{
  "best_metric": 0.23820149898529053,
  "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles12_seed1_q3_DA/checkpoint-2345",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2345,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010660980810234541,
      "grad_norm": 3.491609811782837,
      "learning_rate": 4.989339019189766e-05,
      "loss": 0.1846,
      "step": 5
    },
    {
      "epoch": 0.021321961620469083,
      "grad_norm": 4.594371318817139,
      "learning_rate": 4.978678038379531e-05,
      "loss": 0.127,
      "step": 10
    },
    {
      "epoch": 0.031982942430703626,
      "grad_norm": 4.927212715148926,
      "learning_rate": 4.9680170575692967e-05,
      "loss": 0.1401,
      "step": 15
    },
    {
      "epoch": 0.042643923240938165,
      "grad_norm": 3.7190589904785156,
      "learning_rate": 4.957356076759062e-05,
      "loss": 0.0909,
      "step": 20
    },
    {
      "epoch": 0.053304904051172705,
      "grad_norm": 8.8511323928833,
      "learning_rate": 4.9466950959488276e-05,
      "loss": 0.1589,
      "step": 25
    },
    {
      "epoch": 0.06396588486140725,
      "grad_norm": 4.348031997680664,
      "learning_rate": 4.936034115138593e-05,
      "loss": 0.1815,
      "step": 30
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 5.277684688568115,
      "learning_rate": 4.9253731343283586e-05,
      "loss": 0.133,
      "step": 35
    },
    {
      "epoch": 0.08528784648187633,
      "grad_norm": 3.881178379058838,
      "learning_rate": 4.914712153518124e-05,
      "loss": 0.1968,
      "step": 40
    },
    {
      "epoch": 0.09594882729211088,
      "grad_norm": 1.156666874885559,
      "learning_rate": 4.904051172707889e-05,
      "loss": 0.0783,
      "step": 45
    },
    {
      "epoch": 0.10660980810234541,
      "grad_norm": 10.230473518371582,
      "learning_rate": 4.893390191897655e-05,
      "loss": 0.1686,
      "step": 50
    },
    {
      "epoch": 0.11727078891257996,
      "grad_norm": 4.980260848999023,
      "learning_rate": 4.88272921108742e-05,
      "loss": 0.1415,
      "step": 55
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 2.409019708633423,
      "learning_rate": 4.872068230277186e-05,
      "loss": 0.143,
      "step": 60
    },
    {
      "epoch": 0.13859275053304904,
      "grad_norm": 2.853699207305908,
      "learning_rate": 4.861407249466951e-05,
      "loss": 0.1756,
      "step": 65
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 2.1676573753356934,
      "learning_rate": 4.850746268656717e-05,
      "loss": 0.1306,
      "step": 70
    },
    {
      "epoch": 0.15991471215351813,
      "grad_norm": 2.497457265853882,
      "learning_rate": 4.840085287846482e-05,
      "loss": 0.1959,
      "step": 75
    },
    {
      "epoch": 0.17057569296375266,
      "grad_norm": 1.732093334197998,
      "learning_rate": 4.829424307036248e-05,
      "loss": 0.1187,
      "step": 80
    },
    {
      "epoch": 0.1812366737739872,
      "grad_norm": 3.7211694717407227,
      "learning_rate": 4.8187633262260126e-05,
      "loss": 0.0881,
      "step": 85
    },
    {
      "epoch": 0.19189765458422176,
      "grad_norm": 5.591403484344482,
      "learning_rate": 4.808102345415779e-05,
      "loss": 0.0937,
      "step": 90
    },
    {
      "epoch": 0.2025586353944563,
      "grad_norm": 6.826287746429443,
      "learning_rate": 4.7974413646055436e-05,
      "loss": 0.1656,
      "step": 95
    },
    {
      "epoch": 0.21321961620469082,
      "grad_norm": 2.797285318374634,
      "learning_rate": 4.78678038379531e-05,
      "loss": 0.1693,
      "step": 100
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 9.413431167602539,
      "learning_rate": 4.7761194029850745e-05,
      "loss": 0.0951,
      "step": 105
    },
    {
      "epoch": 0.2345415778251599,
      "grad_norm": 3.8331892490386963,
      "learning_rate": 4.765458422174841e-05,
      "loss": 0.1159,
      "step": 110
    },
    {
      "epoch": 0.24520255863539445,
      "grad_norm": 2.5063424110412598,
      "learning_rate": 4.7547974413646055e-05,
      "loss": 0.1672,
      "step": 115
    },
    {
      "epoch": 0.255863539445629,
      "grad_norm": 2.3521742820739746,
      "learning_rate": 4.7441364605543716e-05,
      "loss": 0.0958,
      "step": 120
    },
    {
      "epoch": 0.26652452025586354,
      "grad_norm": 1.8726422786712646,
      "learning_rate": 4.7334754797441364e-05,
      "loss": 0.1181,
      "step": 125
    },
    {
      "epoch": 0.2771855010660981,
      "grad_norm": 2.66015362739563,
      "learning_rate": 4.7228144989339026e-05,
      "loss": 0.1168,
      "step": 130
    },
    {
      "epoch": 0.2878464818763326,
      "grad_norm": 1.8279051780700684,
      "learning_rate": 4.7121535181236674e-05,
      "loss": 0.0679,
      "step": 135
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 4.7607574462890625,
      "learning_rate": 4.7014925373134335e-05,
      "loss": 0.175,
      "step": 140
    },
    {
      "epoch": 0.3091684434968017,
      "grad_norm": 5.6617937088012695,
      "learning_rate": 4.690831556503198e-05,
      "loss": 0.169,
      "step": 145
    },
    {
      "epoch": 0.31982942430703626,
      "grad_norm": 4.157877445220947,
      "learning_rate": 4.6801705756929645e-05,
      "loss": 0.1028,
      "step": 150
    },
    {
      "epoch": 0.3304904051172708,
      "grad_norm": 4.519536972045898,
      "learning_rate": 4.669509594882729e-05,
      "loss": 0.1641,
      "step": 155
    },
    {
      "epoch": 0.3411513859275053,
      "grad_norm": 3.5423667430877686,
      "learning_rate": 4.658848614072495e-05,
      "loss": 0.1506,
      "step": 160
    },
    {
      "epoch": 0.35181236673773986,
      "grad_norm": 3.687249183654785,
      "learning_rate": 4.64818763326226e-05,
      "loss": 0.1608,
      "step": 165
    },
    {
      "epoch": 0.3624733475479744,
      "grad_norm": 2.339153289794922,
      "learning_rate": 4.637526652452026e-05,
      "loss": 0.1634,
      "step": 170
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 2.263232707977295,
      "learning_rate": 4.626865671641791e-05,
      "loss": 0.09,
      "step": 175
    },
    {
      "epoch": 0.3837953091684435,
      "grad_norm": 3.2844831943511963,
      "learning_rate": 4.6162046908315566e-05,
      "loss": 0.1179,
      "step": 180
    },
    {
      "epoch": 0.39445628997867804,
      "grad_norm": 2.870607852935791,
      "learning_rate": 4.605543710021322e-05,
      "loss": 0.114,
      "step": 185
    },
    {
      "epoch": 0.4051172707889126,
      "grad_norm": 1.0485496520996094,
      "learning_rate": 4.5948827292110876e-05,
      "loss": 0.1216,
      "step": 190
    },
    {
      "epoch": 0.4157782515991471,
      "grad_norm": 2.7309136390686035,
      "learning_rate": 4.584221748400853e-05,
      "loss": 0.1242,
      "step": 195
    },
    {
      "epoch": 0.42643923240938164,
      "grad_norm": 5.80459451675415,
      "learning_rate": 4.5735607675906185e-05,
      "loss": 0.1027,
      "step": 200
    },
    {
      "epoch": 0.43710021321961623,
      "grad_norm": 2.2734932899475098,
      "learning_rate": 4.562899786780384e-05,
      "loss": 0.1902,
      "step": 205
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.6470862030982971,
      "learning_rate": 4.5522388059701495e-05,
      "loss": 0.0667,
      "step": 210
    },
    {
      "epoch": 0.4584221748400853,
      "grad_norm": 5.6269707679748535,
      "learning_rate": 4.541577825159915e-05,
      "loss": 0.1975,
      "step": 215
    },
    {
      "epoch": 0.4690831556503198,
      "grad_norm": 0.48877769708633423,
      "learning_rate": 4.5309168443496804e-05,
      "loss": 0.1044,
      "step": 220
    },
    {
      "epoch": 0.47974413646055436,
      "grad_norm": 4.694411277770996,
      "learning_rate": 4.520255863539446e-05,
      "loss": 0.12,
      "step": 225
    },
    {
      "epoch": 0.4904051172707889,
      "grad_norm": 4.776628017425537,
      "learning_rate": 4.5095948827292114e-05,
      "loss": 0.1086,
      "step": 230
    },
    {
      "epoch": 0.5010660980810234,
      "grad_norm": 6.282637119293213,
      "learning_rate": 4.498933901918977e-05,
      "loss": 0.1327,
      "step": 235
    },
    {
      "epoch": 0.511727078891258,
      "grad_norm": 4.1662068367004395,
      "learning_rate": 4.488272921108742e-05,
      "loss": 0.1005,
      "step": 240
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 4.968635559082031,
      "learning_rate": 4.477611940298508e-05,
      "loss": 0.1838,
      "step": 245
    },
    {
      "epoch": 0.5330490405117271,
      "grad_norm": 0.3055686056613922,
      "learning_rate": 4.466950959488273e-05,
      "loss": 0.1011,
      "step": 250
    },
    {
      "epoch": 0.5437100213219617,
      "grad_norm": 4.217792987823486,
      "learning_rate": 4.456289978678039e-05,
      "loss": 0.1358,
      "step": 255
    },
    {
      "epoch": 0.5543710021321961,
      "grad_norm": 1.9696346521377563,
      "learning_rate": 4.445628997867804e-05,
      "loss": 0.1381,
      "step": 260
    },
    {
      "epoch": 0.5650319829424307,
      "grad_norm": 2.3128340244293213,
      "learning_rate": 4.43496801705757e-05,
      "loss": 0.0912,
      "step": 265
    },
    {
      "epoch": 0.5756929637526652,
      "grad_norm": 2.890897750854492,
      "learning_rate": 4.424307036247335e-05,
      "loss": 0.12,
      "step": 270
    },
    {
      "epoch": 0.5863539445628998,
      "grad_norm": 3.5377447605133057,
      "learning_rate": 4.4136460554371006e-05,
      "loss": 0.0979,
      "step": 275
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 4.772866249084473,
      "learning_rate": 4.402985074626866e-05,
      "loss": 0.1766,
      "step": 280
    },
    {
      "epoch": 0.6076759061833689,
      "grad_norm": 5.789790630340576,
      "learning_rate": 4.3923240938166316e-05,
      "loss": 0.1423,
      "step": 285
    },
    {
      "epoch": 0.6183368869936035,
      "grad_norm": 3.299043893814087,
      "learning_rate": 4.381663113006397e-05,
      "loss": 0.1494,
      "step": 290
    },
    {
      "epoch": 0.6289978678038379,
      "grad_norm": 3.1073977947235107,
      "learning_rate": 4.3710021321961625e-05,
      "loss": 0.0983,
      "step": 295
    },
    {
      "epoch": 0.6396588486140725,
      "grad_norm": 2.49458384513855,
      "learning_rate": 4.360341151385928e-05,
      "loss": 0.1372,
      "step": 300
    },
    {
      "epoch": 0.650319829424307,
      "grad_norm": 2.902350664138794,
      "learning_rate": 4.3496801705756935e-05,
      "loss": 0.1692,
      "step": 305
    },
    {
      "epoch": 0.6609808102345416,
      "grad_norm": 6.115461349487305,
      "learning_rate": 4.339019189765459e-05,
      "loss": 0.0882,
      "step": 310
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 4.318145751953125,
      "learning_rate": 4.328358208955224e-05,
      "loss": 0.2206,
      "step": 315
    },
    {
      "epoch": 0.6823027718550106,
      "grad_norm": 2.468510150909424,
      "learning_rate": 4.31769722814499e-05,
      "loss": 0.118,
      "step": 320
    },
    {
      "epoch": 0.6929637526652452,
      "grad_norm": 1.1873846054077148,
      "learning_rate": 4.307036247334755e-05,
      "loss": 0.1259,
      "step": 325
    },
    {
      "epoch": 0.7036247334754797,
      "grad_norm": 3.2294986248016357,
      "learning_rate": 4.29637526652452e-05,
      "loss": 0.1348,
      "step": 330
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 6.70029878616333,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.2153,
      "step": 335
    },
    {
      "epoch": 0.7249466950959488,
      "grad_norm": 5.352694988250732,
      "learning_rate": 4.275053304904051e-05,
      "loss": 0.1241,
      "step": 340
    },
    {
      "epoch": 0.7356076759061834,
      "grad_norm": 3.538518190383911,
      "learning_rate": 4.2643923240938166e-05,
      "loss": 0.128,
      "step": 345
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 4.7478203773498535,
      "learning_rate": 4.253731343283582e-05,
      "loss": 0.1809,
      "step": 350
    },
    {
      "epoch": 0.7569296375266524,
      "grad_norm": 2.141639232635498,
      "learning_rate": 4.2430703624733475e-05,
      "loss": 0.1317,
      "step": 355
    },
    {
      "epoch": 0.767590618336887,
      "grad_norm": 2.5208847522735596,
      "learning_rate": 4.232409381663113e-05,
      "loss": 0.1192,
      "step": 360
    },
    {
      "epoch": 0.7782515991471215,
      "grad_norm": 3.1750762462615967,
      "learning_rate": 4.2217484008528785e-05,
      "loss": 0.1452,
      "step": 365
    },
    {
      "epoch": 0.7889125799573561,
      "grad_norm": 3.529797077178955,
      "learning_rate": 4.211087420042644e-05,
      "loss": 0.2516,
      "step": 370
    },
    {
      "epoch": 0.7995735607675906,
      "grad_norm": 2.614980936050415,
      "learning_rate": 4.2004264392324094e-05,
      "loss": 0.1302,
      "step": 375
    },
    {
      "epoch": 0.8102345415778252,
      "grad_norm": 3.8845324516296387,
      "learning_rate": 4.189765458422175e-05,
      "loss": 0.1198,
      "step": 380
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 4.767122268676758,
      "learning_rate": 4.1791044776119404e-05,
      "loss": 0.1348,
      "step": 385
    },
    {
      "epoch": 0.8315565031982942,
      "grad_norm": 2.994401216506958,
      "learning_rate": 4.168443496801706e-05,
      "loss": 0.2013,
      "step": 390
    },
    {
      "epoch": 0.8422174840085288,
      "grad_norm": 4.449457168579102,
      "learning_rate": 4.157782515991471e-05,
      "loss": 0.1619,
      "step": 395
    },
    {
      "epoch": 0.8528784648187633,
      "grad_norm": 5.16259765625,
      "learning_rate": 4.147121535181237e-05,
      "loss": 0.1106,
      "step": 400
    },
    {
      "epoch": 0.8635394456289979,
      "grad_norm": 3.81233811378479,
      "learning_rate": 4.136460554371002e-05,
      "loss": 0.1531,
      "step": 405
    },
    {
      "epoch": 0.8742004264392325,
      "grad_norm": 2.9617860317230225,
      "learning_rate": 4.125799573560768e-05,
      "loss": 0.111,
      "step": 410
    },
    {
      "epoch": 0.8848614072494669,
      "grad_norm": 4.039616584777832,
      "learning_rate": 4.115138592750533e-05,
      "loss": 0.1619,
      "step": 415
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 3.124743700027466,
      "learning_rate": 4.104477611940299e-05,
      "loss": 0.2112,
      "step": 420
    },
    {
      "epoch": 0.906183368869936,
      "grad_norm": 4.835186958312988,
      "learning_rate": 4.093816631130064e-05,
      "loss": 0.1188,
      "step": 425
    },
    {
      "epoch": 0.9168443496801706,
      "grad_norm": 3.2836201190948486,
      "learning_rate": 4.0831556503198296e-05,
      "loss": 0.1304,
      "step": 430
    },
    {
      "epoch": 0.9275053304904051,
      "grad_norm": 2.647538661956787,
      "learning_rate": 4.072494669509595e-05,
      "loss": 0.13,
      "step": 435
    },
    {
      "epoch": 0.9381663113006397,
      "grad_norm": 4.8179030418396,
      "learning_rate": 4.0618336886993606e-05,
      "loss": 0.1711,
      "step": 440
    },
    {
      "epoch": 0.9488272921108742,
      "grad_norm": 4.027265548706055,
      "learning_rate": 4.051172707889126e-05,
      "loss": 0.1073,
      "step": 445
    },
    {
      "epoch": 0.9594882729211087,
      "grad_norm": 1.4894267320632935,
      "learning_rate": 4.0405117270788915e-05,
      "loss": 0.0984,
      "step": 450
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 1.5578402280807495,
      "learning_rate": 4.029850746268657e-05,
      "loss": 0.1333,
      "step": 455
    },
    {
      "epoch": 0.9808102345415778,
      "grad_norm": 1.3624298572540283,
      "learning_rate": 4.0191897654584225e-05,
      "loss": 0.0515,
      "step": 460
    },
    {
      "epoch": 0.9914712153518124,
      "grad_norm": 4.03080415725708,
      "learning_rate": 4.008528784648188e-05,
      "loss": 0.1052,
      "step": 465
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9045333333333333,
      "eval_loss": 0.24142488837242126,
      "eval_runtime": 30.3224,
      "eval_samples_per_second": 123.671,
      "eval_steps_per_second": 3.892,
      "step": 469
    },
    {
      "epoch": 1.0021321961620469,
      "grad_norm": 7.205029010772705,
      "learning_rate": 3.997867803837953e-05,
      "loss": 0.189,
      "step": 470
    },
    {
      "epoch": 1.0127931769722816,
      "grad_norm": 2.2204198837280273,
      "learning_rate": 3.987206823027719e-05,
      "loss": 0.0822,
      "step": 475
    },
    {
      "epoch": 1.023454157782516,
      "grad_norm": 1.2866848707199097,
      "learning_rate": 3.976545842217484e-05,
      "loss": 0.1275,
      "step": 480
    },
    {
      "epoch": 1.0341151385927505,
      "grad_norm": 6.5878496170043945,
      "learning_rate": 3.96588486140725e-05,
      "loss": 0.2186,
      "step": 485
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 5.188156604766846,
      "learning_rate": 3.9552238805970146e-05,
      "loss": 0.0854,
      "step": 490
    },
    {
      "epoch": 1.0554371002132197,
      "grad_norm": 10.330496788024902,
      "learning_rate": 3.944562899786781e-05,
      "loss": 0.1507,
      "step": 495
    },
    {
      "epoch": 1.0660980810234542,
      "grad_norm": 3.284458637237549,
      "learning_rate": 3.9339019189765456e-05,
      "loss": 0.1562,
      "step": 500
    },
    {
      "epoch": 1.0767590618336886,
      "grad_norm": 0.40706029534339905,
      "learning_rate": 3.923240938166312e-05,
      "loss": 0.0471,
      "step": 505
    },
    {
      "epoch": 1.0874200426439233,
      "grad_norm": 6.002516269683838,
      "learning_rate": 3.9125799573560765e-05,
      "loss": 0.1202,
      "step": 510
    },
    {
      "epoch": 1.0980810234541578,
      "grad_norm": 3.994664192199707,
      "learning_rate": 3.901918976545843e-05,
      "loss": 0.1728,
      "step": 515
    },
    {
      "epoch": 1.1087420042643923,
      "grad_norm": 7.170152187347412,
      "learning_rate": 3.8912579957356075e-05,
      "loss": 0.084,
      "step": 520
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 4.095866680145264,
      "learning_rate": 3.8805970149253736e-05,
      "loss": 0.1053,
      "step": 525
    },
    {
      "epoch": 1.1300639658848615,
      "grad_norm": 3.0715484619140625,
      "learning_rate": 3.8699360341151384e-05,
      "loss": 0.1913,
      "step": 530
    },
    {
      "epoch": 1.140724946695096,
      "grad_norm": 6.858354568481445,
      "learning_rate": 3.8592750533049046e-05,
      "loss": 0.1349,
      "step": 535
    },
    {
      "epoch": 1.1513859275053304,
      "grad_norm": 1.1626945734024048,
      "learning_rate": 3.8486140724946694e-05,
      "loss": 0.1411,
      "step": 540
    },
    {
      "epoch": 1.1620469083155651,
      "grad_norm": 2.315717935562134,
      "learning_rate": 3.8379530916844355e-05,
      "loss": 0.0802,
      "step": 545
    },
    {
      "epoch": 1.1727078891257996,
      "grad_norm": 5.730618953704834,
      "learning_rate": 3.8272921108742e-05,
      "loss": 0.1233,
      "step": 550
    },
    {
      "epoch": 1.183368869936034,
      "grad_norm": 2.699887752532959,
      "learning_rate": 3.8166311300639665e-05,
      "loss": 0.0705,
      "step": 555
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 2.9159722328186035,
      "learning_rate": 3.805970149253731e-05,
      "loss": 0.1047,
      "step": 560
    },
    {
      "epoch": 1.2046908315565032,
      "grad_norm": 4.452578544616699,
      "learning_rate": 3.7953091684434974e-05,
      "loss": 0.1457,
      "step": 565
    },
    {
      "epoch": 1.2153518123667377,
      "grad_norm": 5.277554988861084,
      "learning_rate": 3.784648187633262e-05,
      "loss": 0.0768,
      "step": 570
    },
    {
      "epoch": 1.2260127931769722,
      "grad_norm": 4.552548408508301,
      "learning_rate": 3.7739872068230284e-05,
      "loss": 0.0988,
      "step": 575
    },
    {
      "epoch": 1.236673773987207,
      "grad_norm": 1.699558138847351,
      "learning_rate": 3.763326226012793e-05,
      "loss": 0.1051,
      "step": 580
    },
    {
      "epoch": 1.2473347547974414,
      "grad_norm": 3.6490185260772705,
      "learning_rate": 3.752665245202559e-05,
      "loss": 0.116,
      "step": 585
    },
    {
      "epoch": 1.2579957356076759,
      "grad_norm": 2.061978578567505,
      "learning_rate": 3.742004264392324e-05,
      "loss": 0.0812,
      "step": 590
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 1.3221495151519775,
      "learning_rate": 3.73134328358209e-05,
      "loss": 0.1751,
      "step": 595
    },
    {
      "epoch": 1.279317697228145,
      "grad_norm": 3.8583619594573975,
      "learning_rate": 3.720682302771855e-05,
      "loss": 0.1326,
      "step": 600
    },
    {
      "epoch": 1.2899786780383795,
      "grad_norm": 7.029452800750732,
      "learning_rate": 3.710021321961621e-05,
      "loss": 0.1488,
      "step": 605
    },
    {
      "epoch": 1.3006396588486142,
      "grad_norm": 3.937187433242798,
      "learning_rate": 3.699360341151386e-05,
      "loss": 0.1504,
      "step": 610
    },
    {
      "epoch": 1.3113006396588487,
      "grad_norm": 3.1681971549987793,
      "learning_rate": 3.6886993603411515e-05,
      "loss": 0.0442,
      "step": 615
    },
    {
      "epoch": 1.3219616204690832,
      "grad_norm": 2.180981159210205,
      "learning_rate": 3.678038379530917e-05,
      "loss": 0.1709,
      "step": 620
    },
    {
      "epoch": 1.3326226012793176,
      "grad_norm": 1.2517527341842651,
      "learning_rate": 3.6673773987206824e-05,
      "loss": 0.0931,
      "step": 625
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 3.145794153213501,
      "learning_rate": 3.656716417910448e-05,
      "loss": 0.1203,
      "step": 630
    },
    {
      "epoch": 1.3539445628997868,
      "grad_norm": 4.349611282348633,
      "learning_rate": 3.6460554371002134e-05,
      "loss": 0.1138,
      "step": 635
    },
    {
      "epoch": 1.3646055437100213,
      "grad_norm": 4.0500288009643555,
      "learning_rate": 3.635394456289979e-05,
      "loss": 0.1625,
      "step": 640
    },
    {
      "epoch": 1.375266524520256,
      "grad_norm": 3.1442723274230957,
      "learning_rate": 3.624733475479744e-05,
      "loss": 0.0791,
      "step": 645
    },
    {
      "epoch": 1.3859275053304905,
      "grad_norm": 3.396623373031616,
      "learning_rate": 3.61407249466951e-05,
      "loss": 0.0883,
      "step": 650
    },
    {
      "epoch": 1.396588486140725,
      "grad_norm": 4.072403430938721,
      "learning_rate": 3.603411513859275e-05,
      "loss": 0.1279,
      "step": 655
    },
    {
      "epoch": 1.4072494669509594,
      "grad_norm": 1.344158411026001,
      "learning_rate": 3.592750533049041e-05,
      "loss": 0.1241,
      "step": 660
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 5.487687587738037,
      "learning_rate": 3.582089552238806e-05,
      "loss": 0.1472,
      "step": 665
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 5.378997802734375,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.0656,
      "step": 670
    },
    {
      "epoch": 1.439232409381663,
      "grad_norm": 1.795943260192871,
      "learning_rate": 3.560767590618337e-05,
      "loss": 0.1115,
      "step": 675
    },
    {
      "epoch": 1.4498933901918978,
      "grad_norm": 3.2164461612701416,
      "learning_rate": 3.5501066098081026e-05,
      "loss": 0.0705,
      "step": 680
    },
    {
      "epoch": 1.4605543710021323,
      "grad_norm": 4.651003837585449,
      "learning_rate": 3.539445628997868e-05,
      "loss": 0.1601,
      "step": 685
    },
    {
      "epoch": 1.4712153518123667,
      "grad_norm": 2.309762477874756,
      "learning_rate": 3.5287846481876336e-05,
      "loss": 0.0672,
      "step": 690
    },
    {
      "epoch": 1.4818763326226012,
      "grad_norm": 3.5963847637176514,
      "learning_rate": 3.518123667377399e-05,
      "loss": 0.1564,
      "step": 695
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 3.1831014156341553,
      "learning_rate": 3.5074626865671645e-05,
      "loss": 0.1347,
      "step": 700
    },
    {
      "epoch": 1.5031982942430704,
      "grad_norm": 1.7468576431274414,
      "learning_rate": 3.496801705756929e-05,
      "loss": 0.0934,
      "step": 705
    },
    {
      "epoch": 1.5138592750533049,
      "grad_norm": 3.377796173095703,
      "learning_rate": 3.4861407249466955e-05,
      "loss": 0.065,
      "step": 710
    },
    {
      "epoch": 1.5245202558635396,
      "grad_norm": 2.401268243789673,
      "learning_rate": 3.47547974413646e-05,
      "loss": 0.0848,
      "step": 715
    },
    {
      "epoch": 1.535181236673774,
      "grad_norm": 6.083065032958984,
      "learning_rate": 3.4648187633262264e-05,
      "loss": 0.1744,
      "step": 720
    },
    {
      "epoch": 1.5458422174840085,
      "grad_norm": 4.216594219207764,
      "learning_rate": 3.454157782515991e-05,
      "loss": 0.1098,
      "step": 725
    },
    {
      "epoch": 1.556503198294243,
      "grad_norm": 1.0128761529922485,
      "learning_rate": 3.4434968017057574e-05,
      "loss": 0.0768,
      "step": 730
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 7.555728912353516,
      "learning_rate": 3.432835820895522e-05,
      "loss": 0.1188,
      "step": 735
    },
    {
      "epoch": 1.5778251599147122,
      "grad_norm": 5.113610744476318,
      "learning_rate": 3.422174840085288e-05,
      "loss": 0.1434,
      "step": 740
    },
    {
      "epoch": 1.5884861407249466,
      "grad_norm": 4.045500755310059,
      "learning_rate": 3.411513859275053e-05,
      "loss": 0.1177,
      "step": 745
    },
    {
      "epoch": 1.5991471215351813,
      "grad_norm": 1.5061864852905273,
      "learning_rate": 3.400852878464819e-05,
      "loss": 0.1334,
      "step": 750
    },
    {
      "epoch": 1.6098081023454158,
      "grad_norm": 2.7121193408966064,
      "learning_rate": 3.390191897654584e-05,
      "loss": 0.0811,
      "step": 755
    },
    {
      "epoch": 1.6204690831556503,
      "grad_norm": 4.889978885650635,
      "learning_rate": 3.37953091684435e-05,
      "loss": 0.1492,
      "step": 760
    },
    {
      "epoch": 1.6311300639658848,
      "grad_norm": 4.828644752502441,
      "learning_rate": 3.368869936034115e-05,
      "loss": 0.1187,
      "step": 765
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 3.8388304710388184,
      "learning_rate": 3.358208955223881e-05,
      "loss": 0.0945,
      "step": 770
    },
    {
      "epoch": 1.652452025586354,
      "grad_norm": 0.8402218222618103,
      "learning_rate": 3.347547974413646e-05,
      "loss": 0.0692,
      "step": 775
    },
    {
      "epoch": 1.6631130063965884,
      "grad_norm": 2.2439775466918945,
      "learning_rate": 3.336886993603412e-05,
      "loss": 0.0342,
      "step": 780
    },
    {
      "epoch": 1.6737739872068231,
      "grad_norm": 2.875929594039917,
      "learning_rate": 3.326226012793177e-05,
      "loss": 0.0654,
      "step": 785
    },
    {
      "epoch": 1.6844349680170576,
      "grad_norm": 1.4052488803863525,
      "learning_rate": 3.3155650319829424e-05,
      "loss": 0.1219,
      "step": 790
    },
    {
      "epoch": 1.695095948827292,
      "grad_norm": 7.4810919761657715,
      "learning_rate": 3.304904051172708e-05,
      "loss": 0.1172,
      "step": 795
    },
    {
      "epoch": 1.7057569296375266,
      "grad_norm": 4.929789066314697,
      "learning_rate": 3.294243070362473e-05,
      "loss": 0.0875,
      "step": 800
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 2.265211582183838,
      "learning_rate": 3.283582089552239e-05,
      "loss": 0.0748,
      "step": 805
    },
    {
      "epoch": 1.7270788912579957,
      "grad_norm": 3.814358711242676,
      "learning_rate": 3.272921108742004e-05,
      "loss": 0.0139,
      "step": 810
    },
    {
      "epoch": 1.7377398720682304,
      "grad_norm": 10.067119598388672,
      "learning_rate": 3.26226012793177e-05,
      "loss": 0.1305,
      "step": 815
    },
    {
      "epoch": 1.748400852878465,
      "grad_norm": 3.1651415824890137,
      "learning_rate": 3.251599147121535e-05,
      "loss": 0.1383,
      "step": 820
    },
    {
      "epoch": 1.7590618336886994,
      "grad_norm": 0.7367953062057495,
      "learning_rate": 3.240938166311301e-05,
      "loss": 0.0839,
      "step": 825
    },
    {
      "epoch": 1.7697228144989339,
      "grad_norm": 0.1356213241815567,
      "learning_rate": 3.230277185501066e-05,
      "loss": 0.0181,
      "step": 830
    },
    {
      "epoch": 1.7803837953091683,
      "grad_norm": 1.255947232246399,
      "learning_rate": 3.2196162046908317e-05,
      "loss": 0.0633,
      "step": 835
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 3.9989399909973145,
      "learning_rate": 3.208955223880597e-05,
      "loss": 0.1973,
      "step": 840
    },
    {
      "epoch": 1.8017057569296375,
      "grad_norm": 7.870970726013184,
      "learning_rate": 3.1982942430703626e-05,
      "loss": 0.1317,
      "step": 845
    },
    {
      "epoch": 1.8123667377398722,
      "grad_norm": 2.217454433441162,
      "learning_rate": 3.187633262260128e-05,
      "loss": 0.0944,
      "step": 850
    },
    {
      "epoch": 1.8230277185501067,
      "grad_norm": 4.237931251525879,
      "learning_rate": 3.1769722814498935e-05,
      "loss": 0.0652,
      "step": 855
    },
    {
      "epoch": 1.8336886993603412,
      "grad_norm": 2.9334123134613037,
      "learning_rate": 3.166311300639659e-05,
      "loss": 0.0644,
      "step": 860
    },
    {
      "epoch": 1.8443496801705757,
      "grad_norm": 0.1925787329673767,
      "learning_rate": 3.1556503198294245e-05,
      "loss": 0.103,
      "step": 865
    },
    {
      "epoch": 1.8550106609808101,
      "grad_norm": 1.2077339887619019,
      "learning_rate": 3.14498933901919e-05,
      "loss": 0.0792,
      "step": 870
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 6.168605327606201,
      "learning_rate": 3.1343283582089554e-05,
      "loss": 0.0988,
      "step": 875
    },
    {
      "epoch": 1.8763326226012793,
      "grad_norm": 3.769495964050293,
      "learning_rate": 3.123667377398721e-05,
      "loss": 0.1281,
      "step": 880
    },
    {
      "epoch": 1.886993603411514,
      "grad_norm": 4.454564571380615,
      "learning_rate": 3.1130063965884864e-05,
      "loss": 0.0653,
      "step": 885
    },
    {
      "epoch": 1.8976545842217485,
      "grad_norm": 4.036810398101807,
      "learning_rate": 3.102345415778252e-05,
      "loss": 0.0763,
      "step": 890
    },
    {
      "epoch": 1.908315565031983,
      "grad_norm": 1.410136342048645,
      "learning_rate": 3.0916844349680173e-05,
      "loss": 0.0813,
      "step": 895
    },
    {
      "epoch": 1.9189765458422174,
      "grad_norm": 3.766080617904663,
      "learning_rate": 3.081023454157783e-05,
      "loss": 0.1819,
      "step": 900
    },
    {
      "epoch": 1.929637526652452,
      "grad_norm": 2.0473406314849854,
      "learning_rate": 3.070362473347548e-05,
      "loss": 0.0906,
      "step": 905
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 1.8353604078292847,
      "learning_rate": 3.059701492537314e-05,
      "loss": 0.1202,
      "step": 910
    },
    {
      "epoch": 1.950959488272921,
      "grad_norm": 3.074885606765747,
      "learning_rate": 3.0490405117270792e-05,
      "loss": 0.1489,
      "step": 915
    },
    {
      "epoch": 1.9616204690831558,
      "grad_norm": 0.35776641964912415,
      "learning_rate": 3.0383795309168444e-05,
      "loss": 0.1441,
      "step": 920
    },
    {
      "epoch": 1.9722814498933903,
      "grad_norm": 1.4147385358810425,
      "learning_rate": 3.0277185501066102e-05,
      "loss": 0.0749,
      "step": 925
    },
    {
      "epoch": 1.9829424307036247,
      "grad_norm": 0.10012034326791763,
      "learning_rate": 3.0170575692963753e-05,
      "loss": 0.0645,
      "step": 930
    },
    {
      "epoch": 1.9936034115138592,
      "grad_norm": 6.8047966957092285,
      "learning_rate": 3.006396588486141e-05,
      "loss": 0.1223,
      "step": 935
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8925333333333333,
      "eval_loss": 0.32056963443756104,
      "eval_runtime": 30.0633,
      "eval_samples_per_second": 124.737,
      "eval_steps_per_second": 3.925,
      "step": 938
    },
    {
      "epoch": 2.0042643923240937,
      "grad_norm": 5.572512626647949,
      "learning_rate": 2.9957356076759063e-05,
      "loss": 0.1118,
      "step": 940
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 8.231983184814453,
      "learning_rate": 2.9850746268656714e-05,
      "loss": 0.1465,
      "step": 945
    },
    {
      "epoch": 2.025586353944563,
      "grad_norm": 3.5708682537078857,
      "learning_rate": 2.9744136460554372e-05,
      "loss": 0.1045,
      "step": 950
    },
    {
      "epoch": 2.0362473347547976,
      "grad_norm": 1.6533825397491455,
      "learning_rate": 2.9637526652452023e-05,
      "loss": 0.1178,
      "step": 955
    },
    {
      "epoch": 2.046908315565032,
      "grad_norm": 2.11965012550354,
      "learning_rate": 2.953091684434968e-05,
      "loss": 0.0821,
      "step": 960
    },
    {
      "epoch": 2.0575692963752665,
      "grad_norm": 2.5297791957855225,
      "learning_rate": 2.9424307036247333e-05,
      "loss": 0.06,
      "step": 965
    },
    {
      "epoch": 2.068230277185501,
      "grad_norm": 0.42102810740470886,
      "learning_rate": 2.931769722814499e-05,
      "loss": 0.1029,
      "step": 970
    },
    {
      "epoch": 2.0788912579957355,
      "grad_norm": 4.654422760009766,
      "learning_rate": 2.9211087420042642e-05,
      "loss": 0.0813,
      "step": 975
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 4.053927898406982,
      "learning_rate": 2.91044776119403e-05,
      "loss": 0.06,
      "step": 980
    },
    {
      "epoch": 2.100213219616205,
      "grad_norm": 0.2010262906551361,
      "learning_rate": 2.8997867803837952e-05,
      "loss": 0.0327,
      "step": 985
    },
    {
      "epoch": 2.1108742004264394,
      "grad_norm": 4.605628967285156,
      "learning_rate": 2.889125799573561e-05,
      "loss": 0.0843,
      "step": 990
    },
    {
      "epoch": 2.121535181236674,
      "grad_norm": 2.55084490776062,
      "learning_rate": 2.878464818763326e-05,
      "loss": 0.0955,
      "step": 995
    },
    {
      "epoch": 2.1321961620469083,
      "grad_norm": 0.5406118035316467,
      "learning_rate": 2.867803837953092e-05,
      "loss": 0.0939,
      "step": 1000
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 1.772225260734558,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0714,
      "step": 1005
    },
    {
      "epoch": 2.1535181236673773,
      "grad_norm": 2.1154720783233643,
      "learning_rate": 2.846481876332623e-05,
      "loss": 0.1357,
      "step": 1010
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 2.614042043685913,
      "learning_rate": 2.835820895522388e-05,
      "loss": 0.0561,
      "step": 1015
    },
    {
      "epoch": 2.1748400852878467,
      "grad_norm": 0.18544137477874756,
      "learning_rate": 2.825159914712154e-05,
      "loss": 0.0881,
      "step": 1020
    },
    {
      "epoch": 2.185501066098081,
      "grad_norm": 1.5250552892684937,
      "learning_rate": 2.814498933901919e-05,
      "loss": 0.0995,
      "step": 1025
    },
    {
      "epoch": 2.1961620469083156,
      "grad_norm": 4.3390092849731445,
      "learning_rate": 2.8038379530916848e-05,
      "loss": 0.107,
      "step": 1030
    },
    {
      "epoch": 2.20682302771855,
      "grad_norm": 2.451718807220459,
      "learning_rate": 2.79317697228145e-05,
      "loss": 0.1038,
      "step": 1035
    },
    {
      "epoch": 2.2174840085287846,
      "grad_norm": 9.413724899291992,
      "learning_rate": 2.7825159914712157e-05,
      "loss": 0.1283,
      "step": 1040
    },
    {
      "epoch": 2.228144989339019,
      "grad_norm": 0.6464016437530518,
      "learning_rate": 2.771855010660981e-05,
      "loss": 0.0968,
      "step": 1045
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 1.6108185052871704,
      "learning_rate": 2.7611940298507467e-05,
      "loss": 0.0631,
      "step": 1050
    },
    {
      "epoch": 2.2494669509594885,
      "grad_norm": 1.4787224531173706,
      "learning_rate": 2.7505330490405118e-05,
      "loss": 0.0657,
      "step": 1055
    },
    {
      "epoch": 2.260127931769723,
      "grad_norm": 4.048679828643799,
      "learning_rate": 2.7398720682302776e-05,
      "loss": 0.1134,
      "step": 1060
    },
    {
      "epoch": 2.2707889125799574,
      "grad_norm": 1.742465615272522,
      "learning_rate": 2.7292110874200428e-05,
      "loss": 0.0967,
      "step": 1065
    },
    {
      "epoch": 2.281449893390192,
      "grad_norm": 1.4644908905029297,
      "learning_rate": 2.7185501066098086e-05,
      "loss": 0.1542,
      "step": 1070
    },
    {
      "epoch": 2.2921108742004264,
      "grad_norm": 2.8316097259521484,
      "learning_rate": 2.7078891257995737e-05,
      "loss": 0.0796,
      "step": 1075
    },
    {
      "epoch": 2.302771855010661,
      "grad_norm": 0.2534070611000061,
      "learning_rate": 2.6972281449893395e-05,
      "loss": 0.1124,
      "step": 1080
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 0.7657762765884399,
      "learning_rate": 2.6865671641791047e-05,
      "loss": 0.0544,
      "step": 1085
    },
    {
      "epoch": 2.3240938166311302,
      "grad_norm": 2.5004377365112305,
      "learning_rate": 2.6759061833688705e-05,
      "loss": 0.0539,
      "step": 1090
    },
    {
      "epoch": 2.3347547974413647,
      "grad_norm": 1.8675498962402344,
      "learning_rate": 2.6652452025586356e-05,
      "loss": 0.0844,
      "step": 1095
    },
    {
      "epoch": 2.345415778251599,
      "grad_norm": 2.6942827701568604,
      "learning_rate": 2.6545842217484007e-05,
      "loss": 0.0635,
      "step": 1100
    },
    {
      "epoch": 2.3560767590618337,
      "grad_norm": 0.1736178696155548,
      "learning_rate": 2.6439232409381666e-05,
      "loss": 0.0414,
      "step": 1105
    },
    {
      "epoch": 2.366737739872068,
      "grad_norm": 0.9560030698776245,
      "learning_rate": 2.6332622601279317e-05,
      "loss": 0.0901,
      "step": 1110
    },
    {
      "epoch": 2.3773987206823026,
      "grad_norm": 1.4020416736602783,
      "learning_rate": 2.6226012793176975e-05,
      "loss": 0.0773,
      "step": 1115
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 1.4743791818618774,
      "learning_rate": 2.6119402985074626e-05,
      "loss": 0.0782,
      "step": 1120
    },
    {
      "epoch": 2.398720682302772,
      "grad_norm": 1.2656854391098022,
      "learning_rate": 2.6012793176972285e-05,
      "loss": 0.0559,
      "step": 1125
    },
    {
      "epoch": 2.4093816631130065,
      "grad_norm": 0.14698302745819092,
      "learning_rate": 2.5906183368869936e-05,
      "loss": 0.1535,
      "step": 1130
    },
    {
      "epoch": 2.420042643923241,
      "grad_norm": 3.3647916316986084,
      "learning_rate": 2.5799573560767594e-05,
      "loss": 0.08,
      "step": 1135
    },
    {
      "epoch": 2.4307036247334755,
      "grad_norm": 2.227022409439087,
      "learning_rate": 2.5692963752665245e-05,
      "loss": 0.1087,
      "step": 1140
    },
    {
      "epoch": 2.44136460554371,
      "grad_norm": 1.343688726425171,
      "learning_rate": 2.5586353944562904e-05,
      "loss": 0.0735,
      "step": 1145
    },
    {
      "epoch": 2.4520255863539444,
      "grad_norm": 1.3545268774032593,
      "learning_rate": 2.5479744136460555e-05,
      "loss": 0.0258,
      "step": 1150
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 5.5803093910217285,
      "learning_rate": 2.537313432835821e-05,
      "loss": 0.143,
      "step": 1155
    },
    {
      "epoch": 2.473347547974414,
      "grad_norm": 5.855179309844971,
      "learning_rate": 2.5266524520255864e-05,
      "loss": 0.1105,
      "step": 1160
    },
    {
      "epoch": 2.4840085287846483,
      "grad_norm": 4.827827453613281,
      "learning_rate": 2.515991471215352e-05,
      "loss": 0.084,
      "step": 1165
    },
    {
      "epoch": 2.4946695095948828,
      "grad_norm": 5.6794753074646,
      "learning_rate": 2.5053304904051174e-05,
      "loss": 0.0789,
      "step": 1170
    },
    {
      "epoch": 2.5053304904051172,
      "grad_norm": 1.1285988092422485,
      "learning_rate": 2.494669509594883e-05,
      "loss": 0.084,
      "step": 1175
    },
    {
      "epoch": 2.5159914712153517,
      "grad_norm": 2.8257339000701904,
      "learning_rate": 2.4840085287846483e-05,
      "loss": 0.0823,
      "step": 1180
    },
    {
      "epoch": 2.526652452025586,
      "grad_norm": 1.478042721748352,
      "learning_rate": 2.4733475479744138e-05,
      "loss": 0.0773,
      "step": 1185
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 1.649285078048706,
      "learning_rate": 2.4626865671641793e-05,
      "loss": 0.1104,
      "step": 1190
    },
    {
      "epoch": 2.5479744136460556,
      "grad_norm": 6.218127250671387,
      "learning_rate": 2.4520255863539444e-05,
      "loss": 0.1224,
      "step": 1195
    },
    {
      "epoch": 2.55863539445629,
      "grad_norm": 1.5206996202468872,
      "learning_rate": 2.44136460554371e-05,
      "loss": 0.068,
      "step": 1200
    },
    {
      "epoch": 2.5692963752665245,
      "grad_norm": 4.296855449676514,
      "learning_rate": 2.4307036247334754e-05,
      "loss": 0.0755,
      "step": 1205
    },
    {
      "epoch": 2.579957356076759,
      "grad_norm": 2.672220230102539,
      "learning_rate": 2.420042643923241e-05,
      "loss": 0.0931,
      "step": 1210
    },
    {
      "epoch": 2.5906183368869935,
      "grad_norm": 2.795624256134033,
      "learning_rate": 2.4093816631130063e-05,
      "loss": 0.057,
      "step": 1215
    },
    {
      "epoch": 2.6012793176972284,
      "grad_norm": 1.5376272201538086,
      "learning_rate": 2.3987206823027718e-05,
      "loss": 0.0346,
      "step": 1220
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 2.280195474624634,
      "learning_rate": 2.3880597014925373e-05,
      "loss": 0.0681,
      "step": 1225
    },
    {
      "epoch": 2.6226012793176974,
      "grad_norm": 0.4716942608356476,
      "learning_rate": 2.3773987206823027e-05,
      "loss": 0.0341,
      "step": 1230
    },
    {
      "epoch": 2.633262260127932,
      "grad_norm": 4.179445266723633,
      "learning_rate": 2.3667377398720682e-05,
      "loss": 0.0737,
      "step": 1235
    },
    {
      "epoch": 2.6439232409381663,
      "grad_norm": 8.277469635009766,
      "learning_rate": 2.3560767590618337e-05,
      "loss": 0.1042,
      "step": 1240
    },
    {
      "epoch": 2.654584221748401,
      "grad_norm": 0.29547402262687683,
      "learning_rate": 2.345415778251599e-05,
      "loss": 0.0617,
      "step": 1245
    },
    {
      "epoch": 2.6652452025586353,
      "grad_norm": 4.500633716583252,
      "learning_rate": 2.3347547974413646e-05,
      "loss": 0.0581,
      "step": 1250
    },
    {
      "epoch": 2.6759061833688698,
      "grad_norm": 1.5811517238616943,
      "learning_rate": 2.32409381663113e-05,
      "loss": 0.0251,
      "step": 1255
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 3.72788143157959,
      "learning_rate": 2.3134328358208956e-05,
      "loss": 0.1252,
      "step": 1260
    },
    {
      "epoch": 2.697228144989339,
      "grad_norm": 0.5226966738700867,
      "learning_rate": 2.302771855010661e-05,
      "loss": 0.1247,
      "step": 1265
    },
    {
      "epoch": 2.7078891257995736,
      "grad_norm": 5.085799694061279,
      "learning_rate": 2.2921108742004265e-05,
      "loss": 0.0892,
      "step": 1270
    },
    {
      "epoch": 2.718550106609808,
      "grad_norm": 4.691335201263428,
      "learning_rate": 2.281449893390192e-05,
      "loss": 0.1249,
      "step": 1275
    },
    {
      "epoch": 2.7292110874200426,
      "grad_norm": 6.774505615234375,
      "learning_rate": 2.2707889125799575e-05,
      "loss": 0.1388,
      "step": 1280
    },
    {
      "epoch": 2.739872068230277,
      "grad_norm": 4.209772109985352,
      "learning_rate": 2.260127931769723e-05,
      "loss": 0.0764,
      "step": 1285
    },
    {
      "epoch": 2.750533049040512,
      "grad_norm": 5.686208248138428,
      "learning_rate": 2.2494669509594884e-05,
      "loss": 0.1428,
      "step": 1290
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 2.8803625106811523,
      "learning_rate": 2.238805970149254e-05,
      "loss": 0.0477,
      "step": 1295
    },
    {
      "epoch": 2.771855010660981,
      "grad_norm": 0.8628917932510376,
      "learning_rate": 2.2281449893390194e-05,
      "loss": 0.0628,
      "step": 1300
    },
    {
      "epoch": 2.7825159914712154,
      "grad_norm": 7.57857608795166,
      "learning_rate": 2.217484008528785e-05,
      "loss": 0.0972,
      "step": 1305
    },
    {
      "epoch": 2.79317697228145,
      "grad_norm": 6.631232738494873,
      "learning_rate": 2.2068230277185503e-05,
      "loss": 0.1018,
      "step": 1310
    },
    {
      "epoch": 2.8038379530916844,
      "grad_norm": 2.567697525024414,
      "learning_rate": 2.1961620469083158e-05,
      "loss": 0.0589,
      "step": 1315
    },
    {
      "epoch": 2.814498933901919,
      "grad_norm": 3.6537904739379883,
      "learning_rate": 2.1855010660980813e-05,
      "loss": 0.0953,
      "step": 1320
    },
    {
      "epoch": 2.8251599147121533,
      "grad_norm": 4.086033821105957,
      "learning_rate": 2.1748400852878467e-05,
      "loss": 0.0576,
      "step": 1325
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 5.503276824951172,
      "learning_rate": 2.164179104477612e-05,
      "loss": 0.1,
      "step": 1330
    },
    {
      "epoch": 2.8464818763326227,
      "grad_norm": 0.1452469676733017,
      "learning_rate": 2.1535181236673773e-05,
      "loss": 0.0551,
      "step": 1335
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 4.335849285125732,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.1458,
      "step": 1340
    },
    {
      "epoch": 2.8678038379530917,
      "grad_norm": 1.4450907707214355,
      "learning_rate": 2.1321961620469083e-05,
      "loss": 0.0377,
      "step": 1345
    },
    {
      "epoch": 2.878464818763326,
      "grad_norm": 1.8356298208236694,
      "learning_rate": 2.1215351812366738e-05,
      "loss": 0.0209,
      "step": 1350
    },
    {
      "epoch": 2.8891257995735606,
      "grad_norm": 5.640768051147461,
      "learning_rate": 2.1108742004264392e-05,
      "loss": 0.1146,
      "step": 1355
    },
    {
      "epoch": 2.8997867803837956,
      "grad_norm": 1.5563876628875732,
      "learning_rate": 2.1002132196162047e-05,
      "loss": 0.1677,
      "step": 1360
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 4.564480304718018,
      "learning_rate": 2.0895522388059702e-05,
      "loss": 0.1178,
      "step": 1365
    },
    {
      "epoch": 2.9211087420042645,
      "grad_norm": 3.673280715942383,
      "learning_rate": 2.0788912579957357e-05,
      "loss": 0.0719,
      "step": 1370
    },
    {
      "epoch": 2.931769722814499,
      "grad_norm": 3.1112773418426514,
      "learning_rate": 2.068230277185501e-05,
      "loss": 0.1587,
      "step": 1375
    },
    {
      "epoch": 2.9424307036247335,
      "grad_norm": 0.9399316906929016,
      "learning_rate": 2.0575692963752666e-05,
      "loss": 0.0859,
      "step": 1380
    },
    {
      "epoch": 2.953091684434968,
      "grad_norm": 4.814915657043457,
      "learning_rate": 2.046908315565032e-05,
      "loss": 0.0935,
      "step": 1385
    },
    {
      "epoch": 2.9637526652452024,
      "grad_norm": 5.766204357147217,
      "learning_rate": 2.0362473347547976e-05,
      "loss": 0.1712,
      "step": 1390
    },
    {
      "epoch": 2.974413646055437,
      "grad_norm": 0.9540748596191406,
      "learning_rate": 2.025586353944563e-05,
      "loss": 0.0408,
      "step": 1395
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 3.0130481719970703,
      "learning_rate": 2.0149253731343285e-05,
      "loss": 0.058,
      "step": 1400
    },
    {
      "epoch": 2.9957356076759063,
      "grad_norm": 3.3353872299194336,
      "learning_rate": 2.004264392324094e-05,
      "loss": 0.0798,
      "step": 1405
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.916,
      "eval_loss": 0.25396573543548584,
      "eval_runtime": 30.3888,
      "eval_samples_per_second": 123.401,
      "eval_steps_per_second": 3.883,
      "step": 1407
    },
    {
      "epoch": 3.0063965884861408,
      "grad_norm": 0.1786058396100998,
      "learning_rate": 1.9936034115138594e-05,
      "loss": 0.0913,
      "step": 1410
    },
    {
      "epoch": 3.0170575692963753,
      "grad_norm": 1.1213440895080566,
      "learning_rate": 1.982942430703625e-05,
      "loss": 0.0768,
      "step": 1415
    },
    {
      "epoch": 3.0277185501066097,
      "grad_norm": 0.8297101855278015,
      "learning_rate": 1.9722814498933904e-05,
      "loss": 0.0237,
      "step": 1420
    },
    {
      "epoch": 3.038379530916844,
      "grad_norm": 0.5259168744087219,
      "learning_rate": 1.961620469083156e-05,
      "loss": 0.0346,
      "step": 1425
    },
    {
      "epoch": 3.0490405117270787,
      "grad_norm": 0.28382739424705505,
      "learning_rate": 1.9509594882729213e-05,
      "loss": 0.1034,
      "step": 1430
    },
    {
      "epoch": 3.0597014925373136,
      "grad_norm": 3.1283106803894043,
      "learning_rate": 1.9402985074626868e-05,
      "loss": 0.0425,
      "step": 1435
    },
    {
      "epoch": 3.070362473347548,
      "grad_norm": 6.615086078643799,
      "learning_rate": 1.9296375266524523e-05,
      "loss": 0.0581,
      "step": 1440
    },
    {
      "epoch": 3.0810234541577826,
      "grad_norm": 7.611501693725586,
      "learning_rate": 1.9189765458422178e-05,
      "loss": 0.1232,
      "step": 1445
    },
    {
      "epoch": 3.091684434968017,
      "grad_norm": 4.48750114440918,
      "learning_rate": 1.9083155650319832e-05,
      "loss": 0.0448,
      "step": 1450
    },
    {
      "epoch": 3.1023454157782515,
      "grad_norm": 3.1024112701416016,
      "learning_rate": 1.8976545842217487e-05,
      "loss": 0.1106,
      "step": 1455
    },
    {
      "epoch": 3.113006396588486,
      "grad_norm": 3.742947816848755,
      "learning_rate": 1.8869936034115142e-05,
      "loss": 0.0582,
      "step": 1460
    },
    {
      "epoch": 3.1236673773987205,
      "grad_norm": 0.9130271077156067,
      "learning_rate": 1.8763326226012797e-05,
      "loss": 0.0275,
      "step": 1465
    },
    {
      "epoch": 3.1343283582089554,
      "grad_norm": 3.0800912380218506,
      "learning_rate": 1.865671641791045e-05,
      "loss": 0.0536,
      "step": 1470
    },
    {
      "epoch": 3.14498933901919,
      "grad_norm": 0.6095083355903625,
      "learning_rate": 1.8550106609808106e-05,
      "loss": 0.1002,
      "step": 1475
    },
    {
      "epoch": 3.1556503198294243,
      "grad_norm": 6.055989742279053,
      "learning_rate": 1.8443496801705757e-05,
      "loss": 0.1311,
      "step": 1480
    },
    {
      "epoch": 3.166311300639659,
      "grad_norm": 0.4931339621543884,
      "learning_rate": 1.8336886993603412e-05,
      "loss": 0.0443,
      "step": 1485
    },
    {
      "epoch": 3.1769722814498933,
      "grad_norm": 5.629963397979736,
      "learning_rate": 1.8230277185501067e-05,
      "loss": 0.0447,
      "step": 1490
    },
    {
      "epoch": 3.1876332622601278,
      "grad_norm": 6.491460800170898,
      "learning_rate": 1.812366737739872e-05,
      "loss": 0.0684,
      "step": 1495
    },
    {
      "epoch": 3.1982942430703627,
      "grad_norm": 1.7566626071929932,
      "learning_rate": 1.8017057569296376e-05,
      "loss": 0.0432,
      "step": 1500
    },
    {
      "epoch": 3.208955223880597,
      "grad_norm": 5.506409645080566,
      "learning_rate": 1.791044776119403e-05,
      "loss": 0.0417,
      "step": 1505
    },
    {
      "epoch": 3.2196162046908317,
      "grad_norm": 0.7198337912559509,
      "learning_rate": 1.7803837953091686e-05,
      "loss": 0.0937,
      "step": 1510
    },
    {
      "epoch": 3.230277185501066,
      "grad_norm": 7.328074932098389,
      "learning_rate": 1.769722814498934e-05,
      "loss": 0.0605,
      "step": 1515
    },
    {
      "epoch": 3.2409381663113006,
      "grad_norm": 8.071795463562012,
      "learning_rate": 1.7590618336886995e-05,
      "loss": 0.0431,
      "step": 1520
    },
    {
      "epoch": 3.251599147121535,
      "grad_norm": 2.7167623043060303,
      "learning_rate": 1.7484008528784647e-05,
      "loss": 0.0579,
      "step": 1525
    },
    {
      "epoch": 3.2622601279317696,
      "grad_norm": 3.20859432220459,
      "learning_rate": 1.73773987206823e-05,
      "loss": 0.0515,
      "step": 1530
    },
    {
      "epoch": 3.272921108742004,
      "grad_norm": 9.579339981079102,
      "learning_rate": 1.7270788912579956e-05,
      "loss": 0.1085,
      "step": 1535
    },
    {
      "epoch": 3.283582089552239,
      "grad_norm": 1.0331966876983643,
      "learning_rate": 1.716417910447761e-05,
      "loss": 0.0817,
      "step": 1540
    },
    {
      "epoch": 3.2942430703624734,
      "grad_norm": 6.120692253112793,
      "learning_rate": 1.7057569296375266e-05,
      "loss": 0.0601,
      "step": 1545
    },
    {
      "epoch": 3.304904051172708,
      "grad_norm": 2.222625255584717,
      "learning_rate": 1.695095948827292e-05,
      "loss": 0.0628,
      "step": 1550
    },
    {
      "epoch": 3.3155650319829424,
      "grad_norm": 3.1484971046447754,
      "learning_rate": 1.6844349680170575e-05,
      "loss": 0.0429,
      "step": 1555
    },
    {
      "epoch": 3.326226012793177,
      "grad_norm": 0.2928318977355957,
      "learning_rate": 1.673773987206823e-05,
      "loss": 0.0416,
      "step": 1560
    },
    {
      "epoch": 3.3368869936034113,
      "grad_norm": 1.8569515943527222,
      "learning_rate": 1.6631130063965885e-05,
      "loss": 0.043,
      "step": 1565
    },
    {
      "epoch": 3.3475479744136463,
      "grad_norm": 0.4534006714820862,
      "learning_rate": 1.652452025586354e-05,
      "loss": 0.0473,
      "step": 1570
    },
    {
      "epoch": 3.3582089552238807,
      "grad_norm": 2.857851982116699,
      "learning_rate": 1.6417910447761194e-05,
      "loss": 0.0529,
      "step": 1575
    },
    {
      "epoch": 3.368869936034115,
      "grad_norm": 7.311967849731445,
      "learning_rate": 1.631130063965885e-05,
      "loss": 0.0619,
      "step": 1580
    },
    {
      "epoch": 3.3795309168443497,
      "grad_norm": 0.1349204182624817,
      "learning_rate": 1.6204690831556504e-05,
      "loss": 0.0823,
      "step": 1585
    },
    {
      "epoch": 3.390191897654584,
      "grad_norm": 0.27863219380378723,
      "learning_rate": 1.6098081023454158e-05,
      "loss": 0.0748,
      "step": 1590
    },
    {
      "epoch": 3.4008528784648187,
      "grad_norm": 5.409698009490967,
      "learning_rate": 1.5991471215351813e-05,
      "loss": 0.0803,
      "step": 1595
    },
    {
      "epoch": 3.411513859275053,
      "grad_norm": 4.55714750289917,
      "learning_rate": 1.5884861407249468e-05,
      "loss": 0.0377,
      "step": 1600
    },
    {
      "epoch": 3.4221748400852876,
      "grad_norm": 6.741073131561279,
      "learning_rate": 1.5778251599147122e-05,
      "loss": 0.0871,
      "step": 1605
    },
    {
      "epoch": 3.4328358208955225,
      "grad_norm": 2.6364269256591797,
      "learning_rate": 1.5671641791044777e-05,
      "loss": 0.0652,
      "step": 1610
    },
    {
      "epoch": 3.443496801705757,
      "grad_norm": 0.656776487827301,
      "learning_rate": 1.5565031982942432e-05,
      "loss": 0.0206,
      "step": 1615
    },
    {
      "epoch": 3.4541577825159915,
      "grad_norm": 0.3795584440231323,
      "learning_rate": 1.5458422174840087e-05,
      "loss": 0.0223,
      "step": 1620
    },
    {
      "epoch": 3.464818763326226,
      "grad_norm": 0.2124588042497635,
      "learning_rate": 1.535181236673774e-05,
      "loss": 0.0389,
      "step": 1625
    },
    {
      "epoch": 3.4754797441364604,
      "grad_norm": 3.543072462081909,
      "learning_rate": 1.5245202558635396e-05,
      "loss": 0.0966,
      "step": 1630
    },
    {
      "epoch": 3.486140724946695,
      "grad_norm": 2.0806832313537598,
      "learning_rate": 1.5138592750533051e-05,
      "loss": 0.1616,
      "step": 1635
    },
    {
      "epoch": 3.49680170575693,
      "grad_norm": 1.8856909275054932,
      "learning_rate": 1.5031982942430706e-05,
      "loss": 0.0764,
      "step": 1640
    },
    {
      "epoch": 3.5074626865671643,
      "grad_norm": 3.8390161991119385,
      "learning_rate": 1.4925373134328357e-05,
      "loss": 0.0622,
      "step": 1645
    },
    {
      "epoch": 3.518123667377399,
      "grad_norm": 1.3755427598953247,
      "learning_rate": 1.4818763326226012e-05,
      "loss": 0.0148,
      "step": 1650
    },
    {
      "epoch": 3.5287846481876333,
      "grad_norm": 0.15994979441165924,
      "learning_rate": 1.4712153518123666e-05,
      "loss": 0.0606,
      "step": 1655
    },
    {
      "epoch": 3.5394456289978677,
      "grad_norm": 0.2989632487297058,
      "learning_rate": 1.4605543710021321e-05,
      "loss": 0.0355,
      "step": 1660
    },
    {
      "epoch": 3.550106609808102,
      "grad_norm": 0.05901180952787399,
      "learning_rate": 1.4498933901918976e-05,
      "loss": 0.0223,
      "step": 1665
    },
    {
      "epoch": 3.5607675906183367,
      "grad_norm": 3.636889696121216,
      "learning_rate": 1.439232409381663e-05,
      "loss": 0.0634,
      "step": 1670
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 4.303390979766846,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.1108,
      "step": 1675
    },
    {
      "epoch": 3.582089552238806,
      "grad_norm": 2.4064323902130127,
      "learning_rate": 1.417910447761194e-05,
      "loss": 0.0647,
      "step": 1680
    },
    {
      "epoch": 3.5927505330490406,
      "grad_norm": 6.4799628257751465,
      "learning_rate": 1.4072494669509595e-05,
      "loss": 0.0525,
      "step": 1685
    },
    {
      "epoch": 3.603411513859275,
      "grad_norm": 4.180606365203857,
      "learning_rate": 1.396588486140725e-05,
      "loss": 0.11,
      "step": 1690
    },
    {
      "epoch": 3.6140724946695095,
      "grad_norm": 0.3562972843647003,
      "learning_rate": 1.3859275053304904e-05,
      "loss": 0.095,
      "step": 1695
    },
    {
      "epoch": 3.624733475479744,
      "grad_norm": 2.4168753623962402,
      "learning_rate": 1.3752665245202559e-05,
      "loss": 0.1005,
      "step": 1700
    },
    {
      "epoch": 3.635394456289979,
      "grad_norm": 5.5596442222595215,
      "learning_rate": 1.3646055437100214e-05,
      "loss": 0.0585,
      "step": 1705
    },
    {
      "epoch": 3.6460554371002134,
      "grad_norm": 5.68841552734375,
      "learning_rate": 1.3539445628997869e-05,
      "loss": 0.1265,
      "step": 1710
    },
    {
      "epoch": 3.656716417910448,
      "grad_norm": 0.07254204154014587,
      "learning_rate": 1.3432835820895523e-05,
      "loss": 0.0481,
      "step": 1715
    },
    {
      "epoch": 3.6673773987206824,
      "grad_norm": 2.572387218475342,
      "learning_rate": 1.3326226012793178e-05,
      "loss": 0.0829,
      "step": 1720
    },
    {
      "epoch": 3.678038379530917,
      "grad_norm": 0.9826696515083313,
      "learning_rate": 1.3219616204690833e-05,
      "loss": 0.0928,
      "step": 1725
    },
    {
      "epoch": 3.6886993603411513,
      "grad_norm": 4.996304988861084,
      "learning_rate": 1.3113006396588488e-05,
      "loss": 0.1367,
      "step": 1730
    },
    {
      "epoch": 3.699360341151386,
      "grad_norm": 0.4510134160518646,
      "learning_rate": 1.3006396588486142e-05,
      "loss": 0.0351,
      "step": 1735
    },
    {
      "epoch": 3.7100213219616203,
      "grad_norm": 4.484724044799805,
      "learning_rate": 1.2899786780383797e-05,
      "loss": 0.074,
      "step": 1740
    },
    {
      "epoch": 3.7206823027718547,
      "grad_norm": 2.331526756286621,
      "learning_rate": 1.2793176972281452e-05,
      "loss": 0.0336,
      "step": 1745
    },
    {
      "epoch": 3.7313432835820897,
      "grad_norm": 0.9993152618408203,
      "learning_rate": 1.2686567164179105e-05,
      "loss": 0.0668,
      "step": 1750
    },
    {
      "epoch": 3.742004264392324,
      "grad_norm": 5.660934925079346,
      "learning_rate": 1.257995735607676e-05,
      "loss": 0.1111,
      "step": 1755
    },
    {
      "epoch": 3.7526652452025586,
      "grad_norm": 0.23702684044837952,
      "learning_rate": 1.2473347547974414e-05,
      "loss": 0.0387,
      "step": 1760
    },
    {
      "epoch": 3.763326226012793,
      "grad_norm": 6.45109224319458,
      "learning_rate": 1.2366737739872069e-05,
      "loss": 0.0511,
      "step": 1765
    },
    {
      "epoch": 3.7739872068230276,
      "grad_norm": 2.480509042739868,
      "learning_rate": 1.2260127931769722e-05,
      "loss": 0.0654,
      "step": 1770
    },
    {
      "epoch": 3.7846481876332625,
      "grad_norm": 1.6727945804595947,
      "learning_rate": 1.2153518123667377e-05,
      "loss": 0.088,
      "step": 1775
    },
    {
      "epoch": 3.795309168443497,
      "grad_norm": 1.9527469873428345,
      "learning_rate": 1.2046908315565032e-05,
      "loss": 0.0608,
      "step": 1780
    },
    {
      "epoch": 3.8059701492537314,
      "grad_norm": 2.770329475402832,
      "learning_rate": 1.1940298507462686e-05,
      "loss": 0.0698,
      "step": 1785
    },
    {
      "epoch": 3.816631130063966,
      "grad_norm": 0.8063251376152039,
      "learning_rate": 1.1833688699360341e-05,
      "loss": 0.0215,
      "step": 1790
    },
    {
      "epoch": 3.8272921108742004,
      "grad_norm": 0.9146325588226318,
      "learning_rate": 1.1727078891257996e-05,
      "loss": 0.1301,
      "step": 1795
    },
    {
      "epoch": 3.837953091684435,
      "grad_norm": 1.2591418027877808,
      "learning_rate": 1.162046908315565e-05,
      "loss": 0.0832,
      "step": 1800
    },
    {
      "epoch": 3.8486140724946694,
      "grad_norm": 2.9084525108337402,
      "learning_rate": 1.1513859275053305e-05,
      "loss": 0.0754,
      "step": 1805
    },
    {
      "epoch": 3.859275053304904,
      "grad_norm": 9.430222511291504,
      "learning_rate": 1.140724946695096e-05,
      "loss": 0.0427,
      "step": 1810
    },
    {
      "epoch": 3.8699360341151388,
      "grad_norm": 0.3675025701522827,
      "learning_rate": 1.1300639658848615e-05,
      "loss": 0.047,
      "step": 1815
    },
    {
      "epoch": 3.8805970149253732,
      "grad_norm": 2.7366576194763184,
      "learning_rate": 1.119402985074627e-05,
      "loss": 0.0582,
      "step": 1820
    },
    {
      "epoch": 3.8912579957356077,
      "grad_norm": 8.85275650024414,
      "learning_rate": 1.1087420042643924e-05,
      "loss": 0.1061,
      "step": 1825
    },
    {
      "epoch": 3.901918976545842,
      "grad_norm": 5.215404987335205,
      "learning_rate": 1.0980810234541579e-05,
      "loss": 0.1126,
      "step": 1830
    },
    {
      "epoch": 3.9125799573560767,
      "grad_norm": 5.9926276206970215,
      "learning_rate": 1.0874200426439234e-05,
      "loss": 0.107,
      "step": 1835
    },
    {
      "epoch": 3.923240938166311,
      "grad_norm": 1.831506371498108,
      "learning_rate": 1.0767590618336887e-05,
      "loss": 0.0836,
      "step": 1840
    },
    {
      "epoch": 3.933901918976546,
      "grad_norm": 3.159712076187134,
      "learning_rate": 1.0660980810234541e-05,
      "loss": 0.0483,
      "step": 1845
    },
    {
      "epoch": 3.9445628997867805,
      "grad_norm": 5.550416946411133,
      "learning_rate": 1.0554371002132196e-05,
      "loss": 0.1224,
      "step": 1850
    },
    {
      "epoch": 3.955223880597015,
      "grad_norm": 4.5900492668151855,
      "learning_rate": 1.0447761194029851e-05,
      "loss": 0.0914,
      "step": 1855
    },
    {
      "epoch": 3.9658848614072495,
      "grad_norm": 7.653269290924072,
      "learning_rate": 1.0341151385927506e-05,
      "loss": 0.049,
      "step": 1860
    },
    {
      "epoch": 3.976545842217484,
      "grad_norm": 10.73240852355957,
      "learning_rate": 1.023454157782516e-05,
      "loss": 0.1036,
      "step": 1865
    },
    {
      "epoch": 3.9872068230277184,
      "grad_norm": 1.4154757261276245,
      "learning_rate": 1.0127931769722815e-05,
      "loss": 0.0723,
      "step": 1870
    },
    {
      "epoch": 3.997867803837953,
      "grad_norm": 0.12970051169395447,
      "learning_rate": 1.002132196162047e-05,
      "loss": 0.0959,
      "step": 1875
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9224,
      "eval_loss": 0.2635798454284668,
      "eval_runtime": 29.9597,
      "eval_samples_per_second": 125.168,
      "eval_steps_per_second": 3.939,
      "step": 1876
    },
    {
      "epoch": 4.008528784648187,
      "grad_norm": 2.3170816898345947,
      "learning_rate": 9.914712153518125e-06,
      "loss": 0.0703,
      "step": 1880
    },
    {
      "epoch": 4.019189765458422,
      "grad_norm": 1.2057290077209473,
      "learning_rate": 9.80810234541578e-06,
      "loss": 0.034,
      "step": 1885
    },
    {
      "epoch": 4.029850746268656,
      "grad_norm": 2.035630702972412,
      "learning_rate": 9.701492537313434e-06,
      "loss": 0.0435,
      "step": 1890
    },
    {
      "epoch": 4.040511727078891,
      "grad_norm": 5.861476421356201,
      "learning_rate": 9.594882729211089e-06,
      "loss": 0.116,
      "step": 1895
    },
    {
      "epoch": 4.051172707889126,
      "grad_norm": 2.190972328186035,
      "learning_rate": 9.488272921108744e-06,
      "loss": 0.0678,
      "step": 1900
    },
    {
      "epoch": 4.061833688699361,
      "grad_norm": 0.36641818284988403,
      "learning_rate": 9.381663113006398e-06,
      "loss": 0.0905,
      "step": 1905
    },
    {
      "epoch": 4.072494669509595,
      "grad_norm": 0.3886570632457733,
      "learning_rate": 9.275053304904053e-06,
      "loss": 0.0077,
      "step": 1910
    },
    {
      "epoch": 4.08315565031983,
      "grad_norm": 5.203346252441406,
      "learning_rate": 9.168443496801706e-06,
      "loss": 0.0671,
      "step": 1915
    },
    {
      "epoch": 4.093816631130064,
      "grad_norm": 0.323293000459671,
      "learning_rate": 9.06183368869936e-06,
      "loss": 0.0426,
      "step": 1920
    },
    {
      "epoch": 4.104477611940299,
      "grad_norm": 3.8653154373168945,
      "learning_rate": 8.955223880597016e-06,
      "loss": 0.03,
      "step": 1925
    },
    {
      "epoch": 4.115138592750533,
      "grad_norm": 0.026840105652809143,
      "learning_rate": 8.84861407249467e-06,
      "loss": 0.0646,
      "step": 1930
    },
    {
      "epoch": 4.1257995735607675,
      "grad_norm": 2.1895546913146973,
      "learning_rate": 8.742004264392323e-06,
      "loss": 0.0452,
      "step": 1935
    },
    {
      "epoch": 4.136460554371002,
      "grad_norm": 2.3772225379943848,
      "learning_rate": 8.635394456289978e-06,
      "loss": 0.0428,
      "step": 1940
    },
    {
      "epoch": 4.1471215351812365,
      "grad_norm": 0.6743361353874207,
      "learning_rate": 8.528784648187633e-06,
      "loss": 0.0098,
      "step": 1945
    },
    {
      "epoch": 4.157782515991471,
      "grad_norm": 4.9721999168396,
      "learning_rate": 8.422174840085288e-06,
      "loss": 0.1001,
      "step": 1950
    },
    {
      "epoch": 4.1684434968017055,
      "grad_norm": 8.063929557800293,
      "learning_rate": 8.315565031982942e-06,
      "loss": 0.0331,
      "step": 1955
    },
    {
      "epoch": 4.17910447761194,
      "grad_norm": 2.5423943996429443,
      "learning_rate": 8.208955223880597e-06,
      "loss": 0.0435,
      "step": 1960
    },
    {
      "epoch": 4.189765458422174,
      "grad_norm": 4.879646301269531,
      "learning_rate": 8.102345415778252e-06,
      "loss": 0.1328,
      "step": 1965
    },
    {
      "epoch": 4.20042643923241,
      "grad_norm": 2.7492663860321045,
      "learning_rate": 7.995735607675907e-06,
      "loss": 0.0212,
      "step": 1970
    },
    {
      "epoch": 4.211087420042644,
      "grad_norm": 2.3990297317504883,
      "learning_rate": 7.889125799573561e-06,
      "loss": 0.0876,
      "step": 1975
    },
    {
      "epoch": 4.221748400852879,
      "grad_norm": 6.531210899353027,
      "learning_rate": 7.782515991471216e-06,
      "loss": 0.091,
      "step": 1980
    },
    {
      "epoch": 4.232409381663113,
      "grad_norm": 1.7070245742797852,
      "learning_rate": 7.67590618336887e-06,
      "loss": 0.0741,
      "step": 1985
    },
    {
      "epoch": 4.243070362473348,
      "grad_norm": 2.82181715965271,
      "learning_rate": 7.5692963752665255e-06,
      "loss": 0.0388,
      "step": 1990
    },
    {
      "epoch": 4.253731343283582,
      "grad_norm": 3.517402410507202,
      "learning_rate": 7.4626865671641785e-06,
      "loss": 0.072,
      "step": 1995
    },
    {
      "epoch": 4.264392324093817,
      "grad_norm": 3.256564140319824,
      "learning_rate": 7.356076759061833e-06,
      "loss": 0.0472,
      "step": 2000
    },
    {
      "epoch": 4.275053304904051,
      "grad_norm": 0.21228185296058655,
      "learning_rate": 7.249466950959488e-06,
      "loss": 0.053,
      "step": 2005
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 5.918868541717529,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.1296,
      "step": 2010
    },
    {
      "epoch": 4.29637526652452,
      "grad_norm": 0.4840411841869354,
      "learning_rate": 7.0362473347547975e-06,
      "loss": 0.0557,
      "step": 2015
    },
    {
      "epoch": 4.3070362473347545,
      "grad_norm": 5.728902816772461,
      "learning_rate": 6.929637526652452e-06,
      "loss": 0.0521,
      "step": 2020
    },
    {
      "epoch": 4.317697228144989,
      "grad_norm": 5.714817523956299,
      "learning_rate": 6.823027718550107e-06,
      "loss": 0.0553,
      "step": 2025
    },
    {
      "epoch": 4.3283582089552235,
      "grad_norm": 0.2205377072095871,
      "learning_rate": 6.716417910447762e-06,
      "loss": 0.0713,
      "step": 2030
    },
    {
      "epoch": 4.339019189765459,
      "grad_norm": 2.263744831085205,
      "learning_rate": 6.609808102345416e-06,
      "loss": 0.0256,
      "step": 2035
    },
    {
      "epoch": 4.349680170575693,
      "grad_norm": 3.861410617828369,
      "learning_rate": 6.503198294243071e-06,
      "loss": 0.0265,
      "step": 2040
    },
    {
      "epoch": 4.360341151385928,
      "grad_norm": 0.3259751498699188,
      "learning_rate": 6.396588486140726e-06,
      "loss": 0.0976,
      "step": 2045
    },
    {
      "epoch": 4.371002132196162,
      "grad_norm": 0.29996100068092346,
      "learning_rate": 6.28997867803838e-06,
      "loss": 0.0144,
      "step": 2050
    },
    {
      "epoch": 4.381663113006397,
      "grad_norm": 0.8312106132507324,
      "learning_rate": 6.1833688699360345e-06,
      "loss": 0.0558,
      "step": 2055
    },
    {
      "epoch": 4.392324093816631,
      "grad_norm": 0.3228684961795807,
      "learning_rate": 6.076759061833688e-06,
      "loss": 0.0949,
      "step": 2060
    },
    {
      "epoch": 4.402985074626866,
      "grad_norm": 0.12889361381530762,
      "learning_rate": 5.970149253731343e-06,
      "loss": 0.0202,
      "step": 2065
    },
    {
      "epoch": 4.4136460554371,
      "grad_norm": 2.6904802322387695,
      "learning_rate": 5.863539445628998e-06,
      "loss": 0.0497,
      "step": 2070
    },
    {
      "epoch": 4.424307036247335,
      "grad_norm": 0.29303914308547974,
      "learning_rate": 5.756929637526653e-06,
      "loss": 0.0068,
      "step": 2075
    },
    {
      "epoch": 4.434968017057569,
      "grad_norm": 3.381070852279663,
      "learning_rate": 5.650319829424307e-06,
      "loss": 0.041,
      "step": 2080
    },
    {
      "epoch": 4.445628997867804,
      "grad_norm": 0.0531914196908474,
      "learning_rate": 5.543710021321962e-06,
      "loss": 0.0212,
      "step": 2085
    },
    {
      "epoch": 4.456289978678038,
      "grad_norm": 3.2930874824523926,
      "learning_rate": 5.437100213219617e-06,
      "loss": 0.0587,
      "step": 2090
    },
    {
      "epoch": 4.466950959488273,
      "grad_norm": 4.437326908111572,
      "learning_rate": 5.330490405117271e-06,
      "loss": 0.1106,
      "step": 2095
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 4.3134589195251465,
      "learning_rate": 5.2238805970149255e-06,
      "loss": 0.0382,
      "step": 2100
    },
    {
      "epoch": 4.4882729211087415,
      "grad_norm": 9.413946151733398,
      "learning_rate": 5.11727078891258e-06,
      "loss": 0.0871,
      "step": 2105
    },
    {
      "epoch": 4.498933901918977,
      "grad_norm": 10.586847305297852,
      "learning_rate": 5.010660980810235e-06,
      "loss": 0.0544,
      "step": 2110
    },
    {
      "epoch": 4.509594882729211,
      "grad_norm": 0.6946657299995422,
      "learning_rate": 4.90405117270789e-06,
      "loss": 0.0712,
      "step": 2115
    },
    {
      "epoch": 4.520255863539446,
      "grad_norm": 7.0809736251831055,
      "learning_rate": 4.797441364605544e-06,
      "loss": 0.1054,
      "step": 2120
    },
    {
      "epoch": 4.53091684434968,
      "grad_norm": 1.441215991973877,
      "learning_rate": 4.690831556503199e-06,
      "loss": 0.0979,
      "step": 2125
    },
    {
      "epoch": 4.541577825159915,
      "grad_norm": 1.6024954319000244,
      "learning_rate": 4.584221748400853e-06,
      "loss": 0.0548,
      "step": 2130
    },
    {
      "epoch": 4.552238805970149,
      "grad_norm": 0.7445539236068726,
      "learning_rate": 4.477611940298508e-06,
      "loss": 0.0639,
      "step": 2135
    },
    {
      "epoch": 4.562899786780384,
      "grad_norm": 11.119247436523438,
      "learning_rate": 4.371002132196162e-06,
      "loss": 0.1372,
      "step": 2140
    },
    {
      "epoch": 4.573560767590618,
      "grad_norm": 1.6980524063110352,
      "learning_rate": 4.264392324093816e-06,
      "loss": 0.0553,
      "step": 2145
    },
    {
      "epoch": 4.584221748400853,
      "grad_norm": 3.8009989261627197,
      "learning_rate": 4.157782515991471e-06,
      "loss": 0.1142,
      "step": 2150
    },
    {
      "epoch": 4.594882729211087,
      "grad_norm": 6.351069450378418,
      "learning_rate": 4.051172707889126e-06,
      "loss": 0.084,
      "step": 2155
    },
    {
      "epoch": 4.605543710021322,
      "grad_norm": 3.482295274734497,
      "learning_rate": 3.944562899786781e-06,
      "loss": 0.0322,
      "step": 2160
    },
    {
      "epoch": 4.616204690831556,
      "grad_norm": 2.991924524307251,
      "learning_rate": 3.837953091684435e-06,
      "loss": 0.0448,
      "step": 2165
    },
    {
      "epoch": 4.6268656716417915,
      "grad_norm": 4.679220199584961,
      "learning_rate": 3.7313432835820893e-06,
      "loss": 0.089,
      "step": 2170
    },
    {
      "epoch": 4.637526652452026,
      "grad_norm": 4.049046039581299,
      "learning_rate": 3.624733475479744e-06,
      "loss": 0.0603,
      "step": 2175
    },
    {
      "epoch": 4.6481876332622605,
      "grad_norm": 0.11563615500926971,
      "learning_rate": 3.5181236673773987e-06,
      "loss": 0.0291,
      "step": 2180
    },
    {
      "epoch": 4.658848614072495,
      "grad_norm": 5.410636901855469,
      "learning_rate": 3.4115138592750535e-06,
      "loss": 0.082,
      "step": 2185
    },
    {
      "epoch": 4.669509594882729,
      "grad_norm": 0.9837836027145386,
      "learning_rate": 3.304904051172708e-06,
      "loss": 0.0144,
      "step": 2190
    },
    {
      "epoch": 4.680170575692964,
      "grad_norm": 1.4047187566757202,
      "learning_rate": 3.198294243070363e-06,
      "loss": 0.0567,
      "step": 2195
    },
    {
      "epoch": 4.690831556503198,
      "grad_norm": 1.9856698513031006,
      "learning_rate": 3.0916844349680173e-06,
      "loss": 0.0352,
      "step": 2200
    },
    {
      "epoch": 4.701492537313433,
      "grad_norm": 6.325466632843018,
      "learning_rate": 2.9850746268656716e-06,
      "loss": 0.0629,
      "step": 2205
    },
    {
      "epoch": 4.712153518123667,
      "grad_norm": 2.1614136695861816,
      "learning_rate": 2.8784648187633263e-06,
      "loss": 0.0482,
      "step": 2210
    },
    {
      "epoch": 4.722814498933902,
      "grad_norm": 3.9222583770751953,
      "learning_rate": 2.771855010660981e-06,
      "loss": 0.0254,
      "step": 2215
    },
    {
      "epoch": 4.733475479744136,
      "grad_norm": 4.831112384796143,
      "learning_rate": 2.6652452025586354e-06,
      "loss": 0.0767,
      "step": 2220
    },
    {
      "epoch": 4.744136460554371,
      "grad_norm": 1.7830005884170532,
      "learning_rate": 2.55863539445629e-06,
      "loss": 0.039,
      "step": 2225
    },
    {
      "epoch": 4.754797441364605,
      "grad_norm": 4.798508644104004,
      "learning_rate": 2.452025586353945e-06,
      "loss": 0.0551,
      "step": 2230
    },
    {
      "epoch": 4.76545842217484,
      "grad_norm": 3.2544662952423096,
      "learning_rate": 2.3454157782515996e-06,
      "loss": 0.0352,
      "step": 2235
    },
    {
      "epoch": 4.776119402985074,
      "grad_norm": 5.712475776672363,
      "learning_rate": 2.238805970149254e-06,
      "loss": 0.0457,
      "step": 2240
    },
    {
      "epoch": 4.786780383795309,
      "grad_norm": 0.1435627043247223,
      "learning_rate": 2.132196162046908e-06,
      "loss": 0.0986,
      "step": 2245
    },
    {
      "epoch": 4.797441364605544,
      "grad_norm": 0.12689536809921265,
      "learning_rate": 2.025586353944563e-06,
      "loss": 0.0506,
      "step": 2250
    },
    {
      "epoch": 4.8081023454157785,
      "grad_norm": 13.168557167053223,
      "learning_rate": 1.9189765458422177e-06,
      "loss": 0.0666,
      "step": 2255
    },
    {
      "epoch": 4.818763326226013,
      "grad_norm": 1.2240266799926758,
      "learning_rate": 1.812366737739872e-06,
      "loss": 0.0658,
      "step": 2260
    },
    {
      "epoch": 4.8294243070362475,
      "grad_norm": 3.136503219604492,
      "learning_rate": 1.7057569296375267e-06,
      "loss": 0.0668,
      "step": 2265
    },
    {
      "epoch": 4.840085287846482,
      "grad_norm": 1.0185257196426392,
      "learning_rate": 1.5991471215351815e-06,
      "loss": 0.0689,
      "step": 2270
    },
    {
      "epoch": 4.850746268656716,
      "grad_norm": 8.60339641571045,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 0.0781,
      "step": 2275
    },
    {
      "epoch": 4.861407249466951,
      "grad_norm": 0.9385230541229248,
      "learning_rate": 1.3859275053304905e-06,
      "loss": 0.0696,
      "step": 2280
    },
    {
      "epoch": 4.872068230277185,
      "grad_norm": 0.7215489745140076,
      "learning_rate": 1.279317697228145e-06,
      "loss": 0.0626,
      "step": 2285
    },
    {
      "epoch": 4.88272921108742,
      "grad_norm": 0.554114043712616,
      "learning_rate": 1.1727078891257998e-06,
      "loss": 0.0096,
      "step": 2290
    },
    {
      "epoch": 4.893390191897654,
      "grad_norm": 0.41608935594558716,
      "learning_rate": 1.066098081023454e-06,
      "loss": 0.0603,
      "step": 2295
    },
    {
      "epoch": 4.904051172707889,
      "grad_norm": 2.0386576652526855,
      "learning_rate": 9.594882729211088e-07,
      "loss": 0.0461,
      "step": 2300
    },
    {
      "epoch": 4.914712153518123,
      "grad_norm": 2.6170732975006104,
      "learning_rate": 8.528784648187634e-07,
      "loss": 0.0821,
      "step": 2305
    },
    {
      "epoch": 4.925373134328359,
      "grad_norm": 4.90107536315918,
      "learning_rate": 7.462686567164179e-07,
      "loss": 0.0373,
      "step": 2310
    },
    {
      "epoch": 4.936034115138593,
      "grad_norm": 0.05737028643488884,
      "learning_rate": 6.396588486140725e-07,
      "loss": 0.0357,
      "step": 2315
    },
    {
      "epoch": 4.946695095948828,
      "grad_norm": 1.5560643672943115,
      "learning_rate": 5.33049040511727e-07,
      "loss": 0.0298,
      "step": 2320
    },
    {
      "epoch": 4.957356076759062,
      "grad_norm": 0.5660206079483032,
      "learning_rate": 4.264392324093817e-07,
      "loss": 0.0056,
      "step": 2325
    },
    {
      "epoch": 4.968017057569297,
      "grad_norm": 4.370354175567627,
      "learning_rate": 3.1982942430703626e-07,
      "loss": 0.0467,
      "step": 2330
    },
    {
      "epoch": 4.978678038379531,
      "grad_norm": 4.627424716949463,
      "learning_rate": 2.1321961620469084e-07,
      "loss": 0.0706,
      "step": 2335
    },
    {
      "epoch": 4.9893390191897655,
      "grad_norm": 3.37949275970459,
      "learning_rate": 1.0660980810234542e-07,
      "loss": 0.0461,
      "step": 2340
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2696599960327148,
      "learning_rate": 0.0,
      "loss": 0.0781,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9272,
      "eval_loss": 0.23820149898529053,
      "eval_runtime": 30.2254,
      "eval_samples_per_second": 124.068,
      "eval_steps_per_second": 3.904,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "step": 2345,
      "total_flos": 5.8118992210944e+18,
      "train_loss": 0.09254299414469234,
      "train_runtime": 1496.6518,
      "train_samples_per_second": 50.112,
      "train_steps_per_second": 1.567
    }
  ],
  "logging_steps": 5,
  "max_steps": 2345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 5.8118992210944e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}