diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4195 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9993227226549273,
+  "eval_steps": 500,
+  "global_step": 2952,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006772773450728074,
+      "grad_norm": 11.0625,
+      "learning_rate": 6.756756756756757e-08,
+      "loss": 2.4477,
+      "step": 1
+    },
+    {
+      "epoch": 0.0033863867253640365,
+      "grad_norm": 11.75,
+      "learning_rate": 3.378378378378379e-07,
+      "loss": 2.4744,
+      "step": 5
+    },
+    {
+      "epoch": 0.006772773450728073,
+      "grad_norm": 12.625,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 2.467,
+      "step": 10
+    },
+    {
+      "epoch": 0.01015916017609211,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.0135135135135136e-06,
+      "loss": 2.4589,
+      "step": 15
+    },
+    {
+      "epoch": 0.013545546901456146,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 2.4357,
+      "step": 20
+    },
+    {
+      "epoch": 0.016931933626820182,
+      "grad_norm": 10.5,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 2.4538,
+      "step": 25
+    },
+    {
+      "epoch": 0.02031832035218422,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 2.4134,
+      "step": 30
+    },
+    {
+      "epoch": 0.023704707077548254,
+      "grad_norm": 5.0625,
+      "learning_rate": 2.364864864864865e-06,
+      "loss": 2.4152,
+      "step": 35
+    },
+    {
+      "epoch": 0.027091093802912292,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 2.4089,
+      "step": 40
+    },
+    {
+      "epoch": 0.03047748052827633,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.040540540540541e-06,
+      "loss": 2.3663,
+      "step": 45
+    },
+    {
+      "epoch": 0.033863867253640365,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 2.348,
+      "step": 50
+    },
+    {
+      "epoch": 0.0372502539790044,
+      "grad_norm": 2.25,
+      "learning_rate": 3.7162162162162162e-06,
+      "loss": 2.3492,
+      "step": 55
+    },
+    {
+      "epoch": 0.04063664070436844,
+      "grad_norm": 1.8359375,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 2.3381,
+      "step": 60
+    },
+    {
+      "epoch": 0.04402302742973248,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.391891891891892e-06,
+      "loss": 2.3151,
+      "step": 65
+    },
+    {
+      "epoch": 0.04740941415509651,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 2.287,
+      "step": 70
+    },
+    {
+      "epoch": 0.05079580088046055,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.067567567567568e-06,
+      "loss": 2.2966,
+      "step": 75
+    },
+    {
+      "epoch": 0.054182187605824585,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 2.2718,
+      "step": 80
+    },
+    {
+      "epoch": 0.05756857433118862,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.743243243243244e-06,
+      "loss": 2.2315,
+      "step": 85
+    },
+    {
+      "epoch": 0.06095496105655266,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 2.2551,
+      "step": 90
+    },
+    {
+      "epoch": 0.06434134778191669,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.41891891891892e-06,
+      "loss": 2.2322,
+      "step": 95
+    },
+    {
+      "epoch": 0.06772773450728073,
+      "grad_norm": 1.421875,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 2.2239,
+      "step": 100
+    },
+    {
+      "epoch": 0.07111412123264477,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.0945945945945946e-06,
+      "loss": 2.2143,
+      "step": 105
+    },
+    {
+      "epoch": 0.0745005079580088,
+      "grad_norm": 1.359375,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 2.2222,
+      "step": 110
+    },
+    {
+      "epoch": 0.07788689468337284,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.77027027027027e-06,
+      "loss": 2.2246,
+      "step": 115
+    },
+    {
+      "epoch": 0.08127328140873688,
+      "grad_norm": 1.3671875,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 2.2054,
+      "step": 120
+    },
+    {
+      "epoch": 0.08465966813410092,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.445945945945948e-06,
+      "loss": 2.189,
+      "step": 125
+    },
+    {
+      "epoch": 0.08804605485946496,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 2.2255,
+      "step": 130
+    },
+    {
+      "epoch": 0.091432441584829,
+      "grad_norm": 1.3359375,
+      "learning_rate": 9.121621621621622e-06,
+      "loss": 2.223,
+      "step": 135
+    },
+    {
+      "epoch": 0.09481882831019302,
+      "grad_norm": 1.328125,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 2.1796,
+      "step": 140
+    },
+    {
+      "epoch": 0.09820521503555706,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.797297297297298e-06,
+      "loss": 2.1801,
+      "step": 145
+    },
+    {
+      "epoch": 0.1015916017609211,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.0135135135135136e-05,
+      "loss": 2.1779,
+      "step": 150
+    },
+    {
+      "epoch": 0.10497798848628513,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.0472972972972975e-05,
+      "loss": 2.1901,
+      "step": 155
+    },
+    {
+      "epoch": 0.10836437521164917,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 2.2081,
+      "step": 160
+    },
+    {
+      "epoch": 0.11175076193701321,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.114864864864865e-05,
+      "loss": 2.1736,
+      "step": 165
+    },
+    {
+      "epoch": 0.11513714866237724,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.1486486486486488e-05,
+      "loss": 2.1732,
+      "step": 170
+    },
+    {
+      "epoch": 0.11852353538774128,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.1824324324324325e-05,
+      "loss": 2.1891,
+      "step": 175
+    },
+    {
+      "epoch": 0.12190992211310532,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.2162162162162164e-05,
+      "loss": 2.1883,
+      "step": 180
+    },
+    {
+      "epoch": 0.12529630883846934,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.25e-05,
+      "loss": 2.1857,
+      "step": 185
+    },
+    {
+      "epoch": 0.12868269556383338,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.283783783783784e-05,
+      "loss": 2.174,
+      "step": 190
+    },
+    {
+      "epoch": 0.13206908228919742,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.3175675675675676e-05,
+      "loss": 2.1994,
+      "step": 195
+    },
+    {
+      "epoch": 0.13545546901456146,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 2.1782,
+      "step": 200
+    },
+    {
+      "epoch": 0.1388418557399255,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3851351351351352e-05,
+      "loss": 2.195,
+      "step": 205
+    },
+    {
+      "epoch": 0.14222824246528953,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4189189189189189e-05,
+      "loss": 2.1801,
+      "step": 210
+    },
+    {
+      "epoch": 0.14561462919065357,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.4527027027027028e-05,
+      "loss": 2.1748,
+      "step": 215
+    },
+    {
+      "epoch": 0.1490010159160176,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.4864864864864865e-05,
+      "loss": 2.1677,
+      "step": 220
+    },
+    {
+      "epoch": 0.15238740264138165,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.5202702702702704e-05,
+      "loss": 2.1867,
+      "step": 225
+    },
+    {
+      "epoch": 0.15577378936674569,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.554054054054054e-05,
+      "loss": 2.1737,
+      "step": 230
+    },
+    {
+      "epoch": 0.15916017609210972,
+      "grad_norm": 1.25,
+      "learning_rate": 1.587837837837838e-05,
+      "loss": 2.1658,
+      "step": 235
+    },
+    {
+      "epoch": 0.16254656281747376,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 2.1719,
+      "step": 240
+    },
+    {
+      "epoch": 0.1659329495428378,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.6554054054054057e-05,
+      "loss": 2.172,
+      "step": 245
+    },
+    {
+      "epoch": 0.16931933626820184,
+      "grad_norm": 1.25,
+      "learning_rate": 1.6891891891891896e-05,
+      "loss": 2.1763,
+      "step": 250
+    },
+    {
+      "epoch": 0.17270572299356587,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.722972972972973e-05,
+      "loss": 2.1797,
+      "step": 255
+    },
+    {
+      "epoch": 0.1760921097189299,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 2.1864,
+      "step": 260
+    },
+    {
+      "epoch": 0.17947849644429395,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.790540540540541e-05,
+      "loss": 2.1847,
+      "step": 265
+    },
+    {
+      "epoch": 0.182864883169658,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8243243243243244e-05,
+      "loss": 2.1668,
+      "step": 270
+    },
+    {
+      "epoch": 0.186251269895022,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.8581081081081082e-05,
+      "loss": 2.1856,
+      "step": 275
+    },
+    {
+      "epoch": 0.18963765662038604,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 2.1586,
+      "step": 280
+    },
+    {
+      "epoch": 0.19302404334575007,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.925675675675676e-05,
+      "loss": 2.1799,
+      "step": 285
+    },
+    {
+      "epoch": 0.1964104300711141,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9594594594594595e-05,
+      "loss": 2.1593,
+      "step": 290
+    },
+    {
+      "epoch": 0.19979681679647815,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9932432432432434e-05,
+      "loss": 2.1456,
+      "step": 295
+    },
+    {
+      "epoch": 0.2031832035218422,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9999888073536733e-05,
+      "loss": 2.1736,
+      "step": 300
+    },
+    {
+      "epoch": 0.20656959024720623,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9999433376573803e-05,
+      "loss": 2.1433,
+      "step": 305
+    },
+    {
+      "epoch": 0.20995597697257026,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9998628929598956e-05,
+      "loss": 2.1948,
+      "step": 310
+    },
+    {
+      "epoch": 0.2133423636979343,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9997474760749314e-05,
+      "loss": 2.1431,
+      "step": 315
+    },
+    {
+      "epoch": 0.21672875042329834,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9995970910394228e-05,
+      "loss": 2.1599,
+      "step": 320
+    },
+    {
+      "epoch": 0.22011513714866238,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9994117431133843e-05,
+      "loss": 2.1537,
+      "step": 325
+    },
+    {
+      "epoch": 0.22350152387402641,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9991914387797266e-05,
+      "loss": 2.1557,
+      "step": 330
+    },
+    {
+      "epoch": 0.22688791059939045,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9989361857440318e-05,
+      "loss": 2.1827,
+      "step": 335
+    },
+    {
+      "epoch": 0.2302742973247545,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.99864599293428e-05,
+      "loss": 2.1506,
+      "step": 340
+    },
+    {
+      "epoch": 0.23366068405011853,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.998320870500541e-05,
+      "loss": 2.1702,
+      "step": 345
+    },
+    {
+      "epoch": 0.23704707077548257,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.997960829814616e-05,
+      "loss": 2.1667,
+      "step": 350
+    },
+    {
+      "epoch": 0.2404334575008466,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.997565883469643e-05,
+      "loss": 2.163,
+      "step": 355
+    },
+    {
+      "epoch": 0.24381984422621064,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9971360452796523e-05,
+      "loss": 2.1578,
+      "step": 360
+    },
+    {
+      "epoch": 0.24720623095157468,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9966713302790875e-05,
+      "loss": 2.176,
+      "step": 365
+    },
+    {
+      "epoch": 0.2505926176769387,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9961717547222775e-05,
+      "loss": 2.1562,
+      "step": 370
+    },
+    {
+      "epoch": 0.25397900440230275,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.995637336082868e-05,
+      "loss": 2.1509,
+      "step": 375
+    },
+    {
+      "epoch": 0.25736539112766676,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9950680930532107e-05,
+      "loss": 2.176,
+      "step": 380
+    },
+    {
+      "epoch": 0.26075177785303083,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.994464045543709e-05,
+      "loss": 2.1551,
+      "step": 385
+    },
+    {
+      "epoch": 0.26413816457839484,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9938252146821236e-05,
+      "loss": 2.1689,
+      "step": 390
+    },
+    {
+      "epoch": 0.2675245513037589,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9931516228128295e-05,
+      "loss": 2.1372,
+      "step": 395
+    },
+    {
+      "epoch": 0.2709109380291229,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9924432934960384e-05,
+      "loss": 2.1622,
+      "step": 400
+    },
+    {
+      "epoch": 0.274297324754487,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9917002515069732e-05,
+      "loss": 2.1516,
+      "step": 405
+    },
+    {
+      "epoch": 0.277683711479851,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9909225228350007e-05,
+      "loss": 2.163,
+      "step": 410
+    },
+    {
+      "epoch": 0.28107009820521506,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9901101346827233e-05,
+      "loss": 2.1407,
+      "step": 415
+    },
+    {
+      "epoch": 0.28445648493057907,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.989263115465028e-05,
+      "loss": 2.153,
+      "step": 420
+    },
+    {
+      "epoch": 0.28784287165594313,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9883814948080918e-05,
+      "loss": 2.1556,
+      "step": 425
+    },
+    {
+      "epoch": 0.29122925838130714,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.987465303548345e-05,
+      "loss": 2.139,
+      "step": 430
+    },
+    {
+      "epoch": 0.2946156451066712,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9865145737313937e-05,
+      "loss": 2.1553,
+      "step": 435
+    },
+    {
+      "epoch": 0.2980020318320352,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9855293386108995e-05,
+      "loss": 2.1489,
+      "step": 440
+    },
+    {
+      "epoch": 0.30138841855739923,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9845096326474135e-05,
+      "loss": 2.1417,
+      "step": 445
+    },
+    {
+      "epoch": 0.3047748052827633,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9834554915071745e-05,
+      "loss": 2.1519,
+      "step": 450
+    },
+    {
+      "epoch": 0.3081611920081273,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.982366952060859e-05,
+      "loss": 2.1289,
+      "step": 455
+    },
+    {
+      "epoch": 0.31154757873349137,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.981244052382293e-05,
+      "loss": 2.1391,
+      "step": 460
+    },
+    {
+      "epoch": 0.3149339654588554,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9800868317471196e-05,
+      "loss": 2.151,
+      "step": 465
+    },
+    {
+      "epoch": 0.31832035218421945,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.978895330631425e-05,
+      "loss": 2.1507,
+      "step": 470
+    },
+    {
+      "epoch": 0.32170673890958346,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.977669590710324e-05,
+      "loss": 2.1597,
+      "step": 475
+    },
+    {
+      "epoch": 0.3250931256349475,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.976409654856501e-05,
+      "loss": 2.1552,
+      "step": 480
+    },
+    {
+      "epoch": 0.32847951236031153,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.975115567138711e-05,
+      "loss": 2.1485,
+      "step": 485
+    },
+    {
+      "epoch": 0.3318658990856756,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9737873728202376e-05,
+      "loss": 2.152,
+      "step": 490
+    },
+    {
+      "epoch": 0.3352522858110396,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.972425118357312e-05,
+      "loss": 2.1666,
+      "step": 495
+    },
+    {
+      "epoch": 0.3386386725364037,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9710288513974846e-05,
+      "loss": 2.1373,
+      "step": 500
+    },
+    {
+      "epoch": 0.3420250592617677,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.969598620777962e-05,
+      "loss": 2.1324,
+      "step": 505
+    },
+    {
+      "epoch": 0.34541144598713175,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9681344765238958e-05,
+      "loss": 2.1416,
+      "step": 510
+    },
+    {
+      "epoch": 0.34879783271249576,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.966636469846635e-05,
+      "loss": 2.1432,
+      "step": 515
+    },
+    {
+      "epoch": 0.3521842194378598,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9651046531419335e-05,
+      "loss": 2.1202,
+      "step": 520
+    },
+    {
+      "epoch": 0.35557060616322383,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9635390799881186e-05,
+      "loss": 2.1234,
+      "step": 525
+    },
+    {
+      "epoch": 0.3589569928885879,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.961939805144216e-05,
+      "loss": 2.15,
+      "step": 530
+    },
+    {
+      "epoch": 0.3623433796139519,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9603068845480347e-05,
+      "loss": 2.1517,
+      "step": 535
+    },
+    {
+      "epoch": 0.365729766339316,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9586403753142104e-05,
+      "loss": 2.1483,
+      "step": 540
+    },
+    {
+      "epoch": 0.36911615306468,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.956940335732209e-05,
+      "loss": 2.1739,
+      "step": 545
+    },
+    {
+      "epoch": 0.372502539790044,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9552068252642858e-05,
+      "loss": 2.1471,
+      "step": 550
+    },
+    {
+      "epoch": 0.37588892651540806,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9534399045434073e-05,
+      "loss": 2.1558,
+      "step": 555
+    },
+    {
+      "epoch": 0.37927531324077207,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9516396353711297e-05,
+      "loss": 2.1578,
+      "step": 560
+    },
+    {
+      "epoch": 0.38266169996613614,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9498060807154368e-05,
+      "loss": 2.1471,
+      "step": 565
+    },
+    {
+      "epoch": 0.38604808669150015,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9479393047085392e-05,
+      "loss": 2.1546,
+      "step": 570
+    },
+    {
+      "epoch": 0.3894344734168642,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.94603937264463e-05,
+      "loss": 2.1526,
+      "step": 575
+    },
+    {
+      "epoch": 0.3928208601422282,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9441063509776003e-05,
+      "loss": 2.1396,
+      "step": 580
+    },
+    {
+      "epoch": 0.3962072468675923,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9421403073187162e-05,
+      "loss": 2.15,
+      "step": 585
+    },
+    {
+      "epoch": 0.3995936335929563,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9401413104342535e-05,
+      "loss": 2.1277,
+      "step": 590
+    },
+    {
+      "epoch": 0.40298002031832036,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.938109430243093e-05,
+      "loss": 2.1419,
+      "step": 595
+    },
+    {
+      "epoch": 0.4063664070436844,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.936044737814273e-05,
+      "loss": 2.1405,
+      "step": 600
+    },
+    {
+      "epoch": 0.40975279376904844,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9339473053645056e-05,
+      "loss": 2.1514,
+      "step": 605
+    },
+    {
+      "epoch": 0.41313918049441245,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.931817206255651e-05,
+      "loss": 2.1326,
+      "step": 610
+    },
+    {
+      "epoch": 0.4165255672197765,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9296545149921488e-05,
+      "loss": 2.1325,
+      "step": 615
+    },
+    {
+      "epoch": 0.4199119539451405,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9274593072184152e-05,
+      "loss": 2.1387,
+      "step": 620
+    },
+    {
+      "epoch": 0.4232983406705046,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9252316597161947e-05,
+      "loss": 2.1507,
+      "step": 625
+    },
+    {
+      "epoch": 0.4266847273958686,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.922971650401877e-05,
+      "loss": 2.1469,
+      "step": 630
+    },
+    {
+      "epoch": 0.43007111412123267,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.920679358323769e-05,
+      "loss": 2.149,
+      "step": 635
+    },
+    {
+      "epoch": 0.4334575008465967,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9183548636593322e-05,
+      "loss": 2.1577,
+      "step": 640
+    },
+    {
+      "epoch": 0.43684388757196074,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9159982477123776e-05,
+      "loss": 2.1464,
+      "step": 645
+    },
+    {
+      "epoch": 0.44023027429732475,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9136095929102204e-05,
+      "loss": 2.1296,
+      "step": 650
+    },
+    {
+      "epoch": 0.4436166610226888,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9111889828007997e-05,
+      "loss": 2.1554,
+      "step": 655
+    },
+    {
+      "epoch": 0.44700304774805283,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.908736502049754e-05,
+      "loss": 2.135,
+      "step": 660
+    },
+    {
+      "epoch": 0.45038943447341684,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9062522364374617e-05,
+      "loss": 2.1337,
+      "step": 665
+    },
+    {
+      "epoch": 0.4537758211987809,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.903736272856038e-05,
+      "loss": 2.1251,
+      "step": 670
+    },
+    {
+      "epoch": 0.4571622079241449,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9011886993062994e-05,
+      "loss": 2.1523,
+      "step": 675
+    },
+    {
+      "epoch": 0.460548594649509,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8986096048946826e-05,
+      "loss": 2.1296,
+      "step": 680
+    },
+    {
+      "epoch": 0.463934981374873,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8959990798301286e-05,
+      "loss": 2.1298,
+      "step": 685
+    },
+    {
+      "epoch": 0.46732136810023706,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.893357215420929e-05,
+      "loss": 2.1453,
+      "step": 690
+    },
+    {
+      "epoch": 0.47070775482560107,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8906841040715304e-05,
+      "loss": 2.1654,
+      "step": 695
+    },
+    {
+      "epoch": 0.47409414155096513,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8879798392793033e-05,
+      "loss": 2.132,
+      "step": 700
+    },
+    {
+      "epoch": 0.47748052827632914,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8852445156312713e-05,
+      "loss": 2.1608,
+      "step": 705
+    },
+    {
+      "epoch": 0.4808669150016932,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8824782288008038e-05,
+      "loss": 2.1316,
+      "step": 710
+    },
+    {
+      "epoch": 0.4842533017270572,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8796810755442675e-05,
+      "loss": 2.1397,
+      "step": 715
+    },
+    {
+      "epoch": 0.4876396884524213,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8768531536976452e-05,
+      "loss": 2.1291,
+      "step": 720
+    },
+    {
+      "epoch": 0.4910260751777853,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8739945621731114e-05,
+      "loss": 2.1432,
+      "step": 725
+    },
+    {
+      "epoch": 0.49441246190314936,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8711054009555736e-05,
+      "loss": 2.1363,
+      "step": 730
+    },
+    {
+      "epoch": 0.49779884862851337,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8681857710991745e-05,
+      "loss": 2.1354,
+      "step": 735
+    },
+    {
+      "epoch": 0.5011852353538774,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.865235774723758e-05,
+      "loss": 2.1494,
+      "step": 740
+    },
+    {
+      "epoch": 0.5045716220792414,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.862255515011298e-05,
+      "loss": 2.1454,
+      "step": 745
+    },
+    {
+      "epoch": 0.5079580088046055,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8592450962022872e-05,
+      "loss": 2.1667,
+      "step": 750
+    },
+    {
+      "epoch": 0.5113443955299696,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8562046235920938e-05,
+      "loss": 2.1121,
+      "step": 755
+    },
+    {
+      "epoch": 0.5147307822553335,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8531342035272768e-05,
+      "loss": 2.1343,
+      "step": 760
+    },
+    {
+      "epoch": 0.5181171689806976,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.850033943401867e-05,
+      "loss": 2.1372,
+      "step": 765
+    },
+    {
+      "epoch": 0.5215035557060617,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.84690395165361e-05,
+      "loss": 2.1472,
+      "step": 770
+    },
+    {
+      "epoch": 0.5248899424314256,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8437443377601736e-05,
+      "loss": 2.1291,
+      "step": 775
+    },
+    {
+      "epoch": 0.5282763291567897,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8405552122353212e-05,
+      "loss": 2.146,
+      "step": 780
+    },
+    {
+      "epoch": 0.5316627158821537,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8373366866250407e-05,
+      "loss": 2.1479,
+      "step": 785
+    },
+    {
+      "epoch": 0.5350491026075178,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8340888735036485e-05,
+      "loss": 2.1379,
+      "step": 790
+    },
+    {
+      "epoch": 0.5384354893328818,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.830811886469849e-05,
+      "loss": 2.1265,
+      "step": 795
+    },
+    {
+      "epoch": 0.5418218760582458,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8275058401427622e-05,
+      "loss": 2.1396,
+      "step": 800
+    },
+    {
+      "epoch": 0.5452082627836099,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8241708501579146e-05,
+      "loss": 2.1362,
+      "step": 805
+    },
+    {
+      "epoch": 0.548594649508974,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8208070331631943e-05,
+      "loss": 2.1365,
+      "step": 810
+    },
+    {
+      "epoch": 0.5519810362343379,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.817414506814772e-05,
+      "loss": 2.1489,
+      "step": 815
+    },
+    {
+      "epoch": 0.555367422959702,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8139933897729833e-05,
+      "loss": 2.1518,
+      "step": 820
+    },
+    {
+      "epoch": 0.558753809685066,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8105438016981816e-05,
+      "loss": 2.1293,
+      "step": 825
+    },
+    {
+      "epoch": 0.5621401964104301,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.807065863246551e-05,
+      "loss": 2.1342,
+      "step": 830
+    },
+    {
+      "epoch": 0.5655265831357941,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8035596960658856e-05,
+      "loss": 2.1393,
+      "step": 835
+    },
+    {
+      "epoch": 0.5689129698611581,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8000254227913346e-05,
+      "loss": 2.1276,
+      "step": 840
+    },
+    {
+      "epoch": 0.5722993565865222,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7964631670411154e-05,
+      "loss": 2.1326,
+      "step": 845
+    },
+    {
+      "epoch": 0.5756857433118863,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7928730534121872e-05,
+      "loss": 2.1374,
+      "step": 850
+    },
+    {
+      "epoch": 0.5790721300372502,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7892552074758932e-05,
+      "loss": 2.1099,
+      "step": 855
+    },
+    {
+      "epoch": 0.5824585167626143,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7856097557735697e-05,
+      "loss": 2.1373,
+      "step": 860
+    },
+    {
+      "epoch": 0.5858449034879784,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.7819368258121188e-05,
+      "loss": 2.1304,
+      "step": 865
+    },
+    {
+      "epoch": 0.5892312902133424,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.77823654605955e-05,
+      "loss": 2.1412,
+      "step": 870
+    },
+    {
+      "epoch": 0.5926176769387064,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7745090459404853e-05,
+      "loss": 2.1393,
+      "step": 875
+    },
+    {
+      "epoch": 0.5960040636640704,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7707544558316332e-05,
+      "loss": 2.1307,
+      "step": 880
+    },
+    {
+      "epoch": 0.5993904503894345,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.766972907057229e-05,
+      "loss": 2.1265,
+      "step": 885
+    },
+    {
+      "epoch": 0.6027768371147985,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.763164531884439e-05,
+      "loss": 2.1137,
+      "step": 890
+    },
+    {
+      "epoch": 0.6061632238401625,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.7593294635187385e-05,
+      "loss": 2.1663,
+      "step": 895
+    },
+    {
+      "epoch": 0.6095496105655266,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7554678360992475e-05,
+      "loss": 2.1131,
+      "step": 900
+    },
+    {
+      "epoch": 0.6129359972908907,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7515797846940435e-05,
+      "loss": 2.129,
+      "step": 905
+    },
+    {
+      "epoch": 0.6163223840162546,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7476654452954354e-05,
+      "loss": 2.1386,
+      "step": 910
+    },
+    {
+      "epoch": 0.6197087707416187,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7437249548152046e-05,
+      "loss": 2.1429,
+      "step": 915
+    },
+    {
+      "epoch": 0.6230951574669827,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7397584510798208e-05,
+      "loss": 2.13,
+      "step": 920
+    },
+    {
+      "epoch": 0.6264815441923468,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.735766072825618e-05,
+      "loss": 2.1204,
+      "step": 925
+    },
+    {
+      "epoch": 0.6298679309177108,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7317479596939426e-05,
+      "loss": 2.1264,
+      "step": 930
+    },
+    {
+      "epoch": 0.6332543176430748,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.72770425222627e-05,
+      "loss": 2.136,
+      "step": 935
+    },
+    {
+      "epoch": 0.6366407043684389,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.723635091859287e-05,
+      "loss": 2.1385,
+      "step": 940
+    },
+    {
+      "epoch": 0.640027091093803,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.7195406209199473e-05,
+      "loss": 2.1392,
+      "step": 945
+    },
+    {
+      "epoch": 0.6434134778191669,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.7154209826204916e-05,
+      "loss": 2.1347,
+      "step": 950
+    },
+    {
+      "epoch": 0.646799864544531,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7112763210534395e-05,
+      "loss": 2.1193,
+      "step": 955
+    },
+    {
+      "epoch": 0.650186251269895,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 2.1334,
+      "step": 960
+    },
+    {
+      "epoch": 0.6535726379952591,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7029125088577418e-05,
+      "loss": 2.1182,
+      "step": 965
+    },
+    {
+      "epoch": 0.6569590247206231,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.698693650770014e-05,
+      "loss": 2.1225,
+      "step": 970
+    },
+    {
+      "epoch": 0.6603454114459871,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6944503544862924e-05,
+      "loss": 2.1439,
+      "step": 975
+    },
+    {
+      "epoch": 0.6637317981713512,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.690182768424279e-05,
+      "loss": 2.1131,
+      "step": 980
+    },
+    {
+      "epoch": 0.6671181848967151,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6858910418512592e-05,
+      "loss": 2.1048,
+      "step": 985
+    },
+    {
+      "epoch": 0.6705045716220792,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6815753248788814e-05,
+      "loss": 2.1454,
+      "step": 990
+    },
+    {
+      "epoch": 0.6738909583474433,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.677235768457904e-05,
+      "loss": 2.1364,
+      "step": 995
+    },
+    {
+      "epoch": 0.6772773450728073,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.672872524372919e-05,
+      "loss": 2.1337,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6806637317981713,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.668485745237041e-05,
+      "loss": 2.147,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6840501185235354,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6640755844865693e-05,
+      "loss": 2.1381,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6874365052488994,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6596421963756225e-05,
+      "loss": 2.119,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6908228919742635,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6551857359707407e-05,
+      "loss": 2.1108,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6942092786996275,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.650706359145464e-05,
+      "loss": 2.1411,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6975956654249915,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6462042225748803e-05,
+      "loss": 2.1335,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7009820521503556,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.641679483730144e-05,
+      "loss": 2.1352,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7043684388757196,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.637132300872969e-05,
+      "loss": 2.1366,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7077548256010836,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6325628330500928e-05,
+      "loss": 2.1233,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7111412123264477,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.627971240087715e-05,
+      "loss": 2.1283,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7145275990518117,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.6233576825859042e-05,
+      "loss": 2.1418,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7179139857771758,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6187223219129825e-05,
+      "loss": 2.1194,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7213003725025398,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.6140653201998833e-05,
+      "loss": 2.1336,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7246867592279038,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.609386840334475e-05,
+      "loss": 2.1405,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7280731459532679,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.60468704595587e-05,
+      "loss": 2.1323,
+      "step": 1075
+    },
+    {
+      "epoch": 0.731459532678632,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5999661014486956e-05,
+      "loss": 2.1257,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7348459194039959,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.595224171937349e-05,
+      "loss": 2.162,
+      "step": 1085
+    },
+    {
+      "epoch": 0.73823230612936,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.5904614232802173e-05,
+      "loss": 2.1532,
+      "step": 1090
+    },
+    {
+      "epoch": 0.741618692854724,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.58567802206388e-05,
+      "loss": 2.1215,
+      "step": 1095
+    },
+    {
+      "epoch": 0.745005079580088,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5808741355972808e-05,
+      "loss": 2.1249,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7483914663054521,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.5760499319058754e-05,
+      "loss": 2.1106,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7517778530308161,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.5712055797257542e-05,
+      "loss": 2.1093,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7551642397561802,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5663412484977417e-05,
+      "loss": 2.1366,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7585506264815441,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5614571083614683e-05,
+      "loss": 2.1394,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7619370132069082,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5565533301494217e-05,
+      "loss": 2.1316,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7653233999322723,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.5516300853809688e-05,
+      "loss": 2.1569,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7687097866576363,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.5466875462563592e-05,
+      "loss": 2.1203,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7720961733830003,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5417258856506995e-05,
+      "loss": 2.1278,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7754825601083644,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.536745277107909e-05,
+      "loss": 2.113,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7788689468337284,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5317458948346482e-05,
+      "loss": 2.1467,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7822553335590925,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.526727913694226e-05,
+      "loss": 2.1434,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7856417202844564,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.5216915092004847e-05,
+      "loss": 2.1192,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7890281070098205,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.516636857511658e-05,
+      "loss": 2.1307,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7924144937351846,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.5115641354242135e-05,
+      "loss": 2.1284,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7958008804605486,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5064735203666644e-05,
+      "loss": 2.1274,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7991872671859126,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5013651903933686e-05,
+      "loss": 2.1549,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8025736539112767,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4962393241782966e-05,
+      "loss": 2.1386,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8059600406366407,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.491096101008784e-05,
+      "loss": 2.1259,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8093464273620048,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.485935700779261e-05,
+      "loss": 2.1174,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8127328140873687,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4807583039849589e-05,
+      "loss": 2.1411,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8161192008127328,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.4755640917155975e-05,
+      "loss": 2.1359,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8195055875380969,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4703532456490518e-05,
+      "loss": 2.1403,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8228919742634608,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.465125948044996e-05,
+      "loss": 2.1396,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8262783609888249,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.45988238173853e-05,
+      "loss": 2.1264,
+      "step": 1220
+    },
+    {
+      "epoch": 0.829664747714189,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4546227301337838e-05,
+      "loss": 2.1368,
+      "step": 1225
+    },
+    {
+      "epoch": 0.833051134439553,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.449347177197503e-05,
+      "loss": 2.1265,
+      "step": 1230
+    },
+    {
+      "epoch": 0.836437521164917,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.4440559074526126e-05,
+      "loss": 2.1437,
+      "step": 1235
+    },
+    {
+      "epoch": 0.839823907890281,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.4387491059717653e-05,
+      "loss": 2.125,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8432102946156451,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.4334269583708664e-05,
+      "loss": 2.1199,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8465966813410092,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4280896508025825e-05,
+      "loss": 2.1056,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8499830680663731,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.4227373699498305e-05,
+      "loss": 2.1412,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8533694547917372,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4173703030192467e-05,
+      "loss": 2.1258,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8567558415171013,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4119886377346402e-05,
+      "loss": 2.1338,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8601422282424653,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4065925623304267e-05,
+      "loss": 2.1335,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8635286149678293,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.4011822655450444e-05,
+      "loss": 2.1066,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8669150016931934,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3957579366143521e-05,
+      "loss": 2.114,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8703013884185574,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.3903197652650114e-05,
+      "loss": 2.1328,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8736877751439215,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3848679417078487e-05,
+      "loss": 2.1254,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8770741618692854,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.379402656631205e-05,
+      "loss": 2.1269,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8804605485946495,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.373924101194264e-05,
+      "loss": 2.1207,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8838469353200136,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3684324670203658e-05,
+      "loss": 2.1297,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8872333220453776,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.3629279461903064e-05,
+      "loss": 2.1379,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8906197087707416,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.3574107312356174e-05,
+      "loss": 2.1275,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8940060954961057,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.351881015131833e-05,
+      "loss": 2.1295,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8973924822214697,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.3463389912917401e-05,
+      "loss": 2.1267,
+      "step": 1325
+    },
+    {
+      "epoch": 0.9007788689468337,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3407848535586122e-05,
+      "loss": 2.1419,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9041652556721977,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.3352187961994309e-05,
+      "loss": 2.1192,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9075516423975618,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3296410138980903e-05,
+      "loss": 2.1367,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9109380291229259,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3240517017485872e-05,
+      "loss": 2.1255,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9143244158482898,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3184510552481988e-05,
+      "loss": 2.129,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9177108025736539,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.3128392702906418e-05,
+      "loss": 2.1271,
+      "step": 1355
+    },
+    {
+      "epoch": 0.921097189299018,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.307216543159225e-05,
+      "loss": 2.1371,
+      "step": 1360
+    },
+    {
+      "epoch": 0.924483576024382,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.301583070519979e-05,
+      "loss": 2.1411,
+      "step": 1365
+    },
+    {
+      "epoch": 0.927869962749746,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2959390494147814e-05,
+      "loss": 2.1348,
+      "step": 1370
+    },
+    {
+      "epoch": 0.93125634947511,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2902846772544625e-05,
+      "loss": 2.1285,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9346427362004741,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2846201518119019e-05,
+      "loss": 2.1163,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9380291229258382,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2789456712151102e-05,
+      "loss": 2.1503,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9414155096512021,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.273261433940299e-05,
+      "loss": 2.1282,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9448018963765662,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.2675676388049394e-05,
+      "loss": 2.1293,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9481882831019303,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.2618644849608068e-05,
+      "loss": 2.135,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9515746698272943,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2561521718870175e-05,
+      "loss": 2.125,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9549610565526583,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.2504308993830488e-05,
+      "loss": 2.1314,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9583474432780223,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2447008675617522e-05,
+      "loss": 2.1242,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9617338300033864,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2389622768423537e-05,
+      "loss": 2.1422,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9651202167287504,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2332153279434438e-05,
+      "loss": 2.1117,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9685066034541144,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.2274602218759569e-05,
+      "loss": 2.1154,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9718929901794785,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.2216971599361402e-05,
+      "loss": 2.1349,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9752793769048426,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2159263436985139e-05,
+      "loss": 2.1118,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9786657636302065,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.2101479750088197e-05,
+      "loss": 2.1256,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9820521503555706,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2043622559769614e-05,
+      "loss": 2.1307,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9854385370809347,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1985693889699359e-05,
+      "loss": 2.1183,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9888249238062987,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.192769576604754e-05,
+      "loss": 2.1364,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9922113105316627,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.186963021741355e-05,
+      "loss": 2.1228,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9955976972570267,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1811499274755099e-05,
+      "loss": 2.1332,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9989840839823908,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1753304971317181e-05,
+      "loss": 2.1211,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9996613613274636,
+      "eval_loss": 2.1251890659332275,
+      "eval_runtime": 86.1037,
+      "eval_samples_per_second": 15.156,
+      "eval_steps_per_second": 1.905,
+      "step": 1476
+    },
+    {
+      "epoch": 1.0023704707077548,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.1695049342560969e-05,
+      "loss": 2.0871,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0057568574331188,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1636734426092597e-05,
+      "loss": 2.099,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0091432441584829,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.1578362261591916e-05,
+      "loss": 2.106,
+      "step": 1490
+    },
+    {
+      "epoch": 1.012529630883847,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1519934890741129e-05,
+      "loss": 2.1098,
+      "step": 1495
+    },
+    {
+      "epoch": 1.015916017609211,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.1461454357153407e-05,
+      "loss": 2.0854,
+      "step": 1500
+    },
+    {
+      "epoch": 1.019302404334575,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1402922706301373e-05,
+      "loss": 2.0965,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0226887910599392,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1344341985445594e-05,
+      "loss": 2.0991,
+      "step": 1510
+    },
+    {
+      "epoch": 1.026075177785303,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1285714243562951e-05,
+      "loss": 2.0845,
+      "step": 1515
+    },
+    {
+      "epoch": 1.029461564510667,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.1227041531274978e-05,
+      "loss": 2.0857,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0328479512360311,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.116832590077614e-05,
+      "loss": 2.1092,
+      "step": 1525
+    },
+    {
+      "epoch": 1.0362343379613952,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.1109569405762051e-05,
+      "loss": 2.1228,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0396207246867593,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1050774101357638e-05,
+      "loss": 2.1078,
+      "step": 1535
+    },
+    {
+      "epoch": 1.0430071114121233,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0991942044045277e-05,
+      "loss": 2.094,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0463934981374874,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0933075291592832e-05,
+      "loss": 2.0997,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0497798848628512,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0874175902981711e-05,
+      "loss": 2.0995,
+      "step": 1550
+    },
+    {
+      "epoch": 1.0531662715882153,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0815245938334833e-05,
+      "loss": 2.0921,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0565526583135794,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.075628745884457e-05,
+      "loss": 2.1105,
+      "step": 1560
+    },
+    {
+      "epoch": 1.0599390450389434,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.069730252670066e-05,
+      "loss": 2.073,
+      "step": 1565
+    },
+    {
+      "epoch": 1.0633254317643075,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0638293205018084e-05,
+      "loss": 2.0949,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0667118184896716,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0579261557764884e-05,
+      "loss": 2.0943,
+      "step": 1575
+    },
+    {
+      "epoch": 1.0700982052150356,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0520209649689979e-05,
+      "loss": 2.1009,
+      "step": 1580
+    },
+    {
+      "epoch": 1.0734845919403997,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.0461139546250969e-05,
+      "loss": 2.1363,
+      "step": 1585
+    },
+    {
+      "epoch": 1.0768709786657635,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.0402053313541854e-05,
+      "loss": 2.0965,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0802573653911276,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.03429530182208e-05,
+      "loss": 2.1188,
+      "step": 1595
+    },
+    {
+      "epoch": 1.0836437521164917,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0283840727437832e-05,
+      "loss": 2.0955,
+      "step": 1600
+    },
+    {
+      "epoch": 1.0870301388418557,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0224718508762549e-05,
+      "loss": 2.0909,
+      "step": 1605
+    },
+    {
+      "epoch": 1.0904165255672198,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0165588430111787e-05,
+      "loss": 2.0914,
+      "step": 1610
+    },
+    {
+      "epoch": 1.0938029122925839,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0106452559677309e-05,
+      "loss": 2.0998,
+      "step": 1615
+    },
+    {
+      "epoch": 1.097189299017948,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.0047312965853456e-05,
+      "loss": 2.0922,
+      "step": 1620
+    },
+    {
+      "epoch": 1.100575685743312,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.988171717164795e-06,
+      "loss": 2.1152,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1039620724686758,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.92903088219378e-06,
+      "loss": 2.1002,
+      "step": 1630
+    },
+    {
+      "epoch": 1.10734845919404,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.8698925295084e-06,
+      "loss": 2.099,
+      "step": 1635
+    },
+    {
+      "epoch": 1.110734845919404,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.810758727589814e-06,
+      "loss": 2.1044,
+      "step": 1640
+    },
+    {
+      "epoch": 1.114121232644768,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.751631544760004e-06,
+      "loss": 2.0954,
+      "step": 1645
+    },
+    {
+      "epoch": 1.117507619370132,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.69251304910945e-06,
+      "loss": 2.0937,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1208940060954962,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.633405308424774e-06,
+      "loss": 2.1045,
+      "step": 1655
+    },
+    {
+      "epoch": 1.1242803928208602,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.574310390116421e-06,
+      "loss": 2.1024,
+      "step": 1660
+    },
+    {
+      "epoch": 1.127666779546224,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.51523036114634e-06,
+      "loss": 2.1049,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1310531662715881,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.45616728795571e-06,
+      "loss": 2.0932,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1344395529969522,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.397123236392647e-06,
+      "loss": 2.1045,
+      "step": 1675
+    },
+    {
+      "epoch": 1.1378259397223163,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.338100271639932e-06,
+      "loss": 2.1071,
+      "step": 1680
+    },
+    {
+      "epoch": 1.1412123264476803,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.279100458142815e-06,
+      "loss": 2.1117,
+      "step": 1685
+    },
+    {
+      "epoch": 1.1445987131730444,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.220125859536766e-06,
+      "loss": 2.1105,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1479850998984085,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.161178538575321e-06,
+      "loss": 2.1038,
+      "step": 1695
+    },
+    {
+      "epoch": 1.1513714866237725,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.102260557057936e-06,
+      "loss": 2.0936,
+      "step": 1700
+    },
+    {
+      "epoch": 1.1547578733491364,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.04337397575785e-06,
+      "loss": 2.1051,
+      "step": 1705
+    },
+    {
+      "epoch": 1.1581442600745004,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.984520854350014e-06,
+      "loss": 2.0896,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1615306467998645,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.925703251339061e-06,
+      "loss": 2.0926,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1649170335252286,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.866923223987303e-06,
+      "loss": 2.0873,
+      "step": 1720
+    },
+    {
+      "epoch": 1.1683034202505926,
+      "grad_norm": 1.25,
+      "learning_rate": 8.808182828242755e-06,
+      "loss": 2.1047,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1716898069759567,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.749484118667242e-06,
+      "loss": 2.0866,
+      "step": 1730
+    },
+    {
+      "epoch": 1.1750761937013208,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.690829148364544e-06,
+      "loss": 2.105,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1784625804266846,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.632219968908556e-06,
+      "loss": 2.097,
+      "step": 1740
+    },
+    {
+      "epoch": 1.1818489671520487,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.573658630271562e-06,
+      "loss": 2.1265,
+      "step": 1745
+    },
+    {
+      "epoch": 1.1852353538774127,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.515147180752504e-06,
+      "loss": 2.0835,
+      "step": 1750
+    },
+    {
+      "epoch": 1.1886217406027768,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.45668766690537e-06,
+      "loss": 2.0991,
+      "step": 1755
+    },
+    {
+      "epoch": 1.1920081273281409,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.398282133467579e-06,
+      "loss": 2.0744,
+      "step": 1760
+    },
+    {
+      "epoch": 1.195394514053505,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.339932623288487e-06,
+      "loss": 2.0952,
+      "step": 1765
+    },
+    {
+      "epoch": 1.198780900778869,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.281641177257929e-06,
+      "loss": 2.1104,
+      "step": 1770
+    },
+    {
+      "epoch": 1.202167287504233,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.223409834234821e-06,
+      "loss": 2.1028,
+      "step": 1775
+    },
+    {
+      "epoch": 1.2055536742295971,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.165240630975863e-06,
+      "loss": 2.0907,
+      "step": 1780
+    },
+    {
+      "epoch": 1.208940060954961,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.107135602064288e-06,
+      "loss": 2.0951,
+      "step": 1785
+    },
+    {
+      "epoch": 1.212326447680325,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 2.1017,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2157128344056891,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.991126194322052e-06,
+      "loss": 2.1146,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2190992211310532,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.93322587315047e-06,
+      "loss": 2.0922,
+      "step": 1800
+    },
+    {
+      "epoch": 1.2224856078564172,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.875397841502536e-06,
+      "loss": 2.0903,
+      "step": 1805
+    },
+    {
+      "epoch": 1.2258719945817813,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.81764412202833e-06,
+      "loss": 2.1242,
+      "step": 1810
+    },
+    {
+      "epoch": 1.2292583813071452,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.75996673477872e-06,
+      "loss": 2.1008,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2326447680325092,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.702367697134701e-06,
+      "loss": 2.1103,
+      "step": 1820
+    },
+    {
+      "epoch": 1.2360311547578733,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.644849023736842e-06,
+      "loss": 2.1026,
+      "step": 1825
+    },
+    {
+      "epoch": 1.2394175414832374,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.5874127264148075e-06,
+      "loss": 2.089,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2428039282086014,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.530060814116992e-06,
+      "loss": 2.113,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2461903149339655,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.4727952928402695e-06,
+      "loss": 2.119,
+      "step": 1840
+    },
+    {
+      "epoch": 1.2495767016593295,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.415618165559799e-06,
+      "loss": 2.0882,
+      "step": 1845
+    },
+    {
+      "epoch": 1.2529630883846936,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.358531432159002e-06,
+      "loss": 2.106,
+      "step": 1850
+    },
+    {
+      "epoch": 1.2563494751100577,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.301537089359576e-06,
+      "loss": 2.0879,
+      "step": 1855
+    },
+    {
+      "epoch": 1.2597358618354215,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.2446371306516935e-06,
+      "loss": 2.1054,
+      "step": 1860
+    },
+    {
+      "epoch": 1.2631222485607856,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.187833546224245e-06,
+      "loss": 2.0957,
+      "step": 1865
+    },
+    {
+      "epoch": 1.2665086352861497,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.1311283228952345e-06,
+      "loss": 2.0789,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2698950220115137,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.074523444042305e-06,
+      "loss": 2.1222,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2732814087368778,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.018020889533348e-06,
+      "loss": 2.117,
+      "step": 1880
+    },
+    {
+      "epoch": 1.2766677954622419,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.961622635657251e-06,
+      "loss": 2.1018,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2800541821876057,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.905330655054782e-06,
+      "loss": 2.0972,
+      "step": 1890
+    },
+    {
+      "epoch": 1.2834405689129698,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.8491469166496e-06,
+      "loss": 2.1186,
+      "step": 1895
+    },
+    {
+      "epoch": 1.2868269556383338,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.793073385579373e-06,
+      "loss": 2.0914,
+      "step": 1900
+    },
+    {
+      "epoch": 1.290213342363698,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.737112023127039e-06,
+      "loss": 2.1059,
+      "step": 1905
+    },
+    {
+      "epoch": 1.293599729089062,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.681264786652236e-06,
+      "loss": 2.1072,
+      "step": 1910
+    },
+    {
+      "epoch": 1.296986115814426,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.6255336295228e-06,
+      "loss": 2.1114,
+      "step": 1915
+    },
+    {
+      "epoch": 1.30037250253979,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.569920501046474e-06,
+      "loss": 2.0876,
+      "step": 1920
+    },
+    {
+      "epoch": 1.3037588892651542,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.514427346402712e-06,
+      "loss": 2.1029,
+      "step": 1925
+    },
+    {
+      "epoch": 1.3071452759905182,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.459056106574646e-06,
+      "loss": 2.1116,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3105316627158823,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.403808718281188e-06,
+      "loss": 2.0932,
+      "step": 1935
+    },
+    {
+      "epoch": 1.3139180494412461,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.348687113909304e-06,
+      "loss": 2.0972,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3173044361666102,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.293693221446424e-06,
+      "loss": 2.1178,
+      "step": 1945
+    },
+    {
+      "epoch": 1.3206908228919743,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.238828964412989e-06,
+      "loss": 2.1112,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3240772096173383,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.184096261795185e-06,
+      "loss": 2.1075,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3274635963427024,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.129497027977829e-06,
+      "loss": 2.1009,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3308499830680665,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.075033172677398e-06,
+      "loss": 2.1045,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3342363697934303,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.020706600875237e-06,
+      "loss": 2.0776,
+      "step": 1970
+    },
+    {
+      "epoch": 1.3376227565187944,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.966519212750923e-06,
+      "loss": 2.0804,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3410091432441584,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.912472903615821e-06,
+      "loss": 2.1139,
+      "step": 1980
+    },
+    {
+      "epoch": 1.3443955299695225,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.858569563846764e-06,
+      "loss": 2.0817,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3477819166948866,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.804811078819956e-06,
+      "loss": 2.1019,
+      "step": 1990
+    },
+    {
+      "epoch": 1.3511683034202506,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.75119932884502e-06,
+      "loss": 2.0934,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3545546901456147,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.69773618909923e-06,
+      "loss": 2.095,
+      "step": 2000
+    },
+    {
+      "epoch": 1.3579410768709788,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.644423529561911e-06,
+      "loss": 2.1009,
+      "step": 2005
+    },
+    {
+      "epoch": 1.3613274635963428,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.591263214949059e-06,
+      "loss": 2.0972,
+      "step": 2010
+    },
+    {
+      "epoch": 1.3647138503217067,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.538257104648095e-06,
+      "loss": 2.1094,
+      "step": 2015
+    },
+    {
+      "epoch": 1.3681002370470707,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.4854070526528445e-06,
+      "loss": 2.0918,
+      "step": 2020
+    },
+    {
+      "epoch": 1.3714866237724348,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.432714907498682e-06,
+      "loss": 2.1138,
+      "step": 2025
+    },
+    {
+      "epoch": 1.3748730104977989,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.380182512197873e-06,
+      "loss": 2.1019,
+      "step": 2030
+    },
+    {
+      "epoch": 1.378259397223163,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.327811704175123e-06,
+      "loss": 2.0865,
+      "step": 2035
+    },
+    {
+      "epoch": 1.381645783948527,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.2756043152032934e-06,
+      "loss": 2.1069,
+      "step": 2040
+    },
+    {
+      "epoch": 1.3850321706738908,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.22356217133934e-06,
+      "loss": 2.1139,
+      "step": 2045
+    },
+    {
+      "epoch": 1.388418557399255,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.171687092860464e-06,
+      "loss": 2.1028,
+      "step": 2050
+    },
+    {
+      "epoch": 1.391804944124619,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.119980894200391e-06,
+      "loss": 2.0999,
+      "step": 2055
+    },
+    {
+      "epoch": 1.395191330849983,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.068445383885961e-06,
+      "loss": 2.1036,
+      "step": 2060
+    },
+    {
+      "epoch": 1.398577717575347,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.017082364473849e-06,
+      "loss": 2.1075,
+      "step": 2065
+    },
+    {
+      "epoch": 1.4019641043007112,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.965893632487511e-06,
+      "loss": 2.1098,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4053504910260752,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.914880978354365e-06,
+      "loss": 2.1129,
+      "step": 2075
+    },
+    {
+      "epoch": 1.4087368777514393,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.86404618634314e-06,
+      "loss": 2.082,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4121232644768034,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.813391034501505e-06,
+      "loss": 2.104,
+      "step": 2085
+    },
+    {
+      "epoch": 1.4155096512021672,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.762917294593851e-06,
+      "loss": 2.0988,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4188960379275313,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.712626732039311e-06,
+      "loss": 2.0984,
+      "step": 2095
+    },
+    {
+      "epoch": 1.4222824246528953,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.662521105850056e-06,
+      "loss": 2.0974,
+      "step": 2100
+    },
+    {
+      "epoch": 1.4256688113782594,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.6126021685697134e-06,
+      "loss": 2.0955,
+      "step": 2105
+    },
+    {
+      "epoch": 1.4290551981036235,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.562871666212107e-06,
+      "loss": 2.1031,
+      "step": 2110
+    },
+    {
+      "epoch": 1.4324415848289875,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.5133313382001865e-06,
+      "loss": 2.0995,
+      "step": 2115
+    },
+    {
+      "epoch": 1.4358279715543514,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.463982917305155e-06,
+      "loss": 2.1049,
+      "step": 2120
+    },
+    {
+      "epoch": 1.4392143582797154,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.414828129585901e-06,
+      "loss": 2.0897,
+      "step": 2125
+    },
+    {
+      "epoch": 1.4426007450050795,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.365868694328599e-06,
+      "loss": 2.0924,
+      "step": 2130
+    },
+    {
+      "epoch": 1.4459871317304436,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.317106323986588e-06,
+      "loss": 2.0942,
+      "step": 2135
+    },
+    {
+      "epoch": 1.4493735184558076,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.2685427241204755e-06,
+      "loss": 2.097,
+      "step": 2140
+    },
+    {
+      "epoch": 1.4527599051811717,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.220179593338459e-06,
+      "loss": 2.0863,
+      "step": 2145
+    },
+    {
+      "epoch": 1.4561462919065358,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.1720186232369595e-06,
+      "loss": 2.1003,
+      "step": 2150
+    },
+    {
+      "epoch": 1.4595326786318998,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.124061498341403e-06,
+      "loss": 2.1006,
+      "step": 2155
+    },
+    {
+      "epoch": 1.462919065357264,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.076309896047337e-06,
+      "loss": 2.0967,
+      "step": 2160
+    },
+    {
+      "epoch": 1.466305452082628,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.028765486561749e-06,
+      "loss": 2.0996,
+      "step": 2165
+    },
+    {
+      "epoch": 1.4696918388079918,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.981429932844641e-06,
+      "loss": 2.1087,
+      "step": 2170
+    },
+    {
+      "epoch": 1.4730782255333559,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.934304890550878e-06,
+      "loss": 2.1034,
+      "step": 2175
+    },
+    {
+      "epoch": 1.47646461225872,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.887392007972265e-06,
+      "loss": 2.0972,
+      "step": 2180
+    },
+    {
+      "epoch": 1.479850998984084,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.840692925979901e-06,
+      "loss": 2.0967,
+      "step": 2185
+    },
+    {
+      "epoch": 1.483237385709448,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.7942092779667904e-06,
+      "loss": 2.1029,
+      "step": 2190
+    },
+    {
+      "epoch": 1.486623772434812,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.7479426897906946e-06,
+      "loss": 2.0948,
+      "step": 2195
+    },
+    {
+      "epoch": 1.490010159160176,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.7018947797172864e-06,
+      "loss": 2.1142,
+      "step": 2200
+    },
+    {
+      "epoch": 1.49339654588554,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.6560671583635467e-06,
+      "loss": 2.1055,
+      "step": 2205
+    },
+    {
+      "epoch": 1.4967829326109041,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.610461428641402e-06,
+      "loss": 2.1109,
+      "step": 2210
+    },
+    {
+      "epoch": 1.5001693193362682,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.56507918570169e-06,
+      "loss": 2.0996,
+      "step": 2215
+    },
+    {
+      "epoch": 1.5035557060616322,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.5199220168783564e-06,
+      "loss": 2.0915,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5069420927869963,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.474991501632928e-06,
+      "loss": 2.1108,
+      "step": 2225
+    },
+    {
+      "epoch": 1.5103284795123604,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.4302892114992814e-06,
+      "loss": 2.101,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5137148662377244,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.385816710028651e-06,
+      "loss": 2.0945,
+      "step": 2235
+    },
+    {
+      "epoch": 1.5171012529630885,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.341575552734978e-06,
+      "loss": 2.0848,
+      "step": 2240
+    },
+    {
+      "epoch": 1.5204876396884526,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.2975672870404717e-06,
+      "loss": 2.1088,
+      "step": 2245
+    },
+    {
+      "epoch": 1.5238740264138164,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.253793452221485e-06,
+      "loss": 2.0804,
+      "step": 2250
+    },
+    {
+      "epoch": 1.5272604131391805,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.210255579354713e-06,
+      "loss": 2.0982,
+      "step": 2255
+    },
+    {
+      "epoch": 1.5306467998645445,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.1669551912635874e-06,
+      "loss": 2.0691,
+      "step": 2260
+    },
+    {
+      "epoch": 1.5340331865899086,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.123893802465051e-06,
+      "loss": 2.1213,
+      "step": 2265
+    },
+    {
+      "epoch": 1.5374195733152725,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.0810729191165722e-06,
+      "loss": 2.115,
+      "step": 2270
+    },
+    {
+      "epoch": 1.5408059600406365,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.0384940389634623e-06,
+      "loss": 2.1023,
+      "step": 2275
+    },
+    {
+      "epoch": 1.5441923467660006,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.9961586512864947e-06,
+      "loss": 2.0765,
+      "step": 2280
+    },
+    {
+      "epoch": 1.5475787334913647,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.9540682368497996e-06,
+      "loss": 2.1124,
+      "step": 2285
+    },
+    {
+      "epoch": 1.5509651202167287,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.9122242678490975e-06,
+      "loss": 2.0982,
+      "step": 2290
+    },
+    {
+      "epoch": 1.5543515069420928,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.8706282078601844e-06,
+      "loss": 2.1077,
+      "step": 2295
+    },
+    {
+      "epoch": 1.5577378936674569,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.829281511787739e-06,
+      "loss": 2.1091,
+      "step": 2300
+    },
+    {
+      "epoch": 1.561124280392821,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.7881856258144557e-06,
+      "loss": 2.0761,
+      "step": 2305
+    },
+    {
+      "epoch": 1.564510667118185,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.747341987350439e-06,
+      "loss": 2.1048,
+      "step": 2310
+    },
+    {
+      "epoch": 1.567897053843549,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.7067520249829425e-06,
+      "loss": 2.0896,
+      "step": 2315
+    },
+    {
+      "epoch": 1.5712834405689131,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.666417158426393e-06,
+      "loss": 2.0974,
+      "step": 2320
+    },
+    {
+      "epoch": 1.574669827294277,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.626338798472733e-06,
+      "loss": 2.0999,
+      "step": 2325
+    },
+    {
+      "epoch": 1.578056214019641,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.586518346942082e-06,
+      "loss": 2.1019,
+      "step": 2330
+    },
+    {
+      "epoch": 1.581442600745005,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.546957196633697e-06,
+      "loss": 2.1049,
+      "step": 2335
+    },
+    {
+      "epoch": 1.5848289874703692,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.5076567312772635e-06,
+      "loss": 2.1062,
+      "step": 2340
+    },
+    {
+      "epoch": 1.588215374195733,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.4686183254844976e-06,
+      "loss": 2.0868,
+      "step": 2345
+    },
+    {
+      "epoch": 1.591601760921097,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.4298433447010495e-06,
+      "loss": 2.0984,
+      "step": 2350
+    },
+    {
+      "epoch": 1.5949881476464611,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.3913331451587663e-06,
+      "loss": 2.103,
+      "step": 2355
+    },
+    {
+      "epoch": 1.5983745343718252,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.353089073828255e-06,
+      "loss": 2.1053,
+      "step": 2360
+    },
+    {
+      "epoch": 1.6017609210971893,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.315112468371742e-06,
+      "loss": 2.0783,
+      "step": 2365
+    },
+    {
+      "epoch": 1.6051473078225533,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.2774046570963147e-06,
+      "loss": 2.0896,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6085336945479174,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.2399669589074526e-06,
+      "loss": 2.0777,
+      "step": 2375
+    },
+    {
+      "epoch": 1.6119200812732815,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.202800683262888e-06,
+      "loss": 2.1108,
+      "step": 2380
+    },
+    {
+      "epoch": 1.6153064679986455,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.165907130126821e-06,
+      "loss": 2.1043,
+      "step": 2385
+    },
+    {
+      "epoch": 1.6186928547240096,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.1292875899244237e-06,
+      "loss": 2.1204,
+      "step": 2390
+    },
+    {
+      "epoch": 1.6220792414493737,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.0929433434967472e-06,
+      "loss": 2.1216,
+      "step": 2395
+    },
+    {
+      "epoch": 1.6254656281747377,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.056875662055874e-06,
+      "loss": 2.0806,
+      "step": 2400
+    },
+    {
+      "epoch": 1.6288520149001016,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.0210858071404883e-06,
+      "loss": 2.1175,
+      "step": 2405
+    },
+    {
+      "epoch": 1.6322384016254656,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.985575030571746e-06,
+      "loss": 2.099,
+      "step": 2410
+    },
+    {
+      "epoch": 1.6356247883508297,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.950344574409471e-06,
+      "loss": 2.1146,
+      "step": 2415
+    },
+    {
+      "epoch": 1.6390111750761935,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.915395670908734e-06,
+      "loss": 2.1048,
+      "step": 2420
+    },
+    {
+      "epoch": 1.6423975618015576,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8807295424767424e-06,
+      "loss": 2.1052,
+      "step": 2425
+    },
+    {
+      "epoch": 1.6457839485269217,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8463474016300831e-06,
+      "loss": 2.1018,
+      "step": 2430
+    },
+    {
+      "epoch": 1.6491703352522857,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8122504509523154e-06,
+      "loss": 2.0773,
+      "step": 2435
+    },
+    {
+      "epoch": 1.6525567219776498,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7784398830519002e-06,
+      "loss": 2.0983,
+      "step": 2440
+    },
+    {
+      "epoch": 1.6559431087030139,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7449168805205041e-06,
+      "loss": 2.1038,
+      "step": 2445
+    },
+    {
+      "epoch": 1.659329495428378,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7116826158916212e-06,
+      "loss": 2.1109,
+      "step": 2450
+    },
+    {
+      "epoch": 1.662715882153742,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.6787382515995565e-06,
+      "loss": 2.1031,
+      "step": 2455
+    },
+    {
+      "epoch": 1.666102268879106,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.6460849399387845e-06,
+      "loss": 2.1019,
+      "step": 2460
+    },
+    {
+      "epoch": 1.6694886556044701,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6137238230236352e-06,
+      "loss": 2.0986,
+      "step": 2465
+    },
+    {
+      "epoch": 1.6728750423298342,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.5816560327483466e-06,
+      "loss": 2.1045,
+      "step": 2470
+    },
+    {
+      "epoch": 1.6762614290551983,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.549882690747475e-06,
+      "loss": 2.103,
+      "step": 2475
+    },
+    {
+      "epoch": 1.679647815780562,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5184049083566688e-06,
+      "loss": 2.0966,
+      "step": 2480
+    },
+    {
+      "epoch": 1.6830342025059262,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.487223786573786e-06,
+      "loss": 2.0874,
+      "step": 2485
+    },
+    {
+      "epoch": 1.6864205892312902,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.4563404160204008e-06,
+      "loss": 2.1299,
+      "step": 2490
+    },
+    {
+      "epoch": 1.6898069759566543,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.4257558769036327e-06,
+      "loss": 2.0836,
+      "step": 2495
+    },
+    {
+      "epoch": 1.6931933626820181,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.3954712389783998e-06,
+      "loss": 2.0968,
+      "step": 2500
+    },
+    {
+      "epoch": 1.6965797494073822,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3654875615099627e-06,
+      "loss": 2.111,
+      "step": 2505
+    },
+    {
+      "epoch": 1.6999661361327463,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.3358058932369012e-06,
+      "loss": 2.1119,
+      "step": 2510
+    },
+    {
+      "epoch": 1.7033525228581103,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.3064272723344263e-06,
+      "loss": 2.0994,
+      "step": 2515
+    },
+    {
+      "epoch": 1.7067389095834744,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.2773527263780626e-06,
+      "loss": 2.1295,
+      "step": 2520
+    },
+    {
+      "epoch": 1.7101252963088385,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.248583272307712e-06,
+      "loss": 2.0949,
+      "step": 2525
+    },
+    {
+      "epoch": 1.7135116830342025,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2201199163920829e-06,
+      "loss": 2.0893,
+      "step": 2530
+    },
+    {
+      "epoch": 1.7168980697595666,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.191963654193493e-06,
+      "loss": 2.0974,
+      "step": 2535
+    },
+    {
+      "epoch": 1.7202844564849307,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.1641154705330503e-06,
+      "loss": 2.0994,
+      "step": 2540
+    },
+    {
+      "epoch": 1.7236708432102947,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1365763394561968e-06,
+      "loss": 2.1048,
+      "step": 2545
+    },
+    {
+      "epoch": 1.7270572299356588,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.1093472241986636e-06,
+      "loss": 2.0906,
+      "step": 2550
+    },
+    {
+      "epoch": 1.7304436166610226,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0824290771527458e-06,
+      "loss": 2.0825,
+      "step": 2555
+    },
+    {
+      "epoch": 1.7338300033863867,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.0558228398340188e-06,
+      "loss": 2.0908,
+      "step": 2560
+    },
+    {
+      "epoch": 1.7372163901117508,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0295294428483993e-06,
+      "loss": 2.1215,
+      "step": 2565
+    },
+    {
+      "epoch": 1.7406027768371148,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.0035498058595838e-06,
+      "loss": 2.115,
+      "step": 2570
+    },
+    {
+      "epoch": 1.7439891635624787,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.778848375568972e-07,
+      "loss": 2.0966,
+      "step": 2575
+    },
+    {
+      "epoch": 1.7473755502878427,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.525354356235006e-07,
+      "loss": 2.0968,
+      "step": 2580
+    },
+    {
+      "epoch": 1.7507619370132068,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.275024867049975e-07,
+      "loss": 2.1054,
+      "step": 2585
+    },
+    {
+      "epoch": 1.7541483237385709,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.027868663784189e-07,
+      "loss": 2.0976,
+      "step": 2590
+    },
+    {
+      "epoch": 1.757534710463935,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.78389439121593e-07,
+      "loss": 2.1175,
+      "step": 2595
+    },
+    {
+      "epoch": 1.760921097189299,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.543110582829272e-07,
+      "loss": 2.1029,
+      "step": 2600
+    },
+    {
+      "epoch": 1.764307483914663,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.305525660515379e-07,
+      "loss": 2.095,
+      "step": 2605
+    },
+    {
+      "epoch": 1.7676938706400271,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.071147934278045e-07,
+      "loss": 2.0987,
+      "step": 2610
+    },
+    {
+      "epoch": 1.7710802573653912,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.839985601943034e-07,
+      "loss": 2.0985,
+      "step": 2615
+    },
+    {
+      "epoch": 1.7744666440907553,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 2.1066,
+      "step": 2620
+    },
+    {
+      "epoch": 1.7778530308161193,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.387339347676348e-07,
+      "loss": 2.0868,
+      "step": 2625
+    },
+    {
+      "epoch": 1.7812394175414832,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.165871257945046e-07,
+      "loss": 2.1022,
+      "step": 2630
+    },
+    {
+      "epoch": 1.7846258042668472,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.947650225963054e-07,
+      "loss": 2.1098,
+      "step": 2635
+    },
+    {
+      "epoch": 1.7880121909922113,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.732683884443736e-07,
+      "loss": 2.1205,
+      "step": 2640
+    },
+    {
+      "epoch": 1.7913985777175754,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.520979752261136e-07,
+      "loss": 2.1009,
+      "step": 2645
+    },
+    {
+      "epoch": 1.7947849644429392,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.312545234187106e-07,
+      "loss": 2.1089,
+      "step": 2650
+    },
+    {
+      "epoch": 1.7981713511683033,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.107387620632299e-07,
+      "loss": 2.0886,
+      "step": 2655
+    },
+    {
+      "epoch": 1.8015577378936674,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.905514087391051e-07,
+      "loss": 2.1024,
+      "step": 2660
+    },
+    {
+      "epoch": 1.8049441246190314,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.706931695390527e-07,
+      "loss": 2.1135,
+      "step": 2665
+    },
+    {
+      "epoch": 1.8083305113443955,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.511647390443697e-07,
+      "loss": 2.1305,
+      "step": 2670
+    },
+    {
+      "epoch": 1.8117168980697596,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.319668003006373e-07,
+      "loss": 2.1208,
+      "step": 2675
+    },
+    {
+      "epoch": 1.8151032847951236,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.131000247938367e-07,
+      "loss": 2.1184,
+      "step": 2680
+    },
+    {
+      "epoch": 1.8184896715204877,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.945650724268514e-07,
+      "loss": 2.1097,
+      "step": 2685
+    },
+    {
+      "epoch": 1.8218760582458517,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.7636259149639983e-07,
+      "loss": 2.0997,
+      "step": 2690
+    },
+    {
+      "epoch": 1.8252624449712158,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.5849321867034944e-07,
+      "loss": 2.1182,
+      "step": 2695
+    },
+    {
+      "epoch": 1.8286488316965799,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.4095757896544747e-07,
+      "loss": 2.1139,
+      "step": 2700
+    },
+    {
+      "epoch": 1.832035218421944,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.2375628572547287e-07,
+      "loss": 2.084,
+      "step": 2705
+    },
+    {
+      "epoch": 1.8354216051473078,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.0688994059976175e-07,
+      "loss": 2.0928,
+      "step": 2710
+    },
+    {
+      "epoch": 1.8388079918726719,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.903591335221857e-07,
+      "loss": 2.0848,
+      "step": 2715
+    },
+    {
+      "epoch": 1.842194378598036,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.7416444269050335e-07,
+      "loss": 2.1022,
+      "step": 2720
+    },
+    {
+      "epoch": 1.8455807653234,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.5830643454614177e-07,
+      "loss": 2.0932,
+      "step": 2725
+    },
+    {
+      "epoch": 1.8489671520487638,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.427856637543825e-07,
+      "loss": 2.1092,
+      "step": 2730
+    },
+    {
+      "epoch": 1.852353538774128,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.276026731849635e-07,
+      "loss": 2.1007,
+      "step": 2735
+    },
+    {
+      "epoch": 1.855739925499492,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.127579938930891e-07,
+      "loss": 2.0757,
+      "step": 2740
+    },
+    {
+      "epoch": 1.859126312224856,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.9825214510085575e-07,
+      "loss": 2.0995,
+      "step": 2745
+    },
+    {
+      "epoch": 1.86251269895022,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.8408563417908875e-07,
+      "loss": 2.0787,
+      "step": 2750
+    },
+    {
+      "epoch": 1.8658990856755842,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.702589566296043e-07,
+      "loss": 2.0853,
+      "step": 2755
+    },
+    {
+      "epoch": 1.8692854724009482,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.5677259606786686e-07,
+      "loss": 2.0985,
+      "step": 2760
+    },
+    {
+      "epoch": 1.8726718591263123,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.436270242060812e-07,
+      "loss": 2.0942,
+      "step": 2765
+    },
+    {
+      "epoch": 1.8760582458516764,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.3082270083669255e-07,
+      "loss": 2.1083,
+      "step": 2770
+    },
+    {
+      "epoch": 1.8794446325770404,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.1836007381630276e-07,
+      "loss": 2.105,
+      "step": 2775
+    },
+    {
+      "epoch": 1.8828310193024045,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.0623957905000603e-07,
+      "loss": 2.1042,
+      "step": 2780
+    },
+    {
+      "epoch": 1.8862174060277683,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9446164047614234e-07,
+      "loss": 2.1035,
+      "step": 2785
+    },
+    {
+      "epoch": 1.8896037927531324,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8302667005146936e-07,
+      "loss": 2.1137,
+      "step": 2790
+    },
+    {
+      "epoch": 1.8929901794784965,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7193506773675594e-07,
+      "loss": 2.0995,
+      "step": 2795
+    },
+    {
+      "epoch": 1.8963765662038605,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.6118722148278586e-07,
+      "loss": 2.0879,
+      "step": 2800
+    },
+    {
+      "epoch": 1.8997629529292244,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5078350721679625e-07,
+      "loss": 2.0781,
+      "step": 2805
+    },
+    {
+      "epoch": 1.9031493396545884,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.40724288829327e-07,
+      "loss": 2.1077,
+      "step": 2810
+    },
+    {
+      "epoch": 1.9065357263799525,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3100991816148657e-07,
+      "loss": 2.1105,
+      "step": 2815
+    },
+    {
+      "epoch": 1.9099221131053166,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2164073499265404e-07,
+      "loss": 2.1119,
+      "step": 2820
+    },
+    {
+      "epoch": 1.9133084998306806,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.1261706702859198e-07,
+      "loss": 2.0975,
+      "step": 2825
+    },
+    {
+      "epoch": 1.9166948865560447,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.0393922988997995e-07,
+      "loss": 2.1107,
+      "step": 2830
+    },
+    {
+      "epoch": 1.9200812732814088,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.560752710138343e-08,
+      "loss": 2.119,
+      "step": 2835
+    },
+    {
+      "epoch": 1.9234676600067728,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.762225008062675e-08,
+      "loss": 2.1086,
+      "step": 2840
+    },
+    {
+      "epoch": 1.926854046732137,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.998367812861007e-08,
+      "loss": 2.0987,
+      "step": 2845
+    },
+    {
+      "epoch": 1.930240433457501,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.269207841953618e-08,
+      "loss": 2.0904,
+      "step": 2850
+    },
+    {
+      "epoch": 1.933626820182865,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.574770599156344e-08,
+      "loss": 2.1,
+      "step": 2855
+    },
+    {
+      "epoch": 1.9370132069082289,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.915080373788962e-08,
+      "loss": 2.087,
+      "step": 2860
+    },
+    {
+      "epoch": 1.940399593633593,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.290160239825315e-08,
+      "loss": 2.0982,
+      "step": 2865
+    },
+    {
+      "epoch": 1.943785980358957,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.700032055086179e-08,
+      "loss": 2.0946,
+      "step": 2870
+    },
+    {
+      "epoch": 1.947172367084321,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.1447164604753175e-08,
+      "loss": 2.1131,
+      "step": 2875
+    },
+    {
+      "epoch": 1.950558753809685,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.6242328792567286e-08,
+      "loss": 2.0808,
+      "step": 2880
+    },
+    {
+      "epoch": 1.953945140535049,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.138599516375962e-08,
+      "loss": 2.0961,
+      "step": 2885
+    },
+    {
+      "epoch": 1.957331527260413,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.6878333578227446e-08,
+      "loss": 2.113,
+      "step": 2890
+    },
+    {
+      "epoch": 1.960717913985777,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.2719501700374514e-08,
+      "loss": 2.0924,
+      "step": 2895
+    },
+    {
+      "epoch": 1.9641043007111412,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.890964499359327e-08,
+      "loss": 2.094,
+      "step": 2900
+    },
+    {
+      "epoch": 1.9674906874365052,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.5448896715177798e-08,
+      "loss": 2.1074,
+      "step": 2905
+    },
+    {
+      "epoch": 1.9708770741618693,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2337377911660897e-08,
+      "loss": 2.0797,
+      "step": 2910
+    },
+    {
+      "epoch": 1.9742634608872334,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.575197414583015e-09,
+      "loss": 2.1091,
+      "step": 2915
+    },
+    {
+      "epoch": 1.9776498476125974,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.162451836685291e-09,
+      "loss": 2.109,
+      "step": 2920
+    },
+    {
+      "epoch": 1.9810362343379615,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.0992255685289315e-09,
+      "loss": 2.0992,
+      "step": 2925
+    },
+    {
+      "epoch": 1.9844226210633256,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.385590775544234e-09,
+      "loss": 2.0804,
+      "step": 2930
+    },
+    {
+      "epoch": 1.9878090077886896,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.0216073955070527e-09,
+      "loss": 2.1013,
+      "step": 2935
+    },
+    {
+      "epoch": 1.9911953945140535,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.007323136438254e-09,
+      "loss": 2.0967,
+      "step": 2940
+    },
+    {
+      "epoch": 1.9945817812394175,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.42773474943936e-10,
+      "loss": 2.1021,
+      "step": 2945
+    },
+    {
+      "epoch": 1.9979681679647816,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.7981654966424686e-11,
+      "loss": 2.0984,
+      "step": 2950
+    },
+    {
+      "epoch": 1.9993227226549273,
+      "eval_loss": 2.1231415271759033,
+      "eval_runtime": 85.7269,
+      "eval_samples_per_second": 15.223,
+      "eval_steps_per_second": 1.913,
+      "step": 2952
+    },
+    {
+      "epoch": 1.9993227226549273,
+      "step": 2952,
+      "total_flos": 6.003537882466222e+17,
+      "train_loss": 2.1295973070591767,
+      "train_runtime": 13069.623,
+      "train_samples_per_second": 3.614,
+      "train_steps_per_second": 0.226
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2952,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.003537882466222e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}