diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28707 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 1275,
+  "global_step": 20392,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019615535504119262,
+      "eval_loss": 13.359358787536621,
+      "eval_runtime": 7.6451,
+      "eval_samples_per_second": 27.338,
+      "eval_steps_per_second": 13.734,
+      "step": 1
+    },
+    {
+      "epoch": 0.000980776775205963,
+      "grad_norm": 261.527587890625,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 13.4315,
+      "step": 5
+    },
+    {
+      "epoch": 0.001961553550411926,
+      "grad_norm": 281.323486328125,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 13.1245,
+      "step": 10
+    },
+    {
+      "epoch": 0.0029423303256178894,
+      "grad_norm": 294.9424133300781,
+      "learning_rate": 1.5e-06,
+      "loss": 13.3844,
+      "step": 15
+    },
+    {
+      "epoch": 0.003923107100823852,
+      "grad_norm": 314.61883544921875,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 12.6558,
+      "step": 20
+    },
+    {
+      "epoch": 0.004903883876029816,
+      "grad_norm": 301.4609375,
+      "learning_rate": 2.5e-06,
+      "loss": 13.2002,
+      "step": 25
+    },
+    {
+      "epoch": 0.005884660651235779,
+      "grad_norm": 265.4776306152344,
+      "learning_rate": 3e-06,
+      "loss": 13.0372,
+      "step": 30
+    },
+    {
+      "epoch": 0.006865437426441742,
+      "grad_norm": 231.08770751953125,
+      "learning_rate": 3.5e-06,
+      "loss": 13.0588,
+      "step": 35
+    },
+    {
+      "epoch": 0.007846214201647704,
+      "grad_norm": 243.46893310546875,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 12.8002,
+      "step": 40
+    },
+    {
+      "epoch": 0.008826990976853669,
+      "grad_norm": 272.3201599121094,
+      "learning_rate": 4.5e-06,
+      "loss": 12.2782,
+      "step": 45
+    },
+    {
+      "epoch": 0.009807767752059632,
+      "grad_norm": 224.776611328125,
+      "learning_rate": 5e-06,
+      "loss": 11.1628,
+      "step": 50
+    },
+    {
+      "epoch": 0.010788544527265595,
+      "grad_norm": 220.5045928955078,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 10.4563,
+      "step": 55
+    },
+    {
+      "epoch": 0.011769321302471557,
+      "grad_norm": 219.2426300048828,
+      "learning_rate": 6e-06,
+      "loss": 9.908,
+      "step": 60
+    },
+    {
+      "epoch": 0.01275009807767752,
+      "grad_norm": 119.27024841308594,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 10.026,
+      "step": 65
+    },
+    {
+      "epoch": 0.013730874852883483,
+      "grad_norm": 91.62039947509766,
+      "learning_rate": 7e-06,
+      "loss": 9.1984,
+      "step": 70
+    },
+    {
+      "epoch": 0.014711651628089448,
+      "grad_norm": 83.55156707763672,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 8.457,
+      "step": 75
+    },
+    {
+      "epoch": 0.01569242840329541,
+      "grad_norm": 67.35411071777344,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 8.3446,
+      "step": 80
+    },
+    {
+      "epoch": 0.016673205178501373,
+      "grad_norm": 50.44355010986328,
+      "learning_rate": 8.5e-06,
+      "loss": 8.1199,
+      "step": 85
+    },
+    {
+      "epoch": 0.017653981953707338,
+      "grad_norm": 45.920475006103516,
+      "learning_rate": 9e-06,
+      "loss": 7.9275,
+      "step": 90
+    },
+    {
+      "epoch": 0.0186347587289133,
+      "grad_norm": 51.882083892822266,
+      "learning_rate": 9.5e-06,
+      "loss": 7.7225,
+      "step": 95
+    },
+    {
+      "epoch": 0.019615535504119264,
+      "grad_norm": 36.69080352783203,
+      "learning_rate": 1e-05,
+      "loss": 7.2467,
+      "step": 100
+    },
+    {
+      "epoch": 0.020596312279325225,
+      "grad_norm": 55.873043060302734,
+      "learning_rate": 9.999998501937153e-06,
+      "loss": 7.2832,
+      "step": 105
+    },
+    {
+      "epoch": 0.02157708905453119,
+      "grad_norm": 28.33873176574707,
+      "learning_rate": 9.999994007749506e-06,
+      "loss": 7.3216,
+      "step": 110
+    },
+    {
+      "epoch": 0.02255786582973715,
+      "grad_norm": 33.89677047729492,
+      "learning_rate": 9.99998651743975e-06,
+      "loss": 7.2139,
+      "step": 115
+    },
+    {
+      "epoch": 0.023538642604943115,
+      "grad_norm": 33.673927307128906,
+      "learning_rate": 9.99997603101238e-06,
+      "loss": 6.7812,
+      "step": 120
+    },
+    {
+      "epoch": 0.02451941938014908,
+      "grad_norm": 35.81504821777344,
+      "learning_rate": 9.999962548473674e-06,
+      "loss": 7.1054,
+      "step": 125
+    },
+    {
+      "epoch": 0.02550019615535504,
+      "grad_norm": 36.798221588134766,
+      "learning_rate": 9.999946069831713e-06,
+      "loss": 7.4411,
+      "step": 130
+    },
+    {
+      "epoch": 0.026480972930561005,
+      "grad_norm": 35.6644172668457,
+      "learning_rate": 9.999926595096373e-06,
+      "loss": 7.3205,
+      "step": 135
+    },
+    {
+      "epoch": 0.027461749705766966,
+      "grad_norm": 20.2699031829834,
+      "learning_rate": 9.999904124279322e-06,
+      "loss": 6.5323,
+      "step": 140
+    },
+    {
+      "epoch": 0.02844252648097293,
+      "grad_norm": 34.727783203125,
+      "learning_rate": 9.999878657394024e-06,
+      "loss": 6.7351,
+      "step": 145
+    },
+    {
+      "epoch": 0.029423303256178895,
+      "grad_norm": 21.760740280151367,
+      "learning_rate": 9.999850194455741e-06,
+      "loss": 6.4626,
+      "step": 150
+    },
+    {
+      "epoch": 0.030404080031384857,
+      "grad_norm": 25.09556770324707,
+      "learning_rate": 9.99981873548153e-06,
+      "loss": 6.5722,
+      "step": 155
+    },
+    {
+      "epoch": 0.03138485680659082,
+      "grad_norm": 19.870086669921875,
+      "learning_rate": 9.999784280490239e-06,
+      "loss": 6.2656,
+      "step": 160
+    },
+    {
+      "epoch": 0.03236563358179678,
+      "grad_norm": 28.291837692260742,
+      "learning_rate": 9.999746829502516e-06,
+      "loss": 6.4729,
+      "step": 165
+    },
+    {
+      "epoch": 0.03334641035700275,
+      "grad_norm": 27.245208740234375,
+      "learning_rate": 9.9997063825408e-06,
+      "loss": 6.3256,
+      "step": 170
+    },
+    {
+      "epoch": 0.03432718713220871,
+      "grad_norm": 25.775121688842773,
+      "learning_rate": 9.999662939629335e-06,
+      "loss": 6.6193,
+      "step": 175
+    },
+    {
+      "epoch": 0.035307963907414676,
+      "grad_norm": 24.102081298828125,
+      "learning_rate": 9.999616500794144e-06,
+      "loss": 6.2076,
+      "step": 180
+    },
+    {
+      "epoch": 0.036288740682620634,
+      "grad_norm": 34.60786437988281,
+      "learning_rate": 9.99956706606306e-06,
+      "loss": 7.2322,
+      "step": 185
+    },
+    {
+      "epoch": 0.0372695174578266,
+      "grad_norm": 21.22093391418457,
+      "learning_rate": 9.999514635465706e-06,
+      "loss": 5.9367,
+      "step": 190
+    },
+    {
+      "epoch": 0.03825029423303256,
+      "grad_norm": 19.872236251831055,
+      "learning_rate": 9.999459209033495e-06,
+      "loss": 6.1443,
+      "step": 195
+    },
+    {
+      "epoch": 0.03923107100823853,
+      "grad_norm": 29.4232234954834,
+      "learning_rate": 9.999400786799644e-06,
+      "loss": 6.2759,
+      "step": 200
+    },
+    {
+      "epoch": 0.040211847783444485,
+      "grad_norm": 30.81177520751953,
+      "learning_rate": 9.99933936879916e-06,
+      "loss": 6.5985,
+      "step": 205
+    },
+    {
+      "epoch": 0.04119262455865045,
+      "grad_norm": 24.815486907958984,
+      "learning_rate": 9.999274955068845e-06,
+      "loss": 5.9052,
+      "step": 210
+    },
+    {
+      "epoch": 0.042173401333856414,
+      "grad_norm": 52.55622482299805,
+      "learning_rate": 9.9992075456473e-06,
+      "loss": 6.3684,
+      "step": 215
+    },
+    {
+      "epoch": 0.04315417810906238,
+      "grad_norm": 24.462528228759766,
+      "learning_rate": 9.999137140574914e-06,
+      "loss": 6.2602,
+      "step": 220
+    },
+    {
+      "epoch": 0.04413495488426834,
+      "grad_norm": 17.4281005859375,
+      "learning_rate": 9.99906373989388e-06,
+      "loss": 6.1871,
+      "step": 225
+    },
+    {
+      "epoch": 0.0451157316594743,
+      "grad_norm": 28.467485427856445,
+      "learning_rate": 9.998987343648182e-06,
+      "loss": 6.0398,
+      "step": 230
+    },
+    {
+      "epoch": 0.046096508434680265,
+      "grad_norm": 28.41315269470215,
+      "learning_rate": 9.998907951883592e-06,
+      "loss": 6.3472,
+      "step": 235
+    },
+    {
+      "epoch": 0.04707728520988623,
+      "grad_norm": 20.66746711730957,
+      "learning_rate": 9.998825564647689e-06,
+      "loss": 6.3148,
+      "step": 240
+    },
+    {
+      "epoch": 0.048058061985092194,
+      "grad_norm": 19.29747200012207,
+      "learning_rate": 9.998740181989842e-06,
+      "loss": 6.0387,
+      "step": 245
+    },
+    {
+      "epoch": 0.04903883876029816,
+      "grad_norm": 29.47091293334961,
+      "learning_rate": 9.998651803961212e-06,
+      "loss": 6.3439,
+      "step": 250
+    },
+    {
+      "epoch": 0.05001961553550412,
+      "grad_norm": 47.9112548828125,
+      "learning_rate": 9.998560430614759e-06,
+      "loss": 6.3999,
+      "step": 255
+    },
+    {
+      "epoch": 0.05100039231071008,
+      "grad_norm": 26.031343460083008,
+      "learning_rate": 9.998466062005234e-06,
+      "loss": 6.4955,
+      "step": 260
+    },
+    {
+      "epoch": 0.051981169085916046,
+      "grad_norm": 28.60124969482422,
+      "learning_rate": 9.998368698189187e-06,
+      "loss": 6.1393,
+      "step": 265
+    },
+    {
+      "epoch": 0.05296194586112201,
+      "grad_norm": 22.979835510253906,
+      "learning_rate": 9.998268339224958e-06,
+      "loss": 6.1923,
+      "step": 270
+    },
+    {
+      "epoch": 0.053942722636327975,
+      "grad_norm": 26.163986206054688,
+      "learning_rate": 9.99816498517269e-06,
+      "loss": 6.4099,
+      "step": 275
+    },
+    {
+      "epoch": 0.05492349941153393,
+      "grad_norm": 30.09125518798828,
+      "learning_rate": 9.998058636094312e-06,
+      "loss": 6.1997,
+      "step": 280
+    },
+    {
+      "epoch": 0.0559042761867399,
+      "grad_norm": 33.007118225097656,
+      "learning_rate": 9.99794929205355e-06,
+      "loss": 6.3884,
+      "step": 285
+    },
+    {
+      "epoch": 0.05688505296194586,
+      "grad_norm": 16.16930389404297,
+      "learning_rate": 9.997836953115927e-06,
+      "loss": 5.9929,
+      "step": 290
+    },
+    {
+      "epoch": 0.057865829737151826,
+      "grad_norm": 15.623757362365723,
+      "learning_rate": 9.99772161934876e-06,
+      "loss": 6.0234,
+      "step": 295
+    },
+    {
+      "epoch": 0.05884660651235779,
+      "grad_norm": 21.748035430908203,
+      "learning_rate": 9.997603290821158e-06,
+      "loss": 5.7948,
+      "step": 300
+    },
+    {
+      "epoch": 0.05982738328756375,
+      "grad_norm": 15.957292556762695,
+      "learning_rate": 9.997481967604028e-06,
+      "loss": 5.8803,
+      "step": 305
+    },
+    {
+      "epoch": 0.06080816006276971,
+      "grad_norm": 16.29456329345703,
+      "learning_rate": 9.99735764977007e-06,
+      "loss": 6.095,
+      "step": 310
+    },
+    {
+      "epoch": 0.06178893683797568,
+      "grad_norm": 13.741049766540527,
+      "learning_rate": 9.997230337393777e-06,
+      "loss": 6.4022,
+      "step": 315
+    },
+    {
+      "epoch": 0.06276971361318164,
+      "grad_norm": 26.837913513183594,
+      "learning_rate": 9.99710003055144e-06,
+      "loss": 6.4859,
+      "step": 320
+    },
+    {
+      "epoch": 0.0637504903883876,
+      "grad_norm": 14.79397964477539,
+      "learning_rate": 9.99696672932114e-06,
+      "loss": 5.7493,
+      "step": 325
+    },
+    {
+      "epoch": 0.06473126716359356,
+      "grad_norm": 14.920068740844727,
+      "learning_rate": 9.996830433782754e-06,
+      "loss": 5.8187,
+      "step": 330
+    },
+    {
+      "epoch": 0.06571204393879954,
+      "grad_norm": 18.98463249206543,
+      "learning_rate": 9.996691144017957e-06,
+      "loss": 5.9097,
+      "step": 335
+    },
+    {
+      "epoch": 0.0666928207140055,
+      "grad_norm": 27.435945510864258,
+      "learning_rate": 9.99654886011021e-06,
+      "loss": 6.0536,
+      "step": 340
+    },
+    {
+      "epoch": 0.06767359748921145,
+      "grad_norm": 13.197905540466309,
+      "learning_rate": 9.99640358214478e-06,
+      "loss": 5.9436,
+      "step": 345
+    },
+    {
+      "epoch": 0.06865437426441742,
+      "grad_norm": 23.347511291503906,
+      "learning_rate": 9.996255310208715e-06,
+      "loss": 6.0234,
+      "step": 350
+    },
+    {
+      "epoch": 0.06963515103962338,
+      "grad_norm": 40.36224365234375,
+      "learning_rate": 9.996104044390866e-06,
+      "loss": 5.834,
+      "step": 355
+    },
+    {
+      "epoch": 0.07061592781482935,
+      "grad_norm": 18.484323501586914,
+      "learning_rate": 9.995949784781873e-06,
+      "loss": 6.143,
+      "step": 360
+    },
+    {
+      "epoch": 0.07159670459003531,
+      "grad_norm": 27.709238052368164,
+      "learning_rate": 9.995792531474175e-06,
+      "loss": 6.0427,
+      "step": 365
+    },
+    {
+      "epoch": 0.07257748136524127,
+      "grad_norm": 12.617931365966797,
+      "learning_rate": 9.995632284562002e-06,
+      "loss": 5.8578,
+      "step": 370
+    },
+    {
+      "epoch": 0.07355825814044724,
+      "grad_norm": 19.61956214904785,
+      "learning_rate": 9.995469044141377e-06,
+      "loss": 5.8423,
+      "step": 375
+    },
+    {
+      "epoch": 0.0745390349156532,
+      "grad_norm": 12.127665519714355,
+      "learning_rate": 9.995302810310116e-06,
+      "loss": 6.1397,
+      "step": 380
+    },
+    {
+      "epoch": 0.07551981169085915,
+      "grad_norm": 23.63418197631836,
+      "learning_rate": 9.995133583167833e-06,
+      "loss": 5.9291,
+      "step": 385
+    },
+    {
+      "epoch": 0.07650058846606513,
+      "grad_norm": 19.848413467407227,
+      "learning_rate": 9.994961362815934e-06,
+      "loss": 5.8147,
+      "step": 390
+    },
+    {
+      "epoch": 0.07748136524127108,
+      "grad_norm": 24.281089782714844,
+      "learning_rate": 9.994786149357614e-06,
+      "loss": 6.1719,
+      "step": 395
+    },
+    {
+      "epoch": 0.07846214201647705,
+      "grad_norm": 26.30230712890625,
+      "learning_rate": 9.99460794289787e-06,
+      "loss": 6.0022,
+      "step": 400
+    },
+    {
+      "epoch": 0.07944291879168301,
+      "grad_norm": 20.717947006225586,
+      "learning_rate": 9.994426743543483e-06,
+      "loss": 6.1346,
+      "step": 405
+    },
+    {
+      "epoch": 0.08042369556688897,
+      "grad_norm": 13.985010147094727,
+      "learning_rate": 9.994242551403036e-06,
+      "loss": 5.7892,
+      "step": 410
+    },
+    {
+      "epoch": 0.08140447234209494,
+      "grad_norm": 26.171998977661133,
+      "learning_rate": 9.9940553665869e-06,
+      "loss": 6.1193,
+      "step": 415
+    },
+    {
+      "epoch": 0.0823852491173009,
+      "grad_norm": 19.03148078918457,
+      "learning_rate": 9.993865189207242e-06,
+      "loss": 5.8301,
+      "step": 420
+    },
+    {
+      "epoch": 0.08336602589250687,
+      "grad_norm": 18.318777084350586,
+      "learning_rate": 9.993672019378017e-06,
+      "loss": 5.7323,
+      "step": 425
+    },
+    {
+      "epoch": 0.08434680266771283,
+      "grad_norm": 15.280516624450684,
+      "learning_rate": 9.993475857214983e-06,
+      "loss": 6.4028,
+      "step": 430
+    },
+    {
+      "epoch": 0.08532757944291879,
+      "grad_norm": 13.854470252990723,
+      "learning_rate": 9.993276702835682e-06,
+      "loss": 5.9028,
+      "step": 435
+    },
+    {
+      "epoch": 0.08630835621812476,
+      "grad_norm": 19.096981048583984,
+      "learning_rate": 9.99307455635945e-06,
+      "loss": 6.3108,
+      "step": 440
+    },
+    {
+      "epoch": 0.08728913299333071,
+      "grad_norm": 20.702112197875977,
+      "learning_rate": 9.992869417907426e-06,
+      "loss": 6.047,
+      "step": 445
+    },
+    {
+      "epoch": 0.08826990976853669,
+      "grad_norm": 14.787785530090332,
+      "learning_rate": 9.992661287602526e-06,
+      "loss": 5.783,
+      "step": 450
+    },
+    {
+      "epoch": 0.08925068654374264,
+      "grad_norm": 26.985492706298828,
+      "learning_rate": 9.99245016556947e-06,
+      "loss": 6.0479,
+      "step": 455
+    },
+    {
+      "epoch": 0.0902314633189486,
+      "grad_norm": 20.202232360839844,
+      "learning_rate": 9.992236051934769e-06,
+      "loss": 5.8905,
+      "step": 460
+    },
+    {
+      "epoch": 0.09121224009415457,
+      "grad_norm": 28.611963272094727,
+      "learning_rate": 9.992018946826723e-06,
+      "loss": 5.669,
+      "step": 465
+    },
+    {
+      "epoch": 0.09219301686936053,
+      "grad_norm": 15.48359203338623,
+      "learning_rate": 9.99179885037543e-06,
+      "loss": 5.8497,
+      "step": 470
+    },
+    {
+      "epoch": 0.0931737936445665,
+      "grad_norm": 30.19912338256836,
+      "learning_rate": 9.991575762712773e-06,
+      "loss": 5.8428,
+      "step": 475
+    },
+    {
+      "epoch": 0.09415457041977246,
+      "grad_norm": 28.23841094970703,
+      "learning_rate": 9.991349683972435e-06,
+      "loss": 5.8517,
+      "step": 480
+    },
+    {
+      "epoch": 0.09513534719497842,
+      "grad_norm": 36.505924224853516,
+      "learning_rate": 9.991120614289887e-06,
+      "loss": 5.5872,
+      "step": 485
+    },
+    {
+      "epoch": 0.09611612397018439,
+      "grad_norm": 27.388525009155273,
+      "learning_rate": 9.99088855380239e-06,
+      "loss": 5.6681,
+      "step": 490
+    },
+    {
+      "epoch": 0.09709690074539035,
+      "grad_norm": 12.868749618530273,
+      "learning_rate": 9.990653502649008e-06,
+      "loss": 5.5927,
+      "step": 495
+    },
+    {
+      "epoch": 0.09807767752059632,
+      "grad_norm": 21.467880249023438,
+      "learning_rate": 9.990415460970584e-06,
+      "loss": 5.6988,
+      "step": 500
+    },
+    {
+      "epoch": 0.09905845429580228,
+      "grad_norm": 27.02879524230957,
+      "learning_rate": 9.99017442890976e-06,
+      "loss": 5.9116,
+      "step": 505
+    },
+    {
+      "epoch": 0.10003923107100823,
+      "grad_norm": 20.046236038208008,
+      "learning_rate": 9.98993040661097e-06,
+      "loss": 5.8835,
+      "step": 510
+    },
+    {
+      "epoch": 0.1010200078462142,
+      "grad_norm": 18.563119888305664,
+      "learning_rate": 9.989683394220435e-06,
+      "loss": 5.9447,
+      "step": 515
+    },
+    {
+      "epoch": 0.10200078462142016,
+      "grad_norm": 19.473636627197266,
+      "learning_rate": 9.989433391886171e-06,
+      "loss": 5.7367,
+      "step": 520
+    },
+    {
+      "epoch": 0.10298156139662613,
+      "grad_norm": 24.5170955657959,
+      "learning_rate": 9.98918039975799e-06,
+      "loss": 5.5517,
+      "step": 525
+    },
+    {
+      "epoch": 0.10396233817183209,
+      "grad_norm": 23.462825775146484,
+      "learning_rate": 9.988924417987489e-06,
+      "loss": 6.7528,
+      "step": 530
+    },
+    {
+      "epoch": 0.10494311494703805,
+      "grad_norm": 30.959993362426758,
+      "learning_rate": 9.988665446728057e-06,
+      "loss": 5.4933,
+      "step": 535
+    },
+    {
+      "epoch": 0.10592389172224402,
+      "grad_norm": 24.289413452148438,
+      "learning_rate": 9.988403486134877e-06,
+      "loss": 6.1583,
+      "step": 540
+    },
+    {
+      "epoch": 0.10690466849744998,
+      "grad_norm": 14.085652351379395,
+      "learning_rate": 9.988138536364922e-06,
+      "loss": 5.9304,
+      "step": 545
+    },
+    {
+      "epoch": 0.10788544527265595,
+      "grad_norm": 21.895252227783203,
+      "learning_rate": 9.987870597576961e-06,
+      "loss": 5.9257,
+      "step": 550
+    },
+    {
+      "epoch": 0.10886622204786191,
+      "grad_norm": 13.12173080444336,
+      "learning_rate": 9.987599669931543e-06,
+      "loss": 5.5943,
+      "step": 555
+    },
+    {
+      "epoch": 0.10984699882306787,
+      "grad_norm": 21.497610092163086,
+      "learning_rate": 9.987325753591019e-06,
+      "loss": 5.4399,
+      "step": 560
+    },
+    {
+      "epoch": 0.11082777559827384,
+      "grad_norm": 19.55420684814453,
+      "learning_rate": 9.987048848719524e-06,
+      "loss": 5.7695,
+      "step": 565
+    },
+    {
+      "epoch": 0.1118085523734798,
+      "grad_norm": 20.084917068481445,
+      "learning_rate": 9.986768955482988e-06,
+      "loss": 5.6785,
+      "step": 570
+    },
+    {
+      "epoch": 0.11278932914868577,
+      "grad_norm": 22.55635643005371,
+      "learning_rate": 9.986486074049131e-06,
+      "loss": 5.9588,
+      "step": 575
+    },
+    {
+      "epoch": 0.11377010592389172,
+      "grad_norm": 19.915632247924805,
+      "learning_rate": 9.98620020458746e-06,
+      "loss": 5.7189,
+      "step": 580
+    },
+    {
+      "epoch": 0.11475088269909768,
+      "grad_norm": 28.99961280822754,
+      "learning_rate": 9.985911347269277e-06,
+      "loss": 5.7044,
+      "step": 585
+    },
+    {
+      "epoch": 0.11573165947430365,
+      "grad_norm": 15.701618194580078,
+      "learning_rate": 9.985619502267671e-06,
+      "loss": 5.4876,
+      "step": 590
+    },
+    {
+      "epoch": 0.11671243624950961,
+      "grad_norm": 11.117427825927734,
+      "learning_rate": 9.985324669757526e-06,
+      "loss": 5.7046,
+      "step": 595
+    },
+    {
+      "epoch": 0.11769321302471558,
+      "grad_norm": 22.775066375732422,
+      "learning_rate": 9.985026849915508e-06,
+      "loss": 6.3677,
+      "step": 600
+    },
+    {
+      "epoch": 0.11867398979992154,
+      "grad_norm": 20.601396560668945,
+      "learning_rate": 9.984726042920085e-06,
+      "loss": 5.8385,
+      "step": 605
+    },
+    {
+      "epoch": 0.1196547665751275,
+      "grad_norm": 11.978520393371582,
+      "learning_rate": 9.984422248951502e-06,
+      "loss": 5.6919,
+      "step": 610
+    },
+    {
+      "epoch": 0.12063554335033347,
+      "grad_norm": 16.846281051635742,
+      "learning_rate": 9.984115468191803e-06,
+      "loss": 5.3983,
+      "step": 615
+    },
+    {
+      "epoch": 0.12161632012553943,
+      "grad_norm": 15.13311767578125,
+      "learning_rate": 9.983805700824816e-06,
+      "loss": 5.3126,
+      "step": 620
+    },
+    {
+      "epoch": 0.12259709690074538,
+      "grad_norm": 18.353530883789062,
+      "learning_rate": 9.983492947036164e-06,
+      "loss": 5.775,
+      "step": 625
+    },
+    {
+      "epoch": 0.12357787367595136,
+      "grad_norm": 28.933935165405273,
+      "learning_rate": 9.983177207013256e-06,
+      "loss": 5.9725,
+      "step": 630
+    },
+    {
+      "epoch": 0.12455865045115731,
+      "grad_norm": 12.9036226272583,
+      "learning_rate": 9.982858480945295e-06,
+      "loss": 5.6471,
+      "step": 635
+    },
+    {
+      "epoch": 0.12553942722636327,
+      "grad_norm": 20.80467414855957,
+      "learning_rate": 9.982536769023262e-06,
+      "loss": 5.5484,
+      "step": 640
+    },
+    {
+      "epoch": 0.12652020400156924,
+      "grad_norm": 16.309709548950195,
+      "learning_rate": 9.982212071439943e-06,
+      "loss": 5.9737,
+      "step": 645
+    },
+    {
+      "epoch": 0.1275009807767752,
+      "grad_norm": 16.632150650024414,
+      "learning_rate": 9.9818843883899e-06,
+      "loss": 5.4386,
+      "step": 650
+    },
+    {
+      "epoch": 0.12848175755198116,
+      "grad_norm": 13.748077392578125,
+      "learning_rate": 9.981553720069487e-06,
+      "loss": 5.5232,
+      "step": 655
+    },
+    {
+      "epoch": 0.12946253432718713,
+      "grad_norm": 11.935937881469727,
+      "learning_rate": 9.981220066676855e-06,
+      "loss": 5.3825,
+      "step": 660
+    },
+    {
+      "epoch": 0.1304433111023931,
+      "grad_norm": 21.656436920166016,
+      "learning_rate": 9.980883428411934e-06,
+      "loss": 5.8827,
+      "step": 665
+    },
+    {
+      "epoch": 0.13142408787759907,
+      "grad_norm": 14.381396293640137,
+      "learning_rate": 9.980543805476447e-06,
+      "loss": 6.2153,
+      "step": 670
+    },
+    {
+      "epoch": 0.13240486465280502,
+      "grad_norm": 18.785852432250977,
+      "learning_rate": 9.980201198073902e-06,
+      "loss": 5.6124,
+      "step": 675
+    },
+    {
+      "epoch": 0.133385641428011,
+      "grad_norm": 15.40206527709961,
+      "learning_rate": 9.9798556064096e-06,
+      "loss": 5.2739,
+      "step": 680
+    },
+    {
+      "epoch": 0.13436641820321696,
+      "grad_norm": 20.313282012939453,
+      "learning_rate": 9.97950703069063e-06,
+      "loss": 5.6705,
+      "step": 685
+    },
+    {
+      "epoch": 0.1353471949784229,
+      "grad_norm": 13.421894073486328,
+      "learning_rate": 9.979155471125866e-06,
+      "loss": 5.9609,
+      "step": 690
+    },
+    {
+      "epoch": 0.13632797175362887,
+      "grad_norm": 19.68906021118164,
+      "learning_rate": 9.97880092792597e-06,
+      "loss": 5.3471,
+      "step": 695
+    },
+    {
+      "epoch": 0.13730874852883485,
+      "grad_norm": 17.42752456665039,
+      "learning_rate": 9.978443401303392e-06,
+      "loss": 5.6414,
+      "step": 700
+    },
+    {
+      "epoch": 0.1382895253040408,
+      "grad_norm": 19.268634796142578,
+      "learning_rate": 9.978082891472376e-06,
+      "loss": 5.6018,
+      "step": 705
+    },
+    {
+      "epoch": 0.13927030207924676,
+      "grad_norm": 14.104660987854004,
+      "learning_rate": 9.977719398648945e-06,
+      "loss": 5.4409,
+      "step": 710
+    },
+    {
+      "epoch": 0.14025107885445273,
+      "grad_norm": 14.374234199523926,
+      "learning_rate": 9.977352923050913e-06,
+      "loss": 5.8209,
+      "step": 715
+    },
+    {
+      "epoch": 0.1412318556296587,
+      "grad_norm": 23.02276611328125,
+      "learning_rate": 9.976983464897882e-06,
+      "loss": 5.7185,
+      "step": 720
+    },
+    {
+      "epoch": 0.14221263240486465,
+      "grad_norm": 16.408750534057617,
+      "learning_rate": 9.976611024411241e-06,
+      "loss": 5.7395,
+      "step": 725
+    },
+    {
+      "epoch": 0.14319340918007062,
+      "grad_norm": 15.385457038879395,
+      "learning_rate": 9.976235601814163e-06,
+      "loss": 5.8248,
+      "step": 730
+    },
+    {
+      "epoch": 0.1441741859552766,
+      "grad_norm": 13.1869478225708,
+      "learning_rate": 9.975857197331617e-06,
+      "loss": 5.7622,
+      "step": 735
+    },
+    {
+      "epoch": 0.14515496273048253,
+      "grad_norm": 18.29127311706543,
+      "learning_rate": 9.975475811190346e-06,
+      "loss": 5.6788,
+      "step": 740
+    },
+    {
+      "epoch": 0.1461357395056885,
+      "grad_norm": 18.088855743408203,
+      "learning_rate": 9.975091443618889e-06,
+      "loss": 5.4656,
+      "step": 745
+    },
+    {
+      "epoch": 0.14711651628089448,
+      "grad_norm": 30.90943145751953,
+      "learning_rate": 9.974704094847568e-06,
+      "loss": 5.525,
+      "step": 750
+    },
+    {
+      "epoch": 0.14809729305610042,
+      "grad_norm": 16.267723083496094,
+      "learning_rate": 9.974313765108492e-06,
+      "loss": 5.3356,
+      "step": 755
+    },
+    {
+      "epoch": 0.1490780698313064,
+      "grad_norm": 15.598844528198242,
+      "learning_rate": 9.973920454635559e-06,
+      "loss": 5.5963,
+      "step": 760
+    },
+    {
+      "epoch": 0.15005884660651236,
+      "grad_norm": 27.659122467041016,
+      "learning_rate": 9.973524163664447e-06,
+      "loss": 5.4735,
+      "step": 765
+    },
+    {
+      "epoch": 0.1510396233817183,
+      "grad_norm": 30.47996711730957,
+      "learning_rate": 9.973124892432626e-06,
+      "loss": 5.6989,
+      "step": 770
+    },
+    {
+      "epoch": 0.15202040015692428,
+      "grad_norm": 18.81101417541504,
+      "learning_rate": 9.972722641179347e-06,
+      "loss": 5.5134,
+      "step": 775
+    },
+    {
+      "epoch": 0.15300117693213025,
+      "grad_norm": 30.00667381286621,
+      "learning_rate": 9.972317410145651e-06,
+      "loss": 5.5737,
+      "step": 780
+    },
+    {
+      "epoch": 0.15398195370733622,
+      "grad_norm": 16.57672119140625,
+      "learning_rate": 9.97190919957436e-06,
+      "loss": 5.369,
+      "step": 785
+    },
+    {
+      "epoch": 0.15496273048254217,
+      "grad_norm": 25.58556365966797,
+      "learning_rate": 9.971498009710088e-06,
+      "loss": 5.7396,
+      "step": 790
+    },
+    {
+      "epoch": 0.15594350725774814,
+      "grad_norm": 30.630939483642578,
+      "learning_rate": 9.971083840799229e-06,
+      "loss": 5.415,
+      "step": 795
+    },
+    {
+      "epoch": 0.1569242840329541,
+      "grad_norm": 12.743330001831055,
+      "learning_rate": 9.97066669308996e-06,
+      "loss": 5.5816,
+      "step": 800
+    },
+    {
+      "epoch": 0.15790506080816005,
+      "grad_norm": 15.443756103515625,
+      "learning_rate": 9.970246566832252e-06,
+      "loss": 5.4506,
+      "step": 805
+    },
+    {
+      "epoch": 0.15888583758336602,
+      "grad_norm": 22.54521942138672,
+      "learning_rate": 9.96982346227785e-06,
+      "loss": 5.6406,
+      "step": 810
+    },
+    {
+      "epoch": 0.159866614358572,
+      "grad_norm": 16.574411392211914,
+      "learning_rate": 9.969397379680293e-06,
+      "loss": 5.4527,
+      "step": 815
+    },
+    {
+      "epoch": 0.16084739113377794,
+      "grad_norm": 24.11166000366211,
+      "learning_rate": 9.968968319294897e-06,
+      "loss": 5.4452,
+      "step": 820
+    },
+    {
+      "epoch": 0.1618281679089839,
+      "grad_norm": 19.735063552856445,
+      "learning_rate": 9.96853628137877e-06,
+      "loss": 5.3691,
+      "step": 825
+    },
+    {
+      "epoch": 0.16280894468418988,
+      "grad_norm": 17.768238067626953,
+      "learning_rate": 9.968101266190795e-06,
+      "loss": 5.3472,
+      "step": 830
+    },
+    {
+      "epoch": 0.16378972145939585,
+      "grad_norm": 23.892215728759766,
+      "learning_rate": 9.967663273991646e-06,
+      "loss": 5.707,
+      "step": 835
+    },
+    {
+      "epoch": 0.1647704982346018,
+      "grad_norm": 31.293302536010742,
+      "learning_rate": 9.96722230504378e-06,
+      "loss": 5.8675,
+      "step": 840
+    },
+    {
+      "epoch": 0.16575127500980777,
+      "grad_norm": 35.70466995239258,
+      "learning_rate": 9.966778359611435e-06,
+      "loss": 6.0157,
+      "step": 845
+    },
+    {
+      "epoch": 0.16673205178501374,
+      "grad_norm": 20.035058975219727,
+      "learning_rate": 9.966331437960636e-06,
+      "loss": 5.6011,
+      "step": 850
+    },
+    {
+      "epoch": 0.16771282856021968,
+      "grad_norm": 18.689212799072266,
+      "learning_rate": 9.96588154035919e-06,
+      "loss": 5.2369,
+      "step": 855
+    },
+    {
+      "epoch": 0.16869360533542566,
+      "grad_norm": 71.4951171875,
+      "learning_rate": 9.965428667076687e-06,
+      "loss": 6.5524,
+      "step": 860
+    },
+    {
+      "epoch": 0.16967438211063163,
+      "grad_norm": 19.083498001098633,
+      "learning_rate": 9.964972818384496e-06,
+      "loss": 5.6899,
+      "step": 865
+    },
+    {
+      "epoch": 0.17065515888583757,
+      "grad_norm": 26.55590057373047,
+      "learning_rate": 9.964513994555778e-06,
+      "loss": 5.4484,
+      "step": 870
+    },
+    {
+      "epoch": 0.17163593566104354,
+      "grad_norm": 18.4299259185791,
+      "learning_rate": 9.964052195865468e-06,
+      "loss": 5.6257,
+      "step": 875
+    },
+    {
+      "epoch": 0.17261671243624951,
+      "grad_norm": 16.13317108154297,
+      "learning_rate": 9.96358742259029e-06,
+      "loss": 5.5983,
+      "step": 880
+    },
+    {
+      "epoch": 0.17359748921145549,
+      "grad_norm": 16.984779357910156,
+      "learning_rate": 9.963119675008748e-06,
+      "loss": 5.48,
+      "step": 885
+    },
+    {
+      "epoch": 0.17457826598666143,
+      "grad_norm": 22.902658462524414,
+      "learning_rate": 9.962648953401125e-06,
+      "loss": 5.9483,
+      "step": 890
+    },
+    {
+      "epoch": 0.1755590427618674,
+      "grad_norm": 24.295028686523438,
+      "learning_rate": 9.962175258049493e-06,
+      "loss": 5.7974,
+      "step": 895
+    },
+    {
+      "epoch": 0.17653981953707337,
+      "grad_norm": 20.873647689819336,
+      "learning_rate": 9.9616985892377e-06,
+      "loss": 5.5721,
+      "step": 900
+    },
+    {
+      "epoch": 0.17752059631227932,
+      "grad_norm": 15.696707725524902,
+      "learning_rate": 9.961218947251378e-06,
+      "loss": 5.4619,
+      "step": 905
+    },
+    {
+      "epoch": 0.1785013730874853,
+      "grad_norm": 20.687946319580078,
+      "learning_rate": 9.96073633237794e-06,
+      "loss": 5.8472,
+      "step": 910
+    },
+    {
+      "epoch": 0.17948214986269126,
+      "grad_norm": 14.133719444274902,
+      "learning_rate": 9.960250744906583e-06,
+      "loss": 5.417,
+      "step": 915
+    },
+    {
+      "epoch": 0.1804629266378972,
+      "grad_norm": 15.220897674560547,
+      "learning_rate": 9.959762185128283e-06,
+      "loss": 5.3647,
+      "step": 920
+    },
+    {
+      "epoch": 0.18144370341310317,
+      "grad_norm": 30.21053695678711,
+      "learning_rate": 9.959270653335795e-06,
+      "loss": 5.7774,
+      "step": 925
+    },
+    {
+      "epoch": 0.18242448018830915,
+      "grad_norm": 21.00077247619629,
+      "learning_rate": 9.958776149823658e-06,
+      "loss": 5.4467,
+      "step": 930
+    },
+    {
+      "epoch": 0.18340525696351512,
+      "grad_norm": 25.028913497924805,
+      "learning_rate": 9.958278674888194e-06,
+      "loss": 5.5723,
+      "step": 935
+    },
+    {
+      "epoch": 0.18438603373872106,
+      "grad_norm": 12.45209789276123,
+      "learning_rate": 9.957778228827499e-06,
+      "loss": 5.5205,
+      "step": 940
+    },
+    {
+      "epoch": 0.18536681051392703,
+      "grad_norm": 12.58310604095459,
+      "learning_rate": 9.957274811941452e-06,
+      "loss": 5.285,
+      "step": 945
+    },
+    {
+      "epoch": 0.186347587289133,
+      "grad_norm": 15.49960708618164,
+      "learning_rate": 9.956768424531717e-06,
+      "loss": 5.5491,
+      "step": 950
+    },
+    {
+      "epoch": 0.18732836406433895,
+      "grad_norm": 14.371564865112305,
+      "learning_rate": 9.956259066901733e-06,
+      "loss": 5.6843,
+      "step": 955
+    },
+    {
+      "epoch": 0.18830914083954492,
+      "grad_norm": 20.336835861206055,
+      "learning_rate": 9.955746739356716e-06,
+      "loss": 5.7197,
+      "step": 960
+    },
+    {
+      "epoch": 0.1892899176147509,
+      "grad_norm": 14.215054512023926,
+      "learning_rate": 9.95523144220367e-06,
+      "loss": 5.7196,
+      "step": 965
+    },
+    {
+      "epoch": 0.19027069438995683,
+      "grad_norm": 20.546175003051758,
+      "learning_rate": 9.954713175751373e-06,
+      "loss": 5.407,
+      "step": 970
+    },
+    {
+      "epoch": 0.1912514711651628,
+      "grad_norm": 14.670819282531738,
+      "learning_rate": 9.954191940310381e-06,
+      "loss": 5.6484,
+      "step": 975
+    },
+    {
+      "epoch": 0.19223224794036878,
+      "grad_norm": 46.397247314453125,
+      "learning_rate": 9.953667736193034e-06,
+      "loss": 5.6878,
+      "step": 980
+    },
+    {
+      "epoch": 0.19321302471557472,
+      "grad_norm": 10.847311973571777,
+      "learning_rate": 9.953140563713448e-06,
+      "loss": 5.4969,
+      "step": 985
+    },
+    {
+      "epoch": 0.1941938014907807,
+      "grad_norm": 17.326196670532227,
+      "learning_rate": 9.952610423187516e-06,
+      "loss": 5.6137,
+      "step": 990
+    },
+    {
+      "epoch": 0.19517457826598666,
+      "grad_norm": 18.91810417175293,
+      "learning_rate": 9.952077314932916e-06,
+      "loss": 5.3132,
+      "step": 995
+    },
+    {
+      "epoch": 0.19615535504119264,
+      "grad_norm": 15.190394401550293,
+      "learning_rate": 9.951541239269093e-06,
+      "loss": 5.4908,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19713613181639858,
+      "grad_norm": 16.841787338256836,
+      "learning_rate": 9.951002196517284e-06,
+      "loss": 5.5744,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19811690859160455,
+      "grad_norm": 17.683765411376953,
+      "learning_rate": 9.950460187000492e-06,
+      "loss": 5.2091,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19909768536681052,
+      "grad_norm": 17.12097930908203,
+      "learning_rate": 9.949915211043504e-06,
+      "loss": 5.4289,
+      "step": 1015
+    },
+    {
+      "epoch": 0.20007846214201647,
+      "grad_norm": 22.892080307006836,
+      "learning_rate": 9.949367268972885e-06,
+      "loss": 5.2564,
+      "step": 1020
+    },
+    {
+      "epoch": 0.20105923891722244,
+      "grad_norm": 15.766369819641113,
+      "learning_rate": 9.948816361116973e-06,
+      "loss": 5.3534,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2020400156924284,
+      "grad_norm": 24.48428726196289,
+      "learning_rate": 9.94826248780589e-06,
+      "loss": 6.1259,
+      "step": 1030
+    },
+    {
+      "epoch": 0.20302079246763435,
+      "grad_norm": 17.7995548248291,
+      "learning_rate": 9.947705649371526e-06,
+      "loss": 5.585,
+      "step": 1035
+    },
+    {
+      "epoch": 0.20400156924284032,
+      "grad_norm": 24.083513259887695,
+      "learning_rate": 9.947145846147555e-06,
+      "loss": 5.7126,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2049823460180463,
+      "grad_norm": 12.388947486877441,
+      "learning_rate": 9.946583078469426e-06,
+      "loss": 5.5251,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20596312279325227,
+      "grad_norm": 12.7921724319458,
+      "learning_rate": 9.946017346674362e-06,
+      "loss": 5.4943,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2069438995684582,
+      "grad_norm": 28.395301818847656,
+      "learning_rate": 9.945448651101365e-06,
+      "loss": 5.7806,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20792467634366418,
+      "grad_norm": 35.039634704589844,
+      "learning_rate": 9.944876992091208e-06,
+      "loss": 5.3988,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20890545311887015,
+      "grad_norm": 12.610820770263672,
+      "learning_rate": 9.944302369986447e-06,
+      "loss": 5.5946,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2098862298940761,
+      "grad_norm": 13.666259765625,
+      "learning_rate": 9.943724785131412e-06,
+      "loss": 5.5101,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21086700666928207,
+      "grad_norm": 19.81101417541504,
+      "learning_rate": 9.943144237872202e-06,
+      "loss": 5.4725,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21184778344448804,
+      "grad_norm": 17.405027389526367,
+      "learning_rate": 9.942560728556696e-06,
+      "loss": 5.4333,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21282856021969399,
+      "grad_norm": 18.48592758178711,
+      "learning_rate": 9.94197425753455e-06,
+      "loss": 5.4893,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21380933699489996,
+      "grad_norm": 26.090227127075195,
+      "learning_rate": 9.94138482515719e-06,
+      "loss": 6.0036,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21479011377010593,
+      "grad_norm": 18.069581985473633,
+      "learning_rate": 9.94079243177782e-06,
+      "loss": 5.3082,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2157708905453119,
+      "grad_norm": 10.112467765808105,
+      "learning_rate": 9.940197077751416e-06,
+      "loss": 5.5434,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21675166732051784,
+      "grad_norm": 20.426862716674805,
+      "learning_rate": 9.93959876343473e-06,
+      "loss": 5.5726,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21773244409572381,
+      "grad_norm": 14.866218566894531,
+      "learning_rate": 9.938997489186287e-06,
+      "loss": 5.7631,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2187132208709298,
+      "grad_norm": 13.593292236328125,
+      "learning_rate": 9.938393255366383e-06,
+      "loss": 5.1905,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21969399764613573,
+      "grad_norm": 15.151540756225586,
+      "learning_rate": 9.937786062337095e-06,
+      "loss": 5.6169,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2206747744213417,
+      "grad_norm": 16.6584529876709,
+      "learning_rate": 9.937175910462264e-06,
+      "loss": 5.1824,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22165555119654767,
+      "grad_norm": 36.43128967285156,
+      "learning_rate": 9.936562800107512e-06,
+      "loss": 5.8551,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22263632797175362,
+      "grad_norm": 20.047487258911133,
+      "learning_rate": 9.935946731640226e-06,
+      "loss": 5.3579,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2236171047469596,
+      "grad_norm": 9.661016464233398,
+      "learning_rate": 9.935327705429572e-06,
+      "loss": 5.4897,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22459788152216556,
+      "grad_norm": 17.872982025146484,
+      "learning_rate": 9.934705721846487e-06,
+      "loss": 5.4356,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22557865829737153,
+      "grad_norm": 14.063901901245117,
+      "learning_rate": 9.934080781263677e-06,
+      "loss": 4.9382,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22655943507257748,
+      "grad_norm": 19.134902954101562,
+      "learning_rate": 9.933452884055625e-06,
+      "loss": 5.4342,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22754021184778345,
+      "grad_norm": 9.560270309448242,
+      "learning_rate": 9.932822030598578e-06,
+      "loss": 5.466,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22852098862298942,
+      "grad_norm": 21.985918045043945,
+      "learning_rate": 9.932188221270564e-06,
+      "loss": 5.2657,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22950176539819536,
+      "grad_norm": 15.274785995483398,
+      "learning_rate": 9.931551456451377e-06,
+      "loss": 5.4104,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23048254217340133,
+      "grad_norm": 28.460533142089844,
+      "learning_rate": 9.93091173652258e-06,
+      "loss": 5.2061,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2314633189486073,
+      "grad_norm": 8.903681755065918,
+      "learning_rate": 9.93026906186751e-06,
+      "loss": 5.3667,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23244409572381325,
+      "grad_norm": 31.3291015625,
+      "learning_rate": 9.929623432871277e-06,
+      "loss": 5.7421,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23342487249901922,
+      "grad_norm": 12.238224983215332,
+      "learning_rate": 9.928974849920752e-06,
+      "loss": 5.2949,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2344056492742252,
+      "grad_norm": 16.755048751831055,
+      "learning_rate": 9.928323313404587e-06,
+      "loss": 5.24,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23538642604943116,
+      "grad_norm": 17.94761848449707,
+      "learning_rate": 9.927668823713197e-06,
+      "loss": 5.5436,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2363672028246371,
+      "grad_norm": 14.653851509094238,
+      "learning_rate": 9.927011381238769e-06,
+      "loss": 5.1157,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23734797959984308,
+      "grad_norm": 23.329797744750977,
+      "learning_rate": 9.926350986375261e-06,
+      "loss": 5.6824,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23832875637504905,
+      "grad_norm": 12.664395332336426,
+      "learning_rate": 9.925687639518395e-06,
+      "loss": 5.4402,
+      "step": 1215
+    },
+    {
+      "epoch": 0.239309533150255,
+      "grad_norm": 23.01675796508789,
+      "learning_rate": 9.925021341065668e-06,
+      "loss": 5.5041,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24029030992546097,
+      "grad_norm": 23.604764938354492,
+      "learning_rate": 9.924352091416342e-06,
+      "loss": 5.3516,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24127108670066694,
+      "grad_norm": 10.142374992370605,
+      "learning_rate": 9.923679890971447e-06,
+      "loss": 5.0815,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24225186347587288,
+      "grad_norm": 46.93704605102539,
+      "learning_rate": 9.923004740133783e-06,
+      "loss": 5.5097,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24323264025107885,
+      "grad_norm": 18.356081008911133,
+      "learning_rate": 9.922326639307918e-06,
+      "loss": 5.367,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24421341702628482,
+      "grad_norm": 17.588830947875977,
+      "learning_rate": 9.921645588900187e-06,
+      "loss": 5.6602,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24519419380149077,
+      "grad_norm": 17.02576446533203,
+      "learning_rate": 9.92096158931869e-06,
+      "loss": 5.2146,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24617497057669674,
+      "grad_norm": 20.34739875793457,
+      "learning_rate": 9.9202746409733e-06,
+      "loss": 5.3317,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2471557473519027,
+      "grad_norm": 11.448872566223145,
+      "learning_rate": 9.919584744275652e-06,
+      "loss": 5.5831,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24813652412710868,
+      "grad_norm": 14.755648612976074,
+      "learning_rate": 9.918891899639151e-06,
+      "loss": 5.3672,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24911730090231463,
+      "grad_norm": 10.774361610412598,
+      "learning_rate": 9.918196107478966e-06,
+      "loss": 4.9776,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25009807767752057,
+      "grad_norm": 13.798211097717285,
+      "learning_rate": 9.917497368212032e-06,
+      "loss": 5.3684,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25009807767752057,
+      "eval_loss": 5.386422157287598,
+      "eval_runtime": 8.0843,
+      "eval_samples_per_second": 25.852,
+      "eval_steps_per_second": 12.988,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25107885445272654,
+      "grad_norm": 14.56896686553955,
+      "learning_rate": 9.916795682257052e-06,
+      "loss": 5.5152,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2520596312279325,
+      "grad_norm": 20.55042839050293,
+      "learning_rate": 9.916091050034496e-06,
+      "loss": 4.9962,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2530404080031385,
+      "grad_norm": 16.28716278076172,
+      "learning_rate": 9.915383471966594e-06,
+      "loss": 5.566,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25402118477834446,
+      "grad_norm": 44.03998565673828,
+      "learning_rate": 9.914672948477347e-06,
+      "loss": 5.226,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2550019615535504,
+      "grad_norm": 10.094902992248535,
+      "learning_rate": 9.913959479992517e-06,
+      "loss": 5.1731,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2559827383287564,
+      "grad_norm": 12.320008277893066,
+      "learning_rate": 9.913243066939631e-06,
+      "loss": 5.8408,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2569635151039623,
+      "grad_norm": 12.670159339904785,
+      "learning_rate": 9.912523709747985e-06,
+      "loss": 5.0899,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2579442918791683,
+      "grad_norm": 10.27662181854248,
+      "learning_rate": 9.911801408848634e-06,
+      "loss": 5.1799,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25892506865437426,
+      "grad_norm": 17.295764923095703,
+      "learning_rate": 9.911076164674401e-06,
+      "loss": 5.104,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25990584542958023,
+      "grad_norm": 21.576318740844727,
+      "learning_rate": 9.910347977659867e-06,
+      "loss": 5.6279,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2608866222047862,
+      "grad_norm": 16.189517974853516,
+      "learning_rate": 9.909616848241382e-06,
+      "loss": 5.2356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26186739897999217,
+      "grad_norm": 15.81222152709961,
+      "learning_rate": 9.908882776857057e-06,
+      "loss": 5.5731,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26284817575519814,
+      "grad_norm": 31.023778915405273,
+      "learning_rate": 9.908145763946766e-06,
+      "loss": 5.2728,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26382895253040406,
+      "grad_norm": 16.400127410888672,
+      "learning_rate": 9.907405809952147e-06,
+      "loss": 5.775,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26480972930561003,
+      "grad_norm": 30.60183334350586,
+      "learning_rate": 9.906662915316595e-06,
+      "loss": 5.4081,
+      "step": 1350
+    },
+    {
+      "epoch": 0.265790506080816,
+      "grad_norm": 12.121785163879395,
+      "learning_rate": 9.905917080485275e-06,
+      "loss": 5.498,
+      "step": 1355
+    },
+    {
+      "epoch": 0.266771282856022,
+      "grad_norm": 14.963290214538574,
+      "learning_rate": 9.905168305905109e-06,
+      "loss": 5.0885,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26775205963122795,
+      "grad_norm": 19.471620559692383,
+      "learning_rate": 9.90441659202478e-06,
+      "loss": 5.2138,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2687328364064339,
+      "grad_norm": 15.10096549987793,
+      "learning_rate": 9.903661939294737e-06,
+      "loss": 5.4012,
+      "step": 1370
+    },
+    {
+      "epoch": 0.26971361318163983,
+      "grad_norm": 14.973530769348145,
+      "learning_rate": 9.902904348167185e-06,
+      "loss": 5.329,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2706943899568458,
+      "grad_norm": 12.646509170532227,
+      "learning_rate": 9.90214381909609e-06,
+      "loss": 5.2064,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2716751667320518,
+      "grad_norm": 10.052018165588379,
+      "learning_rate": 9.901380352537183e-06,
+      "loss": 5.2056,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27265594350725775,
+      "grad_norm": 16.167695999145508,
+      "learning_rate": 9.90061394894795e-06,
+      "loss": 5.4832,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2736367202824637,
+      "grad_norm": 19.09334945678711,
+      "learning_rate": 9.899844608787641e-06,
+      "loss": 5.2711,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2746174970576697,
+      "grad_norm": 17.699079513549805,
+      "learning_rate": 9.899072332517263e-06,
+      "loss": 5.3082,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27559827383287566,
+      "grad_norm": 11.575671195983887,
+      "learning_rate": 9.898297120599585e-06,
+      "loss": 5.1952,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2765790506080816,
+      "grad_norm": 14.866594314575195,
+      "learning_rate": 9.897518973499131e-06,
+      "loss": 5.3446,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27755982738328755,
+      "grad_norm": 26.142955780029297,
+      "learning_rate": 9.89673789168219e-06,
+      "loss": 5.2227,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2785406041584935,
+      "grad_norm": 34.1876220703125,
+      "learning_rate": 9.8959538756168e-06,
+      "loss": 5.5915,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2795213809336995,
+      "grad_norm": 17.60607147216797,
+      "learning_rate": 9.89516692577277e-06,
+      "loss": 5.6551,
+      "step": 1425
+    },
+    {
+      "epoch": 0.28050215770890546,
+      "grad_norm": 15.993223190307617,
+      "learning_rate": 9.894377042621654e-06,
+      "loss": 5.2441,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28148293448411144,
+      "grad_norm": 16.62636375427246,
+      "learning_rate": 9.893584226636773e-06,
+      "loss": 5.2074,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2824637112593174,
+      "grad_norm": 23.400930404663086,
+      "learning_rate": 9.892788478293203e-06,
+      "loss": 5.3432,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2834444880345233,
+      "grad_norm": 11.958523750305176,
+      "learning_rate": 9.891989798067774e-06,
+      "loss": 5.1582,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2844252648097293,
+      "grad_norm": 13.033811569213867,
+      "learning_rate": 9.891188186439077e-06,
+      "loss": 5.2325,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28540604158493527,
+      "grad_norm": 13.869037628173828,
+      "learning_rate": 9.890383643887458e-06,
+      "loss": 5.15,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28638681836014124,
+      "grad_norm": 16.992834091186523,
+      "learning_rate": 9.889576170895016e-06,
+      "loss": 5.3062,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2873675951353472,
+      "grad_norm": 15.610152244567871,
+      "learning_rate": 9.888765767945613e-06,
+      "loss": 5.4888,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2883483719105532,
+      "grad_norm": 24.985816955566406,
+      "learning_rate": 9.887952435524863e-06,
+      "loss": 5.5596,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2893291486857591,
+      "grad_norm": 17.40192222595215,
+      "learning_rate": 9.887136174120132e-06,
+      "loss": 5.1121,
+      "step": 1475
+    },
+    {
+      "epoch": 0.29030992546096507,
+      "grad_norm": 15.199414253234863,
+      "learning_rate": 9.886316984220546e-06,
+      "loss": 4.9102,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29129070223617104,
+      "grad_norm": 17.887176513671875,
+      "learning_rate": 9.885494866316985e-06,
+      "loss": 5.2968,
+      "step": 1485
+    },
+    {
+      "epoch": 0.292271479011377,
+      "grad_norm": 9.527190208435059,
+      "learning_rate": 9.884669820902081e-06,
+      "loss": 5.0361,
+      "step": 1490
+    },
+    {
+      "epoch": 0.293252255786583,
+      "grad_norm": 11.876401901245117,
+      "learning_rate": 9.883841848470222e-06,
+      "loss": 5.2537,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29423303256178895,
+      "grad_norm": 16.37000274658203,
+      "learning_rate": 9.883010949517553e-06,
+      "loss": 5.2722,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2952138093369949,
+      "grad_norm": 15.113862991333008,
+      "learning_rate": 9.882177124541965e-06,
+      "loss": 5.3566,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29619458611220084,
+      "grad_norm": 14.421209335327148,
+      "learning_rate": 9.881340374043111e-06,
+      "loss": 5.1384,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2971753628874068,
+      "grad_norm": 24.657230377197266,
+      "learning_rate": 9.880500698522391e-06,
+      "loss": 5.6017,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2981561396626128,
+      "grad_norm": 12.178624153137207,
+      "learning_rate": 9.879658098482959e-06,
+      "loss": 5.016,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29913691643781876,
+      "grad_norm": 23.32276153564453,
+      "learning_rate": 9.878812574429722e-06,
+      "loss": 5.704,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3001176932130247,
+      "grad_norm": 21.334623336791992,
+      "learning_rate": 9.877964126869341e-06,
+      "loss": 4.914,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3010984699882307,
+      "grad_norm": 16.11152458190918,
+      "learning_rate": 9.877112756310225e-06,
+      "loss": 5.5664,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3020792467634366,
+      "grad_norm": 22.593204498291016,
+      "learning_rate": 9.87625846326254e-06,
+      "loss": 5.2208,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3030600235386426,
+      "grad_norm": 23.615087509155273,
+      "learning_rate": 9.875401248238197e-06,
+      "loss": 5.3042,
+      "step": 1545
+    },
+    {
+      "epoch": 0.30404080031384856,
+      "grad_norm": 14.477107048034668,
+      "learning_rate": 9.874541111750861e-06,
+      "loss": 5.3559,
+      "step": 1550
+    },
+    {
+      "epoch": 0.30502157708905453,
+      "grad_norm": 14.735102653503418,
+      "learning_rate": 9.873678054315949e-06,
+      "loss": 5.4065,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3060023538642605,
+      "grad_norm": 11.800651550292969,
+      "learning_rate": 9.872812076450625e-06,
+      "loss": 5.4283,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3069831306394665,
+      "grad_norm": 24.840158462524414,
+      "learning_rate": 9.871943178673806e-06,
+      "loss": 5.298,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30796390741467244,
+      "grad_norm": 12.651166915893555,
+      "learning_rate": 9.871071361506156e-06,
+      "loss": 5.1233,
+      "step": 1570
+    },
+    {
+      "epoch": 0.30894468418987836,
+      "grad_norm": 30.376554489135742,
+      "learning_rate": 9.87019662547009e-06,
+      "loss": 5.5695,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30992546096508433,
+      "grad_norm": 15.623506546020508,
+      "learning_rate": 9.869318971089774e-06,
+      "loss": 5.3929,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3109062377402903,
+      "grad_norm": 31.46923828125,
+      "learning_rate": 9.868438398891118e-06,
+      "loss": 5.3255,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3118870145154963,
+      "grad_norm": 15.704526901245117,
+      "learning_rate": 9.867554909401785e-06,
+      "loss": 5.1772,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31286779129070225,
+      "grad_norm": 18.602909088134766,
+      "learning_rate": 9.866668503151182e-06,
+      "loss": 4.9734,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3138485680659082,
+      "grad_norm": 20.587440490722656,
+      "learning_rate": 9.865779180670468e-06,
+      "loss": 5.0147,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3148293448411142,
+      "grad_norm": 24.15857696533203,
+      "learning_rate": 9.864886942492543e-06,
+      "loss": 5.2671,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3158101216163201,
+      "grad_norm": 10.229106903076172,
+      "learning_rate": 9.863991789152065e-06,
+      "loss": 5.3938,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3167908983915261,
+      "grad_norm": 19.83039665222168,
+      "learning_rate": 9.86309372118543e-06,
+      "loss": 4.9836,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31777167516673205,
+      "grad_norm": 24.620084762573242,
+      "learning_rate": 9.86219273913078e-06,
+      "loss": 5.1878,
+      "step": 1620
+    },
+    {
+      "epoch": 0.318752451941938,
+      "grad_norm": 10.462682723999023,
+      "learning_rate": 9.86128884352801e-06,
+      "loss": 5.8849,
+      "step": 1625
+    },
+    {
+      "epoch": 0.319733228717144,
+      "grad_norm": 16.412769317626953,
+      "learning_rate": 9.860382034918754e-06,
+      "loss": 5.3453,
+      "step": 1630
+    },
+    {
+      "epoch": 0.32071400549234996,
+      "grad_norm": 27.44397735595703,
+      "learning_rate": 9.859472313846396e-06,
+      "loss": 5.226,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3216947822675559,
+      "grad_norm": 22.453519821166992,
+      "learning_rate": 9.858559680856064e-06,
+      "loss": 5.4033,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32267555904276185,
+      "grad_norm": 15.075695991516113,
+      "learning_rate": 9.857644136494629e-06,
+      "loss": 5.2115,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3236563358179678,
+      "grad_norm": 12.842907905578613,
+      "learning_rate": 9.85672568131071e-06,
+      "loss": 5.0739,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3246371125931738,
+      "grad_norm": 15.345038414001465,
+      "learning_rate": 9.855804315854667e-06,
+      "loss": 5.1001,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32561788936837976,
+      "grad_norm": 12.392874717712402,
+      "learning_rate": 9.854880040678608e-06,
+      "loss": 4.9565,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32659866614358574,
+      "grad_norm": 14.970841407775879,
+      "learning_rate": 9.853952856336377e-06,
+      "loss": 5.2334,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3275794429187917,
+      "grad_norm": 23.218185424804688,
+      "learning_rate": 9.853022763383572e-06,
+      "loss": 5.2103,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3285602196939976,
+      "grad_norm": 16.600784301757812,
+      "learning_rate": 9.852089762377525e-06,
+      "loss": 5.359,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3295409964692036,
+      "grad_norm": 31.103673934936523,
+      "learning_rate": 9.851153853877314e-06,
+      "loss": 5.3389,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33052177324440957,
+      "grad_norm": 19.520828247070312,
+      "learning_rate": 9.850215038443756e-06,
+      "loss": 5.2418,
+      "step": 1685
+    },
+    {
+      "epoch": 0.33150255001961554,
+      "grad_norm": 31.281417846679688,
+      "learning_rate": 9.849273316639418e-06,
+      "loss": 5.5463,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3324833267948215,
+      "grad_norm": 14.541434288024902,
+      "learning_rate": 9.8483286890286e-06,
+      "loss": 5.1112,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3334641035700275,
+      "grad_norm": 20.683828353881836,
+      "learning_rate": 9.847381156177349e-06,
+      "loss": 5.7917,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3344448803452334,
+      "grad_norm": 11.94363784790039,
+      "learning_rate": 9.846430718653449e-06,
+      "loss": 5.3054,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33542565712043937,
+      "grad_norm": 14.712422370910645,
+      "learning_rate": 9.845477377026426e-06,
+      "loss": 5.3564,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33640643389564534,
+      "grad_norm": 20.38669204711914,
+      "learning_rate": 9.844521131867546e-06,
+      "loss": 5.4337,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3373872106708513,
+      "grad_norm": 16.519319534301758,
+      "learning_rate": 9.843561983749816e-06,
+      "loss": 5.4502,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3383679874460573,
+      "grad_norm": 12.27083683013916,
+      "learning_rate": 9.84259993324798e-06,
+      "loss": 5.3244,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33934876422126325,
+      "grad_norm": 15.599812507629395,
+      "learning_rate": 9.841634980938526e-06,
+      "loss": 5.4292,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3403295409964692,
+      "grad_norm": 11.654434204101562,
+      "learning_rate": 9.840667127399675e-06,
+      "loss": 5.1737,
+      "step": 1735
+    },
+    {
+      "epoch": 0.34131031777167514,
+      "grad_norm": 22.475311279296875,
+      "learning_rate": 9.83969637321139e-06,
+      "loss": 5.4079,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3422910945468811,
+      "grad_norm": 10.999152183532715,
+      "learning_rate": 9.838722718955372e-06,
+      "loss": 5.4605,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3432718713220871,
+      "grad_norm": 19.069194793701172,
+      "learning_rate": 9.837746165215057e-06,
+      "loss": 5.6759,
+      "step": 1750
+    },
+    {
+      "epoch": 0.34425264809729306,
+      "grad_norm": 15.041163444519043,
+      "learning_rate": 9.836766712575622e-06,
+      "loss": 5.8595,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34523342487249903,
+      "grad_norm": 12.235005378723145,
+      "learning_rate": 9.83578436162398e-06,
+      "loss": 5.4856,
+      "step": 1760
+    },
+    {
+      "epoch": 0.346214201647705,
+      "grad_norm": 17.395387649536133,
+      "learning_rate": 9.83479911294878e-06,
+      "loss": 5.2461,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34719497842291097,
+      "grad_norm": 11.41357135772705,
+      "learning_rate": 9.833810967140408e-06,
+      "loss": 5.3194,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3481757551981169,
+      "grad_norm": 11.756146430969238,
+      "learning_rate": 9.832819924790986e-06,
+      "loss": 5.3896,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34915653197332286,
+      "grad_norm": 19.872554779052734,
+      "learning_rate": 9.83182598649437e-06,
+      "loss": 5.355,
+      "step": 1780
+    },
+    {
+      "epoch": 0.35013730874852883,
+      "grad_norm": 24.415315628051758,
+      "learning_rate": 9.830829152846154e-06,
+      "loss": 5.2137,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3511180855237348,
+      "grad_norm": 19.011884689331055,
+      "learning_rate": 9.829829424443666e-06,
+      "loss": 5.478,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3520988622989408,
+      "grad_norm": 12.593403816223145,
+      "learning_rate": 9.828826801885967e-06,
+      "loss": 5.7249,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35307963907414674,
+      "grad_norm": 14.620553016662598,
+      "learning_rate": 9.827821285773855e-06,
+      "loss": 5.3739,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35406041584935266,
+      "grad_norm": 26.63345718383789,
+      "learning_rate": 9.826812876709861e-06,
+      "loss": 5.2293,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35504119262455863,
+      "grad_norm": 19.860340118408203,
+      "learning_rate": 9.825801575298248e-06,
+      "loss": 5.567,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3560219693997646,
+      "grad_norm": 18.279808044433594,
+      "learning_rate": 9.824787382145013e-06,
+      "loss": 5.248,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3570027461749706,
+      "grad_norm": 18.21912384033203,
+      "learning_rate": 9.82377029785789e-06,
+      "loss": 5.4724,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35798352295017655,
+      "grad_norm": 18.464963912963867,
+      "learning_rate": 9.822750323046333e-06,
+      "loss": 5.1548,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3589642997253825,
+      "grad_norm": 16.67640495300293,
+      "learning_rate": 9.821727458321544e-06,
+      "loss": 5.2362,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3599450765005885,
+      "grad_norm": 23.646631240844727,
+      "learning_rate": 9.820701704296447e-06,
+      "loss": 5.2346,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3609258532757944,
+      "grad_norm": 24.845317840576172,
+      "learning_rate": 9.819673061585698e-06,
+      "loss": 5.4409,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3619066300510004,
+      "grad_norm": 28.531173706054688,
+      "learning_rate": 9.818641530805688e-06,
+      "loss": 5.3217,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36288740682620635,
+      "grad_norm": 22.022838592529297,
+      "learning_rate": 9.817607112574534e-06,
+      "loss": 5.3782,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3638681836014123,
+      "grad_norm": 17.82651138305664,
+      "learning_rate": 9.816569807512088e-06,
+      "loss": 5.1621,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3648489603766183,
+      "grad_norm": 12.07430362701416,
+      "learning_rate": 9.815529616239927e-06,
+      "loss": 5.5542,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36582973715182426,
+      "grad_norm": 12.271188735961914,
+      "learning_rate": 9.81448653938136e-06,
+      "loss": 5.1587,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36681051392703024,
+      "grad_norm": 13.226649284362793,
+      "learning_rate": 9.813440577561429e-06,
+      "loss": 5.3729,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36779129070223615,
+      "grad_norm": 14.851487159729004,
+      "learning_rate": 9.812391731406893e-06,
+      "loss": 5.1667,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3687720674774421,
+      "grad_norm": 23.42156410217285,
+      "learning_rate": 9.811340001546252e-06,
+      "loss": 5.4899,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3697528442526481,
+      "grad_norm": 11.76750659942627,
+      "learning_rate": 9.81028538860973e-06,
+      "loss": 5.0813,
+      "step": 1885
+    },
+    {
+      "epoch": 0.37073362102785407,
+      "grad_norm": 34.78882598876953,
+      "learning_rate": 9.809227893229273e-06,
+      "loss": 5.2967,
+      "step": 1890
+    },
+    {
+      "epoch": 0.37171439780306004,
+      "grad_norm": 19.89539909362793,
+      "learning_rate": 9.808167516038562e-06,
+      "loss": 5.2345,
+      "step": 1895
+    },
+    {
+      "epoch": 0.372695174578266,
+      "grad_norm": 16.6094913482666,
+      "learning_rate": 9.807104257673003e-06,
+      "loss": 4.9801,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3736759513534719,
+      "grad_norm": 22.73749351501465,
+      "learning_rate": 9.806038118769724e-06,
+      "loss": 5.5207,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3746567281286779,
+      "grad_norm": 10.95235538482666,
+      "learning_rate": 9.804969099967583e-06,
+      "loss": 5.0407,
+      "step": 1910
+    },
+    {
+      "epoch": 0.37563750490388387,
+      "grad_norm": 22.019926071166992,
+      "learning_rate": 9.803897201907164e-06,
+      "loss": 5.0178,
+      "step": 1915
+    },
+    {
+      "epoch": 0.37661828167908984,
+      "grad_norm": 29.805604934692383,
+      "learning_rate": 9.802822425230776e-06,
+      "loss": 5.3743,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3775990584542958,
+      "grad_norm": 16.524503707885742,
+      "learning_rate": 9.801744770582449e-06,
+      "loss": 4.6636,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3785798352295018,
+      "grad_norm": 21.630876541137695,
+      "learning_rate": 9.800664238607942e-06,
+      "loss": 5.465,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37956061200470775,
+      "grad_norm": 22.99994468688965,
+      "learning_rate": 9.799580829954739e-06,
+      "loss": 4.9949,
+      "step": 1935
+    },
+    {
+      "epoch": 0.38054138877991367,
+      "grad_norm": 21.744600296020508,
+      "learning_rate": 9.798494545272044e-06,
+      "loss": 5.2195,
+      "step": 1940
+    },
+    {
+      "epoch": 0.38152216555511964,
+      "grad_norm": 19.225811004638672,
+      "learning_rate": 9.797405385210787e-06,
+      "loss": 5.3427,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3825029423303256,
+      "grad_norm": 16.982257843017578,
+      "learning_rate": 9.796313350423619e-06,
+      "loss": 5.5768,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3834837191055316,
+      "grad_norm": 13.56407642364502,
+      "learning_rate": 9.795218441564914e-06,
+      "loss": 5.4152,
+      "step": 1955
+    },
+    {
+      "epoch": 0.38446449588073756,
+      "grad_norm": 20.912071228027344,
+      "learning_rate": 9.79412065929077e-06,
+      "loss": 5.3516,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3854452726559435,
+      "grad_norm": 18.138721466064453,
+      "learning_rate": 9.793020004259008e-06,
+      "loss": 5.4028,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38642604943114944,
+      "grad_norm": 18.006498336791992,
+      "learning_rate": 9.791916477129165e-06,
+      "loss": 5.0267,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3874068262063554,
+      "grad_norm": 12.000133514404297,
+      "learning_rate": 9.790810078562503e-06,
+      "loss": 5.2237,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3883876029815614,
+      "grad_norm": 21.019866943359375,
+      "learning_rate": 9.789700809222005e-06,
+      "loss": 5.5981,
+      "step": 1980
+    },
+    {
+      "epoch": 0.38936837975676736,
+      "grad_norm": 12.761972427368164,
+      "learning_rate": 9.78858866977237e-06,
+      "loss": 5.0066,
+      "step": 1985
+    },
+    {
+      "epoch": 0.39034915653197333,
+      "grad_norm": 14.214351654052734,
+      "learning_rate": 9.787473660880022e-06,
+      "loss": 5.9138,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3913299333071793,
+      "grad_norm": 15.80001163482666,
+      "learning_rate": 9.786355783213104e-06,
+      "loss": 5.6329,
+      "step": 1995
+    },
+    {
+      "epoch": 0.39231071008238527,
+      "grad_norm": 34.411319732666016,
+      "learning_rate": 9.785235037441473e-06,
+      "loss": 5.4683,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3932914868575912,
+      "grad_norm": 21.931644439697266,
+      "learning_rate": 9.784111424236713e-06,
+      "loss": 5.0713,
+      "step": 2005
+    },
+    {
+      "epoch": 0.39427226363279716,
+      "grad_norm": 15.656501770019531,
+      "learning_rate": 9.782984944272115e-06,
+      "loss": 5.2415,
+      "step": 2010
+    },
+    {
+      "epoch": 0.39525304040800313,
+      "grad_norm": 18.855010986328125,
+      "learning_rate": 9.781855598222698e-06,
+      "loss": 5.1184,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3962338171832091,
+      "grad_norm": 21.040325164794922,
+      "learning_rate": 9.780723386765194e-06,
+      "loss": 5.2587,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3972145939584151,
+      "grad_norm": 21.39097023010254,
+      "learning_rate": 9.779588310578051e-06,
+      "loss": 5.2524,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39819537073362105,
+      "grad_norm": 9.114205360412598,
+      "learning_rate": 9.778450370341439e-06,
+      "loss": 5.4023,
+      "step": 2030
+    },
+    {
+      "epoch": 0.399176147508827,
+      "grad_norm": 19.155128479003906,
+      "learning_rate": 9.777309566737236e-06,
+      "loss": 5.11,
+      "step": 2035
+    },
+    {
+      "epoch": 0.40015692428403293,
+      "grad_norm": 19.29751205444336,
+      "learning_rate": 9.776165900449044e-06,
+      "loss": 4.9788,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4011377010592389,
+      "grad_norm": 25.808748245239258,
+      "learning_rate": 9.775019372162173e-06,
+      "loss": 5.4822,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4021184778344449,
+      "grad_norm": 18.989933013916016,
+      "learning_rate": 9.773869982563653e-06,
+      "loss": 5.6281,
+      "step": 2050
+    },
+    {
+      "epoch": 0.40309925460965085,
+      "grad_norm": 26.764808654785156,
+      "learning_rate": 9.77271773234223e-06,
+      "loss": 5.2508,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4040800313848568,
+      "grad_norm": 24.62432289123535,
+      "learning_rate": 9.771562622188355e-06,
+      "loss": 5.4812,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4050608081600628,
+      "grad_norm": 13.985261917114258,
+      "learning_rate": 9.770404652794206e-06,
+      "loss": 5.2385,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4060415849352687,
+      "grad_norm": 31.23125457763672,
+      "learning_rate": 9.769243824853661e-06,
+      "loss": 5.1801,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4070223617104747,
+      "grad_norm": 25.266263961791992,
+      "learning_rate": 9.768080139062321e-06,
+      "loss": 5.2785,
+      "step": 2075
+    },
+    {
+      "epoch": 0.40800313848568065,
+      "grad_norm": 23.64206886291504,
+      "learning_rate": 9.766913596117497e-06,
+      "loss": 5.5125,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4089839152608866,
+      "grad_norm": 20.972145080566406,
+      "learning_rate": 9.765744196718207e-06,
+      "loss": 5.0472,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4099646920360926,
+      "grad_norm": 22.69337272644043,
+      "learning_rate": 9.764571941565189e-06,
+      "loss": 4.8892,
+      "step": 2090
+    },
+    {
+      "epoch": 0.41094546881129856,
+      "grad_norm": 20.279399871826172,
+      "learning_rate": 9.763396831360884e-06,
+      "loss": 5.4165,
+      "step": 2095
+    },
+    {
+      "epoch": 0.41192624558650454,
+      "grad_norm": 22.26048469543457,
+      "learning_rate": 9.76221886680945e-06,
+      "loss": 5.0247,
+      "step": 2100
+    },
+    {
+      "epoch": 0.41290702236171045,
+      "grad_norm": 23.47907066345215,
+      "learning_rate": 9.76103804861675e-06,
+      "loss": 5.4522,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4138877991369164,
+      "grad_norm": 27.87772560119629,
+      "learning_rate": 9.75985437749036e-06,
+      "loss": 5.1454,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4148685759121224,
+      "grad_norm": 18.217052459716797,
+      "learning_rate": 9.758667854139572e-06,
+      "loss": 5.4936,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41584935268732837,
+      "grad_norm": 18.460439682006836,
+      "learning_rate": 9.757478479275373e-06,
+      "loss": 5.0919,
+      "step": 2120
+    },
+    {
+      "epoch": 0.41683012946253434,
+      "grad_norm": 17.97971534729004,
+      "learning_rate": 9.75628625361047e-06,
+      "loss": 5.1305,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4178109062377403,
+      "grad_norm": 35.104087829589844,
+      "learning_rate": 9.755091177859273e-06,
+      "loss": 5.4199,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4187916830129463,
+      "grad_norm": 13.479544639587402,
+      "learning_rate": 9.753893252737903e-06,
+      "loss": 5.1701,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4197724597881522,
+      "grad_norm": 17.23796844482422,
+      "learning_rate": 9.752692478964186e-06,
+      "loss": 5.186,
+      "step": 2140
+    },
+    {
+      "epoch": 0.42075323656335817,
+      "grad_norm": 68.8131332397461,
+      "learning_rate": 9.751488857257657e-06,
+      "loss": 5.2711,
+      "step": 2145
+    },
+    {
+      "epoch": 0.42173401333856414,
+      "grad_norm": 16.84771156311035,
+      "learning_rate": 9.750282388339554e-06,
+      "loss": 5.0436,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4227147901137701,
+      "grad_norm": 16.67029571533203,
+      "learning_rate": 9.749073072932824e-06,
+      "loss": 5.4642,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4236955668889761,
+      "grad_norm": 13.9814453125,
+      "learning_rate": 9.747860911762122e-06,
+      "loss": 5.7409,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42467634366418205,
+      "grad_norm": 28.299707412719727,
+      "learning_rate": 9.746645905553802e-06,
+      "loss": 5.4507,
+      "step": 2165
+    },
+    {
+      "epoch": 0.42565712043938797,
+      "grad_norm": 27.976619720458984,
+      "learning_rate": 9.745428055035928e-06,
+      "loss": 5.0855,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42663789721459394,
+      "grad_norm": 22.406450271606445,
+      "learning_rate": 9.744207360938267e-06,
+      "loss": 5.2556,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4276186739897999,
+      "grad_norm": 19.76823616027832,
+      "learning_rate": 9.742983823992289e-06,
+      "loss": 4.8967,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4285994507650059,
+      "grad_norm": 19.014881134033203,
+      "learning_rate": 9.741757444931169e-06,
+      "loss": 4.972,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42958022754021186,
+      "grad_norm": 23.06075668334961,
+      "learning_rate": 9.74052822448978e-06,
+      "loss": 5.1589,
+      "step": 2190
+    },
+    {
+      "epoch": 0.43056100431541783,
+      "grad_norm": 14.573113441467285,
+      "learning_rate": 9.739296163404708e-06,
+      "loss": 5.2036,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4315417810906238,
+      "grad_norm": 17.803037643432617,
+      "learning_rate": 9.738061262414232e-06,
+      "loss": 5.0363,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4325225578658297,
+      "grad_norm": 36.00239944458008,
+      "learning_rate": 9.736823522258334e-06,
+      "loss": 5.421,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4335033346410357,
+      "grad_norm": 13.375822067260742,
+      "learning_rate": 9.735582943678701e-06,
+      "loss": 5.034,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43448411141624166,
+      "grad_norm": 13.493207931518555,
+      "learning_rate": 9.73433952741872e-06,
+      "loss": 5.356,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43546488819144763,
+      "grad_norm": 26.28885269165039,
+      "learning_rate": 9.733093274223474e-06,
+      "loss": 5.2359,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4364456649666536,
+      "grad_norm": 25.277320861816406,
+      "learning_rate": 9.731844184839751e-06,
+      "loss": 5.4302,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4374264417418596,
+      "grad_norm": 26.082422256469727,
+      "learning_rate": 9.73059226001604e-06,
+      "loss": 5.0238,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4384072185170655,
+      "grad_norm": 64.33253479003906,
+      "learning_rate": 9.729337500502519e-06,
+      "loss": 5.3079,
+      "step": 2235
+    },
+    {
+      "epoch": 0.43938799529227146,
+      "grad_norm": 28.68381118774414,
+      "learning_rate": 9.728079907051076e-06,
+      "loss": 5.1955,
+      "step": 2240
+    },
+    {
+      "epoch": 0.44036877206747743,
+      "grad_norm": 12.982084274291992,
+      "learning_rate": 9.726819480415292e-06,
+      "loss": 4.99,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4413495488426834,
+      "grad_norm": 27.528892517089844,
+      "learning_rate": 9.725556221350448e-06,
+      "loss": 5.2779,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4423303256178894,
+      "grad_norm": 14.044219970703125,
+      "learning_rate": 9.724290130613518e-06,
+      "loss": 4.8775,
+      "step": 2255
+    },
+    {
+      "epoch": 0.44331110239309535,
+      "grad_norm": 17.11577796936035,
+      "learning_rate": 9.723021208963174e-06,
+      "loss": 5.2466,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4442918791683013,
+      "grad_norm": 19.760128021240234,
+      "learning_rate": 9.72174945715979e-06,
+      "loss": 5.31,
+      "step": 2265
+    },
+    {
+      "epoch": 0.44527265594350723,
+      "grad_norm": 12.064308166503906,
+      "learning_rate": 9.72047487596543e-06,
+      "loss": 5.0058,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4462534327187132,
+      "grad_norm": 21.768482208251953,
+      "learning_rate": 9.71919746614385e-06,
+      "loss": 5.2065,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4472342094939192,
+      "grad_norm": 25.759071350097656,
+      "learning_rate": 9.717917228460516e-06,
+      "loss": 5.234,
+      "step": 2280
+    },
+    {
+      "epoch": 0.44821498626912515,
+      "grad_norm": 12.627178192138672,
+      "learning_rate": 9.71663416368257e-06,
+      "loss": 5.1581,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4491957630443311,
+      "grad_norm": 15.976608276367188,
+      "learning_rate": 9.715348272578861e-06,
+      "loss": 5.026,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4501765398195371,
+      "grad_norm": 16.61644172668457,
+      "learning_rate": 9.714059555919928e-06,
+      "loss": 4.9655,
+      "step": 2295
+    },
+    {
+      "epoch": 0.45115731659474306,
+      "grad_norm": 13.492209434509277,
+      "learning_rate": 9.712768014477997e-06,
+      "loss": 5.3452,
+      "step": 2300
+    },
+    {
+      "epoch": 0.452138093369949,
+      "grad_norm": 18.46678924560547,
+      "learning_rate": 9.711473649027e-06,
+      "loss": 4.9953,
+      "step": 2305
+    },
+    {
+      "epoch": 0.45311887014515495,
+      "grad_norm": 23.91395378112793,
+      "learning_rate": 9.710176460342546e-06,
+      "loss": 5.1415,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4540996469203609,
+      "grad_norm": 40.245357513427734,
+      "learning_rate": 9.708876449201945e-06,
+      "loss": 4.8526,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4550804236955669,
+      "grad_norm": 19.441259384155273,
+      "learning_rate": 9.7075736163842e-06,
+      "loss": 4.808,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45606120047077287,
+      "grad_norm": 19.449247360229492,
+      "learning_rate": 9.706267962669999e-06,
+      "loss": 5.0985,
+      "step": 2325
+    },
+    {
+      "epoch": 0.45704197724597884,
+      "grad_norm": 19.058900833129883,
+      "learning_rate": 9.70495948884172e-06,
+      "loss": 4.9696,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45802275402118475,
+      "grad_norm": 12.461446762084961,
+      "learning_rate": 9.703648195683438e-06,
+      "loss": 5.4223,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4590035307963907,
+      "grad_norm": 19.470163345336914,
+      "learning_rate": 9.70233408398091e-06,
+      "loss": 5.3928,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4599843075715967,
+      "grad_norm": 36.08694076538086,
+      "learning_rate": 9.701017154521584e-06,
+      "loss": 5.2037,
+      "step": 2345
+    },
+    {
+      "epoch": 0.46096508434680267,
+      "grad_norm": 23.371965408325195,
+      "learning_rate": 9.699697408094597e-06,
+      "loss": 5.1928,
+      "step": 2350
+    },
+    {
+      "epoch": 0.46194586112200864,
+      "grad_norm": 24.838356018066406,
+      "learning_rate": 9.698374845490779e-06,
+      "loss": 5.0539,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4629266378972146,
+      "grad_norm": 11.548550605773926,
+      "learning_rate": 9.697049467502637e-06,
+      "loss": 5.2077,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4639074146724206,
+      "grad_norm": 14.973907470703125,
+      "learning_rate": 9.695721274924374e-06,
+      "loss": 5.4514,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4648881914476265,
+      "grad_norm": 16.513147354125977,
+      "learning_rate": 9.694390268551875e-06,
+      "loss": 5.3534,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46586896822283247,
+      "grad_norm": 12.842937469482422,
+      "learning_rate": 9.693056449182714e-06,
+      "loss": 5.1984,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46684974499803844,
+      "grad_norm": 21.4840030670166,
+      "learning_rate": 9.691719817616148e-06,
+      "loss": 5.0132,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4678305217732444,
+      "grad_norm": 35.71097946166992,
+      "learning_rate": 9.690380374653121e-06,
+      "loss": 5.0982,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4688112985484504,
+      "grad_norm": 32.27992630004883,
+      "learning_rate": 9.689038121096259e-06,
+      "loss": 5.1919,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46979207532365636,
+      "grad_norm": 11.394453048706055,
+      "learning_rate": 9.687693057749876e-06,
+      "loss": 4.8216,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4707728520988623,
+      "grad_norm": 15.756589889526367,
+      "learning_rate": 9.686345185419968e-06,
+      "loss": 5.0804,
+      "step": 2400
+    },
+    {
+      "epoch": 0.47175362887406824,
+      "grad_norm": 11.9701509475708,
+      "learning_rate": 9.684994504914212e-06,
+      "loss": 4.9623,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4727344056492742,
+      "grad_norm": 30.16839599609375,
+      "learning_rate": 9.683641017041971e-06,
+      "loss": 5.1943,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4737151824244802,
+      "grad_norm": 27.328493118286133,
+      "learning_rate": 9.68228472261429e-06,
+      "loss": 5.3544,
+      "step": 2415
+    },
+    {
+      "epoch": 0.47469595919968616,
+      "grad_norm": 12.850371360778809,
+      "learning_rate": 9.680925622443893e-06,
+      "loss": 5.3483,
+      "step": 2420
+    },
+    {
+      "epoch": 0.47567673597489213,
+      "grad_norm": 13.248522758483887,
+      "learning_rate": 9.679563717345186e-06,
+      "loss": 5.5595,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4766575127500981,
+      "grad_norm": 16.820730209350586,
+      "learning_rate": 9.67819900813426e-06,
+      "loss": 5.2644,
+      "step": 2430
+    },
+    {
+      "epoch": 0.477638289525304,
+      "grad_norm": 19.460813522338867,
+      "learning_rate": 9.676831495628881e-06,
+      "loss": 5.2494,
+      "step": 2435
+    },
+    {
+      "epoch": 0.47861906630051,
+      "grad_norm": 19.54216766357422,
+      "learning_rate": 9.675461180648498e-06,
+      "loss": 5.3748,
+      "step": 2440
+    },
+    {
+      "epoch": 0.47959984307571596,
+      "grad_norm": 23.862497329711914,
+      "learning_rate": 9.674088064014235e-06,
+      "loss": 5.0946,
+      "step": 2445
+    },
+    {
+      "epoch": 0.48058061985092193,
+      "grad_norm": 10.924250602722168,
+      "learning_rate": 9.672712146548903e-06,
+      "loss": 5.156,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4815613966261279,
+      "grad_norm": 24.398984909057617,
+      "learning_rate": 9.671333429076983e-06,
+      "loss": 5.2259,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4825421734013339,
+      "grad_norm": 12.020353317260742,
+      "learning_rate": 9.669951912424638e-06,
+      "loss": 5.2717,
+      "step": 2460
+    },
+    {
+      "epoch": 0.48352295017653985,
+      "grad_norm": 22.294231414794922,
+      "learning_rate": 9.668567597419708e-06,
+      "loss": 5.2593,
+      "step": 2465
+    },
+    {
+      "epoch": 0.48450372695174576,
+      "grad_norm": 21.997915267944336,
+      "learning_rate": 9.667180484891707e-06,
+      "loss": 5.154,
+      "step": 2470
+    },
+    {
+      "epoch": 0.48548450372695173,
+      "grad_norm": 18.885684967041016,
+      "learning_rate": 9.66579057567183e-06,
+      "loss": 5.0931,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4864652805021577,
+      "grad_norm": 25.34767723083496,
+      "learning_rate": 9.664397870592945e-06,
+      "loss": 4.9524,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4874460572773637,
+      "grad_norm": 15.552264213562012,
+      "learning_rate": 9.663002370489596e-06,
+      "loss": 4.9879,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48842683405256965,
+      "grad_norm": 20.804332733154297,
+      "learning_rate": 9.661604076198003e-06,
+      "loss": 5.2101,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4894076108277756,
+      "grad_norm": 20.767370223999023,
+      "learning_rate": 9.660202988556057e-06,
+      "loss": 4.8171,
+      "step": 2495
+    },
+    {
+      "epoch": 0.49038838760298153,
+      "grad_norm": 25.34084701538086,
+      "learning_rate": 9.658799108403324e-06,
+      "loss": 5.2654,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4913691643781875,
+      "grad_norm": 13.980265617370605,
+      "learning_rate": 9.657392436581049e-06,
+      "loss": 4.9889,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4923499411533935,
+      "grad_norm": 19.053634643554688,
+      "learning_rate": 9.655982973932141e-06,
+      "loss": 5.6044,
+      "step": 2510
+    },
+    {
+      "epoch": 0.49333071792859945,
+      "grad_norm": 26.305130004882812,
+      "learning_rate": 9.654570721301186e-06,
+      "loss": 5.1973,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4943114947038054,
+      "grad_norm": 17.097795486450195,
+      "learning_rate": 9.653155679534441e-06,
+      "loss": 4.9489,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4952922714790114,
+      "grad_norm": 18.81172752380371,
+      "learning_rate": 9.651737849479838e-06,
+      "loss": 5.3182,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49627304825421736,
+      "grad_norm": 20.398967742919922,
+      "learning_rate": 9.650317231986972e-06,
+      "loss": 5.15,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4972538250294233,
+      "grad_norm": 14.472307205200195,
+      "learning_rate": 9.648893827907115e-06,
+      "loss": 5.1667,
+      "step": 2535
+    },
+    {
+      "epoch": 0.49823460180462925,
+      "grad_norm": 18.856861114501953,
+      "learning_rate": 9.647467638093206e-06,
+      "loss": 4.9925,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4992153785798352,
+      "grad_norm": 22.74046516418457,
+      "learning_rate": 9.646038663399854e-06,
+      "loss": 5.1208,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5001961553550411,
+      "grad_norm": 15.445943832397461,
+      "learning_rate": 9.644606904683335e-06,
+      "loss": 5.0133,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5001961553550411,
+      "eval_loss": 5.15736198425293,
+      "eval_runtime": 7.6721,
+      "eval_samples_per_second": 27.242,
+      "eval_steps_per_second": 13.686,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5011769321302472,
+      "grad_norm": 18.68064308166504,
+      "learning_rate": 9.643172362801599e-06,
+      "loss": 4.8855,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5021577089054531,
+      "grad_norm": 18.33717155456543,
+      "learning_rate": 9.641735038614255e-06,
+      "loss": 5.4683,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5031384856806591,
+      "grad_norm": 19.931522369384766,
+      "learning_rate": 9.640294932982585e-06,
+      "loss": 5.3343,
+      "step": 2565
+    },
+    {
+      "epoch": 0.504119262455865,
+      "grad_norm": 23.878559112548828,
+      "learning_rate": 9.63885204676954e-06,
+      "loss": 4.7968,
+      "step": 2570
+    },
+    {
+      "epoch": 0.505100039231071,
+      "grad_norm": 19.06217384338379,
+      "learning_rate": 9.637406380839728e-06,
+      "loss": 5.2138,
+      "step": 2575
+    },
+    {
+      "epoch": 0.506080816006277,
+      "grad_norm": 14.29570484161377,
+      "learning_rate": 9.635957936059432e-06,
+      "loss": 5.4729,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5070615927814829,
+      "grad_norm": 19.553762435913086,
+      "learning_rate": 9.634506713296596e-06,
+      "loss": 5.4094,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5080423695566889,
+      "grad_norm": 15.330031394958496,
+      "learning_rate": 9.633052713420828e-06,
+      "loss": 5.1174,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5090231463318948,
+      "grad_norm": 21.965177536010742,
+      "learning_rate": 9.631595937303402e-06,
+      "loss": 5.2644,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5100039231071009,
+      "grad_norm": 19.100711822509766,
+      "learning_rate": 9.630136385817258e-06,
+      "loss": 5.1775,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5109846998823068,
+      "grad_norm": 14.69621467590332,
+      "learning_rate": 9.62867405983699e-06,
+      "loss": 5.0009,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5119654766575128,
+      "grad_norm": 19.2616024017334,
+      "learning_rate": 9.627208960238864e-06,
+      "loss": 4.8041,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5129462534327187,
+      "grad_norm": 13.647692680358887,
+      "learning_rate": 9.625741087900802e-06,
+      "loss": 4.7495,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5139270302079246,
+      "grad_norm": 14.416105270385742,
+      "learning_rate": 9.624270443702395e-06,
+      "loss": 5.3507,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5149078069831307,
+      "grad_norm": 12.422703742980957,
+      "learning_rate": 9.622797028524885e-06,
+      "loss": 4.8836,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5158885837583366,
+      "grad_norm": 15.866639137268066,
+      "learning_rate": 9.621320843251183e-06,
+      "loss": 5.352,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5168693605335426,
+      "grad_norm": 53.905094146728516,
+      "learning_rate": 9.619841888765853e-06,
+      "loss": 5.4811,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5178501373087485,
+      "grad_norm": 18.568187713623047,
+      "learning_rate": 9.618360165955125e-06,
+      "loss": 5.2916,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5188309140839545,
+      "grad_norm": 31.13910484313965,
+      "learning_rate": 9.61687567570688e-06,
+      "loss": 5.3021,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5198116908591605,
+      "grad_norm": 21.233203887939453,
+      "learning_rate": 9.615388418910668e-06,
+      "loss": 5.2845,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5207924676343664,
+      "grad_norm": 30.170978546142578,
+      "learning_rate": 9.613898396457687e-06,
+      "loss": 5.1301,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5217732444095724,
+      "grad_norm": 23.659400939941406,
+      "learning_rate": 9.612405609240795e-06,
+      "loss": 5.217,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5227540211847783,
+      "grad_norm": 12.635769844055176,
+      "learning_rate": 9.61091005815451e-06,
+      "loss": 5.0976,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5237347979599843,
+      "grad_norm": 23.049379348754883,
+      "learning_rate": 9.609411744095002e-06,
+      "loss": 5.3677,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5247155747351903,
+      "grad_norm": 16.504423141479492,
+      "learning_rate": 9.607910667960098e-06,
+      "loss": 5.0437,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5256963515103963,
+      "grad_norm": 12.01205825805664,
+      "learning_rate": 9.606406830649283e-06,
+      "loss": 4.9374,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5266771282856022,
+      "grad_norm": 16.772647857666016,
+      "learning_rate": 9.604900233063696e-06,
+      "loss": 5.0273,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5276579050608081,
+      "grad_norm": 28.808847427368164,
+      "learning_rate": 9.603390876106123e-06,
+      "loss": 5.305,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5286386818360141,
+      "grad_norm": 13.548622131347656,
+      "learning_rate": 9.60187876068101e-06,
+      "loss": 4.9093,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5296194586112201,
+      "grad_norm": 20.574769973754883,
+      "learning_rate": 9.600363887694455e-06,
+      "loss": 5.088,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5306002353864261,
+      "grad_norm": 16.430707931518555,
+      "learning_rate": 9.598846258054208e-06,
+      "loss": 5.3671,
+      "step": 2705
+    },
+    {
+      "epoch": 0.531581012161632,
+      "grad_norm": 24.375839233398438,
+      "learning_rate": 9.597325872669672e-06,
+      "loss": 5.5251,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5325617889368379,
+      "grad_norm": 19.051464080810547,
+      "learning_rate": 9.5958027324519e-06,
+      "loss": 4.948,
+      "step": 2715
+    },
+    {
+      "epoch": 0.533542565712044,
+      "grad_norm": 24.722259521484375,
+      "learning_rate": 9.594276838313593e-06,
+      "loss": 4.8675,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5345233424872499,
+      "grad_norm": 13.493644714355469,
+      "learning_rate": 9.592748191169107e-06,
+      "loss": 5.3292,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5355041192624559,
+      "grad_norm": 12.233308792114258,
+      "learning_rate": 9.59121679193445e-06,
+      "loss": 5.1377,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5364848960376618,
+      "grad_norm": 15.433117866516113,
+      "learning_rate": 9.589682641527269e-06,
+      "loss": 5.1438,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5374656728128678,
+      "grad_norm": 30.711441040039062,
+      "learning_rate": 9.588145740866866e-06,
+      "loss": 5.4655,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5384464495880738,
+      "grad_norm": 21.23065948486328,
+      "learning_rate": 9.586606090874193e-06,
+      "loss": 5.5321,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5394272263632797,
+      "grad_norm": 14.286324501037598,
+      "learning_rate": 9.585063692471845e-06,
+      "loss": 5.2226,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5404080031384857,
+      "grad_norm": 21.153053283691406,
+      "learning_rate": 9.583518546584069e-06,
+      "loss": 5.2436,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5413887799136916,
+      "grad_norm": 16.570341110229492,
+      "learning_rate": 9.581970654136752e-06,
+      "loss": 5.1266,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5423695566888976,
+      "grad_norm": 17.8549747467041,
+      "learning_rate": 9.580420016057431e-06,
+      "loss": 5.1313,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5433503334641036,
+      "grad_norm": 37.07738494873047,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 5.1527,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5443311102393096,
+      "grad_norm": 19.371015548706055,
+      "learning_rate": 9.577310506721148e-06,
+      "loss": 4.9898,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5453118870145155,
+      "grad_norm": 17.40994644165039,
+      "learning_rate": 9.575751637327481e-06,
+      "loss": 5.1577,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5462926637897214,
+      "grad_norm": 13.387382507324219,
+      "learning_rate": 9.574190026028404e-06,
+      "loss": 5.1941,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5472734405649274,
+      "grad_norm": 21.211164474487305,
+      "learning_rate": 9.57262567375967e-06,
+      "loss": 4.9418,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5482542173401334,
+      "grad_norm": 14.441864967346191,
+      "learning_rate": 9.57105858145868e-06,
+      "loss": 5.2723,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5492349941153394,
+      "grad_norm": 23.936847686767578,
+      "learning_rate": 9.569488750064472e-06,
+      "loss": 5.5534,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5502157708905453,
+      "grad_norm": 14.22829818725586,
+      "learning_rate": 9.567916180517733e-06,
+      "loss": 5.104,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5511965476657513,
+      "grad_norm": 19.610498428344727,
+      "learning_rate": 9.566340873760784e-06,
+      "loss": 5.8822,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5521773244409572,
+      "grad_norm": 32.81593322753906,
+      "learning_rate": 9.564762830737586e-06,
+      "loss": 5.0032,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5531581012161632,
+      "grad_norm": 15.706247329711914,
+      "learning_rate": 9.563182052393747e-06,
+      "loss": 5.1115,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5541388779913692,
+      "grad_norm": 24.40583610534668,
+      "learning_rate": 9.561598539676507e-06,
+      "loss": 5.0677,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5551196547665751,
+      "grad_norm": 13.26285171508789,
+      "learning_rate": 9.560012293534746e-06,
+      "loss": 5.0625,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5561004315417811,
+      "grad_norm": 12.539552688598633,
+      "learning_rate": 9.558423314918982e-06,
+      "loss": 5.0136,
+      "step": 2835
+    },
+    {
+      "epoch": 0.557081208316987,
+      "grad_norm": 22.337364196777344,
+      "learning_rate": 9.556831604781373e-06,
+      "loss": 4.7906,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5580619850921931,
+      "grad_norm": 14.324231147766113,
+      "learning_rate": 9.55523716407571e-06,
+      "loss": 4.7484,
+      "step": 2845
+    },
+    {
+      "epoch": 0.559042761867399,
+      "grad_norm": 16.521678924560547,
+      "learning_rate": 9.553639993757422e-06,
+      "loss": 4.8595,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5600235386426049,
+      "grad_norm": 17.187822341918945,
+      "learning_rate": 9.552040094783575e-06,
+      "loss": 4.8251,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5610043154178109,
+      "grad_norm": 18.47371482849121,
+      "learning_rate": 9.550437468112867e-06,
+      "loss": 5.4023,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5619850921930168,
+      "grad_norm": 22.58086585998535,
+      "learning_rate": 9.548832114705634e-06,
+      "loss": 5.22,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5629658689682229,
+      "grad_norm": 12.957576751708984,
+      "learning_rate": 9.547224035523841e-06,
+      "loss": 5.3866,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5639466457434288,
+      "grad_norm": 23.065975189208984,
+      "learning_rate": 9.545613231531094e-06,
+      "loss": 5.0757,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5649274225186348,
+      "grad_norm": 11.80753231048584,
+      "learning_rate": 9.543999703692624e-06,
+      "loss": 5.2471,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5659081992938407,
+      "grad_norm": 14.523046493530273,
+      "learning_rate": 9.5423834529753e-06,
+      "loss": 5.001,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5668889760690466,
+      "grad_norm": 28.944412231445312,
+      "learning_rate": 9.540764480347616e-06,
+      "loss": 4.9322,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5678697528442527,
+      "grad_norm": 18.60620880126953,
+      "learning_rate": 9.539142786779702e-06,
+      "loss": 5.0668,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5688505296194586,
+      "grad_norm": 16.527719497680664,
+      "learning_rate": 9.537518373243322e-06,
+      "loss": 5.2394,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5698313063946646,
+      "grad_norm": 16.115703582763672,
+      "learning_rate": 9.535891240711861e-06,
+      "loss": 5.0097,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5708120831698705,
+      "grad_norm": 11.602025032043457,
+      "learning_rate": 9.53426139016034e-06,
+      "loss": 5.4735,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5717928599450764,
+      "grad_norm": 14.616000175476074,
+      "learning_rate": 9.532628822565405e-06,
+      "loss": 5.0512,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5727736367202825,
+      "grad_norm": 27.371612548828125,
+      "learning_rate": 9.530993538905332e-06,
+      "loss": 4.9507,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5737544134954884,
+      "grad_norm": 16.049571990966797,
+      "learning_rate": 9.529355540160025e-06,
+      "loss": 5.9872,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5747351902706944,
+      "grad_norm": 43.3659782409668,
+      "learning_rate": 9.527714827311012e-06,
+      "loss": 5.375,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5757159670459003,
+      "grad_norm": 42.861656188964844,
+      "learning_rate": 9.526071401341452e-06,
+      "loss": 4.8584,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5766967438211064,
+      "grad_norm": 36.75173568725586,
+      "learning_rate": 9.524425263236124e-06,
+      "loss": 5.3912,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5776775205963123,
+      "grad_norm": 18.924116134643555,
+      "learning_rate": 9.522776413981438e-06,
+      "loss": 4.8026,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5786582973715182,
+      "grad_norm": 9.715642929077148,
+      "learning_rate": 9.521124854565425e-06,
+      "loss": 5.0732,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5796390741467242,
+      "grad_norm": 10.902597427368164,
+      "learning_rate": 9.51947058597774e-06,
+      "loss": 5.165,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5806198509219301,
+      "grad_norm": 15.857320785522461,
+      "learning_rate": 9.517813609209665e-06,
+      "loss": 5.3345,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5816006276971362,
+      "grad_norm": 18.43710708618164,
+      "learning_rate": 9.5161539252541e-06,
+      "loss": 5.3621,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5825814044723421,
+      "grad_norm": 27.262819290161133,
+      "learning_rate": 9.51449153510557e-06,
+      "loss": 5.2153,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5835621812475481,
+      "grad_norm": 19.492929458618164,
+      "learning_rate": 9.51282643976022e-06,
+      "loss": 5.345,
+      "step": 2975
+    },
+    {
+      "epoch": 0.584542958022754,
+      "grad_norm": 22.923349380493164,
+      "learning_rate": 9.511158640215818e-06,
+      "loss": 4.9704,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5855237347979599,
+      "grad_norm": 12.464946746826172,
+      "learning_rate": 9.509488137471751e-06,
+      "loss": 4.9893,
+      "step": 2985
+    },
+    {
+      "epoch": 0.586504511573166,
+      "grad_norm": 24.526554107666016,
+      "learning_rate": 9.507814932529027e-06,
+      "loss": 4.9967,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5874852883483719,
+      "grad_norm": 24.627796173095703,
+      "learning_rate": 9.50613902639027e-06,
+      "loss": 5.2092,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5884660651235779,
+      "grad_norm": 27.909130096435547,
+      "learning_rate": 9.50446042005973e-06,
+      "loss": 5.2925,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5894468418987838,
+      "grad_norm": 35.04037857055664,
+      "learning_rate": 9.502779114543263e-06,
+      "loss": 5.05,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5904276186739899,
+      "grad_norm": 10.635119438171387,
+      "learning_rate": 9.501095110848357e-06,
+      "loss": 4.9822,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5914083954491958,
+      "grad_norm": 11.250473976135254,
+      "learning_rate": 9.499408409984104e-06,
+      "loss": 4.9801,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5923891722244017,
+      "grad_norm": 19.92534828186035,
+      "learning_rate": 9.49771901296122e-06,
+      "loss": 5.2654,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5933699489996077,
+      "grad_norm": 9.814534187316895,
+      "learning_rate": 9.496026920792034e-06,
+      "loss": 5.1213,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5943507257748136,
+      "grad_norm": 25.265356063842773,
+      "learning_rate": 9.49433213449049e-06,
+      "loss": 5.3733,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5953315025500197,
+      "grad_norm": 13.742667198181152,
+      "learning_rate": 9.492634655072143e-06,
+      "loss": 5.1799,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5963122793252256,
+      "grad_norm": 42.507598876953125,
+      "learning_rate": 9.490934483554173e-06,
+      "loss": 5.215,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5972930561004316,
+      "grad_norm": 32.569759368896484,
+      "learning_rate": 9.48923162095536e-06,
+      "loss": 5.2368,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5982738328756375,
+      "grad_norm": 33.264808654785156,
+      "learning_rate": 9.487526068296102e-06,
+      "loss": 4.9331,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5992546096508434,
+      "grad_norm": 27.735538482666016,
+      "learning_rate": 9.485817826598411e-06,
+      "loss": 5.4973,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6002353864260495,
+      "grad_norm": 26.621362686157227,
+      "learning_rate": 9.48410689688591e-06,
+      "loss": 5.3181,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6012161632012554,
+      "grad_norm": 20.41889762878418,
+      "learning_rate": 9.482393280183827e-06,
+      "loss": 5.168,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6021969399764614,
+      "grad_norm": 12.211244583129883,
+      "learning_rate": 9.480676977519005e-06,
+      "loss": 4.9126,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6031777167516673,
+      "grad_norm": 26.63025665283203,
+      "learning_rate": 9.478957989919897e-06,
+      "loss": 5.3281,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6041584935268732,
+      "grad_norm": 40.21369171142578,
+      "learning_rate": 9.477236318416564e-06,
+      "loss": 5.6442,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6051392703020793,
+      "grad_norm": 18.292333602905273,
+      "learning_rate": 9.475511964040674e-06,
+      "loss": 4.8534,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6061200470772852,
+      "grad_norm": 13.079044342041016,
+      "learning_rate": 9.473784927825503e-06,
+      "loss": 5.0455,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6071008238524912,
+      "grad_norm": 16.62065315246582,
+      "learning_rate": 9.472055210805935e-06,
+      "loss": 4.9956,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6080816006276971,
+      "grad_norm": 23.266950607299805,
+      "learning_rate": 9.47032281401846e-06,
+      "loss": 4.7663,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6090623774029031,
+      "grad_norm": 22.578330993652344,
+      "learning_rate": 9.468587738501176e-06,
+      "loss": 5.2476,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6100431541781091,
+      "grad_norm": 22.685009002685547,
+      "learning_rate": 9.46684998529378e-06,
+      "loss": 5.2655,
+      "step": 3110
+    },
+    {
+      "epoch": 0.611023930953315,
+      "grad_norm": 19.493541717529297,
+      "learning_rate": 9.46510955543758e-06,
+      "loss": 5.34,
+      "step": 3115
+    },
+    {
+      "epoch": 0.612004707728521,
+      "grad_norm": 13.723419189453125,
+      "learning_rate": 9.463366449975483e-06,
+      "loss": 5.2569,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6129854845037269,
+      "grad_norm": 16.924633026123047,
+      "learning_rate": 9.461620669952003e-06,
+      "loss": 5.1248,
+      "step": 3125
+    },
+    {
+      "epoch": 0.613966261278933,
+      "grad_norm": 28.247379302978516,
+      "learning_rate": 9.459872216413255e-06,
+      "loss": 5.4098,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6149470380541389,
+      "grad_norm": 9.836380958557129,
+      "learning_rate": 9.458121090406958e-06,
+      "loss": 5.165,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6159278148293449,
+      "grad_norm": 15.825077056884766,
+      "learning_rate": 9.45636729298243e-06,
+      "loss": 5.1338,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6169085916045508,
+      "grad_norm": 23.045560836791992,
+      "learning_rate": 9.454610825190586e-06,
+      "loss": 5.0819,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6178893683797567,
+      "grad_norm": 16.358854293823242,
+      "learning_rate": 9.452851688083953e-06,
+      "loss": 4.7103,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6188701451549627,
+      "grad_norm": 25.5876522064209,
+      "learning_rate": 9.451089882716644e-06,
+      "loss": 5.0649,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6198509219301687,
+      "grad_norm": 14.046899795532227,
+      "learning_rate": 9.449325410144383e-06,
+      "loss": 4.7795,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6208316987053747,
+      "grad_norm": 16.689111709594727,
+      "learning_rate": 9.44755827142448e-06,
+      "loss": 4.8892,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6218124754805806,
+      "grad_norm": 14.133878707885742,
+      "learning_rate": 9.445788467615852e-06,
+      "loss": 5.3107,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6227932522557866,
+      "grad_norm": 20.716943740844727,
+      "learning_rate": 9.444015999779013e-06,
+      "loss": 5.5117,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6237740290309925,
+      "grad_norm": 14.682984352111816,
+      "learning_rate": 9.442240868976064e-06,
+      "loss": 4.8668,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6247548058061985,
+      "grad_norm": 28.73581886291504,
+      "learning_rate": 9.440463076270713e-06,
+      "loss": 4.9583,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6257355825814045,
+      "grad_norm": 11.658886909484863,
+      "learning_rate": 9.438682622728256e-06,
+      "loss": 5.074,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6267163593566104,
+      "grad_norm": 14.399426460266113,
+      "learning_rate": 9.436899509415586e-06,
+      "loss": 5.2788,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6276971361318164,
+      "grad_norm": 12.162981033325195,
+      "learning_rate": 9.435113737401188e-06,
+      "loss": 5.0695,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6286779129070224,
+      "grad_norm": 28.62458038330078,
+      "learning_rate": 9.433325307755144e-06,
+      "loss": 5.0648,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6296586896822284,
+      "grad_norm": 45.26708984375,
+      "learning_rate": 9.431534221549124e-06,
+      "loss": 5.003,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6306394664574343,
+      "grad_norm": 13.835540771484375,
+      "learning_rate": 9.42974047985639e-06,
+      "loss": 5.1126,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6316202432326402,
+      "grad_norm": 22.305397033691406,
+      "learning_rate": 9.427944083751803e-06,
+      "loss": 5.1913,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6326010200078462,
+      "grad_norm": 26.718841552734375,
+      "learning_rate": 9.426145034311805e-06,
+      "loss": 5.2575,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6335817967830522,
+      "grad_norm": 15.588373184204102,
+      "learning_rate": 9.424343332614432e-06,
+      "loss": 4.7884,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6345625735582582,
+      "grad_norm": 23.680042266845703,
+      "learning_rate": 9.422538979739307e-06,
+      "loss": 5.1369,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6355433503334641,
+      "grad_norm": 21.163063049316406,
+      "learning_rate": 9.420731976767647e-06,
+      "loss": 4.9218,
+      "step": 3240
+    },
+    {
+      "epoch": 0.63652412710867,
+      "grad_norm": 24.679636001586914,
+      "learning_rate": 9.418922324782252e-06,
+      "loss": 5.0978,
+      "step": 3245
+    },
+    {
+      "epoch": 0.637504903883876,
+      "grad_norm": 22.380672454833984,
+      "learning_rate": 9.41711002486751e-06,
+      "loss": 4.9464,
+      "step": 3250
+    },
+    {
+      "epoch": 0.638485680659082,
+      "grad_norm": 37.539794921875,
+      "learning_rate": 9.415295078109398e-06,
+      "loss": 4.9001,
+      "step": 3255
+    },
+    {
+      "epoch": 0.639466457434288,
+      "grad_norm": 21.865497589111328,
+      "learning_rate": 9.413477485595479e-06,
+      "loss": 4.9388,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6404472342094939,
+      "grad_norm": 19.220638275146484,
+      "learning_rate": 9.411657248414898e-06,
+      "loss": 5.1998,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6414280109846999,
+      "grad_norm": 27.21924591064453,
+      "learning_rate": 9.409834367658387e-06,
+      "loss": 5.1988,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6424087877599058,
+      "grad_norm": 15.7559175491333,
+      "learning_rate": 9.408008844418262e-06,
+      "loss": 4.9148,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6433895645351118,
+      "grad_norm": 13.318585395812988,
+      "learning_rate": 9.406180679788423e-06,
+      "loss": 5.2602,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6443703413103178,
+      "grad_norm": 12.137907028198242,
+      "learning_rate": 9.404349874864354e-06,
+      "loss": 4.8963,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6453511180855237,
+      "grad_norm": 16.94658660888672,
+      "learning_rate": 9.402516430743115e-06,
+      "loss": 5.0545,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6463318948607297,
+      "grad_norm": 21.77410316467285,
+      "learning_rate": 9.400680348523356e-06,
+      "loss": 5.4027,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6473126716359356,
+      "grad_norm": 26.694643020629883,
+      "learning_rate": 9.398841629305303e-06,
+      "loss": 4.9592,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6482934484111417,
+      "grad_norm": 17.593299865722656,
+      "learning_rate": 9.397000274190759e-06,
+      "loss": 4.8191,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6492742251863476,
+      "grad_norm": 12.933694839477539,
+      "learning_rate": 9.395156284283113e-06,
+      "loss": 5.2141,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6502550019615535,
+      "grad_norm": 11.42352294921875,
+      "learning_rate": 9.39330966068733e-06,
+      "loss": 5.0004,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6512357787367595,
+      "grad_norm": 17.624004364013672,
+      "learning_rate": 9.391460404509954e-06,
+      "loss": 5.2441,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6522165555119654,
+      "grad_norm": 12.520964622497559,
+      "learning_rate": 9.389608516859106e-06,
+      "loss": 4.8506,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6531973322871715,
+      "grad_norm": 29.8953800201416,
+      "learning_rate": 9.387753998844482e-06,
+      "loss": 5.4652,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6541781090623774,
+      "grad_norm": 24.0643367767334,
+      "learning_rate": 9.385896851577357e-06,
+      "loss": 4.9041,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6551588858375834,
+      "grad_norm": 18.141315460205078,
+      "learning_rate": 9.384037076170578e-06,
+      "loss": 5.2624,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6561396626127893,
+      "grad_norm": 15.43309211730957,
+      "learning_rate": 9.382174673738573e-06,
+      "loss": 5.0105,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6571204393879952,
+      "grad_norm": 15.214509010314941,
+      "learning_rate": 9.380309645397337e-06,
+      "loss": 5.0126,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6581012161632013,
+      "grad_norm": 10.24284553527832,
+      "learning_rate": 9.378441992264444e-06,
+      "loss": 5.022,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6590819929384072,
+      "grad_norm": 20.641220092773438,
+      "learning_rate": 9.376571715459037e-06,
+      "loss": 5.0157,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6600627697136132,
+      "grad_norm": 16.595033645629883,
+      "learning_rate": 9.374698816101836e-06,
+      "loss": 5.3137,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6610435464888191,
+      "grad_norm": 15.82652759552002,
+      "learning_rate": 9.372823295315126e-06,
+      "loss": 5.2748,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6620243232640252,
+      "grad_norm": 18.090869903564453,
+      "learning_rate": 9.370945154222767e-06,
+      "loss": 4.8269,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6630051000392311,
+      "grad_norm": 15.272143363952637,
+      "learning_rate": 9.369064393950189e-06,
+      "loss": 5.1632,
+      "step": 3380
+    },
+    {
+      "epoch": 0.663985876814437,
+      "grad_norm": 12.312703132629395,
+      "learning_rate": 9.367181015624392e-06,
+      "loss": 5.3285,
+      "step": 3385
+    },
+    {
+      "epoch": 0.664966653589643,
+      "grad_norm": 16.462718963623047,
+      "learning_rate": 9.36529502037394e-06,
+      "loss": 5.1033,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6659474303648489,
+      "grad_norm": 12.415363311767578,
+      "learning_rate": 9.363406409328972e-06,
+      "loss": 4.752,
+      "step": 3395
+    },
+    {
+      "epoch": 0.666928207140055,
+      "grad_norm": 16.04773712158203,
+      "learning_rate": 9.361515183621191e-06,
+      "loss": 5.1875,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6679089839152609,
+      "grad_norm": 13.996196746826172,
+      "learning_rate": 9.359621344383867e-06,
+      "loss": 5.1647,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6688897606904668,
+      "grad_norm": 13.507939338684082,
+      "learning_rate": 9.357724892751834e-06,
+      "loss": 4.6931,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6698705374656728,
+      "grad_norm": 27.69146728515625,
+      "learning_rate": 9.355825829861495e-06,
+      "loss": 5.2703,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6708513142408787,
+      "grad_norm": 25.83177375793457,
+      "learning_rate": 9.353924156850816e-06,
+      "loss": 5.6674,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6718320910160848,
+      "grad_norm": 25.738636016845703,
+      "learning_rate": 9.352019874859326e-06,
+      "loss": 4.8759,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6728128677912907,
+      "grad_norm": 8.705573081970215,
+      "learning_rate": 9.350112985028121e-06,
+      "loss": 4.8531,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6737936445664967,
+      "grad_norm": 15.935606956481934,
+      "learning_rate": 9.348203488499858e-06,
+      "loss": 5.1539,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6747744213417026,
+      "grad_norm": 27.665693283081055,
+      "learning_rate": 9.34629138641875e-06,
+      "loss": 5.118,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6757551981169085,
+      "grad_norm": 19.099084854125977,
+      "learning_rate": 9.344376679930585e-06,
+      "loss": 5.0635,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6767359748921146,
+      "grad_norm": 15.758007049560547,
+      "learning_rate": 9.342459370182695e-06,
+      "loss": 4.9812,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6777167516673205,
+      "grad_norm": 13.712491035461426,
+      "learning_rate": 9.340539458323985e-06,
+      "loss": 5.1033,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6786975284425265,
+      "grad_norm": 21.73082733154297,
+      "learning_rate": 9.338616945504913e-06,
+      "loss": 4.9175,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6796783052177324,
+      "grad_norm": 12.152356147766113,
+      "learning_rate": 9.336691832877496e-06,
+      "loss": 4.9905,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6806590819929385,
+      "grad_norm": 27.250158309936523,
+      "learning_rate": 9.334764121595312e-06,
+      "loss": 4.7406,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6816398587681444,
+      "grad_norm": 19.077829360961914,
+      "learning_rate": 9.332833812813494e-06,
+      "loss": 4.907,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6826206355433503,
+      "grad_norm": 34.85506057739258,
+      "learning_rate": 9.330900907688728e-06,
+      "loss": 5.0646,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6836014123185563,
+      "grad_norm": 19.06825828552246,
+      "learning_rate": 9.328965407379265e-06,
+      "loss": 5.1541,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6845821890937622,
+      "grad_norm": 14.873254776000977,
+      "learning_rate": 9.327027313044901e-06,
+      "loss": 4.9209,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6855629658689683,
+      "grad_norm": 20.76241683959961,
+      "learning_rate": 9.325086625846993e-06,
+      "loss": 5.1064,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6865437426441742,
+      "grad_norm": 18.24216651916504,
+      "learning_rate": 9.323143346948449e-06,
+      "loss": 4.974,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6875245194193802,
+      "grad_norm": 17.618261337280273,
+      "learning_rate": 9.32119747751373e-06,
+      "loss": 4.8254,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6885052961945861,
+      "grad_norm": 31.671674728393555,
+      "learning_rate": 9.31924901870885e-06,
+      "loss": 4.8518,
+      "step": 3510
+    },
+    {
+      "epoch": 0.689486072969792,
+      "grad_norm": 20.532033920288086,
+      "learning_rate": 9.317297971701376e-06,
+      "loss": 5.1034,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6904668497449981,
+      "grad_norm": 14.455893516540527,
+      "learning_rate": 9.315344337660422e-06,
+      "loss": 5.1695,
+      "step": 3520
+    },
+    {
+      "epoch": 0.691447626520204,
+      "grad_norm": 15.23082160949707,
+      "learning_rate": 9.313388117756655e-06,
+      "loss": 5.1999,
+      "step": 3525
+    },
+    {
+      "epoch": 0.69242840329541,
+      "grad_norm": 17.498613357543945,
+      "learning_rate": 9.311429313162293e-06,
+      "loss": 5.4637,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6934091800706159,
+      "grad_norm": 14.049081802368164,
+      "learning_rate": 9.309467925051101e-06,
+      "loss": 5.0001,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6943899568458219,
+      "grad_norm": 24.643108367919922,
+      "learning_rate": 9.30750395459839e-06,
+      "loss": 5.2113,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6953707336210279,
+      "grad_norm": 16.181472778320312,
+      "learning_rate": 9.305537402981023e-06,
+      "loss": 4.8661,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6963515103962338,
+      "grad_norm": 32.990962982177734,
+      "learning_rate": 9.303568271377404e-06,
+      "loss": 5.2794,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6973322871714398,
+      "grad_norm": 19.17635726928711,
+      "learning_rate": 9.301596560967488e-06,
+      "loss": 4.8633,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6983130639466457,
+      "grad_norm": 16.803030014038086,
+      "learning_rate": 9.299622272932772e-06,
+      "loss": 4.8278,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6992938407218517,
+      "grad_norm": 11.135583877563477,
+      "learning_rate": 9.297645408456301e-06,
+      "loss": 4.9934,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7002746174970577,
+      "grad_norm": 21.217369079589844,
+      "learning_rate": 9.295665968722663e-06,
+      "loss": 5.3993,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7012553942722637,
+      "grad_norm": 22.149660110473633,
+      "learning_rate": 9.293683954917984e-06,
+      "loss": 5.4629,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7022361710474696,
+      "grad_norm": 30.002939224243164,
+      "learning_rate": 9.29169936822994e-06,
+      "loss": 5.0342,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7032169478226755,
+      "grad_norm": 15.087618827819824,
+      "learning_rate": 9.289712209847745e-06,
+      "loss": 4.9437,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7041977245978815,
+      "grad_norm": 19.03937530517578,
+      "learning_rate": 9.287722480962151e-06,
+      "loss": 4.9014,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7051785013730875,
+      "grad_norm": 16.533588409423828,
+      "learning_rate": 9.285730182765456e-06,
+      "loss": 5.0342,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7061592781482935,
+      "grad_norm": 34.22774887084961,
+      "learning_rate": 9.283735316451497e-06,
+      "loss": 4.8334,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7071400549234994,
+      "grad_norm": 20.17010498046875,
+      "learning_rate": 9.281737883215644e-06,
+      "loss": 5.2429,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7081208316987053,
+      "grad_norm": 18.994447708129883,
+      "learning_rate": 9.279737884254812e-06,
+      "loss": 5.218,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7091016084739113,
+      "grad_norm": 17.285263061523438,
+      "learning_rate": 9.277735320767449e-06,
+      "loss": 5.0856,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7100823852491173,
+      "grad_norm": 61.87449645996094,
+      "learning_rate": 9.275730193953542e-06,
+      "loss": 5.0641,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7110631620243233,
+      "grad_norm": 16.89430046081543,
+      "learning_rate": 9.273722505014615e-06,
+      "loss": 5.146,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7120439387995292,
+      "grad_norm": 18.025007247924805,
+      "learning_rate": 9.271712255153724e-06,
+      "loss": 5.1371,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7130247155747352,
+      "grad_norm": 21.702707290649414,
+      "learning_rate": 9.269699445575462e-06,
+      "loss": 4.9414,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7140054923499412,
+      "grad_norm": 16.395092010498047,
+      "learning_rate": 9.267684077485955e-06,
+      "loss": 5.0574,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7149862691251471,
+      "grad_norm": 26.50773048400879,
+      "learning_rate": 9.26566615209286e-06,
+      "loss": 5.2367,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7159670459003531,
+      "grad_norm": 14.42000675201416,
+      "learning_rate": 9.263645670605373e-06,
+      "loss": 5.1632,
+      "step": 3650
+    },
+    {
+      "epoch": 0.716947822675559,
+      "grad_norm": 17.85245704650879,
+      "learning_rate": 9.261622634234213e-06,
+      "loss": 4.6273,
+      "step": 3655
+    },
+    {
+      "epoch": 0.717928599450765,
+      "grad_norm": 37.26712417602539,
+      "learning_rate": 9.259597044191635e-06,
+      "loss": 4.8634,
+      "step": 3660
+    },
+    {
+      "epoch": 0.718909376225971,
+      "grad_norm": 20.40871238708496,
+      "learning_rate": 9.257568901691428e-06,
+      "loss": 4.8579,
+      "step": 3665
+    },
+    {
+      "epoch": 0.719890153001177,
+      "grad_norm": 13.428717613220215,
+      "learning_rate": 9.2555382079489e-06,
+      "loss": 4.8662,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7208709297763829,
+      "grad_norm": 14.121960639953613,
+      "learning_rate": 9.253504964180897e-06,
+      "loss": 4.957,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7218517065515888,
+      "grad_norm": 23.1538143157959,
+      "learning_rate": 9.25146917160579e-06,
+      "loss": 4.8803,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7228324833267948,
+      "grad_norm": 14.52030086517334,
+      "learning_rate": 9.249430831443474e-06,
+      "loss": 4.8426,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7238132601020008,
+      "grad_norm": 18.916912078857422,
+      "learning_rate": 9.247389944915377e-06,
+      "loss": 4.7263,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7247940368772068,
+      "grad_norm": 22.758358001708984,
+      "learning_rate": 9.245346513244448e-06,
+      "loss": 5.196,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7257748136524127,
+      "grad_norm": 15.67486572265625,
+      "learning_rate": 9.243300537655163e-06,
+      "loss": 5.1512,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7267555904276187,
+      "grad_norm": 13.838863372802734,
+      "learning_rate": 9.241252019373522e-06,
+      "loss": 5.2147,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7277363672028246,
+      "grad_norm": 26.13442611694336,
+      "learning_rate": 9.239200959627048e-06,
+      "loss": 5.0676,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7287171439780306,
+      "grad_norm": 17.461381912231445,
+      "learning_rate": 9.237147359644789e-06,
+      "loss": 5.6134,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7296979207532366,
+      "grad_norm": 19.081663131713867,
+      "learning_rate": 9.235091220657313e-06,
+      "loss": 5.2973,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7306786975284425,
+      "grad_norm": 11.549004554748535,
+      "learning_rate": 9.23303254389671e-06,
+      "loss": 4.9309,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7316594743036485,
+      "grad_norm": 16.581857681274414,
+      "learning_rate": 9.230971330596591e-06,
+      "loss": 4.7896,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7326402510788544,
+      "grad_norm": 23.697467803955078,
+      "learning_rate": 9.228907581992086e-06,
+      "loss": 5.0056,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7336210278540605,
+      "grad_norm": 14.382548332214355,
+      "learning_rate": 9.226841299319846e-06,
+      "loss": 5.1846,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7346018046292664,
+      "grad_norm": 16.23896598815918,
+      "learning_rate": 9.22477248381804e-06,
+      "loss": 5.2869,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7355825814044723,
+      "grad_norm": 16.80748748779297,
+      "learning_rate": 9.222701136726352e-06,
+      "loss": 4.9048,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7365633581796783,
+      "grad_norm": 19.622159957885742,
+      "learning_rate": 9.22062725928599e-06,
+      "loss": 5.2714,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7375441349548842,
+      "grad_norm": 15.167123794555664,
+      "learning_rate": 9.218550852739669e-06,
+      "loss": 5.358,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7385249117300903,
+      "grad_norm": 17.994897842407227,
+      "learning_rate": 9.216471918331625e-06,
+      "loss": 4.9097,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7395056885052962,
+      "grad_norm": 15.874114036560059,
+      "learning_rate": 9.214390457307607e-06,
+      "loss": 4.8598,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7404864652805021,
+      "grad_norm": 18.701860427856445,
+      "learning_rate": 9.212306470914882e-06,
+      "loss": 5.1611,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7414672420557081,
+      "grad_norm": 23.78780746459961,
+      "learning_rate": 9.210219960402223e-06,
+      "loss": 4.8143,
+      "step": 3780
+    },
+    {
+      "epoch": 0.742448018830914,
+      "grad_norm": 27.825786590576172,
+      "learning_rate": 9.208130927019923e-06,
+      "loss": 4.8035,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7434287956061201,
+      "grad_norm": 29.78475570678711,
+      "learning_rate": 9.206039372019779e-06,
+      "loss": 5.0749,
+      "step": 3790
+    },
+    {
+      "epoch": 0.744409572381326,
+      "grad_norm": 38.66997146606445,
+      "learning_rate": 9.203945296655109e-06,
+      "loss": 5.2162,
+      "step": 3795
+    },
+    {
+      "epoch": 0.745390349156532,
+      "grad_norm": 14.391605377197266,
+      "learning_rate": 9.201848702180732e-06,
+      "loss": 5.4503,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7463711259317379,
+      "grad_norm": 25.25947380065918,
+      "learning_rate": 9.19974958985298e-06,
+      "loss": 5.092,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7473519027069438,
+      "grad_norm": 23.04009437561035,
+      "learning_rate": 9.197647960929697e-06,
+      "loss": 5.103,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7483326794821499,
+      "grad_norm": 17.638652801513672,
+      "learning_rate": 9.195543816670228e-06,
+      "loss": 5.0807,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7493134562573558,
+      "grad_norm": 25.059757232666016,
+      "learning_rate": 9.19343715833543e-06,
+      "loss": 4.9474,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7502942330325618,
+      "grad_norm": 21.401994705200195,
+      "learning_rate": 9.191327987187667e-06,
+      "loss": 4.66,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7502942330325618,
+      "eval_loss": 5.044307231903076,
+      "eval_runtime": 7.6237,
+      "eval_samples_per_second": 27.414,
+      "eval_steps_per_second": 13.773,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7512750098077677,
+      "grad_norm": 16.83993911743164,
+      "learning_rate": 9.189216304490806e-06,
+      "loss": 5.4801,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7522557865829738,
+      "grad_norm": 28.338539123535156,
+      "learning_rate": 9.187102111510223e-06,
+      "loss": 5.0957,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7532365633581797,
+      "grad_norm": 16.17990493774414,
+      "learning_rate": 9.184985409512793e-06,
+      "loss": 5.1511,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7542173401333856,
+      "grad_norm": 24.36326026916504,
+      "learning_rate": 9.182866199766898e-06,
+      "loss": 5.1696,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7551981169085916,
+      "grad_norm": 20.284696578979492,
+      "learning_rate": 9.180744483542421e-06,
+      "loss": 5.2461,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7561788936837975,
+      "grad_norm": 14.811017990112305,
+      "learning_rate": 9.178620262110748e-06,
+      "loss": 4.868,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7571596704590036,
+      "grad_norm": 26.946754455566406,
+      "learning_rate": 9.176493536744767e-06,
+      "loss": 5.244,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7581404472342095,
+      "grad_norm": 26.35390281677246,
+      "learning_rate": 9.174364308718862e-06,
+      "loss": 4.9377,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7591212240094155,
+      "grad_norm": 42.68120574951172,
+      "learning_rate": 9.172232579308924e-06,
+      "loss": 5.5446,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7601020007846214,
+      "grad_norm": 16.318851470947266,
+      "learning_rate": 9.170098349792339e-06,
+      "loss": 5.2302,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7610827775598273,
+      "grad_norm": 19.897960662841797,
+      "learning_rate": 9.167961621447984e-06,
+      "loss": 4.8648,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7620635543350334,
+      "grad_norm": 32.09339904785156,
+      "learning_rate": 9.16582239555625e-06,
+      "loss": 4.791,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7630443311102393,
+      "grad_norm": 25.286733627319336,
+      "learning_rate": 9.16368067339901e-06,
+      "loss": 5.1848,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7640251078854453,
+      "grad_norm": 30.160734176635742,
+      "learning_rate": 9.161536456259637e-06,
+      "loss": 5.0082,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7650058846606512,
+      "grad_norm": 27.863746643066406,
+      "learning_rate": 9.159389745423003e-06,
+      "loss": 5.1531,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7659866614358573,
+      "grad_norm": 20.288660049438477,
+      "learning_rate": 9.157240542175468e-06,
+      "loss": 4.9016,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7669674382110632,
+      "grad_norm": 13.439239501953125,
+      "learning_rate": 9.155088847804888e-06,
+      "loss": 5.1312,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7679482149862691,
+      "grad_norm": 31.238624572753906,
+      "learning_rate": 9.152934663600615e-06,
+      "loss": 5.3148,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7689289917614751,
+      "grad_norm": 20.205110549926758,
+      "learning_rate": 9.15077799085349e-06,
+      "loss": 4.9455,
+      "step": 3920
+    },
+    {
+      "epoch": 0.769909768536681,
+      "grad_norm": 29.910932540893555,
+      "learning_rate": 9.148618830855846e-06,
+      "loss": 5.4219,
+      "step": 3925
+    },
+    {
+      "epoch": 0.770890545311887,
+      "grad_norm": 17.717374801635742,
+      "learning_rate": 9.146457184901502e-06,
+      "loss": 5.3079,
+      "step": 3930
+    },
+    {
+      "epoch": 0.771871322087093,
+      "grad_norm": 14.105619430541992,
+      "learning_rate": 9.144293054285776e-06,
+      "loss": 5.173,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7728520988622989,
+      "grad_norm": 18.493331909179688,
+      "learning_rate": 9.142126440305466e-06,
+      "loss": 4.9215,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7738328756375049,
+      "grad_norm": 17.42450523376465,
+      "learning_rate": 9.139957344258863e-06,
+      "loss": 5.0038,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7748136524127108,
+      "grad_norm": 37.434608459472656,
+      "learning_rate": 9.137785767445743e-06,
+      "loss": 4.7231,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7757944291879169,
+      "grad_norm": 18.50735092163086,
+      "learning_rate": 9.135611711167371e-06,
+      "loss": 4.792,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7767752059631228,
+      "grad_norm": 18.376445770263672,
+      "learning_rate": 9.133435176726494e-06,
+      "loss": 4.7746,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7777559827383288,
+      "grad_norm": 13.254705429077148,
+      "learning_rate": 9.131256165427347e-06,
+      "loss": 5.318,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7787367595135347,
+      "grad_norm": 14.707402229309082,
+      "learning_rate": 9.129074678575649e-06,
+      "loss": 5.0746,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7797175362887406,
+      "grad_norm": 17.6321964263916,
+      "learning_rate": 9.1268907174786e-06,
+      "loss": 5.022,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7806983130639467,
+      "grad_norm": 19.831993103027344,
+      "learning_rate": 9.124704283444887e-06,
+      "loss": 4.9905,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7816790898391526,
+      "grad_norm": 19.51742935180664,
+      "learning_rate": 9.122515377784676e-06,
+      "loss": 5.1026,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7826598666143586,
+      "grad_norm": 13.889021873474121,
+      "learning_rate": 9.12032400180961e-06,
+      "loss": 4.7532,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7836406433895645,
+      "grad_norm": 20.075048446655273,
+      "learning_rate": 9.118130156832823e-06,
+      "loss": 4.9112,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7846214201647705,
+      "grad_norm": 43.44004821777344,
+      "learning_rate": 9.115933844168918e-06,
+      "loss": 5.2201,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7856021969399765,
+      "grad_norm": 37.11042404174805,
+      "learning_rate": 9.11373506513398e-06,
+      "loss": 5.1144,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7865829737151824,
+      "grad_norm": 23.289901733398438,
+      "learning_rate": 9.111533821045576e-06,
+      "loss": 4.9628,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7875637504903884,
+      "grad_norm": 12.136765480041504,
+      "learning_rate": 9.109330113222745e-06,
+      "loss": 4.9547,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7885445272655943,
+      "grad_norm": 21.204288482666016,
+      "learning_rate": 9.107123942986003e-06,
+      "loss": 4.8932,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7895253040408003,
+      "grad_norm": 19.411935806274414,
+      "learning_rate": 9.104915311657346e-06,
+      "loss": 4.7352,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7905060808160063,
+      "grad_norm": 19.139673233032227,
+      "learning_rate": 9.102704220560237e-06,
+      "loss": 5.011,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7914868575912123,
+      "grad_norm": 22.491958618164062,
+      "learning_rate": 9.10049067101962e-06,
+      "loss": 5.1561,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7924676343664182,
+      "grad_norm": 23.593151092529297,
+      "learning_rate": 9.09827466436191e-06,
+      "loss": 5.3284,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7934484111416241,
+      "grad_norm": 25.913562774658203,
+      "learning_rate": 9.096056201914993e-06,
+      "loss": 4.8373,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7944291879168301,
+      "grad_norm": 19.09825897216797,
+      "learning_rate": 9.093835285008228e-06,
+      "loss": 5.422,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7954099646920361,
+      "grad_norm": 16.97250747680664,
+      "learning_rate": 9.091611914972443e-06,
+      "loss": 5.1956,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7963907414672421,
+      "grad_norm": 14.669245719909668,
+      "learning_rate": 9.089386093139937e-06,
+      "loss": 4.7867,
+      "step": 4060
+    },
+    {
+      "epoch": 0.797371518242448,
+      "grad_norm": 38.453433990478516,
+      "learning_rate": 9.087157820844482e-06,
+      "loss": 5.1385,
+      "step": 4065
+    },
+    {
+      "epoch": 0.798352295017654,
+      "grad_norm": 13.812004089355469,
+      "learning_rate": 9.08492709942131e-06,
+      "loss": 5.0448,
+      "step": 4070
+    },
+    {
+      "epoch": 0.79933307179286,
+      "grad_norm": 15.264069557189941,
+      "learning_rate": 9.082693930207128e-06,
+      "loss": 5.0391,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8003138485680659,
+      "grad_norm": 28.84646224975586,
+      "learning_rate": 9.080458314540107e-06,
+      "loss": 4.9142,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8012946253432719,
+      "grad_norm": 53.37618637084961,
+      "learning_rate": 9.078220253759884e-06,
+      "loss": 4.6441,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8022754021184778,
+      "grad_norm": 14.650290489196777,
+      "learning_rate": 9.07597974920756e-06,
+      "loss": 4.9577,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8032561788936838,
+      "grad_norm": 20.597881317138672,
+      "learning_rate": 9.073736802225705e-06,
+      "loss": 5.203,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8042369556688898,
+      "grad_norm": 10.912410736083984,
+      "learning_rate": 9.071491414158345e-06,
+      "loss": 5.1236,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8052177324440957,
+      "grad_norm": 19.348920822143555,
+      "learning_rate": 9.069243586350976e-06,
+      "loss": 4.9828,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8061985092193017,
+      "grad_norm": 12.223658561706543,
+      "learning_rate": 9.066993320150552e-06,
+      "loss": 5.5052,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8071792859945076,
+      "grad_norm": 19.24483299255371,
+      "learning_rate": 9.064740616905487e-06,
+      "loss": 5.2803,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8081600627697136,
+      "grad_norm": 18.650951385498047,
+      "learning_rate": 9.062485477965661e-06,
+      "loss": 5.1652,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8091408395449196,
+      "grad_norm": 25.61029052734375,
+      "learning_rate": 9.060227904682408e-06,
+      "loss": 4.7679,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8101216163201256,
+      "grad_norm": 16.603193283081055,
+      "learning_rate": 9.057967898408523e-06,
+      "loss": 4.8387,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8111023930953315,
+      "grad_norm": 14.41788101196289,
+      "learning_rate": 9.055705460498258e-06,
+      "loss": 5.0304,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8120831698705374,
+      "grad_norm": 17.828062057495117,
+      "learning_rate": 9.053440592307322e-06,
+      "loss": 5.3152,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8130639466457434,
+      "grad_norm": 15.405653953552246,
+      "learning_rate": 9.051173295192885e-06,
+      "loss": 4.9053,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8140447234209494,
+      "grad_norm": 18.29840087890625,
+      "learning_rate": 9.048903570513565e-06,
+      "loss": 4.9139,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8150255001961554,
+      "grad_norm": 10.129656791687012,
+      "learning_rate": 9.046631419629438e-06,
+      "loss": 4.8187,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8160062769713613,
+      "grad_norm": 21.516311645507812,
+      "learning_rate": 9.044356843902036e-06,
+      "loss": 4.8428,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8169870537465673,
+      "grad_norm": 15.47047233581543,
+      "learning_rate": 9.042079844694339e-06,
+      "loss": 4.984,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8179678305217732,
+      "grad_norm": 21.687891006469727,
+      "learning_rate": 9.039800423370783e-06,
+      "loss": 5.1401,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8189486072969792,
+      "grad_norm": 28.038644790649414,
+      "learning_rate": 9.037518581297257e-06,
+      "loss": 5.1431,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8199293840721852,
+      "grad_norm": 30.172744750976562,
+      "learning_rate": 9.035234319841095e-06,
+      "loss": 4.8415,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8209101608473911,
+      "grad_norm": 16.931224822998047,
+      "learning_rate": 9.032947640371086e-06,
+      "loss": 4.5958,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8218909376225971,
+      "grad_norm": 23.898576736450195,
+      "learning_rate": 9.030658544257466e-06,
+      "loss": 4.7125,
+      "step": 4190
+    },
+    {
+      "epoch": 0.822871714397803,
+      "grad_norm": 12.608359336853027,
+      "learning_rate": 9.028367032871917e-06,
+      "loss": 5.1428,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8238524911730091,
+      "grad_norm": 12.2660493850708,
+      "learning_rate": 9.026073107587571e-06,
+      "loss": 5.0363,
+      "step": 4200
+    },
+    {
+      "epoch": 0.824833267948215,
+      "grad_norm": 10.48144245147705,
+      "learning_rate": 9.023776769779007e-06,
+      "loss": 5.0101,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8258140447234209,
+      "grad_norm": 26.566226959228516,
+      "learning_rate": 9.021478020822248e-06,
+      "loss": 5.0142,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8267948214986269,
+      "grad_norm": 11.696605682373047,
+      "learning_rate": 9.01917686209476e-06,
+      "loss": 5.2014,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8277755982738328,
+      "grad_norm": 28.729278564453125,
+      "learning_rate": 9.016873294975457e-06,
+      "loss": 5.2874,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8287563750490389,
+      "grad_norm": 21.497310638427734,
+      "learning_rate": 9.014567320844694e-06,
+      "loss": 5.1681,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8297371518242448,
+      "grad_norm": 13.806232452392578,
+      "learning_rate": 9.012258941084269e-06,
+      "loss": 4.664,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8307179285994508,
+      "grad_norm": 15.800619125366211,
+      "learning_rate": 9.009948157077421e-06,
+      "loss": 5.1314,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8316987053746567,
+      "grad_norm": 12.598173141479492,
+      "learning_rate": 9.007634970208829e-06,
+      "loss": 5.0503,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8326794821498626,
+      "grad_norm": 19.550113677978516,
+      "learning_rate": 9.005319381864615e-06,
+      "loss": 5.059,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8336602589250687,
+      "grad_norm": 11.087757110595703,
+      "learning_rate": 9.003001393432334e-06,
+      "loss": 4.9414,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8346410357002746,
+      "grad_norm": 11.731034278869629,
+      "learning_rate": 9.000681006300986e-06,
+      "loss": 4.8664,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8356218124754806,
+      "grad_norm": 21.4744873046875,
+      "learning_rate": 8.998358221861006e-06,
+      "loss": 5.2014,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8366025892506865,
+      "grad_norm": 14.924543380737305,
+      "learning_rate": 8.996033041504262e-06,
+      "loss": 4.8598,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8375833660258926,
+      "grad_norm": 21.695417404174805,
+      "learning_rate": 8.993705466624061e-06,
+      "loss": 4.8761,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8385641428010985,
+      "grad_norm": 22.94925880432129,
+      "learning_rate": 8.991375498615147e-06,
+      "loss": 5.0729,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8395449195763044,
+      "grad_norm": 24.85720443725586,
+      "learning_rate": 8.98904313887369e-06,
+      "loss": 4.8534,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8405256963515104,
+      "grad_norm": 15.123157501220703,
+      "learning_rate": 8.986708388797306e-06,
+      "loss": 4.7483,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8415064731267163,
+      "grad_norm": 17.581621170043945,
+      "learning_rate": 8.984371249785031e-06,
+      "loss": 4.9156,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8424872499019224,
+      "grad_norm": 14.177846908569336,
+      "learning_rate": 8.982031723237338e-06,
+      "loss": 5.0995,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8434680266771283,
+      "grad_norm": 12.062125205993652,
+      "learning_rate": 8.979689810556132e-06,
+      "loss": 4.9483,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8444488034523342,
+      "grad_norm": 21.800048828125,
+      "learning_rate": 8.977345513144743e-06,
+      "loss": 5.163,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8454295802275402,
+      "grad_norm": 13.504755020141602,
+      "learning_rate": 8.974998832407935e-06,
+      "loss": 5.4037,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8464103570027461,
+      "grad_norm": 31.20657730102539,
+      "learning_rate": 8.972649769751897e-06,
+      "loss": 5.3662,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8473911337779522,
+      "grad_norm": 17.510353088378906,
+      "learning_rate": 8.97029832658425e-06,
+      "loss": 5.0793,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8483719105531581,
+      "grad_norm": 35.864559173583984,
+      "learning_rate": 8.967944504314033e-06,
+      "loss": 5.2564,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8493526873283641,
+      "grad_norm": 26.920175552368164,
+      "learning_rate": 8.965588304351716e-06,
+      "loss": 4.9264,
+      "step": 4330
+    },
+    {
+      "epoch": 0.85033346410357,
+      "grad_norm": 12.373115539550781,
+      "learning_rate": 8.963229728109196e-06,
+      "loss": 4.8945,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8513142408787759,
+      "grad_norm": 19.75405502319336,
+      "learning_rate": 8.96086877699979e-06,
+      "loss": 5.0325,
+      "step": 4340
+    },
+    {
+      "epoch": 0.852295017653982,
+      "grad_norm": 25.117504119873047,
+      "learning_rate": 8.95850545243824e-06,
+      "loss": 4.9604,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8532757944291879,
+      "grad_norm": 14.683432579040527,
+      "learning_rate": 8.956139755840706e-06,
+      "loss": 4.8917,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8542565712043939,
+      "grad_norm": 10.729886054992676,
+      "learning_rate": 8.953771688624777e-06,
+      "loss": 4.9119,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8552373479795998,
+      "grad_norm": 18.196348190307617,
+      "learning_rate": 8.951401252209457e-06,
+      "loss": 4.8839,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8562181247548059,
+      "grad_norm": 13.992422103881836,
+      "learning_rate": 8.94902844801517e-06,
+      "loss": 4.9166,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8571989015300118,
+      "grad_norm": 17.236942291259766,
+      "learning_rate": 8.946653277463763e-06,
+      "loss": 4.702,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8581796783052177,
+      "grad_norm": 16.63797950744629,
+      "learning_rate": 8.944275741978495e-06,
+      "loss": 4.9513,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8591604550804237,
+      "grad_norm": 11.08305549621582,
+      "learning_rate": 8.941895842984045e-06,
+      "loss": 5.0286,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8601412318556296,
+      "grad_norm": 20.1209774017334,
+      "learning_rate": 8.939513581906509e-06,
+      "loss": 4.9298,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8611220086308357,
+      "grad_norm": 25.10521125793457,
+      "learning_rate": 8.937128960173399e-06,
+      "loss": 5.3603,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8621027854060416,
+      "grad_norm": 24.15997314453125,
+      "learning_rate": 8.934741979213638e-06,
+      "loss": 5.0146,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8630835621812476,
+      "grad_norm": 21.98234748840332,
+      "learning_rate": 8.932352640457566e-06,
+      "loss": 5.3546,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8640643389564535,
+      "grad_norm": 34.10669708251953,
+      "learning_rate": 8.929960945336936e-06,
+      "loss": 5.2731,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8650451157316594,
+      "grad_norm": 13.953348159790039,
+      "learning_rate": 8.927566895284912e-06,
+      "loss": 5.4325,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8660258925068655,
+      "grad_norm": 32.35000228881836,
+      "learning_rate": 8.925170491736065e-06,
+      "loss": 4.8432,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8670066692820714,
+      "grad_norm": 11.26505184173584,
+      "learning_rate": 8.922771736126384e-06,
+      "loss": 5.1536,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8679874460572774,
+      "grad_norm": 18.455360412597656,
+      "learning_rate": 8.920370629893263e-06,
+      "loss": 5.0085,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8689682228324833,
+      "grad_norm": 15.50165843963623,
+      "learning_rate": 8.917967174475505e-06,
+      "loss": 4.9807,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8699489996076893,
+      "grad_norm": 12.871978759765625,
+      "learning_rate": 8.915561371313321e-06,
+      "loss": 5.2293,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8709297763828953,
+      "grad_norm": 23.6169490814209,
+      "learning_rate": 8.913153221848328e-06,
+      "loss": 5.1125,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8719105531581012,
+      "grad_norm": 32.19282913208008,
+      "learning_rate": 8.91074272752355e-06,
+      "loss": 4.9331,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8728913299333072,
+      "grad_norm": 21.433151245117188,
+      "learning_rate": 8.908329889783418e-06,
+      "loss": 4.9661,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8738721067085131,
+      "grad_norm": 24.601730346679688,
+      "learning_rate": 8.905914710073761e-06,
+      "loss": 4.9997,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8748528834837191,
+      "grad_norm": 27.876657485961914,
+      "learning_rate": 8.903497189841819e-06,
+      "loss": 5.1279,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8758336602589251,
+      "grad_norm": 20.576623916625977,
+      "learning_rate": 8.901077330536228e-06,
+      "loss": 5.1912,
+      "step": 4465
+    },
+    {
+      "epoch": 0.876814437034131,
+      "grad_norm": 20.63385009765625,
+      "learning_rate": 8.89865513360703e-06,
+      "loss": 4.8031,
+      "step": 4470
+    },
+    {
+      "epoch": 0.877795213809337,
+      "grad_norm": 35.024566650390625,
+      "learning_rate": 8.896230600505668e-06,
+      "loss": 5.1266,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8787759905845429,
+      "grad_norm": 21.97711181640625,
+      "learning_rate": 8.893803732684981e-06,
+      "loss": 4.8322,
+      "step": 4480
+    },
+    {
+      "epoch": 0.879756767359749,
+      "grad_norm": 50.84933853149414,
+      "learning_rate": 8.891374531599209e-06,
+      "loss": 5.0662,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8807375441349549,
+      "grad_norm": 17.233617782592773,
+      "learning_rate": 8.88894299870399e-06,
+      "loss": 4.7075,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8817183209101609,
+      "grad_norm": 22.550418853759766,
+      "learning_rate": 8.886509135456362e-06,
+      "loss": 5.4782,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8826990976853668,
+      "grad_norm": 11.513266563415527,
+      "learning_rate": 8.884072943314754e-06,
+      "loss": 5.2502,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8836798744605727,
+      "grad_norm": 40.44765090942383,
+      "learning_rate": 8.881634423738995e-06,
+      "loss": 4.8474,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8846606512357787,
+      "grad_norm": 20.928241729736328,
+      "learning_rate": 8.879193578190311e-06,
+      "loss": 5.2969,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8856414280109847,
+      "grad_norm": 28.95954132080078,
+      "learning_rate": 8.876750408131312e-06,
+      "loss": 4.937,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8866222047861907,
+      "grad_norm": 16.0018253326416,
+      "learning_rate": 8.874304915026012e-06,
+      "loss": 5.0333,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8876029815613966,
+      "grad_norm": 29.66618537902832,
+      "learning_rate": 8.871857100339805e-06,
+      "loss": 5.0874,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8885837583366026,
+      "grad_norm": 21.657180786132812,
+      "learning_rate": 8.869406965539489e-06,
+      "loss": 4.7309,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8895645351118086,
+      "grad_norm": 15.75297737121582,
+      "learning_rate": 8.866954512093246e-06,
+      "loss": 5.0516,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8905453118870145,
+      "grad_norm": 11.395059585571289,
+      "learning_rate": 8.864499741470646e-06,
+      "loss": 4.8039,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8915260886622205,
+      "grad_norm": 20.670623779296875,
+      "learning_rate": 8.86204265514265e-06,
+      "loss": 5.0751,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8925068654374264,
+      "grad_norm": 19.001562118530273,
+      "learning_rate": 8.859583254581604e-06,
+      "loss": 4.948,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8934876422126324,
+      "grad_norm": 17.733688354492188,
+      "learning_rate": 8.857121541261247e-06,
+      "loss": 4.8752,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8944684189878384,
+      "grad_norm": 28.683687210083008,
+      "learning_rate": 8.854657516656697e-06,
+      "loss": 5.201,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8954491957630444,
+      "grad_norm": 18.13066291809082,
+      "learning_rate": 8.852191182244456e-06,
+      "loss": 5.0124,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8964299725382503,
+      "grad_norm": 27.07872200012207,
+      "learning_rate": 8.849722539502419e-06,
+      "loss": 5.3658,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8974107493134562,
+      "grad_norm": 21.85537338256836,
+      "learning_rate": 8.847251589909857e-06,
+      "loss": 5.3605,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8983915260886622,
+      "grad_norm": 21.174556732177734,
+      "learning_rate": 8.844778334947426e-06,
+      "loss": 4.5005,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8993723028638682,
+      "grad_norm": 19.17597007751465,
+      "learning_rate": 8.84230277609716e-06,
+      "loss": 5.2061,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9003530796390742,
+      "grad_norm": 20.14710235595703,
+      "learning_rate": 8.839824914842477e-06,
+      "loss": 5.0189,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9013338564142801,
+      "grad_norm": 13.179183006286621,
+      "learning_rate": 8.837344752668176e-06,
+      "loss": 5.2221,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9023146331894861,
+      "grad_norm": 22.216659545898438,
+      "learning_rate": 8.83486229106043e-06,
+      "loss": 4.9365,
+      "step": 4600
+    },
+    {
+      "epoch": 0.903295409964692,
+      "grad_norm": 26.944229125976562,
+      "learning_rate": 8.832377531506794e-06,
+      "loss": 5.4394,
+      "step": 4605
+    },
+    {
+      "epoch": 0.904276186739898,
+      "grad_norm": 21.65399742126465,
+      "learning_rate": 8.829890475496195e-06,
+      "loss": 5.0546,
+      "step": 4610
+    },
+    {
+      "epoch": 0.905256963515104,
+      "grad_norm": 19.045804977416992,
+      "learning_rate": 8.827401124518945e-06,
+      "loss": 5.0907,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9062377402903099,
+      "grad_norm": 11.208507537841797,
+      "learning_rate": 8.82490948006672e-06,
+      "loss": 4.7015,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9072185170655159,
+      "grad_norm": 26.1632022857666,
+      "learning_rate": 8.82241554363258e-06,
+      "loss": 5.1097,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9081992938407218,
+      "grad_norm": 34.274539947509766,
+      "learning_rate": 8.819919316710954e-06,
+      "loss": 5.1715,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9091800706159278,
+      "grad_norm": 14.437838554382324,
+      "learning_rate": 8.817420800797641e-06,
+      "loss": 4.8171,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9101608473911338,
+      "grad_norm": 19.037561416625977,
+      "learning_rate": 8.814919997389818e-06,
+      "loss": 4.8992,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9111416241663397,
+      "grad_norm": 18.625173568725586,
+      "learning_rate": 8.812416907986027e-06,
+      "loss": 5.0305,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9121224009415457,
+      "grad_norm": 17.157176971435547,
+      "learning_rate": 8.809911534086185e-06,
+      "loss": 4.582,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9131031777167516,
+      "grad_norm": 26.29932975769043,
+      "learning_rate": 8.807403877191572e-06,
+      "loss": 5.138,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9140839544919577,
+      "grad_norm": 20.871774673461914,
+      "learning_rate": 8.804893938804839e-06,
+      "loss": 4.9286,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9150647312671636,
+      "grad_norm": 17.02748680114746,
+      "learning_rate": 8.802381720430006e-06,
+      "loss": 4.9197,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9160455080423695,
+      "grad_norm": 13.891420364379883,
+      "learning_rate": 8.799867223572457e-06,
+      "loss": 5.003,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9170262848175755,
+      "grad_norm": 20.51250457763672,
+      "learning_rate": 8.797350449738941e-06,
+      "loss": 4.9468,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9180070615927814,
+      "grad_norm": 16.219396591186523,
+      "learning_rate": 8.794831400437573e-06,
+      "loss": 4.8844,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9189878383679875,
+      "grad_norm": 23.17514419555664,
+      "learning_rate": 8.79231007717783e-06,
+      "loss": 4.654,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9199686151431934,
+      "grad_norm": 19.793642044067383,
+      "learning_rate": 8.789786481470553e-06,
+      "loss": 4.9788,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9209493919183994,
+      "grad_norm": 17.20256233215332,
+      "learning_rate": 8.787260614827942e-06,
+      "loss": 5.4916,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9219301686936053,
+      "grad_norm": 21.612836837768555,
+      "learning_rate": 8.784732478763562e-06,
+      "loss": 5.2886,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9229109454688113,
+      "grad_norm": 23.609474182128906,
+      "learning_rate": 8.782202074792336e-06,
+      "loss": 4.7733,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9238917222440173,
+      "grad_norm": 25.944740295410156,
+      "learning_rate": 8.779669404430545e-06,
+      "loss": 5.068,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9248724990192232,
+      "grad_norm": 13.784878730773926,
+      "learning_rate": 8.777134469195826e-06,
+      "loss": 4.8825,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9258532757944292,
+      "grad_norm": 21.367332458496094,
+      "learning_rate": 8.77459727060718e-06,
+      "loss": 4.8307,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9268340525696351,
+      "grad_norm": 21.81898307800293,
+      "learning_rate": 8.772057810184957e-06,
+      "loss": 5.2136,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9278148293448412,
+      "grad_norm": 13.044177055358887,
+      "learning_rate": 8.769516089450869e-06,
+      "loss": 4.8561,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9287956061200471,
+      "grad_norm": 21.14825439453125,
+      "learning_rate": 8.766972109927976e-06,
+      "loss": 4.9493,
+      "step": 4735
+    },
+    {
+      "epoch": 0.929776382895253,
+      "grad_norm": 21.38848304748535,
+      "learning_rate": 8.764425873140693e-06,
+      "loss": 5.1738,
+      "step": 4740
+    },
+    {
+      "epoch": 0.930757159670459,
+      "grad_norm": 18.144960403442383,
+      "learning_rate": 8.761877380614796e-06,
+      "loss": 4.8048,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9317379364456649,
+      "grad_norm": 23.306209564208984,
+      "learning_rate": 8.759326633877398e-06,
+      "loss": 4.4888,
+      "step": 4750
+    },
+    {
+      "epoch": 0.932718713220871,
+      "grad_norm": 19.98696517944336,
+      "learning_rate": 8.756773634456975e-06,
+      "loss": 4.7224,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9336994899960769,
+      "grad_norm": 12.257347106933594,
+      "learning_rate": 8.754218383883349e-06,
+      "loss": 4.9588,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9346802667712829,
+      "grad_norm": 12.258370399475098,
+      "learning_rate": 8.751660883687685e-06,
+      "loss": 5.0761,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9356610435464888,
+      "grad_norm": 16.816762924194336,
+      "learning_rate": 8.749101135402508e-06,
+      "loss": 5.1723,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9366418203216947,
+      "grad_norm": 24.14014434814453,
+      "learning_rate": 8.74653914056168e-06,
+      "loss": 4.8346,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9376225970969008,
+      "grad_norm": 18.49698829650879,
+      "learning_rate": 8.743974900700415e-06,
+      "loss": 4.8741,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9386033738721067,
+      "grad_norm": 17.863910675048828,
+      "learning_rate": 8.741408417355264e-06,
+      "loss": 5.0614,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9395841506473127,
+      "grad_norm": 20.26540184020996,
+      "learning_rate": 8.738839692064136e-06,
+      "loss": 4.7714,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9405649274225186,
+      "grad_norm": 19.46352195739746,
+      "learning_rate": 8.736268726366272e-06,
+      "loss": 5.0038,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9415457041977247,
+      "grad_norm": 17.201187133789062,
+      "learning_rate": 8.733695521802259e-06,
+      "loss": 5.0616,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9425264809729306,
+      "grad_norm": 17.212223052978516,
+      "learning_rate": 8.731120079914026e-06,
+      "loss": 4.7626,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9435072577481365,
+      "grad_norm": 27.876567840576172,
+      "learning_rate": 8.728542402244847e-06,
+      "loss": 5.2046,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9444880345233425,
+      "grad_norm": 16.261171340942383,
+      "learning_rate": 8.725962490339323e-06,
+      "loss": 4.6926,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9454688112985484,
+      "grad_norm": 15.20531177520752,
+      "learning_rate": 8.723380345743408e-06,
+      "loss": 4.8294,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9464495880737545,
+      "grad_norm": 23.71307945251465,
+      "learning_rate": 8.720795970004385e-06,
+      "loss": 4.9227,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9474303648489604,
+      "grad_norm": 23.945533752441406,
+      "learning_rate": 8.718209364670881e-06,
+      "loss": 5.343,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9484111416241663,
+      "grad_norm": 32.02574920654297,
+      "learning_rate": 8.71562053129285e-06,
+      "loss": 5.208,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9493919183993723,
+      "grad_norm": 29.752044677734375,
+      "learning_rate": 8.71302947142159e-06,
+      "loss": 4.6208,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9503726951745782,
+      "grad_norm": 10.127486228942871,
+      "learning_rate": 8.710436186609728e-06,
+      "loss": 4.8635,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9513534719497843,
+      "grad_norm": 14.631573677062988,
+      "learning_rate": 8.707840678411223e-06,
+      "loss": 4.8818,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9523342487249902,
+      "grad_norm": 42.126060485839844,
+      "learning_rate": 8.705242948381372e-06,
+      "loss": 5.2032,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9533150255001962,
+      "grad_norm": 21.709590911865234,
+      "learning_rate": 8.702642998076798e-06,
+      "loss": 4.9076,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9542958022754021,
+      "grad_norm": 19.50485610961914,
+      "learning_rate": 8.700040829055458e-06,
+      "loss": 4.8002,
+      "step": 4865
+    },
+    {
+      "epoch": 0.955276579050608,
+      "grad_norm": 17.49419593811035,
+      "learning_rate": 8.697436442876637e-06,
+      "loss": 4.8608,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9562573558258141,
+      "grad_norm": 20.620445251464844,
+      "learning_rate": 8.694829841100946e-06,
+      "loss": 5.0148,
+      "step": 4875
+    },
+    {
+      "epoch": 0.95723813260102,
+      "grad_norm": 12.148300170898438,
+      "learning_rate": 8.69222102529033e-06,
+      "loss": 4.8127,
+      "step": 4880
+    },
+    {
+      "epoch": 0.958218909376226,
+      "grad_norm": 21.182106018066406,
+      "learning_rate": 8.689609997008057e-06,
+      "loss": 5.0137,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9591996861514319,
+      "grad_norm": 11.576159477233887,
+      "learning_rate": 8.686996757818718e-06,
+      "loss": 5.0396,
+      "step": 4890
+    },
+    {
+      "epoch": 0.960180462926638,
+      "grad_norm": 11.674856185913086,
+      "learning_rate": 8.684381309288232e-06,
+      "loss": 4.742,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9611612397018439,
+      "grad_norm": 26.99156951904297,
+      "learning_rate": 8.681763652983846e-06,
+      "loss": 5.0905,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9621420164770498,
+      "grad_norm": 24.627355575561523,
+      "learning_rate": 8.679143790474119e-06,
+      "loss": 4.873,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9631227932522558,
+      "grad_norm": 21.86775016784668,
+      "learning_rate": 8.676521723328942e-06,
+      "loss": 5.1994,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9641035700274617,
+      "grad_norm": 14.598928451538086,
+      "learning_rate": 8.673897453119521e-06,
+      "loss": 4.8057,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9650843468026677,
+      "grad_norm": 27.489707946777344,
+      "learning_rate": 8.67127098141839e-06,
+      "loss": 5.2588,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9660651235778737,
+      "grad_norm": 10.293922424316406,
+      "learning_rate": 8.66864230979939e-06,
+      "loss": 5.0615,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9670459003530797,
+      "grad_norm": 18.869178771972656,
+      "learning_rate": 8.666011439837694e-06,
+      "loss": 5.2834,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9680266771282856,
+      "grad_norm": 40.34228515625,
+      "learning_rate": 8.66337837310978e-06,
+      "loss": 4.9996,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9690074539034915,
+      "grad_norm": 23.41005516052246,
+      "learning_rate": 8.66074311119345e-06,
+      "loss": 5.3371,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9699882306786975,
+      "grad_norm": 33.91444396972656,
+      "learning_rate": 8.65810565566782e-06,
+      "loss": 5.2898,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9709690074539035,
+      "grad_norm": 26.008817672729492,
+      "learning_rate": 8.655466008113318e-06,
+      "loss": 5.0883,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9719497842291095,
+      "grad_norm": 17.434890747070312,
+      "learning_rate": 8.652824170111689e-06,
+      "loss": 4.8602,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9729305610043154,
+      "grad_norm": 13.335190773010254,
+      "learning_rate": 8.650180143245985e-06,
+      "loss": 4.5868,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9739113377795214,
+      "grad_norm": 21.08846664428711,
+      "learning_rate": 8.647533929100577e-06,
+      "loss": 4.6415,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9748921145547274,
+      "grad_norm": 11.147554397583008,
+      "learning_rate": 8.644885529261144e-06,
+      "loss": 5.104,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9758728913299333,
+      "grad_norm": 18.926776885986328,
+      "learning_rate": 8.642234945314671e-06,
+      "loss": 4.9096,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9768536681051393,
+      "grad_norm": 18.32732582092285,
+      "learning_rate": 8.639582178849454e-06,
+      "loss": 4.8699,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9778344448803452,
+      "grad_norm": 16.150941848754883,
+      "learning_rate": 8.6369272314551e-06,
+      "loss": 4.6607,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9788152216555512,
+      "grad_norm": 24.335636138916016,
+      "learning_rate": 8.634270104722518e-06,
+      "loss": 4.7315,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9797959984307572,
+      "grad_norm": 13.760662078857422,
+      "learning_rate": 8.631610800243926e-06,
+      "loss": 4.9446,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9807767752059631,
+      "grad_norm": 27.17913818359375,
+      "learning_rate": 8.628949319612845e-06,
+      "loss": 5.0926,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9817575519811691,
+      "grad_norm": 30.56397247314453,
+      "learning_rate": 8.626285664424104e-06,
+      "loss": 4.9878,
+      "step": 5005
+    },
+    {
+      "epoch": 0.982738328756375,
+      "grad_norm": 22.218719482421875,
+      "learning_rate": 8.62361983627383e-06,
+      "loss": 4.8556,
+      "step": 5010
+    },
+    {
+      "epoch": 0.983719105531581,
+      "grad_norm": 14.08055305480957,
+      "learning_rate": 8.620951836759454e-06,
+      "loss": 5.0962,
+      "step": 5015
+    },
+    {
+      "epoch": 0.984699882306787,
+      "grad_norm": 35.20744705200195,
+      "learning_rate": 8.61828166747971e-06,
+      "loss": 4.5052,
+      "step": 5020
+    },
+    {
+      "epoch": 0.985680659081993,
+      "grad_norm": 52.820247650146484,
+      "learning_rate": 8.615609330034628e-06,
+      "loss": 5.0339,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9866614358571989,
+      "grad_norm": 22.243247985839844,
+      "learning_rate": 8.612934826025542e-06,
+      "loss": 5.1378,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9876422126324048,
+      "grad_norm": 9.840145111083984,
+      "learning_rate": 8.610258157055082e-06,
+      "loss": 4.9607,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9886229894076108,
+      "grad_norm": 15.050167083740234,
+      "learning_rate": 8.607579324727175e-06,
+      "loss": 5.3641,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9896037661828168,
+      "grad_norm": 18.553813934326172,
+      "learning_rate": 8.604898330647043e-06,
+      "loss": 5.3992,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9905845429580228,
+      "grad_norm": 25.090362548828125,
+      "learning_rate": 8.602215176421206e-06,
+      "loss": 4.8542,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9915653197332287,
+      "grad_norm": 24.53265380859375,
+      "learning_rate": 8.59952986365748e-06,
+      "loss": 4.7045,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9925460965084347,
+      "grad_norm": 13.577052116394043,
+      "learning_rate": 8.59684239396497e-06,
+      "loss": 4.8215,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9935268732836406,
+      "grad_norm": 26.031982421875,
+      "learning_rate": 8.594152768954072e-06,
+      "loss": 5.115,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9945076500588466,
+      "grad_norm": 17.36546516418457,
+      "learning_rate": 8.591460990236482e-06,
+      "loss": 4.911,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9954884268340526,
+      "grad_norm": 15.414812088012695,
+      "learning_rate": 8.58876705942518e-06,
+      "loss": 4.9138,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9964692036092585,
+      "grad_norm": 21.61842155456543,
+      "learning_rate": 8.586070978134437e-06,
+      "loss": 4.7276,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9974499803844645,
+      "grad_norm": 20.003103256225586,
+      "learning_rate": 8.583372747979813e-06,
+      "loss": 4.9446,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9984307571596704,
+      "grad_norm": 18.7646541595459,
+      "learning_rate": 8.580672370578152e-06,
+      "loss": 4.8425,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9994115339348765,
+      "grad_norm": 14.617816925048828,
+      "learning_rate": 8.577969847547591e-06,
+      "loss": 4.7787,
+      "step": 5095
+    },
+    {
+      "epoch": 1.0003923107100823,
+      "grad_norm": 25.93317222595215,
+      "learning_rate": 8.575265180507553e-06,
+      "loss": 5.0419,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0003923107100823,
+      "eval_loss": 4.966672420501709,
+      "eval_runtime": 7.6131,
+      "eval_samples_per_second": 27.453,
+      "eval_steps_per_second": 13.792,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0013730874852884,
+      "grad_norm": 19.052959442138672,
+      "learning_rate": 8.572558371078736e-06,
+      "loss": 4.6995,
+      "step": 5105
+    },
+    {
+      "epoch": 1.0023538642604943,
+      "grad_norm": 24.985578536987305,
+      "learning_rate": 8.56984942088313e-06,
+      "loss": 5.0097,
+      "step": 5110
+    },
+    {
+      "epoch": 1.0033346410357002,
+      "grad_norm": 13.549357414245605,
+      "learning_rate": 8.567138331544009e-06,
+      "loss": 4.813,
+      "step": 5115
+    },
+    {
+      "epoch": 1.0043154178109062,
+      "grad_norm": 24.388721466064453,
+      "learning_rate": 8.564425104685926e-06,
+      "loss": 4.9489,
+      "step": 5120
+    },
+    {
+      "epoch": 1.0052961945861123,
+      "grad_norm": 11.15587043762207,
+      "learning_rate": 8.56170974193471e-06,
+      "loss": 4.6149,
+      "step": 5125
+    },
+    {
+      "epoch": 1.0062769713613182,
+      "grad_norm": 15.220309257507324,
+      "learning_rate": 8.55899224491748e-06,
+      "loss": 4.8196,
+      "step": 5130
+    },
+    {
+      "epoch": 1.0072577481365241,
+      "grad_norm": 15.071609497070312,
+      "learning_rate": 8.556272615262623e-06,
+      "loss": 5.3157,
+      "step": 5135
+    },
+    {
+      "epoch": 1.00823852491173,
+      "grad_norm": 16.934743881225586,
+      "learning_rate": 8.553550854599815e-06,
+      "loss": 4.7028,
+      "step": 5140
+    },
+    {
+      "epoch": 1.009219301686936,
+      "grad_norm": 13.945096015930176,
+      "learning_rate": 8.55082696456e-06,
+      "loss": 4.3538,
+      "step": 5145
+    },
+    {
+      "epoch": 1.010200078462142,
+      "grad_norm": 21.077960968017578,
+      "learning_rate": 8.548100946775402e-06,
+      "loss": 4.5671,
+      "step": 5150
+    },
+    {
+      "epoch": 1.011180855237348,
+      "grad_norm": 23.00332260131836,
+      "learning_rate": 8.54537280287952e-06,
+      "loss": 4.8257,
+      "step": 5155
+    },
+    {
+      "epoch": 1.012161632012554,
+      "grad_norm": 12.397933959960938,
+      "learning_rate": 8.542642534507126e-06,
+      "loss": 4.3358,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0131424087877599,
+      "grad_norm": 12.854864120483398,
+      "learning_rate": 8.539910143294265e-06,
+      "loss": 4.7781,
+      "step": 5165
+    },
+    {
+      "epoch": 1.0141231855629658,
+      "grad_norm": 12.894524574279785,
+      "learning_rate": 8.537175630878256e-06,
+      "loss": 4.6318,
+      "step": 5170
+    },
+    {
+      "epoch": 1.015103962338172,
+      "grad_norm": 11.00051498413086,
+      "learning_rate": 8.534438998897686e-06,
+      "loss": 4.6955,
+      "step": 5175
+    },
+    {
+      "epoch": 1.0160847391133778,
+      "grad_norm": 12.633216857910156,
+      "learning_rate": 8.531700248992414e-06,
+      "loss": 4.6592,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0170655158885837,
+      "grad_norm": 17.211999893188477,
+      "learning_rate": 8.52895938280357e-06,
+      "loss": 4.8464,
+      "step": 5185
+    },
+    {
+      "epoch": 1.0180462926637897,
+      "grad_norm": 19.490310668945312,
+      "learning_rate": 8.526216401973546e-06,
+      "loss": 4.5088,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0190270694389958,
+      "grad_norm": 26.104711532592773,
+      "learning_rate": 8.523471308146007e-06,
+      "loss": 4.7509,
+      "step": 5195
+    },
+    {
+      "epoch": 1.0200078462142017,
+      "grad_norm": 19.656057357788086,
+      "learning_rate": 8.520724102965883e-06,
+      "loss": 5.0214,
+      "step": 5200
+    },
+    {
+      "epoch": 1.0209886229894076,
+      "grad_norm": 13.035916328430176,
+      "learning_rate": 8.517974788079369e-06,
+      "loss": 4.5281,
+      "step": 5205
+    },
+    {
+      "epoch": 1.0219693997646135,
+      "grad_norm": 18.815303802490234,
+      "learning_rate": 8.51522336513392e-06,
+      "loss": 4.6767,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0229501765398195,
+      "grad_norm": 16.87318229675293,
+      "learning_rate": 8.512469835778262e-06,
+      "loss": 4.8249,
+      "step": 5215
+    },
+    {
+      "epoch": 1.0239309533150256,
+      "grad_norm": 26.774776458740234,
+      "learning_rate": 8.509714201662377e-06,
+      "loss": 4.8232,
+      "step": 5220
+    },
+    {
+      "epoch": 1.0249117300902315,
+      "grad_norm": 17.386625289916992,
+      "learning_rate": 8.506956464437509e-06,
+      "loss": 5.3507,
+      "step": 5225
+    },
+    {
+      "epoch": 1.0258925068654374,
+      "grad_norm": 19.190387725830078,
+      "learning_rate": 8.504196625756166e-06,
+      "loss": 4.5408,
+      "step": 5230
+    },
+    {
+      "epoch": 1.0268732836406433,
+      "grad_norm": 22.238534927368164,
+      "learning_rate": 8.50143468727211e-06,
+      "loss": 4.6218,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0278540604158493,
+      "grad_norm": 19.188385009765625,
+      "learning_rate": 8.498670650640368e-06,
+      "loss": 4.8509,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0288348371910554,
+      "grad_norm": 17.50986099243164,
+      "learning_rate": 8.495904517517217e-06,
+      "loss": 4.919,
+      "step": 5245
+    },
+    {
+      "epoch": 1.0298156139662613,
+      "grad_norm": 24.868322372436523,
+      "learning_rate": 8.493136289560194e-06,
+      "loss": 4.8401,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0307963907414672,
+      "grad_norm": 19.065523147583008,
+      "learning_rate": 8.49036596842809e-06,
+      "loss": 4.729,
+      "step": 5255
+    },
+    {
+      "epoch": 1.0317771675166731,
+      "grad_norm": 17.544790267944336,
+      "learning_rate": 8.487593555780954e-06,
+      "loss": 4.4417,
+      "step": 5260
+    },
+    {
+      "epoch": 1.032757944291879,
+      "grad_norm": 13.028230667114258,
+      "learning_rate": 8.484819053280082e-06,
+      "loss": 4.7517,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0337387210670852,
+      "grad_norm": 12.053363800048828,
+      "learning_rate": 8.482042462588028e-06,
+      "loss": 4.6513,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0347194978422911,
+      "grad_norm": 16.837976455688477,
+      "learning_rate": 8.479263785368594e-06,
+      "loss": 4.9133,
+      "step": 5275
+    },
+    {
+      "epoch": 1.035700274617497,
+      "grad_norm": 23.320768356323242,
+      "learning_rate": 8.476483023286832e-06,
+      "loss": 4.7035,
+      "step": 5280
+    },
+    {
+      "epoch": 1.036681051392703,
+      "grad_norm": 14.706212043762207,
+      "learning_rate": 8.473700178009047e-06,
+      "loss": 4.7084,
+      "step": 5285
+    },
+    {
+      "epoch": 1.037661828167909,
+      "grad_norm": 16.379053115844727,
+      "learning_rate": 8.470915251202789e-06,
+      "loss": 4.8806,
+      "step": 5290
+    },
+    {
+      "epoch": 1.038642604943115,
+      "grad_norm": 16.73505210876465,
+      "learning_rate": 8.468128244536854e-06,
+      "loss": 4.5875,
+      "step": 5295
+    },
+    {
+      "epoch": 1.039623381718321,
+      "grad_norm": 13.927240371704102,
+      "learning_rate": 8.465339159681291e-06,
+      "loss": 4.5403,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0406041584935268,
+      "grad_norm": 18.347909927368164,
+      "learning_rate": 8.462547998307386e-06,
+      "loss": 4.9035,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0415849352687327,
+      "grad_norm": 35.21794891357422,
+      "learning_rate": 8.459754762087675e-06,
+      "loss": 5.0395,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0425657120439389,
+      "grad_norm": 9.092461585998535,
+      "learning_rate": 8.456959452695934e-06,
+      "loss": 4.8834,
+      "step": 5315
+    },
+    {
+      "epoch": 1.0435464888191448,
+      "grad_norm": 17.34891700744629,
+      "learning_rate": 8.454162071807181e-06,
+      "loss": 4.6001,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0445272655943507,
+      "grad_norm": 19.068683624267578,
+      "learning_rate": 8.45136262109768e-06,
+      "loss": 5.0256,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0455080423695566,
+      "grad_norm": 19.529052734375,
+      "learning_rate": 8.448561102244934e-06,
+      "loss": 4.6542,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0464888191447625,
+      "grad_norm": 28.510236740112305,
+      "learning_rate": 8.445757516927679e-06,
+      "loss": 4.6586,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0474695959199687,
+      "grad_norm": 10.34117603302002,
+      "learning_rate": 8.442951866825898e-06,
+      "loss": 4.8925,
+      "step": 5340
+    },
+    {
+      "epoch": 1.0484503726951746,
+      "grad_norm": 20.325603485107422,
+      "learning_rate": 8.4401441536208e-06,
+      "loss": 4.927,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0494311494703805,
+      "grad_norm": 15.056465148925781,
+      "learning_rate": 8.437334378994846e-06,
+      "loss": 4.807,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0504119262455864,
+      "grad_norm": 23.482568740844727,
+      "learning_rate": 8.434522544631718e-06,
+      "loss": 4.4933,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0513927030207926,
+      "grad_norm": 29.62999725341797,
+      "learning_rate": 8.43170865221634e-06,
+      "loss": 5.0732,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0523734797959985,
+      "grad_norm": 14.750212669372559,
+      "learning_rate": 8.428892703434867e-06,
+      "loss": 4.5473,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0533542565712044,
+      "grad_norm": 16.44927978515625,
+      "learning_rate": 8.426074699974686e-06,
+      "loss": 4.8326,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0543350333464103,
+      "grad_norm": 33.85694122314453,
+      "learning_rate": 8.423254643524415e-06,
+      "loss": 4.9815,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0553158101216162,
+      "grad_norm": 18.947261810302734,
+      "learning_rate": 8.420432535773902e-06,
+      "loss": 4.6267,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0562965868968224,
+      "grad_norm": 21.307687759399414,
+      "learning_rate": 8.417608378414228e-06,
+      "loss": 4.6284,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0572773636720283,
+      "grad_norm": 14.791841506958008,
+      "learning_rate": 8.414782173137697e-06,
+      "loss": 4.7716,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0582581404472342,
+      "grad_norm": 10.958831787109375,
+      "learning_rate": 8.41195392163784e-06,
+      "loss": 4.7179,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0592389172224401,
+      "grad_norm": 17.133495330810547,
+      "learning_rate": 8.409123625609421e-06,
+      "loss": 4.7059,
+      "step": 5400
+    },
+    {
+      "epoch": 1.060219693997646,
+      "grad_norm": 18.041934967041016,
+      "learning_rate": 8.406291286748423e-06,
+      "loss": 4.7454,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0612004707728522,
+      "grad_norm": 20.192089080810547,
+      "learning_rate": 8.403456906752053e-06,
+      "loss": 4.7348,
+      "step": 5410
+    },
+    {
+      "epoch": 1.062181247548058,
+      "grad_norm": 13.601256370544434,
+      "learning_rate": 8.400620487318743e-06,
+      "loss": 4.763,
+      "step": 5415
+    },
+    {
+      "epoch": 1.063162024323264,
+      "grad_norm": 23.853984832763672,
+      "learning_rate": 8.397782030148147e-06,
+      "loss": 4.5687,
+      "step": 5420
+    },
+    {
+      "epoch": 1.06414280109847,
+      "grad_norm": 14.830785751342773,
+      "learning_rate": 8.394941536941141e-06,
+      "loss": 4.9468,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0651235778736758,
+      "grad_norm": 22.439790725708008,
+      "learning_rate": 8.39209900939982e-06,
+      "loss": 4.716,
+      "step": 5430
+    },
+    {
+      "epoch": 1.066104354648882,
+      "grad_norm": 31.342418670654297,
+      "learning_rate": 8.389254449227498e-06,
+      "loss": 4.9564,
+      "step": 5435
+    },
+    {
+      "epoch": 1.067085131424088,
+      "grad_norm": 15.614798545837402,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 4.9951,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0680659081992938,
+      "grad_norm": 23.963844299316406,
+      "learning_rate": 8.383559237809194e-06,
+      "loss": 4.5133,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0690466849744997,
+      "grad_norm": 26.141693115234375,
+      "learning_rate": 8.380708589975923e-06,
+      "loss": 4.7709,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0700274617497059,
+      "grad_norm": 16.767637252807617,
+      "learning_rate": 8.377855916337078e-06,
+      "loss": 4.636,
+      "step": 5455
+    },
+    {
+      "epoch": 1.0710082385249118,
+      "grad_norm": 19.09316635131836,
+      "learning_rate": 8.375001218602053e-06,
+      "loss": 4.671,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0719890153001177,
+      "grad_norm": 14.91485595703125,
+      "learning_rate": 8.372144498481449e-06,
+      "loss": 4.9976,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0729697920753236,
+      "grad_norm": 26.471715927124023,
+      "learning_rate": 8.36928575768709e-06,
+      "loss": 4.8231,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0739505688505295,
+      "grad_norm": 15.676648139953613,
+      "learning_rate": 8.366424997932003e-06,
+      "loss": 4.6148,
+      "step": 5475
+    },
+    {
+      "epoch": 1.0749313456257357,
+      "grad_norm": 20.513164520263672,
+      "learning_rate": 8.363562220930426e-06,
+      "loss": 5.0241,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0759121224009416,
+      "grad_norm": 11.651981353759766,
+      "learning_rate": 8.36069742839781e-06,
+      "loss": 4.6736,
+      "step": 5485
+    },
+    {
+      "epoch": 1.0768928991761475,
+      "grad_norm": 20.40721321105957,
+      "learning_rate": 8.357830622050809e-06,
+      "loss": 4.7698,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0778736759513534,
+      "grad_norm": 17.44342613220215,
+      "learning_rate": 8.354961803607285e-06,
+      "loss": 4.7641,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0788544527265593,
+      "grad_norm": 18.868745803833008,
+      "learning_rate": 8.352090974786305e-06,
+      "loss": 4.8438,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0798352295017655,
+      "grad_norm": 11.686888694763184,
+      "learning_rate": 8.349218137308146e-06,
+      "loss": 4.4356,
+      "step": 5505
+    },
+    {
+      "epoch": 1.0808160062769714,
+      "grad_norm": 13.204122543334961,
+      "learning_rate": 8.34634329289428e-06,
+      "loss": 4.652,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0817967830521773,
+      "grad_norm": 27.09731674194336,
+      "learning_rate": 8.34346644326739e-06,
+      "loss": 4.6513,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0827775598273832,
+      "grad_norm": 16.423181533813477,
+      "learning_rate": 8.340587590151355e-06,
+      "loss": 4.6354,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0837583366025894,
+      "grad_norm": 16.599773406982422,
+      "learning_rate": 8.337706735271252e-06,
+      "loss": 5.013,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0847391133777953,
+      "grad_norm": 18.727693557739258,
+      "learning_rate": 8.334823880353368e-06,
+      "loss": 4.6582,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0857198901530012,
+      "grad_norm": 22.449386596679688,
+      "learning_rate": 8.33193902712518e-06,
+      "loss": 4.8486,
+      "step": 5535
+    },
+    {
+      "epoch": 1.086700666928207,
+      "grad_norm": 10.492510795593262,
+      "learning_rate": 8.329052177315365e-06,
+      "loss": 4.6019,
+      "step": 5540
+    },
+    {
+      "epoch": 1.087681443703413,
+      "grad_norm": 23.221635818481445,
+      "learning_rate": 8.326163332653791e-06,
+      "loss": 4.8868,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0886622204786192,
+      "grad_norm": 15.17660140991211,
+      "learning_rate": 8.323272494871534e-06,
+      "loss": 4.8388,
+      "step": 5550
+    },
+    {
+      "epoch": 1.089642997253825,
+      "grad_norm": 14.142809867858887,
+      "learning_rate": 8.320379665700852e-06,
+      "loss": 4.7848,
+      "step": 5555
+    },
+    {
+      "epoch": 1.090623774029031,
+      "grad_norm": 21.153562545776367,
+      "learning_rate": 8.317484846875202e-06,
+      "loss": 4.7567,
+      "step": 5560
+    },
+    {
+      "epoch": 1.091604550804237,
+      "grad_norm": 20.27756118774414,
+      "learning_rate": 8.314588040129232e-06,
+      "loss": 4.6365,
+      "step": 5565
+    },
+    {
+      "epoch": 1.0925853275794428,
+      "grad_norm": 17.335216522216797,
+      "learning_rate": 8.311689247198783e-06,
+      "loss": 4.8954,
+      "step": 5570
+    },
+    {
+      "epoch": 1.093566104354649,
+      "grad_norm": 12.399637222290039,
+      "learning_rate": 8.308788469820881e-06,
+      "loss": 4.8607,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0945468811298549,
+      "grad_norm": 20.954959869384766,
+      "learning_rate": 8.30588570973375e-06,
+      "loss": 4.5106,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0955276579050608,
+      "grad_norm": 32.353904724121094,
+      "learning_rate": 8.302980968676792e-06,
+      "loss": 4.7893,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0965084346802667,
+      "grad_norm": 16.396020889282227,
+      "learning_rate": 8.300074248390603e-06,
+      "loss": 4.4562,
+      "step": 5590
+    },
+    {
+      "epoch": 1.0974892114554726,
+      "grad_norm": 20.024126052856445,
+      "learning_rate": 8.297165550616964e-06,
+      "loss": 4.486,
+      "step": 5595
+    },
+    {
+      "epoch": 1.0984699882306788,
+      "grad_norm": 26.761669158935547,
+      "learning_rate": 8.294254877098834e-06,
+      "loss": 4.5415,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0994507650058847,
+      "grad_norm": 14.93077564239502,
+      "learning_rate": 8.29134222958037e-06,
+      "loss": 4.7062,
+      "step": 5605
+    },
+    {
+      "epoch": 1.1004315417810906,
+      "grad_norm": 25.9860782623291,
+      "learning_rate": 8.288427609806899e-06,
+      "loss": 4.9213,
+      "step": 5610
+    },
+    {
+      "epoch": 1.1014123185562965,
+      "grad_norm": 15.712201118469238,
+      "learning_rate": 8.285511019524937e-06,
+      "loss": 4.5184,
+      "step": 5615
+    },
+    {
+      "epoch": 1.1023930953315026,
+      "grad_norm": 29.031225204467773,
+      "learning_rate": 8.282592460482175e-06,
+      "loss": 4.6388,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1033738721067086,
+      "grad_norm": 13.145207405090332,
+      "learning_rate": 8.279671934427486e-06,
+      "loss": 4.7952,
+      "step": 5625
+    },
+    {
+      "epoch": 1.1043546488819145,
+      "grad_norm": 15.207327842712402,
+      "learning_rate": 8.276749443110928e-06,
+      "loss": 4.9209,
+      "step": 5630
+    },
+    {
+      "epoch": 1.1053354256571204,
+      "grad_norm": 14.627603530883789,
+      "learning_rate": 8.273824988283727e-06,
+      "loss": 4.5341,
+      "step": 5635
+    },
+    {
+      "epoch": 1.1063162024323263,
+      "grad_norm": 15.305438995361328,
+      "learning_rate": 8.270898571698291e-06,
+      "loss": 4.454,
+      "step": 5640
+    },
+    {
+      "epoch": 1.1072969792075325,
+      "grad_norm": 30.631746292114258,
+      "learning_rate": 8.267970195108204e-06,
+      "loss": 4.8473,
+      "step": 5645
+    },
+    {
+      "epoch": 1.1082777559827384,
+      "grad_norm": 11.449334144592285,
+      "learning_rate": 8.26503986026822e-06,
+      "loss": 4.5177,
+      "step": 5650
+    },
+    {
+      "epoch": 1.1092585327579443,
+      "grad_norm": 22.79661750793457,
+      "learning_rate": 8.262107568934271e-06,
+      "loss": 4.788,
+      "step": 5655
+    },
+    {
+      "epoch": 1.1102393095331502,
+      "grad_norm": 24.50604820251465,
+      "learning_rate": 8.25917332286346e-06,
+      "loss": 4.8102,
+      "step": 5660
+    },
+    {
+      "epoch": 1.1112200863083561,
+      "grad_norm": 22.736764907836914,
+      "learning_rate": 8.256237123814059e-06,
+      "loss": 4.8124,
+      "step": 5665
+    },
+    {
+      "epoch": 1.1122008630835623,
+      "grad_norm": 10.96435260772705,
+      "learning_rate": 8.253298973545516e-06,
+      "loss": 4.3873,
+      "step": 5670
+    },
+    {
+      "epoch": 1.1131816398587682,
+      "grad_norm": 26.919565200805664,
+      "learning_rate": 8.25035887381844e-06,
+      "loss": 4.8894,
+      "step": 5675
+    },
+    {
+      "epoch": 1.114162416633974,
+      "grad_norm": 19.470022201538086,
+      "learning_rate": 8.247416826394616e-06,
+      "loss": 4.7621,
+      "step": 5680
+    },
+    {
+      "epoch": 1.11514319340918,
+      "grad_norm": 29.276655197143555,
+      "learning_rate": 8.24447283303699e-06,
+      "loss": 4.505,
+      "step": 5685
+    },
+    {
+      "epoch": 1.1161239701843861,
+      "grad_norm": 20.404621124267578,
+      "learning_rate": 8.241526895509681e-06,
+      "loss": 4.8768,
+      "step": 5690
+    },
+    {
+      "epoch": 1.117104746959592,
+      "grad_norm": 21.071603775024414,
+      "learning_rate": 8.238579015577966e-06,
+      "loss": 4.6903,
+      "step": 5695
+    },
+    {
+      "epoch": 1.118085523734798,
+      "grad_norm": 24.47282600402832,
+      "learning_rate": 8.235629195008286e-06,
+      "loss": 4.7896,
+      "step": 5700
+    },
+    {
+      "epoch": 1.1190663005100039,
+      "grad_norm": 15.374588966369629,
+      "learning_rate": 8.232677435568252e-06,
+      "loss": 4.9747,
+      "step": 5705
+    },
+    {
+      "epoch": 1.1200470772852098,
+      "grad_norm": 23.620338439941406,
+      "learning_rate": 8.229723739026634e-06,
+      "loss": 4.8028,
+      "step": 5710
+    },
+    {
+      "epoch": 1.121027854060416,
+      "grad_norm": 13.628473281860352,
+      "learning_rate": 8.226768107153356e-06,
+      "loss": 4.6099,
+      "step": 5715
+    },
+    {
+      "epoch": 1.1220086308356219,
+      "grad_norm": 30.455013275146484,
+      "learning_rate": 8.22381054171951e-06,
+      "loss": 4.7731,
+      "step": 5720
+    },
+    {
+      "epoch": 1.1229894076108278,
+      "grad_norm": 16.254222869873047,
+      "learning_rate": 8.220851044497342e-06,
+      "loss": 4.8022,
+      "step": 5725
+    },
+    {
+      "epoch": 1.1239701843860337,
+      "grad_norm": 15.20760440826416,
+      "learning_rate": 8.217889617260257e-06,
+      "loss": 4.5828,
+      "step": 5730
+    },
+    {
+      "epoch": 1.1249509611612396,
+      "grad_norm": 45.388980865478516,
+      "learning_rate": 8.214926261782818e-06,
+      "loss": 4.8385,
+      "step": 5735
+    },
+    {
+      "epoch": 1.1259317379364457,
+      "grad_norm": 29.33622932434082,
+      "learning_rate": 8.211960979840743e-06,
+      "loss": 4.2081,
+      "step": 5740
+    },
+    {
+      "epoch": 1.1269125147116517,
+      "grad_norm": 25.06155014038086,
+      "learning_rate": 8.208993773210903e-06,
+      "loss": 4.7609,
+      "step": 5745
+    },
+    {
+      "epoch": 1.1278932914868576,
+      "grad_norm": 29.778722763061523,
+      "learning_rate": 8.20602464367132e-06,
+      "loss": 4.4514,
+      "step": 5750
+    },
+    {
+      "epoch": 1.1288740682620635,
+      "grad_norm": 21.625354766845703,
+      "learning_rate": 8.203053593001174e-06,
+      "loss": 4.6612,
+      "step": 5755
+    },
+    {
+      "epoch": 1.1298548450372694,
+      "grad_norm": 12.36301040649414,
+      "learning_rate": 8.200080622980793e-06,
+      "loss": 4.8831,
+      "step": 5760
+    },
+    {
+      "epoch": 1.1308356218124755,
+      "grad_norm": 13.567903518676758,
+      "learning_rate": 8.197105735391655e-06,
+      "loss": 4.7941,
+      "step": 5765
+    },
+    {
+      "epoch": 1.1318163985876815,
+      "grad_norm": 33.889305114746094,
+      "learning_rate": 8.194128932016385e-06,
+      "loss": 4.9926,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1327971753628874,
+      "grad_norm": 17.675067901611328,
+      "learning_rate": 8.19115021463876e-06,
+      "loss": 4.6703,
+      "step": 5775
+    },
+    {
+      "epoch": 1.1337779521380933,
+      "grad_norm": 42.091495513916016,
+      "learning_rate": 8.188169585043706e-06,
+      "loss": 4.5128,
+      "step": 5780
+    },
+    {
+      "epoch": 1.1347587289132992,
+      "grad_norm": 37.27928161621094,
+      "learning_rate": 8.185187045017289e-06,
+      "loss": 4.835,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1357395056885053,
+      "grad_norm": 19.49931526184082,
+      "learning_rate": 8.182202596346718e-06,
+      "loss": 4.6441,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1367202824637113,
+      "grad_norm": 21.166391372680664,
+      "learning_rate": 8.179216240820354e-06,
+      "loss": 4.9681,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1377010592389172,
+      "grad_norm": 15.765752792358398,
+      "learning_rate": 8.176227980227693e-06,
+      "loss": 4.809,
+      "step": 5800
+    },
+    {
+      "epoch": 1.138681836014123,
+      "grad_norm": 19.515087127685547,
+      "learning_rate": 8.17323781635938e-06,
+      "loss": 4.8872,
+      "step": 5805
+    },
+    {
+      "epoch": 1.1396626127893292,
+      "grad_norm": 28.068483352661133,
+      "learning_rate": 8.170245751007194e-06,
+      "loss": 4.4617,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1406433895645351,
+      "grad_norm": 13.95158863067627,
+      "learning_rate": 8.167251785964055e-06,
+      "loss": 4.5233,
+      "step": 5815
+    },
+    {
+      "epoch": 1.141624166339741,
+      "grad_norm": 13.904541015625,
+      "learning_rate": 8.164255923024025e-06,
+      "loss": 4.6834,
+      "step": 5820
+    },
+    {
+      "epoch": 1.142604943114947,
+      "grad_norm": 25.7268123626709,
+      "learning_rate": 8.161258163982298e-06,
+      "loss": 4.7352,
+      "step": 5825
+    },
+    {
+      "epoch": 1.1435857198901531,
+      "grad_norm": 21.95464515686035,
+      "learning_rate": 8.158258510635206e-06,
+      "loss": 4.6658,
+      "step": 5830
+    },
+    {
+      "epoch": 1.144566496665359,
+      "grad_norm": 11.288395881652832,
+      "learning_rate": 8.155256964780218e-06,
+      "loss": 4.7413,
+      "step": 5835
+    },
+    {
+      "epoch": 1.145547273440565,
+      "grad_norm": 24.911361694335938,
+      "learning_rate": 8.152253528215937e-06,
+      "loss": 4.67,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1465280502157709,
+      "grad_norm": 10.945342063903809,
+      "learning_rate": 8.149248202742096e-06,
+      "loss": 4.8911,
+      "step": 5845
+    },
+    {
+      "epoch": 1.1475088269909768,
+      "grad_norm": 18.10870361328125,
+      "learning_rate": 8.146240990159558e-06,
+      "loss": 4.9292,
+      "step": 5850
+    },
+    {
+      "epoch": 1.148489603766183,
+      "grad_norm": 23.122905731201172,
+      "learning_rate": 8.143231892270327e-06,
+      "loss": 5.0509,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1494703805413888,
+      "grad_norm": 14.539555549621582,
+      "learning_rate": 8.140220910877529e-06,
+      "loss": 4.8416,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1504511573165948,
+      "grad_norm": 18.56782341003418,
+      "learning_rate": 8.137208047785417e-06,
+      "loss": 4.8764,
+      "step": 5865
+    },
+    {
+      "epoch": 1.1514319340918007,
+      "grad_norm": 22.753496170043945,
+      "learning_rate": 8.134193304799373e-06,
+      "loss": 4.921,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1524127108670066,
+      "grad_norm": 15.556824684143066,
+      "learning_rate": 8.131176683725912e-06,
+      "loss": 5.0195,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1533934876422127,
+      "grad_norm": 16.221290588378906,
+      "learning_rate": 8.128158186372666e-06,
+      "loss": 4.6949,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1543742644174186,
+      "grad_norm": 24.236032485961914,
+      "learning_rate": 8.125137814548394e-06,
+      "loss": 5.0952,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1553550411926246,
+      "grad_norm": 21.219877243041992,
+      "learning_rate": 8.122115570062978e-06,
+      "loss": 4.8158,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1563358179678305,
+      "grad_norm": 23.525745391845703,
+      "learning_rate": 8.119091454727427e-06,
+      "loss": 4.5122,
+      "step": 5895
+    },
+    {
+      "epoch": 1.1573165947430364,
+      "grad_norm": 17.191791534423828,
+      "learning_rate": 8.116065470353863e-06,
+      "loss": 4.5277,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1582973715182425,
+      "grad_norm": 14.015380859375,
+      "learning_rate": 8.113037618755533e-06,
+      "loss": 4.5689,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1592781482934484,
+      "grad_norm": 24.197856903076172,
+      "learning_rate": 8.110007901746804e-06,
+      "loss": 4.7434,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1602589250686544,
+      "grad_norm": 11.606034278869629,
+      "learning_rate": 8.106976321143155e-06,
+      "loss": 4.9903,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1612397018438603,
+      "grad_norm": 10.518669128417969,
+      "learning_rate": 8.103942878761189e-06,
+      "loss": 4.8785,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1622204786190662,
+      "grad_norm": 27.280576705932617,
+      "learning_rate": 8.100907576418616e-06,
+      "loss": 4.7905,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1632012553942723,
+      "grad_norm": 111.20673370361328,
+      "learning_rate": 8.097870415934269e-06,
+      "loss": 4.5365,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1641820321694782,
+      "grad_norm": 17.625286102294922,
+      "learning_rate": 8.094831399128092e-06,
+      "loss": 4.7943,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1651628089446842,
+      "grad_norm": 11.342884063720703,
+      "learning_rate": 8.091790527821138e-06,
+      "loss": 4.6058,
+      "step": 5940
+    },
+    {
+      "epoch": 1.16614358571989,
+      "grad_norm": 20.569257736206055,
+      "learning_rate": 8.088747803835573e-06,
+      "loss": 4.6913,
+      "step": 5945
+    },
+    {
+      "epoch": 1.1671243624950962,
+      "grad_norm": 13.408282279968262,
+      "learning_rate": 8.085703228994674e-06,
+      "loss": 4.9914,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1681051392703021,
+      "grad_norm": 28.350143432617188,
+      "learning_rate": 8.082656805122829e-06,
+      "loss": 4.5938,
+      "step": 5955
+    },
+    {
+      "epoch": 1.169085916045508,
+      "grad_norm": 25.818490982055664,
+      "learning_rate": 8.07960853404553e-06,
+      "loss": 4.4677,
+      "step": 5960
+    },
+    {
+      "epoch": 1.170066692820714,
+      "grad_norm": 16.237733840942383,
+      "learning_rate": 8.07655841758938e-06,
+      "loss": 4.5051,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1710474695959199,
+      "grad_norm": 24.858333587646484,
+      "learning_rate": 8.073506457582082e-06,
+      "loss": 4.574,
+      "step": 5970
+    },
+    {
+      "epoch": 1.172028246371126,
+      "grad_norm": 10.83703327178955,
+      "learning_rate": 8.070452655852445e-06,
+      "loss": 5.1138,
+      "step": 5975
+    },
+    {
+      "epoch": 1.173009023146332,
+      "grad_norm": 22.579744338989258,
+      "learning_rate": 8.067397014230391e-06,
+      "loss": 4.759,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1739897999215378,
+      "grad_norm": 17.268518447875977,
+      "learning_rate": 8.064339534546935e-06,
+      "loss": 4.7689,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1749705766967438,
+      "grad_norm": 34.30562973022461,
+      "learning_rate": 8.061280218634192e-06,
+      "loss": 4.7157,
+      "step": 5990
+    },
+    {
+      "epoch": 1.17595135347195,
+      "grad_norm": 16.77152442932129,
+      "learning_rate": 8.058219068325383e-06,
+      "loss": 4.5394,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1769321302471558,
+      "grad_norm": 10.069709777832031,
+      "learning_rate": 8.055156085454828e-06,
+      "loss": 4.8192,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1779129070223617,
+      "grad_norm": 36.60639190673828,
+      "learning_rate": 8.052091271857942e-06,
+      "loss": 4.5211,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1788936837975676,
+      "grad_norm": 28.5606689453125,
+      "learning_rate": 8.049024629371237e-06,
+      "loss": 4.4561,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1798744605727736,
+      "grad_norm": 18.83449935913086,
+      "learning_rate": 8.045956159832324e-06,
+      "loss": 4.89,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1808552373479797,
+      "grad_norm": 14.824164390563965,
+      "learning_rate": 8.042885865079909e-06,
+      "loss": 4.8188,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1818360141231856,
+      "grad_norm": 17.74277114868164,
+      "learning_rate": 8.039813746953785e-06,
+      "loss": 5.0979,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1828167908983915,
+      "grad_norm": 19.057353973388672,
+      "learning_rate": 8.036739807294844e-06,
+      "loss": 4.915,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1837975676735975,
+      "grad_norm": 24.1026554107666,
+      "learning_rate": 8.03366404794507e-06,
+      "loss": 4.9094,
+      "step": 6035
+    },
+    {
+      "epoch": 1.1847783444488034,
+      "grad_norm": 17.447006225585938,
+      "learning_rate": 8.030586470747535e-06,
+      "loss": 4.7573,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1857591212240095,
+      "grad_norm": 34.93657302856445,
+      "learning_rate": 8.027507077546398e-06,
+      "loss": 4.5797,
+      "step": 6045
+    },
+    {
+      "epoch": 1.1867398979992154,
+      "grad_norm": 19.555604934692383,
+      "learning_rate": 8.024425870186912e-06,
+      "loss": 4.5836,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1877206747744213,
+      "grad_norm": 26.290122985839844,
+      "learning_rate": 8.02134285051541e-06,
+      "loss": 4.7241,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1887014515496273,
+      "grad_norm": 39.10771942138672,
+      "learning_rate": 8.018258020379319e-06,
+      "loss": 4.8729,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1896822283248332,
+      "grad_norm": 12.844032287597656,
+      "learning_rate": 8.015171381627145e-06,
+      "loss": 4.5366,
+      "step": 6065
+    },
+    {
+      "epoch": 1.1906630051000393,
+      "grad_norm": 17.043214797973633,
+      "learning_rate": 8.01208293610848e-06,
+      "loss": 4.7391,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1916437818752452,
+      "grad_norm": 14.163534164428711,
+      "learning_rate": 8.008992685673998e-06,
+      "loss": 5.0134,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1926245586504511,
+      "grad_norm": 18.311864852905273,
+      "learning_rate": 8.005900632175453e-06,
+      "loss": 4.5244,
+      "step": 6080
+    },
+    {
+      "epoch": 1.193605335425657,
+      "grad_norm": 15.40824031829834,
+      "learning_rate": 8.002806777465685e-06,
+      "loss": 5.0018,
+      "step": 6085
+    },
+    {
+      "epoch": 1.194586112200863,
+      "grad_norm": 16.03959846496582,
+      "learning_rate": 7.999711123398607e-06,
+      "loss": 4.8075,
+      "step": 6090
+    },
+    {
+      "epoch": 1.195566888976069,
+      "grad_norm": 12.000391960144043,
+      "learning_rate": 7.996613671829211e-06,
+      "loss": 4.6103,
+      "step": 6095
+    },
+    {
+      "epoch": 1.196547665751275,
+      "grad_norm": 17.460683822631836,
+      "learning_rate": 7.993514424613572e-06,
+      "loss": 4.8899,
+      "step": 6100
+    },
+    {
+      "epoch": 1.197528442526481,
+      "grad_norm": 31.026866912841797,
+      "learning_rate": 7.990413383608833e-06,
+      "loss": 4.7524,
+      "step": 6105
+    },
+    {
+      "epoch": 1.1985092193016869,
+      "grad_norm": 18.090789794921875,
+      "learning_rate": 7.98731055067322e-06,
+      "loss": 4.6904,
+      "step": 6110
+    },
+    {
+      "epoch": 1.199489996076893,
+      "grad_norm": 14.723376274108887,
+      "learning_rate": 7.984205927666023e-06,
+      "loss": 4.5939,
+      "step": 6115
+    },
+    {
+      "epoch": 1.200470772852099,
+      "grad_norm": 25.426658630371094,
+      "learning_rate": 7.981099516447614e-06,
+      "loss": 4.6615,
+      "step": 6120
+    },
+    {
+      "epoch": 1.2014515496273048,
+      "grad_norm": 13.51884651184082,
+      "learning_rate": 7.977991318879432e-06,
+      "loss": 4.9369,
+      "step": 6125
+    },
+    {
+      "epoch": 1.2024323264025107,
+      "grad_norm": 28.351192474365234,
+      "learning_rate": 7.974881336823988e-06,
+      "loss": 4.7937,
+      "step": 6130
+    },
+    {
+      "epoch": 1.2034131031777167,
+      "grad_norm": 39.130531311035156,
+      "learning_rate": 7.971769572144858e-06,
+      "loss": 4.8798,
+      "step": 6135
+    },
+    {
+      "epoch": 1.2043938799529228,
+      "grad_norm": 16.273771286010742,
+      "learning_rate": 7.968656026706693e-06,
+      "loss": 5.1519,
+      "step": 6140
+    },
+    {
+      "epoch": 1.2053746567281287,
+      "grad_norm": 20.515405654907227,
+      "learning_rate": 7.965540702375207e-06,
+      "loss": 4.9034,
+      "step": 6145
+    },
+    {
+      "epoch": 1.2063554335033346,
+      "grad_norm": 21.847187042236328,
+      "learning_rate": 7.962423601017183e-06,
+      "loss": 4.5778,
+      "step": 6150
+    },
+    {
+      "epoch": 1.2073362102785405,
+      "grad_norm": 24.054405212402344,
+      "learning_rate": 7.95930472450046e-06,
+      "loss": 4.8069,
+      "step": 6155
+    },
+    {
+      "epoch": 1.2083169870537467,
+      "grad_norm": 16.962587356567383,
+      "learning_rate": 7.956184074693952e-06,
+      "loss": 4.6251,
+      "step": 6160
+    },
+    {
+      "epoch": 1.2092977638289526,
+      "grad_norm": 37.481475830078125,
+      "learning_rate": 7.953061653467631e-06,
+      "loss": 4.7728,
+      "step": 6165
+    },
+    {
+      "epoch": 1.2102785406041585,
+      "grad_norm": 24.814760208129883,
+      "learning_rate": 7.949937462692528e-06,
+      "loss": 4.7485,
+      "step": 6170
+    },
+    {
+      "epoch": 1.2112593173793644,
+      "grad_norm": 15.355949401855469,
+      "learning_rate": 7.946811504240736e-06,
+      "loss": 4.6448,
+      "step": 6175
+    },
+    {
+      "epoch": 1.2122400941545703,
+      "grad_norm": 30.65787696838379,
+      "learning_rate": 7.943683779985412e-06,
+      "loss": 4.6957,
+      "step": 6180
+    },
+    {
+      "epoch": 1.2132208709297765,
+      "grad_norm": 13.861238479614258,
+      "learning_rate": 7.940554291800766e-06,
+      "loss": 4.4497,
+      "step": 6185
+    },
+    {
+      "epoch": 1.2142016477049824,
+      "grad_norm": 52.959903717041016,
+      "learning_rate": 7.937423041562063e-06,
+      "loss": 4.8312,
+      "step": 6190
+    },
+    {
+      "epoch": 1.2151824244801883,
+      "grad_norm": 20.761072158813477,
+      "learning_rate": 7.934290031145629e-06,
+      "loss": 4.6906,
+      "step": 6195
+    },
+    {
+      "epoch": 1.2161632012553942,
+      "grad_norm": 32.41350173950195,
+      "learning_rate": 7.93115526242884e-06,
+      "loss": 4.9518,
+      "step": 6200
+    },
+    {
+      "epoch": 1.2171439780306001,
+      "grad_norm": 16.019487380981445,
+      "learning_rate": 7.928018737290132e-06,
+      "loss": 4.7748,
+      "step": 6205
+    },
+    {
+      "epoch": 1.2181247548058063,
+      "grad_norm": 12.364213943481445,
+      "learning_rate": 7.924880457608987e-06,
+      "loss": 4.5267,
+      "step": 6210
+    },
+    {
+      "epoch": 1.2191055315810122,
+      "grad_norm": 27.74942970275879,
+      "learning_rate": 7.921740425265944e-06,
+      "loss": 4.6387,
+      "step": 6215
+    },
+    {
+      "epoch": 1.2200863083562181,
+      "grad_norm": 14.577978134155273,
+      "learning_rate": 7.918598642142588e-06,
+      "loss": 4.7566,
+      "step": 6220
+    },
+    {
+      "epoch": 1.221067085131424,
+      "grad_norm": 18.41602325439453,
+      "learning_rate": 7.915455110121553e-06,
+      "loss": 4.613,
+      "step": 6225
+    },
+    {
+      "epoch": 1.22204786190663,
+      "grad_norm": 13.009770393371582,
+      "learning_rate": 7.912309831086522e-06,
+      "loss": 4.8313,
+      "step": 6230
+    },
+    {
+      "epoch": 1.223028638681836,
+      "grad_norm": 16.142860412597656,
+      "learning_rate": 7.909162806922229e-06,
+      "loss": 4.4975,
+      "step": 6235
+    },
+    {
+      "epoch": 1.224009415457042,
+      "grad_norm": 19.88920021057129,
+      "learning_rate": 7.906014039514446e-06,
+      "loss": 4.7498,
+      "step": 6240
+    },
+    {
+      "epoch": 1.224990192232248,
+      "grad_norm": 15.997335433959961,
+      "learning_rate": 7.902863530749995e-06,
+      "loss": 4.8573,
+      "step": 6245
+    },
+    {
+      "epoch": 1.2259709690074538,
+      "grad_norm": 25.197553634643555,
+      "learning_rate": 7.89971128251674e-06,
+      "loss": 4.8058,
+      "step": 6250
+    },
+    {
+      "epoch": 1.2269517457826598,
+      "grad_norm": 25.884624481201172,
+      "learning_rate": 7.896557296703589e-06,
+      "loss": 4.7384,
+      "step": 6255
+    },
+    {
+      "epoch": 1.227932522557866,
+      "grad_norm": 17.546234130859375,
+      "learning_rate": 7.893401575200488e-06,
+      "loss": 4.9061,
+      "step": 6260
+    },
+    {
+      "epoch": 1.2289132993330718,
+      "grad_norm": 15.71548843383789,
+      "learning_rate": 7.890244119898423e-06,
+      "loss": 4.5454,
+      "step": 6265
+    },
+    {
+      "epoch": 1.2298940761082777,
+      "grad_norm": 15.9916353225708,
+      "learning_rate": 7.887084932689424e-06,
+      "loss": 5.0506,
+      "step": 6270
+    },
+    {
+      "epoch": 1.2308748528834836,
+      "grad_norm": 34.314453125,
+      "learning_rate": 7.883924015466554e-06,
+      "loss": 4.604,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2318556296586898,
+      "grad_norm": 15.04297924041748,
+      "learning_rate": 7.880761370123914e-06,
+      "loss": 4.7883,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2328364064338957,
+      "grad_norm": 26.045204162597656,
+      "learning_rate": 7.87759699855664e-06,
+      "loss": 4.6663,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2338171832091016,
+      "grad_norm": 19.145492553710938,
+      "learning_rate": 7.874430902660903e-06,
+      "loss": 4.6448,
+      "step": 6290
+    },
+    {
+      "epoch": 1.2347979599843075,
+      "grad_norm": 12.347979545593262,
+      "learning_rate": 7.87126308433391e-06,
+      "loss": 4.9435,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2357787367595134,
+      "grad_norm": 16.063499450683594,
+      "learning_rate": 7.868093545473891e-06,
+      "loss": 4.6589,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2367595135347196,
+      "grad_norm": 12.78153133392334,
+      "learning_rate": 7.864922287980121e-06,
+      "loss": 4.6432,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2377402903099255,
+      "grad_norm": 15.366764068603516,
+      "learning_rate": 7.861749313752893e-06,
+      "loss": 4.5588,
+      "step": 6310
+    },
+    {
+      "epoch": 1.2387210670851314,
+      "grad_norm": 13.756194114685059,
+      "learning_rate": 7.858574624693533e-06,
+      "loss": 4.5369,
+      "step": 6315
+    },
+    {
+      "epoch": 1.2397018438603373,
+      "grad_norm": 25.25469207763672,
+      "learning_rate": 7.855398222704395e-06,
+      "loss": 5.0234,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2406826206355435,
+      "grad_norm": 20.50889015197754,
+      "learning_rate": 7.85222010968886e-06,
+      "loss": 4.6325,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2416633974107494,
+      "grad_norm": 24.124738693237305,
+      "learning_rate": 7.849040287551331e-06,
+      "loss": 4.8982,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2426441741859553,
+      "grad_norm": 27.7025146484375,
+      "learning_rate": 7.84585875819724e-06,
+      "loss": 5.0759,
+      "step": 6335
+    },
+    {
+      "epoch": 1.2436249509611612,
+      "grad_norm": 11.230903625488281,
+      "learning_rate": 7.842675523533038e-06,
+      "loss": 4.9903,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2446057277363671,
+      "grad_norm": 11.444461822509766,
+      "learning_rate": 7.839490585466198e-06,
+      "loss": 4.6177,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2455865045115733,
+      "grad_norm": 19.044597625732422,
+      "learning_rate": 7.836303945905217e-06,
+      "loss": 4.7733,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2465672812867792,
+      "grad_norm": 15.269647598266602,
+      "learning_rate": 7.833115606759608e-06,
+      "loss": 4.7614,
+      "step": 6355
+    },
+    {
+      "epoch": 1.247548058061985,
+      "grad_norm": 16.577899932861328,
+      "learning_rate": 7.829925569939908e-06,
+      "loss": 4.601,
+      "step": 6360
+    },
+    {
+      "epoch": 1.248528834837191,
+      "grad_norm": 19.186344146728516,
+      "learning_rate": 7.82673383735766e-06,
+      "loss": 4.5914,
+      "step": 6365
+    },
+    {
+      "epoch": 1.249509611612397,
+      "grad_norm": 18.438297271728516,
+      "learning_rate": 7.823540410925434e-06,
+      "loss": 4.7085,
+      "step": 6370
+    },
+    {
+      "epoch": 1.250490388387603,
+      "grad_norm": 15.657546043395996,
+      "learning_rate": 7.820345292556815e-06,
+      "loss": 4.605,
+      "step": 6375
+    },
+    {
+      "epoch": 1.250490388387603,
+      "eval_loss": 4.937655448913574,
+      "eval_runtime": 7.7325,
+      "eval_samples_per_second": 27.029,
+      "eval_steps_per_second": 13.579,
+      "step": 6375
+    },
+    {
+      "epoch": 1.251471165162809,
+      "grad_norm": 24.359493255615234,
+      "learning_rate": 7.817148484166392e-06,
+      "loss": 4.7048,
+      "step": 6380
+    },
+    {
+      "epoch": 1.252451941938015,
+      "grad_norm": 16.691558837890625,
+      "learning_rate": 7.813949987669777e-06,
+      "loss": 4.7663,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2534327187132208,
+      "grad_norm": 26.948749542236328,
+      "learning_rate": 7.81074980498359e-06,
+      "loss": 4.7941,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2544134954884267,
+      "grad_norm": 13.582273483276367,
+      "learning_rate": 7.807547938025458e-06,
+      "loss": 4.6875,
+      "step": 6395
+    },
+    {
+      "epoch": 1.2553942722636329,
+      "grad_norm": 14.236882209777832,
+      "learning_rate": 7.804344388714022e-06,
+      "loss": 4.5558,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2563750490388388,
+      "grad_norm": 30.24952507019043,
+      "learning_rate": 7.801139158968928e-06,
+      "loss": 4.6246,
+      "step": 6405
+    },
+    {
+      "epoch": 1.2573558258140447,
+      "grad_norm": 24.6456298828125,
+      "learning_rate": 7.797932250710832e-06,
+      "loss": 4.6808,
+      "step": 6410
+    },
+    {
+      "epoch": 1.2583366025892506,
+      "grad_norm": 14.85630989074707,
+      "learning_rate": 7.794723665861392e-06,
+      "loss": 4.9929,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2593173793644565,
+      "grad_norm": 20.140844345092773,
+      "learning_rate": 7.791513406343276e-06,
+      "loss": 4.516,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2602981561396627,
+      "grad_norm": 16.46764373779297,
+      "learning_rate": 7.788301474080148e-06,
+      "loss": 4.7002,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2612789329148686,
+      "grad_norm": 16.4792537689209,
+      "learning_rate": 7.785087870996682e-06,
+      "loss": 4.7079,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2622597096900745,
+      "grad_norm": 14.717877388000488,
+      "learning_rate": 7.781872599018547e-06,
+      "loss": 4.5069,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2632404864652804,
+      "grad_norm": 31.211990356445312,
+      "learning_rate": 7.778655660072417e-06,
+      "loss": 4.915,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2642212632404863,
+      "grad_norm": 19.69355010986328,
+      "learning_rate": 7.775437056085961e-06,
+      "loss": 4.7949,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2652020400156925,
+      "grad_norm": 22.162994384765625,
+      "learning_rate": 7.77221678898785e-06,
+      "loss": 4.7987,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2661828167908984,
+      "grad_norm": 57.02979278564453,
+      "learning_rate": 7.768994860707745e-06,
+      "loss": 4.7767,
+      "step": 6455
+    },
+    {
+      "epoch": 1.2671635935661043,
+      "grad_norm": 20.068218231201172,
+      "learning_rate": 7.76577127317631e-06,
+      "loss": 4.4555,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2681443703413104,
+      "grad_norm": 29.125776290893555,
+      "learning_rate": 7.7625460283252e-06,
+      "loss": 4.7459,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2691251471165164,
+      "grad_norm": 13.710939407348633,
+      "learning_rate": 7.759319128087058e-06,
+      "loss": 4.6614,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2701059238917223,
+      "grad_norm": 12.004446029663086,
+      "learning_rate": 7.756090574395528e-06,
+      "loss": 4.6544,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2710867006669282,
+      "grad_norm": 13.436174392700195,
+      "learning_rate": 7.75286036918524e-06,
+      "loss": 4.6053,
+      "step": 6480
+    },
+    {
+      "epoch": 1.272067477442134,
+      "grad_norm": 12.79939079284668,
+      "learning_rate": 7.749628514391814e-06,
+      "loss": 4.9016,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2730482542173402,
+      "grad_norm": 12.1824312210083,
+      "learning_rate": 7.746395011951857e-06,
+      "loss": 4.7288,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2740290309925462,
+      "grad_norm": 17.693960189819336,
+      "learning_rate": 7.743159863802967e-06,
+      "loss": 4.6414,
+      "step": 6495
+    },
+    {
+      "epoch": 1.275009807767752,
+      "grad_norm": 22.020206451416016,
+      "learning_rate": 7.739923071883725e-06,
+      "loss": 4.8802,
+      "step": 6500
+    },
+    {
+      "epoch": 1.275990584542958,
+      "grad_norm": 14.62353801727295,
+      "learning_rate": 7.736684638133699e-06,
+      "loss": 4.7882,
+      "step": 6505
+    },
+    {
+      "epoch": 1.276971361318164,
+      "grad_norm": 22.087949752807617,
+      "learning_rate": 7.73344456449344e-06,
+      "loss": 4.8307,
+      "step": 6510
+    },
+    {
+      "epoch": 1.27795213809337,
+      "grad_norm": 23.798274993896484,
+      "learning_rate": 7.73020285290448e-06,
+      "loss": 4.7251,
+      "step": 6515
+    },
+    {
+      "epoch": 1.278932914868576,
+      "grad_norm": 15.14513874053955,
+      "learning_rate": 7.726959505309335e-06,
+      "loss": 4.7146,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2799136916437819,
+      "grad_norm": 12.246772766113281,
+      "learning_rate": 7.7237145236515e-06,
+      "loss": 4.8786,
+      "step": 6525
+    },
+    {
+      "epoch": 1.2808944684189878,
+      "grad_norm": 25.760866165161133,
+      "learning_rate": 7.720467909875448e-06,
+      "loss": 4.3852,
+      "step": 6530
+    },
+    {
+      "epoch": 1.2818752451941937,
+      "grad_norm": 23.278759002685547,
+      "learning_rate": 7.717219665926635e-06,
+      "loss": 4.7413,
+      "step": 6535
+    },
+    {
+      "epoch": 1.2828560219693999,
+      "grad_norm": 10.611922264099121,
+      "learning_rate": 7.713969793751493e-06,
+      "loss": 4.4106,
+      "step": 6540
+    },
+    {
+      "epoch": 1.2838367987446058,
+      "grad_norm": 21.27843475341797,
+      "learning_rate": 7.710718295297418e-06,
+      "loss": 5.0443,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2848175755198117,
+      "grad_norm": 17.157188415527344,
+      "learning_rate": 7.707465172512798e-06,
+      "loss": 4.9554,
+      "step": 6550
+    },
+    {
+      "epoch": 1.2857983522950176,
+      "grad_norm": 13.657434463500977,
+      "learning_rate": 7.704210427346979e-06,
+      "loss": 4.6168,
+      "step": 6555
+    },
+    {
+      "epoch": 1.2867791290702235,
+      "grad_norm": 20.332624435424805,
+      "learning_rate": 7.700954061750295e-06,
+      "loss": 4.8259,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2877599058454297,
+      "grad_norm": 11.391372680664062,
+      "learning_rate": 7.697696077674032e-06,
+      "loss": 4.7383,
+      "step": 6565
+    },
+    {
+      "epoch": 1.2887406826206356,
+      "grad_norm": 21.024250030517578,
+      "learning_rate": 7.694436477070464e-06,
+      "loss": 4.6618,
+      "step": 6570
+    },
+    {
+      "epoch": 1.2897214593958415,
+      "grad_norm": 17.038434982299805,
+      "learning_rate": 7.691175261892821e-06,
+      "loss": 5.1161,
+      "step": 6575
+    },
+    {
+      "epoch": 1.2907022361710474,
+      "grad_norm": 15.151735305786133,
+      "learning_rate": 7.687912434095306e-06,
+      "loss": 4.7389,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2916830129462533,
+      "grad_norm": 33.51579284667969,
+      "learning_rate": 7.68464799563309e-06,
+      "loss": 4.5458,
+      "step": 6585
+    },
+    {
+      "epoch": 1.2926637897214595,
+      "grad_norm": 22.85167121887207,
+      "learning_rate": 7.681381948462304e-06,
+      "loss": 4.7308,
+      "step": 6590
+    },
+    {
+      "epoch": 1.2936445664966654,
+      "grad_norm": 23.826412200927734,
+      "learning_rate": 7.678114294540046e-06,
+      "loss": 5.0318,
+      "step": 6595
+    },
+    {
+      "epoch": 1.2946253432718713,
+      "grad_norm": 19.270708084106445,
+      "learning_rate": 7.674845035824377e-06,
+      "loss": 4.6337,
+      "step": 6600
+    },
+    {
+      "epoch": 1.2956061200470772,
+      "grad_norm": 22.91469955444336,
+      "learning_rate": 7.671574174274317e-06,
+      "loss": 4.8004,
+      "step": 6605
+    },
+    {
+      "epoch": 1.2965868968222831,
+      "grad_norm": 19.99880027770996,
+      "learning_rate": 7.668301711849852e-06,
+      "loss": 4.7976,
+      "step": 6610
+    },
+    {
+      "epoch": 1.2975676735974893,
+      "grad_norm": 18.21483039855957,
+      "learning_rate": 7.665027650511921e-06,
+      "loss": 4.826,
+      "step": 6615
+    },
+    {
+      "epoch": 1.2985484503726952,
+      "grad_norm": 23.880598068237305,
+      "learning_rate": 7.661751992222425e-06,
+      "loss": 4.335,
+      "step": 6620
+    },
+    {
+      "epoch": 1.299529227147901,
+      "grad_norm": 34.46923065185547,
+      "learning_rate": 7.65847473894422e-06,
+      "loss": 4.6269,
+      "step": 6625
+    },
+    {
+      "epoch": 1.3005100039231072,
+      "grad_norm": 10.862066268920898,
+      "learning_rate": 7.65519589264112e-06,
+      "loss": 4.7542,
+      "step": 6630
+    },
+    {
+      "epoch": 1.3014907806983131,
+      "grad_norm": 18.505020141601562,
+      "learning_rate": 7.65191545527789e-06,
+      "loss": 4.6034,
+      "step": 6635
+    },
+    {
+      "epoch": 1.302471557473519,
+      "grad_norm": 14.459244728088379,
+      "learning_rate": 7.648633428820254e-06,
+      "loss": 4.6442,
+      "step": 6640
+    },
+    {
+      "epoch": 1.303452334248725,
+      "grad_norm": 18.99560546875,
+      "learning_rate": 7.64534981523488e-06,
+      "loss": 4.5086,
+      "step": 6645
+    },
+    {
+      "epoch": 1.304433111023931,
+      "grad_norm": 15.123648643493652,
+      "learning_rate": 7.642064616489394e-06,
+      "loss": 4.9788,
+      "step": 6650
+    },
+    {
+      "epoch": 1.305413887799137,
+      "grad_norm": 30.390419006347656,
+      "learning_rate": 7.638777834552372e-06,
+      "loss": 4.9361,
+      "step": 6655
+    },
+    {
+      "epoch": 1.306394664574343,
+      "grad_norm": 16.628713607788086,
+      "learning_rate": 7.635489471393334e-06,
+      "loss": 4.4967,
+      "step": 6660
+    },
+    {
+      "epoch": 1.3073754413495489,
+      "grad_norm": 23.70064353942871,
+      "learning_rate": 7.632199528982748e-06,
+      "loss": 4.7082,
+      "step": 6665
+    },
+    {
+      "epoch": 1.3083562181247548,
+      "grad_norm": 12.797712326049805,
+      "learning_rate": 7.6289080092920354e-06,
+      "loss": 4.9439,
+      "step": 6670
+    },
+    {
+      "epoch": 1.3093369948999607,
+      "grad_norm": 23.16967010498047,
+      "learning_rate": 7.625614914293553e-06,
+      "loss": 4.6688,
+      "step": 6675
+    },
+    {
+      "epoch": 1.3103177716751668,
+      "grad_norm": 27.266939163208008,
+      "learning_rate": 7.622320245960607e-06,
+      "loss": 4.427,
+      "step": 6680
+    },
+    {
+      "epoch": 1.3112985484503727,
+      "grad_norm": 24.94828224182129,
+      "learning_rate": 7.619024006267448e-06,
+      "loss": 4.6799,
+      "step": 6685
+    },
+    {
+      "epoch": 1.3122793252255787,
+      "grad_norm": 13.128344535827637,
+      "learning_rate": 7.6157261971892626e-06,
+      "loss": 4.4892,
+      "step": 6690
+    },
+    {
+      "epoch": 1.3132601020007846,
+      "grad_norm": 22.916282653808594,
+      "learning_rate": 7.612426820702182e-06,
+      "loss": 4.781,
+      "step": 6695
+    },
+    {
+      "epoch": 1.3142408787759905,
+      "grad_norm": 18.023021697998047,
+      "learning_rate": 7.6091258787832765e-06,
+      "loss": 4.7486,
+      "step": 6700
+    },
+    {
+      "epoch": 1.3152216555511966,
+      "grad_norm": 20.854341506958008,
+      "learning_rate": 7.605823373410553e-06,
+      "loss": 4.8066,
+      "step": 6705
+    },
+    {
+      "epoch": 1.3162024323264025,
+      "grad_norm": 25.060224533081055,
+      "learning_rate": 7.602519306562954e-06,
+      "loss": 4.7733,
+      "step": 6710
+    },
+    {
+      "epoch": 1.3171832091016085,
+      "grad_norm": 34.962406158447266,
+      "learning_rate": 7.599213680220362e-06,
+      "loss": 4.7923,
+      "step": 6715
+    },
+    {
+      "epoch": 1.3181639858768144,
+      "grad_norm": 22.008214950561523,
+      "learning_rate": 7.59590649636359e-06,
+      "loss": 5.0123,
+      "step": 6720
+    },
+    {
+      "epoch": 1.3191447626520203,
+      "grad_norm": 16.601545333862305,
+      "learning_rate": 7.592597756974385e-06,
+      "loss": 4.9437,
+      "step": 6725
+    },
+    {
+      "epoch": 1.3201255394272264,
+      "grad_norm": 24.702821731567383,
+      "learning_rate": 7.589287464035429e-06,
+      "loss": 4.8218,
+      "step": 6730
+    },
+    {
+      "epoch": 1.3211063162024324,
+      "grad_norm": 15.623889923095703,
+      "learning_rate": 7.58597561953033e-06,
+      "loss": 4.7291,
+      "step": 6735
+    },
+    {
+      "epoch": 1.3220870929776383,
+      "grad_norm": 16.421722412109375,
+      "learning_rate": 7.582662225443631e-06,
+      "loss": 4.87,
+      "step": 6740
+    },
+    {
+      "epoch": 1.3230678697528442,
+      "grad_norm": 23.022462844848633,
+      "learning_rate": 7.579347283760801e-06,
+      "loss": 4.9356,
+      "step": 6745
+    },
+    {
+      "epoch": 1.32404864652805,
+      "grad_norm": 26.935955047607422,
+      "learning_rate": 7.576030796468233e-06,
+      "loss": 4.5722,
+      "step": 6750
+    },
+    {
+      "epoch": 1.3250294233032562,
+      "grad_norm": 16.350265502929688,
+      "learning_rate": 7.572712765553254e-06,
+      "loss": 4.7951,
+      "step": 6755
+    },
+    {
+      "epoch": 1.3260102000784622,
+      "grad_norm": 10.10744571685791,
+      "learning_rate": 7.569393193004109e-06,
+      "loss": 4.9494,
+      "step": 6760
+    },
+    {
+      "epoch": 1.326990976853668,
+      "grad_norm": 20.70772933959961,
+      "learning_rate": 7.56607208080997e-06,
+      "loss": 4.4884,
+      "step": 6765
+    },
+    {
+      "epoch": 1.327971753628874,
+      "grad_norm": 18.814462661743164,
+      "learning_rate": 7.562749430960931e-06,
+      "loss": 4.7232,
+      "step": 6770
+    },
+    {
+      "epoch": 1.32895253040408,
+      "grad_norm": 17.414348602294922,
+      "learning_rate": 7.559425245448006e-06,
+      "loss": 4.7775,
+      "step": 6775
+    },
+    {
+      "epoch": 1.329933307179286,
+      "grad_norm": 36.159873962402344,
+      "learning_rate": 7.556099526263132e-06,
+      "loss": 4.5531,
+      "step": 6780
+    },
+    {
+      "epoch": 1.330914083954492,
+      "grad_norm": 22.765953063964844,
+      "learning_rate": 7.552772275399163e-06,
+      "loss": 4.6333,
+      "step": 6785
+    },
+    {
+      "epoch": 1.3318948607296979,
+      "grad_norm": 17.589569091796875,
+      "learning_rate": 7.549443494849872e-06,
+      "loss": 4.7986,
+      "step": 6790
+    },
+    {
+      "epoch": 1.332875637504904,
+      "grad_norm": 19.393115997314453,
+      "learning_rate": 7.5461131866099465e-06,
+      "loss": 4.722,
+      "step": 6795
+    },
+    {
+      "epoch": 1.33385641428011,
+      "grad_norm": 35.45538330078125,
+      "learning_rate": 7.542781352674994e-06,
+      "loss": 4.5966,
+      "step": 6800
+    },
+    {
+      "epoch": 1.3348371910553158,
+      "grad_norm": 11.774484634399414,
+      "learning_rate": 7.539447995041529e-06,
+      "loss": 4.6637,
+      "step": 6805
+    },
+    {
+      "epoch": 1.3358179678305218,
+      "grad_norm": 26.95338249206543,
+      "learning_rate": 7.536113115706987e-06,
+      "loss": 4.8572,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3367987446057277,
+      "grad_norm": 30.311328887939453,
+      "learning_rate": 7.532776716669708e-06,
+      "loss": 4.8531,
+      "step": 6815
+    },
+    {
+      "epoch": 1.3377795213809338,
+      "grad_norm": 24.06073570251465,
+      "learning_rate": 7.52943879992895e-06,
+      "loss": 4.5024,
+      "step": 6820
+    },
+    {
+      "epoch": 1.3387602981561397,
+      "grad_norm": 37.04559326171875,
+      "learning_rate": 7.526099367484871e-06,
+      "loss": 4.9231,
+      "step": 6825
+    },
+    {
+      "epoch": 1.3397410749313456,
+      "grad_norm": 35.2464485168457,
+      "learning_rate": 7.52275842133855e-06,
+      "loss": 5.2958,
+      "step": 6830
+    },
+    {
+      "epoch": 1.3407218517065516,
+      "grad_norm": 19.340726852416992,
+      "learning_rate": 7.519415963491961e-06,
+      "loss": 5.0359,
+      "step": 6835
+    },
+    {
+      "epoch": 1.3417026284817575,
+      "grad_norm": 29.035587310791016,
+      "learning_rate": 7.516071995947991e-06,
+      "loss": 5.0164,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3426834052569636,
+      "grad_norm": 11.302748680114746,
+      "learning_rate": 7.512726520710429e-06,
+      "loss": 4.809,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3436641820321695,
+      "grad_norm": 49.59445571899414,
+      "learning_rate": 7.5093795397839655e-06,
+      "loss": 4.7413,
+      "step": 6850
+    },
+    {
+      "epoch": 1.3446449588073754,
+      "grad_norm": 24.61562728881836,
+      "learning_rate": 7.5060310551741986e-06,
+      "loss": 4.6153,
+      "step": 6855
+    },
+    {
+      "epoch": 1.3456257355825814,
+      "grad_norm": 20.13814353942871,
+      "learning_rate": 7.5026810688876225e-06,
+      "loss": 5.0,
+      "step": 6860
+    },
+    {
+      "epoch": 1.3466065123577873,
+      "grad_norm": 49.62564468383789,
+      "learning_rate": 7.499329582931636e-06,
+      "loss": 5.0259,
+      "step": 6865
+    },
+    {
+      "epoch": 1.3475872891329934,
+      "grad_norm": 18.966625213623047,
+      "learning_rate": 7.495976599314531e-06,
+      "loss": 4.8656,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3485680659081993,
+      "grad_norm": 28.413414001464844,
+      "learning_rate": 7.4926221200455e-06,
+      "loss": 4.6754,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3495488426834052,
+      "grad_norm": 13.881267547607422,
+      "learning_rate": 7.489266147134631e-06,
+      "loss": 4.3717,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3505296194586112,
+      "grad_norm": 29.266584396362305,
+      "learning_rate": 7.485908682592909e-06,
+      "loss": 4.6337,
+      "step": 6885
+    },
+    {
+      "epoch": 1.351510396233817,
+      "grad_norm": 24.69556999206543,
+      "learning_rate": 7.482549728432211e-06,
+      "loss": 4.604,
+      "step": 6890
+    },
+    {
+      "epoch": 1.3524911730090232,
+      "grad_norm": 34.28022003173828,
+      "learning_rate": 7.479189286665305e-06,
+      "loss": 4.3239,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3534719497842291,
+      "grad_norm": 26.678653717041016,
+      "learning_rate": 7.475827359305853e-06,
+      "loss": 4.7429,
+      "step": 6900
+    },
+    {
+      "epoch": 1.354452726559435,
+      "grad_norm": 21.746240615844727,
+      "learning_rate": 7.472463948368407e-06,
+      "loss": 4.7656,
+      "step": 6905
+    },
+    {
+      "epoch": 1.355433503334641,
+      "grad_norm": 23.173107147216797,
+      "learning_rate": 7.469099055868406e-06,
+      "loss": 4.6698,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3564142801098469,
+      "grad_norm": 38.12723159790039,
+      "learning_rate": 7.465732683822182e-06,
+      "loss": 4.7617,
+      "step": 6915
+    },
+    {
+      "epoch": 1.357395056885053,
+      "grad_norm": 17.623767852783203,
+      "learning_rate": 7.462364834246945e-06,
+      "loss": 4.6744,
+      "step": 6920
+    },
+    {
+      "epoch": 1.358375833660259,
+      "grad_norm": 31.644580841064453,
+      "learning_rate": 7.4589955091607954e-06,
+      "loss": 5.1335,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3593566104354649,
+      "grad_norm": 23.06364631652832,
+      "learning_rate": 7.455624710582721e-06,
+      "loss": 4.7955,
+      "step": 6930
+    },
+    {
+      "epoch": 1.360337387210671,
+      "grad_norm": 12.311569213867188,
+      "learning_rate": 7.452252440532587e-06,
+      "loss": 4.7353,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3613181639858767,
+      "grad_norm": 17.3242244720459,
+      "learning_rate": 7.4488787010311425e-06,
+      "loss": 4.639,
+      "step": 6940
+    },
+    {
+      "epoch": 1.3622989407610828,
+      "grad_norm": 25.3474063873291,
+      "learning_rate": 7.445503494100017e-06,
+      "loss": 4.9055,
+      "step": 6945
+    },
+    {
+      "epoch": 1.3632797175362887,
+      "grad_norm": 21.65390396118164,
+      "learning_rate": 7.442126821761719e-06,
+      "loss": 4.7359,
+      "step": 6950
+    },
+    {
+      "epoch": 1.3642604943114947,
+      "grad_norm": 28.077287673950195,
+      "learning_rate": 7.438748686039637e-06,
+      "loss": 4.9454,
+      "step": 6955
+    },
+    {
+      "epoch": 1.3652412710867008,
+      "grad_norm": 19.87175941467285,
+      "learning_rate": 7.435369088958033e-06,
+      "loss": 4.5149,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3662220478619067,
+      "grad_norm": 23.754520416259766,
+      "learning_rate": 7.431988032542048e-06,
+      "loss": 4.5241,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3672028246371126,
+      "grad_norm": 14.080316543579102,
+      "learning_rate": 7.4286055188176945e-06,
+      "loss": 4.7205,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3681836014123185,
+      "grad_norm": 20.002689361572266,
+      "learning_rate": 7.42522154981186e-06,
+      "loss": 4.9367,
+      "step": 6975
+    },
+    {
+      "epoch": 1.3691643781875245,
+      "grad_norm": 18.009675979614258,
+      "learning_rate": 7.4218361275523046e-06,
+      "loss": 5.1465,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3701451549627306,
+      "grad_norm": 18.44239044189453,
+      "learning_rate": 7.418449254067659e-06,
+      "loss": 4.6064,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3711259317379365,
+      "grad_norm": 23.188587188720703,
+      "learning_rate": 7.415060931387422e-06,
+      "loss": 4.4947,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3721067085131424,
+      "grad_norm": 23.407365798950195,
+      "learning_rate": 7.411671161541961e-06,
+      "loss": 4.5955,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3730874852883483,
+      "grad_norm": 17.974319458007812,
+      "learning_rate": 7.408279946562512e-06,
+      "loss": 4.4836,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3740682620635543,
+      "grad_norm": 17.078502655029297,
+      "learning_rate": 7.404887288481177e-06,
+      "loss": 4.8717,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3750490388387604,
+      "grad_norm": 20.545015335083008,
+      "learning_rate": 7.401493189330921e-06,
+      "loss": 4.7096,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3760298156139663,
+      "grad_norm": 22.460535049438477,
+      "learning_rate": 7.398097651145575e-06,
+      "loss": 4.8465,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3770105923891722,
+      "grad_norm": 21.549362182617188,
+      "learning_rate": 7.3947006759598295e-06,
+      "loss": 4.8413,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3779913691643781,
+      "grad_norm": 22.767526626586914,
+      "learning_rate": 7.391302265809237e-06,
+      "loss": 4.7785,
+      "step": 7025
+    },
+    {
+      "epoch": 1.378972145939584,
+      "grad_norm": 21.575889587402344,
+      "learning_rate": 7.387902422730211e-06,
+      "loss": 4.6862,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3799529227147902,
+      "grad_norm": 11.060517311096191,
+      "learning_rate": 7.384501148760024e-06,
+      "loss": 4.8493,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3809336994899961,
+      "grad_norm": 22.244422912597656,
+      "learning_rate": 7.381098445936803e-06,
+      "loss": 4.609,
+      "step": 7040
+    },
+    {
+      "epoch": 1.381914476265202,
+      "grad_norm": 19.83547019958496,
+      "learning_rate": 7.377694316299533e-06,
+      "loss": 4.8589,
+      "step": 7045
+    },
+    {
+      "epoch": 1.382895253040408,
+      "grad_norm": 26.3979434967041,
+      "learning_rate": 7.374288761888056e-06,
+      "loss": 4.6734,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3838760298156139,
+      "grad_norm": 17.394556045532227,
+      "learning_rate": 7.370881784743065e-06,
+      "loss": 4.7569,
+      "step": 7055
+    },
+    {
+      "epoch": 1.38485680659082,
+      "grad_norm": 39.59242630004883,
+      "learning_rate": 7.367473386906106e-06,
+      "loss": 4.6074,
+      "step": 7060
+    },
+    {
+      "epoch": 1.385837583366026,
+      "grad_norm": 10.681520462036133,
+      "learning_rate": 7.364063570419576e-06,
+      "loss": 4.7027,
+      "step": 7065
+    },
+    {
+      "epoch": 1.3868183601412318,
+      "grad_norm": 17.754701614379883,
+      "learning_rate": 7.360652337326725e-06,
+      "loss": 4.5475,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3877991369164377,
+      "grad_norm": 12.982447624206543,
+      "learning_rate": 7.357239689671646e-06,
+      "loss": 4.4779,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3887799136916437,
+      "grad_norm": 22.83730125427246,
+      "learning_rate": 7.353825629499287e-06,
+      "loss": 4.5497,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3897606904668498,
+      "grad_norm": 23.83551025390625,
+      "learning_rate": 7.350410158855437e-06,
+      "loss": 4.7299,
+      "step": 7085
+    },
+    {
+      "epoch": 1.3907414672420557,
+      "grad_norm": 22.732009887695312,
+      "learning_rate": 7.346993279786732e-06,
+      "loss": 4.5305,
+      "step": 7090
+    },
+    {
+      "epoch": 1.3917222440172616,
+      "grad_norm": 19.805837631225586,
+      "learning_rate": 7.343574994340652e-06,
+      "loss": 4.7342,
+      "step": 7095
+    },
+    {
+      "epoch": 1.3927030207924678,
+      "grad_norm": 30.17318344116211,
+      "learning_rate": 7.340155304565518e-06,
+      "loss": 4.6992,
+      "step": 7100
+    },
+    {
+      "epoch": 1.3936837975676735,
+      "grad_norm": 12.482978820800781,
+      "learning_rate": 7.336734212510497e-06,
+      "loss": 4.5203,
+      "step": 7105
+    },
+    {
+      "epoch": 1.3946645743428796,
+      "grad_norm": 24.85907554626465,
+      "learning_rate": 7.333311720225591e-06,
+      "loss": 4.3356,
+      "step": 7110
+    },
+    {
+      "epoch": 1.3956453511180855,
+      "grad_norm": 33.45339584350586,
+      "learning_rate": 7.329887829761645e-06,
+      "loss": 5.0046,
+      "step": 7115
+    },
+    {
+      "epoch": 1.3966261278932914,
+      "grad_norm": 11.957460403442383,
+      "learning_rate": 7.326462543170339e-06,
+      "loss": 4.6839,
+      "step": 7120
+    },
+    {
+      "epoch": 1.3976069046684976,
+      "grad_norm": 27.296606063842773,
+      "learning_rate": 7.323035862504191e-06,
+      "loss": 4.5867,
+      "step": 7125
+    },
+    {
+      "epoch": 1.3985876814437035,
+      "grad_norm": 15.323542594909668,
+      "learning_rate": 7.319607789816555e-06,
+      "loss": 4.6469,
+      "step": 7130
+    },
+    {
+      "epoch": 1.3995684582189094,
+      "grad_norm": 31.46601676940918,
+      "learning_rate": 7.316178327161618e-06,
+      "loss": 4.7618,
+      "step": 7135
+    },
+    {
+      "epoch": 1.4005492349941153,
+      "grad_norm": 16.684764862060547,
+      "learning_rate": 7.3127474765944004e-06,
+      "loss": 4.5262,
+      "step": 7140
+    },
+    {
+      "epoch": 1.4015300117693212,
+      "grad_norm": 17.97416877746582,
+      "learning_rate": 7.309315240170753e-06,
+      "loss": 4.9621,
+      "step": 7145
+    },
+    {
+      "epoch": 1.4025107885445274,
+      "grad_norm": 13.907105445861816,
+      "learning_rate": 7.305881619947359e-06,
+      "loss": 4.7714,
+      "step": 7150
+    },
+    {
+      "epoch": 1.4034915653197333,
+      "grad_norm": 26.989219665527344,
+      "learning_rate": 7.302446617981731e-06,
+      "loss": 4.7963,
+      "step": 7155
+    },
+    {
+      "epoch": 1.4044723420949392,
+      "grad_norm": 11.415125846862793,
+      "learning_rate": 7.2990102363322065e-06,
+      "loss": 4.6166,
+      "step": 7160
+    },
+    {
+      "epoch": 1.4054531188701451,
+      "grad_norm": 15.411608695983887,
+      "learning_rate": 7.295572477057952e-06,
+      "loss": 4.8559,
+      "step": 7165
+    },
+    {
+      "epoch": 1.406433895645351,
+      "grad_norm": 18.74057388305664,
+      "learning_rate": 7.292133342218963e-06,
+      "loss": 4.8423,
+      "step": 7170
+    },
+    {
+      "epoch": 1.4074146724205572,
+      "grad_norm": 15.429743766784668,
+      "learning_rate": 7.28869283387605e-06,
+      "loss": 4.6179,
+      "step": 7175
+    },
+    {
+      "epoch": 1.408395449195763,
+      "grad_norm": 20.775442123413086,
+      "learning_rate": 7.2852509540908546e-06,
+      "loss": 4.5703,
+      "step": 7180
+    },
+    {
+      "epoch": 1.409376225970969,
+      "grad_norm": 10.963190078735352,
+      "learning_rate": 7.281807704925839e-06,
+      "loss": 4.5943,
+      "step": 7185
+    },
+    {
+      "epoch": 1.410357002746175,
+      "grad_norm": 22.696077346801758,
+      "learning_rate": 7.278363088444283e-06,
+      "loss": 4.8727,
+      "step": 7190
+    },
+    {
+      "epoch": 1.4113377795213808,
+      "grad_norm": 22.353727340698242,
+      "learning_rate": 7.2749171067102875e-06,
+      "loss": 4.9767,
+      "step": 7195
+    },
+    {
+      "epoch": 1.412318556296587,
+      "grad_norm": 24.793746948242188,
+      "learning_rate": 7.271469761788772e-06,
+      "loss": 4.4291,
+      "step": 7200
+    },
+    {
+      "epoch": 1.413299333071793,
+      "grad_norm": 24.146011352539062,
+      "learning_rate": 7.2680210557454715e-06,
+      "loss": 4.646,
+      "step": 7205
+    },
+    {
+      "epoch": 1.4142801098469988,
+      "grad_norm": 20.337390899658203,
+      "learning_rate": 7.264570990646938e-06,
+      "loss": 5.1823,
+      "step": 7210
+    },
+    {
+      "epoch": 1.4152608866222047,
+      "grad_norm": 31.162933349609375,
+      "learning_rate": 7.261119568560537e-06,
+      "loss": 4.5848,
+      "step": 7215
+    },
+    {
+      "epoch": 1.4162416633974106,
+      "grad_norm": 17.75670623779297,
+      "learning_rate": 7.257666791554448e-06,
+      "loss": 4.6185,
+      "step": 7220
+    },
+    {
+      "epoch": 1.4172224401726168,
+      "grad_norm": 13.3048734664917,
+      "learning_rate": 7.2542126616976596e-06,
+      "loss": 4.7974,
+      "step": 7225
+    },
+    {
+      "epoch": 1.4182032169478227,
+      "grad_norm": 15.745701789855957,
+      "learning_rate": 7.2507571810599755e-06,
+      "loss": 4.5467,
+      "step": 7230
+    },
+    {
+      "epoch": 1.4191839937230286,
+      "grad_norm": 12.2063627243042,
+      "learning_rate": 7.247300351712007e-06,
+      "loss": 4.6024,
+      "step": 7235
+    },
+    {
+      "epoch": 1.4201647704982345,
+      "grad_norm": 36.18549346923828,
+      "learning_rate": 7.243842175725172e-06,
+      "loss": 4.9873,
+      "step": 7240
+    },
+    {
+      "epoch": 1.4211455472734404,
+      "grad_norm": 26.773792266845703,
+      "learning_rate": 7.240382655171696e-06,
+      "loss": 4.6385,
+      "step": 7245
+    },
+    {
+      "epoch": 1.4221263240486466,
+      "grad_norm": 21.363792419433594,
+      "learning_rate": 7.236921792124611e-06,
+      "loss": 4.647,
+      "step": 7250
+    },
+    {
+      "epoch": 1.4231071008238525,
+      "grad_norm": 15.612807273864746,
+      "learning_rate": 7.233459588657753e-06,
+      "loss": 4.7398,
+      "step": 7255
+    },
+    {
+      "epoch": 1.4240878775990584,
+      "grad_norm": 42.73514175415039,
+      "learning_rate": 7.229996046845762e-06,
+      "loss": 4.9747,
+      "step": 7260
+    },
+    {
+      "epoch": 1.4250686543742646,
+      "grad_norm": 26.65818977355957,
+      "learning_rate": 7.226531168764079e-06,
+      "loss": 5.0051,
+      "step": 7265
+    },
+    {
+      "epoch": 1.4260494311494702,
+      "grad_norm": 15.626964569091797,
+      "learning_rate": 7.223064956488946e-06,
+      "loss": 4.618,
+      "step": 7270
+    },
+    {
+      "epoch": 1.4270302079246764,
+      "grad_norm": 10.92302417755127,
+      "learning_rate": 7.219597412097405e-06,
+      "loss": 5.0778,
+      "step": 7275
+    },
+    {
+      "epoch": 1.4280109846998823,
+      "grad_norm": 19.099761962890625,
+      "learning_rate": 7.216128537667296e-06,
+      "loss": 4.7971,
+      "step": 7280
+    },
+    {
+      "epoch": 1.4289917614750882,
+      "grad_norm": 16.517105102539062,
+      "learning_rate": 7.212658335277255e-06,
+      "loss": 4.3451,
+      "step": 7285
+    },
+    {
+      "epoch": 1.4299725382502944,
+      "grad_norm": 30.557945251464844,
+      "learning_rate": 7.209186807006714e-06,
+      "loss": 4.6488,
+      "step": 7290
+    },
+    {
+      "epoch": 1.4309533150255003,
+      "grad_norm": 10.725804328918457,
+      "learning_rate": 7.205713954935901e-06,
+      "loss": 4.7013,
+      "step": 7295
+    },
+    {
+      "epoch": 1.4319340918007062,
+      "grad_norm": 25.416257858276367,
+      "learning_rate": 7.202239781145834e-06,
+      "loss": 4.814,
+      "step": 7300
+    },
+    {
+      "epoch": 1.432914868575912,
+      "grad_norm": 28.029569625854492,
+      "learning_rate": 7.19876428771833e-06,
+      "loss": 4.6645,
+      "step": 7305
+    },
+    {
+      "epoch": 1.433895645351118,
+      "grad_norm": 11.939929962158203,
+      "learning_rate": 7.195287476735989e-06,
+      "loss": 5.2664,
+      "step": 7310
+    },
+    {
+      "epoch": 1.4348764221263242,
+      "grad_norm": 15.845399856567383,
+      "learning_rate": 7.191809350282204e-06,
+      "loss": 4.6847,
+      "step": 7315
+    },
+    {
+      "epoch": 1.43585719890153,
+      "grad_norm": 15.885704040527344,
+      "learning_rate": 7.188329910441154e-06,
+      "loss": 4.6084,
+      "step": 7320
+    },
+    {
+      "epoch": 1.436837975676736,
+      "grad_norm": 36.25415802001953,
+      "learning_rate": 7.184849159297809e-06,
+      "loss": 5.2779,
+      "step": 7325
+    },
+    {
+      "epoch": 1.437818752451942,
+      "grad_norm": 15.539462089538574,
+      "learning_rate": 7.1813670989379215e-06,
+      "loss": 4.3759,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4387995292271478,
+      "grad_norm": 16.43358612060547,
+      "learning_rate": 7.177883731448031e-06,
+      "loss": 4.4987,
+      "step": 7335
+    },
+    {
+      "epoch": 1.439780306002354,
+      "grad_norm": 14.630133628845215,
+      "learning_rate": 7.174399058915458e-06,
+      "loss": 4.4032,
+      "step": 7340
+    },
+    {
+      "epoch": 1.4407610827775599,
+      "grad_norm": 23.953678131103516,
+      "learning_rate": 7.170913083428306e-06,
+      "loss": 4.9749,
+      "step": 7345
+    },
+    {
+      "epoch": 1.4417418595527658,
+      "grad_norm": 20.43268585205078,
+      "learning_rate": 7.167425807075459e-06,
+      "loss": 4.644,
+      "step": 7350
+    },
+    {
+      "epoch": 1.4427226363279717,
+      "grad_norm": 20.701139450073242,
+      "learning_rate": 7.163937231946581e-06,
+      "loss": 5.2018,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4437034131031776,
+      "grad_norm": 25.42399024963379,
+      "learning_rate": 7.1604473601321125e-06,
+      "loss": 4.5791,
+      "step": 7360
+    },
+    {
+      "epoch": 1.4446841898783838,
+      "grad_norm": 10.516730308532715,
+      "learning_rate": 7.156956193723275e-06,
+      "loss": 4.7935,
+      "step": 7365
+    },
+    {
+      "epoch": 1.4456649666535897,
+      "grad_norm": 31.056705474853516,
+      "learning_rate": 7.153463734812059e-06,
+      "loss": 4.721,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4466457434287956,
+      "grad_norm": 18.38773536682129,
+      "learning_rate": 7.1499699854912385e-06,
+      "loss": 4.4783,
+      "step": 7375
+    },
+    {
+      "epoch": 1.4476265202040015,
+      "grad_norm": 21.014942169189453,
+      "learning_rate": 7.146474947854354e-06,
+      "loss": 4.7552,
+      "step": 7380
+    },
+    {
+      "epoch": 1.4486072969792074,
+      "grad_norm": 26.835750579833984,
+      "learning_rate": 7.1429786239957195e-06,
+      "loss": 4.5055,
+      "step": 7385
+    },
+    {
+      "epoch": 1.4495880737544136,
+      "grad_norm": 23.621606826782227,
+      "learning_rate": 7.13948101601042e-06,
+      "loss": 5.0276,
+      "step": 7390
+    },
+    {
+      "epoch": 1.4505688505296195,
+      "grad_norm": 22.148571014404297,
+      "learning_rate": 7.135982125994311e-06,
+      "loss": 4.6589,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4515496273048254,
+      "grad_norm": 15.912561416625977,
+      "learning_rate": 7.132481956044013e-06,
+      "loss": 4.7373,
+      "step": 7400
+    },
+    {
+      "epoch": 1.4525304040800313,
+      "grad_norm": 15.783080101013184,
+      "learning_rate": 7.128980508256919e-06,
+      "loss": 4.763,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4535111808552372,
+      "grad_norm": 14.95667839050293,
+      "learning_rate": 7.125477784731184e-06,
+      "loss": 4.959,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4544919576304434,
+      "grad_norm": 23.89700698852539,
+      "learning_rate": 7.121973787565727e-06,
+      "loss": 4.8652,
+      "step": 7415
+    },
+    {
+      "epoch": 1.4554727344056493,
+      "grad_norm": 23.417661666870117,
+      "learning_rate": 7.118468518860232e-06,
+      "loss": 4.697,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4564535111808552,
+      "grad_norm": 23.191051483154297,
+      "learning_rate": 7.114961980715142e-06,
+      "loss": 4.4113,
+      "step": 7425
+    },
+    {
+      "epoch": 1.4574342879560613,
+      "grad_norm": 12.547563552856445,
+      "learning_rate": 7.111454175231664e-06,
+      "loss": 4.427,
+      "step": 7430
+    },
+    {
+      "epoch": 1.458415064731267,
+      "grad_norm": 22.79470443725586,
+      "learning_rate": 7.107945104511766e-06,
+      "loss": 4.7032,
+      "step": 7435
+    },
+    {
+      "epoch": 1.4593958415064732,
+      "grad_norm": 21.279233932495117,
+      "learning_rate": 7.1044347706581664e-06,
+      "loss": 4.9074,
+      "step": 7440
+    },
+    {
+      "epoch": 1.460376618281679,
+      "grad_norm": 24.7849178314209,
+      "learning_rate": 7.10092317577435e-06,
+      "loss": 4.7103,
+      "step": 7445
+    },
+    {
+      "epoch": 1.461357395056885,
+      "grad_norm": 12.6660795211792,
+      "learning_rate": 7.09741032196455e-06,
+      "loss": 4.7538,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4623381718320911,
+      "grad_norm": 21.938674926757812,
+      "learning_rate": 7.093896211333757e-06,
+      "loss": 5.1549,
+      "step": 7455
+    },
+    {
+      "epoch": 1.463318948607297,
+      "grad_norm": 25.91387176513672,
+      "learning_rate": 7.090380845987716e-06,
+      "loss": 4.6054,
+      "step": 7460
+    },
+    {
+      "epoch": 1.464299725382503,
+      "grad_norm": 23.84564971923828,
+      "learning_rate": 7.08686422803292e-06,
+      "loss": 4.6331,
+      "step": 7465
+    },
+    {
+      "epoch": 1.4652805021577089,
+      "grad_norm": 18.71299171447754,
+      "learning_rate": 7.083346359576617e-06,
+      "loss": 4.8511,
+      "step": 7470
+    },
+    {
+      "epoch": 1.4662612789329148,
+      "grad_norm": 13.39997673034668,
+      "learning_rate": 7.079827242726801e-06,
+      "loss": 4.7739,
+      "step": 7475
+    },
+    {
+      "epoch": 1.467242055708121,
+      "grad_norm": 21.340524673461914,
+      "learning_rate": 7.076306879592215e-06,
+      "loss": 4.7725,
+      "step": 7480
+    },
+    {
+      "epoch": 1.4682228324833269,
+      "grad_norm": 24.9217529296875,
+      "learning_rate": 7.072785272282351e-06,
+      "loss": 4.7112,
+      "step": 7485
+    },
+    {
+      "epoch": 1.4692036092585328,
+      "grad_norm": 18.006662368774414,
+      "learning_rate": 7.069262422907444e-06,
+      "loss": 4.5367,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4701843860337387,
+      "grad_norm": 17.566797256469727,
+      "learning_rate": 7.065738333578473e-06,
+      "loss": 4.7325,
+      "step": 7495
+    },
+    {
+      "epoch": 1.4711651628089446,
+      "grad_norm": 16.905597686767578,
+      "learning_rate": 7.0622130064071584e-06,
+      "loss": 4.6231,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4721459395841507,
+      "grad_norm": 13.76935863494873,
+      "learning_rate": 7.05868644350597e-06,
+      "loss": 4.7777,
+      "step": 7505
+    },
+    {
+      "epoch": 1.4731267163593567,
+      "grad_norm": 20.6540470123291,
+      "learning_rate": 7.05515864698811e-06,
+      "loss": 4.6186,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4741074931345626,
+      "grad_norm": 19.557889938354492,
+      "learning_rate": 7.051629618967523e-06,
+      "loss": 4.957,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4750882699097685,
+      "grad_norm": 35.486270904541016,
+      "learning_rate": 7.048099361558892e-06,
+      "loss": 4.5927,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4760690466849744,
+      "grad_norm": 23.648691177368164,
+      "learning_rate": 7.044567876877636e-06,
+      "loss": 4.7108,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4770498234601805,
+      "grad_norm": 15.685006141662598,
+      "learning_rate": 7.041035167039909e-06,
+      "loss": 4.689,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4780306002353865,
+      "grad_norm": 21.108903884887695,
+      "learning_rate": 7.037501234162599e-06,
+      "loss": 4.4376,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4790113770105924,
+      "grad_norm": 16.158863067626953,
+      "learning_rate": 7.033966080363328e-06,
+      "loss": 4.6436,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4799921537857983,
+      "grad_norm": 18.32492446899414,
+      "learning_rate": 7.03042970776045e-06,
+      "loss": 4.5336,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4809729305610042,
+      "grad_norm": 15.529178619384766,
+      "learning_rate": 7.026892118473045e-06,
+      "loss": 4.6415,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4819537073362103,
+      "grad_norm": 23.759431838989258,
+      "learning_rate": 7.023353314620931e-06,
+      "loss": 4.6697,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4829344841114163,
+      "grad_norm": 14.494952201843262,
+      "learning_rate": 7.019813298324642e-06,
+      "loss": 5.0709,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4839152608866222,
+      "grad_norm": 23.806947708129883,
+      "learning_rate": 7.016272071705452e-06,
+      "loss": 4.5083,
+      "step": 7565
+    },
+    {
+      "epoch": 1.484896037661828,
+      "grad_norm": 17.701692581176758,
+      "learning_rate": 7.012729636885346e-06,
+      "loss": 4.7351,
+      "step": 7570
+    },
+    {
+      "epoch": 1.485876814437034,
+      "grad_norm": 13.665419578552246,
+      "learning_rate": 7.009185995987042e-06,
+      "loss": 4.5931,
+      "step": 7575
+    },
+    {
+      "epoch": 1.4868575912122401,
+      "grad_norm": 25.867773056030273,
+      "learning_rate": 7.0056411511339805e-06,
+      "loss": 5.2031,
+      "step": 7580
+    },
+    {
+      "epoch": 1.487838367987446,
+      "grad_norm": 23.662742614746094,
+      "learning_rate": 7.002095104450322e-06,
+      "loss": 4.4658,
+      "step": 7585
+    },
+    {
+      "epoch": 1.488819144762652,
+      "grad_norm": 53.169151306152344,
+      "learning_rate": 6.998547858060944e-06,
+      "loss": 5.3187,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4897999215378581,
+      "grad_norm": 15.928996086120605,
+      "learning_rate": 6.994999414091448e-06,
+      "loss": 4.9528,
+      "step": 7595
+    },
+    {
+      "epoch": 1.4907806983130638,
+      "grad_norm": 28.862323760986328,
+      "learning_rate": 6.991449774668149e-06,
+      "loss": 4.8229,
+      "step": 7600
+    },
+    {
+      "epoch": 1.49176147508827,
+      "grad_norm": 23.40715789794922,
+      "learning_rate": 6.987898941918082e-06,
+      "loss": 4.4993,
+      "step": 7605
+    },
+    {
+      "epoch": 1.4927422518634759,
+      "grad_norm": 10.611961364746094,
+      "learning_rate": 6.984346917968994e-06,
+      "loss": 4.3586,
+      "step": 7610
+    },
+    {
+      "epoch": 1.4937230286386818,
+      "grad_norm": 20.525981903076172,
+      "learning_rate": 6.980793704949348e-06,
+      "loss": 4.4613,
+      "step": 7615
+    },
+    {
+      "epoch": 1.494703805413888,
+      "grad_norm": 17.283946990966797,
+      "learning_rate": 6.977239304988318e-06,
+      "loss": 4.8228,
+      "step": 7620
+    },
+    {
+      "epoch": 1.4956845821890938,
+      "grad_norm": 21.752166748046875,
+      "learning_rate": 6.973683720215789e-06,
+      "loss": 4.7708,
+      "step": 7625
+    },
+    {
+      "epoch": 1.4966653589642998,
+      "grad_norm": 20.8831787109375,
+      "learning_rate": 6.970126952762359e-06,
+      "loss": 4.7839,
+      "step": 7630
+    },
+    {
+      "epoch": 1.4976461357395057,
+      "grad_norm": 13.42829418182373,
+      "learning_rate": 6.966569004759331e-06,
+      "loss": 4.7172,
+      "step": 7635
+    },
+    {
+      "epoch": 1.4986269125147116,
+      "grad_norm": 21.750574111938477,
+      "learning_rate": 6.963009878338718e-06,
+      "loss": 4.6011,
+      "step": 7640
+    },
+    {
+      "epoch": 1.4996076892899177,
+      "grad_norm": 19.537803649902344,
+      "learning_rate": 6.959449575633236e-06,
+      "loss": 4.7368,
+      "step": 7645
+    },
+    {
+      "epoch": 1.5005884660651234,
+      "grad_norm": 21.34620475769043,
+      "learning_rate": 6.955888098776308e-06,
+      "loss": 4.6103,
+      "step": 7650
+    },
+    {
+      "epoch": 1.5005884660651234,
+      "eval_loss": 4.907505512237549,
+      "eval_runtime": 7.6604,
+      "eval_samples_per_second": 27.283,
+      "eval_steps_per_second": 13.707,
+      "step": 7650
+    },
+    {
+      "epoch": 1.5015692428403296,
+      "grad_norm": 13.40380573272705,
+      "learning_rate": 6.952325449902062e-06,
+      "loss": 4.6435,
+      "step": 7655
+    },
+    {
+      "epoch": 1.5025500196155355,
+      "grad_norm": 14.482090950012207,
+      "learning_rate": 6.948761631145327e-06,
+      "loss": 4.8812,
+      "step": 7660
+    },
+    {
+      "epoch": 1.5035307963907414,
+      "grad_norm": 22.7137451171875,
+      "learning_rate": 6.945196644641631e-06,
+      "loss": 4.8619,
+      "step": 7665
+    },
+    {
+      "epoch": 1.5045115731659475,
+      "grad_norm": 19.11748504638672,
+      "learning_rate": 6.941630492527205e-06,
+      "loss": 5.1203,
+      "step": 7670
+    },
+    {
+      "epoch": 1.5054923499411534,
+      "grad_norm": 24.688961029052734,
+      "learning_rate": 6.938063176938976e-06,
+      "loss": 4.7891,
+      "step": 7675
+    },
+    {
+      "epoch": 1.5064731267163594,
+      "grad_norm": 20.401952743530273,
+      "learning_rate": 6.934494700014572e-06,
+      "loss": 4.5722,
+      "step": 7680
+    },
+    {
+      "epoch": 1.5074539034915653,
+      "grad_norm": 17.21699333190918,
+      "learning_rate": 6.9309250638923085e-06,
+      "loss": 4.588,
+      "step": 7685
+    },
+    {
+      "epoch": 1.5084346802667712,
+      "grad_norm": 29.611526489257812,
+      "learning_rate": 6.927354270711206e-06,
+      "loss": 4.4101,
+      "step": 7690
+    },
+    {
+      "epoch": 1.5094154570419773,
+      "grad_norm": 19.163101196289062,
+      "learning_rate": 6.923782322610972e-06,
+      "loss": 4.6286,
+      "step": 7695
+    },
+    {
+      "epoch": 1.5103962338171832,
+      "grad_norm": 20.316064834594727,
+      "learning_rate": 6.920209221732007e-06,
+      "loss": 4.7434,
+      "step": 7700
+    },
+    {
+      "epoch": 1.5113770105923892,
+      "grad_norm": 21.396059036254883,
+      "learning_rate": 6.916634970215406e-06,
+      "loss": 4.7623,
+      "step": 7705
+    },
+    {
+      "epoch": 1.5123577873675953,
+      "grad_norm": 33.224063873291016,
+      "learning_rate": 6.913059570202945e-06,
+      "loss": 4.6921,
+      "step": 7710
+    },
+    {
+      "epoch": 1.513338564142801,
+      "grad_norm": 19.221223831176758,
+      "learning_rate": 6.909483023837098e-06,
+      "loss": 4.6417,
+      "step": 7715
+    },
+    {
+      "epoch": 1.5143193409180071,
+      "grad_norm": 14.870227813720703,
+      "learning_rate": 6.905905333261019e-06,
+      "loss": 4.8107,
+      "step": 7720
+    },
+    {
+      "epoch": 1.515300117693213,
+      "grad_norm": 14.234113693237305,
+      "learning_rate": 6.90232650061855e-06,
+      "loss": 4.6731,
+      "step": 7725
+    },
+    {
+      "epoch": 1.516280894468419,
+      "grad_norm": 12.864129066467285,
+      "learning_rate": 6.898746528054221e-06,
+      "loss": 4.7254,
+      "step": 7730
+    },
+    {
+      "epoch": 1.517261671243625,
+      "grad_norm": 32.53975296020508,
+      "learning_rate": 6.895165417713238e-06,
+      "loss": 4.6027,
+      "step": 7735
+    },
+    {
+      "epoch": 1.5182424480188308,
+      "grad_norm": 17.56818199157715,
+      "learning_rate": 6.891583171741494e-06,
+      "loss": 4.6698,
+      "step": 7740
+    },
+    {
+      "epoch": 1.519223224794037,
+      "grad_norm": 24.81783676147461,
+      "learning_rate": 6.88799979228556e-06,
+      "loss": 4.8477,
+      "step": 7745
+    },
+    {
+      "epoch": 1.5202040015692428,
+      "grad_norm": 12.959918975830078,
+      "learning_rate": 6.884415281492686e-06,
+      "loss": 4.8092,
+      "step": 7750
+    },
+    {
+      "epoch": 1.5211847783444488,
+      "grad_norm": 26.625934600830078,
+      "learning_rate": 6.880829641510805e-06,
+      "loss": 4.9008,
+      "step": 7755
+    },
+    {
+      "epoch": 1.522165555119655,
+      "grad_norm": 24.741802215576172,
+      "learning_rate": 6.877242874488518e-06,
+      "loss": 4.4703,
+      "step": 7760
+    },
+    {
+      "epoch": 1.5231463318948606,
+      "grad_norm": 21.69708824157715,
+      "learning_rate": 6.873654982575108e-06,
+      "loss": 4.6701,
+      "step": 7765
+    },
+    {
+      "epoch": 1.5241271086700667,
+      "grad_norm": 30.454282760620117,
+      "learning_rate": 6.8700659679205296e-06,
+      "loss": 4.9591,
+      "step": 7770
+    },
+    {
+      "epoch": 1.5251078854452726,
+      "grad_norm": 25.414043426513672,
+      "learning_rate": 6.866475832675412e-06,
+      "loss": 4.8959,
+      "step": 7775
+    },
+    {
+      "epoch": 1.5260886622204786,
+      "grad_norm": 27.82634925842285,
+      "learning_rate": 6.862884578991054e-06,
+      "loss": 4.9565,
+      "step": 7780
+    },
+    {
+      "epoch": 1.5270694389956847,
+      "grad_norm": 37.48154067993164,
+      "learning_rate": 6.859292209019424e-06,
+      "loss": 4.6826,
+      "step": 7785
+    },
+    {
+      "epoch": 1.5280502157708904,
+      "grad_norm": 14.848986625671387,
+      "learning_rate": 6.85569872491316e-06,
+      "loss": 4.6118,
+      "step": 7790
+    },
+    {
+      "epoch": 1.5290309925460965,
+      "grad_norm": 20.567981719970703,
+      "learning_rate": 6.85210412882557e-06,
+      "loss": 4.7033,
+      "step": 7795
+    },
+    {
+      "epoch": 1.5300117693213025,
+      "grad_norm": 37.882362365722656,
+      "learning_rate": 6.848508422910622e-06,
+      "loss": 4.9443,
+      "step": 7800
+    },
+    {
+      "epoch": 1.5309925460965084,
+      "grad_norm": 15.06411075592041,
+      "learning_rate": 6.8449116093229605e-06,
+      "loss": 4.3824,
+      "step": 7805
+    },
+    {
+      "epoch": 1.5319733228717145,
+      "grad_norm": 11.747258186340332,
+      "learning_rate": 6.841313690217881e-06,
+      "loss": 4.6974,
+      "step": 7810
+    },
+    {
+      "epoch": 1.5329540996469202,
+      "grad_norm": 13.8905611038208,
+      "learning_rate": 6.837714667751351e-06,
+      "loss": 4.7445,
+      "step": 7815
+    },
+    {
+      "epoch": 1.5339348764221263,
+      "grad_norm": 19.0726318359375,
+      "learning_rate": 6.834114544079993e-06,
+      "loss": 4.927,
+      "step": 7820
+    },
+    {
+      "epoch": 1.5349156531973323,
+      "grad_norm": 17.133312225341797,
+      "learning_rate": 6.830513321361089e-06,
+      "loss": 4.8282,
+      "step": 7825
+    },
+    {
+      "epoch": 1.5358964299725382,
+      "grad_norm": 12.729392051696777,
+      "learning_rate": 6.826911001752586e-06,
+      "loss": 4.7162,
+      "step": 7830
+    },
+    {
+      "epoch": 1.5368772067477443,
+      "grad_norm": 30.635669708251953,
+      "learning_rate": 6.823307587413084e-06,
+      "loss": 4.5225,
+      "step": 7835
+    },
+    {
+      "epoch": 1.5378579835229502,
+      "grad_norm": 17.1396427154541,
+      "learning_rate": 6.8197030805018385e-06,
+      "loss": 4.9225,
+      "step": 7840
+    },
+    {
+      "epoch": 1.5388387602981561,
+      "grad_norm": 16.432329177856445,
+      "learning_rate": 6.8160974831787605e-06,
+      "loss": 4.5991,
+      "step": 7845
+    },
+    {
+      "epoch": 1.539819537073362,
+      "grad_norm": 13.991456031799316,
+      "learning_rate": 6.812490797604416e-06,
+      "loss": 4.5396,
+      "step": 7850
+    },
+    {
+      "epoch": 1.540800313848568,
+      "grad_norm": 20.396547317504883,
+      "learning_rate": 6.808883025940019e-06,
+      "loss": 4.8735,
+      "step": 7855
+    },
+    {
+      "epoch": 1.541781090623774,
+      "grad_norm": 18.803667068481445,
+      "learning_rate": 6.805274170347441e-06,
+      "loss": 4.9716,
+      "step": 7860
+    },
+    {
+      "epoch": 1.54276186739898,
+      "grad_norm": 16.86100196838379,
+      "learning_rate": 6.801664232989196e-06,
+      "loss": 4.5278,
+      "step": 7865
+    },
+    {
+      "epoch": 1.543742644174186,
+      "grad_norm": 19.707059860229492,
+      "learning_rate": 6.798053216028448e-06,
+      "loss": 4.719,
+      "step": 7870
+    },
+    {
+      "epoch": 1.544723420949392,
+      "grad_norm": 17.481386184692383,
+      "learning_rate": 6.794441121629013e-06,
+      "loss": 4.5006,
+      "step": 7875
+    },
+    {
+      "epoch": 1.5457041977245978,
+      "grad_norm": 11.351982116699219,
+      "learning_rate": 6.790827951955345e-06,
+      "loss": 4.8233,
+      "step": 7880
+    },
+    {
+      "epoch": 1.546684974499804,
+      "grad_norm": 15.135643005371094,
+      "learning_rate": 6.787213709172551e-06,
+      "loss": 4.9262,
+      "step": 7885
+    },
+    {
+      "epoch": 1.5476657512750098,
+      "grad_norm": 11.407129287719727,
+      "learning_rate": 6.783598395446371e-06,
+      "loss": 5.4497,
+      "step": 7890
+    },
+    {
+      "epoch": 1.5486465280502157,
+      "grad_norm": 13.211956024169922,
+      "learning_rate": 6.779982012943195e-06,
+      "loss": 4.975,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5496273048254219,
+      "grad_norm": 23.667285919189453,
+      "learning_rate": 6.776364563830047e-06,
+      "loss": 4.7498,
+      "step": 7900
+    },
+    {
+      "epoch": 1.5506080816006276,
+      "grad_norm": 16.621427536010742,
+      "learning_rate": 6.772746050274598e-06,
+      "loss": 4.6203,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5515888583758337,
+      "grad_norm": 14.443175315856934,
+      "learning_rate": 6.769126474445149e-06,
+      "loss": 4.4905,
+      "step": 7910
+    },
+    {
+      "epoch": 1.5525696351510396,
+      "grad_norm": 20.291501998901367,
+      "learning_rate": 6.765505838510642e-06,
+      "loss": 4.7088,
+      "step": 7915
+    },
+    {
+      "epoch": 1.5535504119262455,
+      "grad_norm": 19.331323623657227,
+      "learning_rate": 6.761884144640652e-06,
+      "loss": 4.5726,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5545311887014517,
+      "grad_norm": 22.91404914855957,
+      "learning_rate": 6.758261395005391e-06,
+      "loss": 4.7076,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5555119654766574,
+      "grad_norm": 24.691198348999023,
+      "learning_rate": 6.7546375917757e-06,
+      "loss": 4.8906,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5564927422518635,
+      "grad_norm": 28.088998794555664,
+      "learning_rate": 6.751012737123054e-06,
+      "loss": 4.6132,
+      "step": 7935
+    },
+    {
+      "epoch": 1.5574735190270694,
+      "grad_norm": 10.642810821533203,
+      "learning_rate": 6.747386833219556e-06,
+      "loss": 4.9542,
+      "step": 7940
+    },
+    {
+      "epoch": 1.5584542958022753,
+      "grad_norm": 13.011480331420898,
+      "learning_rate": 6.7437598822379405e-06,
+      "loss": 4.631,
+      "step": 7945
+    },
+    {
+      "epoch": 1.5594350725774815,
+      "grad_norm": 29.10993194580078,
+      "learning_rate": 6.740131886351564e-06,
+      "loss": 4.5673,
+      "step": 7950
+    },
+    {
+      "epoch": 1.5604158493526872,
+      "grad_norm": 17.598003387451172,
+      "learning_rate": 6.736502847734417e-06,
+      "loss": 4.6614,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5613966261278933,
+      "grad_norm": 20.10211753845215,
+      "learning_rate": 6.732872768561111e-06,
+      "loss": 4.5551,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5623774029030992,
+      "grad_norm": 38.78089141845703,
+      "learning_rate": 6.729241651006876e-06,
+      "loss": 4.7825,
+      "step": 7965
+    },
+    {
+      "epoch": 1.5633581796783051,
+      "grad_norm": 10.49329662322998,
+      "learning_rate": 6.725609497247573e-06,
+      "loss": 5.0862,
+      "step": 7970
+    },
+    {
+      "epoch": 1.5643389564535113,
+      "grad_norm": 26.494342803955078,
+      "learning_rate": 6.721976309459677e-06,
+      "loss": 4.8202,
+      "step": 7975
+    },
+    {
+      "epoch": 1.565319733228717,
+      "grad_norm": 24.8291015625,
+      "learning_rate": 6.718342089820288e-06,
+      "loss": 4.8502,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5663005100039231,
+      "grad_norm": 22.414064407348633,
+      "learning_rate": 6.714706840507122e-06,
+      "loss": 4.8922,
+      "step": 7985
+    },
+    {
+      "epoch": 1.567281286779129,
+      "grad_norm": 14.861505508422852,
+      "learning_rate": 6.711070563698508e-06,
+      "loss": 4.8528,
+      "step": 7990
+    },
+    {
+      "epoch": 1.568262063554335,
+      "grad_norm": 18.311233520507812,
+      "learning_rate": 6.707433261573399e-06,
+      "loss": 4.4056,
+      "step": 7995
+    },
+    {
+      "epoch": 1.569242840329541,
+      "grad_norm": 22.220447540283203,
+      "learning_rate": 6.703794936311354e-06,
+      "loss": 4.7109,
+      "step": 8000
+    },
+    {
+      "epoch": 1.570223617104747,
+      "grad_norm": 12.350397109985352,
+      "learning_rate": 6.700155590092553e-06,
+      "loss": 4.8512,
+      "step": 8005
+    },
+    {
+      "epoch": 1.571204393879953,
+      "grad_norm": 15.76508903503418,
+      "learning_rate": 6.6965152250977805e-06,
+      "loss": 4.8235,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5721851706551588,
+      "grad_norm": 16.39112663269043,
+      "learning_rate": 6.692873843508436e-06,
+      "loss": 4.5981,
+      "step": 8015
+    },
+    {
+      "epoch": 1.5731659474303648,
+      "grad_norm": 27.190099716186523,
+      "learning_rate": 6.689231447506527e-06,
+      "loss": 4.6361,
+      "step": 8020
+    },
+    {
+      "epoch": 1.574146724205571,
+      "grad_norm": 11.988786697387695,
+      "learning_rate": 6.685588039274666e-06,
+      "loss": 4.5448,
+      "step": 8025
+    },
+    {
+      "epoch": 1.5751275009807768,
+      "grad_norm": 15.980277061462402,
+      "learning_rate": 6.681943620996081e-06,
+      "loss": 4.8218,
+      "step": 8030
+    },
+    {
+      "epoch": 1.5761082777559827,
+      "grad_norm": 16.668498992919922,
+      "learning_rate": 6.678298194854594e-06,
+      "loss": 4.2428,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5770890545311889,
+      "grad_norm": 14.539552688598633,
+      "learning_rate": 6.674651763034636e-06,
+      "loss": 4.7311,
+      "step": 8040
+    },
+    {
+      "epoch": 1.5780698313063946,
+      "grad_norm": 21.027690887451172,
+      "learning_rate": 6.671004327721243e-06,
+      "loss": 4.3581,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5790506080816007,
+      "grad_norm": 14.93703556060791,
+      "learning_rate": 6.667355891100049e-06,
+      "loss": 4.5448,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5800313848568066,
+      "grad_norm": 30.272512435913086,
+      "learning_rate": 6.663706455357288e-06,
+      "loss": 4.6557,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5810121616320125,
+      "grad_norm": 21.9456787109375,
+      "learning_rate": 6.660056022679795e-06,
+      "loss": 4.483,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5819929384072187,
+      "grad_norm": 13.090523719787598,
+      "learning_rate": 6.6564045952549994e-06,
+      "loss": 4.7171,
+      "step": 8065
+    },
+    {
+      "epoch": 1.5829737151824244,
+      "grad_norm": 16.17782974243164,
+      "learning_rate": 6.652752175270933e-06,
+      "loss": 4.4794,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5839544919576305,
+      "grad_norm": 17.59096908569336,
+      "learning_rate": 6.649098764916211e-06,
+      "loss": 4.4093,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5849352687328364,
+      "grad_norm": 22.827747344970703,
+      "learning_rate": 6.64544436638005e-06,
+      "loss": 4.6104,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5859160455080423,
+      "grad_norm": 34.059200286865234,
+      "learning_rate": 6.641788981852262e-06,
+      "loss": 5.1327,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5868968222832485,
+      "grad_norm": 18.246440887451172,
+      "learning_rate": 6.6381326135232415e-06,
+      "loss": 4.7764,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5878775990584542,
+      "grad_norm": 26.03481101989746,
+      "learning_rate": 6.634475263583978e-06,
+      "loss": 4.8857,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5888583758336603,
+      "grad_norm": 11.254276275634766,
+      "learning_rate": 6.630816934226047e-06,
+      "loss": 5.087,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5898391526088662,
+      "grad_norm": 37.02158737182617,
+      "learning_rate": 6.627157627641611e-06,
+      "loss": 4.7483,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5908199293840721,
+      "grad_norm": 11.20740795135498,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 4.747,
+      "step": 8110
+    },
+    {
+      "epoch": 1.5918007061592783,
+      "grad_norm": 17.87250328063965,
+      "learning_rate": 6.619836091564803e-06,
+      "loss": 4.5122,
+      "step": 8115
+    },
+    {
+      "epoch": 1.592781482934484,
+      "grad_norm": 21.036434173583984,
+      "learning_rate": 6.61617386645968e-06,
+      "loss": 4.6486,
+      "step": 8120
+    },
+    {
+      "epoch": 1.59376225970969,
+      "grad_norm": 12.921380996704102,
+      "learning_rate": 6.612510672902545e-06,
+      "loss": 4.5646,
+      "step": 8125
+    },
+    {
+      "epoch": 1.594743036484896,
+      "grad_norm": 23.73027992248535,
+      "learning_rate": 6.608846513088478e-06,
+      "loss": 4.7245,
+      "step": 8130
+    },
+    {
+      "epoch": 1.595723813260102,
+      "grad_norm": 22.765920639038086,
+      "learning_rate": 6.6051813892131355e-06,
+      "loss": 4.8803,
+      "step": 8135
+    },
+    {
+      "epoch": 1.596704590035308,
+      "grad_norm": 13.69679069519043,
+      "learning_rate": 6.601515303472752e-06,
+      "loss": 4.9128,
+      "step": 8140
+    },
+    {
+      "epoch": 1.5976853668105138,
+      "grad_norm": 31.66204071044922,
+      "learning_rate": 6.597848258064138e-06,
+      "loss": 4.7675,
+      "step": 8145
+    },
+    {
+      "epoch": 1.59866614358572,
+      "grad_norm": 37.54737091064453,
+      "learning_rate": 6.594180255184678e-06,
+      "loss": 4.4534,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5996469203609258,
+      "grad_norm": 13.1535005569458,
+      "learning_rate": 6.59051129703233e-06,
+      "loss": 4.8452,
+      "step": 8155
+    },
+    {
+      "epoch": 1.6006276971361317,
+      "grad_norm": 16.961715698242188,
+      "learning_rate": 6.5868413858056315e-06,
+      "loss": 4.4688,
+      "step": 8160
+    },
+    {
+      "epoch": 1.6016084739113379,
+      "grad_norm": 20.165821075439453,
+      "learning_rate": 6.583170523703682e-06,
+      "loss": 4.5493,
+      "step": 8165
+    },
+    {
+      "epoch": 1.6025892506865438,
+      "grad_norm": 16.436073303222656,
+      "learning_rate": 6.579498712926153e-06,
+      "loss": 4.526,
+      "step": 8170
+    },
+    {
+      "epoch": 1.6035700274617497,
+      "grad_norm": 22.747486114501953,
+      "learning_rate": 6.5758259556732896e-06,
+      "loss": 4.8861,
+      "step": 8175
+    },
+    {
+      "epoch": 1.6045508042369556,
+      "grad_norm": 12.746312141418457,
+      "learning_rate": 6.572152254145898e-06,
+      "loss": 4.8588,
+      "step": 8180
+    },
+    {
+      "epoch": 1.6055315810121615,
+      "grad_norm": 23.855030059814453,
+      "learning_rate": 6.568477610545352e-06,
+      "loss": 4.9615,
+      "step": 8185
+    },
+    {
+      "epoch": 1.6065123577873677,
+      "grad_norm": 17.665613174438477,
+      "learning_rate": 6.564802027073592e-06,
+      "loss": 4.7032,
+      "step": 8190
+    },
+    {
+      "epoch": 1.6074931345625736,
+      "grad_norm": 31.55120849609375,
+      "learning_rate": 6.561125505933119e-06,
+      "loss": 4.8295,
+      "step": 8195
+    },
+    {
+      "epoch": 1.6084739113377795,
+      "grad_norm": 14.968411445617676,
+      "learning_rate": 6.557448049326997e-06,
+      "loss": 4.8213,
+      "step": 8200
+    },
+    {
+      "epoch": 1.6094546881129856,
+      "grad_norm": 20.886701583862305,
+      "learning_rate": 6.55376965945885e-06,
+      "loss": 4.6678,
+      "step": 8205
+    },
+    {
+      "epoch": 1.6104354648881913,
+      "grad_norm": 14.710062026977539,
+      "learning_rate": 6.550090338532863e-06,
+      "loss": 4.7658,
+      "step": 8210
+    },
+    {
+      "epoch": 1.6114162416633975,
+      "grad_norm": 13.487741470336914,
+      "learning_rate": 6.546410088753777e-06,
+      "loss": 4.5046,
+      "step": 8215
+    },
+    {
+      "epoch": 1.6123970184386034,
+      "grad_norm": 14.448518753051758,
+      "learning_rate": 6.54272891232689e-06,
+      "loss": 5.073,
+      "step": 8220
+    },
+    {
+      "epoch": 1.6133777952138093,
+      "grad_norm": 22.348711013793945,
+      "learning_rate": 6.539046811458056e-06,
+      "loss": 4.794,
+      "step": 8225
+    },
+    {
+      "epoch": 1.6143585719890154,
+      "grad_norm": 11.976225852966309,
+      "learning_rate": 6.53536378835368e-06,
+      "loss": 4.6712,
+      "step": 8230
+    },
+    {
+      "epoch": 1.6153393487642211,
+      "grad_norm": 19.27861213684082,
+      "learning_rate": 6.531679845220725e-06,
+      "loss": 4.4471,
+      "step": 8235
+    },
+    {
+      "epoch": 1.6163201255394273,
+      "grad_norm": 20.683855056762695,
+      "learning_rate": 6.527994984266702e-06,
+      "loss": 4.8398,
+      "step": 8240
+    },
+    {
+      "epoch": 1.6173009023146332,
+      "grad_norm": 23.678176879882812,
+      "learning_rate": 6.524309207699671e-06,
+      "loss": 5.0151,
+      "step": 8245
+    },
+    {
+      "epoch": 1.618281679089839,
+      "grad_norm": 25.82317543029785,
+      "learning_rate": 6.5206225177282435e-06,
+      "loss": 4.7891,
+      "step": 8250
+    },
+    {
+      "epoch": 1.6192624558650452,
+      "grad_norm": 28.66914939880371,
+      "learning_rate": 6.516934916561575e-06,
+      "loss": 4.8626,
+      "step": 8255
+    },
+    {
+      "epoch": 1.620243232640251,
+      "grad_norm": 11.398292541503906,
+      "learning_rate": 6.513246406409369e-06,
+      "loss": 4.4237,
+      "step": 8260
+    },
+    {
+      "epoch": 1.621224009415457,
+      "grad_norm": 25.888748168945312,
+      "learning_rate": 6.509556989481875e-06,
+      "loss": 4.416,
+      "step": 8265
+    },
+    {
+      "epoch": 1.622204786190663,
+      "grad_norm": 11.734145164489746,
+      "learning_rate": 6.505866667989884e-06,
+      "loss": 4.6947,
+      "step": 8270
+    },
+    {
+      "epoch": 1.623185562965869,
+      "grad_norm": 12.392928123474121,
+      "learning_rate": 6.50217544414473e-06,
+      "loss": 4.4491,
+      "step": 8275
+    },
+    {
+      "epoch": 1.624166339741075,
+      "grad_norm": 11.715088844299316,
+      "learning_rate": 6.498483320158282e-06,
+      "loss": 4.564,
+      "step": 8280
+    },
+    {
+      "epoch": 1.6251471165162807,
+      "grad_norm": 34.25434112548828,
+      "learning_rate": 6.494790298242962e-06,
+      "loss": 4.8514,
+      "step": 8285
+    },
+    {
+      "epoch": 1.6261278932914869,
+      "grad_norm": 25.465301513671875,
+      "learning_rate": 6.491096380611716e-06,
+      "loss": 4.814,
+      "step": 8290
+    },
+    {
+      "epoch": 1.6271086700666928,
+      "grad_norm": 15.51944351196289,
+      "learning_rate": 6.487401569478033e-06,
+      "loss": 4.5486,
+      "step": 8295
+    },
+    {
+      "epoch": 1.6280894468418987,
+      "grad_norm": 26.168691635131836,
+      "learning_rate": 6.483705867055937e-06,
+      "loss": 4.9638,
+      "step": 8300
+    },
+    {
+      "epoch": 1.6290702236171049,
+      "grad_norm": 40.255760192871094,
+      "learning_rate": 6.480009275559985e-06,
+      "loss": 5.1347,
+      "step": 8305
+    },
+    {
+      "epoch": 1.6300510003923105,
+      "grad_norm": 26.99652671813965,
+      "learning_rate": 6.4763117972052704e-06,
+      "loss": 4.3371,
+      "step": 8310
+    },
+    {
+      "epoch": 1.6310317771675167,
+      "grad_norm": 31.933998107910156,
+      "learning_rate": 6.472613434207413e-06,
+      "loss": 4.4121,
+      "step": 8315
+    },
+    {
+      "epoch": 1.6320125539427226,
+      "grad_norm": 22.316579818725586,
+      "learning_rate": 6.4689141887825655e-06,
+      "loss": 4.8148,
+      "step": 8320
+    },
+    {
+      "epoch": 1.6329933307179285,
+      "grad_norm": 27.438989639282227,
+      "learning_rate": 6.465214063147409e-06,
+      "loss": 4.6428,
+      "step": 8325
+    },
+    {
+      "epoch": 1.6339741074931347,
+      "grad_norm": 15.515409469604492,
+      "learning_rate": 6.46151305951915e-06,
+      "loss": 4.9047,
+      "step": 8330
+    },
+    {
+      "epoch": 1.6349548842683406,
+      "grad_norm": 14.11785888671875,
+      "learning_rate": 6.457811180115525e-06,
+      "loss": 4.4316,
+      "step": 8335
+    },
+    {
+      "epoch": 1.6359356610435465,
+      "grad_norm": 13.665918350219727,
+      "learning_rate": 6.454108427154792e-06,
+      "loss": 4.8829,
+      "step": 8340
+    },
+    {
+      "epoch": 1.6369164378187524,
+      "grad_norm": 12.321572303771973,
+      "learning_rate": 6.450404802855734e-06,
+      "loss": 4.4735,
+      "step": 8345
+    },
+    {
+      "epoch": 1.6378972145939583,
+      "grad_norm": 14.074231147766113,
+      "learning_rate": 6.446700309437657e-06,
+      "loss": 4.614,
+      "step": 8350
+    },
+    {
+      "epoch": 1.6388779913691645,
+      "grad_norm": 23.989662170410156,
+      "learning_rate": 6.442994949120385e-06,
+      "loss": 4.4395,
+      "step": 8355
+    },
+    {
+      "epoch": 1.6398587681443704,
+      "grad_norm": 23.025890350341797,
+      "learning_rate": 6.439288724124262e-06,
+      "loss": 4.642,
+      "step": 8360
+    },
+    {
+      "epoch": 1.6408395449195763,
+      "grad_norm": 22.322298049926758,
+      "learning_rate": 6.435581636670154e-06,
+      "loss": 5.0991,
+      "step": 8365
+    },
+    {
+      "epoch": 1.6418203216947824,
+      "grad_norm": 15.972933769226074,
+      "learning_rate": 6.43187368897944e-06,
+      "loss": 4.6842,
+      "step": 8370
+    },
+    {
+      "epoch": 1.6428010984699881,
+      "grad_norm": 18.866586685180664,
+      "learning_rate": 6.4281648832740155e-06,
+      "loss": 4.7774,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6437818752451943,
+      "grad_norm": 20.908037185668945,
+      "learning_rate": 6.424455221776286e-06,
+      "loss": 4.9527,
+      "step": 8380
+    },
+    {
+      "epoch": 1.6447626520204002,
+      "grad_norm": 22.348464965820312,
+      "learning_rate": 6.420744706709181e-06,
+      "loss": 4.6132,
+      "step": 8385
+    },
+    {
+      "epoch": 1.645743428795606,
+      "grad_norm": 27.352182388305664,
+      "learning_rate": 6.417033340296131e-06,
+      "loss": 4.8313,
+      "step": 8390
+    },
+    {
+      "epoch": 1.6467242055708122,
+      "grad_norm": 16.535625457763672,
+      "learning_rate": 6.413321124761082e-06,
+      "loss": 4.9389,
+      "step": 8395
+    },
+    {
+      "epoch": 1.647704982346018,
+      "grad_norm": 17.407255172729492,
+      "learning_rate": 6.409608062328483e-06,
+      "loss": 4.7184,
+      "step": 8400
+    },
+    {
+      "epoch": 1.648685759121224,
+      "grad_norm": 20.512407302856445,
+      "learning_rate": 6.405894155223296e-06,
+      "loss": 4.524,
+      "step": 8405
+    },
+    {
+      "epoch": 1.64966653589643,
+      "grad_norm": 16.161840438842773,
+      "learning_rate": 6.402179405670987e-06,
+      "loss": 4.9194,
+      "step": 8410
+    },
+    {
+      "epoch": 1.650647312671636,
+      "grad_norm": 16.505544662475586,
+      "learning_rate": 6.39846381589753e-06,
+      "loss": 4.9257,
+      "step": 8415
+    },
+    {
+      "epoch": 1.651628089446842,
+      "grad_norm": 21.150869369506836,
+      "learning_rate": 6.394747388129397e-06,
+      "loss": 5.1818,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6526088662220477,
+      "grad_norm": 21.39322853088379,
+      "learning_rate": 6.391030124593567e-06,
+      "loss": 4.4558,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6535896429972539,
+      "grad_norm": 22.720727920532227,
+      "learning_rate": 6.387312027517516e-06,
+      "loss": 5.0587,
+      "step": 8430
+    },
+    {
+      "epoch": 1.6545704197724598,
+      "grad_norm": 24.188188552856445,
+      "learning_rate": 6.383593099129223e-06,
+      "loss": 4.6071,
+      "step": 8435
+    },
+    {
+      "epoch": 1.6555511965476657,
+      "grad_norm": 42.62229537963867,
+      "learning_rate": 6.3798733416571615e-06,
+      "loss": 4.7521,
+      "step": 8440
+    },
+    {
+      "epoch": 1.6565319733228718,
+      "grad_norm": 18.189743041992188,
+      "learning_rate": 6.376152757330305e-06,
+      "loss": 4.9016,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6575127500980775,
+      "grad_norm": 23.545324325561523,
+      "learning_rate": 6.37243134837812e-06,
+      "loss": 4.9774,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6584935268732837,
+      "grad_norm": 12.068930625915527,
+      "learning_rate": 6.368709117030568e-06,
+      "loss": 4.3265,
+      "step": 8455
+    },
+    {
+      "epoch": 1.6594743036484896,
+      "grad_norm": 19.71160316467285,
+      "learning_rate": 6.364986065518106e-06,
+      "loss": 4.7148,
+      "step": 8460
+    },
+    {
+      "epoch": 1.6604550804236955,
+      "grad_norm": 12.18775463104248,
+      "learning_rate": 6.361262196071679e-06,
+      "loss": 4.5165,
+      "step": 8465
+    },
+    {
+      "epoch": 1.6614358571989016,
+      "grad_norm": 20.218761444091797,
+      "learning_rate": 6.357537510922723e-06,
+      "loss": 4.7621,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6624166339741073,
+      "grad_norm": 24.970178604125977,
+      "learning_rate": 6.353812012303162e-06,
+      "loss": 4.9106,
+      "step": 8475
+    },
+    {
+      "epoch": 1.6633974107493135,
+      "grad_norm": 15.730690956115723,
+      "learning_rate": 6.3500857024454085e-06,
+      "loss": 4.616,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6643781875245194,
+      "grad_norm": 17.307218551635742,
+      "learning_rate": 6.346358583582364e-06,
+      "loss": 4.4639,
+      "step": 8485
+    },
+    {
+      "epoch": 1.6653589642997253,
+      "grad_norm": 27.932443618774414,
+      "learning_rate": 6.342630657947409e-06,
+      "loss": 4.9017,
+      "step": 8490
+    },
+    {
+      "epoch": 1.6663397410749314,
+      "grad_norm": 32.948543548583984,
+      "learning_rate": 6.338901927774409e-06,
+      "loss": 4.433,
+      "step": 8495
+    },
+    {
+      "epoch": 1.6673205178501374,
+      "grad_norm": 16.31046485900879,
+      "learning_rate": 6.335172395297716e-06,
+      "loss": 4.6377,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6683012946253433,
+      "grad_norm": 34.126861572265625,
+      "learning_rate": 6.331442062752159e-06,
+      "loss": 5.0827,
+      "step": 8505
+    },
+    {
+      "epoch": 1.6692820714005492,
+      "grad_norm": 25.11467170715332,
+      "learning_rate": 6.327710932373046e-06,
+      "loss": 4.8379,
+      "step": 8510
+    },
+    {
+      "epoch": 1.670262848175755,
+      "grad_norm": 23.186079025268555,
+      "learning_rate": 6.3239790063961635e-06,
+      "loss": 4.4733,
+      "step": 8515
+    },
+    {
+      "epoch": 1.6712436249509612,
+      "grad_norm": 29.186731338500977,
+      "learning_rate": 6.320246287057778e-06,
+      "loss": 4.5714,
+      "step": 8520
+    },
+    {
+      "epoch": 1.6722244017261672,
+      "grad_norm": 17.12517738342285,
+      "learning_rate": 6.316512776594626e-06,
+      "loss": 4.3804,
+      "step": 8525
+    },
+    {
+      "epoch": 1.673205178501373,
+      "grad_norm": 10.104792594909668,
+      "learning_rate": 6.3127784772439215e-06,
+      "loss": 5.0567,
+      "step": 8530
+    },
+    {
+      "epoch": 1.6741859552765792,
+      "grad_norm": 15.159939765930176,
+      "learning_rate": 6.309043391243351e-06,
+      "loss": 4.5438,
+      "step": 8535
+    },
+    {
+      "epoch": 1.675166732051785,
+      "grad_norm": 20.949722290039062,
+      "learning_rate": 6.305307520831075e-06,
+      "loss": 4.4797,
+      "step": 8540
+    },
+    {
+      "epoch": 1.676147508826991,
+      "grad_norm": 14.368339538574219,
+      "learning_rate": 6.3015708682457155e-06,
+      "loss": 4.7613,
+      "step": 8545
+    },
+    {
+      "epoch": 1.677128285602197,
+      "grad_norm": 13.72527027130127,
+      "learning_rate": 6.29783343572637e-06,
+      "loss": 4.9441,
+      "step": 8550
+    },
+    {
+      "epoch": 1.6781090623774029,
+      "grad_norm": 37.21997833251953,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 4.5065,
+      "step": 8555
+    },
+    {
+      "epoch": 1.679089839152609,
+      "grad_norm": 19.89255714416504,
+      "learning_rate": 6.290356239844446e-06,
+      "loss": 4.605,
+      "step": 8560
+    },
+    {
+      "epoch": 1.6800706159278147,
+      "grad_norm": 30.77366828918457,
+      "learning_rate": 6.286616480962392e-06,
+      "loss": 4.7847,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6810513927030208,
+      "grad_norm": 15.917234420776367,
+      "learning_rate": 6.282875951107396e-06,
+      "loss": 4.408,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6820321694782268,
+      "grad_norm": 18.189117431640625,
+      "learning_rate": 6.279134652520881e-06,
+      "loss": 4.7609,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6830129462534327,
+      "grad_norm": 30.72614097595215,
+      "learning_rate": 6.275392587444724e-06,
+      "loss": 4.7158,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6839937230286388,
+      "grad_norm": 11.812881469726562,
+      "learning_rate": 6.271649758121268e-06,
+      "loss": 4.4278,
+      "step": 8585
+    },
+    {
+      "epoch": 1.6849744998038445,
+      "grad_norm": 17.11888313293457,
+      "learning_rate": 6.267906166793306e-06,
+      "loss": 4.3961,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6859552765790506,
+      "grad_norm": 22.773406982421875,
+      "learning_rate": 6.264161815704096e-06,
+      "loss": 4.9114,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6869360533542566,
+      "grad_norm": 26.260541915893555,
+      "learning_rate": 6.260416707097345e-06,
+      "loss": 4.6699,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6879168301294625,
+      "grad_norm": 24.89080238342285,
+      "learning_rate": 6.256670843217217e-06,
+      "loss": 4.6868,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6888976069046686,
+      "grad_norm": 17.895580291748047,
+      "learning_rate": 6.2529242263083305e-06,
+      "loss": 4.6321,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6898783836798743,
+      "grad_norm": 15.826708793640137,
+      "learning_rate": 6.249176858615746e-06,
+      "loss": 4.4789,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6908591604550804,
+      "grad_norm": 20.98354148864746,
+      "learning_rate": 6.2454287423849865e-06,
+      "loss": 4.5294,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6918399372302864,
+      "grad_norm": 12.527328491210938,
+      "learning_rate": 6.241679879862015e-06,
+      "loss": 4.5682,
+      "step": 8625
+    },
+    {
+      "epoch": 1.6928207140054923,
+      "grad_norm": 12.26150131225586,
+      "learning_rate": 6.237930273293244e-06,
+      "loss": 4.6067,
+      "step": 8630
+    },
+    {
+      "epoch": 1.6938014907806984,
+      "grad_norm": 14.805673599243164,
+      "learning_rate": 6.234179924925532e-06,
+      "loss": 4.5776,
+      "step": 8635
+    },
+    {
+      "epoch": 1.6947822675559043,
+      "grad_norm": 27.439050674438477,
+      "learning_rate": 6.230428837006184e-06,
+      "loss": 4.6483,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6957630443311102,
+      "grad_norm": 17.67620277404785,
+      "learning_rate": 6.226677011782944e-06,
+      "loss": 4.6386,
+      "step": 8645
+    },
+    {
+      "epoch": 1.6967438211063162,
+      "grad_norm": 27.96957015991211,
+      "learning_rate": 6.222924451504001e-06,
+      "loss": 4.2794,
+      "step": 8650
+    },
+    {
+      "epoch": 1.697724597881522,
+      "grad_norm": 31.630573272705078,
+      "learning_rate": 6.219171158417981e-06,
+      "loss": 5.1995,
+      "step": 8655
+    },
+    {
+      "epoch": 1.6987053746567282,
+      "grad_norm": 18.49655532836914,
+      "learning_rate": 6.215417134773956e-06,
+      "loss": 4.5782,
+      "step": 8660
+    },
+    {
+      "epoch": 1.6996861514319341,
+      "grad_norm": 14.828500747680664,
+      "learning_rate": 6.211662382821428e-06,
+      "loss": 4.5066,
+      "step": 8665
+    },
+    {
+      "epoch": 1.70066692820714,
+      "grad_norm": 18.147098541259766,
+      "learning_rate": 6.207906904810341e-06,
+      "loss": 4.6078,
+      "step": 8670
+    },
+    {
+      "epoch": 1.701647704982346,
+      "grad_norm": 10.830955505371094,
+      "learning_rate": 6.20415070299107e-06,
+      "loss": 4.5747,
+      "step": 8675
+    },
+    {
+      "epoch": 1.7026284817575519,
+      "grad_norm": 17.871294021606445,
+      "learning_rate": 6.200393779614425e-06,
+      "loss": 4.7012,
+      "step": 8680
+    },
+    {
+      "epoch": 1.703609258532758,
+      "grad_norm": 16.49895477294922,
+      "learning_rate": 6.196636136931652e-06,
+      "loss": 4.5185,
+      "step": 8685
+    },
+    {
+      "epoch": 1.704590035307964,
+      "grad_norm": 15.448025703430176,
+      "learning_rate": 6.192877777194422e-06,
+      "loss": 4.8708,
+      "step": 8690
+    },
+    {
+      "epoch": 1.7055708120831699,
+      "grad_norm": 21.347553253173828,
+      "learning_rate": 6.18911870265484e-06,
+      "loss": 4.6123,
+      "step": 8695
+    },
+    {
+      "epoch": 1.706551588858376,
+      "grad_norm": 13.813310623168945,
+      "learning_rate": 6.185358915565438e-06,
+      "loss": 4.7046,
+      "step": 8700
+    },
+    {
+      "epoch": 1.7075323656335817,
+      "grad_norm": 24.650951385498047,
+      "learning_rate": 6.181598418179173e-06,
+      "loss": 5.0376,
+      "step": 8705
+    },
+    {
+      "epoch": 1.7085131424087878,
+      "grad_norm": 26.78155517578125,
+      "learning_rate": 6.177837212749432e-06,
+      "loss": 4.452,
+      "step": 8710
+    },
+    {
+      "epoch": 1.7094939191839937,
+      "grad_norm": 14.59371566772461,
+      "learning_rate": 6.174075301530024e-06,
+      "loss": 4.9905,
+      "step": 8715
+    },
+    {
+      "epoch": 1.7104746959591997,
+      "grad_norm": 27.0151424407959,
+      "learning_rate": 6.17031268677518e-06,
+      "loss": 4.612,
+      "step": 8720
+    },
+    {
+      "epoch": 1.7114554727344058,
+      "grad_norm": 27.519006729125977,
+      "learning_rate": 6.166549370739553e-06,
+      "loss": 4.7261,
+      "step": 8725
+    },
+    {
+      "epoch": 1.7124362495096115,
+      "grad_norm": 21.014606475830078,
+      "learning_rate": 6.162785355678215e-06,
+      "loss": 4.664,
+      "step": 8730
+    },
+    {
+      "epoch": 1.7134170262848176,
+      "grad_norm": 21.153648376464844,
+      "learning_rate": 6.1590206438466605e-06,
+      "loss": 4.6907,
+      "step": 8735
+    },
+    {
+      "epoch": 1.7143978030600235,
+      "grad_norm": 20.082855224609375,
+      "learning_rate": 6.1552552375008e-06,
+      "loss": 4.4854,
+      "step": 8740
+    },
+    {
+      "epoch": 1.7153785798352295,
+      "grad_norm": 26.873451232910156,
+      "learning_rate": 6.15148913889696e-06,
+      "loss": 5.1138,
+      "step": 8745
+    },
+    {
+      "epoch": 1.7163593566104356,
+      "grad_norm": 12.644761085510254,
+      "learning_rate": 6.147722350291878e-06,
+      "loss": 4.847,
+      "step": 8750
+    },
+    {
+      "epoch": 1.7173401333856413,
+      "grad_norm": 30.004619598388672,
+      "learning_rate": 6.143954873942712e-06,
+      "loss": 4.6217,
+      "step": 8755
+    },
+    {
+      "epoch": 1.7183209101608474,
+      "grad_norm": 15.3263578414917,
+      "learning_rate": 6.140186712107027e-06,
+      "loss": 4.5504,
+      "step": 8760
+    },
+    {
+      "epoch": 1.7193016869360533,
+      "grad_norm": 17.260662078857422,
+      "learning_rate": 6.136417867042801e-06,
+      "loss": 4.3313,
+      "step": 8765
+    },
+    {
+      "epoch": 1.7202824637112593,
+      "grad_norm": 22.777530670166016,
+      "learning_rate": 6.132648341008421e-06,
+      "loss": 4.7004,
+      "step": 8770
+    },
+    {
+      "epoch": 1.7212632404864654,
+      "grad_norm": 9.795730590820312,
+      "learning_rate": 6.128878136262678e-06,
+      "loss": 4.9465,
+      "step": 8775
+    },
+    {
+      "epoch": 1.722244017261671,
+      "grad_norm": 13.725994110107422,
+      "learning_rate": 6.1251072550647775e-06,
+      "loss": 4.4856,
+      "step": 8780
+    },
+    {
+      "epoch": 1.7232247940368772,
+      "grad_norm": 12.93813705444336,
+      "learning_rate": 6.1213356996743265e-06,
+      "loss": 4.4976,
+      "step": 8785
+    },
+    {
+      "epoch": 1.7242055708120831,
+      "grad_norm": 28.854127883911133,
+      "learning_rate": 6.117563472351334e-06,
+      "loss": 4.643,
+      "step": 8790
+    },
+    {
+      "epoch": 1.725186347587289,
+      "grad_norm": 17.60565757751465,
+      "learning_rate": 6.1137905753562155e-06,
+      "loss": 4.8717,
+      "step": 8795
+    },
+    {
+      "epoch": 1.7261671243624952,
+      "grad_norm": 18.80259132385254,
+      "learning_rate": 6.110017010949783e-06,
+      "loss": 4.4856,
+      "step": 8800
+    },
+    {
+      "epoch": 1.7271479011377011,
+      "grad_norm": 8.956795692443848,
+      "learning_rate": 6.106242781393251e-06,
+      "loss": 4.6121,
+      "step": 8805
+    },
+    {
+      "epoch": 1.728128677912907,
+      "grad_norm": 17.677858352661133,
+      "learning_rate": 6.102467888948236e-06,
+      "loss": 4.448,
+      "step": 8810
+    },
+    {
+      "epoch": 1.729109454688113,
+      "grad_norm": 16.494901657104492,
+      "learning_rate": 6.098692335876746e-06,
+      "loss": 4.4203,
+      "step": 8815
+    },
+    {
+      "epoch": 1.7300902314633189,
+      "grad_norm": 14.752246856689453,
+      "learning_rate": 6.0949161244411885e-06,
+      "loss": 4.6896,
+      "step": 8820
+    },
+    {
+      "epoch": 1.731071008238525,
+      "grad_norm": 10.59720516204834,
+      "learning_rate": 6.091139256904363e-06,
+      "loss": 4.4049,
+      "step": 8825
+    },
+    {
+      "epoch": 1.732051785013731,
+      "grad_norm": 22.1069393157959,
+      "learning_rate": 6.087361735529464e-06,
+      "loss": 4.7651,
+      "step": 8830
+    },
+    {
+      "epoch": 1.7330325617889368,
+      "grad_norm": 16.202037811279297,
+      "learning_rate": 6.083583562580078e-06,
+      "loss": 4.6475,
+      "step": 8835
+    },
+    {
+      "epoch": 1.7340133385641427,
+      "grad_norm": 32.283241271972656,
+      "learning_rate": 6.079804740320181e-06,
+      "loss": 4.8508,
+      "step": 8840
+    },
+    {
+      "epoch": 1.7349941153393487,
+      "grad_norm": 15.124344825744629,
+      "learning_rate": 6.076025271014138e-06,
+      "loss": 4.5117,
+      "step": 8845
+    },
+    {
+      "epoch": 1.7359748921145548,
+      "grad_norm": 22.1917724609375,
+      "learning_rate": 6.0722451569267015e-06,
+      "loss": 4.5844,
+      "step": 8850
+    },
+    {
+      "epoch": 1.7369556688897607,
+      "grad_norm": 21.06514549255371,
+      "learning_rate": 6.06846440032301e-06,
+      "loss": 4.647,
+      "step": 8855
+    },
+    {
+      "epoch": 1.7379364456649666,
+      "grad_norm": 29.728837966918945,
+      "learning_rate": 6.064683003468591e-06,
+      "loss": 4.5468,
+      "step": 8860
+    },
+    {
+      "epoch": 1.7389172224401728,
+      "grad_norm": 15.97719669342041,
+      "learning_rate": 6.060900968629352e-06,
+      "loss": 4.7311,
+      "step": 8865
+    },
+    {
+      "epoch": 1.7398979992153785,
+      "grad_norm": 15.47018814086914,
+      "learning_rate": 6.05711829807158e-06,
+      "loss": 4.6125,
+      "step": 8870
+    },
+    {
+      "epoch": 1.7408787759905846,
+      "grad_norm": 22.31346893310547,
+      "learning_rate": 6.05333499406195e-06,
+      "loss": 4.6249,
+      "step": 8875
+    },
+    {
+      "epoch": 1.7418595527657905,
+      "grad_norm": 38.01810836791992,
+      "learning_rate": 6.04955105886751e-06,
+      "loss": 5.4116,
+      "step": 8880
+    },
+    {
+      "epoch": 1.7428403295409964,
+      "grad_norm": 16.22886085510254,
+      "learning_rate": 6.045766494755692e-06,
+      "loss": 4.9962,
+      "step": 8885
+    },
+    {
+      "epoch": 1.7438211063162026,
+      "grad_norm": 25.318523406982422,
+      "learning_rate": 6.0419813039943e-06,
+      "loss": 4.7032,
+      "step": 8890
+    },
+    {
+      "epoch": 1.7448018830914083,
+      "grad_norm": 28.022945404052734,
+      "learning_rate": 6.038195488851515e-06,
+      "loss": 4.5836,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7457826598666144,
+      "grad_norm": 9.770703315734863,
+      "learning_rate": 6.0344090515958946e-06,
+      "loss": 4.6167,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7467634366418203,
+      "grad_norm": 19.886987686157227,
+      "learning_rate": 6.030621994496365e-06,
+      "loss": 4.3776,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7477442134170262,
+      "grad_norm": 14.081059455871582,
+      "learning_rate": 6.026834319822228e-06,
+      "loss": 4.5549,
+      "step": 8910
+    },
+    {
+      "epoch": 1.7487249901922324,
+      "grad_norm": 21.361549377441406,
+      "learning_rate": 6.0230460298431525e-06,
+      "loss": 4.8304,
+      "step": 8915
+    },
+    {
+      "epoch": 1.749705766967438,
+      "grad_norm": 12.771543502807617,
+      "learning_rate": 6.019257126829178e-06,
+      "loss": 4.5428,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7506865437426442,
+      "grad_norm": 36.33089065551758,
+      "learning_rate": 6.015467613050708e-06,
+      "loss": 5.1167,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7506865437426442,
+      "eval_loss": 4.88515567779541,
+      "eval_runtime": 7.8212,
+      "eval_samples_per_second": 26.722,
+      "eval_steps_per_second": 13.425,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7516673205178501,
+      "grad_norm": 15.354480743408203,
+      "learning_rate": 6.0116774907785154e-06,
+      "loss": 5.1371,
+      "step": 8930
+    },
+    {
+      "epoch": 1.752648097293056,
+      "grad_norm": 9.22092056274414,
+      "learning_rate": 6.00788676228374e-06,
+      "loss": 4.2449,
+      "step": 8935
+    },
+    {
+      "epoch": 1.7536288740682622,
+      "grad_norm": 14.879875183105469,
+      "learning_rate": 6.004095429837878e-06,
+      "loss": 4.9914,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7546096508434679,
+      "grad_norm": 20.033912658691406,
+      "learning_rate": 6.000303495712791e-06,
+      "loss": 4.5591,
+      "step": 8945
+    },
+    {
+      "epoch": 1.755590427618674,
+      "grad_norm": 19.408241271972656,
+      "learning_rate": 5.996510962180704e-06,
+      "loss": 4.9889,
+      "step": 8950
+    },
+    {
+      "epoch": 1.75657120439388,
+      "grad_norm": 18.337635040283203,
+      "learning_rate": 5.992717831514196e-06,
+      "loss": 4.8683,
+      "step": 8955
+    },
+    {
+      "epoch": 1.7575519811690858,
+      "grad_norm": 10.698108673095703,
+      "learning_rate": 5.988924105986207e-06,
+      "loss": 4.7408,
+      "step": 8960
+    },
+    {
+      "epoch": 1.758532757944292,
+      "grad_norm": 20.650182723999023,
+      "learning_rate": 5.985129787870032e-06,
+      "loss": 4.748,
+      "step": 8965
+    },
+    {
+      "epoch": 1.759513534719498,
+      "grad_norm": 16.508623123168945,
+      "learning_rate": 5.981334879439324e-06,
+      "loss": 4.7143,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7604943114947038,
+      "grad_norm": 38.822044372558594,
+      "learning_rate": 5.9775393829680865e-06,
+      "loss": 4.8905,
+      "step": 8975
+    },
+    {
+      "epoch": 1.7614750882699097,
+      "grad_norm": 16.134923934936523,
+      "learning_rate": 5.973743300730674e-06,
+      "loss": 4.4573,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7624558650451156,
+      "grad_norm": 29.342491149902344,
+      "learning_rate": 5.9699466350017975e-06,
+      "loss": 4.5743,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7634366418203218,
+      "grad_norm": 30.499624252319336,
+      "learning_rate": 5.9661493880565136e-06,
+      "loss": 4.512,
+      "step": 8990
+    },
+    {
+      "epoch": 1.7644174185955277,
+      "grad_norm": 25.990304946899414,
+      "learning_rate": 5.9623515621702275e-06,
+      "loss": 4.7905,
+      "step": 8995
+    },
+    {
+      "epoch": 1.7653981953707336,
+      "grad_norm": 13.431137084960938,
+      "learning_rate": 5.958553159618693e-06,
+      "loss": 5.4404,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7663789721459395,
+      "grad_norm": 24.7120361328125,
+      "learning_rate": 5.954754182678008e-06,
+      "loss": 4.8807,
+      "step": 9005
+    },
+    {
+      "epoch": 1.7673597489211454,
+      "grad_norm": 22.7612361907959,
+      "learning_rate": 5.9509546336246135e-06,
+      "loss": 4.3851,
+      "step": 9010
+    },
+    {
+      "epoch": 1.7683405256963516,
+      "grad_norm": 20.71037483215332,
+      "learning_rate": 5.947154514735299e-06,
+      "loss": 4.7264,
+      "step": 9015
+    },
+    {
+      "epoch": 1.7693213024715575,
+      "grad_norm": 20.132646560668945,
+      "learning_rate": 5.943353828287185e-06,
+      "loss": 4.525,
+      "step": 9020
+    },
+    {
+      "epoch": 1.7703020792467634,
+      "grad_norm": 21.293230056762695,
+      "learning_rate": 5.939552576557743e-06,
+      "loss": 4.5384,
+      "step": 9025
+    },
+    {
+      "epoch": 1.7712828560219696,
+      "grad_norm": 33.4547233581543,
+      "learning_rate": 5.935750761824777e-06,
+      "loss": 4.6896,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7722636327971752,
+      "grad_norm": 28.96019744873047,
+      "learning_rate": 5.9319483863664306e-06,
+      "loss": 4.4839,
+      "step": 9035
+    },
+    {
+      "epoch": 1.7732444095723814,
+      "grad_norm": 21.133338928222656,
+      "learning_rate": 5.928145452461183e-06,
+      "loss": 4.696,
+      "step": 9040
+    },
+    {
+      "epoch": 1.7742251863475873,
+      "grad_norm": 32.1081657409668,
+      "learning_rate": 5.9243419623878485e-06,
+      "loss": 4.739,
+      "step": 9045
+    },
+    {
+      "epoch": 1.7752059631227932,
+      "grad_norm": 22.64603614807129,
+      "learning_rate": 5.920537918425571e-06,
+      "loss": 4.445,
+      "step": 9050
+    },
+    {
+      "epoch": 1.7761867398979994,
+      "grad_norm": 24.658477783203125,
+      "learning_rate": 5.916733322853831e-06,
+      "loss": 4.4289,
+      "step": 9055
+    },
+    {
+      "epoch": 1.777167516673205,
+      "grad_norm": 20.534879684448242,
+      "learning_rate": 5.912928177952438e-06,
+      "loss": 4.7319,
+      "step": 9060
+    },
+    {
+      "epoch": 1.7781482934484112,
+      "grad_norm": 23.79903793334961,
+      "learning_rate": 5.909122486001531e-06,
+      "loss": 4.5808,
+      "step": 9065
+    },
+    {
+      "epoch": 1.779129070223617,
+      "grad_norm": 13.706748962402344,
+      "learning_rate": 5.905316249281575e-06,
+      "loss": 4.624,
+      "step": 9070
+    },
+    {
+      "epoch": 1.780109846998823,
+      "grad_norm": 37.71672439575195,
+      "learning_rate": 5.901509470073364e-06,
+      "loss": 4.6139,
+      "step": 9075
+    },
+    {
+      "epoch": 1.7810906237740292,
+      "grad_norm": 18.964744567871094,
+      "learning_rate": 5.897702150658015e-06,
+      "loss": 4.6728,
+      "step": 9080
+    },
+    {
+      "epoch": 1.7820714005492349,
+      "grad_norm": 11.068283081054688,
+      "learning_rate": 5.89389429331697e-06,
+      "loss": 4.5614,
+      "step": 9085
+    },
+    {
+      "epoch": 1.783052177324441,
+      "grad_norm": 14.902594566345215,
+      "learning_rate": 5.890085900331991e-06,
+      "loss": 4.6259,
+      "step": 9090
+    },
+    {
+      "epoch": 1.784032954099647,
+      "grad_norm": 26.177350997924805,
+      "learning_rate": 5.8862769739851655e-06,
+      "loss": 4.7789,
+      "step": 9095
+    },
+    {
+      "epoch": 1.7850137308748528,
+      "grad_norm": 11.954874038696289,
+      "learning_rate": 5.882467516558896e-06,
+      "loss": 4.6356,
+      "step": 9100
+    },
+    {
+      "epoch": 1.785994507650059,
+      "grad_norm": 26.418127059936523,
+      "learning_rate": 5.878657530335906e-06,
+      "loss": 4.6969,
+      "step": 9105
+    },
+    {
+      "epoch": 1.7869752844252647,
+      "grad_norm": 24.923782348632812,
+      "learning_rate": 5.874847017599236e-06,
+      "loss": 4.3997,
+      "step": 9110
+    },
+    {
+      "epoch": 1.7879560612004708,
+      "grad_norm": 24.46477699279785,
+      "learning_rate": 5.87103598063224e-06,
+      "loss": 4.5095,
+      "step": 9115
+    },
+    {
+      "epoch": 1.7889368379756767,
+      "grad_norm": 21.268714904785156,
+      "learning_rate": 5.867224421718587e-06,
+      "loss": 4.9617,
+      "step": 9120
+    },
+    {
+      "epoch": 1.7899176147508826,
+      "grad_norm": 16.486682891845703,
+      "learning_rate": 5.863412343142258e-06,
+      "loss": 4.1348,
+      "step": 9125
+    },
+    {
+      "epoch": 1.7908983915260888,
+      "grad_norm": 15.189291954040527,
+      "learning_rate": 5.8595997471875465e-06,
+      "loss": 4.442,
+      "step": 9130
+    },
+    {
+      "epoch": 1.7918791683012947,
+      "grad_norm": 12.895219802856445,
+      "learning_rate": 5.855786636139058e-06,
+      "loss": 4.9543,
+      "step": 9135
+    },
+    {
+      "epoch": 1.7928599450765006,
+      "grad_norm": 13.546518325805664,
+      "learning_rate": 5.8519730122817045e-06,
+      "loss": 4.6226,
+      "step": 9140
+    },
+    {
+      "epoch": 1.7938407218517065,
+      "grad_norm": 16.05805206298828,
+      "learning_rate": 5.848158877900702e-06,
+      "loss": 4.5933,
+      "step": 9145
+    },
+    {
+      "epoch": 1.7948214986269124,
+      "grad_norm": 19.39604377746582,
+      "learning_rate": 5.844344235281578e-06,
+      "loss": 4.5569,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7958022754021186,
+      "grad_norm": 20.664596557617188,
+      "learning_rate": 5.840529086710163e-06,
+      "loss": 4.5215,
+      "step": 9155
+    },
+    {
+      "epoch": 1.7967830521773245,
+      "grad_norm": 24.75821304321289,
+      "learning_rate": 5.836713434472587e-06,
+      "loss": 4.7306,
+      "step": 9160
+    },
+    {
+      "epoch": 1.7977638289525304,
+      "grad_norm": 38.528533935546875,
+      "learning_rate": 5.832897280855289e-06,
+      "loss": 4.8929,
+      "step": 9165
+    },
+    {
+      "epoch": 1.7987446057277363,
+      "grad_norm": 17.18060874938965,
+      "learning_rate": 5.8290806281450004e-06,
+      "loss": 4.6035,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7997253825029422,
+      "grad_norm": 20.632356643676758,
+      "learning_rate": 5.8252634786287574e-06,
+      "loss": 4.5474,
+      "step": 9175
+    },
+    {
+      "epoch": 1.8007061592781484,
+      "grad_norm": 21.149751663208008,
+      "learning_rate": 5.821445834593889e-06,
+      "loss": 4.7935,
+      "step": 9180
+    },
+    {
+      "epoch": 1.8016869360533543,
+      "grad_norm": 21.72342300415039,
+      "learning_rate": 5.817627698328029e-06,
+      "loss": 4.3023,
+      "step": 9185
+    },
+    {
+      "epoch": 1.8026677128285602,
+      "grad_norm": 21.94120216369629,
+      "learning_rate": 5.813809072119098e-06,
+      "loss": 4.719,
+      "step": 9190
+    },
+    {
+      "epoch": 1.8036484896037663,
+      "grad_norm": 21.152067184448242,
+      "learning_rate": 5.80998995825531e-06,
+      "loss": 5.0133,
+      "step": 9195
+    },
+    {
+      "epoch": 1.804629266378972,
+      "grad_norm": 17.194652557373047,
+      "learning_rate": 5.806170359025177e-06,
+      "loss": 4.6929,
+      "step": 9200
+    },
+    {
+      "epoch": 1.8056100431541782,
+      "grad_norm": 22.314577102661133,
+      "learning_rate": 5.802350276717498e-06,
+      "loss": 4.5256,
+      "step": 9205
+    },
+    {
+      "epoch": 1.806590819929384,
+      "grad_norm": 18.893587112426758,
+      "learning_rate": 5.798529713621364e-06,
+      "loss": 4.5493,
+      "step": 9210
+    },
+    {
+      "epoch": 1.80757159670459,
+      "grad_norm": 25.69357681274414,
+      "learning_rate": 5.7947086720261495e-06,
+      "loss": 4.5069,
+      "step": 9215
+    },
+    {
+      "epoch": 1.8085523734797961,
+      "grad_norm": 13.256308555603027,
+      "learning_rate": 5.790887154221521e-06,
+      "loss": 4.6355,
+      "step": 9220
+    },
+    {
+      "epoch": 1.8095331502550018,
+      "grad_norm": 26.34750747680664,
+      "learning_rate": 5.787065162497427e-06,
+      "loss": 4.2925,
+      "step": 9225
+    },
+    {
+      "epoch": 1.810513927030208,
+      "grad_norm": 21.691946029663086,
+      "learning_rate": 5.7832426991441014e-06,
+      "loss": 4.7569,
+      "step": 9230
+    },
+    {
+      "epoch": 1.8114947038054139,
+      "grad_norm": 22.89154052734375,
+      "learning_rate": 5.77941976645206e-06,
+      "loss": 4.5489,
+      "step": 9235
+    },
+    {
+      "epoch": 1.8124754805806198,
+      "grad_norm": 13.551591873168945,
+      "learning_rate": 5.775596366712101e-06,
+      "loss": 4.6459,
+      "step": 9240
+    },
+    {
+      "epoch": 1.813456257355826,
+      "grad_norm": 35.56877899169922,
+      "learning_rate": 5.771772502215301e-06,
+      "loss": 5.1302,
+      "step": 9245
+    },
+    {
+      "epoch": 1.8144370341310316,
+      "grad_norm": 22.027677536010742,
+      "learning_rate": 5.767948175253015e-06,
+      "loss": 4.9661,
+      "step": 9250
+    },
+    {
+      "epoch": 1.8154178109062378,
+      "grad_norm": 10.572216987609863,
+      "learning_rate": 5.764123388116877e-06,
+      "loss": 4.5735,
+      "step": 9255
+    },
+    {
+      "epoch": 1.8163985876814437,
+      "grad_norm": 17.56256866455078,
+      "learning_rate": 5.760298143098797e-06,
+      "loss": 4.725,
+      "step": 9260
+    },
+    {
+      "epoch": 1.8173793644566496,
+      "grad_norm": 36.0407829284668,
+      "learning_rate": 5.756472442490954e-06,
+      "loss": 4.8934,
+      "step": 9265
+    },
+    {
+      "epoch": 1.8183601412318557,
+      "grad_norm": 21.272165298461914,
+      "learning_rate": 5.752646288585808e-06,
+      "loss": 4.488,
+      "step": 9270
+    },
+    {
+      "epoch": 1.8193409180070614,
+      "grad_norm": 19.818387985229492,
+      "learning_rate": 5.748819683676083e-06,
+      "loss": 4.6846,
+      "step": 9275
+    },
+    {
+      "epoch": 1.8203216947822676,
+      "grad_norm": 28.447729110717773,
+      "learning_rate": 5.744992630054779e-06,
+      "loss": 4.6269,
+      "step": 9280
+    },
+    {
+      "epoch": 1.8213024715574735,
+      "grad_norm": 27.741268157958984,
+      "learning_rate": 5.7411651300151624e-06,
+      "loss": 4.8284,
+      "step": 9285
+    },
+    {
+      "epoch": 1.8222832483326794,
+      "grad_norm": 17.49547576904297,
+      "learning_rate": 5.737337185850769e-06,
+      "loss": 4.5584,
+      "step": 9290
+    },
+    {
+      "epoch": 1.8232640251078855,
+      "grad_norm": 24.814552307128906,
+      "learning_rate": 5.733508799855396e-06,
+      "loss": 4.6543,
+      "step": 9295
+    },
+    {
+      "epoch": 1.8242448018830915,
+      "grad_norm": 16.014293670654297,
+      "learning_rate": 5.7296799743231125e-06,
+      "loss": 4.4332,
+      "step": 9300
+    },
+    {
+      "epoch": 1.8252255786582974,
+      "grad_norm": 27.853015899658203,
+      "learning_rate": 5.725850711548242e-06,
+      "loss": 4.3259,
+      "step": 9305
+    },
+    {
+      "epoch": 1.8262063554335033,
+      "grad_norm": 16.20301628112793,
+      "learning_rate": 5.722021013825378e-06,
+      "loss": 4.5385,
+      "step": 9310
+    },
+    {
+      "epoch": 1.8271871322087092,
+      "grad_norm": 26.664472579956055,
+      "learning_rate": 5.718190883449373e-06,
+      "loss": 4.4814,
+      "step": 9315
+    },
+    {
+      "epoch": 1.8281679089839153,
+      "grad_norm": 39.86537170410156,
+      "learning_rate": 5.714360322715335e-06,
+      "loss": 4.6435,
+      "step": 9320
+    },
+    {
+      "epoch": 1.8291486857591213,
+      "grad_norm": 32.14000701904297,
+      "learning_rate": 5.710529333918633e-06,
+      "loss": 4.529,
+      "step": 9325
+    },
+    {
+      "epoch": 1.8301294625343272,
+      "grad_norm": 21.911773681640625,
+      "learning_rate": 5.706697919354892e-06,
+      "loss": 4.801,
+      "step": 9330
+    },
+    {
+      "epoch": 1.8311102393095333,
+      "grad_norm": 17.9941463470459,
+      "learning_rate": 5.702866081319992e-06,
+      "loss": 4.6752,
+      "step": 9335
+    },
+    {
+      "epoch": 1.832091016084739,
+      "grad_norm": 24.887266159057617,
+      "learning_rate": 5.699033822110066e-06,
+      "loss": 4.8302,
+      "step": 9340
+    },
+    {
+      "epoch": 1.8330717928599451,
+      "grad_norm": 18.100421905517578,
+      "learning_rate": 5.6952011440215e-06,
+      "loss": 4.4905,
+      "step": 9345
+    },
+    {
+      "epoch": 1.834052569635151,
+      "grad_norm": 12.726563453674316,
+      "learning_rate": 5.691368049350932e-06,
+      "loss": 5.1001,
+      "step": 9350
+    },
+    {
+      "epoch": 1.835033346410357,
+      "grad_norm": 13.885564804077148,
+      "learning_rate": 5.687534540395247e-06,
+      "loss": 4.3967,
+      "step": 9355
+    },
+    {
+      "epoch": 1.8360141231855631,
+      "grad_norm": 13.186820030212402,
+      "learning_rate": 5.683700619451584e-06,
+      "loss": 4.536,
+      "step": 9360
+    },
+    {
+      "epoch": 1.8369948999607688,
+      "grad_norm": 21.86033058166504,
+      "learning_rate": 5.679866288817321e-06,
+      "loss": 4.5876,
+      "step": 9365
+    },
+    {
+      "epoch": 1.837975676735975,
+      "grad_norm": 23.89190101623535,
+      "learning_rate": 5.676031550790087e-06,
+      "loss": 4.9731,
+      "step": 9370
+    },
+    {
+      "epoch": 1.8389564535111809,
+      "grad_norm": 23.900188446044922,
+      "learning_rate": 5.6721964076677515e-06,
+      "loss": 4.6855,
+      "step": 9375
+    },
+    {
+      "epoch": 1.8399372302863868,
+      "grad_norm": 24.231233596801758,
+      "learning_rate": 5.66836086174843e-06,
+      "loss": 4.4809,
+      "step": 9380
+    },
+    {
+      "epoch": 1.840918007061593,
+      "grad_norm": 20.207731246948242,
+      "learning_rate": 5.664524915330478e-06,
+      "loss": 4.6487,
+      "step": 9385
+    },
+    {
+      "epoch": 1.8418987838367986,
+      "grad_norm": 15.975571632385254,
+      "learning_rate": 5.660688570712492e-06,
+      "loss": 4.8748,
+      "step": 9390
+    },
+    {
+      "epoch": 1.8428795606120048,
+      "grad_norm": 29.445188522338867,
+      "learning_rate": 5.656851830193304e-06,
+      "loss": 4.9362,
+      "step": 9395
+    },
+    {
+      "epoch": 1.8438603373872107,
+      "grad_norm": 16.028915405273438,
+      "learning_rate": 5.653014696071987e-06,
+      "loss": 4.7763,
+      "step": 9400
+    },
+    {
+      "epoch": 1.8448411141624166,
+      "grad_norm": 24.72963523864746,
+      "learning_rate": 5.649177170647847e-06,
+      "loss": 4.6154,
+      "step": 9405
+    },
+    {
+      "epoch": 1.8458218909376227,
+      "grad_norm": 44.67896270751953,
+      "learning_rate": 5.645339256220427e-06,
+      "loss": 4.8692,
+      "step": 9410
+    },
+    {
+      "epoch": 1.8468026677128284,
+      "grad_norm": 31.613073348999023,
+      "learning_rate": 5.641500955089502e-06,
+      "loss": 4.7649,
+      "step": 9415
+    },
+    {
+      "epoch": 1.8477834444880346,
+      "grad_norm": 23.47369956970215,
+      "learning_rate": 5.6376622695550764e-06,
+      "loss": 4.9661,
+      "step": 9420
+    },
+    {
+      "epoch": 1.8487642212632405,
+      "grad_norm": 11.438894271850586,
+      "learning_rate": 5.63382320191739e-06,
+      "loss": 4.7752,
+      "step": 9425
+    },
+    {
+      "epoch": 1.8497449980384464,
+      "grad_norm": 17.62010383605957,
+      "learning_rate": 5.6299837544769046e-06,
+      "loss": 4.7805,
+      "step": 9430
+    },
+    {
+      "epoch": 1.8507257748136525,
+      "grad_norm": 18.699684143066406,
+      "learning_rate": 5.6261439295343175e-06,
+      "loss": 4.5472,
+      "step": 9435
+    },
+    {
+      "epoch": 1.8517065515888582,
+      "grad_norm": 13.261539459228516,
+      "learning_rate": 5.622303729390548e-06,
+      "loss": 4.8652,
+      "step": 9440
+    },
+    {
+      "epoch": 1.8526873283640644,
+      "grad_norm": 17.09125518798828,
+      "learning_rate": 5.61846315634674e-06,
+      "loss": 4.6028,
+      "step": 9445
+    },
+    {
+      "epoch": 1.8536681051392703,
+      "grad_norm": 34.29975891113281,
+      "learning_rate": 5.61462221270426e-06,
+      "loss": 4.392,
+      "step": 9450
+    },
+    {
+      "epoch": 1.8546488819144762,
+      "grad_norm": 16.109909057617188,
+      "learning_rate": 5.6107809007646966e-06,
+      "loss": 4.6932,
+      "step": 9455
+    },
+    {
+      "epoch": 1.8556296586896823,
+      "grad_norm": 18.13594627380371,
+      "learning_rate": 5.606939222829865e-06,
+      "loss": 4.7523,
+      "step": 9460
+    },
+    {
+      "epoch": 1.8566104354648882,
+      "grad_norm": 19.660789489746094,
+      "learning_rate": 5.603097181201793e-06,
+      "loss": 4.4013,
+      "step": 9465
+    },
+    {
+      "epoch": 1.8575912122400942,
+      "grad_norm": 24.52638053894043,
+      "learning_rate": 5.599254778182729e-06,
+      "loss": 5.0819,
+      "step": 9470
+    },
+    {
+      "epoch": 1.8585719890153,
+      "grad_norm": 11.734413146972656,
+      "learning_rate": 5.5954120160751354e-06,
+      "loss": 4.4883,
+      "step": 9475
+    },
+    {
+      "epoch": 1.859552765790506,
+      "grad_norm": 22.79249382019043,
+      "learning_rate": 5.5915688971816955e-06,
+      "loss": 4.6221,
+      "step": 9480
+    },
+    {
+      "epoch": 1.8605335425657121,
+      "grad_norm": 18.433395385742188,
+      "learning_rate": 5.587725423805299e-06,
+      "loss": 4.6917,
+      "step": 9485
+    },
+    {
+      "epoch": 1.861514319340918,
+      "grad_norm": 17.560726165771484,
+      "learning_rate": 5.583881598249054e-06,
+      "loss": 4.6395,
+      "step": 9490
+    },
+    {
+      "epoch": 1.862495096116124,
+      "grad_norm": 18.91305160522461,
+      "learning_rate": 5.5800374228162776e-06,
+      "loss": 4.4603,
+      "step": 9495
+    },
+    {
+      "epoch": 1.86347587289133,
+      "grad_norm": 22.717443466186523,
+      "learning_rate": 5.576192899810495e-06,
+      "loss": 4.8376,
+      "step": 9500
+    },
+    {
+      "epoch": 1.8644566496665358,
+      "grad_norm": 14.52634334564209,
+      "learning_rate": 5.572348031535442e-06,
+      "loss": 4.8505,
+      "step": 9505
+    },
+    {
+      "epoch": 1.865437426441742,
+      "grad_norm": 20.6605167388916,
+      "learning_rate": 5.5685028202950595e-06,
+      "loss": 4.8469,
+      "step": 9510
+    },
+    {
+      "epoch": 1.8664182032169478,
+      "grad_norm": 18.79912757873535,
+      "learning_rate": 5.5646572683934975e-06,
+      "loss": 4.9637,
+      "step": 9515
+    },
+    {
+      "epoch": 1.8673989799921538,
+      "grad_norm": 13.271584510803223,
+      "learning_rate": 5.560811378135104e-06,
+      "loss": 4.8746,
+      "step": 9520
+    },
+    {
+      "epoch": 1.86837975676736,
+      "grad_norm": 15.022200584411621,
+      "learning_rate": 5.556965151824433e-06,
+      "loss": 4.5181,
+      "step": 9525
+    },
+    {
+      "epoch": 1.8693605335425656,
+      "grad_norm": 20.869287490844727,
+      "learning_rate": 5.553118591766241e-06,
+      "loss": 4.4984,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8703413103177717,
+      "grad_norm": 26.764667510986328,
+      "learning_rate": 5.549271700265485e-06,
+      "loss": 4.7425,
+      "step": 9535
+    },
+    {
+      "epoch": 1.8713220870929776,
+      "grad_norm": 20.59969139099121,
+      "learning_rate": 5.5454244796273175e-06,
+      "loss": 4.6921,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8723028638681836,
+      "grad_norm": 19.59276008605957,
+      "learning_rate": 5.54157693215709e-06,
+      "loss": 4.8437,
+      "step": 9545
+    },
+    {
+      "epoch": 1.8732836406433897,
+      "grad_norm": 13.677157402038574,
+      "learning_rate": 5.5377290601603504e-06,
+      "loss": 4.721,
+      "step": 9550
+    },
+    {
+      "epoch": 1.8742644174185954,
+      "grad_norm": 18.6031551361084,
+      "learning_rate": 5.53388086594284e-06,
+      "loss": 4.4678,
+      "step": 9555
+    },
+    {
+      "epoch": 1.8752451941938015,
+      "grad_norm": 28.015052795410156,
+      "learning_rate": 5.5300323518104925e-06,
+      "loss": 4.8587,
+      "step": 9560
+    },
+    {
+      "epoch": 1.8762259709690075,
+      "grad_norm": 23.171489715576172,
+      "learning_rate": 5.526183520069436e-06,
+      "loss": 4.6468,
+      "step": 9565
+    },
+    {
+      "epoch": 1.8772067477442134,
+      "grad_norm": 23.11787223815918,
+      "learning_rate": 5.522334373025986e-06,
+      "loss": 4.5523,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8781875245194195,
+      "grad_norm": 11.215222358703613,
+      "learning_rate": 5.518484912986648e-06,
+      "loss": 4.7702,
+      "step": 9575
+    },
+    {
+      "epoch": 1.8791683012946252,
+      "grad_norm": 22.043649673461914,
+      "learning_rate": 5.514635142258116e-06,
+      "loss": 4.426,
+      "step": 9580
+    },
+    {
+      "epoch": 1.8801490780698313,
+      "grad_norm": 15.956389427185059,
+      "learning_rate": 5.510785063147269e-06,
+      "loss": 4.7553,
+      "step": 9585
+    },
+    {
+      "epoch": 1.8811298548450373,
+      "grad_norm": 31.348739624023438,
+      "learning_rate": 5.506934677961172e-06,
+      "loss": 4.2031,
+      "step": 9590
+    },
+    {
+      "epoch": 1.8821106316202432,
+      "grad_norm": 14.20702838897705,
+      "learning_rate": 5.503083989007072e-06,
+      "loss": 4.6814,
+      "step": 9595
+    },
+    {
+      "epoch": 1.8830914083954493,
+      "grad_norm": 20.319719314575195,
+      "learning_rate": 5.499232998592399e-06,
+      "loss": 4.5755,
+      "step": 9600
+    },
+    {
+      "epoch": 1.884072185170655,
+      "grad_norm": 25.81941795349121,
+      "learning_rate": 5.49538170902476e-06,
+      "loss": 4.801,
+      "step": 9605
+    },
+    {
+      "epoch": 1.8850529619458611,
+      "grad_norm": 24.612640380859375,
+      "learning_rate": 5.49153012261195e-06,
+      "loss": 4.9738,
+      "step": 9610
+    },
+    {
+      "epoch": 1.886033738721067,
+      "grad_norm": 25.011335372924805,
+      "learning_rate": 5.487678241661933e-06,
+      "loss": 4.7212,
+      "step": 9615
+    },
+    {
+      "epoch": 1.887014515496273,
+      "grad_norm": 27.755661010742188,
+      "learning_rate": 5.483826068482854e-06,
+      "loss": 4.6,
+      "step": 9620
+    },
+    {
+      "epoch": 1.887995292271479,
+      "grad_norm": 18.196792602539062,
+      "learning_rate": 5.4799736053830324e-06,
+      "loss": 4.53,
+      "step": 9625
+    },
+    {
+      "epoch": 1.888976069046685,
+      "grad_norm": 13.700377464294434,
+      "learning_rate": 5.476120854670957e-06,
+      "loss": 4.8347,
+      "step": 9630
+    },
+    {
+      "epoch": 1.889956845821891,
+      "grad_norm": 30.03533935546875,
+      "learning_rate": 5.4722678186552995e-06,
+      "loss": 4.7082,
+      "step": 9635
+    },
+    {
+      "epoch": 1.8909376225970969,
+      "grad_norm": 16.419034957885742,
+      "learning_rate": 5.468414499644892e-06,
+      "loss": 4.3029,
+      "step": 9640
+    },
+    {
+      "epoch": 1.8919183993723028,
+      "grad_norm": 14.156940460205078,
+      "learning_rate": 5.4645608999487395e-06,
+      "loss": 4.3517,
+      "step": 9645
+    },
+    {
+      "epoch": 1.892899176147509,
+      "grad_norm": 28.83321762084961,
+      "learning_rate": 5.4607070218760184e-06,
+      "loss": 4.5463,
+      "step": 9650
+    },
+    {
+      "epoch": 1.8938799529227148,
+      "grad_norm": 45.10062026977539,
+      "learning_rate": 5.456852867736067e-06,
+      "loss": 4.8991,
+      "step": 9655
+    },
+    {
+      "epoch": 1.8948607296979207,
+      "grad_norm": 24.96133041381836,
+      "learning_rate": 5.452998439838392e-06,
+      "loss": 4.5104,
+      "step": 9660
+    },
+    {
+      "epoch": 1.8958415064731269,
+      "grad_norm": 11.74023151397705,
+      "learning_rate": 5.449143740492664e-06,
+      "loss": 4.665,
+      "step": 9665
+    },
+    {
+      "epoch": 1.8968222832483326,
+      "grad_norm": 19.273738861083984,
+      "learning_rate": 5.4452887720087165e-06,
+      "loss": 4.7088,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8978030600235387,
+      "grad_norm": 15.15214729309082,
+      "learning_rate": 5.441433536696541e-06,
+      "loss": 4.8068,
+      "step": 9675
+    },
+    {
+      "epoch": 1.8987838367987446,
+      "grad_norm": 18.632030487060547,
+      "learning_rate": 5.437578036866293e-06,
+      "loss": 4.9114,
+      "step": 9680
+    },
+    {
+      "epoch": 1.8997646135739505,
+      "grad_norm": 19.880435943603516,
+      "learning_rate": 5.433722274828286e-06,
+      "loss": 4.7176,
+      "step": 9685
+    },
+    {
+      "epoch": 1.9007453903491567,
+      "grad_norm": 29.079919815063477,
+      "learning_rate": 5.429866252892988e-06,
+      "loss": 3.8863,
+      "step": 9690
+    },
+    {
+      "epoch": 1.9017261671243624,
+      "grad_norm": 20.837482452392578,
+      "learning_rate": 5.426009973371026e-06,
+      "loss": 4.7429,
+      "step": 9695
+    },
+    {
+      "epoch": 1.9027069438995685,
+      "grad_norm": 18.1405029296875,
+      "learning_rate": 5.422153438573176e-06,
+      "loss": 5.1525,
+      "step": 9700
+    },
+    {
+      "epoch": 1.9036877206747744,
+      "grad_norm": 41.7537841796875,
+      "learning_rate": 5.418296650810373e-06,
+      "loss": 4.7584,
+      "step": 9705
+    },
+    {
+      "epoch": 1.9046684974499803,
+      "grad_norm": 13.388349533081055,
+      "learning_rate": 5.414439612393703e-06,
+      "loss": 4.6856,
+      "step": 9710
+    },
+    {
+      "epoch": 1.9056492742251865,
+      "grad_norm": 15.971819877624512,
+      "learning_rate": 5.410582325634397e-06,
+      "loss": 4.8956,
+      "step": 9715
+    },
+    {
+      "epoch": 1.9066300510003922,
+      "grad_norm": 20.830162048339844,
+      "learning_rate": 5.40672479284384e-06,
+      "loss": 4.5245,
+      "step": 9720
+    },
+    {
+      "epoch": 1.9076108277755983,
+      "grad_norm": 12.127934455871582,
+      "learning_rate": 5.402867016333563e-06,
+      "loss": 4.4878,
+      "step": 9725
+    },
+    {
+      "epoch": 1.9085916045508042,
+      "grad_norm": 21.272417068481445,
+      "learning_rate": 5.399008998415242e-06,
+      "loss": 4.5526,
+      "step": 9730
+    },
+    {
+      "epoch": 1.9095723813260101,
+      "grad_norm": 14.217857360839844,
+      "learning_rate": 5.395150741400698e-06,
+      "loss": 4.5744,
+      "step": 9735
+    },
+    {
+      "epoch": 1.9105531581012163,
+      "grad_norm": 24.945274353027344,
+      "learning_rate": 5.3912922476018956e-06,
+      "loss": 4.5555,
+      "step": 9740
+    },
+    {
+      "epoch": 1.911533934876422,
+      "grad_norm": 17.92483139038086,
+      "learning_rate": 5.387433519330941e-06,
+      "loss": 5.0362,
+      "step": 9745
+    },
+    {
+      "epoch": 1.9125147116516281,
+      "grad_norm": 19.815181732177734,
+      "learning_rate": 5.383574558900083e-06,
+      "loss": 4.5256,
+      "step": 9750
+    },
+    {
+      "epoch": 1.913495488426834,
+      "grad_norm": 18.88251304626465,
+      "learning_rate": 5.3797153686217054e-06,
+      "loss": 4.6896,
+      "step": 9755
+    },
+    {
+      "epoch": 1.91447626520204,
+      "grad_norm": 28.06948471069336,
+      "learning_rate": 5.375855950808334e-06,
+      "loss": 4.5653,
+      "step": 9760
+    },
+    {
+      "epoch": 1.915457041977246,
+      "grad_norm": 19.297407150268555,
+      "learning_rate": 5.371996307772628e-06,
+      "loss": 4.6416,
+      "step": 9765
+    },
+    {
+      "epoch": 1.9164378187524518,
+      "grad_norm": 14.056635856628418,
+      "learning_rate": 5.368136441827383e-06,
+      "loss": 4.7386,
+      "step": 9770
+    },
+    {
+      "epoch": 1.917418595527658,
+      "grad_norm": 17.106592178344727,
+      "learning_rate": 5.364276355285527e-06,
+      "loss": 4.5563,
+      "step": 9775
+    },
+    {
+      "epoch": 1.9183993723028638,
+      "grad_norm": 21.838733673095703,
+      "learning_rate": 5.36041605046012e-06,
+      "loss": 4.6798,
+      "step": 9780
+    },
+    {
+      "epoch": 1.9193801490780698,
+      "grad_norm": 17.363008499145508,
+      "learning_rate": 5.3565555296643555e-06,
+      "loss": 4.868,
+      "step": 9785
+    },
+    {
+      "epoch": 1.920360925853276,
+      "grad_norm": 20.810707092285156,
+      "learning_rate": 5.352694795211555e-06,
+      "loss": 4.8239,
+      "step": 9790
+    },
+    {
+      "epoch": 1.9213417026284818,
+      "grad_norm": 29.47817039489746,
+      "learning_rate": 5.348833849415167e-06,
+      "loss": 4.8366,
+      "step": 9795
+    },
+    {
+      "epoch": 1.9223224794036877,
+      "grad_norm": 21.65797233581543,
+      "learning_rate": 5.344972694588766e-06,
+      "loss": 4.3141,
+      "step": 9800
+    },
+    {
+      "epoch": 1.9233032561788936,
+      "grad_norm": 15.772238731384277,
+      "learning_rate": 5.341111333046054e-06,
+      "loss": 4.7682,
+      "step": 9805
+    },
+    {
+      "epoch": 1.9242840329540996,
+      "grad_norm": 17.035268783569336,
+      "learning_rate": 5.337249767100856e-06,
+      "loss": 4.5995,
+      "step": 9810
+    },
+    {
+      "epoch": 1.9252648097293057,
+      "grad_norm": 20.667861938476562,
+      "learning_rate": 5.33338799906712e-06,
+      "loss": 4.7046,
+      "step": 9815
+    },
+    {
+      "epoch": 1.9262455865045116,
+      "grad_norm": 18.487794876098633,
+      "learning_rate": 5.329526031258914e-06,
+      "loss": 4.567,
+      "step": 9820
+    },
+    {
+      "epoch": 1.9272263632797175,
+      "grad_norm": 18.39491844177246,
+      "learning_rate": 5.325663865990425e-06,
+      "loss": 5.004,
+      "step": 9825
+    },
+    {
+      "epoch": 1.9282071400549237,
+      "grad_norm": 17.848005294799805,
+      "learning_rate": 5.321801505575959e-06,
+      "loss": 4.7211,
+      "step": 9830
+    },
+    {
+      "epoch": 1.9291879168301294,
+      "grad_norm": 25.833599090576172,
+      "learning_rate": 5.317938952329943e-06,
+      "loss": 4.4809,
+      "step": 9835
+    },
+    {
+      "epoch": 1.9301686936053355,
+      "grad_norm": 15.303146362304688,
+      "learning_rate": 5.314076208566915e-06,
+      "loss": 4.6026,
+      "step": 9840
+    },
+    {
+      "epoch": 1.9311494703805414,
+      "grad_norm": 16.975481033325195,
+      "learning_rate": 5.310213276601525e-06,
+      "loss": 4.9513,
+      "step": 9845
+    },
+    {
+      "epoch": 1.9321302471557473,
+      "grad_norm": 45.39344024658203,
+      "learning_rate": 5.306350158748544e-06,
+      "loss": 4.7095,
+      "step": 9850
+    },
+    {
+      "epoch": 1.9331110239309535,
+      "grad_norm": 19.3656063079834,
+      "learning_rate": 5.302486857322841e-06,
+      "loss": 4.7902,
+      "step": 9855
+    },
+    {
+      "epoch": 1.9340918007061592,
+      "grad_norm": 18.749727249145508,
+      "learning_rate": 5.298623374639413e-06,
+      "loss": 4.5287,
+      "step": 9860
+    },
+    {
+      "epoch": 1.9350725774813653,
+      "grad_norm": 26.79215431213379,
+      "learning_rate": 5.294759713013351e-06,
+      "loss": 4.5868,
+      "step": 9865
+    },
+    {
+      "epoch": 1.9360533542565712,
+      "grad_norm": 14.35335636138916,
+      "learning_rate": 5.290895874759859e-06,
+      "loss": 5.0044,
+      "step": 9870
+    },
+    {
+      "epoch": 1.9370341310317771,
+      "grad_norm": 12.449397087097168,
+      "learning_rate": 5.287031862194246e-06,
+      "loss": 4.6489,
+      "step": 9875
+    },
+    {
+      "epoch": 1.9380149078069833,
+      "grad_norm": 13.96717357635498,
+      "learning_rate": 5.283167677631926e-06,
+      "loss": 4.772,
+      "step": 9880
+    },
+    {
+      "epoch": 1.938995684582189,
+      "grad_norm": 12.712096214294434,
+      "learning_rate": 5.279303323388413e-06,
+      "loss": 4.7292,
+      "step": 9885
+    },
+    {
+      "epoch": 1.939976461357395,
+      "grad_norm": 18.295469284057617,
+      "learning_rate": 5.275438801779328e-06,
+      "loss": 4.6064,
+      "step": 9890
+    },
+    {
+      "epoch": 1.940957238132601,
+      "grad_norm": 14.68940544128418,
+      "learning_rate": 5.2715741151203895e-06,
+      "loss": 4.7464,
+      "step": 9895
+    },
+    {
+      "epoch": 1.941938014907807,
+      "grad_norm": 27.88214111328125,
+      "learning_rate": 5.267709265727412e-06,
+      "loss": 4.5663,
+      "step": 9900
+    },
+    {
+      "epoch": 1.942918791683013,
+      "grad_norm": 12.562390327453613,
+      "learning_rate": 5.263844255916313e-06,
+      "loss": 4.5185,
+      "step": 9905
+    },
+    {
+      "epoch": 1.9438995684582188,
+      "grad_norm": 17.847429275512695,
+      "learning_rate": 5.259979088003104e-06,
+      "loss": 4.3947,
+      "step": 9910
+    },
+    {
+      "epoch": 1.944880345233425,
+      "grad_norm": 14.35951042175293,
+      "learning_rate": 5.256113764303891e-06,
+      "loss": 4.7351,
+      "step": 9915
+    },
+    {
+      "epoch": 1.9458611220086308,
+      "grad_norm": 18.09912872314453,
+      "learning_rate": 5.252248287134869e-06,
+      "loss": 4.498,
+      "step": 9920
+    },
+    {
+      "epoch": 1.9468418987838367,
+      "grad_norm": 14.626702308654785,
+      "learning_rate": 5.248382658812334e-06,
+      "loss": 4.6051,
+      "step": 9925
+    },
+    {
+      "epoch": 1.9478226755590429,
+      "grad_norm": 15.402030944824219,
+      "learning_rate": 5.2445168816526635e-06,
+      "loss": 4.8815,
+      "step": 9930
+    },
+    {
+      "epoch": 1.9488034523342486,
+      "grad_norm": 30.433496475219727,
+      "learning_rate": 5.2406509579723315e-06,
+      "loss": 4.643,
+      "step": 9935
+    },
+    {
+      "epoch": 1.9497842291094547,
+      "grad_norm": 19.342151641845703,
+      "learning_rate": 5.236784890087897e-06,
+      "loss": 4.3592,
+      "step": 9940
+    },
+    {
+      "epoch": 1.9507650058846606,
+      "grad_norm": 20.78154754638672,
+      "learning_rate": 5.232918680316003e-06,
+      "loss": 4.4561,
+      "step": 9945
+    },
+    {
+      "epoch": 1.9517457826598665,
+      "grad_norm": 14.727789878845215,
+      "learning_rate": 5.229052330973381e-06,
+      "loss": 4.9764,
+      "step": 9950
+    },
+    {
+      "epoch": 1.9527265594350727,
+      "grad_norm": 26.10953712463379,
+      "learning_rate": 5.225185844376842e-06,
+      "loss": 4.7121,
+      "step": 9955
+    },
+    {
+      "epoch": 1.9537073362102786,
+      "grad_norm": 24.18365478515625,
+      "learning_rate": 5.221319222843285e-06,
+      "loss": 4.6004,
+      "step": 9960
+    },
+    {
+      "epoch": 1.9546881129854845,
+      "grad_norm": 23.998422622680664,
+      "learning_rate": 5.217452468689687e-06,
+      "loss": 4.5914,
+      "step": 9965
+    },
+    {
+      "epoch": 1.9556688897606904,
+      "grad_norm": 17.29246711730957,
+      "learning_rate": 5.2135855842331015e-06,
+      "loss": 4.7927,
+      "step": 9970
+    },
+    {
+      "epoch": 1.9566496665358963,
+      "grad_norm": 12.302995681762695,
+      "learning_rate": 5.2097185717906654e-06,
+      "loss": 4.629,
+      "step": 9975
+    },
+    {
+      "epoch": 1.9576304433111025,
+      "grad_norm": 17.5142765045166,
+      "learning_rate": 5.20585143367959e-06,
+      "loss": 4.6272,
+      "step": 9980
+    },
+    {
+      "epoch": 1.9586112200863084,
+      "grad_norm": 59.8892822265625,
+      "learning_rate": 5.201984172217158e-06,
+      "loss": 4.8155,
+      "step": 9985
+    },
+    {
+      "epoch": 1.9595919968615143,
+      "grad_norm": 21.148422241210938,
+      "learning_rate": 5.1981167897207345e-06,
+      "loss": 4.3854,
+      "step": 9990
+    },
+    {
+      "epoch": 1.9605727736367204,
+      "grad_norm": 10.437204360961914,
+      "learning_rate": 5.194249288507749e-06,
+      "loss": 4.3812,
+      "step": 9995
+    },
+    {
+      "epoch": 1.9615535504119261,
+      "grad_norm": 12.685667991638184,
+      "learning_rate": 5.190381670895707e-06,
+      "loss": 4.5556,
+      "step": 10000
+    },
+    {
+      "epoch": 1.9625343271871323,
+      "grad_norm": 120.73411560058594,
+      "learning_rate": 5.18651393920218e-06,
+      "loss": 5.078,
+      "step": 10005
+    },
+    {
+      "epoch": 1.9635151039623382,
+      "grad_norm": 16.686086654663086,
+      "learning_rate": 5.182646095744813e-06,
+      "loss": 4.8024,
+      "step": 10010
+    },
+    {
+      "epoch": 1.964495880737544,
+      "grad_norm": 19.33757209777832,
+      "learning_rate": 5.178778142841315e-06,
+      "loss": 4.6619,
+      "step": 10015
+    },
+    {
+      "epoch": 1.9654766575127502,
+      "grad_norm": 12.6121187210083,
+      "learning_rate": 5.174910082809459e-06,
+      "loss": 5.1033,
+      "step": 10020
+    },
+    {
+      "epoch": 1.966457434287956,
+      "grad_norm": 28.6193790435791,
+      "learning_rate": 5.171041917967083e-06,
+      "loss": 4.3412,
+      "step": 10025
+    },
+    {
+      "epoch": 1.967438211063162,
+      "grad_norm": 14.460850715637207,
+      "learning_rate": 5.16717365063209e-06,
+      "loss": 4.7468,
+      "step": 10030
+    },
+    {
+      "epoch": 1.968418987838368,
+      "grad_norm": 14.494165420532227,
+      "learning_rate": 5.163305283122443e-06,
+      "loss": 4.9953,
+      "step": 10035
+    },
+    {
+      "epoch": 1.969399764613574,
+      "grad_norm": 15.552408218383789,
+      "learning_rate": 5.159436817756166e-06,
+      "loss": 4.7655,
+      "step": 10040
+    },
+    {
+      "epoch": 1.97038054138878,
+      "grad_norm": 14.32104778289795,
+      "learning_rate": 5.155568256851339e-06,
+      "loss": 4.4171,
+      "step": 10045
+    },
+    {
+      "epoch": 1.9713613181639857,
+      "grad_norm": 32.562442779541016,
+      "learning_rate": 5.151699602726101e-06,
+      "loss": 4.7189,
+      "step": 10050
+    },
+    {
+      "epoch": 1.9723420949391919,
+      "grad_norm": 28.78983497619629,
+      "learning_rate": 5.147830857698649e-06,
+      "loss": 4.8201,
+      "step": 10055
+    },
+    {
+      "epoch": 1.9733228717143978,
+      "grad_norm": 39.05999755859375,
+      "learning_rate": 5.143962024087229e-06,
+      "loss": 4.7264,
+      "step": 10060
+    },
+    {
+      "epoch": 1.9743036484896037,
+      "grad_norm": 23.98911476135254,
+      "learning_rate": 5.140093104210147e-06,
+      "loss": 4.383,
+      "step": 10065
+    },
+    {
+      "epoch": 1.9752844252648099,
+      "grad_norm": 21.3807373046875,
+      "learning_rate": 5.136224100385754e-06,
+      "loss": 4.7054,
+      "step": 10070
+    },
+    {
+      "epoch": 1.9762652020400155,
+      "grad_norm": 27.155351638793945,
+      "learning_rate": 5.132355014932455e-06,
+      "loss": 4.6505,
+      "step": 10075
+    },
+    {
+      "epoch": 1.9772459788152217,
+      "grad_norm": 32.85955047607422,
+      "learning_rate": 5.128485850168703e-06,
+      "loss": 4.8904,
+      "step": 10080
+    },
+    {
+      "epoch": 1.9782267555904276,
+      "grad_norm": 16.200801849365234,
+      "learning_rate": 5.124616608413e-06,
+      "loss": 4.6475,
+      "step": 10085
+    },
+    {
+      "epoch": 1.9792075323656335,
+      "grad_norm": 20.49567222595215,
+      "learning_rate": 5.1207472919838945e-06,
+      "loss": 4.8862,
+      "step": 10090
+    },
+    {
+      "epoch": 1.9801883091408397,
+      "grad_norm": 21.595518112182617,
+      "learning_rate": 5.116877903199975e-06,
+      "loss": 4.891,
+      "step": 10095
+    },
+    {
+      "epoch": 1.9811690859160453,
+      "grad_norm": 42.03987121582031,
+      "learning_rate": 5.113008444379878e-06,
+      "loss": 4.8607,
+      "step": 10100
+    },
+    {
+      "epoch": 1.9821498626912515,
+      "grad_norm": 21.260618209838867,
+      "learning_rate": 5.109138917842278e-06,
+      "loss": 4.6659,
+      "step": 10105
+    },
+    {
+      "epoch": 1.9831306394664574,
+      "grad_norm": 23.255855560302734,
+      "learning_rate": 5.105269325905896e-06,
+      "loss": 4.5065,
+      "step": 10110
+    },
+    {
+      "epoch": 1.9841114162416633,
+      "grad_norm": 30.049131393432617,
+      "learning_rate": 5.101399670889489e-06,
+      "loss": 5.2045,
+      "step": 10115
+    },
+    {
+      "epoch": 1.9850921930168695,
+      "grad_norm": 28.775293350219727,
+      "learning_rate": 5.097529955111848e-06,
+      "loss": 4.8581,
+      "step": 10120
+    },
+    {
+      "epoch": 1.9860729697920754,
+      "grad_norm": 25.73450469970703,
+      "learning_rate": 5.093660180891807e-06,
+      "loss": 4.6503,
+      "step": 10125
+    },
+    {
+      "epoch": 1.9870537465672813,
+      "grad_norm": 12.520339965820312,
+      "learning_rate": 5.089790350548232e-06,
+      "loss": 4.793,
+      "step": 10130
+    },
+    {
+      "epoch": 1.9880345233424872,
+      "grad_norm": 10.180679321289062,
+      "learning_rate": 5.085920466400021e-06,
+      "loss": 4.7151,
+      "step": 10135
+    },
+    {
+      "epoch": 1.9890153001176931,
+      "grad_norm": 23.294992446899414,
+      "learning_rate": 5.082050530766105e-06,
+      "loss": 4.5855,
+      "step": 10140
+    },
+    {
+      "epoch": 1.9899960768928993,
+      "grad_norm": 16.03803825378418,
+      "learning_rate": 5.07818054596545e-06,
+      "loss": 4.7725,
+      "step": 10145
+    },
+    {
+      "epoch": 1.9909768536681052,
+      "grad_norm": 16.226547241210938,
+      "learning_rate": 5.074310514317046e-06,
+      "loss": 4.5754,
+      "step": 10150
+    },
+    {
+      "epoch": 1.991957630443311,
+      "grad_norm": 16.746049880981445,
+      "learning_rate": 5.070440438139913e-06,
+      "loss": 5.0154,
+      "step": 10155
+    },
+    {
+      "epoch": 1.9929384072185172,
+      "grad_norm": 27.564208984375,
+      "learning_rate": 5.066570319753099e-06,
+      "loss": 4.4396,
+      "step": 10160
+    },
+    {
+      "epoch": 1.993919183993723,
+      "grad_norm": 24.835065841674805,
+      "learning_rate": 5.0627001614756775e-06,
+      "loss": 4.9128,
+      "step": 10165
+    },
+    {
+      "epoch": 1.994899960768929,
+      "grad_norm": 15.367796897888184,
+      "learning_rate": 5.058829965626742e-06,
+      "loss": 4.5579,
+      "step": 10170
+    },
+    {
+      "epoch": 1.995880737544135,
+      "grad_norm": 15.437244415283203,
+      "learning_rate": 5.054959734525412e-06,
+      "loss": 4.6782,
+      "step": 10175
+    },
+    {
+      "epoch": 1.996861514319341,
+      "grad_norm": 13.151288032531738,
+      "learning_rate": 5.051089470490825e-06,
+      "loss": 4.5289,
+      "step": 10180
+    },
+    {
+      "epoch": 1.997842291094547,
+      "grad_norm": 19.87721824645996,
+      "learning_rate": 5.047219175842146e-06,
+      "loss": 5.0906,
+      "step": 10185
+    },
+    {
+      "epoch": 1.9988230678697527,
+      "grad_norm": 24.418581008911133,
+      "learning_rate": 5.043348852898549e-06,
+      "loss": 4.7632,
+      "step": 10190
+    },
+    {
+      "epoch": 1.9998038446449589,
+      "grad_norm": 34.285369873046875,
+      "learning_rate": 5.03947850397923e-06,
+      "loss": 4.7502,
+      "step": 10195
+    },
+    {
+      "epoch": 2.0007846214201646,
+      "grad_norm": 13.637475967407227,
+      "learning_rate": 5.035608131403397e-06,
+      "loss": 4.6337,
+      "step": 10200
+    },
+    {
+      "epoch": 2.0007846214201646,
+      "eval_loss": 4.867619514465332,
+      "eval_runtime": 7.7779,
+      "eval_samples_per_second": 26.871,
+      "eval_steps_per_second": 13.5,
+      "step": 10200
+    },
+    {
+      "epoch": 2.0017653981953707,
+      "grad_norm": 20.579256057739258,
+      "learning_rate": 5.031737737490278e-06,
+      "loss": 4.2294,
+      "step": 10205
+    },
+    {
+      "epoch": 2.002746174970577,
+      "grad_norm": 22.140090942382812,
+      "learning_rate": 5.027867324559111e-06,
+      "loss": 4.6012,
+      "step": 10210
+    },
+    {
+      "epoch": 2.0037269517457825,
+      "grad_norm": 31.860458374023438,
+      "learning_rate": 5.02399689492914e-06,
+      "loss": 5.1452,
+      "step": 10215
+    },
+    {
+      "epoch": 2.0047077285209887,
+      "grad_norm": 12.973932266235352,
+      "learning_rate": 5.020126450919626e-06,
+      "loss": 4.5311,
+      "step": 10220
+    },
+    {
+      "epoch": 2.005688505296195,
+      "grad_norm": 25.211833953857422,
+      "learning_rate": 5.016255994849837e-06,
+      "loss": 4.8593,
+      "step": 10225
+    },
+    {
+      "epoch": 2.0066692820714005,
+      "grad_norm": 17.093555450439453,
+      "learning_rate": 5.0123855290390465e-06,
+      "loss": 4.8051,
+      "step": 10230
+    },
+    {
+      "epoch": 2.0076500588466066,
+      "grad_norm": 33.78654861450195,
+      "learning_rate": 5.008515055806538e-06,
+      "loss": 4.4996,
+      "step": 10235
+    },
+    {
+      "epoch": 2.0086308356218123,
+      "grad_norm": 18.358837127685547,
+      "learning_rate": 5.004644577471592e-06,
+      "loss": 4.3743,
+      "step": 10240
+    },
+    {
+      "epoch": 2.0096116123970185,
+      "grad_norm": 18.64828109741211,
+      "learning_rate": 5.0007740963535e-06,
+      "loss": 4.4953,
+      "step": 10245
+    },
+    {
+      "epoch": 2.0105923891722246,
+      "grad_norm": 12.689542770385742,
+      "learning_rate": 4.996903614771548e-06,
+      "loss": 4.6645,
+      "step": 10250
+    },
+    {
+      "epoch": 2.0115731659474303,
+      "grad_norm": 35.74906921386719,
+      "learning_rate": 4.99303313504503e-06,
+      "loss": 4.5491,
+      "step": 10255
+    },
+    {
+      "epoch": 2.0125539427226364,
+      "grad_norm": 17.008872985839844,
+      "learning_rate": 4.9891626594932304e-06,
+      "loss": 4.1487,
+      "step": 10260
+    },
+    {
+      "epoch": 2.013534719497842,
+      "grad_norm": 15.55538558959961,
+      "learning_rate": 4.98529219043544e-06,
+      "loss": 4.4378,
+      "step": 10265
+    },
+    {
+      "epoch": 2.0145154962730483,
+      "grad_norm": 15.817734718322754,
+      "learning_rate": 4.981421730190937e-06,
+      "loss": 4.5915,
+      "step": 10270
+    },
+    {
+      "epoch": 2.0154962730482544,
+      "grad_norm": 13.019744873046875,
+      "learning_rate": 4.977551281079001e-06,
+      "loss": 4.4266,
+      "step": 10275
+    },
+    {
+      "epoch": 2.01647704982346,
+      "grad_norm": 18.480030059814453,
+      "learning_rate": 4.973680845418903e-06,
+      "loss": 4.3731,
+      "step": 10280
+    },
+    {
+      "epoch": 2.0174578265986662,
+      "grad_norm": 18.247615814208984,
+      "learning_rate": 4.9698104255299015e-06,
+      "loss": 4.5799,
+      "step": 10285
+    },
+    {
+      "epoch": 2.018438603373872,
+      "grad_norm": 29.975618362426758,
+      "learning_rate": 4.965940023731255e-06,
+      "loss": 4.2928,
+      "step": 10290
+    },
+    {
+      "epoch": 2.019419380149078,
+      "grad_norm": 16.613618850708008,
+      "learning_rate": 4.9620696423422e-06,
+      "loss": 4.6338,
+      "step": 10295
+    },
+    {
+      "epoch": 2.020400156924284,
+      "grad_norm": 11.347552299499512,
+      "learning_rate": 4.958199283681968e-06,
+      "loss": 4.3939,
+      "step": 10300
+    },
+    {
+      "epoch": 2.02138093369949,
+      "grad_norm": 16.170337677001953,
+      "learning_rate": 4.954328950069778e-06,
+      "loss": 4.7013,
+      "step": 10305
+    },
+    {
+      "epoch": 2.022361710474696,
+      "grad_norm": 15.764708518981934,
+      "learning_rate": 4.95045864382483e-06,
+      "loss": 4.6302,
+      "step": 10310
+    },
+    {
+      "epoch": 2.0233424872499017,
+      "grad_norm": 20.800151824951172,
+      "learning_rate": 4.946588367266308e-06,
+      "loss": 4.7581,
+      "step": 10315
+    },
+    {
+      "epoch": 2.024323264025108,
+      "grad_norm": 14.423624038696289,
+      "learning_rate": 4.942718122713377e-06,
+      "loss": 4.5362,
+      "step": 10320
+    },
+    {
+      "epoch": 2.025304040800314,
+      "grad_norm": 27.55731773376465,
+      "learning_rate": 4.938847912485187e-06,
+      "loss": 4.485,
+      "step": 10325
+    },
+    {
+      "epoch": 2.0262848175755197,
+      "grad_norm": 30.498001098632812,
+      "learning_rate": 4.934977738900867e-06,
+      "loss": 4.3044,
+      "step": 10330
+    },
+    {
+      "epoch": 2.027265594350726,
+      "grad_norm": 28.44693374633789,
+      "learning_rate": 4.9311076042795185e-06,
+      "loss": 4.729,
+      "step": 10335
+    },
+    {
+      "epoch": 2.0282463711259315,
+      "grad_norm": 16.130205154418945,
+      "learning_rate": 4.927237510940228e-06,
+      "loss": 4.6165,
+      "step": 10340
+    },
+    {
+      "epoch": 2.0292271479011377,
+      "grad_norm": 14.545047760009766,
+      "learning_rate": 4.9233674612020485e-06,
+      "loss": 4.7173,
+      "step": 10345
+    },
+    {
+      "epoch": 2.030207924676344,
+      "grad_norm": 44.33458709716797,
+      "learning_rate": 4.919497457384012e-06,
+      "loss": 4.7184,
+      "step": 10350
+    },
+    {
+      "epoch": 2.0311887014515495,
+      "grad_norm": 16.259366989135742,
+      "learning_rate": 4.915627501805125e-06,
+      "loss": 4.555,
+      "step": 10355
+    },
+    {
+      "epoch": 2.0321694782267556,
+      "grad_norm": 20.897994995117188,
+      "learning_rate": 4.911757596784358e-06,
+      "loss": 4.2933,
+      "step": 10360
+    },
+    {
+      "epoch": 2.0331502550019613,
+      "grad_norm": 22.479900360107422,
+      "learning_rate": 4.907887744640659e-06,
+      "loss": 4.6158,
+      "step": 10365
+    },
+    {
+      "epoch": 2.0341310317771675,
+      "grad_norm": 21.30280876159668,
+      "learning_rate": 4.9040179476929364e-06,
+      "loss": 4.3196,
+      "step": 10370
+    },
+    {
+      "epoch": 2.0351118085523736,
+      "grad_norm": 14.088079452514648,
+      "learning_rate": 4.900148208260075e-06,
+      "loss": 4.6136,
+      "step": 10375
+    },
+    {
+      "epoch": 2.0360925853275793,
+      "grad_norm": 21.850318908691406,
+      "learning_rate": 4.896278528660916e-06,
+      "loss": 4.6937,
+      "step": 10380
+    },
+    {
+      "epoch": 2.0370733621027854,
+      "grad_norm": 24.06600570678711,
+      "learning_rate": 4.892408911214271e-06,
+      "loss": 4.5812,
+      "step": 10385
+    },
+    {
+      "epoch": 2.0380541388779916,
+      "grad_norm": 17.342010498046875,
+      "learning_rate": 4.888539358238912e-06,
+      "loss": 4.5512,
+      "step": 10390
+    },
+    {
+      "epoch": 2.0390349156531973,
+      "grad_norm": 21.402070999145508,
+      "learning_rate": 4.88466987205357e-06,
+      "loss": 4.3862,
+      "step": 10395
+    },
+    {
+      "epoch": 2.0400156924284034,
+      "grad_norm": 30.924278259277344,
+      "learning_rate": 4.880800454976939e-06,
+      "loss": 4.3662,
+      "step": 10400
+    },
+    {
+      "epoch": 2.040996469203609,
+      "grad_norm": 20.360410690307617,
+      "learning_rate": 4.876931109327675e-06,
+      "loss": 4.5773,
+      "step": 10405
+    },
+    {
+      "epoch": 2.0419772459788152,
+      "grad_norm": 26.364816665649414,
+      "learning_rate": 4.873061837424382e-06,
+      "loss": 4.3995,
+      "step": 10410
+    },
+    {
+      "epoch": 2.0429580227540214,
+      "grad_norm": 18.199127197265625,
+      "learning_rate": 4.869192641585628e-06,
+      "loss": 4.4404,
+      "step": 10415
+    },
+    {
+      "epoch": 2.043938799529227,
+      "grad_norm": 31.47274398803711,
+      "learning_rate": 4.8653235241299315e-06,
+      "loss": 4.5386,
+      "step": 10420
+    },
+    {
+      "epoch": 2.044919576304433,
+      "grad_norm": 36.052207946777344,
+      "learning_rate": 4.861454487375765e-06,
+      "loss": 4.5961,
+      "step": 10425
+    },
+    {
+      "epoch": 2.045900353079639,
+      "grad_norm": 30.397974014282227,
+      "learning_rate": 4.8575855336415536e-06,
+      "loss": 4.6295,
+      "step": 10430
+    },
+    {
+      "epoch": 2.046881129854845,
+      "grad_norm": 13.34601879119873,
+      "learning_rate": 4.853716665245668e-06,
+      "loss": 4.5577,
+      "step": 10435
+    },
+    {
+      "epoch": 2.047861906630051,
+      "grad_norm": 32.35442352294922,
+      "learning_rate": 4.849847884506437e-06,
+      "loss": 4.6545,
+      "step": 10440
+    },
+    {
+      "epoch": 2.048842683405257,
+      "grad_norm": 29.84942054748535,
+      "learning_rate": 4.8459791937421255e-06,
+      "loss": 4.6705,
+      "step": 10445
+    },
+    {
+      "epoch": 2.049823460180463,
+      "grad_norm": 41.33201599121094,
+      "learning_rate": 4.842110595270955e-06,
+      "loss": 4.295,
+      "step": 10450
+    },
+    {
+      "epoch": 2.0508042369556687,
+      "grad_norm": 30.651723861694336,
+      "learning_rate": 4.838242091411085e-06,
+      "loss": 4.4773,
+      "step": 10455
+    },
+    {
+      "epoch": 2.051785013730875,
+      "grad_norm": 17.958133697509766,
+      "learning_rate": 4.83437368448062e-06,
+      "loss": 4.7877,
+      "step": 10460
+    },
+    {
+      "epoch": 2.052765790506081,
+      "grad_norm": 21.545969009399414,
+      "learning_rate": 4.8305053767976075e-06,
+      "loss": 4.3349,
+      "step": 10465
+    },
+    {
+      "epoch": 2.0537465672812867,
+      "grad_norm": 17.481050491333008,
+      "learning_rate": 4.826637170680033e-06,
+      "loss": 4.5657,
+      "step": 10470
+    },
+    {
+      "epoch": 2.054727344056493,
+      "grad_norm": 24.961719512939453,
+      "learning_rate": 4.822769068445824e-06,
+      "loss": 4.6152,
+      "step": 10475
+    },
+    {
+      "epoch": 2.0557081208316985,
+      "grad_norm": 21.356874465942383,
+      "learning_rate": 4.818901072412846e-06,
+      "loss": 4.5025,
+      "step": 10480
+    },
+    {
+      "epoch": 2.0566888976069047,
+      "grad_norm": 19.501522064208984,
+      "learning_rate": 4.8150331848988965e-06,
+      "loss": 4.8735,
+      "step": 10485
+    },
+    {
+      "epoch": 2.057669674382111,
+      "grad_norm": 21.428464889526367,
+      "learning_rate": 4.811165408221715e-06,
+      "loss": 4.5817,
+      "step": 10490
+    },
+    {
+      "epoch": 2.0586504511573165,
+      "grad_norm": 10.177115440368652,
+      "learning_rate": 4.8072977446989665e-06,
+      "loss": 4.3053,
+      "step": 10495
+    },
+    {
+      "epoch": 2.0596312279325226,
+      "grad_norm": 32.65115737915039,
+      "learning_rate": 4.803430196648252e-06,
+      "loss": 3.9838,
+      "step": 10500
+    },
+    {
+      "epoch": 2.0606120047077283,
+      "grad_norm": 18.025699615478516,
+      "learning_rate": 4.799562766387109e-06,
+      "loss": 4.4535,
+      "step": 10505
+    },
+    {
+      "epoch": 2.0615927814829345,
+      "grad_norm": 13.657340049743652,
+      "learning_rate": 4.795695456232993e-06,
+      "loss": 4.7212,
+      "step": 10510
+    },
+    {
+      "epoch": 2.0625735582581406,
+      "grad_norm": 16.720195770263672,
+      "learning_rate": 4.791828268503297e-06,
+      "loss": 4.3234,
+      "step": 10515
+    },
+    {
+      "epoch": 2.0635543350333463,
+      "grad_norm": 25.29947853088379,
+      "learning_rate": 4.7879612055153335e-06,
+      "loss": 4.5086,
+      "step": 10520
+    },
+    {
+      "epoch": 2.0645351118085524,
+      "grad_norm": 15.523651123046875,
+      "learning_rate": 4.784094269586348e-06,
+      "loss": 4.567,
+      "step": 10525
+    },
+    {
+      "epoch": 2.065515888583758,
+      "grad_norm": 38.62691879272461,
+      "learning_rate": 4.780227463033505e-06,
+      "loss": 4.7915,
+      "step": 10530
+    },
+    {
+      "epoch": 2.0664966653589643,
+      "grad_norm": 23.588415145874023,
+      "learning_rate": 4.7763607881738884e-06,
+      "loss": 4.5609,
+      "step": 10535
+    },
+    {
+      "epoch": 2.0674774421341704,
+      "grad_norm": 23.60785484313965,
+      "learning_rate": 4.772494247324512e-06,
+      "loss": 4.8313,
+      "step": 10540
+    },
+    {
+      "epoch": 2.068458218909376,
+      "grad_norm": 26.25421905517578,
+      "learning_rate": 4.7686278428023e-06,
+      "loss": 4.4832,
+      "step": 10545
+    },
+    {
+      "epoch": 2.0694389956845822,
+      "grad_norm": 19.635887145996094,
+      "learning_rate": 4.7647615769241e-06,
+      "loss": 4.5175,
+      "step": 10550
+    },
+    {
+      "epoch": 2.0704197724597884,
+      "grad_norm": 11.955699920654297,
+      "learning_rate": 4.760895452006681e-06,
+      "loss": 4.7148,
+      "step": 10555
+    },
+    {
+      "epoch": 2.071400549234994,
+      "grad_norm": 27.53982162475586,
+      "learning_rate": 4.757029470366716e-06,
+      "loss": 4.3946,
+      "step": 10560
+    },
+    {
+      "epoch": 2.0723813260102,
+      "grad_norm": 14.448670387268066,
+      "learning_rate": 4.753163634320801e-06,
+      "loss": 4.7461,
+      "step": 10565
+    },
+    {
+      "epoch": 2.073362102785406,
+      "grad_norm": 19.246601104736328,
+      "learning_rate": 4.7492979461854405e-06,
+      "loss": 4.349,
+      "step": 10570
+    },
+    {
+      "epoch": 2.074342879560612,
+      "grad_norm": 17.460773468017578,
+      "learning_rate": 4.745432408277053e-06,
+      "loss": 4.1874,
+      "step": 10575
+    },
+    {
+      "epoch": 2.075323656335818,
+      "grad_norm": 21.156417846679688,
+      "learning_rate": 4.741567022911968e-06,
+      "loss": 4.2897,
+      "step": 10580
+    },
+    {
+      "epoch": 2.076304433111024,
+      "grad_norm": 27.80026626586914,
+      "learning_rate": 4.7377017924064175e-06,
+      "loss": 4.3745,
+      "step": 10585
+    },
+    {
+      "epoch": 2.07728520988623,
+      "grad_norm": 12.806593894958496,
+      "learning_rate": 4.733836719076549e-06,
+      "loss": 4.4964,
+      "step": 10590
+    },
+    {
+      "epoch": 2.0782659866614357,
+      "grad_norm": 11.110158920288086,
+      "learning_rate": 4.729971805238407e-06,
+      "loss": 4.4094,
+      "step": 10595
+    },
+    {
+      "epoch": 2.079246763436642,
+      "grad_norm": 14.680946350097656,
+      "learning_rate": 4.72610705320795e-06,
+      "loss": 4.976,
+      "step": 10600
+    },
+    {
+      "epoch": 2.080227540211848,
+      "grad_norm": 28.969358444213867,
+      "learning_rate": 4.722242465301033e-06,
+      "loss": 4.5051,
+      "step": 10605
+    },
+    {
+      "epoch": 2.0812083169870537,
+      "grad_norm": 13.442532539367676,
+      "learning_rate": 4.718378043833411e-06,
+      "loss": 4.4828,
+      "step": 10610
+    },
+    {
+      "epoch": 2.08218909376226,
+      "grad_norm": 14.584362983703613,
+      "learning_rate": 4.714513791120746e-06,
+      "loss": 4.1986,
+      "step": 10615
+    },
+    {
+      "epoch": 2.0831698705374655,
+      "grad_norm": 8.882427215576172,
+      "learning_rate": 4.710649709478593e-06,
+      "loss": 4.6214,
+      "step": 10620
+    },
+    {
+      "epoch": 2.0841506473126716,
+      "grad_norm": 32.854148864746094,
+      "learning_rate": 4.706785801222409e-06,
+      "loss": 4.904,
+      "step": 10625
+    },
+    {
+      "epoch": 2.0851314240878778,
+      "grad_norm": 12.650853157043457,
+      "learning_rate": 4.702922068667546e-06,
+      "loss": 4.6661,
+      "step": 10630
+    },
+    {
+      "epoch": 2.0861122008630835,
+      "grad_norm": 24.82941436767578,
+      "learning_rate": 4.699058514129246e-06,
+      "loss": 4.5173,
+      "step": 10635
+    },
+    {
+      "epoch": 2.0870929776382896,
+      "grad_norm": 32.43788528442383,
+      "learning_rate": 4.695195139922652e-06,
+      "loss": 4.6476,
+      "step": 10640
+    },
+    {
+      "epoch": 2.0880737544134953,
+      "grad_norm": 12.398055076599121,
+      "learning_rate": 4.691331948362789e-06,
+      "loss": 4.5709,
+      "step": 10645
+    },
+    {
+      "epoch": 2.0890545311887014,
+      "grad_norm": 19.99089241027832,
+      "learning_rate": 4.687468941764583e-06,
+      "loss": 4.6279,
+      "step": 10650
+    },
+    {
+      "epoch": 2.0900353079639076,
+      "grad_norm": 19.84067726135254,
+      "learning_rate": 4.683606122442846e-06,
+      "loss": 4.7469,
+      "step": 10655
+    },
+    {
+      "epoch": 2.0910160847391133,
+      "grad_norm": 21.618118286132812,
+      "learning_rate": 4.679743492712273e-06,
+      "loss": 5.2934,
+      "step": 10660
+    },
+    {
+      "epoch": 2.0919968615143194,
+      "grad_norm": 15.103788375854492,
+      "learning_rate": 4.675881054887451e-06,
+      "loss": 4.5272,
+      "step": 10665
+    },
+    {
+      "epoch": 2.092977638289525,
+      "grad_norm": 26.1807861328125,
+      "learning_rate": 4.672018811282849e-06,
+      "loss": 4.5789,
+      "step": 10670
+    },
+    {
+      "epoch": 2.0939584150647312,
+      "grad_norm": 14.807060241699219,
+      "learning_rate": 4.6681567642128195e-06,
+      "loss": 5.14,
+      "step": 10675
+    },
+    {
+      "epoch": 2.0949391918399374,
+      "grad_norm": 22.160566329956055,
+      "learning_rate": 4.664294915991601e-06,
+      "loss": 4.3984,
+      "step": 10680
+    },
+    {
+      "epoch": 2.095919968615143,
+      "grad_norm": 24.023670196533203,
+      "learning_rate": 4.660433268933306e-06,
+      "loss": 4.5185,
+      "step": 10685
+    },
+    {
+      "epoch": 2.096900745390349,
+      "grad_norm": 13.71312427520752,
+      "learning_rate": 4.656571825351936e-06,
+      "loss": 4.5645,
+      "step": 10690
+    },
+    {
+      "epoch": 2.0978815221655553,
+      "grad_norm": 19.164081573486328,
+      "learning_rate": 4.6527105875613574e-06,
+      "loss": 4.2813,
+      "step": 10695
+    },
+    {
+      "epoch": 2.098862298940761,
+      "grad_norm": 22.777475357055664,
+      "learning_rate": 4.6488495578753285e-06,
+      "loss": 4.3725,
+      "step": 10700
+    },
+    {
+      "epoch": 2.099843075715967,
+      "grad_norm": 20.250986099243164,
+      "learning_rate": 4.644988738607471e-06,
+      "loss": 4.4706,
+      "step": 10705
+    },
+    {
+      "epoch": 2.100823852491173,
+      "grad_norm": 11.561543464660645,
+      "learning_rate": 4.641128132071287e-06,
+      "loss": 4.6326,
+      "step": 10710
+    },
+    {
+      "epoch": 2.101804629266379,
+      "grad_norm": 20.83006477355957,
+      "learning_rate": 4.637267740580149e-06,
+      "loss": 4.3374,
+      "step": 10715
+    },
+    {
+      "epoch": 2.102785406041585,
+      "grad_norm": 10.181340217590332,
+      "learning_rate": 4.633407566447297e-06,
+      "loss": 4.5099,
+      "step": 10720
+    },
+    {
+      "epoch": 2.103766182816791,
+      "grad_norm": 12.635647773742676,
+      "learning_rate": 4.629547611985848e-06,
+      "loss": 4.2601,
+      "step": 10725
+    },
+    {
+      "epoch": 2.104746959591997,
+      "grad_norm": 15.131701469421387,
+      "learning_rate": 4.625687879508783e-06,
+      "loss": 4.5051,
+      "step": 10730
+    },
+    {
+      "epoch": 2.1057277363672027,
+      "grad_norm": 16.336679458618164,
+      "learning_rate": 4.62182837132895e-06,
+      "loss": 4.3319,
+      "step": 10735
+    },
+    {
+      "epoch": 2.106708513142409,
+      "grad_norm": 32.087459564208984,
+      "learning_rate": 4.617969089759066e-06,
+      "loss": 4.6912,
+      "step": 10740
+    },
+    {
+      "epoch": 2.107689289917615,
+      "grad_norm": 13.97162914276123,
+      "learning_rate": 4.614110037111706e-06,
+      "loss": 4.6789,
+      "step": 10745
+    },
+    {
+      "epoch": 2.1086700666928206,
+      "grad_norm": 49.92332077026367,
+      "learning_rate": 4.6102512156993116e-06,
+      "loss": 4.7705,
+      "step": 10750
+    },
+    {
+      "epoch": 2.109650843468027,
+      "grad_norm": 31.643251419067383,
+      "learning_rate": 4.6063926278341895e-06,
+      "loss": 4.4707,
+      "step": 10755
+    },
+    {
+      "epoch": 2.1106316202432325,
+      "grad_norm": 17.356582641601562,
+      "learning_rate": 4.602534275828498e-06,
+      "loss": 4.3755,
+      "step": 10760
+    },
+    {
+      "epoch": 2.1116123970184386,
+      "grad_norm": 25.40827178955078,
+      "learning_rate": 4.598676161994262e-06,
+      "loss": 4.5291,
+      "step": 10765
+    },
+    {
+      "epoch": 2.1125931737936448,
+      "grad_norm": 12.720136642456055,
+      "learning_rate": 4.594818288643356e-06,
+      "loss": 4.5491,
+      "step": 10770
+    },
+    {
+      "epoch": 2.1135739505688504,
+      "grad_norm": 17.85308074951172,
+      "learning_rate": 4.59096065808752e-06,
+      "loss": 4.6167,
+      "step": 10775
+    },
+    {
+      "epoch": 2.1145547273440566,
+      "grad_norm": 38.109039306640625,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 4.8616,
+      "step": 10780
+    },
+    {
+      "epoch": 2.1155355041192623,
+      "grad_norm": 19.8563289642334,
+      "learning_rate": 4.583246134607258e-06,
+      "loss": 4.6665,
+      "step": 10785
+    },
+    {
+      "epoch": 2.1165162808944684,
+      "grad_norm": 39.26411437988281,
+      "learning_rate": 4.57938924630557e-06,
+      "loss": 4.3647,
+      "step": 10790
+    },
+    {
+      "epoch": 2.1174970576696746,
+      "grad_norm": 25.38083267211914,
+      "learning_rate": 4.575532610044419e-06,
+      "loss": 4.6439,
+      "step": 10795
+    },
+    {
+      "epoch": 2.1184778344448802,
+      "grad_norm": 18.37013816833496,
+      "learning_rate": 4.571676228134798e-06,
+      "loss": 4.7012,
+      "step": 10800
+    },
+    {
+      "epoch": 2.1194586112200864,
+      "grad_norm": 25.82213020324707,
+      "learning_rate": 4.567820102887552e-06,
+      "loss": 4.4759,
+      "step": 10805
+    },
+    {
+      "epoch": 2.120439387995292,
+      "grad_norm": 18.050235748291016,
+      "learning_rate": 4.563964236613362e-06,
+      "loss": 4.3479,
+      "step": 10810
+    },
+    {
+      "epoch": 2.121420164770498,
+      "grad_norm": 21.06696891784668,
+      "learning_rate": 4.560108631622765e-06,
+      "loss": 4.6009,
+      "step": 10815
+    },
+    {
+      "epoch": 2.1224009415457044,
+      "grad_norm": 26.383195877075195,
+      "learning_rate": 4.556253290226135e-06,
+      "loss": 4.4963,
+      "step": 10820
+    },
+    {
+      "epoch": 2.12338171832091,
+      "grad_norm": 15.887774467468262,
+      "learning_rate": 4.552398214733686e-06,
+      "loss": 4.3878,
+      "step": 10825
+    },
+    {
+      "epoch": 2.124362495096116,
+      "grad_norm": 23.23624038696289,
+      "learning_rate": 4.548543407455482e-06,
+      "loss": 4.7093,
+      "step": 10830
+    },
+    {
+      "epoch": 2.125343271871322,
+      "grad_norm": 17.61455535888672,
+      "learning_rate": 4.544688870701416e-06,
+      "loss": 4.5022,
+      "step": 10835
+    },
+    {
+      "epoch": 2.126324048646528,
+      "grad_norm": 12.441754341125488,
+      "learning_rate": 4.540834606781226e-06,
+      "loss": 4.4838,
+      "step": 10840
+    },
+    {
+      "epoch": 2.127304825421734,
+      "grad_norm": 15.740215301513672,
+      "learning_rate": 4.536980618004481e-06,
+      "loss": 4.3949,
+      "step": 10845
+    },
+    {
+      "epoch": 2.12828560219694,
+      "grad_norm": 13.362967491149902,
+      "learning_rate": 4.533126906680591e-06,
+      "loss": 4.495,
+      "step": 10850
+    },
+    {
+      "epoch": 2.129266378972146,
+      "grad_norm": 11.252303123474121,
+      "learning_rate": 4.529273475118797e-06,
+      "loss": 4.438,
+      "step": 10855
+    },
+    {
+      "epoch": 2.1302471557473517,
+      "grad_norm": 11.080646514892578,
+      "learning_rate": 4.525420325628167e-06,
+      "loss": 4.3683,
+      "step": 10860
+    },
+    {
+      "epoch": 2.131227932522558,
+      "grad_norm": 17.542789459228516,
+      "learning_rate": 4.521567460517612e-06,
+      "loss": 4.7441,
+      "step": 10865
+    },
+    {
+      "epoch": 2.132208709297764,
+      "grad_norm": 14.918002128601074,
+      "learning_rate": 4.517714882095859e-06,
+      "loss": 4.8096,
+      "step": 10870
+    },
+    {
+      "epoch": 2.1331894860729697,
+      "grad_norm": 16.332870483398438,
+      "learning_rate": 4.5138625926714734e-06,
+      "loss": 4.3374,
+      "step": 10875
+    },
+    {
+      "epoch": 2.134170262848176,
+      "grad_norm": 14.32162857055664,
+      "learning_rate": 4.510010594552846e-06,
+      "loss": 4.4616,
+      "step": 10880
+    },
+    {
+      "epoch": 2.1351510396233815,
+      "grad_norm": 26.749788284301758,
+      "learning_rate": 4.506158890048187e-06,
+      "loss": 4.7608,
+      "step": 10885
+    },
+    {
+      "epoch": 2.1361318163985876,
+      "grad_norm": 16.43140983581543,
+      "learning_rate": 4.502307481465536e-06,
+      "loss": 4.4194,
+      "step": 10890
+    },
+    {
+      "epoch": 2.1371125931737938,
+      "grad_norm": 36.98377990722656,
+      "learning_rate": 4.498456371112753e-06,
+      "loss": 4.265,
+      "step": 10895
+    },
+    {
+      "epoch": 2.1380933699489995,
+      "grad_norm": 12.881791114807129,
+      "learning_rate": 4.494605561297521e-06,
+      "loss": 4.7524,
+      "step": 10900
+    },
+    {
+      "epoch": 2.1390741467242056,
+      "grad_norm": 12.371302604675293,
+      "learning_rate": 4.4907550543273436e-06,
+      "loss": 4.4738,
+      "step": 10905
+    },
+    {
+      "epoch": 2.1400549234994117,
+      "grad_norm": 17.81466293334961,
+      "learning_rate": 4.486904852509537e-06,
+      "loss": 4.5088,
+      "step": 10910
+    },
+    {
+      "epoch": 2.1410357002746174,
+      "grad_norm": 14.954310417175293,
+      "learning_rate": 4.483054958151244e-06,
+      "loss": 4.4856,
+      "step": 10915
+    },
+    {
+      "epoch": 2.1420164770498236,
+      "grad_norm": 25.58282470703125,
+      "learning_rate": 4.479205373559415e-06,
+      "loss": 4.5459,
+      "step": 10920
+    },
+    {
+      "epoch": 2.1429972538250293,
+      "grad_norm": 13.035386085510254,
+      "learning_rate": 4.475356101040818e-06,
+      "loss": 4.2163,
+      "step": 10925
+    },
+    {
+      "epoch": 2.1439780306002354,
+      "grad_norm": 21.911907196044922,
+      "learning_rate": 4.471507142902036e-06,
+      "loss": 4.2241,
+      "step": 10930
+    },
+    {
+      "epoch": 2.1449588073754415,
+      "grad_norm": 13.07209300994873,
+      "learning_rate": 4.467658501449458e-06,
+      "loss": 4.6244,
+      "step": 10935
+    },
+    {
+      "epoch": 2.1459395841506472,
+      "grad_norm": 20.160175323486328,
+      "learning_rate": 4.463810178989291e-06,
+      "loss": 4.3972,
+      "step": 10940
+    },
+    {
+      "epoch": 2.1469203609258534,
+      "grad_norm": 15.461299896240234,
+      "learning_rate": 4.459962177827543e-06,
+      "loss": 4.3091,
+      "step": 10945
+    },
+    {
+      "epoch": 2.147901137701059,
+      "grad_norm": 28.049442291259766,
+      "learning_rate": 4.4561145002700325e-06,
+      "loss": 4.393,
+      "step": 10950
+    },
+    {
+      "epoch": 2.148881914476265,
+      "grad_norm": 21.617782592773438,
+      "learning_rate": 4.452267148622389e-06,
+      "loss": 4.5262,
+      "step": 10955
+    },
+    {
+      "epoch": 2.1498626912514713,
+      "grad_norm": 16.35369873046875,
+      "learning_rate": 4.448420125190039e-06,
+      "loss": 4.462,
+      "step": 10960
+    },
+    {
+      "epoch": 2.150843468026677,
+      "grad_norm": 19.262086868286133,
+      "learning_rate": 4.444573432278217e-06,
+      "loss": 4.565,
+      "step": 10965
+    },
+    {
+      "epoch": 2.151824244801883,
+      "grad_norm": 30.79473876953125,
+      "learning_rate": 4.440727072191956e-06,
+      "loss": 4.5552,
+      "step": 10970
+    },
+    {
+      "epoch": 2.152805021577089,
+      "grad_norm": 19.929492950439453,
+      "learning_rate": 4.436881047236092e-06,
+      "loss": 4.6865,
+      "step": 10975
+    },
+    {
+      "epoch": 2.153785798352295,
+      "grad_norm": 22.288394927978516,
+      "learning_rate": 4.433035359715264e-06,
+      "loss": 4.4575,
+      "step": 10980
+    },
+    {
+      "epoch": 2.154766575127501,
+      "grad_norm": 20.23619842529297,
+      "learning_rate": 4.429190011933899e-06,
+      "loss": 4.4639,
+      "step": 10985
+    },
+    {
+      "epoch": 2.155747351902707,
+      "grad_norm": 29.274555206298828,
+      "learning_rate": 4.425345006196231e-06,
+      "loss": 4.3485,
+      "step": 10990
+    },
+    {
+      "epoch": 2.156728128677913,
+      "grad_norm": 17.464130401611328,
+      "learning_rate": 4.421500344806281e-06,
+      "loss": 4.5714,
+      "step": 10995
+    },
+    {
+      "epoch": 2.1577089054531187,
+      "grad_norm": 33.32584762573242,
+      "learning_rate": 4.417656030067866e-06,
+      "loss": 4.645,
+      "step": 11000
+    },
+    {
+      "epoch": 2.158689682228325,
+      "grad_norm": 23.476160049438477,
+      "learning_rate": 4.4138120642846e-06,
+      "loss": 4.5932,
+      "step": 11005
+    },
+    {
+      "epoch": 2.159670459003531,
+      "grad_norm": 30.347623825073242,
+      "learning_rate": 4.409968449759879e-06,
+      "loss": 4.3533,
+      "step": 11010
+    },
+    {
+      "epoch": 2.1606512357787366,
+      "grad_norm": 24.87689781188965,
+      "learning_rate": 4.406125188796898e-06,
+      "loss": 4.6235,
+      "step": 11015
+    },
+    {
+      "epoch": 2.1616320125539428,
+      "grad_norm": 15.419154167175293,
+      "learning_rate": 4.4022822836986315e-06,
+      "loss": 4.4848,
+      "step": 11020
+    },
+    {
+      "epoch": 2.162612789329149,
+      "grad_norm": 21.806236267089844,
+      "learning_rate": 4.3984397367678475e-06,
+      "loss": 4.7521,
+      "step": 11025
+    },
+    {
+      "epoch": 2.1635935661043546,
+      "grad_norm": 29.77591896057129,
+      "learning_rate": 4.394597550307097e-06,
+      "loss": 4.8401,
+      "step": 11030
+    },
+    {
+      "epoch": 2.1645743428795607,
+      "grad_norm": 13.51814079284668,
+      "learning_rate": 4.390755726618714e-06,
+      "loss": 4.5055,
+      "step": 11035
+    },
+    {
+      "epoch": 2.1655551196547664,
+      "grad_norm": 13.01634407043457,
+      "learning_rate": 4.386914268004815e-06,
+      "loss": 4.69,
+      "step": 11040
+    },
+    {
+      "epoch": 2.1665358964299726,
+      "grad_norm": 16.430099487304688,
+      "learning_rate": 4.383073176767299e-06,
+      "loss": 4.8869,
+      "step": 11045
+    },
+    {
+      "epoch": 2.1675166732051787,
+      "grad_norm": 15.964555740356445,
+      "learning_rate": 4.379232455207843e-06,
+      "loss": 4.2059,
+      "step": 11050
+    },
+    {
+      "epoch": 2.1684974499803844,
+      "grad_norm": 12.990334510803223,
+      "learning_rate": 4.375392105627909e-06,
+      "loss": 4.5416,
+      "step": 11055
+    },
+    {
+      "epoch": 2.1694782267555905,
+      "grad_norm": 28.908823013305664,
+      "learning_rate": 4.371552130328725e-06,
+      "loss": 4.3389,
+      "step": 11060
+    },
+    {
+      "epoch": 2.1704590035307962,
+      "grad_norm": 34.378807067871094,
+      "learning_rate": 4.367712531611305e-06,
+      "loss": 4.5571,
+      "step": 11065
+    },
+    {
+      "epoch": 2.1714397803060024,
+      "grad_norm": 18.134485244750977,
+      "learning_rate": 4.3638733117764295e-06,
+      "loss": 4.2235,
+      "step": 11070
+    },
+    {
+      "epoch": 2.1724205570812085,
+      "grad_norm": 25.642366409301758,
+      "learning_rate": 4.360034473124658e-06,
+      "loss": 4.6163,
+      "step": 11075
+    },
+    {
+      "epoch": 2.173401333856414,
+      "grad_norm": 10.750842094421387,
+      "learning_rate": 4.35619601795632e-06,
+      "loss": 4.6841,
+      "step": 11080
+    },
+    {
+      "epoch": 2.1743821106316203,
+      "grad_norm": 12.613737106323242,
+      "learning_rate": 4.3523579485715105e-06,
+      "loss": 4.5547,
+      "step": 11085
+    },
+    {
+      "epoch": 2.175362887406826,
+      "grad_norm": 14.869281768798828,
+      "learning_rate": 4.348520267270102e-06,
+      "loss": 4.6588,
+      "step": 11090
+    },
+    {
+      "epoch": 2.176343664182032,
+      "grad_norm": 21.500011444091797,
+      "learning_rate": 4.344682976351725e-06,
+      "loss": 4.3966,
+      "step": 11095
+    },
+    {
+      "epoch": 2.1773244409572383,
+      "grad_norm": 13.638593673706055,
+      "learning_rate": 4.340846078115784e-06,
+      "loss": 4.429,
+      "step": 11100
+    },
+    {
+      "epoch": 2.178305217732444,
+      "grad_norm": 22.606096267700195,
+      "learning_rate": 4.337009574861443e-06,
+      "loss": 5.035,
+      "step": 11105
+    },
+    {
+      "epoch": 2.17928599450765,
+      "grad_norm": 13.18099594116211,
+      "learning_rate": 4.333173468887632e-06,
+      "loss": 4.4518,
+      "step": 11110
+    },
+    {
+      "epoch": 2.180266771282856,
+      "grad_norm": 10.771079063415527,
+      "learning_rate": 4.329337762493044e-06,
+      "loss": 4.8277,
+      "step": 11115
+    },
+    {
+      "epoch": 2.181247548058062,
+      "grad_norm": 15.482091903686523,
+      "learning_rate": 4.325502457976126e-06,
+      "loss": 4.3848,
+      "step": 11120
+    },
+    {
+      "epoch": 2.182228324833268,
+      "grad_norm": 17.949617385864258,
+      "learning_rate": 4.321667557635092e-06,
+      "loss": 4.3615,
+      "step": 11125
+    },
+    {
+      "epoch": 2.183209101608474,
+      "grad_norm": 18.554176330566406,
+      "learning_rate": 4.317833063767912e-06,
+      "loss": 4.4504,
+      "step": 11130
+    },
+    {
+      "epoch": 2.18418987838368,
+      "grad_norm": 30.101398468017578,
+      "learning_rate": 4.313998978672308e-06,
+      "loss": 4.6963,
+      "step": 11135
+    },
+    {
+      "epoch": 2.1851706551588856,
+      "grad_norm": 13.602922439575195,
+      "learning_rate": 4.310165304645763e-06,
+      "loss": 4.5375,
+      "step": 11140
+    },
+    {
+      "epoch": 2.186151431934092,
+      "grad_norm": 15.757498741149902,
+      "learning_rate": 4.3063320439855085e-06,
+      "loss": 4.5059,
+      "step": 11145
+    },
+    {
+      "epoch": 2.187132208709298,
+      "grad_norm": 23.122838973999023,
+      "learning_rate": 4.302499198988531e-06,
+      "loss": 4.4306,
+      "step": 11150
+    },
+    {
+      "epoch": 2.1881129854845036,
+      "grad_norm": 23.106101989746094,
+      "learning_rate": 4.29866677195157e-06,
+      "loss": 4.1968,
+      "step": 11155
+    },
+    {
+      "epoch": 2.1890937622597098,
+      "grad_norm": 17.85923194885254,
+      "learning_rate": 4.294834765171108e-06,
+      "loss": 4.6488,
+      "step": 11160
+    },
+    {
+      "epoch": 2.1900745390349154,
+      "grad_norm": 8.963134765625,
+      "learning_rate": 4.291003180943385e-06,
+      "loss": 4.3469,
+      "step": 11165
+    },
+    {
+      "epoch": 2.1910553158101216,
+      "grad_norm": 28.65680503845215,
+      "learning_rate": 4.287172021564377e-06,
+      "loss": 4.349,
+      "step": 11170
+    },
+    {
+      "epoch": 2.1920360925853277,
+      "grad_norm": 23.335914611816406,
+      "learning_rate": 4.283341289329815e-06,
+      "loss": 4.6713,
+      "step": 11175
+    },
+    {
+      "epoch": 2.1930168693605334,
+      "grad_norm": 10.928681373596191,
+      "learning_rate": 4.279510986535169e-06,
+      "loss": 4.388,
+      "step": 11180
+    },
+    {
+      "epoch": 2.1939976461357396,
+      "grad_norm": 15.44399642944336,
+      "learning_rate": 4.275681115475651e-06,
+      "loss": 4.4258,
+      "step": 11185
+    },
+    {
+      "epoch": 2.1949784229109452,
+      "grad_norm": 32.2411003112793,
+      "learning_rate": 4.27185167844622e-06,
+      "loss": 4.3384,
+      "step": 11190
+    },
+    {
+      "epoch": 2.1959591996861514,
+      "grad_norm": 18.149885177612305,
+      "learning_rate": 4.268022677741566e-06,
+      "loss": 4.7455,
+      "step": 11195
+    },
+    {
+      "epoch": 2.1969399764613575,
+      "grad_norm": 18.782310485839844,
+      "learning_rate": 4.264194115656124e-06,
+      "loss": 4.6965,
+      "step": 11200
+    },
+    {
+      "epoch": 2.197920753236563,
+      "grad_norm": 11.856428146362305,
+      "learning_rate": 4.260365994484069e-06,
+      "loss": 4.4527,
+      "step": 11205
+    },
+    {
+      "epoch": 2.1989015300117694,
+      "grad_norm": 21.500896453857422,
+      "learning_rate": 4.256538316519303e-06,
+      "loss": 4.3852,
+      "step": 11210
+    },
+    {
+      "epoch": 2.199882306786975,
+      "grad_norm": 16.60177993774414,
+      "learning_rate": 4.252711084055468e-06,
+      "loss": 4.3806,
+      "step": 11215
+    },
+    {
+      "epoch": 2.200863083562181,
+      "grad_norm": 18.734601974487305,
+      "learning_rate": 4.248884299385937e-06,
+      "loss": 4.4284,
+      "step": 11220
+    },
+    {
+      "epoch": 2.2018438603373873,
+      "grad_norm": 21.753625869750977,
+      "learning_rate": 4.245057964803815e-06,
+      "loss": 4.5484,
+      "step": 11225
+    },
+    {
+      "epoch": 2.202824637112593,
+      "grad_norm": 13.715068817138672,
+      "learning_rate": 4.2412320826019425e-06,
+      "loss": 4.5648,
+      "step": 11230
+    },
+    {
+      "epoch": 2.203805413887799,
+      "grad_norm": 28.02129554748535,
+      "learning_rate": 4.237406655072879e-06,
+      "loss": 4.5578,
+      "step": 11235
+    },
+    {
+      "epoch": 2.2047861906630053,
+      "grad_norm": 19.558780670166016,
+      "learning_rate": 4.23358168450892e-06,
+      "loss": 4.7332,
+      "step": 11240
+    },
+    {
+      "epoch": 2.205766967438211,
+      "grad_norm": 27.703575134277344,
+      "learning_rate": 4.229757173202082e-06,
+      "loss": 4.7346,
+      "step": 11245
+    },
+    {
+      "epoch": 2.206747744213417,
+      "grad_norm": 20.822980880737305,
+      "learning_rate": 4.225933123444108e-06,
+      "loss": 4.5904,
+      "step": 11250
+    },
+    {
+      "epoch": 2.207728520988623,
+      "grad_norm": 23.84830665588379,
+      "learning_rate": 4.22210953752647e-06,
+      "loss": 4.6531,
+      "step": 11255
+    },
+    {
+      "epoch": 2.208709297763829,
+      "grad_norm": 10.470589637756348,
+      "learning_rate": 4.218286417740348e-06,
+      "loss": 4.2309,
+      "step": 11260
+    },
+    {
+      "epoch": 2.209690074539035,
+      "grad_norm": 22.220144271850586,
+      "learning_rate": 4.21446376637666e-06,
+      "loss": 4.3394,
+      "step": 11265
+    },
+    {
+      "epoch": 2.210670851314241,
+      "grad_norm": 22.664947509765625,
+      "learning_rate": 4.210641585726029e-06,
+      "loss": 4.6662,
+      "step": 11270
+    },
+    {
+      "epoch": 2.211651628089447,
+      "grad_norm": 18.833683013916016,
+      "learning_rate": 4.206819878078803e-06,
+      "loss": 4.368,
+      "step": 11275
+    },
+    {
+      "epoch": 2.2126324048646526,
+      "grad_norm": 15.988152503967285,
+      "learning_rate": 4.2029986457250495e-06,
+      "loss": 4.3532,
+      "step": 11280
+    },
+    {
+      "epoch": 2.2136131816398588,
+      "grad_norm": 37.39259719848633,
+      "learning_rate": 4.199177890954541e-06,
+      "loss": 4.5568,
+      "step": 11285
+    },
+    {
+      "epoch": 2.214593958415065,
+      "grad_norm": 10.75390625,
+      "learning_rate": 4.195357616056774e-06,
+      "loss": 4.4022,
+      "step": 11290
+    },
+    {
+      "epoch": 2.2155747351902706,
+      "grad_norm": 25.854948043823242,
+      "learning_rate": 4.19153782332095e-06,
+      "loss": 4.4454,
+      "step": 11295
+    },
+    {
+      "epoch": 2.2165555119654767,
+      "grad_norm": 16.350391387939453,
+      "learning_rate": 4.187718515035986e-06,
+      "loss": 4.3026,
+      "step": 11300
+    },
+    {
+      "epoch": 2.2175362887406824,
+      "grad_norm": 11.197614669799805,
+      "learning_rate": 4.18389969349051e-06,
+      "loss": 4.378,
+      "step": 11305
+    },
+    {
+      "epoch": 2.2185170655158886,
+      "grad_norm": 21.271650314331055,
+      "learning_rate": 4.180081360972852e-06,
+      "loss": 4.5606,
+      "step": 11310
+    },
+    {
+      "epoch": 2.2194978422910947,
+      "grad_norm": 31.11151123046875,
+      "learning_rate": 4.176263519771058e-06,
+      "loss": 4.7756,
+      "step": 11315
+    },
+    {
+      "epoch": 2.2204786190663004,
+      "grad_norm": 18.244281768798828,
+      "learning_rate": 4.172446172172868e-06,
+      "loss": 4.5988,
+      "step": 11320
+    },
+    {
+      "epoch": 2.2214593958415065,
+      "grad_norm": 17.656965255737305,
+      "learning_rate": 4.168629320465737e-06,
+      "loss": 4.2405,
+      "step": 11325
+    },
+    {
+      "epoch": 2.2224401726167122,
+      "grad_norm": 17.68485450744629,
+      "learning_rate": 4.164812966936818e-06,
+      "loss": 4.6696,
+      "step": 11330
+    },
+    {
+      "epoch": 2.2234209493919184,
+      "grad_norm": 19.626468658447266,
+      "learning_rate": 4.160997113872964e-06,
+      "loss": 4.6334,
+      "step": 11335
+    },
+    {
+      "epoch": 2.2244017261671245,
+      "grad_norm": 12.249372482299805,
+      "learning_rate": 4.157181763560732e-06,
+      "loss": 4.4522,
+      "step": 11340
+    },
+    {
+      "epoch": 2.22538250294233,
+      "grad_norm": 18.31877326965332,
+      "learning_rate": 4.153366918286374e-06,
+      "loss": 4.6883,
+      "step": 11345
+    },
+    {
+      "epoch": 2.2263632797175363,
+      "grad_norm": 16.244726181030273,
+      "learning_rate": 4.149552580335843e-06,
+      "loss": 4.6329,
+      "step": 11350
+    },
+    {
+      "epoch": 2.2273440564927425,
+      "grad_norm": 38.69582748413086,
+      "learning_rate": 4.1457387519947864e-06,
+      "loss": 4.6677,
+      "step": 11355
+    },
+    {
+      "epoch": 2.228324833267948,
+      "grad_norm": 16.428791046142578,
+      "learning_rate": 4.141925435548545e-06,
+      "loss": 4.4718,
+      "step": 11360
+    },
+    {
+      "epoch": 2.2293056100431543,
+      "grad_norm": 30.045249938964844,
+      "learning_rate": 4.138112633282154e-06,
+      "loss": 4.5068,
+      "step": 11365
+    },
+    {
+      "epoch": 2.23028638681836,
+      "grad_norm": 21.004417419433594,
+      "learning_rate": 4.13430034748034e-06,
+      "loss": 4.5831,
+      "step": 11370
+    },
+    {
+      "epoch": 2.231267163593566,
+      "grad_norm": 18.86147689819336,
+      "learning_rate": 4.13048858042752e-06,
+      "loss": 4.4865,
+      "step": 11375
+    },
+    {
+      "epoch": 2.2322479403687723,
+      "grad_norm": 26.14810562133789,
+      "learning_rate": 4.126677334407804e-06,
+      "loss": 4.5101,
+      "step": 11380
+    },
+    {
+      "epoch": 2.233228717143978,
+      "grad_norm": 23.22881507873535,
+      "learning_rate": 4.122866611704981e-06,
+      "loss": 4.6181,
+      "step": 11385
+    },
+    {
+      "epoch": 2.234209493919184,
+      "grad_norm": 21.121158599853516,
+      "learning_rate": 4.119056414602538e-06,
+      "loss": 4.397,
+      "step": 11390
+    },
+    {
+      "epoch": 2.23519027069439,
+      "grad_norm": 22.4326229095459,
+      "learning_rate": 4.115246745383636e-06,
+      "loss": 4.5258,
+      "step": 11395
+    },
+    {
+      "epoch": 2.236171047469596,
+      "grad_norm": 11.575772285461426,
+      "learning_rate": 4.111437606331126e-06,
+      "loss": 4.5109,
+      "step": 11400
+    },
+    {
+      "epoch": 2.237151824244802,
+      "grad_norm": 34.85311508178711,
+      "learning_rate": 4.107628999727542e-06,
+      "loss": 4.6383,
+      "step": 11405
+    },
+    {
+      "epoch": 2.2381326010200078,
+      "grad_norm": 37.00575256347656,
+      "learning_rate": 4.103820927855092e-06,
+      "loss": 4.5175,
+      "step": 11410
+    },
+    {
+      "epoch": 2.239113377795214,
+      "grad_norm": 15.301085472106934,
+      "learning_rate": 4.1000133929956745e-06,
+      "loss": 4.5078,
+      "step": 11415
+    },
+    {
+      "epoch": 2.2400941545704196,
+      "grad_norm": 19.144397735595703,
+      "learning_rate": 4.096206397430855e-06,
+      "loss": 4.3975,
+      "step": 11420
+    },
+    {
+      "epoch": 2.2410749313456257,
+      "grad_norm": 22.98430633544922,
+      "learning_rate": 4.092399943441884e-06,
+      "loss": 4.6567,
+      "step": 11425
+    },
+    {
+      "epoch": 2.242055708120832,
+      "grad_norm": 16.777803421020508,
+      "learning_rate": 4.088594033309683e-06,
+      "loss": 4.5046,
+      "step": 11430
+    },
+    {
+      "epoch": 2.2430364848960376,
+      "grad_norm": 19.35264015197754,
+      "learning_rate": 4.0847886693148495e-06,
+      "loss": 4.2147,
+      "step": 11435
+    },
+    {
+      "epoch": 2.2440172616712437,
+      "grad_norm": 29.10360336303711,
+      "learning_rate": 4.080983853737654e-06,
+      "loss": 4.7951,
+      "step": 11440
+    },
+    {
+      "epoch": 2.2449980384464494,
+      "grad_norm": 16.541683197021484,
+      "learning_rate": 4.077179588858035e-06,
+      "loss": 4.389,
+      "step": 11445
+    },
+    {
+      "epoch": 2.2459788152216555,
+      "grad_norm": 18.082550048828125,
+      "learning_rate": 4.073375876955606e-06,
+      "loss": 4.8307,
+      "step": 11450
+    },
+    {
+      "epoch": 2.2469595919968617,
+      "grad_norm": 13.92489242553711,
+      "learning_rate": 4.0695727203096466e-06,
+      "loss": 4.599,
+      "step": 11455
+    },
+    {
+      "epoch": 2.2479403687720674,
+      "grad_norm": 22.647361755371094,
+      "learning_rate": 4.065770121199103e-06,
+      "loss": 4.2624,
+      "step": 11460
+    },
+    {
+      "epoch": 2.2489211455472735,
+      "grad_norm": 23.03203773498535,
+      "learning_rate": 4.061968081902591e-06,
+      "loss": 4.4389,
+      "step": 11465
+    },
+    {
+      "epoch": 2.249901922322479,
+      "grad_norm": 22.755691528320312,
+      "learning_rate": 4.058166604698384e-06,
+      "loss": 4.614,
+      "step": 11470
+    },
+    {
+      "epoch": 2.2508826990976853,
+      "grad_norm": 28.409160614013672,
+      "learning_rate": 4.054365691864423e-06,
+      "loss": 4.5129,
+      "step": 11475
+    },
+    {
+      "epoch": 2.2508826990976853,
+      "eval_loss": 4.867314338684082,
+      "eval_runtime": 7.6481,
+      "eval_samples_per_second": 27.327,
+      "eval_steps_per_second": 13.729,
+      "step": 11475
+    },
+    {
+      "epoch": 2.2518634758728915,
+      "grad_norm": 17.517236709594727,
+      "learning_rate": 4.050565345678316e-06,
+      "loss": 4.4421,
+      "step": 11480
+    },
+    {
+      "epoch": 2.252844252648097,
+      "grad_norm": 25.891996383666992,
+      "learning_rate": 4.046765568417318e-06,
+      "loss": 4.382,
+      "step": 11485
+    },
+    {
+      "epoch": 2.2538250294233033,
+      "grad_norm": 42.0782470703125,
+      "learning_rate": 4.042966362358358e-06,
+      "loss": 4.7125,
+      "step": 11490
+    },
+    {
+      "epoch": 2.254805806198509,
+      "grad_norm": 19.0367374420166,
+      "learning_rate": 4.039167729778011e-06,
+      "loss": 4.5896,
+      "step": 11495
+    },
+    {
+      "epoch": 2.255786582973715,
+      "grad_norm": 15.167940139770508,
+      "learning_rate": 4.035369672952516e-06,
+      "loss": 4.5573,
+      "step": 11500
+    },
+    {
+      "epoch": 2.2567673597489213,
+      "grad_norm": 24.878446578979492,
+      "learning_rate": 4.031572194157764e-06,
+      "loss": 4.7611,
+      "step": 11505
+    },
+    {
+      "epoch": 2.257748136524127,
+      "grad_norm": 15.153958320617676,
+      "learning_rate": 4.027775295669297e-06,
+      "loss": 4.6099,
+      "step": 11510
+    },
+    {
+      "epoch": 2.258728913299333,
+      "grad_norm": 16.728437423706055,
+      "learning_rate": 4.023978979762316e-06,
+      "loss": 4.5316,
+      "step": 11515
+    },
+    {
+      "epoch": 2.259709690074539,
+      "grad_norm": 21.663965225219727,
+      "learning_rate": 4.0201832487116655e-06,
+      "loss": 4.9828,
+      "step": 11520
+    },
+    {
+      "epoch": 2.260690466849745,
+      "grad_norm": 18.675153732299805,
+      "learning_rate": 4.016388104791843e-06,
+      "loss": 4.2137,
+      "step": 11525
+    },
+    {
+      "epoch": 2.261671243624951,
+      "grad_norm": 15.859734535217285,
+      "learning_rate": 4.0125935502769984e-06,
+      "loss": 4.3856,
+      "step": 11530
+    },
+    {
+      "epoch": 2.262652020400157,
+      "grad_norm": 25.595434188842773,
+      "learning_rate": 4.00879958744092e-06,
+      "loss": 4.3875,
+      "step": 11535
+    },
+    {
+      "epoch": 2.263632797175363,
+      "grad_norm": 21.921201705932617,
+      "learning_rate": 4.005006218557048e-06,
+      "loss": 4.3901,
+      "step": 11540
+    },
+    {
+      "epoch": 2.2646135739505686,
+      "grad_norm": 25.38811683654785,
+      "learning_rate": 4.001213445898462e-06,
+      "loss": 4.8216,
+      "step": 11545
+    },
+    {
+      "epoch": 2.2655943507257748,
+      "grad_norm": 29.609708786010742,
+      "learning_rate": 3.997421271737887e-06,
+      "loss": 4.3386,
+      "step": 11550
+    },
+    {
+      "epoch": 2.266575127500981,
+      "grad_norm": 22.167301177978516,
+      "learning_rate": 3.993629698347693e-06,
+      "loss": 4.552,
+      "step": 11555
+    },
+    {
+      "epoch": 2.2675559042761866,
+      "grad_norm": 24.351184844970703,
+      "learning_rate": 3.989838727999881e-06,
+      "loss": 4.8661,
+      "step": 11560
+    },
+    {
+      "epoch": 2.2685366810513927,
+      "grad_norm": 16.84194564819336,
+      "learning_rate": 3.9860483629661e-06,
+      "loss": 4.5044,
+      "step": 11565
+    },
+    {
+      "epoch": 2.2695174578265984,
+      "grad_norm": 21.152740478515625,
+      "learning_rate": 3.982258605517627e-06,
+      "loss": 4.558,
+      "step": 11570
+    },
+    {
+      "epoch": 2.2704982346018046,
+      "grad_norm": 16.635719299316406,
+      "learning_rate": 3.978469457925385e-06,
+      "loss": 4.4371,
+      "step": 11575
+    },
+    {
+      "epoch": 2.2714790113770107,
+      "grad_norm": 17.723609924316406,
+      "learning_rate": 3.974680922459926e-06,
+      "loss": 4.5154,
+      "step": 11580
+    },
+    {
+      "epoch": 2.2724597881522164,
+      "grad_norm": 16.740421295166016,
+      "learning_rate": 3.970893001391431e-06,
+      "loss": 4.2428,
+      "step": 11585
+    },
+    {
+      "epoch": 2.2734405649274225,
+      "grad_norm": 27.927316665649414,
+      "learning_rate": 3.967105696989723e-06,
+      "loss": 4.3807,
+      "step": 11590
+    },
+    {
+      "epoch": 2.2744213417026287,
+      "grad_norm": 14.875885963439941,
+      "learning_rate": 3.963319011524246e-06,
+      "loss": 4.3996,
+      "step": 11595
+    },
+    {
+      "epoch": 2.2754021184778344,
+      "grad_norm": 14.314786911010742,
+      "learning_rate": 3.959532947264078e-06,
+      "loss": 4.5584,
+      "step": 11600
+    },
+    {
+      "epoch": 2.2763828952530405,
+      "grad_norm": 17.703577041625977,
+      "learning_rate": 3.955747506477927e-06,
+      "loss": 4.4526,
+      "step": 11605
+    },
+    {
+      "epoch": 2.277363672028246,
+      "grad_norm": 51.20561981201172,
+      "learning_rate": 3.951962691434121e-06,
+      "loss": 4.802,
+      "step": 11610
+    },
+    {
+      "epoch": 2.2783444488034523,
+      "grad_norm": 14.392014503479004,
+      "learning_rate": 3.948178504400619e-06,
+      "loss": 4.8058,
+      "step": 11615
+    },
+    {
+      "epoch": 2.2793252255786585,
+      "grad_norm": 14.023991584777832,
+      "learning_rate": 3.944394947644996e-06,
+      "loss": 4.1792,
+      "step": 11620
+    },
+    {
+      "epoch": 2.280306002353864,
+      "grad_norm": 20.808156967163086,
+      "learning_rate": 3.940612023434459e-06,
+      "loss": 4.5473,
+      "step": 11625
+    },
+    {
+      "epoch": 2.2812867791290703,
+      "grad_norm": 42.275081634521484,
+      "learning_rate": 3.936829734035831e-06,
+      "loss": 4.7274,
+      "step": 11630
+    },
+    {
+      "epoch": 2.2822675559042764,
+      "grad_norm": 21.149211883544922,
+      "learning_rate": 3.933048081715553e-06,
+      "loss": 4.6594,
+      "step": 11635
+    },
+    {
+      "epoch": 2.283248332679482,
+      "grad_norm": 29.46340560913086,
+      "learning_rate": 3.929267068739687e-06,
+      "loss": 4.4786,
+      "step": 11640
+    },
+    {
+      "epoch": 2.2842291094546883,
+      "grad_norm": 13.892107009887695,
+      "learning_rate": 3.925486697373911e-06,
+      "loss": 4.5621,
+      "step": 11645
+    },
+    {
+      "epoch": 2.285209886229894,
+      "grad_norm": 15.604551315307617,
+      "learning_rate": 3.9217069698835175e-06,
+      "loss": 4.7846,
+      "step": 11650
+    },
+    {
+      "epoch": 2.2861906630051,
+      "grad_norm": 27.05708885192871,
+      "learning_rate": 3.917927888533418e-06,
+      "loss": 4.2979,
+      "step": 11655
+    },
+    {
+      "epoch": 2.2871714397803062,
+      "grad_norm": 24.699222564697266,
+      "learning_rate": 3.914149455588127e-06,
+      "loss": 4.6958,
+      "step": 11660
+    },
+    {
+      "epoch": 2.288152216555512,
+      "grad_norm": 15.108890533447266,
+      "learning_rate": 3.910371673311783e-06,
+      "loss": 4.522,
+      "step": 11665
+    },
+    {
+      "epoch": 2.289132993330718,
+      "grad_norm": 24.2240047454834,
+      "learning_rate": 3.906594543968122e-06,
+      "loss": 4.2788,
+      "step": 11670
+    },
+    {
+      "epoch": 2.2901137701059238,
+      "grad_norm": 20.101787567138672,
+      "learning_rate": 3.902818069820498e-06,
+      "loss": 4.5444,
+      "step": 11675
+    },
+    {
+      "epoch": 2.29109454688113,
+      "grad_norm": 16.885448455810547,
+      "learning_rate": 3.8990422531318705e-06,
+      "loss": 4.5266,
+      "step": 11680
+    },
+    {
+      "epoch": 2.292075323656336,
+      "grad_norm": 15.544132232666016,
+      "learning_rate": 3.895267096164802e-06,
+      "loss": 4.984,
+      "step": 11685
+    },
+    {
+      "epoch": 2.2930561004315417,
+      "grad_norm": 19.72736167907715,
+      "learning_rate": 3.891492601181462e-06,
+      "loss": 4.5247,
+      "step": 11690
+    },
+    {
+      "epoch": 2.294036877206748,
+      "grad_norm": 37.8563346862793,
+      "learning_rate": 3.887718770443622e-06,
+      "loss": 4.5902,
+      "step": 11695
+    },
+    {
+      "epoch": 2.2950176539819536,
+      "grad_norm": 14.793488502502441,
+      "learning_rate": 3.883945606212655e-06,
+      "loss": 4.7889,
+      "step": 11700
+    },
+    {
+      "epoch": 2.2959984307571597,
+      "grad_norm": 25.496313095092773,
+      "learning_rate": 3.880173110749541e-06,
+      "loss": 4.6955,
+      "step": 11705
+    },
+    {
+      "epoch": 2.296979207532366,
+      "grad_norm": 21.444076538085938,
+      "learning_rate": 3.876401286314848e-06,
+      "loss": 4.4975,
+      "step": 11710
+    },
+    {
+      "epoch": 2.2979599843075715,
+      "grad_norm": 17.857149124145508,
+      "learning_rate": 3.872630135168753e-06,
+      "loss": 4.4218,
+      "step": 11715
+    },
+    {
+      "epoch": 2.2989407610827777,
+      "grad_norm": 19.443084716796875,
+      "learning_rate": 3.868859659571022e-06,
+      "loss": 4.407,
+      "step": 11720
+    },
+    {
+      "epoch": 2.2999215378579834,
+      "grad_norm": 15.299306869506836,
+      "learning_rate": 3.865089861781017e-06,
+      "loss": 4.7506,
+      "step": 11725
+    },
+    {
+      "epoch": 2.3009023146331895,
+      "grad_norm": 13.797977447509766,
+      "learning_rate": 3.861320744057701e-06,
+      "loss": 4.6852,
+      "step": 11730
+    },
+    {
+      "epoch": 2.3018830914083956,
+      "grad_norm": 28.186782836914062,
+      "learning_rate": 3.857552308659618e-06,
+      "loss": 4.7819,
+      "step": 11735
+    },
+    {
+      "epoch": 2.3028638681836013,
+      "grad_norm": 25.146703720092773,
+      "learning_rate": 3.8537845578449146e-06,
+      "loss": 4.6642,
+      "step": 11740
+    },
+    {
+      "epoch": 2.3038446449588075,
+      "grad_norm": 12.95285415649414,
+      "learning_rate": 3.850017493871317e-06,
+      "loss": 4.4853,
+      "step": 11745
+    },
+    {
+      "epoch": 2.304825421734013,
+      "grad_norm": 21.151214599609375,
+      "learning_rate": 3.846251118996148e-06,
+      "loss": 4.5531,
+      "step": 11750
+    },
+    {
+      "epoch": 2.3058061985092193,
+      "grad_norm": 13.681554794311523,
+      "learning_rate": 3.842485435476313e-06,
+      "loss": 4.5772,
+      "step": 11755
+    },
+    {
+      "epoch": 2.3067869752844254,
+      "grad_norm": 14.076757431030273,
+      "learning_rate": 3.838720445568304e-06,
+      "loss": 4.5151,
+      "step": 11760
+    },
+    {
+      "epoch": 2.307767752059631,
+      "grad_norm": 22.38243865966797,
+      "learning_rate": 3.834956151528198e-06,
+      "loss": 4.3797,
+      "step": 11765
+    },
+    {
+      "epoch": 2.3087485288348373,
+      "grad_norm": 17.647878646850586,
+      "learning_rate": 3.831192555611654e-06,
+      "loss": 4.3728,
+      "step": 11770
+    },
+    {
+      "epoch": 2.309729305610043,
+      "grad_norm": 24.30431365966797,
+      "learning_rate": 3.827429660073913e-06,
+      "loss": 4.6663,
+      "step": 11775
+    },
+    {
+      "epoch": 2.310710082385249,
+      "grad_norm": 16.04897689819336,
+      "learning_rate": 3.8236674671698e-06,
+      "loss": 4.3886,
+      "step": 11780
+    },
+    {
+      "epoch": 2.3116908591604552,
+      "grad_norm": 20.717838287353516,
+      "learning_rate": 3.8199059791537105e-06,
+      "loss": 4.2956,
+      "step": 11785
+    },
+    {
+      "epoch": 2.312671635935661,
+      "grad_norm": 50.36362075805664,
+      "learning_rate": 3.816145198279626e-06,
+      "loss": 4.6332,
+      "step": 11790
+    },
+    {
+      "epoch": 2.313652412710867,
+      "grad_norm": 17.40586280822754,
+      "learning_rate": 3.8123851268011006e-06,
+      "loss": 4.6359,
+      "step": 11795
+    },
+    {
+      "epoch": 2.3146331894860728,
+      "grad_norm": 22.39040756225586,
+      "learning_rate": 3.8086257669712617e-06,
+      "loss": 4.4741,
+      "step": 11800
+    },
+    {
+      "epoch": 2.315613966261279,
+      "grad_norm": 18.884647369384766,
+      "learning_rate": 3.8048671210428157e-06,
+      "loss": 4.3912,
+      "step": 11805
+    },
+    {
+      "epoch": 2.316594743036485,
+      "grad_norm": 29.65871810913086,
+      "learning_rate": 3.8011091912680337e-06,
+      "loss": 4.6412,
+      "step": 11810
+    },
+    {
+      "epoch": 2.3175755198116907,
+      "grad_norm": 21.042131423950195,
+      "learning_rate": 3.7973519798987653e-06,
+      "loss": 4.4512,
+      "step": 11815
+    },
+    {
+      "epoch": 2.318556296586897,
+      "grad_norm": 16.669906616210938,
+      "learning_rate": 3.7935954891864222e-06,
+      "loss": 4.654,
+      "step": 11820
+    },
+    {
+      "epoch": 2.3195370733621026,
+      "grad_norm": 29.70143699645996,
+      "learning_rate": 3.7898397213819916e-06,
+      "loss": 4.3917,
+      "step": 11825
+    },
+    {
+      "epoch": 2.3205178501373087,
+      "grad_norm": 23.048263549804688,
+      "learning_rate": 3.786084678736024e-06,
+      "loss": 4.6097,
+      "step": 11830
+    },
+    {
+      "epoch": 2.321498626912515,
+      "grad_norm": 13.745390892028809,
+      "learning_rate": 3.7823303634986313e-06,
+      "loss": 4.4516,
+      "step": 11835
+    },
+    {
+      "epoch": 2.3224794036877205,
+      "grad_norm": 14.595437049865723,
+      "learning_rate": 3.7785767779194984e-06,
+      "loss": 4.469,
+      "step": 11840
+    },
+    {
+      "epoch": 2.3234601804629267,
+      "grad_norm": 30.718673706054688,
+      "learning_rate": 3.774823924247864e-06,
+      "loss": 4.0418,
+      "step": 11845
+    },
+    {
+      "epoch": 2.3244409572381324,
+      "grad_norm": 20.652172088623047,
+      "learning_rate": 3.771071804732534e-06,
+      "loss": 4.3759,
+      "step": 11850
+    },
+    {
+      "epoch": 2.3254217340133385,
+      "grad_norm": 30.810976028442383,
+      "learning_rate": 3.7673204216218757e-06,
+      "loss": 4.4283,
+      "step": 11855
+    },
+    {
+      "epoch": 2.3264025107885447,
+      "grad_norm": 12.208151817321777,
+      "learning_rate": 3.763569777163808e-06,
+      "loss": 4.5245,
+      "step": 11860
+    },
+    {
+      "epoch": 2.3273832875637503,
+      "grad_norm": 33.74302291870117,
+      "learning_rate": 3.759819873605813e-06,
+      "loss": 4.6234,
+      "step": 11865
+    },
+    {
+      "epoch": 2.3283640643389565,
+      "grad_norm": 22.965965270996094,
+      "learning_rate": 3.7560707131949276e-06,
+      "loss": 4.6614,
+      "step": 11870
+    },
+    {
+      "epoch": 2.329344841114162,
+      "grad_norm": 17.791505813598633,
+      "learning_rate": 3.752322298177741e-06,
+      "loss": 4.6324,
+      "step": 11875
+    },
+    {
+      "epoch": 2.3303256178893683,
+      "grad_norm": 22.28071403503418,
+      "learning_rate": 3.7485746308004013e-06,
+      "loss": 4.6046,
+      "step": 11880
+    },
+    {
+      "epoch": 2.3313063946645745,
+      "grad_norm": 14.98755931854248,
+      "learning_rate": 3.744827713308601e-06,
+      "loss": 4.6156,
+      "step": 11885
+    },
+    {
+      "epoch": 2.33228717143978,
+      "grad_norm": 22.762523651123047,
+      "learning_rate": 3.7410815479475903e-06,
+      "loss": 4.4806,
+      "step": 11890
+    },
+    {
+      "epoch": 2.3332679482149863,
+      "grad_norm": 20.753673553466797,
+      "learning_rate": 3.7373361369621638e-06,
+      "loss": 4.3978,
+      "step": 11895
+    },
+    {
+      "epoch": 2.3342487249901924,
+      "grad_norm": 15.564764022827148,
+      "learning_rate": 3.733591482596667e-06,
+      "loss": 4.728,
+      "step": 11900
+    },
+    {
+      "epoch": 2.335229501765398,
+      "grad_norm": 22.621809005737305,
+      "learning_rate": 3.729847587094991e-06,
+      "loss": 4.5043,
+      "step": 11905
+    },
+    {
+      "epoch": 2.3362102785406043,
+      "grad_norm": 21.3169002532959,
+      "learning_rate": 3.72610445270057e-06,
+      "loss": 4.48,
+      "step": 11910
+    },
+    {
+      "epoch": 2.33719105531581,
+      "grad_norm": 10.993232727050781,
+      "learning_rate": 3.7223620816563884e-06,
+      "loss": 4.6117,
+      "step": 11915
+    },
+    {
+      "epoch": 2.338171832091016,
+      "grad_norm": 22.939523696899414,
+      "learning_rate": 3.7186204762049638e-06,
+      "loss": 4.5433,
+      "step": 11920
+    },
+    {
+      "epoch": 2.3391526088662222,
+      "grad_norm": 24.98711585998535,
+      "learning_rate": 3.714879638588363e-06,
+      "loss": 4.6915,
+      "step": 11925
+    },
+    {
+      "epoch": 2.340133385641428,
+      "grad_norm": 25.30660057067871,
+      "learning_rate": 3.7111395710481924e-06,
+      "loss": 4.3815,
+      "step": 11930
+    },
+    {
+      "epoch": 2.341114162416634,
+      "grad_norm": 19.658090591430664,
+      "learning_rate": 3.70740027582559e-06,
+      "loss": 4.4353,
+      "step": 11935
+    },
+    {
+      "epoch": 2.3420949391918398,
+      "grad_norm": 24.44713592529297,
+      "learning_rate": 3.7036617551612387e-06,
+      "loss": 4.267,
+      "step": 11940
+    },
+    {
+      "epoch": 2.343075715967046,
+      "grad_norm": 12.363530158996582,
+      "learning_rate": 3.699924011295352e-06,
+      "loss": 4.7713,
+      "step": 11945
+    },
+    {
+      "epoch": 2.344056492742252,
+      "grad_norm": 21.272781372070312,
+      "learning_rate": 3.6961870464676796e-06,
+      "loss": 4.9373,
+      "step": 11950
+    },
+    {
+      "epoch": 2.3450372695174577,
+      "grad_norm": 26.136646270751953,
+      "learning_rate": 3.6924508629175083e-06,
+      "loss": 4.5627,
+      "step": 11955
+    },
+    {
+      "epoch": 2.346018046292664,
+      "grad_norm": 18.093294143676758,
+      "learning_rate": 3.6887154628836492e-06,
+      "loss": 4.468,
+      "step": 11960
+    },
+    {
+      "epoch": 2.34699882306787,
+      "grad_norm": 17.701435089111328,
+      "learning_rate": 3.6849808486044515e-06,
+      "loss": 4.6783,
+      "step": 11965
+    },
+    {
+      "epoch": 2.3479795998430757,
+      "grad_norm": 16.679710388183594,
+      "learning_rate": 3.6812470223177865e-06,
+      "loss": 4.7766,
+      "step": 11970
+    },
+    {
+      "epoch": 2.348960376618282,
+      "grad_norm": 17.240333557128906,
+      "learning_rate": 3.6775139862610577e-06,
+      "loss": 5.2171,
+      "step": 11975
+    },
+    {
+      "epoch": 2.3499411533934875,
+      "grad_norm": 16.750885009765625,
+      "learning_rate": 3.6737817426711973e-06,
+      "loss": 4.4293,
+      "step": 11980
+    },
+    {
+      "epoch": 2.3509219301686937,
+      "grad_norm": 32.81517028808594,
+      "learning_rate": 3.6700502937846543e-06,
+      "loss": 4.8392,
+      "step": 11985
+    },
+    {
+      "epoch": 2.3519027069439,
+      "grad_norm": 27.170766830444336,
+      "learning_rate": 3.6663196418374114e-06,
+      "loss": 4.4677,
+      "step": 11990
+    },
+    {
+      "epoch": 2.3528834837191055,
+      "grad_norm": 20.98493766784668,
+      "learning_rate": 3.6625897890649653e-06,
+      "loss": 4.6036,
+      "step": 11995
+    },
+    {
+      "epoch": 2.3538642604943116,
+      "grad_norm": 14.241714477539062,
+      "learning_rate": 3.65886073770234e-06,
+      "loss": 4.6737,
+      "step": 12000
+    },
+    {
+      "epoch": 2.3548450372695173,
+      "grad_norm": 13.720659255981445,
+      "learning_rate": 3.655132489984077e-06,
+      "loss": 4.6364,
+      "step": 12005
+    },
+    {
+      "epoch": 2.3558258140447235,
+      "grad_norm": 20.04840850830078,
+      "learning_rate": 3.6514050481442336e-06,
+      "loss": 4.6231,
+      "step": 12010
+    },
+    {
+      "epoch": 2.3568065908199296,
+      "grad_norm": 26.38751220703125,
+      "learning_rate": 3.64767841441639e-06,
+      "loss": 4.3643,
+      "step": 12015
+    },
+    {
+      "epoch": 2.3577873675951353,
+      "grad_norm": 22.324962615966797,
+      "learning_rate": 3.6439525910336347e-06,
+      "loss": 4.1944,
+      "step": 12020
+    },
+    {
+      "epoch": 2.3587681443703414,
+      "grad_norm": 26.925119400024414,
+      "learning_rate": 3.640227580228577e-06,
+      "loss": 4.7769,
+      "step": 12025
+    },
+    {
+      "epoch": 2.359748921145547,
+      "grad_norm": 14.318288803100586,
+      "learning_rate": 3.6365033842333396e-06,
+      "loss": 4.6321,
+      "step": 12030
+    },
+    {
+      "epoch": 2.3607296979207533,
+      "grad_norm": 30.081195831298828,
+      "learning_rate": 3.6327800052795492e-06,
+      "loss": 4.0996,
+      "step": 12035
+    },
+    {
+      "epoch": 2.3617104746959594,
+      "grad_norm": 14.89295482635498,
+      "learning_rate": 3.6290574455983528e-06,
+      "loss": 4.5598,
+      "step": 12040
+    },
+    {
+      "epoch": 2.362691251471165,
+      "grad_norm": 14.272405624389648,
+      "learning_rate": 3.625335707420399e-06,
+      "loss": 4.402,
+      "step": 12045
+    },
+    {
+      "epoch": 2.3636720282463712,
+      "grad_norm": 15.58284854888916,
+      "learning_rate": 3.621614792975846e-06,
+      "loss": 4.3527,
+      "step": 12050
+    },
+    {
+      "epoch": 2.364652805021577,
+      "grad_norm": 13.304563522338867,
+      "learning_rate": 3.6178947044943636e-06,
+      "loss": 4.573,
+      "step": 12055
+    },
+    {
+      "epoch": 2.365633581796783,
+      "grad_norm": 31.662683486938477,
+      "learning_rate": 3.614175444205116e-06,
+      "loss": 4.69,
+      "step": 12060
+    },
+    {
+      "epoch": 2.366614358571989,
+      "grad_norm": 27.504352569580078,
+      "learning_rate": 3.6104570143367847e-06,
+      "loss": 4.3019,
+      "step": 12065
+    },
+    {
+      "epoch": 2.367595135347195,
+      "grad_norm": 18.727642059326172,
+      "learning_rate": 3.6067394171175397e-06,
+      "loss": 4.0567,
+      "step": 12070
+    },
+    {
+      "epoch": 2.368575912122401,
+      "grad_norm": 16.816965103149414,
+      "learning_rate": 3.6030226547750625e-06,
+      "loss": 4.5963,
+      "step": 12075
+    },
+    {
+      "epoch": 2.3695566888976067,
+      "grad_norm": 18.147640228271484,
+      "learning_rate": 3.5993067295365303e-06,
+      "loss": 4.5701,
+      "step": 12080
+    },
+    {
+      "epoch": 2.370537465672813,
+      "grad_norm": 23.622173309326172,
+      "learning_rate": 3.5955916436286177e-06,
+      "loss": 4.6356,
+      "step": 12085
+    },
+    {
+      "epoch": 2.371518242448019,
+      "grad_norm": 17.842937469482422,
+      "learning_rate": 3.5918773992774996e-06,
+      "loss": 4.3194,
+      "step": 12090
+    },
+    {
+      "epoch": 2.3724990192232247,
+      "grad_norm": 16.464218139648438,
+      "learning_rate": 3.588163998708841e-06,
+      "loss": 4.4403,
+      "step": 12095
+    },
+    {
+      "epoch": 2.373479795998431,
+      "grad_norm": 24.46657943725586,
+      "learning_rate": 3.5844514441478075e-06,
+      "loss": 4.4403,
+      "step": 12100
+    },
+    {
+      "epoch": 2.3744605727736365,
+      "grad_norm": 12.043354034423828,
+      "learning_rate": 3.5807397378190558e-06,
+      "loss": 4.0212,
+      "step": 12105
+    },
+    {
+      "epoch": 2.3754413495488427,
+      "grad_norm": 27.728321075439453,
+      "learning_rate": 3.5770288819467307e-06,
+      "loss": 4.785,
+      "step": 12110
+    },
+    {
+      "epoch": 2.376422126324049,
+      "grad_norm": 15.207083702087402,
+      "learning_rate": 3.573318878754475e-06,
+      "loss": 4.392,
+      "step": 12115
+    },
+    {
+      "epoch": 2.3774029030992545,
+      "grad_norm": 15.682321548461914,
+      "learning_rate": 3.5696097304654107e-06,
+      "loss": 4.6056,
+      "step": 12120
+    },
+    {
+      "epoch": 2.3783836798744606,
+      "grad_norm": 21.045578002929688,
+      "learning_rate": 3.5659014393021547e-06,
+      "loss": 4.5539,
+      "step": 12125
+    },
+    {
+      "epoch": 2.3793644566496663,
+      "grad_norm": 19.962541580200195,
+      "learning_rate": 3.5621940074868105e-06,
+      "loss": 4.5095,
+      "step": 12130
+    },
+    {
+      "epoch": 2.3803452334248725,
+      "grad_norm": 22.4492244720459,
+      "learning_rate": 3.5584874372409605e-06,
+      "loss": 4.1886,
+      "step": 12135
+    },
+    {
+      "epoch": 2.3813260102000786,
+      "grad_norm": 35.58995056152344,
+      "learning_rate": 3.5547817307856792e-06,
+      "loss": 4.7669,
+      "step": 12140
+    },
+    {
+      "epoch": 2.3823067869752843,
+      "grad_norm": 17.150493621826172,
+      "learning_rate": 3.551076890341514e-06,
+      "loss": 4.5894,
+      "step": 12145
+    },
+    {
+      "epoch": 2.3832875637504904,
+      "grad_norm": 26.246667861938477,
+      "learning_rate": 3.547372918128503e-06,
+      "loss": 4.6605,
+      "step": 12150
+    },
+    {
+      "epoch": 2.384268340525696,
+      "grad_norm": 21.408432006835938,
+      "learning_rate": 3.5436698163661578e-06,
+      "loss": 4.2601,
+      "step": 12155
+    },
+    {
+      "epoch": 2.3852491173009023,
+      "grad_norm": 11.675660133361816,
+      "learning_rate": 3.5399675872734687e-06,
+      "loss": 4.3653,
+      "step": 12160
+    },
+    {
+      "epoch": 2.3862298940761084,
+      "grad_norm": 23.07163429260254,
+      "learning_rate": 3.5362662330689067e-06,
+      "loss": 4.7729,
+      "step": 12165
+    },
+    {
+      "epoch": 2.387210670851314,
+      "grad_norm": 28.680078506469727,
+      "learning_rate": 3.532565755970413e-06,
+      "loss": 4.302,
+      "step": 12170
+    },
+    {
+      "epoch": 2.3881914476265202,
+      "grad_norm": 36.65128707885742,
+      "learning_rate": 3.5288661581954097e-06,
+      "loss": 4.129,
+      "step": 12175
+    },
+    {
+      "epoch": 2.389172224401726,
+      "grad_norm": 16.943334579467773,
+      "learning_rate": 3.525167441960789e-06,
+      "loss": 4.5921,
+      "step": 12180
+    },
+    {
+      "epoch": 2.390153001176932,
+      "grad_norm": 23.751453399658203,
+      "learning_rate": 3.521469609482913e-06,
+      "loss": 4.9026,
+      "step": 12185
+    },
+    {
+      "epoch": 2.391133777952138,
+      "grad_norm": 17.941225051879883,
+      "learning_rate": 3.5177726629776155e-06,
+      "loss": 4.5361,
+      "step": 12190
+    },
+    {
+      "epoch": 2.392114554727344,
+      "grad_norm": 18.334062576293945,
+      "learning_rate": 3.5140766046602014e-06,
+      "loss": 4.5451,
+      "step": 12195
+    },
+    {
+      "epoch": 2.39309533150255,
+      "grad_norm": 17.336273193359375,
+      "learning_rate": 3.5103814367454397e-06,
+      "loss": 4.3965,
+      "step": 12200
+    },
+    {
+      "epoch": 2.3940761082777557,
+      "grad_norm": 9.473857879638672,
+      "learning_rate": 3.506687161447571e-06,
+      "loss": 4.504,
+      "step": 12205
+    },
+    {
+      "epoch": 2.395056885052962,
+      "grad_norm": 19.98438262939453,
+      "learning_rate": 3.5029937809802946e-06,
+      "loss": 4.5001,
+      "step": 12210
+    },
+    {
+      "epoch": 2.396037661828168,
+      "grad_norm": 22.35348892211914,
+      "learning_rate": 3.49930129755678e-06,
+      "loss": 4.7322,
+      "step": 12215
+    },
+    {
+      "epoch": 2.3970184386033737,
+      "grad_norm": 20.084745407104492,
+      "learning_rate": 3.4956097133896525e-06,
+      "loss": 4.5124,
+      "step": 12220
+    },
+    {
+      "epoch": 2.39799921537858,
+      "grad_norm": 16.486888885498047,
+      "learning_rate": 3.491919030691005e-06,
+      "loss": 4.2832,
+      "step": 12225
+    },
+    {
+      "epoch": 2.398979992153786,
+      "grad_norm": 12.889225006103516,
+      "learning_rate": 3.488229251672388e-06,
+      "loss": 4.6114,
+      "step": 12230
+    },
+    {
+      "epoch": 2.3999607689289917,
+      "grad_norm": 15.054161071777344,
+      "learning_rate": 3.484540378544806e-06,
+      "loss": 4.4341,
+      "step": 12235
+    },
+    {
+      "epoch": 2.400941545704198,
+      "grad_norm": 21.588682174682617,
+      "learning_rate": 3.4808524135187294e-06,
+      "loss": 4.6551,
+      "step": 12240
+    },
+    {
+      "epoch": 2.4019223224794035,
+      "grad_norm": 18.084217071533203,
+      "learning_rate": 3.4771653588040742e-06,
+      "loss": 4.3948,
+      "step": 12245
+    },
+    {
+      "epoch": 2.4029030992546097,
+      "grad_norm": 18.453102111816406,
+      "learning_rate": 3.4734792166102193e-06,
+      "loss": 4.6146,
+      "step": 12250
+    },
+    {
+      "epoch": 2.403883876029816,
+      "grad_norm": 17.33672332763672,
+      "learning_rate": 3.4697939891459958e-06,
+      "loss": 4.5763,
+      "step": 12255
+    },
+    {
+      "epoch": 2.4048646528050215,
+      "grad_norm": 18.220375061035156,
+      "learning_rate": 3.466109678619681e-06,
+      "loss": 4.3382,
+      "step": 12260
+    },
+    {
+      "epoch": 2.4058454295802276,
+      "grad_norm": 23.422691345214844,
+      "learning_rate": 3.4624262872390092e-06,
+      "loss": 4.8424,
+      "step": 12265
+    },
+    {
+      "epoch": 2.4068262063554333,
+      "grad_norm": 36.47475051879883,
+      "learning_rate": 3.458743817211158e-06,
+      "loss": 4.4286,
+      "step": 12270
+    },
+    {
+      "epoch": 2.4078069831306395,
+      "grad_norm": 30.083444595336914,
+      "learning_rate": 3.455062270742757e-06,
+      "loss": 4.6231,
+      "step": 12275
+    },
+    {
+      "epoch": 2.4087877599058456,
+      "grad_norm": 18.711580276489258,
+      "learning_rate": 3.451381650039885e-06,
+      "loss": 4.4978,
+      "step": 12280
+    },
+    {
+      "epoch": 2.4097685366810513,
+      "grad_norm": 16.53005599975586,
+      "learning_rate": 3.4477019573080572e-06,
+      "loss": 4.503,
+      "step": 12285
+    },
+    {
+      "epoch": 2.4107493134562574,
+      "grad_norm": 15.226875305175781,
+      "learning_rate": 3.4440231947522424e-06,
+      "loss": 4.7863,
+      "step": 12290
+    },
+    {
+      "epoch": 2.4117300902314636,
+      "grad_norm": 22.330272674560547,
+      "learning_rate": 3.440345364576845e-06,
+      "loss": 4.4584,
+      "step": 12295
+    },
+    {
+      "epoch": 2.4127108670066693,
+      "grad_norm": 25.432588577270508,
+      "learning_rate": 3.4366684689857118e-06,
+      "loss": 4.6265,
+      "step": 12300
+    },
+    {
+      "epoch": 2.4136916437818754,
+      "grad_norm": 14.317787170410156,
+      "learning_rate": 3.432992510182136e-06,
+      "loss": 4.6704,
+      "step": 12305
+    },
+    {
+      "epoch": 2.414672420557081,
+      "grad_norm": 26.9719295501709,
+      "learning_rate": 3.429317490368839e-06,
+      "loss": 4.5278,
+      "step": 12310
+    },
+    {
+      "epoch": 2.4156531973322872,
+      "grad_norm": 12.308198928833008,
+      "learning_rate": 3.4256434117479897e-06,
+      "loss": 4.3488,
+      "step": 12315
+    },
+    {
+      "epoch": 2.4166339741074934,
+      "grad_norm": 17.44641876220703,
+      "learning_rate": 3.4219702765211846e-06,
+      "loss": 4.1399,
+      "step": 12320
+    },
+    {
+      "epoch": 2.417614750882699,
+      "grad_norm": 16.461467742919922,
+      "learning_rate": 3.418298086889462e-06,
+      "loss": 4.2916,
+      "step": 12325
+    },
+    {
+      "epoch": 2.418595527657905,
+      "grad_norm": 25.63442611694336,
+      "learning_rate": 3.4146268450532883e-06,
+      "loss": 4.6587,
+      "step": 12330
+    },
+    {
+      "epoch": 2.419576304433111,
+      "grad_norm": 25.120079040527344,
+      "learning_rate": 3.4109565532125645e-06,
+      "loss": 4.7741,
+      "step": 12335
+    },
+    {
+      "epoch": 2.420557081208317,
+      "grad_norm": 27.460861206054688,
+      "learning_rate": 3.4072872135666223e-06,
+      "loss": 4.7769,
+      "step": 12340
+    },
+    {
+      "epoch": 2.421537857983523,
+      "grad_norm": 14.591779708862305,
+      "learning_rate": 3.40361882831422e-06,
+      "loss": 4.3034,
+      "step": 12345
+    },
+    {
+      "epoch": 2.422518634758729,
+      "grad_norm": 18.143795013427734,
+      "learning_rate": 3.399951399653547e-06,
+      "loss": 4.4657,
+      "step": 12350
+    },
+    {
+      "epoch": 2.423499411533935,
+      "grad_norm": 26.938961029052734,
+      "learning_rate": 3.3962849297822225e-06,
+      "loss": 4.5643,
+      "step": 12355
+    },
+    {
+      "epoch": 2.4244801883091407,
+      "grad_norm": 21.520898818969727,
+      "learning_rate": 3.392619420897282e-06,
+      "loss": 4.1642,
+      "step": 12360
+    },
+    {
+      "epoch": 2.425460965084347,
+      "grad_norm": 12.339751243591309,
+      "learning_rate": 3.388954875195195e-06,
+      "loss": 4.4566,
+      "step": 12365
+    },
+    {
+      "epoch": 2.426441741859553,
+      "grad_norm": 15.379558563232422,
+      "learning_rate": 3.3852912948718463e-06,
+      "loss": 4.3242,
+      "step": 12370
+    },
+    {
+      "epoch": 2.4274225186347587,
+      "grad_norm": 30.704980850219727,
+      "learning_rate": 3.3816286821225454e-06,
+      "loss": 4.6186,
+      "step": 12375
+    },
+    {
+      "epoch": 2.428403295409965,
+      "grad_norm": 10.835277557373047,
+      "learning_rate": 3.3779670391420255e-06,
+      "loss": 4.6727,
+      "step": 12380
+    },
+    {
+      "epoch": 2.4293840721851705,
+      "grad_norm": 25.137619018554688,
+      "learning_rate": 3.3743063681244302e-06,
+      "loss": 4.697,
+      "step": 12385
+    },
+    {
+      "epoch": 2.4303648489603766,
+      "grad_norm": 12.802051544189453,
+      "learning_rate": 3.3706466712633302e-06,
+      "loss": 4.5622,
+      "step": 12390
+    },
+    {
+      "epoch": 2.4313456257355828,
+      "grad_norm": 48.51518630981445,
+      "learning_rate": 3.3669879507517034e-06,
+      "loss": 4.4323,
+      "step": 12395
+    },
+    {
+      "epoch": 2.4323264025107885,
+      "grad_norm": 20.264951705932617,
+      "learning_rate": 3.3633302087819507e-06,
+      "loss": 4.3534,
+      "step": 12400
+    },
+    {
+      "epoch": 2.4333071792859946,
+      "grad_norm": 39.245704650878906,
+      "learning_rate": 3.3596734475458815e-06,
+      "loss": 4.7553,
+      "step": 12405
+    },
+    {
+      "epoch": 2.4342879560612003,
+      "grad_norm": 10.262246131896973,
+      "learning_rate": 3.3560176692347198e-06,
+      "loss": 4.3705,
+      "step": 12410
+    },
+    {
+      "epoch": 2.4352687328364064,
+      "grad_norm": 20.34832191467285,
+      "learning_rate": 3.3523628760391e-06,
+      "loss": 4.4396,
+      "step": 12415
+    },
+    {
+      "epoch": 2.4362495096116126,
+      "grad_norm": 22.843975067138672,
+      "learning_rate": 3.3487090701490633e-06,
+      "loss": 4.9229,
+      "step": 12420
+    },
+    {
+      "epoch": 2.4372302863868183,
+      "grad_norm": 19.620826721191406,
+      "learning_rate": 3.3450562537540643e-06,
+      "loss": 4.3282,
+      "step": 12425
+    },
+    {
+      "epoch": 2.4382110631620244,
+      "grad_norm": 23.795570373535156,
+      "learning_rate": 3.3414044290429647e-06,
+      "loss": 4.5254,
+      "step": 12430
+    },
+    {
+      "epoch": 2.43919183993723,
+      "grad_norm": 11.415021896362305,
+      "learning_rate": 3.3377535982040245e-06,
+      "loss": 4.442,
+      "step": 12435
+    },
+    {
+      "epoch": 2.4401726167124362,
+      "grad_norm": 25.729127883911133,
+      "learning_rate": 3.3341037634249185e-06,
+      "loss": 4.5446,
+      "step": 12440
+    },
+    {
+      "epoch": 2.4411533934876424,
+      "grad_norm": 25.259185791015625,
+      "learning_rate": 3.3304549268927163e-06,
+      "loss": 4.6079,
+      "step": 12445
+    },
+    {
+      "epoch": 2.442134170262848,
+      "grad_norm": 22.145832061767578,
+      "learning_rate": 3.3268070907938915e-06,
+      "loss": 4.5964,
+      "step": 12450
+    },
+    {
+      "epoch": 2.443114947038054,
+      "grad_norm": 12.732306480407715,
+      "learning_rate": 3.3231602573143233e-06,
+      "loss": 4.651,
+      "step": 12455
+    },
+    {
+      "epoch": 2.44409572381326,
+      "grad_norm": 16.242197036743164,
+      "learning_rate": 3.3195144286392816e-06,
+      "loss": 4.4981,
+      "step": 12460
+    },
+    {
+      "epoch": 2.445076500588466,
+      "grad_norm": 23.256351470947266,
+      "learning_rate": 3.3158696069534423e-06,
+      "loss": 4.6798,
+      "step": 12465
+    },
+    {
+      "epoch": 2.446057277363672,
+      "grad_norm": 11.750959396362305,
+      "learning_rate": 3.312225794440871e-06,
+      "loss": 5.0038,
+      "step": 12470
+    },
+    {
+      "epoch": 2.447038054138878,
+      "grad_norm": 20.7374267578125,
+      "learning_rate": 3.3085829932850342e-06,
+      "loss": 4.5022,
+      "step": 12475
+    },
+    {
+      "epoch": 2.448018830914084,
+      "grad_norm": 14.744890213012695,
+      "learning_rate": 3.3049412056687895e-06,
+      "loss": 4.5182,
+      "step": 12480
+    },
+    {
+      "epoch": 2.4489996076892897,
+      "grad_norm": 30.004308700561523,
+      "learning_rate": 3.3013004337743857e-06,
+      "loss": 4.5898,
+      "step": 12485
+    },
+    {
+      "epoch": 2.449980384464496,
+      "grad_norm": 16.387863159179688,
+      "learning_rate": 3.2976606797834678e-06,
+      "loss": 4.779,
+      "step": 12490
+    },
+    {
+      "epoch": 2.450961161239702,
+      "grad_norm": 26.760541915893555,
+      "learning_rate": 3.294021945877064e-06,
+      "loss": 4.4479,
+      "step": 12495
+    },
+    {
+      "epoch": 2.4519419380149077,
+      "grad_norm": 18.055334091186523,
+      "learning_rate": 3.290384234235598e-06,
+      "loss": 4.6721,
+      "step": 12500
+    },
+    {
+      "epoch": 2.452922714790114,
+      "grad_norm": 13.502447128295898,
+      "learning_rate": 3.2867475470388793e-06,
+      "loss": 4.907,
+      "step": 12505
+    },
+    {
+      "epoch": 2.4539034915653195,
+      "grad_norm": 21.915271759033203,
+      "learning_rate": 3.2831118864660994e-06,
+      "loss": 4.4258,
+      "step": 12510
+    },
+    {
+      "epoch": 2.4548842683405256,
+      "grad_norm": 24.941059112548828,
+      "learning_rate": 3.279477254695839e-06,
+      "loss": 4.923,
+      "step": 12515
+    },
+    {
+      "epoch": 2.455865045115732,
+      "grad_norm": 14.76949691772461,
+      "learning_rate": 3.27584365390606e-06,
+      "loss": 4.5605,
+      "step": 12520
+    },
+    {
+      "epoch": 2.4568458218909375,
+      "grad_norm": 19.59633445739746,
+      "learning_rate": 3.272211086274107e-06,
+      "loss": 4.9415,
+      "step": 12525
+    },
+    {
+      "epoch": 2.4578265986661436,
+      "grad_norm": 16.15921401977539,
+      "learning_rate": 3.2685795539767084e-06,
+      "loss": 4.7336,
+      "step": 12530
+    },
+    {
+      "epoch": 2.4588073754413493,
+      "grad_norm": 30.89405632019043,
+      "learning_rate": 3.264949059189966e-06,
+      "loss": 4.593,
+      "step": 12535
+    },
+    {
+      "epoch": 2.4597881522165554,
+      "grad_norm": 34.44179916381836,
+      "learning_rate": 3.2613196040893675e-06,
+      "loss": 4.5955,
+      "step": 12540
+    },
+    {
+      "epoch": 2.4607689289917616,
+      "grad_norm": 15.018108367919922,
+      "learning_rate": 3.2576911908497695e-06,
+      "loss": 4.575,
+      "step": 12545
+    },
+    {
+      "epoch": 2.4617497057669673,
+      "grad_norm": 14.546805381774902,
+      "learning_rate": 3.2540638216454114e-06,
+      "loss": 4.1292,
+      "step": 12550
+    },
+    {
+      "epoch": 2.4627304825421734,
+      "grad_norm": 18.152332305908203,
+      "learning_rate": 3.2504374986499044e-06,
+      "loss": 4.3065,
+      "step": 12555
+    },
+    {
+      "epoch": 2.4637112593173796,
+      "grad_norm": 17.869962692260742,
+      "learning_rate": 3.2468122240362287e-06,
+      "loss": 4.409,
+      "step": 12560
+    },
+    {
+      "epoch": 2.4646920360925852,
+      "grad_norm": 12.660460472106934,
+      "learning_rate": 3.2431879999767445e-06,
+      "loss": 4.4719,
+      "step": 12565
+    },
+    {
+      "epoch": 2.4656728128677914,
+      "grad_norm": 18.58382225036621,
+      "learning_rate": 3.2395648286431735e-06,
+      "loss": 4.7638,
+      "step": 12570
+    },
+    {
+      "epoch": 2.466653589642997,
+      "grad_norm": 20.497264862060547,
+      "learning_rate": 3.235942712206614e-06,
+      "loss": 4.3147,
+      "step": 12575
+    },
+    {
+      "epoch": 2.467634366418203,
+      "grad_norm": 16.558277130126953,
+      "learning_rate": 3.2323216528375302e-06,
+      "loss": 4.8391,
+      "step": 12580
+    },
+    {
+      "epoch": 2.4686151431934094,
+      "grad_norm": 19.21946144104004,
+      "learning_rate": 3.2287016527057497e-06,
+      "loss": 4.7922,
+      "step": 12585
+    },
+    {
+      "epoch": 2.469595919968615,
+      "grad_norm": 19.4241886138916,
+      "learning_rate": 3.225082713980468e-06,
+      "loss": 4.6347,
+      "step": 12590
+    },
+    {
+      "epoch": 2.470576696743821,
+      "grad_norm": 36.83026885986328,
+      "learning_rate": 3.2214648388302445e-06,
+      "loss": 4.4627,
+      "step": 12595
+    },
+    {
+      "epoch": 2.471557473519027,
+      "grad_norm": 23.589780807495117,
+      "learning_rate": 3.2178480294229998e-06,
+      "loss": 4.4692,
+      "step": 12600
+    },
+    {
+      "epoch": 2.472538250294233,
+      "grad_norm": 21.215566635131836,
+      "learning_rate": 3.21423228792602e-06,
+      "loss": 4.57,
+      "step": 12605
+    },
+    {
+      "epoch": 2.473519027069439,
+      "grad_norm": 16.17399787902832,
+      "learning_rate": 3.2106176165059444e-06,
+      "loss": 4.7717,
+      "step": 12610
+    },
+    {
+      "epoch": 2.474499803844645,
+      "grad_norm": 25.54176139831543,
+      "learning_rate": 3.207004017328779e-06,
+      "loss": 4.5617,
+      "step": 12615
+    },
+    {
+      "epoch": 2.475480580619851,
+      "grad_norm": 12.003211975097656,
+      "learning_rate": 3.2033914925598796e-06,
+      "loss": 4.7012,
+      "step": 12620
+    },
+    {
+      "epoch": 2.476461357395057,
+      "grad_norm": 11.047526359558105,
+      "learning_rate": 3.199780044363963e-06,
+      "loss": 4.2175,
+      "step": 12625
+    },
+    {
+      "epoch": 2.477442134170263,
+      "grad_norm": 21.22567367553711,
+      "learning_rate": 3.196169674905102e-06,
+      "loss": 4.3798,
+      "step": 12630
+    },
+    {
+      "epoch": 2.478422910945469,
+      "grad_norm": 28.6398983001709,
+      "learning_rate": 3.192560386346717e-06,
+      "loss": 4.4773,
+      "step": 12635
+    },
+    {
+      "epoch": 2.4794036877206747,
+      "grad_norm": 16.08147430419922,
+      "learning_rate": 3.1889521808515888e-06,
+      "loss": 4.2833,
+      "step": 12640
+    },
+    {
+      "epoch": 2.480384464495881,
+      "grad_norm": 16.921104431152344,
+      "learning_rate": 3.1853450605818403e-06,
+      "loss": 4.4575,
+      "step": 12645
+    },
+    {
+      "epoch": 2.481365241271087,
+      "grad_norm": 27.06111717224121,
+      "learning_rate": 3.1817390276989514e-06,
+      "loss": 4.558,
+      "step": 12650
+    },
+    {
+      "epoch": 2.4823460180462926,
+      "grad_norm": 16.07041358947754,
+      "learning_rate": 3.178134084363747e-06,
+      "loss": 4.4276,
+      "step": 12655
+    },
+    {
+      "epoch": 2.4833267948214988,
+      "grad_norm": 17.705095291137695,
+      "learning_rate": 3.1745302327364e-06,
+      "loss": 4.2914,
+      "step": 12660
+    },
+    {
+      "epoch": 2.4843075715967045,
+      "grad_norm": 20.390050888061523,
+      "learning_rate": 3.1709274749764294e-06,
+      "loss": 4.6041,
+      "step": 12665
+    },
+    {
+      "epoch": 2.4852883483719106,
+      "grad_norm": 18.144638061523438,
+      "learning_rate": 3.1673258132426958e-06,
+      "loss": 4.6561,
+      "step": 12670
+    },
+    {
+      "epoch": 2.4862691251471167,
+      "grad_norm": 26.244768142700195,
+      "learning_rate": 3.1637252496934073e-06,
+      "loss": 4.0215,
+      "step": 12675
+    },
+    {
+      "epoch": 2.4872499019223224,
+      "grad_norm": 34.454917907714844,
+      "learning_rate": 3.160125786486114e-06,
+      "loss": 4.517,
+      "step": 12680
+    },
+    {
+      "epoch": 2.4882306786975286,
+      "grad_norm": 24.92203140258789,
+      "learning_rate": 3.1565274257777e-06,
+      "loss": 4.5899,
+      "step": 12685
+    },
+    {
+      "epoch": 2.4892114554727343,
+      "grad_norm": 22.43060874938965,
+      "learning_rate": 3.152930169724399e-06,
+      "loss": 4.267,
+      "step": 12690
+    },
+    {
+      "epoch": 2.4901922322479404,
+      "grad_norm": 18.194143295288086,
+      "learning_rate": 3.1493340204817735e-06,
+      "loss": 4.4478,
+      "step": 12695
+    },
+    {
+      "epoch": 2.4911730090231465,
+      "grad_norm": 22.45499038696289,
+      "learning_rate": 3.145738980204726e-06,
+      "loss": 4.4645,
+      "step": 12700
+    },
+    {
+      "epoch": 2.4921537857983522,
+      "grad_norm": 21.602619171142578,
+      "learning_rate": 3.1421450510474986e-06,
+      "loss": 4.532,
+      "step": 12705
+    },
+    {
+      "epoch": 2.4931345625735584,
+      "grad_norm": 21.310916900634766,
+      "learning_rate": 3.13855223516366e-06,
+      "loss": 4.6983,
+      "step": 12710
+    },
+    {
+      "epoch": 2.494115339348764,
+      "grad_norm": 21.922300338745117,
+      "learning_rate": 3.1349605347061195e-06,
+      "loss": 4.5509,
+      "step": 12715
+    },
+    {
+      "epoch": 2.49509611612397,
+      "grad_norm": 12.221722602844238,
+      "learning_rate": 3.1313699518271113e-06,
+      "loss": 4.5321,
+      "step": 12720
+    },
+    {
+      "epoch": 2.4960768928991763,
+      "grad_norm": 28.056127548217773,
+      "learning_rate": 3.1277804886782043e-06,
+      "loss": 4.4323,
+      "step": 12725
+    },
+    {
+      "epoch": 2.497057669674382,
+      "grad_norm": 13.224735260009766,
+      "learning_rate": 3.1241921474102952e-06,
+      "loss": 4.2532,
+      "step": 12730
+    },
+    {
+      "epoch": 2.498038446449588,
+      "grad_norm": 22.16506004333496,
+      "learning_rate": 3.120604930173608e-06,
+      "loss": 4.454,
+      "step": 12735
+    },
+    {
+      "epoch": 2.499019223224794,
+      "grad_norm": 18.699708938598633,
+      "learning_rate": 3.1170188391176946e-06,
+      "loss": 4.7483,
+      "step": 12740
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.912882804870605,
+      "learning_rate": 3.1134338763914272e-06,
+      "loss": 4.3887,
+      "step": 12745
+    },
+    {
+      "epoch": 2.500980776775206,
+      "grad_norm": 16.283729553222656,
+      "learning_rate": 3.1098500441430085e-06,
+      "loss": 4.3017,
+      "step": 12750
+    },
+    {
+      "epoch": 2.500980776775206,
+      "eval_loss": 4.861581802368164,
+      "eval_runtime": 7.7147,
+      "eval_samples_per_second": 27.091,
+      "eval_steps_per_second": 13.61,
+      "step": 12750
+    },
+    {
+      "epoch": 2.501961553550412,
+      "grad_norm": 19.177406311035156,
+      "learning_rate": 3.1062673445199625e-06,
+      "loss": 4.4072,
+      "step": 12755
+    },
+    {
+      "epoch": 2.502942330325618,
+      "grad_norm": 19.394792556762695,
+      "learning_rate": 3.102685779669129e-06,
+      "loss": 4.3576,
+      "step": 12760
+    },
+    {
+      "epoch": 2.5039231071008237,
+      "grad_norm": 47.282066345214844,
+      "learning_rate": 3.0991053517366753e-06,
+      "loss": 4.5277,
+      "step": 12765
+    },
+    {
+      "epoch": 2.50490388387603,
+      "grad_norm": 11.863676071166992,
+      "learning_rate": 3.095526062868082e-06,
+      "loss": 4.4062,
+      "step": 12770
+    },
+    {
+      "epoch": 2.505884660651236,
+      "grad_norm": 13.989859580993652,
+      "learning_rate": 3.0919479152081468e-06,
+      "loss": 4.7194,
+      "step": 12775
+    },
+    {
+      "epoch": 2.5068654374264416,
+      "grad_norm": 14.908836364746094,
+      "learning_rate": 3.0883709109009907e-06,
+      "loss": 4.6656,
+      "step": 12780
+    },
+    {
+      "epoch": 2.5078462142016478,
+      "grad_norm": 16.357318878173828,
+      "learning_rate": 3.08479505209004e-06,
+      "loss": 4.6039,
+      "step": 12785
+    },
+    {
+      "epoch": 2.5088269909768535,
+      "grad_norm": 19.404542922973633,
+      "learning_rate": 3.081220340918043e-06,
+      "loss": 4.3132,
+      "step": 12790
+    },
+    {
+      "epoch": 2.5098077677520596,
+      "grad_norm": 20.268781661987305,
+      "learning_rate": 3.0776467795270526e-06,
+      "loss": 4.6035,
+      "step": 12795
+    },
+    {
+      "epoch": 2.5107885445272657,
+      "grad_norm": 15.913606643676758,
+      "learning_rate": 3.0740743700584397e-06,
+      "loss": 4.4568,
+      "step": 12800
+    },
+    {
+      "epoch": 2.5117693213024714,
+      "grad_norm": 14.354854583740234,
+      "learning_rate": 3.0705031146528817e-06,
+      "loss": 4.7239,
+      "step": 12805
+    },
+    {
+      "epoch": 2.5127500980776776,
+      "grad_norm": 12.96605396270752,
+      "learning_rate": 3.0669330154503617e-06,
+      "loss": 4.8816,
+      "step": 12810
+    },
+    {
+      "epoch": 2.5137308748528833,
+      "grad_norm": 13.831136703491211,
+      "learning_rate": 3.063364074590177e-06,
+      "loss": 4.5382,
+      "step": 12815
+    },
+    {
+      "epoch": 2.5147116516280894,
+      "grad_norm": 17.773210525512695,
+      "learning_rate": 3.059796294210923e-06,
+      "loss": 4.4677,
+      "step": 12820
+    },
+    {
+      "epoch": 2.5156924284032955,
+      "grad_norm": 14.534049987792969,
+      "learning_rate": 3.056229676450504e-06,
+      "loss": 4.4686,
+      "step": 12825
+    },
+    {
+      "epoch": 2.5166732051785012,
+      "grad_norm": 21.867084503173828,
+      "learning_rate": 3.0526642234461313e-06,
+      "loss": 4.391,
+      "step": 12830
+    },
+    {
+      "epoch": 2.5176539819537074,
+      "grad_norm": 23.599096298217773,
+      "learning_rate": 3.049099937334309e-06,
+      "loss": 4.4712,
+      "step": 12835
+    },
+    {
+      "epoch": 2.518634758728913,
+      "grad_norm": 21.118675231933594,
+      "learning_rate": 3.0455368202508484e-06,
+      "loss": 4.2714,
+      "step": 12840
+    },
+    {
+      "epoch": 2.519615535504119,
+      "grad_norm": 10.092329978942871,
+      "learning_rate": 3.0419748743308595e-06,
+      "loss": 3.9771,
+      "step": 12845
+    },
+    {
+      "epoch": 2.5205963122793253,
+      "grad_norm": 28.692956924438477,
+      "learning_rate": 3.0384141017087483e-06,
+      "loss": 4.6071,
+      "step": 12850
+    },
+    {
+      "epoch": 2.521577089054531,
+      "grad_norm": 20.138452529907227,
+      "learning_rate": 3.034854504518222e-06,
+      "loss": 4.7027,
+      "step": 12855
+    },
+    {
+      "epoch": 2.522557865829737,
+      "grad_norm": 15.83338451385498,
+      "learning_rate": 3.031296084892278e-06,
+      "loss": 4.4956,
+      "step": 12860
+    },
+    {
+      "epoch": 2.523538642604943,
+      "grad_norm": 11.186666488647461,
+      "learning_rate": 3.027738844963213e-06,
+      "loss": 4.7868,
+      "step": 12865
+    },
+    {
+      "epoch": 2.524519419380149,
+      "grad_norm": 13.635848999023438,
+      "learning_rate": 3.024182786862612e-06,
+      "loss": 4.3454,
+      "step": 12870
+    },
+    {
+      "epoch": 2.525500196155355,
+      "grad_norm": 16.45800018310547,
+      "learning_rate": 3.0206279127213565e-06,
+      "loss": 4.4047,
+      "step": 12875
+    },
+    {
+      "epoch": 2.526480972930561,
+      "grad_norm": 35.179325103759766,
+      "learning_rate": 3.017074224669617e-06,
+      "loss": 4.5525,
+      "step": 12880
+    },
+    {
+      "epoch": 2.527461749705767,
+      "grad_norm": 19.517745971679688,
+      "learning_rate": 3.01352172483685e-06,
+      "loss": 4.5151,
+      "step": 12885
+    },
+    {
+      "epoch": 2.5284425264809727,
+      "grad_norm": 32.89816665649414,
+      "learning_rate": 3.0099704153518057e-06,
+      "loss": 4.5045,
+      "step": 12890
+    },
+    {
+      "epoch": 2.529423303256179,
+      "grad_norm": 24.201087951660156,
+      "learning_rate": 3.006420298342515e-06,
+      "loss": 4.6896,
+      "step": 12895
+    },
+    {
+      "epoch": 2.530404080031385,
+      "grad_norm": 19.274837493896484,
+      "learning_rate": 3.002871375936298e-06,
+      "loss": 4.5481,
+      "step": 12900
+    },
+    {
+      "epoch": 2.531384856806591,
+      "grad_norm": 26.81613540649414,
+      "learning_rate": 2.9993236502597624e-06,
+      "loss": 5.2386,
+      "step": 12905
+    },
+    {
+      "epoch": 2.532365633581797,
+      "grad_norm": 12.64597225189209,
+      "learning_rate": 2.99577712343879e-06,
+      "loss": 4.5118,
+      "step": 12910
+    },
+    {
+      "epoch": 2.5333464103570025,
+      "grad_norm": 20.370603561401367,
+      "learning_rate": 2.9922317975985494e-06,
+      "loss": 4.4,
+      "step": 12915
+    },
+    {
+      "epoch": 2.5343271871322086,
+      "grad_norm": 48.84408950805664,
+      "learning_rate": 2.98868767486349e-06,
+      "loss": 4.903,
+      "step": 12920
+    },
+    {
+      "epoch": 2.5353079639074148,
+      "grad_norm": 18.01617431640625,
+      "learning_rate": 2.9851447573573383e-06,
+      "loss": 4.5406,
+      "step": 12925
+    },
+    {
+      "epoch": 2.536288740682621,
+      "grad_norm": 13.745526313781738,
+      "learning_rate": 2.981603047203102e-06,
+      "loss": 4.7924,
+      "step": 12930
+    },
+    {
+      "epoch": 2.5372695174578266,
+      "grad_norm": 19.135507583618164,
+      "learning_rate": 2.9780625465230583e-06,
+      "loss": 4.5221,
+      "step": 12935
+    },
+    {
+      "epoch": 2.5382502942330327,
+      "grad_norm": 13.795884132385254,
+      "learning_rate": 2.97452325743877e-06,
+      "loss": 4.5289,
+      "step": 12940
+    },
+    {
+      "epoch": 2.5392310710082384,
+      "grad_norm": 21.768945693969727,
+      "learning_rate": 2.970985182071063e-06,
+      "loss": 4.381,
+      "step": 12945
+    },
+    {
+      "epoch": 2.5402118477834446,
+      "grad_norm": 22.04071807861328,
+      "learning_rate": 2.9674483225400436e-06,
+      "loss": 4.3116,
+      "step": 12950
+    },
+    {
+      "epoch": 2.5411926245586507,
+      "grad_norm": 24.65144920349121,
+      "learning_rate": 2.9639126809650877e-06,
+      "loss": 4.9135,
+      "step": 12955
+    },
+    {
+      "epoch": 2.5421734013338564,
+      "grad_norm": 15.5100736618042,
+      "learning_rate": 2.9603782594648365e-06,
+      "loss": 4.3342,
+      "step": 12960
+    },
+    {
+      "epoch": 2.5431541781090625,
+      "grad_norm": 20.800643920898438,
+      "learning_rate": 2.9568450601572095e-06,
+      "loss": 4.7446,
+      "step": 12965
+    },
+    {
+      "epoch": 2.544134954884268,
+      "grad_norm": 32.53725814819336,
+      "learning_rate": 2.9533130851593846e-06,
+      "loss": 4.3541,
+      "step": 12970
+    },
+    {
+      "epoch": 2.5451157316594744,
+      "grad_norm": 33.92058181762695,
+      "learning_rate": 2.949782336587812e-06,
+      "loss": 4.3456,
+      "step": 12975
+    },
+    {
+      "epoch": 2.5460965084346805,
+      "grad_norm": 29.062822341918945,
+      "learning_rate": 2.946252816558205e-06,
+      "loss": 4.6352,
+      "step": 12980
+    },
+    {
+      "epoch": 2.547077285209886,
+      "grad_norm": 14.796578407287598,
+      "learning_rate": 2.942724527185539e-06,
+      "loss": 4.263,
+      "step": 12985
+    },
+    {
+      "epoch": 2.5480580619850923,
+      "grad_norm": 17.256603240966797,
+      "learning_rate": 2.939197470584057e-06,
+      "loss": 4.3981,
+      "step": 12990
+    },
+    {
+      "epoch": 2.549038838760298,
+      "grad_norm": 13.124008178710938,
+      "learning_rate": 2.9356716488672556e-06,
+      "loss": 4.4093,
+      "step": 12995
+    },
+    {
+      "epoch": 2.550019615535504,
+      "grad_norm": 19.26356315612793,
+      "learning_rate": 2.9321470641478978e-06,
+      "loss": 4.4876,
+      "step": 13000
+    },
+    {
+      "epoch": 2.5510003923107103,
+      "grad_norm": 9.095273971557617,
+      "learning_rate": 2.928623718538006e-06,
+      "loss": 4.4389,
+      "step": 13005
+    },
+    {
+      "epoch": 2.551981169085916,
+      "grad_norm": 29.93180274963379,
+      "learning_rate": 2.9251016141488532e-06,
+      "loss": 4.2358,
+      "step": 13010
+    },
+    {
+      "epoch": 2.552961945861122,
+      "grad_norm": 12.834431648254395,
+      "learning_rate": 2.921580753090977e-06,
+      "loss": 4.3116,
+      "step": 13015
+    },
+    {
+      "epoch": 2.553942722636328,
+      "grad_norm": 27.98518943786621,
+      "learning_rate": 2.9180611374741623e-06,
+      "loss": 4.6901,
+      "step": 13020
+    },
+    {
+      "epoch": 2.554923499411534,
+      "grad_norm": 15.999112129211426,
+      "learning_rate": 2.914542769407452e-06,
+      "loss": 4.4974,
+      "step": 13025
+    },
+    {
+      "epoch": 2.55590427618674,
+      "grad_norm": 13.242138862609863,
+      "learning_rate": 2.911025650999143e-06,
+      "loss": 4.2905,
+      "step": 13030
+    },
+    {
+      "epoch": 2.556885052961946,
+      "grad_norm": 17.243799209594727,
+      "learning_rate": 2.9075097843567775e-06,
+      "loss": 4.5918,
+      "step": 13035
+    },
+    {
+      "epoch": 2.557865829737152,
+      "grad_norm": 13.571252822875977,
+      "learning_rate": 2.903995171587155e-06,
+      "loss": 4.2669,
+      "step": 13040
+    },
+    {
+      "epoch": 2.5588466065123576,
+      "grad_norm": 47.06708908081055,
+      "learning_rate": 2.900481814796316e-06,
+      "loss": 5.0637,
+      "step": 13045
+    },
+    {
+      "epoch": 2.5598273832875638,
+      "grad_norm": 15.13114070892334,
+      "learning_rate": 2.8969697160895545e-06,
+      "loss": 4.229,
+      "step": 13050
+    },
+    {
+      "epoch": 2.56080816006277,
+      "grad_norm": 21.789676666259766,
+      "learning_rate": 2.893458877571409e-06,
+      "loss": 4.2882,
+      "step": 13055
+    },
+    {
+      "epoch": 2.5617889368379756,
+      "grad_norm": 22.40684700012207,
+      "learning_rate": 2.8899493013456602e-06,
+      "loss": 4.625,
+      "step": 13060
+    },
+    {
+      "epoch": 2.5627697136131817,
+      "grad_norm": 33.69407272338867,
+      "learning_rate": 2.8864409895153365e-06,
+      "loss": 4.523,
+      "step": 13065
+    },
+    {
+      "epoch": 2.5637504903883874,
+      "grad_norm": 18.701000213623047,
+      "learning_rate": 2.8829339441827044e-06,
+      "loss": 4.5509,
+      "step": 13070
+    },
+    {
+      "epoch": 2.5647312671635936,
+      "grad_norm": 14.72784423828125,
+      "learning_rate": 2.879428167449276e-06,
+      "loss": 4.3159,
+      "step": 13075
+    },
+    {
+      "epoch": 2.5657120439387997,
+      "grad_norm": 22.692087173461914,
+      "learning_rate": 2.875923661415799e-06,
+      "loss": 4.5071,
+      "step": 13080
+    },
+    {
+      "epoch": 2.5666928207140054,
+      "grad_norm": 15.203378677368164,
+      "learning_rate": 2.872420428182261e-06,
+      "loss": 4.4571,
+      "step": 13085
+    },
+    {
+      "epoch": 2.5676735974892115,
+      "grad_norm": 24.574600219726562,
+      "learning_rate": 2.86891846984789e-06,
+      "loss": 4.4902,
+      "step": 13090
+    },
+    {
+      "epoch": 2.5686543742644172,
+      "grad_norm": 19.6531982421875,
+      "learning_rate": 2.8654177885111444e-06,
+      "loss": 4.5339,
+      "step": 13095
+    },
+    {
+      "epoch": 2.5696351510396234,
+      "grad_norm": 12.66419792175293,
+      "learning_rate": 2.861918386269721e-06,
+      "loss": 4.267,
+      "step": 13100
+    },
+    {
+      "epoch": 2.5706159278148295,
+      "grad_norm": 20.505992889404297,
+      "learning_rate": 2.8584202652205536e-06,
+      "loss": 4.4909,
+      "step": 13105
+    },
+    {
+      "epoch": 2.571596704590035,
+      "grad_norm": 30.728412628173828,
+      "learning_rate": 2.8549234274597982e-06,
+      "loss": 4.5643,
+      "step": 13110
+    },
+    {
+      "epoch": 2.5725774813652413,
+      "grad_norm": 14.343387603759766,
+      "learning_rate": 2.8514278750828537e-06,
+      "loss": 4.5294,
+      "step": 13115
+    },
+    {
+      "epoch": 2.573558258140447,
+      "grad_norm": 26.706621170043945,
+      "learning_rate": 2.847933610184338e-06,
+      "loss": 4.8221,
+      "step": 13120
+    },
+    {
+      "epoch": 2.574539034915653,
+      "grad_norm": 13.684526443481445,
+      "learning_rate": 2.8444406348581046e-06,
+      "loss": 4.7329,
+      "step": 13125
+    },
+    {
+      "epoch": 2.5755198116908593,
+      "grad_norm": 22.70045280456543,
+      "learning_rate": 2.840948951197234e-06,
+      "loss": 4.6044,
+      "step": 13130
+    },
+    {
+      "epoch": 2.576500588466065,
+      "grad_norm": 10.388460159301758,
+      "learning_rate": 2.8374585612940274e-06,
+      "loss": 4.5919,
+      "step": 13135
+    },
+    {
+      "epoch": 2.577481365241271,
+      "grad_norm": 34.217323303222656,
+      "learning_rate": 2.8339694672400176e-06,
+      "loss": 4.526,
+      "step": 13140
+    },
+    {
+      "epoch": 2.578462142016477,
+      "grad_norm": 19.96867561340332,
+      "learning_rate": 2.8304816711259554e-06,
+      "loss": 4.416,
+      "step": 13145
+    },
+    {
+      "epoch": 2.579442918791683,
+      "grad_norm": 11.595075607299805,
+      "learning_rate": 2.8269951750418144e-06,
+      "loss": 4.4408,
+      "step": 13150
+    },
+    {
+      "epoch": 2.580423695566889,
+      "grad_norm": 17.77700424194336,
+      "learning_rate": 2.823509981076795e-06,
+      "loss": 4.4397,
+      "step": 13155
+    },
+    {
+      "epoch": 2.581404472342095,
+      "grad_norm": 12.028759002685547,
+      "learning_rate": 2.8200260913193077e-06,
+      "loss": 4.7812,
+      "step": 13160
+    },
+    {
+      "epoch": 2.582385249117301,
+      "grad_norm": 22.082979202270508,
+      "learning_rate": 2.816543507856992e-06,
+      "loss": 4.3825,
+      "step": 13165
+    },
+    {
+      "epoch": 2.5833660258925066,
+      "grad_norm": 34.65692901611328,
+      "learning_rate": 2.813062232776695e-06,
+      "loss": 4.3139,
+      "step": 13170
+    },
+    {
+      "epoch": 2.5843468026677128,
+      "grad_norm": 22.31768798828125,
+      "learning_rate": 2.8095822681644864e-06,
+      "loss": 4.5857,
+      "step": 13175
+    },
+    {
+      "epoch": 2.585327579442919,
+      "grad_norm": 22.302383422851562,
+      "learning_rate": 2.8061036161056505e-06,
+      "loss": 4.7143,
+      "step": 13180
+    },
+    {
+      "epoch": 2.5863083562181246,
+      "grad_norm": 20.870906829833984,
+      "learning_rate": 2.802626278684679e-06,
+      "loss": 4.3662,
+      "step": 13185
+    },
+    {
+      "epoch": 2.5872891329933307,
+      "grad_norm": 14.885607719421387,
+      "learning_rate": 2.7991502579852837e-06,
+      "loss": 4.6355,
+      "step": 13190
+    },
+    {
+      "epoch": 2.5882699097685364,
+      "grad_norm": 13.58197021484375,
+      "learning_rate": 2.7956755560903797e-06,
+      "loss": 4.3287,
+      "step": 13195
+    },
+    {
+      "epoch": 2.5892506865437426,
+      "grad_norm": 15.81235122680664,
+      "learning_rate": 2.7922021750820983e-06,
+      "loss": 4.5046,
+      "step": 13200
+    },
+    {
+      "epoch": 2.5902314633189487,
+      "grad_norm": 16.68513298034668,
+      "learning_rate": 2.788730117041778e-06,
+      "loss": 4.5428,
+      "step": 13205
+    },
+    {
+      "epoch": 2.5912122400941544,
+      "grad_norm": 14.253608703613281,
+      "learning_rate": 2.785259384049959e-06,
+      "loss": 4.8577,
+      "step": 13210
+    },
+    {
+      "epoch": 2.5921930168693605,
+      "grad_norm": 25.999387741088867,
+      "learning_rate": 2.7817899781863964e-06,
+      "loss": 4.498,
+      "step": 13215
+    },
+    {
+      "epoch": 2.5931737936445662,
+      "grad_norm": 18.823204040527344,
+      "learning_rate": 2.7783219015300443e-06,
+      "loss": 4.7903,
+      "step": 13220
+    },
+    {
+      "epoch": 2.5941545704197724,
+      "grad_norm": 20.179967880249023,
+      "learning_rate": 2.7748551561590574e-06,
+      "loss": 4.5352,
+      "step": 13225
+    },
+    {
+      "epoch": 2.5951353471949785,
+      "grad_norm": 15.403043746948242,
+      "learning_rate": 2.771389744150802e-06,
+      "loss": 4.6202,
+      "step": 13230
+    },
+    {
+      "epoch": 2.5961161239701847,
+      "grad_norm": 23.73948097229004,
+      "learning_rate": 2.7679256675818357e-06,
+      "loss": 4.9047,
+      "step": 13235
+    },
+    {
+      "epoch": 2.5970969007453903,
+      "grad_norm": 16.40243148803711,
+      "learning_rate": 2.764462928527924e-06,
+      "loss": 4.384,
+      "step": 13240
+    },
+    {
+      "epoch": 2.598077677520596,
+      "grad_norm": 21.48336410522461,
+      "learning_rate": 2.7610015290640237e-06,
+      "loss": 4.3126,
+      "step": 13245
+    },
+    {
+      "epoch": 2.599058454295802,
+      "grad_norm": 26.895301818847656,
+      "learning_rate": 2.7575414712642947e-06,
+      "loss": 4.5225,
+      "step": 13250
+    },
+    {
+      "epoch": 2.6000392310710083,
+      "grad_norm": 28.130233764648438,
+      "learning_rate": 2.754082757202091e-06,
+      "loss": 4.7777,
+      "step": 13255
+    },
+    {
+      "epoch": 2.6010200078462145,
+      "grad_norm": 15.378296852111816,
+      "learning_rate": 2.750625388949959e-06,
+      "loss": 4.1805,
+      "step": 13260
+    },
+    {
+      "epoch": 2.60200078462142,
+      "grad_norm": 23.175745010375977,
+      "learning_rate": 2.7471693685796437e-06,
+      "loss": 4.5967,
+      "step": 13265
+    },
+    {
+      "epoch": 2.6029815613966263,
+      "grad_norm": 17.780309677124023,
+      "learning_rate": 2.7437146981620754e-06,
+      "loss": 4.3283,
+      "step": 13270
+    },
+    {
+      "epoch": 2.603962338171832,
+      "grad_norm": 27.35478401184082,
+      "learning_rate": 2.740261379767382e-06,
+      "loss": 4.6849,
+      "step": 13275
+    },
+    {
+      "epoch": 2.604943114947038,
+      "grad_norm": 15.305866241455078,
+      "learning_rate": 2.7368094154648794e-06,
+      "loss": 4.4812,
+      "step": 13280
+    },
+    {
+      "epoch": 2.6059238917222443,
+      "grad_norm": 20.15849494934082,
+      "learning_rate": 2.7333588073230682e-06,
+      "loss": 4.6729,
+      "step": 13285
+    },
+    {
+      "epoch": 2.60690466849745,
+      "grad_norm": 23.166080474853516,
+      "learning_rate": 2.7299095574096435e-06,
+      "loss": 4.5444,
+      "step": 13290
+    },
+    {
+      "epoch": 2.607885445272656,
+      "grad_norm": 13.613513946533203,
+      "learning_rate": 2.726461667791481e-06,
+      "loss": 4.9226,
+      "step": 13295
+    },
+    {
+      "epoch": 2.608866222047862,
+      "grad_norm": 14.808908462524414,
+      "learning_rate": 2.7230151405346407e-06,
+      "loss": 4.4457,
+      "step": 13300
+    },
+    {
+      "epoch": 2.609846998823068,
+      "grad_norm": 25.118194580078125,
+      "learning_rate": 2.7195699777043723e-06,
+      "loss": 4.2229,
+      "step": 13305
+    },
+    {
+      "epoch": 2.610827775598274,
+      "grad_norm": 26.882368087768555,
+      "learning_rate": 2.7161261813650997e-06,
+      "loss": 4.6098,
+      "step": 13310
+    },
+    {
+      "epoch": 2.6118085523734798,
+      "grad_norm": 26.372621536254883,
+      "learning_rate": 2.7126837535804362e-06,
+      "loss": 4.6734,
+      "step": 13315
+    },
+    {
+      "epoch": 2.612789329148686,
+      "grad_norm": 31.41575050354004,
+      "learning_rate": 2.7092426964131667e-06,
+      "loss": 4.6502,
+      "step": 13320
+    },
+    {
+      "epoch": 2.6137701059238916,
+      "grad_norm": 43.63661193847656,
+      "learning_rate": 2.705803011925262e-06,
+      "loss": 4.8477,
+      "step": 13325
+    },
+    {
+      "epoch": 2.6147508826990977,
+      "grad_norm": 37.062068939208984,
+      "learning_rate": 2.70236470217787e-06,
+      "loss": 4.5201,
+      "step": 13330
+    },
+    {
+      "epoch": 2.615731659474304,
+      "grad_norm": 13.730938911437988,
+      "learning_rate": 2.6989277692313064e-06,
+      "loss": 4.6581,
+      "step": 13335
+    },
+    {
+      "epoch": 2.6167124362495096,
+      "grad_norm": 19.538053512573242,
+      "learning_rate": 2.6954922151450735e-06,
+      "loss": 4.5461,
+      "step": 13340
+    },
+    {
+      "epoch": 2.6176932130247157,
+      "grad_norm": 18.18435287475586,
+      "learning_rate": 2.6920580419778375e-06,
+      "loss": 4.6776,
+      "step": 13345
+    },
+    {
+      "epoch": 2.6186739897999214,
+      "grad_norm": 14.151314735412598,
+      "learning_rate": 2.6886252517874423e-06,
+      "loss": 4.7098,
+      "step": 13350
+    },
+    {
+      "epoch": 2.6196547665751275,
+      "grad_norm": 16.580251693725586,
+      "learning_rate": 2.6851938466309053e-06,
+      "loss": 4.7572,
+      "step": 13355
+    },
+    {
+      "epoch": 2.6206355433503337,
+      "grad_norm": 18.998332977294922,
+      "learning_rate": 2.6817638285644077e-06,
+      "loss": 4.3379,
+      "step": 13360
+    },
+    {
+      "epoch": 2.6216163201255394,
+      "grad_norm": 21.826854705810547,
+      "learning_rate": 2.6783351996433018e-06,
+      "loss": 4.3292,
+      "step": 13365
+    },
+    {
+      "epoch": 2.6225970969007455,
+      "grad_norm": 25.782760620117188,
+      "learning_rate": 2.67490796192211e-06,
+      "loss": 4.8116,
+      "step": 13370
+    },
+    {
+      "epoch": 2.623577873675951,
+      "grad_norm": 17.976646423339844,
+      "learning_rate": 2.671482117454518e-06,
+      "loss": 4.641,
+      "step": 13375
+    },
+    {
+      "epoch": 2.6245586504511573,
+      "grad_norm": 20.79668617248535,
+      "learning_rate": 2.66805766829338e-06,
+      "loss": 4.7579,
+      "step": 13380
+    },
+    {
+      "epoch": 2.6255394272263635,
+      "grad_norm": 14.479522705078125,
+      "learning_rate": 2.6646346164907087e-06,
+      "loss": 4.7085,
+      "step": 13385
+    },
+    {
+      "epoch": 2.626520204001569,
+      "grad_norm": 28.237882614135742,
+      "learning_rate": 2.6612129640976875e-06,
+      "loss": 4.4694,
+      "step": 13390
+    },
+    {
+      "epoch": 2.6275009807767753,
+      "grad_norm": 15.98440170288086,
+      "learning_rate": 2.6577927131646513e-06,
+      "loss": 4.5517,
+      "step": 13395
+    },
+    {
+      "epoch": 2.628481757551981,
+      "grad_norm": 24.405895233154297,
+      "learning_rate": 2.6543738657411033e-06,
+      "loss": 4.3712,
+      "step": 13400
+    },
+    {
+      "epoch": 2.629462534327187,
+      "grad_norm": 14.398210525512695,
+      "learning_rate": 2.650956423875704e-06,
+      "loss": 4.7705,
+      "step": 13405
+    },
+    {
+      "epoch": 2.6304433111023933,
+      "grad_norm": 19.128984451293945,
+      "learning_rate": 2.6475403896162676e-06,
+      "loss": 4.5675,
+      "step": 13410
+    },
+    {
+      "epoch": 2.631424087877599,
+      "grad_norm": 20.786842346191406,
+      "learning_rate": 2.6441257650097705e-06,
+      "loss": 4.3912,
+      "step": 13415
+    },
+    {
+      "epoch": 2.632404864652805,
+      "grad_norm": 19.81397819519043,
+      "learning_rate": 2.6407125521023387e-06,
+      "loss": 4.6613,
+      "step": 13420
+    },
+    {
+      "epoch": 2.633385641428011,
+      "grad_norm": 15.989570617675781,
+      "learning_rate": 2.6373007529392565e-06,
+      "loss": 4.5901,
+      "step": 13425
+    },
+    {
+      "epoch": 2.634366418203217,
+      "grad_norm": 19.9948673248291,
+      "learning_rate": 2.633890369564962e-06,
+      "loss": 4.4191,
+      "step": 13430
+    },
+    {
+      "epoch": 2.635347194978423,
+      "grad_norm": 21.669771194458008,
+      "learning_rate": 2.6304814040230397e-06,
+      "loss": 4.5845,
+      "step": 13435
+    },
+    {
+      "epoch": 2.6363279717536288,
+      "grad_norm": 36.29818344116211,
+      "learning_rate": 2.6270738583562295e-06,
+      "loss": 4.5574,
+      "step": 13440
+    },
+    {
+      "epoch": 2.637308748528835,
+      "grad_norm": 14.819993019104004,
+      "learning_rate": 2.623667734606414e-06,
+      "loss": 4.4312,
+      "step": 13445
+    },
+    {
+      "epoch": 2.6382895253040406,
+      "grad_norm": 13.607355117797852,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 4.5856,
+      "step": 13450
+    },
+    {
+      "epoch": 2.6392703020792467,
+      "grad_norm": 18.39212417602539,
+      "learning_rate": 2.6168597610210673e-06,
+      "loss": 4.4272,
+      "step": 13455
+    },
+    {
+      "epoch": 2.640251078854453,
+      "grad_norm": 10.951640129089355,
+      "learning_rate": 2.613457915265042e-06,
+      "loss": 4.5139,
+      "step": 13460
+    },
+    {
+      "epoch": 2.6412318556296586,
+      "grad_norm": 15.258952140808105,
+      "learning_rate": 2.6100574995850316e-06,
+      "loss": 4.2683,
+      "step": 13465
+    },
+    {
+      "epoch": 2.6422126324048647,
+      "grad_norm": 16.9803466796875,
+      "learning_rate": 2.6066585160186477e-06,
+      "loss": 4.7361,
+      "step": 13470
+    },
+    {
+      "epoch": 2.6431934091800704,
+      "grad_norm": 18.152610778808594,
+      "learning_rate": 2.6032609666026476e-06,
+      "loss": 4.7387,
+      "step": 13475
+    },
+    {
+      "epoch": 2.6441741859552765,
+      "grad_norm": 24.47258949279785,
+      "learning_rate": 2.599864853372931e-06,
+      "loss": 4.4338,
+      "step": 13480
+    },
+    {
+      "epoch": 2.6451549627304827,
+      "grad_norm": 33.78486633300781,
+      "learning_rate": 2.5964701783645296e-06,
+      "loss": 4.4178,
+      "step": 13485
+    },
+    {
+      "epoch": 2.6461357395056884,
+      "grad_norm": 28.668806076049805,
+      "learning_rate": 2.593076943611623e-06,
+      "loss": 4.3034,
+      "step": 13490
+    },
+    {
+      "epoch": 2.6471165162808945,
+      "grad_norm": 14.24481201171875,
+      "learning_rate": 2.5896851511475184e-06,
+      "loss": 4.5715,
+      "step": 13495
+    },
+    {
+      "epoch": 2.6480972930561,
+      "grad_norm": 23.95184326171875,
+      "learning_rate": 2.5862948030046676e-06,
+      "loss": 4.5584,
+      "step": 13500
+    },
+    {
+      "epoch": 2.6490780698313063,
+      "grad_norm": 18.729578018188477,
+      "learning_rate": 2.5829059012146466e-06,
+      "loss": 4.4817,
+      "step": 13505
+    },
+    {
+      "epoch": 2.6500588466065125,
+      "grad_norm": 38.478668212890625,
+      "learning_rate": 2.579518447808177e-06,
+      "loss": 4.7689,
+      "step": 13510
+    },
+    {
+      "epoch": 2.651039623381718,
+      "grad_norm": 29.371217727661133,
+      "learning_rate": 2.5761324448151017e-06,
+      "loss": 4.6578,
+      "step": 13515
+    },
+    {
+      "epoch": 2.6520204001569243,
+      "grad_norm": 11.126766204833984,
+      "learning_rate": 2.572747894264399e-06,
+      "loss": 4.7112,
+      "step": 13520
+    },
+    {
+      "epoch": 2.65300117693213,
+      "grad_norm": 23.9212646484375,
+      "learning_rate": 2.5693647981841766e-06,
+      "loss": 4.4847,
+      "step": 13525
+    },
+    {
+      "epoch": 2.653981953707336,
+      "grad_norm": 16.88579559326172,
+      "learning_rate": 2.565983158601675e-06,
+      "loss": 4.5293,
+      "step": 13530
+    },
+    {
+      "epoch": 2.6549627304825423,
+      "grad_norm": 21.897232055664062,
+      "learning_rate": 2.5626029775432513e-06,
+      "loss": 4.6525,
+      "step": 13535
+    },
+    {
+      "epoch": 2.655943507257748,
+      "grad_norm": 26.479185104370117,
+      "learning_rate": 2.5592242570344008e-06,
+      "loss": 4.477,
+      "step": 13540
+    },
+    {
+      "epoch": 2.656924284032954,
+      "grad_norm": 17.854076385498047,
+      "learning_rate": 2.5558469990997327e-06,
+      "loss": 4.4696,
+      "step": 13545
+    },
+    {
+      "epoch": 2.65790506080816,
+      "grad_norm": 13.446738243103027,
+      "learning_rate": 2.5524712057629867e-06,
+      "loss": 4.8397,
+      "step": 13550
+    },
+    {
+      "epoch": 2.658885837583366,
+      "grad_norm": 14.61733627319336,
+      "learning_rate": 2.549096879047026e-06,
+      "loss": 4.4787,
+      "step": 13555
+    },
+    {
+      "epoch": 2.659866614358572,
+      "grad_norm": 15.360088348388672,
+      "learning_rate": 2.5457240209738278e-06,
+      "loss": 4.4534,
+      "step": 13560
+    },
+    {
+      "epoch": 2.660847391133778,
+      "grad_norm": 14.018891334533691,
+      "learning_rate": 2.5423526335644967e-06,
+      "loss": 4.6227,
+      "step": 13565
+    },
+    {
+      "epoch": 2.661828167908984,
+      "grad_norm": 13.576262474060059,
+      "learning_rate": 2.5389827188392495e-06,
+      "loss": 4.7652,
+      "step": 13570
+    },
+    {
+      "epoch": 2.6628089446841896,
+      "grad_norm": 18.732263565063477,
+      "learning_rate": 2.5356142788174277e-06,
+      "loss": 4.6374,
+      "step": 13575
+    },
+    {
+      "epoch": 2.6637897214593957,
+      "grad_norm": 20.31509017944336,
+      "learning_rate": 2.532247315517481e-06,
+      "loss": 4.3886,
+      "step": 13580
+    },
+    {
+      "epoch": 2.664770498234602,
+      "grad_norm": 12.62197494506836,
+      "learning_rate": 2.528881830956983e-06,
+      "loss": 4.8373,
+      "step": 13585
+    },
+    {
+      "epoch": 2.665751275009808,
+      "grad_norm": 32.32179260253906,
+      "learning_rate": 2.525517827152614e-06,
+      "loss": 4.224,
+      "step": 13590
+    },
+    {
+      "epoch": 2.6667320517850137,
+      "grad_norm": 21.820711135864258,
+      "learning_rate": 2.5221553061201678e-06,
+      "loss": 4.5365,
+      "step": 13595
+    },
+    {
+      "epoch": 2.66771282856022,
+      "grad_norm": 33.68283462524414,
+      "learning_rate": 2.518794269874553e-06,
+      "loss": 4.5051,
+      "step": 13600
+    },
+    {
+      "epoch": 2.6686936053354255,
+      "grad_norm": 12.54172134399414,
+      "learning_rate": 2.5154347204297903e-06,
+      "loss": 4.1707,
+      "step": 13605
+    },
+    {
+      "epoch": 2.6696743821106317,
+      "grad_norm": 10.5189847946167,
+      "learning_rate": 2.512076659799001e-06,
+      "loss": 4.5797,
+      "step": 13610
+    },
+    {
+      "epoch": 2.670655158885838,
+      "grad_norm": 32.68767547607422,
+      "learning_rate": 2.508720089994424e-06,
+      "loss": 4.8186,
+      "step": 13615
+    },
+    {
+      "epoch": 2.6716359356610435,
+      "grad_norm": 19.620351791381836,
+      "learning_rate": 2.505365013027397e-06,
+      "loss": 4.3365,
+      "step": 13620
+    },
+    {
+      "epoch": 2.6726167124362497,
+      "grad_norm": 27.10638999938965,
+      "learning_rate": 2.5020114309083676e-06,
+      "loss": 4.5674,
+      "step": 13625
+    },
+    {
+      "epoch": 2.6735974892114553,
+      "grad_norm": 24.49579620361328,
+      "learning_rate": 2.498659345646888e-06,
+      "loss": 4.3949,
+      "step": 13630
+    },
+    {
+      "epoch": 2.6745782659866615,
+      "grad_norm": 12.413771629333496,
+      "learning_rate": 2.4953087592516088e-06,
+      "loss": 4.4483,
+      "step": 13635
+    },
+    {
+      "epoch": 2.6755590427618676,
+      "grad_norm": 15.775691032409668,
+      "learning_rate": 2.49195967373029e-06,
+      "loss": 4.3685,
+      "step": 13640
+    },
+    {
+      "epoch": 2.6765398195370733,
+      "grad_norm": 14.530928611755371,
+      "learning_rate": 2.4886120910897826e-06,
+      "loss": 4.4426,
+      "step": 13645
+    },
+    {
+      "epoch": 2.6775205963122795,
+      "grad_norm": 36.24016189575195,
+      "learning_rate": 2.485266013336047e-06,
+      "loss": 4.3312,
+      "step": 13650
+    },
+    {
+      "epoch": 2.678501373087485,
+      "grad_norm": 16.918540954589844,
+      "learning_rate": 2.481921442474135e-06,
+      "loss": 4.4554,
+      "step": 13655
+    },
+    {
+      "epoch": 2.6794821498626913,
+      "grad_norm": 21.943265914916992,
+      "learning_rate": 2.478578380508196e-06,
+      "loss": 4.9183,
+      "step": 13660
+    },
+    {
+      "epoch": 2.6804629266378974,
+      "grad_norm": 28.09203338623047,
+      "learning_rate": 2.47523682944148e-06,
+      "loss": 4.5556,
+      "step": 13665
+    },
+    {
+      "epoch": 2.681443703413103,
+      "grad_norm": 12.404359817504883,
+      "learning_rate": 2.471896791276325e-06,
+      "loss": 4.4326,
+      "step": 13670
+    },
+    {
+      "epoch": 2.6824244801883093,
+      "grad_norm": 32.98741912841797,
+      "learning_rate": 2.4685582680141672e-06,
+      "loss": 4.1956,
+      "step": 13675
+    },
+    {
+      "epoch": 2.683405256963515,
+      "grad_norm": 18.23993682861328,
+      "learning_rate": 2.4652212616555367e-06,
+      "loss": 4.6537,
+      "step": 13680
+    },
+    {
+      "epoch": 2.684386033738721,
+      "grad_norm": 20.293107986450195,
+      "learning_rate": 2.4618857742000463e-06,
+      "loss": 4.622,
+      "step": 13685
+    },
+    {
+      "epoch": 2.6853668105139272,
+      "grad_norm": 21.977144241333008,
+      "learning_rate": 2.458551807646409e-06,
+      "loss": 4.631,
+      "step": 13690
+    },
+    {
+      "epoch": 2.686347587289133,
+      "grad_norm": 40.22378158569336,
+      "learning_rate": 2.4552193639924167e-06,
+      "loss": 4.6723,
+      "step": 13695
+    },
+    {
+      "epoch": 2.687328364064339,
+      "grad_norm": 44.497806549072266,
+      "learning_rate": 2.451888445234955e-06,
+      "loss": 5.0051,
+      "step": 13700
+    },
+    {
+      "epoch": 2.6883091408395448,
+      "grad_norm": 18.855010986328125,
+      "learning_rate": 2.4485590533699977e-06,
+      "loss": 4.3175,
+      "step": 13705
+    },
+    {
+      "epoch": 2.689289917614751,
+      "grad_norm": 14.9886474609375,
+      "learning_rate": 2.4452311903925953e-06,
+      "loss": 4.6321,
+      "step": 13710
+    },
+    {
+      "epoch": 2.690270694389957,
+      "grad_norm": 19.263511657714844,
+      "learning_rate": 2.44190485829689e-06,
+      "loss": 4.4146,
+      "step": 13715
+    },
+    {
+      "epoch": 2.6912514711651627,
+      "grad_norm": 15.527201652526855,
+      "learning_rate": 2.4385800590761017e-06,
+      "loss": 4.4363,
+      "step": 13720
+    },
+    {
+      "epoch": 2.692232247940369,
+      "grad_norm": 29.534744262695312,
+      "learning_rate": 2.435256794722536e-06,
+      "loss": 4.7091,
+      "step": 13725
+    },
+    {
+      "epoch": 2.6932130247155746,
+      "grad_norm": 33.34679412841797,
+      "learning_rate": 2.4319350672275743e-06,
+      "loss": 4.456,
+      "step": 13730
+    },
+    {
+      "epoch": 2.6941938014907807,
+      "grad_norm": 18.027267456054688,
+      "learning_rate": 2.428614878581678e-06,
+      "loss": 4.5147,
+      "step": 13735
+    },
+    {
+      "epoch": 2.695174578265987,
+      "grad_norm": 17.677387237548828,
+      "learning_rate": 2.4252962307743922e-06,
+      "loss": 4.5624,
+      "step": 13740
+    },
+    {
+      "epoch": 2.6961553550411925,
+      "grad_norm": 16.738374710083008,
+      "learning_rate": 2.4219791257943287e-06,
+      "loss": 4.4275,
+      "step": 13745
+    },
+    {
+      "epoch": 2.6971361318163987,
+      "grad_norm": 19.49000358581543,
+      "learning_rate": 2.4186635656291834e-06,
+      "loss": 4.2897,
+      "step": 13750
+    },
+    {
+      "epoch": 2.6981169085916044,
+      "grad_norm": 26.61591911315918,
+      "learning_rate": 2.4153495522657246e-06,
+      "loss": 4.5153,
+      "step": 13755
+    },
+    {
+      "epoch": 2.6990976853668105,
+      "grad_norm": 20.397048950195312,
+      "learning_rate": 2.412037087689788e-06,
+      "loss": 4.31,
+      "step": 13760
+    },
+    {
+      "epoch": 2.7000784621420166,
+      "grad_norm": 18.144649505615234,
+      "learning_rate": 2.4087261738862907e-06,
+      "loss": 4.5384,
+      "step": 13765
+    },
+    {
+      "epoch": 2.7010592389172223,
+      "grad_norm": 18.39889144897461,
+      "learning_rate": 2.405416812839211e-06,
+      "loss": 4.4709,
+      "step": 13770
+    },
+    {
+      "epoch": 2.7020400156924285,
+      "grad_norm": 19.8008975982666,
+      "learning_rate": 2.4021090065316026e-06,
+      "loss": 4.2814,
+      "step": 13775
+    },
+    {
+      "epoch": 2.703020792467634,
+      "grad_norm": 12.667327880859375,
+      "learning_rate": 2.3988027569455895e-06,
+      "loss": 4.6457,
+      "step": 13780
+    },
+    {
+      "epoch": 2.7040015692428403,
+      "grad_norm": 23.41883087158203,
+      "learning_rate": 2.3954980660623545e-06,
+      "loss": 4.4892,
+      "step": 13785
+    },
+    {
+      "epoch": 2.7049823460180464,
+      "grad_norm": 23.43563461303711,
+      "learning_rate": 2.392194935862156e-06,
+      "loss": 4.7607,
+      "step": 13790
+    },
+    {
+      "epoch": 2.705963122793252,
+      "grad_norm": 17.52957534790039,
+      "learning_rate": 2.3888933683243105e-06,
+      "loss": 4.3509,
+      "step": 13795
+    },
+    {
+      "epoch": 2.7069438995684583,
+      "grad_norm": 21.38890266418457,
+      "learning_rate": 2.3855933654271986e-06,
+      "loss": 4.6536,
+      "step": 13800
+    },
+    {
+      "epoch": 2.707924676343664,
+      "grad_norm": 21.12211799621582,
+      "learning_rate": 2.382294929148268e-06,
+      "loss": 4.5384,
+      "step": 13805
+    },
+    {
+      "epoch": 2.70890545311887,
+      "grad_norm": 12.217236518859863,
+      "learning_rate": 2.3789980614640212e-06,
+      "loss": 4.6123,
+      "step": 13810
+    },
+    {
+      "epoch": 2.7098862298940762,
+      "grad_norm": 28.40160369873047,
+      "learning_rate": 2.375702764350029e-06,
+      "loss": 4.6543,
+      "step": 13815
+    },
+    {
+      "epoch": 2.710867006669282,
+      "grad_norm": 18.181427001953125,
+      "learning_rate": 2.3724090397809112e-06,
+      "loss": 4.3407,
+      "step": 13820
+    },
+    {
+      "epoch": 2.711847783444488,
+      "grad_norm": 12.905070304870605,
+      "learning_rate": 2.369116889730353e-06,
+      "loss": 4.4545,
+      "step": 13825
+    },
+    {
+      "epoch": 2.7128285602196938,
+      "grad_norm": 14.678855895996094,
+      "learning_rate": 2.3658263161710948e-06,
+      "loss": 4.5105,
+      "step": 13830
+    },
+    {
+      "epoch": 2.7138093369949,
+      "grad_norm": 25.800718307495117,
+      "learning_rate": 2.3625373210749277e-06,
+      "loss": 4.1325,
+      "step": 13835
+    },
+    {
+      "epoch": 2.714790113770106,
+      "grad_norm": 25.922931671142578,
+      "learning_rate": 2.359249906412704e-06,
+      "loss": 4.5417,
+      "step": 13840
+    },
+    {
+      "epoch": 2.7157708905453117,
+      "grad_norm": 16.528545379638672,
+      "learning_rate": 2.3559640741543212e-06,
+      "loss": 4.3443,
+      "step": 13845
+    },
+    {
+      "epoch": 2.716751667320518,
+      "grad_norm": 26.857633590698242,
+      "learning_rate": 2.3526798262687337e-06,
+      "loss": 4.4824,
+      "step": 13850
+    },
+    {
+      "epoch": 2.7177324440957236,
+      "grad_norm": 16.661218643188477,
+      "learning_rate": 2.3493971647239495e-06,
+      "loss": 4.4816,
+      "step": 13855
+    },
+    {
+      "epoch": 2.7187132208709297,
+      "grad_norm": 17.91668128967285,
+      "learning_rate": 2.346116091487016e-06,
+      "loss": 4.8708,
+      "step": 13860
+    },
+    {
+      "epoch": 2.719693997646136,
+      "grad_norm": 18.95901870727539,
+      "learning_rate": 2.3428366085240394e-06,
+      "loss": 4.7818,
+      "step": 13865
+    },
+    {
+      "epoch": 2.720674774421342,
+      "grad_norm": 14.655415534973145,
+      "learning_rate": 2.3395587178001667e-06,
+      "loss": 4.5327,
+      "step": 13870
+    },
+    {
+      "epoch": 2.7216555511965477,
+      "grad_norm": 22.418703079223633,
+      "learning_rate": 2.33628242127959e-06,
+      "loss": 4.3016,
+      "step": 13875
+    },
+    {
+      "epoch": 2.7226363279717534,
+      "grad_norm": 29.357666015625,
+      "learning_rate": 2.333007720925552e-06,
+      "loss": 4.1335,
+      "step": 13880
+    },
+    {
+      "epoch": 2.7236171047469595,
+      "grad_norm": 25.820316314697266,
+      "learning_rate": 2.3297346187003327e-06,
+      "loss": 4.8931,
+      "step": 13885
+    },
+    {
+      "epoch": 2.7245978815221656,
+      "grad_norm": 66.62987518310547,
+      "learning_rate": 2.3264631165652608e-06,
+      "loss": 4.781,
+      "step": 13890
+    },
+    {
+      "epoch": 2.725578658297372,
+      "grad_norm": 12.275607109069824,
+      "learning_rate": 2.323193216480698e-06,
+      "loss": 4.3785,
+      "step": 13895
+    },
+    {
+      "epoch": 2.7265594350725775,
+      "grad_norm": 32.049034118652344,
+      "learning_rate": 2.319924920406054e-06,
+      "loss": 4.359,
+      "step": 13900
+    },
+    {
+      "epoch": 2.727540211847783,
+      "grad_norm": 45.356746673583984,
+      "learning_rate": 2.3166582302997744e-06,
+      "loss": 4.5512,
+      "step": 13905
+    },
+    {
+      "epoch": 2.7285209886229893,
+      "grad_norm": 13.335826873779297,
+      "learning_rate": 2.3133931481193383e-06,
+      "loss": 4.8334,
+      "step": 13910
+    },
+    {
+      "epoch": 2.7295017653981954,
+      "grad_norm": 18.16646957397461,
+      "learning_rate": 2.31012967582127e-06,
+      "loss": 4.5381,
+      "step": 13915
+    },
+    {
+      "epoch": 2.7304825421734016,
+      "grad_norm": 37.283790588378906,
+      "learning_rate": 2.3068678153611195e-06,
+      "loss": 4.7861,
+      "step": 13920
+    },
+    {
+      "epoch": 2.7314633189486073,
+      "grad_norm": 25.240339279174805,
+      "learning_rate": 2.303607568693478e-06,
+      "loss": 4.5222,
+      "step": 13925
+    },
+    {
+      "epoch": 2.7324440957238134,
+      "grad_norm": 23.09556770324707,
+      "learning_rate": 2.3003489377719682e-06,
+      "loss": 4.3437,
+      "step": 13930
+    },
+    {
+      "epoch": 2.733424872499019,
+      "grad_norm": 22.769893646240234,
+      "learning_rate": 2.2970919245492406e-06,
+      "loss": 4.6166,
+      "step": 13935
+    },
+    {
+      "epoch": 2.7344056492742252,
+      "grad_norm": 10.76894760131836,
+      "learning_rate": 2.293836530976984e-06,
+      "loss": 4.4213,
+      "step": 13940
+    },
+    {
+      "epoch": 2.7353864260494314,
+      "grad_norm": 27.2360897064209,
+      "learning_rate": 2.290582759005908e-06,
+      "loss": 4.5921,
+      "step": 13945
+    },
+    {
+      "epoch": 2.736367202824637,
+      "grad_norm": 36.54833984375,
+      "learning_rate": 2.2873306105857546e-06,
+      "loss": 4.7721,
+      "step": 13950
+    },
+    {
+      "epoch": 2.737347979599843,
+      "grad_norm": 28.067516326904297,
+      "learning_rate": 2.2840800876652963e-06,
+      "loss": 4.3667,
+      "step": 13955
+    },
+    {
+      "epoch": 2.738328756375049,
+      "grad_norm": 26.357559204101562,
+      "learning_rate": 2.280831192192324e-06,
+      "loss": 4.6859,
+      "step": 13960
+    },
+    {
+      "epoch": 2.739309533150255,
+      "grad_norm": 37.89749526977539,
+      "learning_rate": 2.2775839261136607e-06,
+      "loss": 4.519,
+      "step": 13965
+    },
+    {
+      "epoch": 2.740290309925461,
+      "grad_norm": 29.265758514404297,
+      "learning_rate": 2.274338291375147e-06,
+      "loss": 4.7713,
+      "step": 13970
+    },
+    {
+      "epoch": 2.741271086700667,
+      "grad_norm": 24.310449600219727,
+      "learning_rate": 2.271094289921651e-06,
+      "loss": 4.5785,
+      "step": 13975
+    },
+    {
+      "epoch": 2.742251863475873,
+      "grad_norm": 24.709442138671875,
+      "learning_rate": 2.2678519236970612e-06,
+      "loss": 4.3531,
+      "step": 13980
+    },
+    {
+      "epoch": 2.7432326402510787,
+      "grad_norm": 21.423248291015625,
+      "learning_rate": 2.2646111946442813e-06,
+      "loss": 4.3444,
+      "step": 13985
+    },
+    {
+      "epoch": 2.744213417026285,
+      "grad_norm": 28.040691375732422,
+      "learning_rate": 2.261372104705241e-06,
+      "loss": 4.2849,
+      "step": 13990
+    },
+    {
+      "epoch": 2.745194193801491,
+      "grad_norm": 22.47325325012207,
+      "learning_rate": 2.2581346558208817e-06,
+      "loss": 4.3244,
+      "step": 13995
+    },
+    {
+      "epoch": 2.7461749705766967,
+      "grad_norm": 16.90743064880371,
+      "learning_rate": 2.2548988499311647e-06,
+      "loss": 4.3961,
+      "step": 14000
+    },
+    {
+      "epoch": 2.747155747351903,
+      "grad_norm": 19.77780532836914,
+      "learning_rate": 2.2516646889750694e-06,
+      "loss": 4.693,
+      "step": 14005
+    },
+    {
+      "epoch": 2.7481365241271085,
+      "grad_norm": 19.31178855895996,
+      "learning_rate": 2.2484321748905835e-06,
+      "loss": 4.2515,
+      "step": 14010
+    },
+    {
+      "epoch": 2.7491173009023147,
+      "grad_norm": 20.418745040893555,
+      "learning_rate": 2.245201309614709e-06,
+      "loss": 4.0248,
+      "step": 14015
+    },
+    {
+      "epoch": 2.750098077677521,
+      "grad_norm": 11.354876518249512,
+      "learning_rate": 2.241972095083466e-06,
+      "loss": 4.429,
+      "step": 14020
+    },
+    {
+      "epoch": 2.7510788544527265,
+      "grad_norm": 19.0582332611084,
+      "learning_rate": 2.238744533231877e-06,
+      "loss": 4.5341,
+      "step": 14025
+    },
+    {
+      "epoch": 2.7510788544527265,
+      "eval_loss": 4.857458114624023,
+      "eval_runtime": 7.7687,
+      "eval_samples_per_second": 26.903,
+      "eval_steps_per_second": 13.516,
+      "step": 14025
+    },
+    {
+      "epoch": 2.7520596312279326,
+      "grad_norm": 24.781587600708008,
+      "learning_rate": 2.235518625993981e-06,
+      "loss": 4.8467,
+      "step": 14030
+    },
+    {
+      "epoch": 2.7530404080031383,
+      "grad_norm": 20.93712615966797,
+      "learning_rate": 2.2322943753028204e-06,
+      "loss": 4.8466,
+      "step": 14035
+    },
+    {
+      "epoch": 2.7540211847783445,
+      "grad_norm": 58.71204376220703,
+      "learning_rate": 2.22907178309045e-06,
+      "loss": 5.1757,
+      "step": 14040
+    },
+    {
+      "epoch": 2.7550019615535506,
+      "grad_norm": 18.296762466430664,
+      "learning_rate": 2.2258508512879246e-06,
+      "loss": 4.4446,
+      "step": 14045
+    },
+    {
+      "epoch": 2.7559827383287563,
+      "grad_norm": 25.872365951538086,
+      "learning_rate": 2.2226315818253097e-06,
+      "loss": 4.7809,
+      "step": 14050
+    },
+    {
+      "epoch": 2.7569635151039624,
+      "grad_norm": 24.24050521850586,
+      "learning_rate": 2.219413976631674e-06,
+      "loss": 4.4556,
+      "step": 14055
+    },
+    {
+      "epoch": 2.757944291879168,
+      "grad_norm": 23.216476440429688,
+      "learning_rate": 2.2161980376350837e-06,
+      "loss": 4.6814,
+      "step": 14060
+    },
+    {
+      "epoch": 2.7589250686543743,
+      "grad_norm": 18.804798126220703,
+      "learning_rate": 2.2129837667626147e-06,
+      "loss": 4.3724,
+      "step": 14065
+    },
+    {
+      "epoch": 2.7599058454295804,
+      "grad_norm": 23.965618133544922,
+      "learning_rate": 2.2097711659403344e-06,
+      "loss": 4.7329,
+      "step": 14070
+    },
+    {
+      "epoch": 2.760886622204786,
+      "grad_norm": 17.891071319580078,
+      "learning_rate": 2.2065602370933153e-06,
+      "loss": 4.8861,
+      "step": 14075
+    },
+    {
+      "epoch": 2.7618673989799922,
+      "grad_norm": 19.460275650024414,
+      "learning_rate": 2.20335098214563e-06,
+      "loss": 4.8461,
+      "step": 14080
+    },
+    {
+      "epoch": 2.762848175755198,
+      "grad_norm": 20.419872283935547,
+      "learning_rate": 2.2001434030203423e-06,
+      "loss": 4.5978,
+      "step": 14085
+    },
+    {
+      "epoch": 2.763828952530404,
+      "grad_norm": 23.175350189208984,
+      "learning_rate": 2.1969375016395138e-06,
+      "loss": 4.3628,
+      "step": 14090
+    },
+    {
+      "epoch": 2.76480972930561,
+      "grad_norm": 15.233471870422363,
+      "learning_rate": 2.1937332799241993e-06,
+      "loss": 4.4651,
+      "step": 14095
+    },
+    {
+      "epoch": 2.765790506080816,
+      "grad_norm": 21.109416961669922,
+      "learning_rate": 2.190530739794452e-06,
+      "loss": 4.4845,
+      "step": 14100
+    },
+    {
+      "epoch": 2.766771282856022,
+      "grad_norm": 26.169544219970703,
+      "learning_rate": 2.187329883169315e-06,
+      "loss": 4.5825,
+      "step": 14105
+    },
+    {
+      "epoch": 2.7677520596312277,
+      "grad_norm": 27.972253799438477,
+      "learning_rate": 2.184130711966819e-06,
+      "loss": 4.7661,
+      "step": 14110
+    },
+    {
+      "epoch": 2.768732836406434,
+      "grad_norm": 15.088165283203125,
+      "learning_rate": 2.180933228103992e-06,
+      "loss": 4.3559,
+      "step": 14115
+    },
+    {
+      "epoch": 2.76971361318164,
+      "grad_norm": 10.408885955810547,
+      "learning_rate": 2.177737433496842e-06,
+      "loss": 4.3335,
+      "step": 14120
+    },
+    {
+      "epoch": 2.7706943899568457,
+      "grad_norm": 12.056416511535645,
+      "learning_rate": 2.1745433300603714e-06,
+      "loss": 4.2495,
+      "step": 14125
+    },
+    {
+      "epoch": 2.771675166732052,
+      "grad_norm": 19.54852294921875,
+      "learning_rate": 2.1713509197085698e-06,
+      "loss": 4.4882,
+      "step": 14130
+    },
+    {
+      "epoch": 2.7726559435072575,
+      "grad_norm": 25.85139274597168,
+      "learning_rate": 2.1681602043544057e-06,
+      "loss": 4.8348,
+      "step": 14135
+    },
+    {
+      "epoch": 2.7736367202824637,
+      "grad_norm": 11.395781517028809,
+      "learning_rate": 2.164971185909839e-06,
+      "loss": 4.3054,
+      "step": 14140
+    },
+    {
+      "epoch": 2.77461749705767,
+      "grad_norm": 24.38686752319336,
+      "learning_rate": 2.1617838662858075e-06,
+      "loss": 4.1594,
+      "step": 14145
+    },
+    {
+      "epoch": 2.7755982738328755,
+      "grad_norm": 28.35773468017578,
+      "learning_rate": 2.158598247392236e-06,
+      "loss": 4.5292,
+      "step": 14150
+    },
+    {
+      "epoch": 2.7765790506080816,
+      "grad_norm": 17.550443649291992,
+      "learning_rate": 2.1554143311380237e-06,
+      "loss": 4.8225,
+      "step": 14155
+    },
+    {
+      "epoch": 2.7775598273832873,
+      "grad_norm": 26.152284622192383,
+      "learning_rate": 2.1522321194310577e-06,
+      "loss": 4.8608,
+      "step": 14160
+    },
+    {
+      "epoch": 2.7785406041584935,
+      "grad_norm": 19.523468017578125,
+      "learning_rate": 2.1490516141781957e-06,
+      "loss": 4.7656,
+      "step": 14165
+    },
+    {
+      "epoch": 2.7795213809336996,
+      "grad_norm": 15.315962791442871,
+      "learning_rate": 2.1458728172852765e-06,
+      "loss": 4.4612,
+      "step": 14170
+    },
+    {
+      "epoch": 2.7805021577089053,
+      "grad_norm": 19.123991012573242,
+      "learning_rate": 2.142695730657116e-06,
+      "loss": 4.499,
+      "step": 14175
+    },
+    {
+      "epoch": 2.7814829344841114,
+      "grad_norm": 47.76284408569336,
+      "learning_rate": 2.139520356197506e-06,
+      "loss": 4.7392,
+      "step": 14180
+    },
+    {
+      "epoch": 2.782463711259317,
+      "grad_norm": 24.149187088012695,
+      "learning_rate": 2.1363466958092077e-06,
+      "loss": 4.4477,
+      "step": 14185
+    },
+    {
+      "epoch": 2.7834444880345233,
+      "grad_norm": 22.08119773864746,
+      "learning_rate": 2.1331747513939615e-06,
+      "loss": 4.5059,
+      "step": 14190
+    },
+    {
+      "epoch": 2.7844252648097294,
+      "grad_norm": 20.25182342529297,
+      "learning_rate": 2.1300045248524724e-06,
+      "loss": 4.7203,
+      "step": 14195
+    },
+    {
+      "epoch": 2.7854060415849355,
+      "grad_norm": 10.127427101135254,
+      "learning_rate": 2.126836018084422e-06,
+      "loss": 4.4238,
+      "step": 14200
+    },
+    {
+      "epoch": 2.7863868183601412,
+      "grad_norm": 14.273734092712402,
+      "learning_rate": 2.123669232988461e-06,
+      "loss": 4.4544,
+      "step": 14205
+    },
+    {
+      "epoch": 2.787367595135347,
+      "grad_norm": 37.149620056152344,
+      "learning_rate": 2.120504171462203e-06,
+      "loss": 4.4163,
+      "step": 14210
+    },
+    {
+      "epoch": 2.788348371910553,
+      "grad_norm": 24.374950408935547,
+      "learning_rate": 2.1173408354022357e-06,
+      "loss": 4.4589,
+      "step": 14215
+    },
+    {
+      "epoch": 2.789329148685759,
+      "grad_norm": 15.419386863708496,
+      "learning_rate": 2.114179226704106e-06,
+      "loss": 4.3327,
+      "step": 14220
+    },
+    {
+      "epoch": 2.7903099254609653,
+      "grad_norm": 21.48487663269043,
+      "learning_rate": 2.1110193472623335e-06,
+      "loss": 4.6511,
+      "step": 14225
+    },
+    {
+      "epoch": 2.791290702236171,
+      "grad_norm": 12.925050735473633,
+      "learning_rate": 2.1078611989703934e-06,
+      "loss": 4.3783,
+      "step": 14230
+    },
+    {
+      "epoch": 2.7922714790113767,
+      "grad_norm": 18.62735939025879,
+      "learning_rate": 2.1047047837207315e-06,
+      "loss": 4.2144,
+      "step": 14235
+    },
+    {
+      "epoch": 2.793252255786583,
+      "grad_norm": 23.844316482543945,
+      "learning_rate": 2.1015501034047486e-06,
+      "loss": 4.7369,
+      "step": 14240
+    },
+    {
+      "epoch": 2.794233032561789,
+      "grad_norm": 24.961200714111328,
+      "learning_rate": 2.0983971599128072e-06,
+      "loss": 4.6055,
+      "step": 14245
+    },
+    {
+      "epoch": 2.795213809336995,
+      "grad_norm": 16.179439544677734,
+      "learning_rate": 2.0952459551342325e-06,
+      "loss": 4.5461,
+      "step": 14250
+    },
+    {
+      "epoch": 2.796194586112201,
+      "grad_norm": 20.85818862915039,
+      "learning_rate": 2.0920964909573065e-06,
+      "loss": 4.8779,
+      "step": 14255
+    },
+    {
+      "epoch": 2.797175362887407,
+      "grad_norm": 13.407557487487793,
+      "learning_rate": 2.0889487692692644e-06,
+      "loss": 4.6044,
+      "step": 14260
+    },
+    {
+      "epoch": 2.7981561396626127,
+      "grad_norm": 20.780235290527344,
+      "learning_rate": 2.0858027919563032e-06,
+      "loss": 4.7772,
+      "step": 14265
+    },
+    {
+      "epoch": 2.799136916437819,
+      "grad_norm": 14.837874412536621,
+      "learning_rate": 2.0826585609035686e-06,
+      "loss": 4.0348,
+      "step": 14270
+    },
+    {
+      "epoch": 2.800117693213025,
+      "grad_norm": 24.567312240600586,
+      "learning_rate": 2.0795160779951645e-06,
+      "loss": 4.277,
+      "step": 14275
+    },
+    {
+      "epoch": 2.8010984699882306,
+      "grad_norm": 21.53330421447754,
+      "learning_rate": 2.076375345114147e-06,
+      "loss": 4.4838,
+      "step": 14280
+    },
+    {
+      "epoch": 2.802079246763437,
+      "grad_norm": 32.00398635864258,
+      "learning_rate": 2.0732363641425197e-06,
+      "loss": 4.5141,
+      "step": 14285
+    },
+    {
+      "epoch": 2.8030600235386425,
+      "grad_norm": 17.834571838378906,
+      "learning_rate": 2.070099136961241e-06,
+      "loss": 4.417,
+      "step": 14290
+    },
+    {
+      "epoch": 2.8040408003138486,
+      "grad_norm": 10.907025337219238,
+      "learning_rate": 2.066963665450214e-06,
+      "loss": 4.8046,
+      "step": 14295
+    },
+    {
+      "epoch": 2.8050215770890548,
+      "grad_norm": 36.39023971557617,
+      "learning_rate": 2.063829951488295e-06,
+      "loss": 4.9091,
+      "step": 14300
+    },
+    {
+      "epoch": 2.8060023538642604,
+      "grad_norm": 33.004146575927734,
+      "learning_rate": 2.0606979969532826e-06,
+      "loss": 4.57,
+      "step": 14305
+    },
+    {
+      "epoch": 2.8069831306394666,
+      "grad_norm": 14.840770721435547,
+      "learning_rate": 2.0575678037219205e-06,
+      "loss": 4.4305,
+      "step": 14310
+    },
+    {
+      "epoch": 2.8079639074146723,
+      "grad_norm": 14.304911613464355,
+      "learning_rate": 2.0544393736699033e-06,
+      "loss": 4.2868,
+      "step": 14315
+    },
+    {
+      "epoch": 2.8089446841898784,
+      "grad_norm": 16.990671157836914,
+      "learning_rate": 2.051312708671861e-06,
+      "loss": 4.448,
+      "step": 14320
+    },
+    {
+      "epoch": 2.8099254609650846,
+      "grad_norm": 33.899173736572266,
+      "learning_rate": 2.048187810601372e-06,
+      "loss": 4.4279,
+      "step": 14325
+    },
+    {
+      "epoch": 2.8109062377402902,
+      "grad_norm": 23.93379020690918,
+      "learning_rate": 2.0450646813309555e-06,
+      "loss": 4.7488,
+      "step": 14330
+    },
+    {
+      "epoch": 2.8118870145154964,
+      "grad_norm": 27.46114730834961,
+      "learning_rate": 2.0419433227320653e-06,
+      "loss": 4.3003,
+      "step": 14335
+    },
+    {
+      "epoch": 2.812867791290702,
+      "grad_norm": 26.478914260864258,
+      "learning_rate": 2.0388237366751005e-06,
+      "loss": 4.8474,
+      "step": 14340
+    },
+    {
+      "epoch": 2.813848568065908,
+      "grad_norm": 26.579689025878906,
+      "learning_rate": 2.035705925029394e-06,
+      "loss": 4.7607,
+      "step": 14345
+    },
+    {
+      "epoch": 2.8148293448411144,
+      "grad_norm": 13.559803009033203,
+      "learning_rate": 2.0325898896632178e-06,
+      "loss": 4.5509,
+      "step": 14350
+    },
+    {
+      "epoch": 2.81581012161632,
+      "grad_norm": 24.959630966186523,
+      "learning_rate": 2.0294756324437804e-06,
+      "loss": 4.2184,
+      "step": 14355
+    },
+    {
+      "epoch": 2.816790898391526,
+      "grad_norm": 14.203490257263184,
+      "learning_rate": 2.026363155237219e-06,
+      "loss": 4.4857,
+      "step": 14360
+    },
+    {
+      "epoch": 2.817771675166732,
+      "grad_norm": 14.8135347366333,
+      "learning_rate": 2.0232524599086116e-06,
+      "loss": 4.4352,
+      "step": 14365
+    },
+    {
+      "epoch": 2.818752451941938,
+      "grad_norm": 19.16025161743164,
+      "learning_rate": 2.0201435483219627e-06,
+      "loss": 4.4608,
+      "step": 14370
+    },
+    {
+      "epoch": 2.819733228717144,
+      "grad_norm": 11.188132286071777,
+      "learning_rate": 2.0170364223402126e-06,
+      "loss": 4.3011,
+      "step": 14375
+    },
+    {
+      "epoch": 2.82071400549235,
+      "grad_norm": 15.542959213256836,
+      "learning_rate": 2.0139310838252283e-06,
+      "loss": 4.668,
+      "step": 14380
+    },
+    {
+      "epoch": 2.821694782267556,
+      "grad_norm": 15.383172988891602,
+      "learning_rate": 2.0108275346378052e-06,
+      "loss": 4.6553,
+      "step": 14385
+    },
+    {
+      "epoch": 2.8226755590427617,
+      "grad_norm": 32.74557113647461,
+      "learning_rate": 2.0077257766376707e-06,
+      "loss": 4.7629,
+      "step": 14390
+    },
+    {
+      "epoch": 2.823656335817968,
+      "grad_norm": 17.852758407592773,
+      "learning_rate": 2.0046258116834725e-06,
+      "loss": 4.6995,
+      "step": 14395
+    },
+    {
+      "epoch": 2.824637112593174,
+      "grad_norm": 17.616985321044922,
+      "learning_rate": 2.00152764163279e-06,
+      "loss": 4.4477,
+      "step": 14400
+    },
+    {
+      "epoch": 2.8256178893683797,
+      "grad_norm": 22.82828712463379,
+      "learning_rate": 1.9984312683421265e-06,
+      "loss": 4.4233,
+      "step": 14405
+    },
+    {
+      "epoch": 2.826598666143586,
+      "grad_norm": 20.8966121673584,
+      "learning_rate": 1.9953366936669023e-06,
+      "loss": 4.3212,
+      "step": 14410
+    },
+    {
+      "epoch": 2.8275794429187915,
+      "grad_norm": 13.764612197875977,
+      "learning_rate": 1.9922439194614686e-06,
+      "loss": 4.5841,
+      "step": 14415
+    },
+    {
+      "epoch": 2.8285602196939976,
+      "grad_norm": 10.778817176818848,
+      "learning_rate": 1.9891529475790894e-06,
+      "loss": 4.5126,
+      "step": 14420
+    },
+    {
+      "epoch": 2.8295409964692038,
+      "grad_norm": 26.92489242553711,
+      "learning_rate": 1.986063779871955e-06,
+      "loss": 4.4893,
+      "step": 14425
+    },
+    {
+      "epoch": 2.8305217732444095,
+      "grad_norm": 22.98575782775879,
+      "learning_rate": 1.9829764181911738e-06,
+      "loss": 4.5547,
+      "step": 14430
+    },
+    {
+      "epoch": 2.8315025500196156,
+      "grad_norm": 23.186853408813477,
+      "learning_rate": 1.979890864386767e-06,
+      "loss": 4.5995,
+      "step": 14435
+    },
+    {
+      "epoch": 2.8324833267948213,
+      "grad_norm": 14.08784294128418,
+      "learning_rate": 1.97680712030768e-06,
+      "loss": 4.9109,
+      "step": 14440
+    },
+    {
+      "epoch": 2.8334641035700274,
+      "grad_norm": 16.856956481933594,
+      "learning_rate": 1.9737251878017678e-06,
+      "loss": 4.7937,
+      "step": 14445
+    },
+    {
+      "epoch": 2.8344448803452336,
+      "grad_norm": 20.808732986450195,
+      "learning_rate": 1.970645068715799e-06,
+      "loss": 4.646,
+      "step": 14450
+    },
+    {
+      "epoch": 2.8354256571204393,
+      "grad_norm": 40.77276611328125,
+      "learning_rate": 1.967566764895464e-06,
+      "loss": 4.9738,
+      "step": 14455
+    },
+    {
+      "epoch": 2.8364064338956454,
+      "grad_norm": 16.21761131286621,
+      "learning_rate": 1.964490278185354e-06,
+      "loss": 4.4132,
+      "step": 14460
+    },
+    {
+      "epoch": 2.837387210670851,
+      "grad_norm": 11.041531562805176,
+      "learning_rate": 1.961415610428983e-06,
+      "loss": 4.6614,
+      "step": 14465
+    },
+    {
+      "epoch": 2.8383679874460572,
+      "grad_norm": 18.46826934814453,
+      "learning_rate": 1.958342763468764e-06,
+      "loss": 4.3458,
+      "step": 14470
+    },
+    {
+      "epoch": 2.8393487642212634,
+      "grad_norm": 25.5513973236084,
+      "learning_rate": 1.955271739146026e-06,
+      "loss": 4.454,
+      "step": 14475
+    },
+    {
+      "epoch": 2.840329540996469,
+      "grad_norm": 21.196474075317383,
+      "learning_rate": 1.952202539301007e-06,
+      "loss": 4.6721,
+      "step": 14480
+    },
+    {
+      "epoch": 2.841310317771675,
+      "grad_norm": 14.46927547454834,
+      "learning_rate": 1.949135165772844e-06,
+      "loss": 4.5196,
+      "step": 14485
+    },
+    {
+      "epoch": 2.842291094546881,
+      "grad_norm": 16.33852767944336,
+      "learning_rate": 1.9460696203995884e-06,
+      "loss": 4.3415,
+      "step": 14490
+    },
+    {
+      "epoch": 2.843271871322087,
+      "grad_norm": 22.176271438598633,
+      "learning_rate": 1.9430059050181883e-06,
+      "loss": 4.2941,
+      "step": 14495
+    },
+    {
+      "epoch": 2.844252648097293,
+      "grad_norm": 14.859818458557129,
+      "learning_rate": 1.9399440214645003e-06,
+      "loss": 4.5169,
+      "step": 14500
+    },
+    {
+      "epoch": 2.845233424872499,
+      "grad_norm": 13.2625732421875,
+      "learning_rate": 1.936883971573285e-06,
+      "loss": 4.424,
+      "step": 14505
+    },
+    {
+      "epoch": 2.846214201647705,
+      "grad_norm": 15.930765151977539,
+      "learning_rate": 1.9338257571781973e-06,
+      "loss": 4.386,
+      "step": 14510
+    },
+    {
+      "epoch": 2.8471949784229107,
+      "grad_norm": 16.069883346557617,
+      "learning_rate": 1.9307693801117983e-06,
+      "loss": 4.5049,
+      "step": 14515
+    },
+    {
+      "epoch": 2.848175755198117,
+      "grad_norm": 17.39498519897461,
+      "learning_rate": 1.9277148422055457e-06,
+      "loss": 4.5397,
+      "step": 14520
+    },
+    {
+      "epoch": 2.849156531973323,
+      "grad_norm": 15.79026985168457,
+      "learning_rate": 1.924662145289793e-06,
+      "loss": 4.4386,
+      "step": 14525
+    },
+    {
+      "epoch": 2.850137308748529,
+      "grad_norm": 28.315719604492188,
+      "learning_rate": 1.921611291193797e-06,
+      "loss": 4.6099,
+      "step": 14530
+    },
+    {
+      "epoch": 2.851118085523735,
+      "grad_norm": 19.82541847229004,
+      "learning_rate": 1.9185622817457024e-06,
+      "loss": 4.8542,
+      "step": 14535
+    },
+    {
+      "epoch": 2.8520988622989405,
+      "grad_norm": 19.061132431030273,
+      "learning_rate": 1.915515118772555e-06,
+      "loss": 4.5148,
+      "step": 14540
+    },
+    {
+      "epoch": 2.8530796390741466,
+      "grad_norm": 26.26479148864746,
+      "learning_rate": 1.912469804100289e-06,
+      "loss": 4.3535,
+      "step": 14545
+    },
+    {
+      "epoch": 2.8540604158493528,
+      "grad_norm": 23.275548934936523,
+      "learning_rate": 1.9094263395537353e-06,
+      "loss": 4.6812,
+      "step": 14550
+    },
+    {
+      "epoch": 2.855041192624559,
+      "grad_norm": 16.343244552612305,
+      "learning_rate": 1.9063847269566154e-06,
+      "loss": 4.2992,
+      "step": 14555
+    },
+    {
+      "epoch": 2.8560219693997646,
+      "grad_norm": 19.14699363708496,
+      "learning_rate": 1.903344968131537e-06,
+      "loss": 4.8218,
+      "step": 14560
+    },
+    {
+      "epoch": 2.8570027461749707,
+      "grad_norm": 15.184374809265137,
+      "learning_rate": 1.9003070649000033e-06,
+      "loss": 4.5549,
+      "step": 14565
+    },
+    {
+      "epoch": 2.8579835229501764,
+      "grad_norm": 32.97745895385742,
+      "learning_rate": 1.897271019082399e-06,
+      "loss": 4.6135,
+      "step": 14570
+    },
+    {
+      "epoch": 2.8589642997253826,
+      "grad_norm": 25.775196075439453,
+      "learning_rate": 1.894236832498001e-06,
+      "loss": 4.4366,
+      "step": 14575
+    },
+    {
+      "epoch": 2.8599450765005887,
+      "grad_norm": 9.466419219970703,
+      "learning_rate": 1.8912045069649709e-06,
+      "loss": 4.6671,
+      "step": 14580
+    },
+    {
+      "epoch": 2.8609258532757944,
+      "grad_norm": 12.89758014678955,
+      "learning_rate": 1.888174044300352e-06,
+      "loss": 4.631,
+      "step": 14585
+    },
+    {
+      "epoch": 2.8619066300510005,
+      "grad_norm": 25.096372604370117,
+      "learning_rate": 1.8851454463200769e-06,
+      "loss": 4.9087,
+      "step": 14590
+    },
+    {
+      "epoch": 2.8628874068262062,
+      "grad_norm": 21.342893600463867,
+      "learning_rate": 1.8821187148389557e-06,
+      "loss": 4.2392,
+      "step": 14595
+    },
+    {
+      "epoch": 2.8638681836014124,
+      "grad_norm": 14.787396430969238,
+      "learning_rate": 1.8790938516706802e-06,
+      "loss": 4.9076,
+      "step": 14600
+    },
+    {
+      "epoch": 2.8648489603766185,
+      "grad_norm": 25.85067367553711,
+      "learning_rate": 1.8760708586278287e-06,
+      "loss": 4.4382,
+      "step": 14605
+    },
+    {
+      "epoch": 2.865829737151824,
+      "grad_norm": 29.25151824951172,
+      "learning_rate": 1.8730497375218504e-06,
+      "loss": 4.4079,
+      "step": 14610
+    },
+    {
+      "epoch": 2.8668105139270303,
+      "grad_norm": 22.0438289642334,
+      "learning_rate": 1.87003049016308e-06,
+      "loss": 4.7012,
+      "step": 14615
+    },
+    {
+      "epoch": 2.867791290702236,
+      "grad_norm": 52.987953186035156,
+      "learning_rate": 1.8670131183607242e-06,
+      "loss": 4.2756,
+      "step": 14620
+    },
+    {
+      "epoch": 2.868772067477442,
+      "grad_norm": 18.41710090637207,
+      "learning_rate": 1.863997623922869e-06,
+      "loss": 4.7143,
+      "step": 14625
+    },
+    {
+      "epoch": 2.8697528442526483,
+      "grad_norm": 22.505746841430664,
+      "learning_rate": 1.8609840086564769e-06,
+      "loss": 4.4031,
+      "step": 14630
+    },
+    {
+      "epoch": 2.870733621027854,
+      "grad_norm": 35.385902404785156,
+      "learning_rate": 1.8579722743673773e-06,
+      "loss": 4.3598,
+      "step": 14635
+    },
+    {
+      "epoch": 2.87171439780306,
+      "grad_norm": 23.900920867919922,
+      "learning_rate": 1.8549624228602815e-06,
+      "loss": 4.5235,
+      "step": 14640
+    },
+    {
+      "epoch": 2.872695174578266,
+      "grad_norm": 22.860639572143555,
+      "learning_rate": 1.8519544559387642e-06,
+      "loss": 4.6174,
+      "step": 14645
+    },
+    {
+      "epoch": 2.873675951353472,
+      "grad_norm": 28.040868759155273,
+      "learning_rate": 1.8489483754052767e-06,
+      "loss": 4.4914,
+      "step": 14650
+    },
+    {
+      "epoch": 2.874656728128678,
+      "grad_norm": 21.25215721130371,
+      "learning_rate": 1.8459441830611402e-06,
+      "loss": 4.4763,
+      "step": 14655
+    },
+    {
+      "epoch": 2.875637504903884,
+      "grad_norm": 19.2554931640625,
+      "learning_rate": 1.8429418807065403e-06,
+      "loss": 4.5422,
+      "step": 14660
+    },
+    {
+      "epoch": 2.87661828167909,
+      "grad_norm": 17.228635787963867,
+      "learning_rate": 1.83994147014053e-06,
+      "loss": 4.385,
+      "step": 14665
+    },
+    {
+      "epoch": 2.8775990584542956,
+      "grad_norm": 27.80978775024414,
+      "learning_rate": 1.8369429531610339e-06,
+      "loss": 4.4902,
+      "step": 14670
+    },
+    {
+      "epoch": 2.878579835229502,
+      "grad_norm": 12.32872486114502,
+      "learning_rate": 1.8339463315648365e-06,
+      "loss": 4.3144,
+      "step": 14675
+    },
+    {
+      "epoch": 2.879560612004708,
+      "grad_norm": 18.406604766845703,
+      "learning_rate": 1.8309516071475909e-06,
+      "loss": 4.0658,
+      "step": 14680
+    },
+    {
+      "epoch": 2.8805413887799136,
+      "grad_norm": 21.1339168548584,
+      "learning_rate": 1.8279587817038086e-06,
+      "loss": 4.4005,
+      "step": 14685
+    },
+    {
+      "epoch": 2.8815221655551198,
+      "grad_norm": 17.803857803344727,
+      "learning_rate": 1.8249678570268697e-06,
+      "loss": 4.5032,
+      "step": 14690
+    },
+    {
+      "epoch": 2.8825029423303254,
+      "grad_norm": 32.61869812011719,
+      "learning_rate": 1.8219788349090067e-06,
+      "loss": 4.7084,
+      "step": 14695
+    },
+    {
+      "epoch": 2.8834837191055316,
+      "grad_norm": 23.566926956176758,
+      "learning_rate": 1.8189917171413196e-06,
+      "loss": 4.2739,
+      "step": 14700
+    },
+    {
+      "epoch": 2.8844644958807377,
+      "grad_norm": 19.303241729736328,
+      "learning_rate": 1.816006505513766e-06,
+      "loss": 4.4533,
+      "step": 14705
+    },
+    {
+      "epoch": 2.8854452726559434,
+      "grad_norm": 15.478785514831543,
+      "learning_rate": 1.8130232018151562e-06,
+      "loss": 4.9631,
+      "step": 14710
+    },
+    {
+      "epoch": 2.8864260494311496,
+      "grad_norm": 17.744274139404297,
+      "learning_rate": 1.8100418078331638e-06,
+      "loss": 4.298,
+      "step": 14715
+    },
+    {
+      "epoch": 2.8874068262063552,
+      "grad_norm": 26.17151641845703,
+      "learning_rate": 1.8070623253543118e-06,
+      "loss": 4.6553,
+      "step": 14720
+    },
+    {
+      "epoch": 2.8883876029815614,
+      "grad_norm": 16.9974308013916,
+      "learning_rate": 1.8040847561639834e-06,
+      "loss": 4.2575,
+      "step": 14725
+    },
+    {
+      "epoch": 2.8893683797567675,
+      "grad_norm": 16.582048416137695,
+      "learning_rate": 1.8011091020464138e-06,
+      "loss": 4.8634,
+      "step": 14730
+    },
+    {
+      "epoch": 2.890349156531973,
+      "grad_norm": 14.952964782714844,
+      "learning_rate": 1.7981353647846883e-06,
+      "loss": 4.4459,
+      "step": 14735
+    },
+    {
+      "epoch": 2.8913299333071794,
+      "grad_norm": 26.1202335357666,
+      "learning_rate": 1.7951635461607453e-06,
+      "loss": 4.4072,
+      "step": 14740
+    },
+    {
+      "epoch": 2.892310710082385,
+      "grad_norm": 27.30790138244629,
+      "learning_rate": 1.792193647955371e-06,
+      "loss": 4.7133,
+      "step": 14745
+    },
+    {
+      "epoch": 2.893291486857591,
+      "grad_norm": 14.769981384277344,
+      "learning_rate": 1.7892256719482053e-06,
+      "loss": 4.2343,
+      "step": 14750
+    },
+    {
+      "epoch": 2.8942722636327973,
+      "grad_norm": 21.78799819946289,
+      "learning_rate": 1.7862596199177351e-06,
+      "loss": 4.8337,
+      "step": 14755
+    },
+    {
+      "epoch": 2.895253040408003,
+      "grad_norm": 16.13930320739746,
+      "learning_rate": 1.783295493641291e-06,
+      "loss": 4.5237,
+      "step": 14760
+    },
+    {
+      "epoch": 2.896233817183209,
+      "grad_norm": 23.837434768676758,
+      "learning_rate": 1.7803332948950542e-06,
+      "loss": 4.5499,
+      "step": 14765
+    },
+    {
+      "epoch": 2.897214593958415,
+      "grad_norm": 20.946630477905273,
+      "learning_rate": 1.777373025454046e-06,
+      "loss": 4.4247,
+      "step": 14770
+    },
+    {
+      "epoch": 2.898195370733621,
+      "grad_norm": 41.92950439453125,
+      "learning_rate": 1.7744146870921357e-06,
+      "loss": 4.4211,
+      "step": 14775
+    },
+    {
+      "epoch": 2.899176147508827,
+      "grad_norm": 35.671695709228516,
+      "learning_rate": 1.7714582815820358e-06,
+      "loss": 4.3364,
+      "step": 14780
+    },
+    {
+      "epoch": 2.900156924284033,
+      "grad_norm": 15.399882316589355,
+      "learning_rate": 1.7685038106952952e-06,
+      "loss": 4.7295,
+      "step": 14785
+    },
+    {
+      "epoch": 2.901137701059239,
+      "grad_norm": 20.41010856628418,
+      "learning_rate": 1.7655512762023108e-06,
+      "loss": 4.4719,
+      "step": 14790
+    },
+    {
+      "epoch": 2.9021184778344447,
+      "grad_norm": 16.088794708251953,
+      "learning_rate": 1.7626006798723121e-06,
+      "loss": 4.2059,
+      "step": 14795
+    },
+    {
+      "epoch": 2.903099254609651,
+      "grad_norm": 25.732389450073242,
+      "learning_rate": 1.7596520234733739e-06,
+      "loss": 4.5793,
+      "step": 14800
+    },
+    {
+      "epoch": 2.904080031384857,
+      "grad_norm": 12.215232849121094,
+      "learning_rate": 1.7567053087724018e-06,
+      "loss": 4.7218,
+      "step": 14805
+    },
+    {
+      "epoch": 2.9050608081600626,
+      "grad_norm": 19.739543914794922,
+      "learning_rate": 1.7537605375351446e-06,
+      "loss": 4.4933,
+      "step": 14810
+    },
+    {
+      "epoch": 2.9060415849352688,
+      "grad_norm": 14.414162635803223,
+      "learning_rate": 1.7508177115261815e-06,
+      "loss": 4.9669,
+      "step": 14815
+    },
+    {
+      "epoch": 2.9070223617104745,
+      "grad_norm": 14.16154956817627,
+      "learning_rate": 1.7478768325089269e-06,
+      "loss": 4.6426,
+      "step": 14820
+    },
+    {
+      "epoch": 2.9080031384856806,
+      "grad_norm": 21.966236114501953,
+      "learning_rate": 1.7449379022456297e-06,
+      "loss": 4.4094,
+      "step": 14825
+    },
+    {
+      "epoch": 2.9089839152608867,
+      "grad_norm": 24.579914093017578,
+      "learning_rate": 1.7420009224973743e-06,
+      "loss": 4.5039,
+      "step": 14830
+    },
+    {
+      "epoch": 2.9099646920360924,
+      "grad_norm": 29.428955078125,
+      "learning_rate": 1.739065895024068e-06,
+      "loss": 4.3724,
+      "step": 14835
+    },
+    {
+      "epoch": 2.9109454688112986,
+      "grad_norm": 15.026422500610352,
+      "learning_rate": 1.736132821584457e-06,
+      "loss": 4.3833,
+      "step": 14840
+    },
+    {
+      "epoch": 2.9119262455865043,
+      "grad_norm": 35.190284729003906,
+      "learning_rate": 1.7332017039361094e-06,
+      "loss": 4.6888,
+      "step": 14845
+    },
+    {
+      "epoch": 2.9129070223617104,
+      "grad_norm": 31.889781951904297,
+      "learning_rate": 1.7302725438354256e-06,
+      "loss": 4.0796,
+      "step": 14850
+    },
+    {
+      "epoch": 2.9138877991369165,
+      "grad_norm": 19.924861907958984,
+      "learning_rate": 1.7273453430376347e-06,
+      "loss": 4.7306,
+      "step": 14855
+    },
+    {
+      "epoch": 2.9148685759121227,
+      "grad_norm": 29.968568801879883,
+      "learning_rate": 1.7244201032967844e-06,
+      "loss": 4.3262,
+      "step": 14860
+    },
+    {
+      "epoch": 2.9158493526873284,
+      "grad_norm": 16.26209831237793,
+      "learning_rate": 1.7214968263657561e-06,
+      "loss": 4.4238,
+      "step": 14865
+    },
+    {
+      "epoch": 2.916830129462534,
+      "grad_norm": 12.584505081176758,
+      "learning_rate": 1.7185755139962473e-06,
+      "loss": 4.6434,
+      "step": 14870
+    },
+    {
+      "epoch": 2.91781090623774,
+      "grad_norm": 13.820115089416504,
+      "learning_rate": 1.7156561679387851e-06,
+      "loss": 4.7432,
+      "step": 14875
+    },
+    {
+      "epoch": 2.9187916830129463,
+      "grad_norm": 23.35183334350586,
+      "learning_rate": 1.7127387899427118e-06,
+      "loss": 4.4374,
+      "step": 14880
+    },
+    {
+      "epoch": 2.9197724597881525,
+      "grad_norm": 10.54651165008545,
+      "learning_rate": 1.7098233817561966e-06,
+      "loss": 4.5436,
+      "step": 14885
+    },
+    {
+      "epoch": 2.920753236563358,
+      "grad_norm": 13.422380447387695,
+      "learning_rate": 1.7069099451262245e-06,
+      "loss": 4.7536,
+      "step": 14890
+    },
+    {
+      "epoch": 2.9217340133385643,
+      "grad_norm": 16.435741424560547,
+      "learning_rate": 1.703998481798597e-06,
+      "loss": 4.6658,
+      "step": 14895
+    },
+    {
+      "epoch": 2.92271479011377,
+      "grad_norm": 38.003475189208984,
+      "learning_rate": 1.7010889935179398e-06,
+      "loss": 4.7723,
+      "step": 14900
+    },
+    {
+      "epoch": 2.923695566888976,
+      "grad_norm": 18.305591583251953,
+      "learning_rate": 1.698181482027691e-06,
+      "loss": 4.6293,
+      "step": 14905
+    },
+    {
+      "epoch": 2.9246763436641823,
+      "grad_norm": 19.372703552246094,
+      "learning_rate": 1.6952759490701021e-06,
+      "loss": 4.5238,
+      "step": 14910
+    },
+    {
+      "epoch": 2.925657120439388,
+      "grad_norm": 15.01822280883789,
+      "learning_rate": 1.6923723963862455e-06,
+      "loss": 4.6348,
+      "step": 14915
+    },
+    {
+      "epoch": 2.926637897214594,
+      "grad_norm": 13.797202110290527,
+      "learning_rate": 1.689470825715998e-06,
+      "loss": 4.3862,
+      "step": 14920
+    },
+    {
+      "epoch": 2.9276186739898,
+      "grad_norm": 13.342180252075195,
+      "learning_rate": 1.686571238798057e-06,
+      "loss": 4.4863,
+      "step": 14925
+    },
+    {
+      "epoch": 2.928599450765006,
+      "grad_norm": 14.439830780029297,
+      "learning_rate": 1.6836736373699286e-06,
+      "loss": 4.6578,
+      "step": 14930
+    },
+    {
+      "epoch": 2.929580227540212,
+      "grad_norm": 14.736735343933105,
+      "learning_rate": 1.680778023167926e-06,
+      "loss": 4.4495,
+      "step": 14935
+    },
+    {
+      "epoch": 2.9305610043154178,
+      "grad_norm": 19.791828155517578,
+      "learning_rate": 1.677884397927176e-06,
+      "loss": 4.5876,
+      "step": 14940
+    },
+    {
+      "epoch": 2.931541781090624,
+      "grad_norm": 27.638660430908203,
+      "learning_rate": 1.6749927633816093e-06,
+      "loss": 4.9591,
+      "step": 14945
+    },
+    {
+      "epoch": 2.9325225578658296,
+      "grad_norm": 28.315677642822266,
+      "learning_rate": 1.6721031212639688e-06,
+      "loss": 4.6651,
+      "step": 14950
+    },
+    {
+      "epoch": 2.9335033346410357,
+      "grad_norm": 22.39848518371582,
+      "learning_rate": 1.6692154733057996e-06,
+      "loss": 4.3891,
+      "step": 14955
+    },
+    {
+      "epoch": 2.934484111416242,
+      "grad_norm": 23.636980056762695,
+      "learning_rate": 1.6663298212374508e-06,
+      "loss": 4.3512,
+      "step": 14960
+    },
+    {
+      "epoch": 2.9354648881914476,
+      "grad_norm": 17.69251251220703,
+      "learning_rate": 1.6634461667880807e-06,
+      "loss": 4.4783,
+      "step": 14965
+    },
+    {
+      "epoch": 2.9364456649666537,
+      "grad_norm": 16.90591812133789,
+      "learning_rate": 1.660564511685645e-06,
+      "loss": 4.7497,
+      "step": 14970
+    },
+    {
+      "epoch": 2.9374264417418594,
+      "grad_norm": 34.280216217041016,
+      "learning_rate": 1.6576848576569054e-06,
+      "loss": 4.5117,
+      "step": 14975
+    },
+    {
+      "epoch": 2.9384072185170655,
+      "grad_norm": 23.570716857910156,
+      "learning_rate": 1.654807206427424e-06,
+      "loss": 4.4814,
+      "step": 14980
+    },
+    {
+      "epoch": 2.9393879952922717,
+      "grad_norm": 13.759195327758789,
+      "learning_rate": 1.651931559721559e-06,
+      "loss": 4.3734,
+      "step": 14985
+    },
+    {
+      "epoch": 2.9403687720674774,
+      "grad_norm": 24.15625,
+      "learning_rate": 1.6490579192624734e-06,
+      "loss": 4.4103,
+      "step": 14990
+    },
+    {
+      "epoch": 2.9413495488426835,
+      "grad_norm": 26.735837936401367,
+      "learning_rate": 1.6461862867721218e-06,
+      "loss": 4.4619,
+      "step": 14995
+    },
+    {
+      "epoch": 2.942330325617889,
+      "grad_norm": 22.7193603515625,
+      "learning_rate": 1.6433166639712594e-06,
+      "loss": 4.5971,
+      "step": 15000
+    },
+    {
+      "epoch": 2.9433111023930953,
+      "grad_norm": 42.432960510253906,
+      "learning_rate": 1.6404490525794392e-06,
+      "loss": 4.5656,
+      "step": 15005
+    },
+    {
+      "epoch": 2.9442918791683015,
+      "grad_norm": 17.02939796447754,
+      "learning_rate": 1.6375834543150015e-06,
+      "loss": 5.0169,
+      "step": 15010
+    },
+    {
+      "epoch": 2.945272655943507,
+      "grad_norm": 35.580562591552734,
+      "learning_rate": 1.6347198708950884e-06,
+      "loss": 4.3484,
+      "step": 15015
+    },
+    {
+      "epoch": 2.9462534327187133,
+      "grad_norm": 28.986095428466797,
+      "learning_rate": 1.6318583040356285e-06,
+      "loss": 4.8381,
+      "step": 15020
+    },
+    {
+      "epoch": 2.947234209493919,
+      "grad_norm": 22.669565200805664,
+      "learning_rate": 1.6289987554513475e-06,
+      "loss": 4.2845,
+      "step": 15025
+    },
+    {
+      "epoch": 2.948214986269125,
+      "grad_norm": 17.930286407470703,
+      "learning_rate": 1.6261412268557564e-06,
+      "loss": 4.7354,
+      "step": 15030
+    },
+    {
+      "epoch": 2.9491957630443313,
+      "grad_norm": 16.47120475769043,
+      "learning_rate": 1.6232857199611579e-06,
+      "loss": 4.3747,
+      "step": 15035
+    },
+    {
+      "epoch": 2.950176539819537,
+      "grad_norm": 25.326810836791992,
+      "learning_rate": 1.6204322364786456e-06,
+      "loss": 4.6454,
+      "step": 15040
+    },
+    {
+      "epoch": 2.951157316594743,
+      "grad_norm": 30.15776252746582,
+      "learning_rate": 1.6175807781180964e-06,
+      "loss": 4.4318,
+      "step": 15045
+    },
+    {
+      "epoch": 2.952138093369949,
+      "grad_norm": 16.590038299560547,
+      "learning_rate": 1.6147313465881758e-06,
+      "loss": 4.4618,
+      "step": 15050
+    },
+    {
+      "epoch": 2.953118870145155,
+      "grad_norm": 17.120126724243164,
+      "learning_rate": 1.6118839435963386e-06,
+      "loss": 4.6643,
+      "step": 15055
+    },
+    {
+      "epoch": 2.954099646920361,
+      "grad_norm": 18.091629028320312,
+      "learning_rate": 1.6090385708488148e-06,
+      "loss": 4.6134,
+      "step": 15060
+    },
+    {
+      "epoch": 2.955080423695567,
+      "grad_norm": 27.78541374206543,
+      "learning_rate": 1.6061952300506285e-06,
+      "loss": 4.6667,
+      "step": 15065
+    },
+    {
+      "epoch": 2.956061200470773,
+      "grad_norm": 18.00626564025879,
+      "learning_rate": 1.6033539229055762e-06,
+      "loss": 4.5732,
+      "step": 15070
+    },
+    {
+      "epoch": 2.9570419772459786,
+      "grad_norm": 23.326290130615234,
+      "learning_rate": 1.6005146511162428e-06,
+      "loss": 4.4773,
+      "step": 15075
+    },
+    {
+      "epoch": 2.9580227540211848,
+      "grad_norm": 23.181758880615234,
+      "learning_rate": 1.5976774163839937e-06,
+      "loss": 4.5546,
+      "step": 15080
+    },
+    {
+      "epoch": 2.959003530796391,
+      "grad_norm": 33.70425796508789,
+      "learning_rate": 1.5948422204089664e-06,
+      "loss": 4.795,
+      "step": 15085
+    },
+    {
+      "epoch": 2.9599843075715966,
+      "grad_norm": 23.166748046875,
+      "learning_rate": 1.5920090648900866e-06,
+      "loss": 4.4982,
+      "step": 15090
+    },
+    {
+      "epoch": 2.9609650843468027,
+      "grad_norm": 17.901594161987305,
+      "learning_rate": 1.5891779515250494e-06,
+      "loss": 4.5362,
+      "step": 15095
+    },
+    {
+      "epoch": 2.9619458611220084,
+      "grad_norm": 15.403538703918457,
+      "learning_rate": 1.586348882010328e-06,
+      "loss": 4.4604,
+      "step": 15100
+    },
+    {
+      "epoch": 2.9629266378972146,
+      "grad_norm": 12.828426361083984,
+      "learning_rate": 1.583521858041175e-06,
+      "loss": 4.4609,
+      "step": 15105
+    },
+    {
+      "epoch": 2.9639074146724207,
+      "grad_norm": 34.96916961669922,
+      "learning_rate": 1.580696881311611e-06,
+      "loss": 4.6684,
+      "step": 15110
+    },
+    {
+      "epoch": 2.9648881914476264,
+      "grad_norm": 19.698801040649414,
+      "learning_rate": 1.5778739535144366e-06,
+      "loss": 4.7109,
+      "step": 15115
+    },
+    {
+      "epoch": 2.9658689682228325,
+      "grad_norm": 17.187973022460938,
+      "learning_rate": 1.5750530763412181e-06,
+      "loss": 4.3427,
+      "step": 15120
+    },
+    {
+      "epoch": 2.966849744998038,
+      "grad_norm": 16.712276458740234,
+      "learning_rate": 1.572234251482297e-06,
+      "loss": 4.5441,
+      "step": 15125
+    },
+    {
+      "epoch": 2.9678305217732444,
+      "grad_norm": 22.001623153686523,
+      "learning_rate": 1.5694174806267854e-06,
+      "loss": 4.4557,
+      "step": 15130
+    },
+    {
+      "epoch": 2.9688112985484505,
+      "grad_norm": 25.948938369750977,
+      "learning_rate": 1.566602765462561e-06,
+      "loss": 4.1687,
+      "step": 15135
+    },
+    {
+      "epoch": 2.969792075323656,
+      "grad_norm": 12.910621643066406,
+      "learning_rate": 1.5637901076762747e-06,
+      "loss": 4.6866,
+      "step": 15140
+    },
+    {
+      "epoch": 2.9707728520988623,
+      "grad_norm": 12.790865898132324,
+      "learning_rate": 1.560979508953338e-06,
+      "loss": 4.5784,
+      "step": 15145
+    },
+    {
+      "epoch": 2.971753628874068,
+      "grad_norm": 13.090932846069336,
+      "learning_rate": 1.5581709709779346e-06,
+      "loss": 4.3458,
+      "step": 15150
+    },
+    {
+      "epoch": 2.972734405649274,
+      "grad_norm": 18.293384552001953,
+      "learning_rate": 1.5553644954330122e-06,
+      "loss": 4.4609,
+      "step": 15155
+    },
+    {
+      "epoch": 2.9737151824244803,
+      "grad_norm": 21.8541202545166,
+      "learning_rate": 1.5525600840002785e-06,
+      "loss": 4.6097,
+      "step": 15160
+    },
+    {
+      "epoch": 2.974695959199686,
+      "grad_norm": 25.977567672729492,
+      "learning_rate": 1.549757738360211e-06,
+      "loss": 4.7444,
+      "step": 15165
+    },
+    {
+      "epoch": 2.975676735974892,
+      "grad_norm": 16.419042587280273,
+      "learning_rate": 1.546957460192043e-06,
+      "loss": 4.3684,
+      "step": 15170
+    },
+    {
+      "epoch": 2.976657512750098,
+      "grad_norm": 36.12578201293945,
+      "learning_rate": 1.5441592511737701e-06,
+      "loss": 4.4426,
+      "step": 15175
+    },
+    {
+      "epoch": 2.977638289525304,
+      "grad_norm": 44.24177932739258,
+      "learning_rate": 1.5413631129821544e-06,
+      "loss": 4.286,
+      "step": 15180
+    },
+    {
+      "epoch": 2.97861906630051,
+      "grad_norm": 15.310659408569336,
+      "learning_rate": 1.5385690472927067e-06,
+      "loss": 4.2994,
+      "step": 15185
+    },
+    {
+      "epoch": 2.9795998430757162,
+      "grad_norm": 17.0074405670166,
+      "learning_rate": 1.5357770557797064e-06,
+      "loss": 4.5799,
+      "step": 15190
+    },
+    {
+      "epoch": 2.980580619850922,
+      "grad_norm": 10.577085494995117,
+      "learning_rate": 1.5329871401161806e-06,
+      "loss": 4.3414,
+      "step": 15195
+    },
+    {
+      "epoch": 2.9815613966261276,
+      "grad_norm": 13.469268798828125,
+      "learning_rate": 1.5301993019739186e-06,
+      "loss": 3.8734,
+      "step": 15200
+    },
+    {
+      "epoch": 2.9825421734013338,
+      "grad_norm": 13.40593433380127,
+      "learning_rate": 1.5274135430234654e-06,
+      "loss": 4.4555,
+      "step": 15205
+    },
+    {
+      "epoch": 2.98352295017654,
+      "grad_norm": 34.490821838378906,
+      "learning_rate": 1.5246298649341146e-06,
+      "loss": 4.7537,
+      "step": 15210
+    },
+    {
+      "epoch": 2.984503726951746,
+      "grad_norm": 12.969808578491211,
+      "learning_rate": 1.5218482693739183e-06,
+      "loss": 5.0805,
+      "step": 15215
+    },
+    {
+      "epoch": 2.9854845037269517,
+      "grad_norm": 15.976425170898438,
+      "learning_rate": 1.5190687580096762e-06,
+      "loss": 4.5393,
+      "step": 15220
+    },
+    {
+      "epoch": 2.986465280502158,
+      "grad_norm": 30.466705322265625,
+      "learning_rate": 1.5162913325069428e-06,
+      "loss": 4.4184,
+      "step": 15225
+    },
+    {
+      "epoch": 2.9874460572773636,
+      "grad_norm": 17.97763442993164,
+      "learning_rate": 1.5135159945300232e-06,
+      "loss": 4.3463,
+      "step": 15230
+    },
+    {
+      "epoch": 2.9884268340525697,
+      "grad_norm": 23.519624710083008,
+      "learning_rate": 1.5107427457419654e-06,
+      "loss": 4.3618,
+      "step": 15235
+    },
+    {
+      "epoch": 2.989407610827776,
+      "grad_norm": 20.899688720703125,
+      "learning_rate": 1.5079715878045737e-06,
+      "loss": 4.6069,
+      "step": 15240
+    },
+    {
+      "epoch": 2.9903883876029815,
+      "grad_norm": 21.9375,
+      "learning_rate": 1.5052025223783944e-06,
+      "loss": 4.5154,
+      "step": 15245
+    },
+    {
+      "epoch": 2.9913691643781877,
+      "grad_norm": 11.83838176727295,
+      "learning_rate": 1.502435551122719e-06,
+      "loss": 4.493,
+      "step": 15250
+    },
+    {
+      "epoch": 2.9923499411533934,
+      "grad_norm": 18.852176666259766,
+      "learning_rate": 1.4996706756955892e-06,
+      "loss": 4.5604,
+      "step": 15255
+    },
+    {
+      "epoch": 2.9933307179285995,
+      "grad_norm": 13.880756378173828,
+      "learning_rate": 1.496907897753785e-06,
+      "loss": 4.6919,
+      "step": 15260
+    },
+    {
+      "epoch": 2.9943114947038056,
+      "grad_norm": 21.89354705810547,
+      "learning_rate": 1.4941472189528356e-06,
+      "loss": 4.6131,
+      "step": 15265
+    },
+    {
+      "epoch": 2.9952922714790113,
+      "grad_norm": 16.394149780273438,
+      "learning_rate": 1.4913886409470062e-06,
+      "loss": 4.3678,
+      "step": 15270
+    },
+    {
+      "epoch": 2.9962730482542175,
+      "grad_norm": 21.523550033569336,
+      "learning_rate": 1.488632165389307e-06,
+      "loss": 4.5263,
+      "step": 15275
+    },
+    {
+      "epoch": 2.997253825029423,
+      "grad_norm": 16.209943771362305,
+      "learning_rate": 1.48587779393149e-06,
+      "loss": 4.3487,
+      "step": 15280
+    },
+    {
+      "epoch": 2.9982346018046293,
+      "grad_norm": 17.297212600708008,
+      "learning_rate": 1.4831255282240397e-06,
+      "loss": 4.755,
+      "step": 15285
+    },
+    {
+      "epoch": 2.9992153785798354,
+      "grad_norm": 15.209367752075195,
+      "learning_rate": 1.4803753699161866e-06,
+      "loss": 4.4553,
+      "step": 15290
+    },
+    {
+      "epoch": 3.000196155355041,
+      "grad_norm": 13.265889167785645,
+      "learning_rate": 1.4776273206558911e-06,
+      "loss": 4.9951,
+      "step": 15295
+    },
+    {
+      "epoch": 3.0011769321302473,
+      "grad_norm": 28.354883193969727,
+      "learning_rate": 1.4748813820898554e-06,
+      "loss": 4.4493,
+      "step": 15300
+    },
+    {
+      "epoch": 3.0011769321302473,
+      "eval_loss": 4.852348804473877,
+      "eval_runtime": 7.8034,
+      "eval_samples_per_second": 26.783,
+      "eval_steps_per_second": 13.456,
+      "step": 15300
+    },
+    {
+      "epoch": 3.002157708905453,
+      "grad_norm": 24.23794174194336,
+      "learning_rate": 1.4721375558635164e-06,
+      "loss": 4.474,
+      "step": 15305
+    },
+    {
+      "epoch": 3.003138485680659,
+      "grad_norm": 11.600415229797363,
+      "learning_rate": 1.4693958436210426e-06,
+      "loss": 4.6386,
+      "step": 15310
+    },
+    {
+      "epoch": 3.0041192624558652,
+      "grad_norm": 30.11103057861328,
+      "learning_rate": 1.466656247005334e-06,
+      "loss": 4.722,
+      "step": 15315
+    },
+    {
+      "epoch": 3.005100039231071,
+      "grad_norm": 24.502426147460938,
+      "learning_rate": 1.4639187676580301e-06,
+      "loss": 4.5352,
+      "step": 15320
+    },
+    {
+      "epoch": 3.006080816006277,
+      "grad_norm": 11.984843254089355,
+      "learning_rate": 1.4611834072194948e-06,
+      "loss": 4.6266,
+      "step": 15325
+    },
+    {
+      "epoch": 3.0070615927814828,
+      "grad_norm": 15.542444229125977,
+      "learning_rate": 1.4584501673288259e-06,
+      "loss": 4.2322,
+      "step": 15330
+    },
+    {
+      "epoch": 3.008042369556689,
+      "grad_norm": 17.654617309570312,
+      "learning_rate": 1.4557190496238483e-06,
+      "loss": 4.5381,
+      "step": 15335
+    },
+    {
+      "epoch": 3.009023146331895,
+      "grad_norm": 29.028491973876953,
+      "learning_rate": 1.452990055741118e-06,
+      "loss": 4.1393,
+      "step": 15340
+    },
+    {
+      "epoch": 3.0100039231071007,
+      "grad_norm": 21.569608688354492,
+      "learning_rate": 1.4502631873159146e-06,
+      "loss": 4.378,
+      "step": 15345
+    },
+    {
+      "epoch": 3.010984699882307,
+      "grad_norm": 22.8667049407959,
+      "learning_rate": 1.4475384459822477e-06,
+      "loss": 4.4697,
+      "step": 15350
+    },
+    {
+      "epoch": 3.0119654766575126,
+      "grad_norm": 22.211538314819336,
+      "learning_rate": 1.444815833372852e-06,
+      "loss": 4.3405,
+      "step": 15355
+    },
+    {
+      "epoch": 3.0129462534327187,
+      "grad_norm": 19.6219539642334,
+      "learning_rate": 1.442095351119182e-06,
+      "loss": 4.4564,
+      "step": 15360
+    },
+    {
+      "epoch": 3.013927030207925,
+      "grad_norm": 28.267377853393555,
+      "learning_rate": 1.4393770008514235e-06,
+      "loss": 4.2524,
+      "step": 15365
+    },
+    {
+      "epoch": 3.0149078069831305,
+      "grad_norm": 12.272148132324219,
+      "learning_rate": 1.436660784198476e-06,
+      "loss": 4.3088,
+      "step": 15370
+    },
+    {
+      "epoch": 3.0158885837583367,
+      "grad_norm": 14.716875076293945,
+      "learning_rate": 1.4339467027879661e-06,
+      "loss": 4.7513,
+      "step": 15375
+    },
+    {
+      "epoch": 3.0168693605335424,
+      "grad_norm": 28.98792266845703,
+      "learning_rate": 1.4312347582462427e-06,
+      "loss": 4.2431,
+      "step": 15380
+    },
+    {
+      "epoch": 3.0178501373087485,
+      "grad_norm": 16.454082489013672,
+      "learning_rate": 1.428524952198368e-06,
+      "loss": 4.5905,
+      "step": 15385
+    },
+    {
+      "epoch": 3.0188309140839547,
+      "grad_norm": 22.37074851989746,
+      "learning_rate": 1.4258172862681268e-06,
+      "loss": 4.6151,
+      "step": 15390
+    },
+    {
+      "epoch": 3.0198116908591603,
+      "grad_norm": 29.454204559326172,
+      "learning_rate": 1.4231117620780188e-06,
+      "loss": 4.6649,
+      "step": 15395
+    },
+    {
+      "epoch": 3.0207924676343665,
+      "grad_norm": 21.081298828125,
+      "learning_rate": 1.4204083812492636e-06,
+      "loss": 4.4562,
+      "step": 15400
+    },
+    {
+      "epoch": 3.021773244409572,
+      "grad_norm": 14.012701988220215,
+      "learning_rate": 1.4177071454017966e-06,
+      "loss": 4.3996,
+      "step": 15405
+    },
+    {
+      "epoch": 3.0227540211847783,
+      "grad_norm": 23.03716278076172,
+      "learning_rate": 1.415008056154263e-06,
+      "loss": 4.4021,
+      "step": 15410
+    },
+    {
+      "epoch": 3.0237347979599845,
+      "grad_norm": 23.56743621826172,
+      "learning_rate": 1.4123111151240283e-06,
+      "loss": 4.7541,
+      "step": 15415
+    },
+    {
+      "epoch": 3.02471557473519,
+      "grad_norm": 14.281283378601074,
+      "learning_rate": 1.4096163239271638e-06,
+      "loss": 4.3043,
+      "step": 15420
+    },
+    {
+      "epoch": 3.0256963515103963,
+      "grad_norm": 34.98441696166992,
+      "learning_rate": 1.4069236841784584e-06,
+      "loss": 4.308,
+      "step": 15425
+    },
+    {
+      "epoch": 3.026677128285602,
+      "grad_norm": 12.55133056640625,
+      "learning_rate": 1.4042331974914103e-06,
+      "loss": 4.2404,
+      "step": 15430
+    },
+    {
+      "epoch": 3.027657905060808,
+      "grad_norm": 15.474343299865723,
+      "learning_rate": 1.4015448654782243e-06,
+      "loss": 4.756,
+      "step": 15435
+    },
+    {
+      "epoch": 3.0286386818360143,
+      "grad_norm": 23.840938568115234,
+      "learning_rate": 1.398858689749819e-06,
+      "loss": 4.5373,
+      "step": 15440
+    },
+    {
+      "epoch": 3.02961945861122,
+      "grad_norm": 21.256649017333984,
+      "learning_rate": 1.3961746719158158e-06,
+      "loss": 4.5571,
+      "step": 15445
+    },
+    {
+      "epoch": 3.030600235386426,
+      "grad_norm": 18.67331886291504,
+      "learning_rate": 1.3934928135845488e-06,
+      "loss": 4.4685,
+      "step": 15450
+    },
+    {
+      "epoch": 3.0315810121616322,
+      "grad_norm": 29.247690200805664,
+      "learning_rate": 1.3908131163630513e-06,
+      "loss": 4.6533,
+      "step": 15455
+    },
+    {
+      "epoch": 3.032561788936838,
+      "grad_norm": 18.554914474487305,
+      "learning_rate": 1.3881355818570691e-06,
+      "loss": 4.444,
+      "step": 15460
+    },
+    {
+      "epoch": 3.033542565712044,
+      "grad_norm": 42.42098617553711,
+      "learning_rate": 1.3854602116710459e-06,
+      "loss": 4.7042,
+      "step": 15465
+    },
+    {
+      "epoch": 3.0345233424872498,
+      "grad_norm": 38.945499420166016,
+      "learning_rate": 1.3827870074081296e-06,
+      "loss": 4.3739,
+      "step": 15470
+    },
+    {
+      "epoch": 3.035504119262456,
+      "grad_norm": 24.96856117248535,
+      "learning_rate": 1.3801159706701727e-06,
+      "loss": 4.1805,
+      "step": 15475
+    },
+    {
+      "epoch": 3.036484896037662,
+      "grad_norm": 15.618642807006836,
+      "learning_rate": 1.3774471030577298e-06,
+      "loss": 4.443,
+      "step": 15480
+    },
+    {
+      "epoch": 3.0374656728128677,
+      "grad_norm": 30.052515029907227,
+      "learning_rate": 1.3747804061700497e-06,
+      "loss": 4.2959,
+      "step": 15485
+    },
+    {
+      "epoch": 3.038446449588074,
+      "grad_norm": 28.82718849182129,
+      "learning_rate": 1.3721158816050872e-06,
+      "loss": 4.6018,
+      "step": 15490
+    },
+    {
+      "epoch": 3.0394272263632796,
+      "grad_norm": 17.50002098083496,
+      "learning_rate": 1.3694535309594903e-06,
+      "loss": 4.5142,
+      "step": 15495
+    },
+    {
+      "epoch": 3.0404080031384857,
+      "grad_norm": 28.0140438079834,
+      "learning_rate": 1.3667933558286067e-06,
+      "loss": 4.2015,
+      "step": 15500
+    },
+    {
+      "epoch": 3.041388779913692,
+      "grad_norm": 11.1800537109375,
+      "learning_rate": 1.3641353578064825e-06,
+      "loss": 4.4448,
+      "step": 15505
+    },
+    {
+      "epoch": 3.0423695566888975,
+      "grad_norm": 24.721168518066406,
+      "learning_rate": 1.3614795384858538e-06,
+      "loss": 4.4718,
+      "step": 15510
+    },
+    {
+      "epoch": 3.0433503334641037,
+      "grad_norm": 13.32351303100586,
+      "learning_rate": 1.3588258994581572e-06,
+      "loss": 4.8924,
+      "step": 15515
+    },
+    {
+      "epoch": 3.0443311102393094,
+      "grad_norm": 29.64592933654785,
+      "learning_rate": 1.3561744423135164e-06,
+      "loss": 4.4433,
+      "step": 15520
+    },
+    {
+      "epoch": 3.0453118870145155,
+      "grad_norm": 36.95175552368164,
+      "learning_rate": 1.3535251686407553e-06,
+      "loss": 4.1702,
+      "step": 15525
+    },
+    {
+      "epoch": 3.0462926637897216,
+      "grad_norm": 19.097549438476562,
+      "learning_rate": 1.3508780800273818e-06,
+      "loss": 4.1353,
+      "step": 15530
+    },
+    {
+      "epoch": 3.0472734405649273,
+      "grad_norm": 41.65642547607422,
+      "learning_rate": 1.3482331780596003e-06,
+      "loss": 4.7218,
+      "step": 15535
+    },
+    {
+      "epoch": 3.0482542173401335,
+      "grad_norm": 16.86134910583496,
+      "learning_rate": 1.3455904643223022e-06,
+      "loss": 4.5133,
+      "step": 15540
+    },
+    {
+      "epoch": 3.049234994115339,
+      "grad_norm": 13.893656730651855,
+      "learning_rate": 1.3429499403990658e-06,
+      "loss": 4.2403,
+      "step": 15545
+    },
+    {
+      "epoch": 3.0502157708905453,
+      "grad_norm": 15.907454490661621,
+      "learning_rate": 1.3403116078721606e-06,
+      "loss": 4.9011,
+      "step": 15550
+    },
+    {
+      "epoch": 3.0511965476657514,
+      "grad_norm": 14.247295379638672,
+      "learning_rate": 1.337675468322544e-06,
+      "loss": 4.5309,
+      "step": 15555
+    },
+    {
+      "epoch": 3.052177324440957,
+      "grad_norm": 24.694496154785156,
+      "learning_rate": 1.3350415233298542e-06,
+      "loss": 4.5389,
+      "step": 15560
+    },
+    {
+      "epoch": 3.0531581012161633,
+      "grad_norm": 33.52488327026367,
+      "learning_rate": 1.332409774472419e-06,
+      "loss": 5.0962,
+      "step": 15565
+    },
+    {
+      "epoch": 3.054138877991369,
+      "grad_norm": 16.590829849243164,
+      "learning_rate": 1.3297802233272473e-06,
+      "loss": 4.4915,
+      "step": 15570
+    },
+    {
+      "epoch": 3.055119654766575,
+      "grad_norm": 15.594505310058594,
+      "learning_rate": 1.3271528714700321e-06,
+      "loss": 4.5788,
+      "step": 15575
+    },
+    {
+      "epoch": 3.0561004315417812,
+      "grad_norm": 23.571426391601562,
+      "learning_rate": 1.3245277204751511e-06,
+      "loss": 4.4476,
+      "step": 15580
+    },
+    {
+      "epoch": 3.057081208316987,
+      "grad_norm": 11.0546236038208,
+      "learning_rate": 1.3219047719156575e-06,
+      "loss": 4.5721,
+      "step": 15585
+    },
+    {
+      "epoch": 3.058061985092193,
+      "grad_norm": 14.763998031616211,
+      "learning_rate": 1.3192840273632907e-06,
+      "loss": 4.628,
+      "step": 15590
+    },
+    {
+      "epoch": 3.0590427618673988,
+      "grad_norm": 20.179122924804688,
+      "learning_rate": 1.3166654883884643e-06,
+      "loss": 4.4939,
+      "step": 15595
+    },
+    {
+      "epoch": 3.060023538642605,
+      "grad_norm": 26.991134643554688,
+      "learning_rate": 1.314049156560276e-06,
+      "loss": 4.7653,
+      "step": 15600
+    },
+    {
+      "epoch": 3.061004315417811,
+      "grad_norm": 16.699296951293945,
+      "learning_rate": 1.3114350334464948e-06,
+      "loss": 4.3669,
+      "step": 15605
+    },
+    {
+      "epoch": 3.0619850921930167,
+      "grad_norm": 13.49559497833252,
+      "learning_rate": 1.308823120613568e-06,
+      "loss": 4.212,
+      "step": 15610
+    },
+    {
+      "epoch": 3.062965868968223,
+      "grad_norm": 26.69697380065918,
+      "learning_rate": 1.3062134196266235e-06,
+      "loss": 4.5457,
+      "step": 15615
+    },
+    {
+      "epoch": 3.063946645743429,
+      "grad_norm": 30.913908004760742,
+      "learning_rate": 1.303605932049456e-06,
+      "loss": 4.6255,
+      "step": 15620
+    },
+    {
+      "epoch": 3.0649274225186347,
+      "grad_norm": 26.240177154541016,
+      "learning_rate": 1.3010006594445384e-06,
+      "loss": 4.6354,
+      "step": 15625
+    },
+    {
+      "epoch": 3.065908199293841,
+      "grad_norm": 17.10197639465332,
+      "learning_rate": 1.2983976033730179e-06,
+      "loss": 4.5503,
+      "step": 15630
+    },
+    {
+      "epoch": 3.0668889760690465,
+      "grad_norm": 20.739166259765625,
+      "learning_rate": 1.2957967653947078e-06,
+      "loss": 4.4489,
+      "step": 15635
+    },
+    {
+      "epoch": 3.0678697528442527,
+      "grad_norm": 16.324705123901367,
+      "learning_rate": 1.2931981470680988e-06,
+      "loss": 4.414,
+      "step": 15640
+    },
+    {
+      "epoch": 3.068850529619459,
+      "grad_norm": 23.186613082885742,
+      "learning_rate": 1.2906017499503454e-06,
+      "loss": 4.4988,
+      "step": 15645
+    },
+    {
+      "epoch": 3.0698313063946645,
+      "grad_norm": 22.5463924407959,
+      "learning_rate": 1.288007575597275e-06,
+      "loss": 4.3282,
+      "step": 15650
+    },
+    {
+      "epoch": 3.0708120831698706,
+      "grad_norm": 12.31671142578125,
+      "learning_rate": 1.2854156255633837e-06,
+      "loss": 4.2331,
+      "step": 15655
+    },
+    {
+      "epoch": 3.0717928599450763,
+      "grad_norm": 21.361194610595703,
+      "learning_rate": 1.2828259014018308e-06,
+      "loss": 4.4753,
+      "step": 15660
+    },
+    {
+      "epoch": 3.0727736367202825,
+      "grad_norm": 11.575060844421387,
+      "learning_rate": 1.2802384046644468e-06,
+      "loss": 4.2673,
+      "step": 15665
+    },
+    {
+      "epoch": 3.0737544134954886,
+      "grad_norm": 13.329882621765137,
+      "learning_rate": 1.2776531369017215e-06,
+      "loss": 4.4678,
+      "step": 15670
+    },
+    {
+      "epoch": 3.0747351902706943,
+      "grad_norm": 27.24415397644043,
+      "learning_rate": 1.275070099662815e-06,
+      "loss": 4.4223,
+      "step": 15675
+    },
+    {
+      "epoch": 3.0757159670459004,
+      "grad_norm": 21.532888412475586,
+      "learning_rate": 1.272489294495548e-06,
+      "loss": 4.4301,
+      "step": 15680
+    },
+    {
+      "epoch": 3.076696743821106,
+      "grad_norm": 17.741239547729492,
+      "learning_rate": 1.2699107229464008e-06,
+      "loss": 4.4361,
+      "step": 15685
+    },
+    {
+      "epoch": 3.0776775205963123,
+      "grad_norm": 18.234464645385742,
+      "learning_rate": 1.2673343865605225e-06,
+      "loss": 4.3779,
+      "step": 15690
+    },
+    {
+      "epoch": 3.0786582973715184,
+      "grad_norm": 29.617433547973633,
+      "learning_rate": 1.264760286881715e-06,
+      "loss": 4.6656,
+      "step": 15695
+    },
+    {
+      "epoch": 3.079639074146724,
+      "grad_norm": 9.898058891296387,
+      "learning_rate": 1.2621884254524452e-06,
+      "loss": 4.8513,
+      "step": 15700
+    },
+    {
+      "epoch": 3.0806198509219302,
+      "grad_norm": 24.25059700012207,
+      "learning_rate": 1.2596188038138385e-06,
+      "loss": 4.5291,
+      "step": 15705
+    },
+    {
+      "epoch": 3.081600627697136,
+      "grad_norm": 19.24030113220215,
+      "learning_rate": 1.2570514235056735e-06,
+      "loss": 4.4476,
+      "step": 15710
+    },
+    {
+      "epoch": 3.082581404472342,
+      "grad_norm": 14.411766052246094,
+      "learning_rate": 1.254486286066393e-06,
+      "loss": 4.7423,
+      "step": 15715
+    },
+    {
+      "epoch": 3.083562181247548,
+      "grad_norm": 19.341766357421875,
+      "learning_rate": 1.2519233930330877e-06,
+      "loss": 4.4205,
+      "step": 15720
+    },
+    {
+      "epoch": 3.084542958022754,
+      "grad_norm": 26.251022338867188,
+      "learning_rate": 1.2493627459415096e-06,
+      "loss": 4.3927,
+      "step": 15725
+    },
+    {
+      "epoch": 3.08552373479796,
+      "grad_norm": 32.140380859375,
+      "learning_rate": 1.2468043463260649e-06,
+      "loss": 4.6213,
+      "step": 15730
+    },
+    {
+      "epoch": 3.0865045115731657,
+      "grad_norm": 21.708030700683594,
+      "learning_rate": 1.2442481957198066e-06,
+      "loss": 4.256,
+      "step": 15735
+    },
+    {
+      "epoch": 3.087485288348372,
+      "grad_norm": 20.06285285949707,
+      "learning_rate": 1.2416942956544486e-06,
+      "loss": 4.4784,
+      "step": 15740
+    },
+    {
+      "epoch": 3.088466065123578,
+      "grad_norm": 17.960481643676758,
+      "learning_rate": 1.2391426476603496e-06,
+      "loss": 4.4885,
+      "step": 15745
+    },
+    {
+      "epoch": 3.0894468418987837,
+      "grad_norm": 19.873380661010742,
+      "learning_rate": 1.236593253266521e-06,
+      "loss": 4.5705,
+      "step": 15750
+    },
+    {
+      "epoch": 3.09042761867399,
+      "grad_norm": 17.436277389526367,
+      "learning_rate": 1.2340461140006255e-06,
+      "loss": 4.1984,
+      "step": 15755
+    },
+    {
+      "epoch": 3.0914083954491955,
+      "grad_norm": 19.593332290649414,
+      "learning_rate": 1.2315012313889708e-06,
+      "loss": 4.3993,
+      "step": 15760
+    },
+    {
+      "epoch": 3.0923891722244017,
+      "grad_norm": 29.571025848388672,
+      "learning_rate": 1.2289586069565174e-06,
+      "loss": 4.6506,
+      "step": 15765
+    },
+    {
+      "epoch": 3.093369948999608,
+      "grad_norm": 15.669720649719238,
+      "learning_rate": 1.2264182422268673e-06,
+      "loss": 4.4139,
+      "step": 15770
+    },
+    {
+      "epoch": 3.0943507257748135,
+      "grad_norm": 12.525907516479492,
+      "learning_rate": 1.2238801387222716e-06,
+      "loss": 4.2745,
+      "step": 15775
+    },
+    {
+      "epoch": 3.0953315025500197,
+      "grad_norm": 31.15891456604004,
+      "learning_rate": 1.221344297963627e-06,
+      "loss": 4.3835,
+      "step": 15780
+    },
+    {
+      "epoch": 3.096312279325226,
+      "grad_norm": 14.809552192687988,
+      "learning_rate": 1.2188107214704714e-06,
+      "loss": 4.3162,
+      "step": 15785
+    },
+    {
+      "epoch": 3.0972930561004315,
+      "grad_norm": 17.6668758392334,
+      "learning_rate": 1.2162794107609888e-06,
+      "loss": 4.1695,
+      "step": 15790
+    },
+    {
+      "epoch": 3.0982738328756376,
+      "grad_norm": 16.730649948120117,
+      "learning_rate": 1.213750367352003e-06,
+      "loss": 4.2216,
+      "step": 15795
+    },
+    {
+      "epoch": 3.0992546096508433,
+      "grad_norm": 19.196990966796875,
+      "learning_rate": 1.2112235927589805e-06,
+      "loss": 4.4758,
+      "step": 15800
+    },
+    {
+      "epoch": 3.1002353864260495,
+      "grad_norm": 18.599266052246094,
+      "learning_rate": 1.2086990884960304e-06,
+      "loss": 4.4006,
+      "step": 15805
+    },
+    {
+      "epoch": 3.1012161632012556,
+      "grad_norm": 17.576534271240234,
+      "learning_rate": 1.2061768560758957e-06,
+      "loss": 4.4374,
+      "step": 15810
+    },
+    {
+      "epoch": 3.1021969399764613,
+      "grad_norm": 21.37404441833496,
+      "learning_rate": 1.2036568970099643e-06,
+      "loss": 4.4077,
+      "step": 15815
+    },
+    {
+      "epoch": 3.1031777167516674,
+      "grad_norm": 10.35575008392334,
+      "learning_rate": 1.2011392128082583e-06,
+      "loss": 4.3657,
+      "step": 15820
+    },
+    {
+      "epoch": 3.104158493526873,
+      "grad_norm": 16.843027114868164,
+      "learning_rate": 1.1986238049794352e-06,
+      "loss": 4.5244,
+      "step": 15825
+    },
+    {
+      "epoch": 3.1051392703020793,
+      "grad_norm": 8.886418342590332,
+      "learning_rate": 1.1961106750307945e-06,
+      "loss": 4.2819,
+      "step": 15830
+    },
+    {
+      "epoch": 3.1061200470772854,
+      "grad_norm": 18.96690559387207,
+      "learning_rate": 1.1935998244682624e-06,
+      "loss": 4.6573,
+      "step": 15835
+    },
+    {
+      "epoch": 3.107100823852491,
+      "grad_norm": 20.823482513427734,
+      "learning_rate": 1.1910912547964076e-06,
+      "loss": 4.5941,
+      "step": 15840
+    },
+    {
+      "epoch": 3.1080816006276972,
+      "grad_norm": 11.613541603088379,
+      "learning_rate": 1.1885849675184252e-06,
+      "loss": 4.3887,
+      "step": 15845
+    },
+    {
+      "epoch": 3.109062377402903,
+      "grad_norm": 12.627098083496094,
+      "learning_rate": 1.186080964136147e-06,
+      "loss": 4.5249,
+      "step": 15850
+    },
+    {
+      "epoch": 3.110043154178109,
+      "grad_norm": 14.541468620300293,
+      "learning_rate": 1.1835792461500357e-06,
+      "loss": 4.2469,
+      "step": 15855
+    },
+    {
+      "epoch": 3.111023930953315,
+      "grad_norm": 15.33710765838623,
+      "learning_rate": 1.1810798150591813e-06,
+      "loss": 4.3271,
+      "step": 15860
+    },
+    {
+      "epoch": 3.112004707728521,
+      "grad_norm": 13.666189193725586,
+      "learning_rate": 1.1785826723613081e-06,
+      "loss": 4.2686,
+      "step": 15865
+    },
+    {
+      "epoch": 3.112985484503727,
+      "grad_norm": 18.96344566345215,
+      "learning_rate": 1.1760878195527642e-06,
+      "loss": 4.3628,
+      "step": 15870
+    },
+    {
+      "epoch": 3.1139662612789327,
+      "grad_norm": 14.437033653259277,
+      "learning_rate": 1.1735952581285299e-06,
+      "loss": 4.4556,
+      "step": 15875
+    },
+    {
+      "epoch": 3.114947038054139,
+      "grad_norm": 20.044998168945312,
+      "learning_rate": 1.1711049895822114e-06,
+      "loss": 4.3907,
+      "step": 15880
+    },
+    {
+      "epoch": 3.115927814829345,
+      "grad_norm": 26.939075469970703,
+      "learning_rate": 1.1686170154060379e-06,
+      "loss": 4.3439,
+      "step": 15885
+    },
+    {
+      "epoch": 3.1169085916045507,
+      "grad_norm": 32.64443588256836,
+      "learning_rate": 1.1661313370908689e-06,
+      "loss": 4.3217,
+      "step": 15890
+    },
+    {
+      "epoch": 3.117889368379757,
+      "grad_norm": 17.43419075012207,
+      "learning_rate": 1.1636479561261832e-06,
+      "loss": 4.7089,
+      "step": 15895
+    },
+    {
+      "epoch": 3.1188701451549625,
+      "grad_norm": 17.65254783630371,
+      "learning_rate": 1.1611668740000848e-06,
+      "loss": 4.3605,
+      "step": 15900
+    },
+    {
+      "epoch": 3.1198509219301687,
+      "grad_norm": 19.919321060180664,
+      "learning_rate": 1.1586880921993022e-06,
+      "loss": 4.3343,
+      "step": 15905
+    },
+    {
+      "epoch": 3.120831698705375,
+      "grad_norm": 31.183734893798828,
+      "learning_rate": 1.156211612209182e-06,
+      "loss": 4.1166,
+      "step": 15910
+    },
+    {
+      "epoch": 3.1218124754805805,
+      "grad_norm": 20.650432586669922,
+      "learning_rate": 1.1537374355136954e-06,
+      "loss": 4.5356,
+      "step": 15915
+    },
+    {
+      "epoch": 3.1227932522557866,
+      "grad_norm": 25.45415687561035,
+      "learning_rate": 1.1512655635954284e-06,
+      "loss": 4.3773,
+      "step": 15920
+    },
+    {
+      "epoch": 3.1237740290309928,
+      "grad_norm": 30.577579498291016,
+      "learning_rate": 1.1487959979355906e-06,
+      "loss": 4.1834,
+      "step": 15925
+    },
+    {
+      "epoch": 3.1247548058061985,
+      "grad_norm": 21.581764221191406,
+      "learning_rate": 1.1463287400140089e-06,
+      "loss": 4.4372,
+      "step": 15930
+    },
+    {
+      "epoch": 3.1257355825814046,
+      "grad_norm": 24.099647521972656,
+      "learning_rate": 1.1438637913091238e-06,
+      "loss": 4.4539,
+      "step": 15935
+    },
+    {
+      "epoch": 3.1267163593566103,
+      "grad_norm": 14.903850555419922,
+      "learning_rate": 1.1414011532979975e-06,
+      "loss": 4.3897,
+      "step": 15940
+    },
+    {
+      "epoch": 3.1276971361318164,
+      "grad_norm": 21.976409912109375,
+      "learning_rate": 1.1389408274563013e-06,
+      "loss": 4.6149,
+      "step": 15945
+    },
+    {
+      "epoch": 3.1286779129070226,
+      "grad_norm": 26.018178939819336,
+      "learning_rate": 1.1364828152583252e-06,
+      "loss": 4.7381,
+      "step": 15950
+    },
+    {
+      "epoch": 3.1296586896822283,
+      "grad_norm": 19.22270965576172,
+      "learning_rate": 1.1340271181769746e-06,
+      "loss": 4.3429,
+      "step": 15955
+    },
+    {
+      "epoch": 3.1306394664574344,
+      "grad_norm": 24.88014030456543,
+      "learning_rate": 1.1315737376837627e-06,
+      "loss": 4.3767,
+      "step": 15960
+    },
+    {
+      "epoch": 3.13162024323264,
+      "grad_norm": 15.811385154724121,
+      "learning_rate": 1.129122675248816e-06,
+      "loss": 4.2861,
+      "step": 15965
+    },
+    {
+      "epoch": 3.1326010200078462,
+      "grad_norm": 17.602039337158203,
+      "learning_rate": 1.1266739323408743e-06,
+      "loss": 4.6436,
+      "step": 15970
+    },
+    {
+      "epoch": 3.1335817967830524,
+      "grad_norm": 17.281982421875,
+      "learning_rate": 1.124227510427286e-06,
+      "loss": 4.6336,
+      "step": 15975
+    },
+    {
+      "epoch": 3.134562573558258,
+      "grad_norm": 10.94583511352539,
+      "learning_rate": 1.1217834109740061e-06,
+      "loss": 3.9648,
+      "step": 15980
+    },
+    {
+      "epoch": 3.135543350333464,
+      "grad_norm": 29.885208129882812,
+      "learning_rate": 1.1193416354456022e-06,
+      "loss": 4.4361,
+      "step": 15985
+    },
+    {
+      "epoch": 3.13652412710867,
+      "grad_norm": 30.1024112701416,
+      "learning_rate": 1.1169021853052491e-06,
+      "loss": 4.9684,
+      "step": 15990
+    },
+    {
+      "epoch": 3.137504903883876,
+      "grad_norm": 18.619029998779297,
+      "learning_rate": 1.114465062014724e-06,
+      "loss": 4.654,
+      "step": 15995
+    },
+    {
+      "epoch": 3.138485680659082,
+      "grad_norm": 17.20469093322754,
+      "learning_rate": 1.1120302670344153e-06,
+      "loss": 4.5966,
+      "step": 16000
+    },
+    {
+      "epoch": 3.139466457434288,
+      "grad_norm": 18.83102035522461,
+      "learning_rate": 1.1095978018233094e-06,
+      "loss": 4.4428,
+      "step": 16005
+    },
+    {
+      "epoch": 3.140447234209494,
+      "grad_norm": 23.325239181518555,
+      "learning_rate": 1.1071676678390036e-06,
+      "loss": 4.3781,
+      "step": 16010
+    },
+    {
+      "epoch": 3.1414280109846997,
+      "grad_norm": 25.824480056762695,
+      "learning_rate": 1.1047398665376956e-06,
+      "loss": 4.5236,
+      "step": 16015
+    },
+    {
+      "epoch": 3.142408787759906,
+      "grad_norm": 21.332595825195312,
+      "learning_rate": 1.102314399374183e-06,
+      "loss": 4.5056,
+      "step": 16020
+    },
+    {
+      "epoch": 3.143389564535112,
+      "grad_norm": 21.27334213256836,
+      "learning_rate": 1.0998912678018685e-06,
+      "loss": 4.3762,
+      "step": 16025
+    },
+    {
+      "epoch": 3.1443703413103177,
+      "grad_norm": 18.111543655395508,
+      "learning_rate": 1.0974704732727514e-06,
+      "loss": 4.5658,
+      "step": 16030
+    },
+    {
+      "epoch": 3.145351118085524,
+      "grad_norm": 15.349037170410156,
+      "learning_rate": 1.0950520172374352e-06,
+      "loss": 4.2495,
+      "step": 16035
+    },
+    {
+      "epoch": 3.1463318948607295,
+      "grad_norm": 13.66447639465332,
+      "learning_rate": 1.0926359011451182e-06,
+      "loss": 4.8234,
+      "step": 16040
+    },
+    {
+      "epoch": 3.1473126716359356,
+      "grad_norm": 24.965930938720703,
+      "learning_rate": 1.0902221264435964e-06,
+      "loss": 4.3959,
+      "step": 16045
+    },
+    {
+      "epoch": 3.148293448411142,
+      "grad_norm": 14.090888977050781,
+      "learning_rate": 1.0878106945792676e-06,
+      "loss": 4.2983,
+      "step": 16050
+    },
+    {
+      "epoch": 3.1492742251863475,
+      "grad_norm": 22.15886878967285,
+      "learning_rate": 1.0854016069971184e-06,
+      "loss": 4.4799,
+      "step": 16055
+    },
+    {
+      "epoch": 3.1502550019615536,
+      "grad_norm": 38.32559585571289,
+      "learning_rate": 1.0829948651407374e-06,
+      "loss": 4.4311,
+      "step": 16060
+    },
+    {
+      "epoch": 3.1512357787367593,
+      "grad_norm": 40.12593460083008,
+      "learning_rate": 1.0805904704523057e-06,
+      "loss": 4.4074,
+      "step": 16065
+    },
+    {
+      "epoch": 3.1522165555119654,
+      "grad_norm": 13.718966484069824,
+      "learning_rate": 1.0781884243725937e-06,
+      "loss": 4.3726,
+      "step": 16070
+    },
+    {
+      "epoch": 3.1531973322871716,
+      "grad_norm": 14.995040893554688,
+      "learning_rate": 1.0757887283409718e-06,
+      "loss": 4.3585,
+      "step": 16075
+    },
+    {
+      "epoch": 3.1541781090623773,
+      "grad_norm": 17.867856979370117,
+      "learning_rate": 1.0733913837953942e-06,
+      "loss": 4.4938,
+      "step": 16080
+    },
+    {
+      "epoch": 3.1551588858375834,
+      "grad_norm": 30.474164962768555,
+      "learning_rate": 1.0709963921724115e-06,
+      "loss": 4.4126,
+      "step": 16085
+    },
+    {
+      "epoch": 3.156139662612789,
+      "grad_norm": 25.54427719116211,
+      "learning_rate": 1.0686037549071648e-06,
+      "loss": 4.7638,
+      "step": 16090
+    },
+    {
+      "epoch": 3.1571204393879952,
+      "grad_norm": 23.775362014770508,
+      "learning_rate": 1.06621347343338e-06,
+      "loss": 4.4845,
+      "step": 16095
+    },
+    {
+      "epoch": 3.1581012161632014,
+      "grad_norm": 23.316543579101562,
+      "learning_rate": 1.063825549183376e-06,
+      "loss": 4.5359,
+      "step": 16100
+    },
+    {
+      "epoch": 3.159081992938407,
+      "grad_norm": 24.66636085510254,
+      "learning_rate": 1.0614399835880545e-06,
+      "loss": 4.537,
+      "step": 16105
+    },
+    {
+      "epoch": 3.160062769713613,
+      "grad_norm": 16.905139923095703,
+      "learning_rate": 1.059056778076909e-06,
+      "loss": 4.4013,
+      "step": 16110
+    },
+    {
+      "epoch": 3.161043546488819,
+      "grad_norm": 22.527822494506836,
+      "learning_rate": 1.0566759340780153e-06,
+      "loss": 4.7759,
+      "step": 16115
+    },
+    {
+      "epoch": 3.162024323264025,
+      "grad_norm": 25.074085235595703,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 4.5525,
+      "step": 16120
+    },
+    {
+      "epoch": 3.163005100039231,
+      "grad_norm": 22.77043342590332,
+      "learning_rate": 1.0519213363222102e-06,
+      "loss": 4.4541,
+      "step": 16125
+    },
+    {
+      "epoch": 3.163985876814437,
+      "grad_norm": 21.752971649169922,
+      "learning_rate": 1.0495475854143738e-06,
+      "loss": 4.0442,
+      "step": 16130
+    },
+    {
+      "epoch": 3.164966653589643,
+      "grad_norm": 33.8589973449707,
+      "learning_rate": 1.0471762017169362e-06,
+      "loss": 4.3213,
+      "step": 16135
+    },
+    {
+      "epoch": 3.165947430364849,
+      "grad_norm": 12.692889213562012,
+      "learning_rate": 1.0448071866508914e-06,
+      "loss": 4.8471,
+      "step": 16140
+    },
+    {
+      "epoch": 3.166928207140055,
+      "grad_norm": 22.040273666381836,
+      "learning_rate": 1.0424405416358096e-06,
+      "loss": 4.3255,
+      "step": 16145
+    },
+    {
+      "epoch": 3.167908983915261,
+      "grad_norm": 32.375091552734375,
+      "learning_rate": 1.0400762680898474e-06,
+      "loss": 4.4573,
+      "step": 16150
+    },
+    {
+      "epoch": 3.1688897606904667,
+      "grad_norm": 16.68021583557129,
+      "learning_rate": 1.037714367429734e-06,
+      "loss": 4.3776,
+      "step": 16155
+    },
+    {
+      "epoch": 3.169870537465673,
+      "grad_norm": 27.631816864013672,
+      "learning_rate": 1.0353548410707815e-06,
+      "loss": 4.7017,
+      "step": 16160
+    },
+    {
+      "epoch": 3.170851314240879,
+      "grad_norm": 21.348148345947266,
+      "learning_rate": 1.0329976904268773e-06,
+      "loss": 4.6062,
+      "step": 16165
+    },
+    {
+      "epoch": 3.1718320910160847,
+      "grad_norm": 23.80921745300293,
+      "learning_rate": 1.0306429169104841e-06,
+      "loss": 4.4043,
+      "step": 16170
+    },
+    {
+      "epoch": 3.172812867791291,
+      "grad_norm": 20.834857940673828,
+      "learning_rate": 1.0282905219326438e-06,
+      "loss": 4.5575,
+      "step": 16175
+    },
+    {
+      "epoch": 3.1737936445664965,
+      "grad_norm": 20.40863609313965,
+      "learning_rate": 1.0259405069029672e-06,
+      "loss": 4.3255,
+      "step": 16180
+    },
+    {
+      "epoch": 3.1747744213417026,
+      "grad_norm": 12.205023765563965,
+      "learning_rate": 1.0235928732296458e-06,
+      "loss": 4.2396,
+      "step": 16185
+    },
+    {
+      "epoch": 3.1757551981169088,
+      "grad_norm": 19.171253204345703,
+      "learning_rate": 1.021247622319439e-06,
+      "loss": 4.7084,
+      "step": 16190
+    },
+    {
+      "epoch": 3.1767359748921145,
+      "grad_norm": 16.924537658691406,
+      "learning_rate": 1.0189047555776787e-06,
+      "loss": 4.2925,
+      "step": 16195
+    },
+    {
+      "epoch": 3.1777167516673206,
+      "grad_norm": 15.352378845214844,
+      "learning_rate": 1.0165642744082726e-06,
+      "loss": 4.4434,
+      "step": 16200
+    },
+    {
+      "epoch": 3.1786975284425263,
+      "grad_norm": 19.03781509399414,
+      "learning_rate": 1.0142261802136931e-06,
+      "loss": 4.5886,
+      "step": 16205
+    },
+    {
+      "epoch": 3.1796783052177324,
+      "grad_norm": 18.974306106567383,
+      "learning_rate": 1.0118904743949865e-06,
+      "loss": 4.8805,
+      "step": 16210
+    },
+    {
+      "epoch": 3.1806590819929386,
+      "grad_norm": 23.474401473999023,
+      "learning_rate": 1.0095571583517665e-06,
+      "loss": 4.3292,
+      "step": 16215
+    },
+    {
+      "epoch": 3.1816398587681443,
+      "grad_norm": 17.590652465820312,
+      "learning_rate": 1.0072262334822142e-06,
+      "loss": 4.3978,
+      "step": 16220
+    },
+    {
+      "epoch": 3.1826206355433504,
+      "grad_norm": 21.562280654907227,
+      "learning_rate": 1.0048977011830791e-06,
+      "loss": 4.1862,
+      "step": 16225
+    },
+    {
+      "epoch": 3.183601412318556,
+      "grad_norm": 35.704036712646484,
+      "learning_rate": 1.0025715628496752e-06,
+      "loss": 4.7851,
+      "step": 16230
+    },
+    {
+      "epoch": 3.1845821890937622,
+      "grad_norm": 11.571843147277832,
+      "learning_rate": 1.000247819875883e-06,
+      "loss": 4.6119,
+      "step": 16235
+    },
+    {
+      "epoch": 3.1855629658689684,
+      "grad_norm": 39.009552001953125,
+      "learning_rate": 9.9792647365415e-07,
+      "loss": 4.5718,
+      "step": 16240
+    },
+    {
+      "epoch": 3.186543742644174,
+      "grad_norm": 17.692380905151367,
+      "learning_rate": 9.956075255754822e-07,
+      "loss": 4.449,
+      "step": 16245
+    },
+    {
+      "epoch": 3.18752451941938,
+      "grad_norm": 14.382763862609863,
+      "learning_rate": 9.932909770294542e-07,
+      "loss": 4.4584,
+      "step": 16250
+    },
+    {
+      "epoch": 3.1885052961945863,
+      "grad_norm": 17.71790313720703,
+      "learning_rate": 9.909768294041989e-07,
+      "loss": 4.5061,
+      "step": 16255
+    },
+    {
+      "epoch": 3.189486072969792,
+      "grad_norm": 24.063451766967773,
+      "learning_rate": 9.886650840864104e-07,
+      "loss": 4.3325,
+      "step": 16260
+    },
+    {
+      "epoch": 3.190466849744998,
+      "grad_norm": 14.177840232849121,
+      "learning_rate": 9.863557424613473e-07,
+      "loss": 4.5417,
+      "step": 16265
+    },
+    {
+      "epoch": 3.191447626520204,
+      "grad_norm": 32.26413345336914,
+      "learning_rate": 9.840488059128228e-07,
+      "loss": 4.4526,
+      "step": 16270
+    },
+    {
+      "epoch": 3.19242840329541,
+      "grad_norm": 30.41230010986328,
+      "learning_rate": 9.817442758232132e-07,
+      "loss": 4.8116,
+      "step": 16275
+    },
+    {
+      "epoch": 3.193409180070616,
+      "grad_norm": 22.243484497070312,
+      "learning_rate": 9.79442153573449e-07,
+      "loss": 4.3898,
+      "step": 16280
+    },
+    {
+      "epoch": 3.194389956845822,
+      "grad_norm": 22.45796775817871,
+      "learning_rate": 9.771424405430196e-07,
+      "loss": 4.5671,
+      "step": 16285
+    },
+    {
+      "epoch": 3.195370733621028,
+      "grad_norm": 17.344751358032227,
+      "learning_rate": 9.748451381099743e-07,
+      "loss": 4.6885,
+      "step": 16290
+    },
+    {
+      "epoch": 3.1963515103962337,
+      "grad_norm": 20.54185676574707,
+      "learning_rate": 9.725502476509102e-07,
+      "loss": 4.5413,
+      "step": 16295
+    },
+    {
+      "epoch": 3.19733228717144,
+      "grad_norm": 19.294301986694336,
+      "learning_rate": 9.702577705409872e-07,
+      "loss": 4.2378,
+      "step": 16300
+    },
+    {
+      "epoch": 3.198313063946646,
+      "grad_norm": 31.30463218688965,
+      "learning_rate": 9.679677081539112e-07,
+      "loss": 4.6142,
+      "step": 16305
+    },
+    {
+      "epoch": 3.1992938407218516,
+      "grad_norm": 27.021711349487305,
+      "learning_rate": 9.656800618619478e-07,
+      "loss": 4.2349,
+      "step": 16310
+    },
+    {
+      "epoch": 3.2002746174970578,
+      "grad_norm": 20.498126983642578,
+      "learning_rate": 9.633948330359122e-07,
+      "loss": 4.4625,
+      "step": 16315
+    },
+    {
+      "epoch": 3.2012553942722635,
+      "grad_norm": 16.401073455810547,
+      "learning_rate": 9.611120230451698e-07,
+      "loss": 4.6958,
+      "step": 16320
+    },
+    {
+      "epoch": 3.2022361710474696,
+      "grad_norm": 21.446414947509766,
+      "learning_rate": 9.588316332576392e-07,
+      "loss": 4.8353,
+      "step": 16325
+    },
+    {
+      "epoch": 3.2032169478226757,
+      "grad_norm": 30.655405044555664,
+      "learning_rate": 9.56553665039786e-07,
+      "loss": 4.0624,
+      "step": 16330
+    },
+    {
+      "epoch": 3.2041977245978814,
+      "grad_norm": 9.359758377075195,
+      "learning_rate": 9.542781197566254e-07,
+      "loss": 4.4915,
+      "step": 16335
+    },
+    {
+      "epoch": 3.2051785013730876,
+      "grad_norm": 13.281656265258789,
+      "learning_rate": 9.52004998771724e-07,
+      "loss": 4.5648,
+      "step": 16340
+    },
+    {
+      "epoch": 3.2061592781482933,
+      "grad_norm": 16.86419677734375,
+      "learning_rate": 9.497343034471896e-07,
+      "loss": 4.3901,
+      "step": 16345
+    },
+    {
+      "epoch": 3.2071400549234994,
+      "grad_norm": 13.59611701965332,
+      "learning_rate": 9.474660351436832e-07,
+      "loss": 4.4795,
+      "step": 16350
+    },
+    {
+      "epoch": 3.2081208316987055,
+      "grad_norm": 19.933231353759766,
+      "learning_rate": 9.452001952204049e-07,
+      "loss": 4.3226,
+      "step": 16355
+    },
+    {
+      "epoch": 3.2091016084739112,
+      "grad_norm": 10.386981010437012,
+      "learning_rate": 9.429367850351051e-07,
+      "loss": 4.2176,
+      "step": 16360
+    },
+    {
+      "epoch": 3.2100823852491174,
+      "grad_norm": 14.15022087097168,
+      "learning_rate": 9.406758059440774e-07,
+      "loss": 4.3851,
+      "step": 16365
+    },
+    {
+      "epoch": 3.211063162024323,
+      "grad_norm": 24.58717918395996,
+      "learning_rate": 9.384172593021534e-07,
+      "loss": 4.1824,
+      "step": 16370
+    },
+    {
+      "epoch": 3.212043938799529,
+      "grad_norm": 18.0743408203125,
+      "learning_rate": 9.361611464627152e-07,
+      "loss": 4.4787,
+      "step": 16375
+    },
+    {
+      "epoch": 3.2130247155747353,
+      "grad_norm": 11.677950859069824,
+      "learning_rate": 9.339074687776789e-07,
+      "loss": 4.3985,
+      "step": 16380
+    },
+    {
+      "epoch": 3.214005492349941,
+      "grad_norm": 11.800333023071289,
+      "learning_rate": 9.316562275975066e-07,
+      "loss": 4.7217,
+      "step": 16385
+    },
+    {
+      "epoch": 3.214986269125147,
+      "grad_norm": 14.704316139221191,
+      "learning_rate": 9.294074242711993e-07,
+      "loss": 4.4862,
+      "step": 16390
+    },
+    {
+      "epoch": 3.215967045900353,
+      "grad_norm": 17.01054573059082,
+      "learning_rate": 9.271610601462955e-07,
+      "loss": 4.533,
+      "step": 16395
+    },
+    {
+      "epoch": 3.216947822675559,
+      "grad_norm": 16.018836975097656,
+      "learning_rate": 9.249171365688714e-07,
+      "loss": 4.3827,
+      "step": 16400
+    },
+    {
+      "epoch": 3.217928599450765,
+      "grad_norm": 24.968673706054688,
+      "learning_rate": 9.226756548835458e-07,
+      "loss": 4.4041,
+      "step": 16405
+    },
+    {
+      "epoch": 3.218909376225971,
+      "grad_norm": 27.278209686279297,
+      "learning_rate": 9.204366164334677e-07,
+      "loss": 4.3926,
+      "step": 16410
+    },
+    {
+      "epoch": 3.219890153001177,
+      "grad_norm": 15.485451698303223,
+      "learning_rate": 9.182000225603282e-07,
+      "loss": 4.4093,
+      "step": 16415
+    },
+    {
+      "epoch": 3.2208709297763827,
+      "grad_norm": 14.306288719177246,
+      "learning_rate": 9.159658746043476e-07,
+      "loss": 4.6559,
+      "step": 16420
+    },
+    {
+      "epoch": 3.221851706551589,
+      "grad_norm": 19.840612411499023,
+      "learning_rate": 9.137341739042859e-07,
+      "loss": 4.4419,
+      "step": 16425
+    },
+    {
+      "epoch": 3.222832483326795,
+      "grad_norm": 36.60581588745117,
+      "learning_rate": 9.115049217974325e-07,
+      "loss": 3.9703,
+      "step": 16430
+    },
+    {
+      "epoch": 3.2238132601020006,
+      "grad_norm": 39.210731506347656,
+      "learning_rate": 9.092781196196121e-07,
+      "loss": 4.9126,
+      "step": 16435
+    },
+    {
+      "epoch": 3.224794036877207,
+      "grad_norm": 14.3563814163208,
+      "learning_rate": 9.070537687051817e-07,
+      "loss": 4.6904,
+      "step": 16440
+    },
+    {
+      "epoch": 3.2257748136524125,
+      "grad_norm": 32.390159606933594,
+      "learning_rate": 9.048318703870263e-07,
+      "loss": 4.4831,
+      "step": 16445
+    },
+    {
+      "epoch": 3.2267555904276186,
+      "grad_norm": 20.18743133544922,
+      "learning_rate": 9.026124259965647e-07,
+      "loss": 4.1097,
+      "step": 16450
+    },
+    {
+      "epoch": 3.2277363672028248,
+      "grad_norm": 14.164064407348633,
+      "learning_rate": 9.003954368637424e-07,
+      "loss": 4.4878,
+      "step": 16455
+    },
+    {
+      "epoch": 3.2287171439780304,
+      "grad_norm": 17.929447174072266,
+      "learning_rate": 8.981809043170353e-07,
+      "loss": 4.4077,
+      "step": 16460
+    },
+    {
+      "epoch": 3.2296979207532366,
+      "grad_norm": 14.398148536682129,
+      "learning_rate": 8.959688296834491e-07,
+      "loss": 4.4003,
+      "step": 16465
+    },
+    {
+      "epoch": 3.2306786975284427,
+      "grad_norm": 20.427139282226562,
+      "learning_rate": 8.937592142885126e-07,
+      "loss": 4.7461,
+      "step": 16470
+    },
+    {
+      "epoch": 3.2316594743036484,
+      "grad_norm": 17.9531192779541,
+      "learning_rate": 8.915520594562821e-07,
+      "loss": 4.4267,
+      "step": 16475
+    },
+    {
+      "epoch": 3.2326402510788546,
+      "grad_norm": 10.198659896850586,
+      "learning_rate": 8.893473665093427e-07,
+      "loss": 4.7172,
+      "step": 16480
+    },
+    {
+      "epoch": 3.2336210278540602,
+      "grad_norm": 25.5524845123291,
+      "learning_rate": 8.871451367687994e-07,
+      "loss": 4.4695,
+      "step": 16485
+    },
+    {
+      "epoch": 3.2346018046292664,
+      "grad_norm": 24.992595672607422,
+      "learning_rate": 8.849453715542855e-07,
+      "loss": 4.5502,
+      "step": 16490
+    },
+    {
+      "epoch": 3.2355825814044725,
+      "grad_norm": 13.642231941223145,
+      "learning_rate": 8.827480721839538e-07,
+      "loss": 4.4869,
+      "step": 16495
+    },
+    {
+      "epoch": 3.236563358179678,
+      "grad_norm": 15.115335464477539,
+      "learning_rate": 8.805532399744837e-07,
+      "loss": 4.375,
+      "step": 16500
+    },
+    {
+      "epoch": 3.2375441349548844,
+      "grad_norm": 17.352785110473633,
+      "learning_rate": 8.783608762410712e-07,
+      "loss": 4.4238,
+      "step": 16505
+    },
+    {
+      "epoch": 3.23852491173009,
+      "grad_norm": 26.560272216796875,
+      "learning_rate": 8.761709822974368e-07,
+      "loss": 4.4159,
+      "step": 16510
+    },
+    {
+      "epoch": 3.239505688505296,
+      "grad_norm": 15.408249855041504,
+      "learning_rate": 8.739835594558216e-07,
+      "loss": 4.538,
+      "step": 16515
+    },
+    {
+      "epoch": 3.2404864652805023,
+      "grad_norm": 15.455544471740723,
+      "learning_rate": 8.71798609026982e-07,
+      "loss": 4.6661,
+      "step": 16520
+    },
+    {
+      "epoch": 3.241467242055708,
+      "grad_norm": 12.074155807495117,
+      "learning_rate": 8.696161323201974e-07,
+      "loss": 4.4038,
+      "step": 16525
+    },
+    {
+      "epoch": 3.242448018830914,
+      "grad_norm": 18.908735275268555,
+      "learning_rate": 8.674361306432599e-07,
+      "loss": 4.458,
+      "step": 16530
+    },
+    {
+      "epoch": 3.24342879560612,
+      "grad_norm": 10.406591415405273,
+      "learning_rate": 8.652586053024836e-07,
+      "loss": 4.5516,
+      "step": 16535
+    },
+    {
+      "epoch": 3.244409572381326,
+      "grad_norm": 32.47367477416992,
+      "learning_rate": 8.630835576026963e-07,
+      "loss": 4.1534,
+      "step": 16540
+    },
+    {
+      "epoch": 3.245390349156532,
+      "grad_norm": 14.95496940612793,
+      "learning_rate": 8.609109888472411e-07,
+      "loss": 4.2641,
+      "step": 16545
+    },
+    {
+      "epoch": 3.246371125931738,
+      "grad_norm": 30.883024215698242,
+      "learning_rate": 8.587409003379754e-07,
+      "loss": 4.4005,
+      "step": 16550
+    },
+    {
+      "epoch": 3.247351902706944,
+      "grad_norm": 17.348379135131836,
+      "learning_rate": 8.565732933752702e-07,
+      "loss": 4.626,
+      "step": 16555
+    },
+    {
+      "epoch": 3.24833267948215,
+      "grad_norm": 15.369430541992188,
+      "learning_rate": 8.544081692580097e-07,
+      "loss": 4.3092,
+      "step": 16560
+    },
+    {
+      "epoch": 3.249313456257356,
+      "grad_norm": 16.423633575439453,
+      "learning_rate": 8.522455292835935e-07,
+      "loss": 4.8843,
+      "step": 16565
+    },
+    {
+      "epoch": 3.250294233032562,
+      "grad_norm": 28.41481590270996,
+      "learning_rate": 8.50085374747927e-07,
+      "loss": 4.6737,
+      "step": 16570
+    },
+    {
+      "epoch": 3.2512750098077676,
+      "grad_norm": 22.398956298828125,
+      "learning_rate": 8.479277069454312e-07,
+      "loss": 4.5422,
+      "step": 16575
+    },
+    {
+      "epoch": 3.2512750098077676,
+      "eval_loss": 4.852513313293457,
+      "eval_runtime": 7.885,
+      "eval_samples_per_second": 26.506,
+      "eval_steps_per_second": 13.317,
+      "step": 16575
+    },
+    {
+      "epoch": 3.2522557865829738,
+      "grad_norm": 15.183259963989258,
+      "learning_rate": 8.457725271690326e-07,
+      "loss": 4.2779,
+      "step": 16580
+    },
+    {
+      "epoch": 3.25323656335818,
+      "grad_norm": 22.774782180786133,
+      "learning_rate": 8.436198367101705e-07,
+      "loss": 4.3453,
+      "step": 16585
+    },
+    {
+      "epoch": 3.2542173401333856,
+      "grad_norm": 27.975404739379883,
+      "learning_rate": 8.414696368587922e-07,
+      "loss": 4.4088,
+      "step": 16590
+    },
+    {
+      "epoch": 3.2551981169085917,
+      "grad_norm": 18.01176643371582,
+      "learning_rate": 8.393219289033489e-07,
+      "loss": 4.5194,
+      "step": 16595
+    },
+    {
+      "epoch": 3.2561788936837974,
+      "grad_norm": 33.312408447265625,
+      "learning_rate": 8.37176714130804e-07,
+      "loss": 4.2223,
+      "step": 16600
+    },
+    {
+      "epoch": 3.2571596704590036,
+      "grad_norm": 14.787341117858887,
+      "learning_rate": 8.350339938266211e-07,
+      "loss": 4.8151,
+      "step": 16605
+    },
+    {
+      "epoch": 3.2581404472342097,
+      "grad_norm": 23.211856842041016,
+      "learning_rate": 8.328937692747757e-07,
+      "loss": 4.7171,
+      "step": 16610
+    },
+    {
+      "epoch": 3.2591212240094154,
+      "grad_norm": 14.314786911010742,
+      "learning_rate": 8.307560417577404e-07,
+      "loss": 4.4659,
+      "step": 16615
+    },
+    {
+      "epoch": 3.2601020007846215,
+      "grad_norm": 15.22576904296875,
+      "learning_rate": 8.286208125564982e-07,
+      "loss": 4.2322,
+      "step": 16620
+    },
+    {
+      "epoch": 3.2610827775598272,
+      "grad_norm": 15.027915954589844,
+      "learning_rate": 8.264880829505312e-07,
+      "loss": 4.5284,
+      "step": 16625
+    },
+    {
+      "epoch": 3.2620635543350334,
+      "grad_norm": 32.40642547607422,
+      "learning_rate": 8.243578542178227e-07,
+      "loss": 4.5117,
+      "step": 16630
+    },
+    {
+      "epoch": 3.2630443311102395,
+      "grad_norm": 17.12091064453125,
+      "learning_rate": 8.222301276348615e-07,
+      "loss": 4.616,
+      "step": 16635
+    },
+    {
+      "epoch": 3.264025107885445,
+      "grad_norm": 18.07270622253418,
+      "learning_rate": 8.201049044766352e-07,
+      "loss": 4.2342,
+      "step": 16640
+    },
+    {
+      "epoch": 3.2650058846606513,
+      "grad_norm": 13.647133827209473,
+      "learning_rate": 8.179821860166288e-07,
+      "loss": 4.4643,
+      "step": 16645
+    },
+    {
+      "epoch": 3.265986661435857,
+      "grad_norm": 13.987504005432129,
+      "learning_rate": 8.158619735268314e-07,
+      "loss": 4.3447,
+      "step": 16650
+    },
+    {
+      "epoch": 3.266967438211063,
+      "grad_norm": 18.850154876708984,
+      "learning_rate": 8.137442682777241e-07,
+      "loss": 4.0809,
+      "step": 16655
+    },
+    {
+      "epoch": 3.2679482149862693,
+      "grad_norm": 20.752836227416992,
+      "learning_rate": 8.116290715382919e-07,
+      "loss": 4.5345,
+      "step": 16660
+    },
+    {
+      "epoch": 3.268928991761475,
+      "grad_norm": 11.551191329956055,
+      "learning_rate": 8.095163845760134e-07,
+      "loss": 4.3558,
+      "step": 16665
+    },
+    {
+      "epoch": 3.269909768536681,
+      "grad_norm": 23.9580135345459,
+      "learning_rate": 8.074062086568629e-07,
+      "loss": 4.5156,
+      "step": 16670
+    },
+    {
+      "epoch": 3.270890545311887,
+      "grad_norm": 15.155923843383789,
+      "learning_rate": 8.052985450453121e-07,
+      "loss": 4.5786,
+      "step": 16675
+    },
+    {
+      "epoch": 3.271871322087093,
+      "grad_norm": 19.49130630493164,
+      "learning_rate": 8.031933950043242e-07,
+      "loss": 4.3139,
+      "step": 16680
+    },
+    {
+      "epoch": 3.272852098862299,
+      "grad_norm": 20.867582321166992,
+      "learning_rate": 8.010907597953604e-07,
+      "loss": 4.2326,
+      "step": 16685
+    },
+    {
+      "epoch": 3.273832875637505,
+      "grad_norm": 18.114887237548828,
+      "learning_rate": 7.989906406783709e-07,
+      "loss": 4.4595,
+      "step": 16690
+    },
+    {
+      "epoch": 3.274813652412711,
+      "grad_norm": 13.96712875366211,
+      "learning_rate": 7.96893038911799e-07,
+      "loss": 4.456,
+      "step": 16695
+    },
+    {
+      "epoch": 3.2757944291879166,
+      "grad_norm": 10.919675827026367,
+      "learning_rate": 7.947979557525832e-07,
+      "loss": 4.3077,
+      "step": 16700
+    },
+    {
+      "epoch": 3.2767752059631228,
+      "grad_norm": 20.295543670654297,
+      "learning_rate": 7.927053924561473e-07,
+      "loss": 4.2028,
+      "step": 16705
+    },
+    {
+      "epoch": 3.277755982738329,
+      "grad_norm": 20.351930618286133,
+      "learning_rate": 7.906153502764085e-07,
+      "loss": 4.6729,
+      "step": 16710
+    },
+    {
+      "epoch": 3.2787367595135346,
+      "grad_norm": 12.80075454711914,
+      "learning_rate": 7.885278304657745e-07,
+      "loss": 4.2542,
+      "step": 16715
+    },
+    {
+      "epoch": 3.2797175362887407,
+      "grad_norm": 19.2437744140625,
+      "learning_rate": 7.864428342751368e-07,
+      "loss": 4.4505,
+      "step": 16720
+    },
+    {
+      "epoch": 3.2806983130639464,
+      "grad_norm": 17.746822357177734,
+      "learning_rate": 7.843603629538804e-07,
+      "loss": 4.5905,
+      "step": 16725
+    },
+    {
+      "epoch": 3.2816790898391526,
+      "grad_norm": 18.832889556884766,
+      "learning_rate": 7.822804177498716e-07,
+      "loss": 4.5141,
+      "step": 16730
+    },
+    {
+      "epoch": 3.2826598666143587,
+      "grad_norm": 12.651679992675781,
+      "learning_rate": 7.802029999094674e-07,
+      "loss": 4.41,
+      "step": 16735
+    },
+    {
+      "epoch": 3.2836406433895644,
+      "grad_norm": 18.479333877563477,
+      "learning_rate": 7.781281106775101e-07,
+      "loss": 4.3111,
+      "step": 16740
+    },
+    {
+      "epoch": 3.2846214201647705,
+      "grad_norm": 17.73377227783203,
+      "learning_rate": 7.760557512973227e-07,
+      "loss": 4.71,
+      "step": 16745
+    },
+    {
+      "epoch": 3.2856021969399762,
+      "grad_norm": 24.678346633911133,
+      "learning_rate": 7.739859230107177e-07,
+      "loss": 4.3451,
+      "step": 16750
+    },
+    {
+      "epoch": 3.2865829737151824,
+      "grad_norm": 10.567961692810059,
+      "learning_rate": 7.719186270579853e-07,
+      "loss": 4.5535,
+      "step": 16755
+    },
+    {
+      "epoch": 3.2875637504903885,
+      "grad_norm": 18.80795669555664,
+      "learning_rate": 7.698538646779047e-07,
+      "loss": 4.211,
+      "step": 16760
+    },
+    {
+      "epoch": 3.288544527265594,
+      "grad_norm": 11.36734676361084,
+      "learning_rate": 7.67791637107731e-07,
+      "loss": 4.1832,
+      "step": 16765
+    },
+    {
+      "epoch": 3.2895253040408003,
+      "grad_norm": 13.719120025634766,
+      "learning_rate": 7.657319455832024e-07,
+      "loss": 4.39,
+      "step": 16770
+    },
+    {
+      "epoch": 3.290506080816006,
+      "grad_norm": 21.944128036499023,
+      "learning_rate": 7.6367479133854e-07,
+      "loss": 4.3722,
+      "step": 16775
+    },
+    {
+      "epoch": 3.291486857591212,
+      "grad_norm": 18.50696563720703,
+      "learning_rate": 7.616201756064401e-07,
+      "loss": 4.7951,
+      "step": 16780
+    },
+    {
+      "epoch": 3.2924676343664183,
+      "grad_norm": 18.831857681274414,
+      "learning_rate": 7.59568099618081e-07,
+      "loss": 4.3203,
+      "step": 16785
+    },
+    {
+      "epoch": 3.293448411141624,
+      "grad_norm": 10.791632652282715,
+      "learning_rate": 7.575185646031197e-07,
+      "loss": 4.4587,
+      "step": 16790
+    },
+    {
+      "epoch": 3.29442918791683,
+      "grad_norm": 21.668577194213867,
+      "learning_rate": 7.554715717896866e-07,
+      "loss": 4.6222,
+      "step": 16795
+    },
+    {
+      "epoch": 3.2954099646920363,
+      "grad_norm": 16.101211547851562,
+      "learning_rate": 7.534271224043932e-07,
+      "loss": 4.4963,
+      "step": 16800
+    },
+    {
+      "epoch": 3.296390741467242,
+      "grad_norm": 18.882286071777344,
+      "learning_rate": 7.513852176723236e-07,
+      "loss": 4.4336,
+      "step": 16805
+    },
+    {
+      "epoch": 3.297371518242448,
+      "grad_norm": 21.985626220703125,
+      "learning_rate": 7.493458588170389e-07,
+      "loss": 4.1264,
+      "step": 16810
+    },
+    {
+      "epoch": 3.298352295017654,
+      "grad_norm": 20.74184226989746,
+      "learning_rate": 7.473090470605754e-07,
+      "loss": 4.3489,
+      "step": 16815
+    },
+    {
+      "epoch": 3.29933307179286,
+      "grad_norm": 22.413835525512695,
+      "learning_rate": 7.452747836234392e-07,
+      "loss": 4.3452,
+      "step": 16820
+    },
+    {
+      "epoch": 3.300313848568066,
+      "grad_norm": 13.036279678344727,
+      "learning_rate": 7.432430697246157e-07,
+      "loss": 4.4649,
+      "step": 16825
+    },
+    {
+      "epoch": 3.301294625343272,
+      "grad_norm": 14.902352333068848,
+      "learning_rate": 7.412139065815555e-07,
+      "loss": 4.4924,
+      "step": 16830
+    },
+    {
+      "epoch": 3.302275402118478,
+      "grad_norm": 24.96709632873535,
+      "learning_rate": 7.39187295410187e-07,
+      "loss": 4.5216,
+      "step": 16835
+    },
+    {
+      "epoch": 3.3032561788936836,
+      "grad_norm": 27.07962989807129,
+      "learning_rate": 7.371632374249049e-07,
+      "loss": 4.7339,
+      "step": 16840
+    },
+    {
+      "epoch": 3.3042369556688898,
+      "grad_norm": 29.041719436645508,
+      "learning_rate": 7.351417338385746e-07,
+      "loss": 4.4904,
+      "step": 16845
+    },
+    {
+      "epoch": 3.305217732444096,
+      "grad_norm": 19.445049285888672,
+      "learning_rate": 7.33122785862535e-07,
+      "loss": 4.3771,
+      "step": 16850
+    },
+    {
+      "epoch": 3.3061985092193016,
+      "grad_norm": 21.5811767578125,
+      "learning_rate": 7.311063947065871e-07,
+      "loss": 4.4102,
+      "step": 16855
+    },
+    {
+      "epoch": 3.3071792859945077,
+      "grad_norm": 21.47755241394043,
+      "learning_rate": 7.290925615790051e-07,
+      "loss": 4.3652,
+      "step": 16860
+    },
+    {
+      "epoch": 3.308160062769714,
+      "grad_norm": 14.804069519042969,
+      "learning_rate": 7.270812876865291e-07,
+      "loss": 4.36,
+      "step": 16865
+    },
+    {
+      "epoch": 3.3091408395449196,
+      "grad_norm": 26.158729553222656,
+      "learning_rate": 7.250725742343629e-07,
+      "loss": 4.8634,
+      "step": 16870
+    },
+    {
+      "epoch": 3.3101216163201257,
+      "grad_norm": 31.665578842163086,
+      "learning_rate": 7.230664224261801e-07,
+      "loss": 4.2634,
+      "step": 16875
+    },
+    {
+      "epoch": 3.3111023930953314,
+      "grad_norm": 23.756200790405273,
+      "learning_rate": 7.210628334641156e-07,
+      "loss": 4.5788,
+      "step": 16880
+    },
+    {
+      "epoch": 3.3120831698705375,
+      "grad_norm": 34.05301284790039,
+      "learning_rate": 7.190618085487705e-07,
+      "loss": 4.1559,
+      "step": 16885
+    },
+    {
+      "epoch": 3.3130639466457437,
+      "grad_norm": 20.167665481567383,
+      "learning_rate": 7.170633488792111e-07,
+      "loss": 4.5557,
+      "step": 16890
+    },
+    {
+      "epoch": 3.3140447234209494,
+      "grad_norm": 12.421319961547852,
+      "learning_rate": 7.150674556529624e-07,
+      "loss": 4.5346,
+      "step": 16895
+    },
+    {
+      "epoch": 3.3150255001961555,
+      "grad_norm": 19.413768768310547,
+      "learning_rate": 7.13074130066016e-07,
+      "loss": 4.3397,
+      "step": 16900
+    },
+    {
+      "epoch": 3.316006276971361,
+      "grad_norm": 17.019386291503906,
+      "learning_rate": 7.110833733128214e-07,
+      "loss": 4.7594,
+      "step": 16905
+    },
+    {
+      "epoch": 3.3169870537465673,
+      "grad_norm": 14.768471717834473,
+      "learning_rate": 7.090951865862888e-07,
+      "loss": 4.3822,
+      "step": 16910
+    },
+    {
+      "epoch": 3.3179678305217735,
+      "grad_norm": 30.920869827270508,
+      "learning_rate": 7.071095710777925e-07,
+      "loss": 4.5661,
+      "step": 16915
+    },
+    {
+      "epoch": 3.318948607296979,
+      "grad_norm": 29.130586624145508,
+      "learning_rate": 7.051265279771602e-07,
+      "loss": 4.4697,
+      "step": 16920
+    },
+    {
+      "epoch": 3.3199293840721853,
+      "grad_norm": 43.41074752807617,
+      "learning_rate": 7.03146058472684e-07,
+      "loss": 4.8567,
+      "step": 16925
+    },
+    {
+      "epoch": 3.320910160847391,
+      "grad_norm": 28.771547317504883,
+      "learning_rate": 7.011681637511092e-07,
+      "loss": 4.5718,
+      "step": 16930
+    },
+    {
+      "epoch": 3.321890937622597,
+      "grad_norm": 13.533172607421875,
+      "learning_rate": 6.991928449976398e-07,
+      "loss": 4.276,
+      "step": 16935
+    },
+    {
+      "epoch": 3.3228717143978033,
+      "grad_norm": 14.783927917480469,
+      "learning_rate": 6.972201033959386e-07,
+      "loss": 4.3144,
+      "step": 16940
+    },
+    {
+      "epoch": 3.323852491173009,
+      "grad_norm": 14.02553939819336,
+      "learning_rate": 6.952499401281199e-07,
+      "loss": 4.3646,
+      "step": 16945
+    },
+    {
+      "epoch": 3.324833267948215,
+      "grad_norm": 35.986019134521484,
+      "learning_rate": 6.932823563747559e-07,
+      "loss": 4.5013,
+      "step": 16950
+    },
+    {
+      "epoch": 3.325814044723421,
+      "grad_norm": 19.50735092163086,
+      "learning_rate": 6.91317353314872e-07,
+      "loss": 4.3468,
+      "step": 16955
+    },
+    {
+      "epoch": 3.326794821498627,
+      "grad_norm": 10.819270133972168,
+      "learning_rate": 6.893549321259468e-07,
+      "loss": 4.3793,
+      "step": 16960
+    },
+    {
+      "epoch": 3.327775598273833,
+      "grad_norm": 31.2229061126709,
+      "learning_rate": 6.873950939839147e-07,
+      "loss": 4.2508,
+      "step": 16965
+    },
+    {
+      "epoch": 3.3287563750490388,
+      "grad_norm": 22.09214973449707,
+      "learning_rate": 6.854378400631573e-07,
+      "loss": 4.7499,
+      "step": 16970
+    },
+    {
+      "epoch": 3.329737151824245,
+      "grad_norm": 27.418867111206055,
+      "learning_rate": 6.834831715365125e-07,
+      "loss": 4.3997,
+      "step": 16975
+    },
+    {
+      "epoch": 3.3307179285994506,
+      "grad_norm": 31.877626419067383,
+      "learning_rate": 6.815310895752658e-07,
+      "loss": 4.2885,
+      "step": 16980
+    },
+    {
+      "epoch": 3.3316987053746567,
+      "grad_norm": 28.25274658203125,
+      "learning_rate": 6.79581595349153e-07,
+      "loss": 4.523,
+      "step": 16985
+    },
+    {
+      "epoch": 3.332679482149863,
+      "grad_norm": 16.39922523498535,
+      "learning_rate": 6.776346900263614e-07,
+      "loss": 4.5928,
+      "step": 16990
+    },
+    {
+      "epoch": 3.3336602589250686,
+      "grad_norm": 45.208892822265625,
+      "learning_rate": 6.756903747735244e-07,
+      "loss": 4.4122,
+      "step": 16995
+    },
+    {
+      "epoch": 3.3346410357002747,
+      "grad_norm": 12.759726524353027,
+      "learning_rate": 6.737486507557262e-07,
+      "loss": 4.3968,
+      "step": 17000
+    },
+    {
+      "epoch": 3.3356218124754804,
+      "grad_norm": 20.541330337524414,
+      "learning_rate": 6.718095191364943e-07,
+      "loss": 4.2077,
+      "step": 17005
+    },
+    {
+      "epoch": 3.3366025892506865,
+      "grad_norm": 18.055675506591797,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 4.5941,
+      "step": 17010
+    },
+    {
+      "epoch": 3.3375833660258927,
+      "grad_norm": 32.806358337402344,
+      "learning_rate": 6.679390377400868e-07,
+      "loss": 4.4608,
+      "step": 17015
+    },
+    {
+      "epoch": 3.3385641428010984,
+      "grad_norm": 14.002641677856445,
+      "learning_rate": 6.660076902821994e-07,
+      "loss": 4.4344,
+      "step": 17020
+    },
+    {
+      "epoch": 3.3395449195763045,
+      "grad_norm": 12.99300765991211,
+      "learning_rate": 6.640789398614588e-07,
+      "loss": 4.5186,
+      "step": 17025
+    },
+    {
+      "epoch": 3.34052569635151,
+      "grad_norm": 9.83618450164795,
+      "learning_rate": 6.621527876336187e-07,
+      "loss": 4.6571,
+      "step": 17030
+    },
+    {
+      "epoch": 3.3415064731267163,
+      "grad_norm": 24.592845916748047,
+      "learning_rate": 6.602292347528794e-07,
+      "loss": 4.2534,
+      "step": 17035
+    },
+    {
+      "epoch": 3.3424872499019225,
+      "grad_norm": 28.41790008544922,
+      "learning_rate": 6.583082823718823e-07,
+      "loss": 4.2271,
+      "step": 17040
+    },
+    {
+      "epoch": 3.343468026677128,
+      "grad_norm": 10.874235153198242,
+      "learning_rate": 6.563899316417099e-07,
+      "loss": 4.5524,
+      "step": 17045
+    },
+    {
+      "epoch": 3.3444488034523343,
+      "grad_norm": 29.61422348022461,
+      "learning_rate": 6.544741837118851e-07,
+      "loss": 4.5697,
+      "step": 17050
+    },
+    {
+      "epoch": 3.34542958022754,
+      "grad_norm": 39.208351135253906,
+      "learning_rate": 6.525610397303739e-07,
+      "loss": 4.6328,
+      "step": 17055
+    },
+    {
+      "epoch": 3.346410357002746,
+      "grad_norm": 16.006139755249023,
+      "learning_rate": 6.506505008435787e-07,
+      "loss": 4.6253,
+      "step": 17060
+    },
+    {
+      "epoch": 3.3473911337779523,
+      "grad_norm": 13.404540061950684,
+      "learning_rate": 6.48742568196345e-07,
+      "loss": 4.4215,
+      "step": 17065
+    },
+    {
+      "epoch": 3.348371910553158,
+      "grad_norm": 17.115400314331055,
+      "learning_rate": 6.468372429319503e-07,
+      "loss": 4.3132,
+      "step": 17070
+    },
+    {
+      "epoch": 3.349352687328364,
+      "grad_norm": 25.940824508666992,
+      "learning_rate": 6.44934526192117e-07,
+      "loss": 4.6653,
+      "step": 17075
+    },
+    {
+      "epoch": 3.35033346410357,
+      "grad_norm": 17.652284622192383,
+      "learning_rate": 6.43034419116998e-07,
+      "loss": 4.4474,
+      "step": 17080
+    },
+    {
+      "epoch": 3.351314240878776,
+      "grad_norm": 12.598196983337402,
+      "learning_rate": 6.411369228451858e-07,
+      "loss": 4.3505,
+      "step": 17085
+    },
+    {
+      "epoch": 3.352295017653982,
+      "grad_norm": 24.629087448120117,
+      "learning_rate": 6.392420385137104e-07,
+      "loss": 4.1629,
+      "step": 17090
+    },
+    {
+      "epoch": 3.3532757944291878,
+      "grad_norm": 13.39428997039795,
+      "learning_rate": 6.373497672580309e-07,
+      "loss": 4.3679,
+      "step": 17095
+    },
+    {
+      "epoch": 3.354256571204394,
+      "grad_norm": 18.239906311035156,
+      "learning_rate": 6.354601102120462e-07,
+      "loss": 4.5945,
+      "step": 17100
+    },
+    {
+      "epoch": 3.3552373479795996,
+      "grad_norm": 12.23827838897705,
+      "learning_rate": 6.335730685080838e-07,
+      "loss": 4.432,
+      "step": 17105
+    },
+    {
+      "epoch": 3.3562181247548057,
+      "grad_norm": 20.083877563476562,
+      "learning_rate": 6.316886432769081e-07,
+      "loss": 4.3537,
+      "step": 17110
+    },
+    {
+      "epoch": 3.357198901530012,
+      "grad_norm": 11.51676082611084,
+      "learning_rate": 6.29806835647715e-07,
+      "loss": 4.5358,
+      "step": 17115
+    },
+    {
+      "epoch": 3.3581796783052176,
+      "grad_norm": 17.398361206054688,
+      "learning_rate": 6.279276467481299e-07,
+      "loss": 4.4557,
+      "step": 17120
+    },
+    {
+      "epoch": 3.3591604550804237,
+      "grad_norm": 33.21310806274414,
+      "learning_rate": 6.260510777042089e-07,
+      "loss": 4.3373,
+      "step": 17125
+    },
+    {
+      "epoch": 3.36014123185563,
+      "grad_norm": 13.65185260772705,
+      "learning_rate": 6.24177129640442e-07,
+      "loss": 4.2739,
+      "step": 17130
+    },
+    {
+      "epoch": 3.3611220086308355,
+      "grad_norm": 20.5279541015625,
+      "learning_rate": 6.22305803679743e-07,
+      "loss": 4.54,
+      "step": 17135
+    },
+    {
+      "epoch": 3.3621027854060417,
+      "grad_norm": 24.423192977905273,
+      "learning_rate": 6.204371009434595e-07,
+      "loss": 4.5603,
+      "step": 17140
+    },
+    {
+      "epoch": 3.3630835621812474,
+      "grad_norm": 25.014257431030273,
+      "learning_rate": 6.185710225513641e-07,
+      "loss": 4.3737,
+      "step": 17145
+    },
+    {
+      "epoch": 3.3640643389564535,
+      "grad_norm": 18.271961212158203,
+      "learning_rate": 6.16707569621659e-07,
+      "loss": 4.2319,
+      "step": 17150
+    },
+    {
+      "epoch": 3.3650451157316597,
+      "grad_norm": 17.02702522277832,
+      "learning_rate": 6.148467432709704e-07,
+      "loss": 4.1899,
+      "step": 17155
+    },
+    {
+      "epoch": 3.3660258925068653,
+      "grad_norm": 14.603148460388184,
+      "learning_rate": 6.129885446143536e-07,
+      "loss": 4.3499,
+      "step": 17160
+    },
+    {
+      "epoch": 3.3670066692820715,
+      "grad_norm": 14.893680572509766,
+      "learning_rate": 6.111329747652884e-07,
+      "loss": 4.7695,
+      "step": 17165
+    },
+    {
+      "epoch": 3.367987446057277,
+      "grad_norm": 22.7106876373291,
+      "learning_rate": 6.092800348356765e-07,
+      "loss": 4.4914,
+      "step": 17170
+    },
+    {
+      "epoch": 3.3689682228324833,
+      "grad_norm": 31.128007888793945,
+      "learning_rate": 6.074297259358492e-07,
+      "loss": 4.3197,
+      "step": 17175
+    },
+    {
+      "epoch": 3.3699489996076895,
+      "grad_norm": 17.588899612426758,
+      "learning_rate": 6.055820491745557e-07,
+      "loss": 4.7125,
+      "step": 17180
+    },
+    {
+      "epoch": 3.370929776382895,
+      "grad_norm": 11.689655303955078,
+      "learning_rate": 6.037370056589709e-07,
+      "loss": 4.4528,
+      "step": 17185
+    },
+    {
+      "epoch": 3.3719105531581013,
+      "grad_norm": 10.81857967376709,
+      "learning_rate": 6.01894596494692e-07,
+      "loss": 4.5021,
+      "step": 17190
+    },
+    {
+      "epoch": 3.3728913299333074,
+      "grad_norm": 18.195873260498047,
+      "learning_rate": 6.000548227857372e-07,
+      "loss": 4.4054,
+      "step": 17195
+    },
+    {
+      "epoch": 3.373872106708513,
+      "grad_norm": 15.955674171447754,
+      "learning_rate": 5.982176856345445e-07,
+      "loss": 4.3278,
+      "step": 17200
+    },
+    {
+      "epoch": 3.3748528834837193,
+      "grad_norm": 24.263078689575195,
+      "learning_rate": 5.963831861419711e-07,
+      "loss": 4.7147,
+      "step": 17205
+    },
+    {
+      "epoch": 3.375833660258925,
+      "grad_norm": 24.826202392578125,
+      "learning_rate": 5.945513254072971e-07,
+      "loss": 4.4723,
+      "step": 17210
+    },
+    {
+      "epoch": 3.376814437034131,
+      "grad_norm": 36.65044021606445,
+      "learning_rate": 5.9272210452822e-07,
+      "loss": 4.6737,
+      "step": 17215
+    },
+    {
+      "epoch": 3.3777952138093372,
+      "grad_norm": 30.517133712768555,
+      "learning_rate": 5.90895524600853e-07,
+      "loss": 4.4498,
+      "step": 17220
+    },
+    {
+      "epoch": 3.378775990584543,
+      "grad_norm": 17.615612030029297,
+      "learning_rate": 5.890715867197305e-07,
+      "loss": 4.2525,
+      "step": 17225
+    },
+    {
+      "epoch": 3.379756767359749,
+      "grad_norm": 24.215167999267578,
+      "learning_rate": 5.872502919778006e-07,
+      "loss": 4.5694,
+      "step": 17230
+    },
+    {
+      "epoch": 3.3807375441349548,
+      "grad_norm": 18.129798889160156,
+      "learning_rate": 5.85431641466429e-07,
+      "loss": 4.2075,
+      "step": 17235
+    },
+    {
+      "epoch": 3.381718320910161,
+      "grad_norm": 14.993080139160156,
+      "learning_rate": 5.836156362753987e-07,
+      "loss": 4.5913,
+      "step": 17240
+    },
+    {
+      "epoch": 3.382699097685367,
+      "grad_norm": 20.029644012451172,
+      "learning_rate": 5.818022774929033e-07,
+      "loss": 4.4598,
+      "step": 17245
+    },
+    {
+      "epoch": 3.3836798744605727,
+      "grad_norm": 24.22998809814453,
+      "learning_rate": 5.799915662055544e-07,
+      "loss": 4.3082,
+      "step": 17250
+    },
+    {
+      "epoch": 3.384660651235779,
+      "grad_norm": 13.657512664794922,
+      "learning_rate": 5.781835034983746e-07,
+      "loss": 4.3457,
+      "step": 17255
+    },
+    {
+      "epoch": 3.3856414280109846,
+      "grad_norm": 17.667686462402344,
+      "learning_rate": 5.763780904548022e-07,
+      "loss": 4.619,
+      "step": 17260
+    },
+    {
+      "epoch": 3.3866222047861907,
+      "grad_norm": 27.960433959960938,
+      "learning_rate": 5.745753281566841e-07,
+      "loss": 4.7395,
+      "step": 17265
+    },
+    {
+      "epoch": 3.387602981561397,
+      "grad_norm": 15.462547302246094,
+      "learning_rate": 5.727752176842827e-07,
+      "loss": 4.609,
+      "step": 17270
+    },
+    {
+      "epoch": 3.3885837583366025,
+      "grad_norm": 12.116942405700684,
+      "learning_rate": 5.70977760116268e-07,
+      "loss": 4.4295,
+      "step": 17275
+    },
+    {
+      "epoch": 3.3895645351118087,
+      "grad_norm": 15.605005264282227,
+      "learning_rate": 5.691829565297219e-07,
+      "loss": 4.3967,
+      "step": 17280
+    },
+    {
+      "epoch": 3.3905453118870144,
+      "grad_norm": 17.77362632751465,
+      "learning_rate": 5.673908080001356e-07,
+      "loss": 3.9033,
+      "step": 17285
+    },
+    {
+      "epoch": 3.3915260886622205,
+      "grad_norm": 13.116978645324707,
+      "learning_rate": 5.656013156014118e-07,
+      "loss": 4.5672,
+      "step": 17290
+    },
+    {
+      "epoch": 3.3925068654374266,
+      "grad_norm": 16.185178756713867,
+      "learning_rate": 5.63814480405856e-07,
+      "loss": 4.6599,
+      "step": 17295
+    },
+    {
+      "epoch": 3.3934876422126323,
+      "grad_norm": 23.99867820739746,
+      "learning_rate": 5.620303034841879e-07,
+      "loss": 4.5349,
+      "step": 17300
+    },
+    {
+      "epoch": 3.3944684189878385,
+      "grad_norm": 17.248830795288086,
+      "learning_rate": 5.602487859055283e-07,
+      "loss": 4.1866,
+      "step": 17305
+    },
+    {
+      "epoch": 3.395449195763044,
+      "grad_norm": 20.530193328857422,
+      "learning_rate": 5.58469928737409e-07,
+      "loss": 4.362,
+      "step": 17310
+    },
+    {
+      "epoch": 3.3964299725382503,
+      "grad_norm": 23.010272979736328,
+      "learning_rate": 5.566937330457667e-07,
+      "loss": 4.0736,
+      "step": 17315
+    },
+    {
+      "epoch": 3.3974107493134564,
+      "grad_norm": 14.467041015625,
+      "learning_rate": 5.549201998949399e-07,
+      "loss": 4.6956,
+      "step": 17320
+    },
+    {
+      "epoch": 3.398391526088662,
+      "grad_norm": 25.341157913208008,
+      "learning_rate": 5.531493303476775e-07,
+      "loss": 4.6609,
+      "step": 17325
+    },
+    {
+      "epoch": 3.3993723028638683,
+      "grad_norm": 16.466978073120117,
+      "learning_rate": 5.513811254651258e-07,
+      "loss": 4.2494,
+      "step": 17330
+    },
+    {
+      "epoch": 3.400353079639074,
+      "grad_norm": 21.759614944458008,
+      "learning_rate": 5.496155863068409e-07,
+      "loss": 4.4691,
+      "step": 17335
+    },
+    {
+      "epoch": 3.40133385641428,
+      "grad_norm": 31.557613372802734,
+      "learning_rate": 5.47852713930776e-07,
+      "loss": 4.4447,
+      "step": 17340
+    },
+    {
+      "epoch": 3.4023146331894862,
+      "grad_norm": 20.74547004699707,
+      "learning_rate": 5.460925093932879e-07,
+      "loss": 4.8401,
+      "step": 17345
+    },
+    {
+      "epoch": 3.403295409964692,
+      "grad_norm": 22.440385818481445,
+      "learning_rate": 5.443349737491377e-07,
+      "loss": 4.5203,
+      "step": 17350
+    },
+    {
+      "epoch": 3.404276186739898,
+      "grad_norm": 15.865456581115723,
+      "learning_rate": 5.425801080514831e-07,
+      "loss": 4.6614,
+      "step": 17355
+    },
+    {
+      "epoch": 3.4052569635151038,
+      "grad_norm": 24.033784866333008,
+      "learning_rate": 5.408279133518846e-07,
+      "loss": 4.6595,
+      "step": 17360
+    },
+    {
+      "epoch": 3.40623774029031,
+      "grad_norm": 16.923505783081055,
+      "learning_rate": 5.390783907003017e-07,
+      "loss": 4.5436,
+      "step": 17365
+    },
+    {
+      "epoch": 3.407218517065516,
+      "grad_norm": 17.031034469604492,
+      "learning_rate": 5.373315411450908e-07,
+      "loss": 4.2987,
+      "step": 17370
+    },
+    {
+      "epoch": 3.4081992938407217,
+      "grad_norm": 21.693077087402344,
+      "learning_rate": 5.355873657330107e-07,
+      "loss": 4.4216,
+      "step": 17375
+    },
+    {
+      "epoch": 3.409180070615928,
+      "grad_norm": 16.699193954467773,
+      "learning_rate": 5.338458655092122e-07,
+      "loss": 4.3596,
+      "step": 17380
+    },
+    {
+      "epoch": 3.4101608473911336,
+      "grad_norm": 21.562950134277344,
+      "learning_rate": 5.321070415172469e-07,
+      "loss": 3.9197,
+      "step": 17385
+    },
+    {
+      "epoch": 3.4111416241663397,
+      "grad_norm": 17.56022834777832,
+      "learning_rate": 5.303708947990638e-07,
+      "loss": 4.3206,
+      "step": 17390
+    },
+    {
+      "epoch": 3.412122400941546,
+      "grad_norm": 20.220117568969727,
+      "learning_rate": 5.286374263950034e-07,
+      "loss": 4.2082,
+      "step": 17395
+    },
+    {
+      "epoch": 3.4131031777167515,
+      "grad_norm": 20.73967933654785,
+      "learning_rate": 5.269066373438048e-07,
+      "loss": 4.5537,
+      "step": 17400
+    },
+    {
+      "epoch": 3.4140839544919577,
+      "grad_norm": 18.557260513305664,
+      "learning_rate": 5.251785286825994e-07,
+      "loss": 4.4783,
+      "step": 17405
+    },
+    {
+      "epoch": 3.4150647312671634,
+      "grad_norm": 26.5538387298584,
+      "learning_rate": 5.23453101446914e-07,
+      "loss": 4.317,
+      "step": 17410
+    },
+    {
+      "epoch": 3.4160455080423695,
+      "grad_norm": 14.261948585510254,
+      "learning_rate": 5.217303566706683e-07,
+      "loss": 4.465,
+      "step": 17415
+    },
+    {
+      "epoch": 3.4170262848175756,
+      "grad_norm": 16.286413192749023,
+      "learning_rate": 5.20010295386173e-07,
+      "loss": 4.1083,
+      "step": 17420
+    },
+    {
+      "epoch": 3.4180070615927813,
+      "grad_norm": 25.792701721191406,
+      "learning_rate": 5.182929186241331e-07,
+      "loss": 4.3977,
+      "step": 17425
+    },
+    {
+      "epoch": 3.4189878383679875,
+      "grad_norm": 30.001115798950195,
+      "learning_rate": 5.165782274136433e-07,
+      "loss": 4.4344,
+      "step": 17430
+    },
+    {
+      "epoch": 3.419968615143193,
+      "grad_norm": 15.186819076538086,
+      "learning_rate": 5.148662227821899e-07,
+      "loss": 4.4817,
+      "step": 17435
+    },
+    {
+      "epoch": 3.4209493919183993,
+      "grad_norm": 14.81648063659668,
+      "learning_rate": 5.131569057556496e-07,
+      "loss": 4.3782,
+      "step": 17440
+    },
+    {
+      "epoch": 3.4219301686936054,
+      "grad_norm": 24.624345779418945,
+      "learning_rate": 5.114502773582875e-07,
+      "loss": 4.6731,
+      "step": 17445
+    },
+    {
+      "epoch": 3.422910945468811,
+      "grad_norm": 19.26167106628418,
+      "learning_rate": 5.097463386127593e-07,
+      "loss": 4.5661,
+      "step": 17450
+    },
+    {
+      "epoch": 3.4238917222440173,
+      "grad_norm": 16.40976333618164,
+      "learning_rate": 5.080450905401057e-07,
+      "loss": 4.3338,
+      "step": 17455
+    },
+    {
+      "epoch": 3.4248724990192234,
+      "grad_norm": 20.465471267700195,
+      "learning_rate": 5.063465341597589e-07,
+      "loss": 4.4763,
+      "step": 17460
+    },
+    {
+      "epoch": 3.425853275794429,
+      "grad_norm": 13.48697280883789,
+      "learning_rate": 5.046506704895376e-07,
+      "loss": 4.3597,
+      "step": 17465
+    },
+    {
+      "epoch": 3.4268340525696352,
+      "grad_norm": 33.226158142089844,
+      "learning_rate": 5.029575005456439e-07,
+      "loss": 4.4754,
+      "step": 17470
+    },
+    {
+      "epoch": 3.427814829344841,
+      "grad_norm": 13.726212501525879,
+      "learning_rate": 5.012670253426699e-07,
+      "loss": 4.4491,
+      "step": 17475
+    },
+    {
+      "epoch": 3.428795606120047,
+      "grad_norm": 16.772018432617188,
+      "learning_rate": 4.995792458935877e-07,
+      "loss": 4.7095,
+      "step": 17480
+    },
+    {
+      "epoch": 3.429776382895253,
+      "grad_norm": 12.425341606140137,
+      "learning_rate": 4.978941632097612e-07,
+      "loss": 4.7728,
+      "step": 17485
+    },
+    {
+      "epoch": 3.430757159670459,
+      "grad_norm": 13.08029842376709,
+      "learning_rate": 4.962117783009313e-07,
+      "loss": 4.5698,
+      "step": 17490
+    },
+    {
+      "epoch": 3.431737936445665,
+      "grad_norm": 13.78771686553955,
+      "learning_rate": 4.945320921752255e-07,
+      "loss": 4.3103,
+      "step": 17495
+    },
+    {
+      "epoch": 3.4327187132208707,
+      "grad_norm": 17.305667877197266,
+      "learning_rate": 4.928551058391556e-07,
+      "loss": 4.458,
+      "step": 17500
+    },
+    {
+      "epoch": 3.433699489996077,
+      "grad_norm": 13.464914321899414,
+      "learning_rate": 4.911808202976121e-07,
+      "loss": 4.3716,
+      "step": 17505
+    },
+    {
+      "epoch": 3.434680266771283,
+      "grad_norm": 18.66707992553711,
+      "learning_rate": 4.895092365538701e-07,
+      "loss": 4.2377,
+      "step": 17510
+    },
+    {
+      "epoch": 3.4356610435464887,
+      "grad_norm": 16.312175750732422,
+      "learning_rate": 4.878403556095851e-07,
+      "loss": 4.6405,
+      "step": 17515
+    },
+    {
+      "epoch": 3.436641820321695,
+      "grad_norm": 18.651382446289062,
+      "learning_rate": 4.86174178464791e-07,
+      "loss": 4.835,
+      "step": 17520
+    },
+    {
+      "epoch": 3.437622597096901,
+      "grad_norm": 18.005874633789062,
+      "learning_rate": 4.845107061179049e-07,
+      "loss": 4.3222,
+      "step": 17525
+    },
+    {
+      "epoch": 3.4386033738721067,
+      "grad_norm": 21.741695404052734,
+      "learning_rate": 4.828499395657194e-07,
+      "loss": 4.4938,
+      "step": 17530
+    },
+    {
+      "epoch": 3.439584150647313,
+      "grad_norm": 27.118772506713867,
+      "learning_rate": 4.811918798034082e-07,
+      "loss": 4.2959,
+      "step": 17535
+    },
+    {
+      "epoch": 3.4405649274225185,
+      "grad_norm": 15.861380577087402,
+      "learning_rate": 4.79536527824524e-07,
+      "loss": 4.3856,
+      "step": 17540
+    },
+    {
+      "epoch": 3.4415457041977247,
+      "grad_norm": 17.277454376220703,
+      "learning_rate": 4.778838846209927e-07,
+      "loss": 4.6898,
+      "step": 17545
+    },
+    {
+      "epoch": 3.442526480972931,
+      "grad_norm": 12.770515441894531,
+      "learning_rate": 4.7623395118312154e-07,
+      "loss": 4.3914,
+      "step": 17550
+    },
+    {
+      "epoch": 3.4435072577481365,
+      "grad_norm": 18.324600219726562,
+      "learning_rate": 4.745867284995914e-07,
+      "loss": 4.308,
+      "step": 17555
+    },
+    {
+      "epoch": 3.4444880345233426,
+      "grad_norm": 49.0174560546875,
+      "learning_rate": 4.7294221755745885e-07,
+      "loss": 4.8907,
+      "step": 17560
+    },
+    {
+      "epoch": 3.4454688112985483,
+      "grad_norm": 23.207433700561523,
+      "learning_rate": 4.7130041934215777e-07,
+      "loss": 4.3127,
+      "step": 17565
+    },
+    {
+      "epoch": 3.4464495880737545,
+      "grad_norm": 17.527206420898438,
+      "learning_rate": 4.6966133483749346e-07,
+      "loss": 4.4622,
+      "step": 17570
+    },
+    {
+      "epoch": 3.4474303648489606,
+      "grad_norm": 14.135499000549316,
+      "learning_rate": 4.680249650256474e-07,
+      "loss": 4.2042,
+      "step": 17575
+    },
+    {
+      "epoch": 3.4484111416241663,
+      "grad_norm": 53.29972457885742,
+      "learning_rate": 4.663913108871726e-07,
+      "loss": 4.5112,
+      "step": 17580
+    },
+    {
+      "epoch": 3.4493919183993724,
+      "grad_norm": 24.076627731323242,
+      "learning_rate": 4.647603734009964e-07,
+      "loss": 4.4795,
+      "step": 17585
+    },
+    {
+      "epoch": 3.450372695174578,
+      "grad_norm": 19.697158813476562,
+      "learning_rate": 4.6313215354441885e-07,
+      "loss": 4.6016,
+      "step": 17590
+    },
+    {
+      "epoch": 3.4513534719497843,
+      "grad_norm": 19.25453758239746,
+      "learning_rate": 4.6150665229310774e-07,
+      "loss": 4.3914,
+      "step": 17595
+    },
+    {
+      "epoch": 3.4523342487249904,
+      "grad_norm": 12.775351524353027,
+      "learning_rate": 4.598838706211062e-07,
+      "loss": 4.7117,
+      "step": 17600
+    },
+    {
+      "epoch": 3.453315025500196,
+      "grad_norm": 25.01823616027832,
+      "learning_rate": 4.5826380950082403e-07,
+      "loss": 4.2695,
+      "step": 17605
+    },
+    {
+      "epoch": 3.4542958022754022,
+      "grad_norm": 16.992021560668945,
+      "learning_rate": 4.5664646990304375e-07,
+      "loss": 4.516,
+      "step": 17610
+    },
+    {
+      "epoch": 3.455276579050608,
+      "grad_norm": 13.946948051452637,
+      "learning_rate": 4.550318527969161e-07,
+      "loss": 4.306,
+      "step": 17615
+    },
+    {
+      "epoch": 3.456257355825814,
+      "grad_norm": 49.40144348144531,
+      "learning_rate": 4.534199591499594e-07,
+      "loss": 4.7331,
+      "step": 17620
+    },
+    {
+      "epoch": 3.45723813260102,
+      "grad_norm": 25.051666259765625,
+      "learning_rate": 4.5181078992806215e-07,
+      "loss": 4.3483,
+      "step": 17625
+    },
+    {
+      "epoch": 3.458218909376226,
+      "grad_norm": 22.993051528930664,
+      "learning_rate": 4.502043460954786e-07,
+      "loss": 4.1843,
+      "step": 17630
+    },
+    {
+      "epoch": 3.459199686151432,
+      "grad_norm": 31.63532829284668,
+      "learning_rate": 4.486006286148287e-07,
+      "loss": 4.4431,
+      "step": 17635
+    },
+    {
+      "epoch": 3.4601804629266377,
+      "grad_norm": 17.333829879760742,
+      "learning_rate": 4.4699963844710203e-07,
+      "loss": 5.1192,
+      "step": 17640
+    },
+    {
+      "epoch": 3.461161239701844,
+      "grad_norm": 14.513242721557617,
+      "learning_rate": 4.454013765516507e-07,
+      "loss": 4.7808,
+      "step": 17645
+    },
+    {
+      "epoch": 3.46214201647705,
+      "grad_norm": 11.146553039550781,
+      "learning_rate": 4.438058438861953e-07,
+      "loss": 4.5906,
+      "step": 17650
+    },
+    {
+      "epoch": 3.4631227932522557,
+      "grad_norm": 17.965736389160156,
+      "learning_rate": 4.4221304140681707e-07,
+      "loss": 4.5514,
+      "step": 17655
+    },
+    {
+      "epoch": 3.464103570027462,
+      "grad_norm": 16.620882034301758,
+      "learning_rate": 4.406229700679643e-07,
+      "loss": 4.7881,
+      "step": 17660
+    },
+    {
+      "epoch": 3.4650843468026675,
+      "grad_norm": 19.04009246826172,
+      "learning_rate": 4.390356308224486e-07,
+      "loss": 4.6736,
+      "step": 17665
+    },
+    {
+      "epoch": 3.4660651235778737,
+      "grad_norm": 21.045454025268555,
+      "learning_rate": 4.3745102462144197e-07,
+      "loss": 4.4628,
+      "step": 17670
+    },
+    {
+      "epoch": 3.46704590035308,
+      "grad_norm": 15.176902770996094,
+      "learning_rate": 4.35869152414482e-07,
+      "loss": 4.3495,
+      "step": 17675
+    },
+    {
+      "epoch": 3.4680266771282855,
+      "grad_norm": 16.883756637573242,
+      "learning_rate": 4.342900151494639e-07,
+      "loss": 4.4987,
+      "step": 17680
+    },
+    {
+      "epoch": 3.4690074539034916,
+      "grad_norm": 17.017446517944336,
+      "learning_rate": 4.327136137726479e-07,
+      "loss": 4.2579,
+      "step": 17685
+    },
+    {
+      "epoch": 3.4699882306786973,
+      "grad_norm": 36.79636001586914,
+      "learning_rate": 4.3113994922865443e-07,
+      "loss": 4.5016,
+      "step": 17690
+    },
+    {
+      "epoch": 3.4709690074539035,
+      "grad_norm": 22.775876998901367,
+      "learning_rate": 4.2956902246046093e-07,
+      "loss": 4.446,
+      "step": 17695
+    },
+    {
+      "epoch": 3.4719497842291096,
+      "grad_norm": 17.92118263244629,
+      "learning_rate": 4.2800083440940663e-07,
+      "loss": 4.582,
+      "step": 17700
+    },
+    {
+      "epoch": 3.4729305610043153,
+      "grad_norm": 25.144628524780273,
+      "learning_rate": 4.264353860151904e-07,
+      "loss": 4.6917,
+      "step": 17705
+    },
+    {
+      "epoch": 3.4739113377795214,
+      "grad_norm": 24.989463806152344,
+      "learning_rate": 4.248726782158663e-07,
+      "loss": 4.7562,
+      "step": 17710
+    },
+    {
+      "epoch": 3.474892114554727,
+      "grad_norm": 21.577003479003906,
+      "learning_rate": 4.233127119478497e-07,
+      "loss": 4.3303,
+      "step": 17715
+    },
+    {
+      "epoch": 3.4758728913299333,
+      "grad_norm": 18.959033966064453,
+      "learning_rate": 4.217554881459107e-07,
+      "loss": 4.4397,
+      "step": 17720
+    },
+    {
+      "epoch": 3.4768536681051394,
+      "grad_norm": 20.839256286621094,
+      "learning_rate": 4.202010077431784e-07,
+      "loss": 4.6779,
+      "step": 17725
+    },
+    {
+      "epoch": 3.477834444880345,
+      "grad_norm": 15.40798282623291,
+      "learning_rate": 4.1864927167113434e-07,
+      "loss": 4.0281,
+      "step": 17730
+    },
+    {
+      "epoch": 3.4788152216555512,
+      "grad_norm": 12.977888107299805,
+      "learning_rate": 4.171002808596192e-07,
+      "loss": 4.7644,
+      "step": 17735
+    },
+    {
+      "epoch": 3.479795998430757,
+      "grad_norm": 18.18838119506836,
+      "learning_rate": 4.155540362368277e-07,
+      "loss": 4.2649,
+      "step": 17740
+    },
+    {
+      "epoch": 3.480776775205963,
+      "grad_norm": 22.027528762817383,
+      "learning_rate": 4.140105387293064e-07,
+      "loss": 4.7378,
+      "step": 17745
+    },
+    {
+      "epoch": 3.481757551981169,
+      "grad_norm": 19.1549015045166,
+      "learning_rate": 4.1246978926196057e-07,
+      "loss": 4.4474,
+      "step": 17750
+    },
+    {
+      "epoch": 3.482738328756375,
+      "grad_norm": 20.467920303344727,
+      "learning_rate": 4.1093178875804384e-07,
+      "loss": 4.4719,
+      "step": 17755
+    },
+    {
+      "epoch": 3.483719105531581,
+      "grad_norm": 13.859403610229492,
+      "learning_rate": 4.093965381391651e-07,
+      "loss": 4.6655,
+      "step": 17760
+    },
+    {
+      "epoch": 3.4846998823067867,
+      "grad_norm": 26.066524505615234,
+      "learning_rate": 4.078640383252869e-07,
+      "loss": 4.3725,
+      "step": 17765
+    },
+    {
+      "epoch": 3.485680659081993,
+      "grad_norm": 22.628177642822266,
+      "learning_rate": 4.0633429023472004e-07,
+      "loss": 4.506,
+      "step": 17770
+    },
+    {
+      "epoch": 3.486661435857199,
+      "grad_norm": 27.378026962280273,
+      "learning_rate": 4.048072947841275e-07,
+      "loss": 4.366,
+      "step": 17775
+    },
+    {
+      "epoch": 3.4876422126324047,
+      "grad_norm": 19.4333553314209,
+      "learning_rate": 4.0328305288852454e-07,
+      "loss": 4.5974,
+      "step": 17780
+    },
+    {
+      "epoch": 3.488622989407611,
+      "grad_norm": 10.120104789733887,
+      "learning_rate": 4.0176156546127443e-07,
+      "loss": 4.6727,
+      "step": 17785
+    },
+    {
+      "epoch": 3.489603766182817,
+      "grad_norm": 31.34845733642578,
+      "learning_rate": 4.0024283341409233e-07,
+      "loss": 4.2804,
+      "step": 17790
+    },
+    {
+      "epoch": 3.4905845429580227,
+      "grad_norm": 18.09394645690918,
+      "learning_rate": 3.98726857657038e-07,
+      "loss": 4.4176,
+      "step": 17795
+    },
+    {
+      "epoch": 3.491565319733229,
+      "grad_norm": 25.27265739440918,
+      "learning_rate": 3.972136390985248e-07,
+      "loss": 4.6796,
+      "step": 17800
+    },
+    {
+      "epoch": 3.4925460965084345,
+      "grad_norm": 24.39628028869629,
+      "learning_rate": 3.9570317864530916e-07,
+      "loss": 4.0912,
+      "step": 17805
+    },
+    {
+      "epoch": 3.4935268732836406,
+      "grad_norm": 36.115509033203125,
+      "learning_rate": 3.941954772024981e-07,
+      "loss": 4.6049,
+      "step": 17810
+    },
+    {
+      "epoch": 3.494507650058847,
+      "grad_norm": 28.326950073242188,
+      "learning_rate": 3.9269053567354497e-07,
+      "loss": 5.1064,
+      "step": 17815
+    },
+    {
+      "epoch": 3.4954884268340525,
+      "grad_norm": 13.671051025390625,
+      "learning_rate": 3.9118835496024685e-07,
+      "loss": 4.6874,
+      "step": 17820
+    },
+    {
+      "epoch": 3.4964692036092586,
+      "grad_norm": 22.13825035095215,
+      "learning_rate": 3.896889359627498e-07,
+      "loss": 4.3539,
+      "step": 17825
+    },
+    {
+      "epoch": 3.4974499803844643,
+      "grad_norm": 20.894668579101562,
+      "learning_rate": 3.8819227957954173e-07,
+      "loss": 4.6068,
+      "step": 17830
+    },
+    {
+      "epoch": 3.4984307571596704,
+      "grad_norm": 21.378633499145508,
+      "learning_rate": 3.866983867074575e-07,
+      "loss": 4.5642,
+      "step": 17835
+    },
+    {
+      "epoch": 3.4994115339348766,
+      "grad_norm": 25.99394416809082,
+      "learning_rate": 3.852072582416766e-07,
+      "loss": 4.5547,
+      "step": 17840
+    },
+    {
+      "epoch": 3.5003923107100823,
+      "grad_norm": 18.724058151245117,
+      "learning_rate": 3.8371889507571925e-07,
+      "loss": 4.501,
+      "step": 17845
+    },
+    {
+      "epoch": 3.5013730874852884,
+      "grad_norm": 23.53910255432129,
+      "learning_rate": 3.8223329810145035e-07,
+      "loss": 4.3075,
+      "step": 17850
+    },
+    {
+      "epoch": 3.5013730874852884,
+      "eval_loss": 4.851827621459961,
+      "eval_runtime": 7.6055,
+      "eval_samples_per_second": 27.48,
+      "eval_steps_per_second": 13.806,
+      "step": 17850
+    },
+    {
+      "epoch": 3.5023538642604946,
+      "grad_norm": 21.512908935546875,
+      "learning_rate": 3.8075046820907623e-07,
+      "loss": 4.3834,
+      "step": 17855
+    },
+    {
+      "epoch": 3.5033346410357002,
+      "grad_norm": 13.204446792602539,
+      "learning_rate": 3.7927040628714663e-07,
+      "loss": 4.4252,
+      "step": 17860
+    },
+    {
+      "epoch": 3.5043154178109064,
+      "grad_norm": 22.651100158691406,
+      "learning_rate": 3.777931132225526e-07,
+      "loss": 4.1096,
+      "step": 17865
+    },
+    {
+      "epoch": 3.505296194586112,
+      "grad_norm": 18.523298263549805,
+      "learning_rate": 3.763185899005234e-07,
+      "loss": 4.5935,
+      "step": 17870
+    },
+    {
+      "epoch": 3.506276971361318,
+      "grad_norm": 21.02290916442871,
+      "learning_rate": 3.7484683720463264e-07,
+      "loss": 4.4319,
+      "step": 17875
+    },
+    {
+      "epoch": 3.5072577481365244,
+      "grad_norm": 19.69526481628418,
+      "learning_rate": 3.7337785601679e-07,
+      "loss": 4.6678,
+      "step": 17880
+    },
+    {
+      "epoch": 3.50823852491173,
+      "grad_norm": 17.148012161254883,
+      "learning_rate": 3.7191164721724573e-07,
+      "loss": 4.4583,
+      "step": 17885
+    },
+    {
+      "epoch": 3.509219301686936,
+      "grad_norm": 17.189929962158203,
+      "learning_rate": 3.704482116845909e-07,
+      "loss": 4.8074,
+      "step": 17890
+    },
+    {
+      "epoch": 3.510200078462142,
+      "grad_norm": 23.94347381591797,
+      "learning_rate": 3.6898755029575016e-07,
+      "loss": 4.4356,
+      "step": 17895
+    },
+    {
+      "epoch": 3.511180855237348,
+      "grad_norm": 25.62436866760254,
+      "learning_rate": 3.675296639259912e-07,
+      "loss": 4.4871,
+      "step": 17900
+    },
+    {
+      "epoch": 3.512161632012554,
+      "grad_norm": 47.62681198120117,
+      "learning_rate": 3.6607455344891464e-07,
+      "loss": 4.6938,
+      "step": 17905
+    },
+    {
+      "epoch": 3.51314240878776,
+      "grad_norm": 12.18612003326416,
+      "learning_rate": 3.646222197364596e-07,
+      "loss": 4.1505,
+      "step": 17910
+    },
+    {
+      "epoch": 3.514123185562966,
+      "grad_norm": 37.44921875,
+      "learning_rate": 3.631726636589006e-07,
+      "loss": 4.2759,
+      "step": 17915
+    },
+    {
+      "epoch": 3.5151039623381717,
+      "grad_norm": 10.986493110656738,
+      "learning_rate": 3.6172588608484936e-07,
+      "loss": 4.4109,
+      "step": 17920
+    },
+    {
+      "epoch": 3.516084739113378,
+      "grad_norm": 29.93390464782715,
+      "learning_rate": 3.602818878812503e-07,
+      "loss": 4.3751,
+      "step": 17925
+    },
+    {
+      "epoch": 3.517065515888584,
+      "grad_norm": 14.553876876831055,
+      "learning_rate": 3.5884066991338283e-07,
+      "loss": 4.3612,
+      "step": 17930
+    },
+    {
+      "epoch": 3.5180462926637897,
+      "grad_norm": 18.669994354248047,
+      "learning_rate": 3.574022330448612e-07,
+      "loss": 4.2946,
+      "step": 17935
+    },
+    {
+      "epoch": 3.519027069438996,
+      "grad_norm": 38.16789627075195,
+      "learning_rate": 3.559665781376348e-07,
+      "loss": 4.5239,
+      "step": 17940
+    },
+    {
+      "epoch": 3.5200078462142015,
+      "grad_norm": 25.857666015625,
+      "learning_rate": 3.5453370605198213e-07,
+      "loss": 4.3706,
+      "step": 17945
+    },
+    {
+      "epoch": 3.5209886229894076,
+      "grad_norm": 15.636646270751953,
+      "learning_rate": 3.531036176465175e-07,
+      "loss": 4.3487,
+      "step": 17950
+    },
+    {
+      "epoch": 3.5219693997646138,
+      "grad_norm": 24.652542114257812,
+      "learning_rate": 3.516763137781842e-07,
+      "loss": 4.7284,
+      "step": 17955
+    },
+    {
+      "epoch": 3.5229501765398195,
+      "grad_norm": 25.52812957763672,
+      "learning_rate": 3.5025179530225995e-07,
+      "loss": 4.369,
+      "step": 17960
+    },
+    {
+      "epoch": 3.5239309533150256,
+      "grad_norm": 24.099510192871094,
+      "learning_rate": 3.4883006307235233e-07,
+      "loss": 4.1875,
+      "step": 17965
+    },
+    {
+      "epoch": 3.5249117300902313,
+      "grad_norm": 20.995121002197266,
+      "learning_rate": 3.474111179403977e-07,
+      "loss": 4.4124,
+      "step": 17970
+    },
+    {
+      "epoch": 3.5258925068654374,
+      "grad_norm": 18.973020553588867,
+      "learning_rate": 3.4599496075666484e-07,
+      "loss": 4.6197,
+      "step": 17975
+    },
+    {
+      "epoch": 3.5268732836406436,
+      "grad_norm": 37.977508544921875,
+      "learning_rate": 3.445815923697499e-07,
+      "loss": 4.5785,
+      "step": 17980
+    },
+    {
+      "epoch": 3.5278540604158493,
+      "grad_norm": 10.61478328704834,
+      "learning_rate": 3.431710136265792e-07,
+      "loss": 4.2225,
+      "step": 17985
+    },
+    {
+      "epoch": 3.5288348371910554,
+      "grad_norm": 21.45575714111328,
+      "learning_rate": 3.4176322537240736e-07,
+      "loss": 4.3511,
+      "step": 17990
+    },
+    {
+      "epoch": 3.529815613966261,
+      "grad_norm": 27.8227596282959,
+      "learning_rate": 3.40358228450815e-07,
+      "loss": 4.5644,
+      "step": 17995
+    },
+    {
+      "epoch": 3.5307963907414672,
+      "grad_norm": 38.02301788330078,
+      "learning_rate": 3.3895602370371374e-07,
+      "loss": 4.4346,
+      "step": 18000
+    },
+    {
+      "epoch": 3.5317771675166734,
+      "grad_norm": 9.561601638793945,
+      "learning_rate": 3.3755661197133747e-07,
+      "loss": 4.3879,
+      "step": 18005
+    },
+    {
+      "epoch": 3.532757944291879,
+      "grad_norm": 14.51889419555664,
+      "learning_rate": 3.361599940922505e-07,
+      "loss": 4.3403,
+      "step": 18010
+    },
+    {
+      "epoch": 3.533738721067085,
+      "grad_norm": 31.724550247192383,
+      "learning_rate": 3.3476617090334174e-07,
+      "loss": 4.2,
+      "step": 18015
+    },
+    {
+      "epoch": 3.534719497842291,
+      "grad_norm": 35.23511505126953,
+      "learning_rate": 3.3337514323982356e-07,
+      "loss": 4.2134,
+      "step": 18020
+    },
+    {
+      "epoch": 3.535700274617497,
+      "grad_norm": 21.88624382019043,
+      "learning_rate": 3.3198691193523593e-07,
+      "loss": 4.7372,
+      "step": 18025
+    },
+    {
+      "epoch": 3.536681051392703,
+      "grad_norm": 22.539533615112305,
+      "learning_rate": 3.3060147782144114e-07,
+      "loss": 4.2245,
+      "step": 18030
+    },
+    {
+      "epoch": 3.537661828167909,
+      "grad_norm": 24.725444793701172,
+      "learning_rate": 3.2921884172862686e-07,
+      "loss": 4.6021,
+      "step": 18035
+    },
+    {
+      "epoch": 3.538642604943115,
+      "grad_norm": 39.704078674316406,
+      "learning_rate": 3.278390044853036e-07,
+      "loss": 4.605,
+      "step": 18040
+    },
+    {
+      "epoch": 3.5396233817183207,
+      "grad_norm": 17.120359420776367,
+      "learning_rate": 3.264619669183033e-07,
+      "loss": 4.4304,
+      "step": 18045
+    },
+    {
+      "epoch": 3.540604158493527,
+      "grad_norm": 20.995365142822266,
+      "learning_rate": 3.250877298527827e-07,
+      "loss": 4.2342,
+      "step": 18050
+    },
+    {
+      "epoch": 3.541584935268733,
+      "grad_norm": 17.42166519165039,
+      "learning_rate": 3.237162941122185e-07,
+      "loss": 4.5169,
+      "step": 18055
+    },
+    {
+      "epoch": 3.5425657120439387,
+      "grad_norm": 31.418668746948242,
+      "learning_rate": 3.2234766051841006e-07,
+      "loss": 4.2783,
+      "step": 18060
+    },
+    {
+      "epoch": 3.543546488819145,
+      "grad_norm": 15.284781455993652,
+      "learning_rate": 3.209818298914763e-07,
+      "loss": 4.4126,
+      "step": 18065
+    },
+    {
+      "epoch": 3.5445272655943505,
+      "grad_norm": 14.738502502441406,
+      "learning_rate": 3.196188030498576e-07,
+      "loss": 4.184,
+      "step": 18070
+    },
+    {
+      "epoch": 3.5455080423695566,
+      "grad_norm": 15.123030662536621,
+      "learning_rate": 3.182585808103139e-07,
+      "loss": 4.4825,
+      "step": 18075
+    },
+    {
+      "epoch": 3.5464888191447628,
+      "grad_norm": 27.406064987182617,
+      "learning_rate": 3.1690116398792435e-07,
+      "loss": 4.2695,
+      "step": 18080
+    },
+    {
+      "epoch": 3.5474695959199685,
+      "grad_norm": 34.00102233886719,
+      "learning_rate": 3.155465533960872e-07,
+      "loss": 4.3696,
+      "step": 18085
+    },
+    {
+      "epoch": 3.5484503726951746,
+      "grad_norm": 23.0728759765625,
+      "learning_rate": 3.1419474984652034e-07,
+      "loss": 4.3947,
+      "step": 18090
+    },
+    {
+      "epoch": 3.5494311494703803,
+      "grad_norm": 23.59056854248047,
+      "learning_rate": 3.128457541492569e-07,
+      "loss": 4.5822,
+      "step": 18095
+    },
+    {
+      "epoch": 3.5504119262455864,
+      "grad_norm": 35.71723175048828,
+      "learning_rate": 3.1149956711265027e-07,
+      "loss": 4.5142,
+      "step": 18100
+    },
+    {
+      "epoch": 3.5513927030207926,
+      "grad_norm": 12.386441230773926,
+      "learning_rate": 3.101561895433686e-07,
+      "loss": 4.4894,
+      "step": 18105
+    },
+    {
+      "epoch": 3.5523734797959983,
+      "grad_norm": 16.826030731201172,
+      "learning_rate": 3.0881562224639726e-07,
+      "loss": 4.2603,
+      "step": 18110
+    },
+    {
+      "epoch": 3.5533542565712044,
+      "grad_norm": 21.321094512939453,
+      "learning_rate": 3.074778660250394e-07,
+      "loss": 4.4259,
+      "step": 18115
+    },
+    {
+      "epoch": 3.55433503334641,
+      "grad_norm": 27.676136016845703,
+      "learning_rate": 3.0614292168091086e-07,
+      "loss": 4.4941,
+      "step": 18120
+    },
+    {
+      "epoch": 3.5553158101216162,
+      "grad_norm": 20.773574829101562,
+      "learning_rate": 3.0481079001394465e-07,
+      "loss": 4.3518,
+      "step": 18125
+    },
+    {
+      "epoch": 3.5562965868968224,
+      "grad_norm": 19.71353530883789,
+      "learning_rate": 3.034814718223861e-07,
+      "loss": 4.2774,
+      "step": 18130
+    },
+    {
+      "epoch": 3.5572773636720285,
+      "grad_norm": 14.87898063659668,
+      "learning_rate": 3.0215496790279853e-07,
+      "loss": 4.5793,
+      "step": 18135
+    },
+    {
+      "epoch": 3.558258140447234,
+      "grad_norm": 11.980874061584473,
+      "learning_rate": 3.0083127905005447e-07,
+      "loss": 4.448,
+      "step": 18140
+    },
+    {
+      "epoch": 3.55923891722244,
+      "grad_norm": 15.427719116210938,
+      "learning_rate": 2.995104060573417e-07,
+      "loss": 4.3879,
+      "step": 18145
+    },
+    {
+      "epoch": 3.560219693997646,
+      "grad_norm": 37.64731979370117,
+      "learning_rate": 2.9819234971616154e-07,
+      "loss": 4.5402,
+      "step": 18150
+    },
+    {
+      "epoch": 3.561200470772852,
+      "grad_norm": 32.419612884521484,
+      "learning_rate": 2.968771108163249e-07,
+      "loss": 4.3902,
+      "step": 18155
+    },
+    {
+      "epoch": 3.5621812475480583,
+      "grad_norm": 20.03985023498535,
+      "learning_rate": 2.9556469014595744e-07,
+      "loss": 4.5519,
+      "step": 18160
+    },
+    {
+      "epoch": 3.563162024323264,
+      "grad_norm": 31.856380462646484,
+      "learning_rate": 2.9425508849149464e-07,
+      "loss": 4.5121,
+      "step": 18165
+    },
+    {
+      "epoch": 3.56414280109847,
+      "grad_norm": 11.391762733459473,
+      "learning_rate": 2.92948306637682e-07,
+      "loss": 4.2529,
+      "step": 18170
+    },
+    {
+      "epoch": 3.565123577873676,
+      "grad_norm": 18.532207489013672,
+      "learning_rate": 2.916443453675766e-07,
+      "loss": 4.4126,
+      "step": 18175
+    },
+    {
+      "epoch": 3.566104354648882,
+      "grad_norm": 20.73578643798828,
+      "learning_rate": 2.903432054625438e-07,
+      "loss": 4.2476,
+      "step": 18180
+    },
+    {
+      "epoch": 3.567085131424088,
+      "grad_norm": 18.467668533325195,
+      "learning_rate": 2.8904488770226003e-07,
+      "loss": 4.5179,
+      "step": 18185
+    },
+    {
+      "epoch": 3.568065908199294,
+      "grad_norm": 15.97587776184082,
+      "learning_rate": 2.877493928647107e-07,
+      "loss": 4.2937,
+      "step": 18190
+    },
+    {
+      "epoch": 3.5690466849745,
+      "grad_norm": 16.574209213256836,
+      "learning_rate": 2.8645672172618766e-07,
+      "loss": 4.6197,
+      "step": 18195
+    },
+    {
+      "epoch": 3.5700274617497056,
+      "grad_norm": 21.534997940063477,
+      "learning_rate": 2.8516687506129294e-07,
+      "loss": 4.3969,
+      "step": 18200
+    },
+    {
+      "epoch": 3.571008238524912,
+      "grad_norm": 21.43780517578125,
+      "learning_rate": 2.8387985364293493e-07,
+      "loss": 4.6026,
+      "step": 18205
+    },
+    {
+      "epoch": 3.571989015300118,
+      "grad_norm": 15.082340240478516,
+      "learning_rate": 2.8259565824232784e-07,
+      "loss": 4.4005,
+      "step": 18210
+    },
+    {
+      "epoch": 3.5729697920753236,
+      "grad_norm": 32.8424186706543,
+      "learning_rate": 2.8131428962899557e-07,
+      "loss": 4.2716,
+      "step": 18215
+    },
+    {
+      "epoch": 3.5739505688505298,
+      "grad_norm": 14.402861595153809,
+      "learning_rate": 2.80035748570765e-07,
+      "loss": 4.2866,
+      "step": 18220
+    },
+    {
+      "epoch": 3.5749313456257354,
+      "grad_norm": 27.641376495361328,
+      "learning_rate": 2.7876003583377165e-07,
+      "loss": 4.5308,
+      "step": 18225
+    },
+    {
+      "epoch": 3.5759121224009416,
+      "grad_norm": 21.16995620727539,
+      "learning_rate": 2.7748715218245346e-07,
+      "loss": 4.6729,
+      "step": 18230
+    },
+    {
+      "epoch": 3.5768928991761477,
+      "grad_norm": 17.894859313964844,
+      "learning_rate": 2.762170983795542e-07,
+      "loss": 4.5135,
+      "step": 18235
+    },
+    {
+      "epoch": 3.5778736759513534,
+      "grad_norm": 13.897953987121582,
+      "learning_rate": 2.749498751861229e-07,
+      "loss": 4.2009,
+      "step": 18240
+    },
+    {
+      "epoch": 3.5788544527265596,
+      "grad_norm": 20.38204002380371,
+      "learning_rate": 2.73685483361511e-07,
+      "loss": 4.7626,
+      "step": 18245
+    },
+    {
+      "epoch": 3.5798352295017652,
+      "grad_norm": 29.557910919189453,
+      "learning_rate": 2.7242392366337465e-07,
+      "loss": 4.2005,
+      "step": 18250
+    },
+    {
+      "epoch": 3.5808160062769714,
+      "grad_norm": 16.029287338256836,
+      "learning_rate": 2.711651968476708e-07,
+      "loss": 4.3986,
+      "step": 18255
+    },
+    {
+      "epoch": 3.5817967830521775,
+      "grad_norm": 20.355484008789062,
+      "learning_rate": 2.6990930366866065e-07,
+      "loss": 4.1751,
+      "step": 18260
+    },
+    {
+      "epoch": 3.582777559827383,
+      "grad_norm": 10.150146484375,
+      "learning_rate": 2.686562448789082e-07,
+      "loss": 4.3862,
+      "step": 18265
+    },
+    {
+      "epoch": 3.5837583366025894,
+      "grad_norm": 40.30896759033203,
+      "learning_rate": 2.674060212292756e-07,
+      "loss": 4.4452,
+      "step": 18270
+    },
+    {
+      "epoch": 3.584739113377795,
+      "grad_norm": 20.16090965270996,
+      "learning_rate": 2.661586334689309e-07,
+      "loss": 4.4507,
+      "step": 18275
+    },
+    {
+      "epoch": 3.585719890153001,
+      "grad_norm": 21.508737564086914,
+      "learning_rate": 2.6491408234533834e-07,
+      "loss": 4.0909,
+      "step": 18280
+    },
+    {
+      "epoch": 3.5867006669282073,
+      "grad_norm": 17.113956451416016,
+      "learning_rate": 2.6367236860426414e-07,
+      "loss": 4.4327,
+      "step": 18285
+    },
+    {
+      "epoch": 3.587681443703413,
+      "grad_norm": 11.977952003479004,
+      "learning_rate": 2.624334929897754e-07,
+      "loss": 4.3403,
+      "step": 18290
+    },
+    {
+      "epoch": 3.588662220478619,
+      "grad_norm": 17.363544464111328,
+      "learning_rate": 2.611974562442365e-07,
+      "loss": 4.3587,
+      "step": 18295
+    },
+    {
+      "epoch": 3.589642997253825,
+      "grad_norm": 41.32529830932617,
+      "learning_rate": 2.599642591083129e-07,
+      "loss": 4.3573,
+      "step": 18300
+    },
+    {
+      "epoch": 3.590623774029031,
+      "grad_norm": 16.15401840209961,
+      "learning_rate": 2.587339023209662e-07,
+      "loss": 4.3301,
+      "step": 18305
+    },
+    {
+      "epoch": 3.591604550804237,
+      "grad_norm": 51.65797424316406,
+      "learning_rate": 2.575063866194577e-07,
+      "loss": 4.4128,
+      "step": 18310
+    },
+    {
+      "epoch": 3.592585327579443,
+      "grad_norm": 21.297771453857422,
+      "learning_rate": 2.5628171273934635e-07,
+      "loss": 4.3364,
+      "step": 18315
+    },
+    {
+      "epoch": 3.593566104354649,
+      "grad_norm": 27.42387580871582,
+      "learning_rate": 2.550598814144861e-07,
+      "loss": 4.3562,
+      "step": 18320
+    },
+    {
+      "epoch": 3.5945468811298547,
+      "grad_norm": 48.64586639404297,
+      "learning_rate": 2.538408933770303e-07,
+      "loss": 3.9871,
+      "step": 18325
+    },
+    {
+      "epoch": 3.595527657905061,
+      "grad_norm": 19.293163299560547,
+      "learning_rate": 2.5262474935742574e-07,
+      "loss": 4.2953,
+      "step": 18330
+    },
+    {
+      "epoch": 3.596508434680267,
+      "grad_norm": 21.172189712524414,
+      "learning_rate": 2.514114500844178e-07,
+      "loss": 4.5203,
+      "step": 18335
+    },
+    {
+      "epoch": 3.5974892114554726,
+      "grad_norm": 16.0252685546875,
+      "learning_rate": 2.5020099628504603e-07,
+      "loss": 4.3533,
+      "step": 18340
+    },
+    {
+      "epoch": 3.5984699882306788,
+      "grad_norm": 27.545793533325195,
+      "learning_rate": 2.4899338868464404e-07,
+      "loss": 4.684,
+      "step": 18345
+    },
+    {
+      "epoch": 3.5994507650058845,
+      "grad_norm": 16.568904876708984,
+      "learning_rate": 2.4778862800684034e-07,
+      "loss": 4.3269,
+      "step": 18350
+    },
+    {
+      "epoch": 3.6004315417810906,
+      "grad_norm": 27.713424682617188,
+      "learning_rate": 2.4658671497355847e-07,
+      "loss": 4.4722,
+      "step": 18355
+    },
+    {
+      "epoch": 3.6014123185562967,
+      "grad_norm": 24.068340301513672,
+      "learning_rate": 2.4538765030501455e-07,
+      "loss": 4.6103,
+      "step": 18360
+    },
+    {
+      "epoch": 3.6023930953315024,
+      "grad_norm": 19.014244079589844,
+      "learning_rate": 2.441914347197194e-07,
+      "loss": 4.4829,
+      "step": 18365
+    },
+    {
+      "epoch": 3.6033738721067086,
+      "grad_norm": 25.846216201782227,
+      "learning_rate": 2.429980689344735e-07,
+      "loss": 4.2449,
+      "step": 18370
+    },
+    {
+      "epoch": 3.6043546488819143,
+      "grad_norm": 17.586687088012695,
+      "learning_rate": 2.4180755366437324e-07,
+      "loss": 4.3699,
+      "step": 18375
+    },
+    {
+      "epoch": 3.6053354256571204,
+      "grad_norm": 17.855295181274414,
+      "learning_rate": 2.406198896228046e-07,
+      "loss": 4.4747,
+      "step": 18380
+    },
+    {
+      "epoch": 3.6063162024323265,
+      "grad_norm": 19.772993087768555,
+      "learning_rate": 2.3943507752144546e-07,
+      "loss": 4.7973,
+      "step": 18385
+    },
+    {
+      "epoch": 3.6072969792075322,
+      "grad_norm": 9.745254516601562,
+      "learning_rate": 2.382531180702663e-07,
+      "loss": 4.494,
+      "step": 18390
+    },
+    {
+      "epoch": 3.6082777559827384,
+      "grad_norm": 25.192859649658203,
+      "learning_rate": 2.3707401197752556e-07,
+      "loss": 4.536,
+      "step": 18395
+    },
+    {
+      "epoch": 3.609258532757944,
+      "grad_norm": 22.296049118041992,
+      "learning_rate": 2.3589775994977416e-07,
+      "loss": 4.4789,
+      "step": 18400
+    },
+    {
+      "epoch": 3.61023930953315,
+      "grad_norm": 10.856245040893555,
+      "learning_rate": 2.3472436269185105e-07,
+      "loss": 4.4375,
+      "step": 18405
+    },
+    {
+      "epoch": 3.6112200863083563,
+      "grad_norm": 13.569968223571777,
+      "learning_rate": 2.3355382090688605e-07,
+      "loss": 4.1242,
+      "step": 18410
+    },
+    {
+      "epoch": 3.612200863083562,
+      "grad_norm": 30.028282165527344,
+      "learning_rate": 2.323861352962975e-07,
+      "loss": 4.7917,
+      "step": 18415
+    },
+    {
+      "epoch": 3.613181639858768,
+      "grad_norm": 21.72928237915039,
+      "learning_rate": 2.3122130655979124e-07,
+      "loss": 4.6841,
+      "step": 18420
+    },
+    {
+      "epoch": 3.614162416633974,
+      "grad_norm": 24.678524017333984,
+      "learning_rate": 2.3005933539536118e-07,
+      "loss": 4.2326,
+      "step": 18425
+    },
+    {
+      "epoch": 3.61514319340918,
+      "grad_norm": 30.336259841918945,
+      "learning_rate": 2.28900222499292e-07,
+      "loss": 3.8564,
+      "step": 18430
+    },
+    {
+      "epoch": 3.616123970184386,
+      "grad_norm": 21.32823944091797,
+      "learning_rate": 2.277439685661509e-07,
+      "loss": 4.8797,
+      "step": 18435
+    },
+    {
+      "epoch": 3.617104746959592,
+      "grad_norm": 29.761159896850586,
+      "learning_rate": 2.2659057428879584e-07,
+      "loss": 4.8908,
+      "step": 18440
+    },
+    {
+      "epoch": 3.618085523734798,
+      "grad_norm": 14.10708236694336,
+      "learning_rate": 2.2544004035836897e-07,
+      "loss": 4.4972,
+      "step": 18445
+    },
+    {
+      "epoch": 3.6190663005100037,
+      "grad_norm": 16.236936569213867,
+      "learning_rate": 2.2429236746429938e-07,
+      "loss": 4.4356,
+      "step": 18450
+    },
+    {
+      "epoch": 3.62004707728521,
+      "grad_norm": 11.860494613647461,
+      "learning_rate": 2.231475562943014e-07,
+      "loss": 4.5097,
+      "step": 18455
+    },
+    {
+      "epoch": 3.621027854060416,
+      "grad_norm": 29.234317779541016,
+      "learning_rate": 2.2200560753437462e-07,
+      "loss": 4.6355,
+      "step": 18460
+    },
+    {
+      "epoch": 3.622008630835622,
+      "grad_norm": 29.146080017089844,
+      "learning_rate": 2.2086652186880386e-07,
+      "loss": 4.8315,
+      "step": 18465
+    },
+    {
+      "epoch": 3.6229894076108278,
+      "grad_norm": 22.21369171142578,
+      "learning_rate": 2.1973029998015703e-07,
+      "loss": 4.7367,
+      "step": 18470
+    },
+    {
+      "epoch": 3.6239701843860335,
+      "grad_norm": 12.365029335021973,
+      "learning_rate": 2.1859694254928844e-07,
+      "loss": 4.5659,
+      "step": 18475
+    },
+    {
+      "epoch": 3.6249509611612396,
+      "grad_norm": 14.387511253356934,
+      "learning_rate": 2.1746645025533198e-07,
+      "loss": 4.2935,
+      "step": 18480
+    },
+    {
+      "epoch": 3.6259317379364457,
+      "grad_norm": 24.00560188293457,
+      "learning_rate": 2.1633882377570913e-07,
+      "loss": 4.4385,
+      "step": 18485
+    },
+    {
+      "epoch": 3.626912514711652,
+      "grad_norm": 20.93753433227539,
+      "learning_rate": 2.1521406378612164e-07,
+      "loss": 4.5372,
+      "step": 18490
+    },
+    {
+      "epoch": 3.6278932914868576,
+      "grad_norm": 35.840152740478516,
+      "learning_rate": 2.1409217096055311e-07,
+      "loss": 4.5068,
+      "step": 18495
+    },
+    {
+      "epoch": 3.6288740682620637,
+      "grad_norm": 19.95467185974121,
+      "learning_rate": 2.1297314597127082e-07,
+      "loss": 4.4501,
+      "step": 18500
+    },
+    {
+      "epoch": 3.6298548450372694,
+      "grad_norm": 15.141322135925293,
+      "learning_rate": 2.118569894888217e-07,
+      "loss": 4.7295,
+      "step": 18505
+    },
+    {
+      "epoch": 3.6308356218124755,
+      "grad_norm": 19.499128341674805,
+      "learning_rate": 2.1074370218203522e-07,
+      "loss": 4.6502,
+      "step": 18510
+    },
+    {
+      "epoch": 3.6318163985876817,
+      "grad_norm": 24.709644317626953,
+      "learning_rate": 2.0963328471802213e-07,
+      "loss": 4.678,
+      "step": 18515
+    },
+    {
+      "epoch": 3.6327971753628874,
+      "grad_norm": 12.919239044189453,
+      "learning_rate": 2.0852573776217078e-07,
+      "loss": 4.3808,
+      "step": 18520
+    },
+    {
+      "epoch": 3.6337779521380935,
+      "grad_norm": 27.91121482849121,
+      "learning_rate": 2.0742106197815304e-07,
+      "loss": 4.6819,
+      "step": 18525
+    },
+    {
+      "epoch": 3.634758728913299,
+      "grad_norm": 26.32551383972168,
+      "learning_rate": 2.0631925802791608e-07,
+      "loss": 4.8111,
+      "step": 18530
+    },
+    {
+      "epoch": 3.6357395056885053,
+      "grad_norm": 10.567716598510742,
+      "learning_rate": 2.0522032657169012e-07,
+      "loss": 4.6578,
+      "step": 18535
+    },
+    {
+      "epoch": 3.6367202824637115,
+      "grad_norm": 17.105424880981445,
+      "learning_rate": 2.0412426826798283e-07,
+      "loss": 4.5804,
+      "step": 18540
+    },
+    {
+      "epoch": 3.637701059238917,
+      "grad_norm": 22.463489532470703,
+      "learning_rate": 2.0303108377357827e-07,
+      "loss": 4.5606,
+      "step": 18545
+    },
+    {
+      "epoch": 3.6386818360141233,
+      "grad_norm": 19.65961265563965,
+      "learning_rate": 2.0194077374354248e-07,
+      "loss": 4.297,
+      "step": 18550
+    },
+    {
+      "epoch": 3.639662612789329,
+      "grad_norm": 12.41259765625,
+      "learning_rate": 2.0085333883121393e-07,
+      "loss": 4.2794,
+      "step": 18555
+    },
+    {
+      "epoch": 3.640643389564535,
+      "grad_norm": 14.857327461242676,
+      "learning_rate": 1.9976877968821306e-07,
+      "loss": 4.2883,
+      "step": 18560
+    },
+    {
+      "epoch": 3.6416241663397413,
+      "grad_norm": 34.63252258300781,
+      "learning_rate": 1.9868709696443334e-07,
+      "loss": 4.5432,
+      "step": 18565
+    },
+    {
+      "epoch": 3.642604943114947,
+      "grad_norm": 17.150300979614258,
+      "learning_rate": 1.9760829130804794e-07,
+      "loss": 4.344,
+      "step": 18570
+    },
+    {
+      "epoch": 3.643585719890153,
+      "grad_norm": 15.355920791625977,
+      "learning_rate": 1.9653236336550363e-07,
+      "loss": 4.5685,
+      "step": 18575
+    },
+    {
+      "epoch": 3.644566496665359,
+      "grad_norm": 13.861881256103516,
+      "learning_rate": 1.954593137815225e-07,
+      "loss": 4.301,
+      "step": 18580
+    },
+    {
+      "epoch": 3.645547273440565,
+      "grad_norm": 52.94959259033203,
+      "learning_rate": 1.943891431991035e-07,
+      "loss": 4.2231,
+      "step": 18585
+    },
+    {
+      "epoch": 3.646528050215771,
+      "grad_norm": 15.73333740234375,
+      "learning_rate": 1.9332185225952092e-07,
+      "loss": 4.49,
+      "step": 18590
+    },
+    {
+      "epoch": 3.647508826990977,
+      "grad_norm": 23.316686630249023,
+      "learning_rate": 1.922574416023204e-07,
+      "loss": 4.7615,
+      "step": 18595
+    },
+    {
+      "epoch": 3.648489603766183,
+      "grad_norm": 15.769940376281738,
+      "learning_rate": 1.9119591186532506e-07,
+      "loss": 4.3419,
+      "step": 18600
+    },
+    {
+      "epoch": 3.6494703805413886,
+      "grad_norm": 19.157127380371094,
+      "learning_rate": 1.9013726368462937e-07,
+      "loss": 4.437,
+      "step": 18605
+    },
+    {
+      "epoch": 3.6504511573165948,
+      "grad_norm": 20.292354583740234,
+      "learning_rate": 1.8908149769460204e-07,
+      "loss": 4.4277,
+      "step": 18610
+    },
+    {
+      "epoch": 3.651431934091801,
+      "grad_norm": 13.844761848449707,
+      "learning_rate": 1.880286145278848e-07,
+      "loss": 4.3454,
+      "step": 18615
+    },
+    {
+      "epoch": 3.6524127108670066,
+      "grad_norm": 16.793840408325195,
+      "learning_rate": 1.8697861481539182e-07,
+      "loss": 4.2819,
+      "step": 18620
+    },
+    {
+      "epoch": 3.6533934876422127,
+      "grad_norm": 14.200879096984863,
+      "learning_rate": 1.8593149918630927e-07,
+      "loss": 4.5943,
+      "step": 18625
+    },
+    {
+      "epoch": 3.6543742644174184,
+      "grad_norm": 15.40435791015625,
+      "learning_rate": 1.848872682680941e-07,
+      "loss": 4.6043,
+      "step": 18630
+    },
+    {
+      "epoch": 3.6553550411926246,
+      "grad_norm": 19.25347328186035,
+      "learning_rate": 1.8384592268647806e-07,
+      "loss": 4.5762,
+      "step": 18635
+    },
+    {
+      "epoch": 3.6563358179678307,
+      "grad_norm": 22.764402389526367,
+      "learning_rate": 1.828074630654597e-07,
+      "loss": 4.8733,
+      "step": 18640
+    },
+    {
+      "epoch": 3.6573165947430364,
+      "grad_norm": 17.222026824951172,
+      "learning_rate": 1.8177189002731021e-07,
+      "loss": 4.3624,
+      "step": 18645
+    },
+    {
+      "epoch": 3.6582973715182425,
+      "grad_norm": 15.655689239501953,
+      "learning_rate": 1.8073920419257208e-07,
+      "loss": 4.3877,
+      "step": 18650
+    },
+    {
+      "epoch": 3.659278148293448,
+      "grad_norm": 15.489755630493164,
+      "learning_rate": 1.7970940618005528e-07,
+      "loss": 4.3674,
+      "step": 18655
+    },
+    {
+      "epoch": 3.6602589250686544,
+      "grad_norm": 21.11366844177246,
+      "learning_rate": 1.7868249660684123e-07,
+      "loss": 4.5259,
+      "step": 18660
+    },
+    {
+      "epoch": 3.6612397018438605,
+      "grad_norm": 23.487411499023438,
+      "learning_rate": 1.7765847608828047e-07,
+      "loss": 4.3651,
+      "step": 18665
+    },
+    {
+      "epoch": 3.662220478619066,
+      "grad_norm": 25.111196517944336,
+      "learning_rate": 1.7663734523799104e-07,
+      "loss": 4.5454,
+      "step": 18670
+    },
+    {
+      "epoch": 3.6632012553942723,
+      "grad_norm": 44.54734420776367,
+      "learning_rate": 1.7561910466786125e-07,
+      "loss": 4.9847,
+      "step": 18675
+    },
+    {
+      "epoch": 3.664182032169478,
+      "grad_norm": 24.96506118774414,
+      "learning_rate": 1.7460375498804527e-07,
+      "loss": 4.2462,
+      "step": 18680
+    },
+    {
+      "epoch": 3.665162808944684,
+      "grad_norm": 14.811649322509766,
+      "learning_rate": 1.7359129680696696e-07,
+      "loss": 4.7993,
+      "step": 18685
+    },
+    {
+      "epoch": 3.6661435857198903,
+      "grad_norm": 20.218128204345703,
+      "learning_rate": 1.7258173073131658e-07,
+      "loss": 4.2195,
+      "step": 18690
+    },
+    {
+      "epoch": 3.667124362495096,
+      "grad_norm": 27.616926193237305,
+      "learning_rate": 1.715750573660513e-07,
+      "loss": 4.3696,
+      "step": 18695
+    },
+    {
+      "epoch": 3.668105139270302,
+      "grad_norm": 19.38068962097168,
+      "learning_rate": 1.7057127731439526e-07,
+      "loss": 4.362,
+      "step": 18700
+    },
+    {
+      "epoch": 3.669085916045508,
+      "grad_norm": 15.382796287536621,
+      "learning_rate": 1.6957039117783848e-07,
+      "loss": 4.6597,
+      "step": 18705
+    },
+    {
+      "epoch": 3.670066692820714,
+      "grad_norm": 14.528923034667969,
+      "learning_rate": 1.6857239955613724e-07,
+      "loss": 4.5035,
+      "step": 18710
+    },
+    {
+      "epoch": 3.67104746959592,
+      "grad_norm": 26.649324417114258,
+      "learning_rate": 1.6757730304731378e-07,
+      "loss": 4.6962,
+      "step": 18715
+    },
+    {
+      "epoch": 3.672028246371126,
+      "grad_norm": 33.269126892089844,
+      "learning_rate": 1.6658510224765333e-07,
+      "loss": 4.4128,
+      "step": 18720
+    },
+    {
+      "epoch": 3.673009023146332,
+      "grad_norm": 17.99794578552246,
+      "learning_rate": 1.655957977517092e-07,
+      "loss": 4.7475,
+      "step": 18725
+    },
+    {
+      "epoch": 3.6739897999215376,
+      "grad_norm": 13.681337356567383,
+      "learning_rate": 1.646093901522966e-07,
+      "loss": 4.6715,
+      "step": 18730
+    },
+    {
+      "epoch": 3.6749705766967438,
+      "grad_norm": 22.046131134033203,
+      "learning_rate": 1.6362588004049606e-07,
+      "loss": 4.3841,
+      "step": 18735
+    },
+    {
+      "epoch": 3.67595135347195,
+      "grad_norm": 13.013222694396973,
+      "learning_rate": 1.6264526800565228e-07,
+      "loss": 4.026,
+      "step": 18740
+    },
+    {
+      "epoch": 3.6769321302471556,
+      "grad_norm": 15.13515853881836,
+      "learning_rate": 1.616675546353713e-07,
+      "loss": 4.3175,
+      "step": 18745
+    },
+    {
+      "epoch": 3.6779129070223617,
+      "grad_norm": 14.422954559326172,
+      "learning_rate": 1.6069274051552453e-07,
+      "loss": 4.3712,
+      "step": 18750
+    },
+    {
+      "epoch": 3.6788936837975674,
+      "grad_norm": 22.292251586914062,
+      "learning_rate": 1.597208262302441e-07,
+      "loss": 4.7974,
+      "step": 18755
+    },
+    {
+      "epoch": 3.6798744605727736,
+      "grad_norm": 23.049213409423828,
+      "learning_rate": 1.5875181236192638e-07,
+      "loss": 4.3128,
+      "step": 18760
+    },
+    {
+      "epoch": 3.6808552373479797,
+      "grad_norm": 14.684503555297852,
+      "learning_rate": 1.5778569949122914e-07,
+      "loss": 4.3054,
+      "step": 18765
+    },
+    {
+      "epoch": 3.6818360141231854,
+      "grad_norm": 24.637575149536133,
+      "learning_rate": 1.5682248819707036e-07,
+      "loss": 4.4834,
+      "step": 18770
+    },
+    {
+      "epoch": 3.6828167908983915,
+      "grad_norm": 24.20187759399414,
+      "learning_rate": 1.558621790566317e-07,
+      "loss": 4.3936,
+      "step": 18775
+    },
+    {
+      "epoch": 3.6837975676735972,
+      "grad_norm": 12.726591110229492,
+      "learning_rate": 1.549047726453534e-07,
+      "loss": 4.504,
+      "step": 18780
+    },
+    {
+      "epoch": 3.6847783444488034,
+      "grad_norm": 20.728734970092773,
+      "learning_rate": 1.5395026953693826e-07,
+      "loss": 4.2898,
+      "step": 18785
+    },
+    {
+      "epoch": 3.6857591212240095,
+      "grad_norm": 20.017847061157227,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 4.4186,
+      "step": 18790
+    },
+    {
+      "epoch": 3.6867398979992156,
+      "grad_norm": 13.246440887451172,
+      "learning_rate": 1.5204997551480527e-07,
+      "loss": 4.5196,
+      "step": 18795
+    },
+    {
+      "epoch": 3.6877206747744213,
+      "grad_norm": 19.194934844970703,
+      "learning_rate": 1.5110418573979157e-07,
+      "loss": 4.4459,
+      "step": 18800
+    },
+    {
+      "epoch": 3.688701451549627,
+      "grad_norm": 14.851069450378418,
+      "learning_rate": 1.501613015450476e-07,
+      "loss": 4.4218,
+      "step": 18805
+    },
+    {
+      "epoch": 3.689682228324833,
+      "grad_norm": 14.522515296936035,
+      "learning_rate": 1.492213234955736e-07,
+      "loss": 4.6519,
+      "step": 18810
+    },
+    {
+      "epoch": 3.6906630051000393,
+      "grad_norm": 16.660293579101562,
+      "learning_rate": 1.482842521546285e-07,
+      "loss": 4.3321,
+      "step": 18815
+    },
+    {
+      "epoch": 3.6916437818752454,
+      "grad_norm": 19.738510131835938,
+      "learning_rate": 1.4735008808372808e-07,
+      "loss": 4.32,
+      "step": 18820
+    },
+    {
+      "epoch": 3.692624558650451,
+      "grad_norm": 35.67677307128906,
+      "learning_rate": 1.4641883184264794e-07,
+      "loss": 4.2388,
+      "step": 18825
+    },
+    {
+      "epoch": 3.6936053354256573,
+      "grad_norm": 22.168577194213867,
+      "learning_rate": 1.454904839894189e-07,
+      "loss": 4.2093,
+      "step": 18830
+    },
+    {
+      "epoch": 3.694586112200863,
+      "grad_norm": 13.524019241333008,
+      "learning_rate": 1.4456504508033152e-07,
+      "loss": 4.4022,
+      "step": 18835
+    },
+    {
+      "epoch": 3.695566888976069,
+      "grad_norm": 15.70751667022705,
+      "learning_rate": 1.4364251566993225e-07,
+      "loss": 4.4863,
+      "step": 18840
+    },
+    {
+      "epoch": 3.6965476657512752,
+      "grad_norm": 11.819734573364258,
+      "learning_rate": 1.4272289631102276e-07,
+      "loss": 4.261,
+      "step": 18845
+    },
+    {
+      "epoch": 3.697528442526481,
+      "grad_norm": 15.136698722839355,
+      "learning_rate": 1.418061875546628e-07,
+      "loss": 4.4006,
+      "step": 18850
+    },
+    {
+      "epoch": 3.698509219301687,
+      "grad_norm": 13.050637245178223,
+      "learning_rate": 1.408923899501674e-07,
+      "loss": 4.5681,
+      "step": 18855
+    },
+    {
+      "epoch": 3.6994899960768928,
+      "grad_norm": 18.412694931030273,
+      "learning_rate": 1.3998150404510635e-07,
+      "loss": 4.6765,
+      "step": 18860
+    },
+    {
+      "epoch": 3.700470772852099,
+      "grad_norm": 20.788856506347656,
+      "learning_rate": 1.390735303853069e-07,
+      "loss": 4.1471,
+      "step": 18865
+    },
+    {
+      "epoch": 3.701451549627305,
+      "grad_norm": 18.045942306518555,
+      "learning_rate": 1.381684695148472e-07,
+      "loss": 4.3271,
+      "step": 18870
+    },
+    {
+      "epoch": 3.7024323264025107,
+      "grad_norm": 14.031227111816406,
+      "learning_rate": 1.3726632197606503e-07,
+      "loss": 4.4308,
+      "step": 18875
+    },
+    {
+      "epoch": 3.703413103177717,
+      "grad_norm": 26.201160430908203,
+      "learning_rate": 1.3636708830954803e-07,
+      "loss": 4.1403,
+      "step": 18880
+    },
+    {
+      "epoch": 3.7043938799529226,
+      "grad_norm": 30.57729721069336,
+      "learning_rate": 1.354707690541407e-07,
+      "loss": 4.4129,
+      "step": 18885
+    },
+    {
+      "epoch": 3.7053746567281287,
+      "grad_norm": 23.976205825805664,
+      "learning_rate": 1.3457736474693949e-07,
+      "loss": 4.3689,
+      "step": 18890
+    },
+    {
+      "epoch": 3.706355433503335,
+      "grad_norm": 24.748432159423828,
+      "learning_rate": 1.336868759232951e-07,
+      "loss": 4.29,
+      "step": 18895
+    },
+    {
+      "epoch": 3.7073362102785405,
+      "grad_norm": 19.802711486816406,
+      "learning_rate": 1.3279930311681123e-07,
+      "loss": 4.6155,
+      "step": 18900
+    },
+    {
+      "epoch": 3.7083169870537467,
+      "grad_norm": 14.154813766479492,
+      "learning_rate": 1.3191464685934241e-07,
+      "loss": 4.4138,
+      "step": 18905
+    },
+    {
+      "epoch": 3.7092977638289524,
+      "grad_norm": 15.545702934265137,
+      "learning_rate": 1.3103290768099796e-07,
+      "loss": 4.571,
+      "step": 18910
+    },
+    {
+      "epoch": 3.7102785406041585,
+      "grad_norm": 56.756019592285156,
+      "learning_rate": 1.30154086110138e-07,
+      "loss": 4.1372,
+      "step": 18915
+    },
+    {
+      "epoch": 3.7112593173793647,
+      "grad_norm": 19.387264251708984,
+      "learning_rate": 1.2927818267337468e-07,
+      "loss": 4.5391,
+      "step": 18920
+    },
+    {
+      "epoch": 3.7122400941545703,
+      "grad_norm": 24.630348205566406,
+      "learning_rate": 1.2840519789557137e-07,
+      "loss": 4.2267,
+      "step": 18925
+    },
+    {
+      "epoch": 3.7132208709297765,
+      "grad_norm": 24.16886329650879,
+      "learning_rate": 1.275351322998425e-07,
+      "loss": 4.6744,
+      "step": 18930
+    },
+    {
+      "epoch": 3.714201647704982,
+      "grad_norm": 21.653339385986328,
+      "learning_rate": 1.2666798640755206e-07,
+      "loss": 4.3333,
+      "step": 18935
+    },
+    {
+      "epoch": 3.7151824244801883,
+      "grad_norm": 17.944013595581055,
+      "learning_rate": 1.2580376073831723e-07,
+      "loss": 4.4333,
+      "step": 18940
+    },
+    {
+      "epoch": 3.7161632012553945,
+      "grad_norm": 20.188127517700195,
+      "learning_rate": 1.249424558100032e-07,
+      "loss": 4.264,
+      "step": 18945
+    },
+    {
+      "epoch": 3.7171439780306,
+      "grad_norm": 25.705101013183594,
+      "learning_rate": 1.2408407213872543e-07,
+      "loss": 4.4369,
+      "step": 18950
+    },
+    {
+      "epoch": 3.7181247548058063,
+      "grad_norm": 21.701995849609375,
+      "learning_rate": 1.2322861023884859e-07,
+      "loss": 4.365,
+      "step": 18955
+    },
+    {
+      "epoch": 3.719105531581012,
+      "grad_norm": 12.588302612304688,
+      "learning_rate": 1.2237607062298762e-07,
+      "loss": 4.0742,
+      "step": 18960
+    },
+    {
+      "epoch": 3.720086308356218,
+      "grad_norm": 24.353504180908203,
+      "learning_rate": 1.215264538020061e-07,
+      "loss": 4.7479,
+      "step": 18965
+    },
+    {
+      "epoch": 3.7210670851314243,
+      "grad_norm": 11.862863540649414,
+      "learning_rate": 1.206797602850146e-07,
+      "loss": 4.3784,
+      "step": 18970
+    },
+    {
+      "epoch": 3.72204786190663,
+      "grad_norm": 23.318727493286133,
+      "learning_rate": 1.198359905793739e-07,
+      "loss": 4.3121,
+      "step": 18975
+    },
+    {
+      "epoch": 3.723028638681836,
+      "grad_norm": 13.879663467407227,
+      "learning_rate": 1.1899514519069177e-07,
+      "loss": 4.0607,
+      "step": 18980
+    },
+    {
+      "epoch": 3.724009415457042,
+      "grad_norm": 14.94577693939209,
+      "learning_rate": 1.1815722462282409e-07,
+      "loss": 4.4307,
+      "step": 18985
+    },
+    {
+      "epoch": 3.724990192232248,
+      "grad_norm": 37.08789825439453,
+      "learning_rate": 1.1732222937787419e-07,
+      "loss": 4.5394,
+      "step": 18990
+    },
+    {
+      "epoch": 3.725970969007454,
+      "grad_norm": 21.196096420288086,
+      "learning_rate": 1.1649015995619184e-07,
+      "loss": 4.6473,
+      "step": 18995
+    },
+    {
+      "epoch": 3.7269517457826598,
+      "grad_norm": 16.924158096313477,
+      "learning_rate": 1.156610168563732e-07,
+      "loss": 4.4108,
+      "step": 19000
+    },
+    {
+      "epoch": 3.727932522557866,
+      "grad_norm": 35.096519470214844,
+      "learning_rate": 1.1483480057526364e-07,
+      "loss": 4.546,
+      "step": 19005
+    },
+    {
+      "epoch": 3.7289132993330716,
+      "grad_norm": 10.689122200012207,
+      "learning_rate": 1.1401151160795044e-07,
+      "loss": 4.3182,
+      "step": 19010
+    },
+    {
+      "epoch": 3.7298940761082777,
+      "grad_norm": 22.799142837524414,
+      "learning_rate": 1.1319115044777063e-07,
+      "loss": 4.3949,
+      "step": 19015
+    },
+    {
+      "epoch": 3.730874852883484,
+      "grad_norm": 14.333352088928223,
+      "learning_rate": 1.1237371758630488e-07,
+      "loss": 4.5137,
+      "step": 19020
+    },
+    {
+      "epoch": 3.7318556296586896,
+      "grad_norm": 40.326988220214844,
+      "learning_rate": 1.1155921351337917e-07,
+      "loss": 4.5281,
+      "step": 19025
+    },
+    {
+      "epoch": 3.7328364064338957,
+      "grad_norm": 21.95504379272461,
+      "learning_rate": 1.1074763871706473e-07,
+      "loss": 4.2508,
+      "step": 19030
+    },
+    {
+      "epoch": 3.7338171832091014,
+      "grad_norm": 18.045551300048828,
+      "learning_rate": 1.0993899368367811e-07,
+      "loss": 4.2403,
+      "step": 19035
+    },
+    {
+      "epoch": 3.7347979599843075,
+      "grad_norm": 11.443171501159668,
+      "learning_rate": 1.0913327889777948e-07,
+      "loss": 4.351,
+      "step": 19040
+    },
+    {
+      "epoch": 3.7357787367595137,
+      "grad_norm": 13.363077163696289,
+      "learning_rate": 1.0833049484217373e-07,
+      "loss": 4.344,
+      "step": 19045
+    },
+    {
+      "epoch": 3.7367595135347194,
+      "grad_norm": 18.82840919494629,
+      "learning_rate": 1.0753064199790886e-07,
+      "loss": 4.2728,
+      "step": 19050
+    },
+    {
+      "epoch": 3.7377402903099255,
+      "grad_norm": 16.456222534179688,
+      "learning_rate": 1.0673372084427646e-07,
+      "loss": 4.3274,
+      "step": 19055
+    },
+    {
+      "epoch": 3.738721067085131,
+      "grad_norm": 15.155776023864746,
+      "learning_rate": 1.0593973185881179e-07,
+      "loss": 4.5719,
+      "step": 19060
+    },
+    {
+      "epoch": 3.7397018438603373,
+      "grad_norm": 14.27761173248291,
+      "learning_rate": 1.0514867551729424e-07,
+      "loss": 4.2692,
+      "step": 19065
+    },
+    {
+      "epoch": 3.7406826206355435,
+      "grad_norm": 12.729808807373047,
+      "learning_rate": 1.0436055229374408e-07,
+      "loss": 4.3371,
+      "step": 19070
+    },
+    {
+      "epoch": 3.741663397410749,
+      "grad_norm": 21.104854583740234,
+      "learning_rate": 1.0357536266042356e-07,
+      "loss": 4.3169,
+      "step": 19075
+    },
+    {
+      "epoch": 3.7426441741859553,
+      "grad_norm": 14.363003730773926,
+      "learning_rate": 1.0279310708783907e-07,
+      "loss": 4.3119,
+      "step": 19080
+    },
+    {
+      "epoch": 3.743624950961161,
+      "grad_norm": 25.326114654541016,
+      "learning_rate": 1.0201378604473677e-07,
+      "loss": 4.7418,
+      "step": 19085
+    },
+    {
+      "epoch": 3.744605727736367,
+      "grad_norm": 14.007926940917969,
+      "learning_rate": 1.01237399998107e-07,
+      "loss": 4.2135,
+      "step": 19090
+    },
+    {
+      "epoch": 3.7455865045115733,
+      "grad_norm": 15.097603797912598,
+      "learning_rate": 1.0046394941317816e-07,
+      "loss": 4.6373,
+      "step": 19095
+    },
+    {
+      "epoch": 3.7465672812867794,
+      "grad_norm": 18.339405059814453,
+      "learning_rate": 9.969343475342285e-08,
+      "loss": 4.8545,
+      "step": 19100
+    },
+    {
+      "epoch": 3.747548058061985,
+      "grad_norm": 10.937397956848145,
+      "learning_rate": 9.89258564805512e-08,
+      "loss": 4.4624,
+      "step": 19105
+    },
+    {
+      "epoch": 3.748528834837191,
+      "grad_norm": 27.80323600769043,
+      "learning_rate": 9.816121505451692e-08,
+      "loss": 4.6097,
+      "step": 19110
+    },
+    {
+      "epoch": 3.749509611612397,
+      "grad_norm": 16.348608016967773,
+      "learning_rate": 9.73995109335113e-08,
+      "loss": 4.2309,
+      "step": 19115
+    },
+    {
+      "epoch": 3.750490388387603,
+      "grad_norm": 19.31072998046875,
+      "learning_rate": 9.664074457396699e-08,
+      "loss": 4.4816,
+      "step": 19120
+    },
+    {
+      "epoch": 3.751471165162809,
+      "grad_norm": 26.25499153137207,
+      "learning_rate": 9.588491643055642e-08,
+      "loss": 4.4299,
+      "step": 19125
+    },
+    {
+      "epoch": 3.751471165162809,
+      "eval_loss": 4.851955413818359,
+      "eval_runtime": 7.7483,
+      "eval_samples_per_second": 26.974,
+      "eval_steps_per_second": 13.551,
+      "step": 19125
+    },
+    {
+      "epoch": 3.752451941938015,
+      "grad_norm": 26.895282745361328,
+      "learning_rate": 9.513202695618951e-08,
+      "loss": 4.6251,
+      "step": 19130
+    },
+    {
+      "epoch": 3.7534327187132206,
+      "grad_norm": 13.861442565917969,
+      "learning_rate": 9.438207660201759e-08,
+      "loss": 4.6765,
+      "step": 19135
+    },
+    {
+      "epoch": 3.7544134954884267,
+      "grad_norm": 38.26374053955078,
+      "learning_rate": 9.363506581742953e-08,
+      "loss": 4.5282,
+      "step": 19140
+    },
+    {
+      "epoch": 3.755394272263633,
+      "grad_norm": 22.872318267822266,
+      "learning_rate": 9.289099505005339e-08,
+      "loss": 4.2434,
+      "step": 19145
+    },
+    {
+      "epoch": 3.756375049038839,
+      "grad_norm": 30.124589920043945,
+      "learning_rate": 9.214986474575471e-08,
+      "loss": 4.2806,
+      "step": 19150
+    },
+    {
+      "epoch": 3.7573558258140447,
+      "grad_norm": 30.88665771484375,
+      "learning_rate": 9.14116753486366e-08,
+      "loss": 4.4856,
+      "step": 19155
+    },
+    {
+      "epoch": 3.758336602589251,
+      "grad_norm": 18.82733917236328,
+      "learning_rate": 9.067642730104132e-08,
+      "loss": 4.4474,
+      "step": 19160
+    },
+    {
+      "epoch": 3.7593173793644565,
+      "grad_norm": 20.41982078552246,
+      "learning_rate": 8.994412104354865e-08,
+      "loss": 4.322,
+      "step": 19165
+    },
+    {
+      "epoch": 3.7602981561396627,
+      "grad_norm": 15.03927993774414,
+      "learning_rate": 8.921475701497373e-08,
+      "loss": 4.3169,
+      "step": 19170
+    },
+    {
+      "epoch": 3.761278932914869,
+      "grad_norm": 14.643938064575195,
+      "learning_rate": 8.84883356523708e-08,
+      "loss": 4.4695,
+      "step": 19175
+    },
+    {
+      "epoch": 3.7622597096900745,
+      "grad_norm": 14.6721830368042,
+      "learning_rate": 8.776485739102947e-08,
+      "loss": 4.3335,
+      "step": 19180
+    },
+    {
+      "epoch": 3.7632404864652806,
+      "grad_norm": 17.07347869873047,
+      "learning_rate": 8.70443226644757e-08,
+      "loss": 4.7187,
+      "step": 19185
+    },
+    {
+      "epoch": 3.7642212632404863,
+      "grad_norm": 11.451022148132324,
+      "learning_rate": 8.632673190447305e-08,
+      "loss": 4.2598,
+      "step": 19190
+    },
+    {
+      "epoch": 3.7652020400156925,
+      "grad_norm": 24.953075408935547,
+      "learning_rate": 8.561208554101863e-08,
+      "loss": 4.3177,
+      "step": 19195
+    },
+    {
+      "epoch": 3.7661828167908986,
+      "grad_norm": 11.898823738098145,
+      "learning_rate": 8.490038400234767e-08,
+      "loss": 4.4786,
+      "step": 19200
+    },
+    {
+      "epoch": 3.7671635935661043,
+      "grad_norm": 29.196277618408203,
+      "learning_rate": 8.41916277149285e-08,
+      "loss": 4.3394,
+      "step": 19205
+    },
+    {
+      "epoch": 3.7681443703413104,
+      "grad_norm": 20.188243865966797,
+      "learning_rate": 8.348581710346692e-08,
+      "loss": 4.3111,
+      "step": 19210
+    },
+    {
+      "epoch": 3.769125147116516,
+      "grad_norm": 16.19799041748047,
+      "learning_rate": 8.27829525909013e-08,
+      "loss": 4.7131,
+      "step": 19215
+    },
+    {
+      "epoch": 3.7701059238917223,
+      "grad_norm": 16.04700469970703,
+      "learning_rate": 8.208303459840694e-08,
+      "loss": 4.2501,
+      "step": 19220
+    },
+    {
+      "epoch": 3.7710867006669284,
+      "grad_norm": 24.354026794433594,
+      "learning_rate": 8.138606354539114e-08,
+      "loss": 4.506,
+      "step": 19225
+    },
+    {
+      "epoch": 3.772067477442134,
+      "grad_norm": 13.918210983276367,
+      "learning_rate": 8.069203984949648e-08,
+      "loss": 4.5426,
+      "step": 19230
+    },
+    {
+      "epoch": 3.7730482542173402,
+      "grad_norm": 15.953398704528809,
+      "learning_rate": 8.000096392660029e-08,
+      "loss": 4.4152,
+      "step": 19235
+    },
+    {
+      "epoch": 3.774029030992546,
+      "grad_norm": 22.89542579650879,
+      "learning_rate": 7.931283619081187e-08,
+      "loss": 4.4485,
+      "step": 19240
+    },
+    {
+      "epoch": 3.775009807767752,
+      "grad_norm": 14.230191230773926,
+      "learning_rate": 7.862765705447528e-08,
+      "loss": 4.6742,
+      "step": 19245
+    },
+    {
+      "epoch": 3.775990584542958,
+      "grad_norm": 32.547611236572266,
+      "learning_rate": 7.794542692816654e-08,
+      "loss": 4.6502,
+      "step": 19250
+    },
+    {
+      "epoch": 3.776971361318164,
+      "grad_norm": 19.713741302490234,
+      "learning_rate": 7.726614622069528e-08,
+      "loss": 4.4383,
+      "step": 19255
+    },
+    {
+      "epoch": 3.77795213809337,
+      "grad_norm": 23.55909538269043,
+      "learning_rate": 7.658981533910315e-08,
+      "loss": 4.8089,
+      "step": 19260
+    },
+    {
+      "epoch": 3.7789329148685757,
+      "grad_norm": 18.59007453918457,
+      "learning_rate": 7.591643468866594e-08,
+      "loss": 4.4116,
+      "step": 19265
+    },
+    {
+      "epoch": 3.779913691643782,
+      "grad_norm": 28.624298095703125,
+      "learning_rate": 7.524600467288923e-08,
+      "loss": 4.6207,
+      "step": 19270
+    },
+    {
+      "epoch": 3.780894468418988,
+      "grad_norm": 35.744075775146484,
+      "learning_rate": 7.457852569351165e-08,
+      "loss": 4.4529,
+      "step": 19275
+    },
+    {
+      "epoch": 3.7818752451941937,
+      "grad_norm": 14.48076057434082,
+      "learning_rate": 7.391399815050326e-08,
+      "loss": 4.3982,
+      "step": 19280
+    },
+    {
+      "epoch": 3.7828560219694,
+      "grad_norm": 15.289562225341797,
+      "learning_rate": 7.32524224420661e-08,
+      "loss": 3.776,
+      "step": 19285
+    },
+    {
+      "epoch": 3.7838367987446055,
+      "grad_norm": 16.431413650512695,
+      "learning_rate": 7.259379896463248e-08,
+      "loss": 5.1205,
+      "step": 19290
+    },
+    {
+      "epoch": 3.7848175755198117,
+      "grad_norm": 22.791608810424805,
+      "learning_rate": 7.193812811286615e-08,
+      "loss": 4.7257,
+      "step": 19295
+    },
+    {
+      "epoch": 3.785798352295018,
+      "grad_norm": 18.99640464782715,
+      "learning_rate": 7.12854102796623e-08,
+      "loss": 4.5685,
+      "step": 19300
+    },
+    {
+      "epoch": 3.7867791290702235,
+      "grad_norm": 14.632157325744629,
+      "learning_rate": 7.063564585614524e-08,
+      "loss": 4.588,
+      "step": 19305
+    },
+    {
+      "epoch": 3.7877599058454297,
+      "grad_norm": 23.26076316833496,
+      "learning_rate": 6.998883523167021e-08,
+      "loss": 4.6019,
+      "step": 19310
+    },
+    {
+      "epoch": 3.7887406826206353,
+      "grad_norm": 15.142358779907227,
+      "learning_rate": 6.934497879382218e-08,
+      "loss": 4.5146,
+      "step": 19315
+    },
+    {
+      "epoch": 3.7897214593958415,
+      "grad_norm": 24.590253829956055,
+      "learning_rate": 6.870407692841696e-08,
+      "loss": 4.2335,
+      "step": 19320
+    },
+    {
+      "epoch": 3.7907022361710476,
+      "grad_norm": 16.87588882446289,
+      "learning_rate": 6.806613001949846e-08,
+      "loss": 4.5,
+      "step": 19325
+    },
+    {
+      "epoch": 3.7916830129462533,
+      "grad_norm": 15.2261323928833,
+      "learning_rate": 6.74311384493398e-08,
+      "loss": 4.4496,
+      "step": 19330
+    },
+    {
+      "epoch": 3.7926637897214595,
+      "grad_norm": 25.586700439453125,
+      "learning_rate": 6.679910259844491e-08,
+      "loss": 5.0063,
+      "step": 19335
+    },
+    {
+      "epoch": 3.793644566496665,
+      "grad_norm": 22.457977294921875,
+      "learning_rate": 6.617002284554585e-08,
+      "loss": 4.1973,
+      "step": 19340
+    },
+    {
+      "epoch": 3.7946253432718713,
+      "grad_norm": 22.65077018737793,
+      "learning_rate": 6.554389956760166e-08,
+      "loss": 4.6877,
+      "step": 19345
+    },
+    {
+      "epoch": 3.7956061200470774,
+      "grad_norm": 23.340280532836914,
+      "learning_rate": 6.492073313980274e-08,
+      "loss": 4.4651,
+      "step": 19350
+    },
+    {
+      "epoch": 3.796586896822283,
+      "grad_norm": 15.831352233886719,
+      "learning_rate": 6.430052393556485e-08,
+      "loss": 4.063,
+      "step": 19355
+    },
+    {
+      "epoch": 3.7975676735974893,
+      "grad_norm": 21.76120948791504,
+      "learning_rate": 6.368327232653349e-08,
+      "loss": 4.4149,
+      "step": 19360
+    },
+    {
+      "epoch": 3.798548450372695,
+      "grad_norm": 20.37674331665039,
+      "learning_rate": 6.306897868258167e-08,
+      "loss": 4.5668,
+      "step": 19365
+    },
+    {
+      "epoch": 3.799529227147901,
+      "grad_norm": 30.422216415405273,
+      "learning_rate": 6.245764337180827e-08,
+      "loss": 4.455,
+      "step": 19370
+    },
+    {
+      "epoch": 3.8005100039231072,
+      "grad_norm": 27.67301368713379,
+      "learning_rate": 6.184926676054192e-08,
+      "loss": 4.6488,
+      "step": 19375
+    },
+    {
+      "epoch": 3.801490780698313,
+      "grad_norm": 15.223337173461914,
+      "learning_rate": 6.124384921333714e-08,
+      "loss": 4.5511,
+      "step": 19380
+    },
+    {
+      "epoch": 3.802471557473519,
+      "grad_norm": 27.290067672729492,
+      "learning_rate": 6.064139109297485e-08,
+      "loss": 4.3251,
+      "step": 19385
+    },
+    {
+      "epoch": 3.8034523342487248,
+      "grad_norm": 12.322364807128906,
+      "learning_rate": 6.004189276046346e-08,
+      "loss": 4.451,
+      "step": 19390
+    },
+    {
+      "epoch": 3.804433111023931,
+      "grad_norm": 20.460012435913086,
+      "learning_rate": 5.944535457503731e-08,
+      "loss": 4.3789,
+      "step": 19395
+    },
+    {
+      "epoch": 3.805413887799137,
+      "grad_norm": 14.362495422363281,
+      "learning_rate": 5.885177689415711e-08,
+      "loss": 4.476,
+      "step": 19400
+    },
+    {
+      "epoch": 3.8063946645743427,
+      "grad_norm": 16.157991409301758,
+      "learning_rate": 5.826116007350946e-08,
+      "loss": 4.6647,
+      "step": 19405
+    },
+    {
+      "epoch": 3.807375441349549,
+      "grad_norm": 21.604549407958984,
+      "learning_rate": 5.7673504467006816e-08,
+      "loss": 4.5372,
+      "step": 19410
+    },
+    {
+      "epoch": 3.8083562181247546,
+      "grad_norm": 26.53864097595215,
+      "learning_rate": 5.708881042678749e-08,
+      "loss": 4.4941,
+      "step": 19415
+    },
+    {
+      "epoch": 3.8093369948999607,
+      "grad_norm": 36.335411071777344,
+      "learning_rate": 5.650707830321456e-08,
+      "loss": 4.4285,
+      "step": 19420
+    },
+    {
+      "epoch": 3.810317771675167,
+      "grad_norm": 17.700517654418945,
+      "learning_rate": 5.59283084448764e-08,
+      "loss": 4.6249,
+      "step": 19425
+    },
+    {
+      "epoch": 3.811298548450373,
+      "grad_norm": 16.858427047729492,
+      "learning_rate": 5.5352501198586705e-08,
+      "loss": 4.4317,
+      "step": 19430
+    },
+    {
+      "epoch": 3.8122793252255787,
+      "grad_norm": 11.746870994567871,
+      "learning_rate": 5.477965690938392e-08,
+      "loss": 4.3457,
+      "step": 19435
+    },
+    {
+      "epoch": 3.8132601020007844,
+      "grad_norm": 15.386157035827637,
+      "learning_rate": 5.420977592053067e-08,
+      "loss": 4.2697,
+      "step": 19440
+    },
+    {
+      "epoch": 3.8142408787759905,
+      "grad_norm": 19.786962509155273,
+      "learning_rate": 5.36428585735127e-08,
+      "loss": 4.7254,
+      "step": 19445
+    },
+    {
+      "epoch": 3.8152216555511966,
+      "grad_norm": 12.420428276062012,
+      "learning_rate": 5.307890520804271e-08,
+      "loss": 4.4631,
+      "step": 19450
+    },
+    {
+      "epoch": 3.8162024323264028,
+      "grad_norm": 19.548391342163086,
+      "learning_rate": 5.251791616205537e-08,
+      "loss": 4.55,
+      "step": 19455
+    },
+    {
+      "epoch": 3.8171832091016085,
+      "grad_norm": 12.915349960327148,
+      "learning_rate": 5.1959891771708456e-08,
+      "loss": 4.5048,
+      "step": 19460
+    },
+    {
+      "epoch": 3.8181639858768146,
+      "grad_norm": 24.07160758972168,
+      "learning_rate": 5.14048323713856e-08,
+      "loss": 4.723,
+      "step": 19465
+    },
+    {
+      "epoch": 3.8191447626520203,
+      "grad_norm": 15.310680389404297,
+      "learning_rate": 5.085273829369186e-08,
+      "loss": 3.9662,
+      "step": 19470
+    },
+    {
+      "epoch": 3.8201255394272264,
+      "grad_norm": 18.540067672729492,
+      "learning_rate": 5.0303609869455375e-08,
+      "loss": 4.6253,
+      "step": 19475
+    },
+    {
+      "epoch": 3.8211063162024326,
+      "grad_norm": 12.179683685302734,
+      "learning_rate": 4.975744742772848e-08,
+      "loss": 4.5506,
+      "step": 19480
+    },
+    {
+      "epoch": 3.8220870929776383,
+      "grad_norm": 20.053329467773438,
+      "learning_rate": 4.9214251295784385e-08,
+      "loss": 4.436,
+      "step": 19485
+    },
+    {
+      "epoch": 3.8230678697528444,
+      "grad_norm": 18.177942276000977,
+      "learning_rate": 4.8674021799121064e-08,
+      "loss": 4.3544,
+      "step": 19490
+    },
+    {
+      "epoch": 3.82404864652805,
+      "grad_norm": 18.189285278320312,
+      "learning_rate": 4.813675926145678e-08,
+      "loss": 4.4312,
+      "step": 19495
+    },
+    {
+      "epoch": 3.8250294233032562,
+      "grad_norm": 40.68314743041992,
+      "learning_rate": 4.760246400473345e-08,
+      "loss": 4.597,
+      "step": 19500
+    },
+    {
+      "epoch": 3.8260102000784624,
+      "grad_norm": 24.781036376953125,
+      "learning_rate": 4.707113634911387e-08,
+      "loss": 4.3725,
+      "step": 19505
+    },
+    {
+      "epoch": 3.826990976853668,
+      "grad_norm": 14.774175643920898,
+      "learning_rate": 4.654277661298223e-08,
+      "loss": 4.2771,
+      "step": 19510
+    },
+    {
+      "epoch": 3.827971753628874,
+      "grad_norm": 21.735511779785156,
+      "learning_rate": 4.60173851129464e-08,
+      "loss": 4.4238,
+      "step": 19515
+    },
+    {
+      "epoch": 3.82895253040408,
+      "grad_norm": 26.12150764465332,
+      "learning_rate": 4.549496216383287e-08,
+      "loss": 4.87,
+      "step": 19520
+    },
+    {
+      "epoch": 3.829933307179286,
+      "grad_norm": 21.527135848999023,
+      "learning_rate": 4.497550807869122e-08,
+      "loss": 4.2331,
+      "step": 19525
+    },
+    {
+      "epoch": 3.830914083954492,
+      "grad_norm": 17.46333122253418,
+      "learning_rate": 4.44590231687908e-08,
+      "loss": 4.7262,
+      "step": 19530
+    },
+    {
+      "epoch": 3.831894860729698,
+      "grad_norm": 13.620034217834473,
+      "learning_rate": 4.394550774362349e-08,
+      "loss": 4.7126,
+      "step": 19535
+    },
+    {
+      "epoch": 3.832875637504904,
+      "grad_norm": 13.067037582397461,
+      "learning_rate": 4.343496211089981e-08,
+      "loss": 4.4245,
+      "step": 19540
+    },
+    {
+      "epoch": 3.8338564142801097,
+      "grad_norm": 33.16718292236328,
+      "learning_rate": 4.292738657655171e-08,
+      "loss": 4.3331,
+      "step": 19545
+    },
+    {
+      "epoch": 3.834837191055316,
+      "grad_norm": 15.34744644165039,
+      "learning_rate": 4.242278144473144e-08,
+      "loss": 4.2482,
+      "step": 19550
+    },
+    {
+      "epoch": 3.835817967830522,
+      "grad_norm": 14.823765754699707,
+      "learning_rate": 4.192114701781047e-08,
+      "loss": 4.3858,
+      "step": 19555
+    },
+    {
+      "epoch": 3.8367987446057277,
+      "grad_norm": 18.384389877319336,
+      "learning_rate": 4.142248359638168e-08,
+      "loss": 4.483,
+      "step": 19560
+    },
+    {
+      "epoch": 3.837779521380934,
+      "grad_norm": 13.538766860961914,
+      "learning_rate": 4.092679147925604e-08,
+      "loss": 4.599,
+      "step": 19565
+    },
+    {
+      "epoch": 3.8387602981561395,
+      "grad_norm": 16.762392044067383,
+      "learning_rate": 4.043407096346486e-08,
+      "loss": 4.2274,
+      "step": 19570
+    },
+    {
+      "epoch": 3.8397410749313456,
+      "grad_norm": 31.2176570892334,
+      "learning_rate": 3.99443223442586e-08,
+      "loss": 4.4158,
+      "step": 19575
+    },
+    {
+      "epoch": 3.840721851706552,
+      "grad_norm": 23.36688804626465,
+      "learning_rate": 3.945754591510698e-08,
+      "loss": 4.5466,
+      "step": 19580
+    },
+    {
+      "epoch": 3.8417026284817575,
+      "grad_norm": 22.09506607055664,
+      "learning_rate": 3.8973741967698874e-08,
+      "loss": 4.5284,
+      "step": 19585
+    },
+    {
+      "epoch": 3.8426834052569636,
+      "grad_norm": 31.000293731689453,
+      "learning_rate": 3.849291079194184e-08,
+      "loss": 4.895,
+      "step": 19590
+    },
+    {
+      "epoch": 3.8436641820321693,
+      "grad_norm": 20.989912033081055,
+      "learning_rate": 3.8015052675961505e-08,
+      "loss": 4.5427,
+      "step": 19595
+    },
+    {
+      "epoch": 3.8446449588073754,
+      "grad_norm": 16.181804656982422,
+      "learning_rate": 3.754016790610271e-08,
+      "loss": 4.3438,
+      "step": 19600
+    },
+    {
+      "epoch": 3.8456257355825816,
+      "grad_norm": 18.582027435302734,
+      "learning_rate": 3.706825676692838e-08,
+      "loss": 4.3791,
+      "step": 19605
+    },
+    {
+      "epoch": 3.8466065123577873,
+      "grad_norm": 42.60034942626953,
+      "learning_rate": 3.659931954121954e-08,
+      "loss": 4.6263,
+      "step": 19610
+    },
+    {
+      "epoch": 3.8475872891329934,
+      "grad_norm": 15.503507614135742,
+      "learning_rate": 3.613335650997585e-08,
+      "loss": 4.6023,
+      "step": 19615
+    },
+    {
+      "epoch": 3.848568065908199,
+      "grad_norm": 16.451189041137695,
+      "learning_rate": 3.56703679524123e-08,
+      "loss": 4.2281,
+      "step": 19620
+    },
+    {
+      "epoch": 3.8495488426834052,
+      "grad_norm": 18.993877410888672,
+      "learning_rate": 3.52103541459653e-08,
+      "loss": 4.5004,
+      "step": 19625
+    },
+    {
+      "epoch": 3.8505296194586114,
+      "grad_norm": 14.785061836242676,
+      "learning_rate": 3.4753315366284904e-08,
+      "loss": 4.1803,
+      "step": 19630
+    },
+    {
+      "epoch": 3.851510396233817,
+      "grad_norm": 19.010860443115234,
+      "learning_rate": 3.429925188724148e-08,
+      "loss": 4.3486,
+      "step": 19635
+    },
+    {
+      "epoch": 3.852491173009023,
+      "grad_norm": 20.439987182617188,
+      "learning_rate": 3.384816398092128e-08,
+      "loss": 4.5895,
+      "step": 19640
+    },
+    {
+      "epoch": 3.853471949784229,
+      "grad_norm": 20.579715728759766,
+      "learning_rate": 3.3400051917626964e-08,
+      "loss": 4.752,
+      "step": 19645
+    },
+    {
+      "epoch": 3.854452726559435,
+      "grad_norm": 35.6832389831543,
+      "learning_rate": 3.295491596587874e-08,
+      "loss": 4.3648,
+      "step": 19650
+    },
+    {
+      "epoch": 3.855433503334641,
+      "grad_norm": 24.63140296936035,
+      "learning_rate": 3.251275639241269e-08,
+      "loss": 4.9887,
+      "step": 19655
+    },
+    {
+      "epoch": 3.856414280109847,
+      "grad_norm": 20.618412017822266,
+      "learning_rate": 3.2073573462182984e-08,
+      "loss": 4.5265,
+      "step": 19660
+    },
+    {
+      "epoch": 3.857395056885053,
+      "grad_norm": 25.951919555664062,
+      "learning_rate": 3.1637367438358544e-08,
+      "loss": 4.9724,
+      "step": 19665
+    },
+    {
+      "epoch": 3.8583758336602587,
+      "grad_norm": 23.954572677612305,
+      "learning_rate": 3.120413858232474e-08,
+      "loss": 4.7667,
+      "step": 19670
+    },
+    {
+      "epoch": 3.859356610435465,
+      "grad_norm": 13.720315933227539,
+      "learning_rate": 3.07738871536839e-08,
+      "loss": 4.3292,
+      "step": 19675
+    },
+    {
+      "epoch": 3.860337387210671,
+      "grad_norm": 15.496363639831543,
+      "learning_rate": 3.034661341025258e-08,
+      "loss": 4.6116,
+      "step": 19680
+    },
+    {
+      "epoch": 3.8613181639858767,
+      "grad_norm": 60.01427459716797,
+      "learning_rate": 2.9922317608064856e-08,
+      "loss": 4.4719,
+      "step": 19685
+    },
+    {
+      "epoch": 3.862298940761083,
+      "grad_norm": 15.243767738342285,
+      "learning_rate": 2.9501000001369018e-08,
+      "loss": 4.3308,
+      "step": 19690
+    },
+    {
+      "epoch": 3.8632797175362885,
+      "grad_norm": 29.687528610229492,
+      "learning_rate": 2.9082660842628674e-08,
+      "loss": 4.7989,
+      "step": 19695
+    },
+    {
+      "epoch": 3.8642604943114947,
+      "grad_norm": 35.74703598022461,
+      "learning_rate": 2.8667300382523855e-08,
+      "loss": 4.586,
+      "step": 19700
+    },
+    {
+      "epoch": 3.865241271086701,
+      "grad_norm": 22.265159606933594,
+      "learning_rate": 2.82549188699488e-08,
+      "loss": 4.8159,
+      "step": 19705
+    },
+    {
+      "epoch": 3.8662220478619065,
+      "grad_norm": 16.983205795288086,
+      "learning_rate": 2.7845516552013064e-08,
+      "loss": 4.362,
+      "step": 19710
+    },
+    {
+      "epoch": 3.8672028246371126,
+      "grad_norm": 15.669620513916016,
+      "learning_rate": 2.7439093674040406e-08,
+      "loss": 4.1492,
+      "step": 19715
+    },
+    {
+      "epoch": 3.8681836014123183,
+      "grad_norm": 11.486489295959473,
+      "learning_rate": 2.7035650479570463e-08,
+      "loss": 4.6597,
+      "step": 19720
+    },
+    {
+      "epoch": 3.8691643781875245,
+      "grad_norm": 19.21707534790039,
+      "learning_rate": 2.6635187210355408e-08,
+      "loss": 4.4448,
+      "step": 19725
+    },
+    {
+      "epoch": 3.8701451549627306,
+      "grad_norm": 28.177640914916992,
+      "learning_rate": 2.6237704106363282e-08,
+      "loss": 4.4331,
+      "step": 19730
+    },
+    {
+      "epoch": 3.8711259317379363,
+      "grad_norm": 19.105148315429688,
+      "learning_rate": 2.584320140577634e-08,
+      "loss": 4.6519,
+      "step": 19735
+    },
+    {
+      "epoch": 3.8721067085131424,
+      "grad_norm": 17.474411010742188,
+      "learning_rate": 2.5451679344989934e-08,
+      "loss": 4.5222,
+      "step": 19740
+    },
+    {
+      "epoch": 3.873087485288348,
+      "grad_norm": 14.374054908752441,
+      "learning_rate": 2.506313815861472e-08,
+      "loss": 4.4389,
+      "step": 19745
+    },
+    {
+      "epoch": 3.8740682620635543,
+      "grad_norm": 29.551166534423828,
+      "learning_rate": 2.467757807947335e-08,
+      "loss": 4.3835,
+      "step": 19750
+    },
+    {
+      "epoch": 3.8750490388387604,
+      "grad_norm": 14.76496696472168,
+      "learning_rate": 2.4294999338604352e-08,
+      "loss": 4.2703,
+      "step": 19755
+    },
+    {
+      "epoch": 3.8760298156139665,
+      "grad_norm": 27.193767547607422,
+      "learning_rate": 2.391540216525712e-08,
+      "loss": 4.7443,
+      "step": 19760
+    },
+    {
+      "epoch": 3.8770105923891722,
+      "grad_norm": 33.37945556640625,
+      "learning_rate": 2.3538786786896918e-08,
+      "loss": 4.8142,
+      "step": 19765
+    },
+    {
+      "epoch": 3.877991369164378,
+      "grad_norm": 36.48703384399414,
+      "learning_rate": 2.316515342920045e-08,
+      "loss": 4.3292,
+      "step": 19770
+    },
+    {
+      "epoch": 3.878972145939584,
+      "grad_norm": 13.171440124511719,
+      "learning_rate": 2.279450231605862e-08,
+      "loss": 4.277,
+      "step": 19775
+    },
+    {
+      "epoch": 3.87995292271479,
+      "grad_norm": 22.090845108032227,
+      "learning_rate": 2.2426833669574875e-08,
+      "loss": 4.2819,
+      "step": 19780
+    },
+    {
+      "epoch": 3.8809336994899963,
+      "grad_norm": 23.142976760864258,
+      "learning_rate": 2.2062147710065208e-08,
+      "loss": 4.5399,
+      "step": 19785
+    },
+    {
+      "epoch": 3.881914476265202,
+      "grad_norm": 21.87250328063965,
+      "learning_rate": 2.170044465605925e-08,
+      "loss": 4.4368,
+      "step": 19790
+    },
+    {
+      "epoch": 3.882895253040408,
+      "grad_norm": 22.828344345092773,
+      "learning_rate": 2.1341724724298073e-08,
+      "loss": 4.2221,
+      "step": 19795
+    },
+    {
+      "epoch": 3.883876029815614,
+      "grad_norm": 18.955265045166016,
+      "learning_rate": 2.0985988129735847e-08,
+      "loss": 4.3943,
+      "step": 19800
+    },
+    {
+      "epoch": 3.88485680659082,
+      "grad_norm": 14.534862518310547,
+      "learning_rate": 2.063323508553816e-08,
+      "loss": 4.3378,
+      "step": 19805
+    },
+    {
+      "epoch": 3.885837583366026,
+      "grad_norm": 22.451967239379883,
+      "learning_rate": 2.028346580308427e-08,
+      "loss": 4.3979,
+      "step": 19810
+    },
+    {
+      "epoch": 3.886818360141232,
+      "grad_norm": 22.964393615722656,
+      "learning_rate": 1.99366804919654e-08,
+      "loss": 4.3687,
+      "step": 19815
+    },
+    {
+      "epoch": 3.887799136916438,
+      "grad_norm": 15.007269859313965,
+      "learning_rate": 1.9592879359981998e-08,
+      "loss": 4.5002,
+      "step": 19820
+    },
+    {
+      "epoch": 3.8887799136916437,
+      "grad_norm": 22.878572463989258,
+      "learning_rate": 1.925206261315038e-08,
+      "loss": 4.9794,
+      "step": 19825
+    },
+    {
+      "epoch": 3.88976069046685,
+      "grad_norm": 32.9472770690918,
+      "learning_rate": 1.8914230455695514e-08,
+      "loss": 4.5072,
+      "step": 19830
+    },
+    {
+      "epoch": 3.890741467242056,
+      "grad_norm": 41.51423645019531,
+      "learning_rate": 1.8579383090054915e-08,
+      "loss": 4.4305,
+      "step": 19835
+    },
+    {
+      "epoch": 3.8917222440172616,
+      "grad_norm": 14.0291748046875,
+      "learning_rate": 1.8247520716878075e-08,
+      "loss": 4.2225,
+      "step": 19840
+    },
+    {
+      "epoch": 3.8927030207924678,
+      "grad_norm": 12.759921073913574,
+      "learning_rate": 1.7918643535024816e-08,
+      "loss": 4.5276,
+      "step": 19845
+    },
+    {
+      "epoch": 3.8936837975676735,
+      "grad_norm": 21.33737564086914,
+      "learning_rate": 1.7592751741566384e-08,
+      "loss": 4.448,
+      "step": 19850
+    },
+    {
+      "epoch": 3.8946645743428796,
+      "grad_norm": 26.02752685546875,
+      "learning_rate": 1.726984553178601e-08,
+      "loss": 4.4764,
+      "step": 19855
+    },
+    {
+      "epoch": 3.8956453511180857,
+      "grad_norm": 15.944296836853027,
+      "learning_rate": 1.69499250991767e-08,
+      "loss": 4.5847,
+      "step": 19860
+    },
+    {
+      "epoch": 3.8966261278932914,
+      "grad_norm": 21.915067672729492,
+      "learning_rate": 1.663299063544288e-08,
+      "loss": 4.9598,
+      "step": 19865
+    },
+    {
+      "epoch": 3.8976069046684976,
+      "grad_norm": 26.551939010620117,
+      "learning_rate": 1.6319042330500413e-08,
+      "loss": 4.7225,
+      "step": 19870
+    },
+    {
+      "epoch": 3.8985876814437033,
+      "grad_norm": 24.626413345336914,
+      "learning_rate": 1.600808037247381e-08,
+      "loss": 4.1449,
+      "step": 19875
+    },
+    {
+      "epoch": 3.8995684582189094,
+      "grad_norm": 9.839727401733398,
+      "learning_rate": 1.570010494769958e-08,
+      "loss": 4.6128,
+      "step": 19880
+    },
+    {
+      "epoch": 3.9005492349941155,
+      "grad_norm": 34.44369888305664,
+      "learning_rate": 1.5395116240725093e-08,
+      "loss": 4.5166,
+      "step": 19885
+    },
+    {
+      "epoch": 3.9015300117693212,
+      "grad_norm": 19.511125564575195,
+      "learning_rate": 1.5093114434306388e-08,
+      "loss": 4.49,
+      "step": 19890
+    },
+    {
+      "epoch": 3.9025107885445274,
+      "grad_norm": 12.833151817321777,
+      "learning_rate": 1.4794099709410925e-08,
+      "loss": 4.4542,
+      "step": 19895
+    },
+    {
+      "epoch": 3.903491565319733,
+      "grad_norm": 12.988509178161621,
+      "learning_rate": 1.4498072245216488e-08,
+      "loss": 4.5994,
+      "step": 19900
+    },
+    {
+      "epoch": 3.904472342094939,
+      "grad_norm": 24.291364669799805,
+      "learning_rate": 1.420503221910896e-08,
+      "loss": 4.2495,
+      "step": 19905
+    },
+    {
+      "epoch": 3.9054531188701453,
+      "grad_norm": 31.334964752197266,
+      "learning_rate": 1.3914979806685659e-08,
+      "loss": 4.5936,
+      "step": 19910
+    },
+    {
+      "epoch": 3.906433895645351,
+      "grad_norm": 38.430110931396484,
+      "learning_rate": 1.3627915181753659e-08,
+      "loss": 4.568,
+      "step": 19915
+    },
+    {
+      "epoch": 3.907414672420557,
+      "grad_norm": 23.217266082763672,
+      "learning_rate": 1.3343838516329255e-08,
+      "loss": 4.4528,
+      "step": 19920
+    },
+    {
+      "epoch": 3.908395449195763,
+      "grad_norm": 18.199256896972656,
+      "learning_rate": 1.3062749980637946e-08,
+      "loss": 4.5258,
+      "step": 19925
+    },
+    {
+      "epoch": 3.909376225970969,
+      "grad_norm": 16.116588592529297,
+      "learning_rate": 1.2784649743115551e-08,
+      "loss": 4.4565,
+      "step": 19930
+    },
+    {
+      "epoch": 3.910357002746175,
+      "grad_norm": 26.552534103393555,
+      "learning_rate": 1.2509537970406549e-08,
+      "loss": 4.6301,
+      "step": 19935
+    },
+    {
+      "epoch": 3.911337779521381,
+      "grad_norm": 15.603949546813965,
+      "learning_rate": 1.2237414827364624e-08,
+      "loss": 4.4856,
+      "step": 19940
+    },
+    {
+      "epoch": 3.912318556296587,
+      "grad_norm": 10.514763832092285,
+      "learning_rate": 1.1968280477052673e-08,
+      "loss": 4.4509,
+      "step": 19945
+    },
+    {
+      "epoch": 3.9132993330717927,
+      "grad_norm": 17.823230743408203,
+      "learning_rate": 1.170213508074336e-08,
+      "loss": 4.0671,
+      "step": 19950
+    },
+    {
+      "epoch": 3.914280109846999,
+      "grad_norm": 23.3159122467041,
+      "learning_rate": 1.1438978797916888e-08,
+      "loss": 4.7467,
+      "step": 19955
+    },
+    {
+      "epoch": 3.915260886622205,
+      "grad_norm": 33.341434478759766,
+      "learning_rate": 1.1178811786263787e-08,
+      "loss": 4.4565,
+      "step": 19960
+    },
+    {
+      "epoch": 3.9162416633974106,
+      "grad_norm": 20.232297897338867,
+      "learning_rate": 1.0921634201682685e-08,
+      "loss": 4.4115,
+      "step": 19965
+    },
+    {
+      "epoch": 3.917222440172617,
+      "grad_norm": 24.05611801147461,
+      "learning_rate": 1.0667446198280307e-08,
+      "loss": 4.2384,
+      "step": 19970
+    },
+    {
+      "epoch": 3.9182032169478225,
+      "grad_norm": 9.7948637008667,
+      "learning_rate": 1.0416247928373147e-08,
+      "loss": 4.2928,
+      "step": 19975
+    },
+    {
+      "epoch": 3.9191839937230286,
+      "grad_norm": 18.230249404907227,
+      "learning_rate": 1.0168039542485242e-08,
+      "loss": 4.7602,
+      "step": 19980
+    },
+    {
+      "epoch": 3.9201647704982348,
+      "grad_norm": 22.710430145263672,
+      "learning_rate": 9.922821189348731e-09,
+      "loss": 4.8469,
+      "step": 19985
+    },
+    {
+      "epoch": 3.9211455472734404,
+      "grad_norm": 19.150442123413086,
+      "learning_rate": 9.680593015905515e-09,
+      "loss": 4.3717,
+      "step": 19990
+    },
+    {
+      "epoch": 3.9221263240486466,
+      "grad_norm": 14.786493301391602,
+      "learning_rate": 9.44135516730449e-09,
+      "loss": 4.3758,
+      "step": 19995
+    },
+    {
+      "epoch": 3.9231071008238523,
+      "grad_norm": 16.019426345825195,
+      "learning_rate": 9.205107786902646e-09,
+      "loss": 4.5163,
+      "step": 20000
+    },
+    {
+      "epoch": 3.9240878775990584,
+      "grad_norm": 22.1610107421875,
+      "learning_rate": 8.971851016265631e-09,
+      "loss": 4.4827,
+      "step": 20005
+    },
+    {
+      "epoch": 3.9250686543742646,
+      "grad_norm": 22.63068199157715,
+      "learning_rate": 8.741584995167195e-09,
+      "loss": 4.2595,
+      "step": 20010
+    },
+    {
+      "epoch": 3.9260494311494702,
+      "grad_norm": 11.803055763244629,
+      "learning_rate": 8.514309861588077e-09,
+      "loss": 4.436,
+      "step": 20015
+    },
+    {
+      "epoch": 3.9270302079246764,
+      "grad_norm": 30.385862350463867,
+      "learning_rate": 8.290025751716558e-09,
+      "loss": 4.6062,
+      "step": 20020
+    },
+    {
+      "epoch": 3.928010984699882,
+      "grad_norm": 18.616378784179688,
+      "learning_rate": 8.068732799950685e-09,
+      "loss": 4.5166,
+      "step": 20025
+    },
+    {
+      "epoch": 3.928991761475088,
+      "grad_norm": 11.325064659118652,
+      "learning_rate": 7.850431138893833e-09,
+      "loss": 4.3665,
+      "step": 20030
+    },
+    {
+      "epoch": 3.9299725382502944,
+      "grad_norm": 20.41812515258789,
+      "learning_rate": 7.635120899358029e-09,
+      "loss": 4.3084,
+      "step": 20035
+    },
+    {
+      "epoch": 3.9309533150255,
+      "grad_norm": 25.305400848388672,
+      "learning_rate": 7.422802210362845e-09,
+      "loss": 4.364,
+      "step": 20040
+    },
+    {
+      "epoch": 3.931934091800706,
+      "grad_norm": 15.921910285949707,
+      "learning_rate": 7.213475199134845e-09,
+      "loss": 4.3632,
+      "step": 20045
+    },
+    {
+      "epoch": 3.932914868575912,
+      "grad_norm": 17.5665283203125,
+      "learning_rate": 7.007139991108136e-09,
+      "loss": 4.1237,
+      "step": 20050
+    },
+    {
+      "epoch": 3.933895645351118,
+      "grad_norm": 15.96419620513916,
+      "learning_rate": 6.8037967099232604e-09,
+      "loss": 4.4997,
+      "step": 20055
+    },
+    {
+      "epoch": 3.934876422126324,
+      "grad_norm": 38.77703094482422,
+      "learning_rate": 6.60344547742997e-09,
+      "loss": 4.9041,
+      "step": 20060
+    },
+    {
+      "epoch": 3.93585719890153,
+      "grad_norm": 16.583595275878906,
+      "learning_rate": 6.406086413682233e-09,
+      "loss": 4.3876,
+      "step": 20065
+    },
+    {
+      "epoch": 3.936837975676736,
+      "grad_norm": 24.678234100341797,
+      "learning_rate": 6.211719636943781e-09,
+      "loss": 4.18,
+      "step": 20070
+    },
+    {
+      "epoch": 3.9378187524519417,
+      "grad_norm": 12.968330383300781,
+      "learning_rate": 6.020345263683114e-09,
+      "loss": 4.8268,
+      "step": 20075
+    },
+    {
+      "epoch": 3.938799529227148,
+      "grad_norm": 17.95810317993164,
+      "learning_rate": 5.83196340857739e-09,
+      "loss": 4.5699,
+      "step": 20080
+    },
+    {
+      "epoch": 3.939780306002354,
+      "grad_norm": 22.08363914489746,
+      "learning_rate": 5.646574184509646e-09,
+      "loss": 4.9704,
+      "step": 20085
+    },
+    {
+      "epoch": 3.94076108277756,
+      "grad_norm": 13.76091480255127,
+      "learning_rate": 5.464177702568796e-09,
+      "loss": 4.4924,
+      "step": 20090
+    },
+    {
+      "epoch": 3.941741859552766,
+      "grad_norm": 19.462295532226562,
+      "learning_rate": 5.2847740720529674e-09,
+      "loss": 4.3258,
+      "step": 20095
+    },
+    {
+      "epoch": 3.9427226363279715,
+      "grad_norm": 11.40530014038086,
+      "learning_rate": 5.108363400463945e-09,
+      "loss": 4.581,
+      "step": 20100
+    },
+    {
+      "epoch": 3.9437034131031776,
+      "grad_norm": 16.985126495361328,
+      "learning_rate": 4.9349457935121695e-09,
+      "loss": 4.7181,
+      "step": 20105
+    },
+    {
+      "epoch": 3.9446841898783838,
+      "grad_norm": 36.759437561035156,
+      "learning_rate": 4.764521355113405e-09,
+      "loss": 4.2881,
+      "step": 20110
+    },
+    {
+      "epoch": 3.94566496665359,
+      "grad_norm": 22.106002807617188,
+      "learning_rate": 4.597090187390407e-09,
+      "loss": 4.3794,
+      "step": 20115
+    },
+    {
+      "epoch": 3.9466457434287956,
+      "grad_norm": 32.104759216308594,
+      "learning_rate": 4.43265239067292e-09,
+      "loss": 4.6535,
+      "step": 20120
+    },
+    {
+      "epoch": 3.9476265202040017,
+      "grad_norm": 21.282289505004883,
+      "learning_rate": 4.2712080634949024e-09,
+      "loss": 4.3232,
+      "step": 20125
+    },
+    {
+      "epoch": 3.9486072969792074,
+      "grad_norm": 25.346208572387695,
+      "learning_rate": 4.112757302598414e-09,
+      "loss": 4.4755,
+      "step": 20130
+    },
+    {
+      "epoch": 3.9495880737544136,
+      "grad_norm": 22.805927276611328,
+      "learning_rate": 3.957300202931391e-09,
+      "loss": 4.6681,
+      "step": 20135
+    },
+    {
+      "epoch": 3.9505688505296197,
+      "grad_norm": 27.887910842895508,
+      "learning_rate": 3.804836857647654e-09,
+      "loss": 4.5999,
+      "step": 20140
+    },
+    {
+      "epoch": 3.9515496273048254,
+      "grad_norm": 22.663606643676758,
+      "learning_rate": 3.655367358106343e-09,
+      "loss": 4.4651,
+      "step": 20145
+    },
+    {
+      "epoch": 3.9525304040800315,
+      "grad_norm": 11.637248992919922,
+      "learning_rate": 3.5088917938741473e-09,
+      "loss": 4.259,
+      "step": 20150
+    },
+    {
+      "epoch": 3.9535111808552372,
+      "grad_norm": 25.06681251525879,
+      "learning_rate": 3.365410252723078e-09,
+      "loss": 4.4804,
+      "step": 20155
+    },
+    {
+      "epoch": 3.9544919576304434,
+      "grad_norm": 34.771358489990234,
+      "learning_rate": 3.2249228206299165e-09,
+      "loss": 4.5491,
+      "step": 20160
+    },
+    {
+      "epoch": 3.9554727344056495,
+      "grad_norm": 14.006319999694824,
+      "learning_rate": 3.087429581778434e-09,
+      "loss": 4.258,
+      "step": 20165
+    },
+    {
+      "epoch": 3.956453511180855,
+      "grad_norm": 22.197072982788086,
+      "learning_rate": 2.952930618558836e-09,
+      "loss": 4.6398,
+      "step": 20170
+    },
+    {
+      "epoch": 3.9574342879560613,
+      "grad_norm": 11.157323837280273,
+      "learning_rate": 2.8214260115655424e-09,
+      "loss": 4.6351,
+      "step": 20175
+    },
+    {
+      "epoch": 3.958415064731267,
+      "grad_norm": 14.1424560546875,
+      "learning_rate": 2.6929158395999634e-09,
+      "loss": 4.5076,
+      "step": 20180
+    },
+    {
+      "epoch": 3.959395841506473,
+      "grad_norm": 24.90903663635254,
+      "learning_rate": 2.567400179667723e-09,
+      "loss": 4.3885,
+      "step": 20185
+    },
+    {
+      "epoch": 3.9603766182816793,
+      "grad_norm": 16.46721076965332,
+      "learning_rate": 2.444879106982545e-09,
+      "loss": 4.8088,
+      "step": 20190
+    },
+    {
+      "epoch": 3.961357395056885,
+      "grad_norm": 32.02701187133789,
+      "learning_rate": 2.325352694960148e-09,
+      "loss": 4.7701,
+      "step": 20195
+    },
+    {
+      "epoch": 3.962338171832091,
+      "grad_norm": 22.428773880004883,
+      "learning_rate": 2.2088210152254596e-09,
+      "loss": 4.4474,
+      "step": 20200
+    },
+    {
+      "epoch": 3.963318948607297,
+      "grad_norm": 27.974700927734375,
+      "learning_rate": 2.0952841376065124e-09,
+      "loss": 4.6453,
+      "step": 20205
+    },
+    {
+      "epoch": 3.964299725382503,
+      "grad_norm": 29.08708381652832,
+      "learning_rate": 1.984742130137218e-09,
+      "loss": 4.4979,
+      "step": 20210
+    },
+    {
+      "epoch": 3.965280502157709,
+      "grad_norm": 12.308250427246094,
+      "learning_rate": 1.8771950590573686e-09,
+      "loss": 4.4553,
+      "step": 20215
+    },
+    {
+      "epoch": 3.966261278932915,
+      "grad_norm": 18.78961944580078,
+      "learning_rate": 1.7726429888120788e-09,
+      "loss": 4.308,
+      "step": 20220
+    },
+    {
+      "epoch": 3.967242055708121,
+      "grad_norm": 11.963844299316406,
+      "learning_rate": 1.6710859820512348e-09,
+      "loss": 4.4348,
+      "step": 20225
+    },
+    {
+      "epoch": 3.9682228324833266,
+      "grad_norm": 19.877851486206055,
+      "learning_rate": 1.5725240996306013e-09,
+      "loss": 4.1762,
+      "step": 20230
+    },
+    {
+      "epoch": 3.9692036092585328,
+      "grad_norm": 30.30404281616211,
+      "learning_rate": 1.4769574006107124e-09,
+      "loss": 4.9034,
+      "step": 20235
+    },
+    {
+      "epoch": 3.970184386033739,
+      "grad_norm": 22.604032516479492,
+      "learning_rate": 1.3843859422574269e-09,
+      "loss": 4.3639,
+      "step": 20240
+    },
+    {
+      "epoch": 3.9711651628089446,
+      "grad_norm": 22.842365264892578,
+      "learning_rate": 1.294809780042483e-09,
+      "loss": 4.1495,
+      "step": 20245
+    },
+    {
+      "epoch": 3.9721459395841507,
+      "grad_norm": 25.694583892822266,
+      "learning_rate": 1.2082289676412784e-09,
+      "loss": 4.9004,
+      "step": 20250
+    },
+    {
+      "epoch": 3.9731267163593564,
+      "grad_norm": 30.319358825683594,
+      "learning_rate": 1.1246435569362002e-09,
+      "loss": 4.5087,
+      "step": 20255
+    },
+    {
+      "epoch": 3.9741074931345626,
+      "grad_norm": 23.130138397216797,
+      "learning_rate": 1.0440535980132948e-09,
+      "loss": 4.0759,
+      "step": 20260
+    },
+    {
+      "epoch": 3.9750882699097687,
+      "grad_norm": 26.70267677307129,
+      "learning_rate": 9.664591391639333e-10,
+      "loss": 4.2694,
+      "step": 20265
+    },
+    {
+      "epoch": 3.9760690466849744,
+      "grad_norm": 21.76305389404297,
+      "learning_rate": 8.918602268848109e-10,
+      "loss": 4.3865,
+      "step": 20270
+    },
+    {
+      "epoch": 3.9770498234601805,
+      "grad_norm": 12.993515968322754,
+      "learning_rate": 8.202569058773924e-10,
+      "loss": 4.225,
+      "step": 20275
+    },
+    {
+      "epoch": 3.9780306002353862,
+      "grad_norm": 17.48309326171875,
+      "learning_rate": 7.516492190479118e-10,
+      "loss": 4.3941,
+      "step": 20280
+    },
+    {
+      "epoch": 3.9790113770105924,
+      "grad_norm": 15.2409029006958,
+      "learning_rate": 6.860372075084831e-10,
+      "loss": 4.4058,
+      "step": 20285
+    },
+    {
+      "epoch": 3.9799921537857985,
+      "grad_norm": 12.336922645568848,
+      "learning_rate": 6.234209105754341e-10,
+      "loss": 4.6124,
+      "step": 20290
+    },
+    {
+      "epoch": 3.980972930561004,
+      "grad_norm": 19.70476722717285,
+      "learning_rate": 5.63800365769307e-10,
+      "loss": 4.2905,
+      "step": 20295
+    },
+    {
+      "epoch": 3.9819537073362103,
+      "grad_norm": 18.989845275878906,
+      "learning_rate": 5.071756088165236e-10,
+      "loss": 4.5843,
+      "step": 20300
+    },
+    {
+      "epoch": 3.982934484111416,
+      "grad_norm": 20.936487197875977,
+      "learning_rate": 4.535466736488303e-10,
+      "loss": 4.546,
+      "step": 20305
+    },
+    {
+      "epoch": 3.983915260886622,
+      "grad_norm": 11.952984809875488,
+      "learning_rate": 4.029135924005223e-10,
+      "loss": 4.3485,
+      "step": 20310
+    },
+    {
+      "epoch": 3.9848960376618283,
+      "grad_norm": 12.227151870727539,
+      "learning_rate": 3.5527639541399486e-10,
+      "loss": 4.4553,
+      "step": 20315
+    },
+    {
+      "epoch": 3.985876814437034,
+      "grad_norm": 38.21388244628906,
+      "learning_rate": 3.1063511123308187e-10,
+      "loss": 4.319,
+      "step": 20320
+    },
+    {
+      "epoch": 3.98685759121224,
+      "grad_norm": 25.145599365234375,
+      "learning_rate": 2.689897666091623e-10,
+      "loss": 4.609,
+      "step": 20325
+    },
+    {
+      "epoch": 3.987838367987446,
+      "grad_norm": 48.00320053100586,
+      "learning_rate": 2.3034038649616398e-10,
+      "loss": 4.4196,
+      "step": 20330
+    },
+    {
+      "epoch": 3.988819144762652,
+      "grad_norm": 34.582763671875,
+      "learning_rate": 1.9468699405444936e-10,
+      "loss": 4.4329,
+      "step": 20335
+    },
+    {
+      "epoch": 3.989799921537858,
+      "grad_norm": 19.579330444335938,
+      "learning_rate": 1.6202961064804013e-10,
+      "loss": 4.5891,
+      "step": 20340
+    },
+    {
+      "epoch": 3.990780698313064,
+      "grad_norm": 31.714948654174805,
+      "learning_rate": 1.3236825584628242e-10,
+      "loss": 4.4694,
+      "step": 20345
+    },
+    {
+      "epoch": 3.99176147508827,
+      "grad_norm": 24.47707176208496,
+      "learning_rate": 1.0570294742329179e-10,
+      "loss": 4.6973,
+      "step": 20350
+    },
+    {
+      "epoch": 3.9927422518634756,
+      "grad_norm": 21.718971252441406,
+      "learning_rate": 8.203370135684286e-11,
+      "loss": 4.7068,
+      "step": 20355
+    },
+    {
+      "epoch": 3.993723028638682,
+      "grad_norm": 22.66058921813965,
+      "learning_rate": 6.136053183058987e-11,
+      "loss": 4.4124,
+      "step": 20360
+    },
+    {
+      "epoch": 3.994703805413888,
+      "grad_norm": 28.243078231811523,
+      "learning_rate": 4.368345123295648e-11,
+      "loss": 4.3122,
+      "step": 20365
+    },
+    {
+      "epoch": 3.9956845821890936,
+      "grad_norm": 12.159256935119629,
+      "learning_rate": 2.900247015547031e-11,
+      "loss": 4.3095,
+      "step": 20370
+    },
+    {
+      "epoch": 3.9966653589642998,
+      "grad_norm": 17.810564041137695,
+      "learning_rate": 1.731759739553862e-11,
+      "loss": 4.7029,
+      "step": 20375
+    },
+    {
+      "epoch": 3.9976461357395054,
+      "grad_norm": 13.572352409362793,
+      "learning_rate": 8.628839955893143e-12,
+      "loss": 4.5484,
+      "step": 20380
+    },
+    {
+      "epoch": 3.9986269125147116,
+      "grad_norm": 18.063383102416992,
+      "learning_rate": 2.936203042369634e-12,
+      "loss": 4.3011,
+      "step": 20385
+    },
+    {
+      "epoch": 3.9996076892899177,
+      "grad_norm": 15.351741790771484,
+      "learning_rate": 2.3969006557322816e-13,
+      "loss": 4.4443,
+      "step": 20390
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 20392,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 102,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.058850979271475e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}