{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4323,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002313208420078649,
      "grad_norm": 1.0980443954467773,
      "learning_rate": 0.0,
      "loss": 3.630915880203247,
      "step": 1
    },
    {
      "epoch": 0.0004626416840157298,
      "grad_norm": 1.1913707256317139,
      "learning_rate": 6.25e-06,
      "loss": 3.6790144443511963,
      "step": 2
    },
    {
      "epoch": 0.0006939625260235947,
      "grad_norm": 1.5132479667663574,
      "learning_rate": 1.25e-05,
      "loss": 3.6685311794281006,
      "step": 3
    },
    {
      "epoch": 0.0009252833680314596,
      "grad_norm": 1.5550132989883423,
      "learning_rate": 1.875e-05,
      "loss": 4.382204055786133,
      "step": 4
    },
    {
      "epoch": 0.0011566042100393246,
      "grad_norm": 1.1964272260665894,
      "learning_rate": 2.5e-05,
      "loss": 3.619861125946045,
      "step": 5
    },
    {
      "epoch": 0.0013879250520471894,
      "grad_norm": 1.6822092533111572,
      "learning_rate": 3.125e-05,
      "loss": 3.8901073932647705,
      "step": 6
    },
    {
      "epoch": 0.0016192458940550544,
      "grad_norm": 1.049480676651001,
      "learning_rate": 3.75e-05,
      "loss": 4.221622467041016,
      "step": 7
    },
    {
      "epoch": 0.0018505667360629193,
      "grad_norm": 1.2558367252349854,
      "learning_rate": 4.375e-05,
      "loss": 4.001527309417725,
      "step": 8
    },
    {
      "epoch": 0.002081887578070784,
      "grad_norm": 1.0762017965316772,
      "learning_rate": 5e-05,
      "loss": 4.221858024597168,
      "step": 9
    },
    {
      "epoch": 0.002313208420078649,
      "grad_norm": 1.5011881589889526,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 3.6632180213928223,
      "step": 10
    },
    {
      "epoch": 0.002544529262086514,
      "grad_norm": 1.3207355737686157,
      "learning_rate": 6.25e-05,
      "loss": 3.64652419090271,
      "step": 11
    },
    {
      "epoch": 0.002775850104094379,
      "grad_norm": 1.455620288848877,
      "learning_rate": 6.875e-05,
      "loss": 3.9262332916259766,
      "step": 12
    },
    {
      "epoch": 0.003007170946102244,
      "grad_norm": 1.1915875673294067,
      "learning_rate": 7.5e-05,
      "loss": 3.996861457824707,
      "step": 13
    },
    {
      "epoch": 0.003238491788110109,
      "grad_norm": 1.163299560546875,
      "learning_rate": 8.125000000000001e-05,
      "loss": 4.0614094734191895,
      "step": 14
    },
    {
      "epoch": 0.0034698126301179735,
      "grad_norm": 1.2658706903457642,
      "learning_rate": 8.75e-05,
      "loss": 3.929135799407959,
      "step": 15
    },
    {
      "epoch": 0.0037011334721258385,
      "grad_norm": 1.3617708683013916,
      "learning_rate": 9.375e-05,
      "loss": 3.9676480293273926,
      "step": 16
    },
    {
      "epoch": 0.003932454314133704,
      "grad_norm": 1.1154868602752686,
      "learning_rate": 0.0001,
      "loss": 3.7171144485473633,
      "step": 17
    },
    {
      "epoch": 0.004163775156141568,
      "grad_norm": 1.0891473293304443,
      "learning_rate": 0.00010625,
      "loss": 3.791104793548584,
      "step": 18
    },
    {
      "epoch": 0.004395095998149434,
      "grad_norm": 1.14180326461792,
      "learning_rate": 0.00011250000000000001,
      "loss": 3.6146225929260254,
      "step": 19
    },
    {
      "epoch": 0.004626416840157298,
      "grad_norm": 1.6510698795318604,
      "learning_rate": 0.00011875,
      "loss": 3.7250571250915527,
      "step": 20
    },
    {
      "epoch": 0.004857737682165163,
      "grad_norm": 3.269594192504883,
      "learning_rate": 0.000125,
      "loss": 4.009101867675781,
      "step": 21
    },
    {
      "epoch": 0.005089058524173028,
      "grad_norm": 1.157247543334961,
      "learning_rate": 0.00013125000000000002,
      "loss": 4.1006855964660645,
      "step": 22
    },
    {
      "epoch": 0.005320379366180893,
      "grad_norm": 1.254481315612793,
      "learning_rate": 0.0001375,
      "loss": 3.843214511871338,
      "step": 23
    },
    {
      "epoch": 0.005551700208188758,
      "grad_norm": 1.3317657709121704,
      "learning_rate": 0.00014375,
      "loss": 3.9063687324523926,
      "step": 24
    },
    {
      "epoch": 0.005783021050196623,
      "grad_norm": 1.9009345769882202,
      "learning_rate": 0.00015,
      "loss": 4.048081398010254,
      "step": 25
    },
    {
      "epoch": 0.006014341892204488,
      "grad_norm": 1.25008225440979,
      "learning_rate": 0.00015625,
      "loss": 3.9562292098999023,
      "step": 26
    },
    {
      "epoch": 0.006245662734212352,
      "grad_norm": 1.450724482536316,
      "learning_rate": 0.00016250000000000002,
      "loss": 4.126107692718506,
      "step": 27
    },
    {
      "epoch": 0.006476983576220218,
      "grad_norm": 1.4373672008514404,
      "learning_rate": 0.00016875,
      "loss": 4.004990577697754,
      "step": 28
    },
    {
      "epoch": 0.006708304418228082,
      "grad_norm": 1.18876051902771,
      "learning_rate": 0.000175,
      "loss": 3.69765305519104,
      "step": 29
    },
    {
      "epoch": 0.006939625260235947,
      "grad_norm": 1.339059829711914,
      "learning_rate": 0.00018125,
      "loss": 4.102259635925293,
      "step": 30
    },
    {
      "epoch": 0.0071709461022438125,
      "grad_norm": 1.0883063077926636,
      "learning_rate": 0.0001875,
      "loss": 4.150142669677734,
      "step": 31
    },
    {
      "epoch": 0.007402266944251677,
      "grad_norm": 1.2718650102615356,
      "learning_rate": 0.00019375000000000002,
      "loss": 3.764147996902466,
      "step": 32
    },
    {
      "epoch": 0.007633587786259542,
      "grad_norm": 1.2919607162475586,
      "learning_rate": 0.0002,
      "loss": 3.4292685985565186,
      "step": 33
    },
    {
      "epoch": 0.007864908628267407,
      "grad_norm": 1.0628468990325928,
      "learning_rate": 0.00020625,
      "loss": 3.781846284866333,
      "step": 34
    },
    {
      "epoch": 0.008096229470275272,
      "grad_norm": 1.4626634120941162,
      "learning_rate": 0.0002125,
      "loss": 3.6269216537475586,
      "step": 35
    },
    {
      "epoch": 0.008327550312283136,
      "grad_norm": 1.4257652759552002,
      "learning_rate": 0.00021875,
      "loss": 4.114569664001465,
      "step": 36
    },
    {
      "epoch": 0.008558871154291001,
      "grad_norm": 1.4326950311660767,
      "learning_rate": 0.00022500000000000002,
      "loss": 3.563236713409424,
      "step": 37
    },
    {
      "epoch": 0.008790191996298867,
      "grad_norm": 1.2933130264282227,
      "learning_rate": 0.00023125,
      "loss": 3.698005199432373,
      "step": 38
    },
    {
      "epoch": 0.009021512838306732,
      "grad_norm": 1.6034282445907593,
      "learning_rate": 0.0002375,
      "loss": 3.3416218757629395,
      "step": 39
    },
    {
      "epoch": 0.009252833680314597,
      "grad_norm": 1.0949491262435913,
      "learning_rate": 0.00024375,
      "loss": 3.58945369720459,
      "step": 40
    },
    {
      "epoch": 0.009484154522322461,
      "grad_norm": 1.3464919328689575,
      "learning_rate": 0.00025,
      "loss": 3.811768054962158,
      "step": 41
    },
    {
      "epoch": 0.009715475364330326,
      "grad_norm": 1.6597055196762085,
      "learning_rate": 0.00025624999999999997,
      "loss": 3.4442296028137207,
      "step": 42
    },
    {
      "epoch": 0.00994679620633819,
      "grad_norm": 1.4191656112670898,
      "learning_rate": 0.00026250000000000004,
      "loss": 3.4365038871765137,
      "step": 43
    },
    {
      "epoch": 0.010178117048346057,
      "grad_norm": 1.1580950021743774,
      "learning_rate": 0.00026875,
      "loss": 3.8890414237976074,
      "step": 44
    },
    {
      "epoch": 0.010409437890353921,
      "grad_norm": 1.2106267213821411,
      "learning_rate": 0.000275,
      "loss": 3.7391891479492188,
      "step": 45
    },
    {
      "epoch": 0.010640758732361786,
      "grad_norm": 1.3837344646453857,
      "learning_rate": 0.00028125000000000003,
      "loss": 3.957463264465332,
      "step": 46
    },
    {
      "epoch": 0.01087207957436965,
      "grad_norm": 1.4573487043380737,
      "learning_rate": 0.0002875,
      "loss": 3.698181390762329,
      "step": 47
    },
    {
      "epoch": 0.011103400416377515,
      "grad_norm": 1.3247969150543213,
      "learning_rate": 0.00029375,
      "loss": 3.68148136138916,
      "step": 48
    },
    {
      "epoch": 0.01133472125838538,
      "grad_norm": 1.314113974571228,
      "learning_rate": 0.0003,
      "loss": 3.595766067504883,
      "step": 49
    },
    {
      "epoch": 0.011566042100393246,
      "grad_norm": 1.345504641532898,
      "learning_rate": 0.00030625000000000004,
      "loss": 3.5795788764953613,
      "step": 50
    },
    {
      "epoch": 0.01179736294240111,
      "grad_norm": 1.3347457647323608,
      "learning_rate": 0.0003125,
      "loss": 3.8384909629821777,
      "step": 51
    },
    {
      "epoch": 0.012028683784408975,
      "grad_norm": 1.1891497373580933,
      "learning_rate": 0.00031874999999999997,
      "loss": 3.6292660236358643,
      "step": 52
    },
    {
      "epoch": 0.01226000462641684,
      "grad_norm": 1.1601086854934692,
      "learning_rate": 0.00032500000000000004,
      "loss": 3.6731700897216797,
      "step": 53
    },
    {
      "epoch": 0.012491325468424705,
      "grad_norm": 1.1167899370193481,
      "learning_rate": 0.00033125,
      "loss": 3.8840532302856445,
      "step": 54
    },
    {
      "epoch": 0.01272264631043257,
      "grad_norm": 1.544752836227417,
      "learning_rate": 0.0003375,
      "loss": 3.54325008392334,
      "step": 55
    },
    {
      "epoch": 0.012953967152440436,
      "grad_norm": 1.2400782108306885,
      "learning_rate": 0.00034375,
      "loss": 3.5325911045074463,
      "step": 56
    },
    {
      "epoch": 0.0131852879944483,
      "grad_norm": 1.5693418979644775,
      "learning_rate": 0.00035,
      "loss": 3.4677767753601074,
      "step": 57
    },
    {
      "epoch": 0.013416608836456165,
      "grad_norm": 1.1658480167388916,
      "learning_rate": 0.00035625,
      "loss": 3.57145357131958,
      "step": 58
    },
    {
      "epoch": 0.01364792967846403,
      "grad_norm": 1.358254075050354,
      "learning_rate": 0.0003625,
      "loss": 3.983759880065918,
      "step": 59
    },
    {
      "epoch": 0.013879250520471894,
      "grad_norm": 1.228789210319519,
      "learning_rate": 0.00036875000000000005,
      "loss": 3.693547248840332,
      "step": 60
    },
    {
      "epoch": 0.014110571362479759,
      "grad_norm": 1.42012619972229,
      "learning_rate": 0.000375,
      "loss": 4.0521111488342285,
      "step": 61
    },
    {
      "epoch": 0.014341892204487625,
      "grad_norm": 1.2809078693389893,
      "learning_rate": 0.00038124999999999997,
      "loss": 3.8445796966552734,
      "step": 62
    },
    {
      "epoch": 0.01457321304649549,
      "grad_norm": 1.2814528942108154,
      "learning_rate": 0.00038750000000000004,
      "loss": 3.58152437210083,
      "step": 63
    },
    {
      "epoch": 0.014804533888503354,
      "grad_norm": 1.350693941116333,
      "learning_rate": 0.00039375,
      "loss": 3.4387731552124023,
      "step": 64
    },
    {
      "epoch": 0.015035854730511219,
      "grad_norm": 1.3500319719314575,
      "learning_rate": 0.0004,
      "loss": 3.6287636756896973,
      "step": 65
    },
    {
      "epoch": 0.015267175572519083,
      "grad_norm": 1.450490117073059,
      "learning_rate": 0.00040625000000000004,
      "loss": 3.7270336151123047,
      "step": 66
    },
    {
      "epoch": 0.01549849641452695,
      "grad_norm": 1.1940770149230957,
      "learning_rate": 0.0004125,
      "loss": 4.285892486572266,
      "step": 67
    },
    {
      "epoch": 0.015729817256534814,
      "grad_norm": 1.1883995532989502,
      "learning_rate": 0.00041875,
      "loss": 3.6085071563720703,
      "step": 68
    },
    {
      "epoch": 0.015961138098542677,
      "grad_norm": 1.2914646863937378,
      "learning_rate": 0.000425,
      "loss": 3.7821691036224365,
      "step": 69
    },
    {
      "epoch": 0.016192458940550544,
      "grad_norm": 1.7341183423995972,
      "learning_rate": 0.00043125000000000005,
      "loss": 3.7667596340179443,
      "step": 70
    },
    {
      "epoch": 0.01642377978255841,
      "grad_norm": 1.1154019832611084,
      "learning_rate": 0.0004375,
      "loss": 3.765030860900879,
      "step": 71
    },
    {
      "epoch": 0.016655100624566273,
      "grad_norm": 1.2280501127243042,
      "learning_rate": 0.00044374999999999997,
      "loss": 4.04004430770874,
      "step": 72
    },
    {
      "epoch": 0.01688642146657414,
      "grad_norm": 2.253908157348633,
      "learning_rate": 0.00045000000000000004,
      "loss": 3.792429208755493,
      "step": 73
    },
    {
      "epoch": 0.017117742308582002,
      "grad_norm": 1.8846203088760376,
      "learning_rate": 0.00045625,
      "loss": 3.8622183799743652,
      "step": 74
    },
    {
      "epoch": 0.01734906315058987,
      "grad_norm": 1.222287893295288,
      "learning_rate": 0.0004625,
      "loss": 3.8600146770477295,
      "step": 75
    },
    {
      "epoch": 0.017580383992597735,
      "grad_norm": 1.440513253211975,
      "learning_rate": 0.00046875,
      "loss": 3.7536892890930176,
      "step": 76
    },
    {
      "epoch": 0.017811704834605598,
      "grad_norm": 1.0279744863510132,
      "learning_rate": 0.000475,
      "loss": 3.894153356552124,
      "step": 77
    },
    {
      "epoch": 0.018043025676613464,
      "grad_norm": 1.4716325998306274,
      "learning_rate": 0.00048125,
      "loss": 4.035886287689209,
      "step": 78
    },
    {
      "epoch": 0.018274346518621327,
      "grad_norm": 1.3427977561950684,
      "learning_rate": 0.0004875,
      "loss": 3.6394386291503906,
      "step": 79
    },
    {
      "epoch": 0.018505667360629193,
      "grad_norm": 1.6241545677185059,
      "learning_rate": 0.00049375,
      "loss": 3.798891544342041,
      "step": 80
    },
    {
      "epoch": 0.018736988202637056,
      "grad_norm": 1.4585132598876953,
      "learning_rate": 0.0005,
      "loss": 3.665830612182617,
      "step": 81
    },
    {
      "epoch": 0.018968309044644922,
      "grad_norm": 1.5634437799453735,
      "learning_rate": 0.0005,
      "loss": 3.9834561347961426,
      "step": 82
    },
    {
      "epoch": 0.01919962988665279,
      "grad_norm": 1.7607983350753784,
      "learning_rate": 0.0005,
      "loss": 3.6046581268310547,
      "step": 83
    },
    {
      "epoch": 0.01943095072866065,
      "grad_norm": 1.4759695529937744,
      "learning_rate": 0.0005,
      "loss": 3.798936367034912,
      "step": 84
    },
    {
      "epoch": 0.019662271570668518,
      "grad_norm": 1.3697260618209839,
      "learning_rate": 0.0005,
      "loss": 3.8513011932373047,
      "step": 85
    },
    {
      "epoch": 0.01989359241267638,
      "grad_norm": 1.1125731468200684,
      "learning_rate": 0.0005,
      "loss": 3.6278319358825684,
      "step": 86
    },
    {
      "epoch": 0.020124913254684247,
      "grad_norm": 1.097348690032959,
      "learning_rate": 0.0005,
      "loss": 3.5492212772369385,
      "step": 87
    },
    {
      "epoch": 0.020356234096692113,
      "grad_norm": 1.3522193431854248,
      "learning_rate": 0.0005,
      "loss": 3.803060531616211,
      "step": 88
    },
    {
      "epoch": 0.020587554938699976,
      "grad_norm": 1.3325737714767456,
      "learning_rate": 0.0005,
      "loss": 3.683772087097168,
      "step": 89
    },
    {
      "epoch": 0.020818875780707843,
      "grad_norm": 1.3217281103134155,
      "learning_rate": 0.0005,
      "loss": 3.905658721923828,
      "step": 90
    },
    {
      "epoch": 0.021050196622715706,
      "grad_norm": 1.9629470109939575,
      "learning_rate": 0.0005,
      "loss": 3.5574240684509277,
      "step": 91
    },
    {
      "epoch": 0.021281517464723572,
      "grad_norm": 1.3187389373779297,
      "learning_rate": 0.0005,
      "loss": 3.5709643363952637,
      "step": 92
    },
    {
      "epoch": 0.021512838306731435,
      "grad_norm": 1.4058637619018555,
      "learning_rate": 0.0005,
      "loss": 3.4003310203552246,
      "step": 93
    },
    {
      "epoch": 0.0217441591487393,
      "grad_norm": 1.0921101570129395,
      "learning_rate": 0.0005,
      "loss": 3.7916738986968994,
      "step": 94
    },
    {
      "epoch": 0.021975479990747167,
      "grad_norm": 1.4782453775405884,
      "learning_rate": 0.0005,
      "loss": 3.7563936710357666,
      "step": 95
    },
    {
      "epoch": 0.02220680083275503,
      "grad_norm": 1.1193164587020874,
      "learning_rate": 0.0005,
      "loss": 3.6431007385253906,
      "step": 96
    },
    {
      "epoch": 0.022438121674762897,
      "grad_norm": 1.1790590286254883,
      "learning_rate": 0.0005,
      "loss": 3.9700825214385986,
      "step": 97
    },
    {
      "epoch": 0.02266944251677076,
      "grad_norm": 1.5634790658950806,
      "learning_rate": 0.0005,
      "loss": 4.013468265533447,
      "step": 98
    },
    {
      "epoch": 0.022900763358778626,
      "grad_norm": 1.3258907794952393,
      "learning_rate": 0.0005,
      "loss": 3.719529628753662,
      "step": 99
    },
    {
      "epoch": 0.023132084200786492,
      "grad_norm": 1.7931456565856934,
      "learning_rate": 0.0005,
      "loss": 3.7560083866119385,
      "step": 100
    },
    {
      "epoch": 0.023363405042794355,
      "grad_norm": 2.4237895011901855,
      "learning_rate": 0.0005,
      "loss": 3.9130711555480957,
      "step": 101
    },
    {
      "epoch": 0.02359472588480222,
      "grad_norm": 2.128514289855957,
      "learning_rate": 0.0005,
      "loss": 3.91335129737854,
      "step": 102
    },
    {
      "epoch": 0.023826046726810084,
      "grad_norm": 1.4869989156723022,
      "learning_rate": 0.0005,
      "loss": 4.018052577972412,
      "step": 103
    },
    {
      "epoch": 0.02405736756881795,
      "grad_norm": 1.3899801969528198,
      "learning_rate": 0.0005,
      "loss": 3.854789972305298,
      "step": 104
    },
    {
      "epoch": 0.024288688410825817,
      "grad_norm": 1.3214784860610962,
      "learning_rate": 0.0005,
      "loss": 3.8494272232055664,
      "step": 105
    },
    {
      "epoch": 0.02452000925283368,
      "grad_norm": 1.5434843301773071,
      "learning_rate": 0.0005,
      "loss": 3.9365949630737305,
      "step": 106
    },
    {
      "epoch": 0.024751330094841546,
      "grad_norm": 1.2162518501281738,
      "learning_rate": 0.0005,
      "loss": 3.727238655090332,
      "step": 107
    },
    {
      "epoch": 0.02498265093684941,
      "grad_norm": 1.3018949031829834,
      "learning_rate": 0.0005,
      "loss": 4.198115348815918,
      "step": 108
    },
    {
      "epoch": 0.025213971778857276,
      "grad_norm": 1.498748779296875,
      "learning_rate": 0.0005,
      "loss": 3.911452531814575,
      "step": 109
    },
    {
      "epoch": 0.02544529262086514,
      "grad_norm": 1.1833291053771973,
      "learning_rate": 0.0005,
      "loss": 3.9802184104919434,
      "step": 110
    },
    {
      "epoch": 0.025676613462873005,
      "grad_norm": 1.3947339057922363,
      "learning_rate": 0.0005,
      "loss": 3.4913883209228516,
      "step": 111
    },
    {
      "epoch": 0.02590793430488087,
      "grad_norm": 1.4024978876113892,
      "learning_rate": 0.0005,
      "loss": 3.6772336959838867,
      "step": 112
    },
    {
      "epoch": 0.026139255146888734,
      "grad_norm": 1.69575035572052,
      "learning_rate": 0.0005,
      "loss": 3.734638214111328,
      "step": 113
    },
    {
      "epoch": 0.0263705759888966,
      "grad_norm": 1.9118092060089111,
      "learning_rate": 0.0005,
      "loss": 3.844080924987793,
      "step": 114
    },
    {
      "epoch": 0.026601896830904463,
      "grad_norm": 1.4185736179351807,
      "learning_rate": 0.0005,
      "loss": 3.6559228897094727,
      "step": 115
    },
    {
      "epoch": 0.02683321767291233,
      "grad_norm": 1.645930290222168,
      "learning_rate": 0.0005,
      "loss": 3.7015647888183594,
      "step": 116
    },
    {
      "epoch": 0.027064538514920196,
      "grad_norm": 1.538001298904419,
      "learning_rate": 0.0005,
      "loss": 3.5293326377868652,
      "step": 117
    },
    {
      "epoch": 0.02729585935692806,
      "grad_norm": 1.844510793685913,
      "learning_rate": 0.0005,
      "loss": 3.7847952842712402,
      "step": 118
    },
    {
      "epoch": 0.027527180198935925,
      "grad_norm": 2.1496009826660156,
      "learning_rate": 0.0005,
      "loss": 4.144896507263184,
      "step": 119
    },
    {
      "epoch": 0.027758501040943788,
      "grad_norm": 1.3942950963974,
      "learning_rate": 0.0005,
      "loss": 4.260809898376465,
      "step": 120
    },
    {
      "epoch": 0.027989821882951654,
      "grad_norm": 1.7220059633255005,
      "learning_rate": 0.0005,
      "loss": 4.069623947143555,
      "step": 121
    },
    {
      "epoch": 0.028221142724959517,
      "grad_norm": 1.5820496082305908,
      "learning_rate": 0.0005,
      "loss": 3.26920747756958,
      "step": 122
    },
    {
      "epoch": 0.028452463566967384,
      "grad_norm": 1.3749690055847168,
      "learning_rate": 0.0005,
      "loss": 3.8135147094726562,
      "step": 123
    },
    {
      "epoch": 0.02868378440897525,
      "grad_norm": 1.2897250652313232,
      "learning_rate": 0.0005,
      "loss": 3.9737792015075684,
      "step": 124
    },
    {
      "epoch": 0.028915105250983113,
      "grad_norm": 2.12416934967041,
      "learning_rate": 0.0005,
      "loss": 4.238428592681885,
      "step": 125
    },
    {
      "epoch": 0.02914642609299098,
      "grad_norm": 2.2014167308807373,
      "learning_rate": 0.0005,
      "loss": 3.8649439811706543,
      "step": 126
    },
    {
      "epoch": 0.029377746934998842,
      "grad_norm": 1.7370049953460693,
      "learning_rate": 0.0005,
      "loss": 4.128363609313965,
      "step": 127
    },
    {
      "epoch": 0.02960906777700671,
      "grad_norm": 1.2678942680358887,
      "learning_rate": 0.0005,
      "loss": 3.5398125648498535,
      "step": 128
    },
    {
      "epoch": 0.029840388619014575,
      "grad_norm": 1.7493218183517456,
      "learning_rate": 0.0005,
      "loss": 3.6316237449645996,
      "step": 129
    },
    {
      "epoch": 0.030071709461022438,
      "grad_norm": 1.4135515689849854,
      "learning_rate": 0.0005,
      "loss": 3.800339460372925,
      "step": 130
    },
    {
      "epoch": 0.030303030303030304,
      "grad_norm": 1.1146379709243774,
      "learning_rate": 0.0005,
      "loss": 4.048551559448242,
      "step": 131
    },
    {
      "epoch": 0.030534351145038167,
      "grad_norm": 1.6295815706253052,
      "learning_rate": 0.0005,
      "loss": 3.54392671585083,
      "step": 132
    },
    {
      "epoch": 0.030765671987046033,
      "grad_norm": 1.2632558345794678,
      "learning_rate": 0.0005,
      "loss": 3.725797176361084,
      "step": 133
    },
    {
      "epoch": 0.0309969928290539,
      "grad_norm": 1.6379729509353638,
      "learning_rate": 0.0005,
      "loss": 3.529371738433838,
      "step": 134
    },
    {
      "epoch": 0.031228313671061762,
      "grad_norm": 1.3477774858474731,
      "learning_rate": 0.0005,
      "loss": 3.5768280029296875,
      "step": 135
    },
    {
      "epoch": 0.03145963451306963,
      "grad_norm": 1.212815523147583,
      "learning_rate": 0.0005,
      "loss": 3.724019765853882,
      "step": 136
    },
    {
      "epoch": 0.031690955355077495,
      "grad_norm": 1.3901617527008057,
      "learning_rate": 0.0005,
      "loss": 3.5941848754882812,
      "step": 137
    },
    {
      "epoch": 0.031922276197085354,
      "grad_norm": 1.3269671201705933,
      "learning_rate": 0.0005,
      "loss": 3.440777540206909,
      "step": 138
    },
    {
      "epoch": 0.03215359703909322,
      "grad_norm": 1.4525771141052246,
      "learning_rate": 0.0005,
      "loss": 3.946812391281128,
      "step": 139
    },
    {
      "epoch": 0.03238491788110109,
      "grad_norm": 1.9113337993621826,
      "learning_rate": 0.0005,
      "loss": 3.8238115310668945,
      "step": 140
    },
    {
      "epoch": 0.03261623872310895,
      "grad_norm": 1.3651715517044067,
      "learning_rate": 0.0005,
      "loss": 3.354266881942749,
      "step": 141
    },
    {
      "epoch": 0.03284755956511682,
      "grad_norm": 1.4754585027694702,
      "learning_rate": 0.0005,
      "loss": 3.3538920879364014,
      "step": 142
    },
    {
      "epoch": 0.03307888040712468,
      "grad_norm": 1.4082255363464355,
      "learning_rate": 0.0005,
      "loss": 3.9997222423553467,
      "step": 143
    },
    {
      "epoch": 0.033310201249132546,
      "grad_norm": 1.2067192792892456,
      "learning_rate": 0.0005,
      "loss": 3.815899610519409,
      "step": 144
    },
    {
      "epoch": 0.03354152209114041,
      "grad_norm": 1.3697025775909424,
      "learning_rate": 0.0005,
      "loss": 3.6782186031341553,
      "step": 145
    },
    {
      "epoch": 0.03377284293314828,
      "grad_norm": 1.2347638607025146,
      "learning_rate": 0.0005,
      "loss": 3.7621185779571533,
      "step": 146
    },
    {
      "epoch": 0.034004163775156145,
      "grad_norm": 1.7865718603134155,
      "learning_rate": 0.0005,
      "loss": 3.6559786796569824,
      "step": 147
    },
    {
      "epoch": 0.034235484617164004,
      "grad_norm": 1.4139426946640015,
      "learning_rate": 0.0005,
      "loss": 3.5772478580474854,
      "step": 148
    },
    {
      "epoch": 0.03446680545917187,
      "grad_norm": 1.063269019126892,
      "learning_rate": 0.0005,
      "loss": 3.7564098834991455,
      "step": 149
    },
    {
      "epoch": 0.03469812630117974,
      "grad_norm": 1.4309043884277344,
      "learning_rate": 0.0005,
      "loss": 3.5627713203430176,
      "step": 150
    },
    {
      "epoch": 0.0349294471431876,
      "grad_norm": 1.8701832294464111,
      "learning_rate": 0.0005,
      "loss": 4.058571815490723,
      "step": 151
    },
    {
      "epoch": 0.03516076798519547,
      "grad_norm": 2.359344482421875,
      "learning_rate": 0.0005,
      "loss": 4.247644424438477,
      "step": 152
    },
    {
      "epoch": 0.03539208882720333,
      "grad_norm": 1.3130327463150024,
      "learning_rate": 0.0005,
      "loss": 4.110618591308594,
      "step": 153
    },
    {
      "epoch": 0.035623409669211195,
      "grad_norm": 1.8730707168579102,
      "learning_rate": 0.0005,
      "loss": 3.463820457458496,
      "step": 154
    },
    {
      "epoch": 0.03585473051121906,
      "grad_norm": 1.5096513032913208,
      "learning_rate": 0.0005,
      "loss": 3.671351432800293,
      "step": 155
    },
    {
      "epoch": 0.03608605135322693,
      "grad_norm": 1.3128407001495361,
      "learning_rate": 0.0005,
      "loss": 3.2561705112457275,
      "step": 156
    },
    {
      "epoch": 0.03631737219523479,
      "grad_norm": 1.644402027130127,
      "learning_rate": 0.0005,
      "loss": 3.641566276550293,
      "step": 157
    },
    {
      "epoch": 0.036548693037242654,
      "grad_norm": 1.8551756143569946,
      "learning_rate": 0.0005,
      "loss": 3.989405632019043,
      "step": 158
    },
    {
      "epoch": 0.03678001387925052,
      "grad_norm": 1.3110793828964233,
      "learning_rate": 0.0005,
      "loss": 3.5289292335510254,
      "step": 159
    },
    {
      "epoch": 0.037011334721258386,
      "grad_norm": 1.4108246564865112,
      "learning_rate": 0.0005,
      "loss": 3.9173481464385986,
      "step": 160
    },
    {
      "epoch": 0.03724265556326625,
      "grad_norm": 1.397367238998413,
      "learning_rate": 0.0005,
      "loss": 3.765411853790283,
      "step": 161
    },
    {
      "epoch": 0.03747397640527411,
      "grad_norm": 1.4032670259475708,
      "learning_rate": 0.0005,
      "loss": 3.8212904930114746,
      "step": 162
    },
    {
      "epoch": 0.03770529724728198,
      "grad_norm": 1.3268886804580688,
      "learning_rate": 0.0005,
      "loss": 3.6402697563171387,
      "step": 163
    },
    {
      "epoch": 0.037936618089289845,
      "grad_norm": 1.3526853322982788,
      "learning_rate": 0.0005,
      "loss": 3.949550151824951,
      "step": 164
    },
    {
      "epoch": 0.03816793893129771,
      "grad_norm": 1.2497812509536743,
      "learning_rate": 0.0005,
      "loss": 3.77891206741333,
      "step": 165
    },
    {
      "epoch": 0.03839925977330558,
      "grad_norm": 1.08015775680542,
      "learning_rate": 0.0005,
      "loss": 3.925262928009033,
      "step": 166
    },
    {
      "epoch": 0.03863058061531344,
      "grad_norm": 1.2234529256820679,
      "learning_rate": 0.0005,
      "loss": 3.651719570159912,
      "step": 167
    },
    {
      "epoch": 0.0388619014573213,
      "grad_norm": 1.3062536716461182,
      "learning_rate": 0.0005,
      "loss": 3.805600643157959,
      "step": 168
    },
    {
      "epoch": 0.03909322229932917,
      "grad_norm": 1.430570125579834,
      "learning_rate": 0.0005,
      "loss": 3.5887022018432617,
      "step": 169
    },
    {
      "epoch": 0.039324543141337036,
      "grad_norm": 1.4754607677459717,
      "learning_rate": 0.0005,
      "loss": 3.3482391834259033,
      "step": 170
    },
    {
      "epoch": 0.0395558639833449,
      "grad_norm": 1.6741350889205933,
      "learning_rate": 0.0005,
      "loss": 3.8094887733459473,
      "step": 171
    },
    {
      "epoch": 0.03978718482535276,
      "grad_norm": 1.6368192434310913,
      "learning_rate": 0.0005,
      "loss": 3.5793089866638184,
      "step": 172
    },
    {
      "epoch": 0.04001850566736063,
      "grad_norm": 1.3067339658737183,
      "learning_rate": 0.0005,
      "loss": 3.7006871700286865,
      "step": 173
    },
    {
      "epoch": 0.040249826509368494,
      "grad_norm": 1.5355005264282227,
      "learning_rate": 0.0005,
      "loss": 3.5927367210388184,
      "step": 174
    },
    {
      "epoch": 0.04048114735137636,
      "grad_norm": 1.2425918579101562,
      "learning_rate": 0.0005,
      "loss": 3.7014102935791016,
      "step": 175
    },
    {
      "epoch": 0.04071246819338423,
      "grad_norm": 1.2921298742294312,
      "learning_rate": 0.0005,
      "loss": 3.4801549911499023,
      "step": 176
    },
    {
      "epoch": 0.040943789035392086,
      "grad_norm": 1.7580575942993164,
      "learning_rate": 0.0005,
      "loss": 3.777881622314453,
      "step": 177
    },
    {
      "epoch": 0.04117510987739995,
      "grad_norm": 2.1360690593719482,
      "learning_rate": 0.0005,
      "loss": 3.8027079105377197,
      "step": 178
    },
    {
      "epoch": 0.04140643071940782,
      "grad_norm": 1.2091397047042847,
      "learning_rate": 0.0005,
      "loss": 3.5656673908233643,
      "step": 179
    },
    {
      "epoch": 0.041637751561415685,
      "grad_norm": 1.146545648574829,
      "learning_rate": 0.0005,
      "loss": 3.938314914703369,
      "step": 180
    },
    {
      "epoch": 0.04186907240342355,
      "grad_norm": 1.151294469833374,
      "learning_rate": 0.0005,
      "loss": 3.7764601707458496,
      "step": 181
    },
    {
      "epoch": 0.04210039324543141,
      "grad_norm": 1.3205286264419556,
      "learning_rate": 0.0005,
      "loss": 3.778503179550171,
      "step": 182
    },
    {
      "epoch": 0.04233171408743928,
      "grad_norm": 1.100386619567871,
      "learning_rate": 0.0005,
      "loss": 3.7893078327178955,
      "step": 183
    },
    {
      "epoch": 0.042563034929447144,
      "grad_norm": 1.3742119073867798,
      "learning_rate": 0.0005,
      "loss": 3.6861634254455566,
      "step": 184
    },
    {
      "epoch": 0.04279435577145501,
      "grad_norm": 1.182211995124817,
      "learning_rate": 0.0005,
      "loss": 3.7670087814331055,
      "step": 185
    },
    {
      "epoch": 0.04302567661346287,
      "grad_norm": 1.3156241178512573,
      "learning_rate": 0.0005,
      "loss": 3.430495262145996,
      "step": 186
    },
    {
      "epoch": 0.043256997455470736,
      "grad_norm": 1.6041662693023682,
      "learning_rate": 0.0005,
      "loss": 3.60441255569458,
      "step": 187
    },
    {
      "epoch": 0.0434883182974786,
      "grad_norm": 1.5299196243286133,
      "learning_rate": 0.0005,
      "loss": 4.2859039306640625,
      "step": 188
    },
    {
      "epoch": 0.04371963913948647,
      "grad_norm": 1.418385624885559,
      "learning_rate": 0.0005,
      "loss": 3.8036727905273438,
      "step": 189
    },
    {
      "epoch": 0.043950959981494335,
      "grad_norm": 1.5277786254882812,
      "learning_rate": 0.0005,
      "loss": 3.8732142448425293,
      "step": 190
    },
    {
      "epoch": 0.044182280823502194,
      "grad_norm": 1.3990788459777832,
      "learning_rate": 0.0005,
      "loss": 3.6172783374786377,
      "step": 191
    },
    {
      "epoch": 0.04441360166551006,
      "grad_norm": 1.2845922708511353,
      "learning_rate": 0.0005,
      "loss": 4.055271625518799,
      "step": 192
    },
    {
      "epoch": 0.04464492250751793,
      "grad_norm": 1.8684473037719727,
      "learning_rate": 0.0005,
      "loss": 3.874866008758545,
      "step": 193
    },
    {
      "epoch": 0.04487624334952579,
      "grad_norm": 1.4793695211410522,
      "learning_rate": 0.0005,
      "loss": 3.6185402870178223,
      "step": 194
    },
    {
      "epoch": 0.04510756419153366,
      "grad_norm": 1.509041428565979,
      "learning_rate": 0.0005,
      "loss": 3.7522335052490234,
      "step": 195
    },
    {
      "epoch": 0.04533888503354152,
      "grad_norm": 1.6722218990325928,
      "learning_rate": 0.0005,
      "loss": 3.6635031700134277,
      "step": 196
    },
    {
      "epoch": 0.045570205875549386,
      "grad_norm": 1.2109665870666504,
      "learning_rate": 0.0005,
      "loss": 3.5829620361328125,
      "step": 197
    },
    {
      "epoch": 0.04580152671755725,
      "grad_norm": 1.2587252855300903,
      "learning_rate": 0.0005,
      "loss": 3.781676769256592,
      "step": 198
    },
    {
      "epoch": 0.04603284755956512,
      "grad_norm": 1.2850024700164795,
      "learning_rate": 0.0005,
      "loss": 3.878889560699463,
      "step": 199
    },
    {
      "epoch": 0.046264168401572985,
      "grad_norm": 1.2542227506637573,
      "learning_rate": 0.0005,
      "loss": 3.734410047531128,
      "step": 200
    },
    {
      "epoch": 0.046495489243580844,
      "grad_norm": 1.2271263599395752,
      "learning_rate": 0.0005,
      "loss": 3.6624338626861572,
      "step": 201
    },
    {
      "epoch": 0.04672681008558871,
      "grad_norm": 1.4292821884155273,
      "learning_rate": 0.0005,
      "loss": 3.812816858291626,
      "step": 202
    },
    {
      "epoch": 0.04695813092759658,
      "grad_norm": 1.223658800125122,
      "learning_rate": 0.0005,
      "loss": 3.4529595375061035,
      "step": 203
    },
    {
      "epoch": 0.04718945176960444,
      "grad_norm": 1.5882270336151123,
      "learning_rate": 0.0005,
      "loss": 3.620861291885376,
      "step": 204
    },
    {
      "epoch": 0.04742077261161231,
      "grad_norm": 1.283645510673523,
      "learning_rate": 0.0005,
      "loss": 3.9295926094055176,
      "step": 205
    },
    {
      "epoch": 0.04765209345362017,
      "grad_norm": 1.4712001085281372,
      "learning_rate": 0.0005,
      "loss": 3.7703897953033447,
      "step": 206
    },
    {
      "epoch": 0.047883414295628035,
      "grad_norm": 1.3562506437301636,
      "learning_rate": 0.0005,
      "loss": 3.592973470687866,
      "step": 207
    },
    {
      "epoch": 0.0481147351376359,
      "grad_norm": 1.8586623668670654,
      "learning_rate": 0.0005,
      "loss": 4.189271926879883,
      "step": 208
    },
    {
      "epoch": 0.04834605597964377,
      "grad_norm": 1.200131893157959,
      "learning_rate": 0.0005,
      "loss": 3.945556163787842,
      "step": 209
    },
    {
      "epoch": 0.048577376821651634,
      "grad_norm": 1.3534471988677979,
      "learning_rate": 0.0005,
      "loss": 3.718233108520508,
      "step": 210
    },
    {
      "epoch": 0.048808697663659494,
      "grad_norm": 1.3967821598052979,
      "learning_rate": 0.0005,
      "loss": 4.0368757247924805,
      "step": 211
    },
    {
      "epoch": 0.04904001850566736,
      "grad_norm": 1.175358533859253,
      "learning_rate": 0.0005,
      "loss": 3.434136152267456,
      "step": 212
    },
    {
      "epoch": 0.049271339347675226,
      "grad_norm": 1.318171501159668,
      "learning_rate": 0.0005,
      "loss": 3.314333200454712,
      "step": 213
    },
    {
      "epoch": 0.04950266018968309,
      "grad_norm": 1.3276971578598022,
      "learning_rate": 0.0005,
      "loss": 3.2912888526916504,
      "step": 214
    },
    {
      "epoch": 0.04973398103169095,
      "grad_norm": 1.6523820161819458,
      "learning_rate": 0.0005,
      "loss": 3.6916208267211914,
      "step": 215
    },
    {
      "epoch": 0.04996530187369882,
      "grad_norm": 1.407034993171692,
      "learning_rate": 0.0005,
      "loss": 3.7394464015960693,
      "step": 216
    },
    {
      "epoch": 0.050196622715706685,
      "grad_norm": 1.742450475692749,
      "learning_rate": 0.0005,
      "loss": 3.626352310180664,
      "step": 217
    },
    {
      "epoch": 0.05042794355771455,
      "grad_norm": 1.86849045753479,
      "learning_rate": 0.0005,
      "loss": 3.605896234512329,
      "step": 218
    },
    {
      "epoch": 0.05065926439972242,
      "grad_norm": 1.4075862169265747,
      "learning_rate": 0.0005,
      "loss": 3.7643206119537354,
      "step": 219
    },
    {
      "epoch": 0.05089058524173028,
      "grad_norm": 1.12716543674469,
      "learning_rate": 0.0005,
      "loss": 3.9002785682678223,
      "step": 220
    },
    {
      "epoch": 0.05112190608373814,
      "grad_norm": 1.3027795553207397,
      "learning_rate": 0.0005,
      "loss": 3.6407394409179688,
      "step": 221
    },
    {
      "epoch": 0.05135322692574601,
      "grad_norm": 1.3612037897109985,
      "learning_rate": 0.0005,
      "loss": 3.598306179046631,
      "step": 222
    },
    {
      "epoch": 0.051584547767753876,
      "grad_norm": 1.4004743099212646,
      "learning_rate": 0.0005,
      "loss": 3.634418487548828,
      "step": 223
    },
    {
      "epoch": 0.05181586860976174,
      "grad_norm": 1.3013031482696533,
      "learning_rate": 0.0005,
      "loss": 3.746474266052246,
      "step": 224
    },
    {
      "epoch": 0.0520471894517696,
      "grad_norm": 1.535317063331604,
      "learning_rate": 0.0005,
      "loss": 3.7848336696624756,
      "step": 225
    },
    {
      "epoch": 0.05227851029377747,
      "grad_norm": 1.6366218328475952,
      "learning_rate": 0.0005,
      "loss": 3.59110689163208,
      "step": 226
    },
    {
      "epoch": 0.052509831135785334,
      "grad_norm": 1.3137116432189941,
      "learning_rate": 0.0005,
      "loss": 4.229291915893555,
      "step": 227
    },
    {
      "epoch": 0.0527411519777932,
      "grad_norm": 1.4145933389663696,
      "learning_rate": 0.0005,
      "loss": 3.814868927001953,
      "step": 228
    },
    {
      "epoch": 0.05297247281980107,
      "grad_norm": 1.5464670658111572,
      "learning_rate": 0.0005,
      "loss": 3.4960403442382812,
      "step": 229
    },
    {
      "epoch": 0.053203793661808926,
      "grad_norm": 1.347408652305603,
      "learning_rate": 0.0005,
      "loss": 4.087603569030762,
      "step": 230
    },
    {
      "epoch": 0.05343511450381679,
      "grad_norm": 1.3174077272415161,
      "learning_rate": 0.0005,
      "loss": 3.691042900085449,
      "step": 231
    },
    {
      "epoch": 0.05366643534582466,
      "grad_norm": 1.3471343517303467,
      "learning_rate": 0.0005,
      "loss": 3.6628832817077637,
      "step": 232
    },
    {
      "epoch": 0.053897756187832525,
      "grad_norm": 1.7511128187179565,
      "learning_rate": 0.0005,
      "loss": 3.6015267372131348,
      "step": 233
    },
    {
      "epoch": 0.05412907702984039,
      "grad_norm": 1.582022786140442,
      "learning_rate": 0.0005,
      "loss": 3.6502480506896973,
      "step": 234
    },
    {
      "epoch": 0.05436039787184825,
      "grad_norm": 1.306855320930481,
      "learning_rate": 0.0005,
      "loss": 3.534210205078125,
      "step": 235
    },
    {
      "epoch": 0.05459171871385612,
      "grad_norm": 1.2672127485275269,
      "learning_rate": 0.0005,
      "loss": 4.003190040588379,
      "step": 236
    },
    {
      "epoch": 0.054823039555863984,
      "grad_norm": 2.1194722652435303,
      "learning_rate": 0.0005,
      "loss": 3.8487000465393066,
      "step": 237
    },
    {
      "epoch": 0.05505436039787185,
      "grad_norm": 1.1987502574920654,
      "learning_rate": 0.0005,
      "loss": 4.016232013702393,
      "step": 238
    },
    {
      "epoch": 0.055285681239879716,
      "grad_norm": 1.3711155652999878,
      "learning_rate": 0.0005,
      "loss": 3.767045497894287,
      "step": 239
    },
    {
      "epoch": 0.055517002081887576,
      "grad_norm": 2.0263686180114746,
      "learning_rate": 0.0005,
      "loss": 3.991227388381958,
      "step": 240
    },
    {
      "epoch": 0.05574832292389544,
      "grad_norm": 1.3462415933609009,
      "learning_rate": 0.0005,
      "loss": 3.773150682449341,
      "step": 241
    },
    {
      "epoch": 0.05597964376590331,
      "grad_norm": 1.231645941734314,
      "learning_rate": 0.0005,
      "loss": 3.487959384918213,
      "step": 242
    },
    {
      "epoch": 0.056210964607911175,
      "grad_norm": 1.586879014968872,
      "learning_rate": 0.0005,
      "loss": 3.6634654998779297,
      "step": 243
    },
    {
      "epoch": 0.056442285449919034,
      "grad_norm": 1.3484405279159546,
      "learning_rate": 0.0005,
      "loss": 3.554858684539795,
      "step": 244
    },
    {
      "epoch": 0.0566736062919269,
      "grad_norm": 1.145593523979187,
      "learning_rate": 0.0005,
      "loss": 3.7756752967834473,
      "step": 245
    },
    {
      "epoch": 0.05690492713393477,
      "grad_norm": 1.3734862804412842,
      "learning_rate": 0.0005,
      "loss": 3.6252613067626953,
      "step": 246
    },
    {
      "epoch": 0.05713624797594263,
      "grad_norm": 1.1822127103805542,
      "learning_rate": 0.0005,
      "loss": 3.782198905944824,
      "step": 247
    },
    {
      "epoch": 0.0573675688179505,
      "grad_norm": 1.1662709712982178,
      "learning_rate": 0.0005,
      "loss": 3.705353260040283,
      "step": 248
    },
    {
      "epoch": 0.05759888965995836,
      "grad_norm": 1.306066632270813,
      "learning_rate": 0.0005,
      "loss": 3.4828691482543945,
      "step": 249
    },
    {
      "epoch": 0.057830210501966225,
      "grad_norm": 1.3461164236068726,
      "learning_rate": 0.0005,
      "loss": 3.3197922706604004,
      "step": 250
    },
    {
      "epoch": 0.05806153134397409,
      "grad_norm": 1.3188163042068481,
      "learning_rate": 0.0005,
      "loss": 3.861231565475464,
      "step": 251
    },
    {
      "epoch": 0.05829285218598196,
      "grad_norm": 1.5569504499435425,
      "learning_rate": 0.0005,
      "loss": 3.7349820137023926,
      "step": 252
    },
    {
      "epoch": 0.058524173027989825,
      "grad_norm": 1.2874164581298828,
      "learning_rate": 0.0005,
      "loss": 3.5939483642578125,
      "step": 253
    },
    {
      "epoch": 0.058755493869997684,
      "grad_norm": 1.4148621559143066,
      "learning_rate": 0.0005,
      "loss": 3.4744114875793457,
      "step": 254
    },
    {
      "epoch": 0.05898681471200555,
      "grad_norm": 1.7827779054641724,
      "learning_rate": 0.0005,
      "loss": 4.120580673217773,
      "step": 255
    },
    {
      "epoch": 0.05921813555401342,
      "grad_norm": 1.4928944110870361,
      "learning_rate": 0.0005,
      "loss": 3.6259243488311768,
      "step": 256
    },
    {
      "epoch": 0.05944945639602128,
      "grad_norm": 1.3590694665908813,
      "learning_rate": 0.0005,
      "loss": 3.558243751525879,
      "step": 257
    },
    {
      "epoch": 0.05968077723802915,
      "grad_norm": 1.300769567489624,
      "learning_rate": 0.0005,
      "loss": 4.025277614593506,
      "step": 258
    },
    {
      "epoch": 0.05991209808003701,
      "grad_norm": 1.321162462234497,
      "learning_rate": 0.0005,
      "loss": 3.370614767074585,
      "step": 259
    },
    {
      "epoch": 0.060143418922044875,
      "grad_norm": 1.897878885269165,
      "learning_rate": 0.0005,
      "loss": 3.610748767852783,
      "step": 260
    },
    {
      "epoch": 0.06037473976405274,
      "grad_norm": 1.7357853651046753,
      "learning_rate": 0.0005,
      "loss": 3.667632818222046,
      "step": 261
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 1.3805984258651733,
      "learning_rate": 0.0005,
      "loss": 3.6551332473754883,
      "step": 262
    },
    {
      "epoch": 0.060837381448068474,
      "grad_norm": 1.225106954574585,
      "learning_rate": 0.0005,
      "loss": 4.044240951538086,
      "step": 263
    },
    {
      "epoch": 0.061068702290076333,
      "grad_norm": 1.3096638917922974,
      "learning_rate": 0.0005,
      "loss": 3.3601255416870117,
      "step": 264
    },
    {
      "epoch": 0.0613000231320842,
      "grad_norm": 1.127160906791687,
      "learning_rate": 0.0005,
      "loss": 3.4653334617614746,
      "step": 265
    },
    {
      "epoch": 0.061531343974092066,
      "grad_norm": 1.1961965560913086,
      "learning_rate": 0.0005,
      "loss": 3.7911367416381836,
      "step": 266
    },
    {
      "epoch": 0.06176266481609993,
      "grad_norm": 2.6379902362823486,
      "learning_rate": 0.0005,
      "loss": 3.4752910137176514,
      "step": 267
    },
    {
      "epoch": 0.0619939856581078,
      "grad_norm": 1.608422040939331,
      "learning_rate": 0.0005,
      "loss": 3.406269073486328,
      "step": 268
    },
    {
      "epoch": 0.06222530650011566,
      "grad_norm": 1.4822862148284912,
      "learning_rate": 0.0005,
      "loss": 3.9536194801330566,
      "step": 269
    },
    {
      "epoch": 0.062456627342123525,
      "grad_norm": 1.455018162727356,
      "learning_rate": 0.0005,
      "loss": 3.961531162261963,
      "step": 270
    },
    {
      "epoch": 0.06268794818413138,
      "grad_norm": 1.2014250755310059,
      "learning_rate": 0.0005,
      "loss": 3.7226462364196777,
      "step": 271
    },
    {
      "epoch": 0.06291926902613926,
      "grad_norm": 1.5859811305999756,
      "learning_rate": 0.0005,
      "loss": 3.9274017810821533,
      "step": 272
    },
    {
      "epoch": 0.06315058986814712,
      "grad_norm": 1.5234670639038086,
      "learning_rate": 0.0005,
      "loss": 3.628068447113037,
      "step": 273
    },
    {
      "epoch": 0.06338191071015499,
      "grad_norm": 1.2802175283432007,
      "learning_rate": 0.0005,
      "loss": 3.6193861961364746,
      "step": 274
    },
    {
      "epoch": 0.06361323155216285,
      "grad_norm": 1.406999111175537,
      "learning_rate": 0.0005,
      "loss": 3.3880386352539062,
      "step": 275
    },
    {
      "epoch": 0.06384455239417071,
      "grad_norm": 1.3304131031036377,
      "learning_rate": 0.0005,
      "loss": 3.593752861022949,
      "step": 276
    },
    {
      "epoch": 0.06407587323617858,
      "grad_norm": 1.401597023010254,
      "learning_rate": 0.0005,
      "loss": 3.732774496078491,
      "step": 277
    },
    {
      "epoch": 0.06430719407818644,
      "grad_norm": 5.8366289138793945,
      "learning_rate": 0.0005,
      "loss": 3.553156852722168,
      "step": 278
    },
    {
      "epoch": 0.06453851492019431,
      "grad_norm": 1.240220069885254,
      "learning_rate": 0.0005,
      "loss": 3.694706678390503,
      "step": 279
    },
    {
      "epoch": 0.06476983576220217,
      "grad_norm": 1.325284481048584,
      "learning_rate": 0.0005,
      "loss": 3.8227486610412598,
      "step": 280
    },
    {
      "epoch": 0.06500115660421003,
      "grad_norm": 2.4205431938171387,
      "learning_rate": 0.0005,
      "loss": 4.103816986083984,
      "step": 281
    },
    {
      "epoch": 0.0652324774462179,
      "grad_norm": 1.6118733882904053,
      "learning_rate": 0.0005,
      "loss": 3.7322304248809814,
      "step": 282
    },
    {
      "epoch": 0.06546379828822577,
      "grad_norm": 1.5773471593856812,
      "learning_rate": 0.0005,
      "loss": 3.3082172870635986,
      "step": 283
    },
    {
      "epoch": 0.06569511913023364,
      "grad_norm": 1.2600442171096802,
      "learning_rate": 0.0005,
      "loss": 3.8940696716308594,
      "step": 284
    },
    {
      "epoch": 0.0659264399722415,
      "grad_norm": 1.169819951057434,
      "learning_rate": 0.0005,
      "loss": 3.909982681274414,
      "step": 285
    },
    {
      "epoch": 0.06615776081424936,
      "grad_norm": 1.377044439315796,
      "learning_rate": 0.0005,
      "loss": 3.677433967590332,
      "step": 286
    },
    {
      "epoch": 0.06638908165625723,
      "grad_norm": 1.288591980934143,
      "learning_rate": 0.0005,
      "loss": 3.5907235145568848,
      "step": 287
    },
    {
      "epoch": 0.06662040249826509,
      "grad_norm": 1.5596824884414673,
      "learning_rate": 0.0005,
      "loss": 4.135848045349121,
      "step": 288
    },
    {
      "epoch": 0.06685172334027296,
      "grad_norm": 1.3136385679244995,
      "learning_rate": 0.0005,
      "loss": 3.7453455924987793,
      "step": 289
    },
    {
      "epoch": 0.06708304418228082,
      "grad_norm": 1.239112377166748,
      "learning_rate": 0.0005,
      "loss": 3.7001702785491943,
      "step": 290
    },
    {
      "epoch": 0.06731436502428868,
      "grad_norm": 1.2838140726089478,
      "learning_rate": 0.0005,
      "loss": 3.844417095184326,
      "step": 291
    },
    {
      "epoch": 0.06754568586629656,
      "grad_norm": 1.4993733167648315,
      "learning_rate": 0.0005,
      "loss": 3.7507834434509277,
      "step": 292
    },
    {
      "epoch": 0.06777700670830442,
      "grad_norm": 1.4118742942810059,
      "learning_rate": 0.0005,
      "loss": 3.639863967895508,
      "step": 293
    },
    {
      "epoch": 0.06800832755031229,
      "grad_norm": 1.3667352199554443,
      "learning_rate": 0.0005,
      "loss": 3.9316112995147705,
      "step": 294
    },
    {
      "epoch": 0.06823964839232015,
      "grad_norm": 1.3142811059951782,
      "learning_rate": 0.0005,
      "loss": 3.6550493240356445,
      "step": 295
    },
    {
      "epoch": 0.06847096923432801,
      "grad_norm": 1.1062250137329102,
      "learning_rate": 0.0005,
      "loss": 4.018372535705566,
      "step": 296
    },
    {
      "epoch": 0.06870229007633588,
      "grad_norm": 1.2298468351364136,
      "learning_rate": 0.0005,
      "loss": 3.6669979095458984,
      "step": 297
    },
    {
      "epoch": 0.06893361091834374,
      "grad_norm": 1.3076558113098145,
      "learning_rate": 0.0005,
      "loss": 3.7758285999298096,
      "step": 298
    },
    {
      "epoch": 0.06916493176035161,
      "grad_norm": 1.2150471210479736,
      "learning_rate": 0.0005,
      "loss": 3.732393264770508,
      "step": 299
    },
    {
      "epoch": 0.06939625260235947,
      "grad_norm": 1.2322394847869873,
      "learning_rate": 0.0005,
      "loss": 3.638270378112793,
      "step": 300
    },
    {
      "epoch": 0.06962757344436733,
      "grad_norm": 1.2531236410140991,
      "learning_rate": 0.0005,
      "loss": 3.497500419616699,
      "step": 301
    },
    {
      "epoch": 0.0698588942863752,
      "grad_norm": 1.415730595588684,
      "learning_rate": 0.0005,
      "loss": 3.737941265106201,
      "step": 302
    },
    {
      "epoch": 0.07009021512838307,
      "grad_norm": 1.03556489944458,
      "learning_rate": 0.0005,
      "loss": 3.6462595462799072,
      "step": 303
    },
    {
      "epoch": 0.07032153597039094,
      "grad_norm": 1.4187175035476685,
      "learning_rate": 0.0005,
      "loss": 4.006418228149414,
      "step": 304
    },
    {
      "epoch": 0.0705528568123988,
      "grad_norm": 1.080061435699463,
      "learning_rate": 0.0005,
      "loss": 3.7075395584106445,
      "step": 305
    },
    {
      "epoch": 0.07078417765440666,
      "grad_norm": 1.2121397256851196,
      "learning_rate": 0.0005,
      "loss": 3.6599998474121094,
      "step": 306
    },
    {
      "epoch": 0.07101549849641453,
      "grad_norm": 1.6774672269821167,
      "learning_rate": 0.0005,
      "loss": 3.661608934402466,
      "step": 307
    },
    {
      "epoch": 0.07124681933842239,
      "grad_norm": 1.6162365674972534,
      "learning_rate": 0.0005,
      "loss": 3.625678777694702,
      "step": 308
    },
    {
      "epoch": 0.07147814018043026,
      "grad_norm": 1.3926607370376587,
      "learning_rate": 0.0005,
      "loss": 4.11833381652832,
      "step": 309
    },
    {
      "epoch": 0.07170946102243812,
      "grad_norm": 1.2871214151382446,
      "learning_rate": 0.0005,
      "loss": 3.4463534355163574,
      "step": 310
    },
    {
      "epoch": 0.07194078186444598,
      "grad_norm": 1.6442217826843262,
      "learning_rate": 0.0005,
      "loss": 3.5558080673217773,
      "step": 311
    },
    {
      "epoch": 0.07217210270645386,
      "grad_norm": 1.3817247152328491,
      "learning_rate": 0.0005,
      "loss": 3.9788758754730225,
      "step": 312
    },
    {
      "epoch": 0.07240342354846172,
      "grad_norm": 1.0941458940505981,
      "learning_rate": 0.0005,
      "loss": 3.78084397315979,
      "step": 313
    },
    {
      "epoch": 0.07263474439046957,
      "grad_norm": 1.6858614683151245,
      "learning_rate": 0.0005,
      "loss": 3.5600128173828125,
      "step": 314
    },
    {
      "epoch": 0.07286606523247745,
      "grad_norm": 1.27935791015625,
      "learning_rate": 0.0005,
      "loss": 3.596074104309082,
      "step": 315
    },
    {
      "epoch": 0.07309738607448531,
      "grad_norm": 1.4701615571975708,
      "learning_rate": 0.0005,
      "loss": 3.7470602989196777,
      "step": 316
    },
    {
      "epoch": 0.07332870691649318,
      "grad_norm": 1.458840250968933,
      "learning_rate": 0.0005,
      "loss": 3.6959388256073,
      "step": 317
    },
    {
      "epoch": 0.07356002775850104,
      "grad_norm": 1.5201733112335205,
      "learning_rate": 0.0005,
      "loss": 3.6289501190185547,
      "step": 318
    },
    {
      "epoch": 0.0737913486005089,
      "grad_norm": 1.063296914100647,
      "learning_rate": 0.0005,
      "loss": 3.425262451171875,
      "step": 319
    },
    {
      "epoch": 0.07402266944251677,
      "grad_norm": 1.3575137853622437,
      "learning_rate": 0.0005,
      "loss": 3.8760452270507812,
      "step": 320
    },
    {
      "epoch": 0.07425399028452463,
      "grad_norm": 1.3390324115753174,
      "learning_rate": 0.0005,
      "loss": 3.6817963123321533,
      "step": 321
    },
    {
      "epoch": 0.0744853111265325,
      "grad_norm": 1.2759087085723877,
      "learning_rate": 0.0005,
      "loss": 3.8902039527893066,
      "step": 322
    },
    {
      "epoch": 0.07471663196854036,
      "grad_norm": 1.3398079872131348,
      "learning_rate": 0.0005,
      "loss": 3.698521137237549,
      "step": 323
    },
    {
      "epoch": 0.07494795281054822,
      "grad_norm": 1.2229738235473633,
      "learning_rate": 0.0005,
      "loss": 3.9286999702453613,
      "step": 324
    },
    {
      "epoch": 0.0751792736525561,
      "grad_norm": 1.4204074144363403,
      "learning_rate": 0.0005,
      "loss": 3.518857955932617,
      "step": 325
    },
    {
      "epoch": 0.07541059449456396,
      "grad_norm": 1.2707080841064453,
      "learning_rate": 0.0005,
      "loss": 3.644885540008545,
      "step": 326
    },
    {
      "epoch": 0.07564191533657183,
      "grad_norm": 1.8028753995895386,
      "learning_rate": 0.0005,
      "loss": 3.7294836044311523,
      "step": 327
    },
    {
      "epoch": 0.07587323617857969,
      "grad_norm": 1.8877991437911987,
      "learning_rate": 0.0005,
      "loss": 3.6860268115997314,
      "step": 328
    },
    {
      "epoch": 0.07610455702058755,
      "grad_norm": 1.758873462677002,
      "learning_rate": 0.0005,
      "loss": 3.5128777027130127,
      "step": 329
    },
    {
      "epoch": 0.07633587786259542,
      "grad_norm": 1.3587888479232788,
      "learning_rate": 0.0005,
      "loss": 3.8641304969787598,
      "step": 330
    },
    {
      "epoch": 0.07656719870460328,
      "grad_norm": 1.3440887928009033,
      "learning_rate": 0.0005,
      "loss": 3.725612163543701,
      "step": 331
    },
    {
      "epoch": 0.07679851954661115,
      "grad_norm": 1.5936044454574585,
      "learning_rate": 0.0005,
      "loss": 3.3491058349609375,
      "step": 332
    },
    {
      "epoch": 0.07702984038861901,
      "grad_norm": 1.3777759075164795,
      "learning_rate": 0.0005,
      "loss": 3.568859338760376,
      "step": 333
    },
    {
      "epoch": 0.07726116123062687,
      "grad_norm": 1.4465068578720093,
      "learning_rate": 0.0005,
      "loss": 3.4838080406188965,
      "step": 334
    },
    {
      "epoch": 0.07749248207263475,
      "grad_norm": 1.6642496585845947,
      "learning_rate": 0.0005,
      "loss": 4.3937177658081055,
      "step": 335
    },
    {
      "epoch": 0.0777238029146426,
      "grad_norm": 1.701362133026123,
      "learning_rate": 0.0005,
      "loss": 3.672417163848877,
      "step": 336
    },
    {
      "epoch": 0.07795512375665048,
      "grad_norm": 1.7232991456985474,
      "learning_rate": 0.0005,
      "loss": 3.9125025272369385,
      "step": 337
    },
    {
      "epoch": 0.07818644459865834,
      "grad_norm": 1.9594602584838867,
      "learning_rate": 0.0005,
      "loss": 3.2357935905456543,
      "step": 338
    },
    {
      "epoch": 0.0784177654406662,
      "grad_norm": 1.565032958984375,
      "learning_rate": 0.0005,
      "loss": 3.8256542682647705,
      "step": 339
    },
    {
      "epoch": 0.07864908628267407,
      "grad_norm": 1.4264545440673828,
      "learning_rate": 0.0005,
      "loss": 3.282503604888916,
      "step": 340
    },
    {
      "epoch": 0.07888040712468193,
      "grad_norm": 1.3649643659591675,
      "learning_rate": 0.0005,
      "loss": 3.678941249847412,
      "step": 341
    },
    {
      "epoch": 0.0791117279666898,
      "grad_norm": 1.3363479375839233,
      "learning_rate": 0.0005,
      "loss": 3.4259791374206543,
      "step": 342
    },
    {
      "epoch": 0.07934304880869766,
      "grad_norm": 1.1161242723464966,
      "learning_rate": 0.0005,
      "loss": 3.600525379180908,
      "step": 343
    },
    {
      "epoch": 0.07957436965070552,
      "grad_norm": 1.2906972169876099,
      "learning_rate": 0.0005,
      "loss": 3.8397669792175293,
      "step": 344
    },
    {
      "epoch": 0.0798056904927134,
      "grad_norm": 1.378008484840393,
      "learning_rate": 0.0005,
      "loss": 3.5513479709625244,
      "step": 345
    },
    {
      "epoch": 0.08003701133472126,
      "grad_norm": 1.2324062585830688,
      "learning_rate": 0.0005,
      "loss": 3.682922840118408,
      "step": 346
    },
    {
      "epoch": 0.08026833217672913,
      "grad_norm": 1.4492748975753784,
      "learning_rate": 0.0005,
      "loss": 3.6088433265686035,
      "step": 347
    },
    {
      "epoch": 0.08049965301873699,
      "grad_norm": 1.2541898488998413,
      "learning_rate": 0.0005,
      "loss": 3.490903377532959,
      "step": 348
    },
    {
      "epoch": 0.08073097386074485,
      "grad_norm": 1.3637157678604126,
      "learning_rate": 0.0005,
      "loss": 3.9015252590179443,
      "step": 349
    },
    {
      "epoch": 0.08096229470275272,
      "grad_norm": 1.3650386333465576,
      "learning_rate": 0.0005,
      "loss": 3.3538854122161865,
      "step": 350
    },
    {
      "epoch": 0.08119361554476058,
      "grad_norm": 1.327065348625183,
      "learning_rate": 0.0005,
      "loss": 3.674595832824707,
      "step": 351
    },
    {
      "epoch": 0.08142493638676845,
      "grad_norm": 1.1148836612701416,
      "learning_rate": 0.0005,
      "loss": 3.5856504440307617,
      "step": 352
    },
    {
      "epoch": 0.08165625722877631,
      "grad_norm": 1.09675133228302,
      "learning_rate": 0.0005,
      "loss": 3.740705966949463,
      "step": 353
    },
    {
      "epoch": 0.08188757807078417,
      "grad_norm": 1.8022757768630981,
      "learning_rate": 0.0005,
      "loss": 3.623914957046509,
      "step": 354
    },
    {
      "epoch": 0.08211889891279205,
      "grad_norm": 1.3861088752746582,
      "learning_rate": 0.0005,
      "loss": 3.806999921798706,
      "step": 355
    },
    {
      "epoch": 0.0823502197547999,
      "grad_norm": 1.0911870002746582,
      "learning_rate": 0.0005,
      "loss": 3.552551746368408,
      "step": 356
    },
    {
      "epoch": 0.08258154059680778,
      "grad_norm": 2.670001268386841,
      "learning_rate": 0.0005,
      "loss": 3.8604211807250977,
      "step": 357
    },
    {
      "epoch": 0.08281286143881564,
      "grad_norm": 1.3252840042114258,
      "learning_rate": 0.0005,
      "loss": 3.5708513259887695,
      "step": 358
    },
    {
      "epoch": 0.0830441822808235,
      "grad_norm": 1.5147367715835571,
      "learning_rate": 0.0005,
      "loss": 3.4792428016662598,
      "step": 359
    },
    {
      "epoch": 0.08327550312283137,
      "grad_norm": 1.2716940641403198,
      "learning_rate": 0.0005,
      "loss": 3.423947334289551,
      "step": 360
    },
    {
      "epoch": 0.08350682396483923,
      "grad_norm": 1.661228060722351,
      "learning_rate": 0.0005,
      "loss": 3.735062599182129,
      "step": 361
    },
    {
      "epoch": 0.0837381448068471,
      "grad_norm": 1.2847388982772827,
      "learning_rate": 0.0005,
      "loss": 3.7856855392456055,
      "step": 362
    },
    {
      "epoch": 0.08396946564885496,
      "grad_norm": 1.163465976715088,
      "learning_rate": 0.0005,
      "loss": 3.694396734237671,
      "step": 363
    },
    {
      "epoch": 0.08420078649086282,
      "grad_norm": 1.2065589427947998,
      "learning_rate": 0.0005,
      "loss": 3.6616930961608887,
      "step": 364
    },
    {
      "epoch": 0.0844321073328707,
      "grad_norm": 1.5104038715362549,
      "learning_rate": 0.0005,
      "loss": 3.42545485496521,
      "step": 365
    },
    {
      "epoch": 0.08466342817487855,
      "grad_norm": 1.2683137655258179,
      "learning_rate": 0.0005,
      "loss": 3.586872100830078,
      "step": 366
    },
    {
      "epoch": 0.08489474901688643,
      "grad_norm": 1.3757925033569336,
      "learning_rate": 0.0005,
      "loss": 3.791553497314453,
      "step": 367
    },
    {
      "epoch": 0.08512606985889429,
      "grad_norm": 1.2433658838272095,
      "learning_rate": 0.0005,
      "loss": 3.7342042922973633,
      "step": 368
    },
    {
      "epoch": 0.08535739070090215,
      "grad_norm": 1.3353224992752075,
      "learning_rate": 0.0005,
      "loss": 3.6095874309539795,
      "step": 369
    },
    {
      "epoch": 0.08558871154291002,
      "grad_norm": 1.2069038152694702,
      "learning_rate": 0.0005,
      "loss": 3.9225027561187744,
      "step": 370
    },
    {
      "epoch": 0.08582003238491788,
      "grad_norm": 1.5850926637649536,
      "learning_rate": 0.0005,
      "loss": 3.5118484497070312,
      "step": 371
    },
    {
      "epoch": 0.08605135322692574,
      "grad_norm": 1.4161951541900635,
      "learning_rate": 0.0005,
      "loss": 3.9600794315338135,
      "step": 372
    },
    {
      "epoch": 0.08628267406893361,
      "grad_norm": 1.2349975109100342,
      "learning_rate": 0.0005,
      "loss": 3.684633255004883,
      "step": 373
    },
    {
      "epoch": 0.08651399491094147,
      "grad_norm": 1.2922697067260742,
      "learning_rate": 0.0005,
      "loss": 3.5526137351989746,
      "step": 374
    },
    {
      "epoch": 0.08674531575294935,
      "grad_norm": 1.6918834447860718,
      "learning_rate": 0.0005,
      "loss": 3.2768774032592773,
      "step": 375
    },
    {
      "epoch": 0.0869766365949572,
      "grad_norm": 1.4540528059005737,
      "learning_rate": 0.0005,
      "loss": 3.176664352416992,
      "step": 376
    },
    {
      "epoch": 0.08720795743696506,
      "grad_norm": 1.1990540027618408,
      "learning_rate": 0.0005,
      "loss": 3.3922128677368164,
      "step": 377
    },
    {
      "epoch": 0.08743927827897294,
      "grad_norm": 1.672174334526062,
      "learning_rate": 0.0005,
      "loss": 3.9305853843688965,
      "step": 378
    },
    {
      "epoch": 0.0876705991209808,
      "grad_norm": 1.475325345993042,
      "learning_rate": 0.0005,
      "loss": 3.5140318870544434,
      "step": 379
    },
    {
      "epoch": 0.08790191996298867,
      "grad_norm": 1.4667911529541016,
      "learning_rate": 0.0005,
      "loss": 3.541513442993164,
      "step": 380
    },
    {
      "epoch": 0.08813324080499653,
      "grad_norm": 1.2083953619003296,
      "learning_rate": 0.0005,
      "loss": 3.756530284881592,
      "step": 381
    },
    {
      "epoch": 0.08836456164700439,
      "grad_norm": 1.122684359550476,
      "learning_rate": 0.0005,
      "loss": 3.6631205081939697,
      "step": 382
    },
    {
      "epoch": 0.08859588248901226,
      "grad_norm": 1.7709808349609375,
      "learning_rate": 0.0005,
      "loss": 3.9572229385375977,
      "step": 383
    },
    {
      "epoch": 0.08882720333102012,
      "grad_norm": 1.5337109565734863,
      "learning_rate": 0.0005,
      "loss": 3.620340347290039,
      "step": 384
    },
    {
      "epoch": 0.089058524173028,
      "grad_norm": 1.5332690477371216,
      "learning_rate": 0.0005,
      "loss": 3.57574462890625,
      "step": 385
    },
    {
      "epoch": 0.08928984501503585,
      "grad_norm": 1.1964465379714966,
      "learning_rate": 0.0005,
      "loss": 3.399348258972168,
      "step": 386
    },
    {
      "epoch": 0.08952116585704371,
      "grad_norm": 1.4862724542617798,
      "learning_rate": 0.0005,
      "loss": 3.5958549976348877,
      "step": 387
    },
    {
      "epoch": 0.08975248669905159,
      "grad_norm": 1.5196958780288696,
      "learning_rate": 0.0005,
      "loss": 3.306023120880127,
      "step": 388
    },
    {
      "epoch": 0.08998380754105945,
      "grad_norm": 1.405994176864624,
      "learning_rate": 0.0005,
      "loss": 3.639401435852051,
      "step": 389
    },
    {
      "epoch": 0.09021512838306732,
      "grad_norm": 1.4028009176254272,
      "learning_rate": 0.0005,
      "loss": 3.5652315616607666,
      "step": 390
    },
    {
      "epoch": 0.09044644922507518,
      "grad_norm": 1.2307472229003906,
      "learning_rate": 0.0005,
      "loss": 3.6880998611450195,
      "step": 391
    },
    {
      "epoch": 0.09067777006708304,
      "grad_norm": 1.230436086654663,
      "learning_rate": 0.0005,
      "loss": 3.8237617015838623,
      "step": 392
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 1.8454370498657227,
      "learning_rate": 0.0005,
      "loss": 3.6699562072753906,
      "step": 393
    },
    {
      "epoch": 0.09114041175109877,
      "grad_norm": 1.201208472251892,
      "learning_rate": 0.0005,
      "loss": 3.4258196353912354,
      "step": 394
    },
    {
      "epoch": 0.09137173259310664,
      "grad_norm": 1.190623164176941,
      "learning_rate": 0.0005,
      "loss": 3.4519739151000977,
      "step": 395
    },
    {
      "epoch": 0.0916030534351145,
      "grad_norm": 1.1560215950012207,
      "learning_rate": 0.0005,
      "loss": 3.6505889892578125,
      "step": 396
    },
    {
      "epoch": 0.09183437427712236,
      "grad_norm": 1.1827839612960815,
      "learning_rate": 0.0005,
      "loss": 3.7127342224121094,
      "step": 397
    },
    {
      "epoch": 0.09206569511913024,
      "grad_norm": 1.3352675437927246,
      "learning_rate": 0.0005,
      "loss": 3.7012996673583984,
      "step": 398
    },
    {
      "epoch": 0.0922970159611381,
      "grad_norm": 1.1233023405075073,
      "learning_rate": 0.0005,
      "loss": 3.6034297943115234,
      "step": 399
    },
    {
      "epoch": 0.09252833680314597,
      "grad_norm": 1.2975887060165405,
      "learning_rate": 0.0005,
      "loss": 3.676487684249878,
      "step": 400
    },
    {
      "epoch": 0.09275965764515383,
      "grad_norm": 1.4117364883422852,
      "learning_rate": 0.0005,
      "loss": 3.7573587894439697,
      "step": 401
    },
    {
      "epoch": 0.09299097848716169,
      "grad_norm": 1.2857729196548462,
      "learning_rate": 0.0005,
      "loss": 3.8555545806884766,
      "step": 402
    },
    {
      "epoch": 0.09322229932916956,
      "grad_norm": 1.2022277116775513,
      "learning_rate": 0.0005,
      "loss": 3.3433384895324707,
      "step": 403
    },
    {
      "epoch": 0.09345362017117742,
      "grad_norm": 1.369255781173706,
      "learning_rate": 0.0005,
      "loss": 3.5251519680023193,
      "step": 404
    },
    {
      "epoch": 0.0936849410131853,
      "grad_norm": 1.5980631113052368,
      "learning_rate": 0.0005,
      "loss": 3.507040023803711,
      "step": 405
    },
    {
      "epoch": 0.09391626185519315,
      "grad_norm": 1.3396875858306885,
      "learning_rate": 0.0005,
      "loss": 3.4226415157318115,
      "step": 406
    },
    {
      "epoch": 0.09414758269720101,
      "grad_norm": 1.6930949687957764,
      "learning_rate": 0.0005,
      "loss": 3.702646255493164,
      "step": 407
    },
    {
      "epoch": 0.09437890353920889,
      "grad_norm": 1.3387141227722168,
      "learning_rate": 0.0005,
      "loss": 3.5502471923828125,
      "step": 408
    },
    {
      "epoch": 0.09461022438121675,
      "grad_norm": 1.43996262550354,
      "learning_rate": 0.0005,
      "loss": 3.7054295539855957,
      "step": 409
    },
    {
      "epoch": 0.09484154522322462,
      "grad_norm": 1.2177127599716187,
      "learning_rate": 0.0005,
      "loss": 4.0196638107299805,
      "step": 410
    },
    {
      "epoch": 0.09507286606523248,
      "grad_norm": 1.3554432392120361,
      "learning_rate": 0.0005,
      "loss": 3.187286138534546,
      "step": 411
    },
    {
      "epoch": 0.09530418690724034,
      "grad_norm": 1.3636268377304077,
      "learning_rate": 0.0005,
      "loss": 3.3692305088043213,
      "step": 412
    },
    {
      "epoch": 0.09553550774924821,
      "grad_norm": 1.3856538534164429,
      "learning_rate": 0.0005,
      "loss": 3.3982558250427246,
      "step": 413
    },
    {
      "epoch": 0.09576682859125607,
      "grad_norm": 1.3439018726348877,
      "learning_rate": 0.0005,
      "loss": 3.5916504859924316,
      "step": 414
    },
    {
      "epoch": 0.09599814943326394,
      "grad_norm": 1.8873322010040283,
      "learning_rate": 0.0005,
      "loss": 3.2825496196746826,
      "step": 415
    },
    {
      "epoch": 0.0962294702752718,
      "grad_norm": 1.1466286182403564,
      "learning_rate": 0.0005,
      "loss": 3.5965757369995117,
      "step": 416
    },
    {
      "epoch": 0.09646079111727966,
      "grad_norm": 1.3826420307159424,
      "learning_rate": 0.0005,
      "loss": 3.63053822517395,
      "step": 417
    },
    {
      "epoch": 0.09669211195928754,
      "grad_norm": 1.1548633575439453,
      "learning_rate": 0.0005,
      "loss": 3.6712985038757324,
      "step": 418
    },
    {
      "epoch": 0.0969234328012954,
      "grad_norm": 1.3228187561035156,
      "learning_rate": 0.0005,
      "loss": 3.5552072525024414,
      "step": 419
    },
    {
      "epoch": 0.09715475364330327,
      "grad_norm": 1.1442086696624756,
      "learning_rate": 0.0005,
      "loss": 3.7503530979156494,
      "step": 420
    },
    {
      "epoch": 0.09738607448531113,
      "grad_norm": 1.8267778158187866,
      "learning_rate": 0.0005,
      "loss": 3.059821128845215,
      "step": 421
    },
    {
      "epoch": 0.09761739532731899,
      "grad_norm": 1.4516521692276,
      "learning_rate": 0.0005,
      "loss": 3.4778149127960205,
      "step": 422
    },
    {
      "epoch": 0.09784871616932686,
      "grad_norm": 1.418588399887085,
      "learning_rate": 0.0005,
      "loss": 3.434175729751587,
      "step": 423
    },
    {
      "epoch": 0.09808003701133472,
      "grad_norm": 1.6750156879425049,
      "learning_rate": 0.0005,
      "loss": 3.6111245155334473,
      "step": 424
    },
    {
      "epoch": 0.09831135785334259,
      "grad_norm": 1.582371711730957,
      "learning_rate": 0.0005,
      "loss": 3.6139276027679443,
      "step": 425
    },
    {
      "epoch": 0.09854267869535045,
      "grad_norm": 1.2942166328430176,
      "learning_rate": 0.0005,
      "loss": 3.4191319942474365,
      "step": 426
    },
    {
      "epoch": 0.09877399953735831,
      "grad_norm": 1.4988442659378052,
      "learning_rate": 0.0005,
      "loss": 3.0288069248199463,
      "step": 427
    },
    {
      "epoch": 0.09900532037936619,
      "grad_norm": 1.3436778783798218,
      "learning_rate": 0.0005,
      "loss": 3.445136547088623,
      "step": 428
    },
    {
      "epoch": 0.09923664122137404,
      "grad_norm": 1.2061022520065308,
      "learning_rate": 0.0005,
      "loss": 3.672032117843628,
      "step": 429
    },
    {
      "epoch": 0.0994679620633819,
      "grad_norm": 2.18245267868042,
      "learning_rate": 0.0005,
      "loss": 3.7760727405548096,
      "step": 430
    },
    {
      "epoch": 0.09969928290538978,
      "grad_norm": 1.532149314880371,
      "learning_rate": 0.0005,
      "loss": 3.5696229934692383,
      "step": 431
    },
    {
      "epoch": 0.09993060374739764,
      "grad_norm": 1.543492317199707,
      "learning_rate": 0.0005,
      "loss": 3.83501935005188,
      "step": 432
    },
    {
      "epoch": 0.10016192458940551,
      "grad_norm": 1.3865993022918701,
      "learning_rate": 0.0005,
      "loss": 3.6268763542175293,
      "step": 433
    },
    {
      "epoch": 0.10039324543141337,
      "grad_norm": 1.528840184211731,
      "learning_rate": 0.0005,
      "loss": 3.4783387184143066,
      "step": 434
    },
    {
      "epoch": 0.10062456627342123,
      "grad_norm": 1.4440724849700928,
      "learning_rate": 0.0005,
      "loss": 3.5219833850860596,
      "step": 435
    },
    {
      "epoch": 0.1008558871154291,
      "grad_norm": 1.1825700998306274,
      "learning_rate": 0.0005,
      "loss": 3.535660982131958,
      "step": 436
    },
    {
      "epoch": 0.10108720795743696,
      "grad_norm": 1.548724889755249,
      "learning_rate": 0.0005,
      "loss": 3.6623353958129883,
      "step": 437
    },
    {
      "epoch": 0.10131852879944483,
      "grad_norm": 1.9923828840255737,
      "learning_rate": 0.0005,
      "loss": 3.7521448135375977,
      "step": 438
    },
    {
      "epoch": 0.1015498496414527,
      "grad_norm": 1.8003686666488647,
      "learning_rate": 0.0005,
      "loss": 3.9304611682891846,
      "step": 439
    },
    {
      "epoch": 0.10178117048346055,
      "grad_norm": 1.4717646837234497,
      "learning_rate": 0.0005,
      "loss": 3.4425268173217773,
      "step": 440
    },
    {
      "epoch": 0.10201249132546843,
      "grad_norm": 1.7308337688446045,
      "learning_rate": 0.0005,
      "loss": 4.016409397125244,
      "step": 441
    },
    {
      "epoch": 0.10224381216747629,
      "grad_norm": 1.3642826080322266,
      "learning_rate": 0.0005,
      "loss": 3.5650429725646973,
      "step": 442
    },
    {
      "epoch": 0.10247513300948416,
      "grad_norm": 1.2597761154174805,
      "learning_rate": 0.0005,
      "loss": 3.834805965423584,
      "step": 443
    },
    {
      "epoch": 0.10270645385149202,
      "grad_norm": 1.6053354740142822,
      "learning_rate": 0.0005,
      "loss": 3.4312596321105957,
      "step": 444
    },
    {
      "epoch": 0.10293777469349988,
      "grad_norm": 1.1870125532150269,
      "learning_rate": 0.0005,
      "loss": 3.7536630630493164,
      "step": 445
    },
    {
      "epoch": 0.10316909553550775,
      "grad_norm": 1.3061825037002563,
      "learning_rate": 0.0005,
      "loss": 3.5362963676452637,
      "step": 446
    },
    {
      "epoch": 0.10340041637751561,
      "grad_norm": 1.4147453308105469,
      "learning_rate": 0.0005,
      "loss": 3.7185490131378174,
      "step": 447
    },
    {
      "epoch": 0.10363173721952348,
      "grad_norm": 2.0918564796447754,
      "learning_rate": 0.0005,
      "loss": 3.83660626411438,
      "step": 448
    },
    {
      "epoch": 0.10386305806153134,
      "grad_norm": 1.1793912649154663,
      "learning_rate": 0.0005,
      "loss": 3.6274092197418213,
      "step": 449
    },
    {
      "epoch": 0.1040943789035392,
      "grad_norm": 1.2286853790283203,
      "learning_rate": 0.0005,
      "loss": 3.575143814086914,
      "step": 450
    },
    {
      "epoch": 0.10432569974554708,
      "grad_norm": 1.3124711513519287,
      "learning_rate": 0.0005,
      "loss": 3.347653865814209,
      "step": 451
    },
    {
      "epoch": 0.10455702058755494,
      "grad_norm": 2.4710493087768555,
      "learning_rate": 0.0005,
      "loss": 3.389542579650879,
      "step": 452
    },
    {
      "epoch": 0.10478834142956281,
      "grad_norm": 1.2729560136795044,
      "learning_rate": 0.0005,
      "loss": 3.820133686065674,
      "step": 453
    },
    {
      "epoch": 0.10501966227157067,
      "grad_norm": 1.3336907625198364,
      "learning_rate": 0.0005,
      "loss": 3.482199192047119,
      "step": 454
    },
    {
      "epoch": 0.10525098311357853,
      "grad_norm": 1.2254596948623657,
      "learning_rate": 0.0005,
      "loss": 3.7696924209594727,
      "step": 455
    },
    {
      "epoch": 0.1054823039555864,
      "grad_norm": 1.3358542919158936,
      "learning_rate": 0.0005,
      "loss": 3.3369340896606445,
      "step": 456
    },
    {
      "epoch": 0.10571362479759426,
      "grad_norm": 1.4292457103729248,
      "learning_rate": 0.0005,
      "loss": 3.539534091949463,
      "step": 457
    },
    {
      "epoch": 0.10594494563960213,
      "grad_norm": 1.106173038482666,
      "learning_rate": 0.0005,
      "loss": 3.335322618484497,
      "step": 458
    },
    {
      "epoch": 0.10617626648161,
      "grad_norm": 1.4242115020751953,
      "learning_rate": 0.0005,
      "loss": 3.737891912460327,
      "step": 459
    },
    {
      "epoch": 0.10640758732361785,
      "grad_norm": 2.113426923751831,
      "learning_rate": 0.0005,
      "loss": 3.884186267852783,
      "step": 460
    },
    {
      "epoch": 0.10663890816562573,
      "grad_norm": 1.1577067375183105,
      "learning_rate": 0.0005,
      "loss": 3.573894500732422,
      "step": 461
    },
    {
      "epoch": 0.10687022900763359,
      "grad_norm": 1.4741352796554565,
      "learning_rate": 0.0005,
      "loss": 3.448483467102051,
      "step": 462
    },
    {
      "epoch": 0.10710154984964146,
      "grad_norm": 1.2678898572921753,
      "learning_rate": 0.0005,
      "loss": 3.460059642791748,
      "step": 463
    },
    {
      "epoch": 0.10733287069164932,
      "grad_norm": 1.1877330541610718,
      "learning_rate": 0.0005,
      "loss": 3.4623775482177734,
      "step": 464
    },
    {
      "epoch": 0.10756419153365718,
      "grad_norm": 1.230708122253418,
      "learning_rate": 0.0005,
      "loss": 3.558189868927002,
      "step": 465
    },
    {
      "epoch": 0.10779551237566505,
      "grad_norm": 1.7426203489303589,
      "learning_rate": 0.0005,
      "loss": 3.957674503326416,
      "step": 466
    },
    {
      "epoch": 0.10802683321767291,
      "grad_norm": 1.200211524963379,
      "learning_rate": 0.0005,
      "loss": 3.515939235687256,
      "step": 467
    },
    {
      "epoch": 0.10825815405968078,
      "grad_norm": 1.262268304824829,
      "learning_rate": 0.0005,
      "loss": 3.432215690612793,
      "step": 468
    },
    {
      "epoch": 0.10848947490168864,
      "grad_norm": 1.3649113178253174,
      "learning_rate": 0.0005,
      "loss": 3.7049007415771484,
      "step": 469
    },
    {
      "epoch": 0.1087207957436965,
      "grad_norm": 1.4979513883590698,
      "learning_rate": 0.0005,
      "loss": 3.5459978580474854,
      "step": 470
    },
    {
      "epoch": 0.10895211658570438,
      "grad_norm": 1.43906831741333,
      "learning_rate": 0.0005,
      "loss": 3.2710604667663574,
      "step": 471
    },
    {
      "epoch": 0.10918343742771223,
      "grad_norm": 1.641688585281372,
      "learning_rate": 0.0005,
      "loss": 3.3360695838928223,
      "step": 472
    },
    {
      "epoch": 0.10941475826972011,
      "grad_norm": 1.5445857048034668,
      "learning_rate": 0.0005,
      "loss": 3.5412087440490723,
      "step": 473
    },
    {
      "epoch": 0.10964607911172797,
      "grad_norm": 1.7726187705993652,
      "learning_rate": 0.0005,
      "loss": 3.4629030227661133,
      "step": 474
    },
    {
      "epoch": 0.10987739995373583,
      "grad_norm": 1.1540158987045288,
      "learning_rate": 0.0005,
      "loss": 3.847960948944092,
      "step": 475
    },
    {
      "epoch": 0.1101087207957437,
      "grad_norm": 1.2866699695587158,
      "learning_rate": 0.0005,
      "loss": 3.984067440032959,
      "step": 476
    },
    {
      "epoch": 0.11034004163775156,
      "grad_norm": 1.153364896774292,
      "learning_rate": 0.0005,
      "loss": 3.8510303497314453,
      "step": 477
    },
    {
      "epoch": 0.11057136247975943,
      "grad_norm": 1.0965770483016968,
      "learning_rate": 0.0005,
      "loss": 3.302612781524658,
      "step": 478
    },
    {
      "epoch": 0.11080268332176729,
      "grad_norm": 1.2102307081222534,
      "learning_rate": 0.0005,
      "loss": 3.571603775024414,
      "step": 479
    },
    {
      "epoch": 0.11103400416377515,
      "grad_norm": 1.1985983848571777,
      "learning_rate": 0.0005,
      "loss": 3.589000701904297,
      "step": 480
    },
    {
      "epoch": 0.11126532500578303,
      "grad_norm": 1.2915465831756592,
      "learning_rate": 0.0005,
      "loss": 3.5414772033691406,
      "step": 481
    },
    {
      "epoch": 0.11149664584779088,
      "grad_norm": 1.2913206815719604,
      "learning_rate": 0.0005,
      "loss": 3.6600546836853027,
      "step": 482
    },
    {
      "epoch": 0.11172796668979876,
      "grad_norm": 1.2274945974349976,
      "learning_rate": 0.0005,
      "loss": 3.6504368782043457,
      "step": 483
    },
    {
      "epoch": 0.11195928753180662,
      "grad_norm": 1.1247657537460327,
      "learning_rate": 0.0005,
      "loss": 3.071497917175293,
      "step": 484
    },
    {
      "epoch": 0.11219060837381448,
      "grad_norm": 1.280872106552124,
      "learning_rate": 0.0005,
      "loss": 3.7456841468811035,
      "step": 485
    },
    {
      "epoch": 0.11242192921582235,
      "grad_norm": 1.3590954542160034,
      "learning_rate": 0.0005,
      "loss": 3.978898763656616,
      "step": 486
    },
    {
      "epoch": 0.11265325005783021,
      "grad_norm": 1.26300847530365,
      "learning_rate": 0.0005,
      "loss": 3.7170886993408203,
      "step": 487
    },
    {
      "epoch": 0.11288457089983807,
      "grad_norm": 1.3795995712280273,
      "learning_rate": 0.0005,
      "loss": 3.2814252376556396,
      "step": 488
    },
    {
      "epoch": 0.11311589174184594,
      "grad_norm": 1.400175929069519,
      "learning_rate": 0.0005,
      "loss": 3.6923089027404785,
      "step": 489
    },
    {
      "epoch": 0.1133472125838538,
      "grad_norm": 1.1268240213394165,
      "learning_rate": 0.0005,
      "loss": 3.373401165008545,
      "step": 490
    },
    {
      "epoch": 0.11357853342586167,
      "grad_norm": 1.3930068016052246,
      "learning_rate": 0.0005,
      "loss": 3.515603542327881,
      "step": 491
    },
    {
      "epoch": 0.11380985426786953,
      "grad_norm": 1.8088526725769043,
      "learning_rate": 0.0005,
      "loss": 3.488384246826172,
      "step": 492
    },
    {
      "epoch": 0.1140411751098774,
      "grad_norm": 1.69031822681427,
      "learning_rate": 0.0005,
      "loss": 3.8375024795532227,
      "step": 493
    },
    {
      "epoch": 0.11427249595188527,
      "grad_norm": 1.3526157140731812,
      "learning_rate": 0.0005,
      "loss": 3.546142339706421,
      "step": 494
    },
    {
      "epoch": 0.11450381679389313,
      "grad_norm": 1.899132490158081,
      "learning_rate": 0.0005,
      "loss": 3.127988576889038,
      "step": 495
    },
    {
      "epoch": 0.114735137635901,
      "grad_norm": 1.472530484199524,
      "learning_rate": 0.0005,
      "loss": 3.471930503845215,
      "step": 496
    },
    {
      "epoch": 0.11496645847790886,
      "grad_norm": 1.3603582382202148,
      "learning_rate": 0.0005,
      "loss": 3.227747917175293,
      "step": 497
    },
    {
      "epoch": 0.11519777931991672,
      "grad_norm": 1.4405471086502075,
      "learning_rate": 0.0005,
      "loss": 3.631979465484619,
      "step": 498
    },
    {
      "epoch": 0.11542910016192459,
      "grad_norm": 1.3604731559753418,
      "learning_rate": 0.0005,
      "loss": 3.568848133087158,
      "step": 499
    },
    {
      "epoch": 0.11566042100393245,
      "grad_norm": 1.450147271156311,
      "learning_rate": 0.0005,
      "loss": 3.6002326011657715,
      "step": 500
    },
    {
      "epoch": 0.11589174184594032,
      "grad_norm": 1.2992982864379883,
      "learning_rate": 0.0005,
      "loss": 3.455394744873047,
      "step": 501
    },
    {
      "epoch": 0.11612306268794818,
      "grad_norm": 2.516400098800659,
      "learning_rate": 0.0005,
      "loss": 3.7785158157348633,
      "step": 502
    },
    {
      "epoch": 0.11635438352995604,
      "grad_norm": 1.3692606687545776,
      "learning_rate": 0.0005,
      "loss": 3.90928316116333,
      "step": 503
    },
    {
      "epoch": 0.11658570437196392,
      "grad_norm": 1.5577261447906494,
      "learning_rate": 0.0005,
      "loss": 3.5327515602111816,
      "step": 504
    },
    {
      "epoch": 0.11681702521397178,
      "grad_norm": 1.6329457759857178,
      "learning_rate": 0.0005,
      "loss": 3.5062522888183594,
      "step": 505
    },
    {
      "epoch": 0.11704834605597965,
      "grad_norm": 1.2849836349487305,
      "learning_rate": 0.0005,
      "loss": 3.400480031967163,
      "step": 506
    },
    {
      "epoch": 0.11727966689798751,
      "grad_norm": 1.8091864585876465,
      "learning_rate": 0.0005,
      "loss": 3.7168798446655273,
      "step": 507
    },
    {
      "epoch": 0.11751098773999537,
      "grad_norm": 2.485085964202881,
      "learning_rate": 0.0005,
      "loss": 3.47760009765625,
      "step": 508
    },
    {
      "epoch": 0.11774230858200324,
      "grad_norm": 1.8858528137207031,
      "learning_rate": 0.0005,
      "loss": 3.786884307861328,
      "step": 509
    },
    {
      "epoch": 0.1179736294240111,
      "grad_norm": 1.174271583557129,
      "learning_rate": 0.0005,
      "loss": 3.439262866973877,
      "step": 510
    },
    {
      "epoch": 0.11820495026601897,
      "grad_norm": 1.1595349311828613,
      "learning_rate": 0.0005,
      "loss": 3.5784926414489746,
      "step": 511
    },
    {
      "epoch": 0.11843627110802683,
      "grad_norm": 1.3752022981643677,
      "learning_rate": 0.0005,
      "loss": 3.7806644439697266,
      "step": 512
    },
    {
      "epoch": 0.11866759195003469,
      "grad_norm": 1.946496844291687,
      "learning_rate": 0.0005,
      "loss": 4.025772571563721,
      "step": 513
    },
    {
      "epoch": 0.11889891279204257,
      "grad_norm": 1.1992785930633545,
      "learning_rate": 0.0005,
      "loss": 3.2654073238372803,
      "step": 514
    },
    {
      "epoch": 0.11913023363405043,
      "grad_norm": 1.3599684238433838,
      "learning_rate": 0.0005,
      "loss": 3.516303062438965,
      "step": 515
    },
    {
      "epoch": 0.1193615544760583,
      "grad_norm": 1.231698989868164,
      "learning_rate": 0.0005,
      "loss": 3.5605268478393555,
      "step": 516
    },
    {
      "epoch": 0.11959287531806616,
      "grad_norm": 1.1288044452667236,
      "learning_rate": 0.0005,
      "loss": 3.302170515060425,
      "step": 517
    },
    {
      "epoch": 0.11982419616007402,
      "grad_norm": 1.2037246227264404,
      "learning_rate": 0.0005,
      "loss": 3.442370891571045,
      "step": 518
    },
    {
      "epoch": 0.12005551700208189,
      "grad_norm": 1.2970250844955444,
      "learning_rate": 0.0005,
      "loss": 3.7239625453948975,
      "step": 519
    },
    {
      "epoch": 0.12028683784408975,
      "grad_norm": 1.3214006423950195,
      "learning_rate": 0.0005,
      "loss": 3.503540277481079,
      "step": 520
    },
    {
      "epoch": 0.12051815868609762,
      "grad_norm": 1.451581597328186,
      "learning_rate": 0.0005,
      "loss": 3.5711562633514404,
      "step": 521
    },
    {
      "epoch": 0.12074947952810548,
      "grad_norm": 1.5786181688308716,
      "learning_rate": 0.0005,
      "loss": 3.590794563293457,
      "step": 522
    },
    {
      "epoch": 0.12098080037011334,
      "grad_norm": 1.3002755641937256,
      "learning_rate": 0.0005,
      "loss": 3.3497262001037598,
      "step": 523
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 1.3194364309310913,
      "learning_rate": 0.0005,
      "loss": 3.7438111305236816,
      "step": 524
    },
    {
      "epoch": 0.12144344205412907,
      "grad_norm": 1.2198050022125244,
      "learning_rate": 0.0005,
      "loss": 3.656121015548706,
      "step": 525
    },
    {
      "epoch": 0.12167476289613695,
      "grad_norm": 1.3048940896987915,
      "learning_rate": 0.0005,
      "loss": 3.3687942028045654,
      "step": 526
    },
    {
      "epoch": 0.12190608373814481,
      "grad_norm": 1.1600085496902466,
      "learning_rate": 0.0005,
      "loss": 3.512143611907959,
      "step": 527
    },
    {
      "epoch": 0.12213740458015267,
      "grad_norm": 1.179613709449768,
      "learning_rate": 0.0005,
      "loss": 3.6573328971862793,
      "step": 528
    },
    {
      "epoch": 0.12236872542216054,
      "grad_norm": 1.7404667139053345,
      "learning_rate": 0.0005,
      "loss": 3.5809154510498047,
      "step": 529
    },
    {
      "epoch": 0.1226000462641684,
      "grad_norm": 1.8906854391098022,
      "learning_rate": 0.0005,
      "loss": 3.3944802284240723,
      "step": 530
    },
    {
      "epoch": 0.12283136710617627,
      "grad_norm": 1.37943696975708,
      "learning_rate": 0.0005,
      "loss": 3.56881046295166,
      "step": 531
    },
    {
      "epoch": 0.12306268794818413,
      "grad_norm": 1.2674839496612549,
      "learning_rate": 0.0005,
      "loss": 3.3836827278137207,
      "step": 532
    },
    {
      "epoch": 0.12329400879019199,
      "grad_norm": 1.8860241174697876,
      "learning_rate": 0.0005,
      "loss": 3.4684255123138428,
      "step": 533
    },
    {
      "epoch": 0.12352532963219987,
      "grad_norm": 1.4313973188400269,
      "learning_rate": 0.0005,
      "loss": 3.4876677989959717,
      "step": 534
    },
    {
      "epoch": 0.12375665047420772,
      "grad_norm": 1.0531870126724243,
      "learning_rate": 0.0005,
      "loss": 3.459484100341797,
      "step": 535
    },
    {
      "epoch": 0.1239879713162156,
      "grad_norm": 1.0197334289550781,
      "learning_rate": 0.0005,
      "loss": 3.4874379634857178,
      "step": 536
    },
    {
      "epoch": 0.12421929215822346,
      "grad_norm": 1.4189910888671875,
      "learning_rate": 0.0005,
      "loss": 3.60015869140625,
      "step": 537
    },
    {
      "epoch": 0.12445061300023132,
      "grad_norm": 1.4724828004837036,
      "learning_rate": 0.0005,
      "loss": 3.5965356826782227,
      "step": 538
    },
    {
      "epoch": 0.12468193384223919,
      "grad_norm": 1.6341608762741089,
      "learning_rate": 0.0005,
      "loss": 3.5278873443603516,
      "step": 539
    },
    {
      "epoch": 0.12491325468424705,
      "grad_norm": 1.3242419958114624,
      "learning_rate": 0.0005,
      "loss": 3.51334285736084,
      "step": 540
    },
    {
      "epoch": 0.1251445755262549,
      "grad_norm": 1.2699661254882812,
      "learning_rate": 0.0005,
      "loss": 3.3776941299438477,
      "step": 541
    },
    {
      "epoch": 0.12537589636826277,
      "grad_norm": 1.3223916292190552,
      "learning_rate": 0.0005,
      "loss": 3.7115068435668945,
      "step": 542
    },
    {
      "epoch": 0.12560721721027066,
      "grad_norm": 1.8581551313400269,
      "learning_rate": 0.0005,
      "loss": 3.9876601696014404,
      "step": 543
    },
    {
      "epoch": 0.12583853805227851,
      "grad_norm": 1.7600761651992798,
      "learning_rate": 0.0005,
      "loss": 3.5500097274780273,
      "step": 544
    },
    {
      "epoch": 0.12606985889428637,
      "grad_norm": 1.6273729801177979,
      "learning_rate": 0.0005,
      "loss": 3.5653741359710693,
      "step": 545
    },
    {
      "epoch": 0.12630117973629423,
      "grad_norm": 1.2580636739730835,
      "learning_rate": 0.0005,
      "loss": 3.5469181537628174,
      "step": 546
    },
    {
      "epoch": 0.1265325005783021,
      "grad_norm": 1.5671559572219849,
      "learning_rate": 0.0005,
      "loss": 3.598200798034668,
      "step": 547
    },
    {
      "epoch": 0.12676382142030998,
      "grad_norm": 1.407334327697754,
      "learning_rate": 0.0005,
      "loss": 3.658641815185547,
      "step": 548
    },
    {
      "epoch": 0.12699514226231784,
      "grad_norm": 1.6668312549591064,
      "learning_rate": 0.0005,
      "loss": 3.218332290649414,
      "step": 549
    },
    {
      "epoch": 0.1272264631043257,
      "grad_norm": 1.5006182193756104,
      "learning_rate": 0.0005,
      "loss": 3.7429261207580566,
      "step": 550
    },
    {
      "epoch": 0.12745778394633356,
      "grad_norm": 1.169505000114441,
      "learning_rate": 0.0005,
      "loss": 3.3945908546447754,
      "step": 551
    },
    {
      "epoch": 0.12768910478834142,
      "grad_norm": 1.230468511581421,
      "learning_rate": 0.0005,
      "loss": 3.812133550643921,
      "step": 552
    },
    {
      "epoch": 0.1279204256303493,
      "grad_norm": 1.5782853364944458,
      "learning_rate": 0.0005,
      "loss": 3.2366127967834473,
      "step": 553
    },
    {
      "epoch": 0.12815174647235716,
      "grad_norm": 1.3537098169326782,
      "learning_rate": 0.0005,
      "loss": 3.331085205078125,
      "step": 554
    },
    {
      "epoch": 0.12838306731436502,
      "grad_norm": 1.5014517307281494,
      "learning_rate": 0.0005,
      "loss": 3.319413185119629,
      "step": 555
    },
    {
      "epoch": 0.12861438815637288,
      "grad_norm": 1.5295897722244263,
      "learning_rate": 0.0005,
      "loss": 3.901078462600708,
      "step": 556
    },
    {
      "epoch": 0.12884570899838074,
      "grad_norm": 1.4237335920333862,
      "learning_rate": 0.0005,
      "loss": 3.6577067375183105,
      "step": 557
    },
    {
      "epoch": 0.12907702984038863,
      "grad_norm": 1.544150710105896,
      "learning_rate": 0.0005,
      "loss": 3.719392776489258,
      "step": 558
    },
    {
      "epoch": 0.1293083506823965,
      "grad_norm": 1.5320570468902588,
      "learning_rate": 0.0005,
      "loss": 3.4504215717315674,
      "step": 559
    },
    {
      "epoch": 0.12953967152440435,
      "grad_norm": 1.3028250932693481,
      "learning_rate": 0.0005,
      "loss": 3.637373208999634,
      "step": 560
    },
    {
      "epoch": 0.1297709923664122,
      "grad_norm": 1.1089057922363281,
      "learning_rate": 0.0005,
      "loss": 3.4153079986572266,
      "step": 561
    },
    {
      "epoch": 0.13000231320842007,
      "grad_norm": 1.170955777168274,
      "learning_rate": 0.0005,
      "loss": 3.3482589721679688,
      "step": 562
    },
    {
      "epoch": 0.13023363405042795,
      "grad_norm": 1.2560747861862183,
      "learning_rate": 0.0005,
      "loss": 3.4238293170928955,
      "step": 563
    },
    {
      "epoch": 0.1304649548924358,
      "grad_norm": 1.3428409099578857,
      "learning_rate": 0.0005,
      "loss": 3.5758283138275146,
      "step": 564
    },
    {
      "epoch": 0.13069627573444367,
      "grad_norm": 1.4138888120651245,
      "learning_rate": 0.0005,
      "loss": 3.435670852661133,
      "step": 565
    },
    {
      "epoch": 0.13092759657645153,
      "grad_norm": 1.1262266635894775,
      "learning_rate": 0.0005,
      "loss": 3.28348970413208,
      "step": 566
    },
    {
      "epoch": 0.1311589174184594,
      "grad_norm": 1.197974443435669,
      "learning_rate": 0.0005,
      "loss": 3.623645305633545,
      "step": 567
    },
    {
      "epoch": 0.13139023826046728,
      "grad_norm": 1.1944067478179932,
      "learning_rate": 0.0005,
      "loss": 3.6314449310302734,
      "step": 568
    },
    {
      "epoch": 0.13162155910247514,
      "grad_norm": 1.8929537534713745,
      "learning_rate": 0.0005,
      "loss": 3.8252482414245605,
      "step": 569
    },
    {
      "epoch": 0.131852879944483,
      "grad_norm": 1.5800718069076538,
      "learning_rate": 0.0005,
      "loss": 3.687775135040283,
      "step": 570
    },
    {
      "epoch": 0.13208420078649086,
      "grad_norm": 1.4166808128356934,
      "learning_rate": 0.0005,
      "loss": 3.643080711364746,
      "step": 571
    },
    {
      "epoch": 0.13231552162849872,
      "grad_norm": 1.2146072387695312,
      "learning_rate": 0.0005,
      "loss": 3.457432746887207,
      "step": 572
    },
    {
      "epoch": 0.1325468424705066,
      "grad_norm": 1.845787763595581,
      "learning_rate": 0.0005,
      "loss": 3.3692545890808105,
      "step": 573
    },
    {
      "epoch": 0.13277816331251446,
      "grad_norm": 1.2476738691329956,
      "learning_rate": 0.0005,
      "loss": 3.3367536067962646,
      "step": 574
    },
    {
      "epoch": 0.13300948415452232,
      "grad_norm": 1.116822361946106,
      "learning_rate": 0.0005,
      "loss": 3.3847694396972656,
      "step": 575
    },
    {
      "epoch": 0.13324080499653018,
      "grad_norm": 1.151185154914856,
      "learning_rate": 0.0005,
      "loss": 3.552168369293213,
      "step": 576
    },
    {
      "epoch": 0.13347212583853804,
      "grad_norm": 1.3237589597702026,
      "learning_rate": 0.0005,
      "loss": 3.321234703063965,
      "step": 577
    },
    {
      "epoch": 0.13370344668054593,
      "grad_norm": 1.3274856805801392,
      "learning_rate": 0.0005,
      "loss": 3.1942849159240723,
      "step": 578
    },
    {
      "epoch": 0.1339347675225538,
      "grad_norm": 1.413801908493042,
      "learning_rate": 0.0005,
      "loss": 3.198539972305298,
      "step": 579
    },
    {
      "epoch": 0.13416608836456165,
      "grad_norm": 1.2394903898239136,
      "learning_rate": 0.0005,
      "loss": 3.288175106048584,
      "step": 580
    },
    {
      "epoch": 0.1343974092065695,
      "grad_norm": 1.2983100414276123,
      "learning_rate": 0.0005,
      "loss": 3.429013729095459,
      "step": 581
    },
    {
      "epoch": 0.13462873004857737,
      "grad_norm": 1.5394004583358765,
      "learning_rate": 0.0005,
      "loss": 3.1730141639709473,
      "step": 582
    },
    {
      "epoch": 0.13486005089058525,
      "grad_norm": 1.5461608171463013,
      "learning_rate": 0.0005,
      "loss": 3.443159580230713,
      "step": 583
    },
    {
      "epoch": 0.1350913717325931,
      "grad_norm": 1.3570305109024048,
      "learning_rate": 0.0005,
      "loss": 3.5281424522399902,
      "step": 584
    },
    {
      "epoch": 0.13532269257460097,
      "grad_norm": 1.1361327171325684,
      "learning_rate": 0.0005,
      "loss": 3.7052693367004395,
      "step": 585
    },
    {
      "epoch": 0.13555401341660883,
      "grad_norm": 1.577549934387207,
      "learning_rate": 0.0005,
      "loss": 3.1001696586608887,
      "step": 586
    },
    {
      "epoch": 0.1357853342586167,
      "grad_norm": 1.8217463493347168,
      "learning_rate": 0.0005,
      "loss": 3.452806234359741,
      "step": 587
    },
    {
      "epoch": 0.13601665510062458,
      "grad_norm": 1.630082368850708,
      "learning_rate": 0.0005,
      "loss": 3.498793601989746,
      "step": 588
    },
    {
      "epoch": 0.13624797594263244,
      "grad_norm": 1.7521365880966187,
      "learning_rate": 0.0005,
      "loss": 3.58341646194458,
      "step": 589
    },
    {
      "epoch": 0.1364792967846403,
      "grad_norm": 1.3720567226409912,
      "learning_rate": 0.0005,
      "loss": 3.376781463623047,
      "step": 590
    },
    {
      "epoch": 0.13671061762664816,
      "grad_norm": 1.384237289428711,
      "learning_rate": 0.0005,
      "loss": 3.4718611240386963,
      "step": 591
    },
    {
      "epoch": 0.13694193846865602,
      "grad_norm": 1.1781089305877686,
      "learning_rate": 0.0005,
      "loss": 3.4557981491088867,
      "step": 592
    },
    {
      "epoch": 0.1371732593106639,
      "grad_norm": 1.716142177581787,
      "learning_rate": 0.0005,
      "loss": 3.525912046432495,
      "step": 593
    },
    {
      "epoch": 0.13740458015267176,
      "grad_norm": 1.7160656452178955,
      "learning_rate": 0.0005,
      "loss": 3.57981276512146,
      "step": 594
    },
    {
      "epoch": 0.13763590099467962,
      "grad_norm": 1.8138409852981567,
      "learning_rate": 0.0005,
      "loss": 3.50663161277771,
      "step": 595
    },
    {
      "epoch": 0.13786722183668748,
      "grad_norm": 2.4383041858673096,
      "learning_rate": 0.0005,
      "loss": 3.660183906555176,
      "step": 596
    },
    {
      "epoch": 0.13809854267869534,
      "grad_norm": 1.1965551376342773,
      "learning_rate": 0.0005,
      "loss": 3.3439743518829346,
      "step": 597
    },
    {
      "epoch": 0.13832986352070323,
      "grad_norm": 1.3914501667022705,
      "learning_rate": 0.0005,
      "loss": 3.466644763946533,
      "step": 598
    },
    {
      "epoch": 0.1385611843627111,
      "grad_norm": 1.8147977590560913,
      "learning_rate": 0.0005,
      "loss": 3.703763008117676,
      "step": 599
    },
    {
      "epoch": 0.13879250520471895,
      "grad_norm": 1.4076337814331055,
      "learning_rate": 0.0005,
      "loss": 3.732654094696045,
      "step": 600
    },
    {
      "epoch": 0.1390238260467268,
      "grad_norm": 1.1649798154830933,
      "learning_rate": 0.0005,
      "loss": 3.6751463413238525,
      "step": 601
    },
    {
      "epoch": 0.13925514688873467,
      "grad_norm": 1.2118433713912964,
      "learning_rate": 0.0005,
      "loss": 3.5957393646240234,
      "step": 602
    },
    {
      "epoch": 0.13948646773074255,
      "grad_norm": 1.2545245885849,
      "learning_rate": 0.0005,
      "loss": 3.38059139251709,
      "step": 603
    },
    {
      "epoch": 0.1397177885727504,
      "grad_norm": 1.1432305574417114,
      "learning_rate": 0.0005,
      "loss": 3.7958054542541504,
      "step": 604
    },
    {
      "epoch": 0.13994910941475827,
      "grad_norm": 1.2346628904342651,
      "learning_rate": 0.0005,
      "loss": 3.5356650352478027,
      "step": 605
    },
    {
      "epoch": 0.14018043025676613,
      "grad_norm": 1.0456628799438477,
      "learning_rate": 0.0005,
      "loss": 3.607987403869629,
      "step": 606
    },
    {
      "epoch": 0.140411751098774,
      "grad_norm": 1.1966588497161865,
      "learning_rate": 0.0005,
      "loss": 3.396857976913452,
      "step": 607
    },
    {
      "epoch": 0.14064307194078188,
      "grad_norm": 1.6543394327163696,
      "learning_rate": 0.0005,
      "loss": 3.4163284301757812,
      "step": 608
    },
    {
      "epoch": 0.14087439278278974,
      "grad_norm": 1.4592249393463135,
      "learning_rate": 0.0005,
      "loss": 3.1427998542785645,
      "step": 609
    },
    {
      "epoch": 0.1411057136247976,
      "grad_norm": 1.1789058446884155,
      "learning_rate": 0.0005,
      "loss": 3.3885440826416016,
      "step": 610
    },
    {
      "epoch": 0.14133703446680546,
      "grad_norm": 1.3298428058624268,
      "learning_rate": 0.0005,
      "loss": 3.4106950759887695,
      "step": 611
    },
    {
      "epoch": 0.14156835530881332,
      "grad_norm": 1.3522566556930542,
      "learning_rate": 0.0005,
      "loss": 3.826760768890381,
      "step": 612
    },
    {
      "epoch": 0.1417996761508212,
      "grad_norm": 1.1625523567199707,
      "learning_rate": 0.0005,
      "loss": 3.5555498600006104,
      "step": 613
    },
    {
      "epoch": 0.14203099699282906,
      "grad_norm": 1.2000396251678467,
      "learning_rate": 0.0005,
      "loss": 3.6337404251098633,
      "step": 614
    },
    {
      "epoch": 0.14226231783483692,
      "grad_norm": 1.2712091207504272,
      "learning_rate": 0.0005,
      "loss": 3.3445560932159424,
      "step": 615
    },
    {
      "epoch": 0.14249363867684478,
      "grad_norm": 1.2803453207015991,
      "learning_rate": 0.0005,
      "loss": 3.748885154724121,
      "step": 616
    },
    {
      "epoch": 0.14272495951885264,
      "grad_norm": 1.4957033395767212,
      "learning_rate": 0.0005,
      "loss": 3.4533586502075195,
      "step": 617
    },
    {
      "epoch": 0.14295628036086053,
      "grad_norm": 1.2208092212677002,
      "learning_rate": 0.0005,
      "loss": 3.5315399169921875,
      "step": 618
    },
    {
      "epoch": 0.1431876012028684,
      "grad_norm": 1.2602993249893188,
      "learning_rate": 0.0005,
      "loss": 3.6447839736938477,
      "step": 619
    },
    {
      "epoch": 0.14341892204487625,
      "grad_norm": 1.5151394605636597,
      "learning_rate": 0.0005,
      "loss": 3.318558931350708,
      "step": 620
    },
    {
      "epoch": 0.1436502428868841,
      "grad_norm": 2.0215747356414795,
      "learning_rate": 0.0005,
      "loss": 3.6402697563171387,
      "step": 621
    },
    {
      "epoch": 0.14388156372889196,
      "grad_norm": 1.1843186616897583,
      "learning_rate": 0.0005,
      "loss": 3.4004905223846436,
      "step": 622
    },
    {
      "epoch": 0.14411288457089985,
      "grad_norm": 1.4607198238372803,
      "learning_rate": 0.0005,
      "loss": 3.399688720703125,
      "step": 623
    },
    {
      "epoch": 0.1443442054129077,
      "grad_norm": 1.022774577140808,
      "learning_rate": 0.0005,
      "loss": 3.3924598693847656,
      "step": 624
    },
    {
      "epoch": 0.14457552625491557,
      "grad_norm": 1.4492988586425781,
      "learning_rate": 0.0005,
      "loss": 3.598701000213623,
      "step": 625
    },
    {
      "epoch": 0.14480684709692343,
      "grad_norm": 1.8712732791900635,
      "learning_rate": 0.0005,
      "loss": 3.689328908920288,
      "step": 626
    },
    {
      "epoch": 0.1450381679389313,
      "grad_norm": 1.610424280166626,
      "learning_rate": 0.0005,
      "loss": 3.5095927715301514,
      "step": 627
    },
    {
      "epoch": 0.14526948878093915,
      "grad_norm": 1.3450469970703125,
      "learning_rate": 0.0005,
      "loss": 3.1837234497070312,
      "step": 628
    },
    {
      "epoch": 0.14550080962294704,
      "grad_norm": 1.3187496662139893,
      "learning_rate": 0.0005,
      "loss": 3.411328077316284,
      "step": 629
    },
    {
      "epoch": 0.1457321304649549,
      "grad_norm": 1.2118916511535645,
      "learning_rate": 0.0005,
      "loss": 3.663698196411133,
      "step": 630
    },
    {
      "epoch": 0.14596345130696275,
      "grad_norm": 1.1627213954925537,
      "learning_rate": 0.0005,
      "loss": 3.471219301223755,
      "step": 631
    },
    {
      "epoch": 0.14619477214897061,
      "grad_norm": 1.628220796585083,
      "learning_rate": 0.0005,
      "loss": 3.504455327987671,
      "step": 632
    },
    {
      "epoch": 0.14642609299097847,
      "grad_norm": 1.8088181018829346,
      "learning_rate": 0.0005,
      "loss": 3.557339668273926,
      "step": 633
    },
    {
      "epoch": 0.14665741383298636,
      "grad_norm": 1.5703692436218262,
      "learning_rate": 0.0005,
      "loss": 3.5600056648254395,
      "step": 634
    },
    {
      "epoch": 0.14688873467499422,
      "grad_norm": 1.2166783809661865,
      "learning_rate": 0.0005,
      "loss": 3.4045751094818115,
      "step": 635
    },
    {
      "epoch": 0.14712005551700208,
      "grad_norm": 1.467523455619812,
      "learning_rate": 0.0005,
      "loss": 3.427823305130005,
      "step": 636
    },
    {
      "epoch": 0.14735137635900994,
      "grad_norm": 1.4250260591506958,
      "learning_rate": 0.0005,
      "loss": 3.211787223815918,
      "step": 637
    },
    {
      "epoch": 0.1475826972010178,
      "grad_norm": 1.1821876764297485,
      "learning_rate": 0.0005,
      "loss": 3.7899017333984375,
      "step": 638
    },
    {
      "epoch": 0.14781401804302569,
      "grad_norm": 1.386946201324463,
      "learning_rate": 0.0005,
      "loss": 3.3122973442077637,
      "step": 639
    },
    {
      "epoch": 0.14804533888503354,
      "grad_norm": 1.0975596904754639,
      "learning_rate": 0.0005,
      "loss": 3.931021213531494,
      "step": 640
    },
    {
      "epoch": 0.1482766597270414,
      "grad_norm": 1.0824191570281982,
      "learning_rate": 0.0005,
      "loss": 3.468280792236328,
      "step": 641
    },
    {
      "epoch": 0.14850798056904926,
      "grad_norm": 1.4306126832962036,
      "learning_rate": 0.0005,
      "loss": 3.513627529144287,
      "step": 642
    },
    {
      "epoch": 0.14873930141105712,
      "grad_norm": 1.239931344985962,
      "learning_rate": 0.0005,
      "loss": 3.4459848403930664,
      "step": 643
    },
    {
      "epoch": 0.148970622253065,
      "grad_norm": 1.206668734550476,
      "learning_rate": 0.0005,
      "loss": 3.382061004638672,
      "step": 644
    },
    {
      "epoch": 0.14920194309507287,
      "grad_norm": 1.296126127243042,
      "learning_rate": 0.0005,
      "loss": 3.374211549758911,
      "step": 645
    },
    {
      "epoch": 0.14943326393708073,
      "grad_norm": 1.2585055828094482,
      "learning_rate": 0.0005,
      "loss": 3.6765336990356445,
      "step": 646
    },
    {
      "epoch": 0.1496645847790886,
      "grad_norm": 1.4006315469741821,
      "learning_rate": 0.0005,
      "loss": 3.5893044471740723,
      "step": 647
    },
    {
      "epoch": 0.14989590562109645,
      "grad_norm": 1.2726784944534302,
      "learning_rate": 0.0005,
      "loss": 3.2051329612731934,
      "step": 648
    },
    {
      "epoch": 0.15012722646310434,
      "grad_norm": 1.410508632659912,
      "learning_rate": 0.0005,
      "loss": 3.3701364994049072,
      "step": 649
    },
    {
      "epoch": 0.1503585473051122,
      "grad_norm": 1.107404112815857,
      "learning_rate": 0.0005,
      "loss": 3.388826370239258,
      "step": 650
    },
    {
      "epoch": 0.15058986814712005,
      "grad_norm": 1.294589638710022,
      "learning_rate": 0.0005,
      "loss": 3.9462227821350098,
      "step": 651
    },
    {
      "epoch": 0.1508211889891279,
      "grad_norm": 1.3366820812225342,
      "learning_rate": 0.0005,
      "loss": 3.353881359100342,
      "step": 652
    },
    {
      "epoch": 0.15105250983113577,
      "grad_norm": 1.1166701316833496,
      "learning_rate": 0.0005,
      "loss": 3.686614513397217,
      "step": 653
    },
    {
      "epoch": 0.15128383067314366,
      "grad_norm": 1.1368229389190674,
      "learning_rate": 0.0005,
      "loss": 3.556337356567383,
      "step": 654
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 1.3830540180206299,
      "learning_rate": 0.0005,
      "loss": 3.245626449584961,
      "step": 655
    },
    {
      "epoch": 0.15174647235715938,
      "grad_norm": 1.4293164014816284,
      "learning_rate": 0.0005,
      "loss": 3.481471538543701,
      "step": 656
    },
    {
      "epoch": 0.15197779319916724,
      "grad_norm": 1.59626042842865,
      "learning_rate": 0.0005,
      "loss": 3.1526200771331787,
      "step": 657
    },
    {
      "epoch": 0.1522091140411751,
      "grad_norm": 1.150602102279663,
      "learning_rate": 0.0005,
      "loss": 3.510307550430298,
      "step": 658
    },
    {
      "epoch": 0.15244043488318298,
      "grad_norm": 1.5180809497833252,
      "learning_rate": 0.0005,
      "loss": 3.510369062423706,
      "step": 659
    },
    {
      "epoch": 0.15267175572519084,
      "grad_norm": 1.26960027217865,
      "learning_rate": 0.0005,
      "loss": 3.466217279434204,
      "step": 660
    },
    {
      "epoch": 0.1529030765671987,
      "grad_norm": 1.581087589263916,
      "learning_rate": 0.0005,
      "loss": 3.602193593978882,
      "step": 661
    },
    {
      "epoch": 0.15313439740920656,
      "grad_norm": 1.4611765146255493,
      "learning_rate": 0.0005,
      "loss": 3.8421154022216797,
      "step": 662
    },
    {
      "epoch": 0.15336571825121442,
      "grad_norm": 1.2155805826187134,
      "learning_rate": 0.0005,
      "loss": 3.553218364715576,
      "step": 663
    },
    {
      "epoch": 0.1535970390932223,
      "grad_norm": 1.3946824073791504,
      "learning_rate": 0.0005,
      "loss": 3.1322572231292725,
      "step": 664
    },
    {
      "epoch": 0.15382835993523017,
      "grad_norm": 1.7121953964233398,
      "learning_rate": 0.0005,
      "loss": 3.6803784370422363,
      "step": 665
    },
    {
      "epoch": 0.15405968077723803,
      "grad_norm": 1.366668462753296,
      "learning_rate": 0.0005,
      "loss": 3.4653992652893066,
      "step": 666
    },
    {
      "epoch": 0.1542910016192459,
      "grad_norm": 1.4021118879318237,
      "learning_rate": 0.0005,
      "loss": 3.317803144454956,
      "step": 667
    },
    {
      "epoch": 0.15452232246125375,
      "grad_norm": 1.330843210220337,
      "learning_rate": 0.0005,
      "loss": 3.5429234504699707,
      "step": 668
    },
    {
      "epoch": 0.15475364330326163,
      "grad_norm": 1.193773627281189,
      "learning_rate": 0.0005,
      "loss": 3.539095163345337,
      "step": 669
    },
    {
      "epoch": 0.1549849641452695,
      "grad_norm": 2.2522695064544678,
      "learning_rate": 0.0005,
      "loss": 3.46757173538208,
      "step": 670
    },
    {
      "epoch": 0.15521628498727735,
      "grad_norm": 1.1951242685317993,
      "learning_rate": 0.0005,
      "loss": 3.6787052154541016,
      "step": 671
    },
    {
      "epoch": 0.1554476058292852,
      "grad_norm": 1.2542895078659058,
      "learning_rate": 0.0005,
      "loss": 3.751192808151245,
      "step": 672
    },
    {
      "epoch": 0.15567892667129307,
      "grad_norm": 1.2557978630065918,
      "learning_rate": 0.0005,
      "loss": 3.5589685440063477,
      "step": 673
    },
    {
      "epoch": 0.15591024751330096,
      "grad_norm": 1.0860798358917236,
      "learning_rate": 0.0005,
      "loss": 3.542313575744629,
      "step": 674
    },
    {
      "epoch": 0.15614156835530882,
      "grad_norm": 1.1598299741744995,
      "learning_rate": 0.0005,
      "loss": 3.3356776237487793,
      "step": 675
    },
    {
      "epoch": 0.15637288919731668,
      "grad_norm": 1.8439626693725586,
      "learning_rate": 0.0005,
      "loss": 3.350156784057617,
      "step": 676
    },
    {
      "epoch": 0.15660421003932454,
      "grad_norm": 1.2392619848251343,
      "learning_rate": 0.0005,
      "loss": 3.303926944732666,
      "step": 677
    },
    {
      "epoch": 0.1568355308813324,
      "grad_norm": 1.409317970275879,
      "learning_rate": 0.0005,
      "loss": 3.32956862449646,
      "step": 678
    },
    {
      "epoch": 0.15706685172334028,
      "grad_norm": 1.3699605464935303,
      "learning_rate": 0.0005,
      "loss": 3.2832956314086914,
      "step": 679
    },
    {
      "epoch": 0.15729817256534814,
      "grad_norm": 1.3618583679199219,
      "learning_rate": 0.0005,
      "loss": 3.516282320022583,
      "step": 680
    },
    {
      "epoch": 0.157529493407356,
      "grad_norm": 1.5556623935699463,
      "learning_rate": 0.0005,
      "loss": 3.5278406143188477,
      "step": 681
    },
    {
      "epoch": 0.15776081424936386,
      "grad_norm": 1.6931558847427368,
      "learning_rate": 0.0005,
      "loss": 3.437887191772461,
      "step": 682
    },
    {
      "epoch": 0.15799213509137172,
      "grad_norm": 1.2844809293746948,
      "learning_rate": 0.0005,
      "loss": 3.4744372367858887,
      "step": 683
    },
    {
      "epoch": 0.1582234559333796,
      "grad_norm": 1.2358312606811523,
      "learning_rate": 0.0005,
      "loss": 3.360698699951172,
      "step": 684
    },
    {
      "epoch": 0.15845477677538747,
      "grad_norm": 1.4391398429870605,
      "learning_rate": 0.0005,
      "loss": 3.318373203277588,
      "step": 685
    },
    {
      "epoch": 0.15868609761739533,
      "grad_norm": 1.5433902740478516,
      "learning_rate": 0.0005,
      "loss": 3.4011247158050537,
      "step": 686
    },
    {
      "epoch": 0.1589174184594032,
      "grad_norm": 1.9060453176498413,
      "learning_rate": 0.0005,
      "loss": 3.610473155975342,
      "step": 687
    },
    {
      "epoch": 0.15914873930141105,
      "grad_norm": 1.82875394821167,
      "learning_rate": 0.0005,
      "loss": 3.3859310150146484,
      "step": 688
    },
    {
      "epoch": 0.15938006014341893,
      "grad_norm": 1.9303629398345947,
      "learning_rate": 0.0005,
      "loss": 3.2969913482666016,
      "step": 689
    },
    {
      "epoch": 0.1596113809854268,
      "grad_norm": 1.1919249296188354,
      "learning_rate": 0.0005,
      "loss": 3.6440837383270264,
      "step": 690
    },
    {
      "epoch": 0.15984270182743465,
      "grad_norm": 1.2019916772842407,
      "learning_rate": 0.0005,
      "loss": 3.5655462741851807,
      "step": 691
    },
    {
      "epoch": 0.1600740226694425,
      "grad_norm": 1.3209847211837769,
      "learning_rate": 0.0005,
      "loss": 3.442246913909912,
      "step": 692
    },
    {
      "epoch": 0.16030534351145037,
      "grad_norm": 1.317398190498352,
      "learning_rate": 0.0005,
      "loss": 3.5496199131011963,
      "step": 693
    },
    {
      "epoch": 0.16053666435345826,
      "grad_norm": 1.4357348680496216,
      "learning_rate": 0.0005,
      "loss": 3.3001394271850586,
      "step": 694
    },
    {
      "epoch": 0.16076798519546612,
      "grad_norm": 1.2578045129776,
      "learning_rate": 0.0005,
      "loss": 3.4422216415405273,
      "step": 695
    },
    {
      "epoch": 0.16099930603747398,
      "grad_norm": 2.5576748847961426,
      "learning_rate": 0.0005,
      "loss": 3.4497244358062744,
      "step": 696
    },
    {
      "epoch": 0.16123062687948184,
      "grad_norm": 1.6001616716384888,
      "learning_rate": 0.0005,
      "loss": 3.493496894836426,
      "step": 697
    },
    {
      "epoch": 0.1614619477214897,
      "grad_norm": 1.1585735082626343,
      "learning_rate": 0.0005,
      "loss": 3.6627087593078613,
      "step": 698
    },
    {
      "epoch": 0.16169326856349758,
      "grad_norm": 1.358856439590454,
      "learning_rate": 0.0005,
      "loss": 3.4448914527893066,
      "step": 699
    },
    {
      "epoch": 0.16192458940550544,
      "grad_norm": 1.552743673324585,
      "learning_rate": 0.0005,
      "loss": 3.76806640625,
      "step": 700
    },
    {
      "epoch": 0.1621559102475133,
      "grad_norm": 1.6100199222564697,
      "learning_rate": 0.0005,
      "loss": 3.3364861011505127,
      "step": 701
    },
    {
      "epoch": 0.16238723108952116,
      "grad_norm": 2.0127086639404297,
      "learning_rate": 0.0005,
      "loss": 3.866227626800537,
      "step": 702
    },
    {
      "epoch": 0.16261855193152902,
      "grad_norm": 1.2368874549865723,
      "learning_rate": 0.0005,
      "loss": 3.0118257999420166,
      "step": 703
    },
    {
      "epoch": 0.1628498727735369,
      "grad_norm": 1.3230034112930298,
      "learning_rate": 0.0005,
      "loss": 3.201079845428467,
      "step": 704
    },
    {
      "epoch": 0.16308119361554477,
      "grad_norm": 1.4140942096710205,
      "learning_rate": 0.0005,
      "loss": 3.328803300857544,
      "step": 705
    },
    {
      "epoch": 0.16331251445755263,
      "grad_norm": 1.4600058794021606,
      "learning_rate": 0.0005,
      "loss": 3.5567398071289062,
      "step": 706
    },
    {
      "epoch": 0.16354383529956049,
      "grad_norm": 1.147301197052002,
      "learning_rate": 0.0005,
      "loss": 3.18296480178833,
      "step": 707
    },
    {
      "epoch": 0.16377515614156835,
      "grad_norm": 1.3739628791809082,
      "learning_rate": 0.0005,
      "loss": 3.955272674560547,
      "step": 708
    },
    {
      "epoch": 0.16400647698357623,
      "grad_norm": 1.0754568576812744,
      "learning_rate": 0.0005,
      "loss": 3.228203773498535,
      "step": 709
    },
    {
      "epoch": 0.1642377978255841,
      "grad_norm": 1.3361977338790894,
      "learning_rate": 0.0005,
      "loss": 3.7236902713775635,
      "step": 710
    },
    {
      "epoch": 0.16446911866759195,
      "grad_norm": 1.120421051979065,
      "learning_rate": 0.0005,
      "loss": 3.2109322547912598,
      "step": 711
    },
    {
      "epoch": 0.1647004395095998,
      "grad_norm": 1.2812739610671997,
      "learning_rate": 0.0005,
      "loss": 3.5024476051330566,
      "step": 712
    },
    {
      "epoch": 0.16493176035160767,
      "grad_norm": 1.16416597366333,
      "learning_rate": 0.0005,
      "loss": 3.192755699157715,
      "step": 713
    },
    {
      "epoch": 0.16516308119361556,
      "grad_norm": 1.2194160223007202,
      "learning_rate": 0.0005,
      "loss": 3.473569393157959,
      "step": 714
    },
    {
      "epoch": 0.16539440203562342,
      "grad_norm": 1.2583849430084229,
      "learning_rate": 0.0005,
      "loss": 3.5146708488464355,
      "step": 715
    },
    {
      "epoch": 0.16562572287763128,
      "grad_norm": 1.275307297706604,
      "learning_rate": 0.0005,
      "loss": 3.4354379177093506,
      "step": 716
    },
    {
      "epoch": 0.16585704371963914,
      "grad_norm": 1.111336350440979,
      "learning_rate": 0.0005,
      "loss": 3.597480535507202,
      "step": 717
    },
    {
      "epoch": 0.166088364561647,
      "grad_norm": 1.525713562965393,
      "learning_rate": 0.0005,
      "loss": 3.5163893699645996,
      "step": 718
    },
    {
      "epoch": 0.16631968540365488,
      "grad_norm": 1.4592721462249756,
      "learning_rate": 0.0005,
      "loss": 3.667609930038452,
      "step": 719
    },
    {
      "epoch": 0.16655100624566274,
      "grad_norm": 1.0633280277252197,
      "learning_rate": 0.0005,
      "loss": 3.8263463973999023,
      "step": 720
    },
    {
      "epoch": 0.1667823270876706,
      "grad_norm": 1.1735336780548096,
      "learning_rate": 0.0005,
      "loss": 3.3484301567077637,
      "step": 721
    },
    {
      "epoch": 0.16701364792967846,
      "grad_norm": 1.1098783016204834,
      "learning_rate": 0.0005,
      "loss": 3.6477911472320557,
      "step": 722
    },
    {
      "epoch": 0.16724496877168632,
      "grad_norm": 1.3095604181289673,
      "learning_rate": 0.0005,
      "loss": 3.5236570835113525,
      "step": 723
    },
    {
      "epoch": 0.1674762896136942,
      "grad_norm": 1.1587885618209839,
      "learning_rate": 0.0005,
      "loss": 3.7001075744628906,
      "step": 724
    },
    {
      "epoch": 0.16770761045570207,
      "grad_norm": 1.0282412767410278,
      "learning_rate": 0.0005,
      "loss": 3.8212666511535645,
      "step": 725
    },
    {
      "epoch": 0.16793893129770993,
      "grad_norm": 1.1814604997634888,
      "learning_rate": 0.0005,
      "loss": 3.255228281021118,
      "step": 726
    },
    {
      "epoch": 0.16817025213971779,
      "grad_norm": 1.3841737508773804,
      "learning_rate": 0.0005,
      "loss": 3.2137503623962402,
      "step": 727
    },
    {
      "epoch": 0.16840157298172564,
      "grad_norm": 1.3895329236984253,
      "learning_rate": 0.0005,
      "loss": 3.302523612976074,
      "step": 728
    },
    {
      "epoch": 0.16863289382373353,
      "grad_norm": 1.2411677837371826,
      "learning_rate": 0.0005,
      "loss": 3.6438121795654297,
      "step": 729
    },
    {
      "epoch": 0.1688642146657414,
      "grad_norm": 1.4807943105697632,
      "learning_rate": 0.0005,
      "loss": 3.385467052459717,
      "step": 730
    },
    {
      "epoch": 0.16909553550774925,
      "grad_norm": 1.4539506435394287,
      "learning_rate": 0.0005,
      "loss": 3.0402908325195312,
      "step": 731
    },
    {
      "epoch": 0.1693268563497571,
      "grad_norm": 1.5664559602737427,
      "learning_rate": 0.0005,
      "loss": 3.851942539215088,
      "step": 732
    },
    {
      "epoch": 0.16955817719176497,
      "grad_norm": 1.4293497800827026,
      "learning_rate": 0.0005,
      "loss": 3.693816661834717,
      "step": 733
    },
    {
      "epoch": 0.16978949803377286,
      "grad_norm": 1.1574290990829468,
      "learning_rate": 0.0005,
      "loss": 3.125894546508789,
      "step": 734
    },
    {
      "epoch": 0.17002081887578072,
      "grad_norm": 1.309543490409851,
      "learning_rate": 0.0005,
      "loss": 3.558418035507202,
      "step": 735
    },
    {
      "epoch": 0.17025213971778858,
      "grad_norm": 1.065603494644165,
      "learning_rate": 0.0005,
      "loss": 3.2787137031555176,
      "step": 736
    },
    {
      "epoch": 0.17048346055979643,
      "grad_norm": 1.2102270126342773,
      "learning_rate": 0.0005,
      "loss": 3.0987093448638916,
      "step": 737
    },
    {
      "epoch": 0.1707147814018043,
      "grad_norm": 1.3059443235397339,
      "learning_rate": 0.0005,
      "loss": 3.4197025299072266,
      "step": 738
    },
    {
      "epoch": 0.17094610224381218,
      "grad_norm": 1.2511227130889893,
      "learning_rate": 0.0005,
      "loss": 3.5008020401000977,
      "step": 739
    },
    {
      "epoch": 0.17117742308582004,
      "grad_norm": 1.1758315563201904,
      "learning_rate": 0.0005,
      "loss": 3.324474573135376,
      "step": 740
    },
    {
      "epoch": 0.1714087439278279,
      "grad_norm": 1.6080198287963867,
      "learning_rate": 0.0005,
      "loss": 3.2113537788391113,
      "step": 741
    },
    {
      "epoch": 0.17164006476983576,
      "grad_norm": 1.4105032682418823,
      "learning_rate": 0.0005,
      "loss": 3.554224729537964,
      "step": 742
    },
    {
      "epoch": 0.17187138561184362,
      "grad_norm": 1.4235833883285522,
      "learning_rate": 0.0005,
      "loss": 3.224155902862549,
      "step": 743
    },
    {
      "epoch": 0.17210270645385148,
      "grad_norm": 1.1241179704666138,
      "learning_rate": 0.0005,
      "loss": 3.1807312965393066,
      "step": 744
    },
    {
      "epoch": 0.17233402729585937,
      "grad_norm": 1.6578682661056519,
      "learning_rate": 0.0005,
      "loss": 3.686790943145752,
      "step": 745
    },
    {
      "epoch": 0.17256534813786722,
      "grad_norm": 1.3406531810760498,
      "learning_rate": 0.0005,
      "loss": 3.7055647373199463,
      "step": 746
    },
    {
      "epoch": 0.17279666897987508,
      "grad_norm": 1.5561472177505493,
      "learning_rate": 0.0005,
      "loss": 3.243936538696289,
      "step": 747
    },
    {
      "epoch": 0.17302798982188294,
      "grad_norm": 1.1578882932662964,
      "learning_rate": 0.0005,
      "loss": 3.2880992889404297,
      "step": 748
    },
    {
      "epoch": 0.1732593106638908,
      "grad_norm": 1.2640702724456787,
      "learning_rate": 0.0005,
      "loss": 3.5301718711853027,
      "step": 749
    },
    {
      "epoch": 0.1734906315058987,
      "grad_norm": 1.6373957395553589,
      "learning_rate": 0.0005,
      "loss": 3.3081307411193848,
      "step": 750
    },
    {
      "epoch": 0.17372195234790655,
      "grad_norm": 1.4935513734817505,
      "learning_rate": 0.0005,
      "loss": 3.6595637798309326,
      "step": 751
    },
    {
      "epoch": 0.1739532731899144,
      "grad_norm": 1.7868094444274902,
      "learning_rate": 0.0005,
      "loss": 2.899893283843994,
      "step": 752
    },
    {
      "epoch": 0.17418459403192227,
      "grad_norm": 1.3438714742660522,
      "learning_rate": 0.0005,
      "loss": 3.5154685974121094,
      "step": 753
    },
    {
      "epoch": 0.17441591487393013,
      "grad_norm": 1.1582934856414795,
      "learning_rate": 0.0005,
      "loss": 3.5184192657470703,
      "step": 754
    },
    {
      "epoch": 0.17464723571593802,
      "grad_norm": 1.2084286212921143,
      "learning_rate": 0.0005,
      "loss": 3.403191566467285,
      "step": 755
    },
    {
      "epoch": 0.17487855655794587,
      "grad_norm": 1.3847509622573853,
      "learning_rate": 0.0005,
      "loss": 3.1444060802459717,
      "step": 756
    },
    {
      "epoch": 0.17510987739995373,
      "grad_norm": 1.1333286762237549,
      "learning_rate": 0.0005,
      "loss": 3.4215755462646484,
      "step": 757
    },
    {
      "epoch": 0.1753411982419616,
      "grad_norm": 1.2393234968185425,
      "learning_rate": 0.0005,
      "loss": 3.5917375087738037,
      "step": 758
    },
    {
      "epoch": 0.17557251908396945,
      "grad_norm": 1.2579638957977295,
      "learning_rate": 0.0005,
      "loss": 3.4015769958496094,
      "step": 759
    },
    {
      "epoch": 0.17580383992597734,
      "grad_norm": 1.1777219772338867,
      "learning_rate": 0.0005,
      "loss": 3.227985382080078,
      "step": 760
    },
    {
      "epoch": 0.1760351607679852,
      "grad_norm": 1.389551043510437,
      "learning_rate": 0.0005,
      "loss": 3.736858367919922,
      "step": 761
    },
    {
      "epoch": 0.17626648160999306,
      "grad_norm": 1.2407143115997314,
      "learning_rate": 0.0005,
      "loss": 3.4365389347076416,
      "step": 762
    },
    {
      "epoch": 0.17649780245200092,
      "grad_norm": 1.4558358192443848,
      "learning_rate": 0.0005,
      "loss": 3.412735939025879,
      "step": 763
    },
    {
      "epoch": 0.17672912329400878,
      "grad_norm": 1.1430675983428955,
      "learning_rate": 0.0005,
      "loss": 3.630014181137085,
      "step": 764
    },
    {
      "epoch": 0.17696044413601666,
      "grad_norm": 1.1815425157546997,
      "learning_rate": 0.0005,
      "loss": 3.2818570137023926,
      "step": 765
    },
    {
      "epoch": 0.17719176497802452,
      "grad_norm": 1.8960416316986084,
      "learning_rate": 0.0005,
      "loss": 3.148496627807617,
      "step": 766
    },
    {
      "epoch": 0.17742308582003238,
      "grad_norm": 1.4406729936599731,
      "learning_rate": 0.0005,
      "loss": 3.3847804069519043,
      "step": 767
    },
    {
      "epoch": 0.17765440666204024,
      "grad_norm": 1.2161833047866821,
      "learning_rate": 0.0005,
      "loss": 3.3453316688537598,
      "step": 768
    },
    {
      "epoch": 0.1778857275040481,
      "grad_norm": 1.5312542915344238,
      "learning_rate": 0.0005,
      "loss": 3.172633647918701,
      "step": 769
    },
    {
      "epoch": 0.178117048346056,
      "grad_norm": 1.3004848957061768,
      "learning_rate": 0.0005,
      "loss": 3.285280227661133,
      "step": 770
    },
    {
      "epoch": 0.17834836918806385,
      "grad_norm": 1.590218424797058,
      "learning_rate": 0.0005,
      "loss": 3.5869007110595703,
      "step": 771
    },
    {
      "epoch": 0.1785796900300717,
      "grad_norm": 1.3461010456085205,
      "learning_rate": 0.0005,
      "loss": 3.3785324096679688,
      "step": 772
    },
    {
      "epoch": 0.17881101087207957,
      "grad_norm": 1.3454742431640625,
      "learning_rate": 0.0005,
      "loss": 3.033769130706787,
      "step": 773
    },
    {
      "epoch": 0.17904233171408743,
      "grad_norm": 1.037401556968689,
      "learning_rate": 0.0005,
      "loss": 3.654369831085205,
      "step": 774
    },
    {
      "epoch": 0.17927365255609531,
      "grad_norm": 1.1313550472259521,
      "learning_rate": 0.0005,
      "loss": 3.5289387702941895,
      "step": 775
    },
    {
      "epoch": 0.17950497339810317,
      "grad_norm": 1.1680599451065063,
      "learning_rate": 0.0005,
      "loss": 3.3604793548583984,
      "step": 776
    },
    {
      "epoch": 0.17973629424011103,
      "grad_norm": 1.4596028327941895,
      "learning_rate": 0.0005,
      "loss": 3.683239221572876,
      "step": 777
    },
    {
      "epoch": 0.1799676150821189,
      "grad_norm": 1.1572253704071045,
      "learning_rate": 0.0005,
      "loss": 3.181474208831787,
      "step": 778
    },
    {
      "epoch": 0.18019893592412675,
      "grad_norm": 1.1261672973632812,
      "learning_rate": 0.0005,
      "loss": 3.7273221015930176,
      "step": 779
    },
    {
      "epoch": 0.18043025676613464,
      "grad_norm": 1.075016736984253,
      "learning_rate": 0.0005,
      "loss": 3.917938470840454,
      "step": 780
    },
    {
      "epoch": 0.1806615776081425,
      "grad_norm": 1.0928972959518433,
      "learning_rate": 0.0005,
      "loss": 3.4608726501464844,
      "step": 781
    },
    {
      "epoch": 0.18089289845015036,
      "grad_norm": 1.5351133346557617,
      "learning_rate": 0.0005,
      "loss": 3.3598597049713135,
      "step": 782
    },
    {
      "epoch": 0.18112421929215822,
      "grad_norm": 1.242697834968567,
      "learning_rate": 0.0005,
      "loss": 3.135894298553467,
      "step": 783
    },
    {
      "epoch": 0.18135554013416608,
      "grad_norm": 1.305405616760254,
      "learning_rate": 0.0005,
      "loss": 3.36740779876709,
      "step": 784
    },
    {
      "epoch": 0.18158686097617396,
      "grad_norm": 1.1421496868133545,
      "learning_rate": 0.0005,
      "loss": 3.2078473567962646,
      "step": 785
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 1.3980684280395508,
      "learning_rate": 0.0005,
      "loss": 3.4267501831054688,
      "step": 786
    },
    {
      "epoch": 0.18204950266018968,
      "grad_norm": 1.3856850862503052,
      "learning_rate": 0.0005,
      "loss": 3.2129626274108887,
      "step": 787
    },
    {
      "epoch": 0.18228082350219754,
      "grad_norm": 1.5438657999038696,
      "learning_rate": 0.0005,
      "loss": 3.38767147064209,
      "step": 788
    },
    {
      "epoch": 0.1825121443442054,
      "grad_norm": 1.3820769786834717,
      "learning_rate": 0.0005,
      "loss": 3.894484519958496,
      "step": 789
    },
    {
      "epoch": 0.1827434651862133,
      "grad_norm": 1.342719316482544,
      "learning_rate": 0.0005,
      "loss": 3.628049612045288,
      "step": 790
    },
    {
      "epoch": 0.18297478602822115,
      "grad_norm": 1.3734360933303833,
      "learning_rate": 0.0005,
      "loss": 3.1985104084014893,
      "step": 791
    },
    {
      "epoch": 0.183206106870229,
      "grad_norm": 1.2835229635238647,
      "learning_rate": 0.0005,
      "loss": 3.3345818519592285,
      "step": 792
    },
    {
      "epoch": 0.18343742771223687,
      "grad_norm": 1.2607874870300293,
      "learning_rate": 0.0005,
      "loss": 3.499333143234253,
      "step": 793
    },
    {
      "epoch": 0.18366874855424473,
      "grad_norm": 1.1735705137252808,
      "learning_rate": 0.0005,
      "loss": 3.5515716075897217,
      "step": 794
    },
    {
      "epoch": 0.1839000693962526,
      "grad_norm": 1.2948222160339355,
      "learning_rate": 0.0005,
      "loss": 3.210883855819702,
      "step": 795
    },
    {
      "epoch": 0.18413139023826047,
      "grad_norm": 1.193611979484558,
      "learning_rate": 0.0005,
      "loss": 3.2236146926879883,
      "step": 796
    },
    {
      "epoch": 0.18436271108026833,
      "grad_norm": 1.250426173210144,
      "learning_rate": 0.0005,
      "loss": 3.3575968742370605,
      "step": 797
    },
    {
      "epoch": 0.1845940319222762,
      "grad_norm": 1.2423840761184692,
      "learning_rate": 0.0005,
      "loss": 3.7292962074279785,
      "step": 798
    },
    {
      "epoch": 0.18482535276428405,
      "grad_norm": 1.2794723510742188,
      "learning_rate": 0.0005,
      "loss": 3.6035101413726807,
      "step": 799
    },
    {
      "epoch": 0.18505667360629194,
      "grad_norm": 2.131720542907715,
      "learning_rate": 0.0005,
      "loss": 3.310133695602417,
      "step": 800
    },
    {
      "epoch": 0.1852879944482998,
      "grad_norm": 1.526192545890808,
      "learning_rate": 0.0005,
      "loss": 3.619142532348633,
      "step": 801
    },
    {
      "epoch": 0.18551931529030766,
      "grad_norm": 1.7537083625793457,
      "learning_rate": 0.0005,
      "loss": 3.2643535137176514,
      "step": 802
    },
    {
      "epoch": 0.18575063613231552,
      "grad_norm": 1.4345309734344482,
      "learning_rate": 0.0005,
      "loss": 3.645946741104126,
      "step": 803
    },
    {
      "epoch": 0.18598195697432338,
      "grad_norm": 1.3943575620651245,
      "learning_rate": 0.0005,
      "loss": 3.6000547409057617,
      "step": 804
    },
    {
      "epoch": 0.18621327781633126,
      "grad_norm": 1.4750263690948486,
      "learning_rate": 0.0005,
      "loss": 3.402128219604492,
      "step": 805
    },
    {
      "epoch": 0.18644459865833912,
      "grad_norm": 1.2649955749511719,
      "learning_rate": 0.0005,
      "loss": 3.2903435230255127,
      "step": 806
    },
    {
      "epoch": 0.18667591950034698,
      "grad_norm": 1.2202366590499878,
      "learning_rate": 0.0005,
      "loss": 3.592350482940674,
      "step": 807
    },
    {
      "epoch": 0.18690724034235484,
      "grad_norm": 1.1613861322402954,
      "learning_rate": 0.0005,
      "loss": 3.6160309314727783,
      "step": 808
    },
    {
      "epoch": 0.1871385611843627,
      "grad_norm": 1.3190535306930542,
      "learning_rate": 0.0005,
      "loss": 3.3191475868225098,
      "step": 809
    },
    {
      "epoch": 0.1873698820263706,
      "grad_norm": 1.385624647140503,
      "learning_rate": 0.0005,
      "loss": 3.1861424446105957,
      "step": 810
    },
    {
      "epoch": 0.18760120286837845,
      "grad_norm": 1.1692125797271729,
      "learning_rate": 0.0005,
      "loss": 3.5817036628723145,
      "step": 811
    },
    {
      "epoch": 0.1878325237103863,
      "grad_norm": 1.2984545230865479,
      "learning_rate": 0.0005,
      "loss": 3.543726921081543,
      "step": 812
    },
    {
      "epoch": 0.18806384455239417,
      "grad_norm": 1.1705005168914795,
      "learning_rate": 0.0005,
      "loss": 3.2759287357330322,
      "step": 813
    },
    {
      "epoch": 0.18829516539440203,
      "grad_norm": 1.578255534172058,
      "learning_rate": 0.0005,
      "loss": 3.4842045307159424,
      "step": 814
    },
    {
      "epoch": 0.1885264862364099,
      "grad_norm": 1.5337426662445068,
      "learning_rate": 0.0005,
      "loss": 3.232771396636963,
      "step": 815
    },
    {
      "epoch": 0.18875780707841777,
      "grad_norm": 1.3476406335830688,
      "learning_rate": 0.0005,
      "loss": 2.620612621307373,
      "step": 816
    },
    {
      "epoch": 0.18898912792042563,
      "grad_norm": 3.3954107761383057,
      "learning_rate": 0.0005,
      "loss": 3.0216474533081055,
      "step": 817
    },
    {
      "epoch": 0.1892204487624335,
      "grad_norm": 1.9412192106246948,
      "learning_rate": 0.0005,
      "loss": 3.4381771087646484,
      "step": 818
    },
    {
      "epoch": 0.18945176960444135,
      "grad_norm": 1.3341830968856812,
      "learning_rate": 0.0005,
      "loss": 3.69888973236084,
      "step": 819
    },
    {
      "epoch": 0.18968309044644924,
      "grad_norm": 1.4314841032028198,
      "learning_rate": 0.0005,
      "loss": 3.6052474975585938,
      "step": 820
    },
    {
      "epoch": 0.1899144112884571,
      "grad_norm": 1.1691125631332397,
      "learning_rate": 0.0005,
      "loss": 3.426856279373169,
      "step": 821
    },
    {
      "epoch": 0.19014573213046496,
      "grad_norm": 1.4397380352020264,
      "learning_rate": 0.0005,
      "loss": 3.458357810974121,
      "step": 822
    },
    {
      "epoch": 0.19037705297247282,
      "grad_norm": 1.46232008934021,
      "learning_rate": 0.0005,
      "loss": 3.284999132156372,
      "step": 823
    },
    {
      "epoch": 0.19060837381448067,
      "grad_norm": 1.161689043045044,
      "learning_rate": 0.0005,
      "loss": 3.3167757987976074,
      "step": 824
    },
    {
      "epoch": 0.19083969465648856,
      "grad_norm": 1.2858994007110596,
      "learning_rate": 0.0005,
      "loss": 3.3888206481933594,
      "step": 825
    },
    {
      "epoch": 0.19107101549849642,
      "grad_norm": 1.1017495393753052,
      "learning_rate": 0.0005,
      "loss": 3.344081401824951,
      "step": 826
    },
    {
      "epoch": 0.19130233634050428,
      "grad_norm": 1.1122946739196777,
      "learning_rate": 0.0005,
      "loss": 2.8374812602996826,
      "step": 827
    },
    {
      "epoch": 0.19153365718251214,
      "grad_norm": 1.1255583763122559,
      "learning_rate": 0.0005,
      "loss": 3.440209150314331,
      "step": 828
    },
    {
      "epoch": 0.19176497802452,
      "grad_norm": 1.3591279983520508,
      "learning_rate": 0.0005,
      "loss": 3.919703483581543,
      "step": 829
    },
    {
      "epoch": 0.1919962988665279,
      "grad_norm": 1.3372340202331543,
      "learning_rate": 0.0005,
      "loss": 3.2685248851776123,
      "step": 830
    },
    {
      "epoch": 0.19222761970853575,
      "grad_norm": 1.5389635562896729,
      "learning_rate": 0.0005,
      "loss": 3.0019938945770264,
      "step": 831
    },
    {
      "epoch": 0.1924589405505436,
      "grad_norm": 1.9476348161697388,
      "learning_rate": 0.0005,
      "loss": 3.414271593093872,
      "step": 832
    },
    {
      "epoch": 0.19269026139255147,
      "grad_norm": 2.0158183574676514,
      "learning_rate": 0.0005,
      "loss": 3.4637885093688965,
      "step": 833
    },
    {
      "epoch": 0.19292158223455932,
      "grad_norm": 1.294486403465271,
      "learning_rate": 0.0005,
      "loss": 3.6038753986358643,
      "step": 834
    },
    {
      "epoch": 0.1931529030765672,
      "grad_norm": 1.2989473342895508,
      "learning_rate": 0.0005,
      "loss": 3.625149726867676,
      "step": 835
    },
    {
      "epoch": 0.19338422391857507,
      "grad_norm": 1.2460615634918213,
      "learning_rate": 0.0005,
      "loss": 3.307340621948242,
      "step": 836
    },
    {
      "epoch": 0.19361554476058293,
      "grad_norm": 1.5849716663360596,
      "learning_rate": 0.0005,
      "loss": 3.405618190765381,
      "step": 837
    },
    {
      "epoch": 0.1938468656025908,
      "grad_norm": 1.1990398168563843,
      "learning_rate": 0.0005,
      "loss": 3.250276565551758,
      "step": 838
    },
    {
      "epoch": 0.19407818644459865,
      "grad_norm": 1.654597282409668,
      "learning_rate": 0.0005,
      "loss": 3.852694272994995,
      "step": 839
    },
    {
      "epoch": 0.19430950728660654,
      "grad_norm": 1.3796526193618774,
      "learning_rate": 0.0005,
      "loss": 3.3903589248657227,
      "step": 840
    },
    {
      "epoch": 0.1945408281286144,
      "grad_norm": 1.3950273990631104,
      "learning_rate": 0.0005,
      "loss": 3.382687568664551,
      "step": 841
    },
    {
      "epoch": 0.19477214897062226,
      "grad_norm": 1.3520289659500122,
      "learning_rate": 0.0005,
      "loss": 3.7285759449005127,
      "step": 842
    },
    {
      "epoch": 0.19500346981263011,
      "grad_norm": 1.2525413036346436,
      "learning_rate": 0.0005,
      "loss": 3.378833055496216,
      "step": 843
    },
    {
      "epoch": 0.19523479065463797,
      "grad_norm": 1.4407323598861694,
      "learning_rate": 0.0005,
      "loss": 3.780548095703125,
      "step": 844
    },
    {
      "epoch": 0.19546611149664586,
      "grad_norm": 1.3795400857925415,
      "learning_rate": 0.0005,
      "loss": 3.447037696838379,
      "step": 845
    },
    {
      "epoch": 0.19569743233865372,
      "grad_norm": 1.2178946733474731,
      "learning_rate": 0.0005,
      "loss": 3.2323875427246094,
      "step": 846
    },
    {
      "epoch": 0.19592875318066158,
      "grad_norm": 1.056085467338562,
      "learning_rate": 0.0005,
      "loss": 3.3524856567382812,
      "step": 847
    },
    {
      "epoch": 0.19616007402266944,
      "grad_norm": 1.4840341806411743,
      "learning_rate": 0.0005,
      "loss": 3.0854012966156006,
      "step": 848
    },
    {
      "epoch": 0.1963913948646773,
      "grad_norm": 1.1161483526229858,
      "learning_rate": 0.0005,
      "loss": 3.5516281127929688,
      "step": 849
    },
    {
      "epoch": 0.19662271570668519,
      "grad_norm": 1.2920902967453003,
      "learning_rate": 0.0005,
      "loss": 3.224722146987915,
      "step": 850
    },
    {
      "epoch": 0.19685403654869305,
      "grad_norm": 1.4371728897094727,
      "learning_rate": 0.0005,
      "loss": 3.3698978424072266,
      "step": 851
    },
    {
      "epoch": 0.1970853573907009,
      "grad_norm": 2.0630788803100586,
      "learning_rate": 0.0005,
      "loss": 3.764411449432373,
      "step": 852
    },
    {
      "epoch": 0.19731667823270876,
      "grad_norm": 1.7193349599838257,
      "learning_rate": 0.0005,
      "loss": 3.6196722984313965,
      "step": 853
    },
    {
      "epoch": 0.19754799907471662,
      "grad_norm": 1.384275197982788,
      "learning_rate": 0.0005,
      "loss": 2.9770150184631348,
      "step": 854
    },
    {
      "epoch": 0.19777931991672448,
      "grad_norm": 1.3151249885559082,
      "learning_rate": 0.0005,
      "loss": 3.084902763366699,
      "step": 855
    },
    {
      "epoch": 0.19801064075873237,
      "grad_norm": 1.1006308794021606,
      "learning_rate": 0.0005,
      "loss": 3.2168045043945312,
      "step": 856
    },
    {
      "epoch": 0.19824196160074023,
      "grad_norm": 1.3362733125686646,
      "learning_rate": 0.0005,
      "loss": 3.3600497245788574,
      "step": 857
    },
    {
      "epoch": 0.1984732824427481,
      "grad_norm": 1.2707017660140991,
      "learning_rate": 0.0005,
      "loss": 3.6282501220703125,
      "step": 858
    },
    {
      "epoch": 0.19870460328475595,
      "grad_norm": 1.368942379951477,
      "learning_rate": 0.0005,
      "loss": 3.399883270263672,
      "step": 859
    },
    {
      "epoch": 0.1989359241267638,
      "grad_norm": 1.1239176988601685,
      "learning_rate": 0.0005,
      "loss": 3.1428401470184326,
      "step": 860
    },
    {
      "epoch": 0.1991672449687717,
      "grad_norm": 1.314601182937622,
      "learning_rate": 0.0005,
      "loss": 3.7059078216552734,
      "step": 861
    },
    {
      "epoch": 0.19939856581077955,
      "grad_norm": 1.152183175086975,
      "learning_rate": 0.0005,
      "loss": 3.1718900203704834,
      "step": 862
    },
    {
      "epoch": 0.1996298866527874,
      "grad_norm": 1.3714467287063599,
      "learning_rate": 0.0005,
      "loss": 3.17504620552063,
      "step": 863
    },
    {
      "epoch": 0.19986120749479527,
      "grad_norm": 1.2511255741119385,
      "learning_rate": 0.0005,
      "loss": 3.3174633979797363,
      "step": 864
    },
    {
      "epoch": 0.20009252833680313,
      "grad_norm": 1.234522819519043,
      "learning_rate": 0.0005,
      "loss": 3.3615524768829346,
      "step": 865
    },
    {
      "epoch": 0.20032384917881102,
      "grad_norm": 1.6721662282943726,
      "learning_rate": 0.0005,
      "loss": 3.386129379272461,
      "step": 866
    },
    {
      "epoch": 0.20055517002081888,
      "grad_norm": 2.0295324325561523,
      "learning_rate": 0.0005,
      "loss": 3.905327320098877,
      "step": 867
    },
    {
      "epoch": 0.20078649086282674,
      "grad_norm": 1.472389578819275,
      "learning_rate": 0.0005,
      "loss": 3.22409987449646,
      "step": 868
    },
    {
      "epoch": 0.2010178117048346,
      "grad_norm": 1.6772366762161255,
      "learning_rate": 0.0005,
      "loss": 3.428314685821533,
      "step": 869
    },
    {
      "epoch": 0.20124913254684246,
      "grad_norm": 1.2793956995010376,
      "learning_rate": 0.0005,
      "loss": 3.472787857055664,
      "step": 870
    },
    {
      "epoch": 0.20148045338885034,
      "grad_norm": 1.1329854726791382,
      "learning_rate": 0.0005,
      "loss": 3.689606189727783,
      "step": 871
    },
    {
      "epoch": 0.2017117742308582,
      "grad_norm": 1.552308201789856,
      "learning_rate": 0.0005,
      "loss": 3.1951022148132324,
      "step": 872
    },
    {
      "epoch": 0.20194309507286606,
      "grad_norm": 1.1298960447311401,
      "learning_rate": 0.0005,
      "loss": 3.659998893737793,
      "step": 873
    },
    {
      "epoch": 0.20217441591487392,
      "grad_norm": 1.068408489227295,
      "learning_rate": 0.0005,
      "loss": 3.1784534454345703,
      "step": 874
    },
    {
      "epoch": 0.20240573675688178,
      "grad_norm": 1.2858362197875977,
      "learning_rate": 0.0005,
      "loss": 3.4379515647888184,
      "step": 875
    },
    {
      "epoch": 0.20263705759888967,
      "grad_norm": 1.1372313499450684,
      "learning_rate": 0.0005,
      "loss": 3.407620429992676,
      "step": 876
    },
    {
      "epoch": 0.20286837844089753,
      "grad_norm": 1.4170178174972534,
      "learning_rate": 0.0005,
      "loss": 3.016397476196289,
      "step": 877
    },
    {
      "epoch": 0.2030996992829054,
      "grad_norm": 1.8991141319274902,
      "learning_rate": 0.0005,
      "loss": 3.223817825317383,
      "step": 878
    },
    {
      "epoch": 0.20333102012491325,
      "grad_norm": 1.2390532493591309,
      "learning_rate": 0.0005,
      "loss": 3.1611900329589844,
      "step": 879
    },
    {
      "epoch": 0.2035623409669211,
      "grad_norm": 1.783341407775879,
      "learning_rate": 0.0005,
      "loss": 3.412729263305664,
      "step": 880
    },
    {
      "epoch": 0.203793661808929,
      "grad_norm": 1.3062633275985718,
      "learning_rate": 0.0005,
      "loss": 3.624297618865967,
      "step": 881
    },
    {
      "epoch": 0.20402498265093685,
      "grad_norm": 1.2796581983566284,
      "learning_rate": 0.0005,
      "loss": 3.5028076171875,
      "step": 882
    },
    {
      "epoch": 0.2042563034929447,
      "grad_norm": 1.1775834560394287,
      "learning_rate": 0.0005,
      "loss": 3.417468786239624,
      "step": 883
    },
    {
      "epoch": 0.20448762433495257,
      "grad_norm": 1.1229547262191772,
      "learning_rate": 0.0005,
      "loss": 3.4526796340942383,
      "step": 884
    },
    {
      "epoch": 0.20471894517696043,
      "grad_norm": 1.6602874994277954,
      "learning_rate": 0.0005,
      "loss": 3.5952091217041016,
      "step": 885
    },
    {
      "epoch": 0.20495026601896832,
      "grad_norm": 1.7711807489395142,
      "learning_rate": 0.0005,
      "loss": 3.2237274646759033,
      "step": 886
    },
    {
      "epoch": 0.20518158686097618,
      "grad_norm": 1.3261595964431763,
      "learning_rate": 0.0005,
      "loss": 3.1638479232788086,
      "step": 887
    },
    {
      "epoch": 0.20541290770298404,
      "grad_norm": 1.1711515188217163,
      "learning_rate": 0.0005,
      "loss": 3.0502207279205322,
      "step": 888
    },
    {
      "epoch": 0.2056442285449919,
      "grad_norm": 1.259407639503479,
      "learning_rate": 0.0005,
      "loss": 3.3595359325408936,
      "step": 889
    },
    {
      "epoch": 0.20587554938699976,
      "grad_norm": 1.3870770931243896,
      "learning_rate": 0.0005,
      "loss": 3.386038303375244,
      "step": 890
    },
    {
      "epoch": 0.20610687022900764,
      "grad_norm": 1.2629940509796143,
      "learning_rate": 0.0005,
      "loss": 3.23954176902771,
      "step": 891
    },
    {
      "epoch": 0.2063381910710155,
      "grad_norm": 1.1016396284103394,
      "learning_rate": 0.0005,
      "loss": 3.5233750343322754,
      "step": 892
    },
    {
      "epoch": 0.20656951191302336,
      "grad_norm": 1.289259433746338,
      "learning_rate": 0.0005,
      "loss": 3.196870803833008,
      "step": 893
    },
    {
      "epoch": 0.20680083275503122,
      "grad_norm": 1.8115335702896118,
      "learning_rate": 0.0005,
      "loss": 3.900942325592041,
      "step": 894
    },
    {
      "epoch": 0.20703215359703908,
      "grad_norm": 1.1750154495239258,
      "learning_rate": 0.0005,
      "loss": 3.0363922119140625,
      "step": 895
    },
    {
      "epoch": 0.20726347443904697,
      "grad_norm": 1.4052568674087524,
      "learning_rate": 0.0005,
      "loss": 3.6268229484558105,
      "step": 896
    },
    {
      "epoch": 0.20749479528105483,
      "grad_norm": 1.4522603750228882,
      "learning_rate": 0.0005,
      "loss": 3.6102027893066406,
      "step": 897
    },
    {
      "epoch": 0.2077261161230627,
      "grad_norm": 1.2221269607543945,
      "learning_rate": 0.0005,
      "loss": 3.6311087608337402,
      "step": 898
    },
    {
      "epoch": 0.20795743696507055,
      "grad_norm": 1.3754582405090332,
      "learning_rate": 0.0005,
      "loss": 3.522120475769043,
      "step": 899
    },
    {
      "epoch": 0.2081887578070784,
      "grad_norm": 1.3548001050949097,
      "learning_rate": 0.0005,
      "loss": 3.328510284423828,
      "step": 900
    },
    {
      "epoch": 0.2084200786490863,
      "grad_norm": 1.3312686681747437,
      "learning_rate": 0.0005,
      "loss": 3.2137393951416016,
      "step": 901
    },
    {
      "epoch": 0.20865139949109415,
      "grad_norm": 1.1516289710998535,
      "learning_rate": 0.0005,
      "loss": 3.5084035396575928,
      "step": 902
    },
    {
      "epoch": 0.208882720333102,
      "grad_norm": 2.2820167541503906,
      "learning_rate": 0.0005,
      "loss": 3.9293699264526367,
      "step": 903
    },
    {
      "epoch": 0.20911404117510987,
      "grad_norm": 1.3666702508926392,
      "learning_rate": 0.0005,
      "loss": 3.675952196121216,
      "step": 904
    },
    {
      "epoch": 0.20934536201711773,
      "grad_norm": 1.1986007690429688,
      "learning_rate": 0.0005,
      "loss": 3.317936897277832,
      "step": 905
    },
    {
      "epoch": 0.20957668285912562,
      "grad_norm": 1.1170021295547485,
      "learning_rate": 0.0005,
      "loss": 3.220417022705078,
      "step": 906
    },
    {
      "epoch": 0.20980800370113348,
      "grad_norm": 1.3528143167495728,
      "learning_rate": 0.0005,
      "loss": 3.546872138977051,
      "step": 907
    },
    {
      "epoch": 0.21003932454314134,
      "grad_norm": 1.1189000606536865,
      "learning_rate": 0.0005,
      "loss": 3.5260727405548096,
      "step": 908
    },
    {
      "epoch": 0.2102706453851492,
      "grad_norm": 1.2407809495925903,
      "learning_rate": 0.0005,
      "loss": 3.1025168895721436,
      "step": 909
    },
    {
      "epoch": 0.21050196622715706,
      "grad_norm": 1.2255491018295288,
      "learning_rate": 0.0005,
      "loss": 3.4136712551116943,
      "step": 910
    },
    {
      "epoch": 0.21073328706916494,
      "grad_norm": 1.2636216878890991,
      "learning_rate": 0.0005,
      "loss": 3.1178250312805176,
      "step": 911
    },
    {
      "epoch": 0.2109646079111728,
      "grad_norm": 1.5337907075881958,
      "learning_rate": 0.0005,
      "loss": 3.4437997341156006,
      "step": 912
    },
    {
      "epoch": 0.21119592875318066,
      "grad_norm": 1.2456074953079224,
      "learning_rate": 0.0005,
      "loss": 3.145988941192627,
      "step": 913
    },
    {
      "epoch": 0.21142724959518852,
      "grad_norm": 1.1621373891830444,
      "learning_rate": 0.0005,
      "loss": 3.1035995483398438,
      "step": 914
    },
    {
      "epoch": 0.21165857043719638,
      "grad_norm": 1.4596118927001953,
      "learning_rate": 0.0005,
      "loss": 3.0277340412139893,
      "step": 915
    },
    {
      "epoch": 0.21188989127920427,
      "grad_norm": 1.223854899406433,
      "learning_rate": 0.0005,
      "loss": 3.36301326751709,
      "step": 916
    },
    {
      "epoch": 0.21212121212121213,
      "grad_norm": 1.2215181589126587,
      "learning_rate": 0.0005,
      "loss": 3.569016695022583,
      "step": 917
    },
    {
      "epoch": 0.21235253296322,
      "grad_norm": 1.306101679801941,
      "learning_rate": 0.0005,
      "loss": 3.728119373321533,
      "step": 918
    },
    {
      "epoch": 0.21258385380522785,
      "grad_norm": 1.2013169527053833,
      "learning_rate": 0.0005,
      "loss": 3.4689834117889404,
      "step": 919
    },
    {
      "epoch": 0.2128151746472357,
      "grad_norm": 1.1922584772109985,
      "learning_rate": 0.0005,
      "loss": 3.264191150665283,
      "step": 920
    },
    {
      "epoch": 0.2130464954892436,
      "grad_norm": 1.1130585670471191,
      "learning_rate": 0.0005,
      "loss": 3.3646645545959473,
      "step": 921
    },
    {
      "epoch": 0.21327781633125145,
      "grad_norm": 1.3225657939910889,
      "learning_rate": 0.0005,
      "loss": 3.238647937774658,
      "step": 922
    },
    {
      "epoch": 0.2135091371732593,
      "grad_norm": 1.6240284442901611,
      "learning_rate": 0.0005,
      "loss": 3.0571556091308594,
      "step": 923
    },
    {
      "epoch": 0.21374045801526717,
      "grad_norm": 1.4139057397842407,
      "learning_rate": 0.0005,
      "loss": 3.2060556411743164,
      "step": 924
    },
    {
      "epoch": 0.21397177885727503,
      "grad_norm": 0.9495165348052979,
      "learning_rate": 0.0005,
      "loss": 3.3778607845306396,
      "step": 925
    },
    {
      "epoch": 0.21420309969928292,
      "grad_norm": 1.1312024593353271,
      "learning_rate": 0.0005,
      "loss": 3.2091851234436035,
      "step": 926
    },
    {
      "epoch": 0.21443442054129078,
      "grad_norm": 1.1423484086990356,
      "learning_rate": 0.0005,
      "loss": 3.3779380321502686,
      "step": 927
    },
    {
      "epoch": 0.21466574138329864,
      "grad_norm": 1.1850559711456299,
      "learning_rate": 0.0005,
      "loss": 3.576171875,
      "step": 928
    },
    {
      "epoch": 0.2148970622253065,
      "grad_norm": 1.3918209075927734,
      "learning_rate": 0.0005,
      "loss": 3.299966812133789,
      "step": 929
    },
    {
      "epoch": 0.21512838306731435,
      "grad_norm": 1.5058798789978027,
      "learning_rate": 0.0005,
      "loss": 3.454754590988159,
      "step": 930
    },
    {
      "epoch": 0.21535970390932224,
      "grad_norm": 1.5802137851715088,
      "learning_rate": 0.0005,
      "loss": 3.440483570098877,
      "step": 931
    },
    {
      "epoch": 0.2155910247513301,
      "grad_norm": 1.4790949821472168,
      "learning_rate": 0.0005,
      "loss": 3.6476974487304688,
      "step": 932
    },
    {
      "epoch": 0.21582234559333796,
      "grad_norm": 1.0978835821151733,
      "learning_rate": 0.0005,
      "loss": 3.5208911895751953,
      "step": 933
    },
    {
      "epoch": 0.21605366643534582,
      "grad_norm": 1.146708369255066,
      "learning_rate": 0.0005,
      "loss": 3.3894882202148438,
      "step": 934
    },
    {
      "epoch": 0.21628498727735368,
      "grad_norm": 1.2045356035232544,
      "learning_rate": 0.0005,
      "loss": 3.037917137145996,
      "step": 935
    },
    {
      "epoch": 0.21651630811936157,
      "grad_norm": 0.9642541408538818,
      "learning_rate": 0.0005,
      "loss": 3.0946121215820312,
      "step": 936
    },
    {
      "epoch": 0.21674762896136943,
      "grad_norm": 1.1410439014434814,
      "learning_rate": 0.0005,
      "loss": 3.5319342613220215,
      "step": 937
    },
    {
      "epoch": 0.21697894980337729,
      "grad_norm": 1.1846345663070679,
      "learning_rate": 0.0005,
      "loss": 2.8672003746032715,
      "step": 938
    },
    {
      "epoch": 0.21721027064538515,
      "grad_norm": 1.2905417680740356,
      "learning_rate": 0.0005,
      "loss": 3.0111899375915527,
      "step": 939
    },
    {
      "epoch": 0.217441591487393,
      "grad_norm": 1.1636649370193481,
      "learning_rate": 0.0005,
      "loss": 3.5684561729431152,
      "step": 940
    },
    {
      "epoch": 0.2176729123294009,
      "grad_norm": 1.4054038524627686,
      "learning_rate": 0.0005,
      "loss": 3.1890220642089844,
      "step": 941
    },
    {
      "epoch": 0.21790423317140875,
      "grad_norm": 1.435172438621521,
      "learning_rate": 0.0005,
      "loss": 3.046447277069092,
      "step": 942
    },
    {
      "epoch": 0.2181355540134166,
      "grad_norm": 1.2941418886184692,
      "learning_rate": 0.0005,
      "loss": 3.3670706748962402,
      "step": 943
    },
    {
      "epoch": 0.21836687485542447,
      "grad_norm": 1.1945080757141113,
      "learning_rate": 0.0005,
      "loss": 2.9522175788879395,
      "step": 944
    },
    {
      "epoch": 0.21859819569743233,
      "grad_norm": 1.1679519414901733,
      "learning_rate": 0.0005,
      "loss": 3.2648508548736572,
      "step": 945
    },
    {
      "epoch": 0.21882951653944022,
      "grad_norm": 1.6646075248718262,
      "learning_rate": 0.0005,
      "loss": 3.7976415157318115,
      "step": 946
    },
    {
      "epoch": 0.21906083738144808,
      "grad_norm": 1.4230842590332031,
      "learning_rate": 0.0005,
      "loss": 3.6382272243499756,
      "step": 947
    },
    {
      "epoch": 0.21929215822345594,
      "grad_norm": 1.1952439546585083,
      "learning_rate": 0.0005,
      "loss": 3.120756149291992,
      "step": 948
    },
    {
      "epoch": 0.2195234790654638,
      "grad_norm": 1.420669436454773,
      "learning_rate": 0.0005,
      "loss": 3.8474764823913574,
      "step": 949
    },
    {
      "epoch": 0.21975479990747165,
      "grad_norm": 1.2769434452056885,
      "learning_rate": 0.0005,
      "loss": 3.278632640838623,
      "step": 950
    },
    {
      "epoch": 0.21998612074947954,
      "grad_norm": 1.1668264865875244,
      "learning_rate": 0.0005,
      "loss": 3.132761001586914,
      "step": 951
    },
    {
      "epoch": 0.2202174415914874,
      "grad_norm": 1.2570068836212158,
      "learning_rate": 0.0005,
      "loss": 3.2342166900634766,
      "step": 952
    },
    {
      "epoch": 0.22044876243349526,
      "grad_norm": 1.2830837965011597,
      "learning_rate": 0.0005,
      "loss": 3.015920400619507,
      "step": 953
    },
    {
      "epoch": 0.22068008327550312,
      "grad_norm": 1.0427294969558716,
      "learning_rate": 0.0005,
      "loss": 3.556793212890625,
      "step": 954
    },
    {
      "epoch": 0.22091140411751098,
      "grad_norm": 1.3168729543685913,
      "learning_rate": 0.0005,
      "loss": 3.513033866882324,
      "step": 955
    },
    {
      "epoch": 0.22114272495951887,
      "grad_norm": 1.098288893699646,
      "learning_rate": 0.0005,
      "loss": 3.7800655364990234,
      "step": 956
    },
    {
      "epoch": 0.22137404580152673,
      "grad_norm": 1.4971730709075928,
      "learning_rate": 0.0005,
      "loss": 3.5587568283081055,
      "step": 957
    },
    {
      "epoch": 0.22160536664353458,
      "grad_norm": 1.2712801694869995,
      "learning_rate": 0.0005,
      "loss": 3.2597053050994873,
      "step": 958
    },
    {
      "epoch": 0.22183668748554244,
      "grad_norm": 1.380296230316162,
      "learning_rate": 0.0005,
      "loss": 3.0989983081817627,
      "step": 959
    },
    {
      "epoch": 0.2220680083275503,
      "grad_norm": 1.4177356958389282,
      "learning_rate": 0.0005,
      "loss": 3.0983223915100098,
      "step": 960
    },
    {
      "epoch": 0.2222993291695582,
      "grad_norm": 1.3509372472763062,
      "learning_rate": 0.0005,
      "loss": 3.545799493789673,
      "step": 961
    },
    {
      "epoch": 0.22253065001156605,
      "grad_norm": 1.4530372619628906,
      "learning_rate": 0.0005,
      "loss": 3.3767001628875732,
      "step": 962
    },
    {
      "epoch": 0.2227619708535739,
      "grad_norm": 1.4533098936080933,
      "learning_rate": 0.0005,
      "loss": 3.7069666385650635,
      "step": 963
    },
    {
      "epoch": 0.22299329169558177,
      "grad_norm": 1.4964178800582886,
      "learning_rate": 0.0005,
      "loss": 3.09688663482666,
      "step": 964
    },
    {
      "epoch": 0.22322461253758963,
      "grad_norm": 4.54449987411499,
      "learning_rate": 0.0005,
      "loss": 2.9823508262634277,
      "step": 965
    },
    {
      "epoch": 0.22345593337959752,
      "grad_norm": 1.4914348125457764,
      "learning_rate": 0.0005,
      "loss": 3.3474597930908203,
      "step": 966
    },
    {
      "epoch": 0.22368725422160537,
      "grad_norm": 1.1218674182891846,
      "learning_rate": 0.0005,
      "loss": 3.5941007137298584,
      "step": 967
    },
    {
      "epoch": 0.22391857506361323,
      "grad_norm": 1.4024368524551392,
      "learning_rate": 0.0005,
      "loss": 3.1586170196533203,
      "step": 968
    },
    {
      "epoch": 0.2241498959056211,
      "grad_norm": 1.295947790145874,
      "learning_rate": 0.0005,
      "loss": 3.1727712154388428,
      "step": 969
    },
    {
      "epoch": 0.22438121674762895,
      "grad_norm": 1.5167282819747925,
      "learning_rate": 0.0005,
      "loss": 3.171097993850708,
      "step": 970
    },
    {
      "epoch": 0.2246125375896368,
      "grad_norm": 1.418994665145874,
      "learning_rate": 0.0005,
      "loss": 3.463712692260742,
      "step": 971
    },
    {
      "epoch": 0.2248438584316447,
      "grad_norm": 1.5810357332229614,
      "learning_rate": 0.0005,
      "loss": 3.3013486862182617,
      "step": 972
    },
    {
      "epoch": 0.22507517927365256,
      "grad_norm": 1.3406604528427124,
      "learning_rate": 0.0005,
      "loss": 3.2388455867767334,
      "step": 973
    },
    {
      "epoch": 0.22530650011566042,
      "grad_norm": 1.4430204629898071,
      "learning_rate": 0.0005,
      "loss": 3.0335869789123535,
      "step": 974
    },
    {
      "epoch": 0.22553782095766828,
      "grad_norm": 1.562465786933899,
      "learning_rate": 0.0005,
      "loss": 3.4881339073181152,
      "step": 975
    },
    {
      "epoch": 0.22576914179967614,
      "grad_norm": 1.0581843852996826,
      "learning_rate": 0.0005,
      "loss": 2.9465768337249756,
      "step": 976
    },
    {
      "epoch": 0.22600046264168402,
      "grad_norm": 1.7096097469329834,
      "learning_rate": 0.0005,
      "loss": 3.1206977367401123,
      "step": 977
    },
    {
      "epoch": 0.22623178348369188,
      "grad_norm": 1.403265357017517,
      "learning_rate": 0.0005,
      "loss": 3.336122751235962,
      "step": 978
    },
    {
      "epoch": 0.22646310432569974,
      "grad_norm": 1.6664992570877075,
      "learning_rate": 0.0005,
      "loss": 2.9909253120422363,
      "step": 979
    },
    {
      "epoch": 0.2266944251677076,
      "grad_norm": 1.1360193490982056,
      "learning_rate": 0.0005,
      "loss": 3.153768301010132,
      "step": 980
    },
    {
      "epoch": 0.22692574600971546,
      "grad_norm": 1.5895129442214966,
      "learning_rate": 0.0005,
      "loss": 3.705353260040283,
      "step": 981
    },
    {
      "epoch": 0.22715706685172335,
      "grad_norm": 1.4042788743972778,
      "learning_rate": 0.0005,
      "loss": 3.415715217590332,
      "step": 982
    },
    {
      "epoch": 0.2273883876937312,
      "grad_norm": 2.6202712059020996,
      "learning_rate": 0.0005,
      "loss": 2.957796335220337,
      "step": 983
    },
    {
      "epoch": 0.22761970853573907,
      "grad_norm": 1.299780011177063,
      "learning_rate": 0.0005,
      "loss": 3.184617519378662,
      "step": 984
    },
    {
      "epoch": 0.22785102937774693,
      "grad_norm": 1.4350942373275757,
      "learning_rate": 0.0005,
      "loss": 3.511678695678711,
      "step": 985
    },
    {
      "epoch": 0.2280823502197548,
      "grad_norm": 2.3756043910980225,
      "learning_rate": 0.0005,
      "loss": 3.268113374710083,
      "step": 986
    },
    {
      "epoch": 0.22831367106176267,
      "grad_norm": 1.2316503524780273,
      "learning_rate": 0.0005,
      "loss": 2.893998861312866,
      "step": 987
    },
    {
      "epoch": 0.22854499190377053,
      "grad_norm": 1.3233299255371094,
      "learning_rate": 0.0005,
      "loss": 3.4298906326293945,
      "step": 988
    },
    {
      "epoch": 0.2287763127457784,
      "grad_norm": 1.5554567575454712,
      "learning_rate": 0.0005,
      "loss": 3.3813815116882324,
      "step": 989
    },
    {
      "epoch": 0.22900763358778625,
      "grad_norm": 1.250089406967163,
      "learning_rate": 0.0005,
      "loss": 2.8828368186950684,
      "step": 990
    },
    {
      "epoch": 0.2292389544297941,
      "grad_norm": 1.5422054529190063,
      "learning_rate": 0.0005,
      "loss": 2.9843788146972656,
      "step": 991
    },
    {
      "epoch": 0.229470275271802,
      "grad_norm": 1.1035627126693726,
      "learning_rate": 0.0005,
      "loss": 3.322521924972534,
      "step": 992
    },
    {
      "epoch": 0.22970159611380986,
      "grad_norm": 1.432809591293335,
      "learning_rate": 0.0005,
      "loss": 3.6337146759033203,
      "step": 993
    },
    {
      "epoch": 0.22993291695581772,
      "grad_norm": 1.159568190574646,
      "learning_rate": 0.0005,
      "loss": 3.7335290908813477,
      "step": 994
    },
    {
      "epoch": 0.23016423779782558,
      "grad_norm": 1.4638853073120117,
      "learning_rate": 0.0005,
      "loss": 3.155982494354248,
      "step": 995
    },
    {
      "epoch": 0.23039555863983344,
      "grad_norm": 1.419293999671936,
      "learning_rate": 0.0005,
      "loss": 3.4275994300842285,
      "step": 996
    },
    {
      "epoch": 0.23062687948184132,
      "grad_norm": 1.327505111694336,
      "learning_rate": 0.0005,
      "loss": 2.9756131172180176,
      "step": 997
    },
    {
      "epoch": 0.23085820032384918,
      "grad_norm": 1.2269933223724365,
      "learning_rate": 0.0005,
      "loss": 3.4540953636169434,
      "step": 998
    },
    {
      "epoch": 0.23108952116585704,
      "grad_norm": 1.723195195198059,
      "learning_rate": 0.0005,
      "loss": 3.2427215576171875,
      "step": 999
    },
    {
      "epoch": 0.2313208420078649,
      "grad_norm": 1.1804683208465576,
      "learning_rate": 0.0005,
      "loss": 3.223611831665039,
      "step": 1000
    },
    {
      "epoch": 0.23155216284987276,
      "grad_norm": 1.2230829000473022,
      "learning_rate": 0.0005,
      "loss": 3.160740852355957,
      "step": 1001
    },
    {
      "epoch": 0.23178348369188065,
      "grad_norm": 1.374667763710022,
      "learning_rate": 0.0005,
      "loss": 3.3505516052246094,
      "step": 1002
    },
    {
      "epoch": 0.2320148045338885,
      "grad_norm": 1.4333975315093994,
      "learning_rate": 0.0005,
      "loss": 3.7704639434814453,
      "step": 1003
    },
    {
      "epoch": 0.23224612537589637,
      "grad_norm": 1.6322598457336426,
      "learning_rate": 0.0005,
      "loss": 3.640878438949585,
      "step": 1004
    },
    {
      "epoch": 0.23247744621790423,
      "grad_norm": 1.3523955345153809,
      "learning_rate": 0.0005,
      "loss": 3.3421473503112793,
      "step": 1005
    },
    {
      "epoch": 0.23270876705991209,
      "grad_norm": 1.1639788150787354,
      "learning_rate": 0.0005,
      "loss": 3.454756736755371,
      "step": 1006
    },
    {
      "epoch": 0.23294008790191997,
      "grad_norm": 1.696256160736084,
      "learning_rate": 0.0005,
      "loss": 3.6419708728790283,
      "step": 1007
    },
    {
      "epoch": 0.23317140874392783,
      "grad_norm": 1.3080165386199951,
      "learning_rate": 0.0005,
      "loss": 3.221817970275879,
      "step": 1008
    },
    {
      "epoch": 0.2334027295859357,
      "grad_norm": 1.588404655456543,
      "learning_rate": 0.0005,
      "loss": 3.3856253623962402,
      "step": 1009
    },
    {
      "epoch": 0.23363405042794355,
      "grad_norm": 1.5002243518829346,
      "learning_rate": 0.0005,
      "loss": 3.59615421295166,
      "step": 1010
    },
    {
      "epoch": 0.2338653712699514,
      "grad_norm": 1.6827869415283203,
      "learning_rate": 0.0005,
      "loss": 3.255915641784668,
      "step": 1011
    },
    {
      "epoch": 0.2340966921119593,
      "grad_norm": 1.2580323219299316,
      "learning_rate": 0.0005,
      "loss": 3.1711349487304688,
      "step": 1012
    },
    {
      "epoch": 0.23432801295396716,
      "grad_norm": 1.3122994899749756,
      "learning_rate": 0.0005,
      "loss": 3.3453450202941895,
      "step": 1013
    },
    {
      "epoch": 0.23455933379597502,
      "grad_norm": 1.5031747817993164,
      "learning_rate": 0.0005,
      "loss": 3.2626161575317383,
      "step": 1014
    },
    {
      "epoch": 0.23479065463798288,
      "grad_norm": 1.209510326385498,
      "learning_rate": 0.0005,
      "loss": 3.2654917240142822,
      "step": 1015
    },
    {
      "epoch": 0.23502197547999074,
      "grad_norm": 1.2995615005493164,
      "learning_rate": 0.0005,
      "loss": 3.5268335342407227,
      "step": 1016
    },
    {
      "epoch": 0.23525329632199862,
      "grad_norm": 1.3090695142745972,
      "learning_rate": 0.0005,
      "loss": 3.0614013671875,
      "step": 1017
    },
    {
      "epoch": 0.23548461716400648,
      "grad_norm": 1.2267520427703857,
      "learning_rate": 0.0005,
      "loss": 3.435124635696411,
      "step": 1018
    },
    {
      "epoch": 0.23571593800601434,
      "grad_norm": 1.0819153785705566,
      "learning_rate": 0.0005,
      "loss": 3.562729597091675,
      "step": 1019
    },
    {
      "epoch": 0.2359472588480222,
      "grad_norm": 1.3162013292312622,
      "learning_rate": 0.0005,
      "loss": 3.1153130531311035,
      "step": 1020
    },
    {
      "epoch": 0.23617857969003006,
      "grad_norm": 1.3530038595199585,
      "learning_rate": 0.0005,
      "loss": 3.5767126083374023,
      "step": 1021
    },
    {
      "epoch": 0.23640990053203795,
      "grad_norm": 1.1248124837875366,
      "learning_rate": 0.0005,
      "loss": 3.446686267852783,
      "step": 1022
    },
    {
      "epoch": 0.2366412213740458,
      "grad_norm": 1.5988579988479614,
      "learning_rate": 0.0005,
      "loss": 3.3966269493103027,
      "step": 1023
    },
    {
      "epoch": 0.23687254221605367,
      "grad_norm": 1.122634768486023,
      "learning_rate": 0.0005,
      "loss": 3.5069737434387207,
      "step": 1024
    },
    {
      "epoch": 0.23710386305806153,
      "grad_norm": 2.040372848510742,
      "learning_rate": 0.0005,
      "loss": 3.673769950866699,
      "step": 1025
    },
    {
      "epoch": 0.23733518390006939,
      "grad_norm": 1.2667126655578613,
      "learning_rate": 0.0005,
      "loss": 3.5998106002807617,
      "step": 1026
    },
    {
      "epoch": 0.23756650474207727,
      "grad_norm": 2.390502691268921,
      "learning_rate": 0.0005,
      "loss": 3.2567808628082275,
      "step": 1027
    },
    {
      "epoch": 0.23779782558408513,
      "grad_norm": 1.0996544361114502,
      "learning_rate": 0.0005,
      "loss": 3.304232120513916,
      "step": 1028
    },
    {
      "epoch": 0.238029146426093,
      "grad_norm": 1.2818455696105957,
      "learning_rate": 0.0005,
      "loss": 3.3601324558258057,
      "step": 1029
    },
    {
      "epoch": 0.23826046726810085,
      "grad_norm": 1.328406572341919,
      "learning_rate": 0.0005,
      "loss": 3.4427566528320312,
      "step": 1030
    },
    {
      "epoch": 0.2384917881101087,
      "grad_norm": 1.6050410270690918,
      "learning_rate": 0.0005,
      "loss": 3.3854212760925293,
      "step": 1031
    },
    {
      "epoch": 0.2387231089521166,
      "grad_norm": 1.5324689149856567,
      "learning_rate": 0.0005,
      "loss": 3.4736790657043457,
      "step": 1032
    },
    {
      "epoch": 0.23895442979412446,
      "grad_norm": 1.3972532749176025,
      "learning_rate": 0.0005,
      "loss": 3.686945915222168,
      "step": 1033
    },
    {
      "epoch": 0.23918575063613232,
      "grad_norm": 1.5567893981933594,
      "learning_rate": 0.0005,
      "loss": 3.0781610012054443,
      "step": 1034
    },
    {
      "epoch": 0.23941707147814018,
      "grad_norm": 1.3917628526687622,
      "learning_rate": 0.0005,
      "loss": 3.2802226543426514,
      "step": 1035
    },
    {
      "epoch": 0.23964839232014803,
      "grad_norm": 1.5110630989074707,
      "learning_rate": 0.0005,
      "loss": 3.546571731567383,
      "step": 1036
    },
    {
      "epoch": 0.23987971316215592,
      "grad_norm": 1.4042903184890747,
      "learning_rate": 0.0005,
      "loss": 3.0510406494140625,
      "step": 1037
    },
    {
      "epoch": 0.24011103400416378,
      "grad_norm": 1.3268613815307617,
      "learning_rate": 0.0005,
      "loss": 3.2640788555145264,
      "step": 1038
    },
    {
      "epoch": 0.24034235484617164,
      "grad_norm": 1.3985273838043213,
      "learning_rate": 0.0005,
      "loss": 3.3427205085754395,
      "step": 1039
    },
    {
      "epoch": 0.2405736756881795,
      "grad_norm": 1.3996264934539795,
      "learning_rate": 0.0005,
      "loss": 3.144099473953247,
      "step": 1040
    },
    {
      "epoch": 0.24080499653018736,
      "grad_norm": 1.1033655405044556,
      "learning_rate": 0.0005,
      "loss": 2.9496498107910156,
      "step": 1041
    },
    {
      "epoch": 0.24103631737219525,
      "grad_norm": 1.1801854372024536,
      "learning_rate": 0.0005,
      "loss": 3.672818183898926,
      "step": 1042
    },
    {
      "epoch": 0.2412676382142031,
      "grad_norm": 1.3128795623779297,
      "learning_rate": 0.0005,
      "loss": 3.0398330688476562,
      "step": 1043
    },
    {
      "epoch": 0.24149895905621097,
      "grad_norm": 1.3259706497192383,
      "learning_rate": 0.0005,
      "loss": 3.444486618041992,
      "step": 1044
    },
    {
      "epoch": 0.24173027989821882,
      "grad_norm": 1.0796968936920166,
      "learning_rate": 0.0005,
      "loss": 3.0411643981933594,
      "step": 1045
    },
    {
      "epoch": 0.24196160074022668,
      "grad_norm": 1.0318078994750977,
      "learning_rate": 0.0005,
      "loss": 3.3738083839416504,
      "step": 1046
    },
    {
      "epoch": 0.24219292158223457,
      "grad_norm": 1.1523339748382568,
      "learning_rate": 0.0005,
      "loss": 3.040283203125,
      "step": 1047
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 1.1596485376358032,
      "learning_rate": 0.0005,
      "loss": 3.390622854232788,
      "step": 1048
    },
    {
      "epoch": 0.2426555632662503,
      "grad_norm": 1.1250061988830566,
      "learning_rate": 0.0005,
      "loss": 3.5730087757110596,
      "step": 1049
    },
    {
      "epoch": 0.24288688410825815,
      "grad_norm": 1.1169986724853516,
      "learning_rate": 0.0005,
      "loss": 3.3942668437957764,
      "step": 1050
    },
    {
      "epoch": 0.243118204950266,
      "grad_norm": 1.1464307308197021,
      "learning_rate": 0.0005,
      "loss": 3.3807363510131836,
      "step": 1051
    },
    {
      "epoch": 0.2433495257922739,
      "grad_norm": 1.0295753479003906,
      "learning_rate": 0.0005,
      "loss": 3.2072248458862305,
      "step": 1052
    },
    {
      "epoch": 0.24358084663428176,
      "grad_norm": 1.5810514688491821,
      "learning_rate": 0.0005,
      "loss": 3.2554967403411865,
      "step": 1053
    },
    {
      "epoch": 0.24381216747628962,
      "grad_norm": 1.089160442352295,
      "learning_rate": 0.0005,
      "loss": 3.235234498977661,
      "step": 1054
    },
    {
      "epoch": 0.24404348831829747,
      "grad_norm": 1.2244008779525757,
      "learning_rate": 0.0005,
      "loss": 3.156797409057617,
      "step": 1055
    },
    {
      "epoch": 0.24427480916030533,
      "grad_norm": 1.1935198307037354,
      "learning_rate": 0.0005,
      "loss": 3.4302873611450195,
      "step": 1056
    },
    {
      "epoch": 0.24450613000231322,
      "grad_norm": 1.5273524522781372,
      "learning_rate": 0.0005,
      "loss": 3.261009693145752,
      "step": 1057
    },
    {
      "epoch": 0.24473745084432108,
      "grad_norm": 1.2646878957748413,
      "learning_rate": 0.0005,
      "loss": 3.5698654651641846,
      "step": 1058
    },
    {
      "epoch": 0.24496877168632894,
      "grad_norm": 1.408781886100769,
      "learning_rate": 0.0005,
      "loss": 3.2740650177001953,
      "step": 1059
    },
    {
      "epoch": 0.2452000925283368,
      "grad_norm": 1.261549949645996,
      "learning_rate": 0.0005,
      "loss": 3.2148635387420654,
      "step": 1060
    },
    {
      "epoch": 0.24543141337034466,
      "grad_norm": 1.2165207862854004,
      "learning_rate": 0.0005,
      "loss": 2.9786276817321777,
      "step": 1061
    },
    {
      "epoch": 0.24566273421235255,
      "grad_norm": 1.2735512256622314,
      "learning_rate": 0.0005,
      "loss": 3.6687374114990234,
      "step": 1062
    },
    {
      "epoch": 0.2458940550543604,
      "grad_norm": 2.3268723487854004,
      "learning_rate": 0.0005,
      "loss": 3.215820789337158,
      "step": 1063
    },
    {
      "epoch": 0.24612537589636826,
      "grad_norm": 1.1616835594177246,
      "learning_rate": 0.0005,
      "loss": 3.5554614067077637,
      "step": 1064
    },
    {
      "epoch": 0.24635669673837612,
      "grad_norm": 1.501289963722229,
      "learning_rate": 0.0005,
      "loss": 3.49428391456604,
      "step": 1065
    },
    {
      "epoch": 0.24658801758038398,
      "grad_norm": 1.1716816425323486,
      "learning_rate": 0.0005,
      "loss": 3.323643207550049,
      "step": 1066
    },
    {
      "epoch": 0.24681933842239187,
      "grad_norm": 1.063179850578308,
      "learning_rate": 0.0005,
      "loss": 3.174877166748047,
      "step": 1067
    },
    {
      "epoch": 0.24705065926439973,
      "grad_norm": 1.1471999883651733,
      "learning_rate": 0.0005,
      "loss": 3.340867519378662,
      "step": 1068
    },
    {
      "epoch": 0.2472819801064076,
      "grad_norm": 1.1026153564453125,
      "learning_rate": 0.0005,
      "loss": 3.619248390197754,
      "step": 1069
    },
    {
      "epoch": 0.24751330094841545,
      "grad_norm": 1.358199119567871,
      "learning_rate": 0.0005,
      "loss": 3.288219690322876,
      "step": 1070
    },
    {
      "epoch": 0.2477446217904233,
      "grad_norm": 1.4378224611282349,
      "learning_rate": 0.0005,
      "loss": 3.3010122776031494,
      "step": 1071
    },
    {
      "epoch": 0.2479759426324312,
      "grad_norm": 1.4097236394882202,
      "learning_rate": 0.0005,
      "loss": 3.6392874717712402,
      "step": 1072
    },
    {
      "epoch": 0.24820726347443905,
      "grad_norm": 1.4374499320983887,
      "learning_rate": 0.0005,
      "loss": 3.0332090854644775,
      "step": 1073
    },
    {
      "epoch": 0.24843858431644691,
      "grad_norm": 1.0834834575653076,
      "learning_rate": 0.0005,
      "loss": 3.3641958236694336,
      "step": 1074
    },
    {
      "epoch": 0.24866990515845477,
      "grad_norm": 1.2454856634140015,
      "learning_rate": 0.0005,
      "loss": 3.659663200378418,
      "step": 1075
    },
    {
      "epoch": 0.24890122600046263,
      "grad_norm": 1.441872239112854,
      "learning_rate": 0.0005,
      "loss": 3.4227678775787354,
      "step": 1076
    },
    {
      "epoch": 0.24913254684247052,
      "grad_norm": 1.4231065511703491,
      "learning_rate": 0.0005,
      "loss": 3.1695427894592285,
      "step": 1077
    },
    {
      "epoch": 0.24936386768447838,
      "grad_norm": 1.2393888235092163,
      "learning_rate": 0.0005,
      "loss": 3.413100481033325,
      "step": 1078
    },
    {
      "epoch": 0.24959518852648624,
      "grad_norm": 1.4108562469482422,
      "learning_rate": 0.0005,
      "loss": 3.3671507835388184,
      "step": 1079
    },
    {
      "epoch": 0.2498265093684941,
      "grad_norm": 1.205104947090149,
      "learning_rate": 0.0005,
      "loss": 3.3907470703125,
      "step": 1080
    },
    {
      "epoch": 0.250057830210502,
      "grad_norm": 1.266671895980835,
      "learning_rate": 0.0005,
      "loss": 3.3375322818756104,
      "step": 1081
    },
    {
      "epoch": 0.2502891510525098,
      "grad_norm": 1.520835041999817,
      "learning_rate": 0.0005,
      "loss": 3.2818117141723633,
      "step": 1082
    },
    {
      "epoch": 0.2505204718945177,
      "grad_norm": 1.3141911029815674,
      "learning_rate": 0.0005,
      "loss": 3.1943955421447754,
      "step": 1083
    },
    {
      "epoch": 0.25075179273652554,
      "grad_norm": 1.3752992153167725,
      "learning_rate": 0.0005,
      "loss": 3.3881561756134033,
      "step": 1084
    },
    {
      "epoch": 0.2509831135785334,
      "grad_norm": 1.170278549194336,
      "learning_rate": 0.0005,
      "loss": 3.1157970428466797,
      "step": 1085
    },
    {
      "epoch": 0.2512144344205413,
      "grad_norm": 1.3096591234207153,
      "learning_rate": 0.0005,
      "loss": 3.0911359786987305,
      "step": 1086
    },
    {
      "epoch": 0.25144575526254914,
      "grad_norm": 1.2331955432891846,
      "learning_rate": 0.0005,
      "loss": 2.870262384414673,
      "step": 1087
    },
    {
      "epoch": 0.25167707610455703,
      "grad_norm": 1.2878189086914062,
      "learning_rate": 0.0005,
      "loss": 3.465139865875244,
      "step": 1088
    },
    {
      "epoch": 0.25190839694656486,
      "grad_norm": 1.1536730527877808,
      "learning_rate": 0.0005,
      "loss": 3.310532569885254,
      "step": 1089
    },
    {
      "epoch": 0.25213971778857275,
      "grad_norm": 1.2876604795455933,
      "learning_rate": 0.0005,
      "loss": 3.130941152572632,
      "step": 1090
    },
    {
      "epoch": 0.25237103863058064,
      "grad_norm": 1.1830722093582153,
      "learning_rate": 0.0005,
      "loss": 3.1009557247161865,
      "step": 1091
    },
    {
      "epoch": 0.25260235947258847,
      "grad_norm": 1.3057347536087036,
      "learning_rate": 0.0005,
      "loss": 3.2172389030456543,
      "step": 1092
    },
    {
      "epoch": 0.25283368031459635,
      "grad_norm": 1.194425106048584,
      "learning_rate": 0.0005,
      "loss": 3.3120436668395996,
      "step": 1093
    },
    {
      "epoch": 0.2530650011566042,
      "grad_norm": 1.5403364896774292,
      "learning_rate": 0.0005,
      "loss": 3.1118311882019043,
      "step": 1094
    },
    {
      "epoch": 0.2532963219986121,
      "grad_norm": 1.5964767932891846,
      "learning_rate": 0.0005,
      "loss": 3.2364768981933594,
      "step": 1095
    },
    {
      "epoch": 0.25352764284061996,
      "grad_norm": 1.2292777299880981,
      "learning_rate": 0.0005,
      "loss": 3.218844413757324,
      "step": 1096
    },
    {
      "epoch": 0.2537589636826278,
      "grad_norm": 1.2295719385147095,
      "learning_rate": 0.0005,
      "loss": 3.163473129272461,
      "step": 1097
    },
    {
      "epoch": 0.2539902845246357,
      "grad_norm": 1.166982889175415,
      "learning_rate": 0.0005,
      "loss": 2.88836932182312,
      "step": 1098
    },
    {
      "epoch": 0.2542216053666435,
      "grad_norm": 1.3220655918121338,
      "learning_rate": 0.0005,
      "loss": 3.2530102729797363,
      "step": 1099
    },
    {
      "epoch": 0.2544529262086514,
      "grad_norm": 1.260011076927185,
      "learning_rate": 0.0005,
      "loss": 2.950857639312744,
      "step": 1100
    },
    {
      "epoch": 0.2546842470506593,
      "grad_norm": 1.1157692670822144,
      "learning_rate": 0.0005,
      "loss": 3.3241512775421143,
      "step": 1101
    },
    {
      "epoch": 0.2549155678926671,
      "grad_norm": 1.4173682928085327,
      "learning_rate": 0.0005,
      "loss": 3.129268169403076,
      "step": 1102
    },
    {
      "epoch": 0.255146888734675,
      "grad_norm": 1.3278262615203857,
      "learning_rate": 0.0005,
      "loss": 3.020172357559204,
      "step": 1103
    },
    {
      "epoch": 0.25537820957668284,
      "grad_norm": 1.2040352821350098,
      "learning_rate": 0.0005,
      "loss": 3.021602153778076,
      "step": 1104
    },
    {
      "epoch": 0.2556095304186907,
      "grad_norm": 1.9509323835372925,
      "learning_rate": 0.0005,
      "loss": 3.2963409423828125,
      "step": 1105
    },
    {
      "epoch": 0.2558408512606986,
      "grad_norm": 1.1637258529663086,
      "learning_rate": 0.0005,
      "loss": 3.40897274017334,
      "step": 1106
    },
    {
      "epoch": 0.25607217210270644,
      "grad_norm": 1.5434391498565674,
      "learning_rate": 0.0005,
      "loss": 2.9815049171447754,
      "step": 1107
    },
    {
      "epoch": 0.25630349294471433,
      "grad_norm": 1.1721303462982178,
      "learning_rate": 0.0005,
      "loss": 3.2520546913146973,
      "step": 1108
    },
    {
      "epoch": 0.25653481378672216,
      "grad_norm": 1.1184732913970947,
      "learning_rate": 0.0005,
      "loss": 3.0944581031799316,
      "step": 1109
    },
    {
      "epoch": 0.25676613462873005,
      "grad_norm": 1.629431128501892,
      "learning_rate": 0.0005,
      "loss": 3.084507942199707,
      "step": 1110
    },
    {
      "epoch": 0.25699745547073793,
      "grad_norm": 1.456087589263916,
      "learning_rate": 0.0005,
      "loss": 3.356407642364502,
      "step": 1111
    },
    {
      "epoch": 0.25722877631274577,
      "grad_norm": 1.2042880058288574,
      "learning_rate": 0.0005,
      "loss": 3.4633007049560547,
      "step": 1112
    },
    {
      "epoch": 0.25746009715475365,
      "grad_norm": 1.3230944871902466,
      "learning_rate": 0.0005,
      "loss": 3.377261161804199,
      "step": 1113
    },
    {
      "epoch": 0.2576914179967615,
      "grad_norm": 1.2492225170135498,
      "learning_rate": 0.0005,
      "loss": 3.1922945976257324,
      "step": 1114
    },
    {
      "epoch": 0.25792273883876937,
      "grad_norm": 1.2920986413955688,
      "learning_rate": 0.0005,
      "loss": 3.290375232696533,
      "step": 1115
    },
    {
      "epoch": 0.25815405968077726,
      "grad_norm": 1.2324081659317017,
      "learning_rate": 0.0005,
      "loss": 3.233351469039917,
      "step": 1116
    },
    {
      "epoch": 0.2583853805227851,
      "grad_norm": 1.08186674118042,
      "learning_rate": 0.0005,
      "loss": 3.6959192752838135,
      "step": 1117
    },
    {
      "epoch": 0.258616701364793,
      "grad_norm": 1.3842319250106812,
      "learning_rate": 0.0005,
      "loss": 3.459743022918701,
      "step": 1118
    },
    {
      "epoch": 0.2588480222068008,
      "grad_norm": 1.4507421255111694,
      "learning_rate": 0.0005,
      "loss": 3.393010377883911,
      "step": 1119
    },
    {
      "epoch": 0.2590793430488087,
      "grad_norm": 1.3291294574737549,
      "learning_rate": 0.0005,
      "loss": 3.1871018409729004,
      "step": 1120
    },
    {
      "epoch": 0.2593106638908166,
      "grad_norm": 1.3129692077636719,
      "learning_rate": 0.0005,
      "loss": 3.0611836910247803,
      "step": 1121
    },
    {
      "epoch": 0.2595419847328244,
      "grad_norm": 1.2324373722076416,
      "learning_rate": 0.0005,
      "loss": 3.180548906326294,
      "step": 1122
    },
    {
      "epoch": 0.2597733055748323,
      "grad_norm": 1.1789973974227905,
      "learning_rate": 0.0005,
      "loss": 3.031808376312256,
      "step": 1123
    },
    {
      "epoch": 0.26000462641684013,
      "grad_norm": 1.4971661567687988,
      "learning_rate": 0.0005,
      "loss": 3.2013516426086426,
      "step": 1124
    },
    {
      "epoch": 0.260235947258848,
      "grad_norm": 1.505798578262329,
      "learning_rate": 0.0005,
      "loss": 3.2836320400238037,
      "step": 1125
    },
    {
      "epoch": 0.2604672681008559,
      "grad_norm": 1.3211143016815186,
      "learning_rate": 0.0005,
      "loss": 2.9679226875305176,
      "step": 1126
    },
    {
      "epoch": 0.26069858894286374,
      "grad_norm": 1.1329344511032104,
      "learning_rate": 0.0005,
      "loss": 2.7298896312713623,
      "step": 1127
    },
    {
      "epoch": 0.2609299097848716,
      "grad_norm": 1.3920300006866455,
      "learning_rate": 0.0005,
      "loss": 3.0924320220947266,
      "step": 1128
    },
    {
      "epoch": 0.26116123062687946,
      "grad_norm": 1.6336477994918823,
      "learning_rate": 0.0005,
      "loss": 2.964970111846924,
      "step": 1129
    },
    {
      "epoch": 0.26139255146888735,
      "grad_norm": 1.1115695238113403,
      "learning_rate": 0.0005,
      "loss": 3.438538074493408,
      "step": 1130
    },
    {
      "epoch": 0.26162387231089523,
      "grad_norm": 1.1277506351470947,
      "learning_rate": 0.0005,
      "loss": 3.0640065670013428,
      "step": 1131
    },
    {
      "epoch": 0.26185519315290307,
      "grad_norm": 1.4297677278518677,
      "learning_rate": 0.0005,
      "loss": 3.0981268882751465,
      "step": 1132
    },
    {
      "epoch": 0.26208651399491095,
      "grad_norm": 1.1232571601867676,
      "learning_rate": 0.0005,
      "loss": 3.611448287963867,
      "step": 1133
    },
    {
      "epoch": 0.2623178348369188,
      "grad_norm": 1.1053366661071777,
      "learning_rate": 0.0005,
      "loss": 2.9159655570983887,
      "step": 1134
    },
    {
      "epoch": 0.26254915567892667,
      "grad_norm": 1.2446787357330322,
      "learning_rate": 0.0005,
      "loss": 3.495081901550293,
      "step": 1135
    },
    {
      "epoch": 0.26278047652093456,
      "grad_norm": 1.357216238975525,
      "learning_rate": 0.0005,
      "loss": 3.335801601409912,
      "step": 1136
    },
    {
      "epoch": 0.2630117973629424,
      "grad_norm": 1.6219544410705566,
      "learning_rate": 0.0005,
      "loss": 3.4432027339935303,
      "step": 1137
    },
    {
      "epoch": 0.2632431182049503,
      "grad_norm": 1.0259383916854858,
      "learning_rate": 0.0005,
      "loss": 2.9671242237091064,
      "step": 1138
    },
    {
      "epoch": 0.2634744390469581,
      "grad_norm": 1.0589148998260498,
      "learning_rate": 0.0005,
      "loss": 3.184131622314453,
      "step": 1139
    },
    {
      "epoch": 0.263705759888966,
      "grad_norm": 1.1374921798706055,
      "learning_rate": 0.0005,
      "loss": 3.3556947708129883,
      "step": 1140
    },
    {
      "epoch": 0.2639370807309739,
      "grad_norm": 1.179592490196228,
      "learning_rate": 0.0005,
      "loss": 3.1362171173095703,
      "step": 1141
    },
    {
      "epoch": 0.2641684015729817,
      "grad_norm": 1.2533694505691528,
      "learning_rate": 0.0005,
      "loss": 3.313783645629883,
      "step": 1142
    },
    {
      "epoch": 0.2643997224149896,
      "grad_norm": 1.16860032081604,
      "learning_rate": 0.0005,
      "loss": 3.47163724899292,
      "step": 1143
    },
    {
      "epoch": 0.26463104325699743,
      "grad_norm": 1.2603760957717896,
      "learning_rate": 0.0005,
      "loss": 3.196251630783081,
      "step": 1144
    },
    {
      "epoch": 0.2648623640990053,
      "grad_norm": 1.16036856174469,
      "learning_rate": 0.0005,
      "loss": 3.1452078819274902,
      "step": 1145
    },
    {
      "epoch": 0.2650936849410132,
      "grad_norm": 1.0490589141845703,
      "learning_rate": 0.0005,
      "loss": 3.1326422691345215,
      "step": 1146
    },
    {
      "epoch": 0.26532500578302104,
      "grad_norm": 1.4013564586639404,
      "learning_rate": 0.0005,
      "loss": 3.3535778522491455,
      "step": 1147
    },
    {
      "epoch": 0.2655563266250289,
      "grad_norm": 1.4892737865447998,
      "learning_rate": 0.0005,
      "loss": 3.397254467010498,
      "step": 1148
    },
    {
      "epoch": 0.26578764746703676,
      "grad_norm": 1.4116557836532593,
      "learning_rate": 0.0005,
      "loss": 3.275075912475586,
      "step": 1149
    },
    {
      "epoch": 0.26601896830904465,
      "grad_norm": 1.2690966129302979,
      "learning_rate": 0.0005,
      "loss": 3.327827215194702,
      "step": 1150
    },
    {
      "epoch": 0.26625028915105253,
      "grad_norm": 1.3405916690826416,
      "learning_rate": 0.0005,
      "loss": 3.4591064453125,
      "step": 1151
    },
    {
      "epoch": 0.26648160999306036,
      "grad_norm": 1.2562642097473145,
      "learning_rate": 0.0005,
      "loss": 3.2578084468841553,
      "step": 1152
    },
    {
      "epoch": 0.26671293083506825,
      "grad_norm": 1.5289608240127563,
      "learning_rate": 0.0005,
      "loss": 3.257894992828369,
      "step": 1153
    },
    {
      "epoch": 0.2669442516770761,
      "grad_norm": 5.584132671356201,
      "learning_rate": 0.0005,
      "loss": 3.0501797199249268,
      "step": 1154
    },
    {
      "epoch": 0.26717557251908397,
      "grad_norm": 1.4138319492340088,
      "learning_rate": 0.0005,
      "loss": 3.1492748260498047,
      "step": 1155
    },
    {
      "epoch": 0.26740689336109186,
      "grad_norm": 1.2322098016738892,
      "learning_rate": 0.0005,
      "loss": 3.230677604675293,
      "step": 1156
    },
    {
      "epoch": 0.2676382142030997,
      "grad_norm": 1.4000775814056396,
      "learning_rate": 0.0005,
      "loss": 2.9374215602874756,
      "step": 1157
    },
    {
      "epoch": 0.2678695350451076,
      "grad_norm": 1.5662838220596313,
      "learning_rate": 0.0005,
      "loss": 3.3929085731506348,
      "step": 1158
    },
    {
      "epoch": 0.2681008558871154,
      "grad_norm": 1.3114527463912964,
      "learning_rate": 0.0005,
      "loss": 3.2573633193969727,
      "step": 1159
    },
    {
      "epoch": 0.2683321767291233,
      "grad_norm": 1.1080505847930908,
      "learning_rate": 0.0005,
      "loss": 2.9537835121154785,
      "step": 1160
    },
    {
      "epoch": 0.2685634975711312,
      "grad_norm": 1.204163670539856,
      "learning_rate": 0.0005,
      "loss": 3.4099509716033936,
      "step": 1161
    },
    {
      "epoch": 0.268794818413139,
      "grad_norm": 1.5695432424545288,
      "learning_rate": 0.0005,
      "loss": 3.28932785987854,
      "step": 1162
    },
    {
      "epoch": 0.2690261392551469,
      "grad_norm": 1.026561975479126,
      "learning_rate": 0.0005,
      "loss": 3.178497791290283,
      "step": 1163
    },
    {
      "epoch": 0.26925746009715473,
      "grad_norm": 1.2063148021697998,
      "learning_rate": 0.0005,
      "loss": 3.4924428462982178,
      "step": 1164
    },
    {
      "epoch": 0.2694887809391626,
      "grad_norm": 1.47279691696167,
      "learning_rate": 0.0005,
      "loss": 3.2789535522460938,
      "step": 1165
    },
    {
      "epoch": 0.2697201017811705,
      "grad_norm": 1.7653790712356567,
      "learning_rate": 0.0005,
      "loss": 3.2730159759521484,
      "step": 1166
    },
    {
      "epoch": 0.26995142262317834,
      "grad_norm": 1.3231468200683594,
      "learning_rate": 0.0005,
      "loss": 3.107302665710449,
      "step": 1167
    },
    {
      "epoch": 0.2701827434651862,
      "grad_norm": 1.2233226299285889,
      "learning_rate": 0.0005,
      "loss": 3.2707042694091797,
      "step": 1168
    },
    {
      "epoch": 0.27041406430719406,
      "grad_norm": 1.1582120656967163,
      "learning_rate": 0.0005,
      "loss": 3.223538875579834,
      "step": 1169
    },
    {
      "epoch": 0.27064538514920194,
      "grad_norm": 1.3475514650344849,
      "learning_rate": 0.0005,
      "loss": 3.3167285919189453,
      "step": 1170
    },
    {
      "epoch": 0.27087670599120983,
      "grad_norm": 1.2003644704818726,
      "learning_rate": 0.0005,
      "loss": 3.2347450256347656,
      "step": 1171
    },
    {
      "epoch": 0.27110802683321766,
      "grad_norm": 1.2234582901000977,
      "learning_rate": 0.0005,
      "loss": 3.0357632637023926,
      "step": 1172
    },
    {
      "epoch": 0.27133934767522555,
      "grad_norm": 1.4838486909866333,
      "learning_rate": 0.0005,
      "loss": 3.1061744689941406,
      "step": 1173
    },
    {
      "epoch": 0.2715706685172334,
      "grad_norm": 1.3180146217346191,
      "learning_rate": 0.0005,
      "loss": 3.0227081775665283,
      "step": 1174
    },
    {
      "epoch": 0.27180198935924127,
      "grad_norm": 1.387650966644287,
      "learning_rate": 0.0005,
      "loss": 3.439058780670166,
      "step": 1175
    },
    {
      "epoch": 0.27203331020124916,
      "grad_norm": 1.9692022800445557,
      "learning_rate": 0.0005,
      "loss": 2.898707866668701,
      "step": 1176
    },
    {
      "epoch": 0.272264631043257,
      "grad_norm": 1.4023510217666626,
      "learning_rate": 0.0005,
      "loss": 3.307377815246582,
      "step": 1177
    },
    {
      "epoch": 0.2724959518852649,
      "grad_norm": 1.139144778251648,
      "learning_rate": 0.0005,
      "loss": 3.387586832046509,
      "step": 1178
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 1.1030267477035522,
      "learning_rate": 0.0005,
      "loss": 3.3409533500671387,
      "step": 1179
    },
    {
      "epoch": 0.2729585935692806,
      "grad_norm": 1.0706242322921753,
      "learning_rate": 0.0005,
      "loss": 3.353806972503662,
      "step": 1180
    },
    {
      "epoch": 0.2731899144112885,
      "grad_norm": 1.1412800550460815,
      "learning_rate": 0.0005,
      "loss": 3.229707717895508,
      "step": 1181
    },
    {
      "epoch": 0.2734212352532963,
      "grad_norm": 1.069780945777893,
      "learning_rate": 0.0005,
      "loss": 3.271146297454834,
      "step": 1182
    },
    {
      "epoch": 0.2736525560953042,
      "grad_norm": 1.5992286205291748,
      "learning_rate": 0.0005,
      "loss": 3.2106873989105225,
      "step": 1183
    },
    {
      "epoch": 0.27388387693731203,
      "grad_norm": 1.5980218648910522,
      "learning_rate": 0.0005,
      "loss": 2.6626977920532227,
      "step": 1184
    },
    {
      "epoch": 0.2741151977793199,
      "grad_norm": 1.0235400199890137,
      "learning_rate": 0.0005,
      "loss": 2.8614509105682373,
      "step": 1185
    },
    {
      "epoch": 0.2743465186213278,
      "grad_norm": 1.2235606908798218,
      "learning_rate": 0.0005,
      "loss": 3.320174217224121,
      "step": 1186
    },
    {
      "epoch": 0.27457783946333564,
      "grad_norm": 1.0804893970489502,
      "learning_rate": 0.0005,
      "loss": 3.0683374404907227,
      "step": 1187
    },
    {
      "epoch": 0.2748091603053435,
      "grad_norm": 1.0841532945632935,
      "learning_rate": 0.0005,
      "loss": 3.246157169342041,
      "step": 1188
    },
    {
      "epoch": 0.27504048114735136,
      "grad_norm": 1.3259845972061157,
      "learning_rate": 0.0005,
      "loss": 2.8285179138183594,
      "step": 1189
    },
    {
      "epoch": 0.27527180198935924,
      "grad_norm": 1.519797682762146,
      "learning_rate": 0.0005,
      "loss": 3.435075044631958,
      "step": 1190
    },
    {
      "epoch": 0.27550312283136713,
      "grad_norm": 1.2391003370285034,
      "learning_rate": 0.0005,
      "loss": 3.180449962615967,
      "step": 1191
    },
    {
      "epoch": 0.27573444367337496,
      "grad_norm": 1.018194556236267,
      "learning_rate": 0.0005,
      "loss": 3.437168598175049,
      "step": 1192
    },
    {
      "epoch": 0.27596576451538285,
      "grad_norm": 1.239357352256775,
      "learning_rate": 0.0005,
      "loss": 3.2565317153930664,
      "step": 1193
    },
    {
      "epoch": 0.2761970853573907,
      "grad_norm": 0.9759475588798523,
      "learning_rate": 0.0005,
      "loss": 3.048020362854004,
      "step": 1194
    },
    {
      "epoch": 0.27642840619939857,
      "grad_norm": 0.9858020544052124,
      "learning_rate": 0.0005,
      "loss": 3.149068832397461,
      "step": 1195
    },
    {
      "epoch": 0.27665972704140646,
      "grad_norm": 2.381641149520874,
      "learning_rate": 0.0005,
      "loss": 3.0767645835876465,
      "step": 1196
    },
    {
      "epoch": 0.2768910478834143,
      "grad_norm": 1.8490805625915527,
      "learning_rate": 0.0005,
      "loss": 3.1023192405700684,
      "step": 1197
    },
    {
      "epoch": 0.2771223687254222,
      "grad_norm": 1.3141049146652222,
      "learning_rate": 0.0005,
      "loss": 3.260498523712158,
      "step": 1198
    },
    {
      "epoch": 0.27735368956743,
      "grad_norm": 1.276157021522522,
      "learning_rate": 0.0005,
      "loss": 3.424809217453003,
      "step": 1199
    },
    {
      "epoch": 0.2775850104094379,
      "grad_norm": 1.1209297180175781,
      "learning_rate": 0.0005,
      "loss": 3.1294775009155273,
      "step": 1200
    },
    {
      "epoch": 0.2778163312514458,
      "grad_norm": 1.2226181030273438,
      "learning_rate": 0.0005,
      "loss": 3.34403657913208,
      "step": 1201
    },
    {
      "epoch": 0.2780476520934536,
      "grad_norm": 1.2854355573654175,
      "learning_rate": 0.0005,
      "loss": 3.1699788570404053,
      "step": 1202
    },
    {
      "epoch": 0.2782789729354615,
      "grad_norm": 1.6900224685668945,
      "learning_rate": 0.0005,
      "loss": 3.38187837600708,
      "step": 1203
    },
    {
      "epoch": 0.27851029377746933,
      "grad_norm": 1.1857268810272217,
      "learning_rate": 0.0005,
      "loss": 3.2581253051757812,
      "step": 1204
    },
    {
      "epoch": 0.2787416146194772,
      "grad_norm": 1.992984414100647,
      "learning_rate": 0.0005,
      "loss": 3.396421432495117,
      "step": 1205
    },
    {
      "epoch": 0.2789729354614851,
      "grad_norm": 1.547516942024231,
      "learning_rate": 0.0005,
      "loss": 3.4177932739257812,
      "step": 1206
    },
    {
      "epoch": 0.27920425630349294,
      "grad_norm": 1.519858479499817,
      "learning_rate": 0.0005,
      "loss": 3.4399869441986084,
      "step": 1207
    },
    {
      "epoch": 0.2794355771455008,
      "grad_norm": 1.1195603609085083,
      "learning_rate": 0.0005,
      "loss": 3.401745557785034,
      "step": 1208
    },
    {
      "epoch": 0.27966689798750866,
      "grad_norm": 1.4782328605651855,
      "learning_rate": 0.0005,
      "loss": 3.1330976486206055,
      "step": 1209
    },
    {
      "epoch": 0.27989821882951654,
      "grad_norm": 1.398840069770813,
      "learning_rate": 0.0005,
      "loss": 3.0247249603271484,
      "step": 1210
    },
    {
      "epoch": 0.28012953967152443,
      "grad_norm": 1.1323293447494507,
      "learning_rate": 0.0005,
      "loss": 3.0981221199035645,
      "step": 1211
    },
    {
      "epoch": 0.28036086051353226,
      "grad_norm": 1.3728053569793701,
      "learning_rate": 0.0005,
      "loss": 3.526243209838867,
      "step": 1212
    },
    {
      "epoch": 0.28059218135554015,
      "grad_norm": 1.2244044542312622,
      "learning_rate": 0.0005,
      "loss": 3.6511569023132324,
      "step": 1213
    },
    {
      "epoch": 0.280823502197548,
      "grad_norm": 1.363146185874939,
      "learning_rate": 0.0005,
      "loss": 3.1477932929992676,
      "step": 1214
    },
    {
      "epoch": 0.28105482303955587,
      "grad_norm": 1.5968201160430908,
      "learning_rate": 0.0005,
      "loss": 3.0563039779663086,
      "step": 1215
    },
    {
      "epoch": 0.28128614388156375,
      "grad_norm": 1.6373809576034546,
      "learning_rate": 0.0005,
      "loss": 3.4186954498291016,
      "step": 1216
    },
    {
      "epoch": 0.2815174647235716,
      "grad_norm": 1.3521523475646973,
      "learning_rate": 0.0005,
      "loss": 3.191148519515991,
      "step": 1217
    },
    {
      "epoch": 0.2817487855655795,
      "grad_norm": 1.4608789682388306,
      "learning_rate": 0.0005,
      "loss": 2.9261832237243652,
      "step": 1218
    },
    {
      "epoch": 0.2819801064075873,
      "grad_norm": 1.432220220565796,
      "learning_rate": 0.0005,
      "loss": 3.3200364112854004,
      "step": 1219
    },
    {
      "epoch": 0.2822114272495952,
      "grad_norm": 1.0008103847503662,
      "learning_rate": 0.0005,
      "loss": 3.3681588172912598,
      "step": 1220
    },
    {
      "epoch": 0.2824427480916031,
      "grad_norm": 1.3237487077713013,
      "learning_rate": 0.0005,
      "loss": 3.021555185317993,
      "step": 1221
    },
    {
      "epoch": 0.2826740689336109,
      "grad_norm": 1.1730817556381226,
      "learning_rate": 0.0005,
      "loss": 3.324402332305908,
      "step": 1222
    },
    {
      "epoch": 0.2829053897756188,
      "grad_norm": 1.6806559562683105,
      "learning_rate": 0.0005,
      "loss": 3.128948450088501,
      "step": 1223
    },
    {
      "epoch": 0.28313671061762663,
      "grad_norm": 1.057220220565796,
      "learning_rate": 0.0005,
      "loss": 3.465264320373535,
      "step": 1224
    },
    {
      "epoch": 0.2833680314596345,
      "grad_norm": 1.222662329673767,
      "learning_rate": 0.0005,
      "loss": 2.9688591957092285,
      "step": 1225
    },
    {
      "epoch": 0.2835993523016424,
      "grad_norm": 1.4469099044799805,
      "learning_rate": 0.0005,
      "loss": 3.3825523853302,
      "step": 1226
    },
    {
      "epoch": 0.28383067314365024,
      "grad_norm": 1.3423645496368408,
      "learning_rate": 0.0005,
      "loss": 3.280404567718506,
      "step": 1227
    },
    {
      "epoch": 0.2840619939856581,
      "grad_norm": 2.2479426860809326,
      "learning_rate": 0.0005,
      "loss": 3.2513084411621094,
      "step": 1228
    },
    {
      "epoch": 0.28429331482766595,
      "grad_norm": 1.3886198997497559,
      "learning_rate": 0.0005,
      "loss": 3.5325767993927,
      "step": 1229
    },
    {
      "epoch": 0.28452463566967384,
      "grad_norm": 1.2187808752059937,
      "learning_rate": 0.0005,
      "loss": 3.1360387802124023,
      "step": 1230
    },
    {
      "epoch": 0.28475595651168173,
      "grad_norm": 1.162785291671753,
      "learning_rate": 0.0005,
      "loss": 3.417485475540161,
      "step": 1231
    },
    {
      "epoch": 0.28498727735368956,
      "grad_norm": 1.2315726280212402,
      "learning_rate": 0.0005,
      "loss": 3.22546124458313,
      "step": 1232
    },
    {
      "epoch": 0.28521859819569745,
      "grad_norm": 1.1704965829849243,
      "learning_rate": 0.0005,
      "loss": 3.353790760040283,
      "step": 1233
    },
    {
      "epoch": 0.2854499190377053,
      "grad_norm": 1.314035177230835,
      "learning_rate": 0.0005,
      "loss": 3.026676654815674,
      "step": 1234
    },
    {
      "epoch": 0.28568123987971317,
      "grad_norm": 1.5981119871139526,
      "learning_rate": 0.0005,
      "loss": 3.1649398803710938,
      "step": 1235
    },
    {
      "epoch": 0.28591256072172105,
      "grad_norm": 1.421086311340332,
      "learning_rate": 0.0005,
      "loss": 3.124689817428589,
      "step": 1236
    },
    {
      "epoch": 0.2861438815637289,
      "grad_norm": 1.5939691066741943,
      "learning_rate": 0.0005,
      "loss": 3.381070613861084,
      "step": 1237
    },
    {
      "epoch": 0.2863752024057368,
      "grad_norm": 1.166144847869873,
      "learning_rate": 0.0005,
      "loss": 3.2327370643615723,
      "step": 1238
    },
    {
      "epoch": 0.2866065232477446,
      "grad_norm": 1.2204712629318237,
      "learning_rate": 0.0005,
      "loss": 3.0171477794647217,
      "step": 1239
    },
    {
      "epoch": 0.2868378440897525,
      "grad_norm": 1.072920799255371,
      "learning_rate": 0.0005,
      "loss": 3.0406813621520996,
      "step": 1240
    },
    {
      "epoch": 0.2870691649317604,
      "grad_norm": 0.9728024005889893,
      "learning_rate": 0.0005,
      "loss": 3.120783805847168,
      "step": 1241
    },
    {
      "epoch": 0.2873004857737682,
      "grad_norm": 1.1056729555130005,
      "learning_rate": 0.0005,
      "loss": 2.9580109119415283,
      "step": 1242
    },
    {
      "epoch": 0.2875318066157761,
      "grad_norm": 1.2078375816345215,
      "learning_rate": 0.0005,
      "loss": 3.621758460998535,
      "step": 1243
    },
    {
      "epoch": 0.28776312745778393,
      "grad_norm": 1.195425033569336,
      "learning_rate": 0.0005,
      "loss": 3.075653314590454,
      "step": 1244
    },
    {
      "epoch": 0.2879944482997918,
      "grad_norm": 1.328898310661316,
      "learning_rate": 0.0005,
      "loss": 3.0828137397766113,
      "step": 1245
    },
    {
      "epoch": 0.2882257691417997,
      "grad_norm": 1.1248579025268555,
      "learning_rate": 0.0005,
      "loss": 3.0048954486846924,
      "step": 1246
    },
    {
      "epoch": 0.28845708998380754,
      "grad_norm": 1.3145240545272827,
      "learning_rate": 0.0005,
      "loss": 2.7876155376434326,
      "step": 1247
    },
    {
      "epoch": 0.2886884108258154,
      "grad_norm": 1.199525237083435,
      "learning_rate": 0.0005,
      "loss": 3.375363349914551,
      "step": 1248
    },
    {
      "epoch": 0.28891973166782325,
      "grad_norm": 1.3971080780029297,
      "learning_rate": 0.0005,
      "loss": 3.122422933578491,
      "step": 1249
    },
    {
      "epoch": 0.28915105250983114,
      "grad_norm": 1.6800496578216553,
      "learning_rate": 0.0005,
      "loss": 3.086254596710205,
      "step": 1250
    },
    {
      "epoch": 0.28938237335183903,
      "grad_norm": 1.444370150566101,
      "learning_rate": 0.0005,
      "loss": 3.5508642196655273,
      "step": 1251
    },
    {
      "epoch": 0.28961369419384686,
      "grad_norm": 1.430405616760254,
      "learning_rate": 0.0005,
      "loss": 3.4870445728302,
      "step": 1252
    },
    {
      "epoch": 0.28984501503585475,
      "grad_norm": 1.0783289670944214,
      "learning_rate": 0.0005,
      "loss": 3.3327481746673584,
      "step": 1253
    },
    {
      "epoch": 0.2900763358778626,
      "grad_norm": 1.2949352264404297,
      "learning_rate": 0.0005,
      "loss": 3.4717178344726562,
      "step": 1254
    },
    {
      "epoch": 0.29030765671987047,
      "grad_norm": 1.775124430656433,
      "learning_rate": 0.0005,
      "loss": 3.125718593597412,
      "step": 1255
    },
    {
      "epoch": 0.2905389775618783,
      "grad_norm": 0.9725856781005859,
      "learning_rate": 0.0005,
      "loss": 2.6971144676208496,
      "step": 1256
    },
    {
      "epoch": 0.2907702984038862,
      "grad_norm": 1.0147504806518555,
      "learning_rate": 0.0005,
      "loss": 2.8955461978912354,
      "step": 1257
    },
    {
      "epoch": 0.29100161924589407,
      "grad_norm": 1.3594554662704468,
      "learning_rate": 0.0005,
      "loss": 3.0055150985717773,
      "step": 1258
    },
    {
      "epoch": 0.2912329400879019,
      "grad_norm": 1.4062579870224,
      "learning_rate": 0.0005,
      "loss": 3.313326358795166,
      "step": 1259
    },
    {
      "epoch": 0.2914642609299098,
      "grad_norm": 1.5649043321609497,
      "learning_rate": 0.0005,
      "loss": 2.9517478942871094,
      "step": 1260
    },
    {
      "epoch": 0.2916955817719176,
      "grad_norm": 1.0257339477539062,
      "learning_rate": 0.0005,
      "loss": 3.267834186553955,
      "step": 1261
    },
    {
      "epoch": 0.2919269026139255,
      "grad_norm": 1.023894190788269,
      "learning_rate": 0.0005,
      "loss": 3.471301317214966,
      "step": 1262
    },
    {
      "epoch": 0.2921582234559334,
      "grad_norm": 1.237943410873413,
      "learning_rate": 0.0005,
      "loss": 3.185333728790283,
      "step": 1263
    },
    {
      "epoch": 0.29238954429794123,
      "grad_norm": 1.1514943838119507,
      "learning_rate": 0.0005,
      "loss": 3.3238515853881836,
      "step": 1264
    },
    {
      "epoch": 0.2926208651399491,
      "grad_norm": 1.291179895401001,
      "learning_rate": 0.0005,
      "loss": 3.3467252254486084,
      "step": 1265
    },
    {
      "epoch": 0.29285218598195695,
      "grad_norm": 1.4303926229476929,
      "learning_rate": 0.0005,
      "loss": 3.205209255218506,
      "step": 1266
    },
    {
      "epoch": 0.29308350682396483,
      "grad_norm": 1.4319360256195068,
      "learning_rate": 0.0005,
      "loss": 3.328725814819336,
      "step": 1267
    },
    {
      "epoch": 0.2933148276659727,
      "grad_norm": 1.1240259408950806,
      "learning_rate": 0.0005,
      "loss": 2.952486991882324,
      "step": 1268
    },
    {
      "epoch": 0.29354614850798055,
      "grad_norm": 1.374759316444397,
      "learning_rate": 0.0005,
      "loss": 3.0496814250946045,
      "step": 1269
    },
    {
      "epoch": 0.29377746934998844,
      "grad_norm": 1.0981404781341553,
      "learning_rate": 0.0005,
      "loss": 2.995978832244873,
      "step": 1270
    },
    {
      "epoch": 0.29400879019199627,
      "grad_norm": 1.2424949407577515,
      "learning_rate": 0.0005,
      "loss": 3.2244677543640137,
      "step": 1271
    },
    {
      "epoch": 0.29424011103400416,
      "grad_norm": 1.1411566734313965,
      "learning_rate": 0.0005,
      "loss": 3.010989189147949,
      "step": 1272
    },
    {
      "epoch": 0.29447143187601205,
      "grad_norm": 1.706262230873108,
      "learning_rate": 0.0005,
      "loss": 3.2062745094299316,
      "step": 1273
    },
    {
      "epoch": 0.2947027527180199,
      "grad_norm": 1.3725091218948364,
      "learning_rate": 0.0005,
      "loss": 3.7240207195281982,
      "step": 1274
    },
    {
      "epoch": 0.29493407356002777,
      "grad_norm": 1.432753324508667,
      "learning_rate": 0.0005,
      "loss": 2.957911729812622,
      "step": 1275
    },
    {
      "epoch": 0.2951653944020356,
      "grad_norm": 1.6965467929840088,
      "learning_rate": 0.0005,
      "loss": 3.0669329166412354,
      "step": 1276
    },
    {
      "epoch": 0.2953967152440435,
      "grad_norm": 1.6764808893203735,
      "learning_rate": 0.0005,
      "loss": 3.044177293777466,
      "step": 1277
    },
    {
      "epoch": 0.29562803608605137,
      "grad_norm": 1.3159476518630981,
      "learning_rate": 0.0005,
      "loss": 3.1747875213623047,
      "step": 1278
    },
    {
      "epoch": 0.2958593569280592,
      "grad_norm": 0.973926305770874,
      "learning_rate": 0.0005,
      "loss": 3.1502935886383057,
      "step": 1279
    },
    {
      "epoch": 0.2960906777700671,
      "grad_norm": 1.394274353981018,
      "learning_rate": 0.0005,
      "loss": 3.421823501586914,
      "step": 1280
    },
    {
      "epoch": 0.2963219986120749,
      "grad_norm": 1.138830304145813,
      "learning_rate": 0.0005,
      "loss": 2.9234232902526855,
      "step": 1281
    },
    {
      "epoch": 0.2965533194540828,
      "grad_norm": 1.5451279878616333,
      "learning_rate": 0.0005,
      "loss": 3.2628636360168457,
      "step": 1282
    },
    {
      "epoch": 0.2967846402960907,
      "grad_norm": 1.5205860137939453,
      "learning_rate": 0.0005,
      "loss": 2.9477124214172363,
      "step": 1283
    },
    {
      "epoch": 0.2970159611380985,
      "grad_norm": 1.5798020362854004,
      "learning_rate": 0.0005,
      "loss": 3.1330502033233643,
      "step": 1284
    },
    {
      "epoch": 0.2972472819801064,
      "grad_norm": 1.3526540994644165,
      "learning_rate": 0.0005,
      "loss": 3.245952844619751,
      "step": 1285
    },
    {
      "epoch": 0.29747860282211425,
      "grad_norm": 1.3480768203735352,
      "learning_rate": 0.0005,
      "loss": 3.27492356300354,
      "step": 1286
    },
    {
      "epoch": 0.29770992366412213,
      "grad_norm": 1.40421462059021,
      "learning_rate": 0.0005,
      "loss": 3.0044760704040527,
      "step": 1287
    },
    {
      "epoch": 0.29794124450613,
      "grad_norm": 1.176737666130066,
      "learning_rate": 0.0005,
      "loss": 3.2626497745513916,
      "step": 1288
    },
    {
      "epoch": 0.29817256534813785,
      "grad_norm": 1.2475249767303467,
      "learning_rate": 0.0005,
      "loss": 3.427155017852783,
      "step": 1289
    },
    {
      "epoch": 0.29840388619014574,
      "grad_norm": 1.0699659585952759,
      "learning_rate": 0.0005,
      "loss": 3.129425525665283,
      "step": 1290
    },
    {
      "epoch": 0.29863520703215357,
      "grad_norm": 1.540431261062622,
      "learning_rate": 0.0005,
      "loss": 3.5037965774536133,
      "step": 1291
    },
    {
      "epoch": 0.29886652787416146,
      "grad_norm": 1.3013168573379517,
      "learning_rate": 0.0005,
      "loss": 3.375152111053467,
      "step": 1292
    },
    {
      "epoch": 0.29909784871616935,
      "grad_norm": 2.1529650688171387,
      "learning_rate": 0.0005,
      "loss": 3.009660243988037,
      "step": 1293
    },
    {
      "epoch": 0.2993291695581772,
      "grad_norm": 1.1371116638183594,
      "learning_rate": 0.0005,
      "loss": 3.5084455013275146,
      "step": 1294
    },
    {
      "epoch": 0.29956049040018506,
      "grad_norm": 1.0585918426513672,
      "learning_rate": 0.0005,
      "loss": 3.4412572383880615,
      "step": 1295
    },
    {
      "epoch": 0.2997918112421929,
      "grad_norm": 1.6140204668045044,
      "learning_rate": 0.0005,
      "loss": 2.991621255874634,
      "step": 1296
    },
    {
      "epoch": 0.3000231320842008,
      "grad_norm": 1.1042619943618774,
      "learning_rate": 0.0005,
      "loss": 3.3750288486480713,
      "step": 1297
    },
    {
      "epoch": 0.30025445292620867,
      "grad_norm": 1.2832707166671753,
      "learning_rate": 0.0005,
      "loss": 3.377115249633789,
      "step": 1298
    },
    {
      "epoch": 0.3004857737682165,
      "grad_norm": 1.142441987991333,
      "learning_rate": 0.0005,
      "loss": 3.139427661895752,
      "step": 1299
    },
    {
      "epoch": 0.3007170946102244,
      "grad_norm": 1.2761564254760742,
      "learning_rate": 0.0005,
      "loss": 3.593144655227661,
      "step": 1300
    },
    {
      "epoch": 0.3009484154522322,
      "grad_norm": 1.6081386804580688,
      "learning_rate": 0.0005,
      "loss": 3.229825973510742,
      "step": 1301
    },
    {
      "epoch": 0.3011797362942401,
      "grad_norm": 1.1479785442352295,
      "learning_rate": 0.0005,
      "loss": 3.19577693939209,
      "step": 1302
    },
    {
      "epoch": 0.301411057136248,
      "grad_norm": 1.2336829900741577,
      "learning_rate": 0.0005,
      "loss": 3.1832218170166016,
      "step": 1303
    },
    {
      "epoch": 0.3016423779782558,
      "grad_norm": 1.1105986833572388,
      "learning_rate": 0.0005,
      "loss": 3.2841451168060303,
      "step": 1304
    },
    {
      "epoch": 0.3018736988202637,
      "grad_norm": 1.5174309015274048,
      "learning_rate": 0.0005,
      "loss": 3.204192638397217,
      "step": 1305
    },
    {
      "epoch": 0.30210501966227155,
      "grad_norm": 1.2912801504135132,
      "learning_rate": 0.0005,
      "loss": 3.5060458183288574,
      "step": 1306
    },
    {
      "epoch": 0.30233634050427943,
      "grad_norm": 1.523983359336853,
      "learning_rate": 0.0005,
      "loss": 3.197134017944336,
      "step": 1307
    },
    {
      "epoch": 0.3025676613462873,
      "grad_norm": 1.446435570716858,
      "learning_rate": 0.0005,
      "loss": 3.190929651260376,
      "step": 1308
    },
    {
      "epoch": 0.30279898218829515,
      "grad_norm": 1.171636939048767,
      "learning_rate": 0.0005,
      "loss": 3.2767934799194336,
      "step": 1309
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.21918523311615,
      "learning_rate": 0.0005,
      "loss": 3.2823128700256348,
      "step": 1310
    },
    {
      "epoch": 0.30326162387231087,
      "grad_norm": 1.0880014896392822,
      "learning_rate": 0.0005,
      "loss": 3.069166421890259,
      "step": 1311
    },
    {
      "epoch": 0.30349294471431876,
      "grad_norm": 1.2632827758789062,
      "learning_rate": 0.0005,
      "loss": 3.2465412616729736,
      "step": 1312
    },
    {
      "epoch": 0.30372426555632664,
      "grad_norm": 1.3344308137893677,
      "learning_rate": 0.0005,
      "loss": 3.0910768508911133,
      "step": 1313
    },
    {
      "epoch": 0.3039555863983345,
      "grad_norm": 1.2249090671539307,
      "learning_rate": 0.0005,
      "loss": 3.1207096576690674,
      "step": 1314
    },
    {
      "epoch": 0.30418690724034236,
      "grad_norm": 1.3131446838378906,
      "learning_rate": 0.0005,
      "loss": 3.3828282356262207,
      "step": 1315
    },
    {
      "epoch": 0.3044182280823502,
      "grad_norm": 1.4495786428451538,
      "learning_rate": 0.0005,
      "loss": 3.1287994384765625,
      "step": 1316
    },
    {
      "epoch": 0.3046495489243581,
      "grad_norm": 1.5483989715576172,
      "learning_rate": 0.0005,
      "loss": 3.15230655670166,
      "step": 1317
    },
    {
      "epoch": 0.30488086976636597,
      "grad_norm": 1.3447291851043701,
      "learning_rate": 0.0005,
      "loss": 3.1192166805267334,
      "step": 1318
    },
    {
      "epoch": 0.3051121906083738,
      "grad_norm": 1.8286561965942383,
      "learning_rate": 0.0005,
      "loss": 3.4487991333007812,
      "step": 1319
    },
    {
      "epoch": 0.3053435114503817,
      "grad_norm": 1.17445707321167,
      "learning_rate": 0.0005,
      "loss": 2.951251983642578,
      "step": 1320
    },
    {
      "epoch": 0.3055748322923895,
      "grad_norm": 1.3444077968597412,
      "learning_rate": 0.0005,
      "loss": 3.3416380882263184,
      "step": 1321
    },
    {
      "epoch": 0.3058061531343974,
      "grad_norm": 1.6999354362487793,
      "learning_rate": 0.0005,
      "loss": 3.3129501342773438,
      "step": 1322
    },
    {
      "epoch": 0.3060374739764053,
      "grad_norm": 1.0530248880386353,
      "learning_rate": 0.0005,
      "loss": 3.3904242515563965,
      "step": 1323
    },
    {
      "epoch": 0.3062687948184131,
      "grad_norm": 1.2150182723999023,
      "learning_rate": 0.0005,
      "loss": 3.0672049522399902,
      "step": 1324
    },
    {
      "epoch": 0.306500115660421,
      "grad_norm": 1.2035467624664307,
      "learning_rate": 0.0005,
      "loss": 3.1089630126953125,
      "step": 1325
    },
    {
      "epoch": 0.30673143650242884,
      "grad_norm": 1.0702667236328125,
      "learning_rate": 0.0005,
      "loss": 3.0529799461364746,
      "step": 1326
    },
    {
      "epoch": 0.30696275734443673,
      "grad_norm": 1.3098417520523071,
      "learning_rate": 0.0005,
      "loss": 3.0539870262145996,
      "step": 1327
    },
    {
      "epoch": 0.3071940781864446,
      "grad_norm": 1.3265126943588257,
      "learning_rate": 0.0005,
      "loss": 3.20477294921875,
      "step": 1328
    },
    {
      "epoch": 0.30742539902845245,
      "grad_norm": 1.2098641395568848,
      "learning_rate": 0.0005,
      "loss": 3.2486023902893066,
      "step": 1329
    },
    {
      "epoch": 0.30765671987046034,
      "grad_norm": 1.0413151979446411,
      "learning_rate": 0.0005,
      "loss": 3.5050837993621826,
      "step": 1330
    },
    {
      "epoch": 0.30788804071246817,
      "grad_norm": 1.1525042057037354,
      "learning_rate": 0.0005,
      "loss": 3.3339805603027344,
      "step": 1331
    },
    {
      "epoch": 0.30811936155447606,
      "grad_norm": 1.3956630229949951,
      "learning_rate": 0.0005,
      "loss": 2.9209909439086914,
      "step": 1332
    },
    {
      "epoch": 0.30835068239648394,
      "grad_norm": 1.3990232944488525,
      "learning_rate": 0.0005,
      "loss": 3.1450421810150146,
      "step": 1333
    },
    {
      "epoch": 0.3085820032384918,
      "grad_norm": 1.1847256422042847,
      "learning_rate": 0.0005,
      "loss": 3.2382254600524902,
      "step": 1334
    },
    {
      "epoch": 0.30881332408049966,
      "grad_norm": 1.1247550249099731,
      "learning_rate": 0.0005,
      "loss": 3.620509624481201,
      "step": 1335
    },
    {
      "epoch": 0.3090446449225075,
      "grad_norm": 1.157487392425537,
      "learning_rate": 0.0005,
      "loss": 2.9485323429107666,
      "step": 1336
    },
    {
      "epoch": 0.3092759657645154,
      "grad_norm": 1.430233359336853,
      "learning_rate": 0.0005,
      "loss": 3.2547178268432617,
      "step": 1337
    },
    {
      "epoch": 0.30950728660652327,
      "grad_norm": 1.3138245344161987,
      "learning_rate": 0.0005,
      "loss": 3.162372589111328,
      "step": 1338
    },
    {
      "epoch": 0.3097386074485311,
      "grad_norm": 1.3935232162475586,
      "learning_rate": 0.0005,
      "loss": 3.4771156311035156,
      "step": 1339
    },
    {
      "epoch": 0.309969928290539,
      "grad_norm": 1.140711784362793,
      "learning_rate": 0.0005,
      "loss": 3.0286507606506348,
      "step": 1340
    },
    {
      "epoch": 0.3102012491325468,
      "grad_norm": 1.5263921022415161,
      "learning_rate": 0.0005,
      "loss": 3.226515769958496,
      "step": 1341
    },
    {
      "epoch": 0.3104325699745547,
      "grad_norm": 1.2876558303833008,
      "learning_rate": 0.0005,
      "loss": 3.4146203994750977,
      "step": 1342
    },
    {
      "epoch": 0.3106638908165626,
      "grad_norm": 1.0212554931640625,
      "learning_rate": 0.0005,
      "loss": 3.366607666015625,
      "step": 1343
    },
    {
      "epoch": 0.3108952116585704,
      "grad_norm": 1.131331443786621,
      "learning_rate": 0.0005,
      "loss": 3.68922758102417,
      "step": 1344
    },
    {
      "epoch": 0.3111265325005783,
      "grad_norm": 1.0341393947601318,
      "learning_rate": 0.0005,
      "loss": 2.921079158782959,
      "step": 1345
    },
    {
      "epoch": 0.31135785334258614,
      "grad_norm": 1.004509687423706,
      "learning_rate": 0.0005,
      "loss": 3.2061691284179688,
      "step": 1346
    },
    {
      "epoch": 0.31158917418459403,
      "grad_norm": 1.1004548072814941,
      "learning_rate": 0.0005,
      "loss": 3.1103672981262207,
      "step": 1347
    },
    {
      "epoch": 0.3118204950266019,
      "grad_norm": 1.2678449153900146,
      "learning_rate": 0.0005,
      "loss": 3.2326836585998535,
      "step": 1348
    },
    {
      "epoch": 0.31205181586860975,
      "grad_norm": 1.1621484756469727,
      "learning_rate": 0.0005,
      "loss": 3.446387767791748,
      "step": 1349
    },
    {
      "epoch": 0.31228313671061764,
      "grad_norm": 1.3611377477645874,
      "learning_rate": 0.0005,
      "loss": 2.9098780155181885,
      "step": 1350
    },
    {
      "epoch": 0.31251445755262547,
      "grad_norm": 1.2912242412567139,
      "learning_rate": 0.0005,
      "loss": 3.0065855979919434,
      "step": 1351
    },
    {
      "epoch": 0.31274577839463336,
      "grad_norm": 1.4964001178741455,
      "learning_rate": 0.0005,
      "loss": 3.176142692565918,
      "step": 1352
    },
    {
      "epoch": 0.31297709923664124,
      "grad_norm": 1.2731488943099976,
      "learning_rate": 0.0005,
      "loss": 3.3747880458831787,
      "step": 1353
    },
    {
      "epoch": 0.3132084200786491,
      "grad_norm": 1.2778011560440063,
      "learning_rate": 0.0005,
      "loss": 3.139071226119995,
      "step": 1354
    },
    {
      "epoch": 0.31343974092065696,
      "grad_norm": 1.3632832765579224,
      "learning_rate": 0.0005,
      "loss": 3.1824865341186523,
      "step": 1355
    },
    {
      "epoch": 0.3136710617626648,
      "grad_norm": 2.175809144973755,
      "learning_rate": 0.0005,
      "loss": 3.1268911361694336,
      "step": 1356
    },
    {
      "epoch": 0.3139023826046727,
      "grad_norm": 1.1061865091323853,
      "learning_rate": 0.0005,
      "loss": 3.275843620300293,
      "step": 1357
    },
    {
      "epoch": 0.31413370344668057,
      "grad_norm": 1.7408573627471924,
      "learning_rate": 0.0005,
      "loss": 3.188948631286621,
      "step": 1358
    },
    {
      "epoch": 0.3143650242886884,
      "grad_norm": 1.2297815084457397,
      "learning_rate": 0.0005,
      "loss": 3.0539402961730957,
      "step": 1359
    },
    {
      "epoch": 0.3145963451306963,
      "grad_norm": 1.114698052406311,
      "learning_rate": 0.0005,
      "loss": 3.141796112060547,
      "step": 1360
    },
    {
      "epoch": 0.3148276659727041,
      "grad_norm": 1.101495623588562,
      "learning_rate": 0.0005,
      "loss": 3.5125510692596436,
      "step": 1361
    },
    {
      "epoch": 0.315058986814712,
      "grad_norm": 1.2046297788619995,
      "learning_rate": 0.0005,
      "loss": 3.070878505706787,
      "step": 1362
    },
    {
      "epoch": 0.3152903076567199,
      "grad_norm": 1.241339921951294,
      "learning_rate": 0.0005,
      "loss": 3.1671507358551025,
      "step": 1363
    },
    {
      "epoch": 0.3155216284987277,
      "grad_norm": 1.2094942331314087,
      "learning_rate": 0.0005,
      "loss": 3.2855002880096436,
      "step": 1364
    },
    {
      "epoch": 0.3157529493407356,
      "grad_norm": 1.050940990447998,
      "learning_rate": 0.0005,
      "loss": 3.0654149055480957,
      "step": 1365
    },
    {
      "epoch": 0.31598427018274344,
      "grad_norm": 1.0174568891525269,
      "learning_rate": 0.0005,
      "loss": 3.3738319873809814,
      "step": 1366
    },
    {
      "epoch": 0.31621559102475133,
      "grad_norm": 1.2496278285980225,
      "learning_rate": 0.0005,
      "loss": 3.3274688720703125,
      "step": 1367
    },
    {
      "epoch": 0.3164469118667592,
      "grad_norm": 1.1135162115097046,
      "learning_rate": 0.0005,
      "loss": 2.9333536624908447,
      "step": 1368
    },
    {
      "epoch": 0.31667823270876705,
      "grad_norm": 1.1941050291061401,
      "learning_rate": 0.0005,
      "loss": 3.0929908752441406,
      "step": 1369
    },
    {
      "epoch": 0.31690955355077494,
      "grad_norm": 1.0905652046203613,
      "learning_rate": 0.0005,
      "loss": 3.1734681129455566,
      "step": 1370
    },
    {
      "epoch": 0.31714087439278277,
      "grad_norm": 1.0661836862564087,
      "learning_rate": 0.0005,
      "loss": 3.0591647624969482,
      "step": 1371
    },
    {
      "epoch": 0.31737219523479065,
      "grad_norm": 1.1765646934509277,
      "learning_rate": 0.0005,
      "loss": 3.306349754333496,
      "step": 1372
    },
    {
      "epoch": 0.31760351607679854,
      "grad_norm": 1.443237543106079,
      "learning_rate": 0.0005,
      "loss": 2.6936728954315186,
      "step": 1373
    },
    {
      "epoch": 0.3178348369188064,
      "grad_norm": 1.3134808540344238,
      "learning_rate": 0.0005,
      "loss": 3.148188591003418,
      "step": 1374
    },
    {
      "epoch": 0.31806615776081426,
      "grad_norm": 1.3947545289993286,
      "learning_rate": 0.0005,
      "loss": 2.8553009033203125,
      "step": 1375
    },
    {
      "epoch": 0.3182974786028221,
      "grad_norm": 1.2184484004974365,
      "learning_rate": 0.0005,
      "loss": 3.450624465942383,
      "step": 1376
    },
    {
      "epoch": 0.31852879944483,
      "grad_norm": 1.2579768896102905,
      "learning_rate": 0.0005,
      "loss": 3.415816307067871,
      "step": 1377
    },
    {
      "epoch": 0.31876012028683787,
      "grad_norm": 1.5383400917053223,
      "learning_rate": 0.0005,
      "loss": 2.7924137115478516,
      "step": 1378
    },
    {
      "epoch": 0.3189914411288457,
      "grad_norm": 1.1943082809448242,
      "learning_rate": 0.0005,
      "loss": 3.144648551940918,
      "step": 1379
    },
    {
      "epoch": 0.3192227619708536,
      "grad_norm": 1.145147681236267,
      "learning_rate": 0.0005,
      "loss": 3.414775848388672,
      "step": 1380
    },
    {
      "epoch": 0.3194540828128614,
      "grad_norm": 1.050492525100708,
      "learning_rate": 0.0005,
      "loss": 3.1314287185668945,
      "step": 1381
    },
    {
      "epoch": 0.3196854036548693,
      "grad_norm": 1.323134422302246,
      "learning_rate": 0.0005,
      "loss": 3.1355361938476562,
      "step": 1382
    },
    {
      "epoch": 0.3199167244968772,
      "grad_norm": 1.1679409742355347,
      "learning_rate": 0.0005,
      "loss": 2.7544848918914795,
      "step": 1383
    },
    {
      "epoch": 0.320148045338885,
      "grad_norm": 1.4017407894134521,
      "learning_rate": 0.0005,
      "loss": 3.281505584716797,
      "step": 1384
    },
    {
      "epoch": 0.3203793661808929,
      "grad_norm": 1.2562780380249023,
      "learning_rate": 0.0005,
      "loss": 2.8811144828796387,
      "step": 1385
    },
    {
      "epoch": 0.32061068702290074,
      "grad_norm": 1.0745558738708496,
      "learning_rate": 0.0005,
      "loss": 3.0780692100524902,
      "step": 1386
    },
    {
      "epoch": 0.32084200786490863,
      "grad_norm": 1.1709364652633667,
      "learning_rate": 0.0005,
      "loss": 3.3026790618896484,
      "step": 1387
    },
    {
      "epoch": 0.3210733287069165,
      "grad_norm": 1.534070611000061,
      "learning_rate": 0.0005,
      "loss": 3.174415349960327,
      "step": 1388
    },
    {
      "epoch": 0.32130464954892435,
      "grad_norm": 1.0831854343414307,
      "learning_rate": 0.0005,
      "loss": 3.064542293548584,
      "step": 1389
    },
    {
      "epoch": 0.32153597039093224,
      "grad_norm": 1.2247260808944702,
      "learning_rate": 0.0005,
      "loss": 2.9613614082336426,
      "step": 1390
    },
    {
      "epoch": 0.32176729123294007,
      "grad_norm": 1.2666188478469849,
      "learning_rate": 0.0005,
      "loss": 3.069507598876953,
      "step": 1391
    },
    {
      "epoch": 0.32199861207494795,
      "grad_norm": 1.0215190649032593,
      "learning_rate": 0.0005,
      "loss": 3.1708309650421143,
      "step": 1392
    },
    {
      "epoch": 0.32222993291695584,
      "grad_norm": 1.2834751605987549,
      "learning_rate": 0.0005,
      "loss": 3.038607120513916,
      "step": 1393
    },
    {
      "epoch": 0.3224612537589637,
      "grad_norm": 1.056019902229309,
      "learning_rate": 0.0005,
      "loss": 3.4219250679016113,
      "step": 1394
    },
    {
      "epoch": 0.32269257460097156,
      "grad_norm": 1.2067043781280518,
      "learning_rate": 0.0005,
      "loss": 2.9522957801818848,
      "step": 1395
    },
    {
      "epoch": 0.3229238954429794,
      "grad_norm": 1.1682119369506836,
      "learning_rate": 0.0005,
      "loss": 3.023416042327881,
      "step": 1396
    },
    {
      "epoch": 0.3231552162849873,
      "grad_norm": 1.6743546724319458,
      "learning_rate": 0.0005,
      "loss": 3.4127275943756104,
      "step": 1397
    },
    {
      "epoch": 0.32338653712699517,
      "grad_norm": 1.0193347930908203,
      "learning_rate": 0.0005,
      "loss": 3.231631278991699,
      "step": 1398
    },
    {
      "epoch": 0.323617857969003,
      "grad_norm": 1.6609177589416504,
      "learning_rate": 0.0005,
      "loss": 3.337124824523926,
      "step": 1399
    },
    {
      "epoch": 0.3238491788110109,
      "grad_norm": 1.191632866859436,
      "learning_rate": 0.0005,
      "loss": 3.2386698722839355,
      "step": 1400
    },
    {
      "epoch": 0.3240804996530187,
      "grad_norm": 1.3665316104888916,
      "learning_rate": 0.0005,
      "loss": 3.1604254245758057,
      "step": 1401
    },
    {
      "epoch": 0.3243118204950266,
      "grad_norm": 0.9643155336380005,
      "learning_rate": 0.0005,
      "loss": 2.86968994140625,
      "step": 1402
    },
    {
      "epoch": 0.3245431413370345,
      "grad_norm": 1.1963080167770386,
      "learning_rate": 0.0005,
      "loss": 3.1948442459106445,
      "step": 1403
    },
    {
      "epoch": 0.3247744621790423,
      "grad_norm": 1.1810458898544312,
      "learning_rate": 0.0005,
      "loss": 2.81738543510437,
      "step": 1404
    },
    {
      "epoch": 0.3250057830210502,
      "grad_norm": 1.2071139812469482,
      "learning_rate": 0.0005,
      "loss": 3.522496461868286,
      "step": 1405
    },
    {
      "epoch": 0.32523710386305804,
      "grad_norm": 1.0815703868865967,
      "learning_rate": 0.0005,
      "loss": 3.0174689292907715,
      "step": 1406
    },
    {
      "epoch": 0.32546842470506593,
      "grad_norm": 1.687819004058838,
      "learning_rate": 0.0005,
      "loss": 3.3207168579101562,
      "step": 1407
    },
    {
      "epoch": 0.3256997455470738,
      "grad_norm": 1.1469552516937256,
      "learning_rate": 0.0005,
      "loss": 3.505368232727051,
      "step": 1408
    },
    {
      "epoch": 0.32593106638908165,
      "grad_norm": 1.7126450538635254,
      "learning_rate": 0.0005,
      "loss": 3.425903081893921,
      "step": 1409
    },
    {
      "epoch": 0.32616238723108953,
      "grad_norm": 1.0775704383850098,
      "learning_rate": 0.0005,
      "loss": 3.302710771560669,
      "step": 1410
    },
    {
      "epoch": 0.32639370807309737,
      "grad_norm": 1.1918904781341553,
      "learning_rate": 0.0005,
      "loss": 3.553454637527466,
      "step": 1411
    },
    {
      "epoch": 0.32662502891510525,
      "grad_norm": 1.0822981595993042,
      "learning_rate": 0.0005,
      "loss": 3.155668258666992,
      "step": 1412
    },
    {
      "epoch": 0.32685634975711314,
      "grad_norm": 1.4440401792526245,
      "learning_rate": 0.0005,
      "loss": 3.0909905433654785,
      "step": 1413
    },
    {
      "epoch": 0.32708767059912097,
      "grad_norm": 1.1819862127304077,
      "learning_rate": 0.0005,
      "loss": 3.1360621452331543,
      "step": 1414
    },
    {
      "epoch": 0.32731899144112886,
      "grad_norm": 1.3448055982589722,
      "learning_rate": 0.0005,
      "loss": 2.9420788288116455,
      "step": 1415
    },
    {
      "epoch": 0.3275503122831367,
      "grad_norm": 1.3920742273330688,
      "learning_rate": 0.0005,
      "loss": 3.8451638221740723,
      "step": 1416
    },
    {
      "epoch": 0.3277816331251446,
      "grad_norm": 1.062727451324463,
      "learning_rate": 0.0005,
      "loss": 3.2774529457092285,
      "step": 1417
    },
    {
      "epoch": 0.32801295396715247,
      "grad_norm": 2.245898962020874,
      "learning_rate": 0.0005,
      "loss": 3.1002025604248047,
      "step": 1418
    },
    {
      "epoch": 0.3282442748091603,
      "grad_norm": 1.520167350769043,
      "learning_rate": 0.0005,
      "loss": 3.2069880962371826,
      "step": 1419
    },
    {
      "epoch": 0.3284755956511682,
      "grad_norm": 1.5909602642059326,
      "learning_rate": 0.0005,
      "loss": 3.220021963119507,
      "step": 1420
    },
    {
      "epoch": 0.328706916493176,
      "grad_norm": 1.2350624799728394,
      "learning_rate": 0.0005,
      "loss": 3.3573007583618164,
      "step": 1421
    },
    {
      "epoch": 0.3289382373351839,
      "grad_norm": 1.1361726522445679,
      "learning_rate": 0.0005,
      "loss": 2.9146509170532227,
      "step": 1422
    },
    {
      "epoch": 0.3291695581771918,
      "grad_norm": 1.2198532819747925,
      "learning_rate": 0.0005,
      "loss": 3.238483428955078,
      "step": 1423
    },
    {
      "epoch": 0.3294008790191996,
      "grad_norm": 1.3708068132400513,
      "learning_rate": 0.0005,
      "loss": 3.593043565750122,
      "step": 1424
    },
    {
      "epoch": 0.3296321998612075,
      "grad_norm": 1.3369581699371338,
      "learning_rate": 0.0005,
      "loss": 3.12886381149292,
      "step": 1425
    },
    {
      "epoch": 0.32986352070321534,
      "grad_norm": 2.621717691421509,
      "learning_rate": 0.0005,
      "loss": 2.945678234100342,
      "step": 1426
    },
    {
      "epoch": 0.3300948415452232,
      "grad_norm": 1.4836429357528687,
      "learning_rate": 0.0005,
      "loss": 3.4355053901672363,
      "step": 1427
    },
    {
      "epoch": 0.3303261623872311,
      "grad_norm": 1.1244056224822998,
      "learning_rate": 0.0005,
      "loss": 2.80545711517334,
      "step": 1428
    },
    {
      "epoch": 0.33055748322923895,
      "grad_norm": 1.2882795333862305,
      "learning_rate": 0.0005,
      "loss": 3.065722942352295,
      "step": 1429
    },
    {
      "epoch": 0.33078880407124683,
      "grad_norm": 1.4062918424606323,
      "learning_rate": 0.0005,
      "loss": 3.1229567527770996,
      "step": 1430
    },
    {
      "epoch": 0.33102012491325467,
      "grad_norm": 1.2004148960113525,
      "learning_rate": 0.0005,
      "loss": 3.0016422271728516,
      "step": 1431
    },
    {
      "epoch": 0.33125144575526255,
      "grad_norm": 1.0852543115615845,
      "learning_rate": 0.0005,
      "loss": 2.9071524143218994,
      "step": 1432
    },
    {
      "epoch": 0.33148276659727044,
      "grad_norm": 1.313629388809204,
      "learning_rate": 0.0005,
      "loss": 3.4881937503814697,
      "step": 1433
    },
    {
      "epoch": 0.33171408743927827,
      "grad_norm": 1.1647531986236572,
      "learning_rate": 0.0005,
      "loss": 2.962773323059082,
      "step": 1434
    },
    {
      "epoch": 0.33194540828128616,
      "grad_norm": 1.297897219657898,
      "learning_rate": 0.0005,
      "loss": 2.946392774581909,
      "step": 1435
    },
    {
      "epoch": 0.332176729123294,
      "grad_norm": 1.3118367195129395,
      "learning_rate": 0.0005,
      "loss": 3.1465072631835938,
      "step": 1436
    },
    {
      "epoch": 0.3324080499653019,
      "grad_norm": 1.6656789779663086,
      "learning_rate": 0.0005,
      "loss": 3.321185827255249,
      "step": 1437
    },
    {
      "epoch": 0.33263937080730976,
      "grad_norm": 1.3304531574249268,
      "learning_rate": 0.0005,
      "loss": 2.853710174560547,
      "step": 1438
    },
    {
      "epoch": 0.3328706916493176,
      "grad_norm": 1.4046498537063599,
      "learning_rate": 0.0005,
      "loss": 3.0763626098632812,
      "step": 1439
    },
    {
      "epoch": 0.3331020124913255,
      "grad_norm": 1.7699674367904663,
      "learning_rate": 0.0005,
      "loss": 3.089660406112671,
      "step": 1440
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.3828877210617065,
      "learning_rate": 0.0005,
      "loss": 3.271857261657715,
      "step": 1441
    },
    {
      "epoch": 0.3335646541753412,
      "grad_norm": 1.3175278902053833,
      "learning_rate": 0.0005,
      "loss": 3.130211353302002,
      "step": 1442
    },
    {
      "epoch": 0.3337959750173491,
      "grad_norm": 1.6795326471328735,
      "learning_rate": 0.0005,
      "loss": 3.043846607208252,
      "step": 1443
    },
    {
      "epoch": 0.3340272958593569,
      "grad_norm": 1.1755154132843018,
      "learning_rate": 0.0005,
      "loss": 3.226527214050293,
      "step": 1444
    },
    {
      "epoch": 0.3342586167013648,
      "grad_norm": 1.1037293672561646,
      "learning_rate": 0.0005,
      "loss": 3.3594601154327393,
      "step": 1445
    },
    {
      "epoch": 0.33448993754337264,
      "grad_norm": 1.4529716968536377,
      "learning_rate": 0.0005,
      "loss": 3.1018567085266113,
      "step": 1446
    },
    {
      "epoch": 0.3347212583853805,
      "grad_norm": 1.3442355394363403,
      "learning_rate": 0.0005,
      "loss": 2.9888062477111816,
      "step": 1447
    },
    {
      "epoch": 0.3349525792273884,
      "grad_norm": 1.2487273216247559,
      "learning_rate": 0.0005,
      "loss": 2.9425415992736816,
      "step": 1448
    },
    {
      "epoch": 0.33518390006939625,
      "grad_norm": 1.194853663444519,
      "learning_rate": 0.0005,
      "loss": 3.397928237915039,
      "step": 1449
    },
    {
      "epoch": 0.33541522091140413,
      "grad_norm": 1.320058822631836,
      "learning_rate": 0.0005,
      "loss": 2.930300235748291,
      "step": 1450
    },
    {
      "epoch": 0.33564654175341196,
      "grad_norm": 1.2254794836044312,
      "learning_rate": 0.0005,
      "loss": 3.0880932807922363,
      "step": 1451
    },
    {
      "epoch": 0.33587786259541985,
      "grad_norm": 1.5135301351547241,
      "learning_rate": 0.0005,
      "loss": 3.5332164764404297,
      "step": 1452
    },
    {
      "epoch": 0.33610918343742774,
      "grad_norm": 1.3983383178710938,
      "learning_rate": 0.0005,
      "loss": 2.872924327850342,
      "step": 1453
    },
    {
      "epoch": 0.33634050427943557,
      "grad_norm": 1.4117497205734253,
      "learning_rate": 0.0005,
      "loss": 2.7184009552001953,
      "step": 1454
    },
    {
      "epoch": 0.33657182512144346,
      "grad_norm": 1.2515404224395752,
      "learning_rate": 0.0005,
      "loss": 3.3150382041931152,
      "step": 1455
    },
    {
      "epoch": 0.3368031459634513,
      "grad_norm": 1.9756156206130981,
      "learning_rate": 0.0005,
      "loss": 3.034447193145752,
      "step": 1456
    },
    {
      "epoch": 0.3370344668054592,
      "grad_norm": 1.9714601039886475,
      "learning_rate": 0.0005,
      "loss": 3.147573709487915,
      "step": 1457
    },
    {
      "epoch": 0.33726578764746706,
      "grad_norm": 1.2999699115753174,
      "learning_rate": 0.0005,
      "loss": 3.0672621726989746,
      "step": 1458
    },
    {
      "epoch": 0.3374971084894749,
      "grad_norm": 1.2206110954284668,
      "learning_rate": 0.0005,
      "loss": 3.2070436477661133,
      "step": 1459
    },
    {
      "epoch": 0.3377284293314828,
      "grad_norm": 1.5712189674377441,
      "learning_rate": 0.0005,
      "loss": 3.018209934234619,
      "step": 1460
    },
    {
      "epoch": 0.3379597501734906,
      "grad_norm": 1.2180523872375488,
      "learning_rate": 0.0005,
      "loss": 3.1503190994262695,
      "step": 1461
    },
    {
      "epoch": 0.3381910710154985,
      "grad_norm": 1.3232924938201904,
      "learning_rate": 0.0005,
      "loss": 2.828289747238159,
      "step": 1462
    },
    {
      "epoch": 0.3384223918575064,
      "grad_norm": 1.5219318866729736,
      "learning_rate": 0.0005,
      "loss": 3.138988733291626,
      "step": 1463
    },
    {
      "epoch": 0.3386537126995142,
      "grad_norm": 1.141318440437317,
      "learning_rate": 0.0005,
      "loss": 3.142061948776245,
      "step": 1464
    },
    {
      "epoch": 0.3388850335415221,
      "grad_norm": 1.0929720401763916,
      "learning_rate": 0.0005,
      "loss": 3.0459747314453125,
      "step": 1465
    },
    {
      "epoch": 0.33911635438352994,
      "grad_norm": 1.3742808103561401,
      "learning_rate": 0.0005,
      "loss": 3.2479135990142822,
      "step": 1466
    },
    {
      "epoch": 0.3393476752255378,
      "grad_norm": 1.6415051221847534,
      "learning_rate": 0.0005,
      "loss": 3.1627302169799805,
      "step": 1467
    },
    {
      "epoch": 0.3395789960675457,
      "grad_norm": 1.279030442237854,
      "learning_rate": 0.0005,
      "loss": 3.3736324310302734,
      "step": 1468
    },
    {
      "epoch": 0.33981031690955354,
      "grad_norm": 1.6342474222183228,
      "learning_rate": 0.0005,
      "loss": 3.134817361831665,
      "step": 1469
    },
    {
      "epoch": 0.34004163775156143,
      "grad_norm": 1.1438325643539429,
      "learning_rate": 0.0005,
      "loss": 3.0203857421875,
      "step": 1470
    },
    {
      "epoch": 0.34027295859356926,
      "grad_norm": 2.85154390335083,
      "learning_rate": 0.0005,
      "loss": 3.147594928741455,
      "step": 1471
    },
    {
      "epoch": 0.34050427943557715,
      "grad_norm": 1.0149723291397095,
      "learning_rate": 0.0005,
      "loss": 2.6150286197662354,
      "step": 1472
    },
    {
      "epoch": 0.34073560027758504,
      "grad_norm": 1.0318825244903564,
      "learning_rate": 0.0005,
      "loss": 3.4051599502563477,
      "step": 1473
    },
    {
      "epoch": 0.34096692111959287,
      "grad_norm": 1.04928457736969,
      "learning_rate": 0.0005,
      "loss": 3.266463041305542,
      "step": 1474
    },
    {
      "epoch": 0.34119824196160076,
      "grad_norm": 1.2813239097595215,
      "learning_rate": 0.0005,
      "loss": 3.1918671131134033,
      "step": 1475
    },
    {
      "epoch": 0.3414295628036086,
      "grad_norm": 1.1319257020950317,
      "learning_rate": 0.0005,
      "loss": 3.1336445808410645,
      "step": 1476
    },
    {
      "epoch": 0.3416608836456165,
      "grad_norm": 1.1512802839279175,
      "learning_rate": 0.0005,
      "loss": 2.9152534008026123,
      "step": 1477
    },
    {
      "epoch": 0.34189220448762436,
      "grad_norm": 1.251184344291687,
      "learning_rate": 0.0005,
      "loss": 2.438809871673584,
      "step": 1478
    },
    {
      "epoch": 0.3421235253296322,
      "grad_norm": 1.1301897764205933,
      "learning_rate": 0.0005,
      "loss": 3.070488929748535,
      "step": 1479
    },
    {
      "epoch": 0.3423548461716401,
      "grad_norm": 1.4217631816864014,
      "learning_rate": 0.0005,
      "loss": 3.1304397583007812,
      "step": 1480
    },
    {
      "epoch": 0.3425861670136479,
      "grad_norm": 1.417410969734192,
      "learning_rate": 0.0005,
      "loss": 3.0047378540039062,
      "step": 1481
    },
    {
      "epoch": 0.3428174878556558,
      "grad_norm": 1.161490559577942,
      "learning_rate": 0.0005,
      "loss": 3.1454524993896484,
      "step": 1482
    },
    {
      "epoch": 0.34304880869766363,
      "grad_norm": 1.1084983348846436,
      "learning_rate": 0.0005,
      "loss": 2.823375940322876,
      "step": 1483
    },
    {
      "epoch": 0.3432801295396715,
      "grad_norm": 1.0128155946731567,
      "learning_rate": 0.0005,
      "loss": 3.4587490558624268,
      "step": 1484
    },
    {
      "epoch": 0.3435114503816794,
      "grad_norm": 1.0262266397476196,
      "learning_rate": 0.0005,
      "loss": 3.210545063018799,
      "step": 1485
    },
    {
      "epoch": 0.34374277122368724,
      "grad_norm": 1.287423014640808,
      "learning_rate": 0.0005,
      "loss": 2.9167206287384033,
      "step": 1486
    },
    {
      "epoch": 0.3439740920656951,
      "grad_norm": 1.2797235250473022,
      "learning_rate": 0.0005,
      "loss": 2.934791088104248,
      "step": 1487
    },
    {
      "epoch": 0.34420541290770296,
      "grad_norm": 1.3244606256484985,
      "learning_rate": 0.0005,
      "loss": 2.9387168884277344,
      "step": 1488
    },
    {
      "epoch": 0.34443673374971084,
      "grad_norm": 1.0077959299087524,
      "learning_rate": 0.0005,
      "loss": 2.847362518310547,
      "step": 1489
    },
    {
      "epoch": 0.34466805459171873,
      "grad_norm": 1.335007905960083,
      "learning_rate": 0.0005,
      "loss": 3.248492956161499,
      "step": 1490
    },
    {
      "epoch": 0.34489937543372656,
      "grad_norm": 1.0614444017410278,
      "learning_rate": 0.0005,
      "loss": 3.234354257583618,
      "step": 1491
    },
    {
      "epoch": 0.34513069627573445,
      "grad_norm": 0.9843588471412659,
      "learning_rate": 0.0005,
      "loss": 3.0124893188476562,
      "step": 1492
    },
    {
      "epoch": 0.3453620171177423,
      "grad_norm": 1.1158933639526367,
      "learning_rate": 0.0005,
      "loss": 3.046618700027466,
      "step": 1493
    },
    {
      "epoch": 0.34559333795975017,
      "grad_norm": 1.0808448791503906,
      "learning_rate": 0.0005,
      "loss": 2.822568416595459,
      "step": 1494
    },
    {
      "epoch": 0.34582465880175806,
      "grad_norm": 1.1727831363677979,
      "learning_rate": 0.0005,
      "loss": 3.4361953735351562,
      "step": 1495
    },
    {
      "epoch": 0.3460559796437659,
      "grad_norm": 1.3698244094848633,
      "learning_rate": 0.0005,
      "loss": 3.2310335636138916,
      "step": 1496
    },
    {
      "epoch": 0.3462873004857738,
      "grad_norm": 1.1472944021224976,
      "learning_rate": 0.0005,
      "loss": 2.921557664871216,
      "step": 1497
    },
    {
      "epoch": 0.3465186213277816,
      "grad_norm": 1.3418079614639282,
      "learning_rate": 0.0005,
      "loss": 3.2759175300598145,
      "step": 1498
    },
    {
      "epoch": 0.3467499421697895,
      "grad_norm": 1.0335627794265747,
      "learning_rate": 0.0005,
      "loss": 2.8749802112579346,
      "step": 1499
    },
    {
      "epoch": 0.3469812630117974,
      "grad_norm": 1.3425272703170776,
      "learning_rate": 0.0005,
      "loss": 3.205101490020752,
      "step": 1500
    },
    {
      "epoch": 0.3472125838538052,
      "grad_norm": 1.139752984046936,
      "learning_rate": 0.0005,
      "loss": 2.8829245567321777,
      "step": 1501
    },
    {
      "epoch": 0.3474439046958131,
      "grad_norm": 1.5821256637573242,
      "learning_rate": 0.0005,
      "loss": 3.047182083129883,
      "step": 1502
    },
    {
      "epoch": 0.34767522553782093,
      "grad_norm": 1.2248609066009521,
      "learning_rate": 0.0005,
      "loss": 3.1795098781585693,
      "step": 1503
    },
    {
      "epoch": 0.3479065463798288,
      "grad_norm": 1.6560155153274536,
      "learning_rate": 0.0005,
      "loss": 3.4362683296203613,
      "step": 1504
    },
    {
      "epoch": 0.3481378672218367,
      "grad_norm": 1.5132521390914917,
      "learning_rate": 0.0005,
      "loss": 2.728476047515869,
      "step": 1505
    },
    {
      "epoch": 0.34836918806384454,
      "grad_norm": 1.4513063430786133,
      "learning_rate": 0.0005,
      "loss": 3.177295446395874,
      "step": 1506
    },
    {
      "epoch": 0.3486005089058524,
      "grad_norm": 1.077427625656128,
      "learning_rate": 0.0005,
      "loss": 3.0937929153442383,
      "step": 1507
    },
    {
      "epoch": 0.34883182974786026,
      "grad_norm": 1.4095661640167236,
      "learning_rate": 0.0005,
      "loss": 2.9422311782836914,
      "step": 1508
    },
    {
      "epoch": 0.34906315058986814,
      "grad_norm": 1.0133827924728394,
      "learning_rate": 0.0005,
      "loss": 2.9552133083343506,
      "step": 1509
    },
    {
      "epoch": 0.34929447143187603,
      "grad_norm": 0.9449451565742493,
      "learning_rate": 0.0005,
      "loss": 2.864503860473633,
      "step": 1510
    },
    {
      "epoch": 0.34952579227388386,
      "grad_norm": 1.0661087036132812,
      "learning_rate": 0.0005,
      "loss": 2.933039903640747,
      "step": 1511
    },
    {
      "epoch": 0.34975711311589175,
      "grad_norm": 1.3728893995285034,
      "learning_rate": 0.0005,
      "loss": 3.2715325355529785,
      "step": 1512
    },
    {
      "epoch": 0.3499884339578996,
      "grad_norm": 1.4795690774917603,
      "learning_rate": 0.0005,
      "loss": 3.2597920894622803,
      "step": 1513
    },
    {
      "epoch": 0.35021975479990747,
      "grad_norm": 1.1209207773208618,
      "learning_rate": 0.0005,
      "loss": 3.3895363807678223,
      "step": 1514
    },
    {
      "epoch": 0.35045107564191535,
      "grad_norm": 0.9153707027435303,
      "learning_rate": 0.0005,
      "loss": 2.779200553894043,
      "step": 1515
    },
    {
      "epoch": 0.3506823964839232,
      "grad_norm": 1.3188759088516235,
      "learning_rate": 0.0005,
      "loss": 2.9432969093322754,
      "step": 1516
    },
    {
      "epoch": 0.3509137173259311,
      "grad_norm": 1.3282471895217896,
      "learning_rate": 0.0005,
      "loss": 2.7740283012390137,
      "step": 1517
    },
    {
      "epoch": 0.3511450381679389,
      "grad_norm": 1.5721924304962158,
      "learning_rate": 0.0005,
      "loss": 3.3936424255371094,
      "step": 1518
    },
    {
      "epoch": 0.3513763590099468,
      "grad_norm": 1.2306652069091797,
      "learning_rate": 0.0005,
      "loss": 2.836007833480835,
      "step": 1519
    },
    {
      "epoch": 0.3516076798519547,
      "grad_norm": 1.5431522130966187,
      "learning_rate": 0.0005,
      "loss": 3.2437095642089844,
      "step": 1520
    },
    {
      "epoch": 0.3518390006939625,
      "grad_norm": 1.0364549160003662,
      "learning_rate": 0.0005,
      "loss": 3.0824968814849854,
      "step": 1521
    },
    {
      "epoch": 0.3520703215359704,
      "grad_norm": 1.2225933074951172,
      "learning_rate": 0.0005,
      "loss": 2.9804623126983643,
      "step": 1522
    },
    {
      "epoch": 0.35230164237797823,
      "grad_norm": 1.5707470178604126,
      "learning_rate": 0.0005,
      "loss": 3.1445324420928955,
      "step": 1523
    },
    {
      "epoch": 0.3525329632199861,
      "grad_norm": 1.057732343673706,
      "learning_rate": 0.0005,
      "loss": 3.090172529220581,
      "step": 1524
    },
    {
      "epoch": 0.352764284061994,
      "grad_norm": 1.6349503993988037,
      "learning_rate": 0.0005,
      "loss": 3.4070355892181396,
      "step": 1525
    },
    {
      "epoch": 0.35299560490400184,
      "grad_norm": 1.285780906677246,
      "learning_rate": 0.0005,
      "loss": 3.1322743892669678,
      "step": 1526
    },
    {
      "epoch": 0.3532269257460097,
      "grad_norm": 1.0872154235839844,
      "learning_rate": 0.0005,
      "loss": 3.107058048248291,
      "step": 1527
    },
    {
      "epoch": 0.35345824658801756,
      "grad_norm": 1.0884796380996704,
      "learning_rate": 0.0005,
      "loss": 3.1601924896240234,
      "step": 1528
    },
    {
      "epoch": 0.35368956743002544,
      "grad_norm": 1.7096613645553589,
      "learning_rate": 0.0005,
      "loss": 3.3237576484680176,
      "step": 1529
    },
    {
      "epoch": 0.35392088827203333,
      "grad_norm": 1.2390614748001099,
      "learning_rate": 0.0005,
      "loss": 3.076786518096924,
      "step": 1530
    },
    {
      "epoch": 0.35415220911404116,
      "grad_norm": 1.218960165977478,
      "learning_rate": 0.0005,
      "loss": 3.1259288787841797,
      "step": 1531
    },
    {
      "epoch": 0.35438352995604905,
      "grad_norm": 1.2450584173202515,
      "learning_rate": 0.0005,
      "loss": 2.807065486907959,
      "step": 1532
    },
    {
      "epoch": 0.3546148507980569,
      "grad_norm": 1.4594529867172241,
      "learning_rate": 0.0005,
      "loss": 3.346928119659424,
      "step": 1533
    },
    {
      "epoch": 0.35484617164006477,
      "grad_norm": 1.102424144744873,
      "learning_rate": 0.0005,
      "loss": 2.956101417541504,
      "step": 1534
    },
    {
      "epoch": 0.35507749248207265,
      "grad_norm": 1.2612708806991577,
      "learning_rate": 0.0005,
      "loss": 2.9849350452423096,
      "step": 1535
    },
    {
      "epoch": 0.3553088133240805,
      "grad_norm": 1.4998220205307007,
      "learning_rate": 0.0005,
      "loss": 3.1445229053497314,
      "step": 1536
    },
    {
      "epoch": 0.3555401341660884,
      "grad_norm": 1.2170608043670654,
      "learning_rate": 0.0005,
      "loss": 2.89797306060791,
      "step": 1537
    },
    {
      "epoch": 0.3557714550080962,
      "grad_norm": 1.2017568349838257,
      "learning_rate": 0.0005,
      "loss": 3.0056729316711426,
      "step": 1538
    },
    {
      "epoch": 0.3560027758501041,
      "grad_norm": 1.2320891618728638,
      "learning_rate": 0.0005,
      "loss": 2.9794859886169434,
      "step": 1539
    },
    {
      "epoch": 0.356234096692112,
      "grad_norm": 1.5600820779800415,
      "learning_rate": 0.0005,
      "loss": 2.973172664642334,
      "step": 1540
    },
    {
      "epoch": 0.3564654175341198,
      "grad_norm": 1.1021522283554077,
      "learning_rate": 0.0005,
      "loss": 2.968681573867798,
      "step": 1541
    },
    {
      "epoch": 0.3566967383761277,
      "grad_norm": 1.293311357498169,
      "learning_rate": 0.0005,
      "loss": 3.0423355102539062,
      "step": 1542
    },
    {
      "epoch": 0.35692805921813553,
      "grad_norm": 1.167078971862793,
      "learning_rate": 0.0005,
      "loss": 3.0517969131469727,
      "step": 1543
    },
    {
      "epoch": 0.3571593800601434,
      "grad_norm": 1.4763938188552856,
      "learning_rate": 0.0005,
      "loss": 3.203969955444336,
      "step": 1544
    },
    {
      "epoch": 0.3573907009021513,
      "grad_norm": 1.851751685142517,
      "learning_rate": 0.0005,
      "loss": 2.988030433654785,
      "step": 1545
    },
    {
      "epoch": 0.35762202174415914,
      "grad_norm": 1.7805277109146118,
      "learning_rate": 0.0005,
      "loss": 3.438692569732666,
      "step": 1546
    },
    {
      "epoch": 0.357853342586167,
      "grad_norm": 1.3089475631713867,
      "learning_rate": 0.0005,
      "loss": 3.005681037902832,
      "step": 1547
    },
    {
      "epoch": 0.35808466342817485,
      "grad_norm": 1.0678905248641968,
      "learning_rate": 0.0005,
      "loss": 3.0043821334838867,
      "step": 1548
    },
    {
      "epoch": 0.35831598427018274,
      "grad_norm": 2.92518949508667,
      "learning_rate": 0.0005,
      "loss": 3.214399814605713,
      "step": 1549
    },
    {
      "epoch": 0.35854730511219063,
      "grad_norm": 1.1556700468063354,
      "learning_rate": 0.0005,
      "loss": 3.194894313812256,
      "step": 1550
    },
    {
      "epoch": 0.35877862595419846,
      "grad_norm": 1.7049579620361328,
      "learning_rate": 0.0005,
      "loss": 3.1557278633117676,
      "step": 1551
    },
    {
      "epoch": 0.35900994679620635,
      "grad_norm": 1.0066232681274414,
      "learning_rate": 0.0005,
      "loss": 3.1193389892578125,
      "step": 1552
    },
    {
      "epoch": 0.3592412676382142,
      "grad_norm": 1.3350480794906616,
      "learning_rate": 0.0005,
      "loss": 2.9548423290252686,
      "step": 1553
    },
    {
      "epoch": 0.35947258848022207,
      "grad_norm": 1.2259612083435059,
      "learning_rate": 0.0005,
      "loss": 3.0409796237945557,
      "step": 1554
    },
    {
      "epoch": 0.35970390932222995,
      "grad_norm": 1.1311084032058716,
      "learning_rate": 0.0005,
      "loss": 2.6494617462158203,
      "step": 1555
    },
    {
      "epoch": 0.3599352301642378,
      "grad_norm": 1.3111547231674194,
      "learning_rate": 0.0005,
      "loss": 2.9492709636688232,
      "step": 1556
    },
    {
      "epoch": 0.36016655100624567,
      "grad_norm": 1.1718143224716187,
      "learning_rate": 0.0005,
      "loss": 3.129262924194336,
      "step": 1557
    },
    {
      "epoch": 0.3603978718482535,
      "grad_norm": 1.2146166563034058,
      "learning_rate": 0.0005,
      "loss": 3.0202531814575195,
      "step": 1558
    },
    {
      "epoch": 0.3606291926902614,
      "grad_norm": 1.4840457439422607,
      "learning_rate": 0.0005,
      "loss": 2.8450560569763184,
      "step": 1559
    },
    {
      "epoch": 0.3608605135322693,
      "grad_norm": 0.9371118545532227,
      "learning_rate": 0.0005,
      "loss": 2.9880266189575195,
      "step": 1560
    },
    {
      "epoch": 0.3610918343742771,
      "grad_norm": 1.2699328660964966,
      "learning_rate": 0.0005,
      "loss": 2.9718661308288574,
      "step": 1561
    },
    {
      "epoch": 0.361323155216285,
      "grad_norm": 1.4330536127090454,
      "learning_rate": 0.0005,
      "loss": 3.008685350418091,
      "step": 1562
    },
    {
      "epoch": 0.36155447605829283,
      "grad_norm": 1.0973408222198486,
      "learning_rate": 0.0005,
      "loss": 3.3698244094848633,
      "step": 1563
    },
    {
      "epoch": 0.3617857969003007,
      "grad_norm": 1.0683631896972656,
      "learning_rate": 0.0005,
      "loss": 3.3512747287750244,
      "step": 1564
    },
    {
      "epoch": 0.3620171177423086,
      "grad_norm": 1.142401933670044,
      "learning_rate": 0.0005,
      "loss": 3.4824395179748535,
      "step": 1565
    },
    {
      "epoch": 0.36224843858431643,
      "grad_norm": 1.101082682609558,
      "learning_rate": 0.0005,
      "loss": 3.241767644882202,
      "step": 1566
    },
    {
      "epoch": 0.3624797594263243,
      "grad_norm": 0.9013984799385071,
      "learning_rate": 0.0005,
      "loss": 2.8051650524139404,
      "step": 1567
    },
    {
      "epoch": 0.36271108026833215,
      "grad_norm": 1.1531168222427368,
      "learning_rate": 0.0005,
      "loss": 2.9034996032714844,
      "step": 1568
    },
    {
      "epoch": 0.36294240111034004,
      "grad_norm": 1.3385740518569946,
      "learning_rate": 0.0005,
      "loss": 3.083590269088745,
      "step": 1569
    },
    {
      "epoch": 0.3631737219523479,
      "grad_norm": 0.9862027764320374,
      "learning_rate": 0.0005,
      "loss": 2.8951778411865234,
      "step": 1570
    },
    {
      "epoch": 0.36340504279435576,
      "grad_norm": 1.5995335578918457,
      "learning_rate": 0.0005,
      "loss": 3.4368205070495605,
      "step": 1571
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.2890877723693848,
      "learning_rate": 0.0005,
      "loss": 2.8963115215301514,
      "step": 1572
    },
    {
      "epoch": 0.3638676844783715,
      "grad_norm": 1.907170057296753,
      "learning_rate": 0.0005,
      "loss": 3.235332727432251,
      "step": 1573
    },
    {
      "epoch": 0.36409900532037937,
      "grad_norm": 1.0678859949111938,
      "learning_rate": 0.0005,
      "loss": 3.0697262287139893,
      "step": 1574
    },
    {
      "epoch": 0.36433032616238725,
      "grad_norm": 1.267791986465454,
      "learning_rate": 0.0005,
      "loss": 2.88132643699646,
      "step": 1575
    },
    {
      "epoch": 0.3645616470043951,
      "grad_norm": 1.2934608459472656,
      "learning_rate": 0.0005,
      "loss": 3.4782094955444336,
      "step": 1576
    },
    {
      "epoch": 0.36479296784640297,
      "grad_norm": 1.3135251998901367,
      "learning_rate": 0.0005,
      "loss": 3.656982898712158,
      "step": 1577
    },
    {
      "epoch": 0.3650242886884108,
      "grad_norm": 1.210749626159668,
      "learning_rate": 0.0005,
      "loss": 2.8959386348724365,
      "step": 1578
    },
    {
      "epoch": 0.3652556095304187,
      "grad_norm": 1.6168482303619385,
      "learning_rate": 0.0005,
      "loss": 3.3424315452575684,
      "step": 1579
    },
    {
      "epoch": 0.3654869303724266,
      "grad_norm": 1.077257513999939,
      "learning_rate": 0.0005,
      "loss": 2.9535253047943115,
      "step": 1580
    },
    {
      "epoch": 0.3657182512144344,
      "grad_norm": 1.3010963201522827,
      "learning_rate": 0.0005,
      "loss": 2.9313442707061768,
      "step": 1581
    },
    {
      "epoch": 0.3659495720564423,
      "grad_norm": 1.1165220737457275,
      "learning_rate": 0.0005,
      "loss": 3.0333781242370605,
      "step": 1582
    },
    {
      "epoch": 0.3661808928984501,
      "grad_norm": 1.2099111080169678,
      "learning_rate": 0.0005,
      "loss": 2.917602062225342,
      "step": 1583
    },
    {
      "epoch": 0.366412213740458,
      "grad_norm": 1.4489421844482422,
      "learning_rate": 0.0005,
      "loss": 3.1976914405822754,
      "step": 1584
    },
    {
      "epoch": 0.3666435345824659,
      "grad_norm": 1.1695160865783691,
      "learning_rate": 0.0005,
      "loss": 3.182417392730713,
      "step": 1585
    },
    {
      "epoch": 0.36687485542447373,
      "grad_norm": 1.178013801574707,
      "learning_rate": 0.0005,
      "loss": 3.2592153549194336,
      "step": 1586
    },
    {
      "epoch": 0.3671061762664816,
      "grad_norm": 1.4020287990570068,
      "learning_rate": 0.0005,
      "loss": 3.2517642974853516,
      "step": 1587
    },
    {
      "epoch": 0.36733749710848945,
      "grad_norm": 1.487388253211975,
      "learning_rate": 0.0005,
      "loss": 2.770918130874634,
      "step": 1588
    },
    {
      "epoch": 0.36756881795049734,
      "grad_norm": 1.192368745803833,
      "learning_rate": 0.0005,
      "loss": 3.116222381591797,
      "step": 1589
    },
    {
      "epoch": 0.3678001387925052,
      "grad_norm": 1.2975555658340454,
      "learning_rate": 0.0005,
      "loss": 3.147392511367798,
      "step": 1590
    },
    {
      "epoch": 0.36803145963451306,
      "grad_norm": 1.410276174545288,
      "learning_rate": 0.0005,
      "loss": 3.0227956771850586,
      "step": 1591
    },
    {
      "epoch": 0.36826278047652095,
      "grad_norm": 1.1654607057571411,
      "learning_rate": 0.0005,
      "loss": 3.304549217224121,
      "step": 1592
    },
    {
      "epoch": 0.3684941013185288,
      "grad_norm": 1.067604899406433,
      "learning_rate": 0.0005,
      "loss": 2.9124412536621094,
      "step": 1593
    },
    {
      "epoch": 0.36872542216053666,
      "grad_norm": 1.6407458782196045,
      "learning_rate": 0.0005,
      "loss": 2.991563558578491,
      "step": 1594
    },
    {
      "epoch": 0.36895674300254455,
      "grad_norm": 1.214957356452942,
      "learning_rate": 0.0005,
      "loss": 3.0949459075927734,
      "step": 1595
    },
    {
      "epoch": 0.3691880638445524,
      "grad_norm": 1.8221988677978516,
      "learning_rate": 0.0005,
      "loss": 3.4974770545959473,
      "step": 1596
    },
    {
      "epoch": 0.36941938468656027,
      "grad_norm": 1.341248631477356,
      "learning_rate": 0.0005,
      "loss": 3.0834054946899414,
      "step": 1597
    },
    {
      "epoch": 0.3696507055285681,
      "grad_norm": 1.0000849962234497,
      "learning_rate": 0.0005,
      "loss": 3.1585612297058105,
      "step": 1598
    },
    {
      "epoch": 0.369882026370576,
      "grad_norm": 1.053768515586853,
      "learning_rate": 0.0005,
      "loss": 2.8906619548797607,
      "step": 1599
    },
    {
      "epoch": 0.3701133472125839,
      "grad_norm": 1.274071216583252,
      "learning_rate": 0.0005,
      "loss": 3.227471351623535,
      "step": 1600
    },
    {
      "epoch": 0.3703446680545917,
      "grad_norm": 1.5218666791915894,
      "learning_rate": 0.0005,
      "loss": 3.234745979309082,
      "step": 1601
    },
    {
      "epoch": 0.3705759888965996,
      "grad_norm": 1.0738176107406616,
      "learning_rate": 0.0005,
      "loss": 2.848217487335205,
      "step": 1602
    },
    {
      "epoch": 0.3708073097386074,
      "grad_norm": 1.067305088043213,
      "learning_rate": 0.0005,
      "loss": 2.862583637237549,
      "step": 1603
    },
    {
      "epoch": 0.3710386305806153,
      "grad_norm": 1.179463267326355,
      "learning_rate": 0.0005,
      "loss": 2.9271557331085205,
      "step": 1604
    },
    {
      "epoch": 0.3712699514226232,
      "grad_norm": 1.0988109111785889,
      "learning_rate": 0.0005,
      "loss": 3.262279987335205,
      "step": 1605
    },
    {
      "epoch": 0.37150127226463103,
      "grad_norm": 1.0158100128173828,
      "learning_rate": 0.0005,
      "loss": 3.1731245517730713,
      "step": 1606
    },
    {
      "epoch": 0.3717325931066389,
      "grad_norm": 1.1508421897888184,
      "learning_rate": 0.0005,
      "loss": 2.81614351272583,
      "step": 1607
    },
    {
      "epoch": 0.37196391394864675,
      "grad_norm": 1.360498309135437,
      "learning_rate": 0.0005,
      "loss": 2.9691433906555176,
      "step": 1608
    },
    {
      "epoch": 0.37219523479065464,
      "grad_norm": 1.1234402656555176,
      "learning_rate": 0.0005,
      "loss": 3.005359172821045,
      "step": 1609
    },
    {
      "epoch": 0.3724265556326625,
      "grad_norm": 1.434147834777832,
      "learning_rate": 0.0005,
      "loss": 2.9940478801727295,
      "step": 1610
    },
    {
      "epoch": 0.37265787647467036,
      "grad_norm": 1.02466881275177,
      "learning_rate": 0.0005,
      "loss": 2.9984350204467773,
      "step": 1611
    },
    {
      "epoch": 0.37288919731667824,
      "grad_norm": 1.0662897825241089,
      "learning_rate": 0.0005,
      "loss": 2.935732841491699,
      "step": 1612
    },
    {
      "epoch": 0.3731205181586861,
      "grad_norm": 1.228594422340393,
      "learning_rate": 0.0005,
      "loss": 3.1094343662261963,
      "step": 1613
    },
    {
      "epoch": 0.37335183900069396,
      "grad_norm": 1.1730656623840332,
      "learning_rate": 0.0005,
      "loss": 3.0182766914367676,
      "step": 1614
    },
    {
      "epoch": 0.37358315984270185,
      "grad_norm": 0.9675710201263428,
      "learning_rate": 0.0005,
      "loss": 3.162346601486206,
      "step": 1615
    },
    {
      "epoch": 0.3738144806847097,
      "grad_norm": 1.0858074426651,
      "learning_rate": 0.0005,
      "loss": 3.0722343921661377,
      "step": 1616
    },
    {
      "epoch": 0.37404580152671757,
      "grad_norm": 1.026370882987976,
      "learning_rate": 0.0005,
      "loss": 3.271085739135742,
      "step": 1617
    },
    {
      "epoch": 0.3742771223687254,
      "grad_norm": 1.0163426399230957,
      "learning_rate": 0.0005,
      "loss": 2.984069347381592,
      "step": 1618
    },
    {
      "epoch": 0.3745084432107333,
      "grad_norm": 1.2470213174819946,
      "learning_rate": 0.0005,
      "loss": 3.3268229961395264,
      "step": 1619
    },
    {
      "epoch": 0.3747397640527412,
      "grad_norm": 1.4101965427398682,
      "learning_rate": 0.0005,
      "loss": 3.174933433532715,
      "step": 1620
    },
    {
      "epoch": 0.374971084894749,
      "grad_norm": 1.11814546585083,
      "learning_rate": 0.0005,
      "loss": 3.003356695175171,
      "step": 1621
    },
    {
      "epoch": 0.3752024057367569,
      "grad_norm": 1.1852647066116333,
      "learning_rate": 0.0005,
      "loss": 2.820481777191162,
      "step": 1622
    },
    {
      "epoch": 0.3754337265787647,
      "grad_norm": 1.799849510192871,
      "learning_rate": 0.0005,
      "loss": 2.966279983520508,
      "step": 1623
    },
    {
      "epoch": 0.3756650474207726,
      "grad_norm": 1.463597059249878,
      "learning_rate": 0.0005,
      "loss": 3.118473529815674,
      "step": 1624
    },
    {
      "epoch": 0.3758963682627805,
      "grad_norm": 1.2065974473953247,
      "learning_rate": 0.0005,
      "loss": 2.8770627975463867,
      "step": 1625
    },
    {
      "epoch": 0.37612768910478833,
      "grad_norm": 1.0095674991607666,
      "learning_rate": 0.0005,
      "loss": 3.240431308746338,
      "step": 1626
    },
    {
      "epoch": 0.3763590099467962,
      "grad_norm": 1.0122381448745728,
      "learning_rate": 0.0005,
      "loss": 3.0730738639831543,
      "step": 1627
    },
    {
      "epoch": 0.37659033078880405,
      "grad_norm": 1.1271899938583374,
      "learning_rate": 0.0005,
      "loss": 3.1085522174835205,
      "step": 1628
    },
    {
      "epoch": 0.37682165163081194,
      "grad_norm": 1.3242143392562866,
      "learning_rate": 0.0005,
      "loss": 3.031447649002075,
      "step": 1629
    },
    {
      "epoch": 0.3770529724728198,
      "grad_norm": 1.1283223628997803,
      "learning_rate": 0.0005,
      "loss": 2.8343803882598877,
      "step": 1630
    },
    {
      "epoch": 0.37728429331482766,
      "grad_norm": 1.1373523473739624,
      "learning_rate": 0.0005,
      "loss": 2.96480655670166,
      "step": 1631
    },
    {
      "epoch": 0.37751561415683554,
      "grad_norm": 1.1435041427612305,
      "learning_rate": 0.0005,
      "loss": 2.713153600692749,
      "step": 1632
    },
    {
      "epoch": 0.3777469349988434,
      "grad_norm": 1.2542370557785034,
      "learning_rate": 0.0005,
      "loss": 3.055950403213501,
      "step": 1633
    },
    {
      "epoch": 0.37797825584085126,
      "grad_norm": 1.6003484725952148,
      "learning_rate": 0.0005,
      "loss": 3.1235315799713135,
      "step": 1634
    },
    {
      "epoch": 0.37820957668285915,
      "grad_norm": 1.3210899829864502,
      "learning_rate": 0.0005,
      "loss": 2.834963798522949,
      "step": 1635
    },
    {
      "epoch": 0.378440897524867,
      "grad_norm": 1.197556972503662,
      "learning_rate": 0.0005,
      "loss": 3.156611442565918,
      "step": 1636
    },
    {
      "epoch": 0.37867221836687487,
      "grad_norm": 1.3068690299987793,
      "learning_rate": 0.0005,
      "loss": 3.151550531387329,
      "step": 1637
    },
    {
      "epoch": 0.3789035392088827,
      "grad_norm": 0.9311863780021667,
      "learning_rate": 0.0005,
      "loss": 3.2856154441833496,
      "step": 1638
    },
    {
      "epoch": 0.3791348600508906,
      "grad_norm": 1.4078738689422607,
      "learning_rate": 0.0005,
      "loss": 3.232626438140869,
      "step": 1639
    },
    {
      "epoch": 0.3793661808928985,
      "grad_norm": 1.2310179471969604,
      "learning_rate": 0.0005,
      "loss": 2.8383853435516357,
      "step": 1640
    },
    {
      "epoch": 0.3795975017349063,
      "grad_norm": 1.435064673423767,
      "learning_rate": 0.0005,
      "loss": 3.0930490493774414,
      "step": 1641
    },
    {
      "epoch": 0.3798288225769142,
      "grad_norm": 1.1112831830978394,
      "learning_rate": 0.0005,
      "loss": 2.845975875854492,
      "step": 1642
    },
    {
      "epoch": 0.380060143418922,
      "grad_norm": 1.3590672016143799,
      "learning_rate": 0.0005,
      "loss": 3.1095170974731445,
      "step": 1643
    },
    {
      "epoch": 0.3802914642609299,
      "grad_norm": 1.7374627590179443,
      "learning_rate": 0.0005,
      "loss": 3.135030508041382,
      "step": 1644
    },
    {
      "epoch": 0.3805227851029378,
      "grad_norm": 1.3098235130310059,
      "learning_rate": 0.0005,
      "loss": 3.1707653999328613,
      "step": 1645
    },
    {
      "epoch": 0.38075410594494563,
      "grad_norm": 1.1622815132141113,
      "learning_rate": 0.0005,
      "loss": 2.711212158203125,
      "step": 1646
    },
    {
      "epoch": 0.3809854267869535,
      "grad_norm": 1.0288848876953125,
      "learning_rate": 0.0005,
      "loss": 2.7095346450805664,
      "step": 1647
    },
    {
      "epoch": 0.38121674762896135,
      "grad_norm": 1.1317514181137085,
      "learning_rate": 0.0005,
      "loss": 2.973914623260498,
      "step": 1648
    },
    {
      "epoch": 0.38144806847096924,
      "grad_norm": 1.53046452999115,
      "learning_rate": 0.0005,
      "loss": 3.0485167503356934,
      "step": 1649
    },
    {
      "epoch": 0.3816793893129771,
      "grad_norm": 1.050736427307129,
      "learning_rate": 0.0005,
      "loss": 3.161381721496582,
      "step": 1650
    },
    {
      "epoch": 0.38191071015498496,
      "grad_norm": 1.8559746742248535,
      "learning_rate": 0.0005,
      "loss": 3.4009056091308594,
      "step": 1651
    },
    {
      "epoch": 0.38214203099699284,
      "grad_norm": 1.216512680053711,
      "learning_rate": 0.0005,
      "loss": 3.0232930183410645,
      "step": 1652
    },
    {
      "epoch": 0.3823733518390007,
      "grad_norm": 1.3841345310211182,
      "learning_rate": 0.0005,
      "loss": 2.9971163272857666,
      "step": 1653
    },
    {
      "epoch": 0.38260467268100856,
      "grad_norm": 1.1749924421310425,
      "learning_rate": 0.0005,
      "loss": 3.1731369495391846,
      "step": 1654
    },
    {
      "epoch": 0.38283599352301645,
      "grad_norm": 1.0460846424102783,
      "learning_rate": 0.0005,
      "loss": 2.802992343902588,
      "step": 1655
    },
    {
      "epoch": 0.3830673143650243,
      "grad_norm": 1.0734285116195679,
      "learning_rate": 0.0005,
      "loss": 3.3006157875061035,
      "step": 1656
    },
    {
      "epoch": 0.38329863520703217,
      "grad_norm": 1.1829512119293213,
      "learning_rate": 0.0005,
      "loss": 2.923532485961914,
      "step": 1657
    },
    {
      "epoch": 0.38352995604904,
      "grad_norm": 1.2524305582046509,
      "learning_rate": 0.0005,
      "loss": 3.163506031036377,
      "step": 1658
    },
    {
      "epoch": 0.3837612768910479,
      "grad_norm": 1.8967950344085693,
      "learning_rate": 0.0005,
      "loss": 2.8594002723693848,
      "step": 1659
    },
    {
      "epoch": 0.3839925977330558,
      "grad_norm": 1.2025753259658813,
      "learning_rate": 0.0005,
      "loss": 3.033905029296875,
      "step": 1660
    },
    {
      "epoch": 0.3842239185750636,
      "grad_norm": 1.1378154754638672,
      "learning_rate": 0.0005,
      "loss": 3.0476789474487305,
      "step": 1661
    },
    {
      "epoch": 0.3844552394170715,
      "grad_norm": 1.2664060592651367,
      "learning_rate": 0.0005,
      "loss": 3.443321704864502,
      "step": 1662
    },
    {
      "epoch": 0.3846865602590793,
      "grad_norm": 1.1941421031951904,
      "learning_rate": 0.0005,
      "loss": 3.29449462890625,
      "step": 1663
    },
    {
      "epoch": 0.3849178811010872,
      "grad_norm": 1.2192659378051758,
      "learning_rate": 0.0005,
      "loss": 3.2052841186523438,
      "step": 1664
    },
    {
      "epoch": 0.3851492019430951,
      "grad_norm": 1.0401248931884766,
      "learning_rate": 0.0005,
      "loss": 2.837794303894043,
      "step": 1665
    },
    {
      "epoch": 0.38538052278510293,
      "grad_norm": 0.9385779500007629,
      "learning_rate": 0.0005,
      "loss": 2.812325954437256,
      "step": 1666
    },
    {
      "epoch": 0.3856118436271108,
      "grad_norm": 1.0194207429885864,
      "learning_rate": 0.0005,
      "loss": 3.138922929763794,
      "step": 1667
    },
    {
      "epoch": 0.38584316446911865,
      "grad_norm": 1.0977956056594849,
      "learning_rate": 0.0005,
      "loss": 3.227851390838623,
      "step": 1668
    },
    {
      "epoch": 0.38607448531112654,
      "grad_norm": 1.1342089176177979,
      "learning_rate": 0.0005,
      "loss": 2.852487564086914,
      "step": 1669
    },
    {
      "epoch": 0.3863058061531344,
      "grad_norm": 1.260693907737732,
      "learning_rate": 0.0005,
      "loss": 2.949669361114502,
      "step": 1670
    },
    {
      "epoch": 0.38653712699514226,
      "grad_norm": 1.578537940979004,
      "learning_rate": 0.0005,
      "loss": 2.968379497528076,
      "step": 1671
    },
    {
      "epoch": 0.38676844783715014,
      "grad_norm": 1.2074934244155884,
      "learning_rate": 0.0005,
      "loss": 3.05387806892395,
      "step": 1672
    },
    {
      "epoch": 0.386999768679158,
      "grad_norm": 1.2265112400054932,
      "learning_rate": 0.0005,
      "loss": 2.962583065032959,
      "step": 1673
    },
    {
      "epoch": 0.38723108952116586,
      "grad_norm": 1.0844931602478027,
      "learning_rate": 0.0005,
      "loss": 3.03916597366333,
      "step": 1674
    },
    {
      "epoch": 0.38746241036317375,
      "grad_norm": 1.2922554016113281,
      "learning_rate": 0.0005,
      "loss": 3.1598191261291504,
      "step": 1675
    },
    {
      "epoch": 0.3876937312051816,
      "grad_norm": 1.2153282165527344,
      "learning_rate": 0.0005,
      "loss": 2.8360719680786133,
      "step": 1676
    },
    {
      "epoch": 0.38792505204718947,
      "grad_norm": 1.238377332687378,
      "learning_rate": 0.0005,
      "loss": 3.2227795124053955,
      "step": 1677
    },
    {
      "epoch": 0.3881563728891973,
      "grad_norm": 0.9348199963569641,
      "learning_rate": 0.0005,
      "loss": 2.7960171699523926,
      "step": 1678
    },
    {
      "epoch": 0.3883876937312052,
      "grad_norm": 0.9884197115898132,
      "learning_rate": 0.0005,
      "loss": 2.645162343978882,
      "step": 1679
    },
    {
      "epoch": 0.3886190145732131,
      "grad_norm": 1.29781174659729,
      "learning_rate": 0.0005,
      "loss": 3.047858238220215,
      "step": 1680
    },
    {
      "epoch": 0.3888503354152209,
      "grad_norm": 1.7625980377197266,
      "learning_rate": 0.0005,
      "loss": 2.8166003227233887,
      "step": 1681
    },
    {
      "epoch": 0.3890816562572288,
      "grad_norm": 1.2046486139297485,
      "learning_rate": 0.0005,
      "loss": 3.174708843231201,
      "step": 1682
    },
    {
      "epoch": 0.3893129770992366,
      "grad_norm": 1.3737125396728516,
      "learning_rate": 0.0005,
      "loss": 2.7828667163848877,
      "step": 1683
    },
    {
      "epoch": 0.3895442979412445,
      "grad_norm": 1.1986579895019531,
      "learning_rate": 0.0005,
      "loss": 3.3036420345306396,
      "step": 1684
    },
    {
      "epoch": 0.3897756187832524,
      "grad_norm": 1.099687933921814,
      "learning_rate": 0.0005,
      "loss": 2.867654800415039,
      "step": 1685
    },
    {
      "epoch": 0.39000693962526023,
      "grad_norm": 1.1114814281463623,
      "learning_rate": 0.0005,
      "loss": 3.208207607269287,
      "step": 1686
    },
    {
      "epoch": 0.3902382604672681,
      "grad_norm": 0.9987950325012207,
      "learning_rate": 0.0005,
      "loss": 2.9426982402801514,
      "step": 1687
    },
    {
      "epoch": 0.39046958130927595,
      "grad_norm": 0.9676096439361572,
      "learning_rate": 0.0005,
      "loss": 2.7542662620544434,
      "step": 1688
    },
    {
      "epoch": 0.39070090215128384,
      "grad_norm": 1.17933988571167,
      "learning_rate": 0.0005,
      "loss": 3.110286235809326,
      "step": 1689
    },
    {
      "epoch": 0.3909322229932917,
      "grad_norm": 1.2053598165512085,
      "learning_rate": 0.0005,
      "loss": 3.2823476791381836,
      "step": 1690
    },
    {
      "epoch": 0.39116354383529955,
      "grad_norm": 1.5013853311538696,
      "learning_rate": 0.0005,
      "loss": 3.077425479888916,
      "step": 1691
    },
    {
      "epoch": 0.39139486467730744,
      "grad_norm": 1.135412335395813,
      "learning_rate": 0.0005,
      "loss": 2.604217052459717,
      "step": 1692
    },
    {
      "epoch": 0.3916261855193153,
      "grad_norm": 1.1300636529922485,
      "learning_rate": 0.0005,
      "loss": 3.0681562423706055,
      "step": 1693
    },
    {
      "epoch": 0.39185750636132316,
      "grad_norm": 1.8808314800262451,
      "learning_rate": 0.0005,
      "loss": 2.82226300239563,
      "step": 1694
    },
    {
      "epoch": 0.39208882720333105,
      "grad_norm": 1.2898973226547241,
      "learning_rate": 0.0005,
      "loss": 2.9084229469299316,
      "step": 1695
    },
    {
      "epoch": 0.3923201480453389,
      "grad_norm": 1.0485836267471313,
      "learning_rate": 0.0005,
      "loss": 3.149623394012451,
      "step": 1696
    },
    {
      "epoch": 0.39255146888734677,
      "grad_norm": 1.0706853866577148,
      "learning_rate": 0.0005,
      "loss": 3.3374829292297363,
      "step": 1697
    },
    {
      "epoch": 0.3927827897293546,
      "grad_norm": 0.9999403357505798,
      "learning_rate": 0.0005,
      "loss": 2.850247383117676,
      "step": 1698
    },
    {
      "epoch": 0.3930141105713625,
      "grad_norm": 1.1436339616775513,
      "learning_rate": 0.0005,
      "loss": 3.141214609146118,
      "step": 1699
    },
    {
      "epoch": 0.39324543141337037,
      "grad_norm": 1.4545338153839111,
      "learning_rate": 0.0005,
      "loss": 3.017820358276367,
      "step": 1700
    },
    {
      "epoch": 0.3934767522553782,
      "grad_norm": 1.3282550573349,
      "learning_rate": 0.0005,
      "loss": 2.820347785949707,
      "step": 1701
    },
    {
      "epoch": 0.3937080730973861,
      "grad_norm": 1.0891824960708618,
      "learning_rate": 0.0005,
      "loss": 3.0825510025024414,
      "step": 1702
    },
    {
      "epoch": 0.3939393939393939,
      "grad_norm": 1.0001051425933838,
      "learning_rate": 0.0005,
      "loss": 3.072523593902588,
      "step": 1703
    },
    {
      "epoch": 0.3941707147814018,
      "grad_norm": 1.0364503860473633,
      "learning_rate": 0.0005,
      "loss": 3.2467586994171143,
      "step": 1704
    },
    {
      "epoch": 0.3944020356234097,
      "grad_norm": 1.2549883127212524,
      "learning_rate": 0.0005,
      "loss": 2.9801764488220215,
      "step": 1705
    },
    {
      "epoch": 0.39463335646541753,
      "grad_norm": 1.1652848720550537,
      "learning_rate": 0.0005,
      "loss": 3.403050422668457,
      "step": 1706
    },
    {
      "epoch": 0.3948646773074254,
      "grad_norm": 1.8023825883865356,
      "learning_rate": 0.0005,
      "loss": 2.6948022842407227,
      "step": 1707
    },
    {
      "epoch": 0.39509599814943325,
      "grad_norm": 1.2493935823440552,
      "learning_rate": 0.0005,
      "loss": 3.168501138687134,
      "step": 1708
    },
    {
      "epoch": 0.39532731899144113,
      "grad_norm": 1.0802021026611328,
      "learning_rate": 0.0005,
      "loss": 2.81286883354187,
      "step": 1709
    },
    {
      "epoch": 0.39555863983344897,
      "grad_norm": 1.1988332271575928,
      "learning_rate": 0.0005,
      "loss": 2.962580919265747,
      "step": 1710
    },
    {
      "epoch": 0.39578996067545685,
      "grad_norm": 1.4122978448867798,
      "learning_rate": 0.0005,
      "loss": 3.278085470199585,
      "step": 1711
    },
    {
      "epoch": 0.39602128151746474,
      "grad_norm": 1.145830512046814,
      "learning_rate": 0.0005,
      "loss": 3.1396539211273193,
      "step": 1712
    },
    {
      "epoch": 0.39625260235947257,
      "grad_norm": 1.1517049074172974,
      "learning_rate": 0.0005,
      "loss": 2.7904739379882812,
      "step": 1713
    },
    {
      "epoch": 0.39648392320148046,
      "grad_norm": 1.3454804420471191,
      "learning_rate": 0.0005,
      "loss": 3.477132797241211,
      "step": 1714
    },
    {
      "epoch": 0.3967152440434883,
      "grad_norm": 1.1856497526168823,
      "learning_rate": 0.0005,
      "loss": 2.7838642597198486,
      "step": 1715
    },
    {
      "epoch": 0.3969465648854962,
      "grad_norm": 1.1586781740188599,
      "learning_rate": 0.0005,
      "loss": 3.2114014625549316,
      "step": 1716
    },
    {
      "epoch": 0.39717788572750407,
      "grad_norm": 2.343670606613159,
      "learning_rate": 0.0005,
      "loss": 2.6191177368164062,
      "step": 1717
    },
    {
      "epoch": 0.3974092065695119,
      "grad_norm": 1.0304070711135864,
      "learning_rate": 0.0005,
      "loss": 3.0797810554504395,
      "step": 1718
    },
    {
      "epoch": 0.3976405274115198,
      "grad_norm": 1.089655876159668,
      "learning_rate": 0.0005,
      "loss": 2.983386993408203,
      "step": 1719
    },
    {
      "epoch": 0.3978718482535276,
      "grad_norm": 1.3689318895339966,
      "learning_rate": 0.0005,
      "loss": 2.9243335723876953,
      "step": 1720
    },
    {
      "epoch": 0.3981031690955355,
      "grad_norm": 1.4721496105194092,
      "learning_rate": 0.0005,
      "loss": 3.0283138751983643,
      "step": 1721
    },
    {
      "epoch": 0.3983344899375434,
      "grad_norm": 1.6989079713821411,
      "learning_rate": 0.0005,
      "loss": 2.9686431884765625,
      "step": 1722
    },
    {
      "epoch": 0.3985658107795512,
      "grad_norm": 1.5694262981414795,
      "learning_rate": 0.0005,
      "loss": 2.864788055419922,
      "step": 1723
    },
    {
      "epoch": 0.3987971316215591,
      "grad_norm": 1.97017240524292,
      "learning_rate": 0.0005,
      "loss": 2.91508150100708,
      "step": 1724
    },
    {
      "epoch": 0.39902845246356694,
      "grad_norm": 1.2538509368896484,
      "learning_rate": 0.0005,
      "loss": 3.089052677154541,
      "step": 1725
    },
    {
      "epoch": 0.3992597733055748,
      "grad_norm": 0.9265540838241577,
      "learning_rate": 0.0005,
      "loss": 3.154934883117676,
      "step": 1726
    },
    {
      "epoch": 0.3994910941475827,
      "grad_norm": 1.203415036201477,
      "learning_rate": 0.0005,
      "loss": 3.2026491165161133,
      "step": 1727
    },
    {
      "epoch": 0.39972241498959055,
      "grad_norm": 1.0990556478500366,
      "learning_rate": 0.0005,
      "loss": 3.0896525382995605,
      "step": 1728
    },
    {
      "epoch": 0.39995373583159843,
      "grad_norm": 1.2292876243591309,
      "learning_rate": 0.0005,
      "loss": 2.500746726989746,
      "step": 1729
    },
    {
      "epoch": 0.40018505667360627,
      "grad_norm": 1.099335789680481,
      "learning_rate": 0.0005,
      "loss": 2.739349603652954,
      "step": 1730
    },
    {
      "epoch": 0.40041637751561415,
      "grad_norm": 1.1808431148529053,
      "learning_rate": 0.0005,
      "loss": 2.7635068893432617,
      "step": 1731
    },
    {
      "epoch": 0.40064769835762204,
      "grad_norm": 1.2843562364578247,
      "learning_rate": 0.0005,
      "loss": 2.827510356903076,
      "step": 1732
    },
    {
      "epoch": 0.40087901919962987,
      "grad_norm": 1.2356902360916138,
      "learning_rate": 0.0005,
      "loss": 2.8398149013519287,
      "step": 1733
    },
    {
      "epoch": 0.40111034004163776,
      "grad_norm": 1.129775047302246,
      "learning_rate": 0.0005,
      "loss": 3.041532516479492,
      "step": 1734
    },
    {
      "epoch": 0.4013416608836456,
      "grad_norm": 1.076898455619812,
      "learning_rate": 0.0005,
      "loss": 3.118582010269165,
      "step": 1735
    },
    {
      "epoch": 0.4015729817256535,
      "grad_norm": 1.0547558069229126,
      "learning_rate": 0.0005,
      "loss": 3.1614933013916016,
      "step": 1736
    },
    {
      "epoch": 0.40180430256766136,
      "grad_norm": 1.2193251848220825,
      "learning_rate": 0.0005,
      "loss": 3.18691349029541,
      "step": 1737
    },
    {
      "epoch": 0.4020356234096692,
      "grad_norm": 1.6131551265716553,
      "learning_rate": 0.0005,
      "loss": 3.1474907398223877,
      "step": 1738
    },
    {
      "epoch": 0.4022669442516771,
      "grad_norm": 1.1126234531402588,
      "learning_rate": 0.0005,
      "loss": 3.194960594177246,
      "step": 1739
    },
    {
      "epoch": 0.4024982650936849,
      "grad_norm": 1.0259069204330444,
      "learning_rate": 0.0005,
      "loss": 3.3009018898010254,
      "step": 1740
    },
    {
      "epoch": 0.4027295859356928,
      "grad_norm": 1.5426480770111084,
      "learning_rate": 0.0005,
      "loss": 2.772508144378662,
      "step": 1741
    },
    {
      "epoch": 0.4029609067777007,
      "grad_norm": 1.6492674350738525,
      "learning_rate": 0.0005,
      "loss": 3.2069172859191895,
      "step": 1742
    },
    {
      "epoch": 0.4031922276197085,
      "grad_norm": 1.1492254734039307,
      "learning_rate": 0.0005,
      "loss": 2.959003448486328,
      "step": 1743
    },
    {
      "epoch": 0.4034235484617164,
      "grad_norm": 1.106619119644165,
      "learning_rate": 0.0005,
      "loss": 2.938015937805176,
      "step": 1744
    },
    {
      "epoch": 0.40365486930372424,
      "grad_norm": 1.0541529655456543,
      "learning_rate": 0.0005,
      "loss": 3.1126742362976074,
      "step": 1745
    },
    {
      "epoch": 0.4038861901457321,
      "grad_norm": 1.0454052686691284,
      "learning_rate": 0.0005,
      "loss": 2.8625917434692383,
      "step": 1746
    },
    {
      "epoch": 0.40411751098774,
      "grad_norm": 1.1596519947052002,
      "learning_rate": 0.0005,
      "loss": 3.1525981426239014,
      "step": 1747
    },
    {
      "epoch": 0.40434883182974785,
      "grad_norm": 1.6356072425842285,
      "learning_rate": 0.0005,
      "loss": 2.9974303245544434,
      "step": 1748
    },
    {
      "epoch": 0.40458015267175573,
      "grad_norm": 1.0457621812820435,
      "learning_rate": 0.0005,
      "loss": 2.677905559539795,
      "step": 1749
    },
    {
      "epoch": 0.40481147351376356,
      "grad_norm": 1.7142647504806519,
      "learning_rate": 0.0005,
      "loss": 2.9570016860961914,
      "step": 1750
    },
    {
      "epoch": 0.40504279435577145,
      "grad_norm": 1.0223463773727417,
      "learning_rate": 0.0005,
      "loss": 2.8324146270751953,
      "step": 1751
    },
    {
      "epoch": 0.40527411519777934,
      "grad_norm": 1.1013010740280151,
      "learning_rate": 0.0005,
      "loss": 3.1302080154418945,
      "step": 1752
    },
    {
      "epoch": 0.40550543603978717,
      "grad_norm": 1.3169959783554077,
      "learning_rate": 0.0005,
      "loss": 2.9378604888916016,
      "step": 1753
    },
    {
      "epoch": 0.40573675688179506,
      "grad_norm": 1.4322824478149414,
      "learning_rate": 0.0005,
      "loss": 3.143839120864868,
      "step": 1754
    },
    {
      "epoch": 0.4059680777238029,
      "grad_norm": 1.2857290506362915,
      "learning_rate": 0.0005,
      "loss": 2.9642605781555176,
      "step": 1755
    },
    {
      "epoch": 0.4061993985658108,
      "grad_norm": 1.203529715538025,
      "learning_rate": 0.0005,
      "loss": 3.2017102241516113,
      "step": 1756
    },
    {
      "epoch": 0.40643071940781866,
      "grad_norm": 1.0212913751602173,
      "learning_rate": 0.0005,
      "loss": 3.1263229846954346,
      "step": 1757
    },
    {
      "epoch": 0.4066620402498265,
      "grad_norm": 1.560394525527954,
      "learning_rate": 0.0005,
      "loss": 3.1229047775268555,
      "step": 1758
    },
    {
      "epoch": 0.4068933610918344,
      "grad_norm": 1.0664572715759277,
      "learning_rate": 0.0005,
      "loss": 3.0230154991149902,
      "step": 1759
    },
    {
      "epoch": 0.4071246819338422,
      "grad_norm": 1.3709369897842407,
      "learning_rate": 0.0005,
      "loss": 3.1659114360809326,
      "step": 1760
    },
    {
      "epoch": 0.4073560027758501,
      "grad_norm": 1.0633220672607422,
      "learning_rate": 0.0005,
      "loss": 2.9556522369384766,
      "step": 1761
    },
    {
      "epoch": 0.407587323617858,
      "grad_norm": 1.399775743484497,
      "learning_rate": 0.0005,
      "loss": 3.198265314102173,
      "step": 1762
    },
    {
      "epoch": 0.4078186444598658,
      "grad_norm": 1.0623465776443481,
      "learning_rate": 0.0005,
      "loss": 2.990295886993408,
      "step": 1763
    },
    {
      "epoch": 0.4080499653018737,
      "grad_norm": 1.0072988271713257,
      "learning_rate": 0.0005,
      "loss": 2.9717206954956055,
      "step": 1764
    },
    {
      "epoch": 0.40828128614388154,
      "grad_norm": 3.667896032333374,
      "learning_rate": 0.0005,
      "loss": 2.9460208415985107,
      "step": 1765
    },
    {
      "epoch": 0.4085126069858894,
      "grad_norm": 1.0178778171539307,
      "learning_rate": 0.0005,
      "loss": 3.0941386222839355,
      "step": 1766
    },
    {
      "epoch": 0.4087439278278973,
      "grad_norm": 1.2377440929412842,
      "learning_rate": 0.0005,
      "loss": 3.0161967277526855,
      "step": 1767
    },
    {
      "epoch": 0.40897524866990514,
      "grad_norm": 1.1768524646759033,
      "learning_rate": 0.0005,
      "loss": 3.108790874481201,
      "step": 1768
    },
    {
      "epoch": 0.40920656951191303,
      "grad_norm": 1.3200794458389282,
      "learning_rate": 0.0005,
      "loss": 3.170577049255371,
      "step": 1769
    },
    {
      "epoch": 0.40943789035392086,
      "grad_norm": 1.32184636592865,
      "learning_rate": 0.0005,
      "loss": 2.982126235961914,
      "step": 1770
    },
    {
      "epoch": 0.40966921119592875,
      "grad_norm": 1.6902034282684326,
      "learning_rate": 0.0005,
      "loss": 2.9419541358947754,
      "step": 1771
    },
    {
      "epoch": 0.40990053203793664,
      "grad_norm": 1.1514171361923218,
      "learning_rate": 0.0005,
      "loss": 2.6534955501556396,
      "step": 1772
    },
    {
      "epoch": 0.41013185287994447,
      "grad_norm": 1.0311367511749268,
      "learning_rate": 0.0005,
      "loss": 3.032714366912842,
      "step": 1773
    },
    {
      "epoch": 0.41036317372195236,
      "grad_norm": 1.4718708992004395,
      "learning_rate": 0.0005,
      "loss": 2.838243007659912,
      "step": 1774
    },
    {
      "epoch": 0.4105944945639602,
      "grad_norm": 1.5322694778442383,
      "learning_rate": 0.0005,
      "loss": 3.1100528240203857,
      "step": 1775
    },
    {
      "epoch": 0.4108258154059681,
      "grad_norm": 1.1683796644210815,
      "learning_rate": 0.0005,
      "loss": 2.850245952606201,
      "step": 1776
    },
    {
      "epoch": 0.41105713624797596,
      "grad_norm": 1.3473597764968872,
      "learning_rate": 0.0005,
      "loss": 3.173161745071411,
      "step": 1777
    },
    {
      "epoch": 0.4112884570899838,
      "grad_norm": 1.2343758344650269,
      "learning_rate": 0.0005,
      "loss": 2.972083806991577,
      "step": 1778
    },
    {
      "epoch": 0.4115197779319917,
      "grad_norm": 0.973402202129364,
      "learning_rate": 0.0005,
      "loss": 2.675546169281006,
      "step": 1779
    },
    {
      "epoch": 0.4117510987739995,
      "grad_norm": 1.090754508972168,
      "learning_rate": 0.0005,
      "loss": 3.1288490295410156,
      "step": 1780
    },
    {
      "epoch": 0.4119824196160074,
      "grad_norm": 1.0868463516235352,
      "learning_rate": 0.0005,
      "loss": 2.8976798057556152,
      "step": 1781
    },
    {
      "epoch": 0.4122137404580153,
      "grad_norm": 1.1326574087142944,
      "learning_rate": 0.0005,
      "loss": 3.022094249725342,
      "step": 1782
    },
    {
      "epoch": 0.4124450613000231,
      "grad_norm": 1.1598533391952515,
      "learning_rate": 0.0005,
      "loss": 3.0355212688446045,
      "step": 1783
    },
    {
      "epoch": 0.412676382142031,
      "grad_norm": 1.215166449546814,
      "learning_rate": 0.0005,
      "loss": 2.715543270111084,
      "step": 1784
    },
    {
      "epoch": 0.41290770298403884,
      "grad_norm": 1.1077922582626343,
      "learning_rate": 0.0005,
      "loss": 3.2031874656677246,
      "step": 1785
    },
    {
      "epoch": 0.4131390238260467,
      "grad_norm": 1.0282927751541138,
      "learning_rate": 0.0005,
      "loss": 3.292854070663452,
      "step": 1786
    },
    {
      "epoch": 0.4133703446680546,
      "grad_norm": 1.0594508647918701,
      "learning_rate": 0.0005,
      "loss": 2.9196059703826904,
      "step": 1787
    },
    {
      "epoch": 0.41360166551006244,
      "grad_norm": 0.9906032085418701,
      "learning_rate": 0.0005,
      "loss": 2.8177285194396973,
      "step": 1788
    },
    {
      "epoch": 0.41383298635207033,
      "grad_norm": 1.3354897499084473,
      "learning_rate": 0.0005,
      "loss": 2.6679797172546387,
      "step": 1789
    },
    {
      "epoch": 0.41406430719407816,
      "grad_norm": 1.027504801750183,
      "learning_rate": 0.0005,
      "loss": 3.1910250186920166,
      "step": 1790
    },
    {
      "epoch": 0.41429562803608605,
      "grad_norm": 0.9538676142692566,
      "learning_rate": 0.0005,
      "loss": 3.4211273193359375,
      "step": 1791
    },
    {
      "epoch": 0.41452694887809394,
      "grad_norm": 1.1236317157745361,
      "learning_rate": 0.0005,
      "loss": 2.9596221446990967,
      "step": 1792
    },
    {
      "epoch": 0.41475826972010177,
      "grad_norm": 1.051421880722046,
      "learning_rate": 0.0005,
      "loss": 2.7609457969665527,
      "step": 1793
    },
    {
      "epoch": 0.41498959056210966,
      "grad_norm": 0.9405477643013,
      "learning_rate": 0.0005,
      "loss": 3.0061535835266113,
      "step": 1794
    },
    {
      "epoch": 0.4152209114041175,
      "grad_norm": 1.247886061668396,
      "learning_rate": 0.0005,
      "loss": 3.0713067054748535,
      "step": 1795
    },
    {
      "epoch": 0.4154522322461254,
      "grad_norm": 1.5034908056259155,
      "learning_rate": 0.0005,
      "loss": 3.289674758911133,
      "step": 1796
    },
    {
      "epoch": 0.41568355308813326,
      "grad_norm": 1.2918133735656738,
      "learning_rate": 0.0005,
      "loss": 3.3951940536499023,
      "step": 1797
    },
    {
      "epoch": 0.4159148739301411,
      "grad_norm": 1.1868888139724731,
      "learning_rate": 0.0005,
      "loss": 2.6869914531707764,
      "step": 1798
    },
    {
      "epoch": 0.416146194772149,
      "grad_norm": 1.2396066188812256,
      "learning_rate": 0.0005,
      "loss": 3.199216842651367,
      "step": 1799
    },
    {
      "epoch": 0.4163775156141568,
      "grad_norm": 1.075848937034607,
      "learning_rate": 0.0005,
      "loss": 2.866366386413574,
      "step": 1800
    },
    {
      "epoch": 0.4166088364561647,
      "grad_norm": 1.4913865327835083,
      "learning_rate": 0.0005,
      "loss": 2.9003703594207764,
      "step": 1801
    },
    {
      "epoch": 0.4168401572981726,
      "grad_norm": 1.4722028970718384,
      "learning_rate": 0.0005,
      "loss": 2.9607131481170654,
      "step": 1802
    },
    {
      "epoch": 0.4170714781401804,
      "grad_norm": 0.9768749475479126,
      "learning_rate": 0.0005,
      "loss": 2.935879707336426,
      "step": 1803
    },
    {
      "epoch": 0.4173027989821883,
      "grad_norm": 1.0282540321350098,
      "learning_rate": 0.0005,
      "loss": 3.0462443828582764,
      "step": 1804
    },
    {
      "epoch": 0.41753411982419614,
      "grad_norm": 1.1294130086898804,
      "learning_rate": 0.0005,
      "loss": 3.0063772201538086,
      "step": 1805
    },
    {
      "epoch": 0.417765440666204,
      "grad_norm": 1.0630571842193604,
      "learning_rate": 0.0005,
      "loss": 2.9361464977264404,
      "step": 1806
    },
    {
      "epoch": 0.4179967615082119,
      "grad_norm": 1.1609752178192139,
      "learning_rate": 0.0005,
      "loss": 3.131221294403076,
      "step": 1807
    },
    {
      "epoch": 0.41822808235021974,
      "grad_norm": 2.684476613998413,
      "learning_rate": 0.0005,
      "loss": 2.832404851913452,
      "step": 1808
    },
    {
      "epoch": 0.41845940319222763,
      "grad_norm": 1.2700884342193604,
      "learning_rate": 0.0005,
      "loss": 2.8456525802612305,
      "step": 1809
    },
    {
      "epoch": 0.41869072403423546,
      "grad_norm": 1.9053221940994263,
      "learning_rate": 0.0005,
      "loss": 2.9667716026306152,
      "step": 1810
    },
    {
      "epoch": 0.41892204487624335,
      "grad_norm": 1.1896969079971313,
      "learning_rate": 0.0005,
      "loss": 2.7021946907043457,
      "step": 1811
    },
    {
      "epoch": 0.41915336571825124,
      "grad_norm": 1.2003213167190552,
      "learning_rate": 0.0005,
      "loss": 3.1079587936401367,
      "step": 1812
    },
    {
      "epoch": 0.41938468656025907,
      "grad_norm": 1.1750375032424927,
      "learning_rate": 0.0005,
      "loss": 2.930210590362549,
      "step": 1813
    },
    {
      "epoch": 0.41961600740226696,
      "grad_norm": 1.3126530647277832,
      "learning_rate": 0.0005,
      "loss": 3.2203845977783203,
      "step": 1814
    },
    {
      "epoch": 0.4198473282442748,
      "grad_norm": 1.1791863441467285,
      "learning_rate": 0.0005,
      "loss": 2.8555660247802734,
      "step": 1815
    },
    {
      "epoch": 0.4200786490862827,
      "grad_norm": 1.137952446937561,
      "learning_rate": 0.0005,
      "loss": 2.916358709335327,
      "step": 1816
    },
    {
      "epoch": 0.42030996992829056,
      "grad_norm": 4.17173957824707,
      "learning_rate": 0.0005,
      "loss": 3.1258721351623535,
      "step": 1817
    },
    {
      "epoch": 0.4205412907702984,
      "grad_norm": 1.2524662017822266,
      "learning_rate": 0.0005,
      "loss": 3.0575122833251953,
      "step": 1818
    },
    {
      "epoch": 0.4207726116123063,
      "grad_norm": 1.0811100006103516,
      "learning_rate": 0.0005,
      "loss": 2.867128849029541,
      "step": 1819
    },
    {
      "epoch": 0.4210039324543141,
      "grad_norm": 1.2936583757400513,
      "learning_rate": 0.0005,
      "loss": 3.075493335723877,
      "step": 1820
    },
    {
      "epoch": 0.421235253296322,
      "grad_norm": 1.564623236656189,
      "learning_rate": 0.0005,
      "loss": 2.9910783767700195,
      "step": 1821
    },
    {
      "epoch": 0.4214665741383299,
      "grad_norm": 1.1579570770263672,
      "learning_rate": 0.0005,
      "loss": 3.063854217529297,
      "step": 1822
    },
    {
      "epoch": 0.4216978949803377,
      "grad_norm": 1.2978461980819702,
      "learning_rate": 0.0005,
      "loss": 3.1770379543304443,
      "step": 1823
    },
    {
      "epoch": 0.4219292158223456,
      "grad_norm": 1.0230000019073486,
      "learning_rate": 0.0005,
      "loss": 2.900660514831543,
      "step": 1824
    },
    {
      "epoch": 0.42216053666435344,
      "grad_norm": 1.0847312211990356,
      "learning_rate": 0.0005,
      "loss": 3.216979742050171,
      "step": 1825
    },
    {
      "epoch": 0.4223918575063613,
      "grad_norm": 1.2612217664718628,
      "learning_rate": 0.0005,
      "loss": 2.9493868350982666,
      "step": 1826
    },
    {
      "epoch": 0.4226231783483692,
      "grad_norm": 1.1240861415863037,
      "learning_rate": 0.0005,
      "loss": 2.859328269958496,
      "step": 1827
    },
    {
      "epoch": 0.42285449919037704,
      "grad_norm": 1.1346099376678467,
      "learning_rate": 0.0005,
      "loss": 2.7930116653442383,
      "step": 1828
    },
    {
      "epoch": 0.42308582003238493,
      "grad_norm": 1.2665483951568604,
      "learning_rate": 0.0005,
      "loss": 3.014063596725464,
      "step": 1829
    },
    {
      "epoch": 0.42331714087439276,
      "grad_norm": 1.0258347988128662,
      "learning_rate": 0.0005,
      "loss": 2.901217460632324,
      "step": 1830
    },
    {
      "epoch": 0.42354846171640065,
      "grad_norm": 1.1013433933258057,
      "learning_rate": 0.0005,
      "loss": 3.214179039001465,
      "step": 1831
    },
    {
      "epoch": 0.42377978255840854,
      "grad_norm": 1.6110204458236694,
      "learning_rate": 0.0005,
      "loss": 2.9952046871185303,
      "step": 1832
    },
    {
      "epoch": 0.42401110340041637,
      "grad_norm": 1.407407522201538,
      "learning_rate": 0.0005,
      "loss": 3.3980917930603027,
      "step": 1833
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 1.5567481517791748,
      "learning_rate": 0.0005,
      "loss": 3.109992504119873,
      "step": 1834
    },
    {
      "epoch": 0.4244737450844321,
      "grad_norm": 1.277703046798706,
      "learning_rate": 0.0005,
      "loss": 2.9557743072509766,
      "step": 1835
    },
    {
      "epoch": 0.42470506592644,
      "grad_norm": 1.3582415580749512,
      "learning_rate": 0.0005,
      "loss": 2.7435460090637207,
      "step": 1836
    },
    {
      "epoch": 0.42493638676844786,
      "grad_norm": 0.9592920541763306,
      "learning_rate": 0.0005,
      "loss": 2.8658053874969482,
      "step": 1837
    },
    {
      "epoch": 0.4251677076104557,
      "grad_norm": 1.4976478815078735,
      "learning_rate": 0.0005,
      "loss": 3.063749313354492,
      "step": 1838
    },
    {
      "epoch": 0.4253990284524636,
      "grad_norm": 1.1895586252212524,
      "learning_rate": 0.0005,
      "loss": 2.966069221496582,
      "step": 1839
    },
    {
      "epoch": 0.4256303492944714,
      "grad_norm": 1.5451922416687012,
      "learning_rate": 0.0005,
      "loss": 2.8164045810699463,
      "step": 1840
    },
    {
      "epoch": 0.4258616701364793,
      "grad_norm": 0.9973535537719727,
      "learning_rate": 0.0005,
      "loss": 3.097810745239258,
      "step": 1841
    },
    {
      "epoch": 0.4260929909784872,
      "grad_norm": 1.2268128395080566,
      "learning_rate": 0.0005,
      "loss": 3.2817883491516113,
      "step": 1842
    },
    {
      "epoch": 0.426324311820495,
      "grad_norm": 1.213911771774292,
      "learning_rate": 0.0005,
      "loss": 2.922227621078491,
      "step": 1843
    },
    {
      "epoch": 0.4265556326625029,
      "grad_norm": 1.231034278869629,
      "learning_rate": 0.0005,
      "loss": 2.8395910263061523,
      "step": 1844
    },
    {
      "epoch": 0.42678695350451074,
      "grad_norm": 1.137141227722168,
      "learning_rate": 0.0005,
      "loss": 2.890186309814453,
      "step": 1845
    },
    {
      "epoch": 0.4270182743465186,
      "grad_norm": 1.2389203310012817,
      "learning_rate": 0.0005,
      "loss": 2.9817159175872803,
      "step": 1846
    },
    {
      "epoch": 0.4272495951885265,
      "grad_norm": 1.2250314950942993,
      "learning_rate": 0.0005,
      "loss": 2.7721288204193115,
      "step": 1847
    },
    {
      "epoch": 0.42748091603053434,
      "grad_norm": 1.2653393745422363,
      "learning_rate": 0.0005,
      "loss": 3.1853601932525635,
      "step": 1848
    },
    {
      "epoch": 0.42771223687254223,
      "grad_norm": 1.2843852043151855,
      "learning_rate": 0.0005,
      "loss": 3.323108196258545,
      "step": 1849
    },
    {
      "epoch": 0.42794355771455006,
      "grad_norm": 0.9607119560241699,
      "learning_rate": 0.0005,
      "loss": 2.996030807495117,
      "step": 1850
    },
    {
      "epoch": 0.42817487855655795,
      "grad_norm": 1.249902606010437,
      "learning_rate": 0.0005,
      "loss": 2.8072104454040527,
      "step": 1851
    },
    {
      "epoch": 0.42840619939856583,
      "grad_norm": 1.0554200410842896,
      "learning_rate": 0.0005,
      "loss": 2.930541753768921,
      "step": 1852
    },
    {
      "epoch": 0.42863752024057367,
      "grad_norm": 1.115580677986145,
      "learning_rate": 0.0005,
      "loss": 2.835753917694092,
      "step": 1853
    },
    {
      "epoch": 0.42886884108258155,
      "grad_norm": 0.9657366275787354,
      "learning_rate": 0.0005,
      "loss": 3.4369454383850098,
      "step": 1854
    },
    {
      "epoch": 0.4291001619245894,
      "grad_norm": 1.1530736684799194,
      "learning_rate": 0.0005,
      "loss": 3.0669798851013184,
      "step": 1855
    },
    {
      "epoch": 0.42933148276659727,
      "grad_norm": 1.169672966003418,
      "learning_rate": 0.0005,
      "loss": 2.586195707321167,
      "step": 1856
    },
    {
      "epoch": 0.42956280360860516,
      "grad_norm": 1.6288901567459106,
      "learning_rate": 0.0005,
      "loss": 2.999290943145752,
      "step": 1857
    },
    {
      "epoch": 0.429794124450613,
      "grad_norm": 1.2174770832061768,
      "learning_rate": 0.0005,
      "loss": 3.19685959815979,
      "step": 1858
    },
    {
      "epoch": 0.4300254452926209,
      "grad_norm": 1.3970694541931152,
      "learning_rate": 0.0005,
      "loss": 3.0184555053710938,
      "step": 1859
    },
    {
      "epoch": 0.4302567661346287,
      "grad_norm": 1.1378686428070068,
      "learning_rate": 0.0005,
      "loss": 2.73228120803833,
      "step": 1860
    },
    {
      "epoch": 0.4304880869766366,
      "grad_norm": 1.3683992624282837,
      "learning_rate": 0.0005,
      "loss": 2.9637856483459473,
      "step": 1861
    },
    {
      "epoch": 0.4307194078186445,
      "grad_norm": 1.2545546293258667,
      "learning_rate": 0.0005,
      "loss": 3.1848747730255127,
      "step": 1862
    },
    {
      "epoch": 0.4309507286606523,
      "grad_norm": 1.714453101158142,
      "learning_rate": 0.0005,
      "loss": 2.759519577026367,
      "step": 1863
    },
    {
      "epoch": 0.4311820495026602,
      "grad_norm": 1.3659026622772217,
      "learning_rate": 0.0005,
      "loss": 2.906376361846924,
      "step": 1864
    },
    {
      "epoch": 0.43141337034466803,
      "grad_norm": 1.3071508407592773,
      "learning_rate": 0.0005,
      "loss": 2.9556572437286377,
      "step": 1865
    },
    {
      "epoch": 0.4316446911866759,
      "grad_norm": 1.3125945329666138,
      "learning_rate": 0.0005,
      "loss": 3.052356719970703,
      "step": 1866
    },
    {
      "epoch": 0.4318760120286838,
      "grad_norm": 0.9443948268890381,
      "learning_rate": 0.0005,
      "loss": 3.2061898708343506,
      "step": 1867
    },
    {
      "epoch": 0.43210733287069164,
      "grad_norm": 1.113062858581543,
      "learning_rate": 0.0005,
      "loss": 2.743974208831787,
      "step": 1868
    },
    {
      "epoch": 0.4323386537126995,
      "grad_norm": 0.9797450304031372,
      "learning_rate": 0.0005,
      "loss": 3.121920585632324,
      "step": 1869
    },
    {
      "epoch": 0.43256997455470736,
      "grad_norm": 1.217185378074646,
      "learning_rate": 0.0005,
      "loss": 3.1232900619506836,
      "step": 1870
    },
    {
      "epoch": 0.43280129539671525,
      "grad_norm": 1.1433438062667847,
      "learning_rate": 0.0005,
      "loss": 2.896543502807617,
      "step": 1871
    },
    {
      "epoch": 0.43303261623872313,
      "grad_norm": 1.1869983673095703,
      "learning_rate": 0.0005,
      "loss": 2.566087245941162,
      "step": 1872
    },
    {
      "epoch": 0.43326393708073097,
      "grad_norm": 1.0442514419555664,
      "learning_rate": 0.0005,
      "loss": 2.9126272201538086,
      "step": 1873
    },
    {
      "epoch": 0.43349525792273885,
      "grad_norm": 1.5252763032913208,
      "learning_rate": 0.0005,
      "loss": 2.834744930267334,
      "step": 1874
    },
    {
      "epoch": 0.4337265787647467,
      "grad_norm": 1.202696681022644,
      "learning_rate": 0.0005,
      "loss": 2.4757680892944336,
      "step": 1875
    },
    {
      "epoch": 0.43395789960675457,
      "grad_norm": 1.1858747005462646,
      "learning_rate": 0.0005,
      "loss": 2.904256820678711,
      "step": 1876
    },
    {
      "epoch": 0.43418922044876246,
      "grad_norm": 1.1012139320373535,
      "learning_rate": 0.0005,
      "loss": 2.9374489784240723,
      "step": 1877
    },
    {
      "epoch": 0.4344205412907703,
      "grad_norm": 1.1283950805664062,
      "learning_rate": 0.0005,
      "loss": 2.616410255432129,
      "step": 1878
    },
    {
      "epoch": 0.4346518621327782,
      "grad_norm": 1.6737542152404785,
      "learning_rate": 0.0005,
      "loss": 2.8024344444274902,
      "step": 1879
    },
    {
      "epoch": 0.434883182974786,
      "grad_norm": 1.3548734188079834,
      "learning_rate": 0.0005,
      "loss": 3.0026416778564453,
      "step": 1880
    },
    {
      "epoch": 0.4351145038167939,
      "grad_norm": 1.0683637857437134,
      "learning_rate": 0.0005,
      "loss": 2.9148106575012207,
      "step": 1881
    },
    {
      "epoch": 0.4353458246588018,
      "grad_norm": 1.0873104333877563,
      "learning_rate": 0.0005,
      "loss": 2.7511250972747803,
      "step": 1882
    },
    {
      "epoch": 0.4355771455008096,
      "grad_norm": 1.3730660676956177,
      "learning_rate": 0.0005,
      "loss": 3.051068067550659,
      "step": 1883
    },
    {
      "epoch": 0.4358084663428175,
      "grad_norm": 1.2197093963623047,
      "learning_rate": 0.0005,
      "loss": 2.8349499702453613,
      "step": 1884
    },
    {
      "epoch": 0.43603978718482533,
      "grad_norm": 1.1285412311553955,
      "learning_rate": 0.0005,
      "loss": 3.2365520000457764,
      "step": 1885
    },
    {
      "epoch": 0.4362711080268332,
      "grad_norm": 1.0170577764511108,
      "learning_rate": 0.0005,
      "loss": 3.041234016418457,
      "step": 1886
    },
    {
      "epoch": 0.4365024288688411,
      "grad_norm": 1.431491732597351,
      "learning_rate": 0.0005,
      "loss": 2.7839481830596924,
      "step": 1887
    },
    {
      "epoch": 0.43673374971084894,
      "grad_norm": 1.5622421503067017,
      "learning_rate": 0.0005,
      "loss": 2.98917818069458,
      "step": 1888
    },
    {
      "epoch": 0.4369650705528568,
      "grad_norm": 1.5894854068756104,
      "learning_rate": 0.0005,
      "loss": 2.982548236846924,
      "step": 1889
    },
    {
      "epoch": 0.43719639139486466,
      "grad_norm": 1.08502197265625,
      "learning_rate": 0.0005,
      "loss": 2.9450321197509766,
      "step": 1890
    },
    {
      "epoch": 0.43742771223687255,
      "grad_norm": 1.3733261823654175,
      "learning_rate": 0.0005,
      "loss": 3.2471938133239746,
      "step": 1891
    },
    {
      "epoch": 0.43765903307888043,
      "grad_norm": 1.8873131275177002,
      "learning_rate": 0.0005,
      "loss": 3.417184352874756,
      "step": 1892
    },
    {
      "epoch": 0.43789035392088826,
      "grad_norm": 1.6279387474060059,
      "learning_rate": 0.0005,
      "loss": 2.7442052364349365,
      "step": 1893
    },
    {
      "epoch": 0.43812167476289615,
      "grad_norm": 1.0197497606277466,
      "learning_rate": 0.0005,
      "loss": 3.046494483947754,
      "step": 1894
    },
    {
      "epoch": 0.438352995604904,
      "grad_norm": 1.253301739692688,
      "learning_rate": 0.0005,
      "loss": 3.370391845703125,
      "step": 1895
    },
    {
      "epoch": 0.43858431644691187,
      "grad_norm": 1.316086769104004,
      "learning_rate": 0.0005,
      "loss": 3.0000600814819336,
      "step": 1896
    },
    {
      "epoch": 0.43881563728891976,
      "grad_norm": 0.9430004358291626,
      "learning_rate": 0.0005,
      "loss": 2.6042561531066895,
      "step": 1897
    },
    {
      "epoch": 0.4390469581309276,
      "grad_norm": 1.7002629041671753,
      "learning_rate": 0.0005,
      "loss": 2.6902143955230713,
      "step": 1898
    },
    {
      "epoch": 0.4392782789729355,
      "grad_norm": 1.012563705444336,
      "learning_rate": 0.0005,
      "loss": 2.660365104675293,
      "step": 1899
    },
    {
      "epoch": 0.4395095998149433,
      "grad_norm": 1.0266791582107544,
      "learning_rate": 0.0005,
      "loss": 3.301248073577881,
      "step": 1900
    },
    {
      "epoch": 0.4397409206569512,
      "grad_norm": 1.6543772220611572,
      "learning_rate": 0.0005,
      "loss": 3.175307273864746,
      "step": 1901
    },
    {
      "epoch": 0.4399722414989591,
      "grad_norm": 1.1234076023101807,
      "learning_rate": 0.0005,
      "loss": 2.921900749206543,
      "step": 1902
    },
    {
      "epoch": 0.4402035623409669,
      "grad_norm": 1.0934021472930908,
      "learning_rate": 0.0005,
      "loss": 2.953101873397827,
      "step": 1903
    },
    {
      "epoch": 0.4404348831829748,
      "grad_norm": 1.1046757698059082,
      "learning_rate": 0.0005,
      "loss": 3.082581043243408,
      "step": 1904
    },
    {
      "epoch": 0.44066620402498263,
      "grad_norm": 1.1305173635482788,
      "learning_rate": 0.0005,
      "loss": 3.0225138664245605,
      "step": 1905
    },
    {
      "epoch": 0.4408975248669905,
      "grad_norm": 2.094237804412842,
      "learning_rate": 0.0005,
      "loss": 3.125181198120117,
      "step": 1906
    },
    {
      "epoch": 0.4411288457089984,
      "grad_norm": 0.9760357737541199,
      "learning_rate": 0.0005,
      "loss": 2.8600258827209473,
      "step": 1907
    },
    {
      "epoch": 0.44136016655100624,
      "grad_norm": 1.5851212739944458,
      "learning_rate": 0.0005,
      "loss": 3.305023431777954,
      "step": 1908
    },
    {
      "epoch": 0.4415914873930141,
      "grad_norm": 1.0617510080337524,
      "learning_rate": 0.0005,
      "loss": 2.649723529815674,
      "step": 1909
    },
    {
      "epoch": 0.44182280823502196,
      "grad_norm": 1.8958123922348022,
      "learning_rate": 0.0005,
      "loss": 3.368607759475708,
      "step": 1910
    },
    {
      "epoch": 0.44205412907702984,
      "grad_norm": 1.2126392126083374,
      "learning_rate": 0.0005,
      "loss": 3.1298789978027344,
      "step": 1911
    },
    {
      "epoch": 0.44228544991903773,
      "grad_norm": 1.1427172422409058,
      "learning_rate": 0.0005,
      "loss": 2.9962661266326904,
      "step": 1912
    },
    {
      "epoch": 0.44251677076104556,
      "grad_norm": 1.1838701963424683,
      "learning_rate": 0.0005,
      "loss": 2.7442469596862793,
      "step": 1913
    },
    {
      "epoch": 0.44274809160305345,
      "grad_norm": 0.9613898396492004,
      "learning_rate": 0.0005,
      "loss": 3.0720741748809814,
      "step": 1914
    },
    {
      "epoch": 0.4429794124450613,
      "grad_norm": 1.013484239578247,
      "learning_rate": 0.0005,
      "loss": 2.869133472442627,
      "step": 1915
    },
    {
      "epoch": 0.44321073328706917,
      "grad_norm": 1.177077054977417,
      "learning_rate": 0.0005,
      "loss": 3.0640010833740234,
      "step": 1916
    },
    {
      "epoch": 0.44344205412907706,
      "grad_norm": 1.524443507194519,
      "learning_rate": 0.0005,
      "loss": 3.000868797302246,
      "step": 1917
    },
    {
      "epoch": 0.4436733749710849,
      "grad_norm": 1.297560214996338,
      "learning_rate": 0.0005,
      "loss": 2.758401393890381,
      "step": 1918
    },
    {
      "epoch": 0.4439046958130928,
      "grad_norm": 1.5744812488555908,
      "learning_rate": 0.0005,
      "loss": 2.925865411758423,
      "step": 1919
    },
    {
      "epoch": 0.4441360166551006,
      "grad_norm": 1.09386146068573,
      "learning_rate": 0.0005,
      "loss": 2.943333625793457,
      "step": 1920
    },
    {
      "epoch": 0.4443673374971085,
      "grad_norm": 0.9905383586883545,
      "learning_rate": 0.0005,
      "loss": 2.6349234580993652,
      "step": 1921
    },
    {
      "epoch": 0.4445986583391164,
      "grad_norm": 1.0227901935577393,
      "learning_rate": 0.0005,
      "loss": 2.6970717906951904,
      "step": 1922
    },
    {
      "epoch": 0.4448299791811242,
      "grad_norm": 1.0486366748809814,
      "learning_rate": 0.0005,
      "loss": 3.1007466316223145,
      "step": 1923
    },
    {
      "epoch": 0.4450613000231321,
      "grad_norm": 1.1590653657913208,
      "learning_rate": 0.0005,
      "loss": 2.941082715988159,
      "step": 1924
    },
    {
      "epoch": 0.44529262086513993,
      "grad_norm": 0.9851685166358948,
      "learning_rate": 0.0005,
      "loss": 2.930830955505371,
      "step": 1925
    },
    {
      "epoch": 0.4455239417071478,
      "grad_norm": 1.3530393838882446,
      "learning_rate": 0.0005,
      "loss": 3.0949835777282715,
      "step": 1926
    },
    {
      "epoch": 0.4457552625491557,
      "grad_norm": 0.9601035118103027,
      "learning_rate": 0.0005,
      "loss": 2.964106798171997,
      "step": 1927
    },
    {
      "epoch": 0.44598658339116354,
      "grad_norm": 1.6062582731246948,
      "learning_rate": 0.0005,
      "loss": 2.8902037143707275,
      "step": 1928
    },
    {
      "epoch": 0.4462179042331714,
      "grad_norm": 1.129815697669983,
      "learning_rate": 0.0005,
      "loss": 3.2455062866210938,
      "step": 1929
    },
    {
      "epoch": 0.44644922507517926,
      "grad_norm": 1.1285823583602905,
      "learning_rate": 0.0005,
      "loss": 2.9233298301696777,
      "step": 1930
    },
    {
      "epoch": 0.44668054591718714,
      "grad_norm": 0.9676403403282166,
      "learning_rate": 0.0005,
      "loss": 3.4582316875457764,
      "step": 1931
    },
    {
      "epoch": 0.44691186675919503,
      "grad_norm": 1.0103694200515747,
      "learning_rate": 0.0005,
      "loss": 3.0649571418762207,
      "step": 1932
    },
    {
      "epoch": 0.44714318760120286,
      "grad_norm": 1.4295060634613037,
      "learning_rate": 0.0005,
      "loss": 3.0681114196777344,
      "step": 1933
    },
    {
      "epoch": 0.44737450844321075,
      "grad_norm": 1.0768355131149292,
      "learning_rate": 0.0005,
      "loss": 2.7707138061523438,
      "step": 1934
    },
    {
      "epoch": 0.4476058292852186,
      "grad_norm": 1.2843633890151978,
      "learning_rate": 0.0005,
      "loss": 2.8757355213165283,
      "step": 1935
    },
    {
      "epoch": 0.44783715012722647,
      "grad_norm": 0.9453499913215637,
      "learning_rate": 0.0005,
      "loss": 2.9213504791259766,
      "step": 1936
    },
    {
      "epoch": 0.4480684709692343,
      "grad_norm": 1.2038161754608154,
      "learning_rate": 0.0005,
      "loss": 2.799403429031372,
      "step": 1937
    },
    {
      "epoch": 0.4482997918112422,
      "grad_norm": 1.0063594579696655,
      "learning_rate": 0.0005,
      "loss": 3.4018394947052,
      "step": 1938
    },
    {
      "epoch": 0.4485311126532501,
      "grad_norm": 1.2253272533416748,
      "learning_rate": 0.0005,
      "loss": 2.788626194000244,
      "step": 1939
    },
    {
      "epoch": 0.4487624334952579,
      "grad_norm": 1.5742402076721191,
      "learning_rate": 0.0005,
      "loss": 2.9076268672943115,
      "step": 1940
    },
    {
      "epoch": 0.4489937543372658,
      "grad_norm": 1.2423608303070068,
      "learning_rate": 0.0005,
      "loss": 2.87022066116333,
      "step": 1941
    },
    {
      "epoch": 0.4492250751792736,
      "grad_norm": 1.5036786794662476,
      "learning_rate": 0.0005,
      "loss": 2.885946750640869,
      "step": 1942
    },
    {
      "epoch": 0.4494563960212815,
      "grad_norm": 1.3141177892684937,
      "learning_rate": 0.0005,
      "loss": 2.9514503479003906,
      "step": 1943
    },
    {
      "epoch": 0.4496877168632894,
      "grad_norm": 1.6568576097488403,
      "learning_rate": 0.0005,
      "loss": 2.692945957183838,
      "step": 1944
    },
    {
      "epoch": 0.44991903770529723,
      "grad_norm": 1.1282480955123901,
      "learning_rate": 0.0005,
      "loss": 2.9836034774780273,
      "step": 1945
    },
    {
      "epoch": 0.4501503585473051,
      "grad_norm": 1.2811155319213867,
      "learning_rate": 0.0005,
      "loss": 2.861119270324707,
      "step": 1946
    },
    {
      "epoch": 0.45038167938931295,
      "grad_norm": 1.7692680358886719,
      "learning_rate": 0.0005,
      "loss": 2.9039313793182373,
      "step": 1947
    },
    {
      "epoch": 0.45061300023132084,
      "grad_norm": 1.4904037714004517,
      "learning_rate": 0.0005,
      "loss": 3.1200084686279297,
      "step": 1948
    },
    {
      "epoch": 0.4508443210733287,
      "grad_norm": 1.5671700239181519,
      "learning_rate": 0.0005,
      "loss": 3.0014071464538574,
      "step": 1949
    },
    {
      "epoch": 0.45107564191533656,
      "grad_norm": 1.5059162378311157,
      "learning_rate": 0.0005,
      "loss": 2.920409679412842,
      "step": 1950
    },
    {
      "epoch": 0.45130696275734444,
      "grad_norm": 1.2761030197143555,
      "learning_rate": 0.0005,
      "loss": 3.1876564025878906,
      "step": 1951
    },
    {
      "epoch": 0.4515382835993523,
      "grad_norm": 4.878520965576172,
      "learning_rate": 0.0005,
      "loss": 3.300065517425537,
      "step": 1952
    },
    {
      "epoch": 0.45176960444136016,
      "grad_norm": 1.4228312969207764,
      "learning_rate": 0.0005,
      "loss": 3.1167402267456055,
      "step": 1953
    },
    {
      "epoch": 0.45200092528336805,
      "grad_norm": 1.7012250423431396,
      "learning_rate": 0.0005,
      "loss": 2.692305326461792,
      "step": 1954
    },
    {
      "epoch": 0.4522322461253759,
      "grad_norm": 1.4080020189285278,
      "learning_rate": 0.0005,
      "loss": 3.072844982147217,
      "step": 1955
    },
    {
      "epoch": 0.45246356696738377,
      "grad_norm": 1.6862938404083252,
      "learning_rate": 0.0005,
      "loss": 3.1924877166748047,
      "step": 1956
    },
    {
      "epoch": 0.4526948878093916,
      "grad_norm": 19.97258186340332,
      "learning_rate": 0.0005,
      "loss": 2.974715232849121,
      "step": 1957
    },
    {
      "epoch": 0.4529262086513995,
      "grad_norm": 1.1131315231323242,
      "learning_rate": 0.0005,
      "loss": 3.151791572570801,
      "step": 1958
    },
    {
      "epoch": 0.4531575294934074,
      "grad_norm": 1.489538311958313,
      "learning_rate": 0.0005,
      "loss": 3.3656015396118164,
      "step": 1959
    },
    {
      "epoch": 0.4533888503354152,
      "grad_norm": 1.6445261240005493,
      "learning_rate": 0.0005,
      "loss": 3.3351831436157227,
      "step": 1960
    },
    {
      "epoch": 0.4536201711774231,
      "grad_norm": 1.0305484533309937,
      "learning_rate": 0.0005,
      "loss": 2.6937291622161865,
      "step": 1961
    },
    {
      "epoch": 0.4538514920194309,
      "grad_norm": 0.9814361333847046,
      "learning_rate": 0.0005,
      "loss": 2.8410418033599854,
      "step": 1962
    },
    {
      "epoch": 0.4540828128614388,
      "grad_norm": 1.0168877840042114,
      "learning_rate": 0.0005,
      "loss": 3.0281782150268555,
      "step": 1963
    },
    {
      "epoch": 0.4543141337034467,
      "grad_norm": 1.1584216356277466,
      "learning_rate": 0.0005,
      "loss": 2.8055419921875,
      "step": 1964
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 1.033128023147583,
      "learning_rate": 0.0005,
      "loss": 2.7840046882629395,
      "step": 1965
    },
    {
      "epoch": 0.4547767753874624,
      "grad_norm": 1.3141881227493286,
      "learning_rate": 0.0005,
      "loss": 3.0874977111816406,
      "step": 1966
    },
    {
      "epoch": 0.45500809622947025,
      "grad_norm": 1.391716718673706,
      "learning_rate": 0.0005,
      "loss": 2.7707297801971436,
      "step": 1967
    },
    {
      "epoch": 0.45523941707147814,
      "grad_norm": 1.1228361129760742,
      "learning_rate": 0.0005,
      "loss": 2.991917371749878,
      "step": 1968
    },
    {
      "epoch": 0.455470737913486,
      "grad_norm": 1.2061949968338013,
      "learning_rate": 0.0005,
      "loss": 2.589421510696411,
      "step": 1969
    },
    {
      "epoch": 0.45570205875549386,
      "grad_norm": 1.0526255369186401,
      "learning_rate": 0.0005,
      "loss": 2.877716541290283,
      "step": 1970
    },
    {
      "epoch": 0.45593337959750174,
      "grad_norm": 1.410862684249878,
      "learning_rate": 0.0005,
      "loss": 2.7860043048858643,
      "step": 1971
    },
    {
      "epoch": 0.4561647004395096,
      "grad_norm": 1.1511516571044922,
      "learning_rate": 0.0005,
      "loss": 2.999361515045166,
      "step": 1972
    },
    {
      "epoch": 0.45639602128151746,
      "grad_norm": 1.1379835605621338,
      "learning_rate": 0.0005,
      "loss": 2.942530632019043,
      "step": 1973
    },
    {
      "epoch": 0.45662734212352535,
      "grad_norm": 1.0045441389083862,
      "learning_rate": 0.0005,
      "loss": 2.7975587844848633,
      "step": 1974
    },
    {
      "epoch": 0.4568586629655332,
      "grad_norm": 1.3116114139556885,
      "learning_rate": 0.0005,
      "loss": 2.8813538551330566,
      "step": 1975
    },
    {
      "epoch": 0.45708998380754107,
      "grad_norm": 1.353808045387268,
      "learning_rate": 0.0005,
      "loss": 2.814310073852539,
      "step": 1976
    },
    {
      "epoch": 0.4573213046495489,
      "grad_norm": 1.4258146286010742,
      "learning_rate": 0.0005,
      "loss": 3.197629690170288,
      "step": 1977
    },
    {
      "epoch": 0.4575526254915568,
      "grad_norm": 1.281784176826477,
      "learning_rate": 0.0005,
      "loss": 2.943809747695923,
      "step": 1978
    },
    {
      "epoch": 0.4577839463335647,
      "grad_norm": 1.2607425451278687,
      "learning_rate": 0.0005,
      "loss": 3.0277342796325684,
      "step": 1979
    },
    {
      "epoch": 0.4580152671755725,
      "grad_norm": 1.4399374723434448,
      "learning_rate": 0.0005,
      "loss": 3.3738045692443848,
      "step": 1980
    },
    {
      "epoch": 0.4582465880175804,
      "grad_norm": 1.2400449514389038,
      "learning_rate": 0.0005,
      "loss": 2.6898539066314697,
      "step": 1981
    },
    {
      "epoch": 0.4584779088595882,
      "grad_norm": 1.1016747951507568,
      "learning_rate": 0.0005,
      "loss": 3.217413902282715,
      "step": 1982
    },
    {
      "epoch": 0.4587092297015961,
      "grad_norm": 1.1827237606048584,
      "learning_rate": 0.0005,
      "loss": 3.0876166820526123,
      "step": 1983
    },
    {
      "epoch": 0.458940550543604,
      "grad_norm": 1.3531177043914795,
      "learning_rate": 0.0005,
      "loss": 2.7696943283081055,
      "step": 1984
    },
    {
      "epoch": 0.45917187138561183,
      "grad_norm": 1.657219648361206,
      "learning_rate": 0.0005,
      "loss": 3.0684542655944824,
      "step": 1985
    },
    {
      "epoch": 0.4594031922276197,
      "grad_norm": 1.111283302307129,
      "learning_rate": 0.0005,
      "loss": 2.707451820373535,
      "step": 1986
    },
    {
      "epoch": 0.45963451306962755,
      "grad_norm": 1.3112314939498901,
      "learning_rate": 0.0005,
      "loss": 2.5050442218780518,
      "step": 1987
    },
    {
      "epoch": 0.45986583391163544,
      "grad_norm": 1.2967215776443481,
      "learning_rate": 0.0005,
      "loss": 2.7490365505218506,
      "step": 1988
    },
    {
      "epoch": 0.4600971547536433,
      "grad_norm": 1.011375069618225,
      "learning_rate": 0.0005,
      "loss": 2.9722297191619873,
      "step": 1989
    },
    {
      "epoch": 0.46032847559565115,
      "grad_norm": 1.2982717752456665,
      "learning_rate": 0.0005,
      "loss": 2.9576754570007324,
      "step": 1990
    },
    {
      "epoch": 0.46055979643765904,
      "grad_norm": 1.1064856052398682,
      "learning_rate": 0.0005,
      "loss": 3.029916286468506,
      "step": 1991
    },
    {
      "epoch": 0.4607911172796669,
      "grad_norm": 1.222534418106079,
      "learning_rate": 0.0005,
      "loss": 2.9185829162597656,
      "step": 1992
    },
    {
      "epoch": 0.46102243812167476,
      "grad_norm": 1.0791336297988892,
      "learning_rate": 0.0005,
      "loss": 3.082594871520996,
      "step": 1993
    },
    {
      "epoch": 0.46125375896368265,
      "grad_norm": 1.024559497833252,
      "learning_rate": 0.0005,
      "loss": 2.825329303741455,
      "step": 1994
    },
    {
      "epoch": 0.4614850798056905,
      "grad_norm": 1.1969221830368042,
      "learning_rate": 0.0005,
      "loss": 2.7330780029296875,
      "step": 1995
    },
    {
      "epoch": 0.46171640064769837,
      "grad_norm": 1.0629560947418213,
      "learning_rate": 0.0005,
      "loss": 2.8715322017669678,
      "step": 1996
    },
    {
      "epoch": 0.4619477214897062,
      "grad_norm": 1.3997224569320679,
      "learning_rate": 0.0005,
      "loss": 2.9726808071136475,
      "step": 1997
    },
    {
      "epoch": 0.4621790423317141,
      "grad_norm": 1.2965781688690186,
      "learning_rate": 0.0005,
      "loss": 3.225109100341797,
      "step": 1998
    },
    {
      "epoch": 0.46241036317372197,
      "grad_norm": 1.1486611366271973,
      "learning_rate": 0.0005,
      "loss": 2.7965526580810547,
      "step": 1999
    },
    {
      "epoch": 0.4626416840157298,
      "grad_norm": 0.9737616181373596,
      "learning_rate": 0.0005,
      "loss": 2.9607818126678467,
      "step": 2000
    },
    {
      "epoch": 0.4628730048577377,
      "grad_norm": 1.126192331314087,
      "learning_rate": 0.0005,
      "loss": 2.670882225036621,
      "step": 2001
    },
    {
      "epoch": 0.4631043256997455,
      "grad_norm": 0.8735986351966858,
      "learning_rate": 0.0005,
      "loss": 2.5953168869018555,
      "step": 2002
    },
    {
      "epoch": 0.4633356465417534,
      "grad_norm": 0.8824494481086731,
      "learning_rate": 0.0005,
      "loss": 3.1221747398376465,
      "step": 2003
    },
    {
      "epoch": 0.4635669673837613,
      "grad_norm": 1.1453927755355835,
      "learning_rate": 0.0005,
      "loss": 2.77219820022583,
      "step": 2004
    },
    {
      "epoch": 0.46379828822576913,
      "grad_norm": 1.1881400346755981,
      "learning_rate": 0.0005,
      "loss": 2.766996383666992,
      "step": 2005
    },
    {
      "epoch": 0.464029609067777,
      "grad_norm": 1.1225436925888062,
      "learning_rate": 0.0005,
      "loss": 3.0918548107147217,
      "step": 2006
    },
    {
      "epoch": 0.46426092990978485,
      "grad_norm": 1.6132302284240723,
      "learning_rate": 0.0005,
      "loss": 2.9135921001434326,
      "step": 2007
    },
    {
      "epoch": 0.46449225075179273,
      "grad_norm": 1.769230842590332,
      "learning_rate": 0.0005,
      "loss": 3.2128005027770996,
      "step": 2008
    },
    {
      "epoch": 0.4647235715938006,
      "grad_norm": 1.444741129875183,
      "learning_rate": 0.0005,
      "loss": 2.7544002532958984,
      "step": 2009
    },
    {
      "epoch": 0.46495489243580845,
      "grad_norm": 1.7592359781265259,
      "learning_rate": 0.0005,
      "loss": 2.809420585632324,
      "step": 2010
    },
    {
      "epoch": 0.46518621327781634,
      "grad_norm": 1.619118332862854,
      "learning_rate": 0.0005,
      "loss": 2.7570555210113525,
      "step": 2011
    },
    {
      "epoch": 0.46541753411982417,
      "grad_norm": 1.3228341341018677,
      "learning_rate": 0.0005,
      "loss": 2.9462928771972656,
      "step": 2012
    },
    {
      "epoch": 0.46564885496183206,
      "grad_norm": 1.5242923498153687,
      "learning_rate": 0.0005,
      "loss": 2.856428384780884,
      "step": 2013
    },
    {
      "epoch": 0.46588017580383995,
      "grad_norm": 1.1026419401168823,
      "learning_rate": 0.0005,
      "loss": 3.002246141433716,
      "step": 2014
    },
    {
      "epoch": 0.4661114966458478,
      "grad_norm": 2.8118133544921875,
      "learning_rate": 0.0005,
      "loss": 2.700284957885742,
      "step": 2015
    },
    {
      "epoch": 0.46634281748785567,
      "grad_norm": 1.0255208015441895,
      "learning_rate": 0.0005,
      "loss": 3.077423572540283,
      "step": 2016
    },
    {
      "epoch": 0.4665741383298635,
      "grad_norm": 1.2800620794296265,
      "learning_rate": 0.0005,
      "loss": 2.9688315391540527,
      "step": 2017
    },
    {
      "epoch": 0.4668054591718714,
      "grad_norm": 1.4125064611434937,
      "learning_rate": 0.0005,
      "loss": 2.500516891479492,
      "step": 2018
    },
    {
      "epoch": 0.46703678001387927,
      "grad_norm": 1.1103630065917969,
      "learning_rate": 0.0005,
      "loss": 2.581714630126953,
      "step": 2019
    },
    {
      "epoch": 0.4672681008558871,
      "grad_norm": 1.0355638265609741,
      "learning_rate": 0.0005,
      "loss": 2.9184937477111816,
      "step": 2020
    },
    {
      "epoch": 0.467499421697895,
      "grad_norm": 1.0110266208648682,
      "learning_rate": 0.0005,
      "loss": 2.9992711544036865,
      "step": 2021
    },
    {
      "epoch": 0.4677307425399028,
      "grad_norm": 1.1157811880111694,
      "learning_rate": 0.0005,
      "loss": 2.3466897010803223,
      "step": 2022
    },
    {
      "epoch": 0.4679620633819107,
      "grad_norm": 1.617424488067627,
      "learning_rate": 0.0005,
      "loss": 2.7748453617095947,
      "step": 2023
    },
    {
      "epoch": 0.4681933842239186,
      "grad_norm": 1.1091361045837402,
      "learning_rate": 0.0005,
      "loss": 2.972975730895996,
      "step": 2024
    },
    {
      "epoch": 0.46842470506592643,
      "grad_norm": 1.2415626049041748,
      "learning_rate": 0.0005,
      "loss": 2.9622104167938232,
      "step": 2025
    },
    {
      "epoch": 0.4686560259079343,
      "grad_norm": 1.1710410118103027,
      "learning_rate": 0.0005,
      "loss": 3.2719292640686035,
      "step": 2026
    },
    {
      "epoch": 0.46888734674994215,
      "grad_norm": 4.483115196228027,
      "learning_rate": 0.0005,
      "loss": 3.2852962017059326,
      "step": 2027
    },
    {
      "epoch": 0.46911866759195003,
      "grad_norm": 0.9667967557907104,
      "learning_rate": 0.0005,
      "loss": 2.9740986824035645,
      "step": 2028
    },
    {
      "epoch": 0.4693499884339579,
      "grad_norm": 1.487488865852356,
      "learning_rate": 0.0005,
      "loss": 2.772350311279297,
      "step": 2029
    },
    {
      "epoch": 0.46958130927596575,
      "grad_norm": 1.1804718971252441,
      "learning_rate": 0.0005,
      "loss": 2.468061685562134,
      "step": 2030
    },
    {
      "epoch": 0.46981263011797364,
      "grad_norm": 1.033950686454773,
      "learning_rate": 0.0005,
      "loss": 2.7585225105285645,
      "step": 2031
    },
    {
      "epoch": 0.47004395095998147,
      "grad_norm": 1.0793988704681396,
      "learning_rate": 0.0005,
      "loss": 3.241961717605591,
      "step": 2032
    },
    {
      "epoch": 0.47027527180198936,
      "grad_norm": 0.8944147825241089,
      "learning_rate": 0.0005,
      "loss": 3.0454957485198975,
      "step": 2033
    },
    {
      "epoch": 0.47050659264399725,
      "grad_norm": 1.0931792259216309,
      "learning_rate": 0.0005,
      "loss": 3.1887729167938232,
      "step": 2034
    },
    {
      "epoch": 0.4707379134860051,
      "grad_norm": 1.2382938861846924,
      "learning_rate": 0.0005,
      "loss": 3.037771224975586,
      "step": 2035
    },
    {
      "epoch": 0.47096923432801296,
      "grad_norm": 1.1009160280227661,
      "learning_rate": 0.0005,
      "loss": 2.8057403564453125,
      "step": 2036
    },
    {
      "epoch": 0.4712005551700208,
      "grad_norm": 22.27155876159668,
      "learning_rate": 0.0005,
      "loss": 2.7956693172454834,
      "step": 2037
    },
    {
      "epoch": 0.4714318760120287,
      "grad_norm": 2.289951801300049,
      "learning_rate": 0.0005,
      "loss": 2.9488697052001953,
      "step": 2038
    },
    {
      "epoch": 0.47166319685403657,
      "grad_norm": 1.0880520343780518,
      "learning_rate": 0.0005,
      "loss": 2.2653117179870605,
      "step": 2039
    },
    {
      "epoch": 0.4718945176960444,
      "grad_norm": 1.011348843574524,
      "learning_rate": 0.0005,
      "loss": 2.835038661956787,
      "step": 2040
    },
    {
      "epoch": 0.4721258385380523,
      "grad_norm": 1.5956063270568848,
      "learning_rate": 0.0005,
      "loss": 2.963616371154785,
      "step": 2041
    },
    {
      "epoch": 0.4723571593800601,
      "grad_norm": 1.499541163444519,
      "learning_rate": 0.0005,
      "loss": 3.3996596336364746,
      "step": 2042
    },
    {
      "epoch": 0.472588480222068,
      "grad_norm": 2.9901087284088135,
      "learning_rate": 0.0005,
      "loss": 3.122716188430786,
      "step": 2043
    },
    {
      "epoch": 0.4728198010640759,
      "grad_norm": 1.7061412334442139,
      "learning_rate": 0.0005,
      "loss": 3.2847321033477783,
      "step": 2044
    },
    {
      "epoch": 0.4730511219060837,
      "grad_norm": 2.2580630779266357,
      "learning_rate": 0.0005,
      "loss": 2.449655294418335,
      "step": 2045
    },
    {
      "epoch": 0.4732824427480916,
      "grad_norm": 1.5530296564102173,
      "learning_rate": 0.0005,
      "loss": 2.921494245529175,
      "step": 2046
    },
    {
      "epoch": 0.47351376359009945,
      "grad_norm": 1.4544548988342285,
      "learning_rate": 0.0005,
      "loss": 3.1381521224975586,
      "step": 2047
    },
    {
      "epoch": 0.47374508443210733,
      "grad_norm": 1.2953139543533325,
      "learning_rate": 0.0005,
      "loss": 3.1688232421875,
      "step": 2048
    },
    {
      "epoch": 0.4739764052741152,
      "grad_norm": 1.5132946968078613,
      "learning_rate": 0.0005,
      "loss": 2.916771650314331,
      "step": 2049
    },
    {
      "epoch": 0.47420772611612305,
      "grad_norm": 1.2658182382583618,
      "learning_rate": 0.0005,
      "loss": 2.8246161937713623,
      "step": 2050
    },
    {
      "epoch": 0.47443904695813094,
      "grad_norm": 1.493757963180542,
      "learning_rate": 0.0005,
      "loss": 2.751248598098755,
      "step": 2051
    },
    {
      "epoch": 0.47467036780013877,
      "grad_norm": 1.1629185676574707,
      "learning_rate": 0.0005,
      "loss": 2.93741512298584,
      "step": 2052
    },
    {
      "epoch": 0.47490168864214666,
      "grad_norm": 3.554882287979126,
      "learning_rate": 0.0005,
      "loss": 2.9729557037353516,
      "step": 2053
    },
    {
      "epoch": 0.47513300948415454,
      "grad_norm": 1.7087420225143433,
      "learning_rate": 0.0005,
      "loss": 2.655716896057129,
      "step": 2054
    },
    {
      "epoch": 0.4753643303261624,
      "grad_norm": 1.1642475128173828,
      "learning_rate": 0.0005,
      "loss": 2.7945213317871094,
      "step": 2055
    },
    {
      "epoch": 0.47559565116817026,
      "grad_norm": 1.0696383714675903,
      "learning_rate": 0.0005,
      "loss": 2.663572311401367,
      "step": 2056
    },
    {
      "epoch": 0.4758269720101781,
      "grad_norm": 1.329378366470337,
      "learning_rate": 0.0005,
      "loss": 2.837507724761963,
      "step": 2057
    },
    {
      "epoch": 0.476058292852186,
      "grad_norm": 1.1134603023529053,
      "learning_rate": 0.0005,
      "loss": 2.8556149005889893,
      "step": 2058
    },
    {
      "epoch": 0.47628961369419387,
      "grad_norm": 1.290236473083496,
      "learning_rate": 0.0005,
      "loss": 2.7974820137023926,
      "step": 2059
    },
    {
      "epoch": 0.4765209345362017,
      "grad_norm": 1.7365562915802002,
      "learning_rate": 0.0005,
      "loss": 2.5990347862243652,
      "step": 2060
    },
    {
      "epoch": 0.4767522553782096,
      "grad_norm": 0.9864060878753662,
      "learning_rate": 0.0005,
      "loss": 2.9332432746887207,
      "step": 2061
    },
    {
      "epoch": 0.4769835762202174,
      "grad_norm": 1.1828892230987549,
      "learning_rate": 0.0005,
      "loss": 2.5993399620056152,
      "step": 2062
    },
    {
      "epoch": 0.4772148970622253,
      "grad_norm": 0.9716320037841797,
      "learning_rate": 0.0005,
      "loss": 2.7861952781677246,
      "step": 2063
    },
    {
      "epoch": 0.4774462179042332,
      "grad_norm": 1.2337440252304077,
      "learning_rate": 0.0005,
      "loss": 3.182185173034668,
      "step": 2064
    },
    {
      "epoch": 0.477677538746241,
      "grad_norm": 0.8843871355056763,
      "learning_rate": 0.0005,
      "loss": 2.585282802581787,
      "step": 2065
    },
    {
      "epoch": 0.4779088595882489,
      "grad_norm": 1.0836907625198364,
      "learning_rate": 0.0005,
      "loss": 3.0417232513427734,
      "step": 2066
    },
    {
      "epoch": 0.47814018043025674,
      "grad_norm": 1.1563645601272583,
      "learning_rate": 0.0005,
      "loss": 3.1799862384796143,
      "step": 2067
    },
    {
      "epoch": 0.47837150127226463,
      "grad_norm": 1.102004885673523,
      "learning_rate": 0.0005,
      "loss": 2.882190465927124,
      "step": 2068
    },
    {
      "epoch": 0.4786028221142725,
      "grad_norm": 1.0996390581130981,
      "learning_rate": 0.0005,
      "loss": 2.6778054237365723,
      "step": 2069
    },
    {
      "epoch": 0.47883414295628035,
      "grad_norm": 0.9827228784561157,
      "learning_rate": 0.0005,
      "loss": 3.1683225631713867,
      "step": 2070
    },
    {
      "epoch": 0.47906546379828824,
      "grad_norm": 1.1089471578598022,
      "learning_rate": 0.0005,
      "loss": 3.017275810241699,
      "step": 2071
    },
    {
      "epoch": 0.47929678464029607,
      "grad_norm": 1.0876867771148682,
      "learning_rate": 0.0005,
      "loss": 2.8270370960235596,
      "step": 2072
    },
    {
      "epoch": 0.47952810548230396,
      "grad_norm": 1.074882984161377,
      "learning_rate": 0.0005,
      "loss": 2.8716320991516113,
      "step": 2073
    },
    {
      "epoch": 0.47975942632431184,
      "grad_norm": 1.1471118927001953,
      "learning_rate": 0.0005,
      "loss": 2.957667350769043,
      "step": 2074
    },
    {
      "epoch": 0.4799907471663197,
      "grad_norm": 1.350597858428955,
      "learning_rate": 0.0005,
      "loss": 2.9280128479003906,
      "step": 2075
    },
    {
      "epoch": 0.48022206800832756,
      "grad_norm": 1.4171125888824463,
      "learning_rate": 0.0005,
      "loss": 2.88076114654541,
      "step": 2076
    },
    {
      "epoch": 0.4804533888503354,
      "grad_norm": 1.3930964469909668,
      "learning_rate": 0.0005,
      "loss": 2.888070821762085,
      "step": 2077
    },
    {
      "epoch": 0.4806847096923433,
      "grad_norm": 1.3202147483825684,
      "learning_rate": 0.0005,
      "loss": 3.0904815196990967,
      "step": 2078
    },
    {
      "epoch": 0.48091603053435117,
      "grad_norm": 1.7260475158691406,
      "learning_rate": 0.0005,
      "loss": 2.9412167072296143,
      "step": 2079
    },
    {
      "epoch": 0.481147351376359,
      "grad_norm": 1.6520347595214844,
      "learning_rate": 0.0005,
      "loss": 2.9688405990600586,
      "step": 2080
    },
    {
      "epoch": 0.4813786722183669,
      "grad_norm": 1.1623741388320923,
      "learning_rate": 0.0005,
      "loss": 2.783679246902466,
      "step": 2081
    },
    {
      "epoch": 0.4816099930603747,
      "grad_norm": 1.341144323348999,
      "learning_rate": 0.0005,
      "loss": 3.353226661682129,
      "step": 2082
    },
    {
      "epoch": 0.4818413139023826,
      "grad_norm": 1.5316678285598755,
      "learning_rate": 0.0005,
      "loss": 2.7322542667388916,
      "step": 2083
    },
    {
      "epoch": 0.4820726347443905,
      "grad_norm": 1.1168220043182373,
      "learning_rate": 0.0005,
      "loss": 3.075490951538086,
      "step": 2084
    },
    {
      "epoch": 0.4823039555863983,
      "grad_norm": 1.1587861776351929,
      "learning_rate": 0.0005,
      "loss": 2.923067808151245,
      "step": 2085
    },
    {
      "epoch": 0.4825352764284062,
      "grad_norm": 1.0132088661193848,
      "learning_rate": 0.0005,
      "loss": 3.0265581607818604,
      "step": 2086
    },
    {
      "epoch": 0.48276659727041404,
      "grad_norm": 1.717628002166748,
      "learning_rate": 0.0005,
      "loss": 2.979793071746826,
      "step": 2087
    },
    {
      "epoch": 0.48299791811242193,
      "grad_norm": 1.1927250623703003,
      "learning_rate": 0.0005,
      "loss": 3.306584358215332,
      "step": 2088
    },
    {
      "epoch": 0.4832292389544298,
      "grad_norm": 1.1680548191070557,
      "learning_rate": 0.0005,
      "loss": 3.0295610427856445,
      "step": 2089
    },
    {
      "epoch": 0.48346055979643765,
      "grad_norm": 1.2981767654418945,
      "learning_rate": 0.0005,
      "loss": 2.5171828269958496,
      "step": 2090
    },
    {
      "epoch": 0.48369188063844554,
      "grad_norm": 1.2043360471725464,
      "learning_rate": 0.0005,
      "loss": 3.2100720405578613,
      "step": 2091
    },
    {
      "epoch": 0.48392320148045337,
      "grad_norm": 1.1872897148132324,
      "learning_rate": 0.0005,
      "loss": 2.635983943939209,
      "step": 2092
    },
    {
      "epoch": 0.48415452232246126,
      "grad_norm": 1.0533509254455566,
      "learning_rate": 0.0005,
      "loss": 2.7765278816223145,
      "step": 2093
    },
    {
      "epoch": 0.48438584316446914,
      "grad_norm": 1.0662355422973633,
      "learning_rate": 0.0005,
      "loss": 2.9922518730163574,
      "step": 2094
    },
    {
      "epoch": 0.484617164006477,
      "grad_norm": 1.1878832578659058,
      "learning_rate": 0.0005,
      "loss": 2.4477386474609375,
      "step": 2095
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.0894804000854492,
      "learning_rate": 0.0005,
      "loss": 2.914473295211792,
      "step": 2096
    },
    {
      "epoch": 0.4850798056904927,
      "grad_norm": 4.747458457946777,
      "learning_rate": 0.0005,
      "loss": 3.3662047386169434,
      "step": 2097
    },
    {
      "epoch": 0.4853111265325006,
      "grad_norm": 1.021360158920288,
      "learning_rate": 0.0005,
      "loss": 2.633328437805176,
      "step": 2098
    },
    {
      "epoch": 0.48554244737450847,
      "grad_norm": 1.0578577518463135,
      "learning_rate": 0.0005,
      "loss": 2.702655076980591,
      "step": 2099
    },
    {
      "epoch": 0.4857737682165163,
      "grad_norm": 1.2040205001831055,
      "learning_rate": 0.0005,
      "loss": 2.4346539974212646,
      "step": 2100
    },
    {
      "epoch": 0.4860050890585242,
      "grad_norm": 0.9380905628204346,
      "learning_rate": 0.0005,
      "loss": 2.85617733001709,
      "step": 2101
    },
    {
      "epoch": 0.486236409900532,
      "grad_norm": 1.0230377912521362,
      "learning_rate": 0.0005,
      "loss": 2.8341546058654785,
      "step": 2102
    },
    {
      "epoch": 0.4864677307425399,
      "grad_norm": 0.9624582529067993,
      "learning_rate": 0.0005,
      "loss": 2.587380886077881,
      "step": 2103
    },
    {
      "epoch": 0.4866990515845478,
      "grad_norm": 1.0511044263839722,
      "learning_rate": 0.0005,
      "loss": 2.771394968032837,
      "step": 2104
    },
    {
      "epoch": 0.4869303724265556,
      "grad_norm": 1.038492202758789,
      "learning_rate": 0.0005,
      "loss": 3.1015334129333496,
      "step": 2105
    },
    {
      "epoch": 0.4871616932685635,
      "grad_norm": 1.183739423751831,
      "learning_rate": 0.0005,
      "loss": 2.805027484893799,
      "step": 2106
    },
    {
      "epoch": 0.48739301411057134,
      "grad_norm": 1.4874118566513062,
      "learning_rate": 0.0005,
      "loss": 3.1288979053497314,
      "step": 2107
    },
    {
      "epoch": 0.48762433495257923,
      "grad_norm": 1.258727788925171,
      "learning_rate": 0.0005,
      "loss": 2.871204137802124,
      "step": 2108
    },
    {
      "epoch": 0.4878556557945871,
      "grad_norm": 1.2975019216537476,
      "learning_rate": 0.0005,
      "loss": 3.1233620643615723,
      "step": 2109
    },
    {
      "epoch": 0.48808697663659495,
      "grad_norm": 1.3420257568359375,
      "learning_rate": 0.0005,
      "loss": 2.719524621963501,
      "step": 2110
    },
    {
      "epoch": 0.48831829747860284,
      "grad_norm": 1.1628448963165283,
      "learning_rate": 0.0005,
      "loss": 2.9457333087921143,
      "step": 2111
    },
    {
      "epoch": 0.48854961832061067,
      "grad_norm": 1.5293282270431519,
      "learning_rate": 0.0005,
      "loss": 3.1999454498291016,
      "step": 2112
    },
    {
      "epoch": 0.48878093916261856,
      "grad_norm": 1.0484507083892822,
      "learning_rate": 0.0005,
      "loss": 2.8175885677337646,
      "step": 2113
    },
    {
      "epoch": 0.48901226000462644,
      "grad_norm": 0.9747511744499207,
      "learning_rate": 0.0005,
      "loss": 2.930011749267578,
      "step": 2114
    },
    {
      "epoch": 0.4892435808466343,
      "grad_norm": 1.168715476989746,
      "learning_rate": 0.0005,
      "loss": 3.23820424079895,
      "step": 2115
    },
    {
      "epoch": 0.48947490168864216,
      "grad_norm": 1.1181437969207764,
      "learning_rate": 0.0005,
      "loss": 3.075005292892456,
      "step": 2116
    },
    {
      "epoch": 0.48970622253065,
      "grad_norm": 3.34824800491333,
      "learning_rate": 0.0005,
      "loss": 2.8069210052490234,
      "step": 2117
    },
    {
      "epoch": 0.4899375433726579,
      "grad_norm": 1.2145410776138306,
      "learning_rate": 0.0005,
      "loss": 3.0430538654327393,
      "step": 2118
    },
    {
      "epoch": 0.49016886421466577,
      "grad_norm": 1.051414966583252,
      "learning_rate": 0.0005,
      "loss": 3.1699838638305664,
      "step": 2119
    },
    {
      "epoch": 0.4904001850566736,
      "grad_norm": 1.0722813606262207,
      "learning_rate": 0.0005,
      "loss": 3.1954634189605713,
      "step": 2120
    },
    {
      "epoch": 0.4906315058986815,
      "grad_norm": 1.2989319562911987,
      "learning_rate": 0.0005,
      "loss": 2.9316210746765137,
      "step": 2121
    },
    {
      "epoch": 0.4908628267406893,
      "grad_norm": 1.152404546737671,
      "learning_rate": 0.0005,
      "loss": 3.2779898643493652,
      "step": 2122
    },
    {
      "epoch": 0.4910941475826972,
      "grad_norm": 1.0751852989196777,
      "learning_rate": 0.0005,
      "loss": 3.37837290763855,
      "step": 2123
    },
    {
      "epoch": 0.4913254684247051,
      "grad_norm": 1.0771812200546265,
      "learning_rate": 0.0005,
      "loss": 2.8712854385375977,
      "step": 2124
    },
    {
      "epoch": 0.4915567892667129,
      "grad_norm": 1.5997787714004517,
      "learning_rate": 0.0005,
      "loss": 2.792119026184082,
      "step": 2125
    },
    {
      "epoch": 0.4917881101087208,
      "grad_norm": 1.0470547676086426,
      "learning_rate": 0.0005,
      "loss": 2.7167344093322754,
      "step": 2126
    },
    {
      "epoch": 0.49201943095072864,
      "grad_norm": 1.3212451934814453,
      "learning_rate": 0.0005,
      "loss": 3.062943935394287,
      "step": 2127
    },
    {
      "epoch": 0.49225075179273653,
      "grad_norm": 1.3685086965560913,
      "learning_rate": 0.0005,
      "loss": 2.953381061553955,
      "step": 2128
    },
    {
      "epoch": 0.4924820726347444,
      "grad_norm": 1.0922226905822754,
      "learning_rate": 0.0005,
      "loss": 2.623741388320923,
      "step": 2129
    },
    {
      "epoch": 0.49271339347675225,
      "grad_norm": 1.7135164737701416,
      "learning_rate": 0.0005,
      "loss": 2.6618053913116455,
      "step": 2130
    },
    {
      "epoch": 0.49294471431876014,
      "grad_norm": 1.3288692235946655,
      "learning_rate": 0.0005,
      "loss": 2.7087979316711426,
      "step": 2131
    },
    {
      "epoch": 0.49317603516076797,
      "grad_norm": 0.9608602523803711,
      "learning_rate": 0.0005,
      "loss": 2.665177345275879,
      "step": 2132
    },
    {
      "epoch": 0.49340735600277585,
      "grad_norm": 1.1703635454177856,
      "learning_rate": 0.0005,
      "loss": 2.683598518371582,
      "step": 2133
    },
    {
      "epoch": 0.49363867684478374,
      "grad_norm": 0.9354174733161926,
      "learning_rate": 0.0005,
      "loss": 3.1179049015045166,
      "step": 2134
    },
    {
      "epoch": 0.4938699976867916,
      "grad_norm": 1.0280354022979736,
      "learning_rate": 0.0005,
      "loss": 3.0158915519714355,
      "step": 2135
    },
    {
      "epoch": 0.49410131852879946,
      "grad_norm": 1.3697696924209595,
      "learning_rate": 0.0005,
      "loss": 2.3837223052978516,
      "step": 2136
    },
    {
      "epoch": 0.4943326393708073,
      "grad_norm": 1.2497742176055908,
      "learning_rate": 0.0005,
      "loss": 2.643599033355713,
      "step": 2137
    },
    {
      "epoch": 0.4945639602128152,
      "grad_norm": 1.7010782957077026,
      "learning_rate": 0.0005,
      "loss": 2.6893434524536133,
      "step": 2138
    },
    {
      "epoch": 0.49479528105482307,
      "grad_norm": 2.379686117172241,
      "learning_rate": 0.0005,
      "loss": 2.8396010398864746,
      "step": 2139
    },
    {
      "epoch": 0.4950266018968309,
      "grad_norm": 1.3428643941879272,
      "learning_rate": 0.0005,
      "loss": 2.8215627670288086,
      "step": 2140
    },
    {
      "epoch": 0.4952579227388388,
      "grad_norm": 1.2618969678878784,
      "learning_rate": 0.0005,
      "loss": 2.9723615646362305,
      "step": 2141
    },
    {
      "epoch": 0.4954892435808466,
      "grad_norm": 0.9213557243347168,
      "learning_rate": 0.0005,
      "loss": 3.106626272201538,
      "step": 2142
    },
    {
      "epoch": 0.4957205644228545,
      "grad_norm": 1.4071643352508545,
      "learning_rate": 0.0005,
      "loss": 2.826174259185791,
      "step": 2143
    },
    {
      "epoch": 0.4959518852648624,
      "grad_norm": 1.1560500860214233,
      "learning_rate": 0.0005,
      "loss": 2.901170015335083,
      "step": 2144
    },
    {
      "epoch": 0.4961832061068702,
      "grad_norm": 1.3039543628692627,
      "learning_rate": 0.0005,
      "loss": 2.781360149383545,
      "step": 2145
    },
    {
      "epoch": 0.4964145269488781,
      "grad_norm": 1.2078993320465088,
      "learning_rate": 0.0005,
      "loss": 3.09147572517395,
      "step": 2146
    },
    {
      "epoch": 0.49664584779088594,
      "grad_norm": 1.0488473176956177,
      "learning_rate": 0.0005,
      "loss": 2.8792405128479004,
      "step": 2147
    },
    {
      "epoch": 0.49687716863289383,
      "grad_norm": 1.547631025314331,
      "learning_rate": 0.0005,
      "loss": 2.9687087535858154,
      "step": 2148
    },
    {
      "epoch": 0.4971084894749017,
      "grad_norm": 1.7284255027770996,
      "learning_rate": 0.0005,
      "loss": 2.868076801300049,
      "step": 2149
    },
    {
      "epoch": 0.49733981031690955,
      "grad_norm": 1.4939844608306885,
      "learning_rate": 0.0005,
      "loss": 2.6445159912109375,
      "step": 2150
    },
    {
      "epoch": 0.49757113115891743,
      "grad_norm": 1.081660509109497,
      "learning_rate": 0.0005,
      "loss": 2.8827762603759766,
      "step": 2151
    },
    {
      "epoch": 0.49780245200092527,
      "grad_norm": 1.041597843170166,
      "learning_rate": 0.0005,
      "loss": 2.9531197547912598,
      "step": 2152
    },
    {
      "epoch": 0.49803377284293315,
      "grad_norm": 1.2461012601852417,
      "learning_rate": 0.0005,
      "loss": 2.463827610015869,
      "step": 2153
    },
    {
      "epoch": 0.49826509368494104,
      "grad_norm": 1.1196093559265137,
      "learning_rate": 0.0005,
      "loss": 2.8610525131225586,
      "step": 2154
    },
    {
      "epoch": 0.49849641452694887,
      "grad_norm": 0.9898086786270142,
      "learning_rate": 0.0005,
      "loss": 2.922744035720825,
      "step": 2155
    },
    {
      "epoch": 0.49872773536895676,
      "grad_norm": 0.9437980055809021,
      "learning_rate": 0.0005,
      "loss": 2.8002419471740723,
      "step": 2156
    },
    {
      "epoch": 0.4989590562109646,
      "grad_norm": 1.1918531656265259,
      "learning_rate": 0.0005,
      "loss": 2.9760639667510986,
      "step": 2157
    },
    {
      "epoch": 0.4991903770529725,
      "grad_norm": 1.0169577598571777,
      "learning_rate": 0.0005,
      "loss": 2.998777151107788,
      "step": 2158
    },
    {
      "epoch": 0.49942169789498037,
      "grad_norm": 1.0824161767959595,
      "learning_rate": 0.0005,
      "loss": 3.083785057067871,
      "step": 2159
    },
    {
      "epoch": 0.4996530187369882,
      "grad_norm": 1.0461435317993164,
      "learning_rate": 0.0005,
      "loss": 2.5291152000427246,
      "step": 2160
    },
    {
      "epoch": 0.4998843395789961,
      "grad_norm": 1.3551535606384277,
      "learning_rate": 0.0005,
      "loss": 2.7477283477783203,
      "step": 2161
    },
    {
      "epoch": 0.500115660421004,
      "grad_norm": 1.157201886177063,
      "learning_rate": 0.0005,
      "loss": 3.1150662899017334,
      "step": 2162
    },
    {
      "epoch": 0.5003469812630118,
      "grad_norm": 0.9744002223014832,
      "learning_rate": 0.0005,
      "loss": 2.896934986114502,
      "step": 2163
    },
    {
      "epoch": 0.5005783021050196,
      "grad_norm": 1.0918633937835693,
      "learning_rate": 0.0005,
      "loss": 2.9255576133728027,
      "step": 2164
    },
    {
      "epoch": 0.5008096229470276,
      "grad_norm": 1.1047357320785522,
      "learning_rate": 0.0005,
      "loss": 2.848034381866455,
      "step": 2165
    },
    {
      "epoch": 0.5010409437890354,
      "grad_norm": 1.100352168083191,
      "learning_rate": 0.0005,
      "loss": 2.932648181915283,
      "step": 2166
    },
    {
      "epoch": 0.5012722646310432,
      "grad_norm": 1.2461562156677246,
      "learning_rate": 0.0005,
      "loss": 3.033261775970459,
      "step": 2167
    },
    {
      "epoch": 0.5015035854730511,
      "grad_norm": 1.1226089000701904,
      "learning_rate": 0.0005,
      "loss": 2.958003044128418,
      "step": 2168
    },
    {
      "epoch": 0.501734906315059,
      "grad_norm": 1.0179725885391235,
      "learning_rate": 0.0005,
      "loss": 2.8220736980438232,
      "step": 2169
    },
    {
      "epoch": 0.5019662271570668,
      "grad_norm": 1.294911503791809,
      "learning_rate": 0.0005,
      "loss": 3.0495216846466064,
      "step": 2170
    },
    {
      "epoch": 0.5021975479990747,
      "grad_norm": 1.1801116466522217,
      "learning_rate": 0.0005,
      "loss": 2.783299446105957,
      "step": 2171
    },
    {
      "epoch": 0.5024288688410826,
      "grad_norm": 1.2071927785873413,
      "learning_rate": 0.0005,
      "loss": 3.068756341934204,
      "step": 2172
    },
    {
      "epoch": 0.5026601896830905,
      "grad_norm": 1.1842073202133179,
      "learning_rate": 0.0005,
      "loss": 2.6930911540985107,
      "step": 2173
    },
    {
      "epoch": 0.5028915105250983,
      "grad_norm": 1.25383722782135,
      "learning_rate": 0.0005,
      "loss": 2.922640800476074,
      "step": 2174
    },
    {
      "epoch": 0.5031228313671062,
      "grad_norm": 1.0922253131866455,
      "learning_rate": 0.0005,
      "loss": 2.844954490661621,
      "step": 2175
    },
    {
      "epoch": 0.5033541522091141,
      "grad_norm": 0.9369425773620605,
      "learning_rate": 0.0005,
      "loss": 2.6047468185424805,
      "step": 2176
    },
    {
      "epoch": 0.5035854730511219,
      "grad_norm": 1.152817726135254,
      "learning_rate": 0.0005,
      "loss": 3.130540609359741,
      "step": 2177
    },
    {
      "epoch": 0.5038167938931297,
      "grad_norm": 1.4914194345474243,
      "learning_rate": 0.0005,
      "loss": 2.924893379211426,
      "step": 2178
    },
    {
      "epoch": 0.5040481147351377,
      "grad_norm": 1.2087739706039429,
      "learning_rate": 0.0005,
      "loss": 2.809565305709839,
      "step": 2179
    },
    {
      "epoch": 0.5042794355771455,
      "grad_norm": 1.1482675075531006,
      "learning_rate": 0.0005,
      "loss": 2.966761589050293,
      "step": 2180
    },
    {
      "epoch": 0.5045107564191533,
      "grad_norm": 1.249955415725708,
      "learning_rate": 0.0005,
      "loss": 2.6588704586029053,
      "step": 2181
    },
    {
      "epoch": 0.5047420772611613,
      "grad_norm": 1.1599558591842651,
      "learning_rate": 0.0005,
      "loss": 2.6910855770111084,
      "step": 2182
    },
    {
      "epoch": 0.5049733981031691,
      "grad_norm": 1.1282578706741333,
      "learning_rate": 0.0005,
      "loss": 3.0268735885620117,
      "step": 2183
    },
    {
      "epoch": 0.5052047189451769,
      "grad_norm": 1.1062144041061401,
      "learning_rate": 0.0005,
      "loss": 3.1838796138763428,
      "step": 2184
    },
    {
      "epoch": 0.5054360397871849,
      "grad_norm": 1.24639892578125,
      "learning_rate": 0.0005,
      "loss": 2.9206383228302,
      "step": 2185
    },
    {
      "epoch": 0.5056673606291927,
      "grad_norm": 0.9643791317939758,
      "learning_rate": 0.0005,
      "loss": 2.5420970916748047,
      "step": 2186
    },
    {
      "epoch": 0.5058986814712005,
      "grad_norm": 1.7811177968978882,
      "learning_rate": 0.0005,
      "loss": 2.6480536460876465,
      "step": 2187
    },
    {
      "epoch": 0.5061300023132084,
      "grad_norm": 1.711782693862915,
      "learning_rate": 0.0005,
      "loss": 2.9906229972839355,
      "step": 2188
    },
    {
      "epoch": 0.5063613231552163,
      "grad_norm": 1.319737434387207,
      "learning_rate": 0.0005,
      "loss": 2.6416237354278564,
      "step": 2189
    },
    {
      "epoch": 0.5065926439972241,
      "grad_norm": 0.9953361749649048,
      "learning_rate": 0.0005,
      "loss": 2.4301466941833496,
      "step": 2190
    },
    {
      "epoch": 0.506823964839232,
      "grad_norm": 0.9210991263389587,
      "learning_rate": 0.0005,
      "loss": 3.215266227722168,
      "step": 2191
    },
    {
      "epoch": 0.5070552856812399,
      "grad_norm": 1.097087025642395,
      "learning_rate": 0.0005,
      "loss": 2.6326098442077637,
      "step": 2192
    },
    {
      "epoch": 0.5072866065232478,
      "grad_norm": 1.1694813966751099,
      "learning_rate": 0.0005,
      "loss": 2.763554811477661,
      "step": 2193
    },
    {
      "epoch": 0.5075179273652556,
      "grad_norm": 1.121677041053772,
      "learning_rate": 0.0005,
      "loss": 2.9251742362976074,
      "step": 2194
    },
    {
      "epoch": 0.5077492482072635,
      "grad_norm": 1.1933038234710693,
      "learning_rate": 0.0005,
      "loss": 2.9715027809143066,
      "step": 2195
    },
    {
      "epoch": 0.5079805690492714,
      "grad_norm": 1.3495237827301025,
      "learning_rate": 0.0005,
      "loss": 2.6909661293029785,
      "step": 2196
    },
    {
      "epoch": 0.5082118898912792,
      "grad_norm": 1.1698696613311768,
      "learning_rate": 0.0005,
      "loss": 2.8928606510162354,
      "step": 2197
    },
    {
      "epoch": 0.508443210733287,
      "grad_norm": 1.6968920230865479,
      "learning_rate": 0.0005,
      "loss": 2.5791587829589844,
      "step": 2198
    },
    {
      "epoch": 0.508674531575295,
      "grad_norm": 1.5181373357772827,
      "learning_rate": 0.0005,
      "loss": 2.997729539871216,
      "step": 2199
    },
    {
      "epoch": 0.5089058524173028,
      "grad_norm": 1.2676154375076294,
      "learning_rate": 0.0005,
      "loss": 2.96243953704834,
      "step": 2200
    },
    {
      "epoch": 0.5091371732593106,
      "grad_norm": 1.1210720539093018,
      "learning_rate": 0.0005,
      "loss": 3.015861988067627,
      "step": 2201
    },
    {
      "epoch": 0.5093684941013186,
      "grad_norm": 0.8843886256217957,
      "learning_rate": 0.0005,
      "loss": 3.007256031036377,
      "step": 2202
    },
    {
      "epoch": 0.5095998149433264,
      "grad_norm": 1.2371565103530884,
      "learning_rate": 0.0005,
      "loss": 2.8039026260375977,
      "step": 2203
    },
    {
      "epoch": 0.5098311357853342,
      "grad_norm": 1.4212512969970703,
      "learning_rate": 0.0005,
      "loss": 3.8085243701934814,
      "step": 2204
    },
    {
      "epoch": 0.5100624566273422,
      "grad_norm": 0.9514058828353882,
      "learning_rate": 0.0005,
      "loss": 2.773331880569458,
      "step": 2205
    },
    {
      "epoch": 0.51029377746935,
      "grad_norm": 1.1249488592147827,
      "learning_rate": 0.0005,
      "loss": 3.0488641262054443,
      "step": 2206
    },
    {
      "epoch": 0.5105250983113578,
      "grad_norm": 1.1889660358428955,
      "learning_rate": 0.0005,
      "loss": 2.803122043609619,
      "step": 2207
    },
    {
      "epoch": 0.5107564191533657,
      "grad_norm": 1.217891812324524,
      "learning_rate": 0.0005,
      "loss": 2.7071533203125,
      "step": 2208
    },
    {
      "epoch": 0.5109877399953736,
      "grad_norm": 1.140529990196228,
      "learning_rate": 0.0005,
      "loss": 3.0904011726379395,
      "step": 2209
    },
    {
      "epoch": 0.5112190608373814,
      "grad_norm": 1.286206841468811,
      "learning_rate": 0.0005,
      "loss": 2.639888048171997,
      "step": 2210
    },
    {
      "epoch": 0.5114503816793893,
      "grad_norm": 1.319097876548767,
      "learning_rate": 0.0005,
      "loss": 3.454268455505371,
      "step": 2211
    },
    {
      "epoch": 0.5116817025213972,
      "grad_norm": 1.12482488155365,
      "learning_rate": 0.0005,
      "loss": 2.992482900619507,
      "step": 2212
    },
    {
      "epoch": 0.511913023363405,
      "grad_norm": 1.4528664350509644,
      "learning_rate": 0.0005,
      "loss": 2.3444013595581055,
      "step": 2213
    },
    {
      "epoch": 0.5121443442054129,
      "grad_norm": 1.4554933309555054,
      "learning_rate": 0.0005,
      "loss": 2.750375270843506,
      "step": 2214
    },
    {
      "epoch": 0.5123756650474208,
      "grad_norm": 1.0578968524932861,
      "learning_rate": 0.0005,
      "loss": 2.4658265113830566,
      "step": 2215
    },
    {
      "epoch": 0.5126069858894287,
      "grad_norm": 0.9592064619064331,
      "learning_rate": 0.0005,
      "loss": 2.703134298324585,
      "step": 2216
    },
    {
      "epoch": 0.5128383067314365,
      "grad_norm": 1.0416947603225708,
      "learning_rate": 0.0005,
      "loss": 2.754439353942871,
      "step": 2217
    },
    {
      "epoch": 0.5130696275734443,
      "grad_norm": 1.9930574893951416,
      "learning_rate": 0.0005,
      "loss": 3.250537157058716,
      "step": 2218
    },
    {
      "epoch": 0.5133009484154523,
      "grad_norm": 1.1697173118591309,
      "learning_rate": 0.0005,
      "loss": 2.763390302658081,
      "step": 2219
    },
    {
      "epoch": 0.5135322692574601,
      "grad_norm": 1.280585765838623,
      "learning_rate": 0.0005,
      "loss": 2.9206671714782715,
      "step": 2220
    },
    {
      "epoch": 0.5137635900994679,
      "grad_norm": 1.0440419912338257,
      "learning_rate": 0.0005,
      "loss": 2.581514358520508,
      "step": 2221
    },
    {
      "epoch": 0.5139949109414759,
      "grad_norm": 0.9510021209716797,
      "learning_rate": 0.0005,
      "loss": 2.8145108222961426,
      "step": 2222
    },
    {
      "epoch": 0.5142262317834837,
      "grad_norm": 1.1229023933410645,
      "learning_rate": 0.0005,
      "loss": 2.5419833660125732,
      "step": 2223
    },
    {
      "epoch": 0.5144575526254915,
      "grad_norm": 1.1480207443237305,
      "learning_rate": 0.0005,
      "loss": 2.7824149131774902,
      "step": 2224
    },
    {
      "epoch": 0.5146888734674995,
      "grad_norm": 0.9494665265083313,
      "learning_rate": 0.0005,
      "loss": 3.162870407104492,
      "step": 2225
    },
    {
      "epoch": 0.5149201943095073,
      "grad_norm": 1.0868747234344482,
      "learning_rate": 0.0005,
      "loss": 3.0034139156341553,
      "step": 2226
    },
    {
      "epoch": 0.5151515151515151,
      "grad_norm": 1.0473127365112305,
      "learning_rate": 0.0005,
      "loss": 2.594031810760498,
      "step": 2227
    },
    {
      "epoch": 0.515382835993523,
      "grad_norm": 1.5030914545059204,
      "learning_rate": 0.0005,
      "loss": 3.140838146209717,
      "step": 2228
    },
    {
      "epoch": 0.5156141568355309,
      "grad_norm": 1.1132320165634155,
      "learning_rate": 0.0005,
      "loss": 2.6596975326538086,
      "step": 2229
    },
    {
      "epoch": 0.5158454776775387,
      "grad_norm": 0.9781486988067627,
      "learning_rate": 0.0005,
      "loss": 2.5303778648376465,
      "step": 2230
    },
    {
      "epoch": 0.5160767985195466,
      "grad_norm": 1.2031699419021606,
      "learning_rate": 0.0005,
      "loss": 3.160937786102295,
      "step": 2231
    },
    {
      "epoch": 0.5163081193615545,
      "grad_norm": 2.2163314819335938,
      "learning_rate": 0.0005,
      "loss": 2.621469259262085,
      "step": 2232
    },
    {
      "epoch": 0.5165394402035624,
      "grad_norm": 0.936947762966156,
      "learning_rate": 0.0005,
      "loss": 2.7060868740081787,
      "step": 2233
    },
    {
      "epoch": 0.5167707610455702,
      "grad_norm": 0.9518850445747375,
      "learning_rate": 0.0005,
      "loss": 2.5333027839660645,
      "step": 2234
    },
    {
      "epoch": 0.5170020818875781,
      "grad_norm": 0.8864775896072388,
      "learning_rate": 0.0005,
      "loss": 2.7255351543426514,
      "step": 2235
    },
    {
      "epoch": 0.517233402729586,
      "grad_norm": 1.2151963710784912,
      "learning_rate": 0.0005,
      "loss": 2.629688262939453,
      "step": 2236
    },
    {
      "epoch": 0.5174647235715938,
      "grad_norm": 1.3056398630142212,
      "learning_rate": 0.0005,
      "loss": 2.8798000812530518,
      "step": 2237
    },
    {
      "epoch": 0.5176960444136016,
      "grad_norm": 1.0471397638320923,
      "learning_rate": 0.0005,
      "loss": 3.1444854736328125,
      "step": 2238
    },
    {
      "epoch": 0.5179273652556096,
      "grad_norm": 1.0116758346557617,
      "learning_rate": 0.0005,
      "loss": 2.7135047912597656,
      "step": 2239
    },
    {
      "epoch": 0.5181586860976174,
      "grad_norm": 0.9508081078529358,
      "learning_rate": 0.0005,
      "loss": 3.007352113723755,
      "step": 2240
    },
    {
      "epoch": 0.5183900069396252,
      "grad_norm": 2.0168612003326416,
      "learning_rate": 0.0005,
      "loss": 2.8335771560668945,
      "step": 2241
    },
    {
      "epoch": 0.5186213277816332,
      "grad_norm": 1.4842344522476196,
      "learning_rate": 0.0005,
      "loss": 3.00825834274292,
      "step": 2242
    },
    {
      "epoch": 0.518852648623641,
      "grad_norm": 1.1148954629898071,
      "learning_rate": 0.0005,
      "loss": 2.749028444290161,
      "step": 2243
    },
    {
      "epoch": 0.5190839694656488,
      "grad_norm": 1.3558809757232666,
      "learning_rate": 0.0005,
      "loss": 2.8180713653564453,
      "step": 2244
    },
    {
      "epoch": 0.5193152903076568,
      "grad_norm": 1.1355781555175781,
      "learning_rate": 0.0005,
      "loss": 2.933138370513916,
      "step": 2245
    },
    {
      "epoch": 0.5195466111496646,
      "grad_norm": 1.061521291732788,
      "learning_rate": 0.0005,
      "loss": 3.1853275299072266,
      "step": 2246
    },
    {
      "epoch": 0.5197779319916724,
      "grad_norm": 1.2233669757843018,
      "learning_rate": 0.0005,
      "loss": 3.003265619277954,
      "step": 2247
    },
    {
      "epoch": 0.5200092528336803,
      "grad_norm": 1.1381475925445557,
      "learning_rate": 0.0005,
      "loss": 2.992149591445923,
      "step": 2248
    },
    {
      "epoch": 0.5202405736756882,
      "grad_norm": 1.250046968460083,
      "learning_rate": 0.0005,
      "loss": 2.8773341178894043,
      "step": 2249
    },
    {
      "epoch": 0.520471894517696,
      "grad_norm": 1.6953279972076416,
      "learning_rate": 0.0005,
      "loss": 3.0221614837646484,
      "step": 2250
    },
    {
      "epoch": 0.5207032153597039,
      "grad_norm": 0.9223484396934509,
      "learning_rate": 0.0005,
      "loss": 2.7524685859680176,
      "step": 2251
    },
    {
      "epoch": 0.5209345362017118,
      "grad_norm": 1.0713310241699219,
      "learning_rate": 0.0005,
      "loss": 2.7628719806671143,
      "step": 2252
    },
    {
      "epoch": 0.5211658570437196,
      "grad_norm": 1.190215826034546,
      "learning_rate": 0.0005,
      "loss": 2.448452949523926,
      "step": 2253
    },
    {
      "epoch": 0.5213971778857275,
      "grad_norm": 1.504746913909912,
      "learning_rate": 0.0005,
      "loss": 3.0716772079467773,
      "step": 2254
    },
    {
      "epoch": 0.5216284987277354,
      "grad_norm": 0.9061456322669983,
      "learning_rate": 0.0005,
      "loss": 2.851485252380371,
      "step": 2255
    },
    {
      "epoch": 0.5218598195697433,
      "grad_norm": 1.6476047039031982,
      "learning_rate": 0.0005,
      "loss": 3.225076913833618,
      "step": 2256
    },
    {
      "epoch": 0.5220911404117511,
      "grad_norm": 1.0130430459976196,
      "learning_rate": 0.0005,
      "loss": 3.066372871398926,
      "step": 2257
    },
    {
      "epoch": 0.5223224612537589,
      "grad_norm": 1.2387386560440063,
      "learning_rate": 0.0005,
      "loss": 2.7285447120666504,
      "step": 2258
    },
    {
      "epoch": 0.5225537820957669,
      "grad_norm": 1.1764425039291382,
      "learning_rate": 0.0005,
      "loss": 2.7698211669921875,
      "step": 2259
    },
    {
      "epoch": 0.5227851029377747,
      "grad_norm": 0.8898482322692871,
      "learning_rate": 0.0005,
      "loss": 2.967592716217041,
      "step": 2260
    },
    {
      "epoch": 0.5230164237797825,
      "grad_norm": 1.2330403327941895,
      "learning_rate": 0.0005,
      "loss": 2.5479025840759277,
      "step": 2261
    },
    {
      "epoch": 0.5232477446217905,
      "grad_norm": 0.970609188079834,
      "learning_rate": 0.0005,
      "loss": 2.923802375793457,
      "step": 2262
    },
    {
      "epoch": 0.5234790654637983,
      "grad_norm": 0.9564231038093567,
      "learning_rate": 0.0005,
      "loss": 2.9175350666046143,
      "step": 2263
    },
    {
      "epoch": 0.5237103863058061,
      "grad_norm": 1.0338531732559204,
      "learning_rate": 0.0005,
      "loss": 3.0968196392059326,
      "step": 2264
    },
    {
      "epoch": 0.5239417071478141,
      "grad_norm": 0.9622048735618591,
      "learning_rate": 0.0005,
      "loss": 3.1984102725982666,
      "step": 2265
    },
    {
      "epoch": 0.5241730279898219,
      "grad_norm": 2.394249200820923,
      "learning_rate": 0.0005,
      "loss": 2.9726462364196777,
      "step": 2266
    },
    {
      "epoch": 0.5244043488318297,
      "grad_norm": 1.1745076179504395,
      "learning_rate": 0.0005,
      "loss": 2.738860607147217,
      "step": 2267
    },
    {
      "epoch": 0.5246356696738376,
      "grad_norm": 1.8730069398880005,
      "learning_rate": 0.0005,
      "loss": 2.81356143951416,
      "step": 2268
    },
    {
      "epoch": 0.5248669905158455,
      "grad_norm": 1.4203554391860962,
      "learning_rate": 0.0005,
      "loss": 2.663050651550293,
      "step": 2269
    },
    {
      "epoch": 0.5250983113578533,
      "grad_norm": 0.8352109789848328,
      "learning_rate": 0.0005,
      "loss": 2.5939648151397705,
      "step": 2270
    },
    {
      "epoch": 0.5253296321998612,
      "grad_norm": 1.4051086902618408,
      "learning_rate": 0.0005,
      "loss": 2.993502616882324,
      "step": 2271
    },
    {
      "epoch": 0.5255609530418691,
      "grad_norm": 1.0195661783218384,
      "learning_rate": 0.0005,
      "loss": 2.930238723754883,
      "step": 2272
    },
    {
      "epoch": 0.525792273883877,
      "grad_norm": 1.1090127229690552,
      "learning_rate": 0.0005,
      "loss": 2.8196918964385986,
      "step": 2273
    },
    {
      "epoch": 0.5260235947258848,
      "grad_norm": 1.013780117034912,
      "learning_rate": 0.0005,
      "loss": 3.1904964447021484,
      "step": 2274
    },
    {
      "epoch": 0.5262549155678927,
      "grad_norm": 1.1658450365066528,
      "learning_rate": 0.0005,
      "loss": 2.707305431365967,
      "step": 2275
    },
    {
      "epoch": 0.5264862364099006,
      "grad_norm": 1.1080819368362427,
      "learning_rate": 0.0005,
      "loss": 3.212353467941284,
      "step": 2276
    },
    {
      "epoch": 0.5267175572519084,
      "grad_norm": 1.1567223072052002,
      "learning_rate": 0.0005,
      "loss": 2.946190595626831,
      "step": 2277
    },
    {
      "epoch": 0.5269488780939162,
      "grad_norm": 1.1808019876480103,
      "learning_rate": 0.0005,
      "loss": 3.2070579528808594,
      "step": 2278
    },
    {
      "epoch": 0.5271801989359242,
      "grad_norm": 1.369611144065857,
      "learning_rate": 0.0005,
      "loss": 2.818838357925415,
      "step": 2279
    },
    {
      "epoch": 0.527411519777932,
      "grad_norm": 1.165634274482727,
      "learning_rate": 0.0005,
      "loss": 2.8922080993652344,
      "step": 2280
    },
    {
      "epoch": 0.5276428406199398,
      "grad_norm": 1.2229293584823608,
      "learning_rate": 0.0005,
      "loss": 3.046262741088867,
      "step": 2281
    },
    {
      "epoch": 0.5278741614619478,
      "grad_norm": 1.1737151145935059,
      "learning_rate": 0.0005,
      "loss": 2.7991275787353516,
      "step": 2282
    },
    {
      "epoch": 0.5281054823039556,
      "grad_norm": 0.9647290706634521,
      "learning_rate": 0.0005,
      "loss": 2.9152073860168457,
      "step": 2283
    },
    {
      "epoch": 0.5283368031459634,
      "grad_norm": 1.1771808862686157,
      "learning_rate": 0.0005,
      "loss": 3.028557300567627,
      "step": 2284
    },
    {
      "epoch": 0.5285681239879713,
      "grad_norm": 1.0536720752716064,
      "learning_rate": 0.0005,
      "loss": 2.879563808441162,
      "step": 2285
    },
    {
      "epoch": 0.5287994448299792,
      "grad_norm": 1.123741865158081,
      "learning_rate": 0.0005,
      "loss": 2.9305148124694824,
      "step": 2286
    },
    {
      "epoch": 0.529030765671987,
      "grad_norm": 1.0867199897766113,
      "learning_rate": 0.0005,
      "loss": 2.5288329124450684,
      "step": 2287
    },
    {
      "epoch": 0.5292620865139949,
      "grad_norm": 1.5831348896026611,
      "learning_rate": 0.0005,
      "loss": 3.184642791748047,
      "step": 2288
    },
    {
      "epoch": 0.5294934073560028,
      "grad_norm": 1.2490427494049072,
      "learning_rate": 0.0005,
      "loss": 2.5559136867523193,
      "step": 2289
    },
    {
      "epoch": 0.5297247281980106,
      "grad_norm": 1.283539056777954,
      "learning_rate": 0.0005,
      "loss": 2.9726736545562744,
      "step": 2290
    },
    {
      "epoch": 0.5299560490400185,
      "grad_norm": 0.9652808308601379,
      "learning_rate": 0.0005,
      "loss": 2.5952117443084717,
      "step": 2291
    },
    {
      "epoch": 0.5301873698820264,
      "grad_norm": 1.179884910583496,
      "learning_rate": 0.0005,
      "loss": 2.947263240814209,
      "step": 2292
    },
    {
      "epoch": 0.5304186907240342,
      "grad_norm": 1.0266879796981812,
      "learning_rate": 0.0005,
      "loss": 2.8922386169433594,
      "step": 2293
    },
    {
      "epoch": 0.5306500115660421,
      "grad_norm": 1.2523564100265503,
      "learning_rate": 0.0005,
      "loss": 3.128903865814209,
      "step": 2294
    },
    {
      "epoch": 0.5308813324080499,
      "grad_norm": 1.0291939973831177,
      "learning_rate": 0.0005,
      "loss": 2.5810747146606445,
      "step": 2295
    },
    {
      "epoch": 0.5311126532500579,
      "grad_norm": 0.9286101460456848,
      "learning_rate": 0.0005,
      "loss": 2.6933679580688477,
      "step": 2296
    },
    {
      "epoch": 0.5313439740920657,
      "grad_norm": 0.937578558921814,
      "learning_rate": 0.0005,
      "loss": 2.793860673904419,
      "step": 2297
    },
    {
      "epoch": 0.5315752949340735,
      "grad_norm": 1.1221287250518799,
      "learning_rate": 0.0005,
      "loss": 2.87252140045166,
      "step": 2298
    },
    {
      "epoch": 0.5318066157760815,
      "grad_norm": 1.0499595403671265,
      "learning_rate": 0.0005,
      "loss": 2.5342867374420166,
      "step": 2299
    },
    {
      "epoch": 0.5320379366180893,
      "grad_norm": 0.9385243654251099,
      "learning_rate": 0.0005,
      "loss": 2.874439001083374,
      "step": 2300
    },
    {
      "epoch": 0.5322692574600971,
      "grad_norm": 1.0520495176315308,
      "learning_rate": 0.0005,
      "loss": 3.048861503601074,
      "step": 2301
    },
    {
      "epoch": 0.5325005783021051,
      "grad_norm": 0.9935257434844971,
      "learning_rate": 0.0005,
      "loss": 2.921250343322754,
      "step": 2302
    },
    {
      "epoch": 0.5327318991441129,
      "grad_norm": 1.008689522743225,
      "learning_rate": 0.0005,
      "loss": 2.8637404441833496,
      "step": 2303
    },
    {
      "epoch": 0.5329632199861207,
      "grad_norm": 1.1718506813049316,
      "learning_rate": 0.0005,
      "loss": 2.9802379608154297,
      "step": 2304
    },
    {
      "epoch": 0.5331945408281286,
      "grad_norm": 1.149470329284668,
      "learning_rate": 0.0005,
      "loss": 3.214754819869995,
      "step": 2305
    },
    {
      "epoch": 0.5334258616701365,
      "grad_norm": 1.2200405597686768,
      "learning_rate": 0.0005,
      "loss": 2.9650821685791016,
      "step": 2306
    },
    {
      "epoch": 0.5336571825121443,
      "grad_norm": 1.2552573680877686,
      "learning_rate": 0.0005,
      "loss": 2.9638023376464844,
      "step": 2307
    },
    {
      "epoch": 0.5338885033541522,
      "grad_norm": 0.9231005311012268,
      "learning_rate": 0.0005,
      "loss": 2.814053535461426,
      "step": 2308
    },
    {
      "epoch": 0.5341198241961601,
      "grad_norm": 2.5617692470550537,
      "learning_rate": 0.0005,
      "loss": 2.801309108734131,
      "step": 2309
    },
    {
      "epoch": 0.5343511450381679,
      "grad_norm": 0.841119647026062,
      "learning_rate": 0.0005,
      "loss": 2.4436843395233154,
      "step": 2310
    },
    {
      "epoch": 0.5345824658801758,
      "grad_norm": 1.138641119003296,
      "learning_rate": 0.0005,
      "loss": 2.9487342834472656,
      "step": 2311
    },
    {
      "epoch": 0.5348137867221837,
      "grad_norm": 1.0076961517333984,
      "learning_rate": 0.0005,
      "loss": 3.0484888553619385,
      "step": 2312
    },
    {
      "epoch": 0.5350451075641915,
      "grad_norm": 1.229603886604309,
      "learning_rate": 0.0005,
      "loss": 2.650613784790039,
      "step": 2313
    },
    {
      "epoch": 0.5352764284061994,
      "grad_norm": 1.6939283609390259,
      "learning_rate": 0.0005,
      "loss": 2.903364896774292,
      "step": 2314
    },
    {
      "epoch": 0.5355077492482072,
      "grad_norm": 1.551100254058838,
      "learning_rate": 0.0005,
      "loss": 2.896671772003174,
      "step": 2315
    },
    {
      "epoch": 0.5357390700902152,
      "grad_norm": 1.2941311597824097,
      "learning_rate": 0.0005,
      "loss": 2.8979597091674805,
      "step": 2316
    },
    {
      "epoch": 0.535970390932223,
      "grad_norm": 1.0584112405776978,
      "learning_rate": 0.0005,
      "loss": 3.0015366077423096,
      "step": 2317
    },
    {
      "epoch": 0.5362017117742308,
      "grad_norm": 1.6227846145629883,
      "learning_rate": 0.0005,
      "loss": 2.6929287910461426,
      "step": 2318
    },
    {
      "epoch": 0.5364330326162388,
      "grad_norm": 1.030278205871582,
      "learning_rate": 0.0005,
      "loss": 2.898015260696411,
      "step": 2319
    },
    {
      "epoch": 0.5366643534582466,
      "grad_norm": 1.2947711944580078,
      "learning_rate": 0.0005,
      "loss": 2.58644962310791,
      "step": 2320
    },
    {
      "epoch": 0.5368956743002544,
      "grad_norm": 1.0182068347930908,
      "learning_rate": 0.0005,
      "loss": 3.14882230758667,
      "step": 2321
    },
    {
      "epoch": 0.5371269951422624,
      "grad_norm": 0.9999816417694092,
      "learning_rate": 0.0005,
      "loss": 2.8586416244506836,
      "step": 2322
    },
    {
      "epoch": 0.5373583159842702,
      "grad_norm": 1.2960972785949707,
      "learning_rate": 0.0005,
      "loss": 3.097811222076416,
      "step": 2323
    },
    {
      "epoch": 0.537589636826278,
      "grad_norm": 1.405161738395691,
      "learning_rate": 0.0005,
      "loss": 2.8314199447631836,
      "step": 2324
    },
    {
      "epoch": 0.5378209576682859,
      "grad_norm": 1.2890950441360474,
      "learning_rate": 0.0005,
      "loss": 2.6993420124053955,
      "step": 2325
    },
    {
      "epoch": 0.5380522785102938,
      "grad_norm": 1.1804769039154053,
      "learning_rate": 0.0005,
      "loss": 2.723414897918701,
      "step": 2326
    },
    {
      "epoch": 0.5382835993523016,
      "grad_norm": 1.2004046440124512,
      "learning_rate": 0.0005,
      "loss": 2.430417537689209,
      "step": 2327
    },
    {
      "epoch": 0.5385149201943095,
      "grad_norm": 1.3336468935012817,
      "learning_rate": 0.0005,
      "loss": 2.781951904296875,
      "step": 2328
    },
    {
      "epoch": 0.5387462410363174,
      "grad_norm": 1.931388020515442,
      "learning_rate": 0.0005,
      "loss": 2.9554643630981445,
      "step": 2329
    },
    {
      "epoch": 0.5389775618783252,
      "grad_norm": 1.1668267250061035,
      "learning_rate": 0.0005,
      "loss": 2.699734687805176,
      "step": 2330
    },
    {
      "epoch": 0.5392088827203331,
      "grad_norm": 1.5284730195999146,
      "learning_rate": 0.0005,
      "loss": 2.5681886672973633,
      "step": 2331
    },
    {
      "epoch": 0.539440203562341,
      "grad_norm": 1.1471797227859497,
      "learning_rate": 0.0005,
      "loss": 2.3175125122070312,
      "step": 2332
    },
    {
      "epoch": 0.5396715244043488,
      "grad_norm": 0.8882282376289368,
      "learning_rate": 0.0005,
      "loss": 2.9410367012023926,
      "step": 2333
    },
    {
      "epoch": 0.5399028452463567,
      "grad_norm": 1.3084282875061035,
      "learning_rate": 0.0005,
      "loss": 2.8413827419281006,
      "step": 2334
    },
    {
      "epoch": 0.5401341660883645,
      "grad_norm": 0.9158526062965393,
      "learning_rate": 0.0005,
      "loss": 2.7562665939331055,
      "step": 2335
    },
    {
      "epoch": 0.5403654869303725,
      "grad_norm": 1.3471273183822632,
      "learning_rate": 0.0005,
      "loss": 2.9276561737060547,
      "step": 2336
    },
    {
      "epoch": 0.5405968077723803,
      "grad_norm": 1.331149697303772,
      "learning_rate": 0.0005,
      "loss": 2.6885786056518555,
      "step": 2337
    },
    {
      "epoch": 0.5408281286143881,
      "grad_norm": 1.1725960969924927,
      "learning_rate": 0.0005,
      "loss": 2.8136940002441406,
      "step": 2338
    },
    {
      "epoch": 0.5410594494563961,
      "grad_norm": 0.9962106943130493,
      "learning_rate": 0.0005,
      "loss": 2.557082176208496,
      "step": 2339
    },
    {
      "epoch": 0.5412907702984039,
      "grad_norm": 0.9937486052513123,
      "learning_rate": 0.0005,
      "loss": 2.653106212615967,
      "step": 2340
    },
    {
      "epoch": 0.5415220911404117,
      "grad_norm": 1.3389521837234497,
      "learning_rate": 0.0005,
      "loss": 2.5343918800354004,
      "step": 2341
    },
    {
      "epoch": 0.5417534119824197,
      "grad_norm": 1.2383208274841309,
      "learning_rate": 0.0005,
      "loss": 2.73309326171875,
      "step": 2342
    },
    {
      "epoch": 0.5419847328244275,
      "grad_norm": 1.2262029647827148,
      "learning_rate": 0.0005,
      "loss": 2.803457260131836,
      "step": 2343
    },
    {
      "epoch": 0.5422160536664353,
      "grad_norm": 1.054585337638855,
      "learning_rate": 0.0005,
      "loss": 2.8399689197540283,
      "step": 2344
    },
    {
      "epoch": 0.5424473745084432,
      "grad_norm": 1.1662418842315674,
      "learning_rate": 0.0005,
      "loss": 2.9026870727539062,
      "step": 2345
    },
    {
      "epoch": 0.5426786953504511,
      "grad_norm": 4.335556507110596,
      "learning_rate": 0.0005,
      "loss": 2.876807451248169,
      "step": 2346
    },
    {
      "epoch": 0.5429100161924589,
      "grad_norm": 1.0874557495117188,
      "learning_rate": 0.0005,
      "loss": 2.644930124282837,
      "step": 2347
    },
    {
      "epoch": 0.5431413370344668,
      "grad_norm": 1.5771198272705078,
      "learning_rate": 0.0005,
      "loss": 2.6812398433685303,
      "step": 2348
    },
    {
      "epoch": 0.5433726578764747,
      "grad_norm": 1.2399379014968872,
      "learning_rate": 0.0005,
      "loss": 2.859846591949463,
      "step": 2349
    },
    {
      "epoch": 0.5436039787184825,
      "grad_norm": 1.050848126411438,
      "learning_rate": 0.0005,
      "loss": 2.803496837615967,
      "step": 2350
    },
    {
      "epoch": 0.5438352995604904,
      "grad_norm": 0.9345716238021851,
      "learning_rate": 0.0005,
      "loss": 2.671353816986084,
      "step": 2351
    },
    {
      "epoch": 0.5440666204024983,
      "grad_norm": 1.0122960805892944,
      "learning_rate": 0.0005,
      "loss": 2.5659563541412354,
      "step": 2352
    },
    {
      "epoch": 0.5442979412445061,
      "grad_norm": 0.9835581183433533,
      "learning_rate": 0.0005,
      "loss": 2.504920244216919,
      "step": 2353
    },
    {
      "epoch": 0.544529262086514,
      "grad_norm": 1.0554358959197998,
      "learning_rate": 0.0005,
      "loss": 3.103712558746338,
      "step": 2354
    },
    {
      "epoch": 0.5447605829285218,
      "grad_norm": 1.0443243980407715,
      "learning_rate": 0.0005,
      "loss": 2.859593391418457,
      "step": 2355
    },
    {
      "epoch": 0.5449919037705298,
      "grad_norm": 1.1094549894332886,
      "learning_rate": 0.0005,
      "loss": 2.8501462936401367,
      "step": 2356
    },
    {
      "epoch": 0.5452232246125376,
      "grad_norm": 1.0642123222351074,
      "learning_rate": 0.0005,
      "loss": 2.786484718322754,
      "step": 2357
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.8130277395248413,
      "learning_rate": 0.0005,
      "loss": 2.699627637863159,
      "step": 2358
    },
    {
      "epoch": 0.5456858662965534,
      "grad_norm": 1.1403156518936157,
      "learning_rate": 0.0005,
      "loss": 2.597580909729004,
      "step": 2359
    },
    {
      "epoch": 0.5459171871385612,
      "grad_norm": 0.9673510789871216,
      "learning_rate": 0.0005,
      "loss": 2.7155556678771973,
      "step": 2360
    },
    {
      "epoch": 0.546148507980569,
      "grad_norm": 1.2416797876358032,
      "learning_rate": 0.0005,
      "loss": 2.667599678039551,
      "step": 2361
    },
    {
      "epoch": 0.546379828822577,
      "grad_norm": 1.2319778203964233,
      "learning_rate": 0.0005,
      "loss": 3.1791090965270996,
      "step": 2362
    },
    {
      "epoch": 0.5466111496645848,
      "grad_norm": 1.0148860216140747,
      "learning_rate": 0.0005,
      "loss": 2.8110456466674805,
      "step": 2363
    },
    {
      "epoch": 0.5468424705065926,
      "grad_norm": 1.1246652603149414,
      "learning_rate": 0.0005,
      "loss": 2.959649085998535,
      "step": 2364
    },
    {
      "epoch": 0.5470737913486005,
      "grad_norm": 2.1458377838134766,
      "learning_rate": 0.0005,
      "loss": 2.9952991008758545,
      "step": 2365
    },
    {
      "epoch": 0.5473051121906084,
      "grad_norm": 1.1566967964172363,
      "learning_rate": 0.0005,
      "loss": 2.77616810798645,
      "step": 2366
    },
    {
      "epoch": 0.5475364330326162,
      "grad_norm": 1.0579733848571777,
      "learning_rate": 0.0005,
      "loss": 3.16495418548584,
      "step": 2367
    },
    {
      "epoch": 0.5477677538746241,
      "grad_norm": 1.5624943971633911,
      "learning_rate": 0.0005,
      "loss": 2.7964906692504883,
      "step": 2368
    },
    {
      "epoch": 0.547999074716632,
      "grad_norm": 1.17946195602417,
      "learning_rate": 0.0005,
      "loss": 2.9575586318969727,
      "step": 2369
    },
    {
      "epoch": 0.5482303955586398,
      "grad_norm": 4.274007320404053,
      "learning_rate": 0.0005,
      "loss": 2.705437660217285,
      "step": 2370
    },
    {
      "epoch": 0.5484617164006477,
      "grad_norm": 1.208528757095337,
      "learning_rate": 0.0005,
      "loss": 2.7578086853027344,
      "step": 2371
    },
    {
      "epoch": 0.5486930372426556,
      "grad_norm": 0.9911776781082153,
      "learning_rate": 0.0005,
      "loss": 2.608609914779663,
      "step": 2372
    },
    {
      "epoch": 0.5489243580846634,
      "grad_norm": 1.1903430223464966,
      "learning_rate": 0.0005,
      "loss": 2.6693038940429688,
      "step": 2373
    },
    {
      "epoch": 0.5491556789266713,
      "grad_norm": 1.4331705570220947,
      "learning_rate": 0.0005,
      "loss": 2.79646635055542,
      "step": 2374
    },
    {
      "epoch": 0.5493869997686791,
      "grad_norm": 0.9781389832496643,
      "learning_rate": 0.0005,
      "loss": 3.022469997406006,
      "step": 2375
    },
    {
      "epoch": 0.549618320610687,
      "grad_norm": 1.1214929819107056,
      "learning_rate": 0.0005,
      "loss": 2.627631187438965,
      "step": 2376
    },
    {
      "epoch": 0.5498496414526949,
      "grad_norm": 1.725888967514038,
      "learning_rate": 0.0005,
      "loss": 3.2159576416015625,
      "step": 2377
    },
    {
      "epoch": 0.5500809622947027,
      "grad_norm": 0.9678700566291809,
      "learning_rate": 0.0005,
      "loss": 2.753154754638672,
      "step": 2378
    },
    {
      "epoch": 0.5503122831367107,
      "grad_norm": 0.9814381003379822,
      "learning_rate": 0.0005,
      "loss": 2.938359260559082,
      "step": 2379
    },
    {
      "epoch": 0.5505436039787185,
      "grad_norm": 1.0171631574630737,
      "learning_rate": 0.0005,
      "loss": 3.0994317531585693,
      "step": 2380
    },
    {
      "epoch": 0.5507749248207263,
      "grad_norm": 1.2213900089263916,
      "learning_rate": 0.0005,
      "loss": 2.791720390319824,
      "step": 2381
    },
    {
      "epoch": 0.5510062456627343,
      "grad_norm": 0.8791614770889282,
      "learning_rate": 0.0005,
      "loss": 2.8891687393188477,
      "step": 2382
    },
    {
      "epoch": 0.5512375665047421,
      "grad_norm": 1.0154515504837036,
      "learning_rate": 0.0005,
      "loss": 2.646531820297241,
      "step": 2383
    },
    {
      "epoch": 0.5514688873467499,
      "grad_norm": 0.9780141711235046,
      "learning_rate": 0.0005,
      "loss": 2.7361278533935547,
      "step": 2384
    },
    {
      "epoch": 0.5517002081887578,
      "grad_norm": 1.1866201162338257,
      "learning_rate": 0.0005,
      "loss": 2.9828104972839355,
      "step": 2385
    },
    {
      "epoch": 0.5519315290307657,
      "grad_norm": 1.0011436939239502,
      "learning_rate": 0.0005,
      "loss": 3.046566963195801,
      "step": 2386
    },
    {
      "epoch": 0.5521628498727735,
      "grad_norm": 1.0122225284576416,
      "learning_rate": 0.0005,
      "loss": 2.7612128257751465,
      "step": 2387
    },
    {
      "epoch": 0.5523941707147814,
      "grad_norm": 1.0915708541870117,
      "learning_rate": 0.0005,
      "loss": 3.339775323867798,
      "step": 2388
    },
    {
      "epoch": 0.5526254915567893,
      "grad_norm": 0.9320783615112305,
      "learning_rate": 0.0005,
      "loss": 2.573397636413574,
      "step": 2389
    },
    {
      "epoch": 0.5528568123987971,
      "grad_norm": 1.2131603956222534,
      "learning_rate": 0.0005,
      "loss": 2.92598295211792,
      "step": 2390
    },
    {
      "epoch": 0.553088133240805,
      "grad_norm": 1.4203954935073853,
      "learning_rate": 0.0005,
      "loss": 2.6867027282714844,
      "step": 2391
    },
    {
      "epoch": 0.5533194540828129,
      "grad_norm": 1.2628108263015747,
      "learning_rate": 0.0005,
      "loss": 2.614347457885742,
      "step": 2392
    },
    {
      "epoch": 0.5535507749248207,
      "grad_norm": 1.0398679971694946,
      "learning_rate": 0.0005,
      "loss": 2.665159225463867,
      "step": 2393
    },
    {
      "epoch": 0.5537820957668286,
      "grad_norm": 1.0272191762924194,
      "learning_rate": 0.0005,
      "loss": 2.784226655960083,
      "step": 2394
    },
    {
      "epoch": 0.5540134166088364,
      "grad_norm": 0.9647221565246582,
      "learning_rate": 0.0005,
      "loss": 2.9758529663085938,
      "step": 2395
    },
    {
      "epoch": 0.5542447374508443,
      "grad_norm": 1.126099944114685,
      "learning_rate": 0.0005,
      "loss": 2.893219232559204,
      "step": 2396
    },
    {
      "epoch": 0.5544760582928522,
      "grad_norm": 1.0507272481918335,
      "learning_rate": 0.0005,
      "loss": 2.720346450805664,
      "step": 2397
    },
    {
      "epoch": 0.55470737913486,
      "grad_norm": 1.342561960220337,
      "learning_rate": 0.0005,
      "loss": 2.5128774642944336,
      "step": 2398
    },
    {
      "epoch": 0.554938699976868,
      "grad_norm": 1.3215749263763428,
      "learning_rate": 0.0005,
      "loss": 2.6156463623046875,
      "step": 2399
    },
    {
      "epoch": 0.5551700208188758,
      "grad_norm": 1.2444835901260376,
      "learning_rate": 0.0005,
      "loss": 2.587195873260498,
      "step": 2400
    },
    {
      "epoch": 0.5554013416608836,
      "grad_norm": 1.4134128093719482,
      "learning_rate": 0.0005,
      "loss": 3.041870594024658,
      "step": 2401
    },
    {
      "epoch": 0.5556326625028916,
      "grad_norm": 0.936686635017395,
      "learning_rate": 0.0005,
      "loss": 2.7858519554138184,
      "step": 2402
    },
    {
      "epoch": 0.5558639833448994,
      "grad_norm": 1.4394543170928955,
      "learning_rate": 0.0005,
      "loss": 2.931321620941162,
      "step": 2403
    },
    {
      "epoch": 0.5560953041869072,
      "grad_norm": 1.5542036294937134,
      "learning_rate": 0.0005,
      "loss": 2.618438482284546,
      "step": 2404
    },
    {
      "epoch": 0.5563266250289151,
      "grad_norm": 1.1510190963745117,
      "learning_rate": 0.0005,
      "loss": 2.865154266357422,
      "step": 2405
    },
    {
      "epoch": 0.556557945870923,
      "grad_norm": 1.0439356565475464,
      "learning_rate": 0.0005,
      "loss": 2.563184976577759,
      "step": 2406
    },
    {
      "epoch": 0.5567892667129308,
      "grad_norm": 1.4094561338424683,
      "learning_rate": 0.0005,
      "loss": 2.9085004329681396,
      "step": 2407
    },
    {
      "epoch": 0.5570205875549387,
      "grad_norm": 1.396929144859314,
      "learning_rate": 0.0005,
      "loss": 3.061628580093384,
      "step": 2408
    },
    {
      "epoch": 0.5572519083969466,
      "grad_norm": 0.9794103503227234,
      "learning_rate": 0.0005,
      "loss": 2.729248046875,
      "step": 2409
    },
    {
      "epoch": 0.5574832292389544,
      "grad_norm": 1.033603310585022,
      "learning_rate": 0.0005,
      "loss": 2.7646689414978027,
      "step": 2410
    },
    {
      "epoch": 0.5577145500809623,
      "grad_norm": 1.86684250831604,
      "learning_rate": 0.0005,
      "loss": 3.0388853549957275,
      "step": 2411
    },
    {
      "epoch": 0.5579458709229702,
      "grad_norm": 1.0644892454147339,
      "learning_rate": 0.0005,
      "loss": 2.503600835800171,
      "step": 2412
    },
    {
      "epoch": 0.558177191764978,
      "grad_norm": 0.9365624785423279,
      "learning_rate": 0.0005,
      "loss": 2.776439666748047,
      "step": 2413
    },
    {
      "epoch": 0.5584085126069859,
      "grad_norm": 1.2917892932891846,
      "learning_rate": 0.0005,
      "loss": 2.658295154571533,
      "step": 2414
    },
    {
      "epoch": 0.5586398334489937,
      "grad_norm": 1.2210866212844849,
      "learning_rate": 0.0005,
      "loss": 2.888216495513916,
      "step": 2415
    },
    {
      "epoch": 0.5588711542910016,
      "grad_norm": 0.9459102749824524,
      "learning_rate": 0.0005,
      "loss": 3.059415340423584,
      "step": 2416
    },
    {
      "epoch": 0.5591024751330095,
      "grad_norm": 1.212292194366455,
      "learning_rate": 0.0005,
      "loss": 2.526564836502075,
      "step": 2417
    },
    {
      "epoch": 0.5593337959750173,
      "grad_norm": 0.9007791876792908,
      "learning_rate": 0.0005,
      "loss": 2.8096582889556885,
      "step": 2418
    },
    {
      "epoch": 0.5595651168170253,
      "grad_norm": 1.2204437255859375,
      "learning_rate": 0.0005,
      "loss": 2.8720078468322754,
      "step": 2419
    },
    {
      "epoch": 0.5597964376590331,
      "grad_norm": 0.9091625213623047,
      "learning_rate": 0.0005,
      "loss": 2.8593063354492188,
      "step": 2420
    },
    {
      "epoch": 0.5600277585010409,
      "grad_norm": 1.1831262111663818,
      "learning_rate": 0.0005,
      "loss": 2.9693684577941895,
      "step": 2421
    },
    {
      "epoch": 0.5602590793430489,
      "grad_norm": 1.0830564498901367,
      "learning_rate": 0.0005,
      "loss": 2.6226654052734375,
      "step": 2422
    },
    {
      "epoch": 0.5604904001850567,
      "grad_norm": 1.3865970373153687,
      "learning_rate": 0.0005,
      "loss": 2.686072826385498,
      "step": 2423
    },
    {
      "epoch": 0.5607217210270645,
      "grad_norm": 1.874845266342163,
      "learning_rate": 0.0005,
      "loss": 2.9980974197387695,
      "step": 2424
    },
    {
      "epoch": 0.5609530418690724,
      "grad_norm": 1.0524183511734009,
      "learning_rate": 0.0005,
      "loss": 2.7647593021392822,
      "step": 2425
    },
    {
      "epoch": 0.5611843627110803,
      "grad_norm": 1.0907026529312134,
      "learning_rate": 0.0005,
      "loss": 2.8397254943847656,
      "step": 2426
    },
    {
      "epoch": 0.5614156835530881,
      "grad_norm": 1.1693967580795288,
      "learning_rate": 0.0005,
      "loss": 2.9416275024414062,
      "step": 2427
    },
    {
      "epoch": 0.561647004395096,
      "grad_norm": 1.5035828351974487,
      "learning_rate": 0.0005,
      "loss": 2.7539782524108887,
      "step": 2428
    },
    {
      "epoch": 0.5618783252371039,
      "grad_norm": 1.0647660493850708,
      "learning_rate": 0.0005,
      "loss": 2.8494718074798584,
      "step": 2429
    },
    {
      "epoch": 0.5621096460791117,
      "grad_norm": 1.7317086458206177,
      "learning_rate": 0.0005,
      "loss": 2.8621785640716553,
      "step": 2430
    },
    {
      "epoch": 0.5623409669211196,
      "grad_norm": 1.1988931894302368,
      "learning_rate": 0.0005,
      "loss": 2.927612543106079,
      "step": 2431
    },
    {
      "epoch": 0.5625722877631275,
      "grad_norm": 1.191699504852295,
      "learning_rate": 0.0005,
      "loss": 2.5371012687683105,
      "step": 2432
    },
    {
      "epoch": 0.5628036086051353,
      "grad_norm": 1.2603962421417236,
      "learning_rate": 0.0005,
      "loss": 2.7663965225219727,
      "step": 2433
    },
    {
      "epoch": 0.5630349294471432,
      "grad_norm": 1.2059307098388672,
      "learning_rate": 0.0005,
      "loss": 2.5641262531280518,
      "step": 2434
    },
    {
      "epoch": 0.563266250289151,
      "grad_norm": 1.4129942655563354,
      "learning_rate": 0.0005,
      "loss": 2.8881921768188477,
      "step": 2435
    },
    {
      "epoch": 0.563497571131159,
      "grad_norm": 1.0789633989334106,
      "learning_rate": 0.0005,
      "loss": 2.8971681594848633,
      "step": 2436
    },
    {
      "epoch": 0.5637288919731668,
      "grad_norm": 1.0585596561431885,
      "learning_rate": 0.0005,
      "loss": 2.495607614517212,
      "step": 2437
    },
    {
      "epoch": 0.5639602128151746,
      "grad_norm": 1.1751344203948975,
      "learning_rate": 0.0005,
      "loss": 3.1557271480560303,
      "step": 2438
    },
    {
      "epoch": 0.5641915336571826,
      "grad_norm": 1.3094792366027832,
      "learning_rate": 0.0005,
      "loss": 2.790959358215332,
      "step": 2439
    },
    {
      "epoch": 0.5644228544991904,
      "grad_norm": 1.0390658378601074,
      "learning_rate": 0.0005,
      "loss": 2.8462464809417725,
      "step": 2440
    },
    {
      "epoch": 0.5646541753411982,
      "grad_norm": 1.0510436296463013,
      "learning_rate": 0.0005,
      "loss": 3.0414934158325195,
      "step": 2441
    },
    {
      "epoch": 0.5648854961832062,
      "grad_norm": 1.3038092851638794,
      "learning_rate": 0.0005,
      "loss": 2.8888795375823975,
      "step": 2442
    },
    {
      "epoch": 0.565116817025214,
      "grad_norm": 2.5309176445007324,
      "learning_rate": 0.0005,
      "loss": 2.8243119716644287,
      "step": 2443
    },
    {
      "epoch": 0.5653481378672218,
      "grad_norm": 1.0010943412780762,
      "learning_rate": 0.0005,
      "loss": 2.960700035095215,
      "step": 2444
    },
    {
      "epoch": 0.5655794587092297,
      "grad_norm": 1.1367229223251343,
      "learning_rate": 0.0005,
      "loss": 3.036039352416992,
      "step": 2445
    },
    {
      "epoch": 0.5658107795512376,
      "grad_norm": 1.20014488697052,
      "learning_rate": 0.0005,
      "loss": 2.753291130065918,
      "step": 2446
    },
    {
      "epoch": 0.5660421003932454,
      "grad_norm": 1.212904453277588,
      "learning_rate": 0.0005,
      "loss": 3.0942039489746094,
      "step": 2447
    },
    {
      "epoch": 0.5662734212352533,
      "grad_norm": 1.1779491901397705,
      "learning_rate": 0.0005,
      "loss": 2.5806350708007812,
      "step": 2448
    },
    {
      "epoch": 0.5665047420772612,
      "grad_norm": 1.4612467288970947,
      "learning_rate": 0.0005,
      "loss": 2.7871789932250977,
      "step": 2449
    },
    {
      "epoch": 0.566736062919269,
      "grad_norm": 1.0955828428268433,
      "learning_rate": 0.0005,
      "loss": 2.869267225265503,
      "step": 2450
    },
    {
      "epoch": 0.5669673837612769,
      "grad_norm": 1.6562256813049316,
      "learning_rate": 0.0005,
      "loss": 2.5949971675872803,
      "step": 2451
    },
    {
      "epoch": 0.5671987046032848,
      "grad_norm": 1.1395574808120728,
      "learning_rate": 0.0005,
      "loss": 2.761671543121338,
      "step": 2452
    },
    {
      "epoch": 0.5674300254452926,
      "grad_norm": 1.1641291379928589,
      "learning_rate": 0.0005,
      "loss": 2.8746719360351562,
      "step": 2453
    },
    {
      "epoch": 0.5676613462873005,
      "grad_norm": 0.9852381348609924,
      "learning_rate": 0.0005,
      "loss": 2.705359697341919,
      "step": 2454
    },
    {
      "epoch": 0.5678926671293083,
      "grad_norm": 0.9623990654945374,
      "learning_rate": 0.0005,
      "loss": 2.7244081497192383,
      "step": 2455
    },
    {
      "epoch": 0.5681239879713162,
      "grad_norm": 1.1263049840927124,
      "learning_rate": 0.0005,
      "loss": 3.1027517318725586,
      "step": 2456
    },
    {
      "epoch": 0.5683553088133241,
      "grad_norm": 0.9090439677238464,
      "learning_rate": 0.0005,
      "loss": 2.684704303741455,
      "step": 2457
    },
    {
      "epoch": 0.5685866296553319,
      "grad_norm": 1.4111824035644531,
      "learning_rate": 0.0005,
      "loss": 2.7325024604797363,
      "step": 2458
    },
    {
      "epoch": 0.5688179504973399,
      "grad_norm": 1.1381522417068481,
      "learning_rate": 0.0005,
      "loss": 2.6853489875793457,
      "step": 2459
    },
    {
      "epoch": 0.5690492713393477,
      "grad_norm": 1.0158426761627197,
      "learning_rate": 0.0005,
      "loss": 2.855421543121338,
      "step": 2460
    },
    {
      "epoch": 0.5692805921813555,
      "grad_norm": 1.0045101642608643,
      "learning_rate": 0.0005,
      "loss": 2.663005828857422,
      "step": 2461
    },
    {
      "epoch": 0.5695119130233635,
      "grad_norm": 0.9888191819190979,
      "learning_rate": 0.0005,
      "loss": 3.272548198699951,
      "step": 2462
    },
    {
      "epoch": 0.5697432338653713,
      "grad_norm": 1.281237006187439,
      "learning_rate": 0.0005,
      "loss": 2.6931543350219727,
      "step": 2463
    },
    {
      "epoch": 0.5699745547073791,
      "grad_norm": 1.1865577697753906,
      "learning_rate": 0.0005,
      "loss": 2.9423584938049316,
      "step": 2464
    },
    {
      "epoch": 0.570205875549387,
      "grad_norm": 1.1157816648483276,
      "learning_rate": 0.0005,
      "loss": 2.9198713302612305,
      "step": 2465
    },
    {
      "epoch": 0.5704371963913949,
      "grad_norm": 1.3804606199264526,
      "learning_rate": 0.0005,
      "loss": 3.163465976715088,
      "step": 2466
    },
    {
      "epoch": 0.5706685172334027,
      "grad_norm": 1.048309087753296,
      "learning_rate": 0.0005,
      "loss": 2.986036777496338,
      "step": 2467
    },
    {
      "epoch": 0.5708998380754106,
      "grad_norm": 1.2075846195220947,
      "learning_rate": 0.0005,
      "loss": 2.837283134460449,
      "step": 2468
    },
    {
      "epoch": 0.5711311589174185,
      "grad_norm": 1.0820218324661255,
      "learning_rate": 0.0005,
      "loss": 2.874784469604492,
      "step": 2469
    },
    {
      "epoch": 0.5713624797594263,
      "grad_norm": 1.0367319583892822,
      "learning_rate": 0.0005,
      "loss": 2.9283828735351562,
      "step": 2470
    },
    {
      "epoch": 0.5715938006014342,
      "grad_norm": 1.0498296022415161,
      "learning_rate": 0.0005,
      "loss": 2.8588271141052246,
      "step": 2471
    },
    {
      "epoch": 0.5718251214434421,
      "grad_norm": 1.0112664699554443,
      "learning_rate": 0.0005,
      "loss": 2.600512981414795,
      "step": 2472
    },
    {
      "epoch": 0.5720564422854499,
      "grad_norm": 1.1916425228118896,
      "learning_rate": 0.0005,
      "loss": 2.9870030879974365,
      "step": 2473
    },
    {
      "epoch": 0.5722877631274578,
      "grad_norm": 1.0077989101409912,
      "learning_rate": 0.0005,
      "loss": 2.8194406032562256,
      "step": 2474
    },
    {
      "epoch": 0.5725190839694656,
      "grad_norm": 1.417346715927124,
      "learning_rate": 0.0005,
      "loss": 3.131787061691284,
      "step": 2475
    },
    {
      "epoch": 0.5727504048114735,
      "grad_norm": 0.9486348032951355,
      "learning_rate": 0.0005,
      "loss": 2.659522294998169,
      "step": 2476
    },
    {
      "epoch": 0.5729817256534814,
      "grad_norm": 1.6360868215560913,
      "learning_rate": 0.0005,
      "loss": 2.7647547721862793,
      "step": 2477
    },
    {
      "epoch": 0.5732130464954892,
      "grad_norm": 1.1977111101150513,
      "learning_rate": 0.0005,
      "loss": 3.125392198562622,
      "step": 2478
    },
    {
      "epoch": 0.5734443673374972,
      "grad_norm": 1.2631661891937256,
      "learning_rate": 0.0005,
      "loss": 2.636992931365967,
      "step": 2479
    },
    {
      "epoch": 0.573675688179505,
      "grad_norm": 0.8961793780326843,
      "learning_rate": 0.0005,
      "loss": 2.3173232078552246,
      "step": 2480
    },
    {
      "epoch": 0.5739070090215128,
      "grad_norm": 1.0735899209976196,
      "learning_rate": 0.0005,
      "loss": 2.725893020629883,
      "step": 2481
    },
    {
      "epoch": 0.5741383298635208,
      "grad_norm": 1.0823265314102173,
      "learning_rate": 0.0005,
      "loss": 2.8088393211364746,
      "step": 2482
    },
    {
      "epoch": 0.5743696507055286,
      "grad_norm": 1.291784644126892,
      "learning_rate": 0.0005,
      "loss": 3.218937397003174,
      "step": 2483
    },
    {
      "epoch": 0.5746009715475364,
      "grad_norm": 1.0870825052261353,
      "learning_rate": 0.0005,
      "loss": 2.8057777881622314,
      "step": 2484
    },
    {
      "epoch": 0.5748322923895443,
      "grad_norm": 1.082149863243103,
      "learning_rate": 0.0005,
      "loss": 2.698996067047119,
      "step": 2485
    },
    {
      "epoch": 0.5750636132315522,
      "grad_norm": 1.1483471393585205,
      "learning_rate": 0.0005,
      "loss": 2.728510856628418,
      "step": 2486
    },
    {
      "epoch": 0.57529493407356,
      "grad_norm": 1.6816531419754028,
      "learning_rate": 0.0005,
      "loss": 2.7282562255859375,
      "step": 2487
    },
    {
      "epoch": 0.5755262549155679,
      "grad_norm": 1.360816478729248,
      "learning_rate": 0.0005,
      "loss": 2.9619498252868652,
      "step": 2488
    },
    {
      "epoch": 0.5757575757575758,
      "grad_norm": 1.1728320121765137,
      "learning_rate": 0.0005,
      "loss": 2.635406732559204,
      "step": 2489
    },
    {
      "epoch": 0.5759888965995836,
      "grad_norm": 0.9593884348869324,
      "learning_rate": 0.0005,
      "loss": 2.7763400077819824,
      "step": 2490
    },
    {
      "epoch": 0.5762202174415915,
      "grad_norm": 1.363776683807373,
      "learning_rate": 0.0005,
      "loss": 2.7869510650634766,
      "step": 2491
    },
    {
      "epoch": 0.5764515382835994,
      "grad_norm": 1.5355697870254517,
      "learning_rate": 0.0005,
      "loss": 2.5871684551239014,
      "step": 2492
    },
    {
      "epoch": 0.5766828591256072,
      "grad_norm": 1.0679339170455933,
      "learning_rate": 0.0005,
      "loss": 2.9225668907165527,
      "step": 2493
    },
    {
      "epoch": 0.5769141799676151,
      "grad_norm": 1.1130441427230835,
      "learning_rate": 0.0005,
      "loss": 2.8395960330963135,
      "step": 2494
    },
    {
      "epoch": 0.5771455008096229,
      "grad_norm": 5.415867805480957,
      "learning_rate": 0.0005,
      "loss": 2.8312060832977295,
      "step": 2495
    },
    {
      "epoch": 0.5773768216516308,
      "grad_norm": 0.939997673034668,
      "learning_rate": 0.0005,
      "loss": 2.637974262237549,
      "step": 2496
    },
    {
      "epoch": 0.5776081424936387,
      "grad_norm": 1.0723942518234253,
      "learning_rate": 0.0005,
      "loss": 2.9388082027435303,
      "step": 2497
    },
    {
      "epoch": 0.5778394633356465,
      "grad_norm": 1.0340087413787842,
      "learning_rate": 0.0005,
      "loss": 3.0473108291625977,
      "step": 2498
    },
    {
      "epoch": 0.5780707841776545,
      "grad_norm": 1.5317078828811646,
      "learning_rate": 0.0005,
      "loss": 2.7049458026885986,
      "step": 2499
    },
    {
      "epoch": 0.5783021050196623,
      "grad_norm": 1.7860435247421265,
      "learning_rate": 0.0005,
      "loss": 2.352327823638916,
      "step": 2500
    },
    {
      "epoch": 0.5785334258616701,
      "grad_norm": 1.8187376260757446,
      "learning_rate": 0.0005,
      "loss": 2.764610767364502,
      "step": 2501
    },
    {
      "epoch": 0.5787647467036781,
      "grad_norm": 1.1862059831619263,
      "learning_rate": 0.0005,
      "loss": 2.535649299621582,
      "step": 2502
    },
    {
      "epoch": 0.5789960675456859,
      "grad_norm": 1.552064299583435,
      "learning_rate": 0.0005,
      "loss": 3.185246467590332,
      "step": 2503
    },
    {
      "epoch": 0.5792273883876937,
      "grad_norm": 1.3048975467681885,
      "learning_rate": 0.0005,
      "loss": 2.7496986389160156,
      "step": 2504
    },
    {
      "epoch": 0.5794587092297016,
      "grad_norm": 1.0424526929855347,
      "learning_rate": 0.0005,
      "loss": 2.729221820831299,
      "step": 2505
    },
    {
      "epoch": 0.5796900300717095,
      "grad_norm": 0.9170109033584595,
      "learning_rate": 0.0005,
      "loss": 2.573848009109497,
      "step": 2506
    },
    {
      "epoch": 0.5799213509137173,
      "grad_norm": 1.1264245510101318,
      "learning_rate": 0.0005,
      "loss": 2.8075919151306152,
      "step": 2507
    },
    {
      "epoch": 0.5801526717557252,
      "grad_norm": 1.069193720817566,
      "learning_rate": 0.0005,
      "loss": 3.000519275665283,
      "step": 2508
    },
    {
      "epoch": 0.5803839925977331,
      "grad_norm": 1.0082100629806519,
      "learning_rate": 0.0005,
      "loss": 2.6555047035217285,
      "step": 2509
    },
    {
      "epoch": 0.5806153134397409,
      "grad_norm": 1.0548955202102661,
      "learning_rate": 0.0005,
      "loss": 3.208566188812256,
      "step": 2510
    },
    {
      "epoch": 0.5808466342817488,
      "grad_norm": 1.1426260471343994,
      "learning_rate": 0.0005,
      "loss": 3.1772053241729736,
      "step": 2511
    },
    {
      "epoch": 0.5810779551237566,
      "grad_norm": 1.1553975343704224,
      "learning_rate": 0.0005,
      "loss": 2.8251898288726807,
      "step": 2512
    },
    {
      "epoch": 0.5813092759657645,
      "grad_norm": 0.9590901732444763,
      "learning_rate": 0.0005,
      "loss": 3.037349224090576,
      "step": 2513
    },
    {
      "epoch": 0.5815405968077724,
      "grad_norm": 1.057792067527771,
      "learning_rate": 0.0005,
      "loss": 2.4818923473358154,
      "step": 2514
    },
    {
      "epoch": 0.5817719176497802,
      "grad_norm": 1.4937834739685059,
      "learning_rate": 0.0005,
      "loss": 2.731018543243408,
      "step": 2515
    },
    {
      "epoch": 0.5820032384917881,
      "grad_norm": 1.0633538961410522,
      "learning_rate": 0.0005,
      "loss": 2.6826226711273193,
      "step": 2516
    },
    {
      "epoch": 0.582234559333796,
      "grad_norm": 0.992741048336029,
      "learning_rate": 0.0005,
      "loss": 2.5522618293762207,
      "step": 2517
    },
    {
      "epoch": 0.5824658801758038,
      "grad_norm": 1.2311128377914429,
      "learning_rate": 0.0005,
      "loss": 3.1327667236328125,
      "step": 2518
    },
    {
      "epoch": 0.5826972010178118,
      "grad_norm": 1.1055374145507812,
      "learning_rate": 0.0005,
      "loss": 2.651343822479248,
      "step": 2519
    },
    {
      "epoch": 0.5829285218598196,
      "grad_norm": 1.518771767616272,
      "learning_rate": 0.0005,
      "loss": 3.3684234619140625,
      "step": 2520
    },
    {
      "epoch": 0.5831598427018274,
      "grad_norm": 0.9719303846359253,
      "learning_rate": 0.0005,
      "loss": 2.7839043140411377,
      "step": 2521
    },
    {
      "epoch": 0.5833911635438352,
      "grad_norm": 0.9589550495147705,
      "learning_rate": 0.0005,
      "loss": 2.8640732765197754,
      "step": 2522
    },
    {
      "epoch": 0.5836224843858432,
      "grad_norm": 2.24433970451355,
      "learning_rate": 0.0005,
      "loss": 2.820765495300293,
      "step": 2523
    },
    {
      "epoch": 0.583853805227851,
      "grad_norm": 1.4217305183410645,
      "learning_rate": 0.0005,
      "loss": 2.592487335205078,
      "step": 2524
    },
    {
      "epoch": 0.5840851260698589,
      "grad_norm": 1.7070348262786865,
      "learning_rate": 0.0005,
      "loss": 2.767451524734497,
      "step": 2525
    },
    {
      "epoch": 0.5843164469118668,
      "grad_norm": 1.008583664894104,
      "learning_rate": 0.0005,
      "loss": 2.9029173851013184,
      "step": 2526
    },
    {
      "epoch": 0.5845477677538746,
      "grad_norm": 1.0679138898849487,
      "learning_rate": 0.0005,
      "loss": 2.769749402999878,
      "step": 2527
    },
    {
      "epoch": 0.5847790885958825,
      "grad_norm": 1.1282850503921509,
      "learning_rate": 0.0005,
      "loss": 2.6560816764831543,
      "step": 2528
    },
    {
      "epoch": 0.5850104094378904,
      "grad_norm": 1.0064282417297363,
      "learning_rate": 0.0005,
      "loss": 3.134699821472168,
      "step": 2529
    },
    {
      "epoch": 0.5852417302798982,
      "grad_norm": 1.2804107666015625,
      "learning_rate": 0.0005,
      "loss": 2.9639525413513184,
      "step": 2530
    },
    {
      "epoch": 0.5854730511219061,
      "grad_norm": 1.0459586381912231,
      "learning_rate": 0.0005,
      "loss": 2.790215253829956,
      "step": 2531
    },
    {
      "epoch": 0.5857043719639139,
      "grad_norm": 1.2770665884017944,
      "learning_rate": 0.0005,
      "loss": 2.7891321182250977,
      "step": 2532
    },
    {
      "epoch": 0.5859356928059218,
      "grad_norm": 0.9290795922279358,
      "learning_rate": 0.0005,
      "loss": 2.656116008758545,
      "step": 2533
    },
    {
      "epoch": 0.5861670136479297,
      "grad_norm": 1.3926193714141846,
      "learning_rate": 0.0005,
      "loss": 2.6332807540893555,
      "step": 2534
    },
    {
      "epoch": 0.5863983344899375,
      "grad_norm": 1.4460535049438477,
      "learning_rate": 0.0005,
      "loss": 2.5055689811706543,
      "step": 2535
    },
    {
      "epoch": 0.5866296553319454,
      "grad_norm": 1.3643585443496704,
      "learning_rate": 0.0005,
      "loss": 2.6814699172973633,
      "step": 2536
    },
    {
      "epoch": 0.5868609761739533,
      "grad_norm": 1.3418818712234497,
      "learning_rate": 0.0005,
      "loss": 3.1779370307922363,
      "step": 2537
    },
    {
      "epoch": 0.5870922970159611,
      "grad_norm": 1.152727723121643,
      "learning_rate": 0.0005,
      "loss": 3.1329421997070312,
      "step": 2538
    },
    {
      "epoch": 0.587323617857969,
      "grad_norm": 1.775425910949707,
      "learning_rate": 0.0005,
      "loss": 2.8143577575683594,
      "step": 2539
    },
    {
      "epoch": 0.5875549386999769,
      "grad_norm": 2.1203134059906006,
      "learning_rate": 0.0005,
      "loss": 3.0997276306152344,
      "step": 2540
    },
    {
      "epoch": 0.5877862595419847,
      "grad_norm": 2.255053997039795,
      "learning_rate": 0.0005,
      "loss": 2.5560150146484375,
      "step": 2541
    },
    {
      "epoch": 0.5880175803839925,
      "grad_norm": 1.0964306592941284,
      "learning_rate": 0.0005,
      "loss": 2.541416645050049,
      "step": 2542
    },
    {
      "epoch": 0.5882489012260005,
      "grad_norm": 0.9385858774185181,
      "learning_rate": 0.0005,
      "loss": 3.03491473197937,
      "step": 2543
    },
    {
      "epoch": 0.5884802220680083,
      "grad_norm": 1.1641738414764404,
      "learning_rate": 0.0005,
      "loss": 2.636094093322754,
      "step": 2544
    },
    {
      "epoch": 0.5887115429100162,
      "grad_norm": 1.0293020009994507,
      "learning_rate": 0.0005,
      "loss": 2.671654224395752,
      "step": 2545
    },
    {
      "epoch": 0.5889428637520241,
      "grad_norm": 1.052390217781067,
      "learning_rate": 0.0005,
      "loss": 3.125439167022705,
      "step": 2546
    },
    {
      "epoch": 0.5891741845940319,
      "grad_norm": 1.022396445274353,
      "learning_rate": 0.0005,
      "loss": 2.8314547538757324,
      "step": 2547
    },
    {
      "epoch": 0.5894055054360398,
      "grad_norm": 1.020233154296875,
      "learning_rate": 0.0005,
      "loss": 2.903904676437378,
      "step": 2548
    },
    {
      "epoch": 0.5896368262780477,
      "grad_norm": 1.0220471620559692,
      "learning_rate": 0.0005,
      "loss": 2.7721993923187256,
      "step": 2549
    },
    {
      "epoch": 0.5898681471200555,
      "grad_norm": 0.9485131502151489,
      "learning_rate": 0.0005,
      "loss": 2.598705291748047,
      "step": 2550
    },
    {
      "epoch": 0.5900994679620634,
      "grad_norm": 1.2352536916732788,
      "learning_rate": 0.0005,
      "loss": 2.963517904281616,
      "step": 2551
    },
    {
      "epoch": 0.5903307888040712,
      "grad_norm": 1.1269716024398804,
      "learning_rate": 0.0005,
      "loss": 2.615105152130127,
      "step": 2552
    },
    {
      "epoch": 0.5905621096460791,
      "grad_norm": 1.2767993211746216,
      "learning_rate": 0.0005,
      "loss": 2.5805938243865967,
      "step": 2553
    },
    {
      "epoch": 0.590793430488087,
      "grad_norm": 1.1324738264083862,
      "learning_rate": 0.0005,
      "loss": 2.6964287757873535,
      "step": 2554
    },
    {
      "epoch": 0.5910247513300948,
      "grad_norm": 0.9193450212478638,
      "learning_rate": 0.0005,
      "loss": 2.8048131465911865,
      "step": 2555
    },
    {
      "epoch": 0.5912560721721027,
      "grad_norm": 1.2563092708587646,
      "learning_rate": 0.0005,
      "loss": 3.207333564758301,
      "step": 2556
    },
    {
      "epoch": 0.5914873930141106,
      "grad_norm": 1.0995508432388306,
      "learning_rate": 0.0005,
      "loss": 2.522019624710083,
      "step": 2557
    },
    {
      "epoch": 0.5917187138561184,
      "grad_norm": 0.950452446937561,
      "learning_rate": 0.0005,
      "loss": 2.5902621746063232,
      "step": 2558
    },
    {
      "epoch": 0.5919500346981263,
      "grad_norm": 1.148219108581543,
      "learning_rate": 0.0005,
      "loss": 2.797649383544922,
      "step": 2559
    },
    {
      "epoch": 0.5921813555401342,
      "grad_norm": 1.0810017585754395,
      "learning_rate": 0.0005,
      "loss": 2.7293901443481445,
      "step": 2560
    },
    {
      "epoch": 0.592412676382142,
      "grad_norm": 1.2471219301223755,
      "learning_rate": 0.0005,
      "loss": 2.8069114685058594,
      "step": 2561
    },
    {
      "epoch": 0.5926439972241498,
      "grad_norm": 1.2513675689697266,
      "learning_rate": 0.0005,
      "loss": 2.865933895111084,
      "step": 2562
    },
    {
      "epoch": 0.5928753180661578,
      "grad_norm": 1.2365846633911133,
      "learning_rate": 0.0005,
      "loss": 2.8485047817230225,
      "step": 2563
    },
    {
      "epoch": 0.5931066389081656,
      "grad_norm": 1.3985341787338257,
      "learning_rate": 0.0005,
      "loss": 2.6588449478149414,
      "step": 2564
    },
    {
      "epoch": 0.5933379597501734,
      "grad_norm": 1.4423253536224365,
      "learning_rate": 0.0005,
      "loss": 2.8770034313201904,
      "step": 2565
    },
    {
      "epoch": 0.5935692805921814,
      "grad_norm": 1.1905666589736938,
      "learning_rate": 0.0005,
      "loss": 2.3599257469177246,
      "step": 2566
    },
    {
      "epoch": 0.5938006014341892,
      "grad_norm": 1.2415924072265625,
      "learning_rate": 0.0005,
      "loss": 2.4460830688476562,
      "step": 2567
    },
    {
      "epoch": 0.594031922276197,
      "grad_norm": 1.0565158128738403,
      "learning_rate": 0.0005,
      "loss": 2.5373892784118652,
      "step": 2568
    },
    {
      "epoch": 0.594263243118205,
      "grad_norm": 0.9479575753211975,
      "learning_rate": 0.0005,
      "loss": 2.831423759460449,
      "step": 2569
    },
    {
      "epoch": 0.5944945639602128,
      "grad_norm": 1.2191009521484375,
      "learning_rate": 0.0005,
      "loss": 2.6137855052948,
      "step": 2570
    },
    {
      "epoch": 0.5947258848022207,
      "grad_norm": 1.4860832691192627,
      "learning_rate": 0.0005,
      "loss": 2.8214468955993652,
      "step": 2571
    },
    {
      "epoch": 0.5949572056442285,
      "grad_norm": 0.8672531843185425,
      "learning_rate": 0.0005,
      "loss": 2.7568929195404053,
      "step": 2572
    },
    {
      "epoch": 0.5951885264862364,
      "grad_norm": 1.0117515325546265,
      "learning_rate": 0.0005,
      "loss": 2.726996898651123,
      "step": 2573
    },
    {
      "epoch": 0.5954198473282443,
      "grad_norm": 0.8863999247550964,
      "learning_rate": 0.0005,
      "loss": 2.6126370429992676,
      "step": 2574
    },
    {
      "epoch": 0.5956511681702521,
      "grad_norm": 1.1724578142166138,
      "learning_rate": 0.0005,
      "loss": 2.829047918319702,
      "step": 2575
    },
    {
      "epoch": 0.59588248901226,
      "grad_norm": 1.069136619567871,
      "learning_rate": 0.0005,
      "loss": 2.8819918632507324,
      "step": 2576
    },
    {
      "epoch": 0.5961138098542679,
      "grad_norm": 0.890373170375824,
      "learning_rate": 0.0005,
      "loss": 2.7255783081054688,
      "step": 2577
    },
    {
      "epoch": 0.5963451306962757,
      "grad_norm": 0.9726929664611816,
      "learning_rate": 0.0005,
      "loss": 2.5761728286743164,
      "step": 2578
    },
    {
      "epoch": 0.5965764515382836,
      "grad_norm": 1.1639220714569092,
      "learning_rate": 0.0005,
      "loss": 2.8621487617492676,
      "step": 2579
    },
    {
      "epoch": 0.5968077723802915,
      "grad_norm": 0.9676002264022827,
      "learning_rate": 0.0005,
      "loss": 3.0763304233551025,
      "step": 2580
    },
    {
      "epoch": 0.5970390932222993,
      "grad_norm": 1.1373780965805054,
      "learning_rate": 0.0005,
      "loss": 2.667144775390625,
      "step": 2581
    },
    {
      "epoch": 0.5972704140643071,
      "grad_norm": 1.0472559928894043,
      "learning_rate": 0.0005,
      "loss": 2.926487445831299,
      "step": 2582
    },
    {
      "epoch": 0.5975017349063151,
      "grad_norm": 1.2213451862335205,
      "learning_rate": 0.0005,
      "loss": 2.991513252258301,
      "step": 2583
    },
    {
      "epoch": 0.5977330557483229,
      "grad_norm": 0.939447820186615,
      "learning_rate": 0.0005,
      "loss": 3.252331256866455,
      "step": 2584
    },
    {
      "epoch": 0.5979643765903307,
      "grad_norm": 0.9895576238632202,
      "learning_rate": 0.0005,
      "loss": 3.236499547958374,
      "step": 2585
    },
    {
      "epoch": 0.5981956974323387,
      "grad_norm": 1.4330819845199585,
      "learning_rate": 0.0005,
      "loss": 2.507061004638672,
      "step": 2586
    },
    {
      "epoch": 0.5984270182743465,
      "grad_norm": 0.9519033432006836,
      "learning_rate": 0.0005,
      "loss": 2.794191837310791,
      "step": 2587
    },
    {
      "epoch": 0.5986583391163544,
      "grad_norm": 0.9552523493766785,
      "learning_rate": 0.0005,
      "loss": 2.522665023803711,
      "step": 2588
    },
    {
      "epoch": 0.5988896599583623,
      "grad_norm": 0.9267489314079285,
      "learning_rate": 0.0005,
      "loss": 2.353032112121582,
      "step": 2589
    },
    {
      "epoch": 0.5991209808003701,
      "grad_norm": 1.10097074508667,
      "learning_rate": 0.0005,
      "loss": 2.4810967445373535,
      "step": 2590
    },
    {
      "epoch": 0.599352301642378,
      "grad_norm": 1.063237190246582,
      "learning_rate": 0.0005,
      "loss": 3.0083811283111572,
      "step": 2591
    },
    {
      "epoch": 0.5995836224843858,
      "grad_norm": 0.8710721135139465,
      "learning_rate": 0.0005,
      "loss": 2.848985433578491,
      "step": 2592
    },
    {
      "epoch": 0.5998149433263937,
      "grad_norm": 1.0494717359542847,
      "learning_rate": 0.0005,
      "loss": 2.830110549926758,
      "step": 2593
    },
    {
      "epoch": 0.6000462641684016,
      "grad_norm": 1.5970529317855835,
      "learning_rate": 0.0005,
      "loss": 2.8158786296844482,
      "step": 2594
    },
    {
      "epoch": 0.6002775850104094,
      "grad_norm": 1.3737138509750366,
      "learning_rate": 0.0005,
      "loss": 2.905698299407959,
      "step": 2595
    },
    {
      "epoch": 0.6005089058524173,
      "grad_norm": 1.164838194847107,
      "learning_rate": 0.0005,
      "loss": 2.748101234436035,
      "step": 2596
    },
    {
      "epoch": 0.6007402266944252,
      "grad_norm": 1.1999906301498413,
      "learning_rate": 0.0005,
      "loss": 2.6555418968200684,
      "step": 2597
    },
    {
      "epoch": 0.600971547536433,
      "grad_norm": 1.2919793128967285,
      "learning_rate": 0.0005,
      "loss": 2.8218064308166504,
      "step": 2598
    },
    {
      "epoch": 0.601202868378441,
      "grad_norm": 1.4727219343185425,
      "learning_rate": 0.0005,
      "loss": 2.9268789291381836,
      "step": 2599
    },
    {
      "epoch": 0.6014341892204488,
      "grad_norm": 1.913244366645813,
      "learning_rate": 0.0005,
      "loss": 3.041548252105713,
      "step": 2600
    },
    {
      "epoch": 0.6016655100624566,
      "grad_norm": 1.2103667259216309,
      "learning_rate": 0.0005,
      "loss": 2.990814685821533,
      "step": 2601
    },
    {
      "epoch": 0.6018968309044644,
      "grad_norm": 1.1918867826461792,
      "learning_rate": 0.0005,
      "loss": 2.5537304878234863,
      "step": 2602
    },
    {
      "epoch": 0.6021281517464724,
      "grad_norm": 1.956162929534912,
      "learning_rate": 0.0005,
      "loss": 2.637526035308838,
      "step": 2603
    },
    {
      "epoch": 0.6023594725884802,
      "grad_norm": 0.9037600159645081,
      "learning_rate": 0.0005,
      "loss": 3.0184340476989746,
      "step": 2604
    },
    {
      "epoch": 0.602590793430488,
      "grad_norm": 1.022444248199463,
      "learning_rate": 0.0005,
      "loss": 2.674156427383423,
      "step": 2605
    },
    {
      "epoch": 0.602822114272496,
      "grad_norm": 1.223500370979309,
      "learning_rate": 0.0005,
      "loss": 2.633180856704712,
      "step": 2606
    },
    {
      "epoch": 0.6030534351145038,
      "grad_norm": 1.0142590999603271,
      "learning_rate": 0.0005,
      "loss": 2.82725191116333,
      "step": 2607
    },
    {
      "epoch": 0.6032847559565117,
      "grad_norm": 0.9866558313369751,
      "learning_rate": 0.0005,
      "loss": 2.996596097946167,
      "step": 2608
    },
    {
      "epoch": 0.6035160767985196,
      "grad_norm": 1.155517339706421,
      "learning_rate": 0.0005,
      "loss": 2.6689224243164062,
      "step": 2609
    },
    {
      "epoch": 0.6037473976405274,
      "grad_norm": 1.073278784751892,
      "learning_rate": 0.0005,
      "loss": 2.566322088241577,
      "step": 2610
    },
    {
      "epoch": 0.6039787184825353,
      "grad_norm": 1.2117828130722046,
      "learning_rate": 0.0005,
      "loss": 3.159907817840576,
      "step": 2611
    },
    {
      "epoch": 0.6042100393245431,
      "grad_norm": 0.9452067017555237,
      "learning_rate": 0.0005,
      "loss": 3.116189956665039,
      "step": 2612
    },
    {
      "epoch": 0.604441360166551,
      "grad_norm": 1.4496549367904663,
      "learning_rate": 0.0005,
      "loss": 3.1436691284179688,
      "step": 2613
    },
    {
      "epoch": 0.6046726810085589,
      "grad_norm": 1.3807733058929443,
      "learning_rate": 0.0005,
      "loss": 2.8923001289367676,
      "step": 2614
    },
    {
      "epoch": 0.6049040018505667,
      "grad_norm": 0.936569333076477,
      "learning_rate": 0.0005,
      "loss": 2.6906728744506836,
      "step": 2615
    },
    {
      "epoch": 0.6051353226925746,
      "grad_norm": 1.1014095544815063,
      "learning_rate": 0.0005,
      "loss": 2.6674447059631348,
      "step": 2616
    },
    {
      "epoch": 0.6053666435345825,
      "grad_norm": 0.9519575238227844,
      "learning_rate": 0.0005,
      "loss": 2.6821627616882324,
      "step": 2617
    },
    {
      "epoch": 0.6055979643765903,
      "grad_norm": 0.8782761693000793,
      "learning_rate": 0.0005,
      "loss": 3.01436185836792,
      "step": 2618
    },
    {
      "epoch": 0.6058292852185982,
      "grad_norm": 1.031514286994934,
      "learning_rate": 0.0005,
      "loss": 2.628753662109375,
      "step": 2619
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.1029613018035889,
      "learning_rate": 0.0005,
      "loss": 2.921133518218994,
      "step": 2620
    },
    {
      "epoch": 0.6062919269026139,
      "grad_norm": 1.0712822675704956,
      "learning_rate": 0.0005,
      "loss": 2.8713977336883545,
      "step": 2621
    },
    {
      "epoch": 0.6065232477446217,
      "grad_norm": 0.8928918838500977,
      "learning_rate": 0.0005,
      "loss": 2.7032856941223145,
      "step": 2622
    },
    {
      "epoch": 0.6067545685866297,
      "grad_norm": 1.0101609230041504,
      "learning_rate": 0.0005,
      "loss": 2.8603365421295166,
      "step": 2623
    },
    {
      "epoch": 0.6069858894286375,
      "grad_norm": 1.2594496011734009,
      "learning_rate": 0.0005,
      "loss": 2.1963067054748535,
      "step": 2624
    },
    {
      "epoch": 0.6072172102706453,
      "grad_norm": 1.279240369796753,
      "learning_rate": 0.0005,
      "loss": 2.677546977996826,
      "step": 2625
    },
    {
      "epoch": 0.6074485311126533,
      "grad_norm": 1.0680856704711914,
      "learning_rate": 0.0005,
      "loss": 2.803037643432617,
      "step": 2626
    },
    {
      "epoch": 0.6076798519546611,
      "grad_norm": 1.2331441640853882,
      "learning_rate": 0.0005,
      "loss": 2.5539462566375732,
      "step": 2627
    },
    {
      "epoch": 0.607911172796669,
      "grad_norm": 1.0729682445526123,
      "learning_rate": 0.0005,
      "loss": 2.904287815093994,
      "step": 2628
    },
    {
      "epoch": 0.6081424936386769,
      "grad_norm": 0.9270147681236267,
      "learning_rate": 0.0005,
      "loss": 2.42378306388855,
      "step": 2629
    },
    {
      "epoch": 0.6083738144806847,
      "grad_norm": 1.0573254823684692,
      "learning_rate": 0.0005,
      "loss": 2.9912304878234863,
      "step": 2630
    },
    {
      "epoch": 0.6086051353226926,
      "grad_norm": 1.1291654109954834,
      "learning_rate": 0.0005,
      "loss": 2.7122201919555664,
      "step": 2631
    },
    {
      "epoch": 0.6088364561647004,
      "grad_norm": 0.9602996706962585,
      "learning_rate": 0.0005,
      "loss": 2.64517879486084,
      "step": 2632
    },
    {
      "epoch": 0.6090677770067083,
      "grad_norm": 1.0447214841842651,
      "learning_rate": 0.0005,
      "loss": 2.684431791305542,
      "step": 2633
    },
    {
      "epoch": 0.6092990978487162,
      "grad_norm": 0.928178608417511,
      "learning_rate": 0.0005,
      "loss": 2.6405534744262695,
      "step": 2634
    },
    {
      "epoch": 0.609530418690724,
      "grad_norm": 1.0126999616622925,
      "learning_rate": 0.0005,
      "loss": 2.4422388076782227,
      "step": 2635
    },
    {
      "epoch": 0.6097617395327319,
      "grad_norm": 1.2465990781784058,
      "learning_rate": 0.0005,
      "loss": 2.9196364879608154,
      "step": 2636
    },
    {
      "epoch": 0.6099930603747398,
      "grad_norm": 1.1301276683807373,
      "learning_rate": 0.0005,
      "loss": 2.620230197906494,
      "step": 2637
    },
    {
      "epoch": 0.6102243812167476,
      "grad_norm": 1.1573024988174438,
      "learning_rate": 0.0005,
      "loss": 2.8638899326324463,
      "step": 2638
    },
    {
      "epoch": 0.6104557020587555,
      "grad_norm": 1.2414475679397583,
      "learning_rate": 0.0005,
      "loss": 2.683709144592285,
      "step": 2639
    },
    {
      "epoch": 0.6106870229007634,
      "grad_norm": 1.2027428150177002,
      "learning_rate": 0.0005,
      "loss": 2.555859327316284,
      "step": 2640
    },
    {
      "epoch": 0.6109183437427712,
      "grad_norm": 1.1767792701721191,
      "learning_rate": 0.0005,
      "loss": 3.117297887802124,
      "step": 2641
    },
    {
      "epoch": 0.611149664584779,
      "grad_norm": 0.9575080871582031,
      "learning_rate": 0.0005,
      "loss": 2.6474204063415527,
      "step": 2642
    },
    {
      "epoch": 0.611380985426787,
      "grad_norm": 1.2371954917907715,
      "learning_rate": 0.0005,
      "loss": 2.695573329925537,
      "step": 2643
    },
    {
      "epoch": 0.6116123062687948,
      "grad_norm": 1.6577377319335938,
      "learning_rate": 0.0005,
      "loss": 3.1686437129974365,
      "step": 2644
    },
    {
      "epoch": 0.6118436271108026,
      "grad_norm": 1.1071139574050903,
      "learning_rate": 0.0005,
      "loss": 2.622023344039917,
      "step": 2645
    },
    {
      "epoch": 0.6120749479528106,
      "grad_norm": 1.0904370546340942,
      "learning_rate": 0.0005,
      "loss": 2.4590232372283936,
      "step": 2646
    },
    {
      "epoch": 0.6123062687948184,
      "grad_norm": 0.9447834491729736,
      "learning_rate": 0.0005,
      "loss": 2.6470038890838623,
      "step": 2647
    },
    {
      "epoch": 0.6125375896368263,
      "grad_norm": 0.9113413691520691,
      "learning_rate": 0.0005,
      "loss": 2.5446577072143555,
      "step": 2648
    },
    {
      "epoch": 0.6127689104788342,
      "grad_norm": 0.8676308393478394,
      "learning_rate": 0.0005,
      "loss": 2.700916290283203,
      "step": 2649
    },
    {
      "epoch": 0.613000231320842,
      "grad_norm": 1.3999196290969849,
      "learning_rate": 0.0005,
      "loss": 2.6002776622772217,
      "step": 2650
    },
    {
      "epoch": 0.6132315521628499,
      "grad_norm": 1.1194103956222534,
      "learning_rate": 0.0005,
      "loss": 2.434490203857422,
      "step": 2651
    },
    {
      "epoch": 0.6134628730048577,
      "grad_norm": 1.0585172176361084,
      "learning_rate": 0.0005,
      "loss": 2.824058771133423,
      "step": 2652
    },
    {
      "epoch": 0.6136941938468656,
      "grad_norm": 1.3830536603927612,
      "learning_rate": 0.0005,
      "loss": 2.86110782623291,
      "step": 2653
    },
    {
      "epoch": 0.6139255146888735,
      "grad_norm": 1.62201988697052,
      "learning_rate": 0.0005,
      "loss": 2.6888625621795654,
      "step": 2654
    },
    {
      "epoch": 0.6141568355308813,
      "grad_norm": 0.9436115622520447,
      "learning_rate": 0.0005,
      "loss": 2.1439592838287354,
      "step": 2655
    },
    {
      "epoch": 0.6143881563728892,
      "grad_norm": 0.8843954205513,
      "learning_rate": 0.0005,
      "loss": 2.7520501613616943,
      "step": 2656
    },
    {
      "epoch": 0.6146194772148971,
      "grad_norm": 1.2447689771652222,
      "learning_rate": 0.0005,
      "loss": 2.8319036960601807,
      "step": 2657
    },
    {
      "epoch": 0.6148507980569049,
      "grad_norm": 1.336094856262207,
      "learning_rate": 0.0005,
      "loss": 3.0611987113952637,
      "step": 2658
    },
    {
      "epoch": 0.6150821188989128,
      "grad_norm": 0.9786916971206665,
      "learning_rate": 0.0005,
      "loss": 2.575920581817627,
      "step": 2659
    },
    {
      "epoch": 0.6153134397409207,
      "grad_norm": 1.4484435319900513,
      "learning_rate": 0.0005,
      "loss": 2.8502674102783203,
      "step": 2660
    },
    {
      "epoch": 0.6155447605829285,
      "grad_norm": 1.5465389490127563,
      "learning_rate": 0.0005,
      "loss": 3.0804080963134766,
      "step": 2661
    },
    {
      "epoch": 0.6157760814249363,
      "grad_norm": 1.1586335897445679,
      "learning_rate": 0.0005,
      "loss": 2.6638941764831543,
      "step": 2662
    },
    {
      "epoch": 0.6160074022669443,
      "grad_norm": 1.1998112201690674,
      "learning_rate": 0.0005,
      "loss": 3.083648920059204,
      "step": 2663
    },
    {
      "epoch": 0.6162387231089521,
      "grad_norm": 1.421269416809082,
      "learning_rate": 0.0005,
      "loss": 2.559542179107666,
      "step": 2664
    },
    {
      "epoch": 0.61647004395096,
      "grad_norm": 0.9830876588821411,
      "learning_rate": 0.0005,
      "loss": 3.1796493530273438,
      "step": 2665
    },
    {
      "epoch": 0.6167013647929679,
      "grad_norm": 1.0015830993652344,
      "learning_rate": 0.0005,
      "loss": 2.641655921936035,
      "step": 2666
    },
    {
      "epoch": 0.6169326856349757,
      "grad_norm": 1.0607770681381226,
      "learning_rate": 0.0005,
      "loss": 2.906266927719116,
      "step": 2667
    },
    {
      "epoch": 0.6171640064769836,
      "grad_norm": 1.092009425163269,
      "learning_rate": 0.0005,
      "loss": 2.795989513397217,
      "step": 2668
    },
    {
      "epoch": 0.6173953273189915,
      "grad_norm": 1.6245139837265015,
      "learning_rate": 0.0005,
      "loss": 3.0838074684143066,
      "step": 2669
    },
    {
      "epoch": 0.6176266481609993,
      "grad_norm": 1.0842170715332031,
      "learning_rate": 0.0005,
      "loss": 2.6047518253326416,
      "step": 2670
    },
    {
      "epoch": 0.6178579690030072,
      "grad_norm": 0.8493924736976624,
      "learning_rate": 0.0005,
      "loss": 2.70835542678833,
      "step": 2671
    },
    {
      "epoch": 0.618089289845015,
      "grad_norm": 1.0887221097946167,
      "learning_rate": 0.0005,
      "loss": 2.8515093326568604,
      "step": 2672
    },
    {
      "epoch": 0.6183206106870229,
      "grad_norm": 1.1508585214614868,
      "learning_rate": 0.0005,
      "loss": 2.8883426189422607,
      "step": 2673
    },
    {
      "epoch": 0.6185519315290308,
      "grad_norm": 0.8603883981704712,
      "learning_rate": 0.0005,
      "loss": 2.9078776836395264,
      "step": 2674
    },
    {
      "epoch": 0.6187832523710386,
      "grad_norm": 1.1175545454025269,
      "learning_rate": 0.0005,
      "loss": 2.967764377593994,
      "step": 2675
    },
    {
      "epoch": 0.6190145732130465,
      "grad_norm": 1.732965111732483,
      "learning_rate": 0.0005,
      "loss": 2.9363532066345215,
      "step": 2676
    },
    {
      "epoch": 0.6192458940550544,
      "grad_norm": 1.0538493394851685,
      "learning_rate": 0.0005,
      "loss": 2.7027177810668945,
      "step": 2677
    },
    {
      "epoch": 0.6194772148970622,
      "grad_norm": 1.0313254594802856,
      "learning_rate": 0.0005,
      "loss": 2.4906165599823,
      "step": 2678
    },
    {
      "epoch": 0.6197085357390701,
      "grad_norm": 1.1992799043655396,
      "learning_rate": 0.0005,
      "loss": 2.4599361419677734,
      "step": 2679
    },
    {
      "epoch": 0.619939856581078,
      "grad_norm": 1.1931405067443848,
      "learning_rate": 0.0005,
      "loss": 2.87491512298584,
      "step": 2680
    },
    {
      "epoch": 0.6201711774230858,
      "grad_norm": 1.0416548252105713,
      "learning_rate": 0.0005,
      "loss": 2.5838637351989746,
      "step": 2681
    },
    {
      "epoch": 0.6204024982650936,
      "grad_norm": 1.284072756767273,
      "learning_rate": 0.0005,
      "loss": 2.47552752494812,
      "step": 2682
    },
    {
      "epoch": 0.6206338191071016,
      "grad_norm": 0.9441689848899841,
      "learning_rate": 0.0005,
      "loss": 2.635018825531006,
      "step": 2683
    },
    {
      "epoch": 0.6208651399491094,
      "grad_norm": 1.3776171207427979,
      "learning_rate": 0.0005,
      "loss": 3.012007713317871,
      "step": 2684
    },
    {
      "epoch": 0.6210964607911172,
      "grad_norm": 1.107731580734253,
      "learning_rate": 0.0005,
      "loss": 2.4396002292633057,
      "step": 2685
    },
    {
      "epoch": 0.6213277816331252,
      "grad_norm": 1.0710947513580322,
      "learning_rate": 0.0005,
      "loss": 2.841435194015503,
      "step": 2686
    },
    {
      "epoch": 0.621559102475133,
      "grad_norm": 1.084213376045227,
      "learning_rate": 0.0005,
      "loss": 2.702084541320801,
      "step": 2687
    },
    {
      "epoch": 0.6217904233171409,
      "grad_norm": 0.8894439935684204,
      "learning_rate": 0.0005,
      "loss": 2.7980847358703613,
      "step": 2688
    },
    {
      "epoch": 0.6220217441591488,
      "grad_norm": 0.794167697429657,
      "learning_rate": 0.0005,
      "loss": 2.3474209308624268,
      "step": 2689
    },
    {
      "epoch": 0.6222530650011566,
      "grad_norm": 1.0126146078109741,
      "learning_rate": 0.0005,
      "loss": 3.2425265312194824,
      "step": 2690
    },
    {
      "epoch": 0.6224843858431645,
      "grad_norm": 1.2897032499313354,
      "learning_rate": 0.0005,
      "loss": 2.667053699493408,
      "step": 2691
    },
    {
      "epoch": 0.6227157066851723,
      "grad_norm": 1.403036117553711,
      "learning_rate": 0.0005,
      "loss": 2.7934138774871826,
      "step": 2692
    },
    {
      "epoch": 0.6229470275271802,
      "grad_norm": 1.082725167274475,
      "learning_rate": 0.0005,
      "loss": 2.6758337020874023,
      "step": 2693
    },
    {
      "epoch": 0.6231783483691881,
      "grad_norm": 1.0720044374465942,
      "learning_rate": 0.0005,
      "loss": 2.879788637161255,
      "step": 2694
    },
    {
      "epoch": 0.6234096692111959,
      "grad_norm": 0.9774022698402405,
      "learning_rate": 0.0005,
      "loss": 2.6996383666992188,
      "step": 2695
    },
    {
      "epoch": 0.6236409900532038,
      "grad_norm": 1.235145926475525,
      "learning_rate": 0.0005,
      "loss": 2.6770811080932617,
      "step": 2696
    },
    {
      "epoch": 0.6238723108952117,
      "grad_norm": 1.0373042821884155,
      "learning_rate": 0.0005,
      "loss": 2.8270654678344727,
      "step": 2697
    },
    {
      "epoch": 0.6241036317372195,
      "grad_norm": 1.0501834154129028,
      "learning_rate": 0.0005,
      "loss": 2.424907922744751,
      "step": 2698
    },
    {
      "epoch": 0.6243349525792274,
      "grad_norm": 1.0993062257766724,
      "learning_rate": 0.0005,
      "loss": 2.346543312072754,
      "step": 2699
    },
    {
      "epoch": 0.6245662734212353,
      "grad_norm": 1.0077568292617798,
      "learning_rate": 0.0005,
      "loss": 2.503931760787964,
      "step": 2700
    },
    {
      "epoch": 0.6247975942632431,
      "grad_norm": 0.9687249064445496,
      "learning_rate": 0.0005,
      "loss": 2.6219887733459473,
      "step": 2701
    },
    {
      "epoch": 0.6250289151052509,
      "grad_norm": 1.0806994438171387,
      "learning_rate": 0.0005,
      "loss": 2.6498448848724365,
      "step": 2702
    },
    {
      "epoch": 0.6252602359472589,
      "grad_norm": 1.1279550790786743,
      "learning_rate": 0.0005,
      "loss": 2.9749302864074707,
      "step": 2703
    },
    {
      "epoch": 0.6254915567892667,
      "grad_norm": 0.9687967300415039,
      "learning_rate": 0.0005,
      "loss": 2.432924270629883,
      "step": 2704
    },
    {
      "epoch": 0.6257228776312745,
      "grad_norm": 1.078627347946167,
      "learning_rate": 0.0005,
      "loss": 2.798460006713867,
      "step": 2705
    },
    {
      "epoch": 0.6259541984732825,
      "grad_norm": 1.0218168497085571,
      "learning_rate": 0.0005,
      "loss": 2.9214277267456055,
      "step": 2706
    },
    {
      "epoch": 0.6261855193152903,
      "grad_norm": 1.1579463481903076,
      "learning_rate": 0.0005,
      "loss": 2.800081491470337,
      "step": 2707
    },
    {
      "epoch": 0.6264168401572981,
      "grad_norm": 1.2412375211715698,
      "learning_rate": 0.0005,
      "loss": 2.5546505451202393,
      "step": 2708
    },
    {
      "epoch": 0.6266481609993061,
      "grad_norm": 0.8998211026191711,
      "learning_rate": 0.0005,
      "loss": 2.6839873790740967,
      "step": 2709
    },
    {
      "epoch": 0.6268794818413139,
      "grad_norm": 1.1378105878829956,
      "learning_rate": 0.0005,
      "loss": 2.718204975128174,
      "step": 2710
    },
    {
      "epoch": 0.6271108026833218,
      "grad_norm": 1.124200701713562,
      "learning_rate": 0.0005,
      "loss": 2.8223273754119873,
      "step": 2711
    },
    {
      "epoch": 0.6273421235253296,
      "grad_norm": 0.9687507748603821,
      "learning_rate": 0.0005,
      "loss": 2.7401556968688965,
      "step": 2712
    },
    {
      "epoch": 0.6275734443673375,
      "grad_norm": 1.0832427740097046,
      "learning_rate": 0.0005,
      "loss": 3.092226266860962,
      "step": 2713
    },
    {
      "epoch": 0.6278047652093454,
      "grad_norm": 0.9154706597328186,
      "learning_rate": 0.0005,
      "loss": 2.6641979217529297,
      "step": 2714
    },
    {
      "epoch": 0.6280360860513532,
      "grad_norm": 0.9713786840438843,
      "learning_rate": 0.0005,
      "loss": 2.683553695678711,
      "step": 2715
    },
    {
      "epoch": 0.6282674068933611,
      "grad_norm": 0.9359776377677917,
      "learning_rate": 0.0005,
      "loss": 2.6841790676116943,
      "step": 2716
    },
    {
      "epoch": 0.628498727735369,
      "grad_norm": 1.0749796628952026,
      "learning_rate": 0.0005,
      "loss": 2.491786003112793,
      "step": 2717
    },
    {
      "epoch": 0.6287300485773768,
      "grad_norm": 1.443023920059204,
      "learning_rate": 0.0005,
      "loss": 2.7144157886505127,
      "step": 2718
    },
    {
      "epoch": 0.6289613694193847,
      "grad_norm": 1.393701434135437,
      "learning_rate": 0.0005,
      "loss": 2.755953788757324,
      "step": 2719
    },
    {
      "epoch": 0.6291926902613926,
      "grad_norm": 1.705281376838684,
      "learning_rate": 0.0005,
      "loss": 2.7760472297668457,
      "step": 2720
    },
    {
      "epoch": 0.6294240111034004,
      "grad_norm": 1.3572591543197632,
      "learning_rate": 0.0005,
      "loss": 2.9982855319976807,
      "step": 2721
    },
    {
      "epoch": 0.6296553319454082,
      "grad_norm": 1.0592186450958252,
      "learning_rate": 0.0005,
      "loss": 2.563931465148926,
      "step": 2722
    },
    {
      "epoch": 0.6298866527874162,
      "grad_norm": 1.1029607057571411,
      "learning_rate": 0.0005,
      "loss": 2.768907070159912,
      "step": 2723
    },
    {
      "epoch": 0.630117973629424,
      "grad_norm": 0.8603317737579346,
      "learning_rate": 0.0005,
      "loss": 2.7488839626312256,
      "step": 2724
    },
    {
      "epoch": 0.6303492944714318,
      "grad_norm": 1.0552241802215576,
      "learning_rate": 0.0005,
      "loss": 2.7158751487731934,
      "step": 2725
    },
    {
      "epoch": 0.6305806153134398,
      "grad_norm": 1.0608376264572144,
      "learning_rate": 0.0005,
      "loss": 2.3444371223449707,
      "step": 2726
    },
    {
      "epoch": 0.6308119361554476,
      "grad_norm": 0.9491751194000244,
      "learning_rate": 0.0005,
      "loss": 2.653473377227783,
      "step": 2727
    },
    {
      "epoch": 0.6310432569974554,
      "grad_norm": 1.1010559797286987,
      "learning_rate": 0.0005,
      "loss": 2.887881278991699,
      "step": 2728
    },
    {
      "epoch": 0.6312745778394634,
      "grad_norm": 1.0888535976409912,
      "learning_rate": 0.0005,
      "loss": 2.78527569770813,
      "step": 2729
    },
    {
      "epoch": 0.6315058986814712,
      "grad_norm": 1.0727955102920532,
      "learning_rate": 0.0005,
      "loss": 2.321718692779541,
      "step": 2730
    },
    {
      "epoch": 0.631737219523479,
      "grad_norm": 1.0376452207565308,
      "learning_rate": 0.0005,
      "loss": 2.8387558460235596,
      "step": 2731
    },
    {
      "epoch": 0.6319685403654869,
      "grad_norm": 0.9738398194313049,
      "learning_rate": 0.0005,
      "loss": 3.2709460258483887,
      "step": 2732
    },
    {
      "epoch": 0.6321998612074948,
      "grad_norm": 1.216416358947754,
      "learning_rate": 0.0005,
      "loss": 2.596132516860962,
      "step": 2733
    },
    {
      "epoch": 0.6324311820495027,
      "grad_norm": 0.9795436859130859,
      "learning_rate": 0.0005,
      "loss": 2.5563440322875977,
      "step": 2734
    },
    {
      "epoch": 0.6326625028915105,
      "grad_norm": 0.8979147672653198,
      "learning_rate": 0.0005,
      "loss": 2.684321165084839,
      "step": 2735
    },
    {
      "epoch": 0.6328938237335184,
      "grad_norm": 1.502485752105713,
      "learning_rate": 0.0005,
      "loss": 2.705533504486084,
      "step": 2736
    },
    {
      "epoch": 0.6331251445755263,
      "grad_norm": 1.3353748321533203,
      "learning_rate": 0.0005,
      "loss": 2.814807891845703,
      "step": 2737
    },
    {
      "epoch": 0.6333564654175341,
      "grad_norm": 1.0745432376861572,
      "learning_rate": 0.0005,
      "loss": 2.570904493331909,
      "step": 2738
    },
    {
      "epoch": 0.6335877862595419,
      "grad_norm": 1.218862771987915,
      "learning_rate": 0.0005,
      "loss": 2.9427568912506104,
      "step": 2739
    },
    {
      "epoch": 0.6338191071015499,
      "grad_norm": 1.1196552515029907,
      "learning_rate": 0.0005,
      "loss": 2.69006609916687,
      "step": 2740
    },
    {
      "epoch": 0.6340504279435577,
      "grad_norm": 1.3261297941207886,
      "learning_rate": 0.0005,
      "loss": 2.447645425796509,
      "step": 2741
    },
    {
      "epoch": 0.6342817487855655,
      "grad_norm": 0.9949273467063904,
      "learning_rate": 0.0005,
      "loss": 2.5309412479400635,
      "step": 2742
    },
    {
      "epoch": 0.6345130696275735,
      "grad_norm": 1.171584129333496,
      "learning_rate": 0.0005,
      "loss": 3.0956029891967773,
      "step": 2743
    },
    {
      "epoch": 0.6347443904695813,
      "grad_norm": 0.8821157217025757,
      "learning_rate": 0.0005,
      "loss": 2.9338436126708984,
      "step": 2744
    },
    {
      "epoch": 0.6349757113115891,
      "grad_norm": 1.1913127899169922,
      "learning_rate": 0.0005,
      "loss": 2.817011594772339,
      "step": 2745
    },
    {
      "epoch": 0.6352070321535971,
      "grad_norm": 0.9840097427368164,
      "learning_rate": 0.0005,
      "loss": 2.6865296363830566,
      "step": 2746
    },
    {
      "epoch": 0.6354383529956049,
      "grad_norm": 0.9328809976577759,
      "learning_rate": 0.0005,
      "loss": 2.683610200881958,
      "step": 2747
    },
    {
      "epoch": 0.6356696738376127,
      "grad_norm": 0.9698749780654907,
      "learning_rate": 0.0005,
      "loss": 2.7287402153015137,
      "step": 2748
    },
    {
      "epoch": 0.6359009946796206,
      "grad_norm": 1.0241920948028564,
      "learning_rate": 0.0005,
      "loss": 2.8235299587249756,
      "step": 2749
    },
    {
      "epoch": 0.6361323155216285,
      "grad_norm": 1.189167857170105,
      "learning_rate": 0.0005,
      "loss": 2.872616767883301,
      "step": 2750
    },
    {
      "epoch": 0.6363636363636364,
      "grad_norm": 1.060959815979004,
      "learning_rate": 0.0005,
      "loss": 3.046642541885376,
      "step": 2751
    },
    {
      "epoch": 0.6365949572056442,
      "grad_norm": 0.9321038722991943,
      "learning_rate": 0.0005,
      "loss": 3.0469367504119873,
      "step": 2752
    },
    {
      "epoch": 0.6368262780476521,
      "grad_norm": 1.0685136318206787,
      "learning_rate": 0.0005,
      "loss": 2.5332984924316406,
      "step": 2753
    },
    {
      "epoch": 0.63705759888966,
      "grad_norm": 1.2043147087097168,
      "learning_rate": 0.0005,
      "loss": 3.040540933609009,
      "step": 2754
    },
    {
      "epoch": 0.6372889197316678,
      "grad_norm": 0.9069989323616028,
      "learning_rate": 0.0005,
      "loss": 2.971829891204834,
      "step": 2755
    },
    {
      "epoch": 0.6375202405736757,
      "grad_norm": 1.1674238443374634,
      "learning_rate": 0.0005,
      "loss": 2.618109703063965,
      "step": 2756
    },
    {
      "epoch": 0.6377515614156836,
      "grad_norm": 0.9751627445220947,
      "learning_rate": 0.0005,
      "loss": 2.6784141063690186,
      "step": 2757
    },
    {
      "epoch": 0.6379828822576914,
      "grad_norm": 1.4948146343231201,
      "learning_rate": 0.0005,
      "loss": 2.6856117248535156,
      "step": 2758
    },
    {
      "epoch": 0.6382142030996992,
      "grad_norm": 1.1590046882629395,
      "learning_rate": 0.0005,
      "loss": 2.826803684234619,
      "step": 2759
    },
    {
      "epoch": 0.6384455239417072,
      "grad_norm": 1.0957998037338257,
      "learning_rate": 0.0005,
      "loss": 2.791102886199951,
      "step": 2760
    },
    {
      "epoch": 0.638676844783715,
      "grad_norm": 1.049924373626709,
      "learning_rate": 0.0005,
      "loss": 2.865090847015381,
      "step": 2761
    },
    {
      "epoch": 0.6389081656257228,
      "grad_norm": 1.0120601654052734,
      "learning_rate": 0.0005,
      "loss": 2.6523079872131348,
      "step": 2762
    },
    {
      "epoch": 0.6391394864677308,
      "grad_norm": 1.1790980100631714,
      "learning_rate": 0.0005,
      "loss": 2.833889961242676,
      "step": 2763
    },
    {
      "epoch": 0.6393708073097386,
      "grad_norm": 1.3849741220474243,
      "learning_rate": 0.0005,
      "loss": 2.7443947792053223,
      "step": 2764
    },
    {
      "epoch": 0.6396021281517464,
      "grad_norm": 1.0889911651611328,
      "learning_rate": 0.0005,
      "loss": 2.501577854156494,
      "step": 2765
    },
    {
      "epoch": 0.6398334489937544,
      "grad_norm": 1.0306029319763184,
      "learning_rate": 0.0005,
      "loss": 2.593264579772949,
      "step": 2766
    },
    {
      "epoch": 0.6400647698357622,
      "grad_norm": 1.004747986793518,
      "learning_rate": 0.0005,
      "loss": 2.8364224433898926,
      "step": 2767
    },
    {
      "epoch": 0.64029609067777,
      "grad_norm": 1.0010242462158203,
      "learning_rate": 0.0005,
      "loss": 2.874990463256836,
      "step": 2768
    },
    {
      "epoch": 0.6405274115197779,
      "grad_norm": 0.926930844783783,
      "learning_rate": 0.0005,
      "loss": 2.735851764678955,
      "step": 2769
    },
    {
      "epoch": 0.6407587323617858,
      "grad_norm": 0.897066056728363,
      "learning_rate": 0.0005,
      "loss": 2.9532675743103027,
      "step": 2770
    },
    {
      "epoch": 0.6409900532037937,
      "grad_norm": 1.0081171989440918,
      "learning_rate": 0.0005,
      "loss": 2.7666618824005127,
      "step": 2771
    },
    {
      "epoch": 0.6412213740458015,
      "grad_norm": 0.9027976393699646,
      "learning_rate": 0.0005,
      "loss": 3.051090717315674,
      "step": 2772
    },
    {
      "epoch": 0.6414526948878094,
      "grad_norm": 0.8843549489974976,
      "learning_rate": 0.0005,
      "loss": 2.568084955215454,
      "step": 2773
    },
    {
      "epoch": 0.6416840157298173,
      "grad_norm": 1.352707028388977,
      "learning_rate": 0.0005,
      "loss": 2.903343439102173,
      "step": 2774
    },
    {
      "epoch": 0.6419153365718251,
      "grad_norm": 1.047837257385254,
      "learning_rate": 0.0005,
      "loss": 2.74333119392395,
      "step": 2775
    },
    {
      "epoch": 0.642146657413833,
      "grad_norm": 1.029860258102417,
      "learning_rate": 0.0005,
      "loss": 2.556300640106201,
      "step": 2776
    },
    {
      "epoch": 0.6423779782558409,
      "grad_norm": 0.8472934365272522,
      "learning_rate": 0.0005,
      "loss": 2.465271472930908,
      "step": 2777
    },
    {
      "epoch": 0.6426092990978487,
      "grad_norm": 1.0382276773452759,
      "learning_rate": 0.0005,
      "loss": 3.2616567611694336,
      "step": 2778
    },
    {
      "epoch": 0.6428406199398565,
      "grad_norm": 0.9873965382575989,
      "learning_rate": 0.0005,
      "loss": 2.5283944606781006,
      "step": 2779
    },
    {
      "epoch": 0.6430719407818645,
      "grad_norm": 1.0313798189163208,
      "learning_rate": 0.0005,
      "loss": 2.880903720855713,
      "step": 2780
    },
    {
      "epoch": 0.6433032616238723,
      "grad_norm": 1.348597764968872,
      "learning_rate": 0.0005,
      "loss": 2.697678565979004,
      "step": 2781
    },
    {
      "epoch": 0.6435345824658801,
      "grad_norm": 1.1002851724624634,
      "learning_rate": 0.0005,
      "loss": 2.4300832748413086,
      "step": 2782
    },
    {
      "epoch": 0.6437659033078881,
      "grad_norm": 1.353371500968933,
      "learning_rate": 0.0005,
      "loss": 2.6451830863952637,
      "step": 2783
    },
    {
      "epoch": 0.6439972241498959,
      "grad_norm": 1.265244483947754,
      "learning_rate": 0.0005,
      "loss": 2.992783784866333,
      "step": 2784
    },
    {
      "epoch": 0.6442285449919037,
      "grad_norm": 1.2379192113876343,
      "learning_rate": 0.0005,
      "loss": 2.6180901527404785,
      "step": 2785
    },
    {
      "epoch": 0.6444598658339117,
      "grad_norm": 1.5660357475280762,
      "learning_rate": 0.0005,
      "loss": 2.6334519386291504,
      "step": 2786
    },
    {
      "epoch": 0.6446911866759195,
      "grad_norm": 1.0503878593444824,
      "learning_rate": 0.0005,
      "loss": 3.016176223754883,
      "step": 2787
    },
    {
      "epoch": 0.6449225075179273,
      "grad_norm": 0.9171252250671387,
      "learning_rate": 0.0005,
      "loss": 2.7855207920074463,
      "step": 2788
    },
    {
      "epoch": 0.6451538283599352,
      "grad_norm": 0.8948739171028137,
      "learning_rate": 0.0005,
      "loss": 2.769641399383545,
      "step": 2789
    },
    {
      "epoch": 0.6453851492019431,
      "grad_norm": 1.0776158571243286,
      "learning_rate": 0.0005,
      "loss": 2.8180811405181885,
      "step": 2790
    },
    {
      "epoch": 0.645616470043951,
      "grad_norm": 1.0885592699050903,
      "learning_rate": 0.0005,
      "loss": 3.0269651412963867,
      "step": 2791
    },
    {
      "epoch": 0.6458477908859588,
      "grad_norm": 0.87969970703125,
      "learning_rate": 0.0005,
      "loss": 2.519507884979248,
      "step": 2792
    },
    {
      "epoch": 0.6460791117279667,
      "grad_norm": 0.9052734971046448,
      "learning_rate": 0.0005,
      "loss": 2.669131278991699,
      "step": 2793
    },
    {
      "epoch": 0.6463104325699746,
      "grad_norm": 1.1694912910461426,
      "learning_rate": 0.0005,
      "loss": 2.7637217044830322,
      "step": 2794
    },
    {
      "epoch": 0.6465417534119824,
      "grad_norm": 0.8147987127304077,
      "learning_rate": 0.0005,
      "loss": 3.0110085010528564,
      "step": 2795
    },
    {
      "epoch": 0.6467730742539903,
      "grad_norm": 0.914856493473053,
      "learning_rate": 0.0005,
      "loss": 2.722717046737671,
      "step": 2796
    },
    {
      "epoch": 0.6470043950959982,
      "grad_norm": 0.9276297688484192,
      "learning_rate": 0.0005,
      "loss": 2.336925983428955,
      "step": 2797
    },
    {
      "epoch": 0.647235715938006,
      "grad_norm": 1.2503201961517334,
      "learning_rate": 0.0005,
      "loss": 2.4987096786499023,
      "step": 2798
    },
    {
      "epoch": 0.6474670367800138,
      "grad_norm": 1.2110745906829834,
      "learning_rate": 0.0005,
      "loss": 2.8506882190704346,
      "step": 2799
    },
    {
      "epoch": 0.6476983576220218,
      "grad_norm": 1.0314686298370361,
      "learning_rate": 0.0005,
      "loss": 2.757363796234131,
      "step": 2800
    },
    {
      "epoch": 0.6479296784640296,
      "grad_norm": 4.113955497741699,
      "learning_rate": 0.0005,
      "loss": 2.7394535541534424,
      "step": 2801
    },
    {
      "epoch": 0.6481609993060374,
      "grad_norm": 0.9900126457214355,
      "learning_rate": 0.0005,
      "loss": 3.0992722511291504,
      "step": 2802
    },
    {
      "epoch": 0.6483923201480454,
      "grad_norm": 1.0958549976348877,
      "learning_rate": 0.0005,
      "loss": 2.485426902770996,
      "step": 2803
    },
    {
      "epoch": 0.6486236409900532,
      "grad_norm": 1.3348708152770996,
      "learning_rate": 0.0005,
      "loss": 2.644649028778076,
      "step": 2804
    },
    {
      "epoch": 0.648854961832061,
      "grad_norm": 0.8386186361312866,
      "learning_rate": 0.0005,
      "loss": 2.3413705825805664,
      "step": 2805
    },
    {
      "epoch": 0.649086282674069,
      "grad_norm": 1.0895154476165771,
      "learning_rate": 0.0005,
      "loss": 2.4575157165527344,
      "step": 2806
    },
    {
      "epoch": 0.6493176035160768,
      "grad_norm": 1.8479535579681396,
      "learning_rate": 0.0005,
      "loss": 2.971235513687134,
      "step": 2807
    },
    {
      "epoch": 0.6495489243580846,
      "grad_norm": 1.044174075126648,
      "learning_rate": 0.0005,
      "loss": 3.0398316383361816,
      "step": 2808
    },
    {
      "epoch": 0.6497802452000925,
      "grad_norm": 0.885717511177063,
      "learning_rate": 0.0005,
      "loss": 2.5691652297973633,
      "step": 2809
    },
    {
      "epoch": 0.6500115660421004,
      "grad_norm": 0.7195972800254822,
      "learning_rate": 0.0005,
      "loss": 2.6712327003479004,
      "step": 2810
    },
    {
      "epoch": 0.6502428868841083,
      "grad_norm": 0.9404656887054443,
      "learning_rate": 0.0005,
      "loss": 2.4675135612487793,
      "step": 2811
    },
    {
      "epoch": 0.6504742077261161,
      "grad_norm": 0.9428961277008057,
      "learning_rate": 0.0005,
      "loss": 2.9590840339660645,
      "step": 2812
    },
    {
      "epoch": 0.650705528568124,
      "grad_norm": 1.2377904653549194,
      "learning_rate": 0.0005,
      "loss": 2.724405288696289,
      "step": 2813
    },
    {
      "epoch": 0.6509368494101319,
      "grad_norm": 1.3294227123260498,
      "learning_rate": 0.0005,
      "loss": 2.7494678497314453,
      "step": 2814
    },
    {
      "epoch": 0.6511681702521397,
      "grad_norm": 1.1800974607467651,
      "learning_rate": 0.0005,
      "loss": 2.45745587348938,
      "step": 2815
    },
    {
      "epoch": 0.6513994910941476,
      "grad_norm": 1.4933582544326782,
      "learning_rate": 0.0005,
      "loss": 2.833977699279785,
      "step": 2816
    },
    {
      "epoch": 0.6516308119361555,
      "grad_norm": 1.0827640295028687,
      "learning_rate": 0.0005,
      "loss": 2.703397512435913,
      "step": 2817
    },
    {
      "epoch": 0.6518621327781633,
      "grad_norm": 0.9968701601028442,
      "learning_rate": 0.0005,
      "loss": 2.702552080154419,
      "step": 2818
    },
    {
      "epoch": 0.6520934536201711,
      "grad_norm": 0.9242246747016907,
      "learning_rate": 0.0005,
      "loss": 2.8308446407318115,
      "step": 2819
    },
    {
      "epoch": 0.6523247744621791,
      "grad_norm": 0.9918711185455322,
      "learning_rate": 0.0005,
      "loss": 2.6659297943115234,
      "step": 2820
    },
    {
      "epoch": 0.6525560953041869,
      "grad_norm": 1.5894951820373535,
      "learning_rate": 0.0005,
      "loss": 2.800945281982422,
      "step": 2821
    },
    {
      "epoch": 0.6527874161461947,
      "grad_norm": 0.8226872086524963,
      "learning_rate": 0.0005,
      "loss": 2.536961078643799,
      "step": 2822
    },
    {
      "epoch": 0.6530187369882027,
      "grad_norm": 0.9347904920578003,
      "learning_rate": 0.0005,
      "loss": 2.7437379360198975,
      "step": 2823
    },
    {
      "epoch": 0.6532500578302105,
      "grad_norm": 0.899600088596344,
      "learning_rate": 0.0005,
      "loss": 2.6721487045288086,
      "step": 2824
    },
    {
      "epoch": 0.6534813786722183,
      "grad_norm": 0.9495742321014404,
      "learning_rate": 0.0005,
      "loss": 3.0417914390563965,
      "step": 2825
    },
    {
      "epoch": 0.6537126995142263,
      "grad_norm": 1.1624658107757568,
      "learning_rate": 0.0005,
      "loss": 2.8932418823242188,
      "step": 2826
    },
    {
      "epoch": 0.6539440203562341,
      "grad_norm": 1.255346417427063,
      "learning_rate": 0.0005,
      "loss": 2.8235228061676025,
      "step": 2827
    },
    {
      "epoch": 0.6541753411982419,
      "grad_norm": 1.1760053634643555,
      "learning_rate": 0.0005,
      "loss": 2.9524402618408203,
      "step": 2828
    },
    {
      "epoch": 0.6544066620402498,
      "grad_norm": 0.939964771270752,
      "learning_rate": 0.0005,
      "loss": 2.9444637298583984,
      "step": 2829
    },
    {
      "epoch": 0.6546379828822577,
      "grad_norm": 1.0088801383972168,
      "learning_rate": 0.0005,
      "loss": 2.7553420066833496,
      "step": 2830
    },
    {
      "epoch": 0.6548693037242656,
      "grad_norm": 1.2141979932785034,
      "learning_rate": 0.0005,
      "loss": 2.641862392425537,
      "step": 2831
    },
    {
      "epoch": 0.6551006245662734,
      "grad_norm": 1.1131055355072021,
      "learning_rate": 0.0005,
      "loss": 2.6620945930480957,
      "step": 2832
    },
    {
      "epoch": 0.6553319454082813,
      "grad_norm": 1.0646474361419678,
      "learning_rate": 0.0005,
      "loss": 2.697303295135498,
      "step": 2833
    },
    {
      "epoch": 0.6555632662502892,
      "grad_norm": 1.1766691207885742,
      "learning_rate": 0.0005,
      "loss": 2.8995614051818848,
      "step": 2834
    },
    {
      "epoch": 0.655794587092297,
      "grad_norm": 1.3574280738830566,
      "learning_rate": 0.0005,
      "loss": 3.012646198272705,
      "step": 2835
    },
    {
      "epoch": 0.6560259079343049,
      "grad_norm": 1.073929786682129,
      "learning_rate": 0.0005,
      "loss": 2.7486331462860107,
      "step": 2836
    },
    {
      "epoch": 0.6562572287763128,
      "grad_norm": 1.113810420036316,
      "learning_rate": 0.0005,
      "loss": 2.8991472721099854,
      "step": 2837
    },
    {
      "epoch": 0.6564885496183206,
      "grad_norm": 1.0139422416687012,
      "learning_rate": 0.0005,
      "loss": 3.009906530380249,
      "step": 2838
    },
    {
      "epoch": 0.6567198704603284,
      "grad_norm": 0.9826388955116272,
      "learning_rate": 0.0005,
      "loss": 3.1883840560913086,
      "step": 2839
    },
    {
      "epoch": 0.6569511913023364,
      "grad_norm": 1.9121702909469604,
      "learning_rate": 0.0005,
      "loss": 3.157569169998169,
      "step": 2840
    },
    {
      "epoch": 0.6571825121443442,
      "grad_norm": 1.0472183227539062,
      "learning_rate": 0.0005,
      "loss": 2.646242141723633,
      "step": 2841
    },
    {
      "epoch": 0.657413832986352,
      "grad_norm": 1.4330823421478271,
      "learning_rate": 0.0005,
      "loss": 2.391875743865967,
      "step": 2842
    },
    {
      "epoch": 0.65764515382836,
      "grad_norm": 1.1063241958618164,
      "learning_rate": 0.0005,
      "loss": 3.048440933227539,
      "step": 2843
    },
    {
      "epoch": 0.6578764746703678,
      "grad_norm": 1.1445704698562622,
      "learning_rate": 0.0005,
      "loss": 3.064850330352783,
      "step": 2844
    },
    {
      "epoch": 0.6581077955123756,
      "grad_norm": 1.0101673603057861,
      "learning_rate": 0.0005,
      "loss": 2.9165241718292236,
      "step": 2845
    },
    {
      "epoch": 0.6583391163543836,
      "grad_norm": 0.8107626438140869,
      "learning_rate": 0.0005,
      "loss": 2.4102120399475098,
      "step": 2846
    },
    {
      "epoch": 0.6585704371963914,
      "grad_norm": 0.9846582412719727,
      "learning_rate": 0.0005,
      "loss": 2.597182035446167,
      "step": 2847
    },
    {
      "epoch": 0.6588017580383992,
      "grad_norm": 1.0199159383773804,
      "learning_rate": 0.0005,
      "loss": 2.808751106262207,
      "step": 2848
    },
    {
      "epoch": 0.6590330788804071,
      "grad_norm": 1.1832913160324097,
      "learning_rate": 0.0005,
      "loss": 2.7026138305664062,
      "step": 2849
    },
    {
      "epoch": 0.659264399722415,
      "grad_norm": 1.431774377822876,
      "learning_rate": 0.0005,
      "loss": 2.92340350151062,
      "step": 2850
    },
    {
      "epoch": 0.6594957205644228,
      "grad_norm": 1.019486665725708,
      "learning_rate": 0.0005,
      "loss": 2.6730799674987793,
      "step": 2851
    },
    {
      "epoch": 0.6597270414064307,
      "grad_norm": 1.3406740427017212,
      "learning_rate": 0.0005,
      "loss": 2.8072609901428223,
      "step": 2852
    },
    {
      "epoch": 0.6599583622484386,
      "grad_norm": 1.004019021987915,
      "learning_rate": 0.0005,
      "loss": 2.790334939956665,
      "step": 2853
    },
    {
      "epoch": 0.6601896830904465,
      "grad_norm": 4.87942361831665,
      "learning_rate": 0.0005,
      "loss": 2.545767307281494,
      "step": 2854
    },
    {
      "epoch": 0.6604210039324543,
      "grad_norm": 1.4210511445999146,
      "learning_rate": 0.0005,
      "loss": 2.9722509384155273,
      "step": 2855
    },
    {
      "epoch": 0.6606523247744622,
      "grad_norm": 1.0618313550949097,
      "learning_rate": 0.0005,
      "loss": 2.746443271636963,
      "step": 2856
    },
    {
      "epoch": 0.6608836456164701,
      "grad_norm": 1.1139687299728394,
      "learning_rate": 0.0005,
      "loss": 2.3803458213806152,
      "step": 2857
    },
    {
      "epoch": 0.6611149664584779,
      "grad_norm": 1.010893702507019,
      "learning_rate": 0.0005,
      "loss": 2.608443260192871,
      "step": 2858
    },
    {
      "epoch": 0.6613462873004857,
      "grad_norm": 1.2493661642074585,
      "learning_rate": 0.0005,
      "loss": 2.801642417907715,
      "step": 2859
    },
    {
      "epoch": 0.6615776081424937,
      "grad_norm": 0.9204928278923035,
      "learning_rate": 0.0005,
      "loss": 2.453909397125244,
      "step": 2860
    },
    {
      "epoch": 0.6618089289845015,
      "grad_norm": 1.0381594896316528,
      "learning_rate": 0.0005,
      "loss": 2.397891044616699,
      "step": 2861
    },
    {
      "epoch": 0.6620402498265093,
      "grad_norm": 1.0866097211837769,
      "learning_rate": 0.0005,
      "loss": 2.609691619873047,
      "step": 2862
    },
    {
      "epoch": 0.6622715706685173,
      "grad_norm": 0.866131603717804,
      "learning_rate": 0.0005,
      "loss": 2.673877239227295,
      "step": 2863
    },
    {
      "epoch": 0.6625028915105251,
      "grad_norm": 1.0465701818466187,
      "learning_rate": 0.0005,
      "loss": 2.7381999492645264,
      "step": 2864
    },
    {
      "epoch": 0.6627342123525329,
      "grad_norm": 0.9719248414039612,
      "learning_rate": 0.0005,
      "loss": 2.607548713684082,
      "step": 2865
    },
    {
      "epoch": 0.6629655331945409,
      "grad_norm": 1.2555229663848877,
      "learning_rate": 0.0005,
      "loss": 2.4745984077453613,
      "step": 2866
    },
    {
      "epoch": 0.6631968540365487,
      "grad_norm": 0.9625316262245178,
      "learning_rate": 0.0005,
      "loss": 2.7290682792663574,
      "step": 2867
    },
    {
      "epoch": 0.6634281748785565,
      "grad_norm": 1.1984882354736328,
      "learning_rate": 0.0005,
      "loss": 2.6769728660583496,
      "step": 2868
    },
    {
      "epoch": 0.6636594957205644,
      "grad_norm": 1.1642922163009644,
      "learning_rate": 0.0005,
      "loss": 2.573852062225342,
      "step": 2869
    },
    {
      "epoch": 0.6638908165625723,
      "grad_norm": 0.9226065874099731,
      "learning_rate": 0.0005,
      "loss": 2.8609986305236816,
      "step": 2870
    },
    {
      "epoch": 0.6641221374045801,
      "grad_norm": 1.078739881515503,
      "learning_rate": 0.0005,
      "loss": 2.733057737350464,
      "step": 2871
    },
    {
      "epoch": 0.664353458246588,
      "grad_norm": 1.276492714881897,
      "learning_rate": 0.0005,
      "loss": 2.6763393878936768,
      "step": 2872
    },
    {
      "epoch": 0.6645847790885959,
      "grad_norm": 0.9267520308494568,
      "learning_rate": 0.0005,
      "loss": 2.4611706733703613,
      "step": 2873
    },
    {
      "epoch": 0.6648160999306038,
      "grad_norm": 1.2088559865951538,
      "learning_rate": 0.0005,
      "loss": 3.023077964782715,
      "step": 2874
    },
    {
      "epoch": 0.6650474207726116,
      "grad_norm": 1.2127453088760376,
      "learning_rate": 0.0005,
      "loss": 2.8799571990966797,
      "step": 2875
    },
    {
      "epoch": 0.6652787416146195,
      "grad_norm": 1.2567999362945557,
      "learning_rate": 0.0005,
      "loss": 2.6214661598205566,
      "step": 2876
    },
    {
      "epoch": 0.6655100624566274,
      "grad_norm": 1.5998873710632324,
      "learning_rate": 0.0005,
      "loss": 2.706652879714966,
      "step": 2877
    },
    {
      "epoch": 0.6657413832986352,
      "grad_norm": 1.158515453338623,
      "learning_rate": 0.0005,
      "loss": 2.343027114868164,
      "step": 2878
    },
    {
      "epoch": 0.665972704140643,
      "grad_norm": 1.137713074684143,
      "learning_rate": 0.0005,
      "loss": 2.911238431930542,
      "step": 2879
    },
    {
      "epoch": 0.666204024982651,
      "grad_norm": 1.3035403490066528,
      "learning_rate": 0.0005,
      "loss": 2.853120803833008,
      "step": 2880
    },
    {
      "epoch": 0.6664353458246588,
      "grad_norm": 1.6850428581237793,
      "learning_rate": 0.0005,
      "loss": 2.953219175338745,
      "step": 2881
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.978403627872467,
      "learning_rate": 0.0005,
      "loss": 2.79744029045105,
      "step": 2882
    },
    {
      "epoch": 0.6668979875086746,
      "grad_norm": 1.0107417106628418,
      "learning_rate": 0.0005,
      "loss": 2.629270553588867,
      "step": 2883
    },
    {
      "epoch": 0.6671293083506824,
      "grad_norm": 0.9910312294960022,
      "learning_rate": 0.0005,
      "loss": 2.7556419372558594,
      "step": 2884
    },
    {
      "epoch": 0.6673606291926902,
      "grad_norm": 1.5713413953781128,
      "learning_rate": 0.0005,
      "loss": 2.574256420135498,
      "step": 2885
    },
    {
      "epoch": 0.6675919500346982,
      "grad_norm": 1.6901730298995972,
      "learning_rate": 0.0005,
      "loss": 2.6972663402557373,
      "step": 2886
    },
    {
      "epoch": 0.667823270876706,
      "grad_norm": 0.852180540561676,
      "learning_rate": 0.0005,
      "loss": 2.418722629547119,
      "step": 2887
    },
    {
      "epoch": 0.6680545917187138,
      "grad_norm": 0.8997339606285095,
      "learning_rate": 0.0005,
      "loss": 2.5791311264038086,
      "step": 2888
    },
    {
      "epoch": 0.6682859125607217,
      "grad_norm": 1.2530120611190796,
      "learning_rate": 0.0005,
      "loss": 2.576345682144165,
      "step": 2889
    },
    {
      "epoch": 0.6685172334027296,
      "grad_norm": 1.4746557474136353,
      "learning_rate": 0.0005,
      "loss": 2.6377952098846436,
      "step": 2890
    },
    {
      "epoch": 0.6687485542447374,
      "grad_norm": 0.9751918315887451,
      "learning_rate": 0.0005,
      "loss": 2.6226882934570312,
      "step": 2891
    },
    {
      "epoch": 0.6689798750867453,
      "grad_norm": 1.2331364154815674,
      "learning_rate": 0.0005,
      "loss": 2.7933194637298584,
      "step": 2892
    },
    {
      "epoch": 0.6692111959287532,
      "grad_norm": 1.0874022245407104,
      "learning_rate": 0.0005,
      "loss": 2.567601203918457,
      "step": 2893
    },
    {
      "epoch": 0.669442516770761,
      "grad_norm": 1.4377484321594238,
      "learning_rate": 0.0005,
      "loss": 2.7516486644744873,
      "step": 2894
    },
    {
      "epoch": 0.6696738376127689,
      "grad_norm": 0.9314296245574951,
      "learning_rate": 0.0005,
      "loss": 2.56339168548584,
      "step": 2895
    },
    {
      "epoch": 0.6699051584547768,
      "grad_norm": 1.4821043014526367,
      "learning_rate": 0.0005,
      "loss": 2.5316433906555176,
      "step": 2896
    },
    {
      "epoch": 0.6701364792967847,
      "grad_norm": 0.9625817537307739,
      "learning_rate": 0.0005,
      "loss": 2.9554145336151123,
      "step": 2897
    },
    {
      "epoch": 0.6703678001387925,
      "grad_norm": 1.0671923160552979,
      "learning_rate": 0.0005,
      "loss": 2.5698800086975098,
      "step": 2898
    },
    {
      "epoch": 0.6705991209808003,
      "grad_norm": 2.7375755310058594,
      "learning_rate": 0.0005,
      "loss": 2.2487411499023438,
      "step": 2899
    },
    {
      "epoch": 0.6708304418228083,
      "grad_norm": 1.1692255735397339,
      "learning_rate": 0.0005,
      "loss": 2.422484874725342,
      "step": 2900
    },
    {
      "epoch": 0.6710617626648161,
      "grad_norm": 1.1675537824630737,
      "learning_rate": 0.0005,
      "loss": 2.5470056533813477,
      "step": 2901
    },
    {
      "epoch": 0.6712930835068239,
      "grad_norm": 1.0532417297363281,
      "learning_rate": 0.0005,
      "loss": 2.562284469604492,
      "step": 2902
    },
    {
      "epoch": 0.6715244043488319,
      "grad_norm": 1.1195751428604126,
      "learning_rate": 0.0005,
      "loss": 2.873121500015259,
      "step": 2903
    },
    {
      "epoch": 0.6717557251908397,
      "grad_norm": 1.1887414455413818,
      "learning_rate": 0.0005,
      "loss": 2.6949961185455322,
      "step": 2904
    },
    {
      "epoch": 0.6719870460328475,
      "grad_norm": 0.9829978346824646,
      "learning_rate": 0.0005,
      "loss": 2.6249656677246094,
      "step": 2905
    },
    {
      "epoch": 0.6722183668748555,
      "grad_norm": 1.1223948001861572,
      "learning_rate": 0.0005,
      "loss": 3.0926127433776855,
      "step": 2906
    },
    {
      "epoch": 0.6724496877168633,
      "grad_norm": 1.2495644092559814,
      "learning_rate": 0.0005,
      "loss": 2.5809431076049805,
      "step": 2907
    },
    {
      "epoch": 0.6726810085588711,
      "grad_norm": 1.12454092502594,
      "learning_rate": 0.0005,
      "loss": 2.734424591064453,
      "step": 2908
    },
    {
      "epoch": 0.672912329400879,
      "grad_norm": 0.8825401067733765,
      "learning_rate": 0.0005,
      "loss": 2.8157639503479004,
      "step": 2909
    },
    {
      "epoch": 0.6731436502428869,
      "grad_norm": 0.9556136727333069,
      "learning_rate": 0.0005,
      "loss": 2.3963239192962646,
      "step": 2910
    },
    {
      "epoch": 0.6733749710848947,
      "grad_norm": 0.9121977090835571,
      "learning_rate": 0.0005,
      "loss": 2.7472574710845947,
      "step": 2911
    },
    {
      "epoch": 0.6736062919269026,
      "grad_norm": 0.9099606871604919,
      "learning_rate": 0.0005,
      "loss": 2.628742218017578,
      "step": 2912
    },
    {
      "epoch": 0.6738376127689105,
      "grad_norm": 1.0511893033981323,
      "learning_rate": 0.0005,
      "loss": 2.4972121715545654,
      "step": 2913
    },
    {
      "epoch": 0.6740689336109184,
      "grad_norm": 1.002907395362854,
      "learning_rate": 0.0005,
      "loss": 2.7130637168884277,
      "step": 2914
    },
    {
      "epoch": 0.6743002544529262,
      "grad_norm": 0.8453859090805054,
      "learning_rate": 0.0005,
      "loss": 2.656473159790039,
      "step": 2915
    },
    {
      "epoch": 0.6745315752949341,
      "grad_norm": 1.1519285440444946,
      "learning_rate": 0.0005,
      "loss": 2.5667848587036133,
      "step": 2916
    },
    {
      "epoch": 0.674762896136942,
      "grad_norm": 1.2578896284103394,
      "learning_rate": 0.0005,
      "loss": 3.0947418212890625,
      "step": 2917
    },
    {
      "epoch": 0.6749942169789498,
      "grad_norm": 0.9974439740180969,
      "learning_rate": 0.0005,
      "loss": 2.592197895050049,
      "step": 2918
    },
    {
      "epoch": 0.6752255378209576,
      "grad_norm": 1.2330403327941895,
      "learning_rate": 0.0005,
      "loss": 2.8278098106384277,
      "step": 2919
    },
    {
      "epoch": 0.6754568586629656,
      "grad_norm": 1.199242353439331,
      "learning_rate": 0.0005,
      "loss": 2.711799144744873,
      "step": 2920
    },
    {
      "epoch": 0.6756881795049734,
      "grad_norm": 1.1592819690704346,
      "learning_rate": 0.0005,
      "loss": 3.2314915657043457,
      "step": 2921
    },
    {
      "epoch": 0.6759195003469812,
      "grad_norm": 1.2453263998031616,
      "learning_rate": 0.0005,
      "loss": 2.7248213291168213,
      "step": 2922
    },
    {
      "epoch": 0.6761508211889892,
      "grad_norm": 1.053748607635498,
      "learning_rate": 0.0005,
      "loss": 2.791405200958252,
      "step": 2923
    },
    {
      "epoch": 0.676382142030997,
      "grad_norm": 1.2396340370178223,
      "learning_rate": 0.0005,
      "loss": 3.0666255950927734,
      "step": 2924
    },
    {
      "epoch": 0.6766134628730048,
      "grad_norm": 1.0115575790405273,
      "learning_rate": 0.0005,
      "loss": 2.3983349800109863,
      "step": 2925
    },
    {
      "epoch": 0.6768447837150128,
      "grad_norm": 0.869499921798706,
      "learning_rate": 0.0005,
      "loss": 2.489428758621216,
      "step": 2926
    },
    {
      "epoch": 0.6770761045570206,
      "grad_norm": 1.010142207145691,
      "learning_rate": 0.0005,
      "loss": 2.4746627807617188,
      "step": 2927
    },
    {
      "epoch": 0.6773074253990284,
      "grad_norm": 1.2546277046203613,
      "learning_rate": 0.0005,
      "loss": 2.734203338623047,
      "step": 2928
    },
    {
      "epoch": 0.6775387462410363,
      "grad_norm": 0.9750205874443054,
      "learning_rate": 0.0005,
      "loss": 2.8080687522888184,
      "step": 2929
    },
    {
      "epoch": 0.6777700670830442,
      "grad_norm": 1.063398838043213,
      "learning_rate": 0.0005,
      "loss": 2.3063559532165527,
      "step": 2930
    },
    {
      "epoch": 0.678001387925052,
      "grad_norm": 1.2705854177474976,
      "learning_rate": 0.0005,
      "loss": 2.608819007873535,
      "step": 2931
    },
    {
      "epoch": 0.6782327087670599,
      "grad_norm": 0.8184959292411804,
      "learning_rate": 0.0005,
      "loss": 2.8473567962646484,
      "step": 2932
    },
    {
      "epoch": 0.6784640296090678,
      "grad_norm": 1.0352144241333008,
      "learning_rate": 0.0005,
      "loss": 2.539304256439209,
      "step": 2933
    },
    {
      "epoch": 0.6786953504510757,
      "grad_norm": 1.0188190937042236,
      "learning_rate": 0.0005,
      "loss": 2.8502583503723145,
      "step": 2934
    },
    {
      "epoch": 0.6789266712930835,
      "grad_norm": 1.264828085899353,
      "learning_rate": 0.0005,
      "loss": 2.7533788681030273,
      "step": 2935
    },
    {
      "epoch": 0.6791579921350914,
      "grad_norm": 0.8886752724647522,
      "learning_rate": 0.0005,
      "loss": 2.74889874458313,
      "step": 2936
    },
    {
      "epoch": 0.6793893129770993,
      "grad_norm": 0.9768177270889282,
      "learning_rate": 0.0005,
      "loss": 2.417835235595703,
      "step": 2937
    },
    {
      "epoch": 0.6796206338191071,
      "grad_norm": 0.9842852354049683,
      "learning_rate": 0.0005,
      "loss": 2.698455810546875,
      "step": 2938
    },
    {
      "epoch": 0.6798519546611149,
      "grad_norm": 1.1037325859069824,
      "learning_rate": 0.0005,
      "loss": 2.683211088180542,
      "step": 2939
    },
    {
      "epoch": 0.6800832755031229,
      "grad_norm": 0.9760487675666809,
      "learning_rate": 0.0005,
      "loss": 2.5954701900482178,
      "step": 2940
    },
    {
      "epoch": 0.6803145963451307,
      "grad_norm": 0.9833692312240601,
      "learning_rate": 0.0005,
      "loss": 2.6103715896606445,
      "step": 2941
    },
    {
      "epoch": 0.6805459171871385,
      "grad_norm": 1.0478403568267822,
      "learning_rate": 0.0005,
      "loss": 2.8771111965179443,
      "step": 2942
    },
    {
      "epoch": 0.6807772380291465,
      "grad_norm": 1.0719157457351685,
      "learning_rate": 0.0005,
      "loss": 2.981964588165283,
      "step": 2943
    },
    {
      "epoch": 0.6810085588711543,
      "grad_norm": 1.1909176111221313,
      "learning_rate": 0.0005,
      "loss": 2.5458436012268066,
      "step": 2944
    },
    {
      "epoch": 0.6812398797131621,
      "grad_norm": 1.0016549825668335,
      "learning_rate": 0.0005,
      "loss": 2.7189154624938965,
      "step": 2945
    },
    {
      "epoch": 0.6814712005551701,
      "grad_norm": 0.883503794670105,
      "learning_rate": 0.0005,
      "loss": 2.4658432006835938,
      "step": 2946
    },
    {
      "epoch": 0.6817025213971779,
      "grad_norm": 0.9818503260612488,
      "learning_rate": 0.0005,
      "loss": 2.6818087100982666,
      "step": 2947
    },
    {
      "epoch": 0.6819338422391857,
      "grad_norm": 1.5152714252471924,
      "learning_rate": 0.0005,
      "loss": 2.598690986633301,
      "step": 2948
    },
    {
      "epoch": 0.6821651630811936,
      "grad_norm": 1.043861746788025,
      "learning_rate": 0.0005,
      "loss": 3.037583351135254,
      "step": 2949
    },
    {
      "epoch": 0.6823964839232015,
      "grad_norm": 0.9761950969696045,
      "learning_rate": 0.0005,
      "loss": 2.408827781677246,
      "step": 2950
    },
    {
      "epoch": 0.6826278047652093,
      "grad_norm": 1.0406057834625244,
      "learning_rate": 0.0005,
      "loss": 2.8123364448547363,
      "step": 2951
    },
    {
      "epoch": 0.6828591256072172,
      "grad_norm": 1.0749030113220215,
      "learning_rate": 0.0005,
      "loss": 2.8990139961242676,
      "step": 2952
    },
    {
      "epoch": 0.6830904464492251,
      "grad_norm": 1.0052489042282104,
      "learning_rate": 0.0005,
      "loss": 2.6169381141662598,
      "step": 2953
    },
    {
      "epoch": 0.683321767291233,
      "grad_norm": 0.8674433827400208,
      "learning_rate": 0.0005,
      "loss": 2.6331207752227783,
      "step": 2954
    },
    {
      "epoch": 0.6835530881332408,
      "grad_norm": 0.900907039642334,
      "learning_rate": 0.0005,
      "loss": 2.9569101333618164,
      "step": 2955
    },
    {
      "epoch": 0.6837844089752487,
      "grad_norm": 0.9379141926765442,
      "learning_rate": 0.0005,
      "loss": 2.66726016998291,
      "step": 2956
    },
    {
      "epoch": 0.6840157298172566,
      "grad_norm": 1.3270894289016724,
      "learning_rate": 0.0005,
      "loss": 2.836293935775757,
      "step": 2957
    },
    {
      "epoch": 0.6842470506592644,
      "grad_norm": 1.157449722290039,
      "learning_rate": 0.0005,
      "loss": 2.352156639099121,
      "step": 2958
    },
    {
      "epoch": 0.6844783715012722,
      "grad_norm": 0.9013752937316895,
      "learning_rate": 0.0005,
      "loss": 2.815078020095825,
      "step": 2959
    },
    {
      "epoch": 0.6847096923432802,
      "grad_norm": 1.1029560565948486,
      "learning_rate": 0.0005,
      "loss": 2.3045387268066406,
      "step": 2960
    },
    {
      "epoch": 0.684941013185288,
      "grad_norm": 0.7667648196220398,
      "learning_rate": 0.0005,
      "loss": 2.586033821105957,
      "step": 2961
    },
    {
      "epoch": 0.6851723340272958,
      "grad_norm": 0.9608504176139832,
      "learning_rate": 0.0005,
      "loss": 2.586122989654541,
      "step": 2962
    },
    {
      "epoch": 0.6854036548693038,
      "grad_norm": 1.027214765548706,
      "learning_rate": 0.0005,
      "loss": 2.585102081298828,
      "step": 2963
    },
    {
      "epoch": 0.6856349757113116,
      "grad_norm": 0.8155549168586731,
      "learning_rate": 0.0005,
      "loss": 2.419696807861328,
      "step": 2964
    },
    {
      "epoch": 0.6858662965533194,
      "grad_norm": 0.9846332669258118,
      "learning_rate": 0.0005,
      "loss": 2.8570470809936523,
      "step": 2965
    },
    {
      "epoch": 0.6860976173953273,
      "grad_norm": 1.1480095386505127,
      "learning_rate": 0.0005,
      "loss": 2.6119444370269775,
      "step": 2966
    },
    {
      "epoch": 0.6863289382373352,
      "grad_norm": 1.3256781101226807,
      "learning_rate": 0.0005,
      "loss": 2.4233789443969727,
      "step": 2967
    },
    {
      "epoch": 0.686560259079343,
      "grad_norm": 1.1157124042510986,
      "learning_rate": 0.0005,
      "loss": 2.541853427886963,
      "step": 2968
    },
    {
      "epoch": 0.6867915799213509,
      "grad_norm": 0.954131007194519,
      "learning_rate": 0.0005,
      "loss": 2.103756904602051,
      "step": 2969
    },
    {
      "epoch": 0.6870229007633588,
      "grad_norm": 1.3580842018127441,
      "learning_rate": 0.0005,
      "loss": 2.4461746215820312,
      "step": 2970
    },
    {
      "epoch": 0.6872542216053666,
      "grad_norm": 1.6533459424972534,
      "learning_rate": 0.0005,
      "loss": 2.631326913833618,
      "step": 2971
    },
    {
      "epoch": 0.6874855424473745,
      "grad_norm": 1.0575635433197021,
      "learning_rate": 0.0005,
      "loss": 2.752051830291748,
      "step": 2972
    },
    {
      "epoch": 0.6877168632893824,
      "grad_norm": 0.9599119424819946,
      "learning_rate": 0.0005,
      "loss": 2.802866220474243,
      "step": 2973
    },
    {
      "epoch": 0.6879481841313903,
      "grad_norm": 1.3662080764770508,
      "learning_rate": 0.0005,
      "loss": 2.846120834350586,
      "step": 2974
    },
    {
      "epoch": 0.6881795049733981,
      "grad_norm": 0.8737412691116333,
      "learning_rate": 0.0005,
      "loss": 2.7083005905151367,
      "step": 2975
    },
    {
      "epoch": 0.6884108258154059,
      "grad_norm": 0.8876683115959167,
      "learning_rate": 0.0005,
      "loss": 2.5685582160949707,
      "step": 2976
    },
    {
      "epoch": 0.6886421466574139,
      "grad_norm": 0.9528459310531616,
      "learning_rate": 0.0005,
      "loss": 2.8540525436401367,
      "step": 2977
    },
    {
      "epoch": 0.6888734674994217,
      "grad_norm": 1.4320002794265747,
      "learning_rate": 0.0005,
      "loss": 2.6102852821350098,
      "step": 2978
    },
    {
      "epoch": 0.6891047883414295,
      "grad_norm": 1.2566128969192505,
      "learning_rate": 0.0005,
      "loss": 2.429354190826416,
      "step": 2979
    },
    {
      "epoch": 0.6893361091834375,
      "grad_norm": 1.8877456188201904,
      "learning_rate": 0.0005,
      "loss": 2.8275692462921143,
      "step": 2980
    },
    {
      "epoch": 0.6895674300254453,
      "grad_norm": 1.0099849700927734,
      "learning_rate": 0.0005,
      "loss": 2.769336223602295,
      "step": 2981
    },
    {
      "epoch": 0.6897987508674531,
      "grad_norm": 1.2073564529418945,
      "learning_rate": 0.0005,
      "loss": 2.310887336730957,
      "step": 2982
    },
    {
      "epoch": 0.6900300717094611,
      "grad_norm": 0.9892210364341736,
      "learning_rate": 0.0005,
      "loss": 2.6110477447509766,
      "step": 2983
    },
    {
      "epoch": 0.6902613925514689,
      "grad_norm": 1.2569458484649658,
      "learning_rate": 0.0005,
      "loss": 2.6420300006866455,
      "step": 2984
    },
    {
      "epoch": 0.6904927133934767,
      "grad_norm": 1.1778749227523804,
      "learning_rate": 0.0005,
      "loss": 2.5060482025146484,
      "step": 2985
    },
    {
      "epoch": 0.6907240342354846,
      "grad_norm": 0.8660236597061157,
      "learning_rate": 0.0005,
      "loss": 2.6599280834198,
      "step": 2986
    },
    {
      "epoch": 0.6909553550774925,
      "grad_norm": 0.8180950880050659,
      "learning_rate": 0.0005,
      "loss": 2.663156032562256,
      "step": 2987
    },
    {
      "epoch": 0.6911866759195003,
      "grad_norm": 0.9437532424926758,
      "learning_rate": 0.0005,
      "loss": 2.765263319015503,
      "step": 2988
    },
    {
      "epoch": 0.6914179967615082,
      "grad_norm": 1.0332728624343872,
      "learning_rate": 0.0005,
      "loss": 2.8768768310546875,
      "step": 2989
    },
    {
      "epoch": 0.6916493176035161,
      "grad_norm": 0.9115749597549438,
      "learning_rate": 0.0005,
      "loss": 2.8645291328430176,
      "step": 2990
    },
    {
      "epoch": 0.6918806384455239,
      "grad_norm": 0.8801731467247009,
      "learning_rate": 0.0005,
      "loss": 2.8804574012756348,
      "step": 2991
    },
    {
      "epoch": 0.6921119592875318,
      "grad_norm": 1.1667524576187134,
      "learning_rate": 0.0005,
      "loss": 2.5930933952331543,
      "step": 2992
    },
    {
      "epoch": 0.6923432801295397,
      "grad_norm": 0.9992888569831848,
      "learning_rate": 0.0005,
      "loss": 2.5723044872283936,
      "step": 2993
    },
    {
      "epoch": 0.6925746009715475,
      "grad_norm": 1.1715298891067505,
      "learning_rate": 0.0005,
      "loss": 2.8642024993896484,
      "step": 2994
    },
    {
      "epoch": 0.6928059218135554,
      "grad_norm": 0.9242631196975708,
      "learning_rate": 0.0005,
      "loss": 2.1091995239257812,
      "step": 2995
    },
    {
      "epoch": 0.6930372426555632,
      "grad_norm": 1.0008606910705566,
      "learning_rate": 0.0005,
      "loss": 2.6097850799560547,
      "step": 2996
    },
    {
      "epoch": 0.6932685634975712,
      "grad_norm": 1.6368290185928345,
      "learning_rate": 0.0005,
      "loss": 2.520597457885742,
      "step": 2997
    },
    {
      "epoch": 0.693499884339579,
      "grad_norm": 1.2467378377914429,
      "learning_rate": 0.0005,
      "loss": 2.7845897674560547,
      "step": 2998
    },
    {
      "epoch": 0.6937312051815868,
      "grad_norm": 1.2053604125976562,
      "learning_rate": 0.0005,
      "loss": 2.658498764038086,
      "step": 2999
    },
    {
      "epoch": 0.6939625260235948,
      "grad_norm": 1.0531961917877197,
      "learning_rate": 0.0005,
      "loss": 2.2620503902435303,
      "step": 3000
    },
    {
      "epoch": 0.6941938468656026,
      "grad_norm": 1.0871690511703491,
      "learning_rate": 0.0005,
      "loss": 2.618229389190674,
      "step": 3001
    },
    {
      "epoch": 0.6944251677076104,
      "grad_norm": 0.8289440274238586,
      "learning_rate": 0.0005,
      "loss": 2.384777784347534,
      "step": 3002
    },
    {
      "epoch": 0.6946564885496184,
      "grad_norm": 1.0430595874786377,
      "learning_rate": 0.0005,
      "loss": 2.87559175491333,
      "step": 3003
    },
    {
      "epoch": 0.6948878093916262,
      "grad_norm": 1.055497646331787,
      "learning_rate": 0.0005,
      "loss": 2.6080989837646484,
      "step": 3004
    },
    {
      "epoch": 0.695119130233634,
      "grad_norm": 1.1283656358718872,
      "learning_rate": 0.0005,
      "loss": 2.8147754669189453,
      "step": 3005
    },
    {
      "epoch": 0.6953504510756419,
      "grad_norm": 1.0762296915054321,
      "learning_rate": 0.0005,
      "loss": 2.453916072845459,
      "step": 3006
    },
    {
      "epoch": 0.6955817719176498,
      "grad_norm": 2.49458646774292,
      "learning_rate": 0.0005,
      "loss": 2.5906410217285156,
      "step": 3007
    },
    {
      "epoch": 0.6958130927596576,
      "grad_norm": 1.410722255706787,
      "learning_rate": 0.0005,
      "loss": 3.401291608810425,
      "step": 3008
    },
    {
      "epoch": 0.6960444136016655,
      "grad_norm": 0.9139392971992493,
      "learning_rate": 0.0005,
      "loss": 2.5835623741149902,
      "step": 3009
    },
    {
      "epoch": 0.6962757344436734,
      "grad_norm": 0.9773948192596436,
      "learning_rate": 0.0005,
      "loss": 2.846747636795044,
      "step": 3010
    },
    {
      "epoch": 0.6965070552856812,
      "grad_norm": 0.8899813294410706,
      "learning_rate": 0.0005,
      "loss": 2.4360151290893555,
      "step": 3011
    },
    {
      "epoch": 0.6967383761276891,
      "grad_norm": 1.1087387800216675,
      "learning_rate": 0.0005,
      "loss": 2.6499602794647217,
      "step": 3012
    },
    {
      "epoch": 0.696969696969697,
      "grad_norm": 0.7989845871925354,
      "learning_rate": 0.0005,
      "loss": 2.52912974357605,
      "step": 3013
    },
    {
      "epoch": 0.6972010178117048,
      "grad_norm": 1.3980717658996582,
      "learning_rate": 0.0005,
      "loss": 2.907236337661743,
      "step": 3014
    },
    {
      "epoch": 0.6974323386537127,
      "grad_norm": 1.0128848552703857,
      "learning_rate": 0.0005,
      "loss": 2.872856855392456,
      "step": 3015
    },
    {
      "epoch": 0.6976636594957205,
      "grad_norm": 1.2422338724136353,
      "learning_rate": 0.0005,
      "loss": 2.981595039367676,
      "step": 3016
    },
    {
      "epoch": 0.6978949803377285,
      "grad_norm": 0.9667390584945679,
      "learning_rate": 0.0005,
      "loss": 2.614107608795166,
      "step": 3017
    },
    {
      "epoch": 0.6981263011797363,
      "grad_norm": 0.9538751244544983,
      "learning_rate": 0.0005,
      "loss": 2.867175579071045,
      "step": 3018
    },
    {
      "epoch": 0.6983576220217441,
      "grad_norm": 0.9963026642799377,
      "learning_rate": 0.0005,
      "loss": 2.608996868133545,
      "step": 3019
    },
    {
      "epoch": 0.6985889428637521,
      "grad_norm": 2.081324577331543,
      "learning_rate": 0.0005,
      "loss": 2.494795799255371,
      "step": 3020
    },
    {
      "epoch": 0.6988202637057599,
      "grad_norm": 1.0255306959152222,
      "learning_rate": 0.0005,
      "loss": 2.5252676010131836,
      "step": 3021
    },
    {
      "epoch": 0.6990515845477677,
      "grad_norm": 1.0381819009780884,
      "learning_rate": 0.0005,
      "loss": 2.6414618492126465,
      "step": 3022
    },
    {
      "epoch": 0.6992829053897757,
      "grad_norm": 0.8539262413978577,
      "learning_rate": 0.0005,
      "loss": 2.635646343231201,
      "step": 3023
    },
    {
      "epoch": 0.6995142262317835,
      "grad_norm": 0.880923330783844,
      "learning_rate": 0.0005,
      "loss": 2.699342727661133,
      "step": 3024
    },
    {
      "epoch": 0.6997455470737913,
      "grad_norm": 1.196013331413269,
      "learning_rate": 0.0005,
      "loss": 2.6476428508758545,
      "step": 3025
    },
    {
      "epoch": 0.6999768679157992,
      "grad_norm": 1.2291066646575928,
      "learning_rate": 0.0005,
      "loss": 2.472972869873047,
      "step": 3026
    },
    {
      "epoch": 0.7002081887578071,
      "grad_norm": 1.2792342901229858,
      "learning_rate": 0.0005,
      "loss": 2.5128159523010254,
      "step": 3027
    },
    {
      "epoch": 0.7004395095998149,
      "grad_norm": 1.0325230360031128,
      "learning_rate": 0.0005,
      "loss": 2.4443235397338867,
      "step": 3028
    },
    {
      "epoch": 0.7006708304418228,
      "grad_norm": 0.9238547086715698,
      "learning_rate": 0.0005,
      "loss": 2.3166391849517822,
      "step": 3029
    },
    {
      "epoch": 0.7009021512838307,
      "grad_norm": 0.9566057920455933,
      "learning_rate": 0.0005,
      "loss": 2.8143200874328613,
      "step": 3030
    },
    {
      "epoch": 0.7011334721258385,
      "grad_norm": 1.0455822944641113,
      "learning_rate": 0.0005,
      "loss": 2.8001227378845215,
      "step": 3031
    },
    {
      "epoch": 0.7013647929678464,
      "grad_norm": 0.8838186264038086,
      "learning_rate": 0.0005,
      "loss": 2.385643243789673,
      "step": 3032
    },
    {
      "epoch": 0.7015961138098543,
      "grad_norm": 0.8946431279182434,
      "learning_rate": 0.0005,
      "loss": 2.279849052429199,
      "step": 3033
    },
    {
      "epoch": 0.7018274346518621,
      "grad_norm": 0.901836097240448,
      "learning_rate": 0.0005,
      "loss": 2.8711156845092773,
      "step": 3034
    },
    {
      "epoch": 0.70205875549387,
      "grad_norm": 1.2575124502182007,
      "learning_rate": 0.0005,
      "loss": 2.6965839862823486,
      "step": 3035
    },
    {
      "epoch": 0.7022900763358778,
      "grad_norm": 0.9458479285240173,
      "learning_rate": 0.0005,
      "loss": 2.698491096496582,
      "step": 3036
    },
    {
      "epoch": 0.7025213971778858,
      "grad_norm": 0.9739621877670288,
      "learning_rate": 0.0005,
      "loss": 2.7795310020446777,
      "step": 3037
    },
    {
      "epoch": 0.7027527180198936,
      "grad_norm": 1.0077656507492065,
      "learning_rate": 0.0005,
      "loss": 2.4631619453430176,
      "step": 3038
    },
    {
      "epoch": 0.7029840388619014,
      "grad_norm": 1.2210323810577393,
      "learning_rate": 0.0005,
      "loss": 2.551043748855591,
      "step": 3039
    },
    {
      "epoch": 0.7032153597039094,
      "grad_norm": 0.9813224077224731,
      "learning_rate": 0.0005,
      "loss": 2.506989002227783,
      "step": 3040
    },
    {
      "epoch": 0.7034466805459172,
      "grad_norm": 1.0497167110443115,
      "learning_rate": 0.0005,
      "loss": 2.8117661476135254,
      "step": 3041
    },
    {
      "epoch": 0.703678001387925,
      "grad_norm": 0.9966554045677185,
      "learning_rate": 0.0005,
      "loss": 2.7995364665985107,
      "step": 3042
    },
    {
      "epoch": 0.703909322229933,
      "grad_norm": 0.8454992175102234,
      "learning_rate": 0.0005,
      "loss": 2.6958067417144775,
      "step": 3043
    },
    {
      "epoch": 0.7041406430719408,
      "grad_norm": 1.026646614074707,
      "learning_rate": 0.0005,
      "loss": 2.734429359436035,
      "step": 3044
    },
    {
      "epoch": 0.7043719639139486,
      "grad_norm": 0.8086413741111755,
      "learning_rate": 0.0005,
      "loss": 2.943697929382324,
      "step": 3045
    },
    {
      "epoch": 0.7046032847559565,
      "grad_norm": 0.9685608744621277,
      "learning_rate": 0.0005,
      "loss": 2.9203641414642334,
      "step": 3046
    },
    {
      "epoch": 0.7048346055979644,
      "grad_norm": 1.067074179649353,
      "learning_rate": 0.0005,
      "loss": 2.319307804107666,
      "step": 3047
    },
    {
      "epoch": 0.7050659264399722,
      "grad_norm": 1.2840242385864258,
      "learning_rate": 0.0005,
      "loss": 2.809937000274658,
      "step": 3048
    },
    {
      "epoch": 0.7052972472819801,
      "grad_norm": 1.2712230682373047,
      "learning_rate": 0.0005,
      "loss": 2.3949508666992188,
      "step": 3049
    },
    {
      "epoch": 0.705528568123988,
      "grad_norm": 0.9890131950378418,
      "learning_rate": 0.0005,
      "loss": 2.614880084991455,
      "step": 3050
    },
    {
      "epoch": 0.7057598889659958,
      "grad_norm": 0.9299209117889404,
      "learning_rate": 0.0005,
      "loss": 3.105769157409668,
      "step": 3051
    },
    {
      "epoch": 0.7059912098080037,
      "grad_norm": 1.3508111238479614,
      "learning_rate": 0.0005,
      "loss": 2.7958390712738037,
      "step": 3052
    },
    {
      "epoch": 0.7062225306500116,
      "grad_norm": 1.0287208557128906,
      "learning_rate": 0.0005,
      "loss": 2.309448480606079,
      "step": 3053
    },
    {
      "epoch": 0.7064538514920194,
      "grad_norm": 1.1320278644561768,
      "learning_rate": 0.0005,
      "loss": 2.8880162239074707,
      "step": 3054
    },
    {
      "epoch": 0.7066851723340273,
      "grad_norm": 0.8398723006248474,
      "learning_rate": 0.0005,
      "loss": 2.4117255210876465,
      "step": 3055
    },
    {
      "epoch": 0.7069164931760351,
      "grad_norm": 2.7176525592803955,
      "learning_rate": 0.0005,
      "loss": 2.9393386840820312,
      "step": 3056
    },
    {
      "epoch": 0.707147814018043,
      "grad_norm": 1.1724191904067993,
      "learning_rate": 0.0005,
      "loss": 2.5679218769073486,
      "step": 3057
    },
    {
      "epoch": 0.7073791348600509,
      "grad_norm": 1.3005256652832031,
      "learning_rate": 0.0005,
      "loss": 2.6148288249969482,
      "step": 3058
    },
    {
      "epoch": 0.7076104557020587,
      "grad_norm": 0.9541012048721313,
      "learning_rate": 0.0005,
      "loss": 2.348041534423828,
      "step": 3059
    },
    {
      "epoch": 0.7078417765440667,
      "grad_norm": 1.0541530847549438,
      "learning_rate": 0.0005,
      "loss": 2.577157974243164,
      "step": 3060
    },
    {
      "epoch": 0.7080730973860745,
      "grad_norm": 1.3119314908981323,
      "learning_rate": 0.0005,
      "loss": 2.6310744285583496,
      "step": 3061
    },
    {
      "epoch": 0.7083044182280823,
      "grad_norm": 1.0142720937728882,
      "learning_rate": 0.0005,
      "loss": 3.07491397857666,
      "step": 3062
    },
    {
      "epoch": 0.7085357390700903,
      "grad_norm": 1.0413224697113037,
      "learning_rate": 0.0005,
      "loss": 2.791804552078247,
      "step": 3063
    },
    {
      "epoch": 0.7087670599120981,
      "grad_norm": 0.854444146156311,
      "learning_rate": 0.0005,
      "loss": 2.5015456676483154,
      "step": 3064
    },
    {
      "epoch": 0.7089983807541059,
      "grad_norm": 1.0009140968322754,
      "learning_rate": 0.0005,
      "loss": 2.6686229705810547,
      "step": 3065
    },
    {
      "epoch": 0.7092297015961138,
      "grad_norm": 0.9561149477958679,
      "learning_rate": 0.0005,
      "loss": 2.5549614429473877,
      "step": 3066
    },
    {
      "epoch": 0.7094610224381217,
      "grad_norm": 1.2876389026641846,
      "learning_rate": 0.0005,
      "loss": 3.280989170074463,
      "step": 3067
    },
    {
      "epoch": 0.7096923432801295,
      "grad_norm": 0.9087967872619629,
      "learning_rate": 0.0005,
      "loss": 2.4769768714904785,
      "step": 3068
    },
    {
      "epoch": 0.7099236641221374,
      "grad_norm": 1.0094730854034424,
      "learning_rate": 0.0005,
      "loss": 2.538090229034424,
      "step": 3069
    },
    {
      "epoch": 0.7101549849641453,
      "grad_norm": 1.2698290348052979,
      "learning_rate": 0.0005,
      "loss": 2.6572022438049316,
      "step": 3070
    },
    {
      "epoch": 0.7103863058061531,
      "grad_norm": 1.1320451498031616,
      "learning_rate": 0.0005,
      "loss": 2.500622272491455,
      "step": 3071
    },
    {
      "epoch": 0.710617626648161,
      "grad_norm": 0.9823915362358093,
      "learning_rate": 0.0005,
      "loss": 2.537024974822998,
      "step": 3072
    },
    {
      "epoch": 0.7108489474901689,
      "grad_norm": 0.9608726501464844,
      "learning_rate": 0.0005,
      "loss": 2.8586485385894775,
      "step": 3073
    },
    {
      "epoch": 0.7110802683321767,
      "grad_norm": 0.9779985547065735,
      "learning_rate": 0.0005,
      "loss": 2.8036856651306152,
      "step": 3074
    },
    {
      "epoch": 0.7113115891741846,
      "grad_norm": 0.8659719824790955,
      "learning_rate": 0.0005,
      "loss": 2.5031352043151855,
      "step": 3075
    },
    {
      "epoch": 0.7115429100161924,
      "grad_norm": 1.05389404296875,
      "learning_rate": 0.0005,
      "loss": 3.105225086212158,
      "step": 3076
    },
    {
      "epoch": 0.7117742308582004,
      "grad_norm": 1.0740808248519897,
      "learning_rate": 0.0005,
      "loss": 2.84952974319458,
      "step": 3077
    },
    {
      "epoch": 0.7120055517002082,
      "grad_norm": 1.4071959257125854,
      "learning_rate": 0.0005,
      "loss": 2.783111572265625,
      "step": 3078
    },
    {
      "epoch": 0.712236872542216,
      "grad_norm": 1.3081272840499878,
      "learning_rate": 0.0005,
      "loss": 2.9380688667297363,
      "step": 3079
    },
    {
      "epoch": 0.712468193384224,
      "grad_norm": 1.03891122341156,
      "learning_rate": 0.0005,
      "loss": 2.522911310195923,
      "step": 3080
    },
    {
      "epoch": 0.7126995142262318,
      "grad_norm": 1.0010560750961304,
      "learning_rate": 0.0005,
      "loss": 2.633699893951416,
      "step": 3081
    },
    {
      "epoch": 0.7129308350682396,
      "grad_norm": 1.5520074367523193,
      "learning_rate": 0.0005,
      "loss": 3.0942225456237793,
      "step": 3082
    },
    {
      "epoch": 0.7131621559102476,
      "grad_norm": 1.229494571685791,
      "learning_rate": 0.0005,
      "loss": 2.623905658721924,
      "step": 3083
    },
    {
      "epoch": 0.7133934767522554,
      "grad_norm": 0.9619930982589722,
      "learning_rate": 0.0005,
      "loss": 2.8827428817749023,
      "step": 3084
    },
    {
      "epoch": 0.7136247975942632,
      "grad_norm": 1.2109826803207397,
      "learning_rate": 0.0005,
      "loss": 2.408156394958496,
      "step": 3085
    },
    {
      "epoch": 0.7138561184362711,
      "grad_norm": 1.181915521621704,
      "learning_rate": 0.0005,
      "loss": 2.5706751346588135,
      "step": 3086
    },
    {
      "epoch": 0.714087439278279,
      "grad_norm": 1.0093320608139038,
      "learning_rate": 0.0005,
      "loss": 2.6376442909240723,
      "step": 3087
    },
    {
      "epoch": 0.7143187601202868,
      "grad_norm": 1.0614924430847168,
      "learning_rate": 0.0005,
      "loss": 2.605867385864258,
      "step": 3088
    },
    {
      "epoch": 0.7145500809622947,
      "grad_norm": 0.9992823600769043,
      "learning_rate": 0.0005,
      "loss": 2.888197898864746,
      "step": 3089
    },
    {
      "epoch": 0.7147814018043026,
      "grad_norm": 0.9373502731323242,
      "learning_rate": 0.0005,
      "loss": 2.9109139442443848,
      "step": 3090
    },
    {
      "epoch": 0.7150127226463104,
      "grad_norm": 1.0949642658233643,
      "learning_rate": 0.0005,
      "loss": 2.723991870880127,
      "step": 3091
    },
    {
      "epoch": 0.7152440434883183,
      "grad_norm": 1.19275963306427,
      "learning_rate": 0.0005,
      "loss": 2.741664409637451,
      "step": 3092
    },
    {
      "epoch": 0.7154753643303262,
      "grad_norm": 1.5496069192886353,
      "learning_rate": 0.0005,
      "loss": 2.5198116302490234,
      "step": 3093
    },
    {
      "epoch": 0.715706685172334,
      "grad_norm": 1.070448637008667,
      "learning_rate": 0.0005,
      "loss": 2.5501441955566406,
      "step": 3094
    },
    {
      "epoch": 0.7159380060143419,
      "grad_norm": 1.0353167057037354,
      "learning_rate": 0.0005,
      "loss": 2.5852739810943604,
      "step": 3095
    },
    {
      "epoch": 0.7161693268563497,
      "grad_norm": 1.191392183303833,
      "learning_rate": 0.0005,
      "loss": 2.988300085067749,
      "step": 3096
    },
    {
      "epoch": 0.7164006476983577,
      "grad_norm": 1.227502703666687,
      "learning_rate": 0.0005,
      "loss": 3.0705533027648926,
      "step": 3097
    },
    {
      "epoch": 0.7166319685403655,
      "grad_norm": 0.8089078068733215,
      "learning_rate": 0.0005,
      "loss": 2.3427183628082275,
      "step": 3098
    },
    {
      "epoch": 0.7168632893823733,
      "grad_norm": 1.497768759727478,
      "learning_rate": 0.0005,
      "loss": 3.089040756225586,
      "step": 3099
    },
    {
      "epoch": 0.7170946102243813,
      "grad_norm": 1.476998209953308,
      "learning_rate": 0.0005,
      "loss": 2.7182013988494873,
      "step": 3100
    },
    {
      "epoch": 0.7173259310663891,
      "grad_norm": 0.9445443153381348,
      "learning_rate": 0.0005,
      "loss": 2.350797176361084,
      "step": 3101
    },
    {
      "epoch": 0.7175572519083969,
      "grad_norm": 1.742164969444275,
      "learning_rate": 0.0005,
      "loss": 2.682727098464966,
      "step": 3102
    },
    {
      "epoch": 0.7177885727504049,
      "grad_norm": 1.0986242294311523,
      "learning_rate": 0.0005,
      "loss": 2.70833158493042,
      "step": 3103
    },
    {
      "epoch": 0.7180198935924127,
      "grad_norm": 1.1269809007644653,
      "learning_rate": 0.0005,
      "loss": 2.6521244049072266,
      "step": 3104
    },
    {
      "epoch": 0.7182512144344205,
      "grad_norm": 1.0490834712982178,
      "learning_rate": 0.0005,
      "loss": 2.793931245803833,
      "step": 3105
    },
    {
      "epoch": 0.7184825352764284,
      "grad_norm": 0.82352215051651,
      "learning_rate": 0.0005,
      "loss": 2.5923523902893066,
      "step": 3106
    },
    {
      "epoch": 0.7187138561184363,
      "grad_norm": 0.9043055176734924,
      "learning_rate": 0.0005,
      "loss": 2.2915501594543457,
      "step": 3107
    },
    {
      "epoch": 0.7189451769604441,
      "grad_norm": 0.9924265146255493,
      "learning_rate": 0.0005,
      "loss": 2.658447265625,
      "step": 3108
    },
    {
      "epoch": 0.719176497802452,
      "grad_norm": 1.0060487985610962,
      "learning_rate": 0.0005,
      "loss": 2.5581021308898926,
      "step": 3109
    },
    {
      "epoch": 0.7194078186444599,
      "grad_norm": 0.8314289450645447,
      "learning_rate": 0.0005,
      "loss": 2.43565034866333,
      "step": 3110
    },
    {
      "epoch": 0.7196391394864677,
      "grad_norm": 0.9115930199623108,
      "learning_rate": 0.0005,
      "loss": 2.642885446548462,
      "step": 3111
    },
    {
      "epoch": 0.7198704603284756,
      "grad_norm": 0.9081886410713196,
      "learning_rate": 0.0005,
      "loss": 2.769235610961914,
      "step": 3112
    },
    {
      "epoch": 0.7201017811704835,
      "grad_norm": 1.2194478511810303,
      "learning_rate": 0.0005,
      "loss": 2.7822680473327637,
      "step": 3113
    },
    {
      "epoch": 0.7203331020124913,
      "grad_norm": 0.9789292812347412,
      "learning_rate": 0.0005,
      "loss": 2.576390266418457,
      "step": 3114
    },
    {
      "epoch": 0.7205644228544992,
      "grad_norm": 1.168184757232666,
      "learning_rate": 0.0005,
      "loss": 2.7468559741973877,
      "step": 3115
    },
    {
      "epoch": 0.720795743696507,
      "grad_norm": 0.9504051208496094,
      "learning_rate": 0.0005,
      "loss": 2.434631109237671,
      "step": 3116
    },
    {
      "epoch": 0.721027064538515,
      "grad_norm": 1.0172737836837769,
      "learning_rate": 0.0005,
      "loss": 2.840898036956787,
      "step": 3117
    },
    {
      "epoch": 0.7212583853805228,
      "grad_norm": 1.07918381690979,
      "learning_rate": 0.0005,
      "loss": 2.4596190452575684,
      "step": 3118
    },
    {
      "epoch": 0.7214897062225306,
      "grad_norm": 1.644507884979248,
      "learning_rate": 0.0005,
      "loss": 2.976287364959717,
      "step": 3119
    },
    {
      "epoch": 0.7217210270645386,
      "grad_norm": 1.4962961673736572,
      "learning_rate": 0.0005,
      "loss": 2.6375715732574463,
      "step": 3120
    },
    {
      "epoch": 0.7219523479065464,
      "grad_norm": 0.8854676485061646,
      "learning_rate": 0.0005,
      "loss": 2.7380707263946533,
      "step": 3121
    },
    {
      "epoch": 0.7221836687485542,
      "grad_norm": 1.018936276435852,
      "learning_rate": 0.0005,
      "loss": 3.1180319786071777,
      "step": 3122
    },
    {
      "epoch": 0.7224149895905622,
      "grad_norm": 1.2472119331359863,
      "learning_rate": 0.0005,
      "loss": 2.4329895973205566,
      "step": 3123
    },
    {
      "epoch": 0.72264631043257,
      "grad_norm": 1.5515074729919434,
      "learning_rate": 0.0005,
      "loss": 2.764310836791992,
      "step": 3124
    },
    {
      "epoch": 0.7228776312745778,
      "grad_norm": 1.0918471813201904,
      "learning_rate": 0.0005,
      "loss": 2.462062120437622,
      "step": 3125
    },
    {
      "epoch": 0.7231089521165857,
      "grad_norm": 1.248299479484558,
      "learning_rate": 0.0005,
      "loss": 2.6576919555664062,
      "step": 3126
    },
    {
      "epoch": 0.7233402729585936,
      "grad_norm": 1.1184595823287964,
      "learning_rate": 0.0005,
      "loss": 2.9254848957061768,
      "step": 3127
    },
    {
      "epoch": 0.7235715938006014,
      "grad_norm": 1.0348035097122192,
      "learning_rate": 0.0005,
      "loss": 2.574259042739868,
      "step": 3128
    },
    {
      "epoch": 0.7238029146426093,
      "grad_norm": 1.368791937828064,
      "learning_rate": 0.0005,
      "loss": 2.562227249145508,
      "step": 3129
    },
    {
      "epoch": 0.7240342354846172,
      "grad_norm": 1.2763590812683105,
      "learning_rate": 0.0005,
      "loss": 2.4923617839813232,
      "step": 3130
    },
    {
      "epoch": 0.724265556326625,
      "grad_norm": 1.188351035118103,
      "learning_rate": 0.0005,
      "loss": 2.9069361686706543,
      "step": 3131
    },
    {
      "epoch": 0.7244968771686329,
      "grad_norm": 0.9745797514915466,
      "learning_rate": 0.0005,
      "loss": 2.4138522148132324,
      "step": 3132
    },
    {
      "epoch": 0.7247281980106408,
      "grad_norm": 1.0992870330810547,
      "learning_rate": 0.0005,
      "loss": 2.764587879180908,
      "step": 3133
    },
    {
      "epoch": 0.7249595188526486,
      "grad_norm": 0.9920317530632019,
      "learning_rate": 0.0005,
      "loss": 2.8728108406066895,
      "step": 3134
    },
    {
      "epoch": 0.7251908396946565,
      "grad_norm": 1.1219844818115234,
      "learning_rate": 0.0005,
      "loss": 2.4123191833496094,
      "step": 3135
    },
    {
      "epoch": 0.7254221605366643,
      "grad_norm": 1.1739152669906616,
      "learning_rate": 0.0005,
      "loss": 2.44513201713562,
      "step": 3136
    },
    {
      "epoch": 0.7256534813786722,
      "grad_norm": 1.108839750289917,
      "learning_rate": 0.0005,
      "loss": 2.6130645275115967,
      "step": 3137
    },
    {
      "epoch": 0.7258848022206801,
      "grad_norm": 1.0285491943359375,
      "learning_rate": 0.0005,
      "loss": 2.8543472290039062,
      "step": 3138
    },
    {
      "epoch": 0.7261161230626879,
      "grad_norm": 1.0005003213882446,
      "learning_rate": 0.0005,
      "loss": 2.6961488723754883,
      "step": 3139
    },
    {
      "epoch": 0.7263474439046959,
      "grad_norm": 0.8955907225608826,
      "learning_rate": 0.0005,
      "loss": 2.25471830368042,
      "step": 3140
    },
    {
      "epoch": 0.7265787647467037,
      "grad_norm": 0.9845394492149353,
      "learning_rate": 0.0005,
      "loss": 2.677025318145752,
      "step": 3141
    },
    {
      "epoch": 0.7268100855887115,
      "grad_norm": 0.9714606404304504,
      "learning_rate": 0.0005,
      "loss": 2.5005977153778076,
      "step": 3142
    },
    {
      "epoch": 0.7270414064307195,
      "grad_norm": 1.0023584365844727,
      "learning_rate": 0.0005,
      "loss": 2.522904872894287,
      "step": 3143
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.1189777851104736,
      "learning_rate": 0.0005,
      "loss": 2.834743022918701,
      "step": 3144
    },
    {
      "epoch": 0.7275040481147351,
      "grad_norm": 1.054254174232483,
      "learning_rate": 0.0005,
      "loss": 2.9507803916931152,
      "step": 3145
    },
    {
      "epoch": 0.727735368956743,
      "grad_norm": 0.9681074023246765,
      "learning_rate": 0.0005,
      "loss": 2.887035608291626,
      "step": 3146
    },
    {
      "epoch": 0.7279666897987509,
      "grad_norm": 1.0714337825775146,
      "learning_rate": 0.0005,
      "loss": 2.761556386947632,
      "step": 3147
    },
    {
      "epoch": 0.7281980106407587,
      "grad_norm": 1.0800261497497559,
      "learning_rate": 0.0005,
      "loss": 2.676625967025757,
      "step": 3148
    },
    {
      "epoch": 0.7284293314827666,
      "grad_norm": 1.0455951690673828,
      "learning_rate": 0.0005,
      "loss": 2.6887264251708984,
      "step": 3149
    },
    {
      "epoch": 0.7286606523247745,
      "grad_norm": 1.1710819005966187,
      "learning_rate": 0.0005,
      "loss": 2.575373649597168,
      "step": 3150
    },
    {
      "epoch": 0.7288919731667823,
      "grad_norm": 1.0663009881973267,
      "learning_rate": 0.0005,
      "loss": 2.2303547859191895,
      "step": 3151
    },
    {
      "epoch": 0.7291232940087902,
      "grad_norm": 1.274186611175537,
      "learning_rate": 0.0005,
      "loss": 2.890730857849121,
      "step": 3152
    },
    {
      "epoch": 0.7293546148507981,
      "grad_norm": 1.0475692749023438,
      "learning_rate": 0.0005,
      "loss": 2.4955618381500244,
      "step": 3153
    },
    {
      "epoch": 0.7295859356928059,
      "grad_norm": 1.3132516145706177,
      "learning_rate": 0.0005,
      "loss": 2.626856803894043,
      "step": 3154
    },
    {
      "epoch": 0.7298172565348138,
      "grad_norm": 3.343494415283203,
      "learning_rate": 0.0005,
      "loss": 2.7456517219543457,
      "step": 3155
    },
    {
      "epoch": 0.7300485773768216,
      "grad_norm": 1.3023834228515625,
      "learning_rate": 0.0005,
      "loss": 2.5423412322998047,
      "step": 3156
    },
    {
      "epoch": 0.7302798982188295,
      "grad_norm": 1.3050811290740967,
      "learning_rate": 0.0005,
      "loss": 2.933232069015503,
      "step": 3157
    },
    {
      "epoch": 0.7305112190608374,
      "grad_norm": 1.1620961427688599,
      "learning_rate": 0.0005,
      "loss": 2.395388603210449,
      "step": 3158
    },
    {
      "epoch": 0.7307425399028452,
      "grad_norm": 1.3042981624603271,
      "learning_rate": 0.0005,
      "loss": 2.346402168273926,
      "step": 3159
    },
    {
      "epoch": 0.7309738607448532,
      "grad_norm": 1.1825052499771118,
      "learning_rate": 0.0005,
      "loss": 2.9201571941375732,
      "step": 3160
    },
    {
      "epoch": 0.731205181586861,
      "grad_norm": 0.9113103747367859,
      "learning_rate": 0.0005,
      "loss": 2.56693172454834,
      "step": 3161
    },
    {
      "epoch": 0.7314365024288688,
      "grad_norm": 0.9462366104125977,
      "learning_rate": 0.0005,
      "loss": 2.5918474197387695,
      "step": 3162
    },
    {
      "epoch": 0.7316678232708768,
      "grad_norm": 1.1273541450500488,
      "learning_rate": 0.0005,
      "loss": 2.546541690826416,
      "step": 3163
    },
    {
      "epoch": 0.7318991441128846,
      "grad_norm": 1.0510565042495728,
      "learning_rate": 0.0005,
      "loss": 2.4990878105163574,
      "step": 3164
    },
    {
      "epoch": 0.7321304649548924,
      "grad_norm": 4.319604873657227,
      "learning_rate": 0.0005,
      "loss": 3.4355218410491943,
      "step": 3165
    },
    {
      "epoch": 0.7323617857969003,
      "grad_norm": 1.0024701356887817,
      "learning_rate": 0.0005,
      "loss": 2.6773324012756348,
      "step": 3166
    },
    {
      "epoch": 0.7325931066389082,
      "grad_norm": 1.3614743947982788,
      "learning_rate": 0.0005,
      "loss": 2.4916934967041016,
      "step": 3167
    },
    {
      "epoch": 0.732824427480916,
      "grad_norm": 0.9151039719581604,
      "learning_rate": 0.0005,
      "loss": 2.421354293823242,
      "step": 3168
    },
    {
      "epoch": 0.7330557483229239,
      "grad_norm": 0.8525654673576355,
      "learning_rate": 0.0005,
      "loss": 2.577869415283203,
      "step": 3169
    },
    {
      "epoch": 0.7332870691649318,
      "grad_norm": 0.8062976002693176,
      "learning_rate": 0.0005,
      "loss": 2.7904038429260254,
      "step": 3170
    },
    {
      "epoch": 0.7335183900069396,
      "grad_norm": 1.3504928350448608,
      "learning_rate": 0.0005,
      "loss": 2.659487724304199,
      "step": 3171
    },
    {
      "epoch": 0.7337497108489475,
      "grad_norm": 1.3228256702423096,
      "learning_rate": 0.0005,
      "loss": 2.5428307056427,
      "step": 3172
    },
    {
      "epoch": 0.7339810316909554,
      "grad_norm": 1.077857255935669,
      "learning_rate": 0.0005,
      "loss": 2.515490770339966,
      "step": 3173
    },
    {
      "epoch": 0.7342123525329632,
      "grad_norm": 1.1054948568344116,
      "learning_rate": 0.0005,
      "loss": 2.619053840637207,
      "step": 3174
    },
    {
      "epoch": 0.7344436733749711,
      "grad_norm": 1.45877206325531,
      "learning_rate": 0.0005,
      "loss": 2.93827486038208,
      "step": 3175
    },
    {
      "epoch": 0.7346749942169789,
      "grad_norm": 1.2474923133850098,
      "learning_rate": 0.0005,
      "loss": 2.775580883026123,
      "step": 3176
    },
    {
      "epoch": 0.7349063150589868,
      "grad_norm": 1.4234157800674438,
      "learning_rate": 0.0005,
      "loss": 3.163055181503296,
      "step": 3177
    },
    {
      "epoch": 0.7351376359009947,
      "grad_norm": 1.3034101724624634,
      "learning_rate": 0.0005,
      "loss": 2.862222671508789,
      "step": 3178
    },
    {
      "epoch": 0.7353689567430025,
      "grad_norm": 1.8292057514190674,
      "learning_rate": 0.0005,
      "loss": 2.5102925300598145,
      "step": 3179
    },
    {
      "epoch": 0.7356002775850105,
      "grad_norm": 1.414795160293579,
      "learning_rate": 0.0005,
      "loss": 2.480480432510376,
      "step": 3180
    },
    {
      "epoch": 0.7358315984270183,
      "grad_norm": 1.9882793426513672,
      "learning_rate": 0.0005,
      "loss": 2.698306083679199,
      "step": 3181
    },
    {
      "epoch": 0.7360629192690261,
      "grad_norm": 1.9420446157455444,
      "learning_rate": 0.0005,
      "loss": 2.6649844646453857,
      "step": 3182
    },
    {
      "epoch": 0.7362942401110341,
      "grad_norm": 1.4975290298461914,
      "learning_rate": 0.0005,
      "loss": 2.8452000617980957,
      "step": 3183
    },
    {
      "epoch": 0.7365255609530419,
      "grad_norm": 0.9742527604103088,
      "learning_rate": 0.0005,
      "loss": 2.8972949981689453,
      "step": 3184
    },
    {
      "epoch": 0.7367568817950497,
      "grad_norm": 1.3763843774795532,
      "learning_rate": 0.0005,
      "loss": 3.226776599884033,
      "step": 3185
    },
    {
      "epoch": 0.7369882026370576,
      "grad_norm": 0.9800048470497131,
      "learning_rate": 0.0005,
      "loss": 2.488297462463379,
      "step": 3186
    },
    {
      "epoch": 0.7372195234790655,
      "grad_norm": 0.9632954597473145,
      "learning_rate": 0.0005,
      "loss": 2.2346243858337402,
      "step": 3187
    },
    {
      "epoch": 0.7374508443210733,
      "grad_norm": 1.0704516172409058,
      "learning_rate": 0.0005,
      "loss": 2.470639705657959,
      "step": 3188
    },
    {
      "epoch": 0.7376821651630812,
      "grad_norm": 1.4650391340255737,
      "learning_rate": 0.0005,
      "loss": 2.675295352935791,
      "step": 3189
    },
    {
      "epoch": 0.7379134860050891,
      "grad_norm": 1.0627213716506958,
      "learning_rate": 0.0005,
      "loss": 2.519925594329834,
      "step": 3190
    },
    {
      "epoch": 0.7381448068470969,
      "grad_norm": 1.1007239818572998,
      "learning_rate": 0.0005,
      "loss": 2.574403762817383,
      "step": 3191
    },
    {
      "epoch": 0.7383761276891048,
      "grad_norm": 1.0764658451080322,
      "learning_rate": 0.0005,
      "loss": 2.9601588249206543,
      "step": 3192
    },
    {
      "epoch": 0.7386074485311126,
      "grad_norm": 0.8523722887039185,
      "learning_rate": 0.0005,
      "loss": 2.4801993370056152,
      "step": 3193
    },
    {
      "epoch": 0.7388387693731205,
      "grad_norm": 0.8951041102409363,
      "learning_rate": 0.0005,
      "loss": 2.7015862464904785,
      "step": 3194
    },
    {
      "epoch": 0.7390700902151284,
      "grad_norm": 1.0809884071350098,
      "learning_rate": 0.0005,
      "loss": 2.492753267288208,
      "step": 3195
    },
    {
      "epoch": 0.7393014110571362,
      "grad_norm": 0.7968887090682983,
      "learning_rate": 0.0005,
      "loss": 2.5524954795837402,
      "step": 3196
    },
    {
      "epoch": 0.7395327318991441,
      "grad_norm": 2.835488796234131,
      "learning_rate": 0.0005,
      "loss": 2.811250686645508,
      "step": 3197
    },
    {
      "epoch": 0.739764052741152,
      "grad_norm": 1.377292275428772,
      "learning_rate": 0.0005,
      "loss": 2.5860595703125,
      "step": 3198
    },
    {
      "epoch": 0.7399953735831598,
      "grad_norm": 1.0741069316864014,
      "learning_rate": 0.0005,
      "loss": 2.5878758430480957,
      "step": 3199
    },
    {
      "epoch": 0.7402266944251678,
      "grad_norm": 1.072657585144043,
      "learning_rate": 0.0005,
      "loss": 2.3702821731567383,
      "step": 3200
    },
    {
      "epoch": 0.7404580152671756,
      "grad_norm": 0.991864800453186,
      "learning_rate": 0.0005,
      "loss": 2.4871106147766113,
      "step": 3201
    },
    {
      "epoch": 0.7406893361091834,
      "grad_norm": 1.0594571828842163,
      "learning_rate": 0.0005,
      "loss": 2.536306142807007,
      "step": 3202
    },
    {
      "epoch": 0.7409206569511912,
      "grad_norm": 0.8501884341239929,
      "learning_rate": 0.0005,
      "loss": 2.8387062549591064,
      "step": 3203
    },
    {
      "epoch": 0.7411519777931992,
      "grad_norm": 1.1427898406982422,
      "learning_rate": 0.0005,
      "loss": 2.4293370246887207,
      "step": 3204
    },
    {
      "epoch": 0.741383298635207,
      "grad_norm": 1.0747101306915283,
      "learning_rate": 0.0005,
      "loss": 2.7607994079589844,
      "step": 3205
    },
    {
      "epoch": 0.7416146194772149,
      "grad_norm": 1.000744342803955,
      "learning_rate": 0.0005,
      "loss": 2.4103589057922363,
      "step": 3206
    },
    {
      "epoch": 0.7418459403192228,
      "grad_norm": 0.9239761829376221,
      "learning_rate": 0.0005,
      "loss": 2.487743854522705,
      "step": 3207
    },
    {
      "epoch": 0.7420772611612306,
      "grad_norm": 0.986859142780304,
      "learning_rate": 0.0005,
      "loss": 2.533998727798462,
      "step": 3208
    },
    {
      "epoch": 0.7423085820032385,
      "grad_norm": 0.9924956560134888,
      "learning_rate": 0.0005,
      "loss": 2.3599984645843506,
      "step": 3209
    },
    {
      "epoch": 0.7425399028452464,
      "grad_norm": 0.9398542046546936,
      "learning_rate": 0.0005,
      "loss": 2.4796359539031982,
      "step": 3210
    },
    {
      "epoch": 0.7427712236872542,
      "grad_norm": 1.3591110706329346,
      "learning_rate": 0.0005,
      "loss": 2.8675918579101562,
      "step": 3211
    },
    {
      "epoch": 0.7430025445292621,
      "grad_norm": 0.9240719676017761,
      "learning_rate": 0.0005,
      "loss": 2.7149012088775635,
      "step": 3212
    },
    {
      "epoch": 0.7432338653712699,
      "grad_norm": 1.0738123655319214,
      "learning_rate": 0.0005,
      "loss": 2.5649194717407227,
      "step": 3213
    },
    {
      "epoch": 0.7434651862132778,
      "grad_norm": 1.1950496435165405,
      "learning_rate": 0.0005,
      "loss": 3.09853196144104,
      "step": 3214
    },
    {
      "epoch": 0.7436965070552857,
      "grad_norm": 1.1256372928619385,
      "learning_rate": 0.0005,
      "loss": 2.5532021522521973,
      "step": 3215
    },
    {
      "epoch": 0.7439278278972935,
      "grad_norm": 1.021433711051941,
      "learning_rate": 0.0005,
      "loss": 2.5823631286621094,
      "step": 3216
    },
    {
      "epoch": 0.7441591487393014,
      "grad_norm": 1.4179120063781738,
      "learning_rate": 0.0005,
      "loss": 2.714665412902832,
      "step": 3217
    },
    {
      "epoch": 0.7443904695813093,
      "grad_norm": 0.9763339757919312,
      "learning_rate": 0.0005,
      "loss": 2.4490973949432373,
      "step": 3218
    },
    {
      "epoch": 0.7446217904233171,
      "grad_norm": 0.8150262236595154,
      "learning_rate": 0.0005,
      "loss": 2.6032166481018066,
      "step": 3219
    },
    {
      "epoch": 0.744853111265325,
      "grad_norm": 1.0126667022705078,
      "learning_rate": 0.0005,
      "loss": 2.7272701263427734,
      "step": 3220
    },
    {
      "epoch": 0.7450844321073329,
      "grad_norm": 0.8625943064689636,
      "learning_rate": 0.0005,
      "loss": 2.343313694000244,
      "step": 3221
    },
    {
      "epoch": 0.7453157529493407,
      "grad_norm": 0.8617326021194458,
      "learning_rate": 0.0005,
      "loss": 2.7909250259399414,
      "step": 3222
    },
    {
      "epoch": 0.7455470737913485,
      "grad_norm": 1.007757306098938,
      "learning_rate": 0.0005,
      "loss": 2.6827447414398193,
      "step": 3223
    },
    {
      "epoch": 0.7457783946333565,
      "grad_norm": 0.9726743102073669,
      "learning_rate": 0.0005,
      "loss": 2.568671703338623,
      "step": 3224
    },
    {
      "epoch": 0.7460097154753643,
      "grad_norm": 0.8588132858276367,
      "learning_rate": 0.0005,
      "loss": 2.5187485218048096,
      "step": 3225
    },
    {
      "epoch": 0.7462410363173722,
      "grad_norm": 0.882318913936615,
      "learning_rate": 0.0005,
      "loss": 2.6862542629241943,
      "step": 3226
    },
    {
      "epoch": 0.7464723571593801,
      "grad_norm": 1.1275405883789062,
      "learning_rate": 0.0005,
      "loss": 2.878538131713867,
      "step": 3227
    },
    {
      "epoch": 0.7467036780013879,
      "grad_norm": 1.0568383932113647,
      "learning_rate": 0.0005,
      "loss": 3.0918452739715576,
      "step": 3228
    },
    {
      "epoch": 0.7469349988433958,
      "grad_norm": 1.0470778942108154,
      "learning_rate": 0.0005,
      "loss": 2.747817277908325,
      "step": 3229
    },
    {
      "epoch": 0.7471663196854037,
      "grad_norm": 1.5071169137954712,
      "learning_rate": 0.0005,
      "loss": 2.5387492179870605,
      "step": 3230
    },
    {
      "epoch": 0.7473976405274115,
      "grad_norm": 1.3354430198669434,
      "learning_rate": 0.0005,
      "loss": 3.110819101333618,
      "step": 3231
    },
    {
      "epoch": 0.7476289613694194,
      "grad_norm": 1.4119387865066528,
      "learning_rate": 0.0005,
      "loss": 3.0142602920532227,
      "step": 3232
    },
    {
      "epoch": 0.7478602822114272,
      "grad_norm": 1.2528281211853027,
      "learning_rate": 0.0005,
      "loss": 2.341670036315918,
      "step": 3233
    },
    {
      "epoch": 0.7480916030534351,
      "grad_norm": 0.9762728810310364,
      "learning_rate": 0.0005,
      "loss": 2.645859479904175,
      "step": 3234
    },
    {
      "epoch": 0.748322923895443,
      "grad_norm": 0.7443507313728333,
      "learning_rate": 0.0005,
      "loss": 2.8181328773498535,
      "step": 3235
    },
    {
      "epoch": 0.7485542447374508,
      "grad_norm": 1.1642298698425293,
      "learning_rate": 0.0005,
      "loss": 2.722169876098633,
      "step": 3236
    },
    {
      "epoch": 0.7487855655794587,
      "grad_norm": 0.9191856384277344,
      "learning_rate": 0.0005,
      "loss": 2.6293294429779053,
      "step": 3237
    },
    {
      "epoch": 0.7490168864214666,
      "grad_norm": 2.3011715412139893,
      "learning_rate": 0.0005,
      "loss": 2.9820470809936523,
      "step": 3238
    },
    {
      "epoch": 0.7492482072634744,
      "grad_norm": 1.1816911697387695,
      "learning_rate": 0.0005,
      "loss": 2.5286192893981934,
      "step": 3239
    },
    {
      "epoch": 0.7494795281054824,
      "grad_norm": 1.2219982147216797,
      "learning_rate": 0.0005,
      "loss": 2.4569091796875,
      "step": 3240
    },
    {
      "epoch": 0.7497108489474902,
      "grad_norm": 1.57146418094635,
      "learning_rate": 0.0005,
      "loss": 2.4638915061950684,
      "step": 3241
    },
    {
      "epoch": 0.749942169789498,
      "grad_norm": 1.1979469060897827,
      "learning_rate": 0.0005,
      "loss": 2.9688191413879395,
      "step": 3242
    },
    {
      "epoch": 0.7501734906315058,
      "grad_norm": 1.0837316513061523,
      "learning_rate": 0.0005,
      "loss": 3.171281337738037,
      "step": 3243
    },
    {
      "epoch": 0.7504048114735138,
      "grad_norm": 1.0095422267913818,
      "learning_rate": 0.0005,
      "loss": 2.6837425231933594,
      "step": 3244
    },
    {
      "epoch": 0.7506361323155216,
      "grad_norm": 1.029741644859314,
      "learning_rate": 0.0005,
      "loss": 2.3007216453552246,
      "step": 3245
    },
    {
      "epoch": 0.7508674531575295,
      "grad_norm": 1.14915931224823,
      "learning_rate": 0.0005,
      "loss": 2.480618476867676,
      "step": 3246
    },
    {
      "epoch": 0.7510987739995374,
      "grad_norm": 1.1430665254592896,
      "learning_rate": 0.0005,
      "loss": 2.951251983642578,
      "step": 3247
    },
    {
      "epoch": 0.7513300948415452,
      "grad_norm": 1.1336946487426758,
      "learning_rate": 0.0005,
      "loss": 2.700138568878174,
      "step": 3248
    },
    {
      "epoch": 0.7515614156835531,
      "grad_norm": 1.1428349018096924,
      "learning_rate": 0.0005,
      "loss": 2.5836219787597656,
      "step": 3249
    },
    {
      "epoch": 0.751792736525561,
      "grad_norm": 0.9291232824325562,
      "learning_rate": 0.0005,
      "loss": 2.560476541519165,
      "step": 3250
    },
    {
      "epoch": 0.7520240573675688,
      "grad_norm": 1.0779929161071777,
      "learning_rate": 0.0005,
      "loss": 2.578278064727783,
      "step": 3251
    },
    {
      "epoch": 0.7522553782095767,
      "grad_norm": 1.125152587890625,
      "learning_rate": 0.0005,
      "loss": 2.5897951126098633,
      "step": 3252
    },
    {
      "epoch": 0.7524866990515845,
      "grad_norm": 1.366114854812622,
      "learning_rate": 0.0005,
      "loss": 2.7593894004821777,
      "step": 3253
    },
    {
      "epoch": 0.7527180198935924,
      "grad_norm": 0.8819989562034607,
      "learning_rate": 0.0005,
      "loss": 3.161375045776367,
      "step": 3254
    },
    {
      "epoch": 0.7529493407356003,
      "grad_norm": 1.0593692064285278,
      "learning_rate": 0.0005,
      "loss": 2.5377144813537598,
      "step": 3255
    },
    {
      "epoch": 0.7531806615776081,
      "grad_norm": 0.8754164576530457,
      "learning_rate": 0.0005,
      "loss": 2.902411460876465,
      "step": 3256
    },
    {
      "epoch": 0.753411982419616,
      "grad_norm": 0.9625754356384277,
      "learning_rate": 0.0005,
      "loss": 2.723503828048706,
      "step": 3257
    },
    {
      "epoch": 0.7536433032616239,
      "grad_norm": 0.8163641691207886,
      "learning_rate": 0.0005,
      "loss": 2.670109272003174,
      "step": 3258
    },
    {
      "epoch": 0.7538746241036317,
      "grad_norm": 0.9905105829238892,
      "learning_rate": 0.0005,
      "loss": 2.576310634613037,
      "step": 3259
    },
    {
      "epoch": 0.7541059449456397,
      "grad_norm": 0.8627825379371643,
      "learning_rate": 0.0005,
      "loss": 2.563453435897827,
      "step": 3260
    },
    {
      "epoch": 0.7543372657876475,
      "grad_norm": 0.9001531600952148,
      "learning_rate": 0.0005,
      "loss": 2.6221327781677246,
      "step": 3261
    },
    {
      "epoch": 0.7545685866296553,
      "grad_norm": 1.0313498973846436,
      "learning_rate": 0.0005,
      "loss": 3.093604803085327,
      "step": 3262
    },
    {
      "epoch": 0.7547999074716631,
      "grad_norm": 0.8068971633911133,
      "learning_rate": 0.0005,
      "loss": 2.364974021911621,
      "step": 3263
    },
    {
      "epoch": 0.7550312283136711,
      "grad_norm": 0.9681753516197205,
      "learning_rate": 0.0005,
      "loss": 2.6928067207336426,
      "step": 3264
    },
    {
      "epoch": 0.7552625491556789,
      "grad_norm": 1.033634066581726,
      "learning_rate": 0.0005,
      "loss": 2.5521798133850098,
      "step": 3265
    },
    {
      "epoch": 0.7554938699976868,
      "grad_norm": 0.9179587960243225,
      "learning_rate": 0.0005,
      "loss": 2.586686849594116,
      "step": 3266
    },
    {
      "epoch": 0.7557251908396947,
      "grad_norm": 1.648945927619934,
      "learning_rate": 0.0005,
      "loss": 2.577328681945801,
      "step": 3267
    },
    {
      "epoch": 0.7559565116817025,
      "grad_norm": 1.015557885169983,
      "learning_rate": 0.0005,
      "loss": 2.7977566719055176,
      "step": 3268
    },
    {
      "epoch": 0.7561878325237104,
      "grad_norm": 1.0981324911117554,
      "learning_rate": 0.0005,
      "loss": 2.8785150051116943,
      "step": 3269
    },
    {
      "epoch": 0.7564191533657183,
      "grad_norm": 1.3903793096542358,
      "learning_rate": 0.0005,
      "loss": 2.6104249954223633,
      "step": 3270
    },
    {
      "epoch": 0.7566504742077261,
      "grad_norm": 1.0777616500854492,
      "learning_rate": 0.0005,
      "loss": 2.506551504135132,
      "step": 3271
    },
    {
      "epoch": 0.756881795049734,
      "grad_norm": 0.89175945520401,
      "learning_rate": 0.0005,
      "loss": 2.5187063217163086,
      "step": 3272
    },
    {
      "epoch": 0.7571131158917418,
      "grad_norm": 0.9645922780036926,
      "learning_rate": 0.0005,
      "loss": 2.6410388946533203,
      "step": 3273
    },
    {
      "epoch": 0.7573444367337497,
      "grad_norm": 0.9750110507011414,
      "learning_rate": 0.0005,
      "loss": 2.858126401901245,
      "step": 3274
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 1.2033922672271729,
      "learning_rate": 0.0005,
      "loss": 2.6073904037475586,
      "step": 3275
    },
    {
      "epoch": 0.7578070784177654,
      "grad_norm": 0.8874656558036804,
      "learning_rate": 0.0005,
      "loss": 2.777784824371338,
      "step": 3276
    },
    {
      "epoch": 0.7580383992597733,
      "grad_norm": 1.069826364517212,
      "learning_rate": 0.0005,
      "loss": 2.336207866668701,
      "step": 3277
    },
    {
      "epoch": 0.7582697201017812,
      "grad_norm": 0.9682414531707764,
      "learning_rate": 0.0005,
      "loss": 2.3817214965820312,
      "step": 3278
    },
    {
      "epoch": 0.758501040943789,
      "grad_norm": 1.0156055688858032,
      "learning_rate": 0.0005,
      "loss": 2.4302124977111816,
      "step": 3279
    },
    {
      "epoch": 0.758732361785797,
      "grad_norm": 0.9634511470794678,
      "learning_rate": 0.0005,
      "loss": 2.741191864013672,
      "step": 3280
    },
    {
      "epoch": 0.7589636826278048,
      "grad_norm": 0.997647762298584,
      "learning_rate": 0.0005,
      "loss": 2.2885313034057617,
      "step": 3281
    },
    {
      "epoch": 0.7591950034698126,
      "grad_norm": 1.2471336126327515,
      "learning_rate": 0.0005,
      "loss": 2.6911182403564453,
      "step": 3282
    },
    {
      "epoch": 0.7594263243118204,
      "grad_norm": 1.3236826658248901,
      "learning_rate": 0.0005,
      "loss": 2.9360179901123047,
      "step": 3283
    },
    {
      "epoch": 0.7596576451538284,
      "grad_norm": 0.976858913898468,
      "learning_rate": 0.0005,
      "loss": 2.807199239730835,
      "step": 3284
    },
    {
      "epoch": 0.7598889659958362,
      "grad_norm": 1.061131238937378,
      "learning_rate": 0.0005,
      "loss": 2.861595869064331,
      "step": 3285
    },
    {
      "epoch": 0.760120286837844,
      "grad_norm": 0.8331253528594971,
      "learning_rate": 0.0005,
      "loss": 2.5990889072418213,
      "step": 3286
    },
    {
      "epoch": 0.760351607679852,
      "grad_norm": 1.1194896697998047,
      "learning_rate": 0.0005,
      "loss": 2.213594913482666,
      "step": 3287
    },
    {
      "epoch": 0.7605829285218598,
      "grad_norm": 0.9331411719322205,
      "learning_rate": 0.0005,
      "loss": 2.243144989013672,
      "step": 3288
    },
    {
      "epoch": 0.7608142493638677,
      "grad_norm": 1.1693227291107178,
      "learning_rate": 0.0005,
      "loss": 2.728522777557373,
      "step": 3289
    },
    {
      "epoch": 0.7610455702058756,
      "grad_norm": 1.0352556705474854,
      "learning_rate": 0.0005,
      "loss": 2.9389240741729736,
      "step": 3290
    },
    {
      "epoch": 0.7612768910478834,
      "grad_norm": 1.2995922565460205,
      "learning_rate": 0.0005,
      "loss": 2.6607589721679688,
      "step": 3291
    },
    {
      "epoch": 0.7615082118898913,
      "grad_norm": 0.9800034761428833,
      "learning_rate": 0.0005,
      "loss": 2.6316349506378174,
      "step": 3292
    },
    {
      "epoch": 0.7617395327318991,
      "grad_norm": 0.9079344868659973,
      "learning_rate": 0.0005,
      "loss": 2.8711085319519043,
      "step": 3293
    },
    {
      "epoch": 0.761970853573907,
      "grad_norm": 1.1910624504089355,
      "learning_rate": 0.0005,
      "loss": 2.788811206817627,
      "step": 3294
    },
    {
      "epoch": 0.7622021744159149,
      "grad_norm": 1.146209478378296,
      "learning_rate": 0.0005,
      "loss": 2.8485660552978516,
      "step": 3295
    },
    {
      "epoch": 0.7624334952579227,
      "grad_norm": 1.0640980005264282,
      "learning_rate": 0.0005,
      "loss": 2.4541752338409424,
      "step": 3296
    },
    {
      "epoch": 0.7626648160999306,
      "grad_norm": 0.8071463108062744,
      "learning_rate": 0.0005,
      "loss": 2.5978283882141113,
      "step": 3297
    },
    {
      "epoch": 0.7628961369419385,
      "grad_norm": 1.170548915863037,
      "learning_rate": 0.0005,
      "loss": 2.6773810386657715,
      "step": 3298
    },
    {
      "epoch": 0.7631274577839463,
      "grad_norm": 1.2071725130081177,
      "learning_rate": 0.0005,
      "loss": 2.282707452774048,
      "step": 3299
    },
    {
      "epoch": 0.7633587786259542,
      "grad_norm": 0.935631513595581,
      "learning_rate": 0.0005,
      "loss": 2.7414920330047607,
      "step": 3300
    },
    {
      "epoch": 0.7635900994679621,
      "grad_norm": 1.5369648933410645,
      "learning_rate": 0.0005,
      "loss": 2.8608388900756836,
      "step": 3301
    },
    {
      "epoch": 0.7638214203099699,
      "grad_norm": 1.0080968141555786,
      "learning_rate": 0.0005,
      "loss": 2.4550609588623047,
      "step": 3302
    },
    {
      "epoch": 0.7640527411519777,
      "grad_norm": 1.0099849700927734,
      "learning_rate": 0.0005,
      "loss": 2.589965581893921,
      "step": 3303
    },
    {
      "epoch": 0.7642840619939857,
      "grad_norm": 1.2949224710464478,
      "learning_rate": 0.0005,
      "loss": 2.2120559215545654,
      "step": 3304
    },
    {
      "epoch": 0.7645153828359935,
      "grad_norm": 1.1756123304367065,
      "learning_rate": 0.0005,
      "loss": 2.886326789855957,
      "step": 3305
    },
    {
      "epoch": 0.7647467036780013,
      "grad_norm": 1.2520380020141602,
      "learning_rate": 0.0005,
      "loss": 2.7878494262695312,
      "step": 3306
    },
    {
      "epoch": 0.7649780245200093,
      "grad_norm": 1.8283963203430176,
      "learning_rate": 0.0005,
      "loss": 2.4826531410217285,
      "step": 3307
    },
    {
      "epoch": 0.7652093453620171,
      "grad_norm": 0.9344798922538757,
      "learning_rate": 0.0005,
      "loss": 2.420175552368164,
      "step": 3308
    },
    {
      "epoch": 0.765440666204025,
      "grad_norm": 1.2762837409973145,
      "learning_rate": 0.0005,
      "loss": 2.5300133228302,
      "step": 3309
    },
    {
      "epoch": 0.7656719870460329,
      "grad_norm": 0.9148538112640381,
      "learning_rate": 0.0005,
      "loss": 2.929372787475586,
      "step": 3310
    },
    {
      "epoch": 0.7659033078880407,
      "grad_norm": 0.9534444808959961,
      "learning_rate": 0.0005,
      "loss": 2.414325475692749,
      "step": 3311
    },
    {
      "epoch": 0.7661346287300486,
      "grad_norm": 0.9930430054664612,
      "learning_rate": 0.0005,
      "loss": 2.477362632751465,
      "step": 3312
    },
    {
      "epoch": 0.7663659495720564,
      "grad_norm": 0.8801076412200928,
      "learning_rate": 0.0005,
      "loss": 2.69382643699646,
      "step": 3313
    },
    {
      "epoch": 0.7665972704140643,
      "grad_norm": 1.1121842861175537,
      "learning_rate": 0.0005,
      "loss": 2.9083802700042725,
      "step": 3314
    },
    {
      "epoch": 0.7668285912560722,
      "grad_norm": 1.0117881298065186,
      "learning_rate": 0.0005,
      "loss": 2.7481250762939453,
      "step": 3315
    },
    {
      "epoch": 0.76705991209808,
      "grad_norm": 0.8082316517829895,
      "learning_rate": 0.0005,
      "loss": 2.875950813293457,
      "step": 3316
    },
    {
      "epoch": 0.7672912329400879,
      "grad_norm": 1.2738010883331299,
      "learning_rate": 0.0005,
      "loss": 2.8098220825195312,
      "step": 3317
    },
    {
      "epoch": 0.7675225537820958,
      "grad_norm": 0.7732531428337097,
      "learning_rate": 0.0005,
      "loss": 2.3104610443115234,
      "step": 3318
    },
    {
      "epoch": 0.7677538746241036,
      "grad_norm": 0.9867333769798279,
      "learning_rate": 0.0005,
      "loss": 2.6207361221313477,
      "step": 3319
    },
    {
      "epoch": 0.7679851954661115,
      "grad_norm": 1.2083113193511963,
      "learning_rate": 0.0005,
      "loss": 2.9351701736450195,
      "step": 3320
    },
    {
      "epoch": 0.7682165163081194,
      "grad_norm": 0.9643878936767578,
      "learning_rate": 0.0005,
      "loss": 2.6936593055725098,
      "step": 3321
    },
    {
      "epoch": 0.7684478371501272,
      "grad_norm": 1.6628506183624268,
      "learning_rate": 0.0005,
      "loss": 2.4968674182891846,
      "step": 3322
    },
    {
      "epoch": 0.768679157992135,
      "grad_norm": 1.543040156364441,
      "learning_rate": 0.0005,
      "loss": 2.375140428543091,
      "step": 3323
    },
    {
      "epoch": 0.768910478834143,
      "grad_norm": 0.9376607537269592,
      "learning_rate": 0.0005,
      "loss": 2.692056655883789,
      "step": 3324
    },
    {
      "epoch": 0.7691417996761508,
      "grad_norm": 1.174622893333435,
      "learning_rate": 0.0005,
      "loss": 2.6493749618530273,
      "step": 3325
    },
    {
      "epoch": 0.7693731205181586,
      "grad_norm": 1.328747272491455,
      "learning_rate": 0.0005,
      "loss": 2.7336769104003906,
      "step": 3326
    },
    {
      "epoch": 0.7696044413601666,
      "grad_norm": 0.9336119890213013,
      "learning_rate": 0.0005,
      "loss": 2.896350383758545,
      "step": 3327
    },
    {
      "epoch": 0.7698357622021744,
      "grad_norm": 0.8297838568687439,
      "learning_rate": 0.0005,
      "loss": 2.7727160453796387,
      "step": 3328
    },
    {
      "epoch": 0.7700670830441823,
      "grad_norm": 0.8436451554298401,
      "learning_rate": 0.0005,
      "loss": 2.9171204566955566,
      "step": 3329
    },
    {
      "epoch": 0.7702984038861902,
      "grad_norm": 0.8717143535614014,
      "learning_rate": 0.0005,
      "loss": 2.3221912384033203,
      "step": 3330
    },
    {
      "epoch": 0.770529724728198,
      "grad_norm": 0.9929223656654358,
      "learning_rate": 0.0005,
      "loss": 2.7240536212921143,
      "step": 3331
    },
    {
      "epoch": 0.7707610455702059,
      "grad_norm": 1.1666538715362549,
      "learning_rate": 0.0005,
      "loss": 2.543321132659912,
      "step": 3332
    },
    {
      "epoch": 0.7709923664122137,
      "grad_norm": 0.9537054300308228,
      "learning_rate": 0.0005,
      "loss": 2.4536523818969727,
      "step": 3333
    },
    {
      "epoch": 0.7712236872542216,
      "grad_norm": 0.8458637595176697,
      "learning_rate": 0.0005,
      "loss": 2.9284229278564453,
      "step": 3334
    },
    {
      "epoch": 0.7714550080962295,
      "grad_norm": 0.9240090250968933,
      "learning_rate": 0.0005,
      "loss": 2.3532614707946777,
      "step": 3335
    },
    {
      "epoch": 0.7716863289382373,
      "grad_norm": 1.2393978834152222,
      "learning_rate": 0.0005,
      "loss": 2.4476537704467773,
      "step": 3336
    },
    {
      "epoch": 0.7719176497802452,
      "grad_norm": 1.0239118337631226,
      "learning_rate": 0.0005,
      "loss": 2.509229898452759,
      "step": 3337
    },
    {
      "epoch": 0.7721489706222531,
      "grad_norm": 0.8968146443367004,
      "learning_rate": 0.0005,
      "loss": 2.437840461730957,
      "step": 3338
    },
    {
      "epoch": 0.7723802914642609,
      "grad_norm": 0.937821626663208,
      "learning_rate": 0.0005,
      "loss": 2.5618410110473633,
      "step": 3339
    },
    {
      "epoch": 0.7726116123062688,
      "grad_norm": 1.054783821105957,
      "learning_rate": 0.0005,
      "loss": 2.680833339691162,
      "step": 3340
    },
    {
      "epoch": 0.7728429331482767,
      "grad_norm": 1.1748250722885132,
      "learning_rate": 0.0005,
      "loss": 2.3820600509643555,
      "step": 3341
    },
    {
      "epoch": 0.7730742539902845,
      "grad_norm": 1.0260555744171143,
      "learning_rate": 0.0005,
      "loss": 2.364128828048706,
      "step": 3342
    },
    {
      "epoch": 0.7733055748322923,
      "grad_norm": 0.8476868867874146,
      "learning_rate": 0.0005,
      "loss": 2.7681384086608887,
      "step": 3343
    },
    {
      "epoch": 0.7735368956743003,
      "grad_norm": 0.9529555439949036,
      "learning_rate": 0.0005,
      "loss": 2.4931836128234863,
      "step": 3344
    },
    {
      "epoch": 0.7737682165163081,
      "grad_norm": 0.7702715992927551,
      "learning_rate": 0.0005,
      "loss": 2.3702926635742188,
      "step": 3345
    },
    {
      "epoch": 0.773999537358316,
      "grad_norm": 1.113584041595459,
      "learning_rate": 0.0005,
      "loss": 2.7020366191864014,
      "step": 3346
    },
    {
      "epoch": 0.7742308582003239,
      "grad_norm": 0.8655537366867065,
      "learning_rate": 0.0005,
      "loss": 2.2919774055480957,
      "step": 3347
    },
    {
      "epoch": 0.7744621790423317,
      "grad_norm": 1.066800594329834,
      "learning_rate": 0.0005,
      "loss": 2.735377550125122,
      "step": 3348
    },
    {
      "epoch": 0.7746934998843396,
      "grad_norm": 1.0938407182693481,
      "learning_rate": 0.0005,
      "loss": 2.648144483566284,
      "step": 3349
    },
    {
      "epoch": 0.7749248207263475,
      "grad_norm": 1.0984686613082886,
      "learning_rate": 0.0005,
      "loss": 2.9926388263702393,
      "step": 3350
    },
    {
      "epoch": 0.7751561415683553,
      "grad_norm": 1.0680999755859375,
      "learning_rate": 0.0005,
      "loss": 2.4913182258605957,
      "step": 3351
    },
    {
      "epoch": 0.7753874624103632,
      "grad_norm": 1.0035889148712158,
      "learning_rate": 0.0005,
      "loss": 2.391313314437866,
      "step": 3352
    },
    {
      "epoch": 0.775618783252371,
      "grad_norm": 0.9088277816772461,
      "learning_rate": 0.0005,
      "loss": 2.464184522628784,
      "step": 3353
    },
    {
      "epoch": 0.7758501040943789,
      "grad_norm": 1.1305999755859375,
      "learning_rate": 0.0005,
      "loss": 2.7138960361480713,
      "step": 3354
    },
    {
      "epoch": 0.7760814249363868,
      "grad_norm": 1.053524136543274,
      "learning_rate": 0.0005,
      "loss": 2.5613789558410645,
      "step": 3355
    },
    {
      "epoch": 0.7763127457783946,
      "grad_norm": 4.4688496589660645,
      "learning_rate": 0.0005,
      "loss": 2.7642600536346436,
      "step": 3356
    },
    {
      "epoch": 0.7765440666204025,
      "grad_norm": 1.1465705633163452,
      "learning_rate": 0.0005,
      "loss": 2.424290180206299,
      "step": 3357
    },
    {
      "epoch": 0.7767753874624104,
      "grad_norm": 0.9321224689483643,
      "learning_rate": 0.0005,
      "loss": 2.749634265899658,
      "step": 3358
    },
    {
      "epoch": 0.7770067083044182,
      "grad_norm": 1.2294650077819824,
      "learning_rate": 0.0005,
      "loss": 2.1330957412719727,
      "step": 3359
    },
    {
      "epoch": 0.7772380291464261,
      "grad_norm": 1.1314997673034668,
      "learning_rate": 0.0005,
      "loss": 2.768425941467285,
      "step": 3360
    },
    {
      "epoch": 0.777469349988434,
      "grad_norm": 0.9637669920921326,
      "learning_rate": 0.0005,
      "loss": 2.4016342163085938,
      "step": 3361
    },
    {
      "epoch": 0.7777006708304418,
      "grad_norm": 1.2593224048614502,
      "learning_rate": 0.0005,
      "loss": 2.371212959289551,
      "step": 3362
    },
    {
      "epoch": 0.7779319916724496,
      "grad_norm": 1.2599923610687256,
      "learning_rate": 0.0005,
      "loss": 2.518383026123047,
      "step": 3363
    },
    {
      "epoch": 0.7781633125144576,
      "grad_norm": 0.9631789326667786,
      "learning_rate": 0.0005,
      "loss": 2.6514527797698975,
      "step": 3364
    },
    {
      "epoch": 0.7783946333564654,
      "grad_norm": 1.151654839515686,
      "learning_rate": 0.0005,
      "loss": 2.7485718727111816,
      "step": 3365
    },
    {
      "epoch": 0.7786259541984732,
      "grad_norm": 1.0902431011199951,
      "learning_rate": 0.0005,
      "loss": 2.385960102081299,
      "step": 3366
    },
    {
      "epoch": 0.7788572750404812,
      "grad_norm": 1.1039632558822632,
      "learning_rate": 0.0005,
      "loss": 2.484128713607788,
      "step": 3367
    },
    {
      "epoch": 0.779088595882489,
      "grad_norm": 1.255305290222168,
      "learning_rate": 0.0005,
      "loss": 2.8014307022094727,
      "step": 3368
    },
    {
      "epoch": 0.7793199167244969,
      "grad_norm": 0.8526194095611572,
      "learning_rate": 0.0005,
      "loss": 2.4529151916503906,
      "step": 3369
    },
    {
      "epoch": 0.7795512375665048,
      "grad_norm": 0.9008347988128662,
      "learning_rate": 0.0005,
      "loss": 2.5666427612304688,
      "step": 3370
    },
    {
      "epoch": 0.7797825584085126,
      "grad_norm": 0.9376559853553772,
      "learning_rate": 0.0005,
      "loss": 2.5318026542663574,
      "step": 3371
    },
    {
      "epoch": 0.7800138792505205,
      "grad_norm": 1.0514476299285889,
      "learning_rate": 0.0005,
      "loss": 2.7378408908843994,
      "step": 3372
    },
    {
      "epoch": 0.7802452000925283,
      "grad_norm": 0.743201494216919,
      "learning_rate": 0.0005,
      "loss": 2.6292929649353027,
      "step": 3373
    },
    {
      "epoch": 0.7804765209345362,
      "grad_norm": 1.0503735542297363,
      "learning_rate": 0.0005,
      "loss": 2.8196630477905273,
      "step": 3374
    },
    {
      "epoch": 0.7807078417765441,
      "grad_norm": 1.1816043853759766,
      "learning_rate": 0.0005,
      "loss": 2.3500046730041504,
      "step": 3375
    },
    {
      "epoch": 0.7809391626185519,
      "grad_norm": 1.34983491897583,
      "learning_rate": 0.0005,
      "loss": 2.3337836265563965,
      "step": 3376
    },
    {
      "epoch": 0.7811704834605598,
      "grad_norm": 1.0109715461730957,
      "learning_rate": 0.0005,
      "loss": 2.545203685760498,
      "step": 3377
    },
    {
      "epoch": 0.7814018043025677,
      "grad_norm": 0.9937698841094971,
      "learning_rate": 0.0005,
      "loss": 2.495079517364502,
      "step": 3378
    },
    {
      "epoch": 0.7816331251445755,
      "grad_norm": 0.9480807185173035,
      "learning_rate": 0.0005,
      "loss": 2.6700258255004883,
      "step": 3379
    },
    {
      "epoch": 0.7818644459865834,
      "grad_norm": 0.8591009378433228,
      "learning_rate": 0.0005,
      "loss": 3.004652500152588,
      "step": 3380
    },
    {
      "epoch": 0.7820957668285913,
      "grad_norm": 0.9571846723556519,
      "learning_rate": 0.0005,
      "loss": 2.3372812271118164,
      "step": 3381
    },
    {
      "epoch": 0.7823270876705991,
      "grad_norm": 1.138134241104126,
      "learning_rate": 0.0005,
      "loss": 2.424769401550293,
      "step": 3382
    },
    {
      "epoch": 0.7825584085126069,
      "grad_norm": 0.8926860094070435,
      "learning_rate": 0.0005,
      "loss": 2.632355213165283,
      "step": 3383
    },
    {
      "epoch": 0.7827897293546149,
      "grad_norm": 1.178519368171692,
      "learning_rate": 0.0005,
      "loss": 2.6069250106811523,
      "step": 3384
    },
    {
      "epoch": 0.7830210501966227,
      "grad_norm": 1.0586429834365845,
      "learning_rate": 0.0005,
      "loss": 2.3455562591552734,
      "step": 3385
    },
    {
      "epoch": 0.7832523710386305,
      "grad_norm": 1.1789884567260742,
      "learning_rate": 0.0005,
      "loss": 2.9924020767211914,
      "step": 3386
    },
    {
      "epoch": 0.7834836918806385,
      "grad_norm": 1.4663877487182617,
      "learning_rate": 0.0005,
      "loss": 2.380715847015381,
      "step": 3387
    },
    {
      "epoch": 0.7837150127226463,
      "grad_norm": 0.9136131405830383,
      "learning_rate": 0.0005,
      "loss": 2.419644355773926,
      "step": 3388
    },
    {
      "epoch": 0.7839463335646542,
      "grad_norm": 0.9443619847297668,
      "learning_rate": 0.0005,
      "loss": 2.650667190551758,
      "step": 3389
    },
    {
      "epoch": 0.7841776544066621,
      "grad_norm": 1.125836730003357,
      "learning_rate": 0.0005,
      "loss": 2.4545340538024902,
      "step": 3390
    },
    {
      "epoch": 0.7844089752486699,
      "grad_norm": 0.8262335062026978,
      "learning_rate": 0.0005,
      "loss": 2.536580801010132,
      "step": 3391
    },
    {
      "epoch": 0.7846402960906778,
      "grad_norm": 1.0352495908737183,
      "learning_rate": 0.0005,
      "loss": 2.306034564971924,
      "step": 3392
    },
    {
      "epoch": 0.7848716169326856,
      "grad_norm": 1.1223042011260986,
      "learning_rate": 0.0005,
      "loss": 2.765120029449463,
      "step": 3393
    },
    {
      "epoch": 0.7851029377746935,
      "grad_norm": 0.868260383605957,
      "learning_rate": 0.0005,
      "loss": 2.911890983581543,
      "step": 3394
    },
    {
      "epoch": 0.7853342586167014,
      "grad_norm": 0.9682134389877319,
      "learning_rate": 0.0005,
      "loss": 2.5211143493652344,
      "step": 3395
    },
    {
      "epoch": 0.7855655794587092,
      "grad_norm": 1.2264583110809326,
      "learning_rate": 0.0005,
      "loss": 2.504434823989868,
      "step": 3396
    },
    {
      "epoch": 0.7857969003007171,
      "grad_norm": 1.1701176166534424,
      "learning_rate": 0.0005,
      "loss": 2.3519344329833984,
      "step": 3397
    },
    {
      "epoch": 0.786028221142725,
      "grad_norm": 0.9221862554550171,
      "learning_rate": 0.0005,
      "loss": 2.5273547172546387,
      "step": 3398
    },
    {
      "epoch": 0.7862595419847328,
      "grad_norm": 0.9610531330108643,
      "learning_rate": 0.0005,
      "loss": 2.3393020629882812,
      "step": 3399
    },
    {
      "epoch": 0.7864908628267407,
      "grad_norm": 1.036123514175415,
      "learning_rate": 0.0005,
      "loss": 2.495117425918579,
      "step": 3400
    },
    {
      "epoch": 0.7867221836687486,
      "grad_norm": 1.4270172119140625,
      "learning_rate": 0.0005,
      "loss": 2.355867624282837,
      "step": 3401
    },
    {
      "epoch": 0.7869535045107564,
      "grad_norm": 1.1610612869262695,
      "learning_rate": 0.0005,
      "loss": 2.817044973373413,
      "step": 3402
    },
    {
      "epoch": 0.7871848253527642,
      "grad_norm": 1.2861049175262451,
      "learning_rate": 0.0005,
      "loss": 2.538658618927002,
      "step": 3403
    },
    {
      "epoch": 0.7874161461947722,
      "grad_norm": 0.8178227543830872,
      "learning_rate": 0.0005,
      "loss": 2.516049861907959,
      "step": 3404
    },
    {
      "epoch": 0.78764746703678,
      "grad_norm": 0.8990283608436584,
      "learning_rate": 0.0005,
      "loss": 2.761366367340088,
      "step": 3405
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9443805813789368,
      "learning_rate": 0.0005,
      "loss": 2.7362871170043945,
      "step": 3406
    },
    {
      "epoch": 0.7881101087207958,
      "grad_norm": 1.003685712814331,
      "learning_rate": 0.0005,
      "loss": 2.3937902450561523,
      "step": 3407
    },
    {
      "epoch": 0.7883414295628036,
      "grad_norm": 1.0421020984649658,
      "learning_rate": 0.0005,
      "loss": 2.5787901878356934,
      "step": 3408
    },
    {
      "epoch": 0.7885727504048115,
      "grad_norm": 0.9703618884086609,
      "learning_rate": 0.0005,
      "loss": 2.2090697288513184,
      "step": 3409
    },
    {
      "epoch": 0.7888040712468194,
      "grad_norm": 1.2425185441970825,
      "learning_rate": 0.0005,
      "loss": 2.5438060760498047,
      "step": 3410
    },
    {
      "epoch": 0.7890353920888272,
      "grad_norm": 1.1922024488449097,
      "learning_rate": 0.0005,
      "loss": 2.101543426513672,
      "step": 3411
    },
    {
      "epoch": 0.7892667129308351,
      "grad_norm": 1.2174519300460815,
      "learning_rate": 0.0005,
      "loss": 2.8214826583862305,
      "step": 3412
    },
    {
      "epoch": 0.7894980337728429,
      "grad_norm": 1.1543552875518799,
      "learning_rate": 0.0005,
      "loss": 2.265944242477417,
      "step": 3413
    },
    {
      "epoch": 0.7897293546148508,
      "grad_norm": 1.0214438438415527,
      "learning_rate": 0.0005,
      "loss": 2.256359577178955,
      "step": 3414
    },
    {
      "epoch": 0.7899606754568587,
      "grad_norm": 0.8817249536514282,
      "learning_rate": 0.0005,
      "loss": 2.5968496799468994,
      "step": 3415
    },
    {
      "epoch": 0.7901919962988665,
      "grad_norm": 6.1144118309021,
      "learning_rate": 0.0005,
      "loss": 2.225923538208008,
      "step": 3416
    },
    {
      "epoch": 0.7904233171408744,
      "grad_norm": 0.9849463701248169,
      "learning_rate": 0.0005,
      "loss": 2.90741229057312,
      "step": 3417
    },
    {
      "epoch": 0.7906546379828823,
      "grad_norm": 0.9359848499298096,
      "learning_rate": 0.0005,
      "loss": 2.518887996673584,
      "step": 3418
    },
    {
      "epoch": 0.7908859588248901,
      "grad_norm": 0.8939897418022156,
      "learning_rate": 0.0005,
      "loss": 2.4300038814544678,
      "step": 3419
    },
    {
      "epoch": 0.7911172796668979,
      "grad_norm": 0.8209629654884338,
      "learning_rate": 0.0005,
      "loss": 2.5174708366394043,
      "step": 3420
    },
    {
      "epoch": 0.7913486005089059,
      "grad_norm": 0.946713387966156,
      "learning_rate": 0.0005,
      "loss": 2.4427919387817383,
      "step": 3421
    },
    {
      "epoch": 0.7915799213509137,
      "grad_norm": 1.244836688041687,
      "learning_rate": 0.0005,
      "loss": 2.2697694301605225,
      "step": 3422
    },
    {
      "epoch": 0.7918112421929215,
      "grad_norm": 1.050675868988037,
      "learning_rate": 0.0005,
      "loss": 2.1847941875457764,
      "step": 3423
    },
    {
      "epoch": 0.7920425630349295,
      "grad_norm": 0.9684081077575684,
      "learning_rate": 0.0005,
      "loss": 2.5724589824676514,
      "step": 3424
    },
    {
      "epoch": 0.7922738838769373,
      "grad_norm": 1.1361744403839111,
      "learning_rate": 0.0005,
      "loss": 2.6711926460266113,
      "step": 3425
    },
    {
      "epoch": 0.7925052047189451,
      "grad_norm": 0.8960947394371033,
      "learning_rate": 0.0005,
      "loss": 2.7904105186462402,
      "step": 3426
    },
    {
      "epoch": 0.7927365255609531,
      "grad_norm": 0.9938209652900696,
      "learning_rate": 0.0005,
      "loss": 2.4194118976593018,
      "step": 3427
    },
    {
      "epoch": 0.7929678464029609,
      "grad_norm": 0.8392536044120789,
      "learning_rate": 0.0005,
      "loss": 2.81235408782959,
      "step": 3428
    },
    {
      "epoch": 0.7931991672449688,
      "grad_norm": 0.9480345845222473,
      "learning_rate": 0.0005,
      "loss": 2.643514633178711,
      "step": 3429
    },
    {
      "epoch": 0.7934304880869766,
      "grad_norm": 1.1756343841552734,
      "learning_rate": 0.0005,
      "loss": 2.7131872177124023,
      "step": 3430
    },
    {
      "epoch": 0.7936618089289845,
      "grad_norm": 0.9259072542190552,
      "learning_rate": 0.0005,
      "loss": 2.5601136684417725,
      "step": 3431
    },
    {
      "epoch": 0.7938931297709924,
      "grad_norm": 0.905392050743103,
      "learning_rate": 0.0005,
      "loss": 2.7977919578552246,
      "step": 3432
    },
    {
      "epoch": 0.7941244506130002,
      "grad_norm": 1.3873435258865356,
      "learning_rate": 0.0005,
      "loss": 2.5387024879455566,
      "step": 3433
    },
    {
      "epoch": 0.7943557714550081,
      "grad_norm": 0.9868394732475281,
      "learning_rate": 0.0005,
      "loss": 2.6337976455688477,
      "step": 3434
    },
    {
      "epoch": 0.794587092297016,
      "grad_norm": 1.0882729291915894,
      "learning_rate": 0.0005,
      "loss": 2.255488395690918,
      "step": 3435
    },
    {
      "epoch": 0.7948184131390238,
      "grad_norm": 1.075338363647461,
      "learning_rate": 0.0005,
      "loss": 2.3900351524353027,
      "step": 3436
    },
    {
      "epoch": 0.7950497339810317,
      "grad_norm": 0.9771377444267273,
      "learning_rate": 0.0005,
      "loss": 2.2973012924194336,
      "step": 3437
    },
    {
      "epoch": 0.7952810548230396,
      "grad_norm": 1.0628435611724854,
      "learning_rate": 0.0005,
      "loss": 2.4536285400390625,
      "step": 3438
    },
    {
      "epoch": 0.7955123756650474,
      "grad_norm": 1.7757115364074707,
      "learning_rate": 0.0005,
      "loss": 2.6981630325317383,
      "step": 3439
    },
    {
      "epoch": 0.7957436965070552,
      "grad_norm": 1.0094529390335083,
      "learning_rate": 0.0005,
      "loss": 2.807600736618042,
      "step": 3440
    },
    {
      "epoch": 0.7959750173490632,
      "grad_norm": 1.3862929344177246,
      "learning_rate": 0.0005,
      "loss": 2.679163932800293,
      "step": 3441
    },
    {
      "epoch": 0.796206338191071,
      "grad_norm": 1.2210203409194946,
      "learning_rate": 0.0005,
      "loss": 2.7363100051879883,
      "step": 3442
    },
    {
      "epoch": 0.7964376590330788,
      "grad_norm": 1.0160380601882935,
      "learning_rate": 0.0005,
      "loss": 2.4106738567352295,
      "step": 3443
    },
    {
      "epoch": 0.7966689798750868,
      "grad_norm": 1.18594491481781,
      "learning_rate": 0.0005,
      "loss": 2.6432292461395264,
      "step": 3444
    },
    {
      "epoch": 0.7969003007170946,
      "grad_norm": 0.9376246333122253,
      "learning_rate": 0.0005,
      "loss": 2.437847137451172,
      "step": 3445
    },
    {
      "epoch": 0.7971316215591024,
      "grad_norm": 1.0266507863998413,
      "learning_rate": 0.0005,
      "loss": 2.414503335952759,
      "step": 3446
    },
    {
      "epoch": 0.7973629424011104,
      "grad_norm": 1.135498285293579,
      "learning_rate": 0.0005,
      "loss": 2.89139986038208,
      "step": 3447
    },
    {
      "epoch": 0.7975942632431182,
      "grad_norm": 0.8644741773605347,
      "learning_rate": 0.0005,
      "loss": 2.7551465034484863,
      "step": 3448
    },
    {
      "epoch": 0.797825584085126,
      "grad_norm": 0.9588108062744141,
      "learning_rate": 0.0005,
      "loss": 2.7132315635681152,
      "step": 3449
    },
    {
      "epoch": 0.7980569049271339,
      "grad_norm": 0.9618892669677734,
      "learning_rate": 0.0005,
      "loss": 2.465730667114258,
      "step": 3450
    },
    {
      "epoch": 0.7982882257691418,
      "grad_norm": 1.665445327758789,
      "learning_rate": 0.0005,
      "loss": 2.715578079223633,
      "step": 3451
    },
    {
      "epoch": 0.7985195466111497,
      "grad_norm": 1.588108777999878,
      "learning_rate": 0.0005,
      "loss": 2.667947292327881,
      "step": 3452
    },
    {
      "epoch": 0.7987508674531575,
      "grad_norm": 0.8812775611877441,
      "learning_rate": 0.0005,
      "loss": 2.7926597595214844,
      "step": 3453
    },
    {
      "epoch": 0.7989821882951654,
      "grad_norm": 0.9721949696540833,
      "learning_rate": 0.0005,
      "loss": 2.8496861457824707,
      "step": 3454
    },
    {
      "epoch": 0.7992135091371733,
      "grad_norm": 0.8274266123771667,
      "learning_rate": 0.0005,
      "loss": 2.7715070247650146,
      "step": 3455
    },
    {
      "epoch": 0.7994448299791811,
      "grad_norm": 1.190194010734558,
      "learning_rate": 0.0005,
      "loss": 2.677807092666626,
      "step": 3456
    },
    {
      "epoch": 0.799676150821189,
      "grad_norm": 0.9335930943489075,
      "learning_rate": 0.0005,
      "loss": 2.0600364208221436,
      "step": 3457
    },
    {
      "epoch": 0.7999074716631969,
      "grad_norm": 0.836424708366394,
      "learning_rate": 0.0005,
      "loss": 2.4793710708618164,
      "step": 3458
    },
    {
      "epoch": 0.8001387925052047,
      "grad_norm": 1.152103066444397,
      "learning_rate": 0.0005,
      "loss": 2.6653833389282227,
      "step": 3459
    },
    {
      "epoch": 0.8003701133472125,
      "grad_norm": 1.0507111549377441,
      "learning_rate": 0.0005,
      "loss": 2.501359462738037,
      "step": 3460
    },
    {
      "epoch": 0.8006014341892205,
      "grad_norm": 1.0831481218338013,
      "learning_rate": 0.0004999983473465486,
      "loss": 2.7081518173217773,
      "step": 3461
    },
    {
      "epoch": 0.8008327550312283,
      "grad_norm": 1.2407417297363281,
      "learning_rate": 0.0004999933894080444,
      "loss": 2.5747833251953125,
      "step": 3462
    },
    {
      "epoch": 0.8010640758732361,
      "grad_norm": 1.0423444509506226,
      "learning_rate": 0.0004999851262500374,
      "loss": 2.441563844680786,
      "step": 3463
    },
    {
      "epoch": 0.8012953967152441,
      "grad_norm": 1.0194469690322876,
      "learning_rate": 0.0004999735579817769,
      "loss": 2.473520040512085,
      "step": 3464
    },
    {
      "epoch": 0.8015267175572519,
      "grad_norm": 0.9420165419578552,
      "learning_rate": 0.0004999586847562094,
      "loss": 2.463444948196411,
      "step": 3465
    },
    {
      "epoch": 0.8017580383992597,
      "grad_norm": 0.9454444646835327,
      "learning_rate": 0.0004999405067699773,
      "loss": 2.887403964996338,
      "step": 3466
    },
    {
      "epoch": 0.8019893592412677,
      "grad_norm": 0.8668871521949768,
      "learning_rate": 0.0004999190242634159,
      "loss": 2.6107254028320312,
      "step": 3467
    },
    {
      "epoch": 0.8022206800832755,
      "grad_norm": 0.7922632098197937,
      "learning_rate": 0.0004998942375205502,
      "loss": 2.532144784927368,
      "step": 3468
    },
    {
      "epoch": 0.8024520009252833,
      "grad_norm": 0.9437198042869568,
      "learning_rate": 0.0004998661468690914,
      "loss": 2.597184658050537,
      "step": 3469
    },
    {
      "epoch": 0.8026833217672912,
      "grad_norm": 1.1162612438201904,
      "learning_rate": 0.0004998347526804325,
      "loss": 2.463031768798828,
      "step": 3470
    },
    {
      "epoch": 0.8029146426092991,
      "grad_norm": 0.8843539953231812,
      "learning_rate": 0.0004998000553696432,
      "loss": 2.5915579795837402,
      "step": 3471
    },
    {
      "epoch": 0.803145963451307,
      "grad_norm": 0.946371853351593,
      "learning_rate": 0.0004997620553954645,
      "loss": 2.6342544555664062,
      "step": 3472
    },
    {
      "epoch": 0.8033772842933148,
      "grad_norm": 0.9874760508537292,
      "learning_rate": 0.0004997207532603026,
      "loss": 2.675211191177368,
      "step": 3473
    },
    {
      "epoch": 0.8036086051353227,
      "grad_norm": 1.522451639175415,
      "learning_rate": 0.0004996761495102226,
      "loss": 2.797607183456421,
      "step": 3474
    },
    {
      "epoch": 0.8038399259773306,
      "grad_norm": 1.4051681756973267,
      "learning_rate": 0.0004996282447349408,
      "loss": 2.695358991622925,
      "step": 3475
    },
    {
      "epoch": 0.8040712468193384,
      "grad_norm": 2.16890025138855,
      "learning_rate": 0.0004995770395678171,
      "loss": 2.4664361476898193,
      "step": 3476
    },
    {
      "epoch": 0.8043025676613463,
      "grad_norm": 1.0176467895507812,
      "learning_rate": 0.0004995225346858467,
      "loss": 2.715359926223755,
      "step": 3477
    },
    {
      "epoch": 0.8045338885033542,
      "grad_norm": 0.8944279551506042,
      "learning_rate": 0.0004994647308096509,
      "loss": 2.1956095695495605,
      "step": 3478
    },
    {
      "epoch": 0.804765209345362,
      "grad_norm": 1.3747470378875732,
      "learning_rate": 0.0004994036287034681,
      "loss": 2.170417070388794,
      "step": 3479
    },
    {
      "epoch": 0.8049965301873698,
      "grad_norm": 1.0621405839920044,
      "learning_rate": 0.0004993392291751431,
      "loss": 2.545851230621338,
      "step": 3480
    },
    {
      "epoch": 0.8052278510293778,
      "grad_norm": 0.8388673067092896,
      "learning_rate": 0.0004992715330761167,
      "loss": 2.4648590087890625,
      "step": 3481
    },
    {
      "epoch": 0.8054591718713856,
      "grad_norm": 1.0293620824813843,
      "learning_rate": 0.0004992005413014144,
      "loss": 2.605078935623169,
      "step": 3482
    },
    {
      "epoch": 0.8056904927133934,
      "grad_norm": 0.9771875143051147,
      "learning_rate": 0.0004991262547896347,
      "loss": 2.670497179031372,
      "step": 3483
    },
    {
      "epoch": 0.8059218135554014,
      "grad_norm": 1.0233173370361328,
      "learning_rate": 0.0004990486745229364,
      "loss": 2.62180233001709,
      "step": 3484
    },
    {
      "epoch": 0.8061531343974092,
      "grad_norm": 1.01175057888031,
      "learning_rate": 0.000498967801527026,
      "loss": 2.4817404747009277,
      "step": 3485
    },
    {
      "epoch": 0.806384455239417,
      "grad_norm": 0.8606128692626953,
      "learning_rate": 0.0004988836368711435,
      "loss": 2.468515396118164,
      "step": 3486
    },
    {
      "epoch": 0.806615776081425,
      "grad_norm": 1.0779433250427246,
      "learning_rate": 0.0004987961816680492,
      "loss": 2.601390838623047,
      "step": 3487
    },
    {
      "epoch": 0.8068470969234328,
      "grad_norm": 0.8898071646690369,
      "learning_rate": 0.0004987054370740083,
      "loss": 2.197514057159424,
      "step": 3488
    },
    {
      "epoch": 0.8070784177654406,
      "grad_norm": 0.936867356300354,
      "learning_rate": 0.0004986114042887756,
      "loss": 2.5288639068603516,
      "step": 3489
    },
    {
      "epoch": 0.8073097386074485,
      "grad_norm": 1.0568296909332275,
      "learning_rate": 0.0004985140845555799,
      "loss": 2.6747055053710938,
      "step": 3490
    },
    {
      "epoch": 0.8075410594494564,
      "grad_norm": 1.0471961498260498,
      "learning_rate": 0.0004984134791611077,
      "loss": 2.3096554279327393,
      "step": 3491
    },
    {
      "epoch": 0.8077723802914643,
      "grad_norm": 0.9581644535064697,
      "learning_rate": 0.0004983095894354857,
      "loss": 2.54168438911438,
      "step": 3492
    },
    {
      "epoch": 0.8080037011334721,
      "grad_norm": 1.063942551612854,
      "learning_rate": 0.0004982024167522638,
      "loss": 2.7450757026672363,
      "step": 3493
    },
    {
      "epoch": 0.80823502197548,
      "grad_norm": 1.2741544246673584,
      "learning_rate": 0.0004980919625283962,
      "loss": 2.4494595527648926,
      "step": 3494
    },
    {
      "epoch": 0.8084663428174879,
      "grad_norm": 0.9622541069984436,
      "learning_rate": 0.0004979782282242234,
      "loss": 2.6459691524505615,
      "step": 3495
    },
    {
      "epoch": 0.8086976636594957,
      "grad_norm": 1.0172996520996094,
      "learning_rate": 0.0004978612153434526,
      "loss": 2.388911485671997,
      "step": 3496
    },
    {
      "epoch": 0.8089289845015036,
      "grad_norm": 0.8554624319076538,
      "learning_rate": 0.0004977409254331377,
      "loss": 2.324234962463379,
      "step": 3497
    },
    {
      "epoch": 0.8091603053435115,
      "grad_norm": 0.9814204573631287,
      "learning_rate": 0.000497617360083659,
      "loss": 3.155993700027466,
      "step": 3498
    },
    {
      "epoch": 0.8093916261855193,
      "grad_norm": 1.1120080947875977,
      "learning_rate": 0.0004974905209287019,
      "loss": 2.604435920715332,
      "step": 3499
    },
    {
      "epoch": 0.8096229470275271,
      "grad_norm": 0.8267092108726501,
      "learning_rate": 0.0004973604096452361,
      "loss": 2.5830509662628174,
      "step": 3500
    },
    {
      "epoch": 0.8098542678695351,
      "grad_norm": 0.8948646783828735,
      "learning_rate": 0.0004972270279534923,
      "loss": 2.5177969932556152,
      "step": 3501
    },
    {
      "epoch": 0.8100855887115429,
      "grad_norm": 0.9137309193611145,
      "learning_rate": 0.0004970903776169402,
      "loss": 2.532268524169922,
      "step": 3502
    },
    {
      "epoch": 0.8103169095535507,
      "grad_norm": 1.3275723457336426,
      "learning_rate": 0.000496950460442265,
      "loss": 2.5860564708709717,
      "step": 3503
    },
    {
      "epoch": 0.8105482303955587,
      "grad_norm": 0.9513399600982666,
      "learning_rate": 0.0004968072782793435,
      "loss": 2.4658360481262207,
      "step": 3504
    },
    {
      "epoch": 0.8107795512375665,
      "grad_norm": 0.8351470828056335,
      "learning_rate": 0.0004966608330212198,
      "loss": 2.3648314476013184,
      "step": 3505
    },
    {
      "epoch": 0.8110108720795743,
      "grad_norm": 0.8730280995368958,
      "learning_rate": 0.0004965111266040798,
      "loss": 2.3630447387695312,
      "step": 3506
    },
    {
      "epoch": 0.8112421929215823,
      "grad_norm": 1.102341651916504,
      "learning_rate": 0.0004963581610072262,
      "loss": 2.5081992149353027,
      "step": 3507
    },
    {
      "epoch": 0.8114735137635901,
      "grad_norm": 1.0641560554504395,
      "learning_rate": 0.000496201938253052,
      "loss": 2.704766273498535,
      "step": 3508
    },
    {
      "epoch": 0.811704834605598,
      "grad_norm": 1.1265352964401245,
      "learning_rate": 0.0004960424604070139,
      "loss": 2.581228017807007,
      "step": 3509
    },
    {
      "epoch": 0.8119361554476058,
      "grad_norm": 0.857427716255188,
      "learning_rate": 0.0004958797295776045,
      "loss": 2.7209343910217285,
      "step": 3510
    },
    {
      "epoch": 0.8121674762896137,
      "grad_norm": 0.9688679575920105,
      "learning_rate": 0.0004957137479163253,
      "loss": 2.4184393882751465,
      "step": 3511
    },
    {
      "epoch": 0.8123987971316216,
      "grad_norm": 1.1740288734436035,
      "learning_rate": 0.0004955445176176577,
      "loss": 2.561614513397217,
      "step": 3512
    },
    {
      "epoch": 0.8126301179736294,
      "grad_norm": 0.9964715242385864,
      "learning_rate": 0.0004953720409190339,
      "loss": 2.6990113258361816,
      "step": 3513
    },
    {
      "epoch": 0.8128614388156373,
      "grad_norm": 0.9284456372261047,
      "learning_rate": 0.0004951963201008077,
      "loss": 2.4783098697662354,
      "step": 3514
    },
    {
      "epoch": 0.8130927596576452,
      "grad_norm": 0.940207839012146,
      "learning_rate": 0.0004950173574862238,
      "loss": 2.5323641300201416,
      "step": 3515
    },
    {
      "epoch": 0.813324080499653,
      "grad_norm": 0.9090897440910339,
      "learning_rate": 0.000494835155441388,
      "loss": 2.6985340118408203,
      "step": 3516
    },
    {
      "epoch": 0.8135554013416609,
      "grad_norm": 1.1738969087600708,
      "learning_rate": 0.0004946497163752346,
      "loss": 2.7783336639404297,
      "step": 3517
    },
    {
      "epoch": 0.8137867221836688,
      "grad_norm": 2.9503610134124756,
      "learning_rate": 0.000494461042739496,
      "loss": 2.973552703857422,
      "step": 3518
    },
    {
      "epoch": 0.8140180430256766,
      "grad_norm": 0.987637996673584,
      "learning_rate": 0.0004942691370286693,
      "loss": 2.6911377906799316,
      "step": 3519
    },
    {
      "epoch": 0.8142493638676844,
      "grad_norm": 1.0339192152023315,
      "learning_rate": 0.0004940740017799833,
      "loss": 2.863420248031616,
      "step": 3520
    },
    {
      "epoch": 0.8144806847096924,
      "grad_norm": 1.0776867866516113,
      "learning_rate": 0.0004938756395733659,
      "loss": 2.394958257675171,
      "step": 3521
    },
    {
      "epoch": 0.8147120055517002,
      "grad_norm": 1.2939170598983765,
      "learning_rate": 0.0004936740530314087,
      "loss": 2.654254198074341,
      "step": 3522
    },
    {
      "epoch": 0.814943326393708,
      "grad_norm": 0.8378993272781372,
      "learning_rate": 0.0004934692448193334,
      "loss": 2.5579636096954346,
      "step": 3523
    },
    {
      "epoch": 0.815174647235716,
      "grad_norm": 1.182410478591919,
      "learning_rate": 0.0004932612176449559,
      "loss": 2.499427080154419,
      "step": 3524
    },
    {
      "epoch": 0.8154059680777238,
      "grad_norm": 0.8447617292404175,
      "learning_rate": 0.000493049974258651,
      "loss": 2.405853748321533,
      "step": 3525
    },
    {
      "epoch": 0.8156372889197316,
      "grad_norm": 5.961202621459961,
      "learning_rate": 0.0004928355174533154,
      "loss": 2.268934965133667,
      "step": 3526
    },
    {
      "epoch": 0.8158686097617396,
      "grad_norm": 1.2925606966018677,
      "learning_rate": 0.0004926178500643313,
      "loss": 2.7069482803344727,
      "step": 3527
    },
    {
      "epoch": 0.8160999306037474,
      "grad_norm": 0.790054976940155,
      "learning_rate": 0.000492396974969529,
      "loss": 2.365891456604004,
      "step": 3528
    },
    {
      "epoch": 0.8163312514457552,
      "grad_norm": 1.158236026763916,
      "learning_rate": 0.0004921728950891483,
      "loss": 2.290088176727295,
      "step": 3529
    },
    {
      "epoch": 0.8165625722877631,
      "grad_norm": 1.548613429069519,
      "learning_rate": 0.0004919456133858002,
      "loss": 2.7725071907043457,
      "step": 3530
    },
    {
      "epoch": 0.816793893129771,
      "grad_norm": 0.996149480342865,
      "learning_rate": 0.000491715132864428,
      "loss": 2.749208927154541,
      "step": 3531
    },
    {
      "epoch": 0.8170252139717789,
      "grad_norm": 1.149382472038269,
      "learning_rate": 0.0004914814565722671,
      "loss": 2.5880441665649414,
      "step": 3532
    },
    {
      "epoch": 0.8172565348137867,
      "grad_norm": 1.065768837928772,
      "learning_rate": 0.0004912445875988048,
      "loss": 2.3160057067871094,
      "step": 3533
    },
    {
      "epoch": 0.8174878556557946,
      "grad_norm": 0.8751046061515808,
      "learning_rate": 0.0004910045290757399,
      "loss": 2.3326728343963623,
      "step": 3534
    },
    {
      "epoch": 0.8177191764978025,
      "grad_norm": 0.7766723036766052,
      "learning_rate": 0.0004907612841769407,
      "loss": 2.514979839324951,
      "step": 3535
    },
    {
      "epoch": 0.8179504973398103,
      "grad_norm": 0.9003658890724182,
      "learning_rate": 0.0004905148561184033,
      "loss": 2.0840260982513428,
      "step": 3536
    },
    {
      "epoch": 0.8181818181818182,
      "grad_norm": 0.8937092423439026,
      "learning_rate": 0.0004902652481582093,
      "loss": 2.9980320930480957,
      "step": 3537
    },
    {
      "epoch": 0.8184131390238261,
      "grad_norm": 0.8817230463027954,
      "learning_rate": 0.0004900124635964822,
      "loss": 2.6815457344055176,
      "step": 3538
    },
    {
      "epoch": 0.8186444598658339,
      "grad_norm": 1.2548199892044067,
      "learning_rate": 0.0004897565057753443,
      "loss": 2.6775379180908203,
      "step": 3539
    },
    {
      "epoch": 0.8188757807078417,
      "grad_norm": 1.2051079273223877,
      "learning_rate": 0.0004894973780788722,
      "loss": 2.3409504890441895,
      "step": 3540
    },
    {
      "epoch": 0.8191071015498497,
      "grad_norm": 0.8308821320533752,
      "learning_rate": 0.0004892350839330522,
      "loss": 2.559457540512085,
      "step": 3541
    },
    {
      "epoch": 0.8193384223918575,
      "grad_norm": 0.9619821310043335,
      "learning_rate": 0.0004889696268057348,
      "loss": 2.6155500411987305,
      "step": 3542
    },
    {
      "epoch": 0.8195697432338653,
      "grad_norm": 0.8715582489967346,
      "learning_rate": 0.0004887010102065893,
      "loss": 2.1600942611694336,
      "step": 3543
    },
    {
      "epoch": 0.8198010640758733,
      "grad_norm": 1.028280258178711,
      "learning_rate": 0.0004884292376870567,
      "loss": 2.0975568294525146,
      "step": 3544
    },
    {
      "epoch": 0.8200323849178811,
      "grad_norm": 1.0315454006195068,
      "learning_rate": 0.00048815431284030344,
      "loss": 2.9563493728637695,
      "step": 3545
    },
    {
      "epoch": 0.8202637057598889,
      "grad_norm": 0.9491914510726929,
      "learning_rate": 0.00048787623930117343,
      "loss": 2.713108539581299,
      "step": 3546
    },
    {
      "epoch": 0.8204950266018969,
      "grad_norm": 0.782070517539978,
      "learning_rate": 0.0004875950207461403,
      "loss": 2.6905150413513184,
      "step": 3547
    },
    {
      "epoch": 0.8207263474439047,
      "grad_norm": 0.844447910785675,
      "learning_rate": 0.0004873106608932585,
      "loss": 2.3384008407592773,
      "step": 3548
    },
    {
      "epoch": 0.8209576682859125,
      "grad_norm": 1.450392723083496,
      "learning_rate": 0.0004870231635021144,
      "loss": 2.327694892883301,
      "step": 3549
    },
    {
      "epoch": 0.8211889891279204,
      "grad_norm": 0.9284141659736633,
      "learning_rate": 0.00048673253237377644,
      "loss": 2.09713077545166,
      "step": 3550
    },
    {
      "epoch": 0.8214203099699283,
      "grad_norm": 1.2618403434753418,
      "learning_rate": 0.0004864387713507449,
      "loss": 2.6832337379455566,
      "step": 3551
    },
    {
      "epoch": 0.8216516308119362,
      "grad_norm": 1.0145411491394043,
      "learning_rate": 0.0004861418843169012,
      "loss": 2.352186679840088,
      "step": 3552
    },
    {
      "epoch": 0.821882951653944,
      "grad_norm": 1.1508512496948242,
      "learning_rate": 0.0004858418751974563,
      "loss": 2.974243640899658,
      "step": 3553
    },
    {
      "epoch": 0.8221142724959519,
      "grad_norm": 1.0078539848327637,
      "learning_rate": 0.0004855387479588991,
      "loss": 2.548313617706299,
      "step": 3554
    },
    {
      "epoch": 0.8223455933379598,
      "grad_norm": 1.0177537202835083,
      "learning_rate": 0.00048523250660894386,
      "loss": 2.4855620861053467,
      "step": 3555
    },
    {
      "epoch": 0.8225769141799676,
      "grad_norm": 1.005082607269287,
      "learning_rate": 0.0004849231551964771,
      "loss": 2.7274577617645264,
      "step": 3556
    },
    {
      "epoch": 0.8228082350219755,
      "grad_norm": 1.4953197240829468,
      "learning_rate": 0.0004846106978115043,
      "loss": 2.3017823696136475,
      "step": 3557
    },
    {
      "epoch": 0.8230395558639834,
      "grad_norm": 1.0145642757415771,
      "learning_rate": 0.0004842951385850956,
      "loss": 2.39874005317688,
      "step": 3558
    },
    {
      "epoch": 0.8232708767059912,
      "grad_norm": 1.2464759349822998,
      "learning_rate": 0.00048397648168933144,
      "loss": 2.5113725662231445,
      "step": 3559
    },
    {
      "epoch": 0.823502197547999,
      "grad_norm": 1.2609920501708984,
      "learning_rate": 0.00048365473133724714,
      "loss": 2.6854982376098633,
      "step": 3560
    },
    {
      "epoch": 0.823733518390007,
      "grad_norm": 1.1913224458694458,
      "learning_rate": 0.0004833298917827773,
      "loss": 2.741799831390381,
      "step": 3561
    },
    {
      "epoch": 0.8239648392320148,
      "grad_norm": 0.9861986041069031,
      "learning_rate": 0.00048300196732069966,
      "loss": 2.0674681663513184,
      "step": 3562
    },
    {
      "epoch": 0.8241961600740226,
      "grad_norm": 0.9248791933059692,
      "learning_rate": 0.00048267096228657815,
      "loss": 2.8137378692626953,
      "step": 3563
    },
    {
      "epoch": 0.8244274809160306,
      "grad_norm": 0.9183050990104675,
      "learning_rate": 0.0004823368810567056,
      "loss": 2.3407399654388428,
      "step": 3564
    },
    {
      "epoch": 0.8246588017580384,
      "grad_norm": 1.5839911699295044,
      "learning_rate": 0.00048199972804804616,
      "loss": 2.6798086166381836,
      "step": 3565
    },
    {
      "epoch": 0.8248901226000462,
      "grad_norm": 1.0041327476501465,
      "learning_rate": 0.00048165950771817637,
      "loss": 3.0000534057617188,
      "step": 3566
    },
    {
      "epoch": 0.8251214434420542,
      "grad_norm": 1.0411717891693115,
      "learning_rate": 0.00048131622456522663,
      "loss": 2.36765193939209,
      "step": 3567
    },
    {
      "epoch": 0.825352764284062,
      "grad_norm": 1.2804536819458008,
      "learning_rate": 0.0004809698831278217,
      "loss": 2.5939176082611084,
      "step": 3568
    },
    {
      "epoch": 0.8255840851260698,
      "grad_norm": 1.060969591140747,
      "learning_rate": 0.00048062048798502057,
      "loss": 3.3311119079589844,
      "step": 3569
    },
    {
      "epoch": 0.8258154059680777,
      "grad_norm": 1.2169747352600098,
      "learning_rate": 0.0004802680437562559,
      "loss": 2.795281410217285,
      "step": 3570
    },
    {
      "epoch": 0.8260467268100856,
      "grad_norm": 1.1788980960845947,
      "learning_rate": 0.0004799125551012731,
      "loss": 2.4407739639282227,
      "step": 3571
    },
    {
      "epoch": 0.8262780476520935,
      "grad_norm": 1.1311455965042114,
      "learning_rate": 0.0004795540267200685,
      "loss": 2.531867504119873,
      "step": 3572
    },
    {
      "epoch": 0.8265093684941013,
      "grad_norm": 1.0104162693023682,
      "learning_rate": 0.00047919246335282754,
      "loss": 2.625241756439209,
      "step": 3573
    },
    {
      "epoch": 0.8267406893361092,
      "grad_norm": 0.9265028834342957,
      "learning_rate": 0.0004788278697798618,
      "loss": 2.40472149848938,
      "step": 3574
    },
    {
      "epoch": 0.8269720101781171,
      "grad_norm": 1.2124131917953491,
      "learning_rate": 0.0004784602508215458,
      "loss": 2.7148823738098145,
      "step": 3575
    },
    {
      "epoch": 0.8272033310201249,
      "grad_norm": 0.8707674741744995,
      "learning_rate": 0.0004780896113382536,
      "loss": 2.4259960651397705,
      "step": 3576
    },
    {
      "epoch": 0.8274346518621328,
      "grad_norm": 1.0462234020233154,
      "learning_rate": 0.000477715956230294,
      "loss": 2.7043063640594482,
      "step": 3577
    },
    {
      "epoch": 0.8276659727041407,
      "grad_norm": 1.0669950246810913,
      "learning_rate": 0.0004773392904378463,
      "loss": 2.7767462730407715,
      "step": 3578
    },
    {
      "epoch": 0.8278972935461485,
      "grad_norm": 0.9099299311637878,
      "learning_rate": 0.0004769596189408947,
      "loss": 2.483011245727539,
      "step": 3579
    },
    {
      "epoch": 0.8281286143881563,
      "grad_norm": 0.8521617650985718,
      "learning_rate": 0.00047657694675916254,
      "loss": 2.8489603996276855,
      "step": 3580
    },
    {
      "epoch": 0.8283599352301643,
      "grad_norm": 1.0129637718200684,
      "learning_rate": 0.00047619127895204565,
      "loss": 2.5119402408599854,
      "step": 3581
    },
    {
      "epoch": 0.8285912560721721,
      "grad_norm": 0.9819126129150391,
      "learning_rate": 0.0004758026206185461,
      "loss": 2.4058265686035156,
      "step": 3582
    },
    {
      "epoch": 0.8288225769141799,
      "grad_norm": 1.5989183187484741,
      "learning_rate": 0.00047541097689720403,
      "loss": 2.8734428882598877,
      "step": 3583
    },
    {
      "epoch": 0.8290538977561879,
      "grad_norm": 0.9106466770172119,
      "learning_rate": 0.0004750163529660303,
      "loss": 2.8457694053649902,
      "step": 3584
    },
    {
      "epoch": 0.8292852185981957,
      "grad_norm": 1.165107250213623,
      "learning_rate": 0.0004746187540424377,
      "loss": 2.794337272644043,
      "step": 3585
    },
    {
      "epoch": 0.8295165394402035,
      "grad_norm": 1.0736173391342163,
      "learning_rate": 0.0004742181853831721,
      "loss": 2.845648765563965,
      "step": 3586
    },
    {
      "epoch": 0.8297478602822115,
      "grad_norm": 0.9469738006591797,
      "learning_rate": 0.0004738146522842429,
      "loss": 2.735745906829834,
      "step": 3587
    },
    {
      "epoch": 0.8299791811242193,
      "grad_norm": 1.1303598880767822,
      "learning_rate": 0.00047340816008085306,
      "loss": 2.2397589683532715,
      "step": 3588
    },
    {
      "epoch": 0.8302105019662271,
      "grad_norm": 1.0027469396591187,
      "learning_rate": 0.00047299871414732854,
      "loss": 2.39047908782959,
      "step": 3589
    },
    {
      "epoch": 0.830441822808235,
      "grad_norm": 0.7516158223152161,
      "learning_rate": 0.00047258631989704727,
      "loss": 2.3732972145080566,
      "step": 3590
    },
    {
      "epoch": 0.8306731436502429,
      "grad_norm": 1.0947380065917969,
      "learning_rate": 0.00047217098278236737,
      "loss": 2.5613720417022705,
      "step": 3591
    },
    {
      "epoch": 0.8309044644922507,
      "grad_norm": 0.7692821025848389,
      "learning_rate": 0.0004717527082945554,
      "loss": 2.4054322242736816,
      "step": 3592
    },
    {
      "epoch": 0.8311357853342586,
      "grad_norm": 0.8873944282531738,
      "learning_rate": 0.00047133150196371367,
      "loss": 2.1720314025878906,
      "step": 3593
    },
    {
      "epoch": 0.8313671061762665,
      "grad_norm": 0.9295334219932556,
      "learning_rate": 0.00047090736935870693,
      "loss": 2.503126621246338,
      "step": 3594
    },
    {
      "epoch": 0.8315984270182744,
      "grad_norm": 0.9697920083999634,
      "learning_rate": 0.00047048031608708875,
      "loss": 2.6778409481048584,
      "step": 3595
    },
    {
      "epoch": 0.8318297478602822,
      "grad_norm": 0.7536936402320862,
      "learning_rate": 0.00047005034779502776,
      "loss": 2.166992664337158,
      "step": 3596
    },
    {
      "epoch": 0.8320610687022901,
      "grad_norm": 0.9847148656845093,
      "learning_rate": 0.00046961747016723265,
      "loss": 2.434565544128418,
      "step": 3597
    },
    {
      "epoch": 0.832292389544298,
      "grad_norm": 0.8747671246528625,
      "learning_rate": 0.00046918168892687696,
      "loss": 2.490204095840454,
      "step": 3598
    },
    {
      "epoch": 0.8325237103863058,
      "grad_norm": 1.5819509029388428,
      "learning_rate": 0.00046874300983552365,
      "loss": 2.5966668128967285,
      "step": 3599
    },
    {
      "epoch": 0.8327550312283136,
      "grad_norm": 1.1442784070968628,
      "learning_rate": 0.00046830143869304897,
      "loss": 2.397045135498047,
      "step": 3600
    },
    {
      "epoch": 0.8329863520703216,
      "grad_norm": 1.102743148803711,
      "learning_rate": 0.00046785698133756535,
      "loss": 2.538593053817749,
      "step": 3601
    },
    {
      "epoch": 0.8332176729123294,
      "grad_norm": 0.7949190735816956,
      "learning_rate": 0.00046740964364534477,
      "loss": 2.449138641357422,
      "step": 3602
    },
    {
      "epoch": 0.8334489937543372,
      "grad_norm": 0.8765080571174622,
      "learning_rate": 0.0004669594315307406,
      "loss": 2.775240898132324,
      "step": 3603
    },
    {
      "epoch": 0.8336803145963452,
      "grad_norm": 1.2061457633972168,
      "learning_rate": 0.00046650635094610973,
      "loss": 2.4939589500427246,
      "step": 3604
    },
    {
      "epoch": 0.833911635438353,
      "grad_norm": 0.9006925225257874,
      "learning_rate": 0.00046605040788173357,
      "loss": 2.545285701751709,
      "step": 3605
    },
    {
      "epoch": 0.8341429562803608,
      "grad_norm": 0.8618360757827759,
      "learning_rate": 0.0004655916083657393,
      "loss": 2.648833751678467,
      "step": 3606
    },
    {
      "epoch": 0.8343742771223688,
      "grad_norm": 0.9049606919288635,
      "learning_rate": 0.00046512995846401974,
      "loss": 2.2042737007141113,
      "step": 3607
    },
    {
      "epoch": 0.8346055979643766,
      "grad_norm": 1.0607131719589233,
      "learning_rate": 0.00046466546428015334,
      "loss": 2.7517309188842773,
      "step": 3608
    },
    {
      "epoch": 0.8348369188063844,
      "grad_norm": 2.0893709659576416,
      "learning_rate": 0.0004641981319553232,
      "loss": 2.714242935180664,
      "step": 3609
    },
    {
      "epoch": 0.8350682396483923,
      "grad_norm": 0.9544740319252014,
      "learning_rate": 0.00046372796766823665,
      "loss": 2.6902456283569336,
      "step": 3610
    },
    {
      "epoch": 0.8352995604904002,
      "grad_norm": 1.0846905708312988,
      "learning_rate": 0.0004632549776350426,
      "loss": 2.424503803253174,
      "step": 3611
    },
    {
      "epoch": 0.835530881332408,
      "grad_norm": 8.299501419067383,
      "learning_rate": 0.0004627791681092499,
      "loss": 2.9855358600616455,
      "step": 3612
    },
    {
      "epoch": 0.8357622021744159,
      "grad_norm": 0.9658156037330627,
      "learning_rate": 0.00046230054538164475,
      "loss": 2.3955368995666504,
      "step": 3613
    },
    {
      "epoch": 0.8359935230164238,
      "grad_norm": 0.9494393467903137,
      "learning_rate": 0.00046181911578020696,
      "loss": 2.4422850608825684,
      "step": 3614
    },
    {
      "epoch": 0.8362248438584317,
      "grad_norm": 1.298826813697815,
      "learning_rate": 0.00046133488567002705,
      "loss": 2.6142306327819824,
      "step": 3615
    },
    {
      "epoch": 0.8364561647004395,
      "grad_norm": 0.8815666437149048,
      "learning_rate": 0.00046084786145322143,
      "loss": 2.4668126106262207,
      "step": 3616
    },
    {
      "epoch": 0.8366874855424474,
      "grad_norm": 0.7683745622634888,
      "learning_rate": 0.00046035804956884817,
      "loss": 2.429126024246216,
      "step": 3617
    },
    {
      "epoch": 0.8369188063844553,
      "grad_norm": 0.975843608379364,
      "learning_rate": 0.00045986545649282163,
      "loss": 2.6156036853790283,
      "step": 3618
    },
    {
      "epoch": 0.8371501272264631,
      "grad_norm": 0.929387092590332,
      "learning_rate": 0.000459370088737827,
      "loss": 2.537489414215088,
      "step": 3619
    },
    {
      "epoch": 0.8373814480684709,
      "grad_norm": 2.1226792335510254,
      "learning_rate": 0.0004588719528532341,
      "loss": 2.448575496673584,
      "step": 3620
    },
    {
      "epoch": 0.8376127689104789,
      "grad_norm": 0.7893987894058228,
      "learning_rate": 0.00045837105542501094,
      "loss": 2.5008621215820312,
      "step": 3621
    },
    {
      "epoch": 0.8378440897524867,
      "grad_norm": 1.0451778173446655,
      "learning_rate": 0.00045786740307563633,
      "loss": 2.6710634231567383,
      "step": 3622
    },
    {
      "epoch": 0.8380754105944945,
      "grad_norm": 1.5001004934310913,
      "learning_rate": 0.00045736100246401265,
      "loss": 2.873936176300049,
      "step": 3623
    },
    {
      "epoch": 0.8383067314365025,
      "grad_norm": 1.117192029953003,
      "learning_rate": 0.00045685186028537764,
      "loss": 2.2562103271484375,
      "step": 3624
    },
    {
      "epoch": 0.8385380522785103,
      "grad_norm": 1.0664620399475098,
      "learning_rate": 0.00045633998327121596,
      "loss": 2.5584185123443604,
      "step": 3625
    },
    {
      "epoch": 0.8387693731205181,
      "grad_norm": 1.0969069004058838,
      "learning_rate": 0.0004558253781891701,
      "loss": 2.3084137439727783,
      "step": 3626
    },
    {
      "epoch": 0.8390006939625261,
      "grad_norm": 1.1107275485992432,
      "learning_rate": 0.00045530805184295096,
      "loss": 2.198458671569824,
      "step": 3627
    },
    {
      "epoch": 0.8392320148045339,
      "grad_norm": 1.1034884452819824,
      "learning_rate": 0.00045478801107224796,
      "loss": 2.6697134971618652,
      "step": 3628
    },
    {
      "epoch": 0.8394633356465417,
      "grad_norm": 0.9088250398635864,
      "learning_rate": 0.00045426526275263847,
      "loss": 2.406587600708008,
      "step": 3629
    },
    {
      "epoch": 0.8396946564885496,
      "grad_norm": 0.8854896426200867,
      "learning_rate": 0.000453739813795497,
      "loss": 2.8114495277404785,
      "step": 3630
    },
    {
      "epoch": 0.8399259773305575,
      "grad_norm": 0.8974097371101379,
      "learning_rate": 0.0004532116711479038,
      "loss": 2.42242431640625,
      "step": 3631
    },
    {
      "epoch": 0.8401572981725653,
      "grad_norm": 2.3899712562561035,
      "learning_rate": 0.0004526808417925531,
      "loss": 2.500736951828003,
      "step": 3632
    },
    {
      "epoch": 0.8403886190145732,
      "grad_norm": 0.9626014232635498,
      "learning_rate": 0.0004521473327476605,
      "loss": 2.8103857040405273,
      "step": 3633
    },
    {
      "epoch": 0.8406199398565811,
      "grad_norm": 1.002421498298645,
      "learning_rate": 0.0004516111510668707,
      "loss": 2.4407596588134766,
      "step": 3634
    },
    {
      "epoch": 0.840851260698589,
      "grad_norm": 0.9323956370353699,
      "learning_rate": 0.0004510723038391635,
      "loss": 2.341843843460083,
      "step": 3635
    },
    {
      "epoch": 0.8410825815405968,
      "grad_norm": 0.8897451758384705,
      "learning_rate": 0.00045053079818876095,
      "loss": 2.942270040512085,
      "step": 3636
    },
    {
      "epoch": 0.8413139023826047,
      "grad_norm": 1.181993842124939,
      "learning_rate": 0.0004499866412750324,
      "loss": 2.234553575515747,
      "step": 3637
    },
    {
      "epoch": 0.8415452232246126,
      "grad_norm": 0.9087205529212952,
      "learning_rate": 0.0004494398402924004,
      "loss": 2.4317545890808105,
      "step": 3638
    },
    {
      "epoch": 0.8417765440666204,
      "grad_norm": 1.0473666191101074,
      "learning_rate": 0.000448890402470245,
      "loss": 2.623710870742798,
      "step": 3639
    },
    {
      "epoch": 0.8420078649086282,
      "grad_norm": 0.8455772399902344,
      "learning_rate": 0.0004483383350728088,
      "loss": 2.501286506652832,
      "step": 3640
    },
    {
      "epoch": 0.8422391857506362,
      "grad_norm": 0.8175027370452881,
      "learning_rate": 0.00044778364539910053,
      "loss": 2.318824052810669,
      "step": 3641
    },
    {
      "epoch": 0.842470506592644,
      "grad_norm": 0.8017075061798096,
      "learning_rate": 0.0004472263407827987,
      "loss": 2.4766626358032227,
      "step": 3642
    },
    {
      "epoch": 0.8427018274346518,
      "grad_norm": 3.62095308303833,
      "learning_rate": 0.00044666642859215425,
      "loss": 2.4265589714050293,
      "step": 3643
    },
    {
      "epoch": 0.8429331482766598,
      "grad_norm": 1.0042214393615723,
      "learning_rate": 0.0004461039162298939,
      "loss": 2.6406335830688477,
      "step": 3644
    },
    {
      "epoch": 0.8431644691186676,
      "grad_norm": 1.037658452987671,
      "learning_rate": 0.00044553881113312165,
      "loss": 2.7273433208465576,
      "step": 3645
    },
    {
      "epoch": 0.8433957899606754,
      "grad_norm": 1.0473436117172241,
      "learning_rate": 0.0004449711207732204,
      "loss": 2.9821696281433105,
      "step": 3646
    },
    {
      "epoch": 0.8436271108026833,
      "grad_norm": 1.0844225883483887,
      "learning_rate": 0.0004444008526557539,
      "loss": 2.7583236694335938,
      "step": 3647
    },
    {
      "epoch": 0.8438584316446912,
      "grad_norm": 0.8221487998962402,
      "learning_rate": 0.00044382801432036646,
      "loss": 2.436981678009033,
      "step": 3648
    },
    {
      "epoch": 0.844089752486699,
      "grad_norm": 0.9718090891838074,
      "learning_rate": 0.0004432526133406842,
      "loss": 2.892427444458008,
      "step": 3649
    },
    {
      "epoch": 0.8443210733287069,
      "grad_norm": 1.4123058319091797,
      "learning_rate": 0.00044267465732421453,
      "loss": 2.3433115482330322,
      "step": 3650
    },
    {
      "epoch": 0.8445523941707148,
      "grad_norm": 0.9683862328529358,
      "learning_rate": 0.0004420941539122454,
      "loss": 2.5762522220611572,
      "step": 3651
    },
    {
      "epoch": 0.8447837150127226,
      "grad_norm": 1.1467581987380981,
      "learning_rate": 0.0004415111107797445,
      "loss": 2.5762147903442383,
      "step": 3652
    },
    {
      "epoch": 0.8450150358547305,
      "grad_norm": 1.2762067317962646,
      "learning_rate": 0.00044092553563525793,
      "loss": 2.193880796432495,
      "step": 3653
    },
    {
      "epoch": 0.8452463566967384,
      "grad_norm": 0.9954598546028137,
      "learning_rate": 0.0004403374362208078,
      "loss": 2.666764497756958,
      "step": 3654
    },
    {
      "epoch": 0.8454776775387463,
      "grad_norm": 0.9621124267578125,
      "learning_rate": 0.0004397468203117905,
      "loss": 2.463890790939331,
      "step": 3655
    },
    {
      "epoch": 0.8457089983807541,
      "grad_norm": 0.8435832858085632,
      "learning_rate": 0.00043915369571687323,
      "loss": 2.5396716594696045,
      "step": 3656
    },
    {
      "epoch": 0.8459403192227619,
      "grad_norm": 1.1270368099212646,
      "learning_rate": 0.00043855807027789133,
      "loss": 2.4958407878875732,
      "step": 3657
    },
    {
      "epoch": 0.8461716400647699,
      "grad_norm": 1.019095778465271,
      "learning_rate": 0.00043795995186974435,
      "loss": 2.858168363571167,
      "step": 3658
    },
    {
      "epoch": 0.8464029609067777,
      "grad_norm": 1.2456121444702148,
      "learning_rate": 0.00043735934840029186,
      "loss": 2.538691520690918,
      "step": 3659
    },
    {
      "epoch": 0.8466342817487855,
      "grad_norm": 1.1929168701171875,
      "learning_rate": 0.000436756267810249,
      "loss": 2.6175832748413086,
      "step": 3660
    },
    {
      "epoch": 0.8468656025907935,
      "grad_norm": 1.4193881750106812,
      "learning_rate": 0.0004361507180730816,
      "loss": 2.396336078643799,
      "step": 3661
    },
    {
      "epoch": 0.8470969234328013,
      "grad_norm": 0.79234778881073,
      "learning_rate": 0.00043554270719490043,
      "loss": 2.5682902336120605,
      "step": 3662
    },
    {
      "epoch": 0.8473282442748091,
      "grad_norm": 0.8695095777511597,
      "learning_rate": 0.0004349322432143559,
      "loss": 2.5138204097747803,
      "step": 3663
    },
    {
      "epoch": 0.8475595651168171,
      "grad_norm": 0.9651882648468018,
      "learning_rate": 0.000434319334202531,
      "loss": 2.523221015930176,
      "step": 3664
    },
    {
      "epoch": 0.8477908859588249,
      "grad_norm": 0.9173825979232788,
      "learning_rate": 0.0004337039882628355,
      "loss": 2.4551029205322266,
      "step": 3665
    },
    {
      "epoch": 0.8480222068008327,
      "grad_norm": 0.8931915760040283,
      "learning_rate": 0.00043308621353089805,
      "loss": 2.697033643722534,
      "step": 3666
    },
    {
      "epoch": 0.8482535276428406,
      "grad_norm": 0.9726338386535645,
      "learning_rate": 0.0004324660181744589,
      "loss": 2.4450478553771973,
      "step": 3667
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8465873003005981,
      "learning_rate": 0.0004318434103932622,
      "loss": 2.7638397216796875,
      "step": 3668
    },
    {
      "epoch": 0.8487161693268563,
      "grad_norm": 0.9197866320610046,
      "learning_rate": 0.000431218398418947,
      "loss": 2.6620895862579346,
      "step": 3669
    },
    {
      "epoch": 0.8489474901688642,
      "grad_norm": 0.93086177110672,
      "learning_rate": 0.0004305909905149389,
      "loss": 2.3699026107788086,
      "step": 3670
    },
    {
      "epoch": 0.8491788110108721,
      "grad_norm": 1.0146310329437256,
      "learning_rate": 0.00042996119497634065,
      "loss": 2.420743942260742,
      "step": 3671
    },
    {
      "epoch": 0.84941013185288,
      "grad_norm": 0.9336373805999756,
      "learning_rate": 0.00042932902012982234,
      "loss": 3.054210662841797,
      "step": 3672
    },
    {
      "epoch": 0.8496414526948878,
      "grad_norm": 0.8738887906074524,
      "learning_rate": 0.0004286944743335116,
      "loss": 2.453946352005005,
      "step": 3673
    },
    {
      "epoch": 0.8498727735368957,
      "grad_norm": 1.0782744884490967,
      "learning_rate": 0.0004280575659768828,
      "loss": 2.49591064453125,
      "step": 3674
    },
    {
      "epoch": 0.8501040943789036,
      "grad_norm": 1.1052106618881226,
      "learning_rate": 0.00042741830348064626,
      "loss": 2.6251533031463623,
      "step": 3675
    },
    {
      "epoch": 0.8503354152209114,
      "grad_norm": 1.0245698690414429,
      "learning_rate": 0.00042677669529663686,
      "loss": 2.395023822784424,
      "step": 3676
    },
    {
      "epoch": 0.8505667360629192,
      "grad_norm": 1.0598498582839966,
      "learning_rate": 0.0004261327499077027,
      "loss": 2.4272427558898926,
      "step": 3677
    },
    {
      "epoch": 0.8507980569049272,
      "grad_norm": 0.9731510877609253,
      "learning_rate": 0.00042548647582759205,
      "loss": 2.657806873321533,
      "step": 3678
    },
    {
      "epoch": 0.851029377746935,
      "grad_norm": 0.7835236191749573,
      "learning_rate": 0.0004248378816008418,
      "loss": 2.4201016426086426,
      "step": 3679
    },
    {
      "epoch": 0.8512606985889428,
      "grad_norm": 0.8556360006332397,
      "learning_rate": 0.0004241869758026637,
      "loss": 2.4634861946105957,
      "step": 3680
    },
    {
      "epoch": 0.8514920194309508,
      "grad_norm": 0.8850479125976562,
      "learning_rate": 0.00042353376703883173,
      "loss": 2.541808605194092,
      "step": 3681
    },
    {
      "epoch": 0.8517233402729586,
      "grad_norm": 0.9373106360435486,
      "learning_rate": 0.0004228782639455674,
      "loss": 2.353647232055664,
      "step": 3682
    },
    {
      "epoch": 0.8519546611149664,
      "grad_norm": 0.8370017409324646,
      "learning_rate": 0.0004222204751894263,
      "loss": 2.5604794025421143,
      "step": 3683
    },
    {
      "epoch": 0.8521859819569744,
      "grad_norm": 0.7752666473388672,
      "learning_rate": 0.00042156040946718344,
      "loss": 2.5041065216064453,
      "step": 3684
    },
    {
      "epoch": 0.8524173027989822,
      "grad_norm": 0.9624239802360535,
      "learning_rate": 0.0004208980755057178,
      "loss": 2.321451187133789,
      "step": 3685
    },
    {
      "epoch": 0.85264862364099,
      "grad_norm": 0.875917375087738,
      "learning_rate": 0.0004202334820618976,
      "loss": 2.454277992248535,
      "step": 3686
    },
    {
      "epoch": 0.8528799444829979,
      "grad_norm": 0.8741204738616943,
      "learning_rate": 0.000419566637922464,
      "loss": 2.334925889968872,
      "step": 3687
    },
    {
      "epoch": 0.8531112653250058,
      "grad_norm": 0.7882574796676636,
      "learning_rate": 0.0004188975519039151,
      "loss": 2.4653427600860596,
      "step": 3688
    },
    {
      "epoch": 0.8533425861670136,
      "grad_norm": 0.7438105940818787,
      "learning_rate": 0.00041822623285238945,
      "loss": 2.203929901123047,
      "step": 3689
    },
    {
      "epoch": 0.8535739070090215,
      "grad_norm": 1.1253949403762817,
      "learning_rate": 0.000417552689643549,
      "loss": 2.3729605674743652,
      "step": 3690
    },
    {
      "epoch": 0.8538052278510294,
      "grad_norm": 0.7312648892402649,
      "learning_rate": 0.0004168769311824618,
      "loss": 2.487443447113037,
      "step": 3691
    },
    {
      "epoch": 0.8540365486930372,
      "grad_norm": 1.1572295427322388,
      "learning_rate": 0.0004161989664034844,
      "loss": 2.5880088806152344,
      "step": 3692
    },
    {
      "epoch": 0.8542678695350451,
      "grad_norm": 1.028674602508545,
      "learning_rate": 0.00041551880427014346,
      "loss": 2.486123561859131,
      "step": 3693
    },
    {
      "epoch": 0.854499190377053,
      "grad_norm": 1.4591281414031982,
      "learning_rate": 0.0004148364537750172,
      "loss": 2.678732395172119,
      "step": 3694
    },
    {
      "epoch": 0.8547305112190609,
      "grad_norm": 0.8032968044281006,
      "learning_rate": 0.000414151923939617,
      "loss": 2.749105930328369,
      "step": 3695
    },
    {
      "epoch": 0.8549618320610687,
      "grad_norm": 1.0870730876922607,
      "learning_rate": 0.0004134652238142674,
      "loss": 2.302546262741089,
      "step": 3696
    },
    {
      "epoch": 0.8551931529030765,
      "grad_norm": 0.8772634267807007,
      "learning_rate": 0.0004127763624779873,
      "loss": 2.44038724899292,
      "step": 3697
    },
    {
      "epoch": 0.8554244737450845,
      "grad_norm": 1.0423297882080078,
      "learning_rate": 0.00041208534903836903,
      "loss": 2.5590014457702637,
      "step": 3698
    },
    {
      "epoch": 0.8556557945870923,
      "grad_norm": 0.8407984375953674,
      "learning_rate": 0.0004113921926314587,
      "loss": 2.5986385345458984,
      "step": 3699
    },
    {
      "epoch": 0.8558871154291001,
      "grad_norm": 0.8307444453239441,
      "learning_rate": 0.0004106969024216348,
      "loss": 2.5112290382385254,
      "step": 3700
    },
    {
      "epoch": 0.8561184362711081,
      "grad_norm": 0.9448078870773315,
      "learning_rate": 0.00040999948760148764,
      "loss": 2.2137770652770996,
      "step": 3701
    },
    {
      "epoch": 0.8563497571131159,
      "grad_norm": 1.9151802062988281,
      "learning_rate": 0.00040929995739169713,
      "loss": 2.8413825035095215,
      "step": 3702
    },
    {
      "epoch": 0.8565810779551237,
      "grad_norm": 0.7700726985931396,
      "learning_rate": 0.0004085983210409114,
      "loss": 2.4562830924987793,
      "step": 3703
    },
    {
      "epoch": 0.8568123987971317,
      "grad_norm": 1.8557795286178589,
      "learning_rate": 0.0004078945878256244,
      "loss": 2.7001166343688965,
      "step": 3704
    },
    {
      "epoch": 0.8570437196391395,
      "grad_norm": 0.8856433033943176,
      "learning_rate": 0.00040718876705005315,
      "loss": 2.3863656520843506,
      "step": 3705
    },
    {
      "epoch": 0.8572750404811473,
      "grad_norm": 1.0283746719360352,
      "learning_rate": 0.0004064808680460148,
      "loss": 2.741550922393799,
      "step": 3706
    },
    {
      "epoch": 0.8575063613231552,
      "grad_norm": 0.8679888844490051,
      "learning_rate": 0.0004057709001728031,
      "loss": 2.559187889099121,
      "step": 3707
    },
    {
      "epoch": 0.8577376821651631,
      "grad_norm": 0.8954722285270691,
      "learning_rate": 0.000405058872817065,
      "loss": 2.395134210586548,
      "step": 3708
    },
    {
      "epoch": 0.8579690030071709,
      "grad_norm": 0.9501569867134094,
      "learning_rate": 0.00040434479539267626,
      "loss": 2.217223644256592,
      "step": 3709
    },
    {
      "epoch": 0.8582003238491788,
      "grad_norm": 0.9212095737457275,
      "learning_rate": 0.0004036286773406169,
      "loss": 2.57997465133667,
      "step": 3710
    },
    {
      "epoch": 0.8584316446911867,
      "grad_norm": 1.1460098028182983,
      "learning_rate": 0.0004029105281288469,
      "loss": 2.4588167667388916,
      "step": 3711
    },
    {
      "epoch": 0.8586629655331945,
      "grad_norm": 0.8100961446762085,
      "learning_rate": 0.0004021903572521802,
      "loss": 2.4284486770629883,
      "step": 3712
    },
    {
      "epoch": 0.8588942863752024,
      "grad_norm": 1.1507349014282227,
      "learning_rate": 0.00040146817423216,
      "loss": 2.613208532333374,
      "step": 3713
    },
    {
      "epoch": 0.8591256072172103,
      "grad_norm": 1.0991507768630981,
      "learning_rate": 0.0004007439886169324,
      "loss": 2.621856927871704,
      "step": 3714
    },
    {
      "epoch": 0.8593569280592182,
      "grad_norm": 0.8631712198257446,
      "learning_rate": 0.0004000178099811203,
      "loss": 2.3836746215820312,
      "step": 3715
    },
    {
      "epoch": 0.859588248901226,
      "grad_norm": 0.9858023524284363,
      "learning_rate": 0.00039928964792569654,
      "loss": 2.875988006591797,
      "step": 3716
    },
    {
      "epoch": 0.8598195697432338,
      "grad_norm": 0.873903751373291,
      "learning_rate": 0.0003985595120778576,
      "loss": 2.775620937347412,
      "step": 3717
    },
    {
      "epoch": 0.8600508905852418,
      "grad_norm": 1.031594157218933,
      "learning_rate": 0.0003978274120908956,
      "loss": 2.539936065673828,
      "step": 3718
    },
    {
      "epoch": 0.8602822114272496,
      "grad_norm": 1.0082811117172241,
      "learning_rate": 0.0003970933576440711,
      "loss": 2.512535810470581,
      "step": 3719
    },
    {
      "epoch": 0.8605135322692574,
      "grad_norm": 0.8367172479629517,
      "learning_rate": 0.0003963573584424852,
      "loss": 2.355748176574707,
      "step": 3720
    },
    {
      "epoch": 0.8607448531112654,
      "grad_norm": 0.8635507822036743,
      "learning_rate": 0.00039561942421695057,
      "loss": 2.3454103469848633,
      "step": 3721
    },
    {
      "epoch": 0.8609761739532732,
      "grad_norm": 0.8449104428291321,
      "learning_rate": 0.00039487956472386375,
      "loss": 2.468264579772949,
      "step": 3722
    },
    {
      "epoch": 0.861207494795281,
      "grad_norm": 0.9374300837516785,
      "learning_rate": 0.00039413778974507546,
      "loss": 2.430351734161377,
      "step": 3723
    },
    {
      "epoch": 0.861438815637289,
      "grad_norm": 1.059836745262146,
      "learning_rate": 0.00039339410908776154,
      "loss": 2.3951005935668945,
      "step": 3724
    },
    {
      "epoch": 0.8616701364792968,
      "grad_norm": 1.051961898803711,
      "learning_rate": 0.00039264853258429323,
      "loss": 2.5079076290130615,
      "step": 3725
    },
    {
      "epoch": 0.8619014573213046,
      "grad_norm": 1.2555360794067383,
      "learning_rate": 0.0003919010700921072,
      "loss": 2.523116111755371,
      "step": 3726
    },
    {
      "epoch": 0.8621327781633125,
      "grad_norm": 1.0143468379974365,
      "learning_rate": 0.00039115173149357514,
      "loss": 2.9256486892700195,
      "step": 3727
    },
    {
      "epoch": 0.8623640990053204,
      "grad_norm": 0.971011221408844,
      "learning_rate": 0.0003904005266958732,
      "loss": 2.730660915374756,
      "step": 3728
    },
    {
      "epoch": 0.8625954198473282,
      "grad_norm": 0.9409152269363403,
      "learning_rate": 0.0003896474656308511,
      "loss": 2.377281665802002,
      "step": 3729
    },
    {
      "epoch": 0.8628267406893361,
      "grad_norm": 5.429323673248291,
      "learning_rate": 0.00038889255825490053,
      "loss": 2.3390088081359863,
      "step": 3730
    },
    {
      "epoch": 0.863058061531344,
      "grad_norm": 1.1225627660751343,
      "learning_rate": 0.00038813581454882375,
      "loss": 2.661938190460205,
      "step": 3731
    },
    {
      "epoch": 0.8632893823733518,
      "grad_norm": 0.859813392162323,
      "learning_rate": 0.0003873772445177015,
      "loss": 2.8507707118988037,
      "step": 3732
    },
    {
      "epoch": 0.8635207032153597,
      "grad_norm": 0.9534983038902283,
      "learning_rate": 0.00038661685819076086,
      "loss": 2.5077786445617676,
      "step": 3733
    },
    {
      "epoch": 0.8637520240573676,
      "grad_norm": 0.8377417325973511,
      "learning_rate": 0.0003858546656212425,
      "loss": 2.4019808769226074,
      "step": 3734
    },
    {
      "epoch": 0.8639833448993754,
      "grad_norm": 1.3150831460952759,
      "learning_rate": 0.0003850906768862679,
      "loss": 2.303025722503662,
      "step": 3735
    },
    {
      "epoch": 0.8642146657413833,
      "grad_norm": 0.8869138360023499,
      "learning_rate": 0.000384324902086706,
      "loss": 2.128614902496338,
      "step": 3736
    },
    {
      "epoch": 0.8644459865833911,
      "grad_norm": 0.8640033602714539,
      "learning_rate": 0.00038355735134703965,
      "loss": 2.4441537857055664,
      "step": 3737
    },
    {
      "epoch": 0.864677307425399,
      "grad_norm": 1.1944787502288818,
      "learning_rate": 0.00038278803481523203,
      "loss": 2.2783889770507812,
      "step": 3738
    },
    {
      "epoch": 0.8649086282674069,
      "grad_norm": 0.9650765657424927,
      "learning_rate": 0.000382016962662592,
      "loss": 2.4927265644073486,
      "step": 3739
    },
    {
      "epoch": 0.8651399491094147,
      "grad_norm": 1.231825828552246,
      "learning_rate": 0.00038124414508364,
      "loss": 2.2456820011138916,
      "step": 3740
    },
    {
      "epoch": 0.8653712699514227,
      "grad_norm": 0.8528977036476135,
      "learning_rate": 0.0003804695922959733,
      "loss": 2.3564465045928955,
      "step": 3741
    },
    {
      "epoch": 0.8656025907934305,
      "grad_norm": 0.8083747625350952,
      "learning_rate": 0.0003796933145401304,
      "loss": 2.506671905517578,
      "step": 3742
    },
    {
      "epoch": 0.8658339116354383,
      "grad_norm": 1.0681536197662354,
      "learning_rate": 0.0003789153220794562,
      "loss": 2.5043740272521973,
      "step": 3743
    },
    {
      "epoch": 0.8660652324774463,
      "grad_norm": 1.0274262428283691,
      "learning_rate": 0.0003781356251999663,
      "loss": 2.403742790222168,
      "step": 3744
    },
    {
      "epoch": 0.8662965533194541,
      "grad_norm": 1.3771651983261108,
      "learning_rate": 0.00037735423421021046,
      "loss": 2.5545053482055664,
      "step": 3745
    },
    {
      "epoch": 0.8665278741614619,
      "grad_norm": 1.1591286659240723,
      "learning_rate": 0.00037657115944113686,
      "loss": 2.9029152393341064,
      "step": 3746
    },
    {
      "epoch": 0.8667591950034698,
      "grad_norm": 0.9638040661811829,
      "learning_rate": 0.0003757864112459551,
      "loss": 2.3273563385009766,
      "step": 3747
    },
    {
      "epoch": 0.8669905158454777,
      "grad_norm": 1.0437066555023193,
      "learning_rate": 0.000375,
      "loss": 2.4700887203216553,
      "step": 3748
    },
    {
      "epoch": 0.8672218366874855,
      "grad_norm": 0.9386389255523682,
      "learning_rate": 0.00037421193610059347,
      "loss": 2.4390640258789062,
      "step": 3749
    },
    {
      "epoch": 0.8674531575294934,
      "grad_norm": 1.0956292152404785,
      "learning_rate": 0.00037342222996690754,
      "loss": 2.2023634910583496,
      "step": 3750
    },
    {
      "epoch": 0.8676844783715013,
      "grad_norm": 1.0017096996307373,
      "learning_rate": 0.000372630892039827,
      "loss": 2.5441346168518066,
      "step": 3751
    },
    {
      "epoch": 0.8679157992135091,
      "grad_norm": 0.8916064500808716,
      "learning_rate": 0.0003718379327818106,
      "loss": 2.396477222442627,
      "step": 3752
    },
    {
      "epoch": 0.868147120055517,
      "grad_norm": 1.6392050981521606,
      "learning_rate": 0.0003710433626767532,
      "loss": 2.5097053050994873,
      "step": 3753
    },
    {
      "epoch": 0.8683784408975249,
      "grad_norm": 1.3312206268310547,
      "learning_rate": 0.00037024719222984693,
      "loss": 2.6524341106414795,
      "step": 3754
    },
    {
      "epoch": 0.8686097617395327,
      "grad_norm": 0.9194204807281494,
      "learning_rate": 0.00036944943196744255,
      "loss": 2.362989902496338,
      "step": 3755
    },
    {
      "epoch": 0.8688410825815406,
      "grad_norm": 0.7632510662078857,
      "learning_rate": 0.0003686500924369101,
      "loss": 2.34548020362854,
      "step": 3756
    },
    {
      "epoch": 0.8690724034235484,
      "grad_norm": 0.9508543610572815,
      "learning_rate": 0.0003678491842064995,
      "loss": 2.5779225826263428,
      "step": 3757
    },
    {
      "epoch": 0.8693037242655564,
      "grad_norm": 0.9595538973808289,
      "learning_rate": 0.0003670467178652005,
      "loss": 2.395709991455078,
      "step": 3758
    },
    {
      "epoch": 0.8695350451075642,
      "grad_norm": 1.0466526746749878,
      "learning_rate": 0.00036624270402260353,
      "loss": 2.254220485687256,
      "step": 3759
    },
    {
      "epoch": 0.869766365949572,
      "grad_norm": 1.0459014177322388,
      "learning_rate": 0.0003654371533087585,
      "loss": 2.068697452545166,
      "step": 3760
    },
    {
      "epoch": 0.86999768679158,
      "grad_norm": 1.5136209726333618,
      "learning_rate": 0.0003646300763740347,
      "loss": 2.200653076171875,
      "step": 3761
    },
    {
      "epoch": 0.8702290076335878,
      "grad_norm": 0.8227434158325195,
      "learning_rate": 0.0003638214838889801,
      "loss": 2.5853679180145264,
      "step": 3762
    },
    {
      "epoch": 0.8704603284755956,
      "grad_norm": 0.8022412061691284,
      "learning_rate": 0.00036301138654417996,
      "loss": 2.3597283363342285,
      "step": 3763
    },
    {
      "epoch": 0.8706916493176036,
      "grad_norm": 1.0858900547027588,
      "learning_rate": 0.0003621997950501156,
      "loss": 2.408733367919922,
      "step": 3764
    },
    {
      "epoch": 0.8709229701596114,
      "grad_norm": 0.7925516963005066,
      "learning_rate": 0.0003613867201370229,
      "loss": 2.260117769241333,
      "step": 3765
    },
    {
      "epoch": 0.8711542910016192,
      "grad_norm": 0.7915555238723755,
      "learning_rate": 0.00036057217255475036,
      "loss": 2.277925491333008,
      "step": 3766
    },
    {
      "epoch": 0.8713856118436271,
      "grad_norm": 1.1268216371536255,
      "learning_rate": 0.000359756163072617,
      "loss": 2.6272196769714355,
      "step": 3767
    },
    {
      "epoch": 0.871616932685635,
      "grad_norm": 0.8657358288764954,
      "learning_rate": 0.00035893870247926985,
      "loss": 2.5068769454956055,
      "step": 3768
    },
    {
      "epoch": 0.8718482535276428,
      "grad_norm": 0.8450140953063965,
      "learning_rate": 0.00035811980158254156,
      "loss": 2.5363306999206543,
      "step": 3769
    },
    {
      "epoch": 0.8720795743696507,
      "grad_norm": 1.09059739112854,
      "learning_rate": 0.0003572994712093073,
      "loss": 2.3347582817077637,
      "step": 3770
    },
    {
      "epoch": 0.8723108952116586,
      "grad_norm": 0.8012298941612244,
      "learning_rate": 0.0003564777222053415,
      "loss": 2.4356772899627686,
      "step": 3771
    },
    {
      "epoch": 0.8725422160536664,
      "grad_norm": 0.9399855732917786,
      "learning_rate": 0.00035565456543517487,
      "loss": 2.3217029571533203,
      "step": 3772
    },
    {
      "epoch": 0.8727735368956743,
      "grad_norm": 0.789210855960846,
      "learning_rate": 0.00035483001178195046,
      "loss": 2.162013530731201,
      "step": 3773
    },
    {
      "epoch": 0.8730048577376822,
      "grad_norm": 0.9742859601974487,
      "learning_rate": 0.0003540040721472798,
      "loss": 2.3903207778930664,
      "step": 3774
    },
    {
      "epoch": 0.87323617857969,
      "grad_norm": 1.0161411762237549,
      "learning_rate": 0.00035317675745109866,
      "loss": 2.7132930755615234,
      "step": 3775
    },
    {
      "epoch": 0.8734674994216979,
      "grad_norm": 0.8332478404045105,
      "learning_rate": 0.0003523480786315231,
      "loss": 2.540842056274414,
      "step": 3776
    },
    {
      "epoch": 0.8736988202637057,
      "grad_norm": 0.9036359786987305,
      "learning_rate": 0.0003515180466447044,
      "loss": 2.530019760131836,
      "step": 3777
    },
    {
      "epoch": 0.8739301411057137,
      "grad_norm": 0.7125075459480286,
      "learning_rate": 0.0003506866724646843,
      "loss": 2.605865955352783,
      "step": 3778
    },
    {
      "epoch": 0.8741614619477215,
      "grad_norm": 0.9379039406776428,
      "learning_rate": 0.00034985396708325014,
      "loss": 2.3725271224975586,
      "step": 3779
    },
    {
      "epoch": 0.8743927827897293,
      "grad_norm": 0.8775745630264282,
      "learning_rate": 0.00034901994150978924,
      "loss": 2.410059928894043,
      "step": 3780
    },
    {
      "epoch": 0.8746241036317373,
      "grad_norm": 0.8294207453727722,
      "learning_rate": 0.00034818460677114356,
      "loss": 2.6230413913726807,
      "step": 3781
    },
    {
      "epoch": 0.8748554244737451,
      "grad_norm": 0.8291412591934204,
      "learning_rate": 0.0003473479739114638,
      "loss": 2.4789061546325684,
      "step": 3782
    },
    {
      "epoch": 0.8750867453157529,
      "grad_norm": 1.1633306741714478,
      "learning_rate": 0.0003465100539920634,
      "loss": 2.467665910720825,
      "step": 3783
    },
    {
      "epoch": 0.8753180661577609,
      "grad_norm": 0.7697573304176331,
      "learning_rate": 0.0003456708580912725,
      "loss": 2.5940206050872803,
      "step": 3784
    },
    {
      "epoch": 0.8755493869997687,
      "grad_norm": 1.0546194314956665,
      "learning_rate": 0.0003448303973042909,
      "loss": 2.4760661125183105,
      "step": 3785
    },
    {
      "epoch": 0.8757807078417765,
      "grad_norm": 0.9733278751373291,
      "learning_rate": 0.000343988682743042,
      "loss": 2.4215803146362305,
      "step": 3786
    },
    {
      "epoch": 0.8760120286837844,
      "grad_norm": 1.139406681060791,
      "learning_rate": 0.00034314572553602573,
      "loss": 2.6262295246124268,
      "step": 3787
    },
    {
      "epoch": 0.8762433495257923,
      "grad_norm": 1.0042310953140259,
      "learning_rate": 0.0003423015368281711,
      "loss": 2.4156076908111572,
      "step": 3788
    },
    {
      "epoch": 0.8764746703678001,
      "grad_norm": 1.2865127325057983,
      "learning_rate": 0.00034145612778068926,
      "loss": 2.608459949493408,
      "step": 3789
    },
    {
      "epoch": 0.876705991209808,
      "grad_norm": 1.2775205373764038,
      "learning_rate": 0.0003406095095709254,
      "loss": 2.690845489501953,
      "step": 3790
    },
    {
      "epoch": 0.8769373120518159,
      "grad_norm": 0.9678350687026978,
      "learning_rate": 0.00033976169339221187,
      "loss": 2.2474145889282227,
      "step": 3791
    },
    {
      "epoch": 0.8771686328938237,
      "grad_norm": 0.9585996270179749,
      "learning_rate": 0.0003389126904537192,
      "loss": 2.592952013015747,
      "step": 3792
    },
    {
      "epoch": 0.8773999537358316,
      "grad_norm": 1.016295075416565,
      "learning_rate": 0.0003380625119803084,
      "loss": 2.3259692192077637,
      "step": 3793
    },
    {
      "epoch": 0.8776312745778395,
      "grad_norm": 0.9069329500198364,
      "learning_rate": 0.00033721116921238273,
      "loss": 2.4505021572113037,
      "step": 3794
    },
    {
      "epoch": 0.8778625954198473,
      "grad_norm": 1.0648678541183472,
      "learning_rate": 0.00033635867340573865,
      "loss": 2.206171989440918,
      "step": 3795
    },
    {
      "epoch": 0.8780939162618552,
      "grad_norm": 1.0603023767471313,
      "learning_rate": 0.0003355050358314172,
      "loss": 2.6930675506591797,
      "step": 3796
    },
    {
      "epoch": 0.878325237103863,
      "grad_norm": 0.9171197414398193,
      "learning_rate": 0.0003346502677755551,
      "loss": 2.762312650680542,
      "step": 3797
    },
    {
      "epoch": 0.878556557945871,
      "grad_norm": 0.8617503643035889,
      "learning_rate": 0.00033379438053923537,
      "loss": 2.5543768405914307,
      "step": 3798
    },
    {
      "epoch": 0.8787878787878788,
      "grad_norm": 0.8791247010231018,
      "learning_rate": 0.0003329373854383381,
      "loss": 2.7155497074127197,
      "step": 3799
    },
    {
      "epoch": 0.8790191996298866,
      "grad_norm": 0.8833146691322327,
      "learning_rate": 0.0003320792938033903,
      "loss": 2.322209596633911,
      "step": 3800
    },
    {
      "epoch": 0.8792505204718946,
      "grad_norm": 0.8770341873168945,
      "learning_rate": 0.00033122011697941695,
      "loss": 2.5289530754089355,
      "step": 3801
    },
    {
      "epoch": 0.8794818413139024,
      "grad_norm": 0.961706817150116,
      "learning_rate": 0.0003303598663257904,
      "loss": 2.5156657695770264,
      "step": 3802
    },
    {
      "epoch": 0.8797131621559102,
      "grad_norm": 1.2916265726089478,
      "learning_rate": 0.0003294985532160804,
      "loss": 2.354957103729248,
      "step": 3803
    },
    {
      "epoch": 0.8799444829979182,
      "grad_norm": 0.9525067210197449,
      "learning_rate": 0.00032863618903790343,
      "loss": 2.2513041496276855,
      "step": 3804
    },
    {
      "epoch": 0.880175803839926,
      "grad_norm": 0.8667315244674683,
      "learning_rate": 0.0003277727851927727,
      "loss": 2.2790849208831787,
      "step": 3805
    },
    {
      "epoch": 0.8804071246819338,
      "grad_norm": 1.1508957147598267,
      "learning_rate": 0.00032690835309594706,
      "loss": 2.466796875,
      "step": 3806
    },
    {
      "epoch": 0.8806384455239417,
      "grad_norm": 0.7677339315414429,
      "learning_rate": 0.00032604290417627993,
      "loss": 2.438441276550293,
      "step": 3807
    },
    {
      "epoch": 0.8808697663659496,
      "grad_norm": 1.362183928489685,
      "learning_rate": 0.00032517644987606825,
      "loss": 3.045461654663086,
      "step": 3808
    },
    {
      "epoch": 0.8811010872079574,
      "grad_norm": 0.9693242311477661,
      "learning_rate": 0.0003243090016509019,
      "loss": 2.410078763961792,
      "step": 3809
    },
    {
      "epoch": 0.8813324080499653,
      "grad_norm": 1.085180640220642,
      "learning_rate": 0.00032344057096951107,
      "loss": 2.3800477981567383,
      "step": 3810
    },
    {
      "epoch": 0.8815637288919732,
      "grad_norm": 0.815904438495636,
      "learning_rate": 0.00032257116931361555,
      "loss": 2.402189254760742,
      "step": 3811
    },
    {
      "epoch": 0.881795049733981,
      "grad_norm": 1.22720205783844,
      "learning_rate": 0.0003217008081777726,
      "loss": 2.5952649116516113,
      "step": 3812
    },
    {
      "epoch": 0.8820263705759889,
      "grad_norm": 1.1446385383605957,
      "learning_rate": 0.0003208294990692247,
      "loss": 2.427324056625366,
      "step": 3813
    },
    {
      "epoch": 0.8822576914179968,
      "grad_norm": 0.9047945141792297,
      "learning_rate": 0.00031995725350774805,
      "loss": 2.428905963897705,
      "step": 3814
    },
    {
      "epoch": 0.8824890122600046,
      "grad_norm": 0.9238738417625427,
      "learning_rate": 0.00031908408302549964,
      "loss": 2.4552388191223145,
      "step": 3815
    },
    {
      "epoch": 0.8827203331020125,
      "grad_norm": 0.9428789615631104,
      "learning_rate": 0.00031820999916686526,
      "loss": 2.289151906967163,
      "step": 3816
    },
    {
      "epoch": 0.8829516539440203,
      "grad_norm": 0.8852382898330688,
      "learning_rate": 0.0003173350134883066,
      "loss": 2.363680601119995,
      "step": 3817
    },
    {
      "epoch": 0.8831829747860283,
      "grad_norm": 0.9560338258743286,
      "learning_rate": 0.0003164591375582082,
      "loss": 2.4804728031158447,
      "step": 3818
    },
    {
      "epoch": 0.8834142956280361,
      "grad_norm": 1.1224981546401978,
      "learning_rate": 0.00031558238295672543,
      "loss": 2.4578425884246826,
      "step": 3819
    },
    {
      "epoch": 0.8836456164700439,
      "grad_norm": 1.0424449443817139,
      "learning_rate": 0.00031470476127563017,
      "loss": 2.6742899417877197,
      "step": 3820
    },
    {
      "epoch": 0.8838769373120519,
      "grad_norm": 0.9035114049911499,
      "learning_rate": 0.00031382628411815865,
      "loss": 2.754335403442383,
      "step": 3821
    },
    {
      "epoch": 0.8841082581540597,
      "grad_norm": 0.8140108585357666,
      "learning_rate": 0.00031294696309885716,
      "loss": 2.7009267807006836,
      "step": 3822
    },
    {
      "epoch": 0.8843395789960675,
      "grad_norm": 1.006603479385376,
      "learning_rate": 0.0003120668098434291,
      "loss": 2.5793824195861816,
      "step": 3823
    },
    {
      "epoch": 0.8845708998380755,
      "grad_norm": 0.8662294745445251,
      "learning_rate": 0.00031118583598858094,
      "loss": 2.6870973110198975,
      "step": 3824
    },
    {
      "epoch": 0.8848022206800833,
      "grad_norm": 1.0194826126098633,
      "learning_rate": 0.00031030405318186865,
      "loss": 2.2210943698883057,
      "step": 3825
    },
    {
      "epoch": 0.8850335415220911,
      "grad_norm": 0.936610221862793,
      "learning_rate": 0.0003094214730815433,
      "loss": 2.3551840782165527,
      "step": 3826
    },
    {
      "epoch": 0.885264862364099,
      "grad_norm": 0.80694979429245,
      "learning_rate": 0.00030853810735639734,
      "loss": 2.2954468727111816,
      "step": 3827
    },
    {
      "epoch": 0.8854961832061069,
      "grad_norm": 0.9137861132621765,
      "learning_rate": 0.00030765396768561005,
      "loss": 2.0335400104522705,
      "step": 3828
    },
    {
      "epoch": 0.8857275040481147,
      "grad_norm": 0.8969683647155762,
      "learning_rate": 0.0003067690657585933,
      "loss": 3.0077028274536133,
      "step": 3829
    },
    {
      "epoch": 0.8859588248901226,
      "grad_norm": 0.8561367392539978,
      "learning_rate": 0.0003058834132748369,
      "loss": 2.3829851150512695,
      "step": 3830
    },
    {
      "epoch": 0.8861901457321305,
      "grad_norm": 0.8684488534927368,
      "learning_rate": 0.0003049970219437539,
      "loss": 2.5856072902679443,
      "step": 3831
    },
    {
      "epoch": 0.8864214665741383,
      "grad_norm": 0.8037282824516296,
      "learning_rate": 0.00030410990348452574,
      "loss": 2.6104702949523926,
      "step": 3832
    },
    {
      "epoch": 0.8866527874161462,
      "grad_norm": 0.7920822501182556,
      "learning_rate": 0.0003032220696259476,
      "loss": 2.283966541290283,
      "step": 3833
    },
    {
      "epoch": 0.8868841082581541,
      "grad_norm": 0.8755449056625366,
      "learning_rate": 0.000302333532106273,
      "loss": 2.43624210357666,
      "step": 3834
    },
    {
      "epoch": 0.887115429100162,
      "grad_norm": 1.0063055753707886,
      "learning_rate": 0.0003014443026730587,
      "loss": 2.1743760108947754,
      "step": 3835
    },
    {
      "epoch": 0.8873467499421698,
      "grad_norm": 0.7992802262306213,
      "learning_rate": 0.0003005543930830095,
      "loss": 2.518136739730835,
      "step": 3836
    },
    {
      "epoch": 0.8875780707841776,
      "grad_norm": 0.9029902815818787,
      "learning_rate": 0.0002996638151018225,
      "loss": 2.3580756187438965,
      "step": 3837
    },
    {
      "epoch": 0.8878093916261856,
      "grad_norm": 0.8472506999969482,
      "learning_rate": 0.0002987725805040321,
      "loss": 2.6207361221313477,
      "step": 3838
    },
    {
      "epoch": 0.8880407124681934,
      "grad_norm": 0.8392115235328674,
      "learning_rate": 0.00029788070107285357,
      "loss": 2.764681816101074,
      "step": 3839
    },
    {
      "epoch": 0.8882720333102012,
      "grad_norm": 0.9169748425483704,
      "learning_rate": 0.00029698818860002794,
      "loss": 2.5486011505126953,
      "step": 3840
    },
    {
      "epoch": 0.8885033541522092,
      "grad_norm": 1.1522047519683838,
      "learning_rate": 0.00029609505488566585,
      "loss": 2.667750835418701,
      "step": 3841
    },
    {
      "epoch": 0.888734674994217,
      "grad_norm": 0.784190833568573,
      "learning_rate": 0.0002952013117380913,
      "loss": 2.3683366775512695,
      "step": 3842
    },
    {
      "epoch": 0.8889659958362248,
      "grad_norm": 0.8930090665817261,
      "learning_rate": 0.00029430697097368604,
      "loss": 2.7058537006378174,
      "step": 3843
    },
    {
      "epoch": 0.8891973166782328,
      "grad_norm": 1.0427663326263428,
      "learning_rate": 0.00029341204441673266,
      "loss": 2.540271043777466,
      "step": 3844
    },
    {
      "epoch": 0.8894286375202406,
      "grad_norm": 1.8979160785675049,
      "learning_rate": 0.0002925165438992589,
      "loss": 2.6416375637054443,
      "step": 3845
    },
    {
      "epoch": 0.8896599583622484,
      "grad_norm": 1.1069841384887695,
      "learning_rate": 0.00029162048126088117,
      "loss": 2.2566184997558594,
      "step": 3846
    },
    {
      "epoch": 0.8898912792042563,
      "grad_norm": 0.9211397767066956,
      "learning_rate": 0.0002907238683486472,
      "loss": 2.338656187057495,
      "step": 3847
    },
    {
      "epoch": 0.8901226000462642,
      "grad_norm": 1.1104730367660522,
      "learning_rate": 0.0002898267170168807,
      "loss": 2.6236400604248047,
      "step": 3848
    },
    {
      "epoch": 0.890353920888272,
      "grad_norm": 0.8718827366828918,
      "learning_rate": 0.00028892903912702345,
      "loss": 2.4739813804626465,
      "step": 3849
    },
    {
      "epoch": 0.8905852417302799,
      "grad_norm": 1.0583229064941406,
      "learning_rate": 0.0002880308465474792,
      "loss": 2.756728172302246,
      "step": 3850
    },
    {
      "epoch": 0.8908165625722878,
      "grad_norm": 1.0797449350357056,
      "learning_rate": 0.00028713215115345645,
      "loss": 2.0734667778015137,
      "step": 3851
    },
    {
      "epoch": 0.8910478834142956,
      "grad_norm": 0.9732435941696167,
      "learning_rate": 0.0002862329648268117,
      "loss": 2.491527557373047,
      "step": 3852
    },
    {
      "epoch": 0.8912792042563035,
      "grad_norm": 1.7976651191711426,
      "learning_rate": 0.0002853332994558919,
      "loss": 2.6771745681762695,
      "step": 3853
    },
    {
      "epoch": 0.8915105250983114,
      "grad_norm": 0.9261050820350647,
      "learning_rate": 0.0002844331669353777,
      "loss": 2.2547812461853027,
      "step": 3854
    },
    {
      "epoch": 0.8917418459403192,
      "grad_norm": 0.9990341663360596,
      "learning_rate": 0.00028353257916612605,
      "loss": 2.216564178466797,
      "step": 3855
    },
    {
      "epoch": 0.8919731667823271,
      "grad_norm": 0.9875020980834961,
      "learning_rate": 0.000282631548055013,
      "loss": 2.5799193382263184,
      "step": 3856
    },
    {
      "epoch": 0.8922044876243349,
      "grad_norm": 0.8052231073379517,
      "learning_rate": 0.0002817300855147756,
      "loss": 2.1756460666656494,
      "step": 3857
    },
    {
      "epoch": 0.8924358084663429,
      "grad_norm": 0.8777169585227966,
      "learning_rate": 0.0002808282034638555,
      "loss": 2.451634407043457,
      "step": 3858
    },
    {
      "epoch": 0.8926671293083507,
      "grad_norm": 0.9548158049583435,
      "learning_rate": 0.00027992591382624063,
      "loss": 2.3214926719665527,
      "step": 3859
    },
    {
      "epoch": 0.8928984501503585,
      "grad_norm": 1.6637603044509888,
      "learning_rate": 0.0002790232285313076,
      "loss": 2.663159132003784,
      "step": 3860
    },
    {
      "epoch": 0.8931297709923665,
      "grad_norm": 0.7767270803451538,
      "learning_rate": 0.0002781201595136641,
      "loss": 2.7490334510803223,
      "step": 3861
    },
    {
      "epoch": 0.8933610918343743,
      "grad_norm": 0.8340877294540405,
      "learning_rate": 0.00027721671871299114,
      "loss": 2.3590927124023438,
      "step": 3862
    },
    {
      "epoch": 0.8935924126763821,
      "grad_norm": 0.9618812203407288,
      "learning_rate": 0.0002763129180738853,
      "loss": 2.6685967445373535,
      "step": 3863
    },
    {
      "epoch": 0.8938237335183901,
      "grad_norm": 0.8903891444206238,
      "learning_rate": 0.0002754087695457005,
      "loss": 2.0844531059265137,
      "step": 3864
    },
    {
      "epoch": 0.8940550543603979,
      "grad_norm": 0.8304296731948853,
      "learning_rate": 0.0002745042850823902,
      "loss": 2.527780532836914,
      "step": 3865
    },
    {
      "epoch": 0.8942863752024057,
      "grad_norm": 0.9685689210891724,
      "learning_rate": 0.00027359947664234936,
      "loss": 2.5386433601379395,
      "step": 3866
    },
    {
      "epoch": 0.8945176960444136,
      "grad_norm": 0.8220421671867371,
      "learning_rate": 0.0002726943561882563,
      "loss": 2.1797595024108887,
      "step": 3867
    },
    {
      "epoch": 0.8947490168864215,
      "grad_norm": 1.163561224937439,
      "learning_rate": 0.0002717889356869146,
      "loss": 2.945173740386963,
      "step": 3868
    },
    {
      "epoch": 0.8949803377284293,
      "grad_norm": 1.0624277591705322,
      "learning_rate": 0.0002708832271090947,
      "loss": 2.282181739807129,
      "step": 3869
    },
    {
      "epoch": 0.8952116585704372,
      "grad_norm": 0.9909601807594299,
      "learning_rate": 0.00026997724242937597,
      "loss": 2.739252805709839,
      "step": 3870
    },
    {
      "epoch": 0.8954429794124451,
      "grad_norm": 0.8104747533798218,
      "learning_rate": 0.0002690709936259881,
      "loss": 2.407168388366699,
      "step": 3871
    },
    {
      "epoch": 0.8956743002544529,
      "grad_norm": 0.9642229676246643,
      "learning_rate": 0.00026816449268065267,
      "loss": 2.527894973754883,
      "step": 3872
    },
    {
      "epoch": 0.8959056210964608,
      "grad_norm": 0.9528427124023438,
      "learning_rate": 0.00026725775157842506,
      "loss": 2.932577133178711,
      "step": 3873
    },
    {
      "epoch": 0.8961369419384686,
      "grad_norm": 0.8989641666412354,
      "learning_rate": 0.0002663507823075358,
      "loss": 2.6585536003112793,
      "step": 3874
    },
    {
      "epoch": 0.8963682627804765,
      "grad_norm": 0.7794551849365234,
      "learning_rate": 0.00026544359685923204,
      "loss": 2.5183401107788086,
      "step": 3875
    },
    {
      "epoch": 0.8965995836224844,
      "grad_norm": 0.8509832620620728,
      "learning_rate": 0.00026453620722761897,
      "loss": 2.420098304748535,
      "step": 3876
    },
    {
      "epoch": 0.8968309044644922,
      "grad_norm": 0.9408103823661804,
      "learning_rate": 0.0002636286254095016,
      "loss": 2.44295072555542,
      "step": 3877
    },
    {
      "epoch": 0.8970622253065001,
      "grad_norm": 0.9224709868431091,
      "learning_rate": 0.0002627208634042257,
      "loss": 2.698448896408081,
      "step": 3878
    },
    {
      "epoch": 0.897293546148508,
      "grad_norm": 0.9945060014724731,
      "learning_rate": 0.00026181293321351926,
      "loss": 2.229306697845459,
      "step": 3879
    },
    {
      "epoch": 0.8975248669905158,
      "grad_norm": 0.7489075660705566,
      "learning_rate": 0.00026090484684133404,
      "loss": 2.3807222843170166,
      "step": 3880
    },
    {
      "epoch": 0.8977561878325238,
      "grad_norm": 0.7649733424186707,
      "learning_rate": 0.0002599966162936866,
      "loss": 2.512479305267334,
      "step": 3881
    },
    {
      "epoch": 0.8979875086745316,
      "grad_norm": 1.0056664943695068,
      "learning_rate": 0.0002590882535785,
      "loss": 2.2374653816223145,
      "step": 3882
    },
    {
      "epoch": 0.8982188295165394,
      "grad_norm": 1.0268003940582275,
      "learning_rate": 0.00025817977070544405,
      "loss": 2.4099156856536865,
      "step": 3883
    },
    {
      "epoch": 0.8984501503585473,
      "grad_norm": 0.9174339175224304,
      "learning_rate": 0.00025727117968577786,
      "loss": 2.1507129669189453,
      "step": 3884
    },
    {
      "epoch": 0.8986814712005552,
      "grad_norm": 0.7978119850158691,
      "learning_rate": 0.00025636249253219005,
      "loss": 2.3239965438842773,
      "step": 3885
    },
    {
      "epoch": 0.898912792042563,
      "grad_norm": 0.7982020378112793,
      "learning_rate": 0.0002554537212586403,
      "loss": 2.3068628311157227,
      "step": 3886
    },
    {
      "epoch": 0.8991441128845709,
      "grad_norm": 0.7778011560440063,
      "learning_rate": 0.00025454487788020044,
      "loss": 2.4839649200439453,
      "step": 3887
    },
    {
      "epoch": 0.8993754337265788,
      "grad_norm": 0.7910434603691101,
      "learning_rate": 0.0002536359744128957,
      "loss": 2.411881446838379,
      "step": 3888
    },
    {
      "epoch": 0.8996067545685866,
      "grad_norm": 1.090920329093933,
      "learning_rate": 0.0002527270228735456,
      "loss": 2.587944507598877,
      "step": 3889
    },
    {
      "epoch": 0.8998380754105945,
      "grad_norm": 0.7916618585586548,
      "learning_rate": 0.00025181803527960547,
      "loss": 2.5098061561584473,
      "step": 3890
    },
    {
      "epoch": 0.9000693962526024,
      "grad_norm": 0.9593981504440308,
      "learning_rate": 0.00025090902364900697,
      "loss": 2.605288505554199,
      "step": 3891
    },
    {
      "epoch": 0.9003007170946102,
      "grad_norm": 0.7411696314811707,
      "learning_rate": 0.00025,
      "loss": 2.7304632663726807,
      "step": 3892
    },
    {
      "epoch": 0.9005320379366181,
      "grad_norm": 0.9958679676055908,
      "learning_rate": 0.000249090976350993,
      "loss": 2.4341647624969482,
      "step": 3893
    },
    {
      "epoch": 0.9007633587786259,
      "grad_norm": 0.9708659052848816,
      "learning_rate": 0.00024818196472039465,
      "loss": 2.659299612045288,
      "step": 3894
    },
    {
      "epoch": 0.9009946796206338,
      "grad_norm": 0.8720176219940186,
      "learning_rate": 0.00024727297712645444,
      "loss": 2.4441981315612793,
      "step": 3895
    },
    {
      "epoch": 0.9012260004626417,
      "grad_norm": 0.8874751925468445,
      "learning_rate": 0.0002463640255871043,
      "loss": 2.503607749938965,
      "step": 3896
    },
    {
      "epoch": 0.9014573213046495,
      "grad_norm": 0.7519695162773132,
      "learning_rate": 0.00024545512211979957,
      "loss": 2.1379222869873047,
      "step": 3897
    },
    {
      "epoch": 0.9016886421466574,
      "grad_norm": 0.830615758895874,
      "learning_rate": 0.00024454627874135975,
      "loss": 2.2974376678466797,
      "step": 3898
    },
    {
      "epoch": 0.9019199629886653,
      "grad_norm": 0.8466845154762268,
      "learning_rate": 0.00024363750746780996,
      "loss": 2.28037166595459,
      "step": 3899
    },
    {
      "epoch": 0.9021512838306731,
      "grad_norm": 0.8215333223342896,
      "learning_rate": 0.00024272882031422215,
      "loss": 2.652790069580078,
      "step": 3900
    },
    {
      "epoch": 0.902382604672681,
      "grad_norm": 0.8306387066841125,
      "learning_rate": 0.00024182022929455598,
      "loss": 2.222747325897217,
      "step": 3901
    },
    {
      "epoch": 0.9026139255146889,
      "grad_norm": 1.1694741249084473,
      "learning_rate": 0.0002409117464215001,
      "loss": 2.365182638168335,
      "step": 3902
    },
    {
      "epoch": 0.9028452463566967,
      "grad_norm": 0.9019613265991211,
      "learning_rate": 0.00024000338370631336,
      "loss": 2.8095755577087402,
      "step": 3903
    },
    {
      "epoch": 0.9030765671987045,
      "grad_norm": 0.9768400192260742,
      "learning_rate": 0.00023909515315866605,
      "loss": 2.4080114364624023,
      "step": 3904
    },
    {
      "epoch": 0.9033078880407125,
      "grad_norm": 1.0192257165908813,
      "learning_rate": 0.0002381870667864808,
      "loss": 2.046928644180298,
      "step": 3905
    },
    {
      "epoch": 0.9035392088827203,
      "grad_norm": 0.9320544004440308,
      "learning_rate": 0.00023727913659577434,
      "loss": 2.3730199337005615,
      "step": 3906
    },
    {
      "epoch": 0.9037705297247282,
      "grad_norm": 1.0305209159851074,
      "learning_rate": 0.0002363713745904984,
      "loss": 1.9734159708023071,
      "step": 3907
    },
    {
      "epoch": 0.9040018505667361,
      "grad_norm": 1.0598901510238647,
      "learning_rate": 0.00023546379277238105,
      "loss": 2.517512798309326,
      "step": 3908
    },
    {
      "epoch": 0.9042331714087439,
      "grad_norm": 0.7834630012512207,
      "learning_rate": 0.00023455640314076805,
      "loss": 2.3766422271728516,
      "step": 3909
    },
    {
      "epoch": 0.9044644922507518,
      "grad_norm": 0.9105112552642822,
      "learning_rate": 0.00023364921769246423,
      "loss": 2.248194694519043,
      "step": 3910
    },
    {
      "epoch": 0.9046958130927597,
      "grad_norm": 1.0027488470077515,
      "learning_rate": 0.00023274224842157503,
      "loss": 2.5750670433044434,
      "step": 3911
    },
    {
      "epoch": 0.9049271339347675,
      "grad_norm": 1.120560884475708,
      "learning_rate": 0.00023183550731934734,
      "loss": 2.411616325378418,
      "step": 3912
    },
    {
      "epoch": 0.9051584547767754,
      "grad_norm": 0.7899209260940552,
      "learning_rate": 0.0002309290063740119,
      "loss": 2.362736225128174,
      "step": 3913
    },
    {
      "epoch": 0.9053897756187832,
      "grad_norm": 0.920970618724823,
      "learning_rate": 0.00023002275757062407,
      "loss": 2.3096413612365723,
      "step": 3914
    },
    {
      "epoch": 0.9056210964607911,
      "grad_norm": 0.7171766757965088,
      "learning_rate": 0.00022911677289090535,
      "loss": 2.209871292114258,
      "step": 3915
    },
    {
      "epoch": 0.905852417302799,
      "grad_norm": 1.0004817247390747,
      "learning_rate": 0.00022821106431308543,
      "loss": 2.1760780811309814,
      "step": 3916
    },
    {
      "epoch": 0.9060837381448068,
      "grad_norm": 1.3297436237335205,
      "learning_rate": 0.00022730564381174374,
      "loss": 2.2564239501953125,
      "step": 3917
    },
    {
      "epoch": 0.9063150589868147,
      "grad_norm": 0.7147906422615051,
      "learning_rate": 0.00022640052335765076,
      "loss": 2.480958938598633,
      "step": 3918
    },
    {
      "epoch": 0.9065463798288226,
      "grad_norm": 1.1415637731552124,
      "learning_rate": 0.00022549571491760985,
      "loss": 2.2810237407684326,
      "step": 3919
    },
    {
      "epoch": 0.9067777006708304,
      "grad_norm": 0.7288040518760681,
      "learning_rate": 0.00022459123045429953,
      "loss": 2.3288321495056152,
      "step": 3920
    },
    {
      "epoch": 0.9070090215128384,
      "grad_norm": 0.7197046875953674,
      "learning_rate": 0.0002236870819261148,
      "loss": 2.2766356468200684,
      "step": 3921
    },
    {
      "epoch": 0.9072403423548462,
      "grad_norm": 1.6805330514907837,
      "learning_rate": 0.00022278328128700893,
      "loss": 2.5595459938049316,
      "step": 3922
    },
    {
      "epoch": 0.907471663196854,
      "grad_norm": 1.3181865215301514,
      "learning_rate": 0.0002218798404863359,
      "loss": 2.2682113647460938,
      "step": 3923
    },
    {
      "epoch": 0.9077029840388618,
      "grad_norm": 0.7811545133590698,
      "learning_rate": 0.0002209767714686924,
      "loss": 2.458062171936035,
      "step": 3924
    },
    {
      "epoch": 0.9079343048808698,
      "grad_norm": 1.084822416305542,
      "learning_rate": 0.00022007408617375943,
      "loss": 2.3810067176818848,
      "step": 3925
    },
    {
      "epoch": 0.9081656257228776,
      "grad_norm": 1.3165817260742188,
      "learning_rate": 0.0002191717965361445,
      "loss": 2.767059087753296,
      "step": 3926
    },
    {
      "epoch": 0.9083969465648855,
      "grad_norm": 0.9802242517471313,
      "learning_rate": 0.00021826991448522438,
      "loss": 2.0790021419525146,
      "step": 3927
    },
    {
      "epoch": 0.9086282674068934,
      "grad_norm": 0.8378863334655762,
      "learning_rate": 0.00021736845194498717,
      "loss": 2.1707353591918945,
      "step": 3928
    },
    {
      "epoch": 0.9088595882489012,
      "grad_norm": 0.9460284113883972,
      "learning_rate": 0.00021646742083387396,
      "loss": 2.3652243614196777,
      "step": 3929
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 1.0450146198272705,
      "learning_rate": 0.00021556683306462232,
      "loss": 2.1305716037750244,
      "step": 3930
    },
    {
      "epoch": 0.909322229932917,
      "grad_norm": 0.8617326617240906,
      "learning_rate": 0.0002146667005441082,
      "loss": 2.475071907043457,
      "step": 3931
    },
    {
      "epoch": 0.9095535507749248,
      "grad_norm": 1.0416898727416992,
      "learning_rate": 0.00021376703517318835,
      "loss": 2.5494861602783203,
      "step": 3932
    },
    {
      "epoch": 0.9097848716169327,
      "grad_norm": 0.8030207753181458,
      "learning_rate": 0.00021286784884654353,
      "loss": 2.4234704971313477,
      "step": 3933
    },
    {
      "epoch": 0.9100161924589405,
      "grad_norm": 1.020963430404663,
      "learning_rate": 0.00021196915345252082,
      "loss": 2.3844852447509766,
      "step": 3934
    },
    {
      "epoch": 0.9102475133009484,
      "grad_norm": 0.8103867173194885,
      "learning_rate": 0.00021107096087297664,
      "loss": 2.5811054706573486,
      "step": 3935
    },
    {
      "epoch": 0.9104788341429563,
      "grad_norm": 0.8936412930488586,
      "learning_rate": 0.00021017328298311938,
      "loss": 2.510867118835449,
      "step": 3936
    },
    {
      "epoch": 0.9107101549849641,
      "grad_norm": 0.9147470593452454,
      "learning_rate": 0.00020927613165135284,
      "loss": 2.4120826721191406,
      "step": 3937
    },
    {
      "epoch": 0.910941475826972,
      "grad_norm": 1.0093547105789185,
      "learning_rate": 0.00020837951873911898,
      "loss": 2.408639430999756,
      "step": 3938
    },
    {
      "epoch": 0.9111727966689799,
      "grad_norm": 1.1293071508407593,
      "learning_rate": 0.00020748345610074114,
      "loss": 2.644878625869751,
      "step": 3939
    },
    {
      "epoch": 0.9114041175109877,
      "grad_norm": 0.7770459055900574,
      "learning_rate": 0.00020658795558326743,
      "loss": 2.312269687652588,
      "step": 3940
    },
    {
      "epoch": 0.9116354383529957,
      "grad_norm": 0.8399761319160461,
      "learning_rate": 0.0002056930290263141,
      "loss": 2.53257417678833,
      "step": 3941
    },
    {
      "epoch": 0.9118667591950035,
      "grad_norm": 0.8278478980064392,
      "learning_rate": 0.00020479868826190871,
      "loss": 2.7358524799346924,
      "step": 3942
    },
    {
      "epoch": 0.9120980800370113,
      "grad_norm": 0.9428454041481018,
      "learning_rate": 0.00020390494511433416,
      "loss": 2.2836484909057617,
      "step": 3943
    },
    {
      "epoch": 0.9123294008790191,
      "grad_norm": 0.8651256561279297,
      "learning_rate": 0.00020301181139997204,
      "loss": 2.5572333335876465,
      "step": 3944
    },
    {
      "epoch": 0.9125607217210271,
      "grad_norm": 0.8333393335342407,
      "learning_rate": 0.00020211929892714647,
      "loss": 2.4302756786346436,
      "step": 3945
    },
    {
      "epoch": 0.9127920425630349,
      "grad_norm": 2.0956809520721436,
      "learning_rate": 0.00020122741949596797,
      "loss": 2.305994987487793,
      "step": 3946
    },
    {
      "epoch": 0.9130233634050428,
      "grad_norm": 1.6422327756881714,
      "learning_rate": 0.00020033618489817749,
      "loss": 2.6639466285705566,
      "step": 3947
    },
    {
      "epoch": 0.9132546842470507,
      "grad_norm": 1.005672574043274,
      "learning_rate": 0.00019944560691699057,
      "loss": 2.458986282348633,
      "step": 3948
    },
    {
      "epoch": 0.9134860050890585,
      "grad_norm": 1.0419694185256958,
      "learning_rate": 0.00019855569732694128,
      "loss": 2.3193936347961426,
      "step": 3949
    },
    {
      "epoch": 0.9137173259310664,
      "grad_norm": 0.8952090740203857,
      "learning_rate": 0.000197666467893727,
      "loss": 2.1199190616607666,
      "step": 3950
    },
    {
      "epoch": 0.9139486467730743,
      "grad_norm": 0.9912616610527039,
      "learning_rate": 0.0001967779303740524,
      "loss": 2.693665027618408,
      "step": 3951
    },
    {
      "epoch": 0.9141799676150821,
      "grad_norm": 0.9002057313919067,
      "learning_rate": 0.0001958900965154743,
      "loss": 2.160792827606201,
      "step": 3952
    },
    {
      "epoch": 0.91441128845709,
      "grad_norm": 1.101587176322937,
      "learning_rate": 0.00019500297805624618,
      "loss": 2.259523391723633,
      "step": 3953
    },
    {
      "epoch": 0.9146426092990978,
      "grad_norm": 0.7717968225479126,
      "learning_rate": 0.00019411658672516315,
      "loss": 2.4388492107391357,
      "step": 3954
    },
    {
      "epoch": 0.9148739301411057,
      "grad_norm": 0.9482707381248474,
      "learning_rate": 0.00019323093424140672,
      "loss": 2.216050863265991,
      "step": 3955
    },
    {
      "epoch": 0.9151052509831136,
      "grad_norm": 0.7285901308059692,
      "learning_rate": 0.00019234603231438997,
      "loss": 2.0985000133514404,
      "step": 3956
    },
    {
      "epoch": 0.9153365718251214,
      "grad_norm": 0.9252454042434692,
      "learning_rate": 0.0001914618926436027,
      "loss": 2.4918994903564453,
      "step": 3957
    },
    {
      "epoch": 0.9155678926671293,
      "grad_norm": 0.8523378372192383,
      "learning_rate": 0.00019057852691845676,
      "loss": 2.958413600921631,
      "step": 3958
    },
    {
      "epoch": 0.9157992135091372,
      "grad_norm": 0.7443702220916748,
      "learning_rate": 0.00018969594681813142,
      "loss": 2.401430606842041,
      "step": 3959
    },
    {
      "epoch": 0.916030534351145,
      "grad_norm": 0.9647448062896729,
      "learning_rate": 0.00018881416401141904,
      "loss": 2.52408766746521,
      "step": 3960
    },
    {
      "epoch": 0.916261855193153,
      "grad_norm": 1.0202221870422363,
      "learning_rate": 0.0001879331901565709,
      "loss": 2.286114454269409,
      "step": 3961
    },
    {
      "epoch": 0.9164931760351608,
      "grad_norm": 0.7862979173660278,
      "learning_rate": 0.00018705303690114288,
      "loss": 2.1823203563690186,
      "step": 3962
    },
    {
      "epoch": 0.9167244968771686,
      "grad_norm": 0.8352393507957458,
      "learning_rate": 0.00018617371588184136,
      "loss": 2.6364643573760986,
      "step": 3963
    },
    {
      "epoch": 0.9169558177191764,
      "grad_norm": 0.8237755298614502,
      "learning_rate": 0.0001852952387243698,
      "loss": 2.068159580230713,
      "step": 3964
    },
    {
      "epoch": 0.9171871385611844,
      "grad_norm": 0.8811264634132385,
      "learning_rate": 0.0001844176170432747,
      "loss": 2.3159985542297363,
      "step": 3965
    },
    {
      "epoch": 0.9174184594031922,
      "grad_norm": 0.9268269538879395,
      "learning_rate": 0.0001835408624417918,
      "loss": 2.444277763366699,
      "step": 3966
    },
    {
      "epoch": 0.9176497802452,
      "grad_norm": 0.8273143172264099,
      "learning_rate": 0.00018266498651169349,
      "loss": 2.4028406143188477,
      "step": 3967
    },
    {
      "epoch": 0.917881101087208,
      "grad_norm": 0.85252845287323,
      "learning_rate": 0.0001817900008331348,
      "loss": 2.422360420227051,
      "step": 3968
    },
    {
      "epoch": 0.9181124219292158,
      "grad_norm": 0.8620074987411499,
      "learning_rate": 0.00018091591697450043,
      "loss": 2.583651542663574,
      "step": 3969
    },
    {
      "epoch": 0.9183437427712237,
      "grad_norm": 1.01572847366333,
      "learning_rate": 0.000180042746492252,
      "loss": 2.5434799194335938,
      "step": 3970
    },
    {
      "epoch": 0.9185750636132316,
      "grad_norm": 1.440956950187683,
      "learning_rate": 0.0001791705009307753,
      "loss": 2.6674742698669434,
      "step": 3971
    },
    {
      "epoch": 0.9188063844552394,
      "grad_norm": 1.047162413597107,
      "learning_rate": 0.00017829919182222752,
      "loss": 2.4528379440307617,
      "step": 3972
    },
    {
      "epoch": 0.9190377052972473,
      "grad_norm": 1.084546685218811,
      "learning_rate": 0.00017742883068638446,
      "loss": 2.256157636642456,
      "step": 3973
    },
    {
      "epoch": 0.9192690261392551,
      "grad_norm": 0.7788847088813782,
      "learning_rate": 0.00017655942903048888,
      "loss": 2.341193199157715,
      "step": 3974
    },
    {
      "epoch": 0.919500346981263,
      "grad_norm": 0.8891081213951111,
      "learning_rate": 0.00017569099834909818,
      "loss": 2.7089457511901855,
      "step": 3975
    },
    {
      "epoch": 0.9197316678232709,
      "grad_norm": 1.2016528844833374,
      "learning_rate": 0.00017482355012393176,
      "loss": 2.7414979934692383,
      "step": 3976
    },
    {
      "epoch": 0.9199629886652787,
      "grad_norm": 0.825495183467865,
      "learning_rate": 0.00017395709582372016,
      "loss": 2.718801498413086,
      "step": 3977
    },
    {
      "epoch": 0.9201943095072866,
      "grad_norm": 0.9608388543128967,
      "learning_rate": 0.00017309164690405292,
      "loss": 2.451542377471924,
      "step": 3978
    },
    {
      "epoch": 0.9204256303492945,
      "grad_norm": 0.9433290362358093,
      "learning_rate": 0.0001722272148072273,
      "loss": 2.352043628692627,
      "step": 3979
    },
    {
      "epoch": 0.9206569511913023,
      "grad_norm": 1.2656444311141968,
      "learning_rate": 0.00017136381096209664,
      "loss": 2.438389301300049,
      "step": 3980
    },
    {
      "epoch": 0.9208882720333103,
      "grad_norm": 1.2357033491134644,
      "learning_rate": 0.00017050144678391963,
      "loss": 2.152191162109375,
      "step": 3981
    },
    {
      "epoch": 0.9211195928753181,
      "grad_norm": 0.9343427419662476,
      "learning_rate": 0.00016964013367420965,
      "loss": 2.4302356243133545,
      "step": 3982
    },
    {
      "epoch": 0.9213509137173259,
      "grad_norm": 0.7278953790664673,
      "learning_rate": 0.00016877988302058312,
      "loss": 2.0976080894470215,
      "step": 3983
    },
    {
      "epoch": 0.9215822345593337,
      "grad_norm": 0.7898283004760742,
      "learning_rate": 0.00016792070619660977,
      "loss": 2.2110958099365234,
      "step": 3984
    },
    {
      "epoch": 0.9218135554013417,
      "grad_norm": 0.6615616083145142,
      "learning_rate": 0.00016706261456166205,
      "loss": 2.3249945640563965,
      "step": 3985
    },
    {
      "epoch": 0.9220448762433495,
      "grad_norm": 0.8639052510261536,
      "learning_rate": 0.0001662056194607646,
      "loss": 2.211409330368042,
      "step": 3986
    },
    {
      "epoch": 0.9222761970853574,
      "grad_norm": 0.9204817414283752,
      "learning_rate": 0.00016534973222444494,
      "loss": 2.5200977325439453,
      "step": 3987
    },
    {
      "epoch": 0.9225075179273653,
      "grad_norm": 0.856096625328064,
      "learning_rate": 0.00016449496416858284,
      "loss": 2.495060920715332,
      "step": 3988
    },
    {
      "epoch": 0.9227388387693731,
      "grad_norm": 0.7338510751724243,
      "learning_rate": 0.00016364132659426144,
      "loss": 2.40138840675354,
      "step": 3989
    },
    {
      "epoch": 0.922970159611381,
      "grad_norm": 0.7220907211303711,
      "learning_rate": 0.00016278883078761733,
      "loss": 2.6308112144470215,
      "step": 3990
    },
    {
      "epoch": 0.9232014804533889,
      "grad_norm": 0.662051796913147,
      "learning_rate": 0.00016193748801969163,
      "loss": 2.583658218383789,
      "step": 3991
    },
    {
      "epoch": 0.9234328012953967,
      "grad_norm": 0.70716392993927,
      "learning_rate": 0.0001610873095462809,
      "loss": 2.629162549972534,
      "step": 3992
    },
    {
      "epoch": 0.9236641221374046,
      "grad_norm": 1.0216474533081055,
      "learning_rate": 0.0001602383066077882,
      "loss": 2.433669090270996,
      "step": 3993
    },
    {
      "epoch": 0.9238954429794124,
      "grad_norm": 1.029405117034912,
      "learning_rate": 0.0001593904904290746,
      "loss": 2.6294188499450684,
      "step": 3994
    },
    {
      "epoch": 0.9241267638214203,
      "grad_norm": 0.8395315408706665,
      "learning_rate": 0.00015854387221931089,
      "loss": 2.3773908615112305,
      "step": 3995
    },
    {
      "epoch": 0.9243580846634282,
      "grad_norm": 0.8215494155883789,
      "learning_rate": 0.00015769846317182891,
      "loss": 2.5592641830444336,
      "step": 3996
    },
    {
      "epoch": 0.924589405505436,
      "grad_norm": 0.8570623397827148,
      "learning_rate": 0.00015685427446397428,
      "loss": 2.0288562774658203,
      "step": 3997
    },
    {
      "epoch": 0.9248207263474439,
      "grad_norm": 0.8921326994895935,
      "learning_rate": 0.000156011317256958,
      "loss": 2.3932008743286133,
      "step": 3998
    },
    {
      "epoch": 0.9250520471894518,
      "grad_norm": 1.0067936182022095,
      "learning_rate": 0.00015516960269570918,
      "loss": 2.4623045921325684,
      "step": 3999
    },
    {
      "epoch": 0.9252833680314596,
      "grad_norm": 0.8305838108062744,
      "learning_rate": 0.00015432914190872756,
      "loss": 2.4627246856689453,
      "step": 4000
    },
    {
      "epoch": 0.9255146888734676,
      "grad_norm": 0.7785557508468628,
      "learning_rate": 0.0001534899460079366,
      "loss": 2.51973557472229,
      "step": 4001
    },
    {
      "epoch": 0.9257460097154754,
      "grad_norm": 0.8259813785552979,
      "learning_rate": 0.00015265202608853628,
      "loss": 2.5055453777313232,
      "step": 4002
    },
    {
      "epoch": 0.9259773305574832,
      "grad_norm": 1.2663073539733887,
      "learning_rate": 0.0001518153932288565,
      "loss": 2.478999137878418,
      "step": 4003
    },
    {
      "epoch": 0.926208651399491,
      "grad_norm": 0.970445990562439,
      "learning_rate": 0.0001509800584902108,
      "loss": 2.272261142730713,
      "step": 4004
    },
    {
      "epoch": 0.926439972241499,
      "grad_norm": 0.8471930027008057,
      "learning_rate": 0.00015014603291674987,
      "loss": 2.352316379547119,
      "step": 4005
    },
    {
      "epoch": 0.9266712930835068,
      "grad_norm": 0.6962822675704956,
      "learning_rate": 0.00014931332753531572,
      "loss": 2.474661350250244,
      "step": 4006
    },
    {
      "epoch": 0.9269026139255147,
      "grad_norm": 0.7644684314727783,
      "learning_rate": 0.0001484819533552956,
      "loss": 2.4444408416748047,
      "step": 4007
    },
    {
      "epoch": 0.9271339347675226,
      "grad_norm": 0.7777777910232544,
      "learning_rate": 0.00014765192136847685,
      "loss": 2.244466781616211,
      "step": 4008
    },
    {
      "epoch": 0.9273652556095304,
      "grad_norm": 0.928949236869812,
      "learning_rate": 0.00014682324254890135,
      "loss": 2.1725573539733887,
      "step": 4009
    },
    {
      "epoch": 0.9275965764515383,
      "grad_norm": 0.7628275156021118,
      "learning_rate": 0.00014599592785272025,
      "loss": 2.4295363426208496,
      "step": 4010
    },
    {
      "epoch": 0.9278278972935462,
      "grad_norm": 0.9692445993423462,
      "learning_rate": 0.0001451699882180496,
      "loss": 2.2722573280334473,
      "step": 4011
    },
    {
      "epoch": 0.928059218135554,
      "grad_norm": 0.9345256090164185,
      "learning_rate": 0.0001443454345648252,
      "loss": 2.3031606674194336,
      "step": 4012
    },
    {
      "epoch": 0.9282905389775619,
      "grad_norm": 0.7951247692108154,
      "learning_rate": 0.00014352227779465855,
      "loss": 2.382739543914795,
      "step": 4013
    },
    {
      "epoch": 0.9285218598195697,
      "grad_norm": 0.7205172777175903,
      "learning_rate": 0.0001427005287906928,
      "loss": 2.648688316345215,
      "step": 4014
    },
    {
      "epoch": 0.9287531806615776,
      "grad_norm": 0.685621440410614,
      "learning_rate": 0.00014188019841745842,
      "loss": 2.1948955059051514,
      "step": 4015
    },
    {
      "epoch": 0.9289845015035855,
      "grad_norm": 0.8640634417533875,
      "learning_rate": 0.00014106129752073022,
      "loss": 2.3959507942199707,
      "step": 4016
    },
    {
      "epoch": 0.9292158223455933,
      "grad_norm": 0.7840452194213867,
      "learning_rate": 0.0001402438369273831,
      "loss": 2.447388172149658,
      "step": 4017
    },
    {
      "epoch": 0.9294471431876012,
      "grad_norm": 0.8442760705947876,
      "learning_rate": 0.00013942782744524973,
      "loss": 2.346501111984253,
      "step": 4018
    },
    {
      "epoch": 0.9296784640296091,
      "grad_norm": 0.7190546989440918,
      "learning_rate": 0.00013861327986297716,
      "loss": 2.1544198989868164,
      "step": 4019
    },
    {
      "epoch": 0.9299097848716169,
      "grad_norm": 0.9004926085472107,
      "learning_rate": 0.00013780020494988447,
      "loss": 2.4718456268310547,
      "step": 4020
    },
    {
      "epoch": 0.9301411057136248,
      "grad_norm": 0.7631949782371521,
      "learning_rate": 0.00013698861345582007,
      "loss": 2.224404811859131,
      "step": 4021
    },
    {
      "epoch": 0.9303724265556327,
      "grad_norm": 0.9251269102096558,
      "learning_rate": 0.00013617851611101995,
      "loss": 2.5472211837768555,
      "step": 4022
    },
    {
      "epoch": 0.9306037473976405,
      "grad_norm": 0.6382388472557068,
      "learning_rate": 0.00013536992362596535,
      "loss": 2.4018664360046387,
      "step": 4023
    },
    {
      "epoch": 0.9308350682396483,
      "grad_norm": 1.3395328521728516,
      "learning_rate": 0.00013456284669124158,
      "loss": 2.3022494316101074,
      "step": 4024
    },
    {
      "epoch": 0.9310663890816563,
      "grad_norm": 0.7757973670959473,
      "learning_rate": 0.0001337572959773965,
      "loss": 2.627894878387451,
      "step": 4025
    },
    {
      "epoch": 0.9312977099236641,
      "grad_norm": 0.7807584404945374,
      "learning_rate": 0.00013295328213479952,
      "loss": 1.9066888093948364,
      "step": 4026
    },
    {
      "epoch": 0.931529030765672,
      "grad_norm": 0.8952817320823669,
      "learning_rate": 0.00013215081579350058,
      "loss": 2.5430490970611572,
      "step": 4027
    },
    {
      "epoch": 0.9317603516076799,
      "grad_norm": 0.787748396396637,
      "learning_rate": 0.0001313499075630899,
      "loss": 2.1401963233947754,
      "step": 4028
    },
    {
      "epoch": 0.9319916724496877,
      "grad_norm": 1.0700440406799316,
      "learning_rate": 0.0001305505680325575,
      "loss": 2.272057294845581,
      "step": 4029
    },
    {
      "epoch": 0.9322229932916956,
      "grad_norm": 0.8480278849601746,
      "learning_rate": 0.00012975280777015313,
      "loss": 2.2362449169158936,
      "step": 4030
    },
    {
      "epoch": 0.9324543141337035,
      "grad_norm": 0.8337933421134949,
      "learning_rate": 0.00012895663732324686,
      "loss": 2.6243302822113037,
      "step": 4031
    },
    {
      "epoch": 0.9326856349757113,
      "grad_norm": 0.9651793241500854,
      "learning_rate": 0.00012816206721818943,
      "loss": 2.407839775085449,
      "step": 4032
    },
    {
      "epoch": 0.9329169558177192,
      "grad_norm": 0.7305319905281067,
      "learning_rate": 0.000127369107960173,
      "loss": 2.432304620742798,
      "step": 4033
    },
    {
      "epoch": 0.933148276659727,
      "grad_norm": 0.9560156464576721,
      "learning_rate": 0.00012657777003309242,
      "loss": 2.653458595275879,
      "step": 4034
    },
    {
      "epoch": 0.9333795975017349,
      "grad_norm": 0.9748632311820984,
      "learning_rate": 0.00012578806389940657,
      "loss": 2.4454362392425537,
      "step": 4035
    },
    {
      "epoch": 0.9336109183437428,
      "grad_norm": 0.9382247924804688,
      "learning_rate": 0.00012500000000000006,
      "loss": 2.722867488861084,
      "step": 4036
    },
    {
      "epoch": 0.9338422391857506,
      "grad_norm": 0.9235392808914185,
      "learning_rate": 0.00012421358875404486,
      "loss": 2.5709500312805176,
      "step": 4037
    },
    {
      "epoch": 0.9340735600277585,
      "grad_norm": 0.8414825201034546,
      "learning_rate": 0.0001234288405588632,
      "loss": 2.7281150817871094,
      "step": 4038
    },
    {
      "epoch": 0.9343048808697664,
      "grad_norm": 0.8032715320587158,
      "learning_rate": 0.00012264576578978955,
      "loss": 2.3216469287872314,
      "step": 4039
    },
    {
      "epoch": 0.9345362017117742,
      "grad_norm": 0.9386684894561768,
      "learning_rate": 0.00012186437480003371,
      "loss": 2.359025239944458,
      "step": 4040
    },
    {
      "epoch": 0.9347675225537821,
      "grad_norm": 0.9633623957633972,
      "learning_rate": 0.00012108467792054372,
      "loss": 2.4444193840026855,
      "step": 4041
    },
    {
      "epoch": 0.93499884339579,
      "grad_norm": 0.8620023131370544,
      "learning_rate": 0.00012030668545986958,
      "loss": 2.19108247756958,
      "step": 4042
    },
    {
      "epoch": 0.9352301642377978,
      "grad_norm": 0.9877381324768066,
      "learning_rate": 0.00011953040770402679,
      "loss": 2.282756805419922,
      "step": 4043
    },
    {
      "epoch": 0.9354614850798056,
      "grad_norm": 0.7292525768280029,
      "learning_rate": 0.00011875585491635998,
      "loss": 2.000946521759033,
      "step": 4044
    },
    {
      "epoch": 0.9356928059218136,
      "grad_norm": 0.9307861328125,
      "learning_rate": 0.00011798303733740801,
      "loss": 2.4987363815307617,
      "step": 4045
    },
    {
      "epoch": 0.9359241267638214,
      "grad_norm": 0.7609665393829346,
      "learning_rate": 0.00011721196518476806,
      "loss": 2.1153643131256104,
      "step": 4046
    },
    {
      "epoch": 0.9361554476058292,
      "grad_norm": 0.7085083723068237,
      "learning_rate": 0.00011644264865296033,
      "loss": 1.9147148132324219,
      "step": 4047
    },
    {
      "epoch": 0.9363867684478372,
      "grad_norm": 3.81083345413208,
      "learning_rate": 0.000115675097913294,
      "loss": 2.2090201377868652,
      "step": 4048
    },
    {
      "epoch": 0.936618089289845,
      "grad_norm": 1.0917704105377197,
      "learning_rate": 0.00011490932311373218,
      "loss": 2.5292739868164062,
      "step": 4049
    },
    {
      "epoch": 0.9368494101318529,
      "grad_norm": 0.8720853328704834,
      "learning_rate": 0.00011414533437875754,
      "loss": 2.2428998947143555,
      "step": 4050
    },
    {
      "epoch": 0.9370807309738608,
      "grad_norm": 0.8865945339202881,
      "learning_rate": 0.00011338314180923916,
      "loss": 2.278707265853882,
      "step": 4051
    },
    {
      "epoch": 0.9373120518158686,
      "grad_norm": 3.662400960922241,
      "learning_rate": 0.0001126227554822985,
      "loss": 2.304396867752075,
      "step": 4052
    },
    {
      "epoch": 0.9375433726578765,
      "grad_norm": 0.9172145128250122,
      "learning_rate": 0.00011186418545117633,
      "loss": 2.3890304565429688,
      "step": 4053
    },
    {
      "epoch": 0.9377746934998843,
      "grad_norm": 0.9231246709823608,
      "learning_rate": 0.00011110744174509952,
      "loss": 2.4497199058532715,
      "step": 4054
    },
    {
      "epoch": 0.9380060143418922,
      "grad_norm": 0.7440099716186523,
      "learning_rate": 0.00011035253436914883,
      "loss": 2.290048122406006,
      "step": 4055
    },
    {
      "epoch": 0.9382373351839001,
      "grad_norm": 1.245893120765686,
      "learning_rate": 0.00010959947330412681,
      "loss": 2.2585811614990234,
      "step": 4056
    },
    {
      "epoch": 0.9384686560259079,
      "grad_norm": 0.9381008744239807,
      "learning_rate": 0.0001088482685064249,
      "loss": 2.195216655731201,
      "step": 4057
    },
    {
      "epoch": 0.9386999768679158,
      "grad_norm": 0.9765157103538513,
      "learning_rate": 0.00010809892990789286,
      "loss": 2.422373056411743,
      "step": 4058
    },
    {
      "epoch": 0.9389312977099237,
      "grad_norm": 0.7481422424316406,
      "learning_rate": 0.00010735146741570679,
      "loss": 2.580026149749756,
      "step": 4059
    },
    {
      "epoch": 0.9391626185519315,
      "grad_norm": 0.8896360397338867,
      "learning_rate": 0.00010660589091223854,
      "loss": 2.1762776374816895,
      "step": 4060
    },
    {
      "epoch": 0.9393939393939394,
      "grad_norm": 1.8389607667922974,
      "learning_rate": 0.00010586221025492462,
      "loss": 2.7603416442871094,
      "step": 4061
    },
    {
      "epoch": 0.9396252602359473,
      "grad_norm": 0.980350136756897,
      "learning_rate": 0.00010512043527613624,
      "loss": 2.3358917236328125,
      "step": 4062
    },
    {
      "epoch": 0.9398565810779551,
      "grad_norm": 0.8883504271507263,
      "learning_rate": 0.0001043805757830495,
      "loss": 2.468141555786133,
      "step": 4063
    },
    {
      "epoch": 0.9400879019199629,
      "grad_norm": 0.7999618649482727,
      "learning_rate": 0.00010364264155751487,
      "loss": 2.6478614807128906,
      "step": 4064
    },
    {
      "epoch": 0.9403192227619709,
      "grad_norm": 1.1417076587677002,
      "learning_rate": 0.00010290664235592886,
      "loss": 3.1159157752990723,
      "step": 4065
    },
    {
      "epoch": 0.9405505436039787,
      "grad_norm": 0.7648515701293945,
      "learning_rate": 0.00010217258790910447,
      "loss": 2.503805637359619,
      "step": 4066
    },
    {
      "epoch": 0.9407818644459865,
      "grad_norm": 0.7064650654792786,
      "learning_rate": 0.00010144048792214247,
      "loss": 2.3881359100341797,
      "step": 4067
    },
    {
      "epoch": 0.9410131852879945,
      "grad_norm": 0.8420358896255493,
      "learning_rate": 0.00010071035207430351,
      "loss": 2.4078726768493652,
      "step": 4068
    },
    {
      "epoch": 0.9412445061300023,
      "grad_norm": 1.1606340408325195,
      "learning_rate": 9.998219001887979e-05,
      "loss": 2.9972541332244873,
      "step": 4069
    },
    {
      "epoch": 0.9414758269720102,
      "grad_norm": 0.8325161933898926,
      "learning_rate": 9.925601138306761e-05,
      "loss": 2.4539289474487305,
      "step": 4070
    },
    {
      "epoch": 0.9417071478140181,
      "grad_norm": 0.8031669855117798,
      "learning_rate": 9.853182576784001e-05,
      "loss": 2.639286994934082,
      "step": 4071
    },
    {
      "epoch": 0.9419384686560259,
      "grad_norm": 0.9343875646591187,
      "learning_rate": 9.780964274781984e-05,
      "loss": 2.2178406715393066,
      "step": 4072
    },
    {
      "epoch": 0.9421697894980338,
      "grad_norm": 0.9195245504379272,
      "learning_rate": 9.708947187115324e-05,
      "loss": 2.400920867919922,
      "step": 4073
    },
    {
      "epoch": 0.9424011103400416,
      "grad_norm": 0.8648451566696167,
      "learning_rate": 9.637132265938314e-05,
      "loss": 2.348653793334961,
      "step": 4074
    },
    {
      "epoch": 0.9426324311820495,
      "grad_norm": 1.2188162803649902,
      "learning_rate": 9.565520460732379e-05,
      "loss": 2.4157731533050537,
      "step": 4075
    },
    {
      "epoch": 0.9428637520240574,
      "grad_norm": 1.0372720956802368,
      "learning_rate": 9.494112718293501e-05,
      "loss": 2.44008207321167,
      "step": 4076
    },
    {
      "epoch": 0.9430950728660652,
      "grad_norm": 0.7168769240379333,
      "learning_rate": 9.42290998271969e-05,
      "loss": 2.3362340927124023,
      "step": 4077
    },
    {
      "epoch": 0.9433263937080731,
      "grad_norm": 0.9248572587966919,
      "learning_rate": 9.351913195398524e-05,
      "loss": 2.3176286220550537,
      "step": 4078
    },
    {
      "epoch": 0.943557714550081,
      "grad_norm": 0.6793690323829651,
      "learning_rate": 9.281123294994684e-05,
      "loss": 2.277588367462158,
      "step": 4079
    },
    {
      "epoch": 0.9437890353920888,
      "grad_norm": 0.7206363677978516,
      "learning_rate": 9.210541217437566e-05,
      "loss": 2.792011260986328,
      "step": 4080
    },
    {
      "epoch": 0.9440203562340967,
      "grad_norm": 0.7940208315849304,
      "learning_rate": 9.140167895908866e-05,
      "loss": 2.7351574897766113,
      "step": 4081
    },
    {
      "epoch": 0.9442516770761046,
      "grad_norm": 0.7669649720191956,
      "learning_rate": 9.070004260830295e-05,
      "loss": 2.5902445316314697,
      "step": 4082
    },
    {
      "epoch": 0.9444829979181124,
      "grad_norm": 0.7665144205093384,
      "learning_rate": 9.000051239851239e-05,
      "loss": 2.220028877258301,
      "step": 4083
    },
    {
      "epoch": 0.9447143187601202,
      "grad_norm": 0.7343670725822449,
      "learning_rate": 8.930309757836516e-05,
      "loss": 2.2529077529907227,
      "step": 4084
    },
    {
      "epoch": 0.9449456396021282,
      "grad_norm": 1.0257489681243896,
      "learning_rate": 8.86078073685413e-05,
      "loss": 2.331200361251831,
      "step": 4085
    },
    {
      "epoch": 0.945176960444136,
      "grad_norm": 0.9697383046150208,
      "learning_rate": 8.791465096163093e-05,
      "loss": 2.4917960166931152,
      "step": 4086
    },
    {
      "epoch": 0.9454082812861438,
      "grad_norm": 0.7217179536819458,
      "learning_rate": 8.722363752201276e-05,
      "loss": 2.3048255443573,
      "step": 4087
    },
    {
      "epoch": 0.9456396021281518,
      "grad_norm": 0.9569346308708191,
      "learning_rate": 8.653477618573261e-05,
      "loss": 2.4198174476623535,
      "step": 4088
    },
    {
      "epoch": 0.9458709229701596,
      "grad_norm": 0.8181317448616028,
      "learning_rate": 8.584807606038306e-05,
      "loss": 2.187650203704834,
      "step": 4089
    },
    {
      "epoch": 0.9461022438121675,
      "grad_norm": 0.8610307574272156,
      "learning_rate": 8.516354622498279e-05,
      "loss": 2.615373134613037,
      "step": 4090
    },
    {
      "epoch": 0.9463335646541754,
      "grad_norm": 0.8135629892349243,
      "learning_rate": 8.448119572985655e-05,
      "loss": 2.2970709800720215,
      "step": 4091
    },
    {
      "epoch": 0.9465648854961832,
      "grad_norm": 0.7447644472122192,
      "learning_rate": 8.380103359651554e-05,
      "loss": 2.10007381439209,
      "step": 4092
    },
    {
      "epoch": 0.9467962063381911,
      "grad_norm": 0.7718713283538818,
      "learning_rate": 8.31230688175382e-05,
      "loss": 2.368683099746704,
      "step": 4093
    },
    {
      "epoch": 0.9470275271801989,
      "grad_norm": 1.0164947509765625,
      "learning_rate": 8.244731035645106e-05,
      "loss": 2.114068031311035,
      "step": 4094
    },
    {
      "epoch": 0.9472588480222068,
      "grad_norm": 1.0296436548233032,
      "learning_rate": 8.177376714761062e-05,
      "loss": 2.5680243968963623,
      "step": 4095
    },
    {
      "epoch": 0.9474901688642147,
      "grad_norm": 1.224376916885376,
      "learning_rate": 8.110244809608495e-05,
      "loss": 2.554395914077759,
      "step": 4096
    },
    {
      "epoch": 0.9477214897062225,
      "grad_norm": 0.8962958455085754,
      "learning_rate": 8.043336207753601e-05,
      "loss": 2.429523229598999,
      "step": 4097
    },
    {
      "epoch": 0.9479528105482304,
      "grad_norm": 0.9987344741821289,
      "learning_rate": 7.976651793810235e-05,
      "loss": 2.55836820602417,
      "step": 4098
    },
    {
      "epoch": 0.9481841313902383,
      "grad_norm": 0.7812055945396423,
      "learning_rate": 7.910192449428217e-05,
      "loss": 2.2010984420776367,
      "step": 4099
    },
    {
      "epoch": 0.9484154522322461,
      "grad_norm": 1.0886379480361938,
      "learning_rate": 7.843959053281663e-05,
      "loss": 2.402850866317749,
      "step": 4100
    },
    {
      "epoch": 0.9486467730742539,
      "grad_norm": 0.7315024137496948,
      "learning_rate": 7.777952481057374e-05,
      "loss": 2.34781551361084,
      "step": 4101
    },
    {
      "epoch": 0.9488780939162619,
      "grad_norm": 0.720155656337738,
      "learning_rate": 7.712173605443268e-05,
      "loss": 2.4006311893463135,
      "step": 4102
    },
    {
      "epoch": 0.9491094147582697,
      "grad_norm": 0.6706374287605286,
      "learning_rate": 7.646623296116839e-05,
      "loss": 2.3750829696655273,
      "step": 4103
    },
    {
      "epoch": 0.9493407356002775,
      "grad_norm": 0.913470983505249,
      "learning_rate": 7.581302419733633e-05,
      "loss": 2.6489131450653076,
      "step": 4104
    },
    {
      "epoch": 0.9495720564422855,
      "grad_norm": 1.013396143913269,
      "learning_rate": 7.51621183991582e-05,
      "loss": 1.7769968509674072,
      "step": 4105
    },
    {
      "epoch": 0.9498033772842933,
      "grad_norm": 0.85032719373703,
      "learning_rate": 7.451352417240792e-05,
      "loss": 2.250688314437866,
      "step": 4106
    },
    {
      "epoch": 0.9500346981263011,
      "grad_norm": 0.8803473711013794,
      "learning_rate": 7.386725009229739e-05,
      "loss": 2.4888553619384766,
      "step": 4107
    },
    {
      "epoch": 0.9502660189683091,
      "grad_norm": 0.8757585287094116,
      "learning_rate": 7.322330470336314e-05,
      "loss": 2.1223223209381104,
      "step": 4108
    },
    {
      "epoch": 0.9504973398103169,
      "grad_norm": 0.9476198554039001,
      "learning_rate": 7.258169651935382e-05,
      "loss": 2.506072998046875,
      "step": 4109
    },
    {
      "epoch": 0.9507286606523248,
      "grad_norm": 2.9364073276519775,
      "learning_rate": 7.194243402311732e-05,
      "loss": 1.9803956747055054,
      "step": 4110
    },
    {
      "epoch": 0.9509599814943326,
      "grad_norm": 0.7409396767616272,
      "learning_rate": 7.130552566648846e-05,
      "loss": 2.195298671722412,
      "step": 4111
    },
    {
      "epoch": 0.9511913023363405,
      "grad_norm": 0.8729414343833923,
      "learning_rate": 7.067097987017762e-05,
      "loss": 2.612546443939209,
      "step": 4112
    },
    {
      "epoch": 0.9514226231783484,
      "grad_norm": 0.8230058550834656,
      "learning_rate": 7.003880502365933e-05,
      "loss": 1.9644180536270142,
      "step": 4113
    },
    {
      "epoch": 0.9516539440203562,
      "grad_norm": 1.0246609449386597,
      "learning_rate": 6.940900948506112e-05,
      "loss": 2.7277374267578125,
      "step": 4114
    },
    {
      "epoch": 0.9518852648623641,
      "grad_norm": 0.8876630067825317,
      "learning_rate": 6.878160158105304e-05,
      "loss": 2.281507968902588,
      "step": 4115
    },
    {
      "epoch": 0.952116585704372,
      "grad_norm": 0.9030771255493164,
      "learning_rate": 6.815658960673781e-05,
      "loss": 2.361034870147705,
      "step": 4116
    },
    {
      "epoch": 0.9523479065463798,
      "grad_norm": 0.795659065246582,
      "learning_rate": 6.753398182554116e-05,
      "loss": 2.39139986038208,
      "step": 4117
    },
    {
      "epoch": 0.9525792273883877,
      "grad_norm": 0.7976886630058289,
      "learning_rate": 6.691378646910204e-05,
      "loss": 2.0600383281707764,
      "step": 4118
    },
    {
      "epoch": 0.9528105482303956,
      "grad_norm": 0.9215635657310486,
      "learning_rate": 6.629601173716453e-05,
      "loss": 2.3904991149902344,
      "step": 4119
    },
    {
      "epoch": 0.9530418690724034,
      "grad_norm": 0.7536221146583557,
      "learning_rate": 6.5680665797469e-05,
      "loss": 2.122069835662842,
      "step": 4120
    },
    {
      "epoch": 0.9532731899144112,
      "grad_norm": 0.840613842010498,
      "learning_rate": 6.506775678564417e-05,
      "loss": 1.8680070638656616,
      "step": 4121
    },
    {
      "epoch": 0.9535045107564192,
      "grad_norm": 0.8091286420822144,
      "learning_rate": 6.445729280509957e-05,
      "loss": 2.228231906890869,
      "step": 4122
    },
    {
      "epoch": 0.953735831598427,
      "grad_norm": 0.9140711426734924,
      "learning_rate": 6.384928192691843e-05,
      "loss": 2.563887596130371,
      "step": 4123
    },
    {
      "epoch": 0.9539671524404348,
      "grad_norm": 0.8048566579818726,
      "learning_rate": 6.324373218975104e-05,
      "loss": 2.294001579284668,
      "step": 4124
    },
    {
      "epoch": 0.9541984732824428,
      "grad_norm": 1.051630973815918,
      "learning_rate": 6.264065159970819e-05,
      "loss": 2.30342960357666,
      "step": 4125
    },
    {
      "epoch": 0.9544297941244506,
      "grad_norm": 0.6677807569503784,
      "learning_rate": 6.204004813025568e-05,
      "loss": 2.1563708782196045,
      "step": 4126
    },
    {
      "epoch": 0.9546611149664584,
      "grad_norm": 0.9236207008361816,
      "learning_rate": 6.144192972210869e-05,
      "loss": 2.507361888885498,
      "step": 4127
    },
    {
      "epoch": 0.9548924358084664,
      "grad_norm": 0.9623232483863831,
      "learning_rate": 6.084630428312679e-05,
      "loss": 2.7207088470458984,
      "step": 4128
    },
    {
      "epoch": 0.9551237566504742,
      "grad_norm": 0.9438007473945618,
      "learning_rate": 6.025317968820954e-05,
      "loss": 2.2299013137817383,
      "step": 4129
    },
    {
      "epoch": 0.955355077492482,
      "grad_norm": 0.8515530228614807,
      "learning_rate": 5.966256377919224e-05,
      "loss": 2.2933411598205566,
      "step": 4130
    },
    {
      "epoch": 0.9555863983344899,
      "grad_norm": 0.6987287998199463,
      "learning_rate": 5.9074464364742145e-05,
      "loss": 2.5261340141296387,
      "step": 4131
    },
    {
      "epoch": 0.9558177191764978,
      "grad_norm": 0.6604101061820984,
      "learning_rate": 5.848888922025553e-05,
      "loss": 2.320600986480713,
      "step": 4132
    },
    {
      "epoch": 0.9560490400185057,
      "grad_norm": 0.7544179558753967,
      "learning_rate": 5.790584608775465e-05,
      "loss": 2.1871495246887207,
      "step": 4133
    },
    {
      "epoch": 0.9562803608605135,
      "grad_norm": 0.9740626811981201,
      "learning_rate": 5.7325342675785484e-05,
      "loss": 2.338984966278076,
      "step": 4134
    },
    {
      "epoch": 0.9565116817025214,
      "grad_norm": 0.7599971890449524,
      "learning_rate": 5.6747386659315755e-05,
      "loss": 2.3877079486846924,
      "step": 4135
    },
    {
      "epoch": 0.9567430025445293,
      "grad_norm": 0.9683665037155151,
      "learning_rate": 5.6171985679633524e-05,
      "loss": 2.213235378265381,
      "step": 4136
    },
    {
      "epoch": 0.9569743233865371,
      "grad_norm": 0.7105645537376404,
      "learning_rate": 5.559914734424617e-05,
      "loss": 2.285961627960205,
      "step": 4137
    },
    {
      "epoch": 0.957205644228545,
      "grad_norm": 1.2386107444763184,
      "learning_rate": 5.50288792267796e-05,
      "loss": 2.55430269241333,
      "step": 4138
    },
    {
      "epoch": 0.9574369650705529,
      "grad_norm": 0.6752905249595642,
      "learning_rate": 5.4461188866878394e-05,
      "loss": 2.204637289047241,
      "step": 4139
    },
    {
      "epoch": 0.9576682859125607,
      "grad_norm": 1.0952448844909668,
      "learning_rate": 5.389608377010607e-05,
      "loss": 2.2518205642700195,
      "step": 4140
    },
    {
      "epoch": 0.9578996067545685,
      "grad_norm": 0.9334809184074402,
      "learning_rate": 5.3333571407845755e-05,
      "loss": 2.336655855178833,
      "step": 4141
    },
    {
      "epoch": 0.9581309275965765,
      "grad_norm": 0.7201266884803772,
      "learning_rate": 5.277365921720137e-05,
      "loss": 2.1424999237060547,
      "step": 4142
    },
    {
      "epoch": 0.9583622484385843,
      "grad_norm": 0.9409101605415344,
      "learning_rate": 5.2216354600899435e-05,
      "loss": 2.4987893104553223,
      "step": 4143
    },
    {
      "epoch": 0.9585935692805921,
      "grad_norm": 1.0663825273513794,
      "learning_rate": 5.1661664927191235e-05,
      "loss": 2.239995002746582,
      "step": 4144
    },
    {
      "epoch": 0.9588248901226001,
      "grad_norm": 0.7853975296020508,
      "learning_rate": 5.1109597529755054e-05,
      "loss": 2.185595989227295,
      "step": 4145
    },
    {
      "epoch": 0.9590562109646079,
      "grad_norm": 0.7104918956756592,
      "learning_rate": 5.056015970759967e-05,
      "loss": 2.3529515266418457,
      "step": 4146
    },
    {
      "epoch": 0.9592875318066157,
      "grad_norm": 0.8236262202262878,
      "learning_rate": 5.0013358724967595e-05,
      "loss": 2.376253128051758,
      "step": 4147
    },
    {
      "epoch": 0.9595188526486237,
      "grad_norm": 0.740924060344696,
      "learning_rate": 4.946920181123904e-05,
      "loss": 2.30171799659729,
      "step": 4148
    },
    {
      "epoch": 0.9597501734906315,
      "grad_norm": 0.9623746871948242,
      "learning_rate": 4.892769616083648e-05,
      "loss": 2.211210012435913,
      "step": 4149
    },
    {
      "epoch": 0.9599814943326394,
      "grad_norm": 0.6152719855308533,
      "learning_rate": 4.838884893312934e-05,
      "loss": 2.323676347732544,
      "step": 4150
    },
    {
      "epoch": 0.9602128151746472,
      "grad_norm": 0.9753477573394775,
      "learning_rate": 4.7852667252339524e-05,
      "loss": 2.2656941413879395,
      "step": 4151
    },
    {
      "epoch": 0.9604441360166551,
      "grad_norm": 0.8853805661201477,
      "learning_rate": 4.731915820744695e-05,
      "loss": 2.3282275199890137,
      "step": 4152
    },
    {
      "epoch": 0.960675456858663,
      "grad_norm": 1.679705262184143,
      "learning_rate": 4.6788328852096216e-05,
      "loss": 2.4778807163238525,
      "step": 4153
    },
    {
      "epoch": 0.9609067777006708,
      "grad_norm": 0.7638845443725586,
      "learning_rate": 4.626018620450309e-05,
      "loss": 2.361994504928589,
      "step": 4154
    },
    {
      "epoch": 0.9611380985426787,
      "grad_norm": 0.8451151847839355,
      "learning_rate": 4.573473724736155e-05,
      "loss": 2.299651622772217,
      "step": 4155
    },
    {
      "epoch": 0.9613694193846866,
      "grad_norm": 0.7118843197822571,
      "learning_rate": 4.521198892775202e-05,
      "loss": 2.4400827884674072,
      "step": 4156
    },
    {
      "epoch": 0.9616007402266944,
      "grad_norm": 0.757322371006012,
      "learning_rate": 4.469194815704908e-05,
      "loss": 2.712153434753418,
      "step": 4157
    },
    {
      "epoch": 0.9618320610687023,
      "grad_norm": 0.7977062463760376,
      "learning_rate": 4.417462181082996e-05,
      "loss": 2.11409592628479,
      "step": 4158
    },
    {
      "epoch": 0.9620633819107102,
      "grad_norm": 1.182490587234497,
      "learning_rate": 4.366001672878406e-05,
      "loss": 2.301971912384033,
      "step": 4159
    },
    {
      "epoch": 0.962294702752718,
      "grad_norm": 0.8897682428359985,
      "learning_rate": 4.3148139714622366e-05,
      "loss": 2.143364906311035,
      "step": 4160
    },
    {
      "epoch": 0.9625260235947258,
      "grad_norm": 0.8320556282997131,
      "learning_rate": 4.2638997535987434e-05,
      "loss": 2.204352855682373,
      "step": 4161
    },
    {
      "epoch": 0.9627573444367338,
      "grad_norm": 0.8832243084907532,
      "learning_rate": 4.213259692436367e-05,
      "loss": 2.430255651473999,
      "step": 4162
    },
    {
      "epoch": 0.9629886652787416,
      "grad_norm": 0.8962817788124084,
      "learning_rate": 4.1628944574989056e-05,
      "loss": 2.525580406188965,
      "step": 4163
    },
    {
      "epoch": 0.9632199861207494,
      "grad_norm": 1.0579367876052856,
      "learning_rate": 4.112804714676593e-05,
      "loss": 2.377018690109253,
      "step": 4164
    },
    {
      "epoch": 0.9634513069627574,
      "grad_norm": 0.7979461550712585,
      "learning_rate": 4.0629911262173054e-05,
      "loss": 2.5037732124328613,
      "step": 4165
    },
    {
      "epoch": 0.9636826278047652,
      "grad_norm": 0.8939361572265625,
      "learning_rate": 4.01345435071784e-05,
      "loss": 2.967282295227051,
      "step": 4166
    },
    {
      "epoch": 0.963913948646773,
      "grad_norm": 0.7626407146453857,
      "learning_rate": 3.9641950431151873e-05,
      "loss": 2.1003479957580566,
      "step": 4167
    },
    {
      "epoch": 0.964145269488781,
      "grad_norm": 0.7976487278938293,
      "learning_rate": 3.9152138546778624e-05,
      "loss": 2.5762228965759277,
      "step": 4168
    },
    {
      "epoch": 0.9643765903307888,
      "grad_norm": 0.8557921051979065,
      "learning_rate": 3.8665114329972996e-05,
      "loss": 2.0963332653045654,
      "step": 4169
    },
    {
      "epoch": 0.9646079111727967,
      "grad_norm": 0.7759171724319458,
      "learning_rate": 3.818088421979299e-05,
      "loss": 2.4680843353271484,
      "step": 4170
    },
    {
      "epoch": 0.9648392320148045,
      "grad_norm": 0.9022979736328125,
      "learning_rate": 3.76994546183553e-05,
      "loss": 2.373114824295044,
      "step": 4171
    },
    {
      "epoch": 0.9650705528568124,
      "grad_norm": 0.8555135130882263,
      "learning_rate": 3.7220831890750065e-05,
      "loss": 2.2365527153015137,
      "step": 4172
    },
    {
      "epoch": 0.9653018736988203,
      "grad_norm": 0.9314596056938171,
      "learning_rate": 3.6745022364957415e-05,
      "loss": 2.364607572555542,
      "step": 4173
    },
    {
      "epoch": 0.9655331945408281,
      "grad_norm": 0.7121050953865051,
      "learning_rate": 3.627203233176341e-05,
      "loss": 2.2487282752990723,
      "step": 4174
    },
    {
      "epoch": 0.965764515382836,
      "grad_norm": 0.7387624979019165,
      "learning_rate": 3.5801868044676806e-05,
      "loss": 2.4037601947784424,
      "step": 4175
    },
    {
      "epoch": 0.9659958362248439,
      "grad_norm": 0.704627275466919,
      "learning_rate": 3.5334535719846764e-05,
      "loss": 2.4436120986938477,
      "step": 4176
    },
    {
      "epoch": 0.9662271570668517,
      "grad_norm": 0.9233609437942505,
      "learning_rate": 3.487004153598028e-05,
      "loss": 2.633993625640869,
      "step": 4177
    },
    {
      "epoch": 0.9664584779088596,
      "grad_norm": 0.7196712493896484,
      "learning_rate": 3.4408391634260686e-05,
      "loss": 2.480276107788086,
      "step": 4178
    },
    {
      "epoch": 0.9666897987508675,
      "grad_norm": 1.0461463928222656,
      "learning_rate": 3.394959211826643e-05,
      "loss": 2.3589494228363037,
      "step": 4179
    },
    {
      "epoch": 0.9669211195928753,
      "grad_norm": 0.8623855113983154,
      "learning_rate": 3.3493649053890325e-05,
      "loss": 2.9438228607177734,
      "step": 4180
    },
    {
      "epoch": 0.9671524404348831,
      "grad_norm": 0.7894553542137146,
      "learning_rate": 3.304056846925949e-05,
      "loss": 2.306276321411133,
      "step": 4181
    },
    {
      "epoch": 0.9673837612768911,
      "grad_norm": 1.0070229768753052,
      "learning_rate": 3.2590356354655296e-05,
      "loss": 2.6192026138305664,
      "step": 4182
    },
    {
      "epoch": 0.9676150821188989,
      "grad_norm": 0.9493356943130493,
      "learning_rate": 3.214301866243469e-05,
      "loss": 2.3622827529907227,
      "step": 4183
    },
    {
      "epoch": 0.9678464029609067,
      "grad_norm": 0.7383671998977661,
      "learning_rate": 3.169856130695106e-05,
      "loss": 2.2197887897491455,
      "step": 4184
    },
    {
      "epoch": 0.9680777238029147,
      "grad_norm": 0.8183758854866028,
      "learning_rate": 3.125699016447634e-05,
      "loss": 2.670337677001953,
      "step": 4185
    },
    {
      "epoch": 0.9683090446449225,
      "grad_norm": 0.6817243695259094,
      "learning_rate": 3.081831107312308e-05,
      "loss": 2.400676965713501,
      "step": 4186
    },
    {
      "epoch": 0.9685403654869303,
      "grad_norm": 0.6524887681007385,
      "learning_rate": 3.0382529832767376e-05,
      "loss": 2.2439403533935547,
      "step": 4187
    },
    {
      "epoch": 0.9687716863289383,
      "grad_norm": 0.6334730386734009,
      "learning_rate": 2.9949652204972254e-05,
      "loss": 2.190138816833496,
      "step": 4188
    },
    {
      "epoch": 0.9690030071709461,
      "grad_norm": 0.8725318908691406,
      "learning_rate": 2.9519683912911265e-05,
      "loss": 2.213852882385254,
      "step": 4189
    },
    {
      "epoch": 0.969234328012954,
      "grad_norm": 0.9005917906761169,
      "learning_rate": 2.9092630641293112e-05,
      "loss": 2.265376567840576,
      "step": 4190
    },
    {
      "epoch": 0.9694656488549618,
      "grad_norm": 0.8520849943161011,
      "learning_rate": 2.8668498036286322e-05,
      "loss": 2.1351089477539062,
      "step": 4191
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.8092983961105347,
      "learning_rate": 2.824729170544457e-05,
      "loss": 2.089890956878662,
      "step": 4192
    },
    {
      "epoch": 0.9699282905389776,
      "grad_norm": 0.9643399715423584,
      "learning_rate": 2.7829017217632645e-05,
      "loss": 2.450809955596924,
      "step": 4193
    },
    {
      "epoch": 0.9701596113809854,
      "grad_norm": 0.8496055603027344,
      "learning_rate": 2.7413680102952755e-05,
      "loss": 2.3418097496032715,
      "step": 4194
    },
    {
      "epoch": 0.9703909322229933,
      "grad_norm": 0.8874192237854004,
      "learning_rate": 2.7001285852671477e-05,
      "loss": 2.3160927295684814,
      "step": 4195
    },
    {
      "epoch": 0.9706222530650012,
      "grad_norm": 0.8462907075881958,
      "learning_rate": 2.659183991914696e-05,
      "loss": 2.5355982780456543,
      "step": 4196
    },
    {
      "epoch": 0.970853573907009,
      "grad_norm": 0.920271098613739,
      "learning_rate": 2.6185347715757118e-05,
      "loss": 2.4476330280303955,
      "step": 4197
    },
    {
      "epoch": 0.9710848947490169,
      "grad_norm": 0.9233880639076233,
      "learning_rate": 2.5781814616827938e-05,
      "loss": 2.363659620285034,
      "step": 4198
    },
    {
      "epoch": 0.9713162155910248,
      "grad_norm": 4.568171977996826,
      "learning_rate": 2.5381245957562304e-05,
      "loss": 2.232478141784668,
      "step": 4199
    },
    {
      "epoch": 0.9715475364330326,
      "grad_norm": 0.9014209508895874,
      "learning_rate": 2.4983647033969715e-05,
      "loss": 2.255443572998047,
      "step": 4200
    },
    {
      "epoch": 0.9717788572750404,
      "grad_norm": 1.3869826793670654,
      "learning_rate": 2.458902310279601e-05,
      "loss": 2.4460763931274414,
      "step": 4201
    },
    {
      "epoch": 0.9720101781170484,
      "grad_norm": 0.903958261013031,
      "learning_rate": 2.4197379381453945e-05,
      "loss": 2.640998125076294,
      "step": 4202
    },
    {
      "epoch": 0.9722414989590562,
      "grad_norm": 0.9297658801078796,
      "learning_rate": 2.380872104795434e-05,
      "loss": 2.0307817459106445,
      "step": 4203
    },
    {
      "epoch": 0.972472819801064,
      "grad_norm": 0.9077376127243042,
      "learning_rate": 2.3423053240837516e-05,
      "loss": 2.607525110244751,
      "step": 4204
    },
    {
      "epoch": 0.972704140643072,
      "grad_norm": 0.8507153987884521,
      "learning_rate": 2.3040381059105304e-05,
      "loss": 2.392317295074463,
      "step": 4205
    },
    {
      "epoch": 0.9729354614850798,
      "grad_norm": 0.6638129353523254,
      "learning_rate": 2.2660709562153713e-05,
      "loss": 2.329800605773926,
      "step": 4206
    },
    {
      "epoch": 0.9731667823270876,
      "grad_norm": 0.923516571521759,
      "learning_rate": 2.2284043769706025e-05,
      "loss": 2.363880157470703,
      "step": 4207
    },
    {
      "epoch": 0.9733981031690956,
      "grad_norm": 0.6924377679824829,
      "learning_rate": 2.1910388661746496e-05,
      "loss": 2.246037483215332,
      "step": 4208
    },
    {
      "epoch": 0.9736294240111034,
      "grad_norm": 0.9507063031196594,
      "learning_rate": 2.153974917845422e-05,
      "loss": 2.399287223815918,
      "step": 4209
    },
    {
      "epoch": 0.9738607448531112,
      "grad_norm": 0.946589469909668,
      "learning_rate": 2.1172130220138224e-05,
      "loss": 2.340837001800537,
      "step": 4210
    },
    {
      "epoch": 0.9740920656951191,
      "grad_norm": 0.7605183720588684,
      "learning_rate": 2.0807536647172496e-05,
      "loss": 2.105867624282837,
      "step": 4211
    },
    {
      "epoch": 0.974323386537127,
      "grad_norm": 0.896282434463501,
      "learning_rate": 2.0445973279931528e-05,
      "loss": 2.4794087409973145,
      "step": 4212
    },
    {
      "epoch": 0.9745547073791349,
      "grad_norm": 0.7852051854133606,
      "learning_rate": 2.0087444898726935e-05,
      "loss": 2.4791226387023926,
      "step": 4213
    },
    {
      "epoch": 0.9747860282211427,
      "grad_norm": 0.957915723323822,
      "learning_rate": 1.9731956243744105e-05,
      "loss": 2.3756301403045654,
      "step": 4214
    },
    {
      "epoch": 0.9750173490631506,
      "grad_norm": 0.8437198996543884,
      "learning_rate": 1.937951201497948e-05,
      "loss": 2.479703664779663,
      "step": 4215
    },
    {
      "epoch": 0.9752486699051585,
      "grad_norm": 0.7148351073265076,
      "learning_rate": 1.9030116872178316e-05,
      "loss": 2.415459632873535,
      "step": 4216
    },
    {
      "epoch": 0.9754799907471663,
      "grad_norm": 0.8804987668991089,
      "learning_rate": 1.8683775434773396e-05,
      "loss": 2.5909199714660645,
      "step": 4217
    },
    {
      "epoch": 0.9757113115891742,
      "grad_norm": 0.8357923030853271,
      "learning_rate": 1.8340492281823718e-05,
      "loss": 2.515697956085205,
      "step": 4218
    },
    {
      "epoch": 0.9759426324311821,
      "grad_norm": 0.7137091159820557,
      "learning_rate": 1.800027195195389e-05,
      "loss": 2.4525909423828125,
      "step": 4219
    },
    {
      "epoch": 0.9761739532731899,
      "grad_norm": 0.7927257418632507,
      "learning_rate": 1.7663118943294365e-05,
      "loss": 2.2859623432159424,
      "step": 4220
    },
    {
      "epoch": 0.9764052741151977,
      "grad_norm": 0.8761951327323914,
      "learning_rate": 1.732903771342187e-05,
      "loss": 2.1955316066741943,
      "step": 4221
    },
    {
      "epoch": 0.9766365949572057,
      "grad_norm": 0.759030818939209,
      "learning_rate": 1.699803267930039e-05,
      "loss": 2.1300365924835205,
      "step": 4222
    },
    {
      "epoch": 0.9768679157992135,
      "grad_norm": 0.8931939005851746,
      "learning_rate": 1.6670108217222728e-05,
      "loss": 2.6935412883758545,
      "step": 4223
    },
    {
      "epoch": 0.9770992366412213,
      "grad_norm": 0.6942476034164429,
      "learning_rate": 1.6345268662752903e-05,
      "loss": 2.3853936195373535,
      "step": 4224
    },
    {
      "epoch": 0.9773305574832293,
      "grad_norm": 0.751437783241272,
      "learning_rate": 1.6023518310668618e-05,
      "loss": 2.2011780738830566,
      "step": 4225
    },
    {
      "epoch": 0.9775618783252371,
      "grad_norm": 0.9451789259910583,
      "learning_rate": 1.570486141490443e-05,
      "loss": 2.4572877883911133,
      "step": 4226
    },
    {
      "epoch": 0.9777931991672449,
      "grad_norm": 0.9787606596946716,
      "learning_rate": 1.538930218849574e-05,
      "loss": 2.5382704734802246,
      "step": 4227
    },
    {
      "epoch": 0.9780245200092529,
      "grad_norm": 0.8224256634712219,
      "learning_rate": 1.5076844803522921e-05,
      "loss": 2.667659044265747,
      "step": 4228
    },
    {
      "epoch": 0.9782558408512607,
      "grad_norm": 0.6807846426963806,
      "learning_rate": 1.4767493391056153e-05,
      "loss": 2.2203445434570312,
      "step": 4229
    },
    {
      "epoch": 0.9784871616932685,
      "grad_norm": 0.7475769519805908,
      "learning_rate": 1.4461252041100882e-05,
      "loss": 2.2921738624572754,
      "step": 4230
    },
    {
      "epoch": 0.9787184825352764,
      "grad_norm": 0.807290256023407,
      "learning_rate": 1.4158124802543692e-05,
      "loss": 2.464876651763916,
      "step": 4231
    },
    {
      "epoch": 0.9789498033772843,
      "grad_norm": 0.8983584046363831,
      "learning_rate": 1.3858115683098832e-05,
      "loss": 2.590867519378662,
      "step": 4232
    },
    {
      "epoch": 0.9791811242192922,
      "grad_norm": 1.331429362297058,
      "learning_rate": 1.3561228649255131e-05,
      "loss": 2.217478036880493,
      "step": 4233
    },
    {
      "epoch": 0.9794124450613,
      "grad_norm": 0.8651931285858154,
      "learning_rate": 1.3267467626223605e-05,
      "loss": 2.3184807300567627,
      "step": 4234
    },
    {
      "epoch": 0.9796437659033079,
      "grad_norm": 0.8260993957519531,
      "learning_rate": 1.297683649788564e-05,
      "loss": 2.3451762199401855,
      "step": 4235
    },
    {
      "epoch": 0.9798750867453158,
      "grad_norm": 0.7564086318016052,
      "learning_rate": 1.2689339106741526e-05,
      "loss": 2.8409130573272705,
      "step": 4236
    },
    {
      "epoch": 0.9801064075873236,
      "grad_norm": 0.7930788397789001,
      "learning_rate": 1.2404979253859721e-05,
      "loss": 2.3696208000183105,
      "step": 4237
    },
    {
      "epoch": 0.9803377284293315,
      "grad_norm": 0.9928672909736633,
      "learning_rate": 1.2123760698826591e-05,
      "loss": 2.3189029693603516,
      "step": 4238
    },
    {
      "epoch": 0.9805690492713394,
      "grad_norm": 0.8837865591049194,
      "learning_rate": 1.1845687159696583e-05,
      "loss": 2.266897678375244,
      "step": 4239
    },
    {
      "epoch": 0.9808003701133472,
      "grad_norm": 0.6769686341285706,
      "learning_rate": 1.1570762312943296e-05,
      "loss": 2.0082318782806396,
      "step": 4240
    },
    {
      "epoch": 0.981031690955355,
      "grad_norm": 0.8064167499542236,
      "learning_rate": 1.129898979341068e-05,
      "loss": 2.5744123458862305,
      "step": 4241
    },
    {
      "epoch": 0.981263011797363,
      "grad_norm": 0.7360134720802307,
      "learning_rate": 1.1030373194265114e-05,
      "loss": 2.296440362930298,
      "step": 4242
    },
    {
      "epoch": 0.9814943326393708,
      "grad_norm": 1.0094934701919556,
      "learning_rate": 1.0764916066947795e-05,
      "loss": 2.353908061981201,
      "step": 4243
    },
    {
      "epoch": 0.9817256534813786,
      "grad_norm": 1.0301473140716553,
      "learning_rate": 1.0502621921127775e-05,
      "loss": 2.5001425743103027,
      "step": 4244
    },
    {
      "epoch": 0.9819569743233866,
      "grad_norm": 0.6950082778930664,
      "learning_rate": 1.0243494224655731e-05,
      "loss": 2.2122654914855957,
      "step": 4245
    },
    {
      "epoch": 0.9821882951653944,
      "grad_norm": 0.861717939376831,
      "learning_rate": 9.98753640351785e-06,
      "loss": 2.3627233505249023,
      "step": 4246
    },
    {
      "epoch": 0.9824196160074022,
      "grad_norm": 0.8476107120513916,
      "learning_rate": 9.734751841790762e-06,
      "loss": 2.4005866050720215,
      "step": 4247
    },
    {
      "epoch": 0.9826509368494102,
      "grad_norm": 0.7861638069152832,
      "learning_rate": 9.485143881596714e-06,
      "loss": 2.2328124046325684,
      "step": 4248
    },
    {
      "epoch": 0.982882257691418,
      "grad_norm": 0.8445808291435242,
      "learning_rate": 9.238715823059324e-06,
      "loss": 2.4838991165161133,
      "step": 4249
    },
    {
      "epoch": 0.9831135785334258,
      "grad_norm": 1.2544138431549072,
      "learning_rate": 8.99547092426009e-06,
      "loss": 2.322256088256836,
      "step": 4250
    },
    {
      "epoch": 0.9833448993754337,
      "grad_norm": 0.7479559183120728,
      "learning_rate": 8.755412401195179e-06,
      "loss": 2.2267751693725586,
      "step": 4251
    },
    {
      "epoch": 0.9835762202174416,
      "grad_norm": 0.9194466471672058,
      "learning_rate": 8.51854342773295e-06,
      "loss": 2.6660993099212646,
      "step": 4252
    },
    {
      "epoch": 0.9838075410594495,
      "grad_norm": 0.8102987408638,
      "learning_rate": 8.284867135572e-06,
      "loss": 2.058830738067627,
      "step": 4253
    },
    {
      "epoch": 0.9840388619014573,
      "grad_norm": 0.9076871275901794,
      "learning_rate": 8.05438661419977e-06,
      "loss": 2.4634838104248047,
      "step": 4254
    },
    {
      "epoch": 0.9842701827434652,
      "grad_norm": 1.02077317237854,
      "learning_rate": 7.827104910851729e-06,
      "loss": 2.7999916076660156,
      "step": 4255
    },
    {
      "epoch": 0.9845015035854731,
      "grad_norm": 0.6663041710853577,
      "learning_rate": 7.6030250304710006e-06,
      "loss": 1.996620774269104,
      "step": 4256
    },
    {
      "epoch": 0.9847328244274809,
      "grad_norm": 0.7432764768600464,
      "learning_rate": 7.382149935668658e-06,
      "loss": 2.2123098373413086,
      "step": 4257
    },
    {
      "epoch": 0.9849641452694888,
      "grad_norm": 1.26542067527771,
      "learning_rate": 7.164482546684642e-06,
      "loss": 2.2879061698913574,
      "step": 4258
    },
    {
      "epoch": 0.9851954661114967,
      "grad_norm": 0.9733278751373291,
      "learning_rate": 6.9500257413490365e-06,
      "loss": 2.2722835540771484,
      "step": 4259
    },
    {
      "epoch": 0.9854267869535045,
      "grad_norm": 1.0605707168579102,
      "learning_rate": 6.7387823550440485e-06,
      "loss": 2.40266489982605,
      "step": 4260
    },
    {
      "epoch": 0.9856581077955123,
      "grad_norm": 0.7988846302032471,
      "learning_rate": 6.530755180666592e-06,
      "loss": 2.504274368286133,
      "step": 4261
    },
    {
      "epoch": 0.9858894286375203,
      "grad_norm": 0.766684353351593,
      "learning_rate": 6.325946968591317e-06,
      "loss": 2.6567397117614746,
      "step": 4262
    },
    {
      "epoch": 0.9861207494795281,
      "grad_norm": 2.0143814086914062,
      "learning_rate": 6.124360426634112e-06,
      "loss": 2.4405412673950195,
      "step": 4263
    },
    {
      "epoch": 0.9863520703215359,
      "grad_norm": 0.6805597543716431,
      "learning_rate": 5.92599822001666e-06,
      "loss": 2.2129037380218506,
      "step": 4264
    },
    {
      "epoch": 0.9865833911635439,
      "grad_norm": 0.8128846883773804,
      "learning_rate": 5.730862971330769e-06,
      "loss": 2.2864432334899902,
      "step": 4265
    },
    {
      "epoch": 0.9868147120055517,
      "grad_norm": 0.9104170799255371,
      "learning_rate": 5.538957260503991e-06,
      "loss": 2.4896178245544434,
      "step": 4266
    },
    {
      "epoch": 0.9870460328475595,
      "grad_norm": 0.7916812300682068,
      "learning_rate": 5.350283624765417e-06,
      "loss": 2.5869486331939697,
      "step": 4267
    },
    {
      "epoch": 0.9872773536895675,
      "grad_norm": 0.789238452911377,
      "learning_rate": 5.1648445586121305e-06,
      "loss": 2.3704261779785156,
      "step": 4268
    },
    {
      "epoch": 0.9875086745315753,
      "grad_norm": 0.8166057467460632,
      "learning_rate": 4.982642513776225e-06,
      "loss": 2.203986644744873,
      "step": 4269
    },
    {
      "epoch": 0.9877399953735831,
      "grad_norm": 0.6872448325157166,
      "learning_rate": 4.803679899192393e-06,
      "loss": 2.2826523780822754,
      "step": 4270
    },
    {
      "epoch": 0.987971316215591,
      "grad_norm": 0.8511764407157898,
      "learning_rate": 4.62795908096611e-06,
      "loss": 2.2907159328460693,
      "step": 4271
    },
    {
      "epoch": 0.9882026370575989,
      "grad_norm": 0.8062317371368408,
      "learning_rate": 4.455482382342335e-06,
      "loss": 2.449483871459961,
      "step": 4272
    },
    {
      "epoch": 0.9884339578996068,
      "grad_norm": 0.8892062306404114,
      "learning_rate": 4.286252083674724e-06,
      "loss": 2.4705843925476074,
      "step": 4273
    },
    {
      "epoch": 0.9886652787416146,
      "grad_norm": 0.9320504665374756,
      "learning_rate": 4.120270422395573e-06,
      "loss": 2.3501529693603516,
      "step": 4274
    },
    {
      "epoch": 0.9888965995836225,
      "grad_norm": 0.9389896988868713,
      "learning_rate": 3.957539592986198e-06,
      "loss": 2.4594645500183105,
      "step": 4275
    },
    {
      "epoch": 0.9891279204256304,
      "grad_norm": 0.8577988743782043,
      "learning_rate": 3.798061746947995e-06,
      "loss": 2.529162883758545,
      "step": 4276
    },
    {
      "epoch": 0.9893592412676382,
      "grad_norm": 0.9080738425254822,
      "learning_rate": 3.641838992773816e-06,
      "loss": 2.2808244228363037,
      "step": 4277
    },
    {
      "epoch": 0.9895905621096461,
      "grad_norm": 0.7033615112304688,
      "learning_rate": 3.488873395920217e-06,
      "loss": 2.3493149280548096,
      "step": 4278
    },
    {
      "epoch": 0.989821882951654,
      "grad_norm": 0.7067438364028931,
      "learning_rate": 3.339166978780256e-06,
      "loss": 2.5133614540100098,
      "step": 4279
    },
    {
      "epoch": 0.9900532037936618,
      "grad_norm": 0.9871717691421509,
      "learning_rate": 3.1927217206564886e-06,
      "loss": 2.2467288970947266,
      "step": 4280
    },
    {
      "epoch": 0.9902845246356696,
      "grad_norm": 1.5355688333511353,
      "learning_rate": 3.049539557735015e-06,
      "loss": 2.393205165863037,
      "step": 4281
    },
    {
      "epoch": 0.9905158454776776,
      "grad_norm": 0.7723457217216492,
      "learning_rate": 2.9096223830598345e-06,
      "loss": 2.6968986988067627,
      "step": 4282
    },
    {
      "epoch": 0.9907471663196854,
      "grad_norm": 0.6708477735519409,
      "learning_rate": 2.7729720465077268e-06,
      "loss": 2.3078346252441406,
      "step": 4283
    },
    {
      "epoch": 0.9909784871616932,
      "grad_norm": 1.3206274509429932,
      "learning_rate": 2.6395903547638823e-06,
      "loss": 2.384979724884033,
      "step": 4284
    },
    {
      "epoch": 0.9912098080037012,
      "grad_norm": 0.6271413564682007,
      "learning_rate": 2.5094790712980318e-06,
      "loss": 2.167762041091919,
      "step": 4285
    },
    {
      "epoch": 0.991441128845709,
      "grad_norm": 0.8140906095504761,
      "learning_rate": 2.3826399163410504e-06,
      "loss": 2.6443891525268555,
      "step": 4286
    },
    {
      "epoch": 0.9916724496877168,
      "grad_norm": 0.8829963803291321,
      "learning_rate": 2.259074566862307e-06,
      "loss": 2.392711639404297,
      "step": 4287
    },
    {
      "epoch": 0.9919037705297248,
      "grad_norm": 0.9750952124595642,
      "learning_rate": 2.1387846565474044e-06,
      "loss": 2.3748655319213867,
      "step": 4288
    },
    {
      "epoch": 0.9921350913717326,
      "grad_norm": 0.699240505695343,
      "learning_rate": 2.0217717757766154e-06,
      "loss": 2.4112446308135986,
      "step": 4289
    },
    {
      "epoch": 0.9923664122137404,
      "grad_norm": 0.8417817950248718,
      "learning_rate": 1.9080374716038408e-06,
      "loss": 2.4967458248138428,
      "step": 4290
    },
    {
      "epoch": 0.9925977330557483,
      "grad_norm": 0.7309555411338806,
      "learning_rate": 1.7975832477362396e-06,
      "loss": 2.3125808238983154,
      "step": 4291
    },
    {
      "epoch": 0.9928290538977562,
      "grad_norm": 0.9992154240608215,
      "learning_rate": 1.6904105645142442e-06,
      "loss": 2.8425817489624023,
      "step": 4292
    },
    {
      "epoch": 0.993060374739764,
      "grad_norm": 0.7402819991111755,
      "learning_rate": 1.5865208388922969e-06,
      "loss": 2.4846842288970947,
      "step": 4293
    },
    {
      "epoch": 0.9932916955817719,
      "grad_norm": 0.737275242805481,
      "learning_rate": 1.4859154444200885e-06,
      "loss": 2.2889626026153564,
      "step": 4294
    },
    {
      "epoch": 0.9935230164237798,
      "grad_norm": 0.8263921737670898,
      "learning_rate": 1.3885957112244607e-06,
      "loss": 1.9346749782562256,
      "step": 4295
    },
    {
      "epoch": 0.9937543372657877,
      "grad_norm": 0.8021543025970459,
      "learning_rate": 1.2945629259917546e-06,
      "loss": 2.7206082344055176,
      "step": 4296
    },
    {
      "epoch": 0.9939856581077955,
      "grad_norm": 0.8753305077552795,
      "learning_rate": 1.2038183319507957e-06,
      "loss": 2.212523937225342,
      "step": 4297
    },
    {
      "epoch": 0.9942169789498034,
      "grad_norm": 0.8570302128791809,
      "learning_rate": 1.116363128856518e-06,
      "loss": 2.4221134185791016,
      "step": 4298
    },
    {
      "epoch": 0.9944482997918113,
      "grad_norm": 0.8654943704605103,
      "learning_rate": 1.0321984729741163e-06,
      "loss": 2.5362162590026855,
      "step": 4299
    },
    {
      "epoch": 0.9946796206338191,
      "grad_norm": 0.7737961411476135,
      "learning_rate": 9.513254770636137e-07,
      "loss": 2.619346857070923,
      "step": 4300
    },
    {
      "epoch": 0.9949109414758269,
      "grad_norm": 0.8589408993721008,
      "learning_rate": 8.737452103653177e-07,
      "loss": 2.4254915714263916,
      "step": 4301
    },
    {
      "epoch": 0.9951422623178349,
      "grad_norm": 1.4383223056793213,
      "learning_rate": 7.99458698585609e-07,
      "loss": 2.622690200805664,
      "step": 4302
    },
    {
      "epoch": 0.9953735831598427,
      "grad_norm": 0.7515788674354553,
      "learning_rate": 7.284669238833419e-07,
      "loss": 2.034594774246216,
      "step": 4303
    },
    {
      "epoch": 0.9956049040018505,
      "grad_norm": 1.0864291191101074,
      "learning_rate": 6.607708248569378e-07,
      "loss": 2.6047637462615967,
      "step": 4304
    },
    {
      "epoch": 0.9958362248438585,
      "grad_norm": 0.925018310546875,
      "learning_rate": 5.963712965319224e-07,
      "loss": 2.24570894241333,
      "step": 4305
    },
    {
      "epoch": 0.9960675456858663,
      "grad_norm": 0.93208247423172,
      "learning_rate": 5.352691903491303e-07,
      "loss": 2.649116039276123,
      "step": 4306
    },
    {
      "epoch": 0.9962988665278741,
      "grad_norm": 0.785614013671875,
      "learning_rate": 4.77465314153408e-07,
      "loss": 2.2825655937194824,
      "step": 4307
    },
    {
      "epoch": 0.9965301873698821,
      "grad_norm": 0.9997379183769226,
      "learning_rate": 4.229604321829561e-07,
      "loss": 2.2297863960266113,
      "step": 4308
    },
    {
      "epoch": 0.9967615082118899,
      "grad_norm": 1.0554535388946533,
      "learning_rate": 3.71755265059226e-07,
      "loss": 2.3953938484191895,
      "step": 4309
    },
    {
      "epoch": 0.9969928290538977,
      "grad_norm": 0.7805792689323425,
      "learning_rate": 3.2385048977739975e-07,
      "loss": 2.4101974964141846,
      "step": 4310
    },
    {
      "epoch": 0.9972241498959056,
      "grad_norm": 0.6927474141120911,
      "learning_rate": 2.792467396973974e-07,
      "loss": 2.266911506652832,
      "step": 4311
    },
    {
      "epoch": 0.9974554707379135,
      "grad_norm": 1.1406919956207275,
      "learning_rate": 2.3794460453555045e-07,
      "loss": 2.3847618103027344,
      "step": 4312
    },
    {
      "epoch": 0.9976867915799214,
      "grad_norm": 0.6707800030708313,
      "learning_rate": 1.9994463035682974e-07,
      "loss": 2.6387171745300293,
      "step": 4313
    },
    {
      "epoch": 0.9979181124219292,
      "grad_norm": 0.6993932127952576,
      "learning_rate": 1.6524731956749084e-07,
      "loss": 2.2121150493621826,
      "step": 4314
    },
    {
      "epoch": 0.9981494332639371,
      "grad_norm": 0.6454387903213501,
      "learning_rate": 1.3385313090857886e-07,
      "loss": 2.201232671737671,
      "step": 4315
    },
    {
      "epoch": 0.998380754105945,
      "grad_norm": 0.8995915055274963,
      "learning_rate": 1.0576247944985018e-07,
      "loss": 2.497774124145508,
      "step": 4316
    },
    {
      "epoch": 0.9986120749479528,
      "grad_norm": 0.6549047231674194,
      "learning_rate": 8.097573658419344e-08,
      "loss": 2.510427713394165,
      "step": 4317
    },
    {
      "epoch": 0.9988433957899607,
      "grad_norm": 0.7686660885810852,
      "learning_rate": 5.949323002271689e-08,
      "loss": 2.4002976417541504,
      "step": 4318
    },
    {
      "epoch": 0.9990747166319686,
      "grad_norm": 0.847687840461731,
      "learning_rate": 4.13152437906128e-08,
      "loss": 2.5880343914031982,
      "step": 4319
    },
    {
      "epoch": 0.9993060374739764,
      "grad_norm": 0.9476446509361267,
      "learning_rate": 2.644201822313286e-08,
      "loss": 2.329092025756836,
      "step": 4320
    },
    {
      "epoch": 0.9995373583159842,
      "grad_norm": 0.8229232430458069,
      "learning_rate": 1.4873749962562855e-08,
      "loss": 2.401028633117676,
      "step": 4321
    },
    {
      "epoch": 0.9997686791579922,
      "grad_norm": 1.1121824979782104,
      "learning_rate": 6.6105919556413985e-09,
      "loss": 2.5340259075164795,
      "step": 4322
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.7289966940879822,
      "learning_rate": 1.6526534514227365e-09,
      "loss": 2.010730504989624,
      "step": 4323
    }
  ],
  "logging_steps": 1,
  "max_steps": 4323,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 432,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.192408860609479e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}