{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 4504,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017761989342806393,
      "grad_norm": 2.08616042137146,
      "learning_rate": 4.9900088809946715e-05,
      "loss": 5.9583,
      "step": 10
    },
    {
      "epoch": 0.035523978685612786,
      "grad_norm": 2.0548300743103027,
      "learning_rate": 4.978907637655418e-05,
      "loss": 3.0383,
      "step": 20
    },
    {
      "epoch": 0.05328596802841918,
      "grad_norm": 1.3866031169891357,
      "learning_rate": 4.967806394316163e-05,
      "loss": 2.0479,
      "step": 30
    },
    {
      "epoch": 0.07104795737122557,
      "grad_norm": 1.5728237628936768,
      "learning_rate": 4.9567051509769094e-05,
      "loss": 1.3301,
      "step": 40
    },
    {
      "epoch": 0.08880994671403197,
      "grad_norm": 1.3490098714828491,
      "learning_rate": 4.9456039076376556e-05,
      "loss": 0.8203,
      "step": 50
    },
    {
      "epoch": 0.10657193605683836,
      "grad_norm": 1.1025753021240234,
      "learning_rate": 4.934502664298402e-05,
      "loss": 0.5509,
      "step": 60
    },
    {
      "epoch": 0.12433392539964476,
      "grad_norm": 1.8544913530349731,
      "learning_rate": 4.923401420959148e-05,
      "loss": 0.3441,
      "step": 70
    },
    {
      "epoch": 0.14209591474245115,
      "grad_norm": 0.7912121415138245,
      "learning_rate": 4.9123001776198936e-05,
      "loss": 0.2242,
      "step": 80
    },
    {
      "epoch": 0.15985790408525755,
      "grad_norm": 0.6116934418678284,
      "learning_rate": 4.90119893428064e-05,
      "loss": 0.1493,
      "step": 90
    },
    {
      "epoch": 0.17761989342806395,
      "grad_norm": 0.627920925617218,
      "learning_rate": 4.890097690941386e-05,
      "loss": 0.123,
      "step": 100
    },
    {
      "epoch": 0.19538188277087035,
      "grad_norm": 0.5851693153381348,
      "learning_rate": 4.8789964476021315e-05,
      "loss": 0.0908,
      "step": 110
    },
    {
      "epoch": 0.21314387211367672,
      "grad_norm": 0.4242522716522217,
      "learning_rate": 4.867895204262878e-05,
      "loss": 0.0668,
      "step": 120
    },
    {
      "epoch": 0.23090586145648312,
      "grad_norm": 0.38480210304260254,
      "learning_rate": 4.856793960923623e-05,
      "loss": 0.0498,
      "step": 130
    },
    {
      "epoch": 0.24866785079928952,
      "grad_norm": 0.4704955816268921,
      "learning_rate": 4.8456927175843695e-05,
      "loss": 0.0478,
      "step": 140
    },
    {
      "epoch": 0.2664298401420959,
      "grad_norm": 0.33954229950904846,
      "learning_rate": 4.834591474245116e-05,
      "loss": 0.0383,
      "step": 150
    },
    {
      "epoch": 0.2841918294849023,
      "grad_norm": 0.3331491947174072,
      "learning_rate": 4.823490230905862e-05,
      "loss": 0.0334,
      "step": 160
    },
    {
      "epoch": 0.3019538188277087,
      "grad_norm": 0.22477354109287262,
      "learning_rate": 4.812388987566608e-05,
      "loss": 0.0285,
      "step": 170
    },
    {
      "epoch": 0.3197158081705151,
      "grad_norm": 0.32435140013694763,
      "learning_rate": 4.8012877442273537e-05,
      "loss": 0.0262,
      "step": 180
    },
    {
      "epoch": 0.33747779751332146,
      "grad_norm": 0.2808849513530731,
      "learning_rate": 4.7901865008881e-05,
      "loss": 0.0234,
      "step": 190
    },
    {
      "epoch": 0.3552397868561279,
      "grad_norm": 0.24697023630142212,
      "learning_rate": 4.779085257548846e-05,
      "loss": 0.0214,
      "step": 200
    },
    {
      "epoch": 0.37300177619893427,
      "grad_norm": 0.1854216456413269,
      "learning_rate": 4.7679840142095916e-05,
      "loss": 0.0188,
      "step": 210
    },
    {
      "epoch": 0.3907637655417407,
      "grad_norm": 0.1791575402021408,
      "learning_rate": 4.756882770870338e-05,
      "loss": 0.0187,
      "step": 220
    },
    {
      "epoch": 0.40852575488454707,
      "grad_norm": 0.2226918488740921,
      "learning_rate": 4.7457815275310833e-05,
      "loss": 0.0169,
      "step": 230
    },
    {
      "epoch": 0.42628774422735344,
      "grad_norm": 0.2505394518375397,
      "learning_rate": 4.7346802841918296e-05,
      "loss": 0.017,
      "step": 240
    },
    {
      "epoch": 0.44404973357015987,
      "grad_norm": 0.1905129998922348,
      "learning_rate": 4.723579040852576e-05,
      "loss": 0.0144,
      "step": 250
    },
    {
      "epoch": 0.46181172291296624,
      "grad_norm": 0.22495870292186737,
      "learning_rate": 4.712477797513321e-05,
      "loss": 0.0151,
      "step": 260
    },
    {
      "epoch": 0.47957371225577267,
      "grad_norm": 0.17531022429466248,
      "learning_rate": 4.701376554174068e-05,
      "loss": 0.0113,
      "step": 270
    },
    {
      "epoch": 0.49733570159857904,
      "grad_norm": 0.18769726157188416,
      "learning_rate": 4.690275310834814e-05,
      "loss": 0.0115,
      "step": 280
    },
    {
      "epoch": 0.5150976909413855,
      "grad_norm": 0.18548175692558289,
      "learning_rate": 4.67917406749556e-05,
      "loss": 0.0106,
      "step": 290
    },
    {
      "epoch": 0.5328596802841918,
      "grad_norm": 0.19990290701389313,
      "learning_rate": 4.668072824156306e-05,
      "loss": 0.0102,
      "step": 300
    },
    {
      "epoch": 0.5506216696269982,
      "grad_norm": 0.12933073937892914,
      "learning_rate": 4.656971580817052e-05,
      "loss": 0.0092,
      "step": 310
    },
    {
      "epoch": 0.5683836589698046,
      "grad_norm": 0.1915041208267212,
      "learning_rate": 4.645870337477798e-05,
      "loss": 0.009,
      "step": 320
    },
    {
      "epoch": 0.5861456483126111,
      "grad_norm": 0.1646175980567932,
      "learning_rate": 4.6347690941385434e-05,
      "loss": 0.0089,
      "step": 330
    },
    {
      "epoch": 0.6039076376554174,
      "grad_norm": 0.13864779472351074,
      "learning_rate": 4.6236678507992896e-05,
      "loss": 0.0078,
      "step": 340
    },
    {
      "epoch": 0.6216696269982238,
      "grad_norm": 0.20968131721019745,
      "learning_rate": 4.612566607460036e-05,
      "loss": 0.0084,
      "step": 350
    },
    {
      "epoch": 0.6394316163410302,
      "grad_norm": 0.1403323858976364,
      "learning_rate": 4.6014653641207814e-05,
      "loss": 0.0067,
      "step": 360
    },
    {
      "epoch": 0.6571936056838366,
      "grad_norm": 0.1056448444724083,
      "learning_rate": 4.5903641207815276e-05,
      "loss": 0.007,
      "step": 370
    },
    {
      "epoch": 0.6749555950266429,
      "grad_norm": 0.09233298897743225,
      "learning_rate": 4.579262877442274e-05,
      "loss": 0.0063,
      "step": 380
    },
    {
      "epoch": 0.6927175843694494,
      "grad_norm": 0.2610642910003662,
      "learning_rate": 4.56816163410302e-05,
      "loss": 0.0061,
      "step": 390
    },
    {
      "epoch": 0.7104795737122558,
      "grad_norm": 0.13106778264045715,
      "learning_rate": 4.557060390763766e-05,
      "loss": 0.0061,
      "step": 400
    },
    {
      "epoch": 0.7282415630550622,
      "grad_norm": 0.11065071076154709,
      "learning_rate": 4.545959147424512e-05,
      "loss": 0.0053,
      "step": 410
    },
    {
      "epoch": 0.7460035523978685,
      "grad_norm": 0.17179323732852936,
      "learning_rate": 4.534857904085258e-05,
      "loss": 0.0064,
      "step": 420
    },
    {
      "epoch": 0.7637655417406749,
      "grad_norm": 0.19529324769973755,
      "learning_rate": 4.5237566607460035e-05,
      "loss": 0.0045,
      "step": 430
    },
    {
      "epoch": 0.7815275310834814,
      "grad_norm": 0.20533151924610138,
      "learning_rate": 4.51265541740675e-05,
      "loss": 0.005,
      "step": 440
    },
    {
      "epoch": 0.7992895204262878,
      "grad_norm": 0.16531281173229218,
      "learning_rate": 4.501554174067496e-05,
      "loss": 0.0041,
      "step": 450
    },
    {
      "epoch": 0.8170515097690941,
      "grad_norm": 0.13664154708385468,
      "learning_rate": 4.4904529307282414e-05,
      "loss": 0.004,
      "step": 460
    },
    {
      "epoch": 0.8348134991119005,
      "grad_norm": 0.11034122854471207,
      "learning_rate": 4.4793516873889876e-05,
      "loss": 0.0039,
      "step": 470
    },
    {
      "epoch": 0.8525754884547069,
      "grad_norm": 0.06444460153579712,
      "learning_rate": 4.468250444049734e-05,
      "loss": 0.0039,
      "step": 480
    },
    {
      "epoch": 0.8703374777975134,
      "grad_norm": 0.2418302446603775,
      "learning_rate": 4.45714920071048e-05,
      "loss": 0.0039,
      "step": 490
    },
    {
      "epoch": 0.8880994671403197,
      "grad_norm": 0.06368648260831833,
      "learning_rate": 4.446047957371226e-05,
      "loss": 0.0035,
      "step": 500
    },
    {
      "epoch": 0.9058614564831261,
      "grad_norm": 0.08606679737567902,
      "learning_rate": 4.434946714031972e-05,
      "loss": 0.0033,
      "step": 510
    },
    {
      "epoch": 0.9236234458259325,
      "grad_norm": 0.06212198734283447,
      "learning_rate": 4.423845470692718e-05,
      "loss": 0.0032,
      "step": 520
    },
    {
      "epoch": 0.9413854351687388,
      "grad_norm": 0.05146817862987518,
      "learning_rate": 4.4127442273534635e-05,
      "loss": 0.0036,
      "step": 530
    },
    {
      "epoch": 0.9591474245115453,
      "grad_norm": 0.10226184129714966,
      "learning_rate": 4.40164298401421e-05,
      "loss": 0.0035,
      "step": 540
    },
    {
      "epoch": 0.9769094138543517,
      "grad_norm": 0.06877221167087555,
      "learning_rate": 4.390541740674956e-05,
      "loss": 0.0034,
      "step": 550
    },
    {
      "epoch": 0.9946714031971581,
      "grad_norm": 0.08828417211771011,
      "learning_rate": 4.3794404973357015e-05,
      "loss": 0.0026,
      "step": 560
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.458987339399755e-05,
      "eval_runtime": 22.2952,
      "eval_samples_per_second": 44.853,
      "eval_steps_per_second": 2.826,
      "step": 563
    },
    {
      "epoch": 1.0124333925399644,
      "grad_norm": 0.09242795407772064,
      "learning_rate": 4.368339253996448e-05,
      "loss": 0.003,
      "step": 570
    },
    {
      "epoch": 1.030195381882771,
      "grad_norm": 0.0889139473438263,
      "learning_rate": 4.357238010657194e-05,
      "loss": 0.0028,
      "step": 580
    },
    {
      "epoch": 1.0479573712255772,
      "grad_norm": 0.12308719754219055,
      "learning_rate": 4.34613676731794e-05,
      "loss": 0.0027,
      "step": 590
    },
    {
      "epoch": 1.0657193605683837,
      "grad_norm": 0.05686354264616966,
      "learning_rate": 4.335035523978686e-05,
      "loss": 0.0024,
      "step": 600
    },
    {
      "epoch": 1.0834813499111902,
      "grad_norm": 0.05174636095762253,
      "learning_rate": 4.323934280639432e-05,
      "loss": 0.0023,
      "step": 610
    },
    {
      "epoch": 1.1012433392539964,
      "grad_norm": 0.1351872682571411,
      "learning_rate": 4.312833037300178e-05,
      "loss": 0.0024,
      "step": 620
    },
    {
      "epoch": 1.119005328596803,
      "grad_norm": 0.11029157787561417,
      "learning_rate": 4.3017317939609236e-05,
      "loss": 0.0026,
      "step": 630
    },
    {
      "epoch": 1.1367673179396092,
      "grad_norm": 0.041344061493873596,
      "learning_rate": 4.29063055062167e-05,
      "loss": 0.0018,
      "step": 640
    },
    {
      "epoch": 1.1545293072824157,
      "grad_norm": 0.15868201851844788,
      "learning_rate": 4.279529307282416e-05,
      "loss": 0.0023,
      "step": 650
    },
    {
      "epoch": 1.1722912966252221,
      "grad_norm": 0.06435976177453995,
      "learning_rate": 4.2684280639431615e-05,
      "loss": 0.0022,
      "step": 660
    },
    {
      "epoch": 1.1900532859680284,
      "grad_norm": 0.05187324061989784,
      "learning_rate": 4.257326820603908e-05,
      "loss": 0.0018,
      "step": 670
    },
    {
      "epoch": 1.2078152753108349,
      "grad_norm": 0.041784290224313736,
      "learning_rate": 4.246225577264654e-05,
      "loss": 0.0023,
      "step": 680
    },
    {
      "epoch": 1.2255772646536411,
      "grad_norm": 0.031757429242134094,
      "learning_rate": 4.2351243339254e-05,
      "loss": 0.0021,
      "step": 690
    },
    {
      "epoch": 1.2433392539964476,
      "grad_norm": 0.044060107320547104,
      "learning_rate": 4.2240230905861464e-05,
      "loss": 0.0019,
      "step": 700
    },
    {
      "epoch": 1.261101243339254,
      "grad_norm": 0.11238252371549606,
      "learning_rate": 4.212921847246892e-05,
      "loss": 0.002,
      "step": 710
    },
    {
      "epoch": 1.2788632326820604,
      "grad_norm": 0.041686106473207474,
      "learning_rate": 4.201820603907638e-05,
      "loss": 0.0015,
      "step": 720
    },
    {
      "epoch": 1.2966252220248669,
      "grad_norm": 0.04672028496861458,
      "learning_rate": 4.1907193605683836e-05,
      "loss": 0.0015,
      "step": 730
    },
    {
      "epoch": 1.3143872113676731,
      "grad_norm": 0.0358208566904068,
      "learning_rate": 4.17961811722913e-05,
      "loss": 0.0016,
      "step": 740
    },
    {
      "epoch": 1.3321492007104796,
      "grad_norm": 0.17647922039031982,
      "learning_rate": 4.168516873889876e-05,
      "loss": 0.0015,
      "step": 750
    },
    {
      "epoch": 1.349911190053286,
      "grad_norm": 0.10813446342945099,
      "learning_rate": 4.1574156305506216e-05,
      "loss": 0.0017,
      "step": 760
    },
    {
      "epoch": 1.3676731793960923,
      "grad_norm": 0.03748601675033569,
      "learning_rate": 4.146314387211368e-05,
      "loss": 0.0018,
      "step": 770
    },
    {
      "epoch": 1.3854351687388988,
      "grad_norm": 0.031867001205682755,
      "learning_rate": 4.135213143872114e-05,
      "loss": 0.0016,
      "step": 780
    },
    {
      "epoch": 1.403197158081705,
      "grad_norm": 0.01816353015601635,
      "learning_rate": 4.12411190053286e-05,
      "loss": 0.0011,
      "step": 790
    },
    {
      "epoch": 1.4209591474245116,
      "grad_norm": 0.05043243616819382,
      "learning_rate": 4.1130106571936064e-05,
      "loss": 0.0014,
      "step": 800
    },
    {
      "epoch": 1.438721136767318,
      "grad_norm": 0.27319490909576416,
      "learning_rate": 4.101909413854352e-05,
      "loss": 0.0017,
      "step": 810
    },
    {
      "epoch": 1.4564831261101243,
      "grad_norm": 0.04838378727436066,
      "learning_rate": 4.090808170515098e-05,
      "loss": 0.0014,
      "step": 820
    },
    {
      "epoch": 1.4742451154529308,
      "grad_norm": 0.05813241004943848,
      "learning_rate": 4.079706927175844e-05,
      "loss": 0.0017,
      "step": 830
    },
    {
      "epoch": 1.492007104795737,
      "grad_norm": 0.06169930100440979,
      "learning_rate": 4.06860568383659e-05,
      "loss": 0.0015,
      "step": 840
    },
    {
      "epoch": 1.5097690941385435,
      "grad_norm": 0.01758798211812973,
      "learning_rate": 4.057504440497336e-05,
      "loss": 0.0012,
      "step": 850
    },
    {
      "epoch": 1.52753108348135,
      "grad_norm": 0.07379154115915298,
      "learning_rate": 4.0464031971580816e-05,
      "loss": 0.0013,
      "step": 860
    },
    {
      "epoch": 1.5452930728241563,
      "grad_norm": 0.1645466685295105,
      "learning_rate": 4.035301953818828e-05,
      "loss": 0.0015,
      "step": 870
    },
    {
      "epoch": 1.5630550621669625,
      "grad_norm": 0.023952888324856758,
      "learning_rate": 4.0242007104795734e-05,
      "loss": 0.0011,
      "step": 880
    },
    {
      "epoch": 1.580817051509769,
      "grad_norm": 0.012347780168056488,
      "learning_rate": 4.0130994671403196e-05,
      "loss": 0.0011,
      "step": 890
    },
    {
      "epoch": 1.5985790408525755,
      "grad_norm": 0.12134626507759094,
      "learning_rate": 4.0019982238010665e-05,
      "loss": 0.0013,
      "step": 900
    },
    {
      "epoch": 1.616341030195382,
      "grad_norm": 0.035019490867853165,
      "learning_rate": 3.990896980461812e-05,
      "loss": 0.001,
      "step": 910
    },
    {
      "epoch": 1.6341030195381883,
      "grad_norm": 0.03914281353354454,
      "learning_rate": 3.979795737122558e-05,
      "loss": 0.001,
      "step": 920
    },
    {
      "epoch": 1.6518650088809945,
      "grad_norm": 0.01977422647178173,
      "learning_rate": 3.968694493783304e-05,
      "loss": 0.0012,
      "step": 930
    },
    {
      "epoch": 1.669626998223801,
      "grad_norm": 0.018485285341739655,
      "learning_rate": 3.95759325044405e-05,
      "loss": 0.0012,
      "step": 940
    },
    {
      "epoch": 1.6873889875666075,
      "grad_norm": 0.2277858406305313,
      "learning_rate": 3.946492007104796e-05,
      "loss": 0.0011,
      "step": 950
    },
    {
      "epoch": 1.705150976909414,
      "grad_norm": 0.012234285473823547,
      "learning_rate": 3.935390763765542e-05,
      "loss": 0.0011,
      "step": 960
    },
    {
      "epoch": 1.7229129662522202,
      "grad_norm": 0.021199282258749008,
      "learning_rate": 3.924289520426288e-05,
      "loss": 0.001,
      "step": 970
    },
    {
      "epoch": 1.7406749555950265,
      "grad_norm": 0.12197419255971909,
      "learning_rate": 3.9131882770870334e-05,
      "loss": 0.0011,
      "step": 980
    },
    {
      "epoch": 1.758436944937833,
      "grad_norm": 0.03652465343475342,
      "learning_rate": 3.9020870337477797e-05,
      "loss": 0.0009,
      "step": 990
    },
    {
      "epoch": 1.7761989342806395,
      "grad_norm": 0.03332207724452019,
      "learning_rate": 3.890985790408526e-05,
      "loss": 0.001,
      "step": 1000
    },
    {
      "epoch": 1.793960923623446,
      "grad_norm": 0.05201757699251175,
      "learning_rate": 3.879884547069272e-05,
      "loss": 0.001,
      "step": 1010
    },
    {
      "epoch": 1.8117229129662522,
      "grad_norm": 0.01703731343150139,
      "learning_rate": 3.868783303730018e-05,
      "loss": 0.001,
      "step": 1020
    },
    {
      "epoch": 1.8294849023090585,
      "grad_norm": 0.027388030663132668,
      "learning_rate": 3.857682060390764e-05,
      "loss": 0.0008,
      "step": 1030
    },
    {
      "epoch": 1.847246891651865,
      "grad_norm": 0.12894806265830994,
      "learning_rate": 3.84658081705151e-05,
      "loss": 0.0009,
      "step": 1040
    },
    {
      "epoch": 1.8650088809946714,
      "grad_norm": 0.030342355370521545,
      "learning_rate": 3.835479573712256e-05,
      "loss": 0.001,
      "step": 1050
    },
    {
      "epoch": 1.882770870337478,
      "grad_norm": 0.043612588196992874,
      "learning_rate": 3.824378330373002e-05,
      "loss": 0.0011,
      "step": 1060
    },
    {
      "epoch": 1.9005328596802842,
      "grad_norm": 0.015847304835915565,
      "learning_rate": 3.813277087033748e-05,
      "loss": 0.0009,
      "step": 1070
    },
    {
      "epoch": 1.9182948490230904,
      "grad_norm": 0.036328867077827454,
      "learning_rate": 3.8021758436944935e-05,
      "loss": 0.0009,
      "step": 1080
    },
    {
      "epoch": 1.936056838365897,
      "grad_norm": 0.013609735295176506,
      "learning_rate": 3.79107460035524e-05,
      "loss": 0.0011,
      "step": 1090
    },
    {
      "epoch": 1.9538188277087034,
      "grad_norm": 0.016604352742433548,
      "learning_rate": 3.779973357015986e-05,
      "loss": 0.0009,
      "step": 1100
    },
    {
      "epoch": 1.97158081705151,
      "grad_norm": 0.030676284804940224,
      "learning_rate": 3.768872113676732e-05,
      "loss": 0.0009,
      "step": 1110
    },
    {
      "epoch": 1.9893428063943162,
      "grad_norm": 0.010764030739665031,
      "learning_rate": 3.7577708703374783e-05,
      "loss": 0.0007,
      "step": 1120
    },
    {
      "epoch": 2.0,
      "eval_loss": 7.020198609097861e-06,
      "eval_runtime": 22.303,
      "eval_samples_per_second": 44.837,
      "eval_steps_per_second": 2.825,
      "step": 1126
    },
    {
      "epoch": 2.0071047957371224,
      "grad_norm": 0.015416032634675503,
      "learning_rate": 3.746669626998224e-05,
      "loss": 0.0008,
      "step": 1130
    },
    {
      "epoch": 2.024866785079929,
      "grad_norm": 0.027968090027570724,
      "learning_rate": 3.73556838365897e-05,
      "loss": 0.0007,
      "step": 1140
    },
    {
      "epoch": 2.0426287744227354,
      "grad_norm": 0.13390642404556274,
      "learning_rate": 3.724467140319716e-05,
      "loss": 0.001,
      "step": 1150
    },
    {
      "epoch": 2.060390763765542,
      "grad_norm": 0.02235797978937626,
      "learning_rate": 3.713365896980462e-05,
      "loss": 0.0007,
      "step": 1160
    },
    {
      "epoch": 2.0781527531083483,
      "grad_norm": 0.018844326958060265,
      "learning_rate": 3.702264653641208e-05,
      "loss": 0.0009,
      "step": 1170
    },
    {
      "epoch": 2.0959147424511544,
      "grad_norm": 0.04496992379426956,
      "learning_rate": 3.6911634103019536e-05,
      "loss": 0.0011,
      "step": 1180
    },
    {
      "epoch": 2.113676731793961,
      "grad_norm": 0.038956914097070694,
      "learning_rate": 3.6800621669627e-05,
      "loss": 0.0008,
      "step": 1190
    },
    {
      "epoch": 2.1314387211367674,
      "grad_norm": 0.03798123076558113,
      "learning_rate": 3.668960923623446e-05,
      "loss": 0.0007,
      "step": 1200
    },
    {
      "epoch": 2.149200710479574,
      "grad_norm": 0.02100321650505066,
      "learning_rate": 3.657859680284192e-05,
      "loss": 0.0006,
      "step": 1210
    },
    {
      "epoch": 2.1669626998223803,
      "grad_norm": 0.021854564547538757,
      "learning_rate": 3.6467584369449384e-05,
      "loss": 0.0008,
      "step": 1220
    },
    {
      "epoch": 2.1847246891651864,
      "grad_norm": 0.011918751522898674,
      "learning_rate": 3.635657193605684e-05,
      "loss": 0.0008,
      "step": 1230
    },
    {
      "epoch": 2.202486678507993,
      "grad_norm": 0.06602652370929718,
      "learning_rate": 3.62455595026643e-05,
      "loss": 0.0006,
      "step": 1240
    },
    {
      "epoch": 2.2202486678507993,
      "grad_norm": 0.007478422950953245,
      "learning_rate": 3.6134547069271763e-05,
      "loss": 0.0006,
      "step": 1250
    },
    {
      "epoch": 2.238010657193606,
      "grad_norm": 0.02918918989598751,
      "learning_rate": 3.602353463587922e-05,
      "loss": 0.0007,
      "step": 1260
    },
    {
      "epoch": 2.2557726465364123,
      "grad_norm": 0.05313832312822342,
      "learning_rate": 3.591252220248668e-05,
      "loss": 0.0008,
      "step": 1270
    },
    {
      "epoch": 2.2735346358792183,
      "grad_norm": 0.04284721240401268,
      "learning_rate": 3.5801509769094136e-05,
      "loss": 0.0009,
      "step": 1280
    },
    {
      "epoch": 2.291296625222025,
      "grad_norm": 0.027897143736481667,
      "learning_rate": 3.56904973357016e-05,
      "loss": 0.0008,
      "step": 1290
    },
    {
      "epoch": 2.3090586145648313,
      "grad_norm": 0.009780477732419968,
      "learning_rate": 3.557948490230906e-05,
      "loss": 0.0006,
      "step": 1300
    },
    {
      "epoch": 2.326820603907638,
      "grad_norm": 0.017684033140540123,
      "learning_rate": 3.546847246891652e-05,
      "loss": 0.0008,
      "step": 1310
    },
    {
      "epoch": 2.3445825932504443,
      "grad_norm": 0.07929302006959915,
      "learning_rate": 3.5357460035523985e-05,
      "loss": 0.0006,
      "step": 1320
    },
    {
      "epoch": 2.3623445825932503,
      "grad_norm": 0.06948840618133545,
      "learning_rate": 3.524644760213144e-05,
      "loss": 0.0006,
      "step": 1330
    },
    {
      "epoch": 2.380106571936057,
      "grad_norm": 0.006130309775471687,
      "learning_rate": 3.51354351687389e-05,
      "loss": 0.0005,
      "step": 1340
    },
    {
      "epoch": 2.3978685612788633,
      "grad_norm": 0.022212188690900803,
      "learning_rate": 3.5024422735346364e-05,
      "loss": 0.0006,
      "step": 1350
    },
    {
      "epoch": 2.4156305506216698,
      "grad_norm": 0.015669459477066994,
      "learning_rate": 3.491341030195382e-05,
      "loss": 0.0006,
      "step": 1360
    },
    {
      "epoch": 2.4333925399644762,
      "grad_norm": 0.026899907737970352,
      "learning_rate": 3.480239786856128e-05,
      "loss": 0.0005,
      "step": 1370
    },
    {
      "epoch": 2.4511545293072823,
      "grad_norm": 0.02007845975458622,
      "learning_rate": 3.469138543516874e-05,
      "loss": 0.0006,
      "step": 1380
    },
    {
      "epoch": 2.4689165186500888,
      "grad_norm": 0.017982805147767067,
      "learning_rate": 3.45803730017762e-05,
      "loss": 0.0005,
      "step": 1390
    },
    {
      "epoch": 2.4866785079928952,
      "grad_norm": 0.05845437943935394,
      "learning_rate": 3.446936056838366e-05,
      "loss": 0.0007,
      "step": 1400
    },
    {
      "epoch": 2.5044404973357017,
      "grad_norm": 0.006352486554533243,
      "learning_rate": 3.4358348134991116e-05,
      "loss": 0.0006,
      "step": 1410
    },
    {
      "epoch": 2.522202486678508,
      "grad_norm": 0.004761823453009129,
      "learning_rate": 3.4247335701598585e-05,
      "loss": 0.0006,
      "step": 1420
    },
    {
      "epoch": 2.5399644760213143,
      "grad_norm": 0.0643671378493309,
      "learning_rate": 3.413632326820604e-05,
      "loss": 0.0005,
      "step": 1430
    },
    {
      "epoch": 2.5577264653641207,
      "grad_norm": 0.2012147605419159,
      "learning_rate": 3.40253108348135e-05,
      "loss": 0.0006,
      "step": 1440
    },
    {
      "epoch": 2.575488454706927,
      "grad_norm": 0.005158168729394674,
      "learning_rate": 3.3914298401420965e-05,
      "loss": 0.0007,
      "step": 1450
    },
    {
      "epoch": 2.5932504440497337,
      "grad_norm": 0.02005215547978878,
      "learning_rate": 3.380328596802842e-05,
      "loss": 0.0005,
      "step": 1460
    },
    {
      "epoch": 2.61101243339254,
      "grad_norm": 0.011687914840877056,
      "learning_rate": 3.369227353463588e-05,
      "loss": 0.0005,
      "step": 1470
    },
    {
      "epoch": 2.6287744227353462,
      "grad_norm": 0.028902605175971985,
      "learning_rate": 3.358126110124334e-05,
      "loss": 0.0007,
      "step": 1480
    },
    {
      "epoch": 2.6465364120781527,
      "grad_norm": 0.004095068667083979,
      "learning_rate": 3.34702486678508e-05,
      "loss": 0.0005,
      "step": 1490
    },
    {
      "epoch": 2.664298401420959,
      "grad_norm": 0.033792633563280106,
      "learning_rate": 3.335923623445826e-05,
      "loss": 0.0005,
      "step": 1500
    },
    {
      "epoch": 2.6820603907637657,
      "grad_norm": 0.010024248622357845,
      "learning_rate": 3.324822380106572e-05,
      "loss": 0.0007,
      "step": 1510
    },
    {
      "epoch": 2.699822380106572,
      "grad_norm": 0.04082302376627922,
      "learning_rate": 3.313721136767318e-05,
      "loss": 0.0005,
      "step": 1520
    },
    {
      "epoch": 2.717584369449378,
      "grad_norm": 0.01643848419189453,
      "learning_rate": 3.302619893428064e-05,
      "loss": 0.0006,
      "step": 1530
    },
    {
      "epoch": 2.7353463587921847,
      "grad_norm": 0.02243112586438656,
      "learning_rate": 3.29151865008881e-05,
      "loss": 0.0005,
      "step": 1540
    },
    {
      "epoch": 2.753108348134991,
      "grad_norm": 0.015440504066646099,
      "learning_rate": 3.2804174067495565e-05,
      "loss": 0.0005,
      "step": 1550
    },
    {
      "epoch": 2.7708703374777977,
      "grad_norm": 0.04903204366564751,
      "learning_rate": 3.269316163410302e-05,
      "loss": 0.0006,
      "step": 1560
    },
    {
      "epoch": 2.788632326820604,
      "grad_norm": 0.04786301776766777,
      "learning_rate": 3.258214920071048e-05,
      "loss": 0.0006,
      "step": 1570
    },
    {
      "epoch": 2.80639431616341,
      "grad_norm": 0.02142539992928505,
      "learning_rate": 3.247113676731794e-05,
      "loss": 0.0008,
      "step": 1580
    },
    {
      "epoch": 2.8241563055062167,
      "grad_norm": 0.005326330661773682,
      "learning_rate": 3.23601243339254e-05,
      "loss": 0.0007,
      "step": 1590
    },
    {
      "epoch": 2.841918294849023,
      "grad_norm": 0.016222188249230385,
      "learning_rate": 3.224911190053286e-05,
      "loss": 0.0005,
      "step": 1600
    },
    {
      "epoch": 2.8596802841918296,
      "grad_norm": 0.005499828606843948,
      "learning_rate": 3.213809946714032e-05,
      "loss": 0.0005,
      "step": 1610
    },
    {
      "epoch": 2.877442273534636,
      "grad_norm": 0.02295517548918724,
      "learning_rate": 3.202708703374778e-05,
      "loss": 0.0005,
      "step": 1620
    },
    {
      "epoch": 2.895204262877442,
      "grad_norm": 0.12497241050004959,
      "learning_rate": 3.191607460035524e-05,
      "loss": 0.0005,
      "step": 1630
    },
    {
      "epoch": 2.9129662522202486,
      "grad_norm": 0.04707463085651398,
      "learning_rate": 3.1805062166962704e-05,
      "loss": 0.0005,
      "step": 1640
    },
    {
      "epoch": 2.930728241563055,
      "grad_norm": 0.030674880370497704,
      "learning_rate": 3.1694049733570166e-05,
      "loss": 0.0008,
      "step": 1650
    },
    {
      "epoch": 2.9484902309058616,
      "grad_norm": 0.1597021222114563,
      "learning_rate": 3.158303730017762e-05,
      "loss": 0.0005,
      "step": 1660
    },
    {
      "epoch": 2.966252220248668,
      "grad_norm": 0.08469359576702118,
      "learning_rate": 3.147202486678508e-05,
      "loss": 0.0005,
      "step": 1670
    },
    {
      "epoch": 2.984014209591474,
      "grad_norm": 0.10349100828170776,
      "learning_rate": 3.136101243339254e-05,
      "loss": 0.0005,
      "step": 1680
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.2967063816613518e-06,
      "eval_runtime": 22.3238,
      "eval_samples_per_second": 44.795,
      "eval_steps_per_second": 2.822,
      "step": 1689
    },
    {
      "epoch": 3.0017761989342806,
      "grad_norm": 0.020849984139204025,
      "learning_rate": 3.125e-05,
      "loss": 0.0005,
      "step": 1690
    },
    {
      "epoch": 3.019538188277087,
      "grad_norm": 0.017703084275126457,
      "learning_rate": 3.113898756660746e-05,
      "loss": 0.0004,
      "step": 1700
    },
    {
      "epoch": 3.0373001776198936,
      "grad_norm": 0.004325501620769501,
      "learning_rate": 3.102797513321492e-05,
      "loss": 0.0003,
      "step": 1710
    },
    {
      "epoch": 3.0550621669626996,
      "grad_norm": 0.007536259479820728,
      "learning_rate": 3.091696269982238e-05,
      "loss": 0.0004,
      "step": 1720
    },
    {
      "epoch": 3.072824156305506,
      "grad_norm": 0.008604285307228565,
      "learning_rate": 3.080595026642984e-05,
      "loss": 0.0004,
      "step": 1730
    },
    {
      "epoch": 3.0905861456483126,
      "grad_norm": 0.011171748861670494,
      "learning_rate": 3.0694937833037304e-05,
      "loss": 0.0005,
      "step": 1740
    },
    {
      "epoch": 3.108348134991119,
      "grad_norm": 0.012916183099150658,
      "learning_rate": 3.0583925399644766e-05,
      "loss": 0.0003,
      "step": 1750
    },
    {
      "epoch": 3.1261101243339255,
      "grad_norm": 0.010790830478072166,
      "learning_rate": 3.0472912966252222e-05,
      "loss": 0.0004,
      "step": 1760
    },
    {
      "epoch": 3.143872113676732,
      "grad_norm": 0.007941185496747494,
      "learning_rate": 3.0361900532859684e-05,
      "loss": 0.0005,
      "step": 1770
    },
    {
      "epoch": 3.161634103019538,
      "grad_norm": 0.022536130622029305,
      "learning_rate": 3.025088809946714e-05,
      "loss": 0.0003,
      "step": 1780
    },
    {
      "epoch": 3.1793960923623446,
      "grad_norm": 0.008146955631673336,
      "learning_rate": 3.01398756660746e-05,
      "loss": 0.0003,
      "step": 1790
    },
    {
      "epoch": 3.197158081705151,
      "grad_norm": 0.01531197503209114,
      "learning_rate": 3.0028863232682063e-05,
      "loss": 0.0005,
      "step": 1800
    },
    {
      "epoch": 3.2149200710479575,
      "grad_norm": 0.007879276759922504,
      "learning_rate": 2.9917850799289522e-05,
      "loss": 0.0003,
      "step": 1810
    },
    {
      "epoch": 3.232682060390764,
      "grad_norm": 0.031879961490631104,
      "learning_rate": 2.9806838365896984e-05,
      "loss": 0.0005,
      "step": 1820
    },
    {
      "epoch": 3.25044404973357,
      "grad_norm": 0.023382920771837234,
      "learning_rate": 2.969582593250444e-05,
      "loss": 0.0005,
      "step": 1830
    },
    {
      "epoch": 3.2682060390763765,
      "grad_norm": 0.011471973732113838,
      "learning_rate": 2.95848134991119e-05,
      "loss": 0.0004,
      "step": 1840
    },
    {
      "epoch": 3.285968028419183,
      "grad_norm": 0.009325550869107246,
      "learning_rate": 2.9473801065719364e-05,
      "loss": 0.0005,
      "step": 1850
    },
    {
      "epoch": 3.3037300177619895,
      "grad_norm": 0.047994911670684814,
      "learning_rate": 2.9362788632326822e-05,
      "loss": 0.0004,
      "step": 1860
    },
    {
      "epoch": 3.321492007104796,
      "grad_norm": 0.004230279475450516,
      "learning_rate": 2.9251776198934284e-05,
      "loss": 0.0004,
      "step": 1870
    },
    {
      "epoch": 3.339253996447602,
      "grad_norm": 0.008561152964830399,
      "learning_rate": 2.914076376554174e-05,
      "loss": 0.0003,
      "step": 1880
    },
    {
      "epoch": 3.3570159857904085,
      "grad_norm": 0.004857722204178572,
      "learning_rate": 2.9029751332149202e-05,
      "loss": 0.0006,
      "step": 1890
    },
    {
      "epoch": 3.374777975133215,
      "grad_norm": 0.012910543009638786,
      "learning_rate": 2.8918738898756664e-05,
      "loss": 0.0003,
      "step": 1900
    },
    {
      "epoch": 3.3925399644760215,
      "grad_norm": 0.010736320167779922,
      "learning_rate": 2.8807726465364123e-05,
      "loss": 0.0005,
      "step": 1910
    },
    {
      "epoch": 3.410301953818828,
      "grad_norm": 0.010209358297288418,
      "learning_rate": 2.8696714031971585e-05,
      "loss": 0.0003,
      "step": 1920
    },
    {
      "epoch": 3.428063943161634,
      "grad_norm": 0.010665121488273144,
      "learning_rate": 2.858570159857904e-05,
      "loss": 0.0004,
      "step": 1930
    },
    {
      "epoch": 3.4458259325044405,
      "grad_norm": 0.008989960886538029,
      "learning_rate": 2.8474689165186502e-05,
      "loss": 0.0003,
      "step": 1940
    },
    {
      "epoch": 3.463587921847247,
      "grad_norm": 0.003941075410693884,
      "learning_rate": 2.8363676731793964e-05,
      "loss": 0.0003,
      "step": 1950
    },
    {
      "epoch": 3.4813499111900534,
      "grad_norm": 0.047086674720048904,
      "learning_rate": 2.825266429840142e-05,
      "loss": 0.0003,
      "step": 1960
    },
    {
      "epoch": 3.49911190053286,
      "grad_norm": 0.009249791502952576,
      "learning_rate": 2.8141651865008885e-05,
      "loss": 0.0004,
      "step": 1970
    },
    {
      "epoch": 3.516873889875666,
      "grad_norm": 0.013919523917138577,
      "learning_rate": 2.803063943161634e-05,
      "loss": 0.0005,
      "step": 1980
    },
    {
      "epoch": 3.5346358792184724,
      "grad_norm": 0.05363338813185692,
      "learning_rate": 2.7919626998223802e-05,
      "loss": 0.0004,
      "step": 1990
    },
    {
      "epoch": 3.552397868561279,
      "grad_norm": 0.027678849175572395,
      "learning_rate": 2.7808614564831264e-05,
      "loss": 0.0003,
      "step": 2000
    },
    {
      "epoch": 3.5701598579040854,
      "grad_norm": 0.02696838229894638,
      "learning_rate": 2.769760213143872e-05,
      "loss": 0.0004,
      "step": 2010
    },
    {
      "epoch": 3.587921847246892,
      "grad_norm": 0.048736572265625,
      "learning_rate": 2.7586589698046182e-05,
      "loss": 0.0003,
      "step": 2020
    },
    {
      "epoch": 3.605683836589698,
      "grad_norm": 0.007270255126059055,
      "learning_rate": 2.747557726465364e-05,
      "loss": 0.0003,
      "step": 2030
    },
    {
      "epoch": 3.6234458259325044,
      "grad_norm": 0.006998795084655285,
      "learning_rate": 2.7364564831261103e-05,
      "loss": 0.0006,
      "step": 2040
    },
    {
      "epoch": 3.641207815275311,
      "grad_norm": 0.004529398865997791,
      "learning_rate": 2.7253552397868565e-05,
      "loss": 0.0002,
      "step": 2050
    },
    {
      "epoch": 3.6589698046181174,
      "grad_norm": 0.009528383612632751,
      "learning_rate": 2.714253996447602e-05,
      "loss": 0.0003,
      "step": 2060
    },
    {
      "epoch": 3.676731793960924,
      "grad_norm": 0.056139055639505386,
      "learning_rate": 2.7031527531083482e-05,
      "loss": 0.0003,
      "step": 2070
    },
    {
      "epoch": 3.69449378330373,
      "grad_norm": 0.014499225653707981,
      "learning_rate": 2.692051509769094e-05,
      "loss": 0.0004,
      "step": 2080
    },
    {
      "epoch": 3.7122557726465364,
      "grad_norm": 0.005347531288862228,
      "learning_rate": 2.6809502664298403e-05,
      "loss": 0.0005,
      "step": 2090
    },
    {
      "epoch": 3.730017761989343,
      "grad_norm": 0.010448618791997433,
      "learning_rate": 2.6698490230905865e-05,
      "loss": 0.0003,
      "step": 2100
    },
    {
      "epoch": 3.7477797513321494,
      "grad_norm": 0.009756255894899368,
      "learning_rate": 2.658747779751332e-05,
      "loss": 0.0004,
      "step": 2110
    },
    {
      "epoch": 3.765541740674956,
      "grad_norm": 0.0032908094581216574,
      "learning_rate": 2.6476465364120782e-05,
      "loss": 0.0006,
      "step": 2120
    },
    {
      "epoch": 3.783303730017762,
      "grad_norm": 0.045266490429639816,
      "learning_rate": 2.636545293072824e-05,
      "loss": 0.0005,
      "step": 2130
    },
    {
      "epoch": 3.8010657193605684,
      "grad_norm": 0.007182170171290636,
      "learning_rate": 2.6254440497335703e-05,
      "loss": 0.0003,
      "step": 2140
    },
    {
      "epoch": 3.818827708703375,
      "grad_norm": 0.00983410980552435,
      "learning_rate": 2.6143428063943165e-05,
      "loss": 0.0004,
      "step": 2150
    },
    {
      "epoch": 3.8365896980461813,
      "grad_norm": 0.00767460698261857,
      "learning_rate": 2.603241563055062e-05,
      "loss": 0.0004,
      "step": 2160
    },
    {
      "epoch": 3.854351687388988,
      "grad_norm": 0.01133924163877964,
      "learning_rate": 2.5921403197158083e-05,
      "loss": 0.0004,
      "step": 2170
    },
    {
      "epoch": 3.872113676731794,
      "grad_norm": 0.005284077022224665,
      "learning_rate": 2.581039076376554e-05,
      "loss": 0.0003,
      "step": 2180
    },
    {
      "epoch": 3.8898756660746003,
      "grad_norm": 0.010912826284766197,
      "learning_rate": 2.5699378330373004e-05,
      "loss": 0.0002,
      "step": 2190
    },
    {
      "epoch": 3.907637655417407,
      "grad_norm": 0.008936459198594093,
      "learning_rate": 2.5588365896980466e-05,
      "loss": 0.0003,
      "step": 2200
    },
    {
      "epoch": 3.9253996447602133,
      "grad_norm": 0.008787784725427628,
      "learning_rate": 2.547735346358792e-05,
      "loss": 0.0003,
      "step": 2210
    },
    {
      "epoch": 3.94316163410302,
      "grad_norm": 0.0062966118566691875,
      "learning_rate": 2.5366341030195383e-05,
      "loss": 0.0003,
      "step": 2220
    },
    {
      "epoch": 3.960923623445826,
      "grad_norm": 0.005867836996912956,
      "learning_rate": 2.5255328596802842e-05,
      "loss": 0.0002,
      "step": 2230
    },
    {
      "epoch": 3.9786856127886323,
      "grad_norm": 0.006783599965274334,
      "learning_rate": 2.5144316163410304e-05,
      "loss": 0.0003,
      "step": 2240
    },
    {
      "epoch": 3.996447602131439,
      "grad_norm": 0.009927009232342243,
      "learning_rate": 2.5033303730017766e-05,
      "loss": 0.0003,
      "step": 2250
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.0054706055816496e-06,
      "eval_runtime": 22.009,
      "eval_samples_per_second": 45.436,
      "eval_steps_per_second": 2.862,
      "step": 2252
    },
    {
      "epoch": 4.014209591474245,
      "grad_norm": 0.026025472208857536,
      "learning_rate": 2.492229129662522e-05,
      "loss": 0.0003,
      "step": 2260
    },
    {
      "epoch": 4.031971580817052,
      "grad_norm": 0.0020729824900627136,
      "learning_rate": 2.4811278863232683e-05,
      "loss": 0.0002,
      "step": 2270
    },
    {
      "epoch": 4.049733570159858,
      "grad_norm": 0.07728759199380875,
      "learning_rate": 2.4700266429840142e-05,
      "loss": 0.0003,
      "step": 2280
    },
    {
      "epoch": 4.067495559502665,
      "grad_norm": 0.006721894256770611,
      "learning_rate": 2.4589253996447604e-05,
      "loss": 0.0003,
      "step": 2290
    },
    {
      "epoch": 4.085257548845471,
      "grad_norm": 0.0029023478273302317,
      "learning_rate": 2.4478241563055063e-05,
      "loss": 0.0005,
      "step": 2300
    },
    {
      "epoch": 4.103019538188277,
      "grad_norm": 0.03222382813692093,
      "learning_rate": 2.436722912966252e-05,
      "loss": 0.0003,
      "step": 2310
    },
    {
      "epoch": 4.120781527531084,
      "grad_norm": 0.015039577148854733,
      "learning_rate": 2.4256216696269984e-05,
      "loss": 0.0003,
      "step": 2320
    },
    {
      "epoch": 4.13854351687389,
      "grad_norm": 0.04993987828493118,
      "learning_rate": 2.4145204262877442e-05,
      "loss": 0.0003,
      "step": 2330
    },
    {
      "epoch": 4.156305506216697,
      "grad_norm": 0.002969494555145502,
      "learning_rate": 2.4034191829484904e-05,
      "loss": 0.0002,
      "step": 2340
    },
    {
      "epoch": 4.174067495559503,
      "grad_norm": 0.019309891387820244,
      "learning_rate": 2.3923179396092363e-05,
      "loss": 0.0003,
      "step": 2350
    },
    {
      "epoch": 4.191829484902309,
      "grad_norm": 0.004141306970268488,
      "learning_rate": 2.3812166962699822e-05,
      "loss": 0.0002,
      "step": 2360
    },
    {
      "epoch": 4.209591474245116,
      "grad_norm": 0.0039856103248894215,
      "learning_rate": 2.3701154529307284e-05,
      "loss": 0.0003,
      "step": 2370
    },
    {
      "epoch": 4.227353463587922,
      "grad_norm": 0.005168092902749777,
      "learning_rate": 2.3590142095914743e-05,
      "loss": 0.0002,
      "step": 2380
    },
    {
      "epoch": 4.245115452930728,
      "grad_norm": 0.009566891007125378,
      "learning_rate": 2.3479129662522205e-05,
      "loss": 0.0002,
      "step": 2390
    },
    {
      "epoch": 4.262877442273535,
      "grad_norm": 0.05170539766550064,
      "learning_rate": 2.3368117229129663e-05,
      "loss": 0.0004,
      "step": 2400
    },
    {
      "epoch": 4.280639431616341,
      "grad_norm": 0.003199364058673382,
      "learning_rate": 2.3257104795737122e-05,
      "loss": 0.0002,
      "step": 2410
    },
    {
      "epoch": 4.298401420959148,
      "grad_norm": 0.11592520773410797,
      "learning_rate": 2.3146092362344584e-05,
      "loss": 0.0005,
      "step": 2420
    },
    {
      "epoch": 4.316163410301954,
      "grad_norm": 0.003450137795880437,
      "learning_rate": 2.3035079928952043e-05,
      "loss": 0.0002,
      "step": 2430
    },
    {
      "epoch": 4.333925399644761,
      "grad_norm": 0.018514085561037064,
      "learning_rate": 2.2924067495559505e-05,
      "loss": 0.0002,
      "step": 2440
    },
    {
      "epoch": 4.351687388987567,
      "grad_norm": 0.023660879582166672,
      "learning_rate": 2.2813055062166964e-05,
      "loss": 0.0003,
      "step": 2450
    },
    {
      "epoch": 4.369449378330373,
      "grad_norm": 0.0032503483816981316,
      "learning_rate": 2.2702042628774422e-05,
      "loss": 0.0002,
      "step": 2460
    },
    {
      "epoch": 4.38721136767318,
      "grad_norm": 0.00737254461273551,
      "learning_rate": 2.2591030195381885e-05,
      "loss": 0.0003,
      "step": 2470
    },
    {
      "epoch": 4.404973357015986,
      "grad_norm": 0.004189962055534124,
      "learning_rate": 2.2480017761989343e-05,
      "loss": 0.0002,
      "step": 2480
    },
    {
      "epoch": 4.422735346358792,
      "grad_norm": 0.00830498244613409,
      "learning_rate": 2.2369005328596805e-05,
      "loss": 0.0004,
      "step": 2490
    },
    {
      "epoch": 4.440497335701599,
      "grad_norm": 0.004922557156533003,
      "learning_rate": 2.2257992895204264e-05,
      "loss": 0.0002,
      "step": 2500
    },
    {
      "epoch": 4.458259325044405,
      "grad_norm": 0.018229391425848007,
      "learning_rate": 2.2146980461811723e-05,
      "loss": 0.0002,
      "step": 2510
    },
    {
      "epoch": 4.476021314387212,
      "grad_norm": 0.006720603443682194,
      "learning_rate": 2.2035968028419185e-05,
      "loss": 0.0002,
      "step": 2520
    },
    {
      "epoch": 4.493783303730018,
      "grad_norm": 0.0020187098998576403,
      "learning_rate": 2.1924955595026644e-05,
      "loss": 0.0003,
      "step": 2530
    },
    {
      "epoch": 4.511545293072825,
      "grad_norm": 0.0030220877379179,
      "learning_rate": 2.1813943161634102e-05,
      "loss": 0.0003,
      "step": 2540
    },
    {
      "epoch": 4.529307282415631,
      "grad_norm": 0.016860356554389,
      "learning_rate": 2.1702930728241564e-05,
      "loss": 0.0003,
      "step": 2550
    },
    {
      "epoch": 4.547069271758437,
      "grad_norm": 0.07945937663316727,
      "learning_rate": 2.1591918294849023e-05,
      "loss": 0.0004,
      "step": 2560
    },
    {
      "epoch": 4.564831261101244,
      "grad_norm": 0.010131836868822575,
      "learning_rate": 2.1480905861456485e-05,
      "loss": 0.0003,
      "step": 2570
    },
    {
      "epoch": 4.58259325044405,
      "grad_norm": 0.016588937491178513,
      "learning_rate": 2.1369893428063944e-05,
      "loss": 0.0002,
      "step": 2580
    },
    {
      "epoch": 4.600355239786856,
      "grad_norm": 0.004532839171588421,
      "learning_rate": 2.1258880994671403e-05,
      "loss": 0.0003,
      "step": 2590
    },
    {
      "epoch": 4.618117229129663,
      "grad_norm": 0.005671780090779066,
      "learning_rate": 2.1147868561278865e-05,
      "loss": 0.0002,
      "step": 2600
    },
    {
      "epoch": 4.635879218472469,
      "grad_norm": 0.1407405436038971,
      "learning_rate": 2.1036856127886323e-05,
      "loss": 0.0004,
      "step": 2610
    },
    {
      "epoch": 4.653641207815276,
      "grad_norm": 0.009644228033721447,
      "learning_rate": 2.0925843694493785e-05,
      "loss": 0.0002,
      "step": 2620
    },
    {
      "epoch": 4.671403197158082,
      "grad_norm": 0.021375441923737526,
      "learning_rate": 2.0814831261101244e-05,
      "loss": 0.0002,
      "step": 2630
    },
    {
      "epoch": 4.6891651865008885,
      "grad_norm": 0.05786694213747978,
      "learning_rate": 2.0703818827708703e-05,
      "loss": 0.0002,
      "step": 2640
    },
    {
      "epoch": 4.706927175843695,
      "grad_norm": 0.01768909953534603,
      "learning_rate": 2.0592806394316165e-05,
      "loss": 0.0005,
      "step": 2650
    },
    {
      "epoch": 4.724689165186501,
      "grad_norm": 0.20010802149772644,
      "learning_rate": 2.0481793960923624e-05,
      "loss": 0.0005,
      "step": 2660
    },
    {
      "epoch": 4.7424511545293075,
      "grad_norm": 0.0016011944971978664,
      "learning_rate": 2.0370781527531086e-05,
      "loss": 0.0002,
      "step": 2670
    },
    {
      "epoch": 4.760213143872114,
      "grad_norm": 0.0031138239428400993,
      "learning_rate": 2.0259769094138544e-05,
      "loss": 0.0003,
      "step": 2680
    },
    {
      "epoch": 4.77797513321492,
      "grad_norm": 0.007894694805145264,
      "learning_rate": 2.0148756660746003e-05,
      "loss": 0.0002,
      "step": 2690
    },
    {
      "epoch": 4.7957371225577266,
      "grad_norm": 0.006601590663194656,
      "learning_rate": 2.0037744227353465e-05,
      "loss": 0.0005,
      "step": 2700
    },
    {
      "epoch": 4.813499111900533,
      "grad_norm": 0.004905109293758869,
      "learning_rate": 1.9926731793960924e-05,
      "loss": 0.0002,
      "step": 2710
    },
    {
      "epoch": 4.8312611012433395,
      "grad_norm": 0.007573388516902924,
      "learning_rate": 1.9815719360568386e-05,
      "loss": 0.0005,
      "step": 2720
    },
    {
      "epoch": 4.849023090586146,
      "grad_norm": 0.004600901156663895,
      "learning_rate": 1.9704706927175845e-05,
      "loss": 0.0002,
      "step": 2730
    },
    {
      "epoch": 4.8667850799289525,
      "grad_norm": 0.004953782539814711,
      "learning_rate": 1.9593694493783303e-05,
      "loss": 0.0002,
      "step": 2740
    },
    {
      "epoch": 4.8845470692717585,
      "grad_norm": 0.005551377311348915,
      "learning_rate": 1.9482682060390765e-05,
      "loss": 0.0002,
      "step": 2750
    },
    {
      "epoch": 4.902309058614565,
      "grad_norm": 0.0048238011077046394,
      "learning_rate": 1.9371669626998224e-05,
      "loss": 0.0002,
      "step": 2760
    },
    {
      "epoch": 4.9200710479573715,
      "grad_norm": 0.009099510498344898,
      "learning_rate": 1.9260657193605686e-05,
      "loss": 0.0002,
      "step": 2770
    },
    {
      "epoch": 4.9378330373001775,
      "grad_norm": 0.0018844065489247441,
      "learning_rate": 1.9149644760213145e-05,
      "loss": 0.0002,
      "step": 2780
    },
    {
      "epoch": 4.955595026642984,
      "grad_norm": 0.0071139088831841946,
      "learning_rate": 1.9038632326820604e-05,
      "loss": 0.0002,
      "step": 2790
    },
    {
      "epoch": 4.9733570159857905,
      "grad_norm": 0.0148090161383152,
      "learning_rate": 1.8927619893428062e-05,
      "loss": 0.0002,
      "step": 2800
    },
    {
      "epoch": 4.9911190053285965,
      "grad_norm": 0.014955924823880196,
      "learning_rate": 1.8816607460035524e-05,
      "loss": 0.0002,
      "step": 2810
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.5623232911821106e-06,
      "eval_runtime": 22.0179,
      "eval_samples_per_second": 45.418,
      "eval_steps_per_second": 2.861,
      "step": 2815
    },
    {
      "epoch": 5.0088809946714035,
      "grad_norm": 0.008464839309453964,
      "learning_rate": 1.8705595026642987e-05,
      "loss": 0.0002,
      "step": 2820
    },
    {
      "epoch": 5.0266429840142095,
      "grad_norm": 0.010644927620887756,
      "learning_rate": 1.8594582593250445e-05,
      "loss": 0.0003,
      "step": 2830
    },
    {
      "epoch": 5.044404973357016,
      "grad_norm": 0.004363817162811756,
      "learning_rate": 1.8483570159857904e-05,
      "loss": 0.0002,
      "step": 2840
    },
    {
      "epoch": 5.0621669626998225,
      "grad_norm": 0.013293485157191753,
      "learning_rate": 1.8372557726465363e-05,
      "loss": 0.0002,
      "step": 2850
    },
    {
      "epoch": 5.0799289520426285,
      "grad_norm": 0.05282105877995491,
      "learning_rate": 1.8261545293072825e-05,
      "loss": 0.0004,
      "step": 2860
    },
    {
      "epoch": 5.097690941385435,
      "grad_norm": 0.002226222539320588,
      "learning_rate": 1.8150532859680287e-05,
      "loss": 0.0003,
      "step": 2870
    },
    {
      "epoch": 5.1154529307282415,
      "grad_norm": 0.004167128819972277,
      "learning_rate": 1.8039520426287746e-05,
      "loss": 0.0002,
      "step": 2880
    },
    {
      "epoch": 5.1332149200710475,
      "grad_norm": 0.10440316051244736,
      "learning_rate": 1.7928507992895204e-05,
      "loss": 0.0003,
      "step": 2890
    },
    {
      "epoch": 5.150976909413854,
      "grad_norm": 0.007494487799704075,
      "learning_rate": 1.7817495559502663e-05,
      "loss": 0.0002,
      "step": 2900
    },
    {
      "epoch": 5.1687388987566605,
      "grad_norm": 0.004491967149078846,
      "learning_rate": 1.7706483126110125e-05,
      "loss": 0.0002,
      "step": 2910
    },
    {
      "epoch": 5.186500888099467,
      "grad_norm": 0.0517878532409668,
      "learning_rate": 1.7595470692717587e-05,
      "loss": 0.0002,
      "step": 2920
    },
    {
      "epoch": 5.2042628774422734,
      "grad_norm": 0.006387591827660799,
      "learning_rate": 1.7484458259325046e-05,
      "loss": 0.0002,
      "step": 2930
    },
    {
      "epoch": 5.22202486678508,
      "grad_norm": 0.008753631263971329,
      "learning_rate": 1.7373445825932505e-05,
      "loss": 0.0003,
      "step": 2940
    },
    {
      "epoch": 5.239786856127886,
      "grad_norm": 0.0052266488783061504,
      "learning_rate": 1.7262433392539963e-05,
      "loss": 0.0003,
      "step": 2950
    },
    {
      "epoch": 5.2575488454706925,
      "grad_norm": 0.004134784918278456,
      "learning_rate": 1.7151420959147425e-05,
      "loss": 0.0002,
      "step": 2960
    },
    {
      "epoch": 5.275310834813499,
      "grad_norm": 0.003984558861702681,
      "learning_rate": 1.7040408525754887e-05,
      "loss": 0.0002,
      "step": 2970
    },
    {
      "epoch": 5.293072824156305,
      "grad_norm": 0.011758877895772457,
      "learning_rate": 1.6929396092362346e-05,
      "loss": 0.0002,
      "step": 2980
    },
    {
      "epoch": 5.3108348134991115,
      "grad_norm": 0.005662502255290747,
      "learning_rate": 1.6818383658969805e-05,
      "loss": 0.0002,
      "step": 2990
    },
    {
      "epoch": 5.328596802841918,
      "grad_norm": 0.0018373008351773024,
      "learning_rate": 1.6707371225577264e-05,
      "loss": 0.0002,
      "step": 3000
    },
    {
      "epoch": 5.346358792184724,
      "grad_norm": 0.0203489288687706,
      "learning_rate": 1.6596358792184726e-05,
      "loss": 0.0002,
      "step": 3010
    },
    {
      "epoch": 5.364120781527531,
      "grad_norm": 0.006496446672827005,
      "learning_rate": 1.6485346358792188e-05,
      "loss": 0.0002,
      "step": 3020
    },
    {
      "epoch": 5.381882770870337,
      "grad_norm": 0.0037881876341998577,
      "learning_rate": 1.6374333925399646e-05,
      "loss": 0.0002,
      "step": 3030
    },
    {
      "epoch": 5.399644760213144,
      "grad_norm": 0.006189519539475441,
      "learning_rate": 1.6263321492007105e-05,
      "loss": 0.0002,
      "step": 3040
    },
    {
      "epoch": 5.41740674955595,
      "grad_norm": 0.016199415549635887,
      "learning_rate": 1.6152309058614564e-05,
      "loss": 0.0002,
      "step": 3050
    },
    {
      "epoch": 5.435168738898756,
      "grad_norm": 0.0038983342237770557,
      "learning_rate": 1.6041296625222023e-05,
      "loss": 0.0002,
      "step": 3060
    },
    {
      "epoch": 5.452930728241563,
      "grad_norm": 0.12250888347625732,
      "learning_rate": 1.5930284191829488e-05,
      "loss": 0.0002,
      "step": 3070
    },
    {
      "epoch": 5.470692717584369,
      "grad_norm": 0.008953145705163479,
      "learning_rate": 1.5819271758436947e-05,
      "loss": 0.0002,
      "step": 3080
    },
    {
      "epoch": 5.488454706927175,
      "grad_norm": 0.0029956751968711615,
      "learning_rate": 1.5708259325044405e-05,
      "loss": 0.0003,
      "step": 3090
    },
    {
      "epoch": 5.506216696269982,
      "grad_norm": 0.011769605800509453,
      "learning_rate": 1.5597246891651864e-05,
      "loss": 0.0003,
      "step": 3100
    },
    {
      "epoch": 5.523978685612788,
      "grad_norm": 0.006393752060830593,
      "learning_rate": 1.5486234458259326e-05,
      "loss": 0.0002,
      "step": 3110
    },
    {
      "epoch": 5.541740674955595,
      "grad_norm": 0.0033896728418767452,
      "learning_rate": 1.537522202486679e-05,
      "loss": 0.0002,
      "step": 3120
    },
    {
      "epoch": 5.559502664298401,
      "grad_norm": 0.004390314687043428,
      "learning_rate": 1.5264209591474247e-05,
      "loss": 0.0002,
      "step": 3130
    },
    {
      "epoch": 5.577264653641208,
      "grad_norm": 0.005353247281163931,
      "learning_rate": 1.5153197158081706e-05,
      "loss": 0.0002,
      "step": 3140
    },
    {
      "epoch": 5.595026642984014,
      "grad_norm": 0.006205048877745867,
      "learning_rate": 1.5042184724689164e-05,
      "loss": 0.0002,
      "step": 3150
    },
    {
      "epoch": 5.61278863232682,
      "grad_norm": 0.0020661200396716595,
      "learning_rate": 1.4931172291296627e-05,
      "loss": 0.0001,
      "step": 3160
    },
    {
      "epoch": 5.630550621669627,
      "grad_norm": 0.004063829779624939,
      "learning_rate": 1.4820159857904087e-05,
      "loss": 0.0001,
      "step": 3170
    },
    {
      "epoch": 5.648312611012433,
      "grad_norm": 0.0014260923489928246,
      "learning_rate": 1.4709147424511546e-05,
      "loss": 0.0001,
      "step": 3180
    },
    {
      "epoch": 5.666074600355239,
      "grad_norm": 0.0031196416821330786,
      "learning_rate": 1.4598134991119006e-05,
      "loss": 0.0002,
      "step": 3190
    },
    {
      "epoch": 5.683836589698046,
      "grad_norm": 0.0019640398677438498,
      "learning_rate": 1.4487122557726465e-05,
      "loss": 0.0003,
      "step": 3200
    },
    {
      "epoch": 5.701598579040852,
      "grad_norm": 0.009311804547905922,
      "learning_rate": 1.4376110124333927e-05,
      "loss": 0.0001,
      "step": 3210
    },
    {
      "epoch": 5.719360568383659,
      "grad_norm": 0.002296882914379239,
      "learning_rate": 1.4265097690941387e-05,
      "loss": 0.0002,
      "step": 3220
    },
    {
      "epoch": 5.737122557726465,
      "grad_norm": 0.008231076411902905,
      "learning_rate": 1.4154085257548846e-05,
      "loss": 0.0001,
      "step": 3230
    },
    {
      "epoch": 5.754884547069272,
      "grad_norm": 0.02695680968463421,
      "learning_rate": 1.4043072824156306e-05,
      "loss": 0.0002,
      "step": 3240
    },
    {
      "epoch": 5.772646536412078,
      "grad_norm": 0.005848235916346312,
      "learning_rate": 1.3932060390763765e-05,
      "loss": 0.0001,
      "step": 3250
    },
    {
      "epoch": 5.790408525754884,
      "grad_norm": 0.00468071224167943,
      "learning_rate": 1.3821047957371227e-05,
      "loss": 0.0001,
      "step": 3260
    },
    {
      "epoch": 5.808170515097691,
      "grad_norm": 0.0027211324777454138,
      "learning_rate": 1.3710035523978688e-05,
      "loss": 0.0002,
      "step": 3270
    },
    {
      "epoch": 5.825932504440497,
      "grad_norm": 0.007977406494319439,
      "learning_rate": 1.3599023090586146e-05,
      "loss": 0.0002,
      "step": 3280
    },
    {
      "epoch": 5.843694493783303,
      "grad_norm": 0.0070778182707726955,
      "learning_rate": 1.3488010657193607e-05,
      "loss": 0.0002,
      "step": 3290
    },
    {
      "epoch": 5.86145648312611,
      "grad_norm": 0.014064283110201359,
      "learning_rate": 1.3376998223801065e-05,
      "loss": 0.0002,
      "step": 3300
    },
    {
      "epoch": 5.879218472468916,
      "grad_norm": 0.0018223852384835482,
      "learning_rate": 1.3265985790408527e-05,
      "loss": 0.0001,
      "step": 3310
    },
    {
      "epoch": 5.896980461811723,
      "grad_norm": 0.070171058177948,
      "learning_rate": 1.3154973357015988e-05,
      "loss": 0.0002,
      "step": 3320
    },
    {
      "epoch": 5.914742451154529,
      "grad_norm": 0.03238040581345558,
      "learning_rate": 1.3043960923623447e-05,
      "loss": 0.0002,
      "step": 3330
    },
    {
      "epoch": 5.932504440497336,
      "grad_norm": 0.003422616748139262,
      "learning_rate": 1.2932948490230907e-05,
      "loss": 0.0002,
      "step": 3340
    },
    {
      "epoch": 5.950266429840142,
      "grad_norm": 0.0016958152409642935,
      "learning_rate": 1.2821936056838366e-05,
      "loss": 0.0001,
      "step": 3350
    },
    {
      "epoch": 5.968028419182948,
      "grad_norm": 0.0023135722149163485,
      "learning_rate": 1.2710923623445828e-05,
      "loss": 0.0001,
      "step": 3360
    },
    {
      "epoch": 5.985790408525755,
      "grad_norm": 0.00846858974546194,
      "learning_rate": 1.2599911190053288e-05,
      "loss": 0.0002,
      "step": 3370
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.150123352999799e-06,
      "eval_runtime": 22.2281,
      "eval_samples_per_second": 44.988,
      "eval_steps_per_second": 2.834,
      "step": 3378
    },
    {
      "epoch": 6.003552397868561,
      "grad_norm": 0.059037767350673676,
      "learning_rate": 1.2488898756660747e-05,
      "loss": 0.0003,
      "step": 3380
    },
    {
      "epoch": 6.021314387211367,
      "grad_norm": 0.010334952734410763,
      "learning_rate": 1.2377886323268206e-05,
      "loss": 0.0002,
      "step": 3390
    },
    {
      "epoch": 6.039076376554174,
      "grad_norm": 0.002199666341766715,
      "learning_rate": 1.2266873889875668e-05,
      "loss": 0.0001,
      "step": 3400
    },
    {
      "epoch": 6.05683836589698,
      "grad_norm": 0.0027150341775268316,
      "learning_rate": 1.2155861456483126e-05,
      "loss": 0.0002,
      "step": 3410
    },
    {
      "epoch": 6.074600355239787,
      "grad_norm": 0.06314147263765335,
      "learning_rate": 1.2044849023090587e-05,
      "loss": 0.0002,
      "step": 3420
    },
    {
      "epoch": 6.092362344582593,
      "grad_norm": 0.10585487633943558,
      "learning_rate": 1.1933836589698047e-05,
      "loss": 0.0002,
      "step": 3430
    },
    {
      "epoch": 6.110124333925399,
      "grad_norm": 0.005490344017744064,
      "learning_rate": 1.1822824156305506e-05,
      "loss": 0.0001,
      "step": 3440
    },
    {
      "epoch": 6.127886323268206,
      "grad_norm": 0.03931137174367905,
      "learning_rate": 1.1711811722912968e-05,
      "loss": 0.0001,
      "step": 3450
    },
    {
      "epoch": 6.145648312611012,
      "grad_norm": 0.001892774016596377,
      "learning_rate": 1.1600799289520427e-05,
      "loss": 0.0001,
      "step": 3460
    },
    {
      "epoch": 6.163410301953819,
      "grad_norm": 0.002348232315853238,
      "learning_rate": 1.1489786856127887e-05,
      "loss": 0.0002,
      "step": 3470
    },
    {
      "epoch": 6.181172291296625,
      "grad_norm": 0.0030624247156083584,
      "learning_rate": 1.1378774422735347e-05,
      "loss": 0.0002,
      "step": 3480
    },
    {
      "epoch": 6.198934280639431,
      "grad_norm": 0.003177167382091284,
      "learning_rate": 1.1267761989342806e-05,
      "loss": 0.0001,
      "step": 3490
    },
    {
      "epoch": 6.216696269982238,
      "grad_norm": 0.003824736224487424,
      "learning_rate": 1.1156749555950268e-05,
      "loss": 0.0002,
      "step": 3500
    },
    {
      "epoch": 6.234458259325044,
      "grad_norm": 0.00209417543374002,
      "learning_rate": 1.1045737122557727e-05,
      "loss": 0.0001,
      "step": 3510
    },
    {
      "epoch": 6.252220248667851,
      "grad_norm": 0.011512521654367447,
      "learning_rate": 1.0934724689165186e-05,
      "loss": 0.0001,
      "step": 3520
    },
    {
      "epoch": 6.269982238010657,
      "grad_norm": 0.01817052625119686,
      "learning_rate": 1.0823712255772648e-05,
      "loss": 0.0002,
      "step": 3530
    },
    {
      "epoch": 6.287744227353464,
      "grad_norm": 0.004132445901632309,
      "learning_rate": 1.0712699822380106e-05,
      "loss": 0.0001,
      "step": 3540
    },
    {
      "epoch": 6.30550621669627,
      "grad_norm": 0.0017249000957235694,
      "learning_rate": 1.0601687388987567e-05,
      "loss": 0.0001,
      "step": 3550
    },
    {
      "epoch": 6.323268206039076,
      "grad_norm": 0.0024557355791330338,
      "learning_rate": 1.0490674955595027e-05,
      "loss": 0.0002,
      "step": 3560
    },
    {
      "epoch": 6.341030195381883,
      "grad_norm": 0.02593265473842621,
      "learning_rate": 1.0379662522202486e-05,
      "loss": 0.0002,
      "step": 3570
    },
    {
      "epoch": 6.358792184724689,
      "grad_norm": 0.00831576343625784,
      "learning_rate": 1.0268650088809948e-05,
      "loss": 0.0001,
      "step": 3580
    },
    {
      "epoch": 6.376554174067495,
      "grad_norm": 0.12613391876220703,
      "learning_rate": 1.0157637655417407e-05,
      "loss": 0.0002,
      "step": 3590
    },
    {
      "epoch": 6.394316163410302,
      "grad_norm": 0.00139847572427243,
      "learning_rate": 1.0046625222024867e-05,
      "loss": 0.0002,
      "step": 3600
    },
    {
      "epoch": 6.412078152753108,
      "grad_norm": 0.0011393679305911064,
      "learning_rate": 9.935612788632327e-06,
      "loss": 0.0001,
      "step": 3610
    },
    {
      "epoch": 6.429840142095915,
      "grad_norm": 0.031851690262556076,
      "learning_rate": 9.824600355239786e-06,
      "loss": 0.0002,
      "step": 3620
    },
    {
      "epoch": 6.447602131438721,
      "grad_norm": 0.002192731713876128,
      "learning_rate": 9.713587921847248e-06,
      "loss": 0.0002,
      "step": 3630
    },
    {
      "epoch": 6.465364120781528,
      "grad_norm": 0.012485062703490257,
      "learning_rate": 9.602575488454707e-06,
      "loss": 0.0002,
      "step": 3640
    },
    {
      "epoch": 6.483126110124334,
      "grad_norm": 0.02442704513669014,
      "learning_rate": 9.491563055062167e-06,
      "loss": 0.0002,
      "step": 3650
    },
    {
      "epoch": 6.50088809946714,
      "grad_norm": 0.0015811693156138062,
      "learning_rate": 9.380550621669628e-06,
      "loss": 0.0001,
      "step": 3660
    },
    {
      "epoch": 6.518650088809947,
      "grad_norm": 0.039409998804330826,
      "learning_rate": 9.269538188277086e-06,
      "loss": 0.0001,
      "step": 3670
    },
    {
      "epoch": 6.536412078152753,
      "grad_norm": 0.002665013074874878,
      "learning_rate": 9.158525754884547e-06,
      "loss": 0.0002,
      "step": 3680
    },
    {
      "epoch": 6.554174067495559,
      "grad_norm": 0.003954344894737005,
      "learning_rate": 9.047513321492007e-06,
      "loss": 0.0002,
      "step": 3690
    },
    {
      "epoch": 6.571936056838366,
      "grad_norm": 0.005066856276243925,
      "learning_rate": 8.936500888099468e-06,
      "loss": 0.0002,
      "step": 3700
    },
    {
      "epoch": 6.589698046181172,
      "grad_norm": 0.007488698698580265,
      "learning_rate": 8.825488454706928e-06,
      "loss": 0.0001,
      "step": 3710
    },
    {
      "epoch": 6.607460035523979,
      "grad_norm": 0.00137425831053406,
      "learning_rate": 8.714476021314387e-06,
      "loss": 0.0002,
      "step": 3720
    },
    {
      "epoch": 6.625222024866785,
      "grad_norm": 0.005670055281370878,
      "learning_rate": 8.603463587921847e-06,
      "loss": 0.0001,
      "step": 3730
    },
    {
      "epoch": 6.642984014209592,
      "grad_norm": 0.0011902772821485996,
      "learning_rate": 8.492451154529308e-06,
      "loss": 0.0001,
      "step": 3740
    },
    {
      "epoch": 6.660746003552398,
      "grad_norm": 0.0016062385402619839,
      "learning_rate": 8.381438721136768e-06,
      "loss": 0.0001,
      "step": 3750
    },
    {
      "epoch": 6.678507992895204,
      "grad_norm": 0.0020727047231048346,
      "learning_rate": 8.270426287744228e-06,
      "loss": 0.0002,
      "step": 3760
    },
    {
      "epoch": 6.696269982238011,
      "grad_norm": 0.0027436264790594578,
      "learning_rate": 8.159413854351687e-06,
      "loss": 0.0004,
      "step": 3770
    },
    {
      "epoch": 6.714031971580817,
      "grad_norm": 0.0016453347634524107,
      "learning_rate": 8.048401420959147e-06,
      "loss": 0.0001,
      "step": 3780
    },
    {
      "epoch": 6.731793960923623,
      "grad_norm": 0.00958060473203659,
      "learning_rate": 7.937388987566608e-06,
      "loss": 0.0002,
      "step": 3790
    },
    {
      "epoch": 6.74955595026643,
      "grad_norm": 0.003439374268054962,
      "learning_rate": 7.826376554174068e-06,
      "loss": 0.0001,
      "step": 3800
    },
    {
      "epoch": 6.767317939609236,
      "grad_norm": 0.007251431699842215,
      "learning_rate": 7.715364120781527e-06,
      "loss": 0.0002,
      "step": 3810
    },
    {
      "epoch": 6.785079928952043,
      "grad_norm": 0.004211882594972849,
      "learning_rate": 7.604351687388989e-06,
      "loss": 0.0001,
      "step": 3820
    },
    {
      "epoch": 6.802841918294849,
      "grad_norm": 0.004078486934304237,
      "learning_rate": 7.493339253996448e-06,
      "loss": 0.0001,
      "step": 3830
    },
    {
      "epoch": 6.820603907637656,
      "grad_norm": 0.0059041669592261314,
      "learning_rate": 7.382326820603907e-06,
      "loss": 0.0001,
      "step": 3840
    },
    {
      "epoch": 6.838365896980462,
      "grad_norm": 0.00534476526081562,
      "learning_rate": 7.2713143872113685e-06,
      "loss": 0.0001,
      "step": 3850
    },
    {
      "epoch": 6.856127886323268,
      "grad_norm": 0.012589712627232075,
      "learning_rate": 7.160301953818828e-06,
      "loss": 0.0003,
      "step": 3860
    },
    {
      "epoch": 6.873889875666075,
      "grad_norm": 0.010835024528205395,
      "learning_rate": 7.0492895204262885e-06,
      "loss": 0.0002,
      "step": 3870
    },
    {
      "epoch": 6.891651865008881,
      "grad_norm": 0.0016127921408042312,
      "learning_rate": 6.938277087033748e-06,
      "loss": 0.0001,
      "step": 3880
    },
    {
      "epoch": 6.909413854351687,
      "grad_norm": 0.00635075569152832,
      "learning_rate": 6.827264653641208e-06,
      "loss": 0.0004,
      "step": 3890
    },
    {
      "epoch": 6.927175843694494,
      "grad_norm": 0.002242429880425334,
      "learning_rate": 6.716252220248669e-06,
      "loss": 0.0001,
      "step": 3900
    },
    {
      "epoch": 6.9449378330373,
      "grad_norm": 0.03803255409002304,
      "learning_rate": 6.605239786856128e-06,
      "loss": 0.0001,
      "step": 3910
    },
    {
      "epoch": 6.962699822380107,
      "grad_norm": 0.0017812428995966911,
      "learning_rate": 6.494227353463589e-06,
      "loss": 0.0001,
      "step": 3920
    },
    {
      "epoch": 6.980461811722913,
      "grad_norm": 0.07593787461519241,
      "learning_rate": 6.383214920071048e-06,
      "loss": 0.0001,
      "step": 3930
    },
    {
      "epoch": 6.99822380106572,
      "grad_norm": 0.004818908404558897,
      "learning_rate": 6.272202486678508e-06,
      "loss": 0.0001,
      "step": 3940
    },
    {
      "epoch": 7.0,
      "eval_loss": 9.495445283391746e-07,
      "eval_runtime": 22.2592,
      "eval_samples_per_second": 44.925,
      "eval_steps_per_second": 2.83,
      "step": 3941
    },
    {
      "epoch": 7.015985790408526,
      "grad_norm": 0.13469889760017395,
      "learning_rate": 6.161190053285968e-06,
      "loss": 0.0002,
      "step": 3950
    },
    {
      "epoch": 7.033747779751332,
      "grad_norm": 0.001866370439529419,
      "learning_rate": 6.050177619893428e-06,
      "loss": 0.0001,
      "step": 3960
    },
    {
      "epoch": 7.051509769094139,
      "grad_norm": 0.0034974406007677317,
      "learning_rate": 5.939165186500888e-06,
      "loss": 0.0001,
      "step": 3970
    },
    {
      "epoch": 7.069271758436945,
      "grad_norm": 0.04713987559080124,
      "learning_rate": 5.828152753108349e-06,
      "loss": 0.0001,
      "step": 3980
    },
    {
      "epoch": 7.087033747779751,
      "grad_norm": 0.01695702224969864,
      "learning_rate": 5.717140319715809e-06,
      "loss": 0.0001,
      "step": 3990
    },
    {
      "epoch": 7.104795737122558,
      "grad_norm": 0.0627860575914383,
      "learning_rate": 5.6061278863232686e-06,
      "loss": 0.0001,
      "step": 4000
    },
    {
      "epoch": 7.122557726465364,
      "grad_norm": 0.002155202440917492,
      "learning_rate": 5.495115452930728e-06,
      "loss": 0.0002,
      "step": 4010
    },
    {
      "epoch": 7.140319715808171,
      "grad_norm": 0.007705147843807936,
      "learning_rate": 5.3841030195381885e-06,
      "loss": 0.0001,
      "step": 4020
    },
    {
      "epoch": 7.158081705150977,
      "grad_norm": 0.0011336603201925755,
      "learning_rate": 5.273090586145649e-06,
      "loss": 0.0002,
      "step": 4030
    },
    {
      "epoch": 7.175843694493783,
      "grad_norm": 0.006822965107858181,
      "learning_rate": 5.1620781527531085e-06,
      "loss": 0.0001,
      "step": 4040
    },
    {
      "epoch": 7.19360568383659,
      "grad_norm": 0.001960517605766654,
      "learning_rate": 5.051065719360568e-06,
      "loss": 0.0001,
      "step": 4050
    },
    {
      "epoch": 7.211367673179396,
      "grad_norm": 0.027150465175509453,
      "learning_rate": 4.940053285968028e-06,
      "loss": 0.0001,
      "step": 4060
    },
    {
      "epoch": 7.229129662522203,
      "grad_norm": 0.011790856719017029,
      "learning_rate": 4.829040852575489e-06,
      "loss": 0.0002,
      "step": 4070
    },
    {
      "epoch": 7.246891651865009,
      "grad_norm": 0.0008871167083270848,
      "learning_rate": 4.718028419182949e-06,
      "loss": 0.0001,
      "step": 4080
    },
    {
      "epoch": 7.264653641207815,
      "grad_norm": 0.001180154737085104,
      "learning_rate": 4.607015985790409e-06,
      "loss": 0.0001,
      "step": 4090
    },
    {
      "epoch": 7.282415630550622,
      "grad_norm": 0.0056028300896286964,
      "learning_rate": 4.496003552397868e-06,
      "loss": 0.0001,
      "step": 4100
    },
    {
      "epoch": 7.300177619893428,
      "grad_norm": 0.009414936415851116,
      "learning_rate": 4.384991119005329e-06,
      "loss": 0.0002,
      "step": 4110
    },
    {
      "epoch": 7.317939609236235,
      "grad_norm": 0.003175571095198393,
      "learning_rate": 4.273978685612789e-06,
      "loss": 0.0001,
      "step": 4120
    },
    {
      "epoch": 7.335701598579041,
      "grad_norm": 0.0012675018515437841,
      "learning_rate": 4.162966252220249e-06,
      "loss": 0.0003,
      "step": 4130
    },
    {
      "epoch": 7.353463587921847,
      "grad_norm": 0.006813838612288237,
      "learning_rate": 4.051953818827709e-06,
      "loss": 0.0001,
      "step": 4140
    },
    {
      "epoch": 7.371225577264654,
      "grad_norm": 0.059542424976825714,
      "learning_rate": 3.940941385435169e-06,
      "loss": 0.0001,
      "step": 4150
    },
    {
      "epoch": 7.38898756660746,
      "grad_norm": 0.12738512456417084,
      "learning_rate": 3.829928952042629e-06,
      "loss": 0.0002,
      "step": 4160
    },
    {
      "epoch": 7.406749555950267,
      "grad_norm": 0.0020928082522004843,
      "learning_rate": 3.718916518650089e-06,
      "loss": 0.0002,
      "step": 4170
    },
    {
      "epoch": 7.424511545293073,
      "grad_norm": 0.03172928839921951,
      "learning_rate": 3.607904085257549e-06,
      "loss": 0.0002,
      "step": 4180
    },
    {
      "epoch": 7.442273534635879,
      "grad_norm": 0.002771994797512889,
      "learning_rate": 3.4968916518650093e-06,
      "loss": 0.0001,
      "step": 4190
    },
    {
      "epoch": 7.460035523978686,
      "grad_norm": 0.002538390224799514,
      "learning_rate": 3.3858792184724693e-06,
      "loss": 0.0002,
      "step": 4200
    },
    {
      "epoch": 7.477797513321492,
      "grad_norm": 0.004362764302641153,
      "learning_rate": 3.274866785079929e-06,
      "loss": 0.0001,
      "step": 4210
    },
    {
      "epoch": 7.495559502664299,
      "grad_norm": 0.0017780765192583203,
      "learning_rate": 3.1638543516873893e-06,
      "loss": 0.0001,
      "step": 4220
    },
    {
      "epoch": 7.513321492007105,
      "grad_norm": 0.004568871576339006,
      "learning_rate": 3.0528419182948492e-06,
      "loss": 0.0001,
      "step": 4230
    },
    {
      "epoch": 7.531083481349912,
      "grad_norm": 0.007695276755839586,
      "learning_rate": 2.9418294849023092e-06,
      "loss": 0.0001,
      "step": 4240
    },
    {
      "epoch": 7.548845470692718,
      "grad_norm": 0.002825086936354637,
      "learning_rate": 2.830817051509769e-06,
      "loss": 0.0001,
      "step": 4250
    },
    {
      "epoch": 7.566607460035524,
      "grad_norm": 0.0010516445618122816,
      "learning_rate": 2.7198046181172296e-06,
      "loss": 0.0001,
      "step": 4260
    },
    {
      "epoch": 7.584369449378331,
      "grad_norm": 0.0022022314369678497,
      "learning_rate": 2.608792184724689e-06,
      "loss": 0.0001,
      "step": 4270
    },
    {
      "epoch": 7.602131438721137,
      "grad_norm": 0.00462978845462203,
      "learning_rate": 2.4977797513321495e-06,
      "loss": 0.0001,
      "step": 4280
    },
    {
      "epoch": 7.619893428063943,
      "grad_norm": 0.002377378987148404,
      "learning_rate": 2.386767317939609e-06,
      "loss": 0.0001,
      "step": 4290
    },
    {
      "epoch": 7.63765541740675,
      "grad_norm": 0.10261301696300507,
      "learning_rate": 2.2757548845470695e-06,
      "loss": 0.0002,
      "step": 4300
    },
    {
      "epoch": 7.655417406749556,
      "grad_norm": 0.005298789124935865,
      "learning_rate": 2.1647424511545295e-06,
      "loss": 0.0002,
      "step": 4310
    },
    {
      "epoch": 7.673179396092363,
      "grad_norm": 0.0017874079057946801,
      "learning_rate": 2.0537300177619894e-06,
      "loss": 0.0001,
      "step": 4320
    },
    {
      "epoch": 7.690941385435169,
      "grad_norm": 0.002110840054228902,
      "learning_rate": 1.9427175843694494e-06,
      "loss": 0.0001,
      "step": 4330
    },
    {
      "epoch": 7.708703374777976,
      "grad_norm": 0.009644666686654091,
      "learning_rate": 1.8317051509769096e-06,
      "loss": 0.0001,
      "step": 4340
    },
    {
      "epoch": 7.726465364120782,
      "grad_norm": 0.005328277125954628,
      "learning_rate": 1.7206927175843696e-06,
      "loss": 0.0001,
      "step": 4350
    },
    {
      "epoch": 7.744227353463588,
      "grad_norm": 0.0009074486442841589,
      "learning_rate": 1.6096802841918297e-06,
      "loss": 0.0001,
      "step": 4360
    },
    {
      "epoch": 7.761989342806395,
      "grad_norm": 0.002376771066337824,
      "learning_rate": 1.4986678507992897e-06,
      "loss": 0.0001,
      "step": 4370
    },
    {
      "epoch": 7.779751332149201,
      "grad_norm": 0.009661192074418068,
      "learning_rate": 1.3876554174067497e-06,
      "loss": 0.0001,
      "step": 4380
    },
    {
      "epoch": 7.797513321492007,
      "grad_norm": 0.0024153704289346933,
      "learning_rate": 1.2766429840142097e-06,
      "loss": 0.0001,
      "step": 4390
    },
    {
      "epoch": 7.815275310834814,
      "grad_norm": 0.008802085183560848,
      "learning_rate": 1.1656305506216696e-06,
      "loss": 0.0001,
      "step": 4400
    },
    {
      "epoch": 7.83303730017762,
      "grad_norm": 0.0018315628403797746,
      "learning_rate": 1.0546181172291296e-06,
      "loss": 0.0001,
      "step": 4410
    },
    {
      "epoch": 7.850799289520427,
      "grad_norm": 0.002987775718793273,
      "learning_rate": 9.436056838365898e-07,
      "loss": 0.0001,
      "step": 4420
    },
    {
      "epoch": 7.868561278863233,
      "grad_norm": 0.010026910342276096,
      "learning_rate": 8.325932504440498e-07,
      "loss": 0.0001,
      "step": 4430
    },
    {
      "epoch": 7.88632326820604,
      "grad_norm": 0.012213947251439095,
      "learning_rate": 7.215808170515097e-07,
      "loss": 0.0001,
      "step": 4440
    },
    {
      "epoch": 7.904085257548846,
      "grad_norm": 0.0036439860705286264,
      "learning_rate": 6.105683836589698e-07,
      "loss": 0.0001,
      "step": 4450
    },
    {
      "epoch": 7.921847246891652,
      "grad_norm": 0.04297882691025734,
      "learning_rate": 4.995559502664299e-07,
      "loss": 0.0001,
      "step": 4460
    },
    {
      "epoch": 7.939609236234459,
      "grad_norm": 0.004080113489180803,
      "learning_rate": 3.885435168738899e-07,
      "loss": 0.0001,
      "step": 4470
    },
    {
      "epoch": 7.957371225577265,
      "grad_norm": 0.0037819864228367805,
      "learning_rate": 2.775310834813499e-07,
      "loss": 0.0001,
      "step": 4480
    },
    {
      "epoch": 7.975133214920071,
      "grad_norm": 0.008480883203446865,
      "learning_rate": 1.6651865008880993e-07,
      "loss": 0.0001,
      "step": 4490
    },
    {
      "epoch": 7.992895204262878,
      "grad_norm": 0.01645360141992569,
      "learning_rate": 5.550621669626998e-08,
      "loss": 0.0002,
      "step": 4500
    },
    {
      "epoch": 8.0,
      "eval_loss": 8.977843890534132e-07,
      "eval_runtime": 21.9896,
      "eval_samples_per_second": 45.476,
      "eval_steps_per_second": 2.865,
      "step": 4504
    }
  ],
  "logging_steps": 10,
  "max_steps": 4504,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.105854208303104e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}