{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9999209511896846,
  "eval_steps": 500,
  "global_step": 1680,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005951910423748122,
      "grad_norm": 0.32421875,
      "learning_rate": 1.1904761904761904e-06,
      "loss": 1.0942,
      "step": 1
    },
    {
      "epoch": 0.0029759552118740614,
      "grad_norm": 0.306640625,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 1.1063,
      "step": 5
    },
    {
      "epoch": 0.005951910423748123,
      "grad_norm": 0.29296875,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 1.0913,
      "step": 10
    },
    {
      "epoch": 0.008927865635622183,
      "grad_norm": 0.294921875,
      "learning_rate": 1.785714285714286e-05,
      "loss": 1.0858,
      "step": 15
    },
    {
      "epoch": 0.011903820847496246,
      "grad_norm": 0.2734375,
      "learning_rate": 2.380952380952381e-05,
      "loss": 1.0707,
      "step": 20
    },
    {
      "epoch": 0.014879776059370306,
      "grad_norm": 0.2451171875,
      "learning_rate": 2.9761904761904762e-05,
      "loss": 1.0274,
      "step": 25
    },
    {
      "epoch": 0.017855731271244367,
      "grad_norm": 0.171875,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.9783,
      "step": 30
    },
    {
      "epoch": 0.02083168648311843,
      "grad_norm": 0.16796875,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.9564,
      "step": 35
    },
    {
      "epoch": 0.02380764169499249,
      "grad_norm": 0.1328125,
      "learning_rate": 4.761904761904762e-05,
      "loss": 0.9325,
      "step": 40
    },
    {
      "epoch": 0.02678359690686655,
      "grad_norm": 0.11767578125,
      "learning_rate": 5.3571428571428575e-05,
      "loss": 0.9242,
      "step": 45
    },
    {
      "epoch": 0.029759552118740613,
      "grad_norm": 0.0859375,
      "learning_rate": 5.9523809523809524e-05,
      "loss": 0.8951,
      "step": 50
    },
    {
      "epoch": 0.03273550733061467,
      "grad_norm": 0.0849609375,
      "learning_rate": 6.547619047619048e-05,
      "loss": 0.8703,
      "step": 55
    },
    {
      "epoch": 0.035711462542488734,
      "grad_norm": 0.076171875,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.8513,
      "step": 60
    },
    {
      "epoch": 0.038687417754362796,
      "grad_norm": 0.06494140625,
      "learning_rate": 7.738095238095239e-05,
      "loss": 0.8488,
      "step": 65
    },
    {
      "epoch": 0.04166337296623686,
      "grad_norm": 0.06494140625,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.8218,
      "step": 70
    },
    {
      "epoch": 0.04463932817811092,
      "grad_norm": 0.068359375,
      "learning_rate": 8.92857142857143e-05,
      "loss": 0.8321,
      "step": 75
    },
    {
      "epoch": 0.04761528338998498,
      "grad_norm": 0.06640625,
      "learning_rate": 9.523809523809524e-05,
      "loss": 0.8271,
      "step": 80
    },
    {
      "epoch": 0.050591238601859045,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0001011904761904762,
      "loss": 0.8097,
      "step": 85
    },
    {
      "epoch": 0.0535671938137331,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00010714285714285715,
      "loss": 0.7871,
      "step": 90
    },
    {
      "epoch": 0.05654314902560716,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00011309523809523809,
      "loss": 0.8205,
      "step": 95
    },
    {
      "epoch": 0.059519104237481225,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00011904761904761905,
      "loss": 0.8013,
      "step": 100
    },
    {
      "epoch": 0.06249505944935529,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.000125,
      "loss": 0.7996,
      "step": 105
    },
    {
      "epoch": 0.06547101466122934,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00013095238095238096,
      "loss": 0.8194,
      "step": 110
    },
    {
      "epoch": 0.0684469698731034,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0001369047619047619,
      "loss": 0.7937,
      "step": 115
    },
    {
      "epoch": 0.07142292508497747,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00014285714285714287,
      "loss": 0.7972,
      "step": 120
    },
    {
      "epoch": 0.07439888029685153,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00014880952380952382,
      "loss": 0.8139,
      "step": 125
    },
    {
      "epoch": 0.07737483550872559,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00015476190476190478,
      "loss": 0.7933,
      "step": 130
    },
    {
      "epoch": 0.08035079072059965,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00016071428571428573,
      "loss": 0.7852,
      "step": 135
    },
    {
      "epoch": 0.08332674593247372,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.8041,
      "step": 140
    },
    {
      "epoch": 0.08630270114434778,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00017261904761904764,
      "loss": 0.8021,
      "step": 145
    },
    {
      "epoch": 0.08927865635622184,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001785714285714286,
      "loss": 0.7883,
      "step": 150
    },
    {
      "epoch": 0.0922546115680959,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00018452380952380955,
      "loss": 0.8034,
      "step": 155
    },
    {
      "epoch": 0.09523056677996997,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00019047619047619048,
      "loss": 0.7865,
      "step": 160
    },
    {
      "epoch": 0.09820652199184403,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00019642857142857144,
      "loss": 0.7945,
      "step": 165
    },
    {
      "epoch": 0.10118247720371809,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0001999991365731819,
      "loss": 0.7968,
      "step": 170
    },
    {
      "epoch": 0.10415843241559214,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00019998942319271077,
      "loss": 0.7942,
      "step": 175
    },
    {
      "epoch": 0.1071343876274662,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00019996891820008164,
      "loss": 0.8003,
      "step": 180
    },
    {
      "epoch": 0.11011034283934026,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00019993762380834785,
      "loss": 0.7792,
      "step": 185
    },
    {
      "epoch": 0.11308629805121433,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00019989554339503612,
      "loss": 0.7829,
      "step": 190
    },
    {
      "epoch": 0.11606225326308839,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00019984268150178167,
      "loss": 0.7946,
      "step": 195
    },
    {
      "epoch": 0.11903820847496245,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0001997790438338385,
      "loss": 0.7961,
      "step": 200
    },
    {
      "epoch": 0.12201416368683651,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00019970463725946336,
      "loss": 0.7896,
      "step": 205
    },
    {
      "epoch": 0.12499011889871058,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00019961946980917456,
      "loss": 0.782,
      "step": 210
    },
    {
      "epoch": 0.12796607411058464,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00019952355067488523,
      "loss": 0.8018,
      "step": 215
    },
    {
      "epoch": 0.1309420293224587,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0001994168902089112,
      "loss": 0.7943,
      "step": 220
    },
    {
      "epoch": 0.13391798453433276,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00019929949992285396,
      "loss": 0.8005,
      "step": 225
    },
    {
      "epoch": 0.1368939397462068,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00019917139248635786,
      "loss": 0.7776,
      "step": 230
    },
    {
      "epoch": 0.1398698949580809,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00019903258172574302,
      "loss": 0.7905,
      "step": 235
    },
    {
      "epoch": 0.14284585016995494,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00019888308262251285,
      "loss": 0.7782,
      "step": 240
    },
    {
      "epoch": 0.145821805381829,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00019872291131173742,
      "loss": 0.7798,
      "step": 245
    },
    {
      "epoch": 0.14879776059370306,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0001985520850803117,
      "loss": 0.7842,
      "step": 250
    },
    {
      "epoch": 0.15177371580557714,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00019837062236509014,
      "loss": 0.7778,
      "step": 255
    },
    {
      "epoch": 0.15474967101745118,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0001981785427508966,
      "loss": 0.7813,
      "step": 260
    },
    {
      "epoch": 0.15772562622932526,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00019797586696841072,
      "loss": 0.7956,
      "step": 265
    },
    {
      "epoch": 0.1607015814411993,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00019776261689193048,
      "loss": 0.8056,
      "step": 270
    },
    {
      "epoch": 0.16367753665307336,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00019753881553701138,
      "loss": 0.7886,
      "step": 275
    },
    {
      "epoch": 0.16665349186494743,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.00019730448705798239,
      "loss": 0.8025,
      "step": 280
    },
    {
      "epoch": 0.16962944707682148,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0001970596567453391,
      "loss": 0.7803,
      "step": 285
    },
    {
      "epoch": 0.17260540228869556,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00019680435102301412,
      "loss": 0.7844,
      "step": 290
    },
    {
      "epoch": 0.1755813575005696,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0001965385974455251,
      "loss": 0.797,
      "step": 295
    },
    {
      "epoch": 0.17855731271244368,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0001962624246950012,
      "loss": 0.7774,
      "step": 300
    },
    {
      "epoch": 0.18153326792431773,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00019597586257808712,
      "loss": 0.7756,
      "step": 305
    },
    {
      "epoch": 0.1845092231361918,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0001956789420227262,
      "loss": 0.7703,
      "step": 310
    },
    {
      "epoch": 0.18748517834806586,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0001953716950748227,
      "loss": 0.7901,
      "step": 315
    },
    {
      "epoch": 0.19046113355993993,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0001950541548947829,
      "loss": 0.78,
      "step": 320
    },
    {
      "epoch": 0.19343708877181398,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0001947263557539363,
      "loss": 0.767,
      "step": 325
    },
    {
      "epoch": 0.19641304398368806,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00019438833303083678,
      "loss": 0.7963,
      "step": 330
    },
    {
      "epoch": 0.1993889991955621,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00019404012320744417,
      "loss": 0.783,
      "step": 335
    },
    {
      "epoch": 0.20236495440743618,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0001936817638651871,
      "loss": 0.7814,
      "step": 340
    },
    {
      "epoch": 0.20534090961931023,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00019331329368090666,
      "loss": 0.7791,
      "step": 345
    },
    {
      "epoch": 0.20831686483118428,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00019293475242268223,
      "loss": 0.7782,
      "step": 350
    },
    {
      "epoch": 0.21129282004305835,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00019254618094553949,
      "loss": 0.7835,
      "step": 355
    },
    {
      "epoch": 0.2142687752549324,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00019214762118704076,
      "loss": 0.7886,
      "step": 360
    },
    {
      "epoch": 0.21724473046680648,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00019173911616275917,
      "loss": 0.7766,
      "step": 365
    },
    {
      "epoch": 0.22022068567868053,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00019132070996163568,
      "loss": 0.7838,
      "step": 370
    },
    {
      "epoch": 0.2231966408905546,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0001908924477412211,
      "loss": 0.7837,
      "step": 375
    },
    {
      "epoch": 0.22617259610242865,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00019045437572280194,
      "loss": 0.7961,
      "step": 380
    },
    {
      "epoch": 0.22914855131430273,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00019000654118641211,
      "loss": 0.7903,
      "step": 385
    },
    {
      "epoch": 0.23212450652617678,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0001895489924657301,
      "loss": 0.7776,
      "step": 390
    },
    {
      "epoch": 0.23510046173805085,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00018908177894286232,
      "loss": 0.7768,
      "step": 395
    },
    {
      "epoch": 0.2380764169499249,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00018860495104301345,
      "loss": 0.7842,
      "step": 400
    },
    {
      "epoch": 0.24105237216179898,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00018811856022904423,
      "loss": 0.7728,
      "step": 405
    },
    {
      "epoch": 0.24402832737367303,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00018762265899591722,
      "loss": 0.7787,
      "step": 410
    },
    {
      "epoch": 0.24700428258554707,
      "grad_norm": 0.05078125,
      "learning_rate": 0.000187117300865031,
      "loss": 0.7893,
      "step": 415
    },
    {
      "epoch": 0.24998023779742115,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.7846,
      "step": 420
    },
    {
      "epoch": 0.2529561930092952,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00018607843309298723,
      "loss": 0.785,
      "step": 425
    },
    {
      "epoch": 0.2559321482211693,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00018554503557426948,
      "loss": 0.7861,
      "step": 430
    },
    {
      "epoch": 0.2589081034330433,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0001850024053905709,
      "loss": 0.7605,
      "step": 435
    },
    {
      "epoch": 0.2618840586449174,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0001844506011066308,
      "loss": 0.7691,
      "step": 440
    },
    {
      "epoch": 0.2648600138567915,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00018388968227732626,
      "loss": 0.7743,
      "step": 445
    },
    {
      "epoch": 0.2678359690686655,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0001833197094412449,
      "loss": 0.7834,
      "step": 450
    },
    {
      "epoch": 0.2708119242805396,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00018274074411415105,
      "loss": 0.7804,
      "step": 455
    },
    {
      "epoch": 0.2737878794924136,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00018215284878234642,
      "loss": 0.7985,
      "step": 460
    },
    {
      "epoch": 0.2767638347042877,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00018155608689592604,
      "loss": 0.7852,
      "step": 465
    },
    {
      "epoch": 0.2797397899161618,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0001809505228619304,
      "loss": 0.7697,
      "step": 470
    },
    {
      "epoch": 0.2827157451280358,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0001803362220373942,
      "loss": 0.772,
      "step": 475
    },
    {
      "epoch": 0.28569170033990987,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00017971325072229226,
      "loss": 0.7613,
      "step": 480
    },
    {
      "epoch": 0.2886676555517839,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00017908167615238415,
      "loss": 0.7737,
      "step": 485
    },
    {
      "epoch": 0.291643610763658,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00017844156649195759,
      "loss": 0.7813,
      "step": 490
    },
    {
      "epoch": 0.29461956597553207,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00017779299082647148,
      "loss": 0.7709,
      "step": 495
    },
    {
      "epoch": 0.2975955211874061,
      "grad_norm": 0.046875,
      "learning_rate": 0.0001771360191551,
      "loss": 0.7838,
      "step": 500
    },
    {
      "epoch": 0.30057147639928017,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00017647072238317728,
      "loss": 0.7638,
      "step": 505
    },
    {
      "epoch": 0.3035474316111543,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0001757971723145453,
      "loss": 0.7838,
      "step": 510
    },
    {
      "epoch": 0.3065233868230283,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00017511544164380388,
      "loss": 0.7852,
      "step": 515
    },
    {
      "epoch": 0.30949934203490237,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00017442560394846516,
      "loss": 0.7664,
      "step": 520
    },
    {
      "epoch": 0.3124752972467764,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0001737277336810124,
      "loss": 0.7715,
      "step": 525
    },
    {
      "epoch": 0.3154512524586505,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00017302190616086464,
      "loss": 0.7474,
      "step": 530
    },
    {
      "epoch": 0.31842720767052457,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0001723081975662476,
      "loss": 0.7906,
      "step": 535
    },
    {
      "epoch": 0.3214031628823986,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00017158668492597186,
      "loss": 0.7694,
      "step": 540
    },
    {
      "epoch": 0.32437911809427267,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00017085744611111957,
      "loss": 0.7855,
      "step": 545
    },
    {
      "epoch": 0.3273550733061467,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0001701205598266398,
      "loss": 0.7674,
      "step": 550
    },
    {
      "epoch": 0.3303310285180208,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00016937610560285418,
      "loss": 0.7619,
      "step": 555
    },
    {
      "epoch": 0.33330698372989487,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0001686241637868734,
      "loss": 0.7777,
      "step": 560
    },
    {
      "epoch": 0.3362829389417689,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00016786481553392548,
      "loss": 0.7792,
      "step": 565
    },
    {
      "epoch": 0.33925889415364296,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00016709814279859702,
      "loss": 0.7601,
      "step": 570
    },
    {
      "epoch": 0.34223484936551707,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00016632422832598795,
      "loss": 0.7817,
      "step": 575
    },
    {
      "epoch": 0.3452108045773911,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.000165543155642781,
      "loss": 0.7905,
      "step": 580
    },
    {
      "epoch": 0.34818675978926517,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00016475500904822706,
      "loss": 0.7581,
      "step": 585
    },
    {
      "epoch": 0.3511627150011392,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00016395987360504668,
      "loss": 0.7693,
      "step": 590
    },
    {
      "epoch": 0.3541386702130133,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00016315783513024977,
      "loss": 0.7861,
      "step": 595
    },
    {
      "epoch": 0.35711462542488737,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00016234898018587337,
      "loss": 0.7765,
      "step": 600
    },
    {
      "epoch": 0.3600905806367614,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0001615333960696393,
      "loss": 0.7682,
      "step": 605
    },
    {
      "epoch": 0.36306653584863546,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00016071117080553236,
      "loss": 0.7768,
      "step": 610
    },
    {
      "epoch": 0.3660424910605095,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00015988239313430004,
      "loss": 0.7684,
      "step": 615
    },
    {
      "epoch": 0.3690184462723836,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00015904715250387498,
      "loss": 0.7677,
      "step": 620
    },
    {
      "epoch": 0.37199440148425766,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0001582055390597212,
      "loss": 0.7866,
      "step": 625
    },
    {
      "epoch": 0.3749703566961317,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0001573576436351046,
      "loss": 0.7753,
      "step": 630
    },
    {
      "epoch": 0.37794631190800576,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00015650355774129,
      "loss": 0.7854,
      "step": 635
    },
    {
      "epoch": 0.38092226711987986,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00015564337355766412,
      "loss": 0.7639,
      "step": 640
    },
    {
      "epoch": 0.3838982223317539,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00015477718392178716,
      "loss": 0.7574,
      "step": 645
    },
    {
      "epoch": 0.38687417754362796,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00015390508231937297,
      "loss": 0.7663,
      "step": 650
    },
    {
      "epoch": 0.389850132755502,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00015302716287419945,
      "loss": 0.7746,
      "step": 655
    },
    {
      "epoch": 0.3928260879673761,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0001521435203379498,
      "loss": 0.7617,
      "step": 660
    },
    {
      "epoch": 0.39580204317925016,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00015125425007998653,
      "loss": 0.7993,
      "step": 665
    },
    {
      "epoch": 0.3987779983911242,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0001503594480770581,
      "loss": 0.7887,
      "step": 670
    },
    {
      "epoch": 0.40175395360299826,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00014945921090294076,
      "loss": 0.7648,
      "step": 675
    },
    {
      "epoch": 0.40472990881487236,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00014855363571801523,
      "loss": 0.7811,
      "step": 680
    },
    {
      "epoch": 0.4077058640267464,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00014764282025878068,
      "loss": 0.7674,
      "step": 685
    },
    {
      "epoch": 0.41068181923862046,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001467268628273062,
      "loss": 0.7684,
      "step": 690
    },
    {
      "epoch": 0.4136577744504945,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00014580586228062122,
      "loss": 0.7957,
      "step": 695
    },
    {
      "epoch": 0.41663372966236856,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00014487991802004623,
      "loss": 0.7778,
      "step": 700
    },
    {
      "epoch": 0.41960968487424266,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0001439491299804645,
      "loss": 0.7593,
      "step": 705
    },
    {
      "epoch": 0.4225856400861167,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001430135986195365,
      "loss": 0.7576,
      "step": 710
    },
    {
      "epoch": 0.42556159529799076,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00014207342490685774,
      "loss": 0.7772,
      "step": 715
    },
    {
      "epoch": 0.4285375505098648,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00014112871031306119,
      "loss": 0.7811,
      "step": 720
    },
    {
      "epoch": 0.4315135057217389,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00014017955679886598,
      "loss": 0.7644,
      "step": 725
    },
    {
      "epoch": 0.43448946093361296,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00013922606680407307,
      "loss": 0.7757,
      "step": 730
    },
    {
      "epoch": 0.437465416145487,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.000138268343236509,
      "loss": 0.7764,
      "step": 735
    },
    {
      "epoch": 0.44044137135736106,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0001373064894609194,
      "loss": 0.7784,
      "step": 740
    },
    {
      "epoch": 0.44341732656923516,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0001363406092878131,
      "loss": 0.7663,
      "step": 745
    },
    {
      "epoch": 0.4463932817811092,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00013537080696225814,
      "loss": 0.7828,
      "step": 750
    },
    {
      "epoch": 0.44936923699298326,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0001343971871526307,
      "loss": 0.7764,
      "step": 755
    },
    {
      "epoch": 0.4523451922048573,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00013341985493931877,
      "loss": 0.7753,
      "step": 760
    },
    {
      "epoch": 0.45532114741673135,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00013243891580338072,
      "loss": 0.7663,
      "step": 765
    },
    {
      "epoch": 0.45829710262860546,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00013145447561516138,
      "loss": 0.7801,
      "step": 770
    },
    {
      "epoch": 0.4612730578404795,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00013046664062286545,
      "loss": 0.7917,
      "step": 775
    },
    {
      "epoch": 0.46424901305235355,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00012947551744109043,
      "loss": 0.7751,
      "step": 780
    },
    {
      "epoch": 0.4672249682642276,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00012848121303932013,
      "loss": 0.7585,
      "step": 785
    },
    {
      "epoch": 0.4702009234761017,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00012748383473037948,
      "loss": 0.7888,
      "step": 790
    },
    {
      "epoch": 0.47317687868797575,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00012648349015885273,
      "loss": 0.7766,
      "step": 795
    },
    {
      "epoch": 0.4761528338998498,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0001254802872894655,
      "loss": 0.7707,
      "step": 800
    },
    {
      "epoch": 0.47912878911172385,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0001244743343954324,
      "loss": 0.7678,
      "step": 805
    },
    {
      "epoch": 0.48210474432359796,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00012346574004677154,
      "loss": 0.7477,
      "step": 810
    },
    {
      "epoch": 0.485080699535472,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001224546130985867,
      "loss": 0.7758,
      "step": 815
    },
    {
      "epoch": 0.48805665474734605,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00012144106267931876,
      "loss": 0.7797,
      "step": 820
    },
    {
      "epoch": 0.4910326099592201,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00012042519817896804,
      "loss": 0.764,
      "step": 825
    },
    {
      "epoch": 0.49400856517109415,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00011940712923728783,
      "loss": 0.7602,
      "step": 830
    },
    {
      "epoch": 0.49698452038296825,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00011838696573195139,
      "loss": 0.7631,
      "step": 835
    },
    {
      "epoch": 0.4999604755948423,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00011736481776669306,
      "loss": 0.7793,
      "step": 840
    },
    {
      "epoch": 0.5029364308067164,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00011634079565942497,
      "loss": 0.7658,
      "step": 845
    },
    {
      "epoch": 0.5059123860185905,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00011531500993033093,
      "loss": 0.7499,
      "step": 850
    },
    {
      "epoch": 0.5088883412304644,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00011428757128993802,
      "loss": 0.7706,
      "step": 855
    },
    {
      "epoch": 0.5118642964423386,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00011325859062716795,
      "loss": 0.7757,
      "step": 860
    },
    {
      "epoch": 0.5148402516542127,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00011222817899736914,
      "loss": 0.7828,
      "step": 865
    },
    {
      "epoch": 0.5178162068660866,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00011119644761033078,
      "loss": 0.7687,
      "step": 870
    },
    {
      "epoch": 0.5207921620779608,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00011016350781828019,
      "loss": 0.775,
      "step": 875
    },
    {
      "epoch": 0.5237681172898347,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00010912947110386484,
      "loss": 0.7664,
      "step": 880
    },
    {
      "epoch": 0.5267440725017088,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00010809444906812033,
      "loss": 0.7962,
      "step": 885
    },
    {
      "epoch": 0.529720027713583,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00010705855341842563,
      "loss": 0.7806,
      "step": 890
    },
    {
      "epoch": 0.532695982925457,
      "grad_norm": 0.046875,
      "learning_rate": 0.0001060218959564466,
      "loss": 0.7626,
      "step": 895
    },
    {
      "epoch": 0.535671938137331,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00010498458856606972,
      "loss": 0.7747,
      "step": 900
    },
    {
      "epoch": 0.538647893349205,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00010394674320132662,
      "loss": 0.7813,
      "step": 905
    },
    {
      "epoch": 0.5416238485610791,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00010290847187431113,
      "loss": 0.7913,
      "step": 910
    },
    {
      "epoch": 0.5445998037729533,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00010186988664309023,
      "loss": 0.7648,
      "step": 915
    },
    {
      "epoch": 0.5475757589848272,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00010083109959960973,
      "loss": 0.7837,
      "step": 920
    },
    {
      "epoch": 0.5505517141967013,
      "grad_norm": 0.04638671875,
      "learning_rate": 9.979222285759651e-05,
      "loss": 0.7756,
      "step": 925
    },
    {
      "epoch": 0.5535276694085755,
      "grad_norm": 0.045166015625,
      "learning_rate": 9.875336854045851e-05,
      "loss": 0.7512,
      "step": 930
    },
    {
      "epoch": 0.5565036246204494,
      "grad_norm": 0.04833984375,
      "learning_rate": 9.771464876918331e-05,
      "loss": 0.7696,
      "step": 935
    },
    {
      "epoch": 0.5594795798323235,
      "grad_norm": 0.82421875,
      "learning_rate": 9.667617565023735e-05,
      "loss": 0.7855,
      "step": 940
    },
    {
      "epoch": 0.5624555350441975,
      "grad_norm": 0.061279296875,
      "learning_rate": 9.563806126346642e-05,
      "loss": 0.7613,
      "step": 945
    },
    {
      "epoch": 0.5654314902560716,
      "grad_norm": 0.049072265625,
      "learning_rate": 9.460041764999928e-05,
      "loss": 0.7782,
      "step": 950
    },
    {
      "epoch": 0.5684074454679457,
      "grad_norm": 0.05029296875,
      "learning_rate": 9.356335680015534e-05,
      "loss": 0.7781,
      "step": 955
    },
    {
      "epoch": 0.5713834006798197,
      "grad_norm": 0.0498046875,
      "learning_rate": 9.252699064135758e-05,
      "loss": 0.7953,
      "step": 960
    },
    {
      "epoch": 0.5743593558916938,
      "grad_norm": 0.049560546875,
      "learning_rate": 9.149143102605294e-05,
      "loss": 0.7649,
      "step": 965
    },
    {
      "epoch": 0.5773353111035678,
      "grad_norm": 0.048583984375,
      "learning_rate": 9.045678971963988e-05,
      "loss": 0.7635,
      "step": 970
    },
    {
      "epoch": 0.5803112663154419,
      "grad_norm": 0.050048828125,
      "learning_rate": 8.942317838840623e-05,
      "loss": 0.7689,
      "step": 975
    },
    {
      "epoch": 0.583287221527316,
      "grad_norm": 0.0517578125,
      "learning_rate": 8.839070858747697e-05,
      "loss": 0.7869,
      "step": 980
    },
    {
      "epoch": 0.58626317673919,
      "grad_norm": 0.0498046875,
      "learning_rate": 8.735949174877466e-05,
      "loss": 0.7606,
      "step": 985
    },
    {
      "epoch": 0.5892391319510641,
      "grad_norm": 0.0498046875,
      "learning_rate": 8.632963916899268e-05,
      "loss": 0.7665,
      "step": 990
    },
    {
      "epoch": 0.5922150871629382,
      "grad_norm": 0.04931640625,
      "learning_rate": 8.530126199758323e-05,
      "loss": 0.7967,
      "step": 995
    },
    {
      "epoch": 0.5951910423748122,
      "grad_norm": 0.052001953125,
      "learning_rate": 8.427447122476148e-05,
      "loss": 0.7784,
      "step": 1000
    },
    {
      "epoch": 0.5981669975866863,
      "grad_norm": 0.048583984375,
      "learning_rate": 8.324937766952638e-05,
      "loss": 0.7651,
      "step": 1005
    },
    {
      "epoch": 0.6011429527985603,
      "grad_norm": 0.05810546875,
      "learning_rate": 8.222609196770036e-05,
      "loss": 0.7773,
      "step": 1010
    },
    {
      "epoch": 0.6041189080104344,
      "grad_norm": 0.05224609375,
      "learning_rate": 8.120472455998882e-05,
      "loss": 0.7509,
      "step": 1015
    },
    {
      "epoch": 0.6070948632223085,
      "grad_norm": 0.048828125,
      "learning_rate": 8.018538568006027e-05,
      "loss": 0.7623,
      "step": 1020
    },
    {
      "epoch": 0.6100708184341825,
      "grad_norm": 0.0693359375,
      "learning_rate": 7.916818534264921e-05,
      "loss": 0.7492,
      "step": 1025
    },
    {
      "epoch": 0.6130467736460566,
      "grad_norm": 0.0478515625,
      "learning_rate": 7.815323333168262e-05,
      "loss": 0.7612,
      "step": 1030
    },
    {
      "epoch": 0.6160227288579306,
      "grad_norm": 0.0556640625,
      "learning_rate": 7.714063918843106e-05,
      "loss": 0.7799,
      "step": 1035
    },
    {
      "epoch": 0.6189986840698047,
      "grad_norm": 0.049072265625,
      "learning_rate": 7.613051219968623e-05,
      "loss": 0.7611,
      "step": 1040
    },
    {
      "epoch": 0.6219746392816788,
      "grad_norm": 0.05029296875,
      "learning_rate": 7.512296138596601e-05,
      "loss": 0.7782,
      "step": 1045
    },
    {
      "epoch": 0.6249505944935528,
      "grad_norm": 0.048095703125,
      "learning_rate": 7.411809548974792e-05,
      "loss": 0.7671,
      "step": 1050
    },
    {
      "epoch": 0.6279265497054269,
      "grad_norm": 0.05078125,
      "learning_rate": 7.31160229637331e-05,
      "loss": 0.7753,
      "step": 1055
    },
    {
      "epoch": 0.630902504917301,
      "grad_norm": 0.05078125,
      "learning_rate": 7.211685195914097e-05,
      "loss": 0.7663,
      "step": 1060
    },
    {
      "epoch": 0.633878460129175,
      "grad_norm": 0.055419921875,
      "learning_rate": 7.112069031403704e-05,
      "loss": 0.7702,
      "step": 1065
    },
    {
      "epoch": 0.6368544153410491,
      "grad_norm": 0.0517578125,
      "learning_rate": 7.012764554169393e-05,
      "loss": 0.7724,
      "step": 1070
    },
    {
      "epoch": 0.6398303705529231,
      "grad_norm": 0.0478515625,
      "learning_rate": 6.913782481898789e-05,
      "loss": 0.7607,
      "step": 1075
    },
    {
      "epoch": 0.6428063257647972,
      "grad_norm": 0.049560546875,
      "learning_rate": 6.815133497483157e-05,
      "loss": 0.7879,
      "step": 1080
    },
    {
      "epoch": 0.6457822809766713,
      "grad_norm": 0.04931640625,
      "learning_rate": 6.71682824786439e-05,
      "loss": 0.7736,
      "step": 1085
    },
    {
      "epoch": 0.6487582361885453,
      "grad_norm": 0.049560546875,
      "learning_rate": 6.618877342885945e-05,
      "loss": 0.7452,
      "step": 1090
    },
    {
      "epoch": 0.6517341914004194,
      "grad_norm": 0.05126953125,
      "learning_rate": 6.521291354147727e-05,
      "loss": 0.7774,
      "step": 1095
    },
    {
      "epoch": 0.6547101466122934,
      "grad_norm": 0.0498046875,
      "learning_rate": 6.424080813865138e-05,
      "loss": 0.7825,
      "step": 1100
    },
    {
      "epoch": 0.6576861018241675,
      "grad_norm": 0.04736328125,
      "learning_rate": 6.327256213732344e-05,
      "loss": 0.7568,
      "step": 1105
    },
    {
      "epoch": 0.6606620570360416,
      "grad_norm": 0.048583984375,
      "learning_rate": 6.230828003789949e-05,
      "loss": 0.7697,
      "step": 1110
    },
    {
      "epoch": 0.6636380122479156,
      "grad_norm": 0.052490234375,
      "learning_rate": 6.134806591297133e-05,
      "loss": 0.7536,
      "step": 1115
    },
    {
      "epoch": 0.6666139674597897,
      "grad_norm": 0.04931640625,
      "learning_rate": 6.039202339608432e-05,
      "loss": 0.7799,
      "step": 1120
    },
    {
      "epoch": 0.6695899226716638,
      "grad_norm": 0.049560546875,
      "learning_rate": 5.944025567055251e-05,
      "loss": 0.7605,
      "step": 1125
    },
    {
      "epoch": 0.6725658778835378,
      "grad_norm": 0.051025390625,
      "learning_rate": 5.849286545832211e-05,
      "loss": 0.7747,
      "step": 1130
    },
    {
      "epoch": 0.6755418330954119,
      "grad_norm": 0.05078125,
      "learning_rate": 5.7549955008885294e-05,
      "loss": 0.7746,
      "step": 1135
    },
    {
      "epoch": 0.6785177883072859,
      "grad_norm": 0.04833984375,
      "learning_rate": 5.6611626088244194e-05,
      "loss": 0.7533,
      "step": 1140
    },
    {
      "epoch": 0.68149374351916,
      "grad_norm": 0.049072265625,
      "learning_rate": 5.567797996792813e-05,
      "loss": 0.7644,
      "step": 1145
    },
    {
      "epoch": 0.6844696987310341,
      "grad_norm": 0.048828125,
      "learning_rate": 5.47491174140631e-05,
      "loss": 0.7807,
      "step": 1150
    },
    {
      "epoch": 0.6874456539429081,
      "grad_norm": 0.047607421875,
      "learning_rate": 5.382513867649663e-05,
      "loss": 0.7398,
      "step": 1155
    },
    {
      "epoch": 0.6904216091547822,
      "grad_norm": 0.055908203125,
      "learning_rate": 5.290614347797802e-05,
      "loss": 0.8002,
      "step": 1160
    },
    {
      "epoch": 0.6933975643666562,
      "grad_norm": 0.047119140625,
      "learning_rate": 5.199223100339539e-05,
      "loss": 0.76,
      "step": 1165
    },
    {
      "epoch": 0.6963735195785303,
      "grad_norm": 0.050048828125,
      "learning_rate": 5.108349988907111e-05,
      "loss": 0.7712,
      "step": 1170
    },
    {
      "epoch": 0.6993494747904044,
      "grad_norm": 0.04931640625,
      "learning_rate": 5.0180048212115924e-05,
      "loss": 0.781,
      "step": 1175
    },
    {
      "epoch": 0.7023254300022784,
      "grad_norm": 0.0478515625,
      "learning_rate": 4.92819734798441e-05,
      "loss": 0.7758,
      "step": 1180
    },
    {
      "epoch": 0.7053013852141525,
      "grad_norm": 0.048828125,
      "learning_rate": 4.8389372619249326e-05,
      "loss": 0.7727,
      "step": 1185
    },
    {
      "epoch": 0.7082773404260266,
      "grad_norm": 0.048095703125,
      "learning_rate": 4.7502341966544e-05,
      "loss": 0.7605,
      "step": 1190
    },
    {
      "epoch": 0.7112532956379006,
      "grad_norm": 0.051513671875,
      "learning_rate": 4.6620977256761514e-05,
      "loss": 0.7649,
      "step": 1195
    },
    {
      "epoch": 0.7142292508497747,
      "grad_norm": 0.0478515625,
      "learning_rate": 4.574537361342407e-05,
      "loss": 0.7786,
      "step": 1200
    },
    {
      "epoch": 0.7172052060616487,
      "grad_norm": 0.04931640625,
      "learning_rate": 4.487562553827622e-05,
      "loss": 0.753,
      "step": 1205
    },
    {
      "epoch": 0.7201811612735228,
      "grad_norm": 0.048583984375,
      "learning_rate": 4.401182690108534e-05,
      "loss": 0.7849,
      "step": 1210
    },
    {
      "epoch": 0.7231571164853969,
      "grad_norm": 0.052978515625,
      "learning_rate": 4.315407092951078e-05,
      "loss": 0.7742,
      "step": 1215
    },
    {
      "epoch": 0.7261330716972709,
      "grad_norm": 0.05224609375,
      "learning_rate": 4.23024501990417e-05,
      "loss": 0.7956,
      "step": 1220
    },
    {
      "epoch": 0.729109026909145,
      "grad_norm": 0.0498046875,
      "learning_rate": 4.145705662300595e-05,
      "loss": 0.7837,
      "step": 1225
    },
    {
      "epoch": 0.732084982121019,
      "grad_norm": 0.05029296875,
      "learning_rate": 4.0617981442649855e-05,
      "loss": 0.7597,
      "step": 1230
    },
    {
      "epoch": 0.7350609373328931,
      "grad_norm": 0.048583984375,
      "learning_rate": 3.978531521729084e-05,
      "loss": 0.7705,
      "step": 1235
    },
    {
      "epoch": 0.7380368925447672,
      "grad_norm": 0.049560546875,
      "learning_rate": 3.89591478145437e-05,
      "loss": 0.7775,
      "step": 1240
    },
    {
      "epoch": 0.7410128477566412,
      "grad_norm": 0.05078125,
      "learning_rate": 3.813956840062118e-05,
      "loss": 0.7534,
      "step": 1245
    },
    {
      "epoch": 0.7439888029685153,
      "grad_norm": 0.04833984375,
      "learning_rate": 3.732666543071079e-05,
      "loss": 0.7725,
      "step": 1250
    },
    {
      "epoch": 0.7469647581803894,
      "grad_norm": 0.047119140625,
      "learning_rate": 3.652052663942769e-05,
      "loss": 0.7661,
      "step": 1255
    },
    {
      "epoch": 0.7499407133922634,
      "grad_norm": 0.05078125,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 0.7778,
      "step": 1260
    },
    {
      "epoch": 0.7529166686041375,
      "grad_norm": 0.048095703125,
      "learning_rate": 3.492888887160866e-05,
      "loss": 0.7704,
      "step": 1265
    },
    {
      "epoch": 0.7558926238160115,
      "grad_norm": 0.04736328125,
      "learning_rate": 3.4143561676616575e-05,
      "loss": 0.7785,
      "step": 1270
    },
    {
      "epoch": 0.7588685790278856,
      "grad_norm": 0.05615234375,
      "learning_rate": 3.336534220479961e-05,
      "loss": 0.7834,
      "step": 1275
    },
    {
      "epoch": 0.7618445342397597,
      "grad_norm": 0.051513671875,
      "learning_rate": 3.259431444746846e-05,
      "loss": 0.7778,
      "step": 1280
    },
    {
      "epoch": 0.7648204894516337,
      "grad_norm": 0.053955078125,
      "learning_rate": 3.1830561619749863e-05,
      "loss": 0.7658,
      "step": 1285
    },
    {
      "epoch": 0.7677964446635078,
      "grad_norm": 0.05029296875,
      "learning_rate": 3.10741661516053e-05,
      "loss": 0.7822,
      "step": 1290
    },
    {
      "epoch": 0.7707723998753819,
      "grad_norm": 0.049560546875,
      "learning_rate": 3.032520967893453e-05,
      "loss": 0.7763,
      "step": 1295
    },
    {
      "epoch": 0.7737483550872559,
      "grad_norm": 0.0498046875,
      "learning_rate": 2.9583773034764826e-05,
      "loss": 0.7855,
      "step": 1300
    },
    {
      "epoch": 0.77672431029913,
      "grad_norm": 0.049560546875,
      "learning_rate": 2.8849936240527008e-05,
      "loss": 0.754,
      "step": 1305
    },
    {
      "epoch": 0.779700265511004,
      "grad_norm": 0.047607421875,
      "learning_rate": 2.8123778497418685e-05,
      "loss": 0.772,
      "step": 1310
    },
    {
      "epoch": 0.7826762207228781,
      "grad_norm": 0.04736328125,
      "learning_rate": 2.74053781778565e-05,
      "loss": 0.7646,
      "step": 1315
    },
    {
      "epoch": 0.7856521759347522,
      "grad_norm": 0.050048828125,
      "learning_rate": 2.669481281701739e-05,
      "loss": 0.7848,
      "step": 1320
    },
    {
      "epoch": 0.7886281311466262,
      "grad_norm": 0.048583984375,
      "learning_rate": 2.5992159104470526e-05,
      "loss": 0.7679,
      "step": 1325
    },
    {
      "epoch": 0.7916040863585003,
      "grad_norm": 0.050537109375,
      "learning_rate": 2.529749287590042e-05,
      "loss": 0.7532,
      "step": 1330
    },
    {
      "epoch": 0.7945800415703743,
      "grad_norm": 0.048095703125,
      "learning_rate": 2.461088910492202e-05,
      "loss": 0.77,
      "step": 1335
    },
    {
      "epoch": 0.7975559967822484,
      "grad_norm": 0.0478515625,
      "learning_rate": 2.3932421894989167e-05,
      "loss": 0.7768,
      "step": 1340
    },
    {
      "epoch": 0.8005319519941225,
      "grad_norm": 0.050048828125,
      "learning_rate": 2.326216447139663e-05,
      "loss": 0.7543,
      "step": 1345
    },
    {
      "epoch": 0.8035079072059965,
      "grad_norm": 0.050048828125,
      "learning_rate": 2.260018917337726e-05,
      "loss": 0.7606,
      "step": 1350
    },
    {
      "epoch": 0.8064838624178706,
      "grad_norm": 0.047607421875,
      "learning_rate": 2.194656744629442e-05,
      "loss": 0.7629,
      "step": 1355
    },
    {
      "epoch": 0.8094598176297447,
      "grad_norm": 0.049560546875,
      "learning_rate": 2.1301369833931117e-05,
      "loss": 0.7619,
      "step": 1360
    },
    {
      "epoch": 0.8124357728416187,
      "grad_norm": 0.04931640625,
      "learning_rate": 2.0664665970876496e-05,
      "loss": 0.7534,
      "step": 1365
    },
    {
      "epoch": 0.8154117280534928,
      "grad_norm": 0.048828125,
      "learning_rate": 2.0036524575010172e-05,
      "loss": 0.7765,
      "step": 1370
    },
    {
      "epoch": 0.8183876832653668,
      "grad_norm": 0.052001953125,
      "learning_rate": 1.9417013440085864e-05,
      "loss": 0.7622,
      "step": 1375
    },
    {
      "epoch": 0.8213636384772409,
      "grad_norm": 0.04833984375,
      "learning_rate": 1.880619942841435e-05,
      "loss": 0.7795,
      "step": 1380
    },
    {
      "epoch": 0.824339593689115,
      "grad_norm": 0.049072265625,
      "learning_rate": 1.8204148463647453e-05,
      "loss": 0.7599,
      "step": 1385
    },
    {
      "epoch": 0.827315548900989,
      "grad_norm": 0.052734375,
      "learning_rate": 1.7610925523662835e-05,
      "loss": 0.7617,
      "step": 1390
    },
    {
      "epoch": 0.8302915041128631,
      "grad_norm": 0.047607421875,
      "learning_rate": 1.702659463355125e-05,
      "loss": 0.7608,
      "step": 1395
    },
    {
      "epoch": 0.8332674593247371,
      "grad_norm": 0.0478515625,
      "learning_rate": 1.6451218858706374e-05,
      "loss": 0.782,
      "step": 1400
    },
    {
      "epoch": 0.8362434145366112,
      "grad_norm": 0.049072265625,
      "learning_rate": 1.5884860298018322e-05,
      "loss": 0.7622,
      "step": 1405
    },
    {
      "epoch": 0.8392193697484853,
      "grad_norm": 0.04736328125,
      "learning_rate": 1.5327580077171587e-05,
      "loss": 0.771,
      "step": 1410
    },
    {
      "epoch": 0.8421953249603593,
      "grad_norm": 0.047119140625,
      "learning_rate": 1.4779438342047713e-05,
      "loss": 0.7684,
      "step": 1415
    },
    {
      "epoch": 0.8451712801722334,
      "grad_norm": 0.048583984375,
      "learning_rate": 1.4240494252234049e-05,
      "loss": 0.7886,
      "step": 1420
    },
    {
      "epoch": 0.8481472353841075,
      "grad_norm": 0.050537109375,
      "learning_rate": 1.3710805974638696e-05,
      "loss": 0.7813,
      "step": 1425
    },
    {
      "epoch": 0.8511231905959815,
      "grad_norm": 0.052978515625,
      "learning_rate": 1.3190430677212794e-05,
      "loss": 0.7557,
      "step": 1430
    },
    {
      "epoch": 0.8540991458078556,
      "grad_norm": 0.048095703125,
      "learning_rate": 1.2679424522780426e-05,
      "loss": 0.7487,
      "step": 1435
    },
    {
      "epoch": 0.8570751010197296,
      "grad_norm": 0.049560546875,
      "learning_rate": 1.2177842662977135e-05,
      "loss": 0.777,
      "step": 1440
    },
    {
      "epoch": 0.8600510562316037,
      "grad_norm": 0.050048828125,
      "learning_rate": 1.1685739232297643e-05,
      "loss": 0.7724,
      "step": 1445
    },
    {
      "epoch": 0.8630270114434778,
      "grad_norm": 0.050537109375,
      "learning_rate": 1.1203167342253062e-05,
      "loss": 0.7689,
      "step": 1450
    },
    {
      "epoch": 0.8660029666553518,
      "grad_norm": 0.054443359375,
      "learning_rate": 1.0730179075638868e-05,
      "loss": 0.7415,
      "step": 1455
    },
    {
      "epoch": 0.8689789218672259,
      "grad_norm": 0.049072265625,
      "learning_rate": 1.0266825480913611e-05,
      "loss": 0.7772,
      "step": 1460
    },
    {
      "epoch": 0.8719548770790999,
      "grad_norm": 0.04833984375,
      "learning_rate": 9.813156566689518e-06,
      "loss": 0.7715,
      "step": 1465
    },
    {
      "epoch": 0.874930832290974,
      "grad_norm": 0.04931640625,
      "learning_rate": 9.369221296335006e-06,
      "loss": 0.7646,
      "step": 1470
    },
    {
      "epoch": 0.8779067875028481,
      "grad_norm": 0.048583984375,
      "learning_rate": 8.935067582690382e-06,
      "loss": 0.7446,
      "step": 1475
    },
    {
      "epoch": 0.8808827427147221,
      "grad_norm": 0.0478515625,
      "learning_rate": 8.510742282896544e-06,
      "loss": 0.7768,
      "step": 1480
    },
    {
      "epoch": 0.8838586979265962,
      "grad_norm": 0.047607421875,
      "learning_rate": 8.096291193337934e-06,
      "loss": 0.7535,
      "step": 1485
    },
    {
      "epoch": 0.8868346531384703,
      "grad_norm": 0.048828125,
      "learning_rate": 7.69175904469982e-06,
      "loss": 0.7691,
      "step": 1490
    },
    {
      "epoch": 0.8898106083503443,
      "grad_norm": 0.050537109375,
      "learning_rate": 7.2971894971405665e-06,
      "loss": 0.7562,
      "step": 1495
    },
    {
      "epoch": 0.8927865635622184,
      "grad_norm": 0.05224609375,
      "learning_rate": 6.9126251355795864e-06,
      "loss": 0.7574,
      "step": 1500
    },
    {
      "epoch": 0.8957625187740924,
      "grad_norm": 0.04931640625,
      "learning_rate": 6.538107465101162e-06,
      "loss": 0.7733,
      "step": 1505
    },
    {
      "epoch": 0.8987384739859665,
      "grad_norm": 0.04736328125,
      "learning_rate": 6.173676906475012e-06,
      "loss": 0.7656,
      "step": 1510
    },
    {
      "epoch": 0.9017144291978406,
      "grad_norm": 0.04931640625,
      "learning_rate": 5.8193727917936536e-06,
      "loss": 0.7604,
      "step": 1515
    },
    {
      "epoch": 0.9046903844097146,
      "grad_norm": 0.04931640625,
      "learning_rate": 5.475233360227516e-06,
      "loss": 0.7688,
      "step": 1520
    },
    {
      "epoch": 0.9076663396215887,
      "grad_norm": 0.0478515625,
      "learning_rate": 5.14129575389779e-06,
      "loss": 0.7556,
      "step": 1525
    },
    {
      "epoch": 0.9106422948334627,
      "grad_norm": 0.0537109375,
      "learning_rate": 4.817596013867764e-06,
      "loss": 0.7765,
      "step": 1530
    },
    {
      "epoch": 0.9136182500453368,
      "grad_norm": 0.0517578125,
      "learning_rate": 4.504169076253084e-06,
      "loss": 0.7788,
      "step": 1535
    },
    {
      "epoch": 0.9165942052572109,
      "grad_norm": 0.04931640625,
      "learning_rate": 4.20104876845111e-06,
      "loss": 0.7723,
      "step": 1540
    },
    {
      "epoch": 0.9195701604690849,
      "grad_norm": 0.04736328125,
      "learning_rate": 3.908267805490051e-06,
      "loss": 0.7748,
      "step": 1545
    },
    {
      "epoch": 0.922546115680959,
      "grad_norm": 0.048095703125,
      "learning_rate": 3.625857786498055e-06,
      "loss": 0.762,
      "step": 1550
    },
    {
      "epoch": 0.9255220708928331,
      "grad_norm": 0.049560546875,
      "learning_rate": 3.3538491912928792e-06,
      "loss": 0.7728,
      "step": 1555
    },
    {
      "epoch": 0.9284980261047071,
      "grad_norm": 0.047119140625,
      "learning_rate": 3.092271377092215e-06,
      "loss": 0.7595,
      "step": 1560
    },
    {
      "epoch": 0.9314739813165812,
      "grad_norm": 0.049072265625,
      "learning_rate": 2.8411525753452185e-06,
      "loss": 0.7776,
      "step": 1565
    },
    {
      "epoch": 0.9344499365284552,
      "grad_norm": 0.07470703125,
      "learning_rate": 2.6005198886856487e-06,
      "loss": 0.7814,
      "step": 1570
    },
    {
      "epoch": 0.9374258917403293,
      "grad_norm": 0.05224609375,
      "learning_rate": 2.3703992880066638e-06,
      "loss": 0.7533,
      "step": 1575
    },
    {
      "epoch": 0.9404018469522034,
      "grad_norm": 0.049072265625,
      "learning_rate": 2.150815609657875e-06,
      "loss": 0.7801,
      "step": 1580
    },
    {
      "epoch": 0.9433778021640774,
      "grad_norm": 0.05126953125,
      "learning_rate": 1.9417925527648096e-06,
      "loss": 0.766,
      "step": 1585
    },
    {
      "epoch": 0.9463537573759515,
      "grad_norm": 0.048095703125,
      "learning_rate": 1.7433526766711728e-06,
      "loss": 0.7911,
      "step": 1590
    },
    {
      "epoch": 0.9493297125878255,
      "grad_norm": 0.051513671875,
      "learning_rate": 1.5555173985039918e-06,
      "loss": 0.7798,
      "step": 1595
    },
    {
      "epoch": 0.9523056677996996,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.378306990862177e-06,
      "loss": 0.7833,
      "step": 1600
    },
    {
      "epoch": 0.9552816230115737,
      "grad_norm": 0.046875,
      "learning_rate": 1.2117405796285286e-06,
      "loss": 0.7826,
      "step": 1605
    },
    {
      "epoch": 0.9582575782234477,
      "grad_norm": 0.055908203125,
      "learning_rate": 1.055836141905553e-06,
      "loss": 0.7792,
      "step": 1610
    },
    {
      "epoch": 0.9612335334353218,
      "grad_norm": 0.0517578125,
      "learning_rate": 9.106105040751822e-07,
      "loss": 0.7687,
      "step": 1615
    },
    {
      "epoch": 0.9642094886471959,
      "grad_norm": 0.048583984375,
      "learning_rate": 7.760793399827937e-07,
      "loss": 0.7452,
      "step": 1620
    },
    {
      "epoch": 0.9671854438590699,
      "grad_norm": 0.0576171875,
      "learning_rate": 6.522571692455736e-07,
      "loss": 0.7721,
      "step": 1625
    },
    {
      "epoch": 0.970161399070944,
      "grad_norm": 0.1337890625,
      "learning_rate": 5.391573556854157e-07,
      "loss": 0.7848,
      "step": 1630
    },
    {
      "epoch": 0.973137354282818,
      "grad_norm": 0.048583984375,
      "learning_rate": 4.3679210588661866e-07,
      "loss": 0.7767,
      "step": 1635
    },
    {
      "epoch": 0.9761133094946921,
      "grad_norm": 0.048095703125,
      "learning_rate": 3.451724678784518e-07,
      "loss": 0.7726,
      "step": 1640
    },
    {
      "epoch": 0.9790892647065662,
      "grad_norm": 0.04931640625,
      "learning_rate": 2.643083299427751e-07,
      "loss": 0.7683,
      "step": 1645
    },
    {
      "epoch": 0.9820652199184402,
      "grad_norm": 0.054931640625,
      "learning_rate": 1.9420841954681525e-07,
      "loss": 0.7589,
      "step": 1650
    },
    {
      "epoch": 0.9850411751303143,
      "grad_norm": 0.05078125,
      "learning_rate": 1.3488030240123017e-07,
      "loss": 0.7689,
      "step": 1655
    },
    {
      "epoch": 0.9880171303421883,
      "grad_norm": 0.052001953125,
      "learning_rate": 8.633038164358454e-08,
      "loss": 0.7738,
      "step": 1660
    },
    {
      "epoch": 0.9909930855540624,
      "grad_norm": 0.07275390625,
      "learning_rate": 4.856389714723575e-08,
      "loss": 0.7728,
      "step": 1665
    },
    {
      "epoch": 0.9939690407659365,
      "grad_norm": 0.05322265625,
      "learning_rate": 2.1584924955819764e-08,
      "loss": 0.7646,
      "step": 1670
    },
    {
      "epoch": 0.9969449959778105,
      "grad_norm": 0.04931640625,
      "learning_rate": 5.396376843369577e-09,
      "loss": 0.7784,
      "step": 1675
    },
    {
      "epoch": 0.9999209511896846,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0,
      "loss": 0.7765,
      "step": 1680
    },
    {
      "epoch": 0.9999209511896846,
      "step": 1680,
      "total_flos": 4.668038789401149e+18,
      "train_loss": 0.0,
      "train_runtime": 0.023,
      "train_samples_per_second": 9337043.614,
      "train_steps_per_second": 72939.887
    }
  ],
  "logging_steps": 5,
  "max_steps": 1680,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "total_flos": 4.668038789401149e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}