_norm": 1.1681993007659912,
+      "learning_rate": 0.001,
+      "loss": 0.6305,
+      "step": 206100
+    },
+    {
+      "epoch": 66.64511958629606,
+      "grad_norm": 1.0052601099014282,
+      "learning_rate": 0.001,
+      "loss": 0.6273,
+      "step": 206200
+    },
+    {
+      "epoch": 66.67744020685197,
+      "grad_norm": 0.967351496219635,
+      "learning_rate": 0.001,
+      "loss": 0.6226,
+      "step": 206300
+    },
+    {
+      "epoch": 66.70976082740789,
+      "grad_norm": 0.9480918049812317,
+      "learning_rate": 0.001,
+      "loss": 0.6232,
+      "step": 206400
+    },
+    {
+      "epoch": 66.7420814479638,
+      "grad_norm": 1.2429511547088623,
+      "learning_rate": 0.001,
+      "loss": 0.6479,
+      "step": 206500
+    },
+    {
+      "epoch": 66.77440206851972,
+      "grad_norm": 0.8635256290435791,
+      "learning_rate": 0.001,
+      "loss": 0.6254,
+      "step": 206600
+    },
+    {
+      "epoch": 66.80672268907563,
+      "grad_norm": 1.0228506326675415,
+      "learning_rate": 0.001,
+      "loss": 0.6335,
+      "step": 206700
+    },
+    {
+      "epoch": 66.83904330963155,
+      "grad_norm": 1.1268036365509033,
+      "learning_rate": 0.001,
+      "loss": 0.6306,
+      "step": 206800
+    },
+    {
+      "epoch": 66.87136393018746,
+      "grad_norm": 0.8553975224494934,
+      "learning_rate": 0.001,
+      "loss": 0.6328,
+      "step": 206900
+    },
+    {
+      "epoch": 66.90368455074338,
+      "grad_norm": 1.1963768005371094,
+      "learning_rate": 0.001,
+      "loss": 0.6224,
+      "step": 207000
+    },
+    {
+      "epoch": 66.93600517129929,
+      "grad_norm": 1.0734686851501465,
+      "learning_rate": 0.001,
+      "loss": 0.6444,
+      "step": 207100
+    },
+    {
+      "epoch": 66.96832579185521,
+      "grad_norm": 1.1173734664916992,
+      "learning_rate": 0.001,
+      "loss": 0.6419,
+      "step": 207200
+    },
+    {
+      "epoch": 67.00064641241111,
+      "grad_norm": 2.1012940406799316,
+      "learning_rate": 0.001,
+      "loss": 0.6206,
+      "step": 207300
+    },
+    {
+      "epoch": 67.03296703296704,
+      "grad_norm": 1.714339017868042,
+      "learning_rate": 0.001,
+      "loss": 0.5702,
+      "step": 207400
+    },
+    {
+      "epoch": 67.06528765352294,
+      "grad_norm": 1.949273705482483,
+      "learning_rate": 0.001,
+      "loss": 0.5714,
+      "step": 207500
+    },
+    {
+      "epoch": 67.09760827407887,
+      "grad_norm": 1.5728001594543457,
+      "learning_rate": 0.001,
+      "loss": 0.5665,
+      "step": 207600
+    },
+    {
+      "epoch": 67.12992889463477,
+      "grad_norm": 1.9394041299819946,
+      "learning_rate": 0.001,
+      "loss": 0.5756,
+      "step": 207700
+    },
+    {
+      "epoch": 67.1622495151907,
+      "grad_norm": 2.8248255252838135,
+      "learning_rate": 0.001,
+      "loss": 0.5865,
+      "step": 207800
+    },
+    {
+      "epoch": 67.1945701357466,
+      "grad_norm": 2.1231889724731445,
+      "learning_rate": 0.001,
+      "loss": 0.5867,
+      "step": 207900
+    },
+    {
+      "epoch": 67.22689075630252,
+      "grad_norm": 1.68788480758667,
+      "learning_rate": 0.001,
+      "loss": 0.5956,
+      "step": 208000
+    },
+    {
+      "epoch": 67.25921137685843,
+      "grad_norm": 2.0351736545562744,
+      "learning_rate": 0.001,
+      "loss": 0.5791,
+      "step": 208100
+    },
+    {
+      "epoch": 67.29153199741435,
+      "grad_norm": 39.95174789428711,
+      "learning_rate": 0.001,
+      "loss": 0.5842,
+      "step": 208200
+    },
+    {
+      "epoch": 67.32385261797026,
+      "grad_norm": 1.5130423307418823,
+      "learning_rate": 0.001,
+      "loss": 0.6005,
+      "step": 208300
+    },
+    {
+      "epoch": 67.35617323852618,
+      "grad_norm": 1.4413126707077026,
+      "learning_rate": 0.001,
+      "loss": 0.6138,
+      "step": 208400
+    },
+    {
+      "epoch": 67.38849385908209,
+      "grad_norm": 1.9939687252044678,
+      "learning_rate": 0.001,
+      "loss": 0.5946,
+      "step": 208500
+    },
+    {
+      "epoch": 67.42081447963801,
+      "grad_norm": 2.0666182041168213,
+      "learning_rate": 0.001,
+      "loss": 0.5939,
+      "step": 208600
+    },
+    {
+      "epoch": 67.45313510019392,
+      "grad_norm": 1.77826726436615,
+      "learning_rate": 0.001,
+      "loss": 0.6037,
+      "step": 208700
+    },
+    {
+      "epoch": 67.48545572074984,
+      "grad_norm": 1.637294888496399,
+      "learning_rate": 0.001,
+      "loss": 0.5942,
+      "step": 208800
+    },
+    {
+      "epoch": 67.51777634130575,
+      "grad_norm": 1.2931591272354126,
+      "learning_rate": 0.001,
+      "loss": 0.6037,
+      "step": 208900
+    },
+    {
+      "epoch": 67.55009696186167,
+      "grad_norm": 1.5712636709213257,
+      "learning_rate": 0.001,
+      "loss": 0.604,
+      "step": 209000
+    },
+    {
+      "epoch": 67.58241758241758,
+      "grad_norm": 2.2092108726501465,
+      "learning_rate": 0.001,
+      "loss": 0.6223,
+      "step": 209100
+    },
+    {
+      "epoch": 67.6147382029735,
+      "grad_norm": 1.896464228630066,
+      "learning_rate": 0.001,
+      "loss": 0.6131,
+      "step": 209200
+    },
+    {
+      "epoch": 67.6470588235294,
+      "grad_norm": 1.501253366470337,
+      "learning_rate": 0.001,
+      "loss": 0.6169,
+      "step": 209300
+    },
+    {
+      "epoch": 67.67937944408533,
+      "grad_norm": 2.152764081954956,
+      "learning_rate": 0.001,
+      "loss": 0.6189,
+      "step": 209400
+    },
+    {
+      "epoch": 67.71170006464124,
+      "grad_norm": 1.7892820835113525,
+      "learning_rate": 0.001,
+      "loss": 0.626,
+      "step": 209500
+    },
+    {
+      "epoch": 67.74402068519716,
+      "grad_norm": 1.704662799835205,
+      "learning_rate": 0.001,
+      "loss": 0.6233,
+      "step": 209600
+    },
+    {
+      "epoch": 67.77634130575306,
+      "grad_norm": 1.6158897876739502,
+      "learning_rate": 0.001,
+      "loss": 0.6373,
+      "step": 209700
+    },
+    {
+      "epoch": 67.80866192630899,
+      "grad_norm": 2.0007736682891846,
+      "learning_rate": 0.001,
+      "loss": 0.6404,
+      "step": 209800
+    },
+    {
+      "epoch": 67.8409825468649,
+      "grad_norm": 2.2105071544647217,
+      "learning_rate": 0.001,
+      "loss": 0.6201,
+      "step": 209900
+    },
+    {
+      "epoch": 67.87330316742081,
+      "grad_norm": 2.1892433166503906,
+      "learning_rate": 0.001,
+      "loss": 0.6174,
+      "step": 210000
+    },
+    {
+      "epoch": 67.90562378797672,
+      "grad_norm": 2.033268451690674,
+      "learning_rate": 0.001,
+      "loss": 0.6263,
+      "step": 210100
+    },
+    {
+      "epoch": 67.93794440853264,
+      "grad_norm": 1.7615630626678467,
+      "learning_rate": 0.001,
+      "loss": 0.6335,
+      "step": 210200
+    },
+    {
+      "epoch": 67.97026502908855,
+      "grad_norm": 2.064373254776001,
+      "learning_rate": 0.001,
+      "loss": 0.6224,
+      "step": 210300
+    },
+    {
+      "epoch": 68.00258564964447,
+      "grad_norm": 1.2215957641601562,
+      "learning_rate": 0.001,
+      "loss": 0.6624,
+      "step": 210400
+    },
+    {
+      "epoch": 68.0349062702004,
+      "grad_norm": 1.3218648433685303,
+      "learning_rate": 0.001,
+      "loss": 0.5683,
+      "step": 210500
+    },
+    {
+      "epoch": 68.0672268907563,
+      "grad_norm": 1.6484256982803345,
+      "learning_rate": 0.001,
+      "loss": 0.5646,
+      "step": 210600
+    },
+    {
+      "epoch": 68.09954751131222,
+      "grad_norm": 2.0356903076171875,
+      "learning_rate": 0.001,
+      "loss": 0.5728,
+      "step": 210700
+    },
+    {
+      "epoch": 68.13186813186813,
+      "grad_norm": 1.6391444206237793,
+      "learning_rate": 0.001,
+      "loss": 0.5682,
+      "step": 210800
+    },
+    {
+      "epoch": 68.16418875242405,
+      "grad_norm": 1.3358436822891235,
+      "learning_rate": 0.001,
+      "loss": 0.572,
+      "step": 210900
+    },
+    {
+      "epoch": 68.19650937297996,
+      "grad_norm": 1.5110633373260498,
+      "learning_rate": 0.001,
+      "loss": 0.5841,
+      "step": 211000
+    },
+    {
+      "epoch": 68.22882999353588,
+      "grad_norm": 1.7556782960891724,
+      "learning_rate": 0.001,
+      "loss": 0.579,
+      "step": 211100
+    },
+    {
+      "epoch": 68.26115061409179,
+      "grad_norm": 1.377893090248108,
+      "learning_rate": 0.001,
+      "loss": 0.5891,
+      "step": 211200
+    },
+    {
+      "epoch": 68.29347123464771,
+      "grad_norm": 1.45765221118927,
+      "learning_rate": 0.001,
+      "loss": 0.5824,
+      "step": 211300
+    },
+    {
+      "epoch": 68.32579185520362,
+      "grad_norm": 1.6839174032211304,
+      "learning_rate": 0.001,
+      "loss": 0.5854,
+      "step": 211400
+    },
+    {
+      "epoch": 68.35811247575954,
+      "grad_norm": 1.3309268951416016,
+      "learning_rate": 0.001,
+      "loss": 0.585,
+      "step": 211500
+    },
+    {
+      "epoch": 68.39043309631545,
+      "grad_norm": 1.6310690641403198,
+      "learning_rate": 0.001,
+      "loss": 0.592,
+      "step": 211600
+    },
+    {
+      "epoch": 68.42275371687137,
+      "grad_norm": 1.7203037738800049,
+      "learning_rate": 0.001,
+      "loss": 0.5917,
+      "step": 211700
+    },
+    {
+      "epoch": 68.45507433742728,
+      "grad_norm": 1.5274381637573242,
+      "learning_rate": 0.001,
+      "loss": 0.6048,
+      "step": 211800
+    },
+    {
+      "epoch": 68.4873949579832,
+      "grad_norm": 1.3097580671310425,
+      "learning_rate": 0.001,
+      "loss": 0.5978,
+      "step": 211900
+    },
+    {
+      "epoch": 68.5197155785391,
+      "grad_norm": 1.3245912790298462,
+      "learning_rate": 0.001,
+      "loss": 0.6101,
+      "step": 212000
+    },
+    {
+      "epoch": 68.55203619909503,
+      "grad_norm": 1.7545794248580933,
+      "learning_rate": 0.001,
+      "loss": 0.6098,
+      "step": 212100
+    },
+    {
+      "epoch": 68.58435681965094,
+      "grad_norm": 1.5376828908920288,
+      "learning_rate": 0.001,
+      "loss": 0.6023,
+      "step": 212200
+    },
+    {
+      "epoch": 68.61667744020686,
+      "grad_norm": 1.4824035167694092,
+      "learning_rate": 0.001,
+      "loss": 0.6075,
+      "step": 212300
+    },
+    {
+      "epoch": 68.64899806076276,
+      "grad_norm": 1.7261457443237305,
+      "learning_rate": 0.001,
+      "loss": 0.62,
+      "step": 212400
+    },
+    {
+      "epoch": 68.68131868131869,
+      "grad_norm": 1.4751150608062744,
+      "learning_rate": 0.001,
+      "loss": 0.595,
+      "step": 212500
+    },
+    {
+      "epoch": 68.7136393018746,
+      "grad_norm": 1.6435458660125732,
+      "learning_rate": 0.001,
+      "loss": 0.6185,
+      "step": 212600
+    },
+    {
+      "epoch": 68.74595992243052,
+      "grad_norm": 1.823643445968628,
+      "learning_rate": 0.001,
+      "loss": 0.6149,
+      "step": 212700
+    },
+    {
+      "epoch": 68.77828054298642,
+      "grad_norm": 1.5763131380081177,
+      "learning_rate": 0.001,
+      "loss": 0.6125,
+      "step": 212800
+    },
+    {
+      "epoch": 68.81060116354234,
+      "grad_norm": 2.119630813598633,
+      "learning_rate": 0.001,
+      "loss": 0.6197,
+      "step": 212900
+    },
+    {
+      "epoch": 68.84292178409825,
+      "grad_norm": 1.5855299234390259,
+      "learning_rate": 0.001,
+      "loss": 0.6174,
+      "step": 213000
+    },
+    {
+      "epoch": 68.87524240465417,
+      "grad_norm": 1.9166241884231567,
+      "learning_rate": 0.001,
+      "loss": 0.6214,
+      "step": 213100
+    },
+    {
+      "epoch": 68.90756302521008,
+      "grad_norm": 1.4405461549758911,
+      "learning_rate": 0.001,
+      "loss": 0.6125,
+      "step": 213200
+    },
+    {
+      "epoch": 68.939883645766,
+      "grad_norm": 1.537550449371338,
+      "learning_rate": 0.001,
+      "loss": 0.6363,
+      "step": 213300
+    },
+    {
+      "epoch": 68.97220426632191,
+      "grad_norm": 1.497914433479309,
+      "learning_rate": 0.001,
+      "loss": 0.6071,
+      "step": 213400
+    },
+    {
+      "epoch": 69.00452488687783,
+      "grad_norm": 1.598833441734314,
+      "learning_rate": 0.001,
+      "loss": 0.6439,
+      "step": 213500
+    },
+    {
+      "epoch": 69.03684550743374,
+      "grad_norm": 1.656296968460083,
+      "learning_rate": 0.001,
+      "loss": 0.5466,
+      "step": 213600
+    },
+    {
+      "epoch": 69.06916612798966,
+      "grad_norm": 1.80128812789917,
+      "learning_rate": 0.001,
+      "loss": 0.5598,
+      "step": 213700
+    },
+    {
+      "epoch": 69.10148674854557,
+      "grad_norm": 1.3732154369354248,
+      "learning_rate": 0.001,
+      "loss": 0.5704,
+      "step": 213800
+    },
+    {
+      "epoch": 69.13380736910149,
+      "grad_norm": 1.4523626565933228,
+      "learning_rate": 0.001,
+      "loss": 0.5613,
+      "step": 213900
+    },
+    {
+      "epoch": 69.1661279896574,
+      "grad_norm": 1.4673432111740112,
+      "learning_rate": 0.001,
+      "loss": 0.592,
+      "step": 214000
+    },
+    {
+      "epoch": 69.19844861021332,
+      "grad_norm": 1.9327133893966675,
+      "learning_rate": 0.001,
+      "loss": 0.5726,
+      "step": 214100
+    },
+    {
+      "epoch": 69.23076923076923,
+      "grad_norm": 1.9804271459579468,
+      "learning_rate": 0.001,
+      "loss": 0.563,
+      "step": 214200
+    },
+    {
+      "epoch": 69.26308985132515,
+      "grad_norm": 1.4549953937530518,
+      "learning_rate": 0.001,
+      "loss": 0.5724,
+      "step": 214300
+    },
+    {
+      "epoch": 69.29541047188106,
+      "grad_norm": 1.4232009649276733,
+      "learning_rate": 0.001,
+      "loss": 0.5779,
+      "step": 214400
+    },
+    {
+      "epoch": 69.32773109243698,
+      "grad_norm": 1.3493443727493286,
+      "learning_rate": 0.001,
+      "loss": 0.5708,
+      "step": 214500
+    },
+    {
+      "epoch": 69.36005171299288,
+      "grad_norm": 1.3888554573059082,
+      "learning_rate": 0.001,
+      "loss": 0.5761,
+      "step": 214600
+    },
+    {
+      "epoch": 69.3923723335488,
+      "grad_norm": 1.9216724634170532,
+      "learning_rate": 0.001,
+      "loss": 0.5856,
+      "step": 214700
+    },
+    {
+      "epoch": 69.42469295410471,
+      "grad_norm": 1.405328631401062,
+      "learning_rate": 0.001,
+      "loss": 0.5776,
+      "step": 214800
+    },
+    {
+      "epoch": 69.45701357466064,
+      "grad_norm": 6.138822078704834,
+      "learning_rate": 0.001,
+      "loss": 0.5921,
+      "step": 214900
+    },
+    {
+      "epoch": 69.48933419521654,
+      "grad_norm": 1.3978585004806519,
+      "learning_rate": 0.001,
+      "loss": 0.6057,
+      "step": 215000
+    },
+    {
+      "epoch": 69.52165481577246,
+      "grad_norm": 1.506648302078247,
+      "learning_rate": 0.001,
+      "loss": 0.6013,
+      "step": 215100
+    },
+    {
+      "epoch": 69.55397543632837,
+      "grad_norm": 1.3363564014434814,
+      "learning_rate": 0.001,
+      "loss": 0.6018,
+      "step": 215200
+    },
+    {
+      "epoch": 69.5862960568843,
+      "grad_norm": 1.5143280029296875,
+      "learning_rate": 0.001,
+      "loss": 0.6043,
+      "step": 215300
+    },
+    {
+      "epoch": 69.6186166774402,
+      "grad_norm": 1.5944569110870361,
+      "learning_rate": 0.001,
+      "loss": 0.6056,
+      "step": 215400
+    },
+    {
+      "epoch": 69.65093729799612,
+      "grad_norm": 1.5265306234359741,
+      "learning_rate": 0.001,
+      "loss": 0.6095,
+      "step": 215500
+    },
+    {
+      "epoch": 69.68325791855203,
+      "grad_norm": 1.4436867237091064,
+      "learning_rate": 0.001,
+      "loss": 0.6039,
+      "step": 215600
+    },
+    {
+      "epoch": 69.71557853910795,
+      "grad_norm": 1.7266173362731934,
+      "learning_rate": 0.001,
+      "loss": 0.6178,
+      "step": 215700
+    },
+    {
+      "epoch": 69.74789915966386,
+      "grad_norm": 1.4980125427246094,
+      "learning_rate": 0.001,
+      "loss": 0.6012,
+      "step": 215800
+    },
+    {
+      "epoch": 69.78021978021978,
+      "grad_norm": 1.6730942726135254,
+      "learning_rate": 0.001,
+      "loss": 0.5914,
+      "step": 215900
+    },
+    {
+      "epoch": 69.81254040077569,
+      "grad_norm": 1.3241289854049683,
+      "learning_rate": 0.001,
+      "loss": 0.6106,
+      "step": 216000
+    },
+    {
+      "epoch": 69.84486102133161,
+      "grad_norm": 1.7309767007827759,
+      "learning_rate": 0.001,
+      "loss": 0.6057,
+      "step": 216100
+    },
+    {
+      "epoch": 69.87718164188752,
+      "grad_norm": 1.775754451751709,
+      "learning_rate": 0.001,
+      "loss": 0.6051,
+      "step": 216200
+    },
+    {
+      "epoch": 69.90950226244344,
+      "grad_norm": 1.5161246061325073,
+      "learning_rate": 0.001,
+      "loss": 0.6182,
+      "step": 216300
+    },
+    {
+      "epoch": 69.94182288299935,
+      "grad_norm": 1.2103925943374634,
+      "learning_rate": 0.001,
+      "loss": 0.6103,
+      "step": 216400
+    },
+    {
+      "epoch": 69.97414350355527,
+      "grad_norm": 1.3319581747055054,
+      "learning_rate": 0.001,
+      "loss": 0.6308,
+      "step": 216500
+    },
+    {
+      "epoch": 70.00646412411119,
+      "grad_norm": 1.5204486846923828,
+      "learning_rate": 0.001,
+      "loss": 0.6271,
+      "step": 216600
+    },
+    {
+      "epoch": 70.0387847446671,
+      "grad_norm": 1.5178135633468628,
+      "learning_rate": 0.001,
+      "loss": 0.56,
+      "step": 216700
+    },
+    {
+      "epoch": 70.07110536522302,
+      "grad_norm": 2.1006743907928467,
+      "learning_rate": 0.001,
+      "loss": 0.5537,
+      "step": 216800
+    },
+    {
+      "epoch": 70.10342598577893,
+      "grad_norm": 1.3505820035934448,
+      "learning_rate": 0.001,
+      "loss": 0.552,
+      "step": 216900
+    },
+    {
+      "epoch": 70.13574660633485,
+      "grad_norm": 1.6325558423995972,
+      "learning_rate": 0.001,
+      "loss": 0.5586,
+      "step": 217000
+    },
+    {
+      "epoch": 70.16806722689076,
+      "grad_norm": 1.2984336614608765,
+      "learning_rate": 0.001,
+      "loss": 0.575,
+      "step": 217100
+    },
+    {
+      "epoch": 70.20038784744668,
+      "grad_norm": 1.3796709775924683,
+      "learning_rate": 0.001,
+      "loss": 0.5591,
+      "step": 217200
+    },
+    {
+      "epoch": 70.23270846800258,
+      "grad_norm": 1.330457091331482,
+      "learning_rate": 0.001,
+      "loss": 0.5662,
+      "step": 217300
+    },
+    {
+      "epoch": 70.2650290885585,
+      "grad_norm": 1.3719303607940674,
+      "learning_rate": 0.001,
+      "loss": 0.5662,
+      "step": 217400
+    },
+    {
+      "epoch": 70.29734970911441,
+      "grad_norm": 1.4298548698425293,
+      "learning_rate": 0.001,
+      "loss": 0.5764,
+      "step": 217500
+    },
+    {
+      "epoch": 70.32967032967034,
+      "grad_norm": 1.5424270629882812,
+      "learning_rate": 0.001,
+      "loss": 0.5796,
+      "step": 217600
+    },
+    {
+      "epoch": 70.36199095022624,
+      "grad_norm": 2.652244806289673,
+      "learning_rate": 0.001,
+      "loss": 0.5843,
+      "step": 217700
+    },
+    {
+      "epoch": 70.39431157078216,
+      "grad_norm": 1.5292531251907349,
+      "learning_rate": 0.001,
+      "loss": 0.5734,
+      "step": 217800
+    },
+    {
+      "epoch": 70.42663219133807,
+      "grad_norm": 1.3583437204360962,
+      "learning_rate": 0.001,
+      "loss": 0.5722,
+      "step": 217900
+    },
+    {
+      "epoch": 70.458952811894,
+      "grad_norm": 1.53679358959198,
+      "learning_rate": 0.001,
+      "loss": 0.5855,
+      "step": 218000
+    },
+    {
+      "epoch": 70.4912734324499,
+      "grad_norm": 1.4344574213027954,
+      "learning_rate": 0.001,
+      "loss": 0.5878,
+      "step": 218100
+    },
+    {
+      "epoch": 70.52359405300582,
+      "grad_norm": 1.6726317405700684,
+      "learning_rate": 0.001,
+      "loss": 0.6018,
+      "step": 218200
+    },
+    {
+      "epoch": 70.55591467356173,
+      "grad_norm": 1.8626093864440918,
+      "learning_rate": 0.001,
+      "loss": 0.5872,
+      "step": 218300
+    },
+    {
+      "epoch": 70.58823529411765,
+      "grad_norm": 1.1871416568756104,
+      "learning_rate": 0.001,
+      "loss": 0.5868,
+      "step": 218400
+    },
+    {
+      "epoch": 70.62055591467356,
+      "grad_norm": 1.326026439666748,
+      "learning_rate": 0.001,
+      "loss": 0.5924,
+      "step": 218500
+    },
+    {
+      "epoch": 70.65287653522948,
+      "grad_norm": 1.5485585927963257,
+      "learning_rate": 0.001,
+      "loss": 0.5912,
+      "step": 218600
+    },
+    {
+      "epoch": 70.68519715578539,
+      "grad_norm": 1.8677207231521606,
+      "learning_rate": 0.001,
+      "loss": 0.5959,
+      "step": 218700
+    },
+    {
+      "epoch": 70.71751777634131,
+      "grad_norm": 1.7053265571594238,
+      "learning_rate": 0.001,
+      "loss": 0.5952,
+      "step": 218800
+    },
+    {
+      "epoch": 70.74983839689722,
+      "grad_norm": 1.3536996841430664,
+      "learning_rate": 0.001,
+      "loss": 0.5907,
+      "step": 218900
+    },
+    {
+      "epoch": 70.78215901745314,
+      "grad_norm": 2.0897066593170166,
+      "learning_rate": 0.001,
+      "loss": 0.6039,
+      "step": 219000
+    },
+    {
+      "epoch": 70.81447963800905,
+      "grad_norm": 1.4968990087509155,
+      "learning_rate": 0.001,
+      "loss": 0.5972,
+      "step": 219100
+    },
+    {
+      "epoch": 70.84680025856497,
+      "grad_norm": 1.3414818048477173,
+      "learning_rate": 0.001,
+      "loss": 0.6101,
+      "step": 219200
+    },
+    {
+      "epoch": 70.87912087912088,
+      "grad_norm": 1.3500386476516724,
+      "learning_rate": 0.001,
+      "loss": 0.6112,
+      "step": 219300
+    },
+    {
+      "epoch": 70.9114414996768,
+      "grad_norm": 1.1630859375,
+      "learning_rate": 0.001,
+      "loss": 0.6123,
+      "step": 219400
+    },
+    {
+      "epoch": 70.9437621202327,
+      "grad_norm": 1.5307071208953857,
+      "learning_rate": 0.001,
+      "loss": 0.6161,
+      "step": 219500
+    },
+    {
+      "epoch": 70.97608274078863,
+      "grad_norm": 1.4781556129455566,
+      "learning_rate": 0.001,
+      "loss": 0.6133,
+      "step": 219600
+    },
+    {
+      "epoch": 71.00840336134453,
+      "grad_norm": 1.4278903007507324,
+      "learning_rate": 0.001,
+      "loss": 0.5894,
+      "step": 219700
+    },
+    {
+      "epoch": 71.04072398190046,
+      "grad_norm": 1.442977786064148,
+      "learning_rate": 0.001,
+      "loss": 0.534,
+      "step": 219800
+    },
+    {
+      "epoch": 71.07304460245636,
+      "grad_norm": 1.3905531167984009,
+      "learning_rate": 0.001,
+      "loss": 0.561,
+      "step": 219900
+    },
+    {
+      "epoch": 71.10536522301229,
+      "grad_norm": 1.9299741983413696,
+      "learning_rate": 0.001,
+      "loss": 0.5489,
+      "step": 220000
+    },
+    {
+      "epoch": 71.13768584356819,
+      "grad_norm": 1.498746633529663,
+      "learning_rate": 0.001,
+      "loss": 0.5542,
+      "step": 220100
+    },
+    {
+      "epoch": 71.17000646412411,
+      "grad_norm": 1.8542300462722778,
+      "learning_rate": 0.001,
+      "loss": 0.5478,
+      "step": 220200
+    },
+    {
+      "epoch": 71.20232708468002,
+      "grad_norm": 1.0949512720108032,
+      "learning_rate": 0.001,
+      "loss": 0.5619,
+      "step": 220300
+    },
+    {
+      "epoch": 71.23464770523594,
+      "grad_norm": 1.170140027999878,
+      "learning_rate": 0.001,
+      "loss": 0.5533,
+      "step": 220400
+    },
+    {
+      "epoch": 71.26696832579185,
+      "grad_norm": 1.5471538305282593,
+      "learning_rate": 0.001,
+      "loss": 0.5618,
+      "step": 220500
+    },
+    {
+      "epoch": 71.29928894634777,
+      "grad_norm": 1.3798595666885376,
+      "learning_rate": 0.001,
+      "loss": 0.5587,
+      "step": 220600
+    },
+    {
+      "epoch": 71.33160956690368,
+      "grad_norm": 1.5981390476226807,
+      "learning_rate": 0.001,
+      "loss": 0.5715,
+      "step": 220700
+    },
+    {
+      "epoch": 71.3639301874596,
+      "grad_norm": 1.3824917078018188,
+      "learning_rate": 0.001,
+      "loss": 0.5809,
+      "step": 220800
+    },
+    {
+      "epoch": 71.39625080801551,
+      "grad_norm": 1.6975314617156982,
+      "learning_rate": 0.001,
+      "loss": 0.5744,
+      "step": 220900
+    },
+    {
+      "epoch": 71.42857142857143,
+      "grad_norm": 1.794413685798645,
+      "learning_rate": 0.001,
+      "loss": 0.5806,
+      "step": 221000
+    },
+    {
+      "epoch": 71.46089204912734,
+      "grad_norm": 1.2590088844299316,
+      "learning_rate": 0.001,
+      "loss": 0.5779,
+      "step": 221100
+    },
+    {
+      "epoch": 71.49321266968326,
+      "grad_norm": 1.3916282653808594,
+      "learning_rate": 0.001,
+      "loss": 0.5704,
+      "step": 221200
+    },
+    {
+      "epoch": 71.52553329023917,
+      "grad_norm": 1.3690452575683594,
+      "learning_rate": 0.001,
+      "loss": 0.5878,
+      "step": 221300
+    },
+    {
+      "epoch": 71.55785391079509,
+      "grad_norm": 1.6590235233306885,
+      "learning_rate": 0.001,
+      "loss": 0.5892,
+      "step": 221400
+    },
+    {
+      "epoch": 71.590174531351,
+      "grad_norm": 1.3054158687591553,
+      "learning_rate": 0.001,
+      "loss": 0.5834,
+      "step": 221500
+    },
+    {
+      "epoch": 71.62249515190692,
+      "grad_norm": 1.5784341096878052,
+      "learning_rate": 0.001,
+      "loss": 0.5712,
+      "step": 221600
+    },
+    {
+      "epoch": 71.65481577246283,
+      "grad_norm": 1.4556595087051392,
+      "learning_rate": 0.001,
+      "loss": 0.5901,
+      "step": 221700
+    },
+    {
+      "epoch": 71.68713639301875,
+      "grad_norm": 1.4017647504806519,
+      "learning_rate": 0.001,
+      "loss": 0.5984,
+      "step": 221800
+    },
+    {
+      "epoch": 71.71945701357465,
+      "grad_norm": 1.2876161336898804,
+      "learning_rate": 0.001,
+      "loss": 0.5852,
+      "step": 221900
+    },
+    {
+      "epoch": 71.75177763413058,
+      "grad_norm": 1.0554882287979126,
+      "learning_rate": 0.001,
+      "loss": 0.6103,
+      "step": 222000
+    },
+    {
+      "epoch": 71.78409825468648,
+      "grad_norm": 1.4979000091552734,
+      "learning_rate": 0.001,
+      "loss": 0.6002,
+      "step": 222100
+    },
+    {
+      "epoch": 71.8164188752424,
+      "grad_norm": 1.4148439168930054,
+      "learning_rate": 0.001,
+      "loss": 0.5982,
+      "step": 222200
+    },
+    {
+      "epoch": 71.84873949579831,
+      "grad_norm": 1.569904088973999,
+      "learning_rate": 0.001,
+      "loss": 0.6052,
+      "step": 222300
+    },
+    {
+      "epoch": 71.88106011635423,
+      "grad_norm": 1.2948265075683594,
+      "learning_rate": 0.001,
+      "loss": 0.6032,
+      "step": 222400
+    },
+    {
+      "epoch": 71.91338073691014,
+      "grad_norm": 1.949517011642456,
+      "learning_rate": 0.001,
+      "loss": 0.5991,
+      "step": 222500
+    },
+    {
+      "epoch": 71.94570135746606,
+      "grad_norm": 1.2780336141586304,
+      "learning_rate": 0.001,
+      "loss": 0.6034,
+      "step": 222600
+    },
+    {
+      "epoch": 71.97802197802197,
+      "grad_norm": 2.284240245819092,
+      "learning_rate": 0.001,
+      "loss": 0.6101,
+      "step": 222700
+    },
+    {
+      "epoch": 72.01034259857789,
+      "grad_norm": 1.7869925498962402,
+      "learning_rate": 0.001,
+      "loss": 0.5696,
+      "step": 222800
+    },
+    {
+      "epoch": 72.04266321913381,
+      "grad_norm": 1.3875548839569092,
+      "learning_rate": 0.001,
+      "loss": 0.545,
+      "step": 222900
+    },
+    {
+      "epoch": 72.07498383968972,
+      "grad_norm": 1.3865928649902344,
+      "learning_rate": 0.001,
+      "loss": 0.5482,
+      "step": 223000
+    },
+    {
+      "epoch": 72.10730446024564,
+      "grad_norm": 1.2828845977783203,
+      "learning_rate": 0.001,
+      "loss": 0.552,
+      "step": 223100
+    },
+    {
+      "epoch": 72.13962508080155,
+      "grad_norm": 1.516929030418396,
+      "learning_rate": 0.001,
+      "loss": 0.5515,
+      "step": 223200
+    },
+    {
+      "epoch": 72.17194570135747,
+      "grad_norm": 1.5746568441390991,
+      "learning_rate": 0.001,
+      "loss": 0.5452,
+      "step": 223300
+    },
+    {
+      "epoch": 72.20426632191338,
+      "grad_norm": 1.2963000535964966,
+      "learning_rate": 0.001,
+      "loss": 0.5619,
+      "step": 223400
+    },
+    {
+      "epoch": 72.2365869424693,
+      "grad_norm": 1.3902140855789185,
+      "learning_rate": 0.001,
+      "loss": 0.5516,
+      "step": 223500
+    },
+    {
+      "epoch": 72.26890756302521,
+      "grad_norm": 1.8248586654663086,
+      "learning_rate": 0.001,
+      "loss": 0.5513,
+      "step": 223600
+    },
+    {
+      "epoch": 72.30122818358113,
+      "grad_norm": 1.8046656847000122,
+      "learning_rate": 0.001,
+      "loss": 0.569,
+      "step": 223700
+    },
+    {
+      "epoch": 72.33354880413704,
+      "grad_norm": 1.6079381704330444,
+      "learning_rate": 0.001,
+      "loss": 0.5721,
+      "step": 223800
+    },
+    {
+      "epoch": 72.36586942469296,
+      "grad_norm": 1.4093081951141357,
+      "learning_rate": 0.001,
+      "loss": 0.5692,
+      "step": 223900
+    },
+    {
+      "epoch": 72.39819004524887,
+      "grad_norm": 1.3620543479919434,
+      "learning_rate": 0.001,
+      "loss": 0.5683,
+      "step": 224000
+    },
+    {
+      "epoch": 72.43051066580479,
+      "grad_norm": 1.451627254486084,
+      "learning_rate": 0.001,
+      "loss": 0.565,
+      "step": 224100
+    },
+    {
+      "epoch": 72.4628312863607,
+      "grad_norm": 1.3643581867218018,
+      "learning_rate": 0.001,
+      "loss": 0.5579,
+      "step": 224200
+    },
+    {
+      "epoch": 72.49515190691662,
+      "grad_norm": 1.289871096611023,
+      "learning_rate": 0.001,
+      "loss": 0.5703,
+      "step": 224300
+    },
+    {
+      "epoch": 72.52747252747253,
+      "grad_norm": 1.699236512184143,
+      "learning_rate": 0.001,
+      "loss": 0.5651,
+      "step": 224400
+    },
+    {
+      "epoch": 72.55979314802845,
+      "grad_norm": 1.1216737031936646,
+      "learning_rate": 0.001,
+      "loss": 0.5717,
+      "step": 224500
+    },
+    {
+      "epoch": 72.59211376858435,
+      "grad_norm": 1.4215786457061768,
+      "learning_rate": 0.001,
+      "loss": 0.5855,
+      "step": 224600
+    },
+    {
+      "epoch": 72.62443438914028,
+      "grad_norm": 1.5842723846435547,
+      "learning_rate": 0.001,
+      "loss": 0.5796,
+      "step": 224700
+    },
+    {
+      "epoch": 72.65675500969618,
+      "grad_norm": 1.7459555864334106,
+      "learning_rate": 0.001,
+      "loss": 0.5807,
+      "step": 224800
+    },
+    {
+      "epoch": 72.6890756302521,
+      "grad_norm": 1.1991825103759766,
+      "learning_rate": 0.001,
+      "loss": 0.5856,
+      "step": 224900
+    },
+    {
+      "epoch": 72.72139625080801,
+      "grad_norm": 1.3949995040893555,
+      "learning_rate": 0.001,
+      "loss": 0.5847,
+      "step": 225000
+    },
+    {
+      "epoch": 72.75371687136393,
+      "grad_norm": 1.186496376991272,
+      "learning_rate": 0.001,
+      "loss": 0.5825,
+      "step": 225100
+    },
+    {
+      "epoch": 72.78603749191984,
+      "grad_norm": 1.3610190153121948,
+      "learning_rate": 0.001,
+      "loss": 0.5915,
+      "step": 225200
+    },
+    {
+      "epoch": 72.81835811247576,
+      "grad_norm": 1.4443519115447998,
+      "learning_rate": 0.001,
+      "loss": 0.5822,
+      "step": 225300
+    },
+    {
+      "epoch": 72.85067873303167,
+      "grad_norm": 1.031691074371338,
+      "learning_rate": 0.001,
+      "loss": 0.5897,
+      "step": 225400
+    },
+    {
+      "epoch": 72.88299935358759,
+      "grad_norm": 1.4198780059814453,
+      "learning_rate": 0.001,
+      "loss": 0.6026,
+      "step": 225500
+    },
+    {
+      "epoch": 72.9153199741435,
+      "grad_norm": 1.3856195211410522,
+      "learning_rate": 0.001,
+      "loss": 0.58,
+      "step": 225600
+    },
+    {
+      "epoch": 72.94764059469942,
+      "grad_norm": 1.4383305311203003,
+      "learning_rate": 0.001,
+      "loss": 0.5963,
+      "step": 225700
+    },
+    {
+      "epoch": 72.97996121525533,
+      "grad_norm": 1.4304252862930298,
+      "learning_rate": 0.001,
+      "loss": 0.6073,
+      "step": 225800
+    },
+    {
+      "epoch": 73.01228183581125,
+      "grad_norm": 1.6289730072021484,
+      "learning_rate": 0.001,
+      "loss": 0.5603,
+      "step": 225900
+    },
+    {
+      "epoch": 73.04460245636716,
+      "grad_norm": 1.623422622680664,
+      "learning_rate": 0.001,
+      "loss": 0.5222,
+      "step": 226000
+    },
+    {
+      "epoch": 73.07692307692308,
+      "grad_norm": 1.2384452819824219,
+      "learning_rate": 0.001,
+      "loss": 0.5311,
+      "step": 226100
+    },
+    {
+      "epoch": 73.10924369747899,
+      "grad_norm": 1.1145490407943726,
+      "learning_rate": 0.001,
+      "loss": 0.5302,
+      "step": 226200
+    },
+    {
+      "epoch": 73.14156431803491,
+      "grad_norm": 1.1442946195602417,
+      "learning_rate": 0.001,
+      "loss": 0.5457,
+      "step": 226300
+    },
+    {
+      "epoch": 73.17388493859082,
+      "grad_norm": 1.4592894315719604,
+      "learning_rate": 0.001,
+      "loss": 0.5369,
+      "step": 226400
+    },
+    {
+      "epoch": 73.20620555914674,
+      "grad_norm": 0.9766262173652649,
+      "learning_rate": 0.001,
+      "loss": 0.5423,
+      "step": 226500
+    },
+    {
+      "epoch": 73.23852617970265,
+      "grad_norm": 1.052048683166504,
+      "learning_rate": 0.001,
+      "loss": 0.5464,
+      "step": 226600
+    },
+    {
+      "epoch": 73.27084680025857,
+      "grad_norm": 1.2609376907348633,
+      "learning_rate": 0.001,
+      "loss": 0.5612,
+      "step": 226700
+    },
+    {
+      "epoch": 73.30316742081448,
+      "grad_norm": 1.1084191799163818,
+      "learning_rate": 0.001,
+      "loss": 0.5519,
+      "step": 226800
+    },
+    {
+      "epoch": 73.3354880413704,
+      "grad_norm": 1.3045639991760254,
+      "learning_rate": 0.001,
+      "loss": 0.5564,
+      "step": 226900
+    },
+    {
+      "epoch": 73.3678086619263,
+      "grad_norm": 1.4363510608673096,
+      "learning_rate": 0.001,
+      "loss": 0.5715,
+      "step": 227000
+    },
+    {
+      "epoch": 73.40012928248223,
+      "grad_norm": 1.3683849573135376,
+      "learning_rate": 0.001,
+      "loss": 0.5619,
+      "step": 227100
+    },
+    {
+      "epoch": 73.43244990303813,
+      "grad_norm": 1.4905529022216797,
+      "learning_rate": 0.001,
+      "loss": 0.5613,
+      "step": 227200
+    },
+    {
+      "epoch": 73.46477052359405,
+      "grad_norm": 1.3460173606872559,
+      "learning_rate": 0.001,
+      "loss": 0.5746,
+      "step": 227300
+    },
+    {
+      "epoch": 73.49709114414996,
+      "grad_norm": 1.5590981245040894,
+      "learning_rate": 0.001,
+      "loss": 0.5597,
+      "step": 227400
+    },
+    {
+      "epoch": 73.52941176470588,
+      "grad_norm": 1.2848957777023315,
+      "learning_rate": 0.001,
+      "loss": 0.5638,
+      "step": 227500
+    },
+    {
+      "epoch": 73.56173238526179,
+      "grad_norm": 1.3358205556869507,
+      "learning_rate": 0.001,
+      "loss": 0.5602,
+      "step": 227600
+    },
+    {
+      "epoch": 73.59405300581771,
+      "grad_norm": 1.6583843231201172,
+      "learning_rate": 0.001,
+      "loss": 0.5654,
+      "step": 227700
+    },
+    {
+      "epoch": 73.62637362637362,
+      "grad_norm": 1.0921430587768555,
+      "learning_rate": 0.001,
+      "loss": 0.5805,
+      "step": 227800
+    },
+    {
+      "epoch": 73.65869424692954,
+      "grad_norm": 1.0443511009216309,
+      "learning_rate": 0.001,
+      "loss": 0.5708,
+      "step": 227900
+    },
+    {
+      "epoch": 73.69101486748545,
+      "grad_norm": 1.2265650033950806,
+      "learning_rate": 0.001,
+      "loss": 0.5858,
+      "step": 228000
+    },
+    {
+      "epoch": 73.72333548804137,
+      "grad_norm": 1.4629805088043213,
+      "learning_rate": 0.001,
+      "loss": 0.5847,
+      "step": 228100
+    },
+    {
+      "epoch": 73.75565610859728,
+      "grad_norm": 1.3702300786972046,
+      "learning_rate": 0.001,
+      "loss": 0.5832,
+      "step": 228200
+    },
+    {
+      "epoch": 73.7879767291532,
+      "grad_norm": 1.6281440258026123,
+      "learning_rate": 0.001,
+      "loss": 0.5861,
+      "step": 228300
+    },
+    {
+      "epoch": 73.82029734970911,
+      "grad_norm": 1.4549918174743652,
+      "learning_rate": 0.001,
+      "loss": 0.5823,
+      "step": 228400
+    },
+    {
+      "epoch": 73.85261797026503,
+      "grad_norm": 1.3140660524368286,
+      "learning_rate": 0.001,
+      "loss": 0.5759,
+      "step": 228500
+    },
+    {
+      "epoch": 73.88493859082094,
+      "grad_norm": 1.5533117055892944,
+      "learning_rate": 0.001,
+      "loss": 0.586,
+      "step": 228600
+    },
+    {
+      "epoch": 73.91725921137686,
+      "grad_norm": 1.5729376077651978,
+      "learning_rate": 0.001,
+      "loss": 0.5965,
+      "step": 228700
+    },
+    {
+      "epoch": 73.94957983193277,
+      "grad_norm": 1.4772778749465942,
+      "learning_rate": 0.001,
+      "loss": 0.5939,
+      "step": 228800
+    },
+    {
+      "epoch": 73.98190045248869,
+      "grad_norm": 1.5759507417678833,
+      "learning_rate": 0.001,
+      "loss": 0.5919,
+      "step": 228900
+    },
+    {
+      "epoch": 74.01422107304461,
+      "grad_norm": 0.9670353531837463,
+      "learning_rate": 0.001,
+      "loss": 0.5414,
+      "step": 229000
+    },
+    {
+      "epoch": 74.04654169360052,
+      "grad_norm": 1.0269129276275635,
+      "learning_rate": 0.001,
+      "loss": 0.5299,
+      "step": 229100
+    },
+    {
+      "epoch": 74.07886231415644,
+      "grad_norm": 1.191085696220398,
+      "learning_rate": 0.001,
+      "loss": 0.5286,
+      "step": 229200
+    },
+    {
+      "epoch": 74.11118293471235,
+      "grad_norm": 1.1801611185073853,
+      "learning_rate": 0.001,
+      "loss": 0.549,
+      "step": 229300
+    },
+    {
+      "epoch": 74.14350355526827,
+      "grad_norm": 0.983653724193573,
+      "learning_rate": 0.001,
+      "loss": 0.5277,
+      "step": 229400
+    },
+    {
+      "epoch": 74.17582417582418,
+      "grad_norm": 1.0740876197814941,
+      "learning_rate": 0.001,
+      "loss": 0.5339,
+      "step": 229500
+    },
+    {
+      "epoch": 74.2081447963801,
+      "grad_norm": 1.0492289066314697,
+      "learning_rate": 0.001,
+      "loss": 0.5553,
+      "step": 229600
+    },
+    {
+      "epoch": 74.240465416936,
+      "grad_norm": 1.4024004936218262,
+      "learning_rate": 0.001,
+      "loss": 0.5512,
+      "step": 229700
+    },
+    {
+      "epoch": 74.27278603749193,
+      "grad_norm": 0.8431639671325684,
+      "learning_rate": 0.001,
+      "loss": 0.538,
+      "step": 229800
+    },
+    {
+      "epoch": 74.30510665804783,
+      "grad_norm": 0.8813731074333191,
+      "learning_rate": 0.001,
+      "loss": 0.5517,
+      "step": 229900
+    },
+    {
+      "epoch": 74.33742727860376,
+      "grad_norm": 1.1411322355270386,
+      "learning_rate": 0.001,
+      "loss": 0.5547,
+      "step": 230000
+    },
+    {
+      "epoch": 74.36974789915966,
+      "grad_norm": 1.0830614566802979,
+      "learning_rate": 0.001,
+      "loss": 0.5547,
+      "step": 230100
+    },
+    {
+      "epoch": 74.40206851971558,
+      "grad_norm": 1.2333221435546875,
+      "learning_rate": 0.001,
+      "loss": 0.5565,
+      "step": 230200
+    },
+    {
+      "epoch": 74.43438914027149,
+      "grad_norm": 1.199706792831421,
+      "learning_rate": 0.001,
+      "loss": 0.5625,
+      "step": 230300
+    },
+    {
+      "epoch": 74.46670976082741,
+      "grad_norm": 1.2191143035888672,
+      "learning_rate": 0.001,
+      "loss": 0.5647,
+      "step": 230400
+    },
+    {
+      "epoch": 74.49903038138332,
+      "grad_norm": 0.8158851861953735,
+      "learning_rate": 0.001,
+      "loss": 0.5592,
+      "step": 230500
+    },
+    {
+      "epoch": 74.53135100193924,
+      "grad_norm": 0.9935479760169983,
+      "learning_rate": 0.001,
+      "loss": 0.558,
+      "step": 230600
+    },
+    {
+      "epoch": 74.56367162249515,
+      "grad_norm": 1.1733251810073853,
+      "learning_rate": 0.001,
+      "loss": 0.5539,
+      "step": 230700
+    },
+    {
+      "epoch": 74.59599224305107,
+      "grad_norm": 1.2278839349746704,
+      "learning_rate": 0.001,
+      "loss": 0.5626,
+      "step": 230800
+    },
+    {
+      "epoch": 74.62831286360698,
+      "grad_norm": 3.4744834899902344,
+      "learning_rate": 0.001,
+      "loss": 0.5678,
+      "step": 230900
+    },
+    {
+      "epoch": 74.6606334841629,
+      "grad_norm": 0.8104879856109619,
+      "learning_rate": 0.001,
+      "loss": 0.5721,
+      "step": 231000
+    },
+    {
+      "epoch": 74.69295410471881,
+      "grad_norm": 0.9081707000732422,
+      "learning_rate": 0.001,
+      "loss": 0.571,
+      "step": 231100
+    },
+    {
+      "epoch": 74.72527472527473,
+      "grad_norm": 1.0820246934890747,
+      "learning_rate": 0.001,
+      "loss": 0.5718,
+      "step": 231200
+    },
+    {
+      "epoch": 74.75759534583064,
+      "grad_norm": 1.142275333404541,
+      "learning_rate": 0.001,
+      "loss": 0.5665,
+      "step": 231300
+    },
+    {
+      "epoch": 74.78991596638656,
+      "grad_norm": 0.747945249080658,
+      "learning_rate": 0.001,
+      "loss": 0.5796,
+      "step": 231400
+    },
+    {
+      "epoch": 74.82223658694247,
+      "grad_norm": 1.0675921440124512,
+      "learning_rate": 0.001,
+      "loss": 0.5735,
+      "step": 231500
+    },
+    {
+      "epoch": 74.85455720749839,
+      "grad_norm": 1.234141230583191,
+      "learning_rate": 0.001,
+      "loss": 0.5672,
+      "step": 231600
+    },
+    {
+      "epoch": 74.8868778280543,
+      "grad_norm": 1.0947504043579102,
+      "learning_rate": 0.001,
+      "loss": 0.5764,
+      "step": 231700
+    },
+    {
+      "epoch": 74.91919844861022,
+      "grad_norm": 1.1369717121124268,
+      "learning_rate": 0.001,
+      "loss": 0.5814,
+      "step": 231800
+    },
+    {
+      "epoch": 74.95151906916612,
+      "grad_norm": 1.0353525876998901,
+      "learning_rate": 0.001,
+      "loss": 0.5731,
+      "step": 231900
+    },
+    {
+      "epoch": 74.98383968972205,
+      "grad_norm": 1.0561206340789795,
+      "learning_rate": 0.001,
+      "loss": 0.5865,
+      "step": 232000
+    },
+    {
+      "epoch": 75.01616031027795,
+      "grad_norm": 1.0930769443511963,
+      "learning_rate": 0.001,
+      "loss": 0.5193,
+      "step": 232100
+    },
+    {
+      "epoch": 75.04848093083388,
+      "grad_norm": 0.1771896481513977,
+      "learning_rate": 0.001,
+      "loss": 0.5157,
+      "step": 232200
+    },
+    {
+      "epoch": 75.08080155138978,
+      "grad_norm": 1.141357660293579,
+      "learning_rate": 0.001,
+      "loss": 0.5361,
+      "step": 232300
+    },
+    {
+      "epoch": 75.1131221719457,
+      "grad_norm": 0.5655999779701233,
+      "learning_rate": 0.001,
+      "loss": 0.5407,
+      "step": 232400
+    },
+    {
+      "epoch": 75.14544279250161,
+      "grad_norm": 0.7469679713249207,
+      "learning_rate": 0.001,
+      "loss": 0.5313,
+      "step": 232500
+    },
+    {
+      "epoch": 75.17776341305753,
+      "grad_norm": 0.29743582010269165,
+      "learning_rate": 0.001,
+      "loss": 0.5311,
+      "step": 232600
+    },
+    {
+      "epoch": 75.21008403361344,
+      "grad_norm": 0.19023145735263824,
+      "learning_rate": 0.001,
+      "loss": 0.5385,
+      "step": 232700
+    },
+    {
+      "epoch": 75.24240465416936,
+      "grad_norm": 0.530585765838623,
+      "learning_rate": 0.001,
+      "loss": 0.5263,
+      "step": 232800
+    },
+    {
+      "epoch": 75.27472527472527,
+      "grad_norm": 0.21099801361560822,
+      "learning_rate": 0.001,
+      "loss": 0.5281,
+      "step": 232900
+    },
+    {
+      "epoch": 75.30704589528119,
+      "grad_norm": 0.460519015789032,
+      "learning_rate": 0.001,
+      "loss": 0.5299,
+      "step": 233000
+    },
+    {
+      "epoch": 75.3393665158371,
+      "grad_norm": 0.435160756111145,
+      "learning_rate": 0.001,
+      "loss": 0.5394,
+      "step": 233100
+    },
+    {
+      "epoch": 75.37168713639302,
+      "grad_norm": 0.1967736929655075,
+      "learning_rate": 0.001,
+      "loss": 0.5525,
+      "step": 233200
+    },
+    {
+      "epoch": 75.40400775694893,
+      "grad_norm": 1.248533844947815,
+      "learning_rate": 0.001,
+      "loss": 0.5491,
+      "step": 233300
+    },
+    {
+      "epoch": 75.43632837750485,
+      "grad_norm": 0.20162171125411987,
+      "learning_rate": 0.001,
+      "loss": 0.5436,
+      "step": 233400
+    },
+    {
+      "epoch": 75.46864899806076,
+      "grad_norm": 0.5278852581977844,
+      "learning_rate": 0.001,
+      "loss": 0.5521,
+      "step": 233500
+    },
+    {
+      "epoch": 75.50096961861668,
+      "grad_norm": 0.3178282380104065,
+      "learning_rate": 0.001,
+      "loss": 0.5498,
+      "step": 233600
+    },
+    {
+      "epoch": 75.53329023917259,
+      "grad_norm": 0.2286689579486847,
+      "learning_rate": 0.001,
+      "loss": 0.5596,
+      "step": 233700
+    },
+    {
+      "epoch": 75.56561085972851,
+      "grad_norm": 0.18539604544639587,
+      "learning_rate": 0.001,
+      "loss": 0.5556,
+      "step": 233800
+    },
+    {
+      "epoch": 75.59793148028442,
+      "grad_norm": 0.07810671627521515,
+      "learning_rate": 0.001,
+      "loss": 0.5644,
+      "step": 233900
+    },
+    {
+      "epoch": 75.63025210084034,
+      "grad_norm": 0.5242074131965637,
+      "learning_rate": 0.001,
+      "loss": 0.5799,
+      "step": 234000
+    },
+    {
+      "epoch": 75.66257272139624,
+      "grad_norm": 0.18300612270832062,
+      "learning_rate": 0.001,
+      "loss": 0.5625,
+      "step": 234100
+    },
+    {
+      "epoch": 75.69489334195217,
+      "grad_norm": 0.3278989791870117,
+      "learning_rate": 0.001,
+      "loss": 0.5732,
+      "step": 234200
+    },
+    {
+      "epoch": 75.72721396250807,
+      "grad_norm": 0.913730800151825,
+      "learning_rate": 0.001,
+      "loss": 0.5817,
+      "step": 234300
+    },
+    {
+      "epoch": 75.759534583064,
+      "grad_norm": 0.3457682132720947,
+      "learning_rate": 0.001,
+      "loss": 0.5654,
+      "step": 234400
+    },
+    {
+      "epoch": 75.7918552036199,
+      "grad_norm": 0.430754154920578,
+      "learning_rate": 0.001,
+      "loss": 0.5754,
+      "step": 234500
+    },
+    {
+      "epoch": 75.82417582417582,
+      "grad_norm": 0.433608740568161,
+      "learning_rate": 0.001,
+      "loss": 0.5754,
+      "step": 234600
+    },
+    {
+      "epoch": 75.85649644473173,
+      "grad_norm": 0.19627642631530762,
+      "learning_rate": 0.001,
+      "loss": 0.5867,
+      "step": 234700
+    },
+    {
+      "epoch": 75.88881706528765,
+      "grad_norm": 0.11609924584627151,
+      "learning_rate": 0.001,
+      "loss": 0.5757,
+      "step": 234800
+    },
+    {
+      "epoch": 75.92113768584356,
+      "grad_norm": 0.7167121171951294,
+      "learning_rate": 0.001,
+      "loss": 0.5651,
+      "step": 234900
+    },
+    {
+      "epoch": 75.95345830639948,
+      "grad_norm": 0.4575307369232178,
+      "learning_rate": 0.001,
+      "loss": 0.5579,
+      "step": 235000
+    },
+    {
+      "epoch": 75.98577892695539,
+      "grad_norm": 0.15593650937080383,
+      "learning_rate": 0.001,
+      "loss": 0.5828,
+      "step": 235100
+    },
+    {
+      "epoch": 76.01809954751131,
+      "grad_norm": 1.413036584854126,
+      "learning_rate": 0.001,
+      "loss": 0.5612,
+      "step": 235200
+    },
+    {
+      "epoch": 76.05042016806723,
+      "grad_norm": 1.6599562168121338,
+      "learning_rate": 0.001,
+      "loss": 0.5069,
+      "step": 235300
+    },
+    {
+      "epoch": 76.08274078862314,
+      "grad_norm": 1.707543969154358,
+      "learning_rate": 0.001,
+      "loss": 0.5186,
+      "step": 235400
+    },
+    {
+      "epoch": 76.11506140917906,
+      "grad_norm": 1.6075223684310913,
+      "learning_rate": 0.001,
+      "loss": 0.5235,
+      "step": 235500
+    },
+    {
+      "epoch": 76.14738202973497,
+      "grad_norm": 1.6342514753341675,
+      "learning_rate": 0.001,
+      "loss": 0.5195,
+      "step": 235600
+    },
+    {
+      "epoch": 76.17970265029089,
+      "grad_norm": 2.0759618282318115,
+      "learning_rate": 0.001,
+      "loss": 0.5269,
+      "step": 235700
+    },
+    {
+      "epoch": 76.2120232708468,
+      "grad_norm": 1.9232611656188965,
+      "learning_rate": 0.001,
+      "loss": 0.5439,
+      "step": 235800
+    },
+    {
+      "epoch": 76.24434389140272,
+      "grad_norm": 1.413246989250183,
+      "learning_rate": 0.001,
+      "loss": 0.5142,
+      "step": 235900
+    },
+    {
+      "epoch": 76.27666451195863,
+      "grad_norm": 1.647636890411377,
+      "learning_rate": 0.001,
+      "loss": 0.5399,
+      "step": 236000
+    },
+    {
+      "epoch": 76.30898513251455,
+      "grad_norm": 1.4785175323486328,
+      "learning_rate": 0.001,
+      "loss": 0.5389,
+      "step": 236100
+    },
+    {
+      "epoch": 76.34130575307046,
+      "grad_norm": 1.6953210830688477,
+      "learning_rate": 0.001,
+      "loss": 0.5401,
+      "step": 236200
+    },
+    {
+      "epoch": 76.37362637362638,
+      "grad_norm": 1.4326900243759155,
+      "learning_rate": 0.001,
+      "loss": 0.5289,
+      "step": 236300
+    },
+    {
+      "epoch": 76.40594699418229,
+      "grad_norm": 1.54912531375885,
+      "learning_rate": 0.001,
+      "loss": 0.5453,
+      "step": 236400
+    },
+    {
+      "epoch": 76.43826761473821,
+      "grad_norm": 1.8962860107421875,
+      "learning_rate": 0.001,
+      "loss": 0.5377,
+      "step": 236500
+    },
+    {
+      "epoch": 76.47058823529412,
+      "grad_norm": 1.2893810272216797,
+      "learning_rate": 0.001,
+      "loss": 0.5392,
+      "step": 236600
+    },
+    {
+      "epoch": 76.50290885585004,
+      "grad_norm": 1.5049962997436523,
+      "learning_rate": 0.001,
+      "loss": 0.5516,
+      "step": 236700
+    },
+    {
+      "epoch": 76.53522947640595,
+      "grad_norm": 1.5820329189300537,
+      "learning_rate": 0.001,
+      "loss": 0.5614,
+      "step": 236800
+    },
+    {
+      "epoch": 76.56755009696187,
+      "grad_norm": 1.3666566610336304,
+      "learning_rate": 0.001,
+      "loss": 0.5431,
+      "step": 236900
+    },
+    {
+      "epoch": 76.59987071751777,
+      "grad_norm": 1.2955440282821655,
+      "learning_rate": 0.001,
+      "loss": 0.5587,
+      "step": 237000
+    },
+    {
+      "epoch": 76.6321913380737,
+      "grad_norm": 1.5790258646011353,
+      "learning_rate": 0.001,
+      "loss": 0.5448,
+      "step": 237100
+    },
+    {
+      "epoch": 76.6645119586296,
+      "grad_norm": 1.5983870029449463,
+      "learning_rate": 0.001,
+      "loss": 0.5588,
+      "step": 237200
+    },
+    {
+      "epoch": 76.69683257918552,
+      "grad_norm": 1.6741951704025269,
+      "learning_rate": 0.001,
+      "loss": 0.5467,
+      "step": 237300
+    },
+    {
+      "epoch": 76.72915319974143,
+      "grad_norm": 1.8480662107467651,
+      "learning_rate": 0.001,
+      "loss": 0.5635,
+      "step": 237400
+    },
+    {
+      "epoch": 76.76147382029735,
+      "grad_norm": 1.6002708673477173,
+      "learning_rate": 0.001,
+      "loss": 0.5654,
+      "step": 237500
+    },
+    {
+      "epoch": 76.79379444085326,
+      "grad_norm": 1.3160319328308105,
+      "learning_rate": 0.001,
+      "loss": 0.561,
+      "step": 237600
+    },
+    {
+      "epoch": 76.82611506140918,
+      "grad_norm": 1.7721960544586182,
+      "learning_rate": 0.001,
+      "loss": 0.5754,
+      "step": 237700
+    },
+    {
+      "epoch": 76.85843568196509,
+      "grad_norm": 1.5011405944824219,
+      "learning_rate": 0.001,
+      "loss": 0.5592,
+      "step": 237800
+    },
+    {
+      "epoch": 76.89075630252101,
+      "grad_norm": 1.771674633026123,
+      "learning_rate": 0.001,
+      "loss": 0.5749,
+      "step": 237900
+    },
+    {
+      "epoch": 76.92307692307692,
+      "grad_norm": 1.3502850532531738,
+      "learning_rate": 0.001,
+      "loss": 0.5739,
+      "step": 238000
+    },
+    {
+      "epoch": 76.95539754363284,
+      "grad_norm": 1.7754833698272705,
+      "learning_rate": 0.001,
+      "loss": 0.5752,
+      "step": 238100
+    },
+    {
+      "epoch": 76.98771816418875,
+      "grad_norm": 1.694458246231079,
+      "learning_rate": 0.001,
+      "loss": 0.5826,
+      "step": 238200
+    },
+    {
+      "epoch": 77.02003878474467,
+      "grad_norm": 1.396079421043396,
+      "learning_rate": 0.001,
+      "loss": 0.5537,
+      "step": 238300
+    },
+    {
+      "epoch": 77.05235940530058,
+      "grad_norm": 1.532308578491211,
+      "learning_rate": 0.001,
+      "loss": 0.5143,
+      "step": 238400
+    },
+    {
+      "epoch": 77.0846800258565,
+      "grad_norm": 1.5896692276000977,
+      "learning_rate": 0.001,
+      "loss": 0.5114,
+      "step": 238500
+    },
+    {
+      "epoch": 77.11700064641241,
+      "grad_norm": 1.4605082273483276,
+      "learning_rate": 0.001,
+      "loss": 0.512,
+      "step": 238600
+    },
+    {
+      "epoch": 77.14932126696833,
+      "grad_norm": 1.233646035194397,
+      "learning_rate": 0.001,
+      "loss": 0.515,
+      "step": 238700
+    },
+    {
+      "epoch": 77.18164188752424,
+      "grad_norm": 1.7335985898971558,
+      "learning_rate": 0.001,
+      "loss": 0.5205,
+      "step": 238800
+    },
+    {
+      "epoch": 77.21396250808016,
+      "grad_norm": 1.4268879890441895,
+      "learning_rate": 0.001,
+      "loss": 0.5347,
+      "step": 238900
+    },
+    {
+      "epoch": 77.24628312863607,
+      "grad_norm": 1.8350880146026611,
+      "learning_rate": 0.001,
+      "loss": 0.5227,
+      "step": 239000
+    },
+    {
+      "epoch": 77.27860374919199,
+      "grad_norm": 1.3111212253570557,
+      "learning_rate": 0.001,
+      "loss": 0.512,
+      "step": 239100
+    },
+    {
+      "epoch": 77.3109243697479,
+      "grad_norm": 1.2222663164138794,
+      "learning_rate": 0.001,
+      "loss": 0.5264,
+      "step": 239200
+    },
+    {
+      "epoch": 77.34324499030382,
+      "grad_norm": 1.6778181791305542,
+      "learning_rate": 0.001,
+      "loss": 0.525,
+      "step": 239300
+    },
+    {
+      "epoch": 77.37556561085972,
+      "grad_norm": 1.714513897895813,
+      "learning_rate": 0.001,
+      "loss": 0.53,
+      "step": 239400
+    },
+    {
+      "epoch": 77.40788623141565,
+      "grad_norm": 1.4410483837127686,
+      "learning_rate": 0.001,
+      "loss": 0.5352,
+      "step": 239500
+    },
+    {
+      "epoch": 77.44020685197155,
+      "grad_norm": 1.4495421648025513,
+      "learning_rate": 0.001,
+      "loss": 0.5429,
+      "step": 239600
+    },
+    {
+      "epoch": 77.47252747252747,
+      "grad_norm": 1.4672760963439941,
+      "learning_rate": 0.001,
+      "loss": 0.5369,
+      "step": 239700
+    },
+    {
+      "epoch": 77.50484809308338,
+      "grad_norm": 1.186777114868164,
+      "learning_rate": 0.001,
+      "loss": 0.5525,
+      "step": 239800
+    },
+    {
+      "epoch": 77.5371687136393,
+      "grad_norm": 1.2865841388702393,
+      "learning_rate": 0.001,
+      "loss": 0.5406,
+      "step": 239900
+    },
+    {
+      "epoch": 77.56948933419521,
+      "grad_norm": 1.964085340499878,
+      "learning_rate": 0.001,
+      "loss": 0.5471,
+      "step": 240000
+    },
+    {
+      "epoch": 77.60180995475113,
+      "grad_norm": 1.2692760229110718,
+      "learning_rate": 0.001,
+      "loss": 0.5511,
+      "step": 240100
+    },
+    {
+      "epoch": 77.63413057530704,
+      "grad_norm": 1.427521824836731,
+      "learning_rate": 0.001,
+      "loss": 0.5463,
+      "step": 240200
+    },
+    {
+      "epoch": 77.66645119586296,
+      "grad_norm": 1.5266354084014893,
+      "learning_rate": 0.001,
+      "loss": 0.5682,
+      "step": 240300
+    },
+    {
+      "epoch": 77.69877181641887,
+      "grad_norm": 1.3747498989105225,
+      "learning_rate": 0.001,
+      "loss": 0.5447,
+      "step": 240400
+    },
+    {
+      "epoch": 77.73109243697479,
+      "grad_norm": 1.8401381969451904,
+      "learning_rate": 0.001,
+      "loss": 0.5509,
+      "step": 240500
+    },
+    {
+      "epoch": 77.7634130575307,
+      "grad_norm": 1.593461513519287,
+      "learning_rate": 0.001,
+      "loss": 0.569,
+      "step": 240600
+    },
+    {
+      "epoch": 77.79573367808662,
+      "grad_norm": 1.2960352897644043,
+      "learning_rate": 0.001,
+      "loss": 0.5557,
+      "step": 240700
+    },
+    {
+      "epoch": 77.82805429864253,
+      "grad_norm": 1.2029883861541748,
+      "learning_rate": 0.001,
+      "loss": 0.5646,
+      "step": 240800
+    },
+    {
+      "epoch": 77.86037491919845,
+      "grad_norm": 1.672709584236145,
+      "learning_rate": 0.001,
+      "loss": 0.5518,
+      "step": 240900
+    },
+    {
+      "epoch": 77.89269553975436,
+      "grad_norm": 1.5223016738891602,
+      "learning_rate": 0.001,
+      "loss": 0.5597,
+      "step": 241000
+    },
+    {
+      "epoch": 77.92501616031028,
+      "grad_norm": 1.2413277626037598,
+      "learning_rate": 0.001,
+      "loss": 0.5648,
+      "step": 241100
+    },
+    {
+      "epoch": 77.95733678086619,
+      "grad_norm": 1.4993832111358643,
+      "learning_rate": 0.001,
+      "loss": 0.571,
+      "step": 241200
+    },
+    {
+      "epoch": 77.98965740142211,
+      "grad_norm": 1.328671932220459,
+      "learning_rate": 0.001,
+      "loss": 0.5648,
+      "step": 241300
+    },
+    {
+      "epoch": 78.02197802197803,
+      "grad_norm": 1.6180659532546997,
+      "learning_rate": 0.001,
+      "loss": 0.5366,
+      "step": 241400
+    },
+    {
+      "epoch": 78.05429864253394,
+      "grad_norm": 1.5706161260604858,
+      "learning_rate": 0.001,
+      "loss": 0.5075,
+      "step": 241500
+    },
+    {
+      "epoch": 78.08661926308986,
+      "grad_norm": 1.5466357469558716,
+      "learning_rate": 0.001,
+      "loss": 0.4995,
+      "step": 241600
+    },
+    {
+      "epoch": 78.11893988364577,
+      "grad_norm": 1.4233200550079346,
+      "learning_rate": 0.001,
+      "loss": 0.507,
+      "step": 241700
+    },
+    {
+      "epoch": 78.15126050420169,
+      "grad_norm": 3.640650749206543,
+      "learning_rate": 0.001,
+      "loss": 0.5213,
+      "step": 241800
+    },
+    {
+      "epoch": 78.1835811247576,
+      "grad_norm": 1.5065131187438965,
+      "learning_rate": 0.001,
+      "loss": 0.5084,
+      "step": 241900
+    },
+    {
+      "epoch": 78.21590174531352,
+      "grad_norm": 1.3846442699432373,
+      "learning_rate": 0.001,
+      "loss": 0.5188,
+      "step": 242000
+    },
+    {
+      "epoch": 78.24822236586942,
+      "grad_norm": 1.592302918434143,
+      "learning_rate": 0.001,
+      "loss": 0.5206,
+      "step": 242100
+    },
+    {
+      "epoch": 78.28054298642535,
+      "grad_norm": 1.4627457857131958,
+      "learning_rate": 0.001,
+      "loss": 0.5306,
+      "step": 242200
+    },
+    {
+      "epoch": 78.31286360698125,
+      "grad_norm": 1.4785966873168945,
+      "learning_rate": 0.001,
+      "loss": 0.5259,
+      "step": 242300
+    },
+    {
+      "epoch": 78.34518422753717,
+      "grad_norm": 1.6304181814193726,
+      "learning_rate": 0.001,
+      "loss": 0.5293,
+      "step": 242400
+    },
+    {
+      "epoch": 78.37750484809308,
+      "grad_norm": 1.4955110549926758,
+      "learning_rate": 0.001,
+      "loss": 0.5267,
+      "step": 242500
+    },
+    {
+      "epoch": 78.409825468649,
+      "grad_norm": 1.3782621622085571,
+      "learning_rate": 0.001,
+      "loss": 0.5309,
+      "step": 242600
+    },
+    {
+      "epoch": 78.44214608920491,
+      "grad_norm": 1.5828373432159424,
+      "learning_rate": 0.001,
+      "loss": 0.5314,
+      "step": 242700
+    },
+    {
+      "epoch": 78.47446670976083,
+      "grad_norm": 2.112483024597168,
+      "learning_rate": 0.001,
+      "loss": 0.5381,
+      "step": 242800
+    },
+    {
+      "epoch": 78.50678733031674,
+      "grad_norm": 1.565582036972046,
+      "learning_rate": 0.001,
+      "loss": 0.5339,
+      "step": 242900
+    },
+    {
+      "epoch": 78.53910795087266,
+      "grad_norm": 1.747968316078186,
+      "learning_rate": 0.001,
+      "loss": 0.5372,
+      "step": 243000
+    },
+    {
+      "epoch": 78.57142857142857,
+      "grad_norm": 1.6491109132766724,
+      "learning_rate": 0.001,
+      "loss": 0.5333,
+      "step": 243100
+    },
+    {
+      "epoch": 78.60374919198449,
+      "grad_norm": 1.4201022386550903,
+      "learning_rate": 0.001,
+      "loss": 0.5317,
+      "step": 243200
+    },
+    {
+      "epoch": 78.6360698125404,
+      "grad_norm": 1.7670608758926392,
+      "learning_rate": 0.001,
+      "loss": 0.5339,
+      "step": 243300
+    },
+    {
+      "epoch": 78.66839043309632,
+      "grad_norm": 1.3140902519226074,
+      "learning_rate": 0.001,
+      "loss": 0.539,
+      "step": 243400
+    },
+    {
+      "epoch": 78.70071105365223,
+      "grad_norm": 1.4319005012512207,
+      "learning_rate": 0.001,
+      "loss": 0.5504,
+      "step": 243500
+    },
+    {
+      "epoch": 78.73303167420815,
+      "grad_norm": 1.4455088376998901,
+      "learning_rate": 0.001,
+      "loss": 0.5456,
+      "step": 243600
+    },
+    {
+      "epoch": 78.76535229476406,
+      "grad_norm": 1.309688687324524,
+      "learning_rate": 0.001,
+      "loss": 0.557,
+      "step": 243700
+    },
+    {
+      "epoch": 78.79767291531998,
+      "grad_norm": 1.356952428817749,
+      "learning_rate": 0.001,
+      "loss": 0.5575,
+      "step": 243800
+    },
+    {
+      "epoch": 78.82999353587589,
+      "grad_norm": 1.2032933235168457,
+      "learning_rate": 0.001,
+      "loss": 0.5477,
+      "step": 243900
+    },
+    {
+      "epoch": 78.86231415643181,
+      "grad_norm": 1.314985990524292,
+      "learning_rate": 0.001,
+      "loss": 0.554,
+      "step": 244000
+    },
+    {
+      "epoch": 78.89463477698771,
+      "grad_norm": 1.465287685394287,
+      "learning_rate": 0.001,
+      "loss": 0.5621,
+      "step": 244100
+    },
+    {
+      "epoch": 78.92695539754364,
+      "grad_norm": 1.4359557628631592,
+      "learning_rate": 0.001,
+      "loss": 0.5535,
+      "step": 244200
+    },
+    {
+      "epoch": 78.95927601809954,
+      "grad_norm": 1.966088056564331,
+      "learning_rate": 0.001,
+      "loss": 0.5639,
+      "step": 244300
+    },
+    {
+      "epoch": 78.99159663865547,
+      "grad_norm": 1.3312709331512451,
+      "learning_rate": 0.001,
+      "loss": 0.5657,
+      "step": 244400
+    },
+    {
+      "epoch": 79.02391725921137,
+      "grad_norm": 1.2351542711257935,
+      "learning_rate": 0.001,
+      "loss": 0.5102,
+      "step": 244500
+    },
+    {
+      "epoch": 79.0562378797673,
+      "grad_norm": 1.573776364326477,
+      "learning_rate": 0.001,
+      "loss": 0.4925,
+      "step": 244600
+    },
+    {
+      "epoch": 79.0885585003232,
+      "grad_norm": 1.472129225730896,
+      "learning_rate": 0.001,
+      "loss": 0.4985,
+      "step": 244700
+    },
+    {
+      "epoch": 79.12087912087912,
+      "grad_norm": 1.5297859907150269,
+      "learning_rate": 0.001,
+      "loss": 0.5091,
+      "step": 244800
+    },
+    {
+      "epoch": 79.15319974143503,
+      "grad_norm": 1.306641697883606,
+      "learning_rate": 0.001,
+      "loss": 0.5116,
+      "step": 244900
+    },
+    {
+      "epoch": 79.18552036199095,
+      "grad_norm": 1.8457226753234863,
+      "learning_rate": 0.001,
+      "loss": 0.5082,
+      "step": 245000
+    },
+    {
+      "epoch": 79.21784098254686,
+      "grad_norm": 1.9189332723617554,
+      "learning_rate": 0.001,
+      "loss": 0.5053,
+      "step": 245100
+    },
+    {
+      "epoch": 79.25016160310278,
+      "grad_norm": 1.3402392864227295,
+      "learning_rate": 0.001,
+      "loss": 0.5216,
+      "step": 245200
+    },
+    {
+      "epoch": 79.28248222365869,
+      "grad_norm": 1.5878645181655884,
+      "learning_rate": 0.001,
+      "loss": 0.5042,
+      "step": 245300
+    },
+    {
+      "epoch": 79.31480284421461,
+      "grad_norm": 1.192588448524475,
+      "learning_rate": 0.001,
+      "loss": 0.5133,
+      "step": 245400
+    },
+    {
+      "epoch": 79.34712346477052,
+      "grad_norm": 1.310906171798706,
+      "learning_rate": 0.001,
+      "loss": 0.522,
+      "step": 245500
+    },
+    {
+      "epoch": 79.37944408532644,
+      "grad_norm": 1.6999276876449585,
+      "learning_rate": 0.001,
+      "loss": 0.5244,
+      "step": 245600
+    },
+    {
+      "epoch": 79.41176470588235,
+      "grad_norm": 1.4118130207061768,
+      "learning_rate": 0.001,
+      "loss": 0.5215,
+      "step": 245700
+    },
+    {
+      "epoch": 79.44408532643827,
+      "grad_norm": 1.5291193723678589,
+      "learning_rate": 0.001,
+      "loss": 0.5291,
+      "step": 245800
+    },
+    {
+      "epoch": 79.47640594699418,
+      "grad_norm": 1.4727531671524048,
+      "learning_rate": 0.001,
+      "loss": 0.5372,
+      "step": 245900
+    },
+    {
+      "epoch": 79.5087265675501,
+      "grad_norm": 1.576590657234192,
+      "learning_rate": 0.001,
+      "loss": 0.5243,
+      "step": 246000
+    },
+    {
+      "epoch": 79.541047188106,
+      "grad_norm": 1.39322030544281,
+      "learning_rate": 0.001,
+      "loss": 0.5249,
+      "step": 246100
+    },
+    {
+      "epoch": 79.57336780866193,
+      "grad_norm": 1.2230325937271118,
+      "learning_rate": 0.001,
+      "loss": 0.5388,
+      "step": 246200
+    },
+    {
+      "epoch": 79.60568842921784,
+      "grad_norm": 1.5058960914611816,
+      "learning_rate": 0.001,
+      "loss": 0.5271,
+      "step": 246300
+    },
+    {
+      "epoch": 79.63800904977376,
+      "grad_norm": 1.4781763553619385,
+      "learning_rate": 0.001,
+      "loss": 0.5506,
+      "step": 246400
+    },
+    {
+      "epoch": 79.67032967032966,
+      "grad_norm": 1.458723783493042,
+      "learning_rate": 0.001,
+      "loss": 0.5491,
+      "step": 246500
+    },
+    {
+      "epoch": 79.70265029088559,
+      "grad_norm": 1.423794150352478,
+      "learning_rate": 0.001,
+      "loss": 0.5513,
+      "step": 246600
+    },
+    {
+      "epoch": 79.7349709114415,
+      "grad_norm": 1.9729347229003906,
+      "learning_rate": 0.001,
+      "loss": 0.5496,
+      "step": 246700
+    },
+    {
+      "epoch": 79.76729153199742,
+      "grad_norm": 1.5807349681854248,
+      "learning_rate": 0.001,
+      "loss": 0.5364,
+      "step": 246800
+    },
+    {
+      "epoch": 79.79961215255332,
+      "grad_norm": 1.4756208658218384,
+      "learning_rate": 0.001,
+      "loss": 0.5361,
+      "step": 246900
+    },
+    {
+      "epoch": 79.83193277310924,
+      "grad_norm": 2.0747814178466797,
+      "learning_rate": 0.001,
+      "loss": 0.5479,
+      "step": 247000
+    },
+    {
+      "epoch": 79.86425339366515,
+      "grad_norm": 1.1575489044189453,
+      "learning_rate": 0.001,
+      "loss": 0.5413,
+      "step": 247100
+    },
+    {
+      "epoch": 79.89657401422107,
+      "grad_norm": 1.3140865564346313,
+      "learning_rate": 0.001,
+      "loss": 0.561,
+      "step": 247200
+    },
+    {
+      "epoch": 79.92889463477698,
+      "grad_norm": 1.5108518600463867,
+      "learning_rate": 0.001,
+      "loss": 0.5559,
+      "step": 247300
+    },
+    {
+      "epoch": 79.9612152553329,
+      "grad_norm": 1.3009579181671143,
+      "learning_rate": 0.001,
+      "loss": 0.5639,
+      "step": 247400
+    },
+    {
+      "epoch": 79.99353587588882,
+      "grad_norm": 1.4407790899276733,
+      "learning_rate": 0.001,
+      "loss": 0.5556,
+      "step": 247500
+    },
+    {
+      "epoch": 80.02585649644473,
+      "grad_norm": 4.573031902313232,
+      "learning_rate": 0.001,
+      "loss": 0.5021,
+      "step": 247600
+    },
+    {
+      "epoch": 80.05817711700065,
+      "grad_norm": 1.7188622951507568,
+      "learning_rate": 0.001,
+      "loss": 0.4948,
+      "step": 247700
+    },
+    {
+      "epoch": 80.09049773755656,
+      "grad_norm": 1.1180098056793213,
+      "learning_rate": 0.001,
+      "loss": 0.4918,
+      "step": 247800
+    },
+    {
+      "epoch": 80.12281835811248,
+      "grad_norm": 1.9229891300201416,
+      "learning_rate": 0.001,
+      "loss": 0.5011,
+      "step": 247900
+    },
+    {
+      "epoch": 80.15513897866839,
+      "grad_norm": 1.5299087762832642,
+      "learning_rate": 0.001,
+      "loss": 0.4973,
+      "step": 248000
+    },
+    {
+      "epoch": 80.18745959922431,
+      "grad_norm": 1.6426825523376465,
+      "learning_rate": 0.001,
+      "loss": 0.5023,
+      "step": 248100
+    },
+    {
+      "epoch": 80.21978021978022,
+      "grad_norm": 1.409515142440796,
+      "learning_rate": 0.001,
+      "loss": 0.5165,
+      "step": 248200
+    },
+    {
+      "epoch": 80.25210084033614,
+      "grad_norm": 1.6629232168197632,
+      "learning_rate": 0.001,
+      "loss": 0.5056,
+      "step": 248300
+    },
+    {
+      "epoch": 80.28442146089205,
+      "grad_norm": 1.4850993156433105,
+      "learning_rate": 0.001,
+      "loss": 0.5186,
+      "step": 248400
+    },
+    {
+      "epoch": 80.31674208144797,
+      "grad_norm": 1.3737995624542236,
+      "learning_rate": 0.001,
+      "loss": 0.52,
+      "step": 248500
+    },
+    {
+      "epoch": 80.34906270200388,
+      "grad_norm": 1.550413966178894,
+      "learning_rate": 0.001,
+      "loss": 0.5109,
+      "step": 248600
+    },
+    {
+      "epoch": 80.3813833225598,
+      "grad_norm": 1.4993878602981567,
+      "learning_rate": 0.001,
+      "loss": 0.521,
+      "step": 248700
+    },
+    {
+      "epoch": 80.4137039431157,
+      "grad_norm": 1.1930394172668457,
+      "learning_rate": 0.001,
+      "loss": 0.5255,
+      "step": 248800
+    },
+    {
+      "epoch": 80.44602456367163,
+      "grad_norm": 1.3800441026687622,
+      "learning_rate": 0.001,
+      "loss": 0.525,
+      "step": 248900
+    },
+    {
+      "epoch": 80.47834518422754,
+      "grad_norm": 1.4232189655303955,
+      "learning_rate": 0.001,
+      "loss": 0.522,
+      "step": 249000
+    },
+    {
+      "epoch": 80.51066580478346,
+      "grad_norm": 1.3573640584945679,
+      "learning_rate": 0.001,
+      "loss": 0.524,
+      "step": 249100
+    },
+    {
+      "epoch": 80.54298642533936,
+      "grad_norm": 1.6436176300048828,
+      "learning_rate": 0.001,
+      "loss": 0.5277,
+      "step": 249200
+    },
+    {
+      "epoch": 80.57530704589529,
+      "grad_norm": 1.6628504991531372,
+      "learning_rate": 0.001,
+      "loss": 0.5448,
+      "step": 249300
+    },
+    {
+      "epoch": 80.6076276664512,
+      "grad_norm": 1.403755784034729,
+      "learning_rate": 0.001,
+      "loss": 0.5198,
+      "step": 249400
+    },
+    {
+      "epoch": 80.63994828700712,
+      "grad_norm": 1.6308602094650269,
+      "learning_rate": 0.001,
+      "loss": 0.5291,
+      "step": 249500
+    },
+    {
+      "epoch": 80.67226890756302,
+      "grad_norm": 1.2523367404937744,
+      "learning_rate": 0.001,
+      "loss": 0.5306,
+      "step": 249600
+    },
+    {
+      "epoch": 80.70458952811894,
+      "grad_norm": 1.445035457611084,
+      "learning_rate": 0.001,
+      "loss": 0.542,
+      "step": 249700
+    },
+    {
+      "epoch": 80.73691014867485,
+      "grad_norm": 1.7217998504638672,
+      "learning_rate": 0.001,
+      "loss": 0.5308,
+      "step": 249800
+    },
+    {
+      "epoch": 80.76923076923077,
+      "grad_norm": 1.1785924434661865,
+      "learning_rate": 0.001,
+      "loss": 0.5486,
+      "step": 249900
+    },
+    {
+      "epoch": 80.80155138978668,
+      "grad_norm": 1.978576898574829,
+      "learning_rate": 0.001,
+      "loss": 0.5426,
+      "step": 250000
+    },
+    {
+      "epoch": 80.8338720103426,
+      "grad_norm": 1.4352853298187256,
+      "learning_rate": 0.001,
+      "loss": 0.5462,
+      "step": 250100
+    },
+    {
+      "epoch": 80.86619263089851,
+      "grad_norm": 1.4232949018478394,
+      "learning_rate": 0.001,
+      "loss": 0.5443,
+      "step": 250200
+    },
+    {
+      "epoch": 80.89851325145443,
+      "grad_norm": 2.1108646392822266,
+      "learning_rate": 0.001,
+      "loss": 0.5451,
+      "step": 250300
+    },
+    {
+      "epoch": 80.93083387201034,
+      "grad_norm": 1.9700771570205688,
+      "learning_rate": 0.001,
+      "loss": 0.5406,
+      "step": 250400
+    },
+    {
+      "epoch": 80.96315449256626,
+      "grad_norm": 1.5097615718841553,
+      "learning_rate": 0.001,
+      "loss": 0.5569,
+      "step": 250500
+    },
+    {
+      "epoch": 80.99547511312217,
+      "grad_norm": 1.2994375228881836,
+      "learning_rate": 0.001,
+      "loss": 0.5285,
+      "step": 250600
+    },
+    {
+      "epoch": 81.02779573367809,
+      "grad_norm": 1.177480697631836,
+      "learning_rate": 0.001,
+      "loss": 0.4835,
+      "step": 250700
+    },
+    {
+      "epoch": 81.060116354234,
+      "grad_norm": 1.5064599514007568,
+      "learning_rate": 0.001,
+      "loss": 0.488,
+      "step": 250800
+    },
+    {
+      "epoch": 81.09243697478992,
+      "grad_norm": 1.557698369026184,
+      "learning_rate": 0.001,
+      "loss": 0.49,
+      "step": 250900
+    },
+    {
+      "epoch": 81.12475759534583,
+      "grad_norm": 1.3125524520874023,
+      "learning_rate": 0.001,
+      "loss": 0.5003,
+      "step": 251000
+    },
+    {
+      "epoch": 81.15707821590175,
+      "grad_norm": 1.453548550605774,
+      "learning_rate": 0.001,
+      "loss": 0.5017,
+      "step": 251100
+    },
+    {
+      "epoch": 81.18939883645766,
+      "grad_norm": 1.3550654649734497,
+      "learning_rate": 0.001,
+      "loss": 0.5051,
+      "step": 251200
+    },
+    {
+      "epoch": 81.22171945701358,
+      "grad_norm": 1.2782098054885864,
+      "learning_rate": 0.001,
+      "loss": 0.4999,
+      "step": 251300
+    },
+    {
+      "epoch": 81.25404007756948,
+      "grad_norm": 1.5780470371246338,
+      "learning_rate": 0.001,
+      "loss": 0.5014,
+      "step": 251400
+    },
+    {
+      "epoch": 81.2863606981254,
+      "grad_norm": 1.6237266063690186,
+      "learning_rate": 0.001,
+      "loss": 0.5,
+      "step": 251500
+    },
+    {
+      "epoch": 81.31868131868131,
+      "grad_norm": 1.1814064979553223,
+      "learning_rate": 0.001,
+      "loss": 0.5206,
+      "step": 251600
+    },
+    {
+      "epoch": 81.35100193923724,
+      "grad_norm": 1.1381033658981323,
+      "learning_rate": 0.001,
+      "loss": 0.4945,
+      "step": 251700
+    },
+    {
+      "epoch": 81.38332255979314,
+      "grad_norm": 1.3088628053665161,
+      "learning_rate": 0.001,
+      "loss": 0.505,
+      "step": 251800
+    },
+    {
+      "epoch": 81.41564318034906,
+      "grad_norm": 1.3342161178588867,
+      "learning_rate": 0.001,
+      "loss": 0.5175,
+      "step": 251900
+    },
+    {
+      "epoch": 81.44796380090497,
+      "grad_norm": 1.3151096105575562,
+      "learning_rate": 0.001,
+      "loss": 0.5176,
+      "step": 252000
+    },
+    {
+      "epoch": 81.4802844214609,
+      "grad_norm": 1.4550706148147583,
+      "learning_rate": 0.001,
+      "loss": 0.5305,
+      "step": 252100
+    },
+    {
+      "epoch": 81.5126050420168,
+      "grad_norm": 1.0697120428085327,
+      "learning_rate": 0.001,
+      "loss": 0.5091,
+      "step": 252200
+    },
+    {
+      "epoch": 81.54492566257272,
+      "grad_norm": 1.358964443206787,
+      "learning_rate": 0.001,
+      "loss": 0.5143,
+      "step": 252300
+    },
+    {
+      "epoch": 81.57724628312863,
+      "grad_norm": 1.603599190711975,
+      "learning_rate": 0.001,
+      "loss": 0.5222,
+      "step": 252400
+    },
+    {
+      "epoch": 81.60956690368455,
+      "grad_norm": 1.2112364768981934,
+      "learning_rate": 0.001,
+      "loss": 0.5381,
+      "step": 252500
+    },
+    {
+      "epoch": 81.64188752424046,
+      "grad_norm": 1.5382639169692993,
+      "learning_rate": 0.001,
+      "loss": 0.52,
+      "step": 252600
+    },
+    {
+      "epoch": 81.67420814479638,
+      "grad_norm": 1.3610308170318604,
+      "learning_rate": 0.001,
+      "loss": 0.5281,
+      "step": 252700
+    },
+    {
+      "epoch": 81.70652876535229,
+      "grad_norm": 1.3466453552246094,
+      "learning_rate": 0.001,
+      "loss": 0.5268,
+      "step": 252800
+    },
+    {
+      "epoch": 81.73884938590821,
+      "grad_norm": 1.0718330144882202,
+      "learning_rate": 0.001,
+      "loss": 0.539,
+      "step": 252900
+    },
+    {
+      "epoch": 81.77117000646412,
+      "grad_norm": 1.501024603843689,
+      "learning_rate": 0.001,
+      "loss": 0.5298,
+      "step": 253000
+    },
+    {
+      "epoch": 81.80349062702004,
+      "grad_norm": 1.89411461353302,
+      "learning_rate": 0.001,
+      "loss": 0.5463,
+      "step": 253100
+    },
+    {
+      "epoch": 81.83581124757595,
+      "grad_norm": 1.5171462297439575,
+      "learning_rate": 0.001,
+      "loss": 0.5408,
+      "step": 253200
+    },
+    {
+      "epoch": 81.86813186813187,
+      "grad_norm": 1.353501796722412,
+      "learning_rate": 0.001,
+      "loss": 0.5524,
+      "step": 253300
+    },
+    {
+      "epoch": 81.90045248868778,
+      "grad_norm": 1.2344218492507935,
+      "learning_rate": 0.001,
+      "loss": 0.5385,
+      "step": 253400
+    },
+    {
+      "epoch": 81.9327731092437,
+      "grad_norm": 1.0975984334945679,
+      "learning_rate": 0.001,
+      "loss": 0.5289,
+      "step": 253500
+    },
+    {
+      "epoch": 81.9650937297996,
+      "grad_norm": 1.1510794162750244,
+      "learning_rate": 0.001,
+      "loss": 0.5447,
+      "step": 253600
+    },
+    {
+      "epoch": 81.99741435035553,
+      "grad_norm": 1.4088082313537598,
+      "learning_rate": 0.001,
+      "loss": 0.5271,
+      "step": 253700
+    },
+    {
+      "epoch": 82.02973497091145,
+      "grad_norm": 1.0249762535095215,
+      "learning_rate": 0.001,
+      "loss": 0.4862,
+      "step": 253800
+    },
+    {
+      "epoch": 82.06205559146736,
+      "grad_norm": 1.7136085033416748,
+      "learning_rate": 0.001,
+      "loss": 0.4803,
+      "step": 253900
+    },
+    {
+      "epoch": 82.09437621202328,
+      "grad_norm": 1.3353995084762573,
+      "learning_rate": 0.001,
+      "loss": 0.4792,
+      "step": 254000
+    },
+    {
+      "epoch": 82.12669683257919,
+      "grad_norm": 1.2580580711364746,
+      "learning_rate": 0.001,
+      "loss": 0.4859,
+      "step": 254100
+    },
+    {
+      "epoch": 82.1590174531351,
+      "grad_norm": 1.9082170724868774,
+      "learning_rate": 0.001,
+      "loss": 0.4976,
+      "step": 254200
+    },
+    {
+      "epoch": 82.19133807369101,
+      "grad_norm": 0.9331086277961731,
+      "learning_rate": 0.001,
+      "loss": 0.4908,
+      "step": 254300
+    },
+    {
+      "epoch": 82.22365869424694,
+      "grad_norm": 1.1894890069961548,
+      "learning_rate": 0.001,
+      "loss": 0.5025,
+      "step": 254400
+    },
+    {
+      "epoch": 82.25597931480284,
+      "grad_norm": 2.082413673400879,
+      "learning_rate": 0.001,
+      "loss": 0.4951,
+      "step": 254500
+    },
+    {
+      "epoch": 82.28829993535876,
+      "grad_norm": 1.085007905960083,
+      "learning_rate": 0.001,
+      "loss": 0.5051,
+      "step": 254600
+    },
+    {
+      "epoch": 82.32062055591467,
+      "grad_norm": 4.038010120391846,
+      "learning_rate": 0.001,
+      "loss": 0.4978,
+      "step": 254700
+    },
+    {
+      "epoch": 82.3529411764706,
+      "grad_norm": 0.947473406791687,
+      "learning_rate": 0.001,
+      "loss": 0.504,
+      "step": 254800
+    },
+    {
+      "epoch": 82.3852617970265,
+      "grad_norm": 0.9688652157783508,
+      "learning_rate": 0.001,
+      "loss": 0.515,
+      "step": 254900
+    },
+    {
+      "epoch": 82.41758241758242,
+      "grad_norm": 1.1633764505386353,
+      "learning_rate": 0.001,
+      "loss": 0.516,
+      "step": 255000
+    },
+    {
+      "epoch": 82.44990303813833,
+      "grad_norm": 1.1820343732833862,
+      "learning_rate": 0.001,
+      "loss": 0.5087,
+      "step": 255100
+    },
+    {
+      "epoch": 82.48222365869425,
+      "grad_norm": 0.9853097796440125,
+      "learning_rate": 0.001,
+      "loss": 0.5236,
+      "step": 255200
+    },
+    {
+      "epoch": 82.51454427925016,
+      "grad_norm": 1.0773764848709106,
+      "learning_rate": 0.001,
+      "loss": 0.5222,
+      "step": 255300
+    },
+    {
+      "epoch": 82.54686489980608,
+      "grad_norm": 1.3068816661834717,
+      "learning_rate": 0.001,
+      "loss": 0.5314,
+      "step": 255400
+    },
+    {
+      "epoch": 82.57918552036199,
+      "grad_norm": 1.1652721166610718,
+      "learning_rate": 0.001,
+      "loss": 0.5161,
+      "step": 255500
+    },
+    {
+      "epoch": 82.61150614091791,
+      "grad_norm": 1.4878023862838745,
+      "learning_rate": 0.001,
+      "loss": 0.5206,
+      "step": 255600
+    },
+    {
+      "epoch": 82.64382676147382,
+      "grad_norm": 0.9717551469802856,
+      "learning_rate": 0.001,
+      "loss": 0.5143,
+      "step": 255700
+    },
+    {
+      "epoch": 82.67614738202974,
+      "grad_norm": 0.9180939793586731,
+      "learning_rate": 0.001,
+      "loss": 0.5245,
+      "step": 255800
+    },
+    {
+      "epoch": 82.70846800258565,
+      "grad_norm": 1.2813619375228882,
+      "learning_rate": 0.001,
+      "loss": 0.5479,
+      "step": 255900
+    },
+    {
+      "epoch": 82.74078862314157,
+      "grad_norm": 1.2449421882629395,
+      "learning_rate": 0.001,
+      "loss": 0.5229,
+      "step": 256000
+    },
+    {
+      "epoch": 82.77310924369748,
+      "grad_norm": 1.3184751272201538,
+      "learning_rate": 0.001,
+      "loss": 0.5303,
+      "step": 256100
+    },
+    {
+      "epoch": 82.8054298642534,
+      "grad_norm": 0.9855522513389587,
+      "learning_rate": 0.001,
+      "loss": 0.5216,
+      "step": 256200
+    },
+    {
+      "epoch": 82.8377504848093,
+      "grad_norm": 1.078248143196106,
+      "learning_rate": 0.001,
+      "loss": 0.5323,
+      "step": 256300
+    },
+    {
+      "epoch": 82.87007110536523,
+      "grad_norm": 1.2535841464996338,
+      "learning_rate": 0.001,
+      "loss": 0.5222,
+      "step": 256400
+    },
+    {
+      "epoch": 82.90239172592113,
+      "grad_norm": 1.2945135831832886,
+      "learning_rate": 0.001,
+      "loss": 0.5212,
+      "step": 256500
+    },
+    {
+      "epoch": 82.93471234647706,
+      "grad_norm": 1.130628228187561,
+      "learning_rate": 0.001,
+      "loss": 0.5398,
+      "step": 256600
+    },
+    {
+      "epoch": 82.96703296703296,
+      "grad_norm": 1.1353565454483032,
+      "learning_rate": 0.001,
+      "loss": 0.5363,
+      "step": 256700
+    },
+    {
+      "epoch": 82.99935358758889,
+      "grad_norm": 0.877821683883667,
+      "learning_rate": 0.001,
+      "loss": 0.5173,
+      "step": 256800
+    },
+    {
+      "epoch": 83.03167420814479,
+      "grad_norm": 1.0781742334365845,
+      "learning_rate": 0.001,
+      "loss": 0.4592,
+      "step": 256900
+    },
+    {
+      "epoch": 83.06399482870071,
+      "grad_norm": 0.8791617751121521,
+      "learning_rate": 0.001,
+      "loss": 0.4768,
+      "step": 257000
+    },
+    {
+      "epoch": 83.09631544925662,
+      "grad_norm": 0.7136475443840027,
+      "learning_rate": 0.001,
+      "loss": 0.4807,
+      "step": 257100
+    },
+    {
+      "epoch": 83.12863606981254,
+      "grad_norm": 1.038556456565857,
+      "learning_rate": 0.001,
+      "loss": 0.4869,
+      "step": 257200
+    },
+    {
+      "epoch": 83.16095669036845,
+      "grad_norm": 0.8827602863311768,
+      "learning_rate": 0.001,
+      "loss": 0.4861,
+      "step": 257300
+    },
+    {
+      "epoch": 83.19327731092437,
+      "grad_norm": 0.8706760406494141,
+      "learning_rate": 0.001,
+      "loss": 0.5013,
+      "step": 257400
+    },
+    {
+      "epoch": 83.22559793148028,
+      "grad_norm": 0.7551383376121521,
+      "learning_rate": 0.001,
+      "loss": 0.4835,
+      "step": 257500
+    },
+    {
+      "epoch": 83.2579185520362,
+      "grad_norm": 0.6588653922080994,
+      "learning_rate": 0.001,
+      "loss": 0.4866,
+      "step": 257600
+    },
+    {
+      "epoch": 83.29023917259211,
+      "grad_norm": 0.7341609597206116,
+      "learning_rate": 0.001,
+      "loss": 0.4937,
+      "step": 257700
+    },
+    {
+      "epoch": 83.32255979314803,
+      "grad_norm": 0.7104642391204834,
+      "learning_rate": 0.001,
+      "loss": 0.4991,
+      "step": 257800
+    },
+    {
+      "epoch": 83.35488041370394,
+      "grad_norm": 1.0989702939987183,
+      "learning_rate": 0.001,
+      "loss": 0.4976,
+      "step": 257900
+    },
+    {
+      "epoch": 83.38720103425986,
+      "grad_norm": 0.7935035228729248,
+      "learning_rate": 0.001,
+      "loss": 0.5084,
+      "step": 258000
+    },
+    {
+      "epoch": 83.41952165481577,
+      "grad_norm": 0.5782608985900879,
+      "learning_rate": 0.001,
+      "loss": 0.5151,
+      "step": 258100
+    },
+    {
+      "epoch": 83.45184227537169,
+      "grad_norm": 0.8032583594322205,
+      "learning_rate": 0.001,
+      "loss": 0.5042,
+      "step": 258200
+    },
+    {
+      "epoch": 83.4841628959276,
+      "grad_norm": 0.7582470774650574,
+      "learning_rate": 0.001,
+      "loss": 0.514,
+      "step": 258300
+    },
+    {
+      "epoch": 83.51648351648352,
+      "grad_norm": 0.8967429995536804,
+      "learning_rate": 0.001,
+      "loss": 0.5185,
+      "step": 258400
+    },
+    {
+      "epoch": 83.54880413703943,
+      "grad_norm": 0.7805576920509338,
+      "learning_rate": 0.001,
+      "loss": 0.5115,
+      "step": 258500
+    },
+    {
+      "epoch": 83.58112475759535,
+      "grad_norm": 1.0321794748306274,
+      "learning_rate": 0.001,
+      "loss": 0.5181,
+      "step": 258600
+    },
+    {
+      "epoch": 83.61344537815125,
+      "grad_norm": 0.6753956079483032,
+      "learning_rate": 0.001,
+      "loss": 0.514,
+      "step": 258700
+    },
+    {
+      "epoch": 83.64576599870718,
+      "grad_norm": 0.8881549835205078,
+      "learning_rate": 0.001,
+      "loss": 0.5151,
+      "step": 258800
+    },
+    {
+      "epoch": 83.67808661926308,
+      "grad_norm": 0.6611385941505432,
+      "learning_rate": 0.001,
+      "loss": 0.5282,
+      "step": 258900
+    },
+    {
+      "epoch": 83.710407239819,
+      "grad_norm": 0.7606088519096375,
+      "learning_rate": 0.001,
+      "loss": 0.5317,
+      "step": 259000
+    },
+    {
+      "epoch": 83.74272786037491,
+      "grad_norm": 0.8321859836578369,
+      "learning_rate": 0.001,
+      "loss": 0.5248,
+      "step": 259100
+    },
+    {
+      "epoch": 83.77504848093083,
+      "grad_norm": 0.7359297275543213,
+      "learning_rate": 0.001,
+      "loss": 0.5207,
+      "step": 259200
+    },
+    {
+      "epoch": 83.80736910148674,
+      "grad_norm": 0.6839292049407959,
+      "learning_rate": 0.001,
+      "loss": 0.5175,
+      "step": 259300
+    },
+    {
+      "epoch": 83.83968972204266,
+      "grad_norm": 0.5971709489822388,
+      "learning_rate": 0.001,
+      "loss": 0.5355,
+      "step": 259400
+    },
+    {
+      "epoch": 83.87201034259857,
+      "grad_norm": 0.5360345244407654,
+      "learning_rate": 0.001,
+      "loss": 0.5233,
+      "step": 259500
+    },
+    {
+      "epoch": 83.9043309631545,
+      "grad_norm": 0.551607608795166,
+      "learning_rate": 0.001,
+      "loss": 0.5298,
+      "step": 259600
+    },
+    {
+      "epoch": 83.9366515837104,
+      "grad_norm": 0.7951961755752563,
+      "learning_rate": 0.001,
+      "loss": 0.5336,
+      "step": 259700
+    },
+    {
+      "epoch": 83.96897220426632,
+      "grad_norm": 0.7808154821395874,
+      "learning_rate": 0.001,
+      "loss": 0.5337,
+      "step": 259800
+    },
+    {
+      "epoch": 84.00129282482224,
+      "grad_norm": 1.4088801145553589,
+      "learning_rate": 0.001,
+      "loss": 0.5438,
+      "step": 259900
+    },
+    {
+      "epoch": 84.03361344537815,
+      "grad_norm": 1.5833579301834106,
+      "learning_rate": 0.001,
+      "loss": 0.4798,
+      "step": 260000
+    },
+    {
+      "epoch": 84.06593406593407,
+      "grad_norm": 1.3921915292739868,
+      "learning_rate": 0.001,
+      "loss": 0.4727,
+      "step": 260100
+    },
+    {
+      "epoch": 84.09825468648998,
+      "grad_norm": 1.3180691003799438,
+      "learning_rate": 0.001,
+      "loss": 0.4633,
+      "step": 260200
+    },
+    {
+      "epoch": 84.1305753070459,
+      "grad_norm": 1.685084581375122,
+      "learning_rate": 0.001,
+      "loss": 0.4799,
+      "step": 260300
+    },
+    {
+      "epoch": 84.16289592760181,
+      "grad_norm": 1.4940513372421265,
+      "learning_rate": 0.001,
+      "loss": 0.4854,
+      "step": 260400
+    },
+    {
+      "epoch": 84.19521654815773,
+      "grad_norm": 1.7479921579360962,
+      "learning_rate": 0.001,
+      "loss": 0.4912,
+      "step": 260500
+    },
+    {
+      "epoch": 84.22753716871364,
+      "grad_norm": 2.0624806880950928,
+      "learning_rate": 0.001,
+      "loss": 0.4887,
+      "step": 260600
+    },
+    {
+      "epoch": 84.25985778926956,
+      "grad_norm": 1.9271317720413208,
+      "learning_rate": 0.001,
+      "loss": 0.4773,
+      "step": 260700
+    },
+    {
+      "epoch": 84.29217840982547,
+      "grad_norm": 4.249568939208984,
+      "learning_rate": 0.001,
+      "loss": 0.4867,
+      "step": 260800
+    },
+    {
+      "epoch": 84.32449903038139,
+      "grad_norm": 1.7943352460861206,
+      "learning_rate": 0.001,
+      "loss": 0.5005,
+      "step": 260900
+    },
+    {
+      "epoch": 84.3568196509373,
+      "grad_norm": 2.092435359954834,
+      "learning_rate": 0.001,
+      "loss": 0.5074,
+      "step": 261000
+    },
+    {
+      "epoch": 84.38914027149322,
+      "grad_norm": 1.368276834487915,
+      "learning_rate": 0.001,
+      "loss": 0.4962,
+      "step": 261100
+    },
+    {
+      "epoch": 84.42146089204913,
+      "grad_norm": 1.6893104314804077,
+      "learning_rate": 0.001,
+      "loss": 0.503,
+      "step": 261200
+    },
+    {
+      "epoch": 84.45378151260505,
+      "grad_norm": 1.8402940034866333,
+      "learning_rate": 0.001,
+      "loss": 0.5013,
+      "step": 261300
+    },
+    {
+      "epoch": 84.48610213316095,
+      "grad_norm": 1.8994182348251343,
+      "learning_rate": 0.001,
+      "loss": 0.5002,
+      "step": 261400
+    },
+    {
+      "epoch": 84.51842275371688,
+      "grad_norm": 1.7885549068450928,
+      "learning_rate": 0.001,
+      "loss": 0.5087,
+      "step": 261500
+    },
+    {
+      "epoch": 84.55074337427278,
+      "grad_norm": 2.087735652923584,
+      "learning_rate": 0.001,
+      "loss": 0.4951,
+      "step": 261600
+    },
+    {
+      "epoch": 84.5830639948287,
+      "grad_norm": 1.4527853727340698,
+      "learning_rate": 0.001,
+      "loss": 0.5014,
+      "step": 261700
+    },
+    {
+      "epoch": 84.61538461538461,
+      "grad_norm": 1.629301905632019,
+      "learning_rate": 0.001,
+      "loss": 0.5102,
+      "step": 261800
+    },
+    {
+      "epoch": 84.64770523594053,
+      "grad_norm": 1.9578490257263184,
+      "learning_rate": 0.001,
+      "loss": 0.5024,
+      "step": 261900
+    },
+    {
+      "epoch": 84.68002585649644,
+      "grad_norm": 2.0275087356567383,
+      "learning_rate": 0.001,
+      "loss": 0.5099,
+      "step": 262000
+    },
+    {
+      "epoch": 84.71234647705236,
+      "grad_norm": 1.6318986415863037,
+      "learning_rate": 0.001,
+      "loss": 0.5113,
+      "step": 262100
+    },
+    {
+      "epoch": 84.74466709760827,
+      "grad_norm": 1.8131710290908813,
+      "learning_rate": 0.001,
+      "loss": 0.5119,
+      "step": 262200
+    },
+    {
+      "epoch": 84.7769877181642,
+      "grad_norm": 1.5382099151611328,
+      "learning_rate": 0.001,
+      "loss": 0.5303,
+      "step": 262300
+    },
+    {
+      "epoch": 84.8093083387201,
+      "grad_norm": 2.2519476413726807,
+      "learning_rate": 0.001,
+      "loss": 0.5204,
+      "step": 262400
+    },
+    {
+      "epoch": 84.84162895927602,
+      "grad_norm": 1.4845337867736816,
+      "learning_rate": 0.001,
+      "loss": 0.5277,
+      "step": 262500
+    },
+    {
+      "epoch": 84.87394957983193,
+      "grad_norm": 1.4768366813659668,
+      "learning_rate": 0.001,
+      "loss": 0.5232,
+      "step": 262600
+    },
+    {
+      "epoch": 84.90627020038785,
+      "grad_norm": 1.7839581966400146,
+      "learning_rate": 0.001,
+      "loss": 0.5319,
+      "step": 262700
+    },
+    {
+      "epoch": 84.93859082094376,
+      "grad_norm": 1.6787962913513184,
+      "learning_rate": 0.001,
+      "loss": 0.5212,
+      "step": 262800
+    },
+    {
+      "epoch": 84.97091144149968,
+      "grad_norm": 1.7544941902160645,
+      "learning_rate": 0.001,
+      "loss": 0.5338,
+      "step": 262900
+    },
+    {
+      "epoch": 85.00323206205559,
+      "grad_norm": 1.494774580001831,
+      "learning_rate": 0.001,
+      "loss": 0.545,
+      "step": 263000
+    },
+    {
+      "epoch": 85.03555268261151,
+      "grad_norm": 1.3960219621658325,
+      "learning_rate": 0.001,
+      "loss": 0.4712,
+      "step": 263100
+    },
+    {
+      "epoch": 85.06787330316742,
+      "grad_norm": 1.553849220275879,
+      "learning_rate": 0.001,
+      "loss": 0.4741,
+      "step": 263200
+    },
+    {
+      "epoch": 85.10019392372334,
+      "grad_norm": 1.8072491884231567,
+      "learning_rate": 0.001,
+      "loss": 0.4699,
+      "step": 263300
+    },
+    {
+      "epoch": 85.13251454427925,
+      "grad_norm": 1.5213582515716553,
+      "learning_rate": 0.001,
+      "loss": 0.4671,
+      "step": 263400
+    },
+    {
+      "epoch": 85.16483516483517,
+      "grad_norm": 1.9114036560058594,
+      "learning_rate": 0.001,
+      "loss": 0.4783,
+      "step": 263500
+    },
+    {
+      "epoch": 85.19715578539108,
+      "grad_norm": 1.4559093713760376,
+      "learning_rate": 0.001,
+      "loss": 0.4804,
+      "step": 263600
+    },
+    {
+      "epoch": 85.229476405947,
+      "grad_norm": 1.2537195682525635,
+      "learning_rate": 0.001,
+      "loss": 0.4838,
+      "step": 263700
+    },
+    {
+      "epoch": 85.2617970265029,
+      "grad_norm": 1.655987024307251,
+      "learning_rate": 0.001,
+      "loss": 0.5079,
+      "step": 263800
+    },
+    {
+      "epoch": 85.29411764705883,
+      "grad_norm": 1.7664257287979126,
+      "learning_rate": 0.001,
+      "loss": 0.4813,
+      "step": 263900
+    },
+    {
+      "epoch": 85.32643826761473,
+      "grad_norm": 1.795916199684143,
+      "learning_rate": 0.001,
+      "loss": 0.4974,
+      "step": 264000
+    },
+    {
+      "epoch": 85.35875888817066,
+      "grad_norm": 1.5820717811584473,
+      "learning_rate": 0.001,
+      "loss": 0.5014,
+      "step": 264100
+    },
+    {
+      "epoch": 85.39107950872656,
+      "grad_norm": 1.4357985258102417,
+      "learning_rate": 0.001,
+      "loss": 0.496,
+      "step": 264200
+    },
+    {
+      "epoch": 85.42340012928248,
+      "grad_norm": 1.449780821800232,
+      "learning_rate": 0.001,
+      "loss": 0.4962,
+      "step": 264300
+    },
+    {
+      "epoch": 85.45572074983839,
+      "grad_norm": 1.4229458570480347,
+      "learning_rate": 0.001,
+      "loss": 0.5044,
+      "step": 264400
+    },
+    {
+      "epoch": 85.48804137039431,
+      "grad_norm": 1.443252682685852,
+      "learning_rate": 0.001,
+      "loss": 0.4841,
+      "step": 264500
+    },
+    {
+      "epoch": 85.52036199095022,
+      "grad_norm": 1.1594971418380737,
+      "learning_rate": 0.001,
+      "loss": 0.5087,
+      "step": 264600
+    },
+    {
+      "epoch": 85.55268261150614,
+      "grad_norm": 1.7895407676696777,
+      "learning_rate": 0.001,
+      "loss": 0.5107,
+      "step": 264700
+    },
+    {
+      "epoch": 85.58500323206205,
+      "grad_norm": 1.5300350189208984,
+      "learning_rate": 0.001,
+      "loss": 0.5063,
+      "step": 264800
+    },
+    {
+      "epoch": 85.61732385261797,
+      "grad_norm": 1.501360297203064,
+      "learning_rate": 0.001,
+      "loss": 0.5023,
+      "step": 264900
+    },
+    {
+      "epoch": 85.64964447317388,
+      "grad_norm": 1.5320264101028442,
+      "learning_rate": 0.001,
+      "loss": 0.4965,
+      "step": 265000
+    },
+    {
+      "epoch": 85.6819650937298,
+      "grad_norm": 1.4581434726715088,
+      "learning_rate": 0.001,
+      "loss": 0.5048,
+      "step": 265100
+    },
+    {
+      "epoch": 85.71428571428571,
+      "grad_norm": 1.3853094577789307,
+      "learning_rate": 0.001,
+      "loss": 0.4939,
+      "step": 265200
+    },
+    {
+      "epoch": 85.74660633484163,
+      "grad_norm": 1.4571151733398438,
+      "learning_rate": 0.001,
+      "loss": 0.5211,
+      "step": 265300
+    },
+    {
+      "epoch": 85.77892695539754,
+      "grad_norm": 1.6544904708862305,
+      "learning_rate": 0.001,
+      "loss": 0.51,
+      "step": 265400
+    },
+    {
+      "epoch": 85.81124757595346,
+      "grad_norm": 1.4322007894515991,
+      "learning_rate": 0.001,
+      "loss": 0.521,
+      "step": 265500
+    },
+    {
+      "epoch": 85.84356819650937,
+      "grad_norm": 1.7621874809265137,
+      "learning_rate": 0.001,
+      "loss": 0.5152,
+      "step": 265600
+    },
+    {
+      "epoch": 85.87588881706529,
+      "grad_norm": 1.492745041847229,
+      "learning_rate": 0.001,
+      "loss": 0.5191,
+      "step": 265700
+    },
+    {
+      "epoch": 85.9082094376212,
+      "grad_norm": 1.6130776405334473,
+      "learning_rate": 0.001,
+      "loss": 0.5192,
+      "step": 265800
+    },
+    {
+      "epoch": 85.94053005817712,
+      "grad_norm": 1.2372713088989258,
+      "learning_rate": 0.001,
+      "loss": 0.524,
+      "step": 265900
+    },
+    {
+      "epoch": 85.97285067873302,
+      "grad_norm": 1.493804693222046,
+      "learning_rate": 0.001,
+      "loss": 0.5237,
+      "step": 266000
+    },
+    {
+      "epoch": 86.00517129928895,
+      "grad_norm": 1.6226726770401,
+      "learning_rate": 0.001,
+      "loss": 0.5214,
+      "step": 266100
+    },
+    {
+      "epoch": 86.03749191984487,
+      "grad_norm": 1.511785864830017,
+      "learning_rate": 0.001,
+      "loss": 0.4658,
+      "step": 266200
+    },
+    {
+      "epoch": 86.06981254040078,
+      "grad_norm": 1.5849074125289917,
+      "learning_rate": 0.001,
+      "loss": 0.4684,
+      "step": 266300
+    },
+    {
+      "epoch": 86.1021331609567,
+      "grad_norm": 1.510827898979187,
+      "learning_rate": 0.001,
+      "loss": 0.4636,
+      "step": 266400
+    },
+    {
+      "epoch": 86.1344537815126,
+      "grad_norm": 1.6693753004074097,
+      "learning_rate": 0.001,
+      "loss": 0.4764,
+      "step": 266500
+    },
+    {
+      "epoch": 86.16677440206853,
+      "grad_norm": 1.2760614156723022,
+      "learning_rate": 0.001,
+      "loss": 0.4754,
+      "step": 266600
+    },
+    {
+      "epoch": 86.19909502262443,
+      "grad_norm": 1.7338852882385254,
+      "learning_rate": 0.001,
+      "loss": 0.4715,
+      "step": 266700
+    },
+    {
+      "epoch": 86.23141564318036,
+      "grad_norm": 1.388390302658081,
+      "learning_rate": 0.001,
+      "loss": 0.4788,
+      "step": 266800
+    },
+    {
+      "epoch": 86.26373626373626,
+      "grad_norm": 1.734383225440979,
+      "learning_rate": 0.001,
+      "loss": 0.4847,
+      "step": 266900
+    },
+    {
+      "epoch": 86.29605688429218,
+      "grad_norm": 1.4146814346313477,
+      "learning_rate": 0.001,
+      "loss": 0.4865,
+      "step": 267000
+    },
+    {
+      "epoch": 86.32837750484809,
+      "grad_norm": 1.3522993326187134,
+      "learning_rate": 0.001,
+      "loss": 0.4895,
+      "step": 267100
+    },
+    {
+      "epoch": 86.36069812540401,
+      "grad_norm": 1.469128131866455,
+      "learning_rate": 0.001,
+      "loss": 0.4917,
+      "step": 267200
+    },
+    {
+      "epoch": 86.39301874595992,
+      "grad_norm": 1.8461087942123413,
+      "learning_rate": 0.001,
+      "loss": 0.4969,
+      "step": 267300
+    },
+    {
+      "epoch": 86.42533936651584,
+      "grad_norm": 1.894525170326233,
+      "learning_rate": 0.001,
+      "loss": 0.4877,
+      "step": 267400
+    },
+    {
+      "epoch": 86.45765998707175,
+      "grad_norm": 2.118194341659546,
+      "learning_rate": 0.001,
+      "loss": 0.48,
+      "step": 267500
+    },
+    {
+      "epoch": 86.48998060762767,
+      "grad_norm": 1.5822690725326538,
+      "learning_rate": 0.001,
+      "loss": 0.4902,
+      "step": 267600
+    },
+    {
+      "epoch": 86.52230122818358,
+      "grad_norm": 1.3895262479782104,
+      "learning_rate": 0.001,
+      "loss": 0.4972,
+      "step": 267700
+    },
+    {
+      "epoch": 86.5546218487395,
+      "grad_norm": 1.4661526679992676,
+      "learning_rate": 0.001,
+      "loss": 0.4966,
+      "step": 267800
+    },
+    {
+      "epoch": 86.58694246929541,
+      "grad_norm": 1.634231448173523,
+      "learning_rate": 0.001,
+      "loss": 0.5081,
+      "step": 267900
+    },
+    {
+      "epoch": 86.61926308985133,
+      "grad_norm": 1.520247459411621,
+      "learning_rate": 0.001,
+      "loss": 0.4971,
+      "step": 268000
+    },
+    {
+      "epoch": 86.65158371040724,
+      "grad_norm": 1.878395438194275,
+      "learning_rate": 0.001,
+      "loss": 0.5079,
+      "step": 268100
+    },
+    {
+      "epoch": 86.68390433096316,
+      "grad_norm": 1.7085822820663452,
+      "learning_rate": 0.001,
+      "loss": 0.5052,
+      "step": 268200
+    },
+    {
+      "epoch": 86.71622495151907,
+      "grad_norm": 1.3339471817016602,
+      "learning_rate": 0.001,
+      "loss": 0.4948,
+      "step": 268300
+    },
+    {
+      "epoch": 86.74854557207499,
+      "grad_norm": 1.9712716341018677,
+      "learning_rate": 0.001,
+      "loss": 0.5098,
+      "step": 268400
+    },
+    {
+      "epoch": 86.7808661926309,
+      "grad_norm": 2.210573434829712,
+      "learning_rate": 0.001,
+      "loss": 0.5105,
+      "step": 268500
+    },
+    {
+      "epoch": 86.81318681318682,
+      "grad_norm": 1.4131146669387817,
+      "learning_rate": 0.001,
+      "loss": 0.5096,
+      "step": 268600
+    },
+    {
+      "epoch": 86.84550743374272,
+      "grad_norm": 1.391532063484192,
+      "learning_rate": 0.001,
+      "loss": 0.5151,
+      "step": 268700
+    },
+    {
+      "epoch": 86.87782805429865,
+      "grad_norm": 1.9204423427581787,
+      "learning_rate": 0.001,
+      "loss": 0.5061,
+      "step": 268800
+    },
+    {
+      "epoch": 86.91014867485455,
+      "grad_norm": 2.0518362522125244,
+      "learning_rate": 0.001,
+      "loss": 0.5058,
+      "step": 268900
+    },
+    {
+      "epoch": 86.94246929541048,
+      "grad_norm": 1.5138390064239502,
+      "learning_rate": 0.001,
+      "loss": 0.5072,
+      "step": 269000
+    },
+    {
+      "epoch": 86.97478991596638,
+      "grad_norm": 1.3717081546783447,
+      "learning_rate": 0.001,
+      "loss": 0.5203,
+      "step": 269100
+    },
+    {
+      "epoch": 87.0071105365223,
+      "grad_norm": 1.4673140048980713,
+      "learning_rate": 0.001,
+      "loss": 0.5119,
+      "step": 269200
+    },
+    {
+      "epoch": 87.03943115707821,
+      "grad_norm": 1.371171474456787,
+      "learning_rate": 0.001,
+      "loss": 0.4482,
+      "step": 269300
+    },
+    {
+      "epoch": 87.07175177763413,
+      "grad_norm": 1.397260308265686,
+      "learning_rate": 0.001,
+      "loss": 0.4578,
+      "step": 269400
+    },
+    {
+      "epoch": 87.10407239819004,
+      "grad_norm": 1.6206611394882202,
+      "learning_rate": 0.001,
+      "loss": 0.4609,
+      "step": 269500
+    },
+    {
+      "epoch": 87.13639301874596,
+      "grad_norm": 1.3245476484298706,
+      "learning_rate": 0.001,
+      "loss": 0.4701,
+      "step": 269600
+    },
+    {
+      "epoch": 87.16871363930187,
+      "grad_norm": 1.6914180517196655,
+      "learning_rate": 0.001,
+      "loss": 0.4584,
+      "step": 269700
+    },
+    {
+      "epoch": 87.20103425985779,
+      "grad_norm": 2.0249228477478027,
+      "learning_rate": 0.001,
+      "loss": 0.4669,
+      "step": 269800
+    },
+    {
+      "epoch": 87.2333548804137,
+      "grad_norm": 1.3308566808700562,
+      "learning_rate": 0.001,
+      "loss": 0.4724,
+      "step": 269900
+    },
+    {
+      "epoch": 87.26567550096962,
+      "grad_norm": 1.2931392192840576,
+      "learning_rate": 0.001,
+      "loss": 0.4759,
+      "step": 270000
+    },
+    {
+      "epoch": 87.29799612152553,
+      "grad_norm": 1.290625810623169,
+      "learning_rate": 0.001,
+      "loss": 0.4803,
+      "step": 270100
+    },
+    {
+      "epoch": 87.33031674208145,
+      "grad_norm": 1.3367546796798706,
+      "learning_rate": 0.001,
+      "loss": 0.4847,
+      "step": 270200
+    },
+    {
+      "epoch": 87.36263736263736,
+      "grad_norm": 1.2759326696395874,
+      "learning_rate": 0.001,
+      "loss": 0.4872,
+      "step": 270300
+    },
+    {
+      "epoch": 87.39495798319328,
+      "grad_norm": 1.3992459774017334,
+      "learning_rate": 0.001,
+      "loss": 0.4876,
+      "step": 270400
+    },
+    {
+      "epoch": 87.42727860374919,
+      "grad_norm": 1.4173572063446045,
+      "learning_rate": 0.001,
+      "loss": 0.4712,
+      "step": 270500
+    },
+    {
+      "epoch": 87.45959922430511,
+      "grad_norm": 1.8669795989990234,
+      "learning_rate": 0.001,
+      "loss": 0.4922,
+      "step": 270600
+    },
+    {
+      "epoch": 87.49191984486102,
+      "grad_norm": 1.6839419603347778,
+      "learning_rate": 0.001,
+      "loss": 0.495,
+      "step": 270700
+    },
+    {
+      "epoch": 87.52424046541694,
+      "grad_norm": 1.2317043542861938,
+      "learning_rate": 0.001,
+      "loss": 0.4771,
+      "step": 270800
+    },
+    {
+      "epoch": 87.55656108597285,
+      "grad_norm": 1.3039339780807495,
+      "learning_rate": 0.001,
+      "loss": 0.4837,
+      "step": 270900
+    },
+    {
+      "epoch": 87.58888170652877,
+      "grad_norm": 1.291479229927063,
+      "learning_rate": 0.001,
+      "loss": 0.4902,
+      "step": 271000
+    },
+    {
+      "epoch": 87.62120232708467,
+      "grad_norm": 1.3181146383285522,
+      "learning_rate": 0.001,
+      "loss": 0.4975,
+      "step": 271100
+    },
+    {
+      "epoch": 87.6535229476406,
+      "grad_norm": 1.7635709047317505,
+      "learning_rate": 0.001,
+      "loss": 0.4954,
+      "step": 271200
+    },
+    {
+      "epoch": 87.6858435681965,
+      "grad_norm": 1.4242125749588013,
+      "learning_rate": 0.001,
+      "loss": 0.509,
+      "step": 271300
+    },
+    {
+      "epoch": 87.71816418875243,
+      "grad_norm": 1.4490323066711426,
+      "learning_rate": 0.001,
+      "loss": 0.5068,
+      "step": 271400
+    },
+    {
+      "epoch": 87.75048480930833,
+      "grad_norm": 1.5760105848312378,
+      "learning_rate": 0.001,
+      "loss": 0.4997,
+      "step": 271500
+    },
+    {
+      "epoch": 87.78280542986425,
+      "grad_norm": 1.3230277299880981,
+      "learning_rate": 0.001,
+      "loss": 0.5041,
+      "step": 271600
+    },
+    {
+      "epoch": 87.81512605042016,
+      "grad_norm": 1.6378422975540161,
+      "learning_rate": 0.001,
+      "loss": 0.5138,
+      "step": 271700
+    },
+    {
+      "epoch": 87.84744667097608,
+      "grad_norm": 1.4305038452148438,
+      "learning_rate": 0.001,
+      "loss": 0.5073,
+      "step": 271800
+    },
+    {
+      "epoch": 87.87976729153199,
+      "grad_norm": 1.6552515029907227,
+      "learning_rate": 0.001,
+      "loss": 0.4978,
+      "step": 271900
+    },
+    {
+      "epoch": 87.91208791208791,
+      "grad_norm": 1.555861473083496,
+      "learning_rate": 0.001,
+      "loss": 0.511,
+      "step": 272000
+    },
+    {
+      "epoch": 87.94440853264382,
+      "grad_norm": 1.9281134605407715,
+      "learning_rate": 0.001,
+      "loss": 0.5197,
+      "step": 272100
+    },
+    {
+      "epoch": 87.97672915319974,
+      "grad_norm": 2.146212100982666,
+      "learning_rate": 0.001,
+      "loss": 0.5133,
+      "step": 272200
+    },
+    {
+      "epoch": 88.00904977375566,
+      "grad_norm": 1.4571073055267334,
+      "learning_rate": 0.001,
+      "loss": 0.4881,
+      "step": 272300
+    },
+    {
+      "epoch": 88.04137039431157,
+      "grad_norm": 1.42752206325531,
+      "learning_rate": 0.001,
+      "loss": 0.4532,
+      "step": 272400
+    },
+    {
+      "epoch": 88.07369101486749,
+      "grad_norm": 1.7515978813171387,
+      "learning_rate": 0.001,
+      "loss": 0.4482,
+      "step": 272500
+    },
+    {
+      "epoch": 88.1060116354234,
+      "grad_norm": 1.3939858675003052,
+      "learning_rate": 0.001,
+      "loss": 0.4564,
+      "step": 272600
+    },
+    {
+      "epoch": 88.13833225597932,
+      "grad_norm": 1.2333893775939941,
+      "learning_rate": 0.001,
+      "loss": 0.4716,
+      "step": 272700
+    },
+    {
+      "epoch": 88.17065287653523,
+      "grad_norm": 1.5755928754806519,
+      "learning_rate": 0.001,
+      "loss": 0.4726,
+      "step": 272800
+    },
+    {
+      "epoch": 88.20297349709115,
+      "grad_norm": 2.824388265609741,
+      "learning_rate": 0.001,
+      "loss": 0.4669,
+      "step": 272900
+    },
+    {
+      "epoch": 88.23529411764706,
+      "grad_norm": 1.1696257591247559,
+      "learning_rate": 0.001,
+      "loss": 0.4707,
+      "step": 273000
+    },
+    {
+      "epoch": 88.26761473820298,
+      "grad_norm": 1.5280286073684692,
+      "learning_rate": 0.001,
+      "loss": 0.4658,
+      "step": 273100
+    },
+    {
+      "epoch": 88.29993535875889,
+      "grad_norm": 1.493591070175171,
+      "learning_rate": 0.001,
+      "loss": 0.4746,
+      "step": 273200
+    },
+    {
+      "epoch": 88.33225597931481,
+      "grad_norm": 1.619083046913147,
+      "learning_rate": 0.001,
+      "loss": 0.4755,
+      "step": 273300
+    },
+    {
+      "epoch": 88.36457659987072,
+      "grad_norm": 3.416677236557007,
+      "learning_rate": 0.001,
+      "loss": 0.4689,
+      "step": 273400
+    },
+    {
+      "epoch": 88.39689722042664,
+      "grad_norm": 1.3980712890625,
+      "learning_rate": 0.001,
+      "loss": 0.4713,
+      "step": 273500
+    },
+    {
+      "epoch": 88.42921784098255,
+      "grad_norm": 1.4884861707687378,
+      "learning_rate": 0.001,
+      "loss": 0.4793,
+      "step": 273600
+    },
+    {
+      "epoch": 88.46153846153847,
+      "grad_norm": 1.4330198764801025,
+      "learning_rate": 0.001,
+      "loss": 0.4845,
+      "step": 273700
+    },
+    {
+      "epoch": 88.49385908209437,
+      "grad_norm": 1.7243083715438843,
+      "learning_rate": 0.001,
+      "loss": 0.4794,
+      "step": 273800
+    },
+    {
+      "epoch": 88.5261797026503,
+      "grad_norm": 1.2517062425613403,
+      "learning_rate": 0.001,
+      "loss": 0.4865,
+      "step": 273900
+    },
+    {
+      "epoch": 88.5585003232062,
+      "grad_norm": 1.5816173553466797,
+      "learning_rate": 0.001,
+      "loss": 0.4873,
+      "step": 274000
+    },
+    {
+      "epoch": 88.59082094376213,
+      "grad_norm": 1.3292566537857056,
+      "learning_rate": 0.001,
+      "loss": 0.4888,
+      "step": 274100
+    },
+    {
+      "epoch": 88.62314156431803,
+      "grad_norm": 1.2126435041427612,
+      "learning_rate": 0.001,
+      "loss": 0.4922,
+      "step": 274200
+    },
+    {
+      "epoch": 88.65546218487395,
+      "grad_norm": 1.7256187200546265,
+      "learning_rate": 0.001,
+      "loss": 0.4931,
+      "step": 274300
+    },
+    {
+      "epoch": 88.68778280542986,
+      "grad_norm": 1.8610934019088745,
+      "learning_rate": 0.001,
+      "loss": 0.5008,
+      "step": 274400
+    },
+    {
+      "epoch": 88.72010342598578,
+      "grad_norm": 1.6567224264144897,
+      "learning_rate": 0.001,
+      "loss": 0.4895,
+      "step": 274500
+    },
+    {
+      "epoch": 88.75242404654169,
+      "grad_norm": 1.3405829668045044,
+      "learning_rate": 0.001,
+      "loss": 0.5039,
+      "step": 274600
+    },
+    {
+      "epoch": 88.78474466709761,
+      "grad_norm": 1.4917628765106201,
+      "learning_rate": 0.001,
+      "loss": 0.4968,
+      "step": 274700
+    },
+    {
+      "epoch": 88.81706528765352,
+      "grad_norm": 1.517630696296692,
+      "learning_rate": 0.001,
+      "loss": 0.4977,
+      "step": 274800
+    },
+    {
+      "epoch": 88.84938590820944,
+      "grad_norm": 1.293230652809143,
+      "learning_rate": 0.001,
+      "loss": 0.4992,
+      "step": 274900
+    },
+    {
+      "epoch": 88.88170652876535,
+      "grad_norm": 1.6211521625518799,
+      "learning_rate": 0.001,
+      "loss": 0.5038,
+      "step": 275000
+    },
+    {
+      "epoch": 88.91402714932127,
+      "grad_norm": 1.3522111177444458,
+      "learning_rate": 0.001,
+      "loss": 0.5073,
+      "step": 275100
+    },
+    {
+      "epoch": 88.94634776987718,
+      "grad_norm": 1.6885173320770264,
+      "learning_rate": 0.001,
+      "loss": 0.5128,
+      "step": 275200
+    },
+    {
+      "epoch": 88.9786683904331,
+      "grad_norm": 1.1733167171478271,
+      "learning_rate": 0.001,
+      "loss": 0.5152,
+      "step": 275300
+    },
+    {
+      "epoch": 89.01098901098901,
+      "grad_norm": 1.6227352619171143,
+      "learning_rate": 0.001,
+      "loss": 0.4826,
+      "step": 275400
+    },
+    {
+      "epoch": 89.04330963154493,
+      "grad_norm": 1.5030772686004639,
+      "learning_rate": 0.001,
+      "loss": 0.4584,
+      "step": 275500
+    },
+    {
+      "epoch": 89.07563025210084,
+      "grad_norm": 1.3889706134796143,
+      "learning_rate": 0.001,
+      "loss": 0.4535,
+      "step": 275600
+    },
+    {
+      "epoch": 89.10795087265676,
+      "grad_norm": 2.0631189346313477,
+      "learning_rate": 0.001,
+      "loss": 0.4577,
+      "step": 275700
+    },
+    {
+      "epoch": 89.14027149321267,
+      "grad_norm": 1.3928669691085815,
+      "learning_rate": 0.001,
+      "loss": 0.4504,
+      "step": 275800
+    },
+    {
+      "epoch": 89.17259211376859,
+      "grad_norm": 1.517021656036377,
+      "learning_rate": 0.001,
+      "loss": 0.4644,
+      "step": 275900
+    },
+    {
+      "epoch": 89.2049127343245,
+      "grad_norm": 1.4841409921646118,
+      "learning_rate": 0.001,
+      "loss": 0.4523,
+      "step": 276000
+    },
+    {
+      "epoch": 89.23723335488042,
+      "grad_norm": 1.5518596172332764,
+      "learning_rate": 0.001,
+      "loss": 0.4628,
+      "step": 276100
+    },
+    {
+      "epoch": 89.26955397543632,
+      "grad_norm": 1.3948451280593872,
+      "learning_rate": 0.001,
+      "loss": 0.4553,
+      "step": 276200
+    },
+    {
+      "epoch": 89.30187459599225,
+      "grad_norm": 1.4886418581008911,
+      "learning_rate": 0.001,
+      "loss": 0.4834,
+      "step": 276300
+    },
+    {
+      "epoch": 89.33419521654815,
+      "grad_norm": 1.31479012966156,
+      "learning_rate": 0.001,
+      "loss": 0.4627,
+      "step": 276400
+    },
+    {
+      "epoch": 89.36651583710407,
+      "grad_norm": 1.0634099245071411,
+      "learning_rate": 0.001,
+      "loss": 0.4594,
+      "step": 276500
+    },
+    {
+      "epoch": 89.39883645765998,
+      "grad_norm": 1.2459696531295776,
+      "learning_rate": 0.001,
+      "loss": 0.479,
+      "step": 276600
+    },
+    {
+      "epoch": 89.4311570782159,
+      "grad_norm": 1.0318691730499268,
+      "learning_rate": 0.001,
+      "loss": 0.4676,
+      "step": 276700
+    },
+    {
+      "epoch": 89.46347769877181,
+      "grad_norm": 1.7067400217056274,
+      "learning_rate": 0.001,
+      "loss": 0.4723,
+      "step": 276800
+    },
+    {
+      "epoch": 89.49579831932773,
+      "grad_norm": 1.4222896099090576,
+      "learning_rate": 0.001,
+      "loss": 0.4702,
+      "step": 276900
+    },
+    {
+      "epoch": 89.52811893988364,
+      "grad_norm": 1.3543707132339478,
+      "learning_rate": 0.001,
+      "loss": 0.4773,
+      "step": 277000
+    },
+    {
+      "epoch": 89.56043956043956,
+      "grad_norm": 1.1513952016830444,
+      "learning_rate": 0.001,
+      "loss": 0.4819,
+      "step": 277100
+    },
+    {
+      "epoch": 89.59276018099547,
+      "grad_norm": 1.3153300285339355,
+      "learning_rate": 0.001,
+      "loss": 0.4861,
+      "step": 277200
+    },
+    {
+      "epoch": 89.62508080155139,
+      "grad_norm": 1.3194316625595093,
+      "learning_rate": 0.001,
+      "loss": 0.4933,
+      "step": 277300
+    },
+    {
+      "epoch": 89.6574014221073,
+      "grad_norm": 1.3721164464950562,
+      "learning_rate": 0.001,
+      "loss": 0.49,
+      "step": 277400
+    },
+    {
+      "epoch": 89.68972204266322,
+      "grad_norm": 1.4286446571350098,
+      "learning_rate": 0.001,
+      "loss": 0.4892,
+      "step": 277500
+    },
+    {
+      "epoch": 89.72204266321913,
+      "grad_norm": 1.2708386182785034,
+      "learning_rate": 0.001,
+      "loss": 0.4953,
+      "step": 277600
+    },
+    {
+      "epoch": 89.75436328377505,
+      "grad_norm": 1.0783087015151978,
+      "learning_rate": 0.001,
+      "loss": 0.4883,
+      "step": 277700
+    },
+    {
+      "epoch": 89.78668390433096,
+      "grad_norm": 1.4168130159378052,
+      "learning_rate": 0.001,
+      "loss": 0.4907,
+      "step": 277800
+    },
+    {
+      "epoch": 89.81900452488688,
+      "grad_norm": 1.315295934677124,
+      "learning_rate": 0.001,
+      "loss": 0.498,
+      "step": 277900
+    },
+    {
+      "epoch": 89.85132514544279,
+      "grad_norm": 1.4645394086837769,
+      "learning_rate": 0.001,
+      "loss": 0.5051,
+      "step": 278000
+    },
+    {
+      "epoch": 89.88364576599871,
+      "grad_norm": 1.6038243770599365,
+      "learning_rate": 0.001,
+      "loss": 0.5035,
+      "step": 278100
+    },
+    {
+      "epoch": 89.91596638655462,
+      "grad_norm": 1.3226242065429688,
+      "learning_rate": 0.001,
+      "loss": 0.5009,
+      "step": 278200
+    },
+    {
+      "epoch": 89.94828700711054,
+      "grad_norm": 1.4041608572006226,
+      "learning_rate": 0.001,
+      "loss": 0.5034,
+      "step": 278300
+    },
+    {
+      "epoch": 89.98060762766644,
+      "grad_norm": 1.4587429761886597,
+      "learning_rate": 0.001,
+      "loss": 0.5081,
+      "step": 278400
+    },
+    {
+      "epoch": 90.01292824822237,
+      "grad_norm": 1.2513530254364014,
+      "learning_rate": 0.001,
+      "loss": 0.4572,
+      "step": 278500
+    },
+    {
+      "epoch": 90.04524886877829,
+      "grad_norm": 1.0929393768310547,
+      "learning_rate": 0.001,
+      "loss": 0.4435,
+      "step": 278600
+    },
+    {
+      "epoch": 90.0775694893342,
+      "grad_norm": 1.1157305240631104,
+      "learning_rate": 0.001,
+      "loss": 0.4525,
+      "step": 278700
+    },
+    {
+      "epoch": 90.10989010989012,
+      "grad_norm": 1.1245415210723877,
+      "learning_rate": 0.001,
+      "loss": 0.452,
+      "step": 278800
+    },
+    {
+      "epoch": 90.14221073044602,
+      "grad_norm": 1.493154764175415,
+      "learning_rate": 0.001,
+      "loss": 0.4605,
+      "step": 278900
+    },
+    {
+      "epoch": 90.17453135100195,
+      "grad_norm": 0.9956198334693909,
+      "learning_rate": 0.001,
+      "loss": 0.4653,
+      "step": 279000
+    },
+    {
+      "epoch": 90.20685197155785,
+      "grad_norm": 1.3177974224090576,
+      "learning_rate": 0.001,
+      "loss": 0.4651,
+      "step": 279100
+    },
+    {
+      "epoch": 90.23917259211377,
+      "grad_norm": 1.2235926389694214,
+      "learning_rate": 0.001,
+      "loss": 0.46,
+      "step": 279200
+    },
+    {
+      "epoch": 90.27149321266968,
+      "grad_norm": 1.260908603668213,
+      "learning_rate": 0.001,
+      "loss": 0.4564,
+      "step": 279300
+    },
+    {
+      "epoch": 90.3038138332256,
+      "grad_norm": 1.13411545753479,
+      "learning_rate": 0.001,
+      "loss": 0.4617,
+      "step": 279400
+    },
+    {
+      "epoch": 90.33613445378151,
+      "grad_norm": 1.1787828207015991,
+      "learning_rate": 0.001,
+      "loss": 0.4699,
+      "step": 279500
+    },
+    {
+      "epoch": 90.36845507433743,
+      "grad_norm": 1.1601965427398682,
+      "learning_rate": 0.001,
+      "loss": 0.4677,
+      "step": 279600
+    },
+    {
+      "epoch": 90.40077569489334,
+      "grad_norm": 1.2311824560165405,
+      "learning_rate": 0.001,
+      "loss": 0.4751,
+      "step": 279700
+    },
+    {
+      "epoch": 90.43309631544926,
+      "grad_norm": 1.132232427597046,
+      "learning_rate": 0.001,
+      "loss": 0.472,
+      "step": 279800
+    },
+    {
+      "epoch": 90.46541693600517,
+      "grad_norm": 0.876586377620697,
+      "learning_rate": 0.001,
+      "loss": 0.4705,
+      "step": 279900
+    },
+    {
+      "epoch": 90.49773755656109,
+      "grad_norm": 1.2030634880065918,
+      "learning_rate": 0.001,
+      "loss": 0.477,
+      "step": 280000
+    },
+    {
+      "epoch": 90.530058177117,
+      "grad_norm": 0.9745743274688721,
+      "learning_rate": 0.001,
+      "loss": 0.4403,
+      "step": 280100
+    },
+    {
+      "epoch": 90.56237879767292,
+      "grad_norm": 1.3243677616119385,
+      "learning_rate": 0.001,
+      "loss": 0.4505,
+      "step": 280200
+    },
+    {
+      "epoch": 90.59469941822883,
+      "grad_norm": 1.1608860492706299,
+      "learning_rate": 0.001,
+      "loss": 0.4497,
+      "step": 280300
+    },
+    {
+      "epoch": 90.62702003878475,
+      "grad_norm": 1.1231688261032104,
+      "learning_rate": 0.001,
+      "loss": 0.4549,
+      "step": 280400
+    },
+    {
+      "epoch": 90.65934065934066,
+      "grad_norm": 1.5799304246902466,
+      "learning_rate": 0.001,
+      "loss": 0.4487,
+      "step": 280500
+    },
+    {
+      "epoch": 90.69166127989658,
+      "grad_norm": 1.247395396232605,
+      "learning_rate": 0.001,
+      "loss": 0.4572,
+      "step": 280600
+    },
+    {
+      "epoch": 90.72398190045249,
+      "grad_norm": 1.3597042560577393,
+      "learning_rate": 0.001,
+      "loss": 0.4627,
+      "step": 280700
+    },
+    {
+      "epoch": 90.75630252100841,
+      "grad_norm": 1.445366621017456,
+      "learning_rate": 0.001,
+      "loss": 0.4587,
+      "step": 280800
+    },
+    {
+      "epoch": 90.78862314156432,
+      "grad_norm": 1.2129086256027222,
+      "learning_rate": 0.001,
+      "loss": 0.4598,
+      "step": 280900
+    },
+    {
+      "epoch": 90.82094376212024,
+      "grad_norm": 1.0306206941604614,
+      "learning_rate": 0.001,
+      "loss": 0.4561,
+      "step": 281000
+    },
+    {
+      "epoch": 90.85326438267614,
+      "grad_norm": 1.1839573383331299,
+      "learning_rate": 0.001,
+      "loss": 0.4791,
+      "step": 281100
+    },
+    {
+      "epoch": 90.88558500323207,
+      "grad_norm": 0.9946609735488892,
+      "learning_rate": 0.001,
+      "loss": 0.4646,
+      "step": 281200
+    },
+    {
+      "epoch": 90.91790562378797,
+      "grad_norm": 1.1721656322479248,
+      "learning_rate": 0.001,
+      "loss": 0.4819,
+      "step": 281300
+    },
+    {
+      "epoch": 90.9502262443439,
+      "grad_norm": 1.1597821712493896,
+      "learning_rate": 0.001,
+      "loss": 0.4703,
+      "step": 281400
+    },
+    {
+      "epoch": 90.9825468648998,
+      "grad_norm": 1.4225759506225586,
+      "learning_rate": 0.001,
+      "loss": 0.4751,
+      "step": 281500
+    },
+    {
+      "epoch": 91.01486748545572,
+      "grad_norm": 9.793893814086914,
+      "learning_rate": 0.001,
+      "loss": 0.434,
+      "step": 281600
+    },
+    {
+      "epoch": 91.04718810601163,
+      "grad_norm": 1.1725051403045654,
+      "learning_rate": 0.001,
+      "loss": 0.4464,
+      "step": 281700
+    },
+    {
+      "epoch": 91.07950872656755,
+      "grad_norm": 0.8536701202392578,
+      "learning_rate": 0.001,
+      "loss": 0.439,
+      "step": 281800
+    },
+    {
+      "epoch": 91.11182934712346,
+      "grad_norm": 0.939294695854187,
+      "learning_rate": 0.001,
+      "loss": 0.4568,
+      "step": 281900
+    },
+    {
+      "epoch": 91.14414996767938,
+      "grad_norm": 1.0028142929077148,
+      "learning_rate": 0.001,
+      "loss": 0.452,
+      "step": 282000
+    },
+    {
+      "epoch": 91.17647058823529,
+      "grad_norm": 0.818120539188385,
+      "learning_rate": 0.001,
+      "loss": 0.4583,
+      "step": 282100
+    },
+    {
+      "epoch": 91.20879120879121,
+      "grad_norm": 0.8494078516960144,
+      "learning_rate": 0.001,
+      "loss": 0.4485,
+      "step": 282200
+    },
+    {
+      "epoch": 91.24111182934712,
+      "grad_norm": 0.7314313054084778,
+      "learning_rate": 0.001,
+      "loss": 0.4577,
+      "step": 282300
+    },
+    {
+      "epoch": 91.27343244990304,
+      "grad_norm": 0.6735559105873108,
+      "learning_rate": 0.001,
+      "loss": 0.465,
+      "step": 282400
+    },
+    {
+      "epoch": 91.30575307045895,
+      "grad_norm": 0.8712020516395569,
+      "learning_rate": 0.001,
+      "loss": 0.4652,
+      "step": 282500
+    },
+    {
+      "epoch": 91.33807369101487,
+      "grad_norm": 1.028043270111084,
+      "learning_rate": 0.001,
+      "loss": 0.4593,
+      "step": 282600
+    },
+    {
+      "epoch": 91.37039431157078,
+      "grad_norm": 0.8021206855773926,
+      "learning_rate": 0.001,
+      "loss": 0.4723,
+      "step": 282700
+    },
+    {
+      "epoch": 91.4027149321267,
+      "grad_norm": 0.8332772850990295,
+      "learning_rate": 0.001,
+      "loss": 0.47,
+      "step": 282800
+    },
+    {
+      "epoch": 91.4350355526826,
+      "grad_norm": 0.8631690740585327,
+      "learning_rate": 0.001,
+      "loss": 0.4683,
+      "step": 282900
+    },
+    {
+      "epoch": 91.46735617323853,
+      "grad_norm": 0.806066632270813,
+      "learning_rate": 0.001,
+      "loss": 0.47,
+      "step": 283000
+    },
+    {
+      "epoch": 91.49967679379444,
+      "grad_norm": 0.9106870889663696,
+      "learning_rate": 0.001,
+      "loss": 0.4687,
+      "step": 283100
+    },
+    {
+      "epoch": 91.53199741435036,
+      "grad_norm": 0.8315029740333557,
+      "learning_rate": 0.001,
+      "loss": 0.4618,
+      "step": 283200
+    },
+    {
+      "epoch": 91.56431803490626,
+      "grad_norm": 0.8321149945259094,
+      "learning_rate": 0.001,
+      "loss": 0.4793,
+      "step": 283300
+    },
+    {
+      "epoch": 91.59663865546219,
+      "grad_norm": 0.6962825059890747,
+      "learning_rate": 0.001,
+      "loss": 0.4825,
+      "step": 283400
+    },
+    {
+      "epoch": 91.6289592760181,
+      "grad_norm": 1.0555658340454102,
+      "learning_rate": 0.001,
+      "loss": 0.4724,
+      "step": 283500
+    },
+    {
+      "epoch": 91.66127989657402,
+      "grad_norm": 0.9735150337219238,
+      "learning_rate": 0.001,
+      "loss": 0.51,
+      "step": 283600
+    },
+    {
+      "epoch": 91.69360051712992,
+      "grad_norm": 1.2564635276794434,
+      "learning_rate": 0.001,
+      "loss": 0.4918,
+      "step": 283700
+    },
+    {
+      "epoch": 91.72592113768584,
+      "grad_norm": 0.7897902131080627,
+      "learning_rate": 0.001,
+      "loss": 0.486,
+      "step": 283800
+    },
+    {
+      "epoch": 91.75824175824175,
+      "grad_norm": 1.0546153783798218,
+      "learning_rate": 0.001,
+      "loss": 0.4935,
+      "step": 283900
+    },
+    {
+      "epoch": 91.79056237879767,
+      "grad_norm": 0.8487304449081421,
+      "learning_rate": 0.001,
+      "loss": 0.4855,
+      "step": 284000
+    },
+    {
+      "epoch": 91.82288299935358,
+      "grad_norm": 0.8980190753936768,
+      "learning_rate": 0.001,
+      "loss": 0.4905,
+      "step": 284100
+    },
+    {
+      "epoch": 91.8552036199095,
+      "grad_norm": 1.1748449802398682,
+      "learning_rate": 0.001,
+      "loss": 0.4818,
+      "step": 284200
+    },
+    {
+      "epoch": 91.88752424046541,
+      "grad_norm": 0.9486263394355774,
+      "learning_rate": 0.001,
+      "loss": 0.4965,
+      "step": 284300
+    },
+    {
+      "epoch": 91.91984486102133,
+      "grad_norm": 1.0960006713867188,
+      "learning_rate": 0.001,
+      "loss": 0.502,
+      "step": 284400
+    },
+    {
+      "epoch": 91.95216548157724,
+      "grad_norm": 1.006560206413269,
+      "learning_rate": 0.001,
+      "loss": 0.4868,
+      "step": 284500
+    },
+    {
+      "epoch": 91.98448610213316,
+      "grad_norm": 1.080833911895752,
+      "learning_rate": 0.001,
+      "loss": 0.502,
+      "step": 284600
+    },
+    {
+      "epoch": 92.01680672268908,
+      "grad_norm": 2.0345189571380615,
+      "learning_rate": 0.001,
+      "loss": 0.4498,
+      "step": 284700
+    },
+    {
+      "epoch": 92.04912734324499,
+      "grad_norm": 1.6295733451843262,
+      "learning_rate": 0.001,
+      "loss": 0.4319,
+      "step": 284800
+    },
+    {
+      "epoch": 92.08144796380091,
+      "grad_norm": 1.82840096950531,
+      "learning_rate": 0.001,
+      "loss": 0.4405,
+      "step": 284900
+    },
+    {
+      "epoch": 92.11376858435682,
+      "grad_norm": 3.0892772674560547,
+      "learning_rate": 0.001,
+      "loss": 0.4414,
+      "step": 285000
+    },
+    {
+      "epoch": 92.14608920491274,
+      "grad_norm": 2.1895627975463867,
+      "learning_rate": 0.001,
+      "loss": 0.4482,
+      "step": 285100
+    },
+    {
+      "epoch": 92.17840982546865,
+      "grad_norm": 1.6839579343795776,
+      "learning_rate": 0.001,
+      "loss": 0.4534,
+      "step": 285200
+    },
+    {
+      "epoch": 92.21073044602457,
+      "grad_norm": 2.260793924331665,
+      "learning_rate": 0.001,
+      "loss": 0.453,
+      "step": 285300
+    },
+    {
+      "epoch": 92.24305106658048,
+      "grad_norm": 42.66197967529297,
+      "learning_rate": 0.001,
+      "loss": 0.4458,
+      "step": 285400
+    },
+    {
+      "epoch": 92.2753716871364,
+      "grad_norm": 1.6944698095321655,
+      "learning_rate": 0.001,
+      "loss": 0.4503,
+      "step": 285500
+    },
+    {
+      "epoch": 92.3076923076923,
+      "grad_norm": 1.6756781339645386,
+      "learning_rate": 0.001,
+      "loss": 0.4563,
+      "step": 285600
+    },
+    {
+      "epoch": 92.34001292824823,
+      "grad_norm": 1.8374137878417969,
+      "learning_rate": 0.001,
+      "loss": 0.4468,
+      "step": 285700
+    },
+    {
+      "epoch": 92.37233354880414,
+      "grad_norm": 2.041393518447876,
+      "learning_rate": 0.001,
+      "loss": 0.4505,
+      "step": 285800
+    },
+    {
+      "epoch": 92.40465416936006,
+      "grad_norm": 1.8268532752990723,
+      "learning_rate": 0.001,
+      "loss": 0.4608,
+      "step": 285900
+    },
+    {
+      "epoch": 92.43697478991596,
+      "grad_norm": 1.7450902462005615,
+      "learning_rate": 0.001,
+      "loss": 0.4634,
+      "step": 286000
+    },
+    {
+      "epoch": 92.46929541047189,
+      "grad_norm": 1.7353445291519165,
+      "learning_rate": 0.001,
+      "loss": 0.4794,
+      "step": 286100
+    },
+    {
+      "epoch": 92.5016160310278,
+      "grad_norm": 2.340031385421753,
+      "learning_rate": 0.001,
+      "loss": 0.471,
+      "step": 286200
+    },
+    {
+      "epoch": 92.53393665158372,
+      "grad_norm": 1.8648983240127563,
+      "learning_rate": 0.001,
+      "loss": 0.4723,
+      "step": 286300
+    },
+    {
+      "epoch": 92.56625727213962,
+      "grad_norm": 2.6980106830596924,
+      "learning_rate": 0.001,
+      "loss": 0.4644,
+      "step": 286400
+    },
+    {
+      "epoch": 92.59857789269554,
+      "grad_norm": 1.8055412769317627,
+      "learning_rate": 0.001,
+      "loss": 0.468,
+      "step": 286500
+    },
+    {
+      "epoch": 92.63089851325145,
+      "grad_norm": 2.3196511268615723,
+      "learning_rate": 0.001,
+      "loss": 0.4766,
+      "step": 286600
+    },
+    {
+      "epoch": 92.66321913380737,
+      "grad_norm": 1.4777075052261353,
+      "learning_rate": 0.001,
+      "loss": 0.4736,
+      "step": 286700
+    },
+    {
+      "epoch": 92.69553975436328,
+      "grad_norm": 2.1524486541748047,
+      "learning_rate": 0.001,
+      "loss": 0.4834,
+      "step": 286800
+    },
+    {
+      "epoch": 92.7278603749192,
+      "grad_norm": 2.242614507675171,
+      "learning_rate": 0.001,
+      "loss": 0.4772,
+      "step": 286900
+    },
+    {
+      "epoch": 92.76018099547511,
+      "grad_norm": 1.7869904041290283,
+      "learning_rate": 0.001,
+      "loss": 0.4739,
+      "step": 287000
+    },
+    {
+      "epoch": 92.79250161603103,
+      "grad_norm": 2.129345417022705,
+      "learning_rate": 0.001,
+      "loss": 0.4887,
+      "step": 287100
+    },
+    {
+      "epoch": 92.82482223658694,
+      "grad_norm": 1.5137324333190918,
+      "learning_rate": 0.001,
+      "loss": 0.4807,
+      "step": 287200
+    },
+    {
+      "epoch": 92.85714285714286,
+      "grad_norm": 2.0685720443725586,
+      "learning_rate": 0.001,
+      "loss": 0.487,
+      "step": 287300
+    },
+    {
+      "epoch": 92.88946347769877,
+      "grad_norm": 1.8368549346923828,
+      "learning_rate": 0.001,
+      "loss": 0.5025,
+      "step": 287400
+    },
+    {
+      "epoch": 92.92178409825469,
+      "grad_norm": 1.7226216793060303,
+      "learning_rate": 0.001,
+      "loss": 0.4917,
+      "step": 287500
+    },
+    {
+      "epoch": 92.9541047188106,
+      "grad_norm": 2.0484440326690674,
+      "learning_rate": 0.001,
+      "loss": 0.4885,
+      "step": 287600
+    },
+    {
+      "epoch": 92.98642533936652,
+      "grad_norm": 1.8843029737472534,
+      "learning_rate": 0.001,
+      "loss": 0.4872,
+      "step": 287700
+    },
+    {
+      "epoch": 93.01874595992243,
+      "grad_norm": 2.280056953430176,
+      "learning_rate": 0.001,
+      "loss": 0.4684,
+      "step": 287800
+    },
+    {
+      "epoch": 93.05106658047835,
+      "grad_norm": 1.6946696043014526,
+      "learning_rate": 0.001,
+      "loss": 0.4274,
+      "step": 287900
+    },
+    {
+      "epoch": 93.08338720103426,
+      "grad_norm": 2.300701856613159,
+      "learning_rate": 0.001,
+      "loss": 0.4325,
+      "step": 288000
+    },
+    {
+      "epoch": 93.11570782159018,
+      "grad_norm": 1.729552984237671,
+      "learning_rate": 0.001,
+      "loss": 0.4413,
+      "step": 288100
+    },
+    {
+      "epoch": 93.14802844214609,
+      "grad_norm": 2.0106143951416016,
+      "learning_rate": 0.001,
+      "loss": 0.4507,
+      "step": 288200
+    },
+    {
+      "epoch": 93.180349062702,
+      "grad_norm": 1.5920679569244385,
+      "learning_rate": 0.001,
+      "loss": 0.4397,
+      "step": 288300
+    },
+    {
+      "epoch": 93.21266968325791,
+      "grad_norm": 1.4083905220031738,
+      "learning_rate": 0.001,
+      "loss": 0.4442,
+      "step": 288400
+    },
+    {
+      "epoch": 93.24499030381384,
+      "grad_norm": 1.5656646490097046,
+      "learning_rate": 0.001,
+      "loss": 0.444,
+      "step": 288500
+    },
+    {
+      "epoch": 93.27731092436974,
+      "grad_norm": 1.4155176877975464,
+      "learning_rate": 0.001,
+      "loss": 0.4594,
+      "step": 288600
+    },
+    {
+      "epoch": 93.30963154492567,
+      "grad_norm": 1.781264305114746,
+      "learning_rate": 0.001,
+      "loss": 0.4541,
+      "step": 288700
+    },
+    {
+      "epoch": 93.34195216548157,
+      "grad_norm": 1.7372766733169556,
+      "learning_rate": 0.001,
+      "loss": 0.4577,
+      "step": 288800
+    },
+    {
+      "epoch": 93.3742727860375,
+      "grad_norm": 1.7941893339157104,
+      "learning_rate": 0.001,
+      "loss": 0.4646,
+      "step": 288900
+    },
+    {
+      "epoch": 93.4065934065934,
+      "grad_norm": 1.5748519897460938,
+      "learning_rate": 0.001,
+      "loss": 0.4587,
+      "step": 289000
+    },
+    {
+      "epoch": 93.43891402714932,
+      "grad_norm": 2.3406195640563965,
+      "learning_rate": 0.001,
+      "loss": 0.4608,
+      "step": 289100
+    },
+    {
+      "epoch": 93.47123464770523,
+      "grad_norm": 1.817068099975586,
+      "learning_rate": 0.001,
+      "loss": 0.4598,
+      "step": 289200
+    },
+    {
+      "epoch": 93.50355526826115,
+      "grad_norm": 1.4823451042175293,
+      "learning_rate": 0.001,
+      "loss": 0.4561,
+      "step": 289300
+    },
+    {
+      "epoch": 93.53587588881706,
+      "grad_norm": 1.5961811542510986,
+      "learning_rate": 0.001,
+      "loss": 0.4662,
+      "step": 289400
+    },
+    {
+      "epoch": 93.56819650937298,
+      "grad_norm": 1.4638841152191162,
+      "learning_rate": 0.001,
+      "loss": 0.4713,
+      "step": 289500
+    },
+    {
+      "epoch": 93.60051712992889,
+      "grad_norm": 2.1063473224639893,
+      "learning_rate": 0.001,
+      "loss": 0.4644,
+      "step": 289600
+    },
+    {
+      "epoch": 93.63283775048481,
+      "grad_norm": 1.989016056060791,
+      "learning_rate": 0.001,
+      "loss": 0.4611,
+      "step": 289700
+    },
+    {
+      "epoch": 93.66515837104072,
+      "grad_norm": 1.9754928350448608,
+      "learning_rate": 0.001,
+      "loss": 0.4624,
+      "step": 289800
+    },
+    {
+      "epoch": 93.69747899159664,
+      "grad_norm": 1.4447102546691895,
+      "learning_rate": 0.001,
+      "loss": 0.4636,
+      "step": 289900
+    },
+    {
+      "epoch": 93.72979961215255,
+      "grad_norm": 1.5371551513671875,
+      "learning_rate": 0.001,
+      "loss": 0.4768,
+      "step": 290000
+    },
+    {
+      "epoch": 93.76212023270847,
+      "grad_norm": 1.5299981832504272,
+      "learning_rate": 0.001,
+      "loss": 0.4783,
+      "step": 290100
+    },
+    {
+      "epoch": 93.79444085326438,
+      "grad_norm": 1.286318063735962,
+      "learning_rate": 0.001,
+      "loss": 0.4764,
+      "step": 290200
+    },
+    {
+      "epoch": 93.8267614738203,
+      "grad_norm": 2.0626156330108643,
+      "learning_rate": 0.001,
+      "loss": 0.4776,
+      "step": 290300
+    },
+    {
+      "epoch": 93.8590820943762,
+      "grad_norm": 1.6997871398925781,
+      "learning_rate": 0.001,
+      "loss": 0.4695,
+      "step": 290400
+    },
+    {
+      "epoch": 93.89140271493213,
+      "grad_norm": 3.9364075660705566,
+      "learning_rate": 0.001,
+      "loss": 0.4801,
+      "step": 290500
+    },
+    {
+      "epoch": 93.92372333548803,
+      "grad_norm": 2.015624523162842,
+      "learning_rate": 0.001,
+      "loss": 0.4962,
+      "step": 290600
+    },
+    {
+      "epoch": 93.95604395604396,
+      "grad_norm": 344.4596862792969,
+      "learning_rate": 0.001,
+      "loss": 0.4812,
+      "step": 290700
+    },
+    {
+      "epoch": 93.98836457659988,
+      "grad_norm": 1.8616305589675903,
+      "learning_rate": 0.001,
+      "loss": 0.4814,
+      "step": 290800
+    },
+    {
+      "epoch": 94.02068519715579,
+      "grad_norm": 1.8940582275390625,
+      "learning_rate": 0.001,
+      "loss": 0.464,
+      "step": 290900
+    },
+    {
+      "epoch": 94.0530058177117,
+      "grad_norm": 1.938902735710144,
+      "learning_rate": 0.001,
+      "loss": 0.4213,
+      "step": 291000
+    },
+    {
+      "epoch": 94.08532643826761,
+      "grad_norm": 1.4678521156311035,
+      "learning_rate": 0.001,
+      "loss": 0.4337,
+      "step": 291100
+    },
+    {
+      "epoch": 94.11764705882354,
+      "grad_norm": 1.5415394306182861,
+      "learning_rate": 0.001,
+      "loss": 0.4445,
+      "step": 291200
+    },
+    {
+      "epoch": 94.14996767937944,
+      "grad_norm": 1.8836512565612793,
+      "learning_rate": 0.001,
+      "loss": 0.4274,
+      "step": 291300
+    },
+    {
+      "epoch": 94.18228829993537,
+      "grad_norm": 1.2818994522094727,
+      "learning_rate": 0.001,
+      "loss": 0.4448,
+      "step": 291400
+    },
+    {
+      "epoch": 94.21460892049127,
+      "grad_norm": 2.0979812145233154,
+      "learning_rate": 0.001,
+      "loss": 0.4396,
+      "step": 291500
+    },
+    {
+      "epoch": 94.2469295410472,
+      "grad_norm": 1.6995420455932617,
+      "learning_rate": 0.001,
+      "loss": 0.4488,
+      "step": 291600
+    },
+    {
+      "epoch": 94.2792501616031,
+      "grad_norm": 1.7920165061950684,
+      "learning_rate": 0.001,
+      "loss": 0.448,
+      "step": 291700
+    },
+    {
+      "epoch": 94.31157078215902,
+      "grad_norm": 1.6216713190078735,
+      "learning_rate": 0.001,
+      "loss": 0.452,
+      "step": 291800
+    },
+    {
+      "epoch": 94.34389140271493,
+      "grad_norm": 1.4908044338226318,
+      "learning_rate": 0.001,
+      "loss": 0.4468,
+      "step": 291900
+    },
+    {
+      "epoch": 94.37621202327085,
+      "grad_norm": 1.6962876319885254,
+      "learning_rate": 0.001,
+      "loss": 0.45,
+      "step": 292000
+    },
+    {
+      "epoch": 94.40853264382676,
+      "grad_norm": 1.460558533668518,
+      "learning_rate": 0.001,
+      "loss": 0.446,
+      "step": 292100
+    },
+    {
+      "epoch": 94.44085326438268,
+      "grad_norm": 1.4219996929168701,
+      "learning_rate": 0.001,
+      "loss": 0.4533,
+      "step": 292200
+    },
+    {
+      "epoch": 94.47317388493859,
+      "grad_norm": 1.758070468902588,
+      "learning_rate": 0.001,
+      "loss": 0.4511,
+      "step": 292300
+    },
+    {
+      "epoch": 94.50549450549451,
+      "grad_norm": 1.518767237663269,
+      "learning_rate": 0.001,
+      "loss": 0.4601,
+      "step": 292400
+    },
+    {
+      "epoch": 94.53781512605042,
+      "grad_norm": 1.4330819845199585,
+      "learning_rate": 0.001,
+      "loss": 0.464,
+      "step": 292500
+    },
+    {
+      "epoch": 94.57013574660634,
+      "grad_norm": 1.7060296535491943,
+      "learning_rate": 0.001,
+      "loss": 0.4593,
+      "step": 292600
+    },
+    {
+      "epoch": 94.60245636716225,
+      "grad_norm": 1.3387072086334229,
+      "learning_rate": 0.001,
+      "loss": 0.4582,
+      "step": 292700
+    },
+    {
+      "epoch": 94.63477698771817,
+      "grad_norm": 1.6963450908660889,
+      "learning_rate": 0.001,
+      "loss": 0.4619,
+      "step": 292800
+    },
+    {
+      "epoch": 94.66709760827408,
+      "grad_norm": 2.0844035148620605,
+      "learning_rate": 0.001,
+      "loss": 0.4665,
+      "step": 292900
+    },
+    {
+      "epoch": 94.69941822883,
+      "grad_norm": 1.3902792930603027,
+      "learning_rate": 0.001,
+      "loss": 0.4783,
+      "step": 293000
+    },
+    {
+      "epoch": 94.7317388493859,
+      "grad_norm": 1.3973846435546875,
+      "learning_rate": 0.001,
+      "loss": 0.4714,
+      "step": 293100
+    },
+    {
+      "epoch": 94.76405946994183,
+      "grad_norm": 1.479785442352295,
+      "learning_rate": 0.001,
+      "loss": 0.471,
+      "step": 293200
+    },
+    {
+      "epoch": 94.79638009049773,
+      "grad_norm": 1.6167840957641602,
+      "learning_rate": 0.001,
+      "loss": 0.4683,
+      "step": 293300
+    },
+    {
+      "epoch": 94.82870071105366,
+      "grad_norm": 1.5192022323608398,
+      "learning_rate": 0.001,
+      "loss": 0.484,
+      "step": 293400
+    },
+    {
+      "epoch": 94.86102133160956,
+      "grad_norm": 1.438111424446106,
+      "learning_rate": 0.001,
+      "loss": 0.4732,
+      "step": 293500
+    },
+    {
+      "epoch": 94.89334195216549,
+      "grad_norm": 1.3876760005950928,
+      "learning_rate": 0.001,
+      "loss": 0.4824,
+      "step": 293600
+    },
+    {
+      "epoch": 94.9256625727214,
+      "grad_norm": 1.9509048461914062,
+      "learning_rate": 0.001,
+      "loss": 0.4725,
+      "step": 293700
+    },
+    {
+      "epoch": 94.95798319327731,
+      "grad_norm": 1.4013237953186035,
+      "learning_rate": 0.001,
+      "loss": 0.4802,
+      "step": 293800
+    },
+    {
+      "epoch": 94.99030381383322,
+      "grad_norm": 1.737876534461975,
+      "learning_rate": 0.001,
+      "loss": 0.4793,
+      "step": 293900
+    },
+    {
+      "epoch": 95.02262443438914,
+      "grad_norm": 1.3785439729690552,
+      "learning_rate": 0.001,
+      "loss": 0.4504,
+      "step": 294000
+    },
+    {
+      "epoch": 95.05494505494505,
+      "grad_norm": 1.9236104488372803,
+      "learning_rate": 0.001,
+      "loss": 0.4349,
+      "step": 294100
+    },
+    {
+      "epoch": 95.08726567550097,
+      "grad_norm": 1.619763970375061,
+      "learning_rate": 0.001,
+      "loss": 0.4403,
+      "step": 294200
+    },
+    {
+      "epoch": 95.11958629605688,
+      "grad_norm": 1.5868083238601685,
+      "learning_rate": 0.001,
+      "loss": 0.4352,
+      "step": 294300
+    },
+    {
+      "epoch": 95.1519069166128,
+      "grad_norm": 1.8232312202453613,
+      "learning_rate": 0.001,
+      "loss": 0.4315,
+      "step": 294400
+    },
+    {
+      "epoch": 95.18422753716871,
+      "grad_norm": 1.5176035165786743,
+      "learning_rate": 0.001,
+      "loss": 0.434,
+      "step": 294500
+    },
+    {
+      "epoch": 95.21654815772463,
+      "grad_norm": 1.5614092350006104,
+      "learning_rate": 0.001,
+      "loss": 0.4418,
+      "step": 294600
+    },
+    {
+      "epoch": 95.24886877828054,
+      "grad_norm": 1.5651224851608276,
+      "learning_rate": 0.001,
+      "loss": 0.434,
+      "step": 294700
+    },
+    {
+      "epoch": 95.28118939883646,
+      "grad_norm": 1.3997880220413208,
+      "learning_rate": 0.001,
+      "loss": 0.4466,
+      "step": 294800
+    },
+    {
+      "epoch": 95.31351001939237,
+      "grad_norm": 1.8400006294250488,
+      "learning_rate": 0.001,
+      "loss": 0.4402,
+      "step": 294900
+    },
+    {
+      "epoch": 95.34583063994829,
+      "grad_norm": 1.32888662815094,
+      "learning_rate": 0.001,
+      "loss": 0.4404,
+      "step": 295000
+    },
+    {
+      "epoch": 95.3781512605042,
+      "grad_norm": 1.6747912168502808,
+      "learning_rate": 0.001,
+      "loss": 0.4522,
+      "step": 295100
+    },
+    {
+      "epoch": 95.41047188106012,
+      "grad_norm": 1.6857593059539795,
+      "learning_rate": 0.001,
+      "loss": 0.444,
+      "step": 295200
+    },
+    {
+      "epoch": 95.44279250161603,
+      "grad_norm": 1.3832392692565918,
+      "learning_rate": 0.001,
+      "loss": 0.4434,
+      "step": 295300
+    },
+    {
+      "epoch": 95.47511312217195,
+      "grad_norm": 1.914587140083313,
+      "learning_rate": 0.001,
+      "loss": 0.454,
+      "step": 295400
+    },
+    {
+      "epoch": 95.50743374272786,
+      "grad_norm": 1.6839224100112915,
+      "learning_rate": 0.001,
+      "loss": 0.4537,
+      "step": 295500
+    },
+    {
+      "epoch": 95.53975436328378,
+      "grad_norm": 1.4552383422851562,
+      "learning_rate": 0.001,
+      "loss": 0.4591,
+      "step": 295600
+    },
+    {
+      "epoch": 95.57207498383968,
+      "grad_norm": 1.5288134813308716,
+      "learning_rate": 0.001,
+      "loss": 0.4533,
+      "step": 295700
+    },
+    {
+      "epoch": 95.6043956043956,
+      "grad_norm": 1.8305370807647705,
+      "learning_rate": 0.001,
+      "loss": 0.4655,
+      "step": 295800
+    },
+    {
+      "epoch": 95.63671622495151,
+      "grad_norm": 1.2696152925491333,
+      "learning_rate": 0.001,
+      "loss": 0.4595,
+      "step": 295900
+    },
+    {
+      "epoch": 95.66903684550743,
+      "grad_norm": 1.5350581407546997,
+      "learning_rate": 0.001,
+      "loss": 0.4685,
+      "step": 296000
+    },
+    {
+      "epoch": 95.70135746606334,
+      "grad_norm": 1.1999725103378296,
+      "learning_rate": 0.001,
+      "loss": 0.4595,
+      "step": 296100
+    },
+    {
+      "epoch": 95.73367808661926,
+      "grad_norm": 1.2505929470062256,
+      "learning_rate": 0.001,
+      "loss": 0.4678,
+      "step": 296200
+    },
+    {
+      "epoch": 95.76599870717517,
+      "grad_norm": 1.3994829654693604,
+      "learning_rate": 0.001,
+      "loss": 0.4646,
+      "step": 296300
+    },
+    {
+      "epoch": 95.7983193277311,
+      "grad_norm": 1.6253557205200195,
+      "learning_rate": 0.001,
+      "loss": 0.4723,
+      "step": 296400
+    },
+    {
+      "epoch": 95.830639948287,
+      "grad_norm": 1.6554275751113892,
+      "learning_rate": 0.001,
+      "loss": 0.4714,
+      "step": 296500
+    },
+    {
+      "epoch": 95.86296056884292,
+      "grad_norm": 1.4614795446395874,
+      "learning_rate": 0.001,
+      "loss": 0.4793,
+      "step": 296600
+    },
+    {
+      "epoch": 95.89528118939883,
+      "grad_norm": 1.41798996925354,
+      "learning_rate": 0.001,
+      "loss": 0.464,
+      "step": 296700
+    },
+    {
+      "epoch": 95.92760180995475,
+      "grad_norm": 1.7458993196487427,
+      "learning_rate": 0.001,
+      "loss": 0.4661,
+      "step": 296800
+    },
+    {
+      "epoch": 95.95992243051066,
+      "grad_norm": 2.5169103145599365,
+      "learning_rate": 0.001,
+      "loss": 0.4737,
+      "step": 296900
+    },
+    {
+      "epoch": 95.99224305106658,
+      "grad_norm": 1.4173320531845093,
+      "learning_rate": 0.001,
+      "loss": 0.4853,
+      "step": 297000
+    },
+    {
+      "epoch": 96.0245636716225,
+      "grad_norm": 1.596639633178711,
+      "learning_rate": 0.001,
+      "loss": 0.4336,
+      "step": 297100
+    },
+    {
+      "epoch": 96.05688429217841,
+      "grad_norm": 1.4442991018295288,
+      "learning_rate": 0.001,
+      "loss": 0.4185,
+      "step": 297200
+    },
+    {
+      "epoch": 96.08920491273433,
+      "grad_norm": 1.3611177206039429,
+      "learning_rate": 0.001,
+      "loss": 0.4274,
+      "step": 297300
+    },
+    {
+      "epoch": 96.12152553329024,
+      "grad_norm": 1.1800782680511475,
+      "learning_rate": 0.001,
+      "loss": 0.4365,
+      "step": 297400
+    },
+    {
+      "epoch": 96.15384615384616,
+      "grad_norm": 1.8172197341918945,
+      "learning_rate": 0.001,
+      "loss": 0.4372,
+      "step": 297500
+    },
+    {
+      "epoch": 96.18616677440207,
+      "grad_norm": 1.4317479133605957,
+      "learning_rate": 0.001,
+      "loss": 0.4323,
+      "step": 297600
+    },
+    {
+      "epoch": 96.21848739495799,
+      "grad_norm": 1.2062143087387085,
+      "learning_rate": 0.001,
+      "loss": 0.4232,
+      "step": 297700
+    },
+    {
+      "epoch": 96.2508080155139,
+      "grad_norm": 1.6967133283615112,
+      "learning_rate": 0.001,
+      "loss": 0.4363,
+      "step": 297800
+    },
+    {
+      "epoch": 96.28312863606982,
+      "grad_norm": 1.3945610523223877,
+      "learning_rate": 0.001,
+      "loss": 0.4354,
+      "step": 297900
+    },
+    {
+      "epoch": 96.31544925662573,
+      "grad_norm": 1.4094641208648682,
+      "learning_rate": 0.001,
+      "loss": 0.4441,
+      "step": 298000
+    },
+    {
+      "epoch": 96.34776987718165,
+      "grad_norm": 1.2168558835983276,
+      "learning_rate": 0.001,
+      "loss": 0.4382,
+      "step": 298100
+    },
+    {
+      "epoch": 96.38009049773756,
+      "grad_norm": 1.9572267532348633,
+      "learning_rate": 0.001,
+      "loss": 0.4441,
+      "step": 298200
+    },
+    {
+      "epoch": 96.41241111829348,
+      "grad_norm": 1.3450571298599243,
+      "learning_rate": 0.001,
+      "loss": 0.439,
+      "step": 298300
+    },
+    {
+      "epoch": 96.44473173884938,
+      "grad_norm": 1.5765080451965332,
+      "learning_rate": 0.001,
+      "loss": 0.4468,
+      "step": 298400
+    },
+    {
+      "epoch": 96.4770523594053,
+      "grad_norm": 1.4704385995864868,
+      "learning_rate": 0.001,
+      "loss": 0.4472,
+      "step": 298500
+    },
+    {
+      "epoch": 96.50937297996121,
+      "grad_norm": 1.5044299364089966,
+      "learning_rate": 0.001,
+      "loss": 0.4453,
+      "step": 298600
+    },
+    {
+      "epoch": 96.54169360051714,
+      "grad_norm": 1.373509407043457,
+      "learning_rate": 0.001,
+      "loss": 0.4599,
+      "step": 298700
+    },
+    {
+      "epoch": 96.57401422107304,
+      "grad_norm": 1.180299162864685,
+      "learning_rate": 0.001,
+      "loss": 0.4476,
+      "step": 298800
+    },
+    {
+      "epoch": 96.60633484162896,
+      "grad_norm": 2.22157621383667,
+      "learning_rate": 0.001,
+      "loss": 0.451,
+      "step": 298900
+    },
+    {
+      "epoch": 96.63865546218487,
+      "grad_norm": 1.5506231784820557,
+      "learning_rate": 0.001,
+      "loss": 0.461,
+      "step": 299000
+    },
+    {
+      "epoch": 96.6709760827408,
+      "grad_norm": 1.7377060651779175,
+      "learning_rate": 0.001,
+      "loss": 0.4565,
+      "step": 299100
+    },
+    {
+      "epoch": 96.7032967032967,
+      "grad_norm": 1.1429818868637085,
+      "learning_rate": 0.001,
+      "loss": 0.4613,
+      "step": 299200
+    },
+    {
+      "epoch": 96.73561732385262,
+      "grad_norm": 1.4461809396743774,
+      "learning_rate": 0.001,
+      "loss": 0.4636,
+      "step": 299300
+    },
+    {
+      "epoch": 96.76793794440853,
+      "grad_norm": 1.6371991634368896,
+      "learning_rate": 0.001,
+      "loss": 0.4725,
+      "step": 299400
+    },
+    {
+      "epoch": 96.80025856496445,
+      "grad_norm": 2.10194993019104,
+      "learning_rate": 0.001,
+      "loss": 0.4548,
+      "step": 299500
+    },
+    {
+      "epoch": 96.83257918552036,
+      "grad_norm": 1.3943737745285034,
+      "learning_rate": 0.001,
+      "loss": 0.4732,
+      "step": 299600
+    },
+    {
+      "epoch": 96.86489980607628,
+      "grad_norm": 1.4672203063964844,
+      "learning_rate": 0.001,
+      "loss": 0.465,
+      "step": 299700
+    },
+    {
+      "epoch": 96.89722042663219,
+      "grad_norm": 1.6431360244750977,
+      "learning_rate": 0.001,
+      "loss": 0.4576,
+      "step": 299800
+    },
+    {
+      "epoch": 96.92954104718811,
+      "grad_norm": 1.731074333190918,
+      "learning_rate": 0.001,
+      "loss": 0.4686,
+      "step": 299900
+    },
+    {
+      "epoch": 96.96186166774402,
+      "grad_norm": 1.5479676723480225,
+      "learning_rate": 0.001,
+      "loss": 0.4716,
+      "step": 300000
+    },
+    {
+      "epoch": 96.99418228829994,
+      "grad_norm": 1.5700945854187012,
+      "learning_rate": 0.001,
+      "loss": 0.4677,
+      "step": 300100
+    },
+    {
+      "epoch": 97.02650290885585,
+      "grad_norm": 1.269909381866455,
+      "learning_rate": 0.001,
+      "loss": 0.4298,
+      "step": 300200
+    },
+    {
+      "epoch": 97.05882352941177,
+      "grad_norm": 1.5230191946029663,
+      "learning_rate": 0.001,
+      "loss": 0.4174,
+      "step": 300300
+    },
+    {
+      "epoch": 97.09114414996768,
+      "grad_norm": 1.2741731405258179,
+      "learning_rate": 0.001,
+      "loss": 0.4206,
+      "step": 300400
+    },
+    {
+      "epoch": 97.1234647705236,
+      "grad_norm": 1.440308690071106,
+      "learning_rate": 0.001,
+      "loss": 0.4147,
+      "step": 300500
+    },
+    {
+      "epoch": 97.1557853910795,
+      "grad_norm": 1.1169484853744507,
+      "learning_rate": 0.001,
+      "loss": 0.4224,
+      "step": 300600
+    },
+    {
+      "epoch": 97.18810601163543,
+      "grad_norm": 1.6060082912445068,
+      "learning_rate": 0.001,
+      "loss": 0.4322,
+      "step": 300700
+    },
+    {
+      "epoch": 97.22042663219133,
+      "grad_norm": 1.4612475633621216,
+      "learning_rate": 0.001,
+      "loss": 0.4353,
+      "step": 300800
+    },
+    {
+      "epoch": 97.25274725274726,
+      "grad_norm": 1.2991474866867065,
+      "learning_rate": 0.001,
+      "loss": 0.4241,
+      "step": 300900
+    },
+    {
+      "epoch": 97.28506787330316,
+      "grad_norm": 1.3116704225540161,
+      "learning_rate": 0.001,
+      "loss": 0.4366,
+      "step": 301000
+    },
+    {
+      "epoch": 97.31738849385908,
+      "grad_norm": 1.2389549016952515,
+      "learning_rate": 0.001,
+      "loss": 0.4384,
+      "step": 301100
+    },
+    {
+      "epoch": 97.34970911441499,
+      "grad_norm": 1.2728149890899658,
+      "learning_rate": 0.001,
+      "loss": 0.4372,
+      "step": 301200
+    },
+    {
+      "epoch": 97.38202973497091,
+      "grad_norm": 1.3004136085510254,
+      "learning_rate": 0.001,
+      "loss": 0.4373,
+      "step": 301300
+    },
+    {
+      "epoch": 97.41435035552682,
+      "grad_norm": 1.227472186088562,
+      "learning_rate": 0.001,
+      "loss": 0.4422,
+      "step": 301400
+    },
+    {
+      "epoch": 97.44667097608274,
+      "grad_norm": 1.1583950519561768,
+      "learning_rate": 0.001,
+      "loss": 0.4505,
+      "step": 301500
+    },
+    {
+      "epoch": 97.47899159663865,
+      "grad_norm": 1.5739110708236694,
+      "learning_rate": 0.001,
+      "loss": 0.4521,
+      "step": 301600
+    },
+    {
+      "epoch": 97.51131221719457,
+      "grad_norm": 1.3445355892181396,
+      "learning_rate": 0.001,
+      "loss": 0.4532,
+      "step": 301700
+    },
+    {
+      "epoch": 97.54363283775048,
+      "grad_norm": 1.2330920696258545,
+      "learning_rate": 0.001,
+      "loss": 0.4477,
+      "step": 301800
+    },
+    {
+      "epoch": 97.5759534583064,
+      "grad_norm": 1.5380816459655762,
+      "learning_rate": 0.001,
+      "loss": 0.4513,
+      "step": 301900
+    },
+    {
+      "epoch": 97.60827407886231,
+      "grad_norm": 1.2839477062225342,
+      "learning_rate": 0.001,
+      "loss": 0.4455,
+      "step": 302000
+    },
+    {
+      "epoch": 97.64059469941823,
+      "grad_norm": 1.4159269332885742,
+      "learning_rate": 0.001,
+      "loss": 0.4491,
+      "step": 302100
+    },
+    {
+      "epoch": 97.67291531997414,
+      "grad_norm": 1.401963472366333,
+      "learning_rate": 0.001,
+      "loss": 0.4503,
+      "step": 302200
+    },
+    {
+      "epoch": 97.70523594053006,
+      "grad_norm": 1.7864422798156738,
+      "learning_rate": 0.001,
+      "loss": 0.4675,
+      "step": 302300
+    },
+    {
+      "epoch": 97.73755656108597,
+      "grad_norm": 1.2578034400939941,
+      "learning_rate": 0.001,
+      "loss": 0.4583,
+      "step": 302400
+    },
+    {
+      "epoch": 97.76987718164189,
+      "grad_norm": 1.2712514400482178,
+      "learning_rate": 0.001,
+      "loss": 0.4646,
+      "step": 302500
+    },
+    {
+      "epoch": 97.8021978021978,
+      "grad_norm": 1.465354561805725,
+      "learning_rate": 0.001,
+      "loss": 0.4548,
+      "step": 302600
+    },
+    {
+      "epoch": 97.83451842275372,
+      "grad_norm": 1.1005356311798096,
+      "learning_rate": 0.001,
+      "loss": 0.4517,
+      "step": 302700
+    },
+    {
+      "epoch": 97.86683904330962,
+      "grad_norm": 1.7632066011428833,
+      "learning_rate": 0.001,
+      "loss": 0.4521,
+      "step": 302800
+    },
+    {
+      "epoch": 97.89915966386555,
+      "grad_norm": 1.5967885255813599,
+      "learning_rate": 0.001,
+      "loss": 0.473,
+      "step": 302900
+    },
+    {
+      "epoch": 97.93148028442145,
+      "grad_norm": 1.3136255741119385,
+      "learning_rate": 0.001,
+      "loss": 0.4713,
+      "step": 303000
+    },
+    {
+      "epoch": 97.96380090497738,
+      "grad_norm": 1.0595479011535645,
+      "learning_rate": 0.001,
+      "loss": 0.4598,
+      "step": 303100
+    },
+    {
+      "epoch": 97.99612152553328,
+      "grad_norm": 1.0992523431777954,
+      "learning_rate": 0.001,
+      "loss": 0.464,
+      "step": 303200
+    },
+    {
+      "epoch": 98.0284421460892,
+      "grad_norm": 1.4200688600540161,
+      "learning_rate": 0.001,
+      "loss": 0.4209,
+      "step": 303300
+    },
+    {
+      "epoch": 98.06076276664513,
+      "grad_norm": 1.107372760772705,
+      "learning_rate": 0.001,
+      "loss": 0.418,
+      "step": 303400
+    },
+    {
+      "epoch": 98.09308338720103,
+      "grad_norm": 0.9656426310539246,
+      "learning_rate": 0.001,
+      "loss": 0.4113,
+      "step": 303500
+    },
+    {
+      "epoch": 98.12540400775696,
+      "grad_norm": 0.9549182057380676,
+      "learning_rate": 0.001,
+      "loss": 0.4301,
+      "step": 303600
+    },
+    {
+      "epoch": 98.15772462831286,
+      "grad_norm": 1.4307749271392822,
+      "learning_rate": 0.001,
+      "loss": 0.4327,
+      "step": 303700
+    },
+    {
+      "epoch": 98.19004524886878,
+      "grad_norm": 1.2919740676879883,
+      "learning_rate": 0.001,
+      "loss": 0.4276,
+      "step": 303800
+    },
+    {
+      "epoch": 98.22236586942469,
+      "grad_norm": 1.20173978805542,
+      "learning_rate": 0.001,
+      "loss": 0.4228,
+      "step": 303900
+    },
+    {
+      "epoch": 98.25468648998061,
+      "grad_norm": 1.2167638540267944,
+      "learning_rate": 0.001,
+      "loss": 0.4178,
+      "step": 304000
+    },
+    {
+      "epoch": 98.28700711053652,
+      "grad_norm": 1.0864397287368774,
+      "learning_rate": 0.001,
+      "loss": 0.4273,
+      "step": 304100
+    },
+    {
+      "epoch": 98.31932773109244,
+      "grad_norm": 1.4295344352722168,
+      "learning_rate": 0.001,
+      "loss": 0.4379,
+      "step": 304200
+    },
+    {
+      "epoch": 98.35164835164835,
+      "grad_norm": 1.2204785346984863,
+      "learning_rate": 0.001,
+      "loss": 0.4396,
+      "step": 304300
+    },
+    {
+      "epoch": 98.38396897220427,
+      "grad_norm": 1.4378937482833862,
+      "learning_rate": 0.001,
+      "loss": 0.4384,
+      "step": 304400
+    },
+    {
+      "epoch": 98.41628959276018,
+      "grad_norm": 1.0496668815612793,
+      "learning_rate": 0.001,
+      "loss": 0.4438,
+      "step": 304500
+    },
+    {
+      "epoch": 98.4486102133161,
+      "grad_norm": 1.357386827468872,
+      "learning_rate": 0.001,
+      "loss": 0.4592,
+      "step": 304600
+    },
+    {
+      "epoch": 98.48093083387201,
+      "grad_norm": 1.6617597341537476,
+      "learning_rate": 0.001,
+      "loss": 0.4409,
+      "step": 304700
+    },
+    {
+      "epoch": 98.51325145442793,
+      "grad_norm": 1.351397156715393,
+      "learning_rate": 0.001,
+      "loss": 0.4471,
+      "step": 304800
+    },
+    {
+      "epoch": 98.54557207498384,
+      "grad_norm": 1.240196704864502,
+      "learning_rate": 0.001,
+      "loss": 0.4375,
+      "step": 304900
+    },
+    {
+      "epoch": 98.57789269553976,
+      "grad_norm": 1.221144437789917,
+      "learning_rate": 0.001,
+      "loss": 0.4476,
+      "step": 305000
+    },
+    {
+      "epoch": 98.61021331609567,
+      "grad_norm": 1.5508735179901123,
+      "learning_rate": 0.001,
+      "loss": 0.4431,
+      "step": 305100
+    },
+    {
+      "epoch": 98.64253393665159,
+      "grad_norm": 1.0978624820709229,
+      "learning_rate": 0.001,
+      "loss": 0.4461,
+      "step": 305200
+    },
+    {
+      "epoch": 98.6748545572075,
+      "grad_norm": 1.2614808082580566,
+      "learning_rate": 0.001,
+      "loss": 0.4397,
+      "step": 305300
+    },
+    {
+      "epoch": 98.70717517776342,
+      "grad_norm": 1.1213574409484863,
+      "learning_rate": 0.001,
+      "loss": 0.439,
+      "step": 305400
+    },
+    {
+      "epoch": 98.73949579831933,
+      "grad_norm": 1.5465046167373657,
+      "learning_rate": 0.001,
+      "loss": 0.4592,
+      "step": 305500
+    },
+    {
+      "epoch": 98.77181641887525,
+      "grad_norm": 1.1399887800216675,
+      "learning_rate": 0.001,
+      "loss": 0.4663,
+      "step": 305600
+    },
+    {
+      "epoch": 98.80413703943115,
+      "grad_norm": 1.1044340133666992,
+      "learning_rate": 0.001,
+      "loss": 0.4492,
+      "step": 305700
+    },
+    {
+      "epoch": 98.83645765998708,
+      "grad_norm": 1.1779178380966187,
+      "learning_rate": 0.001,
+      "loss": 0.4516,
+      "step": 305800
+    },
+    {
+      "epoch": 98.86877828054298,
+      "grad_norm": 1.3684269189834595,
+      "learning_rate": 0.001,
+      "loss": 0.4532,
+      "step": 305900
+    },
+    {
+      "epoch": 98.9010989010989,
+      "grad_norm": 1.3492311239242554,
+      "learning_rate": 0.001,
+      "loss": 0.4511,
+      "step": 306000
+    },
+    {
+      "epoch": 98.93341952165481,
+      "grad_norm": 1.1889764070510864,
+      "learning_rate": 0.001,
+      "loss": 0.4664,
+      "step": 306100
+    },
+    {
+      "epoch": 98.96574014221073,
+      "grad_norm": 0.8841199278831482,
+      "learning_rate": 0.001,
+      "loss": 0.4649,
+      "step": 306200
+    },
+    {
+      "epoch": 98.99806076276664,
+      "grad_norm": 0.9290673136711121,
+      "learning_rate": 0.001,
+      "loss": 0.4531,
+      "step": 306300
+    },
+    {
+      "epoch": 99.03038138332256,
+      "grad_norm": 1.022675633430481,
+      "learning_rate": 0.001,
+      "loss": 0.42,
+      "step": 306400
+    },
+    {
+      "epoch": 99.06270200387847,
+      "grad_norm": 1.138304352760315,
+      "learning_rate": 0.001,
+      "loss": 0.4125,
+      "step": 306500
+    },
+    {
+      "epoch": 99.09502262443439,
+      "grad_norm": 1.1132817268371582,
+      "learning_rate": 0.001,
+      "loss": 0.4205,
+      "step": 306600
+    },
+    {
+      "epoch": 99.1273432449903,
+      "grad_norm": 1.1500136852264404,
+      "learning_rate": 0.001,
+      "loss": 0.4136,
+      "step": 306700
+    },
+    {
+      "epoch": 99.15966386554622,
+      "grad_norm": 1.1773744821548462,
+      "learning_rate": 0.001,
+      "loss": 0.4212,
+      "step": 306800
+    },
+    {
+      "epoch": 99.19198448610213,
+      "grad_norm": 1.4039117097854614,
+      "learning_rate": 0.001,
+      "loss": 0.4338,
+      "step": 306900
+    },
+    {
+      "epoch": 99.22430510665805,
+      "grad_norm": 1.0985853672027588,
+      "learning_rate": 0.001,
+      "loss": 0.4278,
+      "step": 307000
+    },
+    {
+      "epoch": 99.25662572721396,
+      "grad_norm": 0.7000985741615295,
+      "learning_rate": 0.001,
+      "loss": 0.424,
+      "step": 307100
+    },
+    {
+      "epoch": 99.28894634776988,
+      "grad_norm": 1.5359601974487305,
+      "learning_rate": 0.001,
+      "loss": 0.4353,
+      "step": 307200
+    },
+    {
+      "epoch": 99.32126696832579,
+      "grad_norm": 0.8994166254997253,
+      "learning_rate": 0.001,
+      "loss": 0.4382,
+      "step": 307300
+    },
+    {
+      "epoch": 99.35358758888171,
+      "grad_norm": 0.724768340587616,
+      "learning_rate": 0.001,
+      "loss": 0.4355,
+      "step": 307400
+    },
+    {
+      "epoch": 99.38590820943762,
+      "grad_norm": 0.8465492129325867,
+      "learning_rate": 0.001,
+      "loss": 0.444,
+      "step": 307500
+    },
+    {
+      "epoch": 99.41822882999354,
+      "grad_norm": 1.830216884613037,
+      "learning_rate": 0.001,
+      "loss": 0.4336,
+      "step": 307600
+    },
+    {
+      "epoch": 99.45054945054945,
+      "grad_norm": 1.0542271137237549,
+      "learning_rate": 0.001,
+      "loss": 0.4336,
+      "step": 307700
+    },
+    {
+      "epoch": 99.48287007110537,
+      "grad_norm": 0.8846477270126343,
+      "learning_rate": 0.001,
+      "loss": 0.4516,
+      "step": 307800
+    },
+    {
+      "epoch": 99.51519069166127,
+      "grad_norm": 0.8210715055465698,
+      "learning_rate": 0.001,
+      "loss": 0.4412,
+      "step": 307900
+    },
+    {
+      "epoch": 99.5475113122172,
+      "grad_norm": 0.9523438215255737,
+      "learning_rate": 0.001,
+      "loss": 0.4341,
+      "step": 308000
+    },
+    {
+      "epoch": 99.5798319327731,
+      "grad_norm": 1.0389193296432495,
+      "learning_rate": 0.001,
+      "loss": 0.4382,
+      "step": 308100
+    },
+    {
+      "epoch": 99.61215255332903,
+      "grad_norm": 0.9580590128898621,
+      "learning_rate": 0.001,
+      "loss": 0.4358,
+      "step": 308200
+    },
+    {
+      "epoch": 99.64447317388493,
+      "grad_norm": 1.2587848901748657,
+      "learning_rate": 0.001,
+      "loss": 0.4293,
+      "step": 308300
+    },
+    {
+      "epoch": 99.67679379444085,
+      "grad_norm": 1.1644697189331055,
+      "learning_rate": 0.001,
+      "loss": 0.4547,
+      "step": 308400
+    },
+    {
+      "epoch": 99.70911441499676,
+      "grad_norm": 1.3537368774414062,
+      "learning_rate": 0.001,
+      "loss": 0.4463,
+      "step": 308500
+    },
+    {
+      "epoch": 99.74143503555268,
+      "grad_norm": 1.1971321105957031,
+      "learning_rate": 0.001,
+      "loss": 0.4542,
+      "step": 308600
+    },
+    {
+      "epoch": 99.77375565610859,
+      "grad_norm": 438.0790710449219,
+      "learning_rate": 0.001,
+      "loss": 0.5571,
+      "step": 308700
+    },
+    {
+      "epoch": 99.80607627666451,
+      "grad_norm": 1.3546308279037476,
+      "learning_rate": 0.001,
+      "loss": 0.4643,
+      "step": 308800
+    },
+    {
+      "epoch": 99.83839689722042,
+      "grad_norm": 0.8377397060394287,
+      "learning_rate": 0.001,
+      "loss": 0.444,
+      "step": 308900
+    },
+    {
+      "epoch": 99.87071751777634,
+      "grad_norm": 1.2964484691619873,
+      "learning_rate": 0.001,
+      "loss": 0.4493,
+      "step": 309000
+    },
+    {
+      "epoch": 99.90303813833225,
+      "grad_norm": 1.27469801902771,
+      "learning_rate": 0.001,
+      "loss": 0.4468,
+      "step": 309100
+    },
+    {
+      "epoch": 99.93535875888817,
+      "grad_norm": 0.9581504464149475,
+      "learning_rate": 0.001,
+      "loss": 0.4506,
+      "step": 309200
+    },
+    {
+      "epoch": 99.96767937944408,
+      "grad_norm": 1.0009796619415283,
+      "learning_rate": 0.001,
+      "loss": 0.4404,
+      "step": 309300
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 0.019829602912068367,
+      "learning_rate": 0.001,
+      "loss": 0.4208,
+      "step": 309400
+    },
+    {
+      "epoch": 100.0,
+      "step": 309400,
+      "total_flos": 3.569919021005046e+17,
+      "train_loss": 0.043210694436959164,
+      "train_runtime": 3964.1937,
+      "train_samples_per_second": 2497.355,
+      "train_steps_per_second": 78.049
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 309400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 100,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.569919021005046e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}