{
  "best_global_step": 33400,
  "best_metric": 0.06660357117652893,
  "best_model_checkpoint": "saves/ia3/llama-3-8b-instruct/train_sst2_1744902621/checkpoint-33400",
  "epoch": 10.55694866041969,
  "eval_steps": 200,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0013197835554968984,
      "grad_norm": 3.1349356174468994,
      "learning_rate": 4.999999876629946e-05,
      "loss": 0.9631,
      "num_input_tokens_seen": 4480,
      "step": 5
    },
    {
      "epoch": 0.002639567110993797,
      "grad_norm": 2.9442992210388184,
      "learning_rate": 4.999999375439123e-05,
      "loss": 0.9189,
      "num_input_tokens_seen": 9024,
      "step": 10
    },
    {
      "epoch": 0.0039593506664906955,
      "grad_norm": 3.237074136734009,
      "learning_rate": 4.9999984887169785e-05,
      "loss": 0.9132,
      "num_input_tokens_seen": 14176,
      "step": 15
    },
    {
      "epoch": 0.005279134221987594,
      "grad_norm": 2.991049289703369,
      "learning_rate": 4.9999972164636506e-05,
      "loss": 1.098,
      "num_input_tokens_seen": 18528,
      "step": 20
    },
    {
      "epoch": 0.006598917777484493,
      "grad_norm": 2.9857192039489746,
      "learning_rate": 4.999995558679334e-05,
      "loss": 1.026,
      "num_input_tokens_seen": 22848,
      "step": 25
    },
    {
      "epoch": 0.007918701332981391,
      "grad_norm": 3.1116230487823486,
      "learning_rate": 4.999993515364287e-05,
      "loss": 0.9916,
      "num_input_tokens_seen": 27648,
      "step": 30
    },
    {
      "epoch": 0.00923848488847829,
      "grad_norm": 2.9589507579803467,
      "learning_rate": 4.999991086518822e-05,
      "loss": 0.9495,
      "num_input_tokens_seen": 32256,
      "step": 35
    },
    {
      "epoch": 0.010558268443975187,
      "grad_norm": 2.8274903297424316,
      "learning_rate": 4.999988272143315e-05,
      "loss": 0.9066,
      "num_input_tokens_seen": 36832,
      "step": 40
    },
    {
      "epoch": 0.011878051999472087,
      "grad_norm": 2.7399206161499023,
      "learning_rate": 4.999985072238199e-05,
      "loss": 0.8802,
      "num_input_tokens_seen": 41216,
      "step": 45
    },
    {
      "epoch": 0.013197835554968985,
      "grad_norm": 1.9083120822906494,
      "learning_rate": 4.999981486803969e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 45536,
      "step": 50
    },
    {
      "epoch": 0.014517619110465884,
      "grad_norm": 1.5509576797485352,
      "learning_rate": 4.999977515841176e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 50112,
      "step": 55
    },
    {
      "epoch": 0.015837402665962782,
      "grad_norm": 0.4766233265399933,
      "learning_rate": 4.9999731593504344e-05,
      "loss": 0.22,
      "num_input_tokens_seen": 54336,
      "step": 60
    },
    {
      "epoch": 0.017157186221459682,
      "grad_norm": 0.9432277679443359,
      "learning_rate": 4.999968417332415e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 58688,
      "step": 65
    },
    {
      "epoch": 0.01847696977695658,
      "grad_norm": 0.5134417414665222,
      "learning_rate": 4.999963289787848e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 63456,
      "step": 70
    },
    {
      "epoch": 0.019796753332453478,
      "grad_norm": 0.6916096806526184,
      "learning_rate": 4.999957776717526e-05,
      "loss": 0.1623,
      "num_input_tokens_seen": 67840,
      "step": 75
    },
    {
      "epoch": 0.021116536887950375,
      "grad_norm": 0.42286863923072815,
      "learning_rate": 4.9999518781222984e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 72192,
      "step": 80
    },
    {
      "epoch": 0.022436320443447275,
      "grad_norm": 0.8365755677223206,
      "learning_rate": 4.9999455940030746e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 76960,
      "step": 85
    },
    {
      "epoch": 0.023756103998944175,
      "grad_norm": 0.22348973155021667,
      "learning_rate": 4.999938924360824e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 81504,
      "step": 90
    },
    {
      "epoch": 0.02507588755444107,
      "grad_norm": 0.5905721783638,
      "learning_rate": 4.999931869196575e-05,
      "loss": 0.1214,
      "num_input_tokens_seen": 85696,
      "step": 95
    },
    {
      "epoch": 0.02639567110993797,
      "grad_norm": 0.593951940536499,
      "learning_rate": 4.999924428511416e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 90016,
      "step": 100
    },
    {
      "epoch": 0.027715454665434867,
      "grad_norm": 0.1924770474433899,
      "learning_rate": 4.999916602306494e-05,
      "loss": 0.1354,
      "num_input_tokens_seen": 94528,
      "step": 105
    },
    {
      "epoch": 0.029035238220931767,
      "grad_norm": 0.9475557208061218,
      "learning_rate": 4.999908390583016e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 98848,
      "step": 110
    },
    {
      "epoch": 0.030355021776428667,
      "grad_norm": 0.4281267821788788,
      "learning_rate": 4.999899793342247e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 103584,
      "step": 115
    },
    {
      "epoch": 0.031674805331925564,
      "grad_norm": 0.3587918281555176,
      "learning_rate": 4.999890810585516e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 108160,
      "step": 120
    },
    {
      "epoch": 0.03299458888742246,
      "grad_norm": 0.9242098331451416,
      "learning_rate": 4.999881442314206e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 112640,
      "step": 125
    },
    {
      "epoch": 0.034314372442919364,
      "grad_norm": 0.08543769270181656,
      "learning_rate": 4.9998716885297617e-05,
      "loss": 0.178,
      "num_input_tokens_seen": 116992,
      "step": 130
    },
    {
      "epoch": 0.03563415599841626,
      "grad_norm": 0.4117920398712158,
      "learning_rate": 4.999861549233688e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 121440,
      "step": 135
    },
    {
      "epoch": 0.03695393955391316,
      "grad_norm": 0.25010940432548523,
      "learning_rate": 4.999851024427548e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 126016,
      "step": 140
    },
    {
      "epoch": 0.03827372310941006,
      "grad_norm": 0.4828072488307953,
      "learning_rate": 4.999840114112965e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 130304,
      "step": 145
    },
    {
      "epoch": 0.039593506664906956,
      "grad_norm": 0.5798388123512268,
      "learning_rate": 4.999828818291621e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 134688,
      "step": 150
    },
    {
      "epoch": 0.04091329022040385,
      "grad_norm": 0.533508837223053,
      "learning_rate": 4.999817136965259e-05,
      "loss": 0.1878,
      "num_input_tokens_seen": 138784,
      "step": 155
    },
    {
      "epoch": 0.04223307377590075,
      "grad_norm": 0.2037423849105835,
      "learning_rate": 4.9998050701356794e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 143392,
      "step": 160
    },
    {
      "epoch": 0.04355285733139765,
      "grad_norm": 0.3969312310218811,
      "learning_rate": 4.999792617804744e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 147520,
      "step": 165
    },
    {
      "epoch": 0.04487264088689455,
      "grad_norm": 0.5613162517547607,
      "learning_rate": 4.9997797799743724e-05,
      "loss": 0.1199,
      "num_input_tokens_seen": 151840,
      "step": 170
    },
    {
      "epoch": 0.046192424442391446,
      "grad_norm": 0.7909346222877502,
      "learning_rate": 4.999766556646545e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 156160,
      "step": 175
    },
    {
      "epoch": 0.04751220799788835,
      "grad_norm": 0.7198038101196289,
      "learning_rate": 4.9997529478232996e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 160480,
      "step": 180
    },
    {
      "epoch": 0.048831991553385246,
      "grad_norm": 0.8182074427604675,
      "learning_rate": 4.9997389535067365e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 164992,
      "step": 185
    },
    {
      "epoch": 0.05015177510888214,
      "grad_norm": 0.6495415568351746,
      "learning_rate": 4.999724573699012e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 169600,
      "step": 190
    },
    {
      "epoch": 0.05147155866437904,
      "grad_norm": 0.4754011929035187,
      "learning_rate": 4.9997098084023457e-05,
      "loss": 0.1368,
      "num_input_tokens_seen": 174048,
      "step": 195
    },
    {
      "epoch": 0.05279134221987594,
      "grad_norm": 0.1435583531856537,
      "learning_rate": 4.999694657619013e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 178464,
      "step": 200
    },
    {
      "epoch": 0.05279134221987594,
      "eval_loss": 0.12938973307609558,
      "eval_runtime": 64.7462,
      "eval_samples_per_second": 104.022,
      "eval_steps_per_second": 26.009,
      "num_input_tokens_seen": 178464,
      "step": 200
    },
    {
      "epoch": 0.05411112577537284,
      "grad_norm": 0.25542405247688293,
      "learning_rate": 4.999679121351352e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 182784,
      "step": 205
    },
    {
      "epoch": 0.055430909330869735,
      "grad_norm": 0.7231872081756592,
      "learning_rate": 4.9996631996017565e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 187616,
      "step": 210
    },
    {
      "epoch": 0.05675069288636664,
      "grad_norm": 0.4120265245437622,
      "learning_rate": 4.9996468923726835e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 192224,
      "step": 215
    },
    {
      "epoch": 0.058070476441863535,
      "grad_norm": 0.32709211111068726,
      "learning_rate": 4.999630199666647e-05,
      "loss": 0.1156,
      "num_input_tokens_seen": 196896,
      "step": 220
    },
    {
      "epoch": 0.05939025999736043,
      "grad_norm": 0.28501302003860474,
      "learning_rate": 4.999613121486222e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 201184,
      "step": 225
    },
    {
      "epoch": 0.060710043552857335,
      "grad_norm": 0.29709622263908386,
      "learning_rate": 4.999595657834041e-05,
      "loss": 0.1614,
      "num_input_tokens_seen": 205664,
      "step": 230
    },
    {
      "epoch": 0.06202982710835423,
      "grad_norm": 0.8627525568008423,
      "learning_rate": 4.999577808712798e-05,
      "loss": 0.1521,
      "num_input_tokens_seen": 210112,
      "step": 235
    },
    {
      "epoch": 0.06334961066385113,
      "grad_norm": 0.5421628952026367,
      "learning_rate": 4.999559574125244e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 214656,
      "step": 240
    },
    {
      "epoch": 0.06466939421934803,
      "grad_norm": 0.393868625164032,
      "learning_rate": 4.9995409540741934e-05,
      "loss": 0.0978,
      "num_input_tokens_seen": 219072,
      "step": 245
    },
    {
      "epoch": 0.06598917777484492,
      "grad_norm": 0.28029927611351013,
      "learning_rate": 4.999521948562516e-05,
      "loss": 0.1211,
      "num_input_tokens_seen": 223520,
      "step": 250
    },
    {
      "epoch": 0.06730896133034182,
      "grad_norm": 0.622859001159668,
      "learning_rate": 4.999502557593143e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 227936,
      "step": 255
    },
    {
      "epoch": 0.06862874488583873,
      "grad_norm": 0.47244149446487427,
      "learning_rate": 4.999482781169066e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 232128,
      "step": 260
    },
    {
      "epoch": 0.06994852844133562,
      "grad_norm": 0.2308342307806015,
      "learning_rate": 4.9994626192933324e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 236800,
      "step": 265
    },
    {
      "epoch": 0.07126831199683252,
      "grad_norm": 0.16735781729221344,
      "learning_rate": 4.999442071969054e-05,
      "loss": 0.1109,
      "num_input_tokens_seen": 241280,
      "step": 270
    },
    {
      "epoch": 0.07258809555232942,
      "grad_norm": 0.29746511578559875,
      "learning_rate": 4.999421139199397e-05,
      "loss": 0.1201,
      "num_input_tokens_seen": 245888,
      "step": 275
    },
    {
      "epoch": 0.07390787910782631,
      "grad_norm": 0.47787079215049744,
      "learning_rate": 4.999399820987592e-05,
      "loss": 0.1735,
      "num_input_tokens_seen": 250496,
      "step": 280
    },
    {
      "epoch": 0.07522766266332322,
      "grad_norm": 0.6780644059181213,
      "learning_rate": 4.999378117336924e-05,
      "loss": 0.1371,
      "num_input_tokens_seen": 254944,
      "step": 285
    },
    {
      "epoch": 0.07654744621882012,
      "grad_norm": 0.8207125663757324,
      "learning_rate": 4.9993560282507415e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 259392,
      "step": 290
    },
    {
      "epoch": 0.07786722977431701,
      "grad_norm": 0.19681544601917267,
      "learning_rate": 4.9993335537324495e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 263872,
      "step": 295
    },
    {
      "epoch": 0.07918701332981391,
      "grad_norm": 0.2524445950984955,
      "learning_rate": 4.999310693785516e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 268384,
      "step": 300
    },
    {
      "epoch": 0.0805067968853108,
      "grad_norm": 0.28487133979797363,
      "learning_rate": 4.9992874484134653e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 272736,
      "step": 305
    },
    {
      "epoch": 0.0818265804408077,
      "grad_norm": 0.6542393565177917,
      "learning_rate": 4.999263817619882e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 277376,
      "step": 310
    },
    {
      "epoch": 0.08314636399630461,
      "grad_norm": 0.25458741188049316,
      "learning_rate": 4.9992398014084105e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 281664,
      "step": 315
    },
    {
      "epoch": 0.0844661475518015,
      "grad_norm": 0.16027264297008514,
      "learning_rate": 4.999215399782754e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 286304,
      "step": 320
    },
    {
      "epoch": 0.0857859311072984,
      "grad_norm": 0.4422883987426758,
      "learning_rate": 4.999190612746675e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 290368,
      "step": 325
    },
    {
      "epoch": 0.0871057146627953,
      "grad_norm": 0.8368133902549744,
      "learning_rate": 4.999165440303998e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 294624,
      "step": 330
    },
    {
      "epoch": 0.0884254982182922,
      "grad_norm": 0.23252594470977783,
      "learning_rate": 4.999139882458603e-05,
      "loss": 0.1385,
      "num_input_tokens_seen": 299168,
      "step": 335
    },
    {
      "epoch": 0.0897452817737891,
      "grad_norm": 0.4841640293598175,
      "learning_rate": 4.9991139392144314e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 303936,
      "step": 340
    },
    {
      "epoch": 0.091065065329286,
      "grad_norm": 0.1294574737548828,
      "learning_rate": 4.999087610575485e-05,
      "loss": 0.0741,
      "num_input_tokens_seen": 308384,
      "step": 345
    },
    {
      "epoch": 0.09238484888478289,
      "grad_norm": 0.4516092836856842,
      "learning_rate": 4.999060896545824e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 312768,
      "step": 350
    },
    {
      "epoch": 0.0937046324402798,
      "grad_norm": 0.5316590070724487,
      "learning_rate": 4.999033797129568e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 317248,
      "step": 355
    },
    {
      "epoch": 0.0950244159957767,
      "grad_norm": 0.5188689827919006,
      "learning_rate": 4.999006312330894e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 321760,
      "step": 360
    },
    {
      "epoch": 0.09634419955127359,
      "grad_norm": 0.2947482466697693,
      "learning_rate": 4.998978442154043e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 326176,
      "step": 365
    },
    {
      "epoch": 0.09766398310677049,
      "grad_norm": 0.44138023257255554,
      "learning_rate": 4.9989501866033125e-05,
      "loss": 0.1289,
      "num_input_tokens_seen": 330592,
      "step": 370
    },
    {
      "epoch": 0.0989837666622674,
      "grad_norm": 0.3255839943885803,
      "learning_rate": 4.998921545683059e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 334912,
      "step": 375
    },
    {
      "epoch": 0.10030355021776428,
      "grad_norm": 0.4621589183807373,
      "learning_rate": 4.9988925193976996e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 339840,
      "step": 380
    },
    {
      "epoch": 0.10162333377326119,
      "grad_norm": 0.33085736632347107,
      "learning_rate": 4.998863107751711e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 344064,
      "step": 385
    },
    {
      "epoch": 0.10294311732875808,
      "grad_norm": 0.5871715545654297,
      "learning_rate": 4.998833310749629e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 348512,
      "step": 390
    },
    {
      "epoch": 0.10426290088425498,
      "grad_norm": 0.07381974905729294,
      "learning_rate": 4.998803128396047e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 352960,
      "step": 395
    },
    {
      "epoch": 0.10558268443975188,
      "grad_norm": 0.5194829702377319,
      "learning_rate": 4.9987725606956215e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 357184,
      "step": 400
    },
    {
      "epoch": 0.10558268443975188,
      "eval_loss": 0.12012168020009995,
      "eval_runtime": 64.7271,
      "eval_samples_per_second": 104.052,
      "eval_steps_per_second": 26.017,
      "num_input_tokens_seen": 357184,
      "step": 400
    },
    {
      "epoch": 0.10690246799524877,
      "grad_norm": 0.3671155869960785,
      "learning_rate": 4.998741607653066e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 361888,
      "step": 405
    },
    {
      "epoch": 0.10822225155074568,
      "grad_norm": 0.12836100161075592,
      "learning_rate": 4.9987102692731523e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 366464,
      "step": 410
    },
    {
      "epoch": 0.10954203510624258,
      "grad_norm": 0.20675607025623322,
      "learning_rate": 4.9986785455607157e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 371008,
      "step": 415
    },
    {
      "epoch": 0.11086181866173947,
      "grad_norm": 0.10741525143384933,
      "learning_rate": 4.9986464365206456e-05,
      "loss": 0.0859,
      "num_input_tokens_seen": 375424,
      "step": 420
    },
    {
      "epoch": 0.11218160221723637,
      "grad_norm": 0.44301462173461914,
      "learning_rate": 4.9986139421578956e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 379872,
      "step": 425
    },
    {
      "epoch": 0.11350138577273328,
      "grad_norm": 0.22900453209877014,
      "learning_rate": 4.998581062477477e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 384320,
      "step": 430
    },
    {
      "epoch": 0.11482116932823017,
      "grad_norm": 0.24082441627979279,
      "learning_rate": 4.998547797484458e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 388448,
      "step": 435
    },
    {
      "epoch": 0.11614095288372707,
      "grad_norm": 0.38517895340919495,
      "learning_rate": 4.9985141471839706e-05,
      "loss": 0.1165,
      "num_input_tokens_seen": 393280,
      "step": 440
    },
    {
      "epoch": 0.11746073643922397,
      "grad_norm": 0.15212547779083252,
      "learning_rate": 4.998480111581203e-05,
      "loss": 0.2165,
      "num_input_tokens_seen": 397792,
      "step": 445
    },
    {
      "epoch": 0.11878051999472086,
      "grad_norm": 0.19099442660808563,
      "learning_rate": 4.998445690681405e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 402208,
      "step": 450
    },
    {
      "epoch": 0.12010030355021777,
      "grad_norm": 0.8132048845291138,
      "learning_rate": 4.9984108844898834e-05,
      "loss": 0.2056,
      "num_input_tokens_seen": 406720,
      "step": 455
    },
    {
      "epoch": 0.12142008710571467,
      "grad_norm": 0.10390175133943558,
      "learning_rate": 4.9983756930120076e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 411040,
      "step": 460
    },
    {
      "epoch": 0.12273987066121156,
      "grad_norm": 0.234686940908432,
      "learning_rate": 4.9983401162532025e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 415648,
      "step": 465
    },
    {
      "epoch": 0.12405965421670846,
      "grad_norm": 0.6214133501052856,
      "learning_rate": 4.998304154218955e-05,
      "loss": 0.125,
      "num_input_tokens_seen": 420224,
      "step": 470
    },
    {
      "epoch": 0.12537943777220537,
      "grad_norm": 0.15896718204021454,
      "learning_rate": 4.998267806914812e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 424768,
      "step": 475
    },
    {
      "epoch": 0.12669922132770225,
      "grad_norm": 0.8189268708229065,
      "learning_rate": 4.998231074346378e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 429280,
      "step": 480
    },
    {
      "epoch": 0.12801900488319914,
      "grad_norm": 0.1861373335123062,
      "learning_rate": 4.998193956519317e-05,
      "loss": 0.1155,
      "num_input_tokens_seen": 433600,
      "step": 485
    },
    {
      "epoch": 0.12933878843869606,
      "grad_norm": 0.4415060579776764,
      "learning_rate": 4.9981564534393545e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 438048,
      "step": 490
    },
    {
      "epoch": 0.13065857199419295,
      "grad_norm": 0.5352500677108765,
      "learning_rate": 4.998118565112272e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 442304,
      "step": 495
    },
    {
      "epoch": 0.13197835554968984,
      "grad_norm": 0.433537095785141,
      "learning_rate": 4.998080291543914e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 446976,
      "step": 500
    },
    {
      "epoch": 0.13329813910518676,
      "grad_norm": 0.5941283702850342,
      "learning_rate": 4.9980416327401826e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 451104,
      "step": 505
    },
    {
      "epoch": 0.13461792266068365,
      "grad_norm": 0.21125994622707367,
      "learning_rate": 4.998002588707038e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 455712,
      "step": 510
    },
    {
      "epoch": 0.13593770621618054,
      "grad_norm": 0.39476466178894043,
      "learning_rate": 4.997963159450503e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 460416,
      "step": 515
    },
    {
      "epoch": 0.13725748977167745,
      "grad_norm": 0.11853265762329102,
      "learning_rate": 4.9979233449766575e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 464768,
      "step": 520
    },
    {
      "epoch": 0.13857727332717434,
      "grad_norm": 0.4697645902633667,
      "learning_rate": 4.997883145291641e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 469120,
      "step": 525
    },
    {
      "epoch": 0.13989705688267123,
      "grad_norm": 0.7779855132102966,
      "learning_rate": 4.9978425604016536e-05,
      "loss": 0.1142,
      "num_input_tokens_seen": 473376,
      "step": 530
    },
    {
      "epoch": 0.14121684043816815,
      "grad_norm": 0.25538647174835205,
      "learning_rate": 4.9978015903129536e-05,
      "loss": 0.0926,
      "num_input_tokens_seen": 477824,
      "step": 535
    },
    {
      "epoch": 0.14253662399366504,
      "grad_norm": 0.15424871444702148,
      "learning_rate": 4.997760235031859e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 482304,
      "step": 540
    },
    {
      "epoch": 0.14385640754916193,
      "grad_norm": 0.3539102375507355,
      "learning_rate": 4.9977184945647473e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 486720,
      "step": 545
    },
    {
      "epoch": 0.14517619110465885,
      "grad_norm": 0.6351729035377502,
      "learning_rate": 4.997676368918055e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 491328,
      "step": 550
    },
    {
      "epoch": 0.14649597466015574,
      "grad_norm": 0.5514878630638123,
      "learning_rate": 4.9976338580982794e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 495616,
      "step": 555
    },
    {
      "epoch": 0.14781575821565263,
      "grad_norm": 0.1769968867301941,
      "learning_rate": 4.9975909621119755e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 499936,
      "step": 560
    },
    {
      "epoch": 0.14913554177114954,
      "grad_norm": 0.4349355101585388,
      "learning_rate": 4.997547680965758e-05,
      "loss": 0.1023,
      "num_input_tokens_seen": 504864,
      "step": 565
    },
    {
      "epoch": 0.15045532532664643,
      "grad_norm": 0.3109443783760071,
      "learning_rate": 4.997504014666302e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 509312,
      "step": 570
    },
    {
      "epoch": 0.15177510888214332,
      "grad_norm": 0.27755749225616455,
      "learning_rate": 4.997459963220342e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 513824,
      "step": 575
    },
    {
      "epoch": 0.15309489243764024,
      "grad_norm": 0.654373824596405,
      "learning_rate": 4.997415526634671e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 518208,
      "step": 580
    },
    {
      "epoch": 0.15441467599313713,
      "grad_norm": 0.65253746509552,
      "learning_rate": 4.99737070491614e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 522592,
      "step": 585
    },
    {
      "epoch": 0.15573445954863402,
      "grad_norm": 0.7733871340751648,
      "learning_rate": 4.997325498071663e-05,
      "loss": 0.2244,
      "num_input_tokens_seen": 527040,
      "step": 590
    },
    {
      "epoch": 0.15705424310413094,
      "grad_norm": 0.7715486288070679,
      "learning_rate": 4.997279906108211e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 531296,
      "step": 595
    },
    {
      "epoch": 0.15837402665962783,
      "grad_norm": 0.0990673080086708,
      "learning_rate": 4.9972339290328155e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 535488,
      "step": 600
    },
    {
      "epoch": 0.15837402665962783,
      "eval_loss": 0.1138034537434578,
      "eval_runtime": 64.6906,
      "eval_samples_per_second": 104.111,
      "eval_steps_per_second": 26.032,
      "num_input_tokens_seen": 535488,
      "step": 600
    },
    {
      "epoch": 0.15969381021512472,
      "grad_norm": 0.33363252878189087,
      "learning_rate": 4.9971875668525646e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 540000,
      "step": 605
    },
    {
      "epoch": 0.1610135937706216,
      "grad_norm": 0.49440857768058777,
      "learning_rate": 4.997140819574609e-05,
      "loss": 0.1298,
      "num_input_tokens_seen": 544256,
      "step": 610
    },
    {
      "epoch": 0.16233337732611852,
      "grad_norm": 0.2718019485473633,
      "learning_rate": 4.997093687206159e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 548704,
      "step": 615
    },
    {
      "epoch": 0.1636531608816154,
      "grad_norm": 0.13265176117420197,
      "learning_rate": 4.997046169754482e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 553184,
      "step": 620
    },
    {
      "epoch": 0.1649729444371123,
      "grad_norm": 0.25224459171295166,
      "learning_rate": 4.996998267226905e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 557664,
      "step": 625
    },
    {
      "epoch": 0.16629272799260922,
      "grad_norm": 0.31581202149391174,
      "learning_rate": 4.996949979630817e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 562080,
      "step": 630
    },
    {
      "epoch": 0.1676125115481061,
      "grad_norm": 0.4302205443382263,
      "learning_rate": 4.996901306973663e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 566528,
      "step": 635
    },
    {
      "epoch": 0.168932295103603,
      "grad_norm": 0.27955928444862366,
      "learning_rate": 4.996852249262949e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 571040,
      "step": 640
    },
    {
      "epoch": 0.17025207865909991,
      "grad_norm": 0.2058146893978119,
      "learning_rate": 4.996802806506241e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 575488,
      "step": 645
    },
    {
      "epoch": 0.1715718622145968,
      "grad_norm": 0.29260969161987305,
      "learning_rate": 4.996752978711164e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 580192,
      "step": 650
    },
    {
      "epoch": 0.1728916457700937,
      "grad_norm": 0.3859860897064209,
      "learning_rate": 4.996702765885401e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 584768,
      "step": 655
    },
    {
      "epoch": 0.1742114293255906,
      "grad_norm": 0.34305039048194885,
      "learning_rate": 4.9966521680366964e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 588928,
      "step": 660
    },
    {
      "epoch": 0.1755312128810875,
      "grad_norm": 0.5077118277549744,
      "learning_rate": 4.9966011851728524e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 593248,
      "step": 665
    },
    {
      "epoch": 0.1768509964365844,
      "grad_norm": 0.15490347146987915,
      "learning_rate": 4.996549817301731e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 597728,
      "step": 670
    },
    {
      "epoch": 0.1781707799920813,
      "grad_norm": 0.15187938511371613,
      "learning_rate": 4.9964980644312544e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 601888,
      "step": 675
    },
    {
      "epoch": 0.1794905635475782,
      "grad_norm": 0.3256392478942871,
      "learning_rate": 4.996445926569403e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 606336,
      "step": 680
    },
    {
      "epoch": 0.1808103471030751,
      "grad_norm": 0.2006968855857849,
      "learning_rate": 4.996393403724218e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 610912,
      "step": 685
    },
    {
      "epoch": 0.182130130658572,
      "grad_norm": 0.24015803635120392,
      "learning_rate": 4.9963404959037985e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 615392,
      "step": 690
    },
    {
      "epoch": 0.1834499142140689,
      "grad_norm": 0.7320178151130676,
      "learning_rate": 4.996287203116303e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 620000,
      "step": 695
    },
    {
      "epoch": 0.18476969776956578,
      "grad_norm": 0.22626900672912598,
      "learning_rate": 4.996233525369951e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 624704,
      "step": 700
    },
    {
      "epoch": 0.1860894813250627,
      "grad_norm": 0.4822400212287903,
      "learning_rate": 4.99617946267302e-05,
      "loss": 0.1091,
      "num_input_tokens_seen": 629088,
      "step": 705
    },
    {
      "epoch": 0.1874092648805596,
      "grad_norm": 0.17198865115642548,
      "learning_rate": 4.996125015033846e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 633632,
      "step": 710
    },
    {
      "epoch": 0.18872904843605648,
      "grad_norm": 0.2185218632221222,
      "learning_rate": 4.996070182460827e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 637888,
      "step": 715
    },
    {
      "epoch": 0.1900488319915534,
      "grad_norm": 0.12374971807003021,
      "learning_rate": 4.996014964962418e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 642240,
      "step": 720
    },
    {
      "epoch": 0.19136861554705029,
      "grad_norm": 0.33605867624282837,
      "learning_rate": 4.9959593625471344e-05,
      "loss": 0.1513,
      "num_input_tokens_seen": 646784,
      "step": 725
    },
    {
      "epoch": 0.19268839910254718,
      "grad_norm": 0.5107312798500061,
      "learning_rate": 4.995903375223552e-05,
      "loss": 0.115,
      "num_input_tokens_seen": 651424,
      "step": 730
    },
    {
      "epoch": 0.1940081826580441,
      "grad_norm": 0.3866291046142578,
      "learning_rate": 4.995847003000302e-05,
      "loss": 0.1214,
      "num_input_tokens_seen": 655712,
      "step": 735
    },
    {
      "epoch": 0.19532796621354098,
      "grad_norm": 0.19646111130714417,
      "learning_rate": 4.9957902458860804e-05,
      "loss": 0.1104,
      "num_input_tokens_seen": 660448,
      "step": 740
    },
    {
      "epoch": 0.19664774976903787,
      "grad_norm": 0.40927743911743164,
      "learning_rate": 4.995733103889639e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 664960,
      "step": 745
    },
    {
      "epoch": 0.1979675333245348,
      "grad_norm": 0.1660161167383194,
      "learning_rate": 4.99567557701979e-05,
      "loss": 0.1034,
      "num_input_tokens_seen": 669536,
      "step": 750
    },
    {
      "epoch": 0.19928731688003168,
      "grad_norm": 0.4235718250274658,
      "learning_rate": 4.995617665285403e-05,
      "loss": 0.0837,
      "num_input_tokens_seen": 674016,
      "step": 755
    },
    {
      "epoch": 0.20060710043552857,
      "grad_norm": 0.45117032527923584,
      "learning_rate": 4.99555936869541e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 678592,
      "step": 760
    },
    {
      "epoch": 0.20192688399102549,
      "grad_norm": 0.2279418706893921,
      "learning_rate": 4.995500687258803e-05,
      "loss": 0.1328,
      "num_input_tokens_seen": 682944,
      "step": 765
    },
    {
      "epoch": 0.20324666754652237,
      "grad_norm": 0.4901421070098877,
      "learning_rate": 4.995441620984628e-05,
      "loss": 0.1006,
      "num_input_tokens_seen": 687328,
      "step": 770
    },
    {
      "epoch": 0.20456645110201926,
      "grad_norm": 0.7021390199661255,
      "learning_rate": 4.995382169881996e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 691968,
      "step": 775
    },
    {
      "epoch": 0.20588623465751615,
      "grad_norm": 0.3093251585960388,
      "learning_rate": 4.9953223339600755e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 696512,
      "step": 780
    },
    {
      "epoch": 0.20720601821301307,
      "grad_norm": 0.23922395706176758,
      "learning_rate": 4.995262113228091e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 701088,
      "step": 785
    },
    {
      "epoch": 0.20852580176850996,
      "grad_norm": 0.4129209518432617,
      "learning_rate": 4.995201507695332e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 705472,
      "step": 790
    },
    {
      "epoch": 0.20984558532400685,
      "grad_norm": 0.3629891276359558,
      "learning_rate": 4.995140517371144e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 709984,
      "step": 795
    },
    {
      "epoch": 0.21116536887950377,
      "grad_norm": 0.18731938302516937,
      "learning_rate": 4.995079142264932e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 714592,
      "step": 800
    },
    {
      "epoch": 0.21116536887950377,
      "eval_loss": 0.10932759195566177,
      "eval_runtime": 64.7822,
      "eval_samples_per_second": 103.964,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 714592,
      "step": 800
    },
    {
      "epoch": 0.21248515243500066,
      "grad_norm": 0.2243635207414627,
      "learning_rate": 4.995017382386162e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 719168,
      "step": 805
    },
    {
      "epoch": 0.21380493599049755,
      "grad_norm": 0.24593189358711243,
      "learning_rate": 4.994955237744356e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 723840,
      "step": 810
    },
    {
      "epoch": 0.21512471954599446,
      "grad_norm": 0.21937479078769684,
      "learning_rate": 4.994892708349101e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 728096,
      "step": 815
    },
    {
      "epoch": 0.21644450310149135,
      "grad_norm": 0.24053339660167694,
      "learning_rate": 4.994829794210035e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 732736,
      "step": 820
    },
    {
      "epoch": 0.21776428665698824,
      "grad_norm": 0.2171311229467392,
      "learning_rate": 4.994766495336864e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 737088,
      "step": 825
    },
    {
      "epoch": 0.21908407021248516,
      "grad_norm": 0.2992972135543823,
      "learning_rate": 4.994702811739348e-05,
      "loss": 0.1271,
      "num_input_tokens_seen": 741664,
      "step": 830
    },
    {
      "epoch": 0.22040385376798205,
      "grad_norm": 0.3149014413356781,
      "learning_rate": 4.994638743427308e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 746048,
      "step": 835
    },
    {
      "epoch": 0.22172363732347894,
      "grad_norm": 0.17598281800746918,
      "learning_rate": 4.994574290410624e-05,
      "loss": 0.1007,
      "num_input_tokens_seen": 750880,
      "step": 840
    },
    {
      "epoch": 0.22304342087897586,
      "grad_norm": 0.20840731263160706,
      "learning_rate": 4.9945094526992364e-05,
      "loss": 0.1269,
      "num_input_tokens_seen": 755232,
      "step": 845
    },
    {
      "epoch": 0.22436320443447275,
      "grad_norm": 0.24480533599853516,
      "learning_rate": 4.994444230303142e-05,
      "loss": 0.1159,
      "num_input_tokens_seen": 759712,
      "step": 850
    },
    {
      "epoch": 0.22568298798996964,
      "grad_norm": 0.2270325869321823,
      "learning_rate": 4.994378623232402e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 764064,
      "step": 855
    },
    {
      "epoch": 0.22700277154546655,
      "grad_norm": 0.4034910798072815,
      "learning_rate": 4.99431263149713e-05,
      "loss": 0.1317,
      "num_input_tokens_seen": 768544,
      "step": 860
    },
    {
      "epoch": 0.22832255510096344,
      "grad_norm": 0.2449798285961151,
      "learning_rate": 4.9942462551075056e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 773152,
      "step": 865
    },
    {
      "epoch": 0.22964233865646033,
      "grad_norm": 0.5677728056907654,
      "learning_rate": 4.994179494073764e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 777440,
      "step": 870
    },
    {
      "epoch": 0.23096212221195725,
      "grad_norm": 0.22239822149276733,
      "learning_rate": 4.9941123484062e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 782016,
      "step": 875
    },
    {
      "epoch": 0.23228190576745414,
      "grad_norm": 0.5216113328933716,
      "learning_rate": 4.99404481811517e-05,
      "loss": 0.1556,
      "num_input_tokens_seen": 786336,
      "step": 880
    },
    {
      "epoch": 0.23360168932295103,
      "grad_norm": 0.14891619980335236,
      "learning_rate": 4.9939769032110864e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 790976,
      "step": 885
    },
    {
      "epoch": 0.23492147287844795,
      "grad_norm": 0.16395199298858643,
      "learning_rate": 4.993908603704423e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 795456,
      "step": 890
    },
    {
      "epoch": 0.23624125643394484,
      "grad_norm": 0.5704664587974548,
      "learning_rate": 4.9938399196057126e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 799808,
      "step": 895
    },
    {
      "epoch": 0.23756103998944172,
      "grad_norm": 0.23360532522201538,
      "learning_rate": 4.993770850925547e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 804576,
      "step": 900
    },
    {
      "epoch": 0.23888082354493864,
      "grad_norm": 0.27320852875709534,
      "learning_rate": 4.993701397674577e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 809248,
      "step": 905
    },
    {
      "epoch": 0.24020060710043553,
      "grad_norm": 0.40846580266952515,
      "learning_rate": 4.993631559863515e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 813888,
      "step": 910
    },
    {
      "epoch": 0.24152039065593242,
      "grad_norm": 0.4350108802318573,
      "learning_rate": 4.9935613375031283e-05,
      "loss": 0.1104,
      "num_input_tokens_seen": 817952,
      "step": 915
    },
    {
      "epoch": 0.24284017421142934,
      "grad_norm": 0.4193384051322937,
      "learning_rate": 4.993490730604248e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 822368,
      "step": 920
    },
    {
      "epoch": 0.24415995776692623,
      "grad_norm": 0.25415298342704773,
      "learning_rate": 4.993419739177761e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 826816,
      "step": 925
    },
    {
      "epoch": 0.24547974132242312,
      "grad_norm": 0.10651128739118576,
      "learning_rate": 4.9933483632346164e-05,
      "loss": 0.1071,
      "num_input_tokens_seen": 831104,
      "step": 930
    },
    {
      "epoch": 0.24679952487792003,
      "grad_norm": 0.2735975980758667,
      "learning_rate": 4.993276602785821e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 835680,
      "step": 935
    },
    {
      "epoch": 0.24811930843341692,
      "grad_norm": 0.3815348148345947,
      "learning_rate": 4.993204457842441e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 840352,
      "step": 940
    },
    {
      "epoch": 0.2494390919889138,
      "grad_norm": 0.4771551191806793,
      "learning_rate": 4.993131928415602e-05,
      "loss": 0.0788,
      "num_input_tokens_seen": 844800,
      "step": 945
    },
    {
      "epoch": 0.25075887554441073,
      "grad_norm": 0.24847100675106049,
      "learning_rate": 4.993059014516489e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 849152,
      "step": 950
    },
    {
      "epoch": 0.2520786590999076,
      "grad_norm": 0.16576330363750458,
      "learning_rate": 4.9929857161563464e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 853632,
      "step": 955
    },
    {
      "epoch": 0.2533984426554045,
      "grad_norm": 0.12331245094537735,
      "learning_rate": 4.992912033346477e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 858336,
      "step": 960
    },
    {
      "epoch": 0.2547182262109014,
      "grad_norm": 0.28777411580085754,
      "learning_rate": 4.992837966098245e-05,
      "loss": 0.1248,
      "num_input_tokens_seen": 862528,
      "step": 965
    },
    {
      "epoch": 0.2560380097663983,
      "grad_norm": 0.22176648676395416,
      "learning_rate": 4.992763514423071e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 866976,
      "step": 970
    },
    {
      "epoch": 0.25735779332189523,
      "grad_norm": 0.42294949293136597,
      "learning_rate": 4.992688678332437e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 871456,
      "step": 975
    },
    {
      "epoch": 0.2586775768773921,
      "grad_norm": 0.44421321153640747,
      "learning_rate": 4.992613457837884e-05,
      "loss": 0.1199,
      "num_input_tokens_seen": 876128,
      "step": 980
    },
    {
      "epoch": 0.259997360432889,
      "grad_norm": 0.5526481866836548,
      "learning_rate": 4.992537852951011e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 880352,
      "step": 985
    },
    {
      "epoch": 0.2613171439883859,
      "grad_norm": 0.32688695192337036,
      "learning_rate": 4.9924618636834785e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 884672,
      "step": 990
    },
    {
      "epoch": 0.2626369275438828,
      "grad_norm": 0.17850558459758759,
      "learning_rate": 4.9923854900470046e-05,
      "loss": 0.1902,
      "num_input_tokens_seen": 888800,
      "step": 995
    },
    {
      "epoch": 0.2639567110993797,
      "grad_norm": 0.15969976782798767,
      "learning_rate": 4.992308732053367e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 893216,
      "step": 1000
    },
    {
      "epoch": 0.2639567110993797,
      "eval_loss": 0.10668887197971344,
      "eval_runtime": 64.7406,
      "eval_samples_per_second": 104.031,
      "eval_steps_per_second": 26.012,
      "num_input_tokens_seen": 893216,
      "step": 1000
    },
    {
      "epoch": 0.2652764946548766,
      "grad_norm": 0.27260082960128784,
      "learning_rate": 4.992231589714402e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 897600,
      "step": 1005
    },
    {
      "epoch": 0.2665962782103735,
      "grad_norm": 0.22051632404327393,
      "learning_rate": 4.992154063042007e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 902112,
      "step": 1010
    },
    {
      "epoch": 0.2679160617658704,
      "grad_norm": 0.2527467608451843,
      "learning_rate": 4.992076152048136e-05,
      "loss": 0.1315,
      "num_input_tokens_seen": 906560,
      "step": 1015
    },
    {
      "epoch": 0.2692358453213673,
      "grad_norm": 0.22719652950763702,
      "learning_rate": 4.991997856744807e-05,
      "loss": 0.0928,
      "num_input_tokens_seen": 910720,
      "step": 1020
    },
    {
      "epoch": 0.2705556288768642,
      "grad_norm": 0.3573048412799835,
      "learning_rate": 4.9919191771440905e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 915040,
      "step": 1025
    },
    {
      "epoch": 0.2718754124323611,
      "grad_norm": 0.16758829355239868,
      "learning_rate": 4.991840113258122e-05,
      "loss": 0.1287,
      "num_input_tokens_seen": 919712,
      "step": 1030
    },
    {
      "epoch": 0.27319519598785796,
      "grad_norm": 0.3242450952529907,
      "learning_rate": 4.9917606650990933e-05,
      "loss": 0.0859,
      "num_input_tokens_seen": 924224,
      "step": 1035
    },
    {
      "epoch": 0.2745149795433549,
      "grad_norm": 0.08949166536331177,
      "learning_rate": 4.9916808326792566e-05,
      "loss": 0.1408,
      "num_input_tokens_seen": 928416,
      "step": 1040
    },
    {
      "epoch": 0.2758347630988518,
      "grad_norm": 0.29730647802352905,
      "learning_rate": 4.9916006160109235e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 932800,
      "step": 1045
    },
    {
      "epoch": 0.2771545466543487,
      "grad_norm": 0.5288802981376648,
      "learning_rate": 4.991520015106464e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 937440,
      "step": 1050
    },
    {
      "epoch": 0.2784743302098456,
      "grad_norm": 0.1059739887714386,
      "learning_rate": 4.991439029978308e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 941952,
      "step": 1055
    },
    {
      "epoch": 0.27979411376534247,
      "grad_norm": 1.870263934135437,
      "learning_rate": 4.9913576606389434e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 946496,
      "step": 1060
    },
    {
      "epoch": 0.28111389732083936,
      "grad_norm": 0.20395955443382263,
      "learning_rate": 4.991275907100919e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 950496,
      "step": 1065
    },
    {
      "epoch": 0.2824336808763363,
      "grad_norm": 0.20579855144023895,
      "learning_rate": 4.9911937693768434e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 955104,
      "step": 1070
    },
    {
      "epoch": 0.2837534644318332,
      "grad_norm": 0.34967049956321716,
      "learning_rate": 4.991111247479382e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 959648,
      "step": 1075
    },
    {
      "epoch": 0.2850732479873301,
      "grad_norm": 0.33796805143356323,
      "learning_rate": 4.9910283414212605e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 964256,
      "step": 1080
    },
    {
      "epoch": 0.28639303154282697,
      "grad_norm": 0.38363730907440186,
      "learning_rate": 4.990945051215265e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 968608,
      "step": 1085
    },
    {
      "epoch": 0.28771281509832386,
      "grad_norm": 0.11821810901165009,
      "learning_rate": 4.99086137687424e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 973376,
      "step": 1090
    },
    {
      "epoch": 0.28903259865382075,
      "grad_norm": 0.5425934195518494,
      "learning_rate": 4.9907773184110874e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 977696,
      "step": 1095
    },
    {
      "epoch": 0.2903523822093177,
      "grad_norm": 0.24836581945419312,
      "learning_rate": 4.9906928758387715e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 981856,
      "step": 1100
    },
    {
      "epoch": 0.2916721657648146,
      "grad_norm": 0.2982899844646454,
      "learning_rate": 4.9906080491703146e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 986368,
      "step": 1105
    },
    {
      "epoch": 0.2929919493203115,
      "grad_norm": 0.2044144719839096,
      "learning_rate": 4.990522838418797e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 990848,
      "step": 1110
    },
    {
      "epoch": 0.29431173287580836,
      "grad_norm": 0.34760767221450806,
      "learning_rate": 4.9904372435973604e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 995488,
      "step": 1115
    },
    {
      "epoch": 0.29563151643130525,
      "grad_norm": 0.6031702160835266,
      "learning_rate": 4.990351264719203e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 999872,
      "step": 1120
    },
    {
      "epoch": 0.29695129998680214,
      "grad_norm": 0.30428245663642883,
      "learning_rate": 4.990264901797586e-05,
      "loss": 0.1183,
      "num_input_tokens_seen": 1004256,
      "step": 1125
    },
    {
      "epoch": 0.2982710835422991,
      "grad_norm": 0.26574578881263733,
      "learning_rate": 4.990178154845826e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 1008864,
      "step": 1130
    },
    {
      "epoch": 0.299590867097796,
      "grad_norm": 0.15483444929122925,
      "learning_rate": 4.9900910238773014e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 1013472,
      "step": 1135
    },
    {
      "epoch": 0.30091065065329287,
      "grad_norm": 0.17342835664749146,
      "learning_rate": 4.990003508905448e-05,
      "loss": 0.0752,
      "num_input_tokens_seen": 1018016,
      "step": 1140
    },
    {
      "epoch": 0.30223043420878976,
      "grad_norm": 0.15781110525131226,
      "learning_rate": 4.989915609943763e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 1022272,
      "step": 1145
    },
    {
      "epoch": 0.30355021776428665,
      "grad_norm": 0.28851866722106934,
      "learning_rate": 4.9898273270058e-05,
      "loss": 0.0783,
      "num_input_tokens_seen": 1026816,
      "step": 1150
    },
    {
      "epoch": 0.30487000131978353,
      "grad_norm": 0.0742795318365097,
      "learning_rate": 4.989738660105174e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 1031616,
      "step": 1155
    },
    {
      "epoch": 0.3061897848752805,
      "grad_norm": 0.460222065448761,
      "learning_rate": 4.989649609255559e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 1036320,
      "step": 1160
    },
    {
      "epoch": 0.30750956843077737,
      "grad_norm": 0.6675683259963989,
      "learning_rate": 4.989560174470687e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 1041248,
      "step": 1165
    },
    {
      "epoch": 0.30882935198627426,
      "grad_norm": 0.09712900221347809,
      "learning_rate": 4.989470355764351e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 1045824,
      "step": 1170
    },
    {
      "epoch": 0.31014913554177115,
      "grad_norm": 0.21646857261657715,
      "learning_rate": 4.9893801531504e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 1050368,
      "step": 1175
    },
    {
      "epoch": 0.31146891909726804,
      "grad_norm": 0.22017158567905426,
      "learning_rate": 4.9892895666427475e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 1054880,
      "step": 1180
    },
    {
      "epoch": 0.3127887026527649,
      "grad_norm": 0.2171737104654312,
      "learning_rate": 4.9891985962553606e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 1059168,
      "step": 1185
    },
    {
      "epoch": 0.3141084862082619,
      "grad_norm": 0.35547348856925964,
      "learning_rate": 4.989107242002269e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 1063584,
      "step": 1190
    },
    {
      "epoch": 0.31542826976375876,
      "grad_norm": 0.3880825936794281,
      "learning_rate": 4.989015503897561e-05,
      "loss": 0.104,
      "num_input_tokens_seen": 1067936,
      "step": 1195
    },
    {
      "epoch": 0.31674805331925565,
      "grad_norm": 0.09029817581176758,
      "learning_rate": 4.988923381955383e-05,
      "loss": 0.1027,
      "num_input_tokens_seen": 1072832,
      "step": 1200
    },
    {
      "epoch": 0.31674805331925565,
      "eval_loss": 0.10287895798683167,
      "eval_runtime": 64.7186,
      "eval_samples_per_second": 104.066,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 1072832,
      "step": 1200
    },
    {
      "epoch": 0.31806783687475254,
      "grad_norm": 0.47186365723609924,
      "learning_rate": 4.988830876189942e-05,
      "loss": 0.127,
      "num_input_tokens_seen": 1077440,
      "step": 1205
    },
    {
      "epoch": 0.31938762043024943,
      "grad_norm": 0.2922489047050476,
      "learning_rate": 4.988737986615503e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 1081696,
      "step": 1210
    },
    {
      "epoch": 0.3207074039857463,
      "grad_norm": 0.19668005406856537,
      "learning_rate": 4.988644713246391e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 1086112,
      "step": 1215
    },
    {
      "epoch": 0.3220271875412432,
      "grad_norm": 0.3127315044403076,
      "learning_rate": 4.988551056096991e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 1090528,
      "step": 1220
    },
    {
      "epoch": 0.32334697109674015,
      "grad_norm": 0.529601514339447,
      "learning_rate": 4.988457015181743e-05,
      "loss": 0.127,
      "num_input_tokens_seen": 1094944,
      "step": 1225
    },
    {
      "epoch": 0.32466675465223704,
      "grad_norm": 0.31953519582748413,
      "learning_rate": 4.988362590515153e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 1099296,
      "step": 1230
    },
    {
      "epoch": 0.32598653820773393,
      "grad_norm": 0.22914747893810272,
      "learning_rate": 4.9882677821117805e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 1103552,
      "step": 1235
    },
    {
      "epoch": 0.3273063217632308,
      "grad_norm": 0.19504457712173462,
      "learning_rate": 4.988172589986246e-05,
      "loss": 0.176,
      "num_input_tokens_seen": 1108096,
      "step": 1240
    },
    {
      "epoch": 0.3286261053187277,
      "grad_norm": 0.25285378098487854,
      "learning_rate": 4.9880770141532304e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 1112576,
      "step": 1245
    },
    {
      "epoch": 0.3299458888742246,
      "grad_norm": 0.04385167732834816,
      "learning_rate": 4.987981054627472e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 1116832,
      "step": 1250
    },
    {
      "epoch": 0.33126567242972155,
      "grad_norm": 0.31458359956741333,
      "learning_rate": 4.987884711423769e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 1121280,
      "step": 1255
    },
    {
      "epoch": 0.33258545598521844,
      "grad_norm": 0.6382271647453308,
      "learning_rate": 4.9877879845569784e-05,
      "loss": 0.1226,
      "num_input_tokens_seen": 1125728,
      "step": 1260
    },
    {
      "epoch": 0.3339052395407153,
      "grad_norm": 0.3020384907722473,
      "learning_rate": 4.9876908740420175e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 1130368,
      "step": 1265
    },
    {
      "epoch": 0.3352250230962122,
      "grad_norm": 0.15630574524402618,
      "learning_rate": 4.987593379893861e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 1134944,
      "step": 1270
    },
    {
      "epoch": 0.3365448066517091,
      "grad_norm": 0.13506488502025604,
      "learning_rate": 4.987495502127545e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 1139168,
      "step": 1275
    },
    {
      "epoch": 0.337864590207206,
      "grad_norm": 0.617756724357605,
      "learning_rate": 4.987397240758162e-05,
      "loss": 0.1222,
      "num_input_tokens_seen": 1143392,
      "step": 1280
    },
    {
      "epoch": 0.33918437376270294,
      "grad_norm": 0.395846962928772,
      "learning_rate": 4.9872985958008664e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 1147648,
      "step": 1285
    },
    {
      "epoch": 0.34050415731819983,
      "grad_norm": 0.5758973956108093,
      "learning_rate": 4.987199567270871e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 1151744,
      "step": 1290
    },
    {
      "epoch": 0.3418239408736967,
      "grad_norm": 0.5372743606567383,
      "learning_rate": 4.9871001551834444e-05,
      "loss": 0.125,
      "num_input_tokens_seen": 1156256,
      "step": 1295
    },
    {
      "epoch": 0.3431437244291936,
      "grad_norm": 0.3537611663341522,
      "learning_rate": 4.98700035955392e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 1160768,
      "step": 1300
    },
    {
      "epoch": 0.3444635079846905,
      "grad_norm": 0.30623751878738403,
      "learning_rate": 4.986900180397686e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 1165248,
      "step": 1305
    },
    {
      "epoch": 0.3457832915401874,
      "grad_norm": 0.3520481288433075,
      "learning_rate": 4.9867996177301926e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 1169504,
      "step": 1310
    },
    {
      "epoch": 0.34710307509568433,
      "grad_norm": 0.18603025376796722,
      "learning_rate": 4.9866986715669464e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 1173856,
      "step": 1315
    },
    {
      "epoch": 0.3484228586511812,
      "grad_norm": 0.11228612810373306,
      "learning_rate": 4.9865973419235155e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 1178432,
      "step": 1320
    },
    {
      "epoch": 0.3497426422066781,
      "grad_norm": 0.319702684879303,
      "learning_rate": 4.986495628815526e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 1182944,
      "step": 1325
    },
    {
      "epoch": 0.351062425762175,
      "grad_norm": 0.23791475594043732,
      "learning_rate": 4.986393532258663e-05,
      "loss": 0.1379,
      "num_input_tokens_seen": 1187456,
      "step": 1330
    },
    {
      "epoch": 0.3523822093176719,
      "grad_norm": 0.19384010136127472,
      "learning_rate": 4.986291052268671e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 1191968,
      "step": 1335
    },
    {
      "epoch": 0.3537019928731688,
      "grad_norm": 0.20502661168575287,
      "learning_rate": 4.986188188861355e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 1196416,
      "step": 1340
    },
    {
      "epoch": 0.3550217764286657,
      "grad_norm": 0.1767132580280304,
      "learning_rate": 4.9860849420525766e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 1201120,
      "step": 1345
    },
    {
      "epoch": 0.3563415599841626,
      "grad_norm": 0.17463809251785278,
      "learning_rate": 4.9859813118582575e-05,
      "loss": 0.0865,
      "num_input_tokens_seen": 1205952,
      "step": 1350
    },
    {
      "epoch": 0.3576613435396595,
      "grad_norm": 0.13529054820537567,
      "learning_rate": 4.98587729829438e-05,
      "loss": 0.1058,
      "num_input_tokens_seen": 1210368,
      "step": 1355
    },
    {
      "epoch": 0.3589811270951564,
      "grad_norm": 0.37380582094192505,
      "learning_rate": 4.985772901376983e-05,
      "loss": 0.1259,
      "num_input_tokens_seen": 1214848,
      "step": 1360
    },
    {
      "epoch": 0.3603009106506533,
      "grad_norm": 0.45343899726867676,
      "learning_rate": 4.9856681211221666e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 1219328,
      "step": 1365
    },
    {
      "epoch": 0.3616206942061502,
      "grad_norm": 0.5038014054298401,
      "learning_rate": 4.985562957546089e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 1224128,
      "step": 1370
    },
    {
      "epoch": 0.36294047776164706,
      "grad_norm": 0.09238982200622559,
      "learning_rate": 4.9854574106649686e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 1228512,
      "step": 1375
    },
    {
      "epoch": 0.364260261317144,
      "grad_norm": 0.17974753677845,
      "learning_rate": 4.985351480495081e-05,
      "loss": 0.1263,
      "num_input_tokens_seen": 1232864,
      "step": 1380
    },
    {
      "epoch": 0.3655800448726409,
      "grad_norm": 0.12978380918502808,
      "learning_rate": 4.985245167052762e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 1237248,
      "step": 1385
    },
    {
      "epoch": 0.3668998284281378,
      "grad_norm": 0.29572758078575134,
      "learning_rate": 4.9851384703544066e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 1241952,
      "step": 1390
    },
    {
      "epoch": 0.3682196119836347,
      "grad_norm": 0.227921724319458,
      "learning_rate": 4.985031390416469e-05,
      "loss": 0.0706,
      "num_input_tokens_seen": 1246048,
      "step": 1395
    },
    {
      "epoch": 0.36953939553913157,
      "grad_norm": 0.09136226773262024,
      "learning_rate": 4.984923927255461e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 1250688,
      "step": 1400
    },
    {
      "epoch": 0.36953939553913157,
      "eval_loss": 0.10004811733961105,
      "eval_runtime": 64.7302,
      "eval_samples_per_second": 104.047,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 1250688,
      "step": 1400
    },
    {
      "epoch": 0.37085917909462845,
      "grad_norm": 0.12683330476284027,
      "learning_rate": 4.984816080887958e-05,
      "loss": 0.1363,
      "num_input_tokens_seen": 1255200,
      "step": 1405
    },
    {
      "epoch": 0.3721789626501254,
      "grad_norm": 0.45369547605514526,
      "learning_rate": 4.9847078513305875e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 1259872,
      "step": 1410
    },
    {
      "epoch": 0.3734987462056223,
      "grad_norm": 0.33496251702308655,
      "learning_rate": 4.984599238600043e-05,
      "loss": 0.1219,
      "num_input_tokens_seen": 1264352,
      "step": 1415
    },
    {
      "epoch": 0.3748185297611192,
      "grad_norm": 0.16501431167125702,
      "learning_rate": 4.9844902427130716e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 1268736,
      "step": 1420
    },
    {
      "epoch": 0.37613831331661607,
      "grad_norm": 0.21569593250751495,
      "learning_rate": 4.984380863686482e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 1273376,
      "step": 1425
    },
    {
      "epoch": 0.37745809687211296,
      "grad_norm": 0.45600253343582153,
      "learning_rate": 4.984271101537143e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 1277664,
      "step": 1430
    },
    {
      "epoch": 0.37877788042760985,
      "grad_norm": 0.16238021850585938,
      "learning_rate": 4.9841609562819816e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 1281856,
      "step": 1435
    },
    {
      "epoch": 0.3800976639831068,
      "grad_norm": 0.2542300224304199,
      "learning_rate": 4.984050427937983e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 1286112,
      "step": 1440
    },
    {
      "epoch": 0.3814174475386037,
      "grad_norm": 0.2706325948238373,
      "learning_rate": 4.983939516522191e-05,
      "loss": 0.1064,
      "num_input_tokens_seen": 1290208,
      "step": 1445
    },
    {
      "epoch": 0.38273723109410057,
      "grad_norm": 0.28565552830696106,
      "learning_rate": 4.983828222051711e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 1294656,
      "step": 1450
    },
    {
      "epoch": 0.38405701464959746,
      "grad_norm": 0.1407063603401184,
      "learning_rate": 4.983716544543705e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 1299296,
      "step": 1455
    },
    {
      "epoch": 0.38537679820509435,
      "grad_norm": 0.40513288974761963,
      "learning_rate": 4.983604484015395e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 1303776,
      "step": 1460
    },
    {
      "epoch": 0.38669658176059124,
      "grad_norm": 0.10866474360227585,
      "learning_rate": 4.983492040484064e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 1308160,
      "step": 1465
    },
    {
      "epoch": 0.3880163653160882,
      "grad_norm": 0.6641761064529419,
      "learning_rate": 4.98337921396705e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 1312736,
      "step": 1470
    },
    {
      "epoch": 0.3893361488715851,
      "grad_norm": 0.3927798569202423,
      "learning_rate": 4.983266004481753e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 1317248,
      "step": 1475
    },
    {
      "epoch": 0.39065593242708196,
      "grad_norm": 0.1425616443157196,
      "learning_rate": 4.9831524120456316e-05,
      "loss": 0.1212,
      "num_input_tokens_seen": 1321696,
      "step": 1480
    },
    {
      "epoch": 0.39197571598257885,
      "grad_norm": 0.22967137396335602,
      "learning_rate": 4.9830384366762026e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 1326080,
      "step": 1485
    },
    {
      "epoch": 0.39329549953807574,
      "grad_norm": 0.20993387699127197,
      "learning_rate": 4.9829240783910436e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 1330688,
      "step": 1490
    },
    {
      "epoch": 0.39461528309357263,
      "grad_norm": 0.36247509717941284,
      "learning_rate": 4.982809337207789e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 1335200,
      "step": 1495
    },
    {
      "epoch": 0.3959350666490696,
      "grad_norm": 0.30440714955329895,
      "learning_rate": 4.9826942131441337e-05,
      "loss": 0.1143,
      "num_input_tokens_seen": 1339520,
      "step": 1500
    },
    {
      "epoch": 0.39725485020456647,
      "grad_norm": 0.1928398609161377,
      "learning_rate": 4.9825787062178315e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 1343872,
      "step": 1505
    },
    {
      "epoch": 0.39857463376006336,
      "grad_norm": 0.17904697358608246,
      "learning_rate": 4.9824628164466945e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 1348288,
      "step": 1510
    },
    {
      "epoch": 0.39989441731556025,
      "grad_norm": 0.2993811070919037,
      "learning_rate": 4.982346543848595e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 1352928,
      "step": 1515
    },
    {
      "epoch": 0.40121420087105714,
      "grad_norm": 0.4796639680862427,
      "learning_rate": 4.9822298884414626e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 1357312,
      "step": 1520
    },
    {
      "epoch": 0.402533984426554,
      "grad_norm": 0.19794157147407532,
      "learning_rate": 4.982112850243288e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 1361664,
      "step": 1525
    },
    {
      "epoch": 0.40385376798205097,
      "grad_norm": 0.0979548767209053,
      "learning_rate": 4.98199542927212e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 1366048,
      "step": 1530
    },
    {
      "epoch": 0.40517355153754786,
      "grad_norm": 0.8280842900276184,
      "learning_rate": 4.981877625546066e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 1370720,
      "step": 1535
    },
    {
      "epoch": 0.40649333509304475,
      "grad_norm": 0.051977045834064484,
      "learning_rate": 4.981759439083293e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 1375232,
      "step": 1540
    },
    {
      "epoch": 0.40781311864854164,
      "grad_norm": 0.25704362988471985,
      "learning_rate": 4.981640869902027e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 1379904,
      "step": 1545
    },
    {
      "epoch": 0.40913290220403853,
      "grad_norm": 0.3540825843811035,
      "learning_rate": 4.9815219180205517e-05,
      "loss": 0.1048,
      "num_input_tokens_seen": 1384544,
      "step": 1550
    },
    {
      "epoch": 0.4104526857595354,
      "grad_norm": 0.1316901594400406,
      "learning_rate": 4.9814025834572126e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 1388960,
      "step": 1555
    },
    {
      "epoch": 0.4117724693150323,
      "grad_norm": 0.1883913278579712,
      "learning_rate": 4.981282866230411e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 1393408,
      "step": 1560
    },
    {
      "epoch": 0.41309225287052925,
      "grad_norm": 0.5797374248504639,
      "learning_rate": 4.981162766358611e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 1398016,
      "step": 1565
    },
    {
      "epoch": 0.41441203642602614,
      "grad_norm": 0.45674562454223633,
      "learning_rate": 4.9810422838603316e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 1403008,
      "step": 1570
    },
    {
      "epoch": 0.41573181998152303,
      "grad_norm": 0.22351659834384918,
      "learning_rate": 4.9809214187541533e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 1407328,
      "step": 1575
    },
    {
      "epoch": 0.4170516035370199,
      "grad_norm": 0.30450570583343506,
      "learning_rate": 4.980800171058715e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 1411776,
      "step": 1580
    },
    {
      "epoch": 0.4183713870925168,
      "grad_norm": 0.49518051743507385,
      "learning_rate": 4.980678540792715e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 1415968,
      "step": 1585
    },
    {
      "epoch": 0.4196911706480137,
      "grad_norm": 0.3628959655761719,
      "learning_rate": 4.980556527974909e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 1420288,
      "step": 1590
    },
    {
      "epoch": 0.42101095420351065,
      "grad_norm": 0.20368343591690063,
      "learning_rate": 4.980434132624114e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 1425216,
      "step": 1595
    },
    {
      "epoch": 0.42233073775900754,
      "grad_norm": 0.42046883702278137,
      "learning_rate": 4.980311354759205e-05,
      "loss": 0.1518,
      "num_input_tokens_seen": 1429824,
      "step": 1600
    },
    {
      "epoch": 0.42233073775900754,
      "eval_loss": 0.09779907763004303,
      "eval_runtime": 64.8213,
      "eval_samples_per_second": 103.901,
      "eval_steps_per_second": 25.979,
      "num_input_tokens_seen": 1429824,
      "step": 1600
    },
    {
      "epoch": 0.4236505213145044,
      "grad_norm": 0.19637933373451233,
      "learning_rate": 4.980188194399116e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 1434144,
      "step": 1605
    },
    {
      "epoch": 0.4249703048700013,
      "grad_norm": 0.05599268525838852,
      "learning_rate": 4.9800646515628384e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 1438240,
      "step": 1610
    },
    {
      "epoch": 0.4262900884254982,
      "grad_norm": 0.7357270121574402,
      "learning_rate": 4.979940726269426e-05,
      "loss": 0.133,
      "num_input_tokens_seen": 1442784,
      "step": 1615
    },
    {
      "epoch": 0.4276098719809951,
      "grad_norm": 0.1672726571559906,
      "learning_rate": 4.979816418537988e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 1447200,
      "step": 1620
    },
    {
      "epoch": 0.42892965553649204,
      "grad_norm": 0.10277122259140015,
      "learning_rate": 4.979691728387696e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 1451552,
      "step": 1625
    },
    {
      "epoch": 0.43024943909198893,
      "grad_norm": 0.38115474581718445,
      "learning_rate": 4.979566655837776e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 1455904,
      "step": 1630
    },
    {
      "epoch": 0.4315692226474858,
      "grad_norm": 0.1787000149488449,
      "learning_rate": 4.9794412009075184e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 1460288,
      "step": 1635
    },
    {
      "epoch": 0.4328890062029827,
      "grad_norm": 0.04687885567545891,
      "learning_rate": 4.979315363616269e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 1464544,
      "step": 1640
    },
    {
      "epoch": 0.4342087897584796,
      "grad_norm": 0.3999486565589905,
      "learning_rate": 4.979189143983434e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 1468896,
      "step": 1645
    },
    {
      "epoch": 0.4355285733139765,
      "grad_norm": 0.32199081778526306,
      "learning_rate": 4.979062542028478e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 1473568,
      "step": 1650
    },
    {
      "epoch": 0.43684835686947343,
      "grad_norm": 0.7778474688529968,
      "learning_rate": 4.978935557770923e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 1478016,
      "step": 1655
    },
    {
      "epoch": 0.4381681404249703,
      "grad_norm": 0.12072020769119263,
      "learning_rate": 4.978808191230353e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 1482688,
      "step": 1660
    },
    {
      "epoch": 0.4394879239804672,
      "grad_norm": 0.34124454855918884,
      "learning_rate": 4.9786804424264085e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 1487104,
      "step": 1665
    },
    {
      "epoch": 0.4408077075359641,
      "grad_norm": 0.4942472577095032,
      "learning_rate": 4.978552311378792e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 1491712,
      "step": 1670
    },
    {
      "epoch": 0.442127491091461,
      "grad_norm": 0.2944345772266388,
      "learning_rate": 4.978423798107261e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 1495968,
      "step": 1675
    },
    {
      "epoch": 0.4434472746469579,
      "grad_norm": 0.5714668035507202,
      "learning_rate": 4.978294902631635e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 1500544,
      "step": 1680
    },
    {
      "epoch": 0.4447670582024548,
      "grad_norm": 0.13224996626377106,
      "learning_rate": 4.9781656249717914e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 1504864,
      "step": 1685
    },
    {
      "epoch": 0.4460868417579517,
      "grad_norm": 0.3308027684688568,
      "learning_rate": 4.9780359651476645e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 1509184,
      "step": 1690
    },
    {
      "epoch": 0.4474066253134486,
      "grad_norm": 0.5095565319061279,
      "learning_rate": 4.977905923179251e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 1513536,
      "step": 1695
    },
    {
      "epoch": 0.4487264088689455,
      "grad_norm": 0.18884052336215973,
      "learning_rate": 4.977775499086606e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 1517920,
      "step": 1700
    },
    {
      "epoch": 0.4500461924244424,
      "grad_norm": 0.0852644145488739,
      "learning_rate": 4.97764469288984e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 1522688,
      "step": 1705
    },
    {
      "epoch": 0.45136597597993927,
      "grad_norm": 0.6086108684539795,
      "learning_rate": 4.977513504609127e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 1527360,
      "step": 1710
    },
    {
      "epoch": 0.45268575953543616,
      "grad_norm": 0.30944541096687317,
      "learning_rate": 4.9773819342646965e-05,
      "loss": 0.096,
      "num_input_tokens_seen": 1531776,
      "step": 1715
    },
    {
      "epoch": 0.4540055430909331,
      "grad_norm": 0.5952541828155518,
      "learning_rate": 4.97724998187684e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 1536384,
      "step": 1720
    },
    {
      "epoch": 0.45532532664643,
      "grad_norm": 0.259219229221344,
      "learning_rate": 4.9771176474659045e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 1541184,
      "step": 1725
    },
    {
      "epoch": 0.4566451102019269,
      "grad_norm": 0.25308936834335327,
      "learning_rate": 4.976984931052299e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 1545568,
      "step": 1730
    },
    {
      "epoch": 0.4579648937574238,
      "grad_norm": 0.2521771788597107,
      "learning_rate": 4.976851832656489e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 1550048,
      "step": 1735
    },
    {
      "epoch": 0.45928467731292066,
      "grad_norm": 0.11577163636684418,
      "learning_rate": 4.9767183522990004e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 1554656,
      "step": 1740
    },
    {
      "epoch": 0.46060446086841755,
      "grad_norm": 0.647648811340332,
      "learning_rate": 4.9765844900004176e-05,
      "loss": 0.1031,
      "num_input_tokens_seen": 1559232,
      "step": 1745
    },
    {
      "epoch": 0.4619242444239145,
      "grad_norm": 0.08472394943237305,
      "learning_rate": 4.9764502457813834e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 1564192,
      "step": 1750
    },
    {
      "epoch": 0.4632440279794114,
      "grad_norm": 0.07346078753471375,
      "learning_rate": 4.9763156196626005e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 1568672,
      "step": 1755
    },
    {
      "epoch": 0.4645638115349083,
      "grad_norm": 0.22509826719760895,
      "learning_rate": 4.97618061166483e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 1573184,
      "step": 1760
    },
    {
      "epoch": 0.46588359509040517,
      "grad_norm": 0.28262823820114136,
      "learning_rate": 4.9760452218088915e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 1577568,
      "step": 1765
    },
    {
      "epoch": 0.46720337864590206,
      "grad_norm": 0.3671574294567108,
      "learning_rate": 4.975909450115663e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 1582144,
      "step": 1770
    },
    {
      "epoch": 0.46852316220139895,
      "grad_norm": 0.07467343658208847,
      "learning_rate": 4.975773296606084e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 1586656,
      "step": 1775
    },
    {
      "epoch": 0.4698429457568959,
      "grad_norm": 0.33608803153038025,
      "learning_rate": 4.97563676130115e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 1590976,
      "step": 1780
    },
    {
      "epoch": 0.4711627293123928,
      "grad_norm": 0.1629789173603058,
      "learning_rate": 4.9754998442219166e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 1595360,
      "step": 1785
    },
    {
      "epoch": 0.47248251286788967,
      "grad_norm": 0.5906822681427002,
      "learning_rate": 4.9753625453894984e-05,
      "loss": 0.1074,
      "num_input_tokens_seen": 1599744,
      "step": 1790
    },
    {
      "epoch": 0.47380229642338656,
      "grad_norm": 0.07769432663917542,
      "learning_rate": 4.975224864825068e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 1604096,
      "step": 1795
    },
    {
      "epoch": 0.47512207997888345,
      "grad_norm": 0.136910080909729,
      "learning_rate": 4.9750868025498576e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 1608736,
      "step": 1800
    },
    {
      "epoch": 0.47512207997888345,
      "eval_loss": 0.09629397094249725,
      "eval_runtime": 64.7424,
      "eval_samples_per_second": 104.028,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 1608736,
      "step": 1800
    },
    {
      "epoch": 0.47644186353438034,
      "grad_norm": 0.3345334529876709,
      "learning_rate": 4.974948358585158e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 1613440,
      "step": 1805
    },
    {
      "epoch": 0.4777616470898773,
      "grad_norm": 0.3737996816635132,
      "learning_rate": 4.9748095329523205e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 1618144,
      "step": 1810
    },
    {
      "epoch": 0.4790814306453742,
      "grad_norm": 0.14992079138755798,
      "learning_rate": 4.974670325672752e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 1622560,
      "step": 1815
    },
    {
      "epoch": 0.48040121420087106,
      "grad_norm": 0.22753696143627167,
      "learning_rate": 4.974530736767921e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 1626976,
      "step": 1820
    },
    {
      "epoch": 0.48172099775636795,
      "grad_norm": 0.48472854495048523,
      "learning_rate": 4.9743907662593524e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 1631616,
      "step": 1825
    },
    {
      "epoch": 0.48304078131186484,
      "grad_norm": 0.19427478313446045,
      "learning_rate": 4.974250414168633e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 1636224,
      "step": 1830
    },
    {
      "epoch": 0.48436056486736173,
      "grad_norm": 0.18898725509643555,
      "learning_rate": 4.974109680517407e-05,
      "loss": 0.1095,
      "num_input_tokens_seen": 1640896,
      "step": 1835
    },
    {
      "epoch": 0.4856803484228587,
      "grad_norm": 0.4109775424003601,
      "learning_rate": 4.973968565327376e-05,
      "loss": 0.1206,
      "num_input_tokens_seen": 1645504,
      "step": 1840
    },
    {
      "epoch": 0.48700013197835557,
      "grad_norm": 0.25475284457206726,
      "learning_rate": 4.973827068620303e-05,
      "loss": 0.1392,
      "num_input_tokens_seen": 1649952,
      "step": 1845
    },
    {
      "epoch": 0.48831991553385246,
      "grad_norm": 0.3834966719150543,
      "learning_rate": 4.973685190418008e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 1654432,
      "step": 1850
    },
    {
      "epoch": 0.48963969908934935,
      "grad_norm": 0.32690882682800293,
      "learning_rate": 4.97354293074237e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 1658720,
      "step": 1855
    },
    {
      "epoch": 0.49095948264484623,
      "grad_norm": 0.20059968531131744,
      "learning_rate": 4.9734002896153276e-05,
      "loss": 0.1034,
      "num_input_tokens_seen": 1663040,
      "step": 1860
    },
    {
      "epoch": 0.4922792662003431,
      "grad_norm": 0.22899504005908966,
      "learning_rate": 4.973257267058877e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 1667776,
      "step": 1865
    },
    {
      "epoch": 0.49359904975584007,
      "grad_norm": 0.24791018664836884,
      "learning_rate": 4.973113863095076e-05,
      "loss": 0.093,
      "num_input_tokens_seen": 1672384,
      "step": 1870
    },
    {
      "epoch": 0.49491883331133696,
      "grad_norm": 0.15553689002990723,
      "learning_rate": 4.9729700777460384e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 1676768,
      "step": 1875
    },
    {
      "epoch": 0.49623861686683385,
      "grad_norm": 0.3527602553367615,
      "learning_rate": 4.972825911033937e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 1681344,
      "step": 1880
    },
    {
      "epoch": 0.49755840042233074,
      "grad_norm": 0.6908682584762573,
      "learning_rate": 4.9726813629810056e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 1685824,
      "step": 1885
    },
    {
      "epoch": 0.4988781839778276,
      "grad_norm": 0.21874700486660004,
      "learning_rate": 4.9725364336095326e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 1690208,
      "step": 1890
    },
    {
      "epoch": 0.5001979675333246,
      "grad_norm": 0.4724104106426239,
      "learning_rate": 4.972391122941871e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 1694784,
      "step": 1895
    },
    {
      "epoch": 0.5015177510888215,
      "grad_norm": 0.49158525466918945,
      "learning_rate": 4.972245431000428e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 1699232,
      "step": 1900
    },
    {
      "epoch": 0.5028375346443184,
      "grad_norm": 0.3119906187057495,
      "learning_rate": 4.972099357807671e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 1703712,
      "step": 1905
    },
    {
      "epoch": 0.5041573181998152,
      "grad_norm": 0.11862573027610779,
      "learning_rate": 4.971952903386127e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 1707968,
      "step": 1910
    },
    {
      "epoch": 0.5054771017553121,
      "grad_norm": 0.3632994592189789,
      "learning_rate": 4.971806067758381e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 1712224,
      "step": 1915
    },
    {
      "epoch": 0.506796885310809,
      "grad_norm": 0.4220863878726959,
      "learning_rate": 4.971658850947076e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 1716416,
      "step": 1920
    },
    {
      "epoch": 0.5081166688663059,
      "grad_norm": 0.13918663561344147,
      "learning_rate": 4.9715112529749165e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 1720992,
      "step": 1925
    },
    {
      "epoch": 0.5094364524218028,
      "grad_norm": 0.20145854353904724,
      "learning_rate": 4.9713632738646624e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 1725184,
      "step": 1930
    },
    {
      "epoch": 0.5107562359772997,
      "grad_norm": 0.2000434249639511,
      "learning_rate": 4.971214913639134e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 1729664,
      "step": 1935
    },
    {
      "epoch": 0.5120760195327966,
      "grad_norm": 0.13887614011764526,
      "learning_rate": 4.9710661723212104e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 1734208,
      "step": 1940
    },
    {
      "epoch": 0.5133958030882935,
      "grad_norm": 0.06072315573692322,
      "learning_rate": 4.9709170499338295e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 1738688,
      "step": 1945
    },
    {
      "epoch": 0.5147155866437905,
      "grad_norm": 0.4405849575996399,
      "learning_rate": 4.9707675464999895e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 1743296,
      "step": 1950
    },
    {
      "epoch": 0.5160353701992874,
      "grad_norm": 0.22208745777606964,
      "learning_rate": 4.970617662042743e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 1747776,
      "step": 1955
    },
    {
      "epoch": 0.5173551537547842,
      "grad_norm": 0.21280348300933838,
      "learning_rate": 4.970467396585206e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 1752096,
      "step": 1960
    },
    {
      "epoch": 0.5186749373102811,
      "grad_norm": 0.24638284742832184,
      "learning_rate": 4.97031675015055e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 1756416,
      "step": 1965
    },
    {
      "epoch": 0.519994720865778,
      "grad_norm": 0.10897671431303024,
      "learning_rate": 4.9701657227620075e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 1760864,
      "step": 1970
    },
    {
      "epoch": 0.5213145044212749,
      "grad_norm": 0.05196571350097656,
      "learning_rate": 4.9700143144428685e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 1765312,
      "step": 1975
    },
    {
      "epoch": 0.5226342879767718,
      "grad_norm": 0.3574368953704834,
      "learning_rate": 4.969862525216482e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 1769824,
      "step": 1980
    },
    {
      "epoch": 0.5239540715322687,
      "grad_norm": 0.22640113532543182,
      "learning_rate": 4.9697103551062556e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 1774208,
      "step": 1985
    },
    {
      "epoch": 0.5252738550877656,
      "grad_norm": 0.16414359211921692,
      "learning_rate": 4.9695578041356565e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 1778528,
      "step": 1990
    },
    {
      "epoch": 0.5265936386432625,
      "grad_norm": 0.6334508657455444,
      "learning_rate": 4.969404872328209e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 1783008,
      "step": 1995
    },
    {
      "epoch": 0.5279134221987594,
      "grad_norm": 0.840112030506134,
      "learning_rate": 4.969251559707498e-05,
      "loss": 0.1294,
      "num_input_tokens_seen": 1787552,
      "step": 2000
    },
    {
      "epoch": 0.5279134221987594,
      "eval_loss": 0.09432559460401535,
      "eval_runtime": 64.7151,
      "eval_samples_per_second": 104.072,
      "eval_steps_per_second": 26.022,
      "num_input_tokens_seen": 1787552,
      "step": 2000
    },
    {
      "epoch": 0.5292332057542563,
      "grad_norm": 0.39028042554855347,
      "learning_rate": 4.9690978662971674e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 1791936,
      "step": 2005
    },
    {
      "epoch": 0.5305529893097533,
      "grad_norm": 0.4104963541030884,
      "learning_rate": 4.968943792120916e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 1796480,
      "step": 2010
    },
    {
      "epoch": 0.5318727728652501,
      "grad_norm": 0.1568603515625,
      "learning_rate": 4.9687893372025046e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 1800416,
      "step": 2015
    },
    {
      "epoch": 0.533192556420747,
      "grad_norm": 0.5351195335388184,
      "learning_rate": 4.9686345015657535e-05,
      "loss": 0.0945,
      "num_input_tokens_seen": 1804896,
      "step": 2020
    },
    {
      "epoch": 0.5345123399762439,
      "grad_norm": 0.24083472788333893,
      "learning_rate": 4.968479285234538e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 1809568,
      "step": 2025
    },
    {
      "epoch": 0.5358321235317408,
      "grad_norm": 0.09859143197536469,
      "learning_rate": 4.9683236882327974e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 1814624,
      "step": 2030
    },
    {
      "epoch": 0.5371519070872377,
      "grad_norm": 0.3288732171058655,
      "learning_rate": 4.968167710584526e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 1819136,
      "step": 2035
    },
    {
      "epoch": 0.5384716906427346,
      "grad_norm": 0.16594676673412323,
      "learning_rate": 4.968011352313775e-05,
      "loss": 0.1078,
      "num_input_tokens_seen": 1823776,
      "step": 2040
    },
    {
      "epoch": 0.5397914741982315,
      "grad_norm": 0.090375155210495,
      "learning_rate": 4.967854613444659e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 1828576,
      "step": 2045
    },
    {
      "epoch": 0.5411112577537284,
      "grad_norm": 0.17054040729999542,
      "learning_rate": 4.967697494001349e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 1833152,
      "step": 2050
    },
    {
      "epoch": 0.5424310413092253,
      "grad_norm": 0.2564355432987213,
      "learning_rate": 4.9675399940080736e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 1837376,
      "step": 2055
    },
    {
      "epoch": 0.5437508248647221,
      "grad_norm": 0.09312942624092102,
      "learning_rate": 4.9673821134891226e-05,
      "loss": 0.1044,
      "num_input_tokens_seen": 1841824,
      "step": 2060
    },
    {
      "epoch": 0.545070608420219,
      "grad_norm": 0.27085360884666443,
      "learning_rate": 4.967223852468842e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 1846240,
      "step": 2065
    },
    {
      "epoch": 0.5463903919757159,
      "grad_norm": 0.1996983140707016,
      "learning_rate": 4.967065210971639e-05,
      "loss": 0.1298,
      "num_input_tokens_seen": 1850816,
      "step": 2070
    },
    {
      "epoch": 0.5477101755312129,
      "grad_norm": 0.10482937097549438,
      "learning_rate": 4.966906189021977e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 1855328,
      "step": 2075
    },
    {
      "epoch": 0.5490299590867098,
      "grad_norm": 0.3350965678691864,
      "learning_rate": 4.966746786644379e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 1859424,
      "step": 2080
    },
    {
      "epoch": 0.5503497426422067,
      "grad_norm": 0.1711457520723343,
      "learning_rate": 4.966587003863429e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 1864000,
      "step": 2085
    },
    {
      "epoch": 0.5516695261977036,
      "grad_norm": 0.42763251066207886,
      "learning_rate": 4.966426840703765e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 1868416,
      "step": 2090
    },
    {
      "epoch": 0.5529893097532005,
      "grad_norm": 0.22510138154029846,
      "learning_rate": 4.9662662971900875e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 1872864,
      "step": 2095
    },
    {
      "epoch": 0.5543090933086974,
      "grad_norm": 0.08934833109378815,
      "learning_rate": 4.9661053733471534e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 1877312,
      "step": 2100
    },
    {
      "epoch": 0.5556288768641943,
      "grad_norm": 0.11473236232995987,
      "learning_rate": 4.965944069199781e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 1881824,
      "step": 2105
    },
    {
      "epoch": 0.5569486604196912,
      "grad_norm": 0.2178570181131363,
      "learning_rate": 4.965782384772842e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 1886400,
      "step": 2110
    },
    {
      "epoch": 0.558268443975188,
      "grad_norm": 0.4115614593029022,
      "learning_rate": 4.9656203200912734e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 1891072,
      "step": 2115
    },
    {
      "epoch": 0.5595882275306849,
      "grad_norm": 0.08171983063220978,
      "learning_rate": 4.965457875180067e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 1895328,
      "step": 2120
    },
    {
      "epoch": 0.5609080110861818,
      "grad_norm": 0.08847381174564362,
      "learning_rate": 4.9652950500642724e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 1899904,
      "step": 2125
    },
    {
      "epoch": 0.5622277946416787,
      "grad_norm": 0.12006539106369019,
      "learning_rate": 4.965131844769001e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 1904288,
      "step": 2130
    },
    {
      "epoch": 0.5635475781971757,
      "grad_norm": 0.24875101447105408,
      "learning_rate": 4.96496825931942e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 1908960,
      "step": 2135
    },
    {
      "epoch": 0.5648673617526726,
      "grad_norm": 0.2574158310890198,
      "learning_rate": 4.9648042937407566e-05,
      "loss": 0.1143,
      "num_input_tokens_seen": 1913248,
      "step": 2140
    },
    {
      "epoch": 0.5661871453081695,
      "grad_norm": 0.358491986989975,
      "learning_rate": 4.964639948058297e-05,
      "loss": 0.1472,
      "num_input_tokens_seen": 1917792,
      "step": 2145
    },
    {
      "epoch": 0.5675069288636664,
      "grad_norm": 0.5776029229164124,
      "learning_rate": 4.9644752222973846e-05,
      "loss": 0.1022,
      "num_input_tokens_seen": 1922464,
      "step": 2150
    },
    {
      "epoch": 0.5688267124191633,
      "grad_norm": 0.192543625831604,
      "learning_rate": 4.964310116483422e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 1927168,
      "step": 2155
    },
    {
      "epoch": 0.5701464959746602,
      "grad_norm": 0.4438026249408722,
      "learning_rate": 4.964144630641872e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 1932096,
      "step": 2160
    },
    {
      "epoch": 0.571466279530157,
      "grad_norm": 0.303320050239563,
      "learning_rate": 4.9639787647982525e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 1936384,
      "step": 2165
    },
    {
      "epoch": 0.5727860630856539,
      "grad_norm": 0.22063741087913513,
      "learning_rate": 4.963812518978143e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 1940960,
      "step": 2170
    },
    {
      "epoch": 0.5741058466411508,
      "grad_norm": 0.4069156348705292,
      "learning_rate": 4.963645893207182e-05,
      "loss": 0.1086,
      "num_input_tokens_seen": 1945408,
      "step": 2175
    },
    {
      "epoch": 0.5754256301966477,
      "grad_norm": 0.16308940947055817,
      "learning_rate": 4.963478887511063e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 1949888,
      "step": 2180
    },
    {
      "epoch": 0.5767454137521446,
      "grad_norm": 0.1811479777097702,
      "learning_rate": 4.963311501915542e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 1954304,
      "step": 2185
    },
    {
      "epoch": 0.5780651973076415,
      "grad_norm": 0.08718302100896835,
      "learning_rate": 4.963143736446432e-05,
      "loss": 0.0984,
      "num_input_tokens_seen": 1958912,
      "step": 2190
    },
    {
      "epoch": 0.5793849808631385,
      "grad_norm": 0.35911914706230164,
      "learning_rate": 4.962975591129603e-05,
      "loss": 0.0937,
      "num_input_tokens_seen": 1963456,
      "step": 2195
    },
    {
      "epoch": 0.5807047644186354,
      "grad_norm": 0.18520918488502502,
      "learning_rate": 4.962807065990986e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 1968064,
      "step": 2200
    },
    {
      "epoch": 0.5807047644186354,
      "eval_loss": 0.09268565475940704,
      "eval_runtime": 64.744,
      "eval_samples_per_second": 104.025,
      "eval_steps_per_second": 26.01,
      "num_input_tokens_seen": 1968064,
      "step": 2200
    },
    {
      "epoch": 0.5820245479741323,
      "grad_norm": 0.27053436636924744,
      "learning_rate": 4.9626381610565714e-05,
      "loss": 0.0981,
      "num_input_tokens_seen": 1972480,
      "step": 2205
    },
    {
      "epoch": 0.5833443315296292,
      "grad_norm": 0.1914701610803604,
      "learning_rate": 4.9624688763524043e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 1976832,
      "step": 2210
    },
    {
      "epoch": 0.5846641150851261,
      "grad_norm": 0.03107849694788456,
      "learning_rate": 4.962299211904591e-05,
      "loss": 0.0698,
      "num_input_tokens_seen": 1981344,
      "step": 2215
    },
    {
      "epoch": 0.585983898640623,
      "grad_norm": 0.2835747003555298,
      "learning_rate": 4.962129167739296e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 1986080,
      "step": 2220
    },
    {
      "epoch": 0.5873036821961198,
      "grad_norm": 0.6673892736434937,
      "learning_rate": 4.961958743882742e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 1990496,
      "step": 2225
    },
    {
      "epoch": 0.5886234657516167,
      "grad_norm": 0.4252258241176605,
      "learning_rate": 4.961787940361211e-05,
      "loss": 0.0917,
      "num_input_tokens_seen": 1994880,
      "step": 2230
    },
    {
      "epoch": 0.5899432493071136,
      "grad_norm": 0.29872697591781616,
      "learning_rate": 4.961616757201043e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 1999232,
      "step": 2235
    },
    {
      "epoch": 0.5912630328626105,
      "grad_norm": 0.15087249875068665,
      "learning_rate": 4.961445194428637e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 2003680,
      "step": 2240
    },
    {
      "epoch": 0.5925828164181074,
      "grad_norm": 0.46493667364120483,
      "learning_rate": 4.9612732520704486e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 2008128,
      "step": 2245
    },
    {
      "epoch": 0.5939025999736043,
      "grad_norm": 0.12040568888187408,
      "learning_rate": 4.961100930152994e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 2012864,
      "step": 2250
    },
    {
      "epoch": 0.5952223835291012,
      "grad_norm": 0.3718259632587433,
      "learning_rate": 4.960928228702849e-05,
      "loss": 0.115,
      "num_input_tokens_seen": 2017248,
      "step": 2255
    },
    {
      "epoch": 0.5965421670845982,
      "grad_norm": 0.4353402853012085,
      "learning_rate": 4.960755147746645e-05,
      "loss": 0.1249,
      "num_input_tokens_seen": 2021216,
      "step": 2260
    },
    {
      "epoch": 0.5978619506400951,
      "grad_norm": 0.4015991985797882,
      "learning_rate": 4.9605816873110736e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 2025600,
      "step": 2265
    },
    {
      "epoch": 0.599181734195592,
      "grad_norm": 0.47197839617729187,
      "learning_rate": 4.960407847422883e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 2029760,
      "step": 2270
    },
    {
      "epoch": 0.6005015177510888,
      "grad_norm": 0.09070291370153427,
      "learning_rate": 4.960233628108885e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 2034272,
      "step": 2275
    },
    {
      "epoch": 0.6018213013065857,
      "grad_norm": 0.39558684825897217,
      "learning_rate": 4.960059029395942e-05,
      "loss": 0.1061,
      "num_input_tokens_seen": 2038464,
      "step": 2280
    },
    {
      "epoch": 0.6031410848620826,
      "grad_norm": 0.33149096369743347,
      "learning_rate": 4.959884051310983e-05,
      "loss": 0.1393,
      "num_input_tokens_seen": 2042528,
      "step": 2285
    },
    {
      "epoch": 0.6044608684175795,
      "grad_norm": 0.1789817363023758,
      "learning_rate": 4.959708693880991e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 2047008,
      "step": 2290
    },
    {
      "epoch": 0.6057806519730764,
      "grad_norm": 0.29706957936286926,
      "learning_rate": 4.9595329571330074e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 2051776,
      "step": 2295
    },
    {
      "epoch": 0.6071004355285733,
      "grad_norm": 0.22138841450214386,
      "learning_rate": 4.9593568410941326e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 2056384,
      "step": 2300
    },
    {
      "epoch": 0.6084202190840702,
      "grad_norm": 0.46790894865989685,
      "learning_rate": 4.959180345791528e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 2060512,
      "step": 2305
    },
    {
      "epoch": 0.6097400026395671,
      "grad_norm": 0.30988621711730957,
      "learning_rate": 4.9590034712524086e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 2064864,
      "step": 2310
    },
    {
      "epoch": 0.611059786195064,
      "grad_norm": 0.35101160407066345,
      "learning_rate": 4.958826217504053e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 2068896,
      "step": 2315
    },
    {
      "epoch": 0.612379569750561,
      "grad_norm": 0.3100353181362152,
      "learning_rate": 4.958648584573795e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 2073152,
      "step": 2320
    },
    {
      "epoch": 0.6136993533060578,
      "grad_norm": 0.3215196430683136,
      "learning_rate": 4.958470572489028e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 2077664,
      "step": 2325
    },
    {
      "epoch": 0.6150191368615547,
      "grad_norm": 0.1023888811469078,
      "learning_rate": 4.958292181277203e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 2082496,
      "step": 2330
    },
    {
      "epoch": 0.6163389204170516,
      "grad_norm": 0.7924720644950867,
      "learning_rate": 4.958113410965832e-05,
      "loss": 0.1278,
      "num_input_tokens_seen": 2086880,
      "step": 2335
    },
    {
      "epoch": 0.6176587039725485,
      "grad_norm": 0.07269216328859329,
      "learning_rate": 4.957934261582481e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 2091488,
      "step": 2340
    },
    {
      "epoch": 0.6189784875280454,
      "grad_norm": 0.12217990309000015,
      "learning_rate": 4.95775473315478e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 2096000,
      "step": 2345
    },
    {
      "epoch": 0.6202982710835423,
      "grad_norm": 0.33540695905685425,
      "learning_rate": 4.9575748257104124e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 2100544,
      "step": 2350
    },
    {
      "epoch": 0.6216180546390392,
      "grad_norm": 0.2551947236061096,
      "learning_rate": 4.9573945392771224e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 2104928,
      "step": 2355
    },
    {
      "epoch": 0.6229378381945361,
      "grad_norm": 0.26239874958992004,
      "learning_rate": 4.9572138738827134e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 2109120,
      "step": 2360
    },
    {
      "epoch": 0.624257621750033,
      "grad_norm": 0.15608693659305573,
      "learning_rate": 4.957032829555046e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 2113632,
      "step": 2365
    },
    {
      "epoch": 0.6255774053055299,
      "grad_norm": 0.15318377315998077,
      "learning_rate": 4.956851406322039e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 2118496,
      "step": 2370
    },
    {
      "epoch": 0.6268971888610267,
      "grad_norm": 0.17658647894859314,
      "learning_rate": 4.9566696042116704e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 2123008,
      "step": 2375
    },
    {
      "epoch": 0.6282169724165237,
      "grad_norm": 0.28433656692504883,
      "learning_rate": 4.9564874232519766e-05,
      "loss": 0.097,
      "num_input_tokens_seen": 2127488,
      "step": 2380
    },
    {
      "epoch": 0.6295367559720206,
      "grad_norm": 0.4256756007671356,
      "learning_rate": 4.9563048634710516e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 2132032,
      "step": 2385
    },
    {
      "epoch": 0.6308565395275175,
      "grad_norm": 0.1269175261259079,
      "learning_rate": 4.956121924897049e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 2136416,
      "step": 2390
    },
    {
      "epoch": 0.6321763230830144,
      "grad_norm": 0.11275062710046768,
      "learning_rate": 4.955938607558181e-05,
      "loss": 0.1297,
      "num_input_tokens_seen": 2140608,
      "step": 2395
    },
    {
      "epoch": 0.6334961066385113,
      "grad_norm": 0.26657289266586304,
      "learning_rate": 4.955754911482715e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 2145056,
      "step": 2400
    },
    {
      "epoch": 0.6334961066385113,
      "eval_loss": 0.09172502905130386,
      "eval_runtime": 64.8864,
      "eval_samples_per_second": 103.797,
      "eval_steps_per_second": 25.953,
      "num_input_tokens_seen": 2145056,
      "step": 2400
    },
    {
      "epoch": 0.6348158901940082,
      "grad_norm": 0.19227822124958038,
      "learning_rate": 4.9555708366989804e-05,
      "loss": 0.1678,
      "num_input_tokens_seen": 2149376,
      "step": 2405
    },
    {
      "epoch": 0.6361356737495051,
      "grad_norm": 0.25120624899864197,
      "learning_rate": 4.9553863832353655e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 2153728,
      "step": 2410
    },
    {
      "epoch": 0.637455457305002,
      "grad_norm": 0.24276836216449738,
      "learning_rate": 4.955201551120313e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 2158464,
      "step": 2415
    },
    {
      "epoch": 0.6387752408604989,
      "grad_norm": 0.2187170386314392,
      "learning_rate": 4.955016340382328e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 2162912,
      "step": 2420
    },
    {
      "epoch": 0.6400950244159958,
      "grad_norm": 0.3755386769771576,
      "learning_rate": 4.954830751049972e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 2167552,
      "step": 2425
    },
    {
      "epoch": 0.6414148079714926,
      "grad_norm": 0.48159822821617126,
      "learning_rate": 4.954644783151864e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 2171904,
      "step": 2430
    },
    {
      "epoch": 0.6427345915269895,
      "grad_norm": 0.2776750922203064,
      "learning_rate": 4.954458436716684e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 2176480,
      "step": 2435
    },
    {
      "epoch": 0.6440543750824864,
      "grad_norm": 0.2558130621910095,
      "learning_rate": 4.954271711773168e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 2181024,
      "step": 2440
    },
    {
      "epoch": 0.6453741586379834,
      "grad_norm": 0.21499189734458923,
      "learning_rate": 4.9540846083501115e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 2185376,
      "step": 2445
    },
    {
      "epoch": 0.6466939421934803,
      "grad_norm": 0.32245105504989624,
      "learning_rate": 4.953897126476369e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 2189792,
      "step": 2450
    },
    {
      "epoch": 0.6480137257489772,
      "grad_norm": 0.1765044629573822,
      "learning_rate": 4.9537092661808514e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 2194272,
      "step": 2455
    },
    {
      "epoch": 0.6493335093044741,
      "grad_norm": 0.21801163256168365,
      "learning_rate": 4.9535210274925306e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 2198496,
      "step": 2460
    },
    {
      "epoch": 0.650653292859971,
      "grad_norm": 0.18894007802009583,
      "learning_rate": 4.953332410440435e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 2202720,
      "step": 2465
    },
    {
      "epoch": 0.6519730764154679,
      "grad_norm": 0.30771195888519287,
      "learning_rate": 4.9531434150536496e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 2207488,
      "step": 2470
    },
    {
      "epoch": 0.6532928599709648,
      "grad_norm": 0.23108595609664917,
      "learning_rate": 4.952954041361322e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 2211904,
      "step": 2475
    },
    {
      "epoch": 0.6546126435264616,
      "grad_norm": 0.16950362920761108,
      "learning_rate": 4.952764289392655e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 2216128,
      "step": 2480
    },
    {
      "epoch": 0.6559324270819585,
      "grad_norm": 0.21451860666275024,
      "learning_rate": 4.952574159176912e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 2220736,
      "step": 2485
    },
    {
      "epoch": 0.6572522106374554,
      "grad_norm": 0.22987647354602814,
      "learning_rate": 4.952383650743413e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 2225248,
      "step": 2490
    },
    {
      "epoch": 0.6585719941929523,
      "grad_norm": 0.28814026713371277,
      "learning_rate": 4.952192764121536e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 2229664,
      "step": 2495
    },
    {
      "epoch": 0.6598917777484492,
      "grad_norm": 0.151066392660141,
      "learning_rate": 4.9520014993407185e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 2233984,
      "step": 2500
    },
    {
      "epoch": 0.6612115613039462,
      "grad_norm": 0.11902409046888351,
      "learning_rate": 4.951809856430456e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 2238176,
      "step": 2505
    },
    {
      "epoch": 0.6625313448594431,
      "grad_norm": 0.16067589819431305,
      "learning_rate": 4.951617835420303e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 2242528,
      "step": 2510
    },
    {
      "epoch": 0.66385112841494,
      "grad_norm": 0.3467065393924713,
      "learning_rate": 4.951425436339869e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 2247136,
      "step": 2515
    },
    {
      "epoch": 0.6651709119704369,
      "grad_norm": 0.06977517902851105,
      "learning_rate": 4.9512326592188274e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 2251328,
      "step": 2520
    },
    {
      "epoch": 0.6664906955259338,
      "grad_norm": 0.18142037093639374,
      "learning_rate": 4.9510395040869054e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 2255744,
      "step": 2525
    },
    {
      "epoch": 0.6678104790814307,
      "grad_norm": 0.30557191371917725,
      "learning_rate": 4.9508459709738905e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 2260384,
      "step": 2530
    },
    {
      "epoch": 0.6691302626369275,
      "grad_norm": 0.22357675433158875,
      "learning_rate": 4.950652059909627e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 2264800,
      "step": 2535
    },
    {
      "epoch": 0.6704500461924244,
      "grad_norm": 0.1345997303724289,
      "learning_rate": 4.95045777092402e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 2269376,
      "step": 2540
    },
    {
      "epoch": 0.6717698297479213,
      "grad_norm": 0.26949450373649597,
      "learning_rate": 4.950263104047031e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 2273888,
      "step": 2545
    },
    {
      "epoch": 0.6730896133034182,
      "grad_norm": 0.07178498059511185,
      "learning_rate": 4.9500680593086775e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 2278624,
      "step": 2550
    },
    {
      "epoch": 0.6744093968589151,
      "grad_norm": 0.1158762127161026,
      "learning_rate": 4.94987263673904e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 2283232,
      "step": 2555
    },
    {
      "epoch": 0.675729180414412,
      "grad_norm": 0.23397794365882874,
      "learning_rate": 4.949676836368256e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 2287456,
      "step": 2560
    },
    {
      "epoch": 0.677048963969909,
      "grad_norm": 0.27922961115837097,
      "learning_rate": 4.949480658226518e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 2292320,
      "step": 2565
    },
    {
      "epoch": 0.6783687475254059,
      "grad_norm": 0.10881134122610092,
      "learning_rate": 4.949284102344082e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 2296736,
      "step": 2570
    },
    {
      "epoch": 0.6796885310809028,
      "grad_norm": 0.12265644967556,
      "learning_rate": 4.9490871687512565e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 2301312,
      "step": 2575
    },
    {
      "epoch": 0.6810083146363997,
      "grad_norm": 0.22348056733608246,
      "learning_rate": 4.948889857478413e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 2305792,
      "step": 2580
    },
    {
      "epoch": 0.6823280981918965,
      "grad_norm": 0.1187608391046524,
      "learning_rate": 4.948692168555978e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 2310272,
      "step": 2585
    },
    {
      "epoch": 0.6836478817473934,
      "grad_norm": 0.18306167423725128,
      "learning_rate": 4.94849410201444e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 2314400,
      "step": 2590
    },
    {
      "epoch": 0.6849676653028903,
      "grad_norm": 0.34867310523986816,
      "learning_rate": 4.948295657884341e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 2319200,
      "step": 2595
    },
    {
      "epoch": 0.6862874488583872,
      "grad_norm": 0.10164482146501541,
      "learning_rate": 4.9480968361962835e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 2323552,
      "step": 2600
    },
    {
      "epoch": 0.6862874488583872,
      "eval_loss": 0.09086666256189346,
      "eval_runtime": 64.7204,
      "eval_samples_per_second": 104.063,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 2323552,
      "step": 2600
    },
    {
      "epoch": 0.6876072324138841,
      "grad_norm": 0.39820751547813416,
      "learning_rate": 4.9478976369809305e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 2328352,
      "step": 2605
    },
    {
      "epoch": 0.688927015969381,
      "grad_norm": 0.2730591297149658,
      "learning_rate": 4.947698060268999e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 2332864,
      "step": 2610
    },
    {
      "epoch": 0.6902467995248779,
      "grad_norm": 0.07695958018302917,
      "learning_rate": 4.9474981060912665e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 2337056,
      "step": 2615
    },
    {
      "epoch": 0.6915665830803748,
      "grad_norm": 0.05772676318883896,
      "learning_rate": 4.94729777447857e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 2341664,
      "step": 2620
    },
    {
      "epoch": 0.6928863666358717,
      "grad_norm": 0.5159198641777039,
      "learning_rate": 4.947097065461801e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 2346144,
      "step": 2625
    },
    {
      "epoch": 0.6942061501913687,
      "grad_norm": 0.27272090315818787,
      "learning_rate": 4.9468959790719125e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 2350496,
      "step": 2630
    },
    {
      "epoch": 0.6955259337468656,
      "grad_norm": 0.20330604910850525,
      "learning_rate": 4.9466945153399146e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 2354976,
      "step": 2635
    },
    {
      "epoch": 0.6968457173023624,
      "grad_norm": 0.3083922564983368,
      "learning_rate": 4.9464926742968755e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 2359520,
      "step": 2640
    },
    {
      "epoch": 0.6981655008578593,
      "grad_norm": 0.40600186586380005,
      "learning_rate": 4.946290455973921e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 2363808,
      "step": 2645
    },
    {
      "epoch": 0.6994852844133562,
      "grad_norm": 0.44406139850616455,
      "learning_rate": 4.9460878604022365e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 2368000,
      "step": 2650
    },
    {
      "epoch": 0.7008050679688531,
      "grad_norm": 0.37707364559173584,
      "learning_rate": 4.945884887613065e-05,
      "loss": 0.1023,
      "num_input_tokens_seen": 2372352,
      "step": 2655
    },
    {
      "epoch": 0.70212485152435,
      "grad_norm": 0.06486539542675018,
      "learning_rate": 4.9456815376377055e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 2376928,
      "step": 2660
    },
    {
      "epoch": 0.7034446350798469,
      "grad_norm": 0.13737806677818298,
      "learning_rate": 4.9454778105075195e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 2381568,
      "step": 2665
    },
    {
      "epoch": 0.7047644186353438,
      "grad_norm": 0.17064864933490753,
      "learning_rate": 4.945273706253924e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 2386144,
      "step": 2670
    },
    {
      "epoch": 0.7060842021908407,
      "grad_norm": 0.20768266916275024,
      "learning_rate": 4.9450692249083925e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 2390432,
      "step": 2675
    },
    {
      "epoch": 0.7074039857463376,
      "grad_norm": 0.2147536724805832,
      "learning_rate": 4.9448643665024605e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 2395008,
      "step": 2680
    },
    {
      "epoch": 0.7087237693018344,
      "grad_norm": 0.31932583451271057,
      "learning_rate": 4.944659131067719e-05,
      "loss": 0.0752,
      "num_input_tokens_seen": 2399424,
      "step": 2685
    },
    {
      "epoch": 0.7100435528573315,
      "grad_norm": 0.21582411229610443,
      "learning_rate": 4.944453518635818e-05,
      "loss": 0.1005,
      "num_input_tokens_seen": 2403840,
      "step": 2690
    },
    {
      "epoch": 0.7113633364128283,
      "grad_norm": 0.07626861333847046,
      "learning_rate": 4.944247529238465e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 2408000,
      "step": 2695
    },
    {
      "epoch": 0.7126831199683252,
      "grad_norm": 0.2858969569206238,
      "learning_rate": 4.944041162907427e-05,
      "loss": 0.0944,
      "num_input_tokens_seen": 2412320,
      "step": 2700
    },
    {
      "epoch": 0.7140029035238221,
      "grad_norm": 0.26525992155075073,
      "learning_rate": 4.943834419674529e-05,
      "loss": 0.0927,
      "num_input_tokens_seen": 2416864,
      "step": 2705
    },
    {
      "epoch": 0.715322687079319,
      "grad_norm": 0.31241464614868164,
      "learning_rate": 4.9436272995716506e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 2420992,
      "step": 2710
    },
    {
      "epoch": 0.7166424706348159,
      "grad_norm": 0.12966488301753998,
      "learning_rate": 4.943419802630735e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 2425408,
      "step": 2715
    },
    {
      "epoch": 0.7179622541903128,
      "grad_norm": 0.35114187002182007,
      "learning_rate": 4.94321192888378e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 2429888,
      "step": 2720
    },
    {
      "epoch": 0.7192820377458097,
      "grad_norm": 0.4297981858253479,
      "learning_rate": 4.943003678362842e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 2434144,
      "step": 2725
    },
    {
      "epoch": 0.7206018213013066,
      "grad_norm": 0.321590781211853,
      "learning_rate": 4.942795051100036e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 2438656,
      "step": 2730
    },
    {
      "epoch": 0.7219216048568035,
      "grad_norm": 0.1313585788011551,
      "learning_rate": 4.942586047127536e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 2443072,
      "step": 2735
    },
    {
      "epoch": 0.7232413884123003,
      "grad_norm": 0.33832108974456787,
      "learning_rate": 4.942376666477571e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 2447392,
      "step": 2740
    },
    {
      "epoch": 0.7245611719677972,
      "grad_norm": 0.3430827260017395,
      "learning_rate": 4.9421669091824304e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 2451744,
      "step": 2745
    },
    {
      "epoch": 0.7258809555232941,
      "grad_norm": 0.2370171844959259,
      "learning_rate": 4.9419567752744634e-05,
      "loss": 0.0927,
      "num_input_tokens_seen": 2456032,
      "step": 2750
    },
    {
      "epoch": 0.7272007390787911,
      "grad_norm": 0.3264555335044861,
      "learning_rate": 4.941746264786074e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 2460832,
      "step": 2755
    },
    {
      "epoch": 0.728520522634288,
      "grad_norm": 0.2719355821609497,
      "learning_rate": 4.9415353777497254e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 2465120,
      "step": 2760
    },
    {
      "epoch": 0.7298403061897849,
      "grad_norm": 0.3218131363391876,
      "learning_rate": 4.9413241141979394e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 2469760,
      "step": 2765
    },
    {
      "epoch": 0.7311600897452818,
      "grad_norm": 0.09677737951278687,
      "learning_rate": 4.9411124741632956e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 2474208,
      "step": 2770
    },
    {
      "epoch": 0.7324798733007787,
      "grad_norm": 0.22909219563007355,
      "learning_rate": 4.940900457678431e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 2478656,
      "step": 2775
    },
    {
      "epoch": 0.7337996568562756,
      "grad_norm": 0.18849419057369232,
      "learning_rate": 4.9406880647760425e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 2483232,
      "step": 2780
    },
    {
      "epoch": 0.7351194404117725,
      "grad_norm": 0.14013653993606567,
      "learning_rate": 4.9404752954888824e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 2487872,
      "step": 2785
    },
    {
      "epoch": 0.7364392239672694,
      "grad_norm": 0.11565928161144257,
      "learning_rate": 4.940262149849762e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 2492640,
      "step": 2790
    },
    {
      "epoch": 0.7377590075227662,
      "grad_norm": 0.1926334649324417,
      "learning_rate": 4.9400486278915526e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 2497120,
      "step": 2795
    },
    {
      "epoch": 0.7390787910782631,
      "grad_norm": 0.62828528881073,
      "learning_rate": 4.939834729647181e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 2501632,
      "step": 2800
    },
    {
      "epoch": 0.7390787910782631,
      "eval_loss": 0.08939369022846222,
      "eval_runtime": 64.7515,
      "eval_samples_per_second": 104.013,
      "eval_steps_per_second": 26.007,
      "num_input_tokens_seen": 2501632,
      "step": 2800
    },
    {
      "epoch": 0.74039857463376,
      "grad_norm": 0.3150688409805298,
      "learning_rate": 4.9396204551496326e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 2506080,
      "step": 2805
    },
    {
      "epoch": 0.7417183581892569,
      "grad_norm": 0.17197035253047943,
      "learning_rate": 4.939405804431952e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 2510272,
      "step": 2810
    },
    {
      "epoch": 0.7430381417447539,
      "grad_norm": 0.08846244215965271,
      "learning_rate": 4.9391907775272414e-05,
      "loss": 0.1341,
      "num_input_tokens_seen": 2514944,
      "step": 2815
    },
    {
      "epoch": 0.7443579253002508,
      "grad_norm": 0.3886677026748657,
      "learning_rate": 4.9389753744686604e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 2519424,
      "step": 2820
    },
    {
      "epoch": 0.7456777088557477,
      "grad_norm": 0.4647468626499176,
      "learning_rate": 4.938759595289426e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 2524000,
      "step": 2825
    },
    {
      "epoch": 0.7469974924112446,
      "grad_norm": 0.25056251883506775,
      "learning_rate": 4.938543440022815e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 2528448,
      "step": 2830
    },
    {
      "epoch": 0.7483172759667415,
      "grad_norm": 0.44536638259887695,
      "learning_rate": 4.938326908702161e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 2532672,
      "step": 2835
    },
    {
      "epoch": 0.7496370595222384,
      "grad_norm": 0.18656006455421448,
      "learning_rate": 4.9381100013608554e-05,
      "loss": 0.1024,
      "num_input_tokens_seen": 2536928,
      "step": 2840
    },
    {
      "epoch": 0.7509568430777352,
      "grad_norm": 0.32168400287628174,
      "learning_rate": 4.9378927180323485e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 2541280,
      "step": 2845
    },
    {
      "epoch": 0.7522766266332321,
      "grad_norm": 0.34983646869659424,
      "learning_rate": 4.937675058750148e-05,
      "loss": 0.0927,
      "num_input_tokens_seen": 2545856,
      "step": 2850
    },
    {
      "epoch": 0.753596410188729,
      "grad_norm": 0.3630096912384033,
      "learning_rate": 4.937457023547819e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 2550432,
      "step": 2855
    },
    {
      "epoch": 0.7549161937442259,
      "grad_norm": 0.1443314403295517,
      "learning_rate": 4.9372386124589876e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 2555008,
      "step": 2860
    },
    {
      "epoch": 0.7562359772997228,
      "grad_norm": 0.12975271046161652,
      "learning_rate": 4.937019825517333e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 2559424,
      "step": 2865
    },
    {
      "epoch": 0.7575557608552197,
      "grad_norm": 0.09849181026220322,
      "learning_rate": 4.9368006627565954e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 2563840,
      "step": 2870
    },
    {
      "epoch": 0.7588755444107167,
      "grad_norm": 0.2932078242301941,
      "learning_rate": 4.936581124210573e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 2568416,
      "step": 2875
    },
    {
      "epoch": 0.7601953279662136,
      "grad_norm": 0.13444513082504272,
      "learning_rate": 4.9363612099131216e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 2572768,
      "step": 2880
    },
    {
      "epoch": 0.7615151115217105,
      "grad_norm": 0.5697705149650574,
      "learning_rate": 4.936140919898155e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 2577120,
      "step": 2885
    },
    {
      "epoch": 0.7628348950772074,
      "grad_norm": 0.2412351816892624,
      "learning_rate": 4.9359202541996426e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 2581824,
      "step": 2890
    },
    {
      "epoch": 0.7641546786327043,
      "grad_norm": 0.2971256971359253,
      "learning_rate": 4.935699212851616e-05,
      "loss": 0.0955,
      "num_input_tokens_seen": 2586464,
      "step": 2895
    },
    {
      "epoch": 0.7654744621882011,
      "grad_norm": 0.3475034534931183,
      "learning_rate": 4.935477795888162e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 2591456,
      "step": 2900
    },
    {
      "epoch": 0.766794245743698,
      "grad_norm": 0.06647197157144547,
      "learning_rate": 4.935256003343426e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 2595840,
      "step": 2905
    },
    {
      "epoch": 0.7681140292991949,
      "grad_norm": 0.5598300695419312,
      "learning_rate": 4.93503383525161e-05,
      "loss": 0.0917,
      "num_input_tokens_seen": 2600096,
      "step": 2910
    },
    {
      "epoch": 0.7694338128546918,
      "grad_norm": 0.2598221004009247,
      "learning_rate": 4.934811291646977e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 2604608,
      "step": 2915
    },
    {
      "epoch": 0.7707535964101887,
      "grad_norm": 0.29899170994758606,
      "learning_rate": 4.934588372563845e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 2609056,
      "step": 2920
    },
    {
      "epoch": 0.7720733799656856,
      "grad_norm": 0.4705277383327484,
      "learning_rate": 4.93436507803659e-05,
      "loss": 0.1212,
      "num_input_tokens_seen": 2613440,
      "step": 2925
    },
    {
      "epoch": 0.7733931635211825,
      "grad_norm": 0.43246787786483765,
      "learning_rate": 4.934141408099649e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 2617920,
      "step": 2930
    },
    {
      "epoch": 0.7747129470766794,
      "grad_norm": 0.10244724154472351,
      "learning_rate": 4.9339173627875135e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 2622560,
      "step": 2935
    },
    {
      "epoch": 0.7760327306321764,
      "grad_norm": 0.2124890238046646,
      "learning_rate": 4.9336929421347335e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 2627200,
      "step": 2940
    },
    {
      "epoch": 0.7773525141876733,
      "grad_norm": 0.31476423144340515,
      "learning_rate": 4.933468146175918e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 2631904,
      "step": 2945
    },
    {
      "epoch": 0.7786722977431701,
      "grad_norm": 0.21239709854125977,
      "learning_rate": 4.933242974945734e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 2636608,
      "step": 2950
    },
    {
      "epoch": 0.779992081298667,
      "grad_norm": 0.3374054431915283,
      "learning_rate": 4.933017428478906e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 2641184,
      "step": 2955
    },
    {
      "epoch": 0.7813118648541639,
      "grad_norm": 0.11836996674537659,
      "learning_rate": 4.932791506810214e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 2645792,
      "step": 2960
    },
    {
      "epoch": 0.7826316484096608,
      "grad_norm": 0.29594671726226807,
      "learning_rate": 4.932565209974499e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 2650080,
      "step": 2965
    },
    {
      "epoch": 0.7839514319651577,
      "grad_norm": 0.10724084079265594,
      "learning_rate": 4.93233853800666e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 2654560,
      "step": 2970
    },
    {
      "epoch": 0.7852712155206546,
      "grad_norm": 0.1988878697156906,
      "learning_rate": 4.932111490941651e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 2659264,
      "step": 2975
    },
    {
      "epoch": 0.7865909990761515,
      "grad_norm": 0.4171641767024994,
      "learning_rate": 4.9318840688144876e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 2663744,
      "step": 2980
    },
    {
      "epoch": 0.7879107826316484,
      "grad_norm": 0.5788004994392395,
      "learning_rate": 4.9316562716602387e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 2668128,
      "step": 2985
    },
    {
      "epoch": 0.7892305661871453,
      "grad_norm": 0.30139848589897156,
      "learning_rate": 4.9314280995140346e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 2672544,
      "step": 2990
    },
    {
      "epoch": 0.7905503497426422,
      "grad_norm": 0.19543936848640442,
      "learning_rate": 4.931199552411063e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 2677120,
      "step": 2995
    },
    {
      "epoch": 0.7918701332981392,
      "grad_norm": 0.575272262096405,
      "learning_rate": 4.930970630386568e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 2681600,
      "step": 3000
    },
    {
      "epoch": 0.7918701332981392,
      "eval_loss": 0.08825664222240448,
      "eval_runtime": 64.717,
      "eval_samples_per_second": 104.068,
      "eval_steps_per_second": 26.021,
      "num_input_tokens_seen": 2681600,
      "step": 3000
    },
    {
      "epoch": 0.793189916853636,
      "grad_norm": 0.3676469922065735,
      "learning_rate": 4.9307413334758524e-05,
      "loss": 0.1311,
      "num_input_tokens_seen": 2685856,
      "step": 3005
    },
    {
      "epoch": 0.7945097004091329,
      "grad_norm": 0.5000125765800476,
      "learning_rate": 4.930511661714276e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 2690080,
      "step": 3010
    },
    {
      "epoch": 0.7958294839646298,
      "grad_norm": 0.32019558548927307,
      "learning_rate": 4.9302816151372576e-05,
      "loss": 0.1112,
      "num_input_tokens_seen": 2694208,
      "step": 3015
    },
    {
      "epoch": 0.7971492675201267,
      "grad_norm": 0.3597044050693512,
      "learning_rate": 4.930051193780274e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 2698880,
      "step": 3020
    },
    {
      "epoch": 0.7984690510756236,
      "grad_norm": 0.38136425614356995,
      "learning_rate": 4.929820397678858e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 2703424,
      "step": 3025
    },
    {
      "epoch": 0.7997888346311205,
      "grad_norm": 0.24444857239723206,
      "learning_rate": 4.9295892268686015e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 2708032,
      "step": 3030
    },
    {
      "epoch": 0.8011086181866174,
      "grad_norm": 0.07816655188798904,
      "learning_rate": 4.9293576813851536e-05,
      "loss": 0.1342,
      "num_input_tokens_seen": 2712320,
      "step": 3035
    },
    {
      "epoch": 0.8024284017421143,
      "grad_norm": 0.21107934415340424,
      "learning_rate": 4.929125761264223e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 2716896,
      "step": 3040
    },
    {
      "epoch": 0.8037481852976112,
      "grad_norm": 0.13296520709991455,
      "learning_rate": 4.928893466541573e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 2721568,
      "step": 3045
    },
    {
      "epoch": 0.805067968853108,
      "grad_norm": 0.1991877406835556,
      "learning_rate": 4.928660797253027e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 2726144,
      "step": 3050
    },
    {
      "epoch": 0.8063877524086049,
      "grad_norm": 0.32446032762527466,
      "learning_rate": 4.928427753434467e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 2730496,
      "step": 3055
    },
    {
      "epoch": 0.8077075359641019,
      "grad_norm": 0.5705958604812622,
      "learning_rate": 4.9281943351218286e-05,
      "loss": 0.1265,
      "num_input_tokens_seen": 2735104,
      "step": 3060
    },
    {
      "epoch": 0.8090273195195988,
      "grad_norm": 0.3894815444946289,
      "learning_rate": 4.9279605423511095e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 2739776,
      "step": 3065
    },
    {
      "epoch": 0.8103471030750957,
      "grad_norm": 0.202248215675354,
      "learning_rate": 4.927726375158363e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 2743968,
      "step": 3070
    },
    {
      "epoch": 0.8116668866305926,
      "grad_norm": 0.2518572509288788,
      "learning_rate": 4.9274918335797004e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 2748512,
      "step": 3075
    },
    {
      "epoch": 0.8129866701860895,
      "grad_norm": 0.5856395363807678,
      "learning_rate": 4.927256917651292e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 2752800,
      "step": 3080
    },
    {
      "epoch": 0.8143064537415864,
      "grad_norm": 0.42110559344291687,
      "learning_rate": 4.927021627409364e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 2757344,
      "step": 3085
    },
    {
      "epoch": 0.8156262372970833,
      "grad_norm": 0.16488933563232422,
      "learning_rate": 4.9267859628902005e-05,
      "loss": 0.1161,
      "num_input_tokens_seen": 2761760,
      "step": 3090
    },
    {
      "epoch": 0.8169460208525802,
      "grad_norm": 0.24452261626720428,
      "learning_rate": 4.9265499241301454e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 2766016,
      "step": 3095
    },
    {
      "epoch": 0.8182658044080771,
      "grad_norm": 0.42030489444732666,
      "learning_rate": 4.926313511165598e-05,
      "loss": 0.0999,
      "num_input_tokens_seen": 2770592,
      "step": 3100
    },
    {
      "epoch": 0.819585587963574,
      "grad_norm": 0.3904484510421753,
      "learning_rate": 4.926076724033016e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 2774720,
      "step": 3105
    },
    {
      "epoch": 0.8209053715190708,
      "grad_norm": 0.40583816170692444,
      "learning_rate": 4.9258395627689146e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 2779104,
      "step": 3110
    },
    {
      "epoch": 0.8222251550745677,
      "grad_norm": 0.08078955113887787,
      "learning_rate": 4.925602027409868e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 2783488,
      "step": 3115
    },
    {
      "epoch": 0.8235449386300646,
      "grad_norm": 0.4615092873573303,
      "learning_rate": 4.925364117992507e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 2787712,
      "step": 3120
    },
    {
      "epoch": 0.8248647221855616,
      "grad_norm": 0.3310774266719818,
      "learning_rate": 4.92512583455352e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 2791776,
      "step": 3125
    },
    {
      "epoch": 0.8261845057410585,
      "grad_norm": 0.12238745391368866,
      "learning_rate": 4.9248871771296536e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 2796448,
      "step": 3130
    },
    {
      "epoch": 0.8275042892965554,
      "grad_norm": 0.20484626293182373,
      "learning_rate": 4.924648145757711e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 2800992,
      "step": 3135
    },
    {
      "epoch": 0.8288240728520523,
      "grad_norm": 0.7171216607093811,
      "learning_rate": 4.924408740474554e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 2805184,
      "step": 3140
    },
    {
      "epoch": 0.8301438564075492,
      "grad_norm": 0.09341349452733994,
      "learning_rate": 4.924168961317103e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 2809824,
      "step": 3145
    },
    {
      "epoch": 0.8314636399630461,
      "grad_norm": 0.7684889435768127,
      "learning_rate": 4.9239288083223334e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 2814432,
      "step": 3150
    },
    {
      "epoch": 0.832783423518543,
      "grad_norm": 0.15415146946907043,
      "learning_rate": 4.9236882815272803e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 2818688,
      "step": 3155
    },
    {
      "epoch": 0.8341032070740398,
      "grad_norm": 0.10995927453041077,
      "learning_rate": 4.9234473809690365e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 2823200,
      "step": 3160
    },
    {
      "epoch": 0.8354229906295367,
      "grad_norm": 0.5262004137039185,
      "learning_rate": 4.923206106684752e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 2827616,
      "step": 3165
    },
    {
      "epoch": 0.8367427741850336,
      "grad_norm": 0.22380705177783966,
      "learning_rate": 4.922964458711634e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 2832320,
      "step": 3170
    },
    {
      "epoch": 0.8380625577405305,
      "grad_norm": 0.5028089284896851,
      "learning_rate": 4.9227224370869474e-05,
      "loss": 0.1208,
      "num_input_tokens_seen": 2836736,
      "step": 3175
    },
    {
      "epoch": 0.8393823412960274,
      "grad_norm": 0.20682218670845032,
      "learning_rate": 4.9224800418480155e-05,
      "loss": 0.096,
      "num_input_tokens_seen": 2841376,
      "step": 3180
    },
    {
      "epoch": 0.8407021248515244,
      "grad_norm": 0.3023107647895813,
      "learning_rate": 4.9222372730322176e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 2845920,
      "step": 3185
    },
    {
      "epoch": 0.8420219084070213,
      "grad_norm": 0.0918809026479721,
      "learning_rate": 4.921994130676993e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 2850144,
      "step": 3190
    },
    {
      "epoch": 0.8433416919625182,
      "grad_norm": 0.1666431427001953,
      "learning_rate": 4.9217506148198366e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 2854880,
      "step": 3195
    },
    {
      "epoch": 0.8446614755180151,
      "grad_norm": 0.41103827953338623,
      "learning_rate": 4.921506725498302e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 2859456,
      "step": 3200
    },
    {
      "epoch": 0.8446614755180151,
      "eval_loss": 0.08669708669185638,
      "eval_runtime": 64.7141,
      "eval_samples_per_second": 104.073,
      "eval_steps_per_second": 26.022,
      "num_input_tokens_seen": 2859456,
      "step": 3200
    },
    {
      "epoch": 0.845981259073512,
      "grad_norm": 0.8006001710891724,
      "learning_rate": 4.9212624627499994e-05,
      "loss": 0.1255,
      "num_input_tokens_seen": 2864320,
      "step": 3205
    },
    {
      "epoch": 0.8473010426290088,
      "grad_norm": 0.29559987783432007,
      "learning_rate": 4.921017826612597e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 2868928,
      "step": 3210
    },
    {
      "epoch": 0.8486208261845057,
      "grad_norm": 0.14707621932029724,
      "learning_rate": 4.9207728171238223e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 2873536,
      "step": 3215
    },
    {
      "epoch": 0.8499406097400026,
      "grad_norm": 0.357501357793808,
      "learning_rate": 4.920527434321458e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 2878112,
      "step": 3220
    },
    {
      "epoch": 0.8512603932954995,
      "grad_norm": 0.3020641803741455,
      "learning_rate": 4.920281678243345e-05,
      "loss": 0.1027,
      "num_input_tokens_seen": 2882240,
      "step": 3225
    },
    {
      "epoch": 0.8525801768509964,
      "grad_norm": 0.19769532978534698,
      "learning_rate": 4.920035548927381e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 2886720,
      "step": 3230
    },
    {
      "epoch": 0.8538999604064933,
      "grad_norm": 0.29588621854782104,
      "learning_rate": 4.919789046411525e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 2891072,
      "step": 3235
    },
    {
      "epoch": 0.8552197439619902,
      "grad_norm": 0.31500303745269775,
      "learning_rate": 4.919542170733787e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 2895584,
      "step": 3240
    },
    {
      "epoch": 0.8565395275174872,
      "grad_norm": 0.25430646538734436,
      "learning_rate": 4.919294921932242e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 2899808,
      "step": 3245
    },
    {
      "epoch": 0.8578593110729841,
      "grad_norm": 0.29362353682518005,
      "learning_rate": 4.919047300045016e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 2904288,
      "step": 3250
    },
    {
      "epoch": 0.859179094628481,
      "grad_norm": 0.16907300055027008,
      "learning_rate": 4.918799305110299e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 2908608,
      "step": 3255
    },
    {
      "epoch": 0.8604988781839779,
      "grad_norm": 0.07650022208690643,
      "learning_rate": 4.918550937166331e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 2912768,
      "step": 3260
    },
    {
      "epoch": 0.8618186617394747,
      "grad_norm": 0.1158524677157402,
      "learning_rate": 4.918302196251415e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 2917280,
      "step": 3265
    },
    {
      "epoch": 0.8631384452949716,
      "grad_norm": 0.2968883216381073,
      "learning_rate": 4.91805308240391e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 2922112,
      "step": 3270
    },
    {
      "epoch": 0.8644582288504685,
      "grad_norm": 0.40005528926849365,
      "learning_rate": 4.9178035956622326e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 2926496,
      "step": 3275
    },
    {
      "epoch": 0.8657780124059654,
      "grad_norm": 0.4341690242290497,
      "learning_rate": 4.917553736064857e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 2931040,
      "step": 3280
    },
    {
      "epoch": 0.8670977959614623,
      "grad_norm": 0.26169392466545105,
      "learning_rate": 4.917303503650314e-05,
      "loss": 0.0982,
      "num_input_tokens_seen": 2935648,
      "step": 3285
    },
    {
      "epoch": 0.8684175795169592,
      "grad_norm": 0.31261658668518066,
      "learning_rate": 4.917052898457194e-05,
      "loss": 0.1074,
      "num_input_tokens_seen": 2939968,
      "step": 3290
    },
    {
      "epoch": 0.8697373630724561,
      "grad_norm": 0.20674723386764526,
      "learning_rate": 4.916801920524141e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 2944288,
      "step": 3295
    },
    {
      "epoch": 0.871057146627953,
      "grad_norm": 0.2808683216571808,
      "learning_rate": 4.916550569889862e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 2948960,
      "step": 3300
    },
    {
      "epoch": 0.8723769301834499,
      "grad_norm": 0.6423003077507019,
      "learning_rate": 4.916298846593116e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 2953472,
      "step": 3305
    },
    {
      "epoch": 0.8736967137389469,
      "grad_norm": 0.15420837700366974,
      "learning_rate": 4.916046750672722e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 2957920,
      "step": 3310
    },
    {
      "epoch": 0.8750164972944438,
      "grad_norm": 0.42586788535118103,
      "learning_rate": 4.915794282167559e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 2962240,
      "step": 3315
    },
    {
      "epoch": 0.8763362808499406,
      "grad_norm": 0.2385115772485733,
      "learning_rate": 4.915541441116558e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 2966912,
      "step": 3320
    },
    {
      "epoch": 0.8776560644054375,
      "grad_norm": 0.29325252771377563,
      "learning_rate": 4.915288227558711e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 2971712,
      "step": 3325
    },
    {
      "epoch": 0.8789758479609344,
      "grad_norm": 0.06891842186450958,
      "learning_rate": 4.915034641533066e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 2976416,
      "step": 3330
    },
    {
      "epoch": 0.8802956315164313,
      "grad_norm": 0.5288883447647095,
      "learning_rate": 4.914780683078731e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 2981120,
      "step": 3335
    },
    {
      "epoch": 0.8816154150719282,
      "grad_norm": 0.24800622463226318,
      "learning_rate": 4.9145263522348695e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 2985472,
      "step": 3340
    },
    {
      "epoch": 0.8829351986274251,
      "grad_norm": 0.4597092866897583,
      "learning_rate": 4.9142716490407e-05,
      "loss": 0.115,
      "num_input_tokens_seen": 2989952,
      "step": 3345
    },
    {
      "epoch": 0.884254982182922,
      "grad_norm": 0.39968809485435486,
      "learning_rate": 4.914016573535504e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 2994624,
      "step": 3350
    },
    {
      "epoch": 0.8855747657384189,
      "grad_norm": 0.17291751503944397,
      "learning_rate": 4.9137611257586154e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 2999232,
      "step": 3355
    },
    {
      "epoch": 0.8868945492939158,
      "grad_norm": 0.35235458612442017,
      "learning_rate": 4.9135053057494274e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 3003648,
      "step": 3360
    },
    {
      "epoch": 0.8882143328494126,
      "grad_norm": 0.12908470630645752,
      "learning_rate": 4.913249113547392e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 3008448,
      "step": 3365
    },
    {
      "epoch": 0.8895341164049096,
      "grad_norm": 0.17904619872570038,
      "learning_rate": 4.912992549192016e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 3012960,
      "step": 3370
    },
    {
      "epoch": 0.8908538999604065,
      "grad_norm": 0.06685513257980347,
      "learning_rate": 4.9127356127228665e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 3017408,
      "step": 3375
    },
    {
      "epoch": 0.8921736835159034,
      "grad_norm": 0.19979850947856903,
      "learning_rate": 4.912478304179564e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 3021728,
      "step": 3380
    },
    {
      "epoch": 0.8934934670714003,
      "grad_norm": 0.378202348947525,
      "learning_rate": 4.9122206236017896e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 3026208,
      "step": 3385
    },
    {
      "epoch": 0.8948132506268972,
      "grad_norm": 0.09463414549827576,
      "learning_rate": 4.911962571029282e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 3030688,
      "step": 3390
    },
    {
      "epoch": 0.8961330341823941,
      "grad_norm": 0.45749804377555847,
      "learning_rate": 4.9117041465018353e-05,
      "loss": 0.127,
      "num_input_tokens_seen": 3035040,
      "step": 3395
    },
    {
      "epoch": 0.897452817737891,
      "grad_norm": 0.072332464158535,
      "learning_rate": 4.911445350059302e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 3039712,
      "step": 3400
    },
    {
      "epoch": 0.897452817737891,
      "eval_loss": 0.08539655804634094,
      "eval_runtime": 64.8209,
      "eval_samples_per_second": 103.902,
      "eval_steps_per_second": 25.979,
      "num_input_tokens_seen": 3039712,
      "step": 3400
    },
    {
      "epoch": 0.8987726012933879,
      "grad_norm": 0.5369530320167542,
      "learning_rate": 4.9111861817415905e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 3043904,
      "step": 3405
    },
    {
      "epoch": 0.9000923848488848,
      "grad_norm": 0.24395689368247986,
      "learning_rate": 4.91092664158867e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 3048384,
      "step": 3410
    },
    {
      "epoch": 0.9014121684043817,
      "grad_norm": 0.11107509583234787,
      "learning_rate": 4.910666729640563e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 3052736,
      "step": 3415
    },
    {
      "epoch": 0.9027319519598785,
      "grad_norm": 0.15976864099502563,
      "learning_rate": 4.910406445937353e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 3057376,
      "step": 3420
    },
    {
      "epoch": 0.9040517355153754,
      "grad_norm": 0.6188194751739502,
      "learning_rate": 4.9101457905191774e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 3062208,
      "step": 3425
    },
    {
      "epoch": 0.9053715190708723,
      "grad_norm": 0.2900412380695343,
      "learning_rate": 4.909884763426233e-05,
      "loss": 0.1013,
      "num_input_tokens_seen": 3066304,
      "step": 3430
    },
    {
      "epoch": 0.9066913026263693,
      "grad_norm": 0.12687575817108154,
      "learning_rate": 4.9096233646987736e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 3070720,
      "step": 3435
    },
    {
      "epoch": 0.9080110861818662,
      "grad_norm": 0.15802253782749176,
      "learning_rate": 4.9093615943771104e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 3075168,
      "step": 3440
    },
    {
      "epoch": 0.9093308697373631,
      "grad_norm": 0.42126262187957764,
      "learning_rate": 4.909099452501611e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 3079488,
      "step": 3445
    },
    {
      "epoch": 0.91065065329286,
      "grad_norm": 0.41974180936813354,
      "learning_rate": 4.908836939112702e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 3084224,
      "step": 3450
    },
    {
      "epoch": 0.9119704368483569,
      "grad_norm": 0.26655495166778564,
      "learning_rate": 4.908574054250865e-05,
      "loss": 0.1134,
      "num_input_tokens_seen": 3088800,
      "step": 3455
    },
    {
      "epoch": 0.9132902204038538,
      "grad_norm": 0.36356255412101746,
      "learning_rate": 4.9083107979566414e-05,
      "loss": 0.1157,
      "num_input_tokens_seen": 3093344,
      "step": 3460
    },
    {
      "epoch": 0.9146100039593507,
      "grad_norm": 0.2589225769042969,
      "learning_rate": 4.908047170270628e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 3097600,
      "step": 3465
    },
    {
      "epoch": 0.9159297875148475,
      "grad_norm": 0.34415295720100403,
      "learning_rate": 4.9077831712334784e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 3101792,
      "step": 3470
    },
    {
      "epoch": 0.9172495710703444,
      "grad_norm": 0.5704537034034729,
      "learning_rate": 4.907518800885907e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 3106240,
      "step": 3475
    },
    {
      "epoch": 0.9185693546258413,
      "grad_norm": 0.37569931149482727,
      "learning_rate": 4.907254059268681e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 3110688,
      "step": 3480
    },
    {
      "epoch": 0.9198891381813382,
      "grad_norm": 0.14082498848438263,
      "learning_rate": 4.906988946422628e-05,
      "loss": 0.1023,
      "num_input_tokens_seen": 3115520,
      "step": 3485
    },
    {
      "epoch": 0.9212089217368351,
      "grad_norm": 0.2145296186208725,
      "learning_rate": 4.9067234623886315e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 3120128,
      "step": 3490
    },
    {
      "epoch": 0.9225287052923321,
      "grad_norm": 0.2338973879814148,
      "learning_rate": 4.9064576072076316e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 3124448,
      "step": 3495
    },
    {
      "epoch": 0.923848488847829,
      "grad_norm": 0.6379575729370117,
      "learning_rate": 4.906191380920628e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 3128960,
      "step": 3500
    },
    {
      "epoch": 0.9251682724033259,
      "grad_norm": 0.18910695612430573,
      "learning_rate": 4.905924783568675e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 3133312,
      "step": 3505
    },
    {
      "epoch": 0.9264880559588228,
      "grad_norm": 0.33646896481513977,
      "learning_rate": 4.905657815192886e-05,
      "loss": 0.0937,
      "num_input_tokens_seen": 3137792,
      "step": 3510
    },
    {
      "epoch": 0.9278078395143197,
      "grad_norm": 0.22252024710178375,
      "learning_rate": 4.90539047583443e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 3142304,
      "step": 3515
    },
    {
      "epoch": 0.9291276230698166,
      "grad_norm": 0.1181967705488205,
      "learning_rate": 4.905122765534534e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 3146912,
      "step": 3520
    },
    {
      "epoch": 0.9304474066253134,
      "grad_norm": 0.4838852882385254,
      "learning_rate": 4.9048546843344846e-05,
      "loss": 0.0949,
      "num_input_tokens_seen": 3151424,
      "step": 3525
    },
    {
      "epoch": 0.9317671901808103,
      "grad_norm": 0.2745245397090912,
      "learning_rate": 4.9045862322756206e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 3155616,
      "step": 3530
    },
    {
      "epoch": 0.9330869737363072,
      "grad_norm": 0.10697958618402481,
      "learning_rate": 4.904317409399342e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 3159936,
      "step": 3535
    },
    {
      "epoch": 0.9344067572918041,
      "grad_norm": 0.09313882142305374,
      "learning_rate": 4.904048215747104e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 3164736,
      "step": 3540
    },
    {
      "epoch": 0.935726540847301,
      "grad_norm": 0.3413710594177246,
      "learning_rate": 4.90377865136042e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 3169120,
      "step": 3545
    },
    {
      "epoch": 0.9370463244027979,
      "grad_norm": 0.18707430362701416,
      "learning_rate": 4.90350871628086e-05,
      "loss": 0.1152,
      "num_input_tokens_seen": 3173984,
      "step": 3550
    },
    {
      "epoch": 0.9383661079582949,
      "grad_norm": 0.38687068223953247,
      "learning_rate": 4.903238410550052e-05,
      "loss": 0.1112,
      "num_input_tokens_seen": 3178400,
      "step": 3555
    },
    {
      "epoch": 0.9396858915137918,
      "grad_norm": 0.18848969042301178,
      "learning_rate": 4.90296773420968e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 3182656,
      "step": 3560
    },
    {
      "epoch": 0.9410056750692887,
      "grad_norm": 0.196629598736763,
      "learning_rate": 4.902696687301486e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 3187232,
      "step": 3565
    },
    {
      "epoch": 0.9423254586247856,
      "grad_norm": 0.1749841868877411,
      "learning_rate": 4.902425269867268e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 3191616,
      "step": 3570
    },
    {
      "epoch": 0.9436452421802825,
      "grad_norm": 0.10314325988292694,
      "learning_rate": 4.902153481948883e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 3196096,
      "step": 3575
    },
    {
      "epoch": 0.9449650257357793,
      "grad_norm": 0.11345010250806808,
      "learning_rate": 4.901881323588244e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 3200640,
      "step": 3580
    },
    {
      "epoch": 0.9462848092912762,
      "grad_norm": 0.3500674068927765,
      "learning_rate": 4.90160879482732e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 3205088,
      "step": 3585
    },
    {
      "epoch": 0.9476045928467731,
      "grad_norm": 0.2619893252849579,
      "learning_rate": 4.9013358957081405e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 3209536,
      "step": 3590
    },
    {
      "epoch": 0.94892437640227,
      "grad_norm": 0.30071645975112915,
      "learning_rate": 4.901062626272789e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 3213856,
      "step": 3595
    },
    {
      "epoch": 0.9502441599577669,
      "grad_norm": 0.0862569659948349,
      "learning_rate": 4.900788986563406e-05,
      "loss": 0.1093,
      "num_input_tokens_seen": 3218400,
      "step": 3600
    },
    {
      "epoch": 0.9502441599577669,
      "eval_loss": 0.08537984639406204,
      "eval_runtime": 64.7456,
      "eval_samples_per_second": 104.022,
      "eval_steps_per_second": 26.009,
      "num_input_tokens_seen": 3218400,
      "step": 3600
    },
    {
      "epoch": 0.9515639435132638,
      "grad_norm": 0.18426357209682465,
      "learning_rate": 4.9005149766221915e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 3222912,
      "step": 3605
    },
    {
      "epoch": 0.9528837270687607,
      "grad_norm": 0.12116490304470062,
      "learning_rate": 4.9002405964914e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 3227104,
      "step": 3610
    },
    {
      "epoch": 0.9542035106242576,
      "grad_norm": 0.3477870523929596,
      "learning_rate": 4.899965846213346e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 3231552,
      "step": 3615
    },
    {
      "epoch": 0.9555232941797546,
      "grad_norm": 0.4415113627910614,
      "learning_rate": 4.899690725830399e-05,
      "loss": 0.0896,
      "num_input_tokens_seen": 3236448,
      "step": 3620
    },
    {
      "epoch": 0.9568430777352515,
      "grad_norm": 0.35103827714920044,
      "learning_rate": 4.899415235384985e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 3240576,
      "step": 3625
    },
    {
      "epoch": 0.9581628612907483,
      "grad_norm": 0.15636710822582245,
      "learning_rate": 4.899139374919589e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 3244896,
      "step": 3630
    },
    {
      "epoch": 0.9594826448462452,
      "grad_norm": 0.2798616886138916,
      "learning_rate": 4.898863144476752e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 3249376,
      "step": 3635
    },
    {
      "epoch": 0.9608024284017421,
      "grad_norm": 0.5134257078170776,
      "learning_rate": 4.898586544099072e-05,
      "loss": 0.1171,
      "num_input_tokens_seen": 3254080,
      "step": 3640
    },
    {
      "epoch": 0.962122211957239,
      "grad_norm": 0.36010757088661194,
      "learning_rate": 4.898309573829204e-05,
      "loss": 0.1185,
      "num_input_tokens_seen": 3258592,
      "step": 3645
    },
    {
      "epoch": 0.9634419955127359,
      "grad_norm": 0.38726165890693665,
      "learning_rate": 4.898032233709862e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 3263360,
      "step": 3650
    },
    {
      "epoch": 0.9647617790682328,
      "grad_norm": 0.12077130377292633,
      "learning_rate": 4.8977545237838123e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 3267872,
      "step": 3655
    },
    {
      "epoch": 0.9660815626237297,
      "grad_norm": 0.16140489280223846,
      "learning_rate": 4.8974764440938836e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 3272480,
      "step": 3660
    },
    {
      "epoch": 0.9674013461792266,
      "grad_norm": 0.20887038111686707,
      "learning_rate": 4.897197994682959e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 3276960,
      "step": 3665
    },
    {
      "epoch": 0.9687211297347235,
      "grad_norm": 0.12402015179395676,
      "learning_rate": 4.8969191755939786e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 3281216,
      "step": 3670
    },
    {
      "epoch": 0.9700409132902204,
      "grad_norm": 0.18823190033435822,
      "learning_rate": 4.8966399868699396e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 3285856,
      "step": 3675
    },
    {
      "epoch": 0.9713606968457174,
      "grad_norm": 0.23505936563014984,
      "learning_rate": 4.8963604285538965e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 3290208,
      "step": 3680
    },
    {
      "epoch": 0.9726804804012142,
      "grad_norm": 0.19444938004016876,
      "learning_rate": 4.8960805006889604e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 3294464,
      "step": 3685
    },
    {
      "epoch": 0.9740002639567111,
      "grad_norm": 0.04791402816772461,
      "learning_rate": 4.8958002033183004e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 3298784,
      "step": 3690
    },
    {
      "epoch": 0.975320047512208,
      "grad_norm": 0.11152095347642899,
      "learning_rate": 4.8955195364851414e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 3303424,
      "step": 3695
    },
    {
      "epoch": 0.9766398310677049,
      "grad_norm": 0.14665816724300385,
      "learning_rate": 4.895238500232766e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 3308032,
      "step": 3700
    },
    {
      "epoch": 0.9779596146232018,
      "grad_norm": 0.09416782855987549,
      "learning_rate": 4.8949570946045143e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 3312640,
      "step": 3705
    },
    {
      "epoch": 0.9792793981786987,
      "grad_norm": 0.1455061435699463,
      "learning_rate": 4.89467531964378e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 3317024,
      "step": 3710
    },
    {
      "epoch": 0.9805991817341956,
      "grad_norm": 0.08683057129383087,
      "learning_rate": 4.894393175394019e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 3321568,
      "step": 3715
    },
    {
      "epoch": 0.9819189652896925,
      "grad_norm": 0.0437285378575325,
      "learning_rate": 4.8941106618987406e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 3325856,
      "step": 3720
    },
    {
      "epoch": 0.9832387488451894,
      "grad_norm": 0.47485217452049255,
      "learning_rate": 4.893827779201512e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 3330240,
      "step": 3725
    },
    {
      "epoch": 0.9845585324006862,
      "grad_norm": 0.24226413667201996,
      "learning_rate": 4.893544527345957e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 3334560,
      "step": 3730
    },
    {
      "epoch": 0.9858783159561831,
      "grad_norm": 0.04881235957145691,
      "learning_rate": 4.8932609063757563e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 3338656,
      "step": 3735
    },
    {
      "epoch": 0.9871980995116801,
      "grad_norm": 0.17872962355613708,
      "learning_rate": 4.8929769163346484e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 3343104,
      "step": 3740
    },
    {
      "epoch": 0.988517883067177,
      "grad_norm": 0.2904002070426941,
      "learning_rate": 4.892692557266429e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 3347456,
      "step": 3745
    },
    {
      "epoch": 0.9898376666226739,
      "grad_norm": 0.11871351301670074,
      "learning_rate": 4.8924078292149464e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 3351712,
      "step": 3750
    },
    {
      "epoch": 0.9911574501781708,
      "grad_norm": 0.5551296472549438,
      "learning_rate": 4.892122732224114e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 3356192,
      "step": 3755
    },
    {
      "epoch": 0.9924772337336677,
      "grad_norm": 0.1894797831773758,
      "learning_rate": 4.8918372663378944e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 3360704,
      "step": 3760
    },
    {
      "epoch": 0.9937970172891646,
      "grad_norm": 0.3918653726577759,
      "learning_rate": 4.89155143160031e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 3365216,
      "step": 3765
    },
    {
      "epoch": 0.9951168008446615,
      "grad_norm": 0.5060731768608093,
      "learning_rate": 4.891265228055441e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 3369696,
      "step": 3770
    },
    {
      "epoch": 0.9964365844001584,
      "grad_norm": 0.1691993623971939,
      "learning_rate": 4.890978655747424e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 3373952,
      "step": 3775
    },
    {
      "epoch": 0.9977563679556553,
      "grad_norm": 0.2829606831073761,
      "learning_rate": 4.89069171472045e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 3378432,
      "step": 3780
    },
    {
      "epoch": 0.9990761515111521,
      "grad_norm": 0.07150144129991531,
      "learning_rate": 4.890404405018772e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 3382528,
      "step": 3785
    },
    {
      "epoch": 1.0002639567110994,
      "grad_norm": 0.6697476506233215,
      "learning_rate": 4.8901167266866934e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 3386576,
      "step": 3790
    },
    {
      "epoch": 1.0015837402665964,
      "grad_norm": 0.48284265398979187,
      "learning_rate": 4.88982867976858e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 3391376,
      "step": 3795
    },
    {
      "epoch": 1.0029035238220931,
      "grad_norm": 0.13684047758579254,
      "learning_rate": 4.889540264308852e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 3395632,
      "step": 3800
    },
    {
      "epoch": 1.0029035238220931,
      "eval_loss": 0.0841970220208168,
      "eval_runtime": 64.6968,
      "eval_samples_per_second": 104.101,
      "eval_steps_per_second": 26.029,
      "num_input_tokens_seen": 3395632,
      "step": 3800
    },
    {
      "epoch": 1.0042233073775901,
      "grad_norm": 0.40840673446655273,
      "learning_rate": 4.889251480351986e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 3400144,
      "step": 3805
    },
    {
      "epoch": 1.005543090933087,
      "grad_norm": 0.1282190978527069,
      "learning_rate": 4.888962327942517e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 3404752,
      "step": 3810
    },
    {
      "epoch": 1.006862874488584,
      "grad_norm": 0.18420645594596863,
      "learning_rate": 4.8886728071250356e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 3409456,
      "step": 3815
    },
    {
      "epoch": 1.0081826580440807,
      "grad_norm": 0.176153764128685,
      "learning_rate": 4.8883829179441884e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 3413840,
      "step": 3820
    },
    {
      "epoch": 1.0095024415995777,
      "grad_norm": 0.4257528781890869,
      "learning_rate": 4.888092660444682e-05,
      "loss": 0.0789,
      "num_input_tokens_seen": 3418160,
      "step": 3825
    },
    {
      "epoch": 1.0108222251550745,
      "grad_norm": 0.13079734146595,
      "learning_rate": 4.887802034671276e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 3422672,
      "step": 3830
    },
    {
      "epoch": 1.0121420087105715,
      "grad_norm": 0.40562328696250916,
      "learning_rate": 4.88751104066879e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 3427216,
      "step": 3835
    },
    {
      "epoch": 1.0134617922660685,
      "grad_norm": 0.2332824319601059,
      "learning_rate": 4.887219678482098e-05,
      "loss": 0.1141,
      "num_input_tokens_seen": 3431952,
      "step": 3840
    },
    {
      "epoch": 1.0147815758215653,
      "grad_norm": 0.5409038066864014,
      "learning_rate": 4.8869279481561316e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 3436368,
      "step": 3845
    },
    {
      "epoch": 1.0161013593770623,
      "grad_norm": 0.28266409039497375,
      "learning_rate": 4.88663584973588e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 3440624,
      "step": 3850
    },
    {
      "epoch": 1.017421142932559,
      "grad_norm": 0.1674271821975708,
      "learning_rate": 4.8863433832663874e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 3445296,
      "step": 3855
    },
    {
      "epoch": 1.018740926488056,
      "grad_norm": 0.3472689986228943,
      "learning_rate": 4.886050548792757e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 3449552,
      "step": 3860
    },
    {
      "epoch": 1.0200607100435528,
      "grad_norm": 0.5386594533920288,
      "learning_rate": 4.8857573463601465e-05,
      "loss": 0.1133,
      "num_input_tokens_seen": 3454384,
      "step": 3865
    },
    {
      "epoch": 1.0213804935990498,
      "grad_norm": 0.4810366928577423,
      "learning_rate": 4.885463776013772e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 3458896,
      "step": 3870
    },
    {
      "epoch": 1.0227002771545466,
      "grad_norm": 0.3793335556983948,
      "learning_rate": 4.8851698377989056e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 3463184,
      "step": 3875
    },
    {
      "epoch": 1.0240200607100436,
      "grad_norm": 0.3431379497051239,
      "learning_rate": 4.884875531760876e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 3467824,
      "step": 3880
    },
    {
      "epoch": 1.0253398442655404,
      "grad_norm": 0.24647897481918335,
      "learning_rate": 4.88458085794507e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 3471984,
      "step": 3885
    },
    {
      "epoch": 1.0266596278210374,
      "grad_norm": 0.20652002096176147,
      "learning_rate": 4.884285816396929e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 3476560,
      "step": 3890
    },
    {
      "epoch": 1.0279794113765341,
      "grad_norm": 0.2015364170074463,
      "learning_rate": 4.8839904071619526e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 3480912,
      "step": 3895
    },
    {
      "epoch": 1.0292991949320311,
      "grad_norm": 0.26247185468673706,
      "learning_rate": 4.8836946302856955e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 3485360,
      "step": 3900
    },
    {
      "epoch": 1.0306189784875281,
      "grad_norm": 0.33072471618652344,
      "learning_rate": 4.8833984858137715e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 3490000,
      "step": 3905
    },
    {
      "epoch": 1.031938762043025,
      "grad_norm": 0.10270272940397263,
      "learning_rate": 4.8831019737918494e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 3494480,
      "step": 3910
    },
    {
      "epoch": 1.033258545598522,
      "grad_norm": 0.23943404853343964,
      "learning_rate": 4.882805094265655e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 3499152,
      "step": 3915
    },
    {
      "epoch": 1.0345783291540187,
      "grad_norm": 0.42587146162986755,
      "learning_rate": 4.8825078472809706e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 3503568,
      "step": 3920
    },
    {
      "epoch": 1.0358981127095157,
      "grad_norm": 0.3739672899246216,
      "learning_rate": 4.882210232883635e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 3508016,
      "step": 3925
    },
    {
      "epoch": 1.0372178962650125,
      "grad_norm": 0.43589022755622864,
      "learning_rate": 4.881912251119546e-05,
      "loss": 0.0896,
      "num_input_tokens_seen": 3512208,
      "step": 3930
    },
    {
      "epoch": 1.0385376798205095,
      "grad_norm": 0.20134888589382172,
      "learning_rate": 4.881613902034654e-05,
      "loss": 0.0452,
      "num_input_tokens_seen": 3516816,
      "step": 3935
    },
    {
      "epoch": 1.0398574633760063,
      "grad_norm": 0.24141134321689606,
      "learning_rate": 4.88131518567497e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 3521264,
      "step": 3940
    },
    {
      "epoch": 1.0411772469315033,
      "grad_norm": 0.16694284975528717,
      "learning_rate": 4.881016102086558e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 3525872,
      "step": 3945
    },
    {
      "epoch": 1.042497030487,
      "grad_norm": 0.3523333668708801,
      "learning_rate": 4.8807166513155425e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 3530576,
      "step": 3950
    },
    {
      "epoch": 1.043816814042497,
      "grad_norm": 0.13776417076587677,
      "learning_rate": 4.8804168334081004e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 3535056,
      "step": 3955
    },
    {
      "epoch": 1.0451365975979938,
      "grad_norm": 0.08046849071979523,
      "learning_rate": 4.880116648410468e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 3539248,
      "step": 3960
    },
    {
      "epoch": 1.0464563811534908,
      "grad_norm": 0.21321848034858704,
      "learning_rate": 4.879816096368939e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 3543344,
      "step": 3965
    },
    {
      "epoch": 1.0477761647089878,
      "grad_norm": 0.10912656038999557,
      "learning_rate": 4.879515177329861e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 3547696,
      "step": 3970
    },
    {
      "epoch": 1.0490959482644846,
      "grad_norm": 0.07434891909360886,
      "learning_rate": 4.8792138913396394e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 3552624,
      "step": 3975
    },
    {
      "epoch": 1.0504157318199816,
      "grad_norm": 0.35398176312446594,
      "learning_rate": 4.8789122384447374e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 3556848,
      "step": 3980
    },
    {
      "epoch": 1.0517355153754784,
      "grad_norm": 0.707976758480072,
      "learning_rate": 4.878610218691673e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 3561616,
      "step": 3985
    },
    {
      "epoch": 1.0530552989309754,
      "grad_norm": 0.2761719226837158,
      "learning_rate": 4.87830783212702e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 3566224,
      "step": 3990
    },
    {
      "epoch": 1.0543750824864722,
      "grad_norm": 0.15643230080604553,
      "learning_rate": 4.878005078797413e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 3570416,
      "step": 3995
    },
    {
      "epoch": 1.0556948660419692,
      "grad_norm": 0.2590215802192688,
      "learning_rate": 4.877701958749539e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 3575248,
      "step": 4000
    },
    {
      "epoch": 1.0556948660419692,
      "eval_loss": 0.08392225950956345,
      "eval_runtime": 64.7543,
      "eval_samples_per_second": 104.009,
      "eval_steps_per_second": 26.006,
      "num_input_tokens_seen": 3575248,
      "step": 4000
    },
    {
      "epoch": 1.057014649597466,
      "grad_norm": 0.5423316955566406,
      "learning_rate": 4.877398472030142e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 3579856,
      "step": 4005
    },
    {
      "epoch": 1.058334433152963,
      "grad_norm": 0.13931109011173248,
      "learning_rate": 4.877094618686024e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 3584304,
      "step": 4010
    },
    {
      "epoch": 1.0596542167084597,
      "grad_norm": 0.7526949644088745,
      "learning_rate": 4.876790398764045e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 3588784,
      "step": 4015
    },
    {
      "epoch": 1.0609740002639567,
      "grad_norm": 0.25323137640953064,
      "learning_rate": 4.8764858123111167e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 3593168,
      "step": 4020
    },
    {
      "epoch": 1.0622937838194537,
      "grad_norm": 0.1867348849773407,
      "learning_rate": 4.876180859374212e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 3597584,
      "step": 4025
    },
    {
      "epoch": 1.0636135673749505,
      "grad_norm": 0.5692929625511169,
      "learning_rate": 4.875875540000357e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 3602224,
      "step": 4030
    },
    {
      "epoch": 1.0649333509304475,
      "grad_norm": 0.27245578169822693,
      "learning_rate": 4.8755698542366376e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 3606576,
      "step": 4035
    },
    {
      "epoch": 1.0662531344859443,
      "grad_norm": 0.3179202973842621,
      "learning_rate": 4.875263802130193e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 3611216,
      "step": 4040
    },
    {
      "epoch": 1.0675729180414413,
      "grad_norm": 0.45173290371894836,
      "learning_rate": 4.8749573837282207e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 3615568,
      "step": 4045
    },
    {
      "epoch": 1.068892701596938,
      "grad_norm": 0.5114203095436096,
      "learning_rate": 4.874650599077974e-05,
      "loss": 0.1048,
      "num_input_tokens_seen": 3619984,
      "step": 4050
    },
    {
      "epoch": 1.070212485152435,
      "grad_norm": 0.08957991749048233,
      "learning_rate": 4.874343448226764e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 3624432,
      "step": 4055
    },
    {
      "epoch": 1.0715322687079318,
      "grad_norm": 0.19655869901180267,
      "learning_rate": 4.874035931221955e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 3628944,
      "step": 4060
    },
    {
      "epoch": 1.0728520522634288,
      "grad_norm": 0.14045877754688263,
      "learning_rate": 4.8737280481109724e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 3633648,
      "step": 4065
    },
    {
      "epoch": 1.0741718358189256,
      "grad_norm": 0.0934409350156784,
      "learning_rate": 4.873419798941294e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 3638192,
      "step": 4070
    },
    {
      "epoch": 1.0754916193744226,
      "grad_norm": 0.249934583902359,
      "learning_rate": 4.873111183760458e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 3642576,
      "step": 4075
    },
    {
      "epoch": 1.0768114029299194,
      "grad_norm": 0.16090010106563568,
      "learning_rate": 4.8728022026160537e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 3647088,
      "step": 4080
    },
    {
      "epoch": 1.0781311864854164,
      "grad_norm": 0.44176578521728516,
      "learning_rate": 4.872492855555732e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 3651664,
      "step": 4085
    },
    {
      "epoch": 1.0794509700409134,
      "grad_norm": 0.21694275736808777,
      "learning_rate": 4.8721831426271956e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 3656208,
      "step": 4090
    },
    {
      "epoch": 1.0807707535964102,
      "grad_norm": 0.20099543035030365,
      "learning_rate": 4.87187306387821e-05,
      "loss": 0.0768,
      "num_input_tokens_seen": 3660816,
      "step": 4095
    },
    {
      "epoch": 1.0820905371519072,
      "grad_norm": 0.16338351368904114,
      "learning_rate": 4.87156261935659e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 3665328,
      "step": 4100
    },
    {
      "epoch": 1.083410320707404,
      "grad_norm": 0.43167057633399963,
      "learning_rate": 4.871251809110211e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 3670032,
      "step": 4105
    },
    {
      "epoch": 1.084730104262901,
      "grad_norm": 0.5297998785972595,
      "learning_rate": 4.8709406331870044e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 3674256,
      "step": 4110
    },
    {
      "epoch": 1.0860498878183977,
      "grad_norm": 0.08996012806892395,
      "learning_rate": 4.8706290916349574e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 3678608,
      "step": 4115
    },
    {
      "epoch": 1.0873696713738947,
      "grad_norm": 0.08041474968194962,
      "learning_rate": 4.8703171845021134e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 3683120,
      "step": 4120
    },
    {
      "epoch": 1.0886894549293915,
      "grad_norm": 0.2204785794019699,
      "learning_rate": 4.870004911836572e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 3687568,
      "step": 4125
    },
    {
      "epoch": 1.0900092384848885,
      "grad_norm": 0.1558510959148407,
      "learning_rate": 4.869692273686489e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 3692112,
      "step": 4130
    },
    {
      "epoch": 1.0913290220403853,
      "grad_norm": 0.1930859535932541,
      "learning_rate": 4.869379270100079e-05,
      "loss": 0.1289,
      "num_input_tokens_seen": 3696688,
      "step": 4135
    },
    {
      "epoch": 1.0926488055958823,
      "grad_norm": 0.1563756763935089,
      "learning_rate": 4.86906590112561e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 3701456,
      "step": 4140
    },
    {
      "epoch": 1.093968589151379,
      "grad_norm": 0.2564115822315216,
      "learning_rate": 4.8687521668114064e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 3706064,
      "step": 4145
    },
    {
      "epoch": 1.095288372706876,
      "grad_norm": 0.10277988761663437,
      "learning_rate": 4.868438067205853e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 3710768,
      "step": 4150
    },
    {
      "epoch": 1.096608156262373,
      "grad_norm": 0.31413015723228455,
      "learning_rate": 4.8681236023573844e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 3715216,
      "step": 4155
    },
    {
      "epoch": 1.0979279398178698,
      "grad_norm": 0.18851597607135773,
      "learning_rate": 4.867808772314497e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 3719632,
      "step": 4160
    },
    {
      "epoch": 1.0992477233733668,
      "grad_norm": 0.22655442357063293,
      "learning_rate": 4.867493577125741e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 3723600,
      "step": 4165
    },
    {
      "epoch": 1.1005675069288636,
      "grad_norm": 0.42436203360557556,
      "learning_rate": 4.867178016839725e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 3728080,
      "step": 4170
    },
    {
      "epoch": 1.1018872904843606,
      "grad_norm": 0.3642787039279938,
      "learning_rate": 4.8668620915051094e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 3732816,
      "step": 4175
    },
    {
      "epoch": 1.1032070740398574,
      "grad_norm": 0.2699391543865204,
      "learning_rate": 4.866545801170616e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 3737008,
      "step": 4180
    },
    {
      "epoch": 1.1045268575953544,
      "grad_norm": 0.4498985707759857,
      "learning_rate": 4.86622914588502e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 3741616,
      "step": 4185
    },
    {
      "epoch": 1.1058466411508512,
      "grad_norm": 0.16889449954032898,
      "learning_rate": 4.865912125697154e-05,
      "loss": 0.1105,
      "num_input_tokens_seen": 3746064,
      "step": 4190
    },
    {
      "epoch": 1.1071664247063482,
      "grad_norm": 0.5030496120452881,
      "learning_rate": 4.865594740655907e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 3750800,
      "step": 4195
    },
    {
      "epoch": 1.108486208261845,
      "grad_norm": 0.16665053367614746,
      "learning_rate": 4.865276990810222e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 3754960,
      "step": 4200
    },
    {
      "epoch": 1.108486208261845,
      "eval_loss": 0.08246643096208572,
      "eval_runtime": 64.8096,
      "eval_samples_per_second": 103.92,
      "eval_steps_per_second": 25.984,
      "num_input_tokens_seen": 3754960,
      "step": 4200
    },
    {
      "epoch": 1.109805991817342,
      "grad_norm": 0.1580665558576584,
      "learning_rate": 4.8649588762091016e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 3759248,
      "step": 4205
    },
    {
      "epoch": 1.1111257753728387,
      "grad_norm": 0.3454453647136688,
      "learning_rate": 4.8646403969016016e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 3763824,
      "step": 4210
    },
    {
      "epoch": 1.1124455589283357,
      "grad_norm": 0.16657036542892456,
      "learning_rate": 4.864321552936838e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 3768080,
      "step": 4215
    },
    {
      "epoch": 1.1137653424838327,
      "grad_norm": 0.12236739695072174,
      "learning_rate": 4.864002344363978e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 3772336,
      "step": 4220
    },
    {
      "epoch": 1.1150851260393295,
      "grad_norm": 0.19526484608650208,
      "learning_rate": 4.863682771232248e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 3777008,
      "step": 4225
    },
    {
      "epoch": 1.1164049095948265,
      "grad_norm": 0.11880189925432205,
      "learning_rate": 4.8633628335909324e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 3781520,
      "step": 4230
    },
    {
      "epoch": 1.1177246931503233,
      "grad_norm": 0.3267318904399872,
      "learning_rate": 4.8630425314893676e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 3785648,
      "step": 4235
    },
    {
      "epoch": 1.1190444767058203,
      "grad_norm": 0.14763163030147552,
      "learning_rate": 4.862721864976948e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 3790032,
      "step": 4240
    },
    {
      "epoch": 1.120364260261317,
      "grad_norm": 0.3275463581085205,
      "learning_rate": 4.862400834103125e-05,
      "loss": 0.0801,
      "num_input_tokens_seen": 3794384,
      "step": 4245
    },
    {
      "epoch": 1.121684043816814,
      "grad_norm": 0.19305244088172913,
      "learning_rate": 4.862079438917406e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 3798704,
      "step": 4250
    },
    {
      "epoch": 1.1230038273723109,
      "grad_norm": 0.16991020739078522,
      "learning_rate": 4.8617576794693536e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 3803152,
      "step": 4255
    },
    {
      "epoch": 1.1243236109278079,
      "grad_norm": 0.4030729830265045,
      "learning_rate": 4.8614355558085875e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 3807792,
      "step": 4260
    },
    {
      "epoch": 1.1256433944833049,
      "grad_norm": 0.13737830519676208,
      "learning_rate": 4.861113067984783e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 3812240,
      "step": 4265
    },
    {
      "epoch": 1.1269631780388016,
      "grad_norm": 0.06664368510246277,
      "learning_rate": 4.860790216047671e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 3816912,
      "step": 4270
    },
    {
      "epoch": 1.1282829615942984,
      "grad_norm": 0.5525670647621155,
      "learning_rate": 4.860467000047041e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 3821392,
      "step": 4275
    },
    {
      "epoch": 1.1296027451497954,
      "grad_norm": 0.16312049329280853,
      "learning_rate": 4.860143420032737e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 3825904,
      "step": 4280
    },
    {
      "epoch": 1.1309225287052924,
      "grad_norm": 0.19475403428077698,
      "learning_rate": 4.859819476054657e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 3830224,
      "step": 4285
    },
    {
      "epoch": 1.1322423122607892,
      "grad_norm": 0.11151828616857529,
      "learning_rate": 4.859495168162758e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 3834864,
      "step": 4290
    },
    {
      "epoch": 1.1335620958162862,
      "grad_norm": 0.14119647443294525,
      "learning_rate": 4.859170496407054e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 3839216,
      "step": 4295
    },
    {
      "epoch": 1.134881879371783,
      "grad_norm": 0.141139954328537,
      "learning_rate": 4.8588454608376114e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 3844048,
      "step": 4300
    },
    {
      "epoch": 1.13620166292728,
      "grad_norm": 0.05400727689266205,
      "learning_rate": 4.8585200615045555e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 3848528,
      "step": 4305
    },
    {
      "epoch": 1.1375214464827768,
      "grad_norm": 0.18937702476978302,
      "learning_rate": 4.8581942984580674e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 3852720,
      "step": 4310
    },
    {
      "epoch": 1.1388412300382738,
      "grad_norm": 0.144100159406662,
      "learning_rate": 4.857868171748384e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 3857232,
      "step": 4315
    },
    {
      "epoch": 1.1401610135937705,
      "grad_norm": 0.21508780121803284,
      "learning_rate": 4.8575416814257976e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 3861680,
      "step": 4320
    },
    {
      "epoch": 1.1414807971492675,
      "grad_norm": 0.2570151388645172,
      "learning_rate": 4.857214827540657e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 3866096,
      "step": 4325
    },
    {
      "epoch": 1.1428005807047645,
      "grad_norm": 0.13804972171783447,
      "learning_rate": 4.856887610143367e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 3870480,
      "step": 4330
    },
    {
      "epoch": 1.1441203642602613,
      "grad_norm": 0.21129173040390015,
      "learning_rate": 4.8565600292843896e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 3874736,
      "step": 4335
    },
    {
      "epoch": 1.145440147815758,
      "grad_norm": 0.42242226004600525,
      "learning_rate": 4.856232085014241e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 3879024,
      "step": 4340
    },
    {
      "epoch": 1.146759931371255,
      "grad_norm": 0.15966832637786865,
      "learning_rate": 4.855903777383495e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 3883504,
      "step": 4345
    },
    {
      "epoch": 1.148079714926752,
      "grad_norm": 0.2897903323173523,
      "learning_rate": 4.85557510644278e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 3888144,
      "step": 4350
    },
    {
      "epoch": 1.1493994984822489,
      "grad_norm": 0.09813579171895981,
      "learning_rate": 4.855246072242782e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 3892432,
      "step": 4355
    },
    {
      "epoch": 1.1507192820377459,
      "grad_norm": 0.15595395863056183,
      "learning_rate": 4.8549166748342414e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 3897104,
      "step": 4360
    },
    {
      "epoch": 1.1520390655932427,
      "grad_norm": 0.31183916330337524,
      "learning_rate": 4.8545869142679556e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 3901552,
      "step": 4365
    },
    {
      "epoch": 1.1533588491487397,
      "grad_norm": 0.1998717039823532,
      "learning_rate": 4.8542567905947776e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 3906096,
      "step": 4370
    },
    {
      "epoch": 1.1546786327042364,
      "grad_norm": 0.1302712857723236,
      "learning_rate": 4.853926303865618e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 3910480,
      "step": 4375
    },
    {
      "epoch": 1.1559984162597334,
      "grad_norm": 0.45480620861053467,
      "learning_rate": 4.853595454131441e-05,
      "loss": 0.1382,
      "num_input_tokens_seen": 3914832,
      "step": 4380
    },
    {
      "epoch": 1.1573181998152302,
      "grad_norm": 0.34436386823654175,
      "learning_rate": 4.8532642414432674e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 3919280,
      "step": 4385
    },
    {
      "epoch": 1.1586379833707272,
      "grad_norm": 0.08240567892789841,
      "learning_rate": 4.8529326658521754e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 3924112,
      "step": 4390
    },
    {
      "epoch": 1.1599577669262242,
      "grad_norm": 0.21992577612400055,
      "learning_rate": 4.8526007274092965e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 3928464,
      "step": 4395
    },
    {
      "epoch": 1.161277550481721,
      "grad_norm": 0.22887659072875977,
      "learning_rate": 4.852268426165822e-05,
      "loss": 0.1007,
      "num_input_tokens_seen": 3932752,
      "step": 4400
    },
    {
      "epoch": 1.161277550481721,
      "eval_loss": 0.08230160176753998,
      "eval_runtime": 64.7534,
      "eval_samples_per_second": 104.01,
      "eval_steps_per_second": 26.006,
      "num_input_tokens_seen": 3932752,
      "step": 4400
    },
    {
      "epoch": 1.162597334037218,
      "grad_norm": 0.08518625050783157,
      "learning_rate": 4.851935762172995e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 3936976,
      "step": 4405
    },
    {
      "epoch": 1.1639171175927148,
      "grad_norm": 0.18197834491729736,
      "learning_rate": 4.8516027354821175e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 3941264,
      "step": 4410
    },
    {
      "epoch": 1.1652369011482118,
      "grad_norm": 0.21768225729465485,
      "learning_rate": 4.851269346144546e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 3945616,
      "step": 4415
    },
    {
      "epoch": 1.1665566847037085,
      "grad_norm": 0.25081053376197815,
      "learning_rate": 4.850935594211693e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 3950128,
      "step": 4420
    },
    {
      "epoch": 1.1678764682592055,
      "grad_norm": 0.42511430382728577,
      "learning_rate": 4.850601479735029e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 3954576,
      "step": 4425
    },
    {
      "epoch": 1.1691962518147023,
      "grad_norm": 0.2510102689266205,
      "learning_rate": 4.850267002766076e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 3958960,
      "step": 4430
    },
    {
      "epoch": 1.1705160353701993,
      "grad_norm": 0.18003112077713013,
      "learning_rate": 4.849932163356417e-05,
      "loss": 0.1153,
      "num_input_tokens_seen": 3963024,
      "step": 4435
    },
    {
      "epoch": 1.171835818925696,
      "grad_norm": 0.13992653787136078,
      "learning_rate": 4.8495969615576864e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 3967632,
      "step": 4440
    },
    {
      "epoch": 1.173155602481193,
      "grad_norm": 0.06868690997362137,
      "learning_rate": 4.849261397421577e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 3972400,
      "step": 4445
    },
    {
      "epoch": 1.1744753860366899,
      "grad_norm": 0.1193857491016388,
      "learning_rate": 4.848925470999839e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 3976784,
      "step": 4450
    },
    {
      "epoch": 1.1757951695921869,
      "grad_norm": 0.27131497859954834,
      "learning_rate": 4.848589182344273e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 3981296,
      "step": 4455
    },
    {
      "epoch": 1.1771149531476839,
      "grad_norm": 0.14180465042591095,
      "learning_rate": 4.848252531506742e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 3985712,
      "step": 4460
    },
    {
      "epoch": 1.1784347367031807,
      "grad_norm": 0.35425296425819397,
      "learning_rate": 4.847915518539161e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 3989872,
      "step": 4465
    },
    {
      "epoch": 1.1797545202586777,
      "grad_norm": 0.2901477515697479,
      "learning_rate": 4.847578143493501e-05,
      "loss": 0.1293,
      "num_input_tokens_seen": 3994352,
      "step": 4470
    },
    {
      "epoch": 1.1810743038141744,
      "grad_norm": 0.24459728598594666,
      "learning_rate": 4.847240406421789e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 3999056,
      "step": 4475
    },
    {
      "epoch": 1.1823940873696714,
      "grad_norm": 0.3255689740180969,
      "learning_rate": 4.84690230737611e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 4003504,
      "step": 4480
    },
    {
      "epoch": 1.1837138709251682,
      "grad_norm": 0.8029049038887024,
      "learning_rate": 4.846563846408602e-05,
      "loss": 0.1275,
      "num_input_tokens_seen": 4008112,
      "step": 4485
    },
    {
      "epoch": 1.1850336544806652,
      "grad_norm": 0.13837052881717682,
      "learning_rate": 4.84622502357146e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 4012496,
      "step": 4490
    },
    {
      "epoch": 1.186353438036162,
      "grad_norm": 0.15304633975028992,
      "learning_rate": 4.8458858389169345e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 4016720,
      "step": 4495
    },
    {
      "epoch": 1.187673221591659,
      "grad_norm": 0.3612307012081146,
      "learning_rate": 4.8455462924973334e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 4021584,
      "step": 4500
    },
    {
      "epoch": 1.1889930051471558,
      "grad_norm": 0.11760736256837845,
      "learning_rate": 4.845206384365018e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 4026032,
      "step": 4505
    },
    {
      "epoch": 1.1903127887026528,
      "grad_norm": 0.19630874693393707,
      "learning_rate": 4.844866114572405e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 4030512,
      "step": 4510
    },
    {
      "epoch": 1.1916325722581496,
      "grad_norm": 0.4415590763092041,
      "learning_rate": 4.8445254831719706e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 4035056,
      "step": 4515
    },
    {
      "epoch": 1.1929523558136466,
      "grad_norm": 0.2646600008010864,
      "learning_rate": 4.8441844902162434e-05,
      "loss": 0.1213,
      "num_input_tokens_seen": 4039408,
      "step": 4520
    },
    {
      "epoch": 1.1942721393691436,
      "grad_norm": 0.07634265720844269,
      "learning_rate": 4.843843135757809e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 4043408,
      "step": 4525
    },
    {
      "epoch": 1.1955919229246403,
      "grad_norm": 0.02797778695821762,
      "learning_rate": 4.843501419849308e-05,
      "loss": 0.0536,
      "num_input_tokens_seen": 4048560,
      "step": 4530
    },
    {
      "epoch": 1.1969117064801373,
      "grad_norm": 0.10018549859523773,
      "learning_rate": 4.8431593425434386e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 4053360,
      "step": 4535
    },
    {
      "epoch": 1.1982314900356341,
      "grad_norm": 0.4374690651893616,
      "learning_rate": 4.8428169038929526e-05,
      "loss": 0.0937,
      "num_input_tokens_seen": 4057744,
      "step": 4540
    },
    {
      "epoch": 1.1995512735911311,
      "grad_norm": 0.19169427454471588,
      "learning_rate": 4.8424741039506575e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 4062416,
      "step": 4545
    },
    {
      "epoch": 1.200871057146628,
      "grad_norm": 0.13905975222587585,
      "learning_rate": 4.842130942769419e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 4067152,
      "step": 4550
    },
    {
      "epoch": 1.202190840702125,
      "grad_norm": 0.3515130579471588,
      "learning_rate": 4.841787420402156e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 4071760,
      "step": 4555
    },
    {
      "epoch": 1.2035106242576217,
      "grad_norm": 0.3192265033721924,
      "learning_rate": 4.841443536901844e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 4076240,
      "step": 4560
    },
    {
      "epoch": 1.2048304078131187,
      "grad_norm": 0.5361230373382568,
      "learning_rate": 4.841099292321514e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 4080432,
      "step": 4565
    },
    {
      "epoch": 1.2061501913686155,
      "grad_norm": 0.267772912979126,
      "learning_rate": 4.8407546867142525e-05,
      "loss": 0.1007,
      "num_input_tokens_seen": 4084976,
      "step": 4570
    },
    {
      "epoch": 1.2074699749241125,
      "grad_norm": 0.09927910566329956,
      "learning_rate": 4.840409720133203e-05,
      "loss": 0.0945,
      "num_input_tokens_seen": 4089936,
      "step": 4575
    },
    {
      "epoch": 1.2087897584796092,
      "grad_norm": 0.22060927748680115,
      "learning_rate": 4.8400643926315634e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 4094608,
      "step": 4580
    },
    {
      "epoch": 1.2101095420351062,
      "grad_norm": 0.4117361605167389,
      "learning_rate": 4.839718704262587e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 4099088,
      "step": 4585
    },
    {
      "epoch": 1.2114293255906032,
      "grad_norm": 0.5627230405807495,
      "learning_rate": 4.839372655079585e-05,
      "loss": 0.0969,
      "num_input_tokens_seen": 4103440,
      "step": 4590
    },
    {
      "epoch": 1.2127491091461,
      "grad_norm": 0.27293699979782104,
      "learning_rate": 4.83902624513592e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 4108336,
      "step": 4595
    },
    {
      "epoch": 1.214068892701597,
      "grad_norm": 0.4113752245903015,
      "learning_rate": 4.838679474485014e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 4112272,
      "step": 4600
    },
    {
      "epoch": 1.214068892701597,
      "eval_loss": 0.08132775127887726,
      "eval_runtime": 64.7147,
      "eval_samples_per_second": 104.072,
      "eval_steps_per_second": 26.022,
      "num_input_tokens_seen": 4112272,
      "step": 4600
    },
    {
      "epoch": 1.2153886762570938,
      "grad_norm": 0.3161044716835022,
      "learning_rate": 4.838332343180343e-05,
      "loss": 0.1105,
      "num_input_tokens_seen": 4116496,
      "step": 4605
    },
    {
      "epoch": 1.2167084598125908,
      "grad_norm": 0.3680439889431,
      "learning_rate": 4.83798485127544e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 4121072,
      "step": 4610
    },
    {
      "epoch": 1.2180282433680876,
      "grad_norm": 0.7212788462638855,
      "learning_rate": 4.837636998823892e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 4125488,
      "step": 4615
    },
    {
      "epoch": 1.2193480269235846,
      "grad_norm": 0.18974103033542633,
      "learning_rate": 4.8372887858793414e-05,
      "loss": 0.1187,
      "num_input_tokens_seen": 4129936,
      "step": 4620
    },
    {
      "epoch": 1.2206678104790813,
      "grad_norm": 0.1769970804452896,
      "learning_rate": 4.836940212495489e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 4133968,
      "step": 4625
    },
    {
      "epoch": 1.2219875940345784,
      "grad_norm": 0.443345844745636,
      "learning_rate": 4.836591278726087e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 4138544,
      "step": 4630
    },
    {
      "epoch": 1.2233073775900754,
      "grad_norm": 0.442806601524353,
      "learning_rate": 4.836241984624947e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 4143024,
      "step": 4635
    },
    {
      "epoch": 1.2246271611455721,
      "grad_norm": 0.2790873944759369,
      "learning_rate": 4.8358923302459336e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 4147472,
      "step": 4640
    },
    {
      "epoch": 1.225946944701069,
      "grad_norm": 0.18140792846679688,
      "learning_rate": 4.835542315642968e-05,
      "loss": 0.0984,
      "num_input_tokens_seen": 4152080,
      "step": 4645
    },
    {
      "epoch": 1.227266728256566,
      "grad_norm": 0.1695387065410614,
      "learning_rate": 4.8351919408700274e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 4156816,
      "step": 4650
    },
    {
      "epoch": 1.228586511812063,
      "grad_norm": 0.11536339670419693,
      "learning_rate": 4.834841205981144e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 4161072,
      "step": 4655
    },
    {
      "epoch": 1.2299062953675597,
      "grad_norm": 0.0577889047563076,
      "learning_rate": 4.8344901110304054e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 4165232,
      "step": 4660
    },
    {
      "epoch": 1.2312260789230567,
      "grad_norm": 0.27206772565841675,
      "learning_rate": 4.8341386560719534e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 4169840,
      "step": 4665
    },
    {
      "epoch": 1.2325458624785535,
      "grad_norm": 0.08307453989982605,
      "learning_rate": 4.833786841159989e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 4174640,
      "step": 4670
    },
    {
      "epoch": 1.2338656460340505,
      "grad_norm": 0.167721226811409,
      "learning_rate": 4.833434666348765e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 4179344,
      "step": 4675
    },
    {
      "epoch": 1.2351854295895472,
      "grad_norm": 0.15354494750499725,
      "learning_rate": 4.833082131692592e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 4184144,
      "step": 4680
    },
    {
      "epoch": 1.2365052131450442,
      "grad_norm": 0.19773054122924805,
      "learning_rate": 4.832729237245835e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 4189040,
      "step": 4685
    },
    {
      "epoch": 1.237824996700541,
      "grad_norm": 0.10577980428934097,
      "learning_rate": 4.8323759830629145e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 4193648,
      "step": 4690
    },
    {
      "epoch": 1.239144780256038,
      "grad_norm": 0.22638793289661407,
      "learning_rate": 4.8320223691983066e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 4197968,
      "step": 4695
    },
    {
      "epoch": 1.240464563811535,
      "grad_norm": 0.6550453305244446,
      "learning_rate": 4.831668395706544e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 4202512,
      "step": 4700
    },
    {
      "epoch": 1.2417843473670318,
      "grad_norm": 0.3708903193473816,
      "learning_rate": 4.8313140626422125e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 4207024,
      "step": 4705
    },
    {
      "epoch": 1.2431041309225286,
      "grad_norm": 0.26724153757095337,
      "learning_rate": 4.830959370059956e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 4211440,
      "step": 4710
    },
    {
      "epoch": 1.2444239144780256,
      "grad_norm": 0.09195451438426971,
      "learning_rate": 4.830604318014472e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 4215760,
      "step": 4715
    },
    {
      "epoch": 1.2457436980335226,
      "grad_norm": 0.30698201060295105,
      "learning_rate": 4.830248906560514e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 4220400,
      "step": 4720
    },
    {
      "epoch": 1.2470634815890194,
      "grad_norm": 0.3779836595058441,
      "learning_rate": 4.829893135752891e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 4224752,
      "step": 4725
    },
    {
      "epoch": 1.2483832651445164,
      "grad_norm": 0.24843882024288177,
      "learning_rate": 4.829537005646466e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 4229264,
      "step": 4730
    },
    {
      "epoch": 1.2497030487000131,
      "grad_norm": 0.3456835150718689,
      "learning_rate": 4.8291805162961615e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 4233648,
      "step": 4735
    },
    {
      "epoch": 1.2510228322555101,
      "grad_norm": 0.3580072224140167,
      "learning_rate": 4.82882366775695e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 4237968,
      "step": 4740
    },
    {
      "epoch": 1.252342615811007,
      "grad_norm": 0.15434615314006805,
      "learning_rate": 4.828466460083864e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 4242448,
      "step": 4745
    },
    {
      "epoch": 1.253662399366504,
      "grad_norm": 0.213471457362175,
      "learning_rate": 4.8281088933319877e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 4246896,
      "step": 4750
    },
    {
      "epoch": 1.2549821829220007,
      "grad_norm": 0.22187449038028717,
      "learning_rate": 4.827750967556464e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 4251312,
      "step": 4755
    },
    {
      "epoch": 1.2563019664774977,
      "grad_norm": 0.20207275450229645,
      "learning_rate": 4.827392682812488e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 4256016,
      "step": 4760
    },
    {
      "epoch": 1.2576217500329947,
      "grad_norm": 0.280529648065567,
      "learning_rate": 4.827034039155312e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 4260656,
      "step": 4765
    },
    {
      "epoch": 1.2589415335884915,
      "grad_norm": 0.10901498794555664,
      "learning_rate": 4.8266750366402445e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 4264848,
      "step": 4770
    },
    {
      "epoch": 1.2602613171439883,
      "grad_norm": 0.16844114661216736,
      "learning_rate": 4.8263156753226476e-05,
      "loss": 0.0859,
      "num_input_tokens_seen": 4269520,
      "step": 4775
    },
    {
      "epoch": 1.2615811006994853,
      "grad_norm": 0.1793612539768219,
      "learning_rate": 4.8259559552579394e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 4274128,
      "step": 4780
    },
    {
      "epoch": 1.2629008842549823,
      "grad_norm": 0.37080785632133484,
      "learning_rate": 4.825595876501593e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 4278416,
      "step": 4785
    },
    {
      "epoch": 1.264220667810479,
      "grad_norm": 0.26953259110450745,
      "learning_rate": 4.825235439109137e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 4282928,
      "step": 4790
    },
    {
      "epoch": 1.265540451365976,
      "grad_norm": 0.4278879761695862,
      "learning_rate": 4.824874643136156e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 4287216,
      "step": 4795
    },
    {
      "epoch": 1.2668602349214728,
      "grad_norm": 0.23623058199882507,
      "learning_rate": 4.824513488638288e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 4291792,
      "step": 4800
    },
    {
      "epoch": 1.2668602349214728,
      "eval_loss": 0.08161582052707672,
      "eval_runtime": 64.7596,
      "eval_samples_per_second": 104.0,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 4291792,
      "step": 4800
    },
    {
      "epoch": 1.2681800184769698,
      "grad_norm": 0.13478226959705353,
      "learning_rate": 4.8241519756712293e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 4296176,
      "step": 4805
    },
    {
      "epoch": 1.2694998020324666,
      "grad_norm": 0.28952741622924805,
      "learning_rate": 4.8237901042907285e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 4300656,
      "step": 4810
    },
    {
      "epoch": 1.2708195855879636,
      "grad_norm": 0.21401876211166382,
      "learning_rate": 4.823427874552591e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 4305232,
      "step": 4815
    },
    {
      "epoch": 1.2721393691434604,
      "grad_norm": 0.1189020425081253,
      "learning_rate": 4.823065286512677e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 4310192,
      "step": 4820
    },
    {
      "epoch": 1.2734591526989574,
      "grad_norm": 0.03214819356799126,
      "learning_rate": 4.8227023402269025e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 4314768,
      "step": 4825
    },
    {
      "epoch": 1.2747789362544544,
      "grad_norm": 0.09765693545341492,
      "learning_rate": 4.822339035751239e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 4319120,
      "step": 4830
    },
    {
      "epoch": 1.2760987198099512,
      "grad_norm": 0.3685755431652069,
      "learning_rate": 4.8219753731417104e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 4323664,
      "step": 4835
    },
    {
      "epoch": 1.277418503365448,
      "grad_norm": 0.467263787984848,
      "learning_rate": 4.821611352454401e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 4328368,
      "step": 4840
    },
    {
      "epoch": 1.278738286920945,
      "grad_norm": 0.29263240098953247,
      "learning_rate": 4.8212469737454444e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 4333136,
      "step": 4845
    },
    {
      "epoch": 1.280058070476442,
      "grad_norm": 0.09253711998462677,
      "learning_rate": 4.820882237071035e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 4337616,
      "step": 4850
    },
    {
      "epoch": 1.2813778540319387,
      "grad_norm": 0.08086097985506058,
      "learning_rate": 4.820517142487417e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 4342192,
      "step": 4855
    },
    {
      "epoch": 1.2826976375874357,
      "grad_norm": 0.6061115860939026,
      "learning_rate": 4.8201516900508956e-05,
      "loss": 0.1437,
      "num_input_tokens_seen": 4346736,
      "step": 4860
    },
    {
      "epoch": 1.2840174211429325,
      "grad_norm": 0.1166522204875946,
      "learning_rate": 4.819785879817827e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 4351216,
      "step": 4865
    },
    {
      "epoch": 1.2853372046984295,
      "grad_norm": 0.17058229446411133,
      "learning_rate": 4.8194197118446226e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 4355408,
      "step": 4870
    },
    {
      "epoch": 1.2866569882539265,
      "grad_norm": 0.24322225153446198,
      "learning_rate": 4.819053186187752e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 4359888,
      "step": 4875
    },
    {
      "epoch": 1.2879767718094233,
      "grad_norm": 0.21464328467845917,
      "learning_rate": 4.818686302903736e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 4364272,
      "step": 4880
    },
    {
      "epoch": 1.28929655536492,
      "grad_norm": 0.20886161923408508,
      "learning_rate": 4.818319062049154e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 4368912,
      "step": 4885
    },
    {
      "epoch": 1.290616338920417,
      "grad_norm": 0.7577788233757019,
      "learning_rate": 4.817951463680639e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 4373232,
      "step": 4890
    },
    {
      "epoch": 1.291936122475914,
      "grad_norm": 0.03194143995642662,
      "learning_rate": 4.817583507854879e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 4377584,
      "step": 4895
    },
    {
      "epoch": 1.2932559060314108,
      "grad_norm": 0.16970330476760864,
      "learning_rate": 4.817215194628617e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 4381968,
      "step": 4900
    },
    {
      "epoch": 1.2945756895869078,
      "grad_norm": 0.18699966371059418,
      "learning_rate": 4.816846524058653e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 4386768,
      "step": 4905
    },
    {
      "epoch": 1.2958954731424046,
      "grad_norm": 0.26557132601737976,
      "learning_rate": 4.816477496201839e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 4391312,
      "step": 4910
    },
    {
      "epoch": 1.2972152566979016,
      "grad_norm": 0.4463260769844055,
      "learning_rate": 4.8161081111150845e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 4395856,
      "step": 4915
    },
    {
      "epoch": 1.2985350402533984,
      "grad_norm": 0.0483197383582592,
      "learning_rate": 4.815738368855354e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 4400240,
      "step": 4920
    },
    {
      "epoch": 1.2998548238088954,
      "grad_norm": 0.3342621922492981,
      "learning_rate": 4.815368269479664e-05,
      "loss": 0.1099,
      "num_input_tokens_seen": 4405008,
      "step": 4925
    },
    {
      "epoch": 1.3011746073643922,
      "grad_norm": 0.4109770655632019,
      "learning_rate": 4.814997813045092e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 4409872,
      "step": 4930
    },
    {
      "epoch": 1.3024943909198892,
      "grad_norm": 0.11418696492910385,
      "learning_rate": 4.814626999608764e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 4414256,
      "step": 4935
    },
    {
      "epoch": 1.3038141744753862,
      "grad_norm": 0.11664246767759323,
      "learning_rate": 4.814255829227865e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 4419024,
      "step": 4940
    },
    {
      "epoch": 1.305133958030883,
      "grad_norm": 0.6612749099731445,
      "learning_rate": 4.813884301959635e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 4423952,
      "step": 4945
    },
    {
      "epoch": 1.3064537415863797,
      "grad_norm": 0.6188772320747375,
      "learning_rate": 4.813512417861368e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 4428304,
      "step": 4950
    },
    {
      "epoch": 1.3077735251418767,
      "grad_norm": 0.6771420836448669,
      "learning_rate": 4.813140176990411e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 4432784,
      "step": 4955
    },
    {
      "epoch": 1.3090933086973737,
      "grad_norm": 0.16383981704711914,
      "learning_rate": 4.8127675794041714e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 4437456,
      "step": 4960
    },
    {
      "epoch": 1.3104130922528705,
      "grad_norm": 0.3050370514392853,
      "learning_rate": 4.812394625160107e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 4442320,
      "step": 4965
    },
    {
      "epoch": 1.3117328758083675,
      "grad_norm": 0.3932335078716278,
      "learning_rate": 4.812021314315732e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 4446832,
      "step": 4970
    },
    {
      "epoch": 1.3130526593638643,
      "grad_norm": 0.5249871611595154,
      "learning_rate": 4.811647646928616e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 4451184,
      "step": 4975
    },
    {
      "epoch": 1.3143724429193613,
      "grad_norm": 0.26924532651901245,
      "learning_rate": 4.8112736230563814e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 4455728,
      "step": 4980
    },
    {
      "epoch": 1.315692226474858,
      "grad_norm": 0.28766703605651855,
      "learning_rate": 4.81089924275671e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 4459920,
      "step": 4985
    },
    {
      "epoch": 1.317012010030355,
      "grad_norm": 0.16108350455760956,
      "learning_rate": 4.810524506087335e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 4464336,
      "step": 4990
    },
    {
      "epoch": 1.3183317935858518,
      "grad_norm": 0.11470439285039902,
      "learning_rate": 4.810149413106044e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 4468528,
      "step": 4995
    },
    {
      "epoch": 1.3196515771413488,
      "grad_norm": 0.27432647347450256,
      "learning_rate": 4.809773963870684e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 4472784,
      "step": 5000
    },
    {
      "epoch": 1.3196515771413488,
      "eval_loss": 0.08041132986545563,
      "eval_runtime": 64.7123,
      "eval_samples_per_second": 104.076,
      "eval_steps_per_second": 26.023,
      "num_input_tokens_seen": 4472784,
      "step": 5000
    },
    {
      "epoch": 1.3209713606968458,
      "grad_norm": 0.20007339119911194,
      "learning_rate": 4.809398158439151e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 4477232,
      "step": 5005
    },
    {
      "epoch": 1.3222911442523426,
      "grad_norm": 0.3440200686454773,
      "learning_rate": 4.8090219968694005e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 4481840,
      "step": 5010
    },
    {
      "epoch": 1.3236109278078394,
      "grad_norm": 0.2975916266441345,
      "learning_rate": 4.808645479219442e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 4486416,
      "step": 5015
    },
    {
      "epoch": 1.3249307113633364,
      "grad_norm": 0.22423741221427917,
      "learning_rate": 4.8082686055473375e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 4490928,
      "step": 5020
    },
    {
      "epoch": 1.3262504949188334,
      "grad_norm": 0.3804813027381897,
      "learning_rate": 4.8078913759112066e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 4495344,
      "step": 5025
    },
    {
      "epoch": 1.3275702784743302,
      "grad_norm": 0.6519297361373901,
      "learning_rate": 4.807513790369223e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 4499504,
      "step": 5030
    },
    {
      "epoch": 1.3288900620298272,
      "grad_norm": 0.2891515791416168,
      "learning_rate": 4.8071358489796145e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 4504016,
      "step": 5035
    },
    {
      "epoch": 1.330209845585324,
      "grad_norm": 0.22226184606552124,
      "learning_rate": 4.806757551800665e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 4508144,
      "step": 5040
    },
    {
      "epoch": 1.331529629140821,
      "grad_norm": 0.08077623695135117,
      "learning_rate": 4.806378898890713e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 4512592,
      "step": 5045
    },
    {
      "epoch": 1.3328494126963177,
      "grad_norm": 0.5659134387969971,
      "learning_rate": 4.80599989030815e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 4516848,
      "step": 5050
    },
    {
      "epoch": 1.3341691962518147,
      "grad_norm": 0.1509067714214325,
      "learning_rate": 4.805620526111426e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 4521296,
      "step": 5055
    },
    {
      "epoch": 1.3354889798073115,
      "grad_norm": 0.2895316481590271,
      "learning_rate": 4.805240806359042e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 4525584,
      "step": 5060
    },
    {
      "epoch": 1.3368087633628085,
      "grad_norm": 0.3272639811038971,
      "learning_rate": 4.804860731109557e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 4530320,
      "step": 5065
    },
    {
      "epoch": 1.3381285469183055,
      "grad_norm": 0.2112981677055359,
      "learning_rate": 4.804480300421581e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 4534640,
      "step": 5070
    },
    {
      "epoch": 1.3394483304738023,
      "grad_norm": 0.3563102185726166,
      "learning_rate": 4.804099514353784e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 4538928,
      "step": 5075
    },
    {
      "epoch": 1.340768114029299,
      "grad_norm": 0.10504698008298874,
      "learning_rate": 4.8037183729648867e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 4543280,
      "step": 5080
    },
    {
      "epoch": 1.342087897584796,
      "grad_norm": 0.2180953025817871,
      "learning_rate": 4.803336876313666e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 4547504,
      "step": 5085
    },
    {
      "epoch": 1.343407681140293,
      "grad_norm": 0.2923080027103424,
      "learning_rate": 4.802955024458953e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 4551984,
      "step": 5090
    },
    {
      "epoch": 1.3447274646957899,
      "grad_norm": 0.06934521347284317,
      "learning_rate": 4.802572817459634e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 4556784,
      "step": 5095
    },
    {
      "epoch": 1.3460472482512869,
      "grad_norm": 0.0800878182053566,
      "learning_rate": 4.802190255374651e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 4561520,
      "step": 5100
    },
    {
      "epoch": 1.3473670318067836,
      "grad_norm": 0.06487201899290085,
      "learning_rate": 4.801807338263e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 4565648,
      "step": 5105
    },
    {
      "epoch": 1.3486868153622806,
      "grad_norm": 0.43343186378479004,
      "learning_rate": 4.8014240661837306e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 4570352,
      "step": 5110
    },
    {
      "epoch": 1.3500065989177774,
      "grad_norm": 0.24935777485370636,
      "learning_rate": 4.80104043919595e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 4575088,
      "step": 5115
    },
    {
      "epoch": 1.3513263824732744,
      "grad_norm": 0.165144145488739,
      "learning_rate": 4.800656457358815e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 4579664,
      "step": 5120
    },
    {
      "epoch": 1.3526461660287712,
      "grad_norm": 0.19699345529079437,
      "learning_rate": 4.800272120731544e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 4584144,
      "step": 5125
    },
    {
      "epoch": 1.3539659495842682,
      "grad_norm": 0.10191094130277634,
      "learning_rate": 4.799887429373404e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 4588720,
      "step": 5130
    },
    {
      "epoch": 1.3552857331397652,
      "grad_norm": 0.11645718663930893,
      "learning_rate": 4.79950238334372e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 4593072,
      "step": 5135
    },
    {
      "epoch": 1.356605516695262,
      "grad_norm": 0.33851832151412964,
      "learning_rate": 4.799116982701872e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 4597680,
      "step": 5140
    },
    {
      "epoch": 1.3579253002507587,
      "grad_norm": 0.6722334027290344,
      "learning_rate": 4.7987312275072926e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 4602064,
      "step": 5145
    },
    {
      "epoch": 1.3592450838062557,
      "grad_norm": 0.2841826379299164,
      "learning_rate": 4.79834511781947e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 4606416,
      "step": 5150
    },
    {
      "epoch": 1.3605648673617527,
      "grad_norm": 0.3335546553134918,
      "learning_rate": 4.797958653697947e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 4610896,
      "step": 5155
    },
    {
      "epoch": 1.3618846509172495,
      "grad_norm": 0.030426545068621635,
      "learning_rate": 4.7975718352023225e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 4615376,
      "step": 5160
    },
    {
      "epoch": 1.3632044344727465,
      "grad_norm": 0.26046568155288696,
      "learning_rate": 4.7971846623922476e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 4619536,
      "step": 5165
    },
    {
      "epoch": 1.3645242180282433,
      "grad_norm": 0.0727534219622612,
      "learning_rate": 4.7967971353274294e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 4624080,
      "step": 5170
    },
    {
      "epoch": 1.3658440015837403,
      "grad_norm": 0.1809157282114029,
      "learning_rate": 4.79640925406763e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 4628432,
      "step": 5175
    },
    {
      "epoch": 1.367163785139237,
      "grad_norm": 0.2920337915420532,
      "learning_rate": 4.796021018672664e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 4633232,
      "step": 5180
    },
    {
      "epoch": 1.368483568694734,
      "grad_norm": 0.783295214176178,
      "learning_rate": 4.795632429202405e-05,
      "loss": 0.132,
      "num_input_tokens_seen": 4637840,
      "step": 5185
    },
    {
      "epoch": 1.3698033522502309,
      "grad_norm": 0.08794157207012177,
      "learning_rate": 4.795243485716775e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 4642544,
      "step": 5190
    },
    {
      "epoch": 1.3711231358057279,
      "grad_norm": 0.20806986093521118,
      "learning_rate": 4.794854188275757e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 4647280,
      "step": 5195
    },
    {
      "epoch": 1.3724429193612249,
      "grad_norm": 0.23969388008117676,
      "learning_rate": 4.794464536939384e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 4651696,
      "step": 5200
    },
    {
      "epoch": 1.3724429193612249,
      "eval_loss": 0.07964098453521729,
      "eval_runtime": 64.6576,
      "eval_samples_per_second": 104.164,
      "eval_steps_per_second": 26.045,
      "num_input_tokens_seen": 4651696,
      "step": 5200
    },
    {
      "epoch": 1.3737627029167216,
      "grad_norm": 0.22421321272850037,
      "learning_rate": 4.794074531767745e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 4656144,
      "step": 5205
    },
    {
      "epoch": 1.3750824864722184,
      "grad_norm": 0.16617850959300995,
      "learning_rate": 4.7936841728209834e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 4660624,
      "step": 5210
    },
    {
      "epoch": 1.3764022700277154,
      "grad_norm": 0.020756810903549194,
      "learning_rate": 4.7932934601593e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 4665200,
      "step": 5215
    },
    {
      "epoch": 1.3777220535832124,
      "grad_norm": 0.21342386305332184,
      "learning_rate": 4.792902393842943e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 4669552,
      "step": 5220
    },
    {
      "epoch": 1.3790418371387092,
      "grad_norm": 0.46818751096725464,
      "learning_rate": 4.792510973932225e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 4673712,
      "step": 5225
    },
    {
      "epoch": 1.3803616206942062,
      "grad_norm": 0.33836469054222107,
      "learning_rate": 4.7921192004875036e-05,
      "loss": 0.0986,
      "num_input_tokens_seen": 4678160,
      "step": 5230
    },
    {
      "epoch": 1.381681404249703,
      "grad_norm": 0.3834356665611267,
      "learning_rate": 4.791727073569198e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 4682960,
      "step": 5235
    },
    {
      "epoch": 1.3830011878052,
      "grad_norm": 0.18832901120185852,
      "learning_rate": 4.7913345932377775e-05,
      "loss": 0.096,
      "num_input_tokens_seen": 4687408,
      "step": 5240
    },
    {
      "epoch": 1.384320971360697,
      "grad_norm": 0.16693943738937378,
      "learning_rate": 4.790941759553769e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 4691888,
      "step": 5245
    },
    {
      "epoch": 1.3856407549161938,
      "grad_norm": 0.07065894454717636,
      "learning_rate": 4.79054857257775e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 4696240,
      "step": 5250
    },
    {
      "epoch": 1.3869605384716905,
      "grad_norm": 0.2353706657886505,
      "learning_rate": 4.790155032370357e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 4700816,
      "step": 5255
    },
    {
      "epoch": 1.3882803220271875,
      "grad_norm": 0.03853423520922661,
      "learning_rate": 4.789761138992278e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 4705360,
      "step": 5260
    },
    {
      "epoch": 1.3896001055826845,
      "grad_norm": 0.14937782287597656,
      "learning_rate": 4.7893668925042565e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 4709712,
      "step": 5265
    },
    {
      "epoch": 1.3909198891381813,
      "grad_norm": 0.4335743188858032,
      "learning_rate": 4.78897229296709e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 4714032,
      "step": 5270
    },
    {
      "epoch": 1.392239672693678,
      "grad_norm": 0.2282271832227707,
      "learning_rate": 4.7885773404416315e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 4718384,
      "step": 5275
    },
    {
      "epoch": 1.393559456249175,
      "grad_norm": 0.11093328893184662,
      "learning_rate": 4.788182034988786e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 4722832,
      "step": 5280
    },
    {
      "epoch": 1.394879239804672,
      "grad_norm": 0.38227182626724243,
      "learning_rate": 4.787786376669516e-05,
      "loss": 0.1125,
      "num_input_tokens_seen": 4726992,
      "step": 5285
    },
    {
      "epoch": 1.3961990233601689,
      "grad_norm": 0.14109593629837036,
      "learning_rate": 4.787390365544837e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 4731216,
      "step": 5290
    },
    {
      "epoch": 1.3975188069156659,
      "grad_norm": 0.060291215777397156,
      "learning_rate": 4.786994001675818e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 4735504,
      "step": 5295
    },
    {
      "epoch": 1.3988385904711627,
      "grad_norm": 0.12635262310504913,
      "learning_rate": 4.786597285123584e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 4740144,
      "step": 5300
    },
    {
      "epoch": 1.4001583740266597,
      "grad_norm": 0.3112564980983734,
      "learning_rate": 4.7862002159493135e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 4744624,
      "step": 5305
    },
    {
      "epoch": 1.4014781575821567,
      "grad_norm": 0.34095343947410583,
      "learning_rate": 4.785802794214239e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 4749136,
      "step": 5310
    },
    {
      "epoch": 1.4027979411376534,
      "grad_norm": 0.32172417640686035,
      "learning_rate": 4.7854050199796495e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 4753552,
      "step": 5315
    },
    {
      "epoch": 1.4041177246931502,
      "grad_norm": 0.24310630559921265,
      "learning_rate": 4.7850068933068845e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 4758256,
      "step": 5320
    },
    {
      "epoch": 1.4054375082486472,
      "grad_norm": 0.32415255904197693,
      "learning_rate": 4.7846084142573425e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 4762832,
      "step": 5325
    },
    {
      "epoch": 1.4067572918041442,
      "grad_norm": 0.11996932327747345,
      "learning_rate": 4.7842095828924725e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 4766960,
      "step": 5330
    },
    {
      "epoch": 1.408077075359641,
      "grad_norm": 0.23023372888565063,
      "learning_rate": 4.783810399273779e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 4771248,
      "step": 5335
    },
    {
      "epoch": 1.409396858915138,
      "grad_norm": 0.1694478690624237,
      "learning_rate": 4.7834108634628226e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 4775440,
      "step": 5340
    },
    {
      "epoch": 1.4107166424706348,
      "grad_norm": 0.19602078199386597,
      "learning_rate": 4.783010975521216e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 4780080,
      "step": 5345
    },
    {
      "epoch": 1.4120364260261318,
      "grad_norm": 0.08027384430170059,
      "learning_rate": 4.782610735510626e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 4784688,
      "step": 5350
    },
    {
      "epoch": 1.4133562095816286,
      "grad_norm": 0.16327489912509918,
      "learning_rate": 4.782210143492776e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 4789040,
      "step": 5355
    },
    {
      "epoch": 1.4146759931371256,
      "grad_norm": 0.18734019994735718,
      "learning_rate": 4.781809199529442e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 4793520,
      "step": 5360
    },
    {
      "epoch": 1.4159957766926223,
      "grad_norm": 0.16671977937221527,
      "learning_rate": 4.781407903682454e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 4797968,
      "step": 5365
    },
    {
      "epoch": 1.4173155602481193,
      "grad_norm": 0.1296880692243576,
      "learning_rate": 4.781006256013698e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 4802544,
      "step": 5370
    },
    {
      "epoch": 1.4186353438036163,
      "grad_norm": 0.36691993474960327,
      "learning_rate": 4.7806042565851115e-05,
      "loss": 0.1043,
      "num_input_tokens_seen": 4806800,
      "step": 5375
    },
    {
      "epoch": 1.4199551273591131,
      "grad_norm": 0.20597416162490845,
      "learning_rate": 4.7802019054586895e-05,
      "loss": 0.121,
      "num_input_tokens_seen": 4811344,
      "step": 5380
    },
    {
      "epoch": 1.42127491091461,
      "grad_norm": 0.10593754798173904,
      "learning_rate": 4.779799202696479e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 4816144,
      "step": 5385
    },
    {
      "epoch": 1.422594694470107,
      "grad_norm": 0.3683891296386719,
      "learning_rate": 4.779396148360581e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 4820528,
      "step": 5390
    },
    {
      "epoch": 1.423914478025604,
      "grad_norm": 0.3037063479423523,
      "learning_rate": 4.7789927425131517e-05,
      "loss": 0.1908,
      "num_input_tokens_seen": 4824976,
      "step": 5395
    },
    {
      "epoch": 1.4252342615811007,
      "grad_norm": 0.24169357120990753,
      "learning_rate": 4.778588985216403e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 4829360,
      "step": 5400
    },
    {
      "epoch": 1.4252342615811007,
      "eval_loss": 0.07894226163625717,
      "eval_runtime": 64.7772,
      "eval_samples_per_second": 103.972,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 4829360,
      "step": 5400
    },
    {
      "epoch": 1.4265540451365977,
      "grad_norm": 0.20239225029945374,
      "learning_rate": 4.778184876532598e-05,
      "loss": 0.1145,
      "num_input_tokens_seen": 4833680,
      "step": 5405
    },
    {
      "epoch": 1.4278738286920944,
      "grad_norm": 0.20283399522304535,
      "learning_rate": 4.7777804165240556e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 4838000,
      "step": 5410
    },
    {
      "epoch": 1.4291936122475914,
      "grad_norm": 0.46933332085609436,
      "learning_rate": 4.7773756052531485e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 4842384,
      "step": 5415
    },
    {
      "epoch": 1.4305133958030882,
      "grad_norm": 0.16172149777412415,
      "learning_rate": 4.7769704427823035e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 4846768,
      "step": 5420
    },
    {
      "epoch": 1.4318331793585852,
      "grad_norm": 0.3040544092655182,
      "learning_rate": 4.776564929174003e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 4851440,
      "step": 5425
    },
    {
      "epoch": 1.433152962914082,
      "grad_norm": 0.16845110058784485,
      "learning_rate": 4.7761590644907806e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 4856304,
      "step": 5430
    },
    {
      "epoch": 1.434472746469579,
      "grad_norm": 0.33745089173316956,
      "learning_rate": 4.7757528487952263e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 4860784,
      "step": 5435
    },
    {
      "epoch": 1.435792530025076,
      "grad_norm": 0.2157873809337616,
      "learning_rate": 4.7753462821499836e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 4865104,
      "step": 5440
    },
    {
      "epoch": 1.4371123135805728,
      "grad_norm": 0.07641094923019409,
      "learning_rate": 4.774939364617751e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 4869680,
      "step": 5445
    },
    {
      "epoch": 1.4384320971360696,
      "grad_norm": 0.14259940385818481,
      "learning_rate": 4.7745320962612795e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 4873936,
      "step": 5450
    },
    {
      "epoch": 1.4397518806915666,
      "grad_norm": 0.1825931966304779,
      "learning_rate": 4.7741244771433756e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 4878352,
      "step": 5455
    },
    {
      "epoch": 1.4410716642470636,
      "grad_norm": 0.4214733839035034,
      "learning_rate": 4.7737165073268985e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 4882704,
      "step": 5460
    },
    {
      "epoch": 1.4423914478025603,
      "grad_norm": 0.17257174849510193,
      "learning_rate": 4.7733081868747626e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 4887152,
      "step": 5465
    },
    {
      "epoch": 1.4437112313580573,
      "grad_norm": 0.5272191166877747,
      "learning_rate": 4.772899515849936e-05,
      "loss": 0.1172,
      "num_input_tokens_seen": 4891280,
      "step": 5470
    },
    {
      "epoch": 1.4450310149135541,
      "grad_norm": 0.28314852714538574,
      "learning_rate": 4.7724904943154414e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 4895664,
      "step": 5475
    },
    {
      "epoch": 1.4463507984690511,
      "grad_norm": 0.2088296264410019,
      "learning_rate": 4.772081122334354e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 4900208,
      "step": 5480
    },
    {
      "epoch": 1.447670582024548,
      "grad_norm": 0.3795498013496399,
      "learning_rate": 4.771671399969806e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 4904496,
      "step": 5485
    },
    {
      "epoch": 1.448990365580045,
      "grad_norm": 0.19824357330799103,
      "learning_rate": 4.7712613272849794e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 4908912,
      "step": 5490
    },
    {
      "epoch": 1.4503101491355417,
      "grad_norm": 0.3328114449977875,
      "learning_rate": 4.770850904343114e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 4913456,
      "step": 5495
    },
    {
      "epoch": 1.4516299326910387,
      "grad_norm": 0.1506216675043106,
      "learning_rate": 4.770440131207502e-05,
      "loss": 0.0326,
      "num_input_tokens_seen": 4918128,
      "step": 5500
    },
    {
      "epoch": 1.4529497162465357,
      "grad_norm": 0.1752959042787552,
      "learning_rate": 4.7700290079414896e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 4922576,
      "step": 5505
    },
    {
      "epoch": 1.4542694998020325,
      "grad_norm": 0.08696412295103073,
      "learning_rate": 4.769617534608477e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 4926928,
      "step": 5510
    },
    {
      "epoch": 1.4555892833575292,
      "grad_norm": 0.09538356959819794,
      "learning_rate": 4.7692057112719193e-05,
      "loss": 0.0304,
      "num_input_tokens_seen": 4931344,
      "step": 5515
    },
    {
      "epoch": 1.4569090669130262,
      "grad_norm": 0.058067869395017624,
      "learning_rate": 4.7687935379953234e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 4936016,
      "step": 5520
    },
    {
      "epoch": 1.4582288504685232,
      "grad_norm": 0.03800162300467491,
      "learning_rate": 4.7683810148422534e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 4940848,
      "step": 5525
    },
    {
      "epoch": 1.45954863402402,
      "grad_norm": 0.41322988271713257,
      "learning_rate": 4.767968141876324e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 4945200,
      "step": 5530
    },
    {
      "epoch": 1.460868417579517,
      "grad_norm": 0.21540051698684692,
      "learning_rate": 4.767554919161207e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 4949936,
      "step": 5535
    },
    {
      "epoch": 1.4621882011350138,
      "grad_norm": 0.0661291778087616,
      "learning_rate": 4.767141346760624e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 4954480,
      "step": 5540
    },
    {
      "epoch": 1.4635079846905108,
      "grad_norm": 0.41380563378334045,
      "learning_rate": 4.766727424738356e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 4958832,
      "step": 5545
    },
    {
      "epoch": 1.4648277682460076,
      "grad_norm": 0.19801203906536102,
      "learning_rate": 4.7663131531582325e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 4963504,
      "step": 5550
    },
    {
      "epoch": 1.4661475518015046,
      "grad_norm": 0.43371960520744324,
      "learning_rate": 4.765898532084142e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 4968208,
      "step": 5555
    },
    {
      "epoch": 1.4674673353570014,
      "grad_norm": 0.1632298082113266,
      "learning_rate": 4.765483561580022e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 4972528,
      "step": 5560
    },
    {
      "epoch": 1.4687871189124984,
      "grad_norm": 0.39230212569236755,
      "learning_rate": 4.7650682417098666e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 4977072,
      "step": 5565
    },
    {
      "epoch": 1.4701069024679954,
      "grad_norm": 0.1705838292837143,
      "learning_rate": 4.7646525725377244e-05,
      "loss": 0.155,
      "num_input_tokens_seen": 4981520,
      "step": 5570
    },
    {
      "epoch": 1.4714266860234921,
      "grad_norm": 0.17158518731594086,
      "learning_rate": 4.764236554127696e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 4986000,
      "step": 5575
    },
    {
      "epoch": 1.472746469578989,
      "grad_norm": 0.1324654370546341,
      "learning_rate": 4.7638201865439356e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 4990032,
      "step": 5580
    },
    {
      "epoch": 1.474066253134486,
      "grad_norm": 0.08171235024929047,
      "learning_rate": 4.7634034698506545e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 4994256,
      "step": 5585
    },
    {
      "epoch": 1.475386036689983,
      "grad_norm": 0.20350681245326996,
      "learning_rate": 4.762986404112115e-05,
      "loss": 0.1045,
      "num_input_tokens_seen": 4998672,
      "step": 5590
    },
    {
      "epoch": 1.4767058202454797,
      "grad_norm": 0.31480562686920166,
      "learning_rate": 4.762568989392633e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 5003312,
      "step": 5595
    },
    {
      "epoch": 1.4780256038009767,
      "grad_norm": 0.24262858927249908,
      "learning_rate": 4.76215122575658e-05,
      "loss": 0.0919,
      "num_input_tokens_seen": 5007920,
      "step": 5600
    },
    {
      "epoch": 1.4780256038009767,
      "eval_loss": 0.0785464346408844,
      "eval_runtime": 64.7571,
      "eval_samples_per_second": 104.004,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 5007920,
      "step": 5600
    },
    {
      "epoch": 1.4793453873564735,
      "grad_norm": 0.10564015805721283,
      "learning_rate": 4.7617331132683795e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 5012144,
      "step": 5605
    },
    {
      "epoch": 1.4806651709119705,
      "grad_norm": 0.06095464155077934,
      "learning_rate": 4.7613146519925105e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 5016624,
      "step": 5610
    },
    {
      "epoch": 1.4819849544674673,
      "grad_norm": 0.3032098710536957,
      "learning_rate": 4.7608958419935045e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 5021008,
      "step": 5615
    },
    {
      "epoch": 1.4833047380229643,
      "grad_norm": 0.30858275294303894,
      "learning_rate": 4.760476683335948e-05,
      "loss": 0.1104,
      "num_input_tokens_seen": 5025392,
      "step": 5620
    },
    {
      "epoch": 1.484624521578461,
      "grad_norm": 0.2372182309627533,
      "learning_rate": 4.760057176084479e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 5030096,
      "step": 5625
    },
    {
      "epoch": 1.485944305133958,
      "grad_norm": 0.10320360958576202,
      "learning_rate": 4.759637320303793e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 5035120,
      "step": 5630
    },
    {
      "epoch": 1.487264088689455,
      "grad_norm": 0.49019500613212585,
      "learning_rate": 4.759217116058635e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 5039408,
      "step": 5635
    },
    {
      "epoch": 1.4885838722449518,
      "grad_norm": 0.352451354265213,
      "learning_rate": 4.758796563413807e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 5043728,
      "step": 5640
    },
    {
      "epoch": 1.4899036558004486,
      "grad_norm": 0.05368480831384659,
      "learning_rate": 4.758375662434163e-05,
      "loss": 0.0223,
      "num_input_tokens_seen": 5048272,
      "step": 5645
    },
    {
      "epoch": 1.4912234393559456,
      "grad_norm": 0.27110958099365234,
      "learning_rate": 4.7579544131846114e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 5052880,
      "step": 5650
    },
    {
      "epoch": 1.4925432229114426,
      "grad_norm": 0.08626504987478256,
      "learning_rate": 4.757532815730114e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 5057328,
      "step": 5655
    },
    {
      "epoch": 1.4938630064669394,
      "grad_norm": 0.23050181567668915,
      "learning_rate": 4.7571108701356865e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 5061936,
      "step": 5660
    },
    {
      "epoch": 1.4951827900224364,
      "grad_norm": 0.4139077365398407,
      "learning_rate": 4.756688576466398e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 5066416,
      "step": 5665
    },
    {
      "epoch": 1.4965025735779331,
      "grad_norm": 0.3134812116622925,
      "learning_rate": 4.756265934787372e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 5070832,
      "step": 5670
    },
    {
      "epoch": 1.4978223571334301,
      "grad_norm": 0.06357229501008987,
      "learning_rate": 4.755842945163785e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 5075440,
      "step": 5675
    },
    {
      "epoch": 1.4991421406889271,
      "grad_norm": 0.07387790828943253,
      "learning_rate": 4.755419607660867e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 5079888,
      "step": 5680
    },
    {
      "epoch": 1.500461924244424,
      "grad_norm": 0.18835237622261047,
      "learning_rate": 4.7549959223439016e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 5084272,
      "step": 5685
    },
    {
      "epoch": 1.5017817077999207,
      "grad_norm": 0.11256110668182373,
      "learning_rate": 4.754571889278228e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 5089040,
      "step": 5690
    },
    {
      "epoch": 1.5031014913554177,
      "grad_norm": 0.07079589366912842,
      "learning_rate": 4.754147508529235e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 5093552,
      "step": 5695
    },
    {
      "epoch": 1.5044212749109147,
      "grad_norm": 0.18633604049682617,
      "learning_rate": 4.75372278016237e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 5098128,
      "step": 5700
    },
    {
      "epoch": 1.5057410584664115,
      "grad_norm": 0.07297640293836594,
      "learning_rate": 4.753297704243129e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 5102576,
      "step": 5705
    },
    {
      "epoch": 1.5070608420219083,
      "grad_norm": 0.20948012173175812,
      "learning_rate": 4.752872280837066e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 5106960,
      "step": 5710
    },
    {
      "epoch": 1.5083806255774053,
      "grad_norm": 0.24794456362724304,
      "learning_rate": 4.752446510009786e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 5111600,
      "step": 5715
    },
    {
      "epoch": 1.5097004091329023,
      "grad_norm": 0.273430198431015,
      "learning_rate": 4.7520203918269476e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 5116144,
      "step": 5720
    },
    {
      "epoch": 1.5110201926883993,
      "grad_norm": 0.20416802167892456,
      "learning_rate": 4.751593926354265e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 5120624,
      "step": 5725
    },
    {
      "epoch": 1.512339976243896,
      "grad_norm": 0.3313671052455902,
      "learning_rate": 4.751167113657503e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 5124944,
      "step": 5730
    },
    {
      "epoch": 1.5136597597993928,
      "grad_norm": 0.22932900488376617,
      "learning_rate": 4.7507399538024834e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 5129744,
      "step": 5735
    },
    {
      "epoch": 1.5149795433548898,
      "grad_norm": 0.16310812532901764,
      "learning_rate": 4.750312446855077e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 5134000,
      "step": 5740
    },
    {
      "epoch": 1.5162993269103868,
      "grad_norm": 0.1686880737543106,
      "learning_rate": 4.749884592881212e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 5138608,
      "step": 5745
    },
    {
      "epoch": 1.5176191104658836,
      "grad_norm": 0.03517255559563637,
      "learning_rate": 4.74945639194687e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 5143056,
      "step": 5750
    },
    {
      "epoch": 1.5189388940213804,
      "grad_norm": 0.26234811544418335,
      "learning_rate": 4.749027844118083e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 5147696,
      "step": 5755
    },
    {
      "epoch": 1.5202586775768774,
      "grad_norm": 0.02521154098212719,
      "learning_rate": 4.7485989494609395e-05,
      "loss": 0.0238,
      "num_input_tokens_seen": 5152208,
      "step": 5760
    },
    {
      "epoch": 1.5215784611323744,
      "grad_norm": 0.22014053165912628,
      "learning_rate": 4.748169708041581e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 5156592,
      "step": 5765
    },
    {
      "epoch": 1.5228982446878712,
      "grad_norm": 0.31146240234375,
      "learning_rate": 4.7477401199262004e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 5161328,
      "step": 5770
    },
    {
      "epoch": 1.524218028243368,
      "grad_norm": 0.21080772578716278,
      "learning_rate": 4.747310185181048e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 5165968,
      "step": 5775
    },
    {
      "epoch": 1.525537811798865,
      "grad_norm": 0.05414314568042755,
      "learning_rate": 4.746879903872422e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 5170448,
      "step": 5780
    },
    {
      "epoch": 1.526857595354362,
      "grad_norm": 0.34524157643318176,
      "learning_rate": 4.746449276066679e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 5174928,
      "step": 5785
    },
    {
      "epoch": 1.528177378909859,
      "grad_norm": 0.15992657840251923,
      "learning_rate": 4.746018301830227e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 5179376,
      "step": 5790
    },
    {
      "epoch": 1.5294971624653557,
      "grad_norm": 0.376916766166687,
      "learning_rate": 4.7455869812295275e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 5183664,
      "step": 5795
    },
    {
      "epoch": 1.5308169460208525,
      "grad_norm": 0.34162667393684387,
      "learning_rate": 4.7451553143310964e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 5188208,
      "step": 5800
    },
    {
      "epoch": 1.5308169460208525,
      "eval_loss": 0.07894091308116913,
      "eval_runtime": 64.7573,
      "eval_samples_per_second": 104.004,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 5188208,
      "step": 5800
    },
    {
      "epoch": 1.5321367295763495,
      "grad_norm": 0.20969893038272858,
      "learning_rate": 4.744723301201501e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 5192656,
      "step": 5805
    },
    {
      "epoch": 1.5334565131318465,
      "grad_norm": 0.30848294496536255,
      "learning_rate": 4.744290941907364e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 5197296,
      "step": 5810
    },
    {
      "epoch": 1.5347762966873433,
      "grad_norm": 0.49358293414115906,
      "learning_rate": 4.7438582365153594e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 5202128,
      "step": 5815
    },
    {
      "epoch": 1.53609608024284,
      "grad_norm": 0.09449385851621628,
      "learning_rate": 4.743425185092217e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 5206704,
      "step": 5820
    },
    {
      "epoch": 1.537415863798337,
      "grad_norm": 0.0785050019621849,
      "learning_rate": 4.742991787704719e-05,
      "loss": 0.0917,
      "num_input_tokens_seen": 5211088,
      "step": 5825
    },
    {
      "epoch": 1.538735647353834,
      "grad_norm": 0.2837215065956116,
      "learning_rate": 4.7425580444196994e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 5215568,
      "step": 5830
    },
    {
      "epoch": 1.5400554309093308,
      "grad_norm": 0.06769419461488724,
      "learning_rate": 4.742123955304048e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 5219760,
      "step": 5835
    },
    {
      "epoch": 1.5413752144648276,
      "grad_norm": 0.37260809540748596,
      "learning_rate": 4.741689520424706e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 5224016,
      "step": 5840
    },
    {
      "epoch": 1.5426949980203246,
      "grad_norm": 0.1141723021864891,
      "learning_rate": 4.741254739848669e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 5228592,
      "step": 5845
    },
    {
      "epoch": 1.5440147815758216,
      "grad_norm": 0.08904385566711426,
      "learning_rate": 4.740819613642987e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 5232848,
      "step": 5850
    },
    {
      "epoch": 1.5453345651313186,
      "grad_norm": 0.055948033928871155,
      "learning_rate": 4.74038414187476e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 5237296,
      "step": 5855
    },
    {
      "epoch": 1.5466543486868154,
      "grad_norm": 0.3357619643211365,
      "learning_rate": 4.739948324611144e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 5241744,
      "step": 5860
    },
    {
      "epoch": 1.5479741322423122,
      "grad_norm": 0.26810574531555176,
      "learning_rate": 4.7395121619193465e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 5246352,
      "step": 5865
    },
    {
      "epoch": 1.5492939157978092,
      "grad_norm": 0.20834662020206451,
      "learning_rate": 4.7390756538666313e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 5250608,
      "step": 5870
    },
    {
      "epoch": 1.5506136993533062,
      "grad_norm": 0.26430609822273254,
      "learning_rate": 4.738638800520311e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 5254864,
      "step": 5875
    },
    {
      "epoch": 1.551933482908803,
      "grad_norm": 0.34068745374679565,
      "learning_rate": 4.738201601947757e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 5259024,
      "step": 5880
    },
    {
      "epoch": 1.5532532664642997,
      "grad_norm": 0.3501797914505005,
      "learning_rate": 4.7377640582163876e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 5263312,
      "step": 5885
    },
    {
      "epoch": 1.5545730500197967,
      "grad_norm": 0.20551666617393494,
      "learning_rate": 4.7373261693936786e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 5268016,
      "step": 5890
    },
    {
      "epoch": 1.5558928335752937,
      "grad_norm": 0.2281952202320099,
      "learning_rate": 4.7368879355471595e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 5272432,
      "step": 5895
    },
    {
      "epoch": 1.5572126171307905,
      "grad_norm": 0.15988720953464508,
      "learning_rate": 4.736449356744409e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 5276944,
      "step": 5900
    },
    {
      "epoch": 1.5585324006862873,
      "grad_norm": 0.292346715927124,
      "learning_rate": 4.736010433053064e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 5281360,
      "step": 5905
    },
    {
      "epoch": 1.5598521842417843,
      "grad_norm": 0.4188351035118103,
      "learning_rate": 4.73557116454081e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 5285776,
      "step": 5910
    },
    {
      "epoch": 1.5611719677972813,
      "grad_norm": 0.2938206195831299,
      "learning_rate": 4.735131551275389e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 5290224,
      "step": 5915
    },
    {
      "epoch": 1.5624917513527783,
      "grad_norm": 0.3782735764980316,
      "learning_rate": 4.734691593324594e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 5294448,
      "step": 5920
    },
    {
      "epoch": 1.563811534908275,
      "grad_norm": 0.25608259439468384,
      "learning_rate": 4.734251290756272e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 5298928,
      "step": 5925
    },
    {
      "epoch": 1.5651313184637718,
      "grad_norm": 0.16561755537986755,
      "learning_rate": 4.7338106436383246e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 5303280,
      "step": 5930
    },
    {
      "epoch": 1.5664511020192688,
      "grad_norm": 0.3275984227657318,
      "learning_rate": 4.733369652038703e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 5307888,
      "step": 5935
    },
    {
      "epoch": 1.5677708855747658,
      "grad_norm": 0.3632076382637024,
      "learning_rate": 4.7329283160254156e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 5312240,
      "step": 5940
    },
    {
      "epoch": 1.5690906691302626,
      "grad_norm": 0.24520328640937805,
      "learning_rate": 4.732486635666521e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 5317008,
      "step": 5945
    },
    {
      "epoch": 1.5704104526857594,
      "grad_norm": 0.5132139325141907,
      "learning_rate": 4.732044611030132e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 5321712,
      "step": 5950
    },
    {
      "epoch": 1.5717302362412564,
      "grad_norm": 0.11438081413507462,
      "learning_rate": 4.731602242184414e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 5325968,
      "step": 5955
    },
    {
      "epoch": 1.5730500197967534,
      "grad_norm": 0.1153215691447258,
      "learning_rate": 4.7311595291975864e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 5330480,
      "step": 5960
    },
    {
      "epoch": 1.5743698033522502,
      "grad_norm": 0.1654602736234665,
      "learning_rate": 4.7307164721379216e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 5335184,
      "step": 5965
    },
    {
      "epoch": 1.575689586907747,
      "grad_norm": 0.06799542903900146,
      "learning_rate": 4.730273071073743e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 5339600,
      "step": 5970
    },
    {
      "epoch": 1.577009370463244,
      "grad_norm": 0.17956531047821045,
      "learning_rate": 4.729829326073429e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 5344336,
      "step": 5975
    },
    {
      "epoch": 1.578329154018741,
      "grad_norm": 0.17375290393829346,
      "learning_rate": 4.7293852372054126e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 5348688,
      "step": 5980
    },
    {
      "epoch": 1.579648937574238,
      "grad_norm": 0.0627119392156601,
      "learning_rate": 4.728940804538176e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 5353200,
      "step": 5985
    },
    {
      "epoch": 1.5809687211297347,
      "grad_norm": 0.11077764630317688,
      "learning_rate": 4.7284960281402556e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 5357616,
      "step": 5990
    },
    {
      "epoch": 1.5822885046852315,
      "grad_norm": 0.6453601717948914,
      "learning_rate": 4.728050908080244e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 5362288,
      "step": 5995
    },
    {
      "epoch": 1.5836082882407285,
      "grad_norm": 0.21794354915618896,
      "learning_rate": 4.727605444426782e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 5366384,
      "step": 6000
    },
    {
      "epoch": 1.5836082882407285,
      "eval_loss": 0.07824458926916122,
      "eval_runtime": 64.7806,
      "eval_samples_per_second": 103.966,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 5366384,
      "step": 6000
    },
    {
      "epoch": 1.5849280717962255,
      "grad_norm": 0.29802024364471436,
      "learning_rate": 4.727159637248567e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 5370576,
      "step": 6005
    },
    {
      "epoch": 1.5862478553517223,
      "grad_norm": 0.21445690095424652,
      "learning_rate": 4.7267134866143474e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 5375216,
      "step": 6010
    },
    {
      "epoch": 1.587567638907219,
      "grad_norm": 0.2629227638244629,
      "learning_rate": 4.726266992592926e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 5379568,
      "step": 6015
    },
    {
      "epoch": 1.588887422462716,
      "grad_norm": 0.2240363210439682,
      "learning_rate": 4.725820155253157e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 5383984,
      "step": 6020
    },
    {
      "epoch": 1.590207206018213,
      "grad_norm": 0.41963574290275574,
      "learning_rate": 4.725372974663948e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 5388528,
      "step": 6025
    },
    {
      "epoch": 1.5915269895737099,
      "grad_norm": 0.3622005879878998,
      "learning_rate": 4.724925450894262e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 5392784,
      "step": 6030
    },
    {
      "epoch": 1.5928467731292069,
      "grad_norm": 0.14770478010177612,
      "learning_rate": 4.72447758401311e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 5397488,
      "step": 6035
    },
    {
      "epoch": 1.5941665566847036,
      "grad_norm": 0.41996148228645325,
      "learning_rate": 4.7240293740895616e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 5401904,
      "step": 6040
    },
    {
      "epoch": 1.5954863402402006,
      "grad_norm": 0.18501776456832886,
      "learning_rate": 4.723580821192733e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 5406416,
      "step": 6045
    },
    {
      "epoch": 1.5968061237956976,
      "grad_norm": 0.2960822284221649,
      "learning_rate": 4.7231319253917996e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 5410864,
      "step": 6050
    },
    {
      "epoch": 1.5981259073511944,
      "grad_norm": 0.2859703600406647,
      "learning_rate": 4.722682686755986e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 5415312,
      "step": 6055
    },
    {
      "epoch": 1.5994456909066912,
      "grad_norm": 0.23860345780849457,
      "learning_rate": 4.722233105354569e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 5419824,
      "step": 6060
    },
    {
      "epoch": 1.6007654744621882,
      "grad_norm": 0.11686433106660843,
      "learning_rate": 4.7217831812568815e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 5424208,
      "step": 6065
    },
    {
      "epoch": 1.6020852580176852,
      "grad_norm": 0.26035457849502563,
      "learning_rate": 4.721332914532307e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 5428816,
      "step": 6070
    },
    {
      "epoch": 1.603405041573182,
      "grad_norm": 0.15029872953891754,
      "learning_rate": 4.720882305250281e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 5433488,
      "step": 6075
    },
    {
      "epoch": 1.6047248251286788,
      "grad_norm": 0.06308599561452866,
      "learning_rate": 4.720431353480295e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 5437840,
      "step": 6080
    },
    {
      "epoch": 1.6060446086841758,
      "grad_norm": 0.27090874314308167,
      "learning_rate": 4.719980059291891e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 5442096,
      "step": 6085
    },
    {
      "epoch": 1.6073643922396728,
      "grad_norm": 0.45265886187553406,
      "learning_rate": 4.7195284227546634e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 5446384,
      "step": 6090
    },
    {
      "epoch": 1.6086841757951698,
      "grad_norm": 0.08372284471988678,
      "learning_rate": 4.7190764439382604e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 5450736,
      "step": 6095
    },
    {
      "epoch": 1.6100039593506665,
      "grad_norm": 0.046336691826581955,
      "learning_rate": 4.7186241229123826e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 5455248,
      "step": 6100
    },
    {
      "epoch": 1.6113237429061633,
      "grad_norm": 0.16240645945072174,
      "learning_rate": 4.718171459746785e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 5459440,
      "step": 6105
    },
    {
      "epoch": 1.6126435264616603,
      "grad_norm": 0.13885074853897095,
      "learning_rate": 4.717718454511273e-05,
      "loss": 0.098,
      "num_input_tokens_seen": 5464336,
      "step": 6110
    },
    {
      "epoch": 1.6139633100171573,
      "grad_norm": 0.10553596168756485,
      "learning_rate": 4.7172651072757056e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 5468656,
      "step": 6115
    },
    {
      "epoch": 1.615283093572654,
      "grad_norm": 0.12766800820827484,
      "learning_rate": 4.7168114181099945e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 5473136,
      "step": 6120
    },
    {
      "epoch": 1.6166028771281509,
      "grad_norm": 0.2556128203868866,
      "learning_rate": 4.716357387084105e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 5477168,
      "step": 6125
    },
    {
      "epoch": 1.6179226606836479,
      "grad_norm": 0.3456389904022217,
      "learning_rate": 4.715903014268054e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 5481776,
      "step": 6130
    },
    {
      "epoch": 1.6192424442391449,
      "grad_norm": 0.1864997148513794,
      "learning_rate": 4.715448299731911e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 5485968,
      "step": 6135
    },
    {
      "epoch": 1.6205622277946417,
      "grad_norm": 0.28685256838798523,
      "learning_rate": 4.7149932435457986e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 5490640,
      "step": 6140
    },
    {
      "epoch": 1.6218820113501384,
      "grad_norm": 0.3350217640399933,
      "learning_rate": 4.714537845779894e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 5495024,
      "step": 6145
    },
    {
      "epoch": 1.6232017949056354,
      "grad_norm": 0.2627733051776886,
      "learning_rate": 4.714082106504423e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 5499472,
      "step": 6150
    },
    {
      "epoch": 1.6245215784611324,
      "grad_norm": 0.07647845149040222,
      "learning_rate": 4.713626025789667e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 5504176,
      "step": 6155
    },
    {
      "epoch": 1.6258413620166294,
      "grad_norm": 0.1756701022386551,
      "learning_rate": 4.7131696037059606e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 5508656,
      "step": 6160
    },
    {
      "epoch": 1.6271611455721262,
      "grad_norm": 0.20765821635723114,
      "learning_rate": 4.712712840323689e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 5512976,
      "step": 6165
    },
    {
      "epoch": 1.628480929127623,
      "grad_norm": 0.19779954850673676,
      "learning_rate": 4.71225573571329e-05,
      "loss": 0.0706,
      "num_input_tokens_seen": 5517616,
      "step": 6170
    },
    {
      "epoch": 1.62980071268312,
      "grad_norm": 0.28386735916137695,
      "learning_rate": 4.711798289945256e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 5522192,
      "step": 6175
    },
    {
      "epoch": 1.631120496238617,
      "grad_norm": 0.3031769096851349,
      "learning_rate": 4.71134050309013e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 5526768,
      "step": 6180
    },
    {
      "epoch": 1.6324402797941138,
      "grad_norm": 0.30241701006889343,
      "learning_rate": 4.710882375218509e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 5531152,
      "step": 6185
    },
    {
      "epoch": 1.6337600633496105,
      "grad_norm": 0.08450118452310562,
      "learning_rate": 4.7104239064010424e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 5535312,
      "step": 6190
    },
    {
      "epoch": 1.6350798469051075,
      "grad_norm": 0.18649108707904816,
      "learning_rate": 4.709965096708432e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 5539760,
      "step": 6195
    },
    {
      "epoch": 1.6363996304606045,
      "grad_norm": 0.6834298372268677,
      "learning_rate": 4.709505946211431e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 5544176,
      "step": 6200
    },
    {
      "epoch": 1.6363996304606045,
      "eval_loss": 0.07735671103000641,
      "eval_runtime": 64.7177,
      "eval_samples_per_second": 104.067,
      "eval_steps_per_second": 26.021,
      "num_input_tokens_seen": 5544176,
      "step": 6200
    },
    {
      "epoch": 1.6377194140161013,
      "grad_norm": 0.10068556666374207,
      "learning_rate": 4.709046454980846e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 5548592,
      "step": 6205
    },
    {
      "epoch": 1.639039197571598,
      "grad_norm": 0.5610377788543701,
      "learning_rate": 4.708586623087538e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 5553072,
      "step": 6210
    },
    {
      "epoch": 1.640358981127095,
      "grad_norm": 0.07748523354530334,
      "learning_rate": 4.708126450602418e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 5557456,
      "step": 6215
    },
    {
      "epoch": 1.641678764682592,
      "grad_norm": 0.37537017464637756,
      "learning_rate": 4.7076659375964495e-05,
      "loss": 0.125,
      "num_input_tokens_seen": 5561712,
      "step": 6220
    },
    {
      "epoch": 1.642998548238089,
      "grad_norm": 0.6133891940116882,
      "learning_rate": 4.707205084140651e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 5566064,
      "step": 6225
    },
    {
      "epoch": 1.6443183317935859,
      "grad_norm": 0.14687713980674744,
      "learning_rate": 4.7067438903060904e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 5570704,
      "step": 6230
    },
    {
      "epoch": 1.6456381153490827,
      "grad_norm": 0.2580854892730713,
      "learning_rate": 4.70628235616389e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 5574928,
      "step": 6235
    },
    {
      "epoch": 1.6469578989045797,
      "grad_norm": 0.2071758210659027,
      "learning_rate": 4.7058204817852256e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 5579408,
      "step": 6240
    },
    {
      "epoch": 1.6482776824600767,
      "grad_norm": 0.323658287525177,
      "learning_rate": 4.705358267241322e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 5584176,
      "step": 6245
    },
    {
      "epoch": 1.6495974660155734,
      "grad_norm": 0.31353524327278137,
      "learning_rate": 4.704895712603459e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 5588688,
      "step": 6250
    },
    {
      "epoch": 1.6509172495710702,
      "grad_norm": 0.15379095077514648,
      "learning_rate": 4.704432817942969e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 5593264,
      "step": 6255
    },
    {
      "epoch": 1.6522370331265672,
      "grad_norm": 0.13582763075828552,
      "learning_rate": 4.703969583331236e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 5597840,
      "step": 6260
    },
    {
      "epoch": 1.6535568166820642,
      "grad_norm": 0.5205592513084412,
      "learning_rate": 4.7035060088396965e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 5602512,
      "step": 6265
    },
    {
      "epoch": 1.654876600237561,
      "grad_norm": 0.2897705137729645,
      "learning_rate": 4.703042094539839e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 5607088,
      "step": 6270
    },
    {
      "epoch": 1.6561963837930578,
      "grad_norm": 0.21647261083126068,
      "learning_rate": 4.702577840503206e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 5611600,
      "step": 6275
    },
    {
      "epoch": 1.6575161673485548,
      "grad_norm": 0.15820325911045074,
      "learning_rate": 4.70211324680139e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 5616304,
      "step": 6280
    },
    {
      "epoch": 1.6588359509040518,
      "grad_norm": 0.14594845473766327,
      "learning_rate": 4.7016483135060386e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 5620688,
      "step": 6285
    },
    {
      "epoch": 1.6601557344595488,
      "grad_norm": 0.21176597476005554,
      "learning_rate": 4.701183040688849e-05,
      "loss": 0.1025,
      "num_input_tokens_seen": 5625520,
      "step": 6290
    },
    {
      "epoch": 1.6614755180150456,
      "grad_norm": 0.24600651860237122,
      "learning_rate": 4.700717428421573e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 5630032,
      "step": 6295
    },
    {
      "epoch": 1.6627953015705423,
      "grad_norm": 0.12306217849254608,
      "learning_rate": 4.700251476776014e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 5634448,
      "step": 6300
    },
    {
      "epoch": 1.6641150851260393,
      "grad_norm": 0.29385313391685486,
      "learning_rate": 4.699785185824026e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 5639024,
      "step": 6305
    },
    {
      "epoch": 1.6654348686815363,
      "grad_norm": 0.1173398345708847,
      "learning_rate": 4.699318555637519e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 5643504,
      "step": 6310
    },
    {
      "epoch": 1.6667546522370331,
      "grad_norm": 0.0948636531829834,
      "learning_rate": 4.6988515862884525e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 5648048,
      "step": 6315
    },
    {
      "epoch": 1.66807443579253,
      "grad_norm": 0.28425514698028564,
      "learning_rate": 4.698384277848838e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 5652368,
      "step": 6320
    },
    {
      "epoch": 1.669394219348027,
      "grad_norm": 0.17787136137485504,
      "learning_rate": 4.6979166303907425e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 5656880,
      "step": 6325
    },
    {
      "epoch": 1.670714002903524,
      "grad_norm": 0.4983280599117279,
      "learning_rate": 4.697448643986281e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 5661392,
      "step": 6330
    },
    {
      "epoch": 1.6720337864590207,
      "grad_norm": 0.07861651480197906,
      "learning_rate": 4.696980318707624e-05,
      "loss": 0.0259,
      "num_input_tokens_seen": 5665808,
      "step": 6335
    },
    {
      "epoch": 1.6733535700145175,
      "grad_norm": 0.23327556252479553,
      "learning_rate": 4.6965116546269924e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 5670384,
      "step": 6340
    },
    {
      "epoch": 1.6746733535700145,
      "grad_norm": 0.3305080533027649,
      "learning_rate": 4.6960426518166615e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 5674736,
      "step": 6345
    },
    {
      "epoch": 1.6759931371255115,
      "grad_norm": 0.13714560866355896,
      "learning_rate": 4.6955733103489556e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 5679088,
      "step": 6350
    },
    {
      "epoch": 1.6773129206810085,
      "grad_norm": 0.3895653486251831,
      "learning_rate": 4.695103630296255e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 5683536,
      "step": 6355
    },
    {
      "epoch": 1.6786327042365052,
      "grad_norm": 0.34837237000465393,
      "learning_rate": 4.694633611730988e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 5687856,
      "step": 6360
    },
    {
      "epoch": 1.679952487792002,
      "grad_norm": 0.23766452074050903,
      "learning_rate": 4.694163254725639e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 5692496,
      "step": 6365
    },
    {
      "epoch": 1.681272271347499,
      "grad_norm": 0.08897238224744797,
      "learning_rate": 4.693692559352743e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 5696688,
      "step": 6370
    },
    {
      "epoch": 1.682592054902996,
      "grad_norm": 0.17052972316741943,
      "learning_rate": 4.693221525684886e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 5701136,
      "step": 6375
    },
    {
      "epoch": 1.6839118384584928,
      "grad_norm": 0.08918283879756927,
      "learning_rate": 4.6927501537947084e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 5705328,
      "step": 6380
    },
    {
      "epoch": 1.6852316220139896,
      "grad_norm": 0.06502197682857513,
      "learning_rate": 4.692278443754901e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 5709680,
      "step": 6385
    },
    {
      "epoch": 1.6865514055694866,
      "grad_norm": 0.17533433437347412,
      "learning_rate": 4.691806395638208e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 5714128,
      "step": 6390
    },
    {
      "epoch": 1.6878711891249836,
      "grad_norm": 0.13499194383621216,
      "learning_rate": 4.6913340095174255e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 5718640,
      "step": 6395
    },
    {
      "epoch": 1.6891909726804804,
      "grad_norm": 0.37257638573646545,
      "learning_rate": 4.690861285465399e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 5723216,
      "step": 6400
    },
    {
      "epoch": 1.6891909726804804,
      "eval_loss": 0.07763779163360596,
      "eval_runtime": 64.718,
      "eval_samples_per_second": 104.067,
      "eval_steps_per_second": 26.021,
      "num_input_tokens_seen": 5723216,
      "step": 6400
    },
    {
      "epoch": 1.6905107562359774,
      "grad_norm": 0.45880937576293945,
      "learning_rate": 4.690388223555031e-05,
      "loss": 0.1096,
      "num_input_tokens_seen": 5727984,
      "step": 6405
    },
    {
      "epoch": 1.6918305397914741,
      "grad_norm": 0.40928998589515686,
      "learning_rate": 4.689914823859273e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 5732816,
      "step": 6410
    },
    {
      "epoch": 1.6931503233469711,
      "grad_norm": 0.04590551182627678,
      "learning_rate": 4.689441086451129e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 5737424,
      "step": 6415
    },
    {
      "epoch": 1.6944701069024681,
      "grad_norm": 0.15273034572601318,
      "learning_rate": 4.688967011403655e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 5741616,
      "step": 6420
    },
    {
      "epoch": 1.695789890457965,
      "grad_norm": 0.12678445875644684,
      "learning_rate": 4.68849259878996e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 5746448,
      "step": 6425
    },
    {
      "epoch": 1.6971096740134617,
      "grad_norm": 0.18746735155582428,
      "learning_rate": 4.6880178486832036e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 5750768,
      "step": 6430
    },
    {
      "epoch": 1.6984294575689587,
      "grad_norm": 0.09261399507522583,
      "learning_rate": 4.687542761156598e-05,
      "loss": 0.0241,
      "num_input_tokens_seen": 5755088,
      "step": 6435
    },
    {
      "epoch": 1.6997492411244557,
      "grad_norm": 0.21981015801429749,
      "learning_rate": 4.6870673362834096e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 5759408,
      "step": 6440
    },
    {
      "epoch": 1.7010690246799525,
      "grad_norm": 0.21868005394935608,
      "learning_rate": 4.6865915741369526e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 5763824,
      "step": 6445
    },
    {
      "epoch": 1.7023888082354492,
      "grad_norm": 0.16330015659332275,
      "learning_rate": 4.686115474790597e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 5768240,
      "step": 6450
    },
    {
      "epoch": 1.7037085917909462,
      "grad_norm": 0.09738574922084808,
      "learning_rate": 4.685639038317762e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 5772784,
      "step": 6455
    },
    {
      "epoch": 1.7050283753464432,
      "grad_norm": 0.04816265404224396,
      "learning_rate": 4.685162264791921e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 5777200,
      "step": 6460
    },
    {
      "epoch": 1.70634815890194,
      "grad_norm": 0.037401869893074036,
      "learning_rate": 4.684685154286599e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 5781968,
      "step": 6465
    },
    {
      "epoch": 1.707667942457437,
      "grad_norm": 0.407725989818573,
      "learning_rate": 4.684207706875371e-05,
      "loss": 0.1491,
      "num_input_tokens_seen": 5786704,
      "step": 6470
    },
    {
      "epoch": 1.7089877260129338,
      "grad_norm": 0.5893194675445557,
      "learning_rate": 4.683729922631866e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 5791216,
      "step": 6475
    },
    {
      "epoch": 1.7103075095684308,
      "grad_norm": 0.13526876270771027,
      "learning_rate": 4.683251801629765e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 5795376,
      "step": 6480
    },
    {
      "epoch": 1.7116272931239278,
      "grad_norm": 0.14884258806705475,
      "learning_rate": 4.6827733439428e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 5799664,
      "step": 6485
    },
    {
      "epoch": 1.7129470766794246,
      "grad_norm": 0.3277275860309601,
      "learning_rate": 4.682294549644754e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 5804080,
      "step": 6490
    },
    {
      "epoch": 1.7142668602349214,
      "grad_norm": 0.27445316314697266,
      "learning_rate": 4.681815418809464e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 5808592,
      "step": 6495
    },
    {
      "epoch": 1.7155866437904184,
      "grad_norm": 0.37290483713150024,
      "learning_rate": 4.681335951510819e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 5813072,
      "step": 6500
    },
    {
      "epoch": 1.7169064273459154,
      "grad_norm": 0.35742345452308655,
      "learning_rate": 4.6808561478227576e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 5817648,
      "step": 6505
    },
    {
      "epoch": 1.7182262109014121,
      "grad_norm": 0.18662406504154205,
      "learning_rate": 4.680376007819271e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 5822320,
      "step": 6510
    },
    {
      "epoch": 1.719545994456909,
      "grad_norm": 0.1383635401725769,
      "learning_rate": 4.679895531574405e-05,
      "loss": 0.0191,
      "num_input_tokens_seen": 5826512,
      "step": 6515
    },
    {
      "epoch": 1.720865778012406,
      "grad_norm": 0.19470007717609406,
      "learning_rate": 4.679414719162253e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 5831088,
      "step": 6520
    },
    {
      "epoch": 1.722185561567903,
      "grad_norm": 0.15924309194087982,
      "learning_rate": 4.6789335706569635e-05,
      "loss": 0.1296,
      "num_input_tokens_seen": 5835344,
      "step": 6525
    },
    {
      "epoch": 1.7235053451234,
      "grad_norm": 0.1498899608850479,
      "learning_rate": 4.678452086132734e-05,
      "loss": 0.1469,
      "num_input_tokens_seen": 5839728,
      "step": 6530
    },
    {
      "epoch": 1.7248251286788967,
      "grad_norm": 0.1980186104774475,
      "learning_rate": 4.677970265663818e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 5844176,
      "step": 6535
    },
    {
      "epoch": 1.7261449122343935,
      "grad_norm": 0.08850986510515213,
      "learning_rate": 4.677488109324517e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 5848880,
      "step": 6540
    },
    {
      "epoch": 1.7274646957898905,
      "grad_norm": 0.5354941487312317,
      "learning_rate": 4.6770056171891846e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 5853136,
      "step": 6545
    },
    {
      "epoch": 1.7287844793453875,
      "grad_norm": 0.17719906568527222,
      "learning_rate": 4.6765227893322286e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 5857392,
      "step": 6550
    },
    {
      "epoch": 1.7301042629008843,
      "grad_norm": 0.12624286115169525,
      "learning_rate": 4.676039625828107e-05,
      "loss": 0.0788,
      "num_input_tokens_seen": 5862000,
      "step": 6555
    },
    {
      "epoch": 1.731424046456381,
      "grad_norm": 0.07651223987340927,
      "learning_rate": 4.675556126751328e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 5866672,
      "step": 6560
    },
    {
      "epoch": 1.732743830011878,
      "grad_norm": 0.35980457067489624,
      "learning_rate": 4.6750722921764556e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 5871152,
      "step": 6565
    },
    {
      "epoch": 1.734063613567375,
      "grad_norm": 0.19881732761859894,
      "learning_rate": 4.674588122178102e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 5875536,
      "step": 6570
    },
    {
      "epoch": 1.7353833971228718,
      "grad_norm": 0.10140056908130646,
      "learning_rate": 4.674103616830931e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 5880048,
      "step": 6575
    },
    {
      "epoch": 1.7367031806783686,
      "grad_norm": 0.3566274046897888,
      "learning_rate": 4.673618776209663e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 5884752,
      "step": 6580
    },
    {
      "epoch": 1.7380229642338656,
      "grad_norm": 0.3141302466392517,
      "learning_rate": 4.673133600389063e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 5889008,
      "step": 6585
    },
    {
      "epoch": 1.7393427477893626,
      "grad_norm": 0.1597874015569687,
      "learning_rate": 4.672648089443953e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 5893456,
      "step": 6590
    },
    {
      "epoch": 1.7406625313448596,
      "grad_norm": 0.23406513035297394,
      "learning_rate": 4.672162243449204e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 5898288,
      "step": 6595
    },
    {
      "epoch": 1.7419823149003564,
      "grad_norm": 0.2751595079898834,
      "learning_rate": 4.67167606247974e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 5902896,
      "step": 6600
    },
    {
      "epoch": 1.7419823149003564,
      "eval_loss": 0.07670366764068604,
      "eval_runtime": 64.778,
      "eval_samples_per_second": 103.97,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 5902896,
      "step": 6600
    },
    {
      "epoch": 1.7433020984558532,
      "grad_norm": 0.06079903244972229,
      "learning_rate": 4.671189546610536e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 5907536,
      "step": 6605
    },
    {
      "epoch": 1.7446218820113502,
      "grad_norm": 0.28400281071662903,
      "learning_rate": 4.67070269591662e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 5912048,
      "step": 6610
    },
    {
      "epoch": 1.7459416655668472,
      "grad_norm": 0.1595672369003296,
      "learning_rate": 4.670215510473068e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 5916496,
      "step": 6615
    },
    {
      "epoch": 1.747261449122344,
      "grad_norm": 0.10201507806777954,
      "learning_rate": 4.669727990355013e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 5920784,
      "step": 6620
    },
    {
      "epoch": 1.7485812326778407,
      "grad_norm": 0.2828303277492523,
      "learning_rate": 4.669240135637635e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 5925680,
      "step": 6625
    },
    {
      "epoch": 1.7499010162333377,
      "grad_norm": 0.06887925416231155,
      "learning_rate": 4.6687519463961675e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 5930064,
      "step": 6630
    },
    {
      "epoch": 1.7512207997888347,
      "grad_norm": 0.22175997495651245,
      "learning_rate": 4.668263422705896e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 5934352,
      "step": 6635
    },
    {
      "epoch": 1.7525405833443315,
      "grad_norm": 0.05469990894198418,
      "learning_rate": 4.667774564642156e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 5938928,
      "step": 6640
    },
    {
      "epoch": 1.7538603668998283,
      "grad_norm": 0.06107835844159126,
      "learning_rate": 4.6672853722803365e-05,
      "loss": 0.0536,
      "num_input_tokens_seen": 5943760,
      "step": 6645
    },
    {
      "epoch": 1.7551801504553253,
      "grad_norm": 0.20193369686603546,
      "learning_rate": 4.666795845695877e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 5948496,
      "step": 6650
    },
    {
      "epoch": 1.7564999340108223,
      "grad_norm": 0.18798863887786865,
      "learning_rate": 4.666305984964269e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 5952624,
      "step": 6655
    },
    {
      "epoch": 1.7578197175663193,
      "grad_norm": 0.22653409838676453,
      "learning_rate": 4.6658157901610535e-05,
      "loss": 0.0441,
      "num_input_tokens_seen": 5957040,
      "step": 6660
    },
    {
      "epoch": 1.759139501121816,
      "grad_norm": 0.09003376215696335,
      "learning_rate": 4.665325261361826e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 5961328,
      "step": 6665
    },
    {
      "epoch": 1.7604592846773128,
      "grad_norm": 0.3066350817680359,
      "learning_rate": 4.664834398642232e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 5966032,
      "step": 6670
    },
    {
      "epoch": 1.7617790682328098,
      "grad_norm": 0.9687243103981018,
      "learning_rate": 4.6643432020779686e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 5970000,
      "step": 6675
    },
    {
      "epoch": 1.7630988517883068,
      "grad_norm": 0.4614083468914032,
      "learning_rate": 4.663851671744786e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 5974256,
      "step": 6680
    },
    {
      "epoch": 1.7644186353438036,
      "grad_norm": 0.034824226051568985,
      "learning_rate": 4.6633598077184815e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 5978384,
      "step": 6685
    },
    {
      "epoch": 1.7657384188993004,
      "grad_norm": 0.09816955029964447,
      "learning_rate": 4.662867610074908e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 5982896,
      "step": 6690
    },
    {
      "epoch": 1.7670582024547974,
      "grad_norm": 0.037977058440446854,
      "learning_rate": 4.6623750788899696e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 5987152,
      "step": 6695
    },
    {
      "epoch": 1.7683779860102944,
      "grad_norm": 0.14718440175056458,
      "learning_rate": 4.6618822142396195e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 5991632,
      "step": 6700
    },
    {
      "epoch": 1.7696977695657912,
      "grad_norm": 0.2078886479139328,
      "learning_rate": 4.661389016199864e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 5996208,
      "step": 6705
    },
    {
      "epoch": 1.771017553121288,
      "grad_norm": 0.3250770568847656,
      "learning_rate": 4.660895484846761e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 6000912,
      "step": 6710
    },
    {
      "epoch": 1.772337336676785,
      "grad_norm": 0.23032908141613007,
      "learning_rate": 4.660401620256418e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 6004848,
      "step": 6715
    },
    {
      "epoch": 1.773657120232282,
      "grad_norm": 0.23126226663589478,
      "learning_rate": 4.659907422504997e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 6009584,
      "step": 6720
    },
    {
      "epoch": 1.774976903787779,
      "grad_norm": 0.08596619963645935,
      "learning_rate": 4.6594128916687074e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 6013936,
      "step": 6725
    },
    {
      "epoch": 1.7762966873432757,
      "grad_norm": 0.08057559281587601,
      "learning_rate": 4.658918027823813e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 6018160,
      "step": 6730
    },
    {
      "epoch": 1.7776164708987725,
      "grad_norm": 0.04312612861394882,
      "learning_rate": 4.658422831046628e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 6022896,
      "step": 6735
    },
    {
      "epoch": 1.7789362544542695,
      "grad_norm": 0.2413734644651413,
      "learning_rate": 4.657927301413518e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 6027536,
      "step": 6740
    },
    {
      "epoch": 1.7802560380097665,
      "grad_norm": 0.05529560148715973,
      "learning_rate": 4.657431439000901e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 6031952,
      "step": 6745
    },
    {
      "epoch": 1.7815758215652633,
      "grad_norm": 0.0844552218914032,
      "learning_rate": 4.656935243885243e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 6036656,
      "step": 6750
    },
    {
      "epoch": 1.78289560512076,
      "grad_norm": 0.28549104928970337,
      "learning_rate": 4.656438716143066e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 6041136,
      "step": 6755
    },
    {
      "epoch": 1.784215388676257,
      "grad_norm": 0.15741398930549622,
      "learning_rate": 4.6559418558509384e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 6045872,
      "step": 6760
    },
    {
      "epoch": 1.785535172231754,
      "grad_norm": 0.28424903750419617,
      "learning_rate": 4.6554446630854833e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 6050384,
      "step": 6765
    },
    {
      "epoch": 1.7868549557872508,
      "grad_norm": 0.09712236374616623,
      "learning_rate": 4.654947137923374e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 6054896,
      "step": 6770
    },
    {
      "epoch": 1.7881747393427478,
      "grad_norm": 0.40514037013053894,
      "learning_rate": 4.654449280441335e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 6059440,
      "step": 6775
    },
    {
      "epoch": 1.7894945228982446,
      "grad_norm": 0.5149907469749451,
      "learning_rate": 4.653951090716143e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 6063664,
      "step": 6780
    },
    {
      "epoch": 1.7908143064537416,
      "grad_norm": 0.21506710350513458,
      "learning_rate": 4.653452568824625e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 6068208,
      "step": 6785
    },
    {
      "epoch": 1.7921340900092386,
      "grad_norm": 0.4691607356071472,
      "learning_rate": 4.6529537148436585e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 6072656,
      "step": 6790
    },
    {
      "epoch": 1.7934538735647354,
      "grad_norm": 0.05693565309047699,
      "learning_rate": 4.6524545288501734e-05,
      "loss": 0.1345,
      "num_input_tokens_seen": 6076784,
      "step": 6795
    },
    {
      "epoch": 1.7947736571202322,
      "grad_norm": 0.3669026792049408,
      "learning_rate": 4.6519550109211506e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 6081040,
      "step": 6800
    },
    {
      "epoch": 1.7947736571202322,
      "eval_loss": 0.07645101845264435,
      "eval_runtime": 64.7837,
      "eval_samples_per_second": 103.961,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 6081040,
      "step": 6800
    },
    {
      "epoch": 1.7960934406757292,
      "grad_norm": 0.2074628472328186,
      "learning_rate": 4.651455161133622e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 6085680,
      "step": 6805
    },
    {
      "epoch": 1.7974132242312262,
      "grad_norm": 0.1918603479862213,
      "learning_rate": 4.6509549795646704e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 6090320,
      "step": 6810
    },
    {
      "epoch": 1.798733007786723,
      "grad_norm": 0.23296843469142914,
      "learning_rate": 4.6504544662914306e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 6094608,
      "step": 6815
    },
    {
      "epoch": 1.8000527913422197,
      "grad_norm": 0.15890564024448395,
      "learning_rate": 4.6499536213910876e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 6099280,
      "step": 6820
    },
    {
      "epoch": 1.8013725748977167,
      "grad_norm": 0.24076932668685913,
      "learning_rate": 4.6494524449408786e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 6103888,
      "step": 6825
    },
    {
      "epoch": 1.8026923584532137,
      "grad_norm": 0.1924286186695099,
      "learning_rate": 4.6489509370180903e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 6108176,
      "step": 6830
    },
    {
      "epoch": 1.8040121420087105,
      "grad_norm": 0.14083442091941833,
      "learning_rate": 4.648449097700063e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 6112560,
      "step": 6835
    },
    {
      "epoch": 1.8053319255642075,
      "grad_norm": 0.0990385189652443,
      "learning_rate": 4.647946927064185e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 6117008,
      "step": 6840
    },
    {
      "epoch": 1.8066517091197043,
      "grad_norm": 0.2637682557106018,
      "learning_rate": 4.647444425187898e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 6121264,
      "step": 6845
    },
    {
      "epoch": 1.8079714926752013,
      "grad_norm": 0.5847938656806946,
      "learning_rate": 4.646941592148695e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 6125744,
      "step": 6850
    },
    {
      "epoch": 1.8092912762306983,
      "grad_norm": 0.3137251138687134,
      "learning_rate": 4.646438428024117e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 6130480,
      "step": 6855
    },
    {
      "epoch": 1.810611059786195,
      "grad_norm": 0.24401280283927917,
      "learning_rate": 4.64593493289176e-05,
      "loss": 0.0263,
      "num_input_tokens_seen": 6134928,
      "step": 6860
    },
    {
      "epoch": 1.8119308433416919,
      "grad_norm": 0.21498775482177734,
      "learning_rate": 4.64543110682927e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 6139504,
      "step": 6865
    },
    {
      "epoch": 1.8132506268971889,
      "grad_norm": 0.47805729508399963,
      "learning_rate": 4.644926949914341e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 6143920,
      "step": 6870
    },
    {
      "epoch": 1.8145704104526859,
      "grad_norm": 0.11170867830514908,
      "learning_rate": 4.644422462224722e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 6148400,
      "step": 6875
    },
    {
      "epoch": 1.8158901940081826,
      "grad_norm": 0.11510006338357925,
      "learning_rate": 4.643917643838211e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 6152720,
      "step": 6880
    },
    {
      "epoch": 1.8172099775636794,
      "grad_norm": 0.27835461497306824,
      "learning_rate": 4.6434124948326564e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 6157424,
      "step": 6885
    },
    {
      "epoch": 1.8185297611191764,
      "grad_norm": 0.36345091462135315,
      "learning_rate": 4.6429070152859594e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 6161840,
      "step": 6890
    },
    {
      "epoch": 1.8198495446746734,
      "grad_norm": 0.31958088278770447,
      "learning_rate": 4.6424012052760714e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 6166352,
      "step": 6895
    },
    {
      "epoch": 1.8211693282301704,
      "grad_norm": 0.061690185219049454,
      "learning_rate": 4.6418950648809945e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 6170640,
      "step": 6900
    },
    {
      "epoch": 1.8224891117856672,
      "grad_norm": 0.15668371319770813,
      "learning_rate": 4.641388594178782e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 6175056,
      "step": 6905
    },
    {
      "epoch": 1.823808895341164,
      "grad_norm": 0.10498514026403427,
      "learning_rate": 4.640881793247538e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 6179472,
      "step": 6910
    },
    {
      "epoch": 1.825128678896661,
      "grad_norm": 0.32304131984710693,
      "learning_rate": 4.6403746621654173e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 6183664,
      "step": 6915
    },
    {
      "epoch": 1.826448462452158,
      "grad_norm": 0.06814830005168915,
      "learning_rate": 4.639867201010626e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 6188272,
      "step": 6920
    },
    {
      "epoch": 1.8277682460076548,
      "grad_norm": 0.4275568723678589,
      "learning_rate": 4.6393594098614204e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 6192336,
      "step": 6925
    },
    {
      "epoch": 1.8290880295631515,
      "grad_norm": 0.39397117495536804,
      "learning_rate": 4.63885128879611e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 6196624,
      "step": 6930
    },
    {
      "epoch": 1.8304078131186485,
      "grad_norm": 0.4087122976779938,
      "learning_rate": 4.638342837893052e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 6201424,
      "step": 6935
    },
    {
      "epoch": 1.8317275966741455,
      "grad_norm": 0.03938461095094681,
      "learning_rate": 4.6378340572306565e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 6205776,
      "step": 6940
    },
    {
      "epoch": 1.8330473802296423,
      "grad_norm": 0.17708829045295715,
      "learning_rate": 4.6373249468873833e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 6210416,
      "step": 6945
    },
    {
      "epoch": 1.834367163785139,
      "grad_norm": 0.23982881009578705,
      "learning_rate": 4.636815506941744e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 6214736,
      "step": 6950
    },
    {
      "epoch": 1.835686947340636,
      "grad_norm": 0.22591456770896912,
      "learning_rate": 4.6363057374723004e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 6219088,
      "step": 6955
    },
    {
      "epoch": 1.837006730896133,
      "grad_norm": 0.06749723851680756,
      "learning_rate": 4.635795638557666e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 6223568,
      "step": 6960
    },
    {
      "epoch": 1.83832651445163,
      "grad_norm": 0.039544425904750824,
      "learning_rate": 4.635285210276504e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 6227920,
      "step": 6965
    },
    {
      "epoch": 1.8396462980071269,
      "grad_norm": 0.20393970608711243,
      "learning_rate": 4.6347744527075295e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 6232144,
      "step": 6970
    },
    {
      "epoch": 1.8409660815626236,
      "grad_norm": 0.1472388356924057,
      "learning_rate": 4.634263365929506e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 6236624,
      "step": 6975
    },
    {
      "epoch": 1.8422858651181206,
      "grad_norm": 0.11189628392457962,
      "learning_rate": 4.6337519500212515e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 6241328,
      "step": 6980
    },
    {
      "epoch": 1.8436056486736176,
      "grad_norm": 0.25500476360321045,
      "learning_rate": 4.633240205061632e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 6245744,
      "step": 6985
    },
    {
      "epoch": 1.8449254322291144,
      "grad_norm": 0.1959000676870346,
      "learning_rate": 4.632728131129565e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 6250064,
      "step": 6990
    },
    {
      "epoch": 1.8462452157846112,
      "grad_norm": 0.12271429598331451,
      "learning_rate": 4.632215728304018e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 6254544,
      "step": 6995
    },
    {
      "epoch": 1.8475649993401082,
      "grad_norm": 0.38256895542144775,
      "learning_rate": 4.63170299666401e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 6259056,
      "step": 7000
    },
    {
      "epoch": 1.8475649993401082,
      "eval_loss": 0.0762862041592598,
      "eval_runtime": 64.7861,
      "eval_samples_per_second": 103.958,
      "eval_steps_per_second": 25.993,
      "num_input_tokens_seen": 6259056,
      "step": 7000
    },
    {
      "epoch": 1.8488847828956052,
      "grad_norm": 0.0857502892613411,
      "learning_rate": 4.631189936288612e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 6263632,
      "step": 7005
    },
    {
      "epoch": 1.850204566451102,
      "grad_norm": 0.13003812730312347,
      "learning_rate": 4.630676547256944e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 6268304,
      "step": 7010
    },
    {
      "epoch": 1.8515243500065988,
      "grad_norm": 0.014970618300139904,
      "learning_rate": 4.630162829648176e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 6272816,
      "step": 7015
    },
    {
      "epoch": 1.8528441335620958,
      "grad_norm": 0.13892458379268646,
      "learning_rate": 4.629648783541531e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 6277264,
      "step": 7020
    },
    {
      "epoch": 1.8541639171175928,
      "grad_norm": 0.1330774426460266,
      "learning_rate": 4.6291344090162804e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 6281552,
      "step": 7025
    },
    {
      "epoch": 1.8554837006730898,
      "grad_norm": 0.6068649291992188,
      "learning_rate": 4.628619706151748e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 6285712,
      "step": 7030
    },
    {
      "epoch": 1.8568034842285865,
      "grad_norm": 0.14965306222438812,
      "learning_rate": 4.628104675027306e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 6289840,
      "step": 7035
    },
    {
      "epoch": 1.8581232677840833,
      "grad_norm": 0.30804044008255005,
      "learning_rate": 4.6275893157223805e-05,
      "loss": 0.1643,
      "num_input_tokens_seen": 6294416,
      "step": 7040
    },
    {
      "epoch": 1.8594430513395803,
      "grad_norm": 0.11024698615074158,
      "learning_rate": 4.627073628316445e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 6298928,
      "step": 7045
    },
    {
      "epoch": 1.8607628348950773,
      "grad_norm": 0.3184722661972046,
      "learning_rate": 4.626557612889026e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 6303568,
      "step": 7050
    },
    {
      "epoch": 1.862082618450574,
      "grad_norm": 0.22575193643569946,
      "learning_rate": 4.626041269519699e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 6308048,
      "step": 7055
    },
    {
      "epoch": 1.8634024020060709,
      "grad_norm": 0.12449900805950165,
      "learning_rate": 4.6255245982880905e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 6312720,
      "step": 7060
    },
    {
      "epoch": 1.8647221855615679,
      "grad_norm": 0.2679610848426819,
      "learning_rate": 4.625007599273879e-05,
      "loss": 0.1217,
      "num_input_tokens_seen": 6317296,
      "step": 7065
    },
    {
      "epoch": 1.8660419691170649,
      "grad_norm": 0.37533795833587646,
      "learning_rate": 4.6244902725567895e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 6321936,
      "step": 7070
    },
    {
      "epoch": 1.8673617526725617,
      "grad_norm": 0.22294484078884125,
      "learning_rate": 4.6239726182166024e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 6326640,
      "step": 7075
    },
    {
      "epoch": 1.8686815362280584,
      "grad_norm": 0.2395455688238144,
      "learning_rate": 4.623454636333147e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 6331344,
      "step": 7080
    },
    {
      "epoch": 1.8700013197835554,
      "grad_norm": 0.3017948567867279,
      "learning_rate": 4.622936326986301e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 6335600,
      "step": 7085
    },
    {
      "epoch": 1.8713211033390524,
      "grad_norm": 0.1621217131614685,
      "learning_rate": 4.6224176902559946e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 6339888,
      "step": 7090
    },
    {
      "epoch": 1.8726408868945494,
      "grad_norm": 0.14675092697143555,
      "learning_rate": 4.621898726222209e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 6344400,
      "step": 7095
    },
    {
      "epoch": 1.8739606704500462,
      "grad_norm": 0.33293139934539795,
      "learning_rate": 4.6213794349649744e-05,
      "loss": 0.1457,
      "num_input_tokens_seen": 6349008,
      "step": 7100
    },
    {
      "epoch": 1.875280454005543,
      "grad_norm": 0.4207361042499542,
      "learning_rate": 4.6208598165643715e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 6353072,
      "step": 7105
    },
    {
      "epoch": 1.87660023756104,
      "grad_norm": 0.28801441192626953,
      "learning_rate": 4.620339871100533e-05,
      "loss": 0.11,
      "num_input_tokens_seen": 6357968,
      "step": 7110
    },
    {
      "epoch": 1.877920021116537,
      "grad_norm": 0.21969226002693176,
      "learning_rate": 4.6198195986536394e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 6362736,
      "step": 7115
    },
    {
      "epoch": 1.8792398046720338,
      "grad_norm": 0.10012209415435791,
      "learning_rate": 4.619298999303926e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 6367216,
      "step": 7120
    },
    {
      "epoch": 1.8805595882275306,
      "grad_norm": 0.1143481582403183,
      "learning_rate": 4.618778073131673e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 6371696,
      "step": 7125
    },
    {
      "epoch": 1.8818793717830276,
      "grad_norm": 0.03170226514339447,
      "learning_rate": 4.618256820217215e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 6375888,
      "step": 7130
    },
    {
      "epoch": 1.8831991553385246,
      "grad_norm": 0.07537917792797089,
      "learning_rate": 4.617735240640936e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 6380176,
      "step": 7135
    },
    {
      "epoch": 1.8845189388940213,
      "grad_norm": 0.06016154587268829,
      "learning_rate": 4.6172133344832705e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 6384880,
      "step": 7140
    },
    {
      "epoch": 1.8858387224495183,
      "grad_norm": 0.17361405491828918,
      "learning_rate": 4.6166911018247004e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 6389296,
      "step": 7145
    },
    {
      "epoch": 1.8871585060050151,
      "grad_norm": 0.10163392871618271,
      "learning_rate": 4.616168542745764e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 6393712,
      "step": 7150
    },
    {
      "epoch": 1.8884782895605121,
      "grad_norm": 0.09090641140937805,
      "learning_rate": 4.6156456573270446e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 6398000,
      "step": 7155
    },
    {
      "epoch": 1.8897980731160091,
      "grad_norm": 0.12246877700090408,
      "learning_rate": 4.615122445649177e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 6402640,
      "step": 7160
    },
    {
      "epoch": 1.891117856671506,
      "grad_norm": 0.051915451884269714,
      "learning_rate": 4.6145989077928486e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 6407184,
      "step": 7165
    },
    {
      "epoch": 1.8924376402270027,
      "grad_norm": 0.08832618594169617,
      "learning_rate": 4.6140750438387953e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 6411568,
      "step": 7170
    },
    {
      "epoch": 1.8937574237824997,
      "grad_norm": 0.2811886966228485,
      "learning_rate": 4.613550853867803e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 6415760,
      "step": 7175
    },
    {
      "epoch": 1.8950772073379967,
      "grad_norm": 0.030251650139689445,
      "learning_rate": 4.613026337960708e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 6419952,
      "step": 7180
    },
    {
      "epoch": 1.8963969908934935,
      "grad_norm": 0.2344350516796112,
      "learning_rate": 4.612501496198398e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 6424592,
      "step": 7185
    },
    {
      "epoch": 1.8977167744489902,
      "grad_norm": 0.08863913267850876,
      "learning_rate": 4.61197632866181e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 6429040,
      "step": 7190
    },
    {
      "epoch": 1.8990365580044872,
      "grad_norm": 0.19203996658325195,
      "learning_rate": 4.611450835431931e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 6433552,
      "step": 7195
    },
    {
      "epoch": 1.9003563415599842,
      "grad_norm": 0.17268742620944977,
      "learning_rate": 4.6109250165898e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 6437904,
      "step": 7200
    },
    {
      "epoch": 1.9003563415599842,
      "eval_loss": 0.07598280161619186,
      "eval_runtime": 64.5813,
      "eval_samples_per_second": 104.287,
      "eval_steps_per_second": 26.076,
      "num_input_tokens_seen": 6437904,
      "step": 7200
    },
    {
      "epoch": 1.901676125115481,
      "grad_norm": 0.10872519016265869,
      "learning_rate": 4.610398872216503e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 6442672,
      "step": 7205
    },
    {
      "epoch": 1.902995908670978,
      "grad_norm": 0.28377804160118103,
      "learning_rate": 4.6098724023931796e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 6447376,
      "step": 7210
    },
    {
      "epoch": 1.9043156922264748,
      "grad_norm": 0.31565606594085693,
      "learning_rate": 4.609345607201017e-05,
      "loss": 0.1399,
      "num_input_tokens_seen": 6451792,
      "step": 7215
    },
    {
      "epoch": 1.9056354757819718,
      "grad_norm": 0.25704529881477356,
      "learning_rate": 4.608818486721254e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 6456336,
      "step": 7220
    },
    {
      "epoch": 1.9069552593374688,
      "grad_norm": 0.047139327973127365,
      "learning_rate": 4.608291041035179e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 6460720,
      "step": 7225
    },
    {
      "epoch": 1.9082750428929656,
      "grad_norm": 0.06276936084032059,
      "learning_rate": 4.607763270224132e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 6465360,
      "step": 7230
    },
    {
      "epoch": 1.9095948264484623,
      "grad_norm": 0.1673480123281479,
      "learning_rate": 4.6072351743695e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 6469744,
      "step": 7235
    },
    {
      "epoch": 1.9109146100039593,
      "grad_norm": 0.07516109198331833,
      "learning_rate": 4.606706753552723e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 6474096,
      "step": 7240
    },
    {
      "epoch": 1.9122343935594563,
      "grad_norm": 0.06936344504356384,
      "learning_rate": 4.6061780078552906e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 6478576,
      "step": 7245
    },
    {
      "epoch": 1.9135541771149531,
      "grad_norm": 0.09838573634624481,
      "learning_rate": 4.605648937358742e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 6482928,
      "step": 7250
    },
    {
      "epoch": 1.91487396067045,
      "grad_norm": 0.09534476697444916,
      "learning_rate": 4.605119542144665e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 6487632,
      "step": 7255
    },
    {
      "epoch": 1.916193744225947,
      "grad_norm": 0.1923447996377945,
      "learning_rate": 4.604589822294701e-05,
      "loss": 0.0984,
      "num_input_tokens_seen": 6491952,
      "step": 7260
    },
    {
      "epoch": 1.917513527781444,
      "grad_norm": 0.041288550943136215,
      "learning_rate": 4.604059777890537e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 6496496,
      "step": 7265
    },
    {
      "epoch": 1.918833311336941,
      "grad_norm": 0.2149384319782257,
      "learning_rate": 4.6035294090139145e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 6500752,
      "step": 7270
    },
    {
      "epoch": 1.9201530948924377,
      "grad_norm": 0.1374235600233078,
      "learning_rate": 4.6029987157466226e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 6505008,
      "step": 7275
    },
    {
      "epoch": 1.9214728784479345,
      "grad_norm": 0.3556566536426544,
      "learning_rate": 4.602467698170502e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 6509392,
      "step": 7280
    },
    {
      "epoch": 1.9227926620034315,
      "grad_norm": 0.28320884704589844,
      "learning_rate": 4.601936356367439e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 6513936,
      "step": 7285
    },
    {
      "epoch": 1.9241124455589285,
      "grad_norm": 0.2886987626552582,
      "learning_rate": 4.601404690419377e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 6518256,
      "step": 7290
    },
    {
      "epoch": 1.9254322291144252,
      "grad_norm": 0.20514874160289764,
      "learning_rate": 4.600872700408303e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 6522704,
      "step": 7295
    },
    {
      "epoch": 1.926752012669922,
      "grad_norm": 0.3322639763355255,
      "learning_rate": 4.600340386416258e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 6527408,
      "step": 7300
    },
    {
      "epoch": 1.928071796225419,
      "grad_norm": 0.16277071833610535,
      "learning_rate": 4.5998077485253296e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 6531856,
      "step": 7305
    },
    {
      "epoch": 1.929391579780916,
      "grad_norm": 0.21233443915843964,
      "learning_rate": 4.59927478681766e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 6536368,
      "step": 7310
    },
    {
      "epoch": 1.9307113633364128,
      "grad_norm": 0.36338627338409424,
      "learning_rate": 4.5987415013754366e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 6540688,
      "step": 7315
    },
    {
      "epoch": 1.9320311468919096,
      "grad_norm": 0.2929827868938446,
      "learning_rate": 4.598207892280899e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 6545200,
      "step": 7320
    },
    {
      "epoch": 1.9333509304474066,
      "grad_norm": 0.33678942918777466,
      "learning_rate": 4.597673959616337e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 6549808,
      "step": 7325
    },
    {
      "epoch": 1.9346707140029036,
      "grad_norm": 0.09622807800769806,
      "learning_rate": 4.597139703464089e-05,
      "loss": 0.0452,
      "num_input_tokens_seen": 6553968,
      "step": 7330
    },
    {
      "epoch": 1.9359904975584006,
      "grad_norm": 0.1935141384601593,
      "learning_rate": 4.596605123906545e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 6558544,
      "step": 7335
    },
    {
      "epoch": 1.9373102811138974,
      "grad_norm": 0.30104488134384155,
      "learning_rate": 4.596070221026143e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 6562896,
      "step": 7340
    },
    {
      "epoch": 1.9386300646693941,
      "grad_norm": 0.44369760155677795,
      "learning_rate": 4.595534994905372e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 6567184,
      "step": 7345
    },
    {
      "epoch": 1.9399498482248911,
      "grad_norm": 0.4172000586986542,
      "learning_rate": 4.594999445626771e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 6571632,
      "step": 7350
    },
    {
      "epoch": 1.9412696317803881,
      "grad_norm": 0.16598327457904816,
      "learning_rate": 4.5944635732729276e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 6576208,
      "step": 7355
    },
    {
      "epoch": 1.942589415335885,
      "grad_norm": 0.4370569586753845,
      "learning_rate": 4.5939273779264804e-05,
      "loss": 0.1078,
      "num_input_tokens_seen": 6580944,
      "step": 7360
    },
    {
      "epoch": 1.9439091988913817,
      "grad_norm": 0.29714450240135193,
      "learning_rate": 4.593390859670118e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 6585168,
      "step": 7365
    },
    {
      "epoch": 1.9452289824468787,
      "grad_norm": 0.12176475673913956,
      "learning_rate": 4.5928540185865776e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 6589840,
      "step": 7370
    },
    {
      "epoch": 1.9465487660023757,
      "grad_norm": 0.15068376064300537,
      "learning_rate": 4.592316854758648e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 6594128,
      "step": 7375
    },
    {
      "epoch": 1.9478685495578725,
      "grad_norm": 0.09998653829097748,
      "learning_rate": 4.5917793682691646e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 6598640,
      "step": 7380
    },
    {
      "epoch": 1.9491883331133693,
      "grad_norm": 0.39897748827934265,
      "learning_rate": 4.5912415592010164e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 6603120,
      "step": 7385
    },
    {
      "epoch": 1.9505081166688663,
      "grad_norm": 0.25645193457603455,
      "learning_rate": 4.5907034276371386e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 6607632,
      "step": 7390
    },
    {
      "epoch": 1.9518279002243633,
      "grad_norm": 0.28610509634017944,
      "learning_rate": 4.5901649736605196e-05,
      "loss": 0.1295,
      "num_input_tokens_seen": 6611952,
      "step": 7395
    },
    {
      "epoch": 1.9531476837798603,
      "grad_norm": 0.6255216598510742,
      "learning_rate": 4.589626197354195e-05,
      "loss": 0.1075,
      "num_input_tokens_seen": 6616176,
      "step": 7400
    },
    {
      "epoch": 1.9531476837798603,
      "eval_loss": 0.07558133453130722,
      "eval_runtime": 64.7185,
      "eval_samples_per_second": 104.066,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 6616176,
      "step": 7400
    },
    {
      "epoch": 1.954467467335357,
      "grad_norm": 0.08739921450614929,
      "learning_rate": 4.5890870988012504e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 6620592,
      "step": 7405
    },
    {
      "epoch": 1.9557872508908538,
      "grad_norm": 0.10449250042438507,
      "learning_rate": 4.5885476780848226e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 6625136,
      "step": 7410
    },
    {
      "epoch": 1.9571070344463508,
      "grad_norm": 0.11688117682933807,
      "learning_rate": 4.5880079352880964e-05,
      "loss": 0.0819,
      "num_input_tokens_seen": 6629840,
      "step": 7415
    },
    {
      "epoch": 1.9584268180018478,
      "grad_norm": 0.20546916127204895,
      "learning_rate": 4.5874678704943065e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 6634512,
      "step": 7420
    },
    {
      "epoch": 1.9597466015573446,
      "grad_norm": 0.20570893585681915,
      "learning_rate": 4.5869274837867394e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 6638928,
      "step": 7425
    },
    {
      "epoch": 1.9610663851128414,
      "grad_norm": 0.2569142282009125,
      "learning_rate": 4.5863867752487275e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 6643472,
      "step": 7430
    },
    {
      "epoch": 1.9623861686683384,
      "grad_norm": 0.13570162653923035,
      "learning_rate": 4.5858457449636554e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 6648112,
      "step": 7435
    },
    {
      "epoch": 1.9637059522238354,
      "grad_norm": 0.16203832626342773,
      "learning_rate": 4.5853043930149574e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 6652496,
      "step": 7440
    },
    {
      "epoch": 1.9650257357793321,
      "grad_norm": 0.05498882755637169,
      "learning_rate": 4.584762719486117e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 6656656,
      "step": 7445
    },
    {
      "epoch": 1.966345519334829,
      "grad_norm": 0.16412782669067383,
      "learning_rate": 4.584220724460665e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 6661200,
      "step": 7450
    },
    {
      "epoch": 1.967665302890326,
      "grad_norm": 0.20030273497104645,
      "learning_rate": 4.5836784080221865e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 6665680,
      "step": 7455
    },
    {
      "epoch": 1.968985086445823,
      "grad_norm": 0.3757839798927307,
      "learning_rate": 4.583135770254312e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 6670128,
      "step": 7460
    },
    {
      "epoch": 1.97030487000132,
      "grad_norm": 0.24799558520317078,
      "learning_rate": 4.5825928112407236e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 6674352,
      "step": 7465
    },
    {
      "epoch": 1.9716246535568167,
      "grad_norm": 0.045685745775699615,
      "learning_rate": 4.582049531065152e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 6679088,
      "step": 7470
    },
    {
      "epoch": 1.9729444371123135,
      "grad_norm": 0.46948033571243286,
      "learning_rate": 4.5815059298113783e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 6683568,
      "step": 7475
    },
    {
      "epoch": 1.9742642206678105,
      "grad_norm": 0.12226312607526779,
      "learning_rate": 4.580962007563232e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 6688016,
      "step": 7480
    },
    {
      "epoch": 1.9755840042233075,
      "grad_norm": 0.09945758432149887,
      "learning_rate": 4.5804177644045935e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 6692240,
      "step": 7485
    },
    {
      "epoch": 1.9769037877788043,
      "grad_norm": 0.15116862952709198,
      "learning_rate": 4.579873200419391e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 6696592,
      "step": 7490
    },
    {
      "epoch": 1.978223571334301,
      "grad_norm": 0.1248839721083641,
      "learning_rate": 4.5793283156916046e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 6700784,
      "step": 7495
    },
    {
      "epoch": 1.979543354889798,
      "grad_norm": 0.4491346776485443,
      "learning_rate": 4.578783110305261e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 6705200,
      "step": 7500
    },
    {
      "epoch": 1.980863138445295,
      "grad_norm": 0.350924015045166,
      "learning_rate": 4.578237584344438e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 6709648,
      "step": 7505
    },
    {
      "epoch": 1.9821829220007918,
      "grad_norm": 0.21357512474060059,
      "learning_rate": 4.577691737893263e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 6714288,
      "step": 7510
    },
    {
      "epoch": 1.9835027055562886,
      "grad_norm": 0.389934778213501,
      "learning_rate": 4.577145571035912e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 6718896,
      "step": 7515
    },
    {
      "epoch": 1.9848224891117856,
      "grad_norm": 0.10635025054216385,
      "learning_rate": 4.576599083856611e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 6723248,
      "step": 7520
    },
    {
      "epoch": 1.9861422726672826,
      "grad_norm": 0.15491513907909393,
      "learning_rate": 4.576052276439635e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 6727856,
      "step": 7525
    },
    {
      "epoch": 1.9874620562227796,
      "grad_norm": 0.23030245304107666,
      "learning_rate": 4.575505148869308e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 6732496,
      "step": 7530
    },
    {
      "epoch": 1.9887818397782764,
      "grad_norm": 0.1251135915517807,
      "learning_rate": 4.574957701230006e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 6736976,
      "step": 7535
    },
    {
      "epoch": 1.9901016233337732,
      "grad_norm": 0.39714178442955017,
      "learning_rate": 4.57440993360615e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 6741456,
      "step": 7540
    },
    {
      "epoch": 1.9914214068892702,
      "grad_norm": 0.2053503841161728,
      "learning_rate": 4.5738618460822134e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 6745776,
      "step": 7545
    },
    {
      "epoch": 1.9927411904447672,
      "grad_norm": 0.3128725588321686,
      "learning_rate": 4.573313438742719e-05,
      "loss": 0.1104,
      "num_input_tokens_seen": 6750256,
      "step": 7550
    },
    {
      "epoch": 1.994060974000264,
      "grad_norm": 0.08910363167524338,
      "learning_rate": 4.5727647116722374e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 6754704,
      "step": 7555
    },
    {
      "epoch": 1.9953807575557607,
      "grad_norm": 0.209807351231575,
      "learning_rate": 4.5722156649553884e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 6758896,
      "step": 7560
    },
    {
      "epoch": 1.9967005411112577,
      "grad_norm": 0.1922292411327362,
      "learning_rate": 4.571666298676843e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 6763248,
      "step": 7565
    },
    {
      "epoch": 1.9980203246667547,
      "grad_norm": 0.11855418980121613,
      "learning_rate": 4.571116612921321e-05,
      "loss": 0.1161,
      "num_input_tokens_seen": 6767472,
      "step": 7570
    },
    {
      "epoch": 1.9993401082222515,
      "grad_norm": 0.36413517594337463,
      "learning_rate": 4.57056660777359e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 6771952,
      "step": 7575
    },
    {
      "epoch": 2.0005279134221987,
      "grad_norm": 0.25645843148231506,
      "learning_rate": 4.5700162833184666e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 6775856,
      "step": 7580
    },
    {
      "epoch": 2.0018476969776957,
      "grad_norm": 0.4639410376548767,
      "learning_rate": 4.5694656396408195e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 6780080,
      "step": 7585
    },
    {
      "epoch": 2.0031674805331927,
      "grad_norm": 0.3608251214027405,
      "learning_rate": 4.5689146768255646e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 6784528,
      "step": 7590
    },
    {
      "epoch": 2.0044872640886893,
      "grad_norm": 0.07005508244037628,
      "learning_rate": 4.568363394957667e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 6789168,
      "step": 7595
    },
    {
      "epoch": 2.0058070476441863,
      "grad_norm": 0.060909345746040344,
      "learning_rate": 4.567811794122141e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 6793648,
      "step": 7600
    },
    {
      "epoch": 2.0058070476441863,
      "eval_loss": 0.07510706037282944,
      "eval_runtime": 64.7268,
      "eval_samples_per_second": 104.053,
      "eval_steps_per_second": 26.017,
      "num_input_tokens_seen": 6793648,
      "step": 7600
    },
    {
      "epoch": 2.0071268311996833,
      "grad_norm": 0.21742381155490875,
      "learning_rate": 4.56725987440405e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 6798160,
      "step": 7605
    },
    {
      "epoch": 2.0084466147551803,
      "grad_norm": 0.2879336476325989,
      "learning_rate": 4.566707635888508e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 6803024,
      "step": 7610
    },
    {
      "epoch": 2.009766398310677,
      "grad_norm": 0.21169108152389526,
      "learning_rate": 4.566155078660677e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 6807376,
      "step": 7615
    },
    {
      "epoch": 2.011086181866174,
      "grad_norm": 0.12386344373226166,
      "learning_rate": 4.565602202805768e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 6811920,
      "step": 7620
    },
    {
      "epoch": 2.012405965421671,
      "grad_norm": 0.2798001170158386,
      "learning_rate": 4.56504900840904e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 6816208,
      "step": 7625
    },
    {
      "epoch": 2.013725748977168,
      "grad_norm": 0.4686562716960907,
      "learning_rate": 4.564495495555805e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 6820528,
      "step": 7630
    },
    {
      "epoch": 2.015045532532665,
      "grad_norm": 0.3561771810054779,
      "learning_rate": 4.5639416643314204e-05,
      "loss": 0.117,
      "num_input_tokens_seen": 6825200,
      "step": 7635
    },
    {
      "epoch": 2.0163653160881614,
      "grad_norm": 0.33920371532440186,
      "learning_rate": 4.5633875148212946e-05,
      "loss": 0.1183,
      "num_input_tokens_seen": 6829712,
      "step": 7640
    },
    {
      "epoch": 2.0176850996436584,
      "grad_norm": 0.36639538407325745,
      "learning_rate": 4.562833047110883e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 6834224,
      "step": 7645
    },
    {
      "epoch": 2.0190048831991554,
      "grad_norm": 0.06148790940642357,
      "learning_rate": 4.5622782612856923e-05,
      "loss": 0.0191,
      "num_input_tokens_seen": 6839152,
      "step": 7650
    },
    {
      "epoch": 2.0203246667546524,
      "grad_norm": 0.08195067942142487,
      "learning_rate": 4.561723157431278e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 6843568,
      "step": 7655
    },
    {
      "epoch": 2.021644450310149,
      "grad_norm": 0.2522927522659302,
      "learning_rate": 4.5611677356332435e-05,
      "loss": 0.1466,
      "num_input_tokens_seen": 6848112,
      "step": 7660
    },
    {
      "epoch": 2.022964233865646,
      "grad_norm": 0.03633113205432892,
      "learning_rate": 4.560611995977242e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 6852592,
      "step": 7665
    },
    {
      "epoch": 2.024284017421143,
      "grad_norm": 0.25907647609710693,
      "learning_rate": 4.560055938548975e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 6856784,
      "step": 7670
    },
    {
      "epoch": 2.02560380097664,
      "grad_norm": 0.20321422815322876,
      "learning_rate": 4.5594995634341944e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 6861296,
      "step": 7675
    },
    {
      "epoch": 2.026923584532137,
      "grad_norm": 0.5927806496620178,
      "learning_rate": 4.5589428707187e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 6865840,
      "step": 7680
    },
    {
      "epoch": 2.0282433680876335,
      "grad_norm": 0.16285324096679688,
      "learning_rate": 4.55838586048834e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 6870160,
      "step": 7685
    },
    {
      "epoch": 2.0295631516431305,
      "grad_norm": 0.07795558124780655,
      "learning_rate": 4.557828532829013e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 6874992,
      "step": 7690
    },
    {
      "epoch": 2.0308829351986275,
      "grad_norm": 0.10244237631559372,
      "learning_rate": 4.557270887826667e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 6879312,
      "step": 7695
    },
    {
      "epoch": 2.0322027187541245,
      "grad_norm": 0.20857179164886475,
      "learning_rate": 4.556712925567296e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 6883984,
      "step": 7700
    },
    {
      "epoch": 2.033522502309621,
      "grad_norm": 0.4328763782978058,
      "learning_rate": 4.5561546461369454e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 6888272,
      "step": 7705
    },
    {
      "epoch": 2.034842285865118,
      "grad_norm": 0.34909024834632874,
      "learning_rate": 4.55559604962171e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 6892688,
      "step": 7710
    },
    {
      "epoch": 2.036162069420615,
      "grad_norm": 0.40270885825157166,
      "learning_rate": 4.55503713610773e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 6897360,
      "step": 7715
    },
    {
      "epoch": 2.037481852976112,
      "grad_norm": 0.26024842262268066,
      "learning_rate": 4.5544779056812e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 6901840,
      "step": 7720
    },
    {
      "epoch": 2.0388016365316086,
      "grad_norm": 0.1849614530801773,
      "learning_rate": 4.553918358428358e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 6906160,
      "step": 7725
    },
    {
      "epoch": 2.0401214200871056,
      "grad_norm": 0.2401839643716812,
      "learning_rate": 4.553358494435494e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 6910704,
      "step": 7730
    },
    {
      "epoch": 2.0414412036426026,
      "grad_norm": 0.2517256438732147,
      "learning_rate": 4.5527983137889464e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 6915280,
      "step": 7735
    },
    {
      "epoch": 2.0427609871980996,
      "grad_norm": 0.1625821739435196,
      "learning_rate": 4.5522378165751015e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 6919824,
      "step": 7740
    },
    {
      "epoch": 2.0440807707535966,
      "grad_norm": 0.10548185557126999,
      "learning_rate": 4.5516770028803954e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 6924432,
      "step": 7745
    },
    {
      "epoch": 2.045400554309093,
      "grad_norm": 0.15737095475196838,
      "learning_rate": 4.5511158727913116e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 6929232,
      "step": 7750
    },
    {
      "epoch": 2.04672033786459,
      "grad_norm": 0.33843058347702026,
      "learning_rate": 4.5505544263943856e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 6933744,
      "step": 7755
    },
    {
      "epoch": 2.048040121420087,
      "grad_norm": 0.2520791292190552,
      "learning_rate": 4.549992663776197e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 6938128,
      "step": 7760
    },
    {
      "epoch": 2.049359904975584,
      "grad_norm": 0.04124077409505844,
      "learning_rate": 4.5494305850233786e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 6942512,
      "step": 7765
    },
    {
      "epoch": 2.0506796885310807,
      "grad_norm": 0.05362673103809357,
      "learning_rate": 4.5488681902226094e-05,
      "loss": 0.0542,
      "num_input_tokens_seen": 6946832,
      "step": 7770
    },
    {
      "epoch": 2.0519994720865777,
      "grad_norm": 0.20405462384223938,
      "learning_rate": 4.5483054794606174e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 6951184,
      "step": 7775
    },
    {
      "epoch": 2.0533192556420747,
      "grad_norm": 0.09370773285627365,
      "learning_rate": 4.547742452824179e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 6955696,
      "step": 7780
    },
    {
      "epoch": 2.0546390391975717,
      "grad_norm": 0.3412418067455292,
      "learning_rate": 4.5471791104001215e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 6960016,
      "step": 7785
    },
    {
      "epoch": 2.0559588227530683,
      "grad_norm": 0.784838080406189,
      "learning_rate": 4.546615452275319e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 6964496,
      "step": 7790
    },
    {
      "epoch": 2.0572786063085653,
      "grad_norm": 0.16238315403461456,
      "learning_rate": 4.5460514785366944e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 6969168,
      "step": 7795
    },
    {
      "epoch": 2.0585983898640623,
      "grad_norm": 0.10413187742233276,
      "learning_rate": 4.545487189271219e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 6973744,
      "step": 7800
    },
    {
      "epoch": 2.0585983898640623,
      "eval_loss": 0.0754496306180954,
      "eval_runtime": 64.7725,
      "eval_samples_per_second": 103.979,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 6973744,
      "step": 7800
    },
    {
      "epoch": 2.0599181734195593,
      "grad_norm": 0.3889513313770294,
      "learning_rate": 4.544922584565914e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 6977968,
      "step": 7805
    },
    {
      "epoch": 2.0612379569750563,
      "grad_norm": 0.25308260321617126,
      "learning_rate": 4.544357664507848e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 6982416,
      "step": 7810
    },
    {
      "epoch": 2.062557740530553,
      "grad_norm": 0.13888294994831085,
      "learning_rate": 4.54379242918414e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 6986512,
      "step": 7815
    },
    {
      "epoch": 2.06387752408605,
      "grad_norm": 0.489170640707016,
      "learning_rate": 4.543226878681955e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 6990576,
      "step": 7820
    },
    {
      "epoch": 2.065197307641547,
      "grad_norm": 0.16548322141170502,
      "learning_rate": 4.5426610130885087e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 6994640,
      "step": 7825
    },
    {
      "epoch": 2.066517091197044,
      "grad_norm": 0.08078751713037491,
      "learning_rate": 4.542094832491064e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 6998864,
      "step": 7830
    },
    {
      "epoch": 2.0678368747525404,
      "grad_norm": 0.253131628036499,
      "learning_rate": 4.541528336976934e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 7003248,
      "step": 7835
    },
    {
      "epoch": 2.0691566583080374,
      "grad_norm": 0.1995379477739334,
      "learning_rate": 4.540961526633479e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 7007664,
      "step": 7840
    },
    {
      "epoch": 2.0704764418635344,
      "grad_norm": 0.29386454820632935,
      "learning_rate": 4.540394401548108e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 7012144,
      "step": 7845
    },
    {
      "epoch": 2.0717962254190314,
      "grad_norm": 0.12137898057699203,
      "learning_rate": 4.539826961808279e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 7016816,
      "step": 7850
    },
    {
      "epoch": 2.073116008974528,
      "grad_norm": 0.506098747253418,
      "learning_rate": 4.5392592075014994e-05,
      "loss": 0.0963,
      "num_input_tokens_seen": 7021072,
      "step": 7855
    },
    {
      "epoch": 2.074435792530025,
      "grad_norm": 0.09510201215744019,
      "learning_rate": 4.538691138715322e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 7025488,
      "step": 7860
    },
    {
      "epoch": 2.075755576085522,
      "grad_norm": 0.07047756761312485,
      "learning_rate": 4.5381227555373516e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 7030224,
      "step": 7865
    },
    {
      "epoch": 2.077075359641019,
      "grad_norm": 0.04694641754031181,
      "learning_rate": 4.537554058055239e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 7034448,
      "step": 7870
    },
    {
      "epoch": 2.078395143196516,
      "grad_norm": 0.24805070459842682,
      "learning_rate": 4.5369850463566865e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 7039280,
      "step": 7875
    },
    {
      "epoch": 2.0797149267520125,
      "grad_norm": 0.05336683616042137,
      "learning_rate": 4.5364157205294404e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 7043760,
      "step": 7880
    },
    {
      "epoch": 2.0810347103075095,
      "grad_norm": 0.3307097554206848,
      "learning_rate": 4.5358460806612996e-05,
      "loss": 0.1189,
      "num_input_tokens_seen": 7047856,
      "step": 7885
    },
    {
      "epoch": 2.0823544938630065,
      "grad_norm": 0.15268053114414215,
      "learning_rate": 4.535276126840109e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 7052400,
      "step": 7890
    },
    {
      "epoch": 2.0836742774185035,
      "grad_norm": 0.09284672886133194,
      "learning_rate": 4.5347058591537626e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 7056976,
      "step": 7895
    },
    {
      "epoch": 2.084994060974,
      "grad_norm": 0.49223774671554565,
      "learning_rate": 4.534135277690203e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 7061712,
      "step": 7900
    },
    {
      "epoch": 2.086313844529497,
      "grad_norm": 0.31551679968833923,
      "learning_rate": 4.533564382537421e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 7066096,
      "step": 7905
    },
    {
      "epoch": 2.087633628084994,
      "grad_norm": 0.06261837482452393,
      "learning_rate": 4.532993173783456e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 7070544,
      "step": 7910
    },
    {
      "epoch": 2.088953411640491,
      "grad_norm": 0.23710490763187408,
      "learning_rate": 4.5324216515163954e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 7074736,
      "step": 7915
    },
    {
      "epoch": 2.0902731951959876,
      "grad_norm": 0.2738867402076721,
      "learning_rate": 4.531849815824375e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 7079216,
      "step": 7920
    },
    {
      "epoch": 2.0915929787514846,
      "grad_norm": 0.08283576369285583,
      "learning_rate": 4.5312776667955795e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 7083856,
      "step": 7925
    },
    {
      "epoch": 2.0929127623069816,
      "grad_norm": 0.09812027215957642,
      "learning_rate": 4.5307052045182405e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 7088272,
      "step": 7930
    },
    {
      "epoch": 2.0942325458624786,
      "grad_norm": 0.10546258091926575,
      "learning_rate": 4.53013242908064e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 7092688,
      "step": 7935
    },
    {
      "epoch": 2.0955523294179756,
      "grad_norm": 0.16834400594234467,
      "learning_rate": 4.529559340571107e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 7096720,
      "step": 7940
    },
    {
      "epoch": 2.096872112973472,
      "grad_norm": 0.28702494502067566,
      "learning_rate": 4.528985939078018e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 7101232,
      "step": 7945
    },
    {
      "epoch": 2.098191896528969,
      "grad_norm": 0.06454991549253464,
      "learning_rate": 4.5284122246898e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 7106032,
      "step": 7950
    },
    {
      "epoch": 2.099511680084466,
      "grad_norm": 0.5221126079559326,
      "learning_rate": 4.527838197494926e-05,
      "loss": 0.1118,
      "num_input_tokens_seen": 7110800,
      "step": 7955
    },
    {
      "epoch": 2.100831463639963,
      "grad_norm": 0.3111208379268646,
      "learning_rate": 4.527263857581918e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 7115408,
      "step": 7960
    },
    {
      "epoch": 2.1021512471954598,
      "grad_norm": 0.5192667245864868,
      "learning_rate": 4.526689205039347e-05,
      "loss": 0.0788,
      "num_input_tokens_seen": 7120112,
      "step": 7965
    },
    {
      "epoch": 2.1034710307509568,
      "grad_norm": 0.3179950416088104,
      "learning_rate": 4.5261142399558324e-05,
      "loss": 0.1121,
      "num_input_tokens_seen": 7124624,
      "step": 7970
    },
    {
      "epoch": 2.1047908143064538,
      "grad_norm": 0.3168044090270996,
      "learning_rate": 4.525538962420041e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 7128784,
      "step": 7975
    },
    {
      "epoch": 2.1061105978619508,
      "grad_norm": 0.1840326189994812,
      "learning_rate": 4.524963372520685e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 7133168,
      "step": 7980
    },
    {
      "epoch": 2.1074303814174478,
      "grad_norm": 0.2840273976325989,
      "learning_rate": 4.524387470346531e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 7137616,
      "step": 7985
    },
    {
      "epoch": 2.1087501649729443,
      "grad_norm": 0.2445802539587021,
      "learning_rate": 4.5238112559863885e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 7142096,
      "step": 7990
    },
    {
      "epoch": 2.1100699485284413,
      "grad_norm": 0.40466511249542236,
      "learning_rate": 4.5232347295291175e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 7146256,
      "step": 7995
    },
    {
      "epoch": 2.1113897320839383,
      "grad_norm": 0.06845315545797348,
      "learning_rate": 4.522657891063626e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 7150896,
      "step": 8000
    },
    {
      "epoch": 2.1113897320839383,
      "eval_loss": 0.07489334791898727,
      "eval_runtime": 64.7648,
      "eval_samples_per_second": 103.992,
      "eval_steps_per_second": 26.002,
      "num_input_tokens_seen": 7150896,
      "step": 8000
    },
    {
      "epoch": 2.1127095156394353,
      "grad_norm": 0.4186022877693176,
      "learning_rate": 4.52208074067887e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 7155792,
      "step": 8005
    },
    {
      "epoch": 2.114029299194932,
      "grad_norm": 0.1848691701889038,
      "learning_rate": 4.5215032784638516e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 7160272,
      "step": 8010
    },
    {
      "epoch": 2.115349082750429,
      "grad_norm": 0.24331854283809662,
      "learning_rate": 4.5209255045076245e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 7164592,
      "step": 8015
    },
    {
      "epoch": 2.116668866305926,
      "grad_norm": 0.9791966080665588,
      "learning_rate": 4.5203474188992875e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 7168944,
      "step": 8020
    },
    {
      "epoch": 2.117988649861423,
      "grad_norm": 0.1498843878507614,
      "learning_rate": 4.51976902172799e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 7173200,
      "step": 8025
    },
    {
      "epoch": 2.1193084334169194,
      "grad_norm": 0.13253352046012878,
      "learning_rate": 4.519190313082927e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 7177648,
      "step": 8030
    },
    {
      "epoch": 2.1206282169724164,
      "grad_norm": 0.18419434130191803,
      "learning_rate": 4.518611293053343e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 7182256,
      "step": 8035
    },
    {
      "epoch": 2.1219480005279134,
      "grad_norm": 0.22218145430088043,
      "learning_rate": 4.51803196172853e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 7186800,
      "step": 8040
    },
    {
      "epoch": 2.1232677840834104,
      "grad_norm": 0.2679457664489746,
      "learning_rate": 4.517452319197828e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 7191184,
      "step": 8045
    },
    {
      "epoch": 2.1245875676389074,
      "grad_norm": 0.30709266662597656,
      "learning_rate": 4.5168723655506265e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 7195664,
      "step": 8050
    },
    {
      "epoch": 2.125907351194404,
      "grad_norm": 0.27189576625823975,
      "learning_rate": 4.51629210087636e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 7200080,
      "step": 8055
    },
    {
      "epoch": 2.127227134749901,
      "grad_norm": 0.2236274629831314,
      "learning_rate": 4.515711525264513e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 7204528,
      "step": 8060
    },
    {
      "epoch": 2.128546918305398,
      "grad_norm": 0.047212280333042145,
      "learning_rate": 4.5151306388046175e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 7209136,
      "step": 8065
    },
    {
      "epoch": 2.129866701860895,
      "grad_norm": 0.23198921978473663,
      "learning_rate": 4.514549441586255e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 7213648,
      "step": 8070
    },
    {
      "epoch": 2.1311864854163916,
      "grad_norm": 0.2576185464859009,
      "learning_rate": 4.513967933699051e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 7218480,
      "step": 8075
    },
    {
      "epoch": 2.1325062689718886,
      "grad_norm": 0.18668513000011444,
      "learning_rate": 4.513386115232684e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 7222960,
      "step": 8080
    },
    {
      "epoch": 2.1338260525273856,
      "grad_norm": 0.3644115626811981,
      "learning_rate": 4.5128039862768745e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 7227536,
      "step": 8085
    },
    {
      "epoch": 2.1351458360828826,
      "grad_norm": 0.18272079527378082,
      "learning_rate": 4.512221546921397e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 7232048,
      "step": 8090
    },
    {
      "epoch": 2.136465619638379,
      "grad_norm": 0.10349217802286148,
      "learning_rate": 4.5116387972560694e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 7236432,
      "step": 8095
    },
    {
      "epoch": 2.137785403193876,
      "grad_norm": 0.2785578966140747,
      "learning_rate": 4.511055737370759e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 7240752,
      "step": 8100
    },
    {
      "epoch": 2.139105186749373,
      "grad_norm": 0.17527374625205994,
      "learning_rate": 4.510472367355383e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 7245552,
      "step": 8105
    },
    {
      "epoch": 2.14042497030487,
      "grad_norm": 0.22334282100200653,
      "learning_rate": 4.509888687299901e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 7250128,
      "step": 8110
    },
    {
      "epoch": 2.141744753860367,
      "grad_norm": 0.2183559089899063,
      "learning_rate": 4.5093046972943266e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 7254256,
      "step": 8115
    },
    {
      "epoch": 2.1430645374158637,
      "grad_norm": 0.44457024335861206,
      "learning_rate": 4.508720397428717e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 7258736,
      "step": 8120
    },
    {
      "epoch": 2.1443843209713607,
      "grad_norm": 0.07362023741006851,
      "learning_rate": 4.508135787793178e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 7263280,
      "step": 8125
    },
    {
      "epoch": 2.1457041045268577,
      "grad_norm": 0.21433889865875244,
      "learning_rate": 4.5075508684778664e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 7267760,
      "step": 8130
    },
    {
      "epoch": 2.1470238880823547,
      "grad_norm": 0.07051128894090652,
      "learning_rate": 4.506965639572982e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 7272688,
      "step": 8135
    },
    {
      "epoch": 2.1483436716378512,
      "grad_norm": 0.36079487204551697,
      "learning_rate": 4.506380101168774e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 7277136,
      "step": 8140
    },
    {
      "epoch": 2.1496634551933482,
      "grad_norm": 0.1364288032054901,
      "learning_rate": 4.505794253355542e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 7282064,
      "step": 8145
    },
    {
      "epoch": 2.1509832387488452,
      "grad_norm": 0.4952329397201538,
      "learning_rate": 4.5052080962236286e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 7286416,
      "step": 8150
    },
    {
      "epoch": 2.1523030223043422,
      "grad_norm": 0.20140352845191956,
      "learning_rate": 4.504621629863428e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 7290640,
      "step": 8155
    },
    {
      "epoch": 2.153622805859839,
      "grad_norm": 0.10488418489694595,
      "learning_rate": 4.504034854365381e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 7294800,
      "step": 8160
    },
    {
      "epoch": 2.154942589415336,
      "grad_norm": 0.37058648467063904,
      "learning_rate": 4.503447769819974e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 7299312,
      "step": 8165
    },
    {
      "epoch": 2.156262372970833,
      "grad_norm": 0.21084456145763397,
      "learning_rate": 4.502860376317745e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 7303408,
      "step": 8170
    },
    {
      "epoch": 2.15758215652633,
      "grad_norm": 0.09452219307422638,
      "learning_rate": 4.502272673949276e-05,
      "loss": 0.0984,
      "num_input_tokens_seen": 7307920,
      "step": 8175
    },
    {
      "epoch": 2.158901940081827,
      "grad_norm": 0.038731399923563004,
      "learning_rate": 4.501684662805199e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 7312176,
      "step": 8180
    },
    {
      "epoch": 2.1602217236373233,
      "grad_norm": 0.28049367666244507,
      "learning_rate": 4.5010963429761924e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 7316848,
      "step": 8185
    },
    {
      "epoch": 2.1615415071928203,
      "grad_norm": 0.26468542218208313,
      "learning_rate": 4.500507714552982e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 7321264,
      "step": 8190
    },
    {
      "epoch": 2.1628612907483173,
      "grad_norm": 0.0818038359284401,
      "learning_rate": 4.499918777626342e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 7325872,
      "step": 8195
    },
    {
      "epoch": 2.1641810743038143,
      "grad_norm": 0.7742751836776733,
      "learning_rate": 4.499329532287093e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 7330032,
      "step": 8200
    },
    {
      "epoch": 2.1641810743038143,
      "eval_loss": 0.0747479721903801,
      "eval_runtime": 64.7526,
      "eval_samples_per_second": 104.011,
      "eval_steps_per_second": 26.007,
      "num_input_tokens_seen": 7330032,
      "step": 8200
    },
    {
      "epoch": 2.165500857859311,
      "grad_norm": 0.15500406920909882,
      "learning_rate": 4.4987399786261064e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 7334320,
      "step": 8205
    },
    {
      "epoch": 2.166820641414808,
      "grad_norm": 0.06565006822347641,
      "learning_rate": 4.498150116734297e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 7338640,
      "step": 8210
    },
    {
      "epoch": 2.168140424970305,
      "grad_norm": 0.18759655952453613,
      "learning_rate": 4.4975599467026294e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 7343312,
      "step": 8215
    },
    {
      "epoch": 2.169460208525802,
      "grad_norm": 0.1542542725801468,
      "learning_rate": 4.496969468622114e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 7347824,
      "step": 8220
    },
    {
      "epoch": 2.1707799920812985,
      "grad_norm": 0.270298033952713,
      "learning_rate": 4.496378682583813e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 7352528,
      "step": 8225
    },
    {
      "epoch": 2.1720997756367955,
      "grad_norm": 0.03728686273097992,
      "learning_rate": 4.495787588678829e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 7357008,
      "step": 8230
    },
    {
      "epoch": 2.1734195591922925,
      "grad_norm": 0.18597032129764557,
      "learning_rate": 4.4951961869983196e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 7361552,
      "step": 8235
    },
    {
      "epoch": 2.1747393427477895,
      "grad_norm": 0.13053740561008453,
      "learning_rate": 4.494604477633485e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 7366128,
      "step": 8240
    },
    {
      "epoch": 2.1760591263032865,
      "grad_norm": 0.38580843806266785,
      "learning_rate": 4.4940124606755734e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 7370384,
      "step": 8245
    },
    {
      "epoch": 2.177378909858783,
      "grad_norm": 0.24125556647777557,
      "learning_rate": 4.493420136215882e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 7374800,
      "step": 8250
    },
    {
      "epoch": 2.17869869341428,
      "grad_norm": 0.06357249617576599,
      "learning_rate": 4.492827504345756e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 7379152,
      "step": 8255
    },
    {
      "epoch": 2.180018476969777,
      "grad_norm": 0.07770431041717529,
      "learning_rate": 4.492234565156584e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 7383568,
      "step": 8260
    },
    {
      "epoch": 2.181338260525274,
      "grad_norm": 0.15502195060253143,
      "learning_rate": 4.491641318739807e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 7387760,
      "step": 8265
    },
    {
      "epoch": 2.1826580440807706,
      "grad_norm": 0.2789313793182373,
      "learning_rate": 4.4910477651869096e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 7391888,
      "step": 8270
    },
    {
      "epoch": 2.1839778276362676,
      "grad_norm": 0.22910848259925842,
      "learning_rate": 4.4904539045894254e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 7396144,
      "step": 8275
    },
    {
      "epoch": 2.1852976111917646,
      "grad_norm": 0.3092767894268036,
      "learning_rate": 4.4898597370389364e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 7400560,
      "step": 8280
    },
    {
      "epoch": 2.1866173947472616,
      "grad_norm": 0.25779837369918823,
      "learning_rate": 4.489265262627069e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 7405296,
      "step": 8285
    },
    {
      "epoch": 2.187937178302758,
      "grad_norm": 0.05394850671291351,
      "learning_rate": 4.488670481445499e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 7409680,
      "step": 8290
    },
    {
      "epoch": 2.189256961858255,
      "grad_norm": 0.22546350955963135,
      "learning_rate": 4.488075393585951e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 7414000,
      "step": 8295
    },
    {
      "epoch": 2.190576745413752,
      "grad_norm": 0.2602360248565674,
      "learning_rate": 4.487479999140193e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 7418768,
      "step": 8300
    },
    {
      "epoch": 2.191896528969249,
      "grad_norm": 0.11775390803813934,
      "learning_rate": 4.4868842982000425e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 7423184,
      "step": 8305
    },
    {
      "epoch": 2.193216312524746,
      "grad_norm": 0.24081440269947052,
      "learning_rate": 4.486288290857365e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 7427696,
      "step": 8310
    },
    {
      "epoch": 2.1945360960802427,
      "grad_norm": 0.23181308805942535,
      "learning_rate": 4.4856919772040715e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 7432272,
      "step": 8315
    },
    {
      "epoch": 2.1958558796357397,
      "grad_norm": 0.13985151052474976,
      "learning_rate": 4.485095357332122e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 7437072,
      "step": 8320
    },
    {
      "epoch": 2.1971756631912367,
      "grad_norm": 0.04570583254098892,
      "learning_rate": 4.484498431333521e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 7441488,
      "step": 8325
    },
    {
      "epoch": 2.1984954467467337,
      "grad_norm": 0.1922800987958908,
      "learning_rate": 4.4839011993003245e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 7446192,
      "step": 8330
    },
    {
      "epoch": 2.1998152303022303,
      "grad_norm": 0.39609408378601074,
      "learning_rate": 4.4833036613246305e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 7450512,
      "step": 8335
    },
    {
      "epoch": 2.2011350138577273,
      "grad_norm": 0.1092604473233223,
      "learning_rate": 4.482705817498589e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 7455024,
      "step": 8340
    },
    {
      "epoch": 2.2024547974132243,
      "grad_norm": 0.052072856575250626,
      "learning_rate": 4.4821076679143934e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 7459536,
      "step": 8345
    },
    {
      "epoch": 2.2037745809687213,
      "grad_norm": 0.19168980419635773,
      "learning_rate": 4.481509212664288e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 7463920,
      "step": 8350
    },
    {
      "epoch": 2.205094364524218,
      "grad_norm": 0.1736547201871872,
      "learning_rate": 4.480910451840559e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 7468688,
      "step": 8355
    },
    {
      "epoch": 2.206414148079715,
      "grad_norm": 0.37127041816711426,
      "learning_rate": 4.480311385535546e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 7473456,
      "step": 8360
    },
    {
      "epoch": 2.207733931635212,
      "grad_norm": 0.1207113191485405,
      "learning_rate": 4.47971201384163e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 7478096,
      "step": 8365
    },
    {
      "epoch": 2.209053715190709,
      "grad_norm": 0.14862358570098877,
      "learning_rate": 4.4791123368512446e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 7482256,
      "step": 8370
    },
    {
      "epoch": 2.210373498746206,
      "grad_norm": 0.04348155856132507,
      "learning_rate": 4.478512354656864e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 7486608,
      "step": 8375
    },
    {
      "epoch": 2.2116932823017024,
      "grad_norm": 0.11444752663373947,
      "learning_rate": 4.477912067351016e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 7491184,
      "step": 8380
    },
    {
      "epoch": 2.2130130658571994,
      "grad_norm": 0.25760114192962646,
      "learning_rate": 4.477311475026271e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 7495728,
      "step": 8385
    },
    {
      "epoch": 2.2143328494126964,
      "grad_norm": 0.6964402198791504,
      "learning_rate": 4.476710577775248e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 7500336,
      "step": 8390
    },
    {
      "epoch": 2.2156526329681934,
      "grad_norm": 0.1507468968629837,
      "learning_rate": 4.476109375690612e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 7504592,
      "step": 8395
    },
    {
      "epoch": 2.21697241652369,
      "grad_norm": 0.3025268018245697,
      "learning_rate": 4.4755078688650784e-05,
      "loss": 0.1147,
      "num_input_tokens_seen": 7508816,
      "step": 8400
    },
    {
      "epoch": 2.21697241652369,
      "eval_loss": 0.07559145987033844,
      "eval_runtime": 64.7224,
      "eval_samples_per_second": 104.06,
      "eval_steps_per_second": 26.019,
      "num_input_tokens_seen": 7508816,
      "step": 8400
    },
    {
      "epoch": 2.218292200079187,
      "grad_norm": 0.4827629625797272,
      "learning_rate": 4.474906057391406e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 7513232,
      "step": 8405
    },
    {
      "epoch": 2.219611983634684,
      "grad_norm": 0.38947463035583496,
      "learning_rate": 4.4743039413624e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 7518000,
      "step": 8410
    },
    {
      "epoch": 2.220931767190181,
      "grad_norm": 0.2349194586277008,
      "learning_rate": 4.473701520870916e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 7522544,
      "step": 8415
    },
    {
      "epoch": 2.2222515507456775,
      "grad_norm": 0.3964787423610687,
      "learning_rate": 4.4730987960098544e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 7526960,
      "step": 8420
    },
    {
      "epoch": 2.2235713343011745,
      "grad_norm": 0.044484756886959076,
      "learning_rate": 4.4724957668721635e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 7531280,
      "step": 8425
    },
    {
      "epoch": 2.2248911178566715,
      "grad_norm": 0.11051148921251297,
      "learning_rate": 4.471892433550836e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 7535728,
      "step": 8430
    },
    {
      "epoch": 2.2262109014121685,
      "grad_norm": 0.12696214020252228,
      "learning_rate": 4.471288796138916e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 7540560,
      "step": 8435
    },
    {
      "epoch": 2.2275306849676655,
      "grad_norm": 0.3541504144668579,
      "learning_rate": 4.470684854729491e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 7544912,
      "step": 8440
    },
    {
      "epoch": 2.228850468523162,
      "grad_norm": 0.07317245006561279,
      "learning_rate": 4.4700806094156955e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 7549552,
      "step": 8445
    },
    {
      "epoch": 2.230170252078659,
      "grad_norm": 0.271212100982666,
      "learning_rate": 4.469476060290713e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 7553904,
      "step": 8450
    },
    {
      "epoch": 2.231490035634156,
      "grad_norm": 0.25180765986442566,
      "learning_rate": 4.468871207447772e-05,
      "loss": 0.1045,
      "num_input_tokens_seen": 7558640,
      "step": 8455
    },
    {
      "epoch": 2.232809819189653,
      "grad_norm": 0.22470711171627045,
      "learning_rate": 4.4682660509801486e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 7562704,
      "step": 8460
    },
    {
      "epoch": 2.2341296027451496,
      "grad_norm": 0.043757762759923935,
      "learning_rate": 4.467660590981165e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 7567216,
      "step": 8465
    },
    {
      "epoch": 2.2354493863006466,
      "grad_norm": 0.24988308548927307,
      "learning_rate": 4.467054827544191e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 7571760,
      "step": 8470
    },
    {
      "epoch": 2.2367691698561436,
      "grad_norm": 0.13088427484035492,
      "learning_rate": 4.4664487607626434e-05,
      "loss": 0.0999,
      "num_input_tokens_seen": 7576144,
      "step": 8475
    },
    {
      "epoch": 2.2380889534116406,
      "grad_norm": 0.11584965139627457,
      "learning_rate": 4.4658423907299845e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 7580208,
      "step": 8480
    },
    {
      "epoch": 2.239408736967137,
      "grad_norm": 0.2219480723142624,
      "learning_rate": 4.465235717539725e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 7584880,
      "step": 8485
    },
    {
      "epoch": 2.240728520522634,
      "grad_norm": 0.28186845779418945,
      "learning_rate": 4.464628741285421e-05,
      "loss": 0.0965,
      "num_input_tokens_seen": 7589168,
      "step": 8490
    },
    {
      "epoch": 2.242048304078131,
      "grad_norm": 0.25733256340026855,
      "learning_rate": 4.4640214620606754e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 7593456,
      "step": 8495
    },
    {
      "epoch": 2.243368087633628,
      "grad_norm": 0.3034327030181885,
      "learning_rate": 4.46341387995914e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 7597936,
      "step": 8500
    },
    {
      "epoch": 2.244687871189125,
      "grad_norm": 0.7294071316719055,
      "learning_rate": 4.4628059950745106e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 7602256,
      "step": 8505
    },
    {
      "epoch": 2.2460076547446217,
      "grad_norm": 0.13351532816886902,
      "learning_rate": 4.4621978075005297e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 7607152,
      "step": 8510
    },
    {
      "epoch": 2.2473274383001187,
      "grad_norm": 0.1253606230020523,
      "learning_rate": 4.461589317330989e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 7611632,
      "step": 8515
    },
    {
      "epoch": 2.2486472218556157,
      "grad_norm": 0.2611004412174225,
      "learning_rate": 4.460980524659724e-05,
      "loss": 0.0967,
      "num_input_tokens_seen": 7616208,
      "step": 8520
    },
    {
      "epoch": 2.2499670054111127,
      "grad_norm": 0.06876882165670395,
      "learning_rate": 4.46037142958062e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 7620720,
      "step": 8525
    },
    {
      "epoch": 2.2512867889666097,
      "grad_norm": 0.049842096865177155,
      "learning_rate": 4.4597620321876046e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 7624720,
      "step": 8530
    },
    {
      "epoch": 2.2526065725221063,
      "grad_norm": 0.13951657712459564,
      "learning_rate": 4.459152332574656e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 7629072,
      "step": 8535
    },
    {
      "epoch": 2.2539263560776033,
      "grad_norm": 0.1306658834218979,
      "learning_rate": 4.4585423308357985e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 7633584,
      "step": 8540
    },
    {
      "epoch": 2.2552461396331003,
      "grad_norm": 0.07734635472297668,
      "learning_rate": 4.457932027065102e-05,
      "loss": 0.1035,
      "num_input_tokens_seen": 7637904,
      "step": 8545
    },
    {
      "epoch": 2.256565923188597,
      "grad_norm": 0.18720126152038574,
      "learning_rate": 4.45732142135668e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 7642352,
      "step": 8550
    },
    {
      "epoch": 2.257885706744094,
      "grad_norm": 0.18094687163829803,
      "learning_rate": 4.4567105138046986e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 7646640,
      "step": 8555
    },
    {
      "epoch": 2.259205490299591,
      "grad_norm": 0.4723154902458191,
      "learning_rate": 4.456099304503365e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 7651088,
      "step": 8560
    },
    {
      "epoch": 2.260525273855088,
      "grad_norm": 0.1067311093211174,
      "learning_rate": 4.455487793546939e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 7655216,
      "step": 8565
    },
    {
      "epoch": 2.261845057410585,
      "grad_norm": 0.2313871830701828,
      "learning_rate": 4.454875981029719e-05,
      "loss": 0.1012,
      "num_input_tokens_seen": 7659824,
      "step": 8570
    },
    {
      "epoch": 2.2631648409660814,
      "grad_norm": 0.05004528909921646,
      "learning_rate": 4.454263867046057e-05,
      "loss": 0.108,
      "num_input_tokens_seen": 7664176,
      "step": 8575
    },
    {
      "epoch": 2.2644846245215784,
      "grad_norm": 0.15142562985420227,
      "learning_rate": 4.4536514516903484e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 7668592,
      "step": 8580
    },
    {
      "epoch": 2.2658044080770754,
      "grad_norm": 0.1581783890724182,
      "learning_rate": 4.453038735057034e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 7672912,
      "step": 8585
    },
    {
      "epoch": 2.2671241916325724,
      "grad_norm": 0.03706010431051254,
      "learning_rate": 4.4524257172406034e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 7677456,
      "step": 8590
    },
    {
      "epoch": 2.2684439751880694,
      "grad_norm": 0.1639820784330368,
      "learning_rate": 4.451812398335592e-05,
      "loss": 0.1045,
      "num_input_tokens_seen": 7681744,
      "step": 8595
    },
    {
      "epoch": 2.269763758743566,
      "grad_norm": 0.08108969777822495,
      "learning_rate": 4.4511987784365805e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 7686352,
      "step": 8600
    },
    {
      "epoch": 2.269763758743566,
      "eval_loss": 0.07421866804361343,
      "eval_runtime": 64.7842,
      "eval_samples_per_second": 103.961,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 7686352,
      "step": 8600
    },
    {
      "epoch": 2.271083542299063,
      "grad_norm": 0.1125345379114151,
      "learning_rate": 4.450584857638197e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 7690896,
      "step": 8605
    },
    {
      "epoch": 2.27240332585456,
      "grad_norm": 0.15789508819580078,
      "learning_rate": 4.449970636035116e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 7695536,
      "step": 8610
    },
    {
      "epoch": 2.2737231094100565,
      "grad_norm": 0.5257883667945862,
      "learning_rate": 4.4493561137220574e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 7700048,
      "step": 8615
    },
    {
      "epoch": 2.2750428929655535,
      "grad_norm": 0.24736924469470978,
      "learning_rate": 4.44874129079379e-05,
      "loss": 0.0223,
      "num_input_tokens_seen": 7704528,
      "step": 8620
    },
    {
      "epoch": 2.2763626765210505,
      "grad_norm": 0.3959580361843109,
      "learning_rate": 4.4481261673451255e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 7708848,
      "step": 8625
    },
    {
      "epoch": 2.2776824600765475,
      "grad_norm": 0.29168596863746643,
      "learning_rate": 4.4475107434709245e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 7713552,
      "step": 8630
    },
    {
      "epoch": 2.2790022436320445,
      "grad_norm": 0.15896978974342346,
      "learning_rate": 4.446895019266093e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 7717808,
      "step": 8635
    },
    {
      "epoch": 2.280322027187541,
      "grad_norm": 0.4824647009372711,
      "learning_rate": 4.446278994825583e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 7722128,
      "step": 8640
    },
    {
      "epoch": 2.281641810743038,
      "grad_norm": 0.3201243281364441,
      "learning_rate": 4.445662670244394e-05,
      "loss": 0.1022,
      "num_input_tokens_seen": 7726320,
      "step": 8645
    },
    {
      "epoch": 2.282961594298535,
      "grad_norm": 0.20121006667613983,
      "learning_rate": 4.44504604561757e-05,
      "loss": 0.0697,
      "num_input_tokens_seen": 7730704,
      "step": 8650
    },
    {
      "epoch": 2.284281377854032,
      "grad_norm": 0.15775150060653687,
      "learning_rate": 4.4444291210402035e-05,
      "loss": 0.1065,
      "num_input_tokens_seen": 7735024,
      "step": 8655
    },
    {
      "epoch": 2.285601161409529,
      "grad_norm": 0.12930023670196533,
      "learning_rate": 4.443811896607431e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 7739472,
      "step": 8660
    },
    {
      "epoch": 2.2869209449650256,
      "grad_norm": 0.24057447910308838,
      "learning_rate": 4.443194372414436e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 7743952,
      "step": 8665
    },
    {
      "epoch": 2.2882407285205226,
      "grad_norm": 0.5563846826553345,
      "learning_rate": 4.442576548556449e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 7748464,
      "step": 8670
    },
    {
      "epoch": 2.2895605120760196,
      "grad_norm": 0.24843387305736542,
      "learning_rate": 4.441958425128747e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 7753168,
      "step": 8675
    },
    {
      "epoch": 2.290880295631516,
      "grad_norm": 0.20776672661304474,
      "learning_rate": 4.4413400022266515e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 7757712,
      "step": 8680
    },
    {
      "epoch": 2.292200079187013,
      "grad_norm": 0.0841105505824089,
      "learning_rate": 4.4407212799455313e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 7761904,
      "step": 8685
    },
    {
      "epoch": 2.29351986274251,
      "grad_norm": 0.10613290965557098,
      "learning_rate": 4.4401022583808003e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 7766480,
      "step": 8690
    },
    {
      "epoch": 2.294839646298007,
      "grad_norm": 0.29400667548179626,
      "learning_rate": 4.439482937627921e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 7771120,
      "step": 8695
    },
    {
      "epoch": 2.296159429853504,
      "grad_norm": 0.1684807986021042,
      "learning_rate": 4.4388633177824004e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 7775600,
      "step": 8700
    },
    {
      "epoch": 2.2974792134090007,
      "grad_norm": 0.4519040882587433,
      "learning_rate": 4.4382433989397895e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 7780112,
      "step": 8705
    },
    {
      "epoch": 2.2987989969644977,
      "grad_norm": 0.032076336443424225,
      "learning_rate": 4.4376231811956895e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 7784400,
      "step": 8710
    },
    {
      "epoch": 2.3001187805199947,
      "grad_norm": 0.5452390909194946,
      "learning_rate": 4.437002664645745e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 7789200,
      "step": 8715
    },
    {
      "epoch": 2.3014385640754917,
      "grad_norm": 0.13484683632850647,
      "learning_rate": 4.436381849385649e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 7793744,
      "step": 8720
    },
    {
      "epoch": 2.3027583476309887,
      "grad_norm": 0.029698144644498825,
      "learning_rate": 4.435760735511136e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 7798416,
      "step": 8725
    },
    {
      "epoch": 2.3040781311864853,
      "grad_norm": 0.07295767217874527,
      "learning_rate": 4.435139323117992e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 7802960,
      "step": 8730
    },
    {
      "epoch": 2.3053979147419823,
      "grad_norm": 0.1978251338005066,
      "learning_rate": 4.434517612302046e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 7807472,
      "step": 8735
    },
    {
      "epoch": 2.3067176982974793,
      "grad_norm": 0.20524822175502777,
      "learning_rate": 4.433895603159174e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 7811856,
      "step": 8740
    },
    {
      "epoch": 2.3080374818529763,
      "grad_norm": 0.24219156801700592,
      "learning_rate": 4.433273295785296e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 7816240,
      "step": 8745
    },
    {
      "epoch": 2.309357265408473,
      "grad_norm": 0.49947288632392883,
      "learning_rate": 4.432650690276382e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 7820336,
      "step": 8750
    },
    {
      "epoch": 2.31067704896397,
      "grad_norm": 0.5283124446868896,
      "learning_rate": 4.4320277867284435e-05,
      "loss": 0.108,
      "num_input_tokens_seen": 7825040,
      "step": 8755
    },
    {
      "epoch": 2.311996832519467,
      "grad_norm": 0.3377077281475067,
      "learning_rate": 4.431404585237541e-05,
      "loss": 0.0955,
      "num_input_tokens_seen": 7829616,
      "step": 8760
    },
    {
      "epoch": 2.313316616074964,
      "grad_norm": 0.18836532533168793,
      "learning_rate": 4.43078108589978e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 7833968,
      "step": 8765
    },
    {
      "epoch": 2.3146363996304604,
      "grad_norm": 0.1997460573911667,
      "learning_rate": 4.4301572888113116e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 7838320,
      "step": 8770
    },
    {
      "epoch": 2.3159561831859574,
      "grad_norm": 0.3314087986946106,
      "learning_rate": 4.4295331940683337e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 7842352,
      "step": 8775
    },
    {
      "epoch": 2.3172759667414544,
      "grad_norm": 0.19467756152153015,
      "learning_rate": 4.428908801767089e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 7846480,
      "step": 8780
    },
    {
      "epoch": 2.3185957502969514,
      "grad_norm": 0.11510217934846878,
      "learning_rate": 4.428284112003868e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 7850992,
      "step": 8785
    },
    {
      "epoch": 2.3199155338524484,
      "grad_norm": 0.3122459948062897,
      "learning_rate": 4.4276591248750033e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 7855248,
      "step": 8790
    },
    {
      "epoch": 2.321235317407945,
      "grad_norm": 0.38800936937332153,
      "learning_rate": 4.4270338404768774e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 7859536,
      "step": 8795
    },
    {
      "epoch": 2.322555100963442,
      "grad_norm": 0.8272725939750671,
      "learning_rate": 4.426408258905917e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 7864080,
      "step": 8800
    },
    {
      "epoch": 2.322555100963442,
      "eval_loss": 0.07382544875144958,
      "eval_runtime": 64.7374,
      "eval_samples_per_second": 104.036,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 7864080,
      "step": 8800
    },
    {
      "epoch": 2.323874884518939,
      "grad_norm": 0.22693680226802826,
      "learning_rate": 4.425782380258594e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 7868720,
      "step": 8805
    },
    {
      "epoch": 2.325194668074436,
      "grad_norm": 0.24192887544631958,
      "learning_rate": 4.425156204631427e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 7873200,
      "step": 8810
    },
    {
      "epoch": 2.3265144516299325,
      "grad_norm": 0.17911675572395325,
      "learning_rate": 4.424529732120981e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 7877840,
      "step": 8815
    },
    {
      "epoch": 2.3278342351854295,
      "grad_norm": 0.16295243799686432,
      "learning_rate": 4.423902962823864e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 7882416,
      "step": 8820
    },
    {
      "epoch": 2.3291540187409265,
      "grad_norm": 0.3257666230201721,
      "learning_rate": 4.423275896836733e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 7886864,
      "step": 8825
    },
    {
      "epoch": 2.3304738022964235,
      "grad_norm": 0.18956910073757172,
      "learning_rate": 4.42264853425629e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 7891024,
      "step": 8830
    },
    {
      "epoch": 2.33179358585192,
      "grad_norm": 0.2562328577041626,
      "learning_rate": 4.4220208751792816e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 7895280,
      "step": 8835
    },
    {
      "epoch": 2.333113369407417,
      "grad_norm": 0.23893992602825165,
      "learning_rate": 4.421392919702499e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 7900176,
      "step": 8840
    },
    {
      "epoch": 2.334433152962914,
      "grad_norm": 0.07824353128671646,
      "learning_rate": 4.4207646679227846e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 7904560,
      "step": 8845
    },
    {
      "epoch": 2.335752936518411,
      "grad_norm": 0.15438447892665863,
      "learning_rate": 4.42013611993702e-05,
      "loss": 0.0857,
      "num_input_tokens_seen": 7909072,
      "step": 8850
    },
    {
      "epoch": 2.337072720073908,
      "grad_norm": 0.16404542326927185,
      "learning_rate": 4.419507275842135e-05,
      "loss": 0.118,
      "num_input_tokens_seen": 7913680,
      "step": 8855
    },
    {
      "epoch": 2.3383925036294047,
      "grad_norm": 0.23074950277805328,
      "learning_rate": 4.418878135735106e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 7918032,
      "step": 8860
    },
    {
      "epoch": 2.3397122871849017,
      "grad_norm": 0.13195401430130005,
      "learning_rate": 4.418248699712955e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 7922704,
      "step": 8865
    },
    {
      "epoch": 2.3410320707403987,
      "grad_norm": 0.15457263588905334,
      "learning_rate": 4.417618967872748e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 7927184,
      "step": 8870
    },
    {
      "epoch": 2.3423518542958957,
      "grad_norm": 0.263943612575531,
      "learning_rate": 4.4169889403115985e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 7931152,
      "step": 8875
    },
    {
      "epoch": 2.343671637851392,
      "grad_norm": 0.062111757695674896,
      "learning_rate": 4.4163586171266627e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 7935536,
      "step": 8880
    },
    {
      "epoch": 2.344991421406889,
      "grad_norm": 0.31245023012161255,
      "learning_rate": 4.415727998415147e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 7939952,
      "step": 8885
    },
    {
      "epoch": 2.346311204962386,
      "grad_norm": 0.32612842321395874,
      "learning_rate": 4.4150970842742985e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 7944304,
      "step": 8890
    },
    {
      "epoch": 2.347630988517883,
      "grad_norm": 0.1825343668460846,
      "learning_rate": 4.4144658748014134e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 7948912,
      "step": 8895
    },
    {
      "epoch": 2.3489507720733798,
      "grad_norm": 0.16124267876148224,
      "learning_rate": 4.413834370093831e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 7953232,
      "step": 8900
    },
    {
      "epoch": 2.3502705556288768,
      "grad_norm": 0.28025346994400024,
      "learning_rate": 4.413202570248939e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 7957488,
      "step": 8905
    },
    {
      "epoch": 2.3515903391843738,
      "grad_norm": 0.23958459496498108,
      "learning_rate": 4.412570475364167e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 7962064,
      "step": 8910
    },
    {
      "epoch": 2.3529101227398708,
      "grad_norm": 0.3731449842453003,
      "learning_rate": 4.411938085536994e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 7966640,
      "step": 8915
    },
    {
      "epoch": 2.3542299062953678,
      "grad_norm": 0.08741065859794617,
      "learning_rate": 4.41130540086494e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 7971216,
      "step": 8920
    },
    {
      "epoch": 2.3555496898508643,
      "grad_norm": 0.17499606311321259,
      "learning_rate": 4.4106724214455754e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 7975952,
      "step": 8925
    },
    {
      "epoch": 2.3568694734063613,
      "grad_norm": 0.2939094305038452,
      "learning_rate": 4.4100391473765115e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 7980464,
      "step": 8930
    },
    {
      "epoch": 2.3581892569618583,
      "grad_norm": 0.2896695137023926,
      "learning_rate": 4.409405578755408e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 7984912,
      "step": 8935
    },
    {
      "epoch": 2.3595090405173553,
      "grad_norm": 0.34834980964660645,
      "learning_rate": 4.4087717156799705e-05,
      "loss": 0.0281,
      "num_input_tokens_seen": 7989616,
      "step": 8940
    },
    {
      "epoch": 2.360828824072852,
      "grad_norm": 0.2766324579715729,
      "learning_rate": 4.408137558247946e-05,
      "loss": 0.0783,
      "num_input_tokens_seen": 7994032,
      "step": 8945
    },
    {
      "epoch": 2.362148607628349,
      "grad_norm": 0.13774330914020538,
      "learning_rate": 4.4075031065571306e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 7998672,
      "step": 8950
    },
    {
      "epoch": 2.363468391183846,
      "grad_norm": 0.1742262840270996,
      "learning_rate": 4.406868360705366e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 8003088,
      "step": 8955
    },
    {
      "epoch": 2.364788174739343,
      "grad_norm": 0.2502993643283844,
      "learning_rate": 4.406233320790536e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 8007440,
      "step": 8960
    },
    {
      "epoch": 2.3661079582948394,
      "grad_norm": 0.1978713572025299,
      "learning_rate": 4.4055979869105734e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 8011728,
      "step": 8965
    },
    {
      "epoch": 2.3674277418503364,
      "grad_norm": 0.13060395419597626,
      "learning_rate": 4.404962359163454e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 8016112,
      "step": 8970
    },
    {
      "epoch": 2.3687475254058334,
      "grad_norm": 0.19114427268505096,
      "learning_rate": 4.404326437647199e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 8020944,
      "step": 8975
    },
    {
      "epoch": 2.3700673089613304,
      "grad_norm": 0.06981606036424637,
      "learning_rate": 4.403690222459877e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 8025392,
      "step": 8980
    },
    {
      "epoch": 2.3713870925168274,
      "grad_norm": 0.20018890500068665,
      "learning_rate": 4.4030537136995984e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 8029744,
      "step": 8985
    },
    {
      "epoch": 2.372706876072324,
      "grad_norm": 0.05702696740627289,
      "learning_rate": 4.402416911464523e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 8033968,
      "step": 8990
    },
    {
      "epoch": 2.374026659627821,
      "grad_norm": 0.11251654475927353,
      "learning_rate": 4.4017798158528516e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 8038448,
      "step": 8995
    },
    {
      "epoch": 2.375346443183318,
      "grad_norm": 0.2353343665599823,
      "learning_rate": 4.401142426962834e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 8042928,
      "step": 9000
    },
    {
      "epoch": 2.375346443183318,
      "eval_loss": 0.07366812229156494,
      "eval_runtime": 64.7068,
      "eval_samples_per_second": 104.085,
      "eval_steps_per_second": 26.025,
      "num_input_tokens_seen": 8042928,
      "step": 9000
    },
    {
      "epoch": 2.376666226738815,
      "grad_norm": 0.458644300699234,
      "learning_rate": 4.400504744892763e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 8047472,
      "step": 9005
    },
    {
      "epoch": 2.3779860102943116,
      "grad_norm": 0.26188164949417114,
      "learning_rate": 4.399866769740975e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 8051792,
      "step": 9010
    },
    {
      "epoch": 2.3793057938498086,
      "grad_norm": 0.17202936112880707,
      "learning_rate": 4.399228501605859e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 8056432,
      "step": 9015
    },
    {
      "epoch": 2.3806255774053056,
      "grad_norm": 0.41313886642456055,
      "learning_rate": 4.398589940585839e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 8060784,
      "step": 9020
    },
    {
      "epoch": 2.3819453609608026,
      "grad_norm": 0.20247070491313934,
      "learning_rate": 4.3979510867793917e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 8065136,
      "step": 9025
    },
    {
      "epoch": 2.383265144516299,
      "grad_norm": 0.3748520612716675,
      "learning_rate": 4.3973119402850346e-05,
      "loss": 0.0865,
      "num_input_tokens_seen": 8069200,
      "step": 9030
    },
    {
      "epoch": 2.384584928071796,
      "grad_norm": 0.16561299562454224,
      "learning_rate": 4.396672501201334e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 8073744,
      "step": 9035
    },
    {
      "epoch": 2.385904711627293,
      "grad_norm": 0.17081308364868164,
      "learning_rate": 4.396032769626899e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 8078128,
      "step": 9040
    },
    {
      "epoch": 2.38722449518279,
      "grad_norm": 0.07910822331905365,
      "learning_rate": 4.395392745660384e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 8082544,
      "step": 9045
    },
    {
      "epoch": 2.388544278738287,
      "grad_norm": 0.33812302350997925,
      "learning_rate": 4.394752429400488e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 8086960,
      "step": 9050
    },
    {
      "epoch": 2.3898640622937837,
      "grad_norm": 0.341682106256485,
      "learning_rate": 4.394111820945957e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 8091504,
      "step": 9055
    },
    {
      "epoch": 2.3911838458492807,
      "grad_norm": 0.2222287654876709,
      "learning_rate": 4.393470920395579e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 8095632,
      "step": 9060
    },
    {
      "epoch": 2.3925036294047777,
      "grad_norm": 0.2893441617488861,
      "learning_rate": 4.392829727848192e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 8100048,
      "step": 9065
    },
    {
      "epoch": 2.3938234129602747,
      "grad_norm": 0.24630434811115265,
      "learning_rate": 4.392188243402673e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 8104752,
      "step": 9070
    },
    {
      "epoch": 2.3951431965157712,
      "grad_norm": 0.30344417691230774,
      "learning_rate": 4.391546467157949e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 8109200,
      "step": 9075
    },
    {
      "epoch": 2.3964629800712682,
      "grad_norm": 0.42786845564842224,
      "learning_rate": 4.390904399212988e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 8113712,
      "step": 9080
    },
    {
      "epoch": 2.3977827636267652,
      "grad_norm": 0.49901145696640015,
      "learning_rate": 4.390262039666807e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 8118256,
      "step": 9085
    },
    {
      "epoch": 2.3991025471822622,
      "grad_norm": 0.32027795910835266,
      "learning_rate": 4.389619388618464e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 8122768,
      "step": 9090
    },
    {
      "epoch": 2.400422330737759,
      "grad_norm": 0.43628621101379395,
      "learning_rate": 4.3889764461670655e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 8127216,
      "step": 9095
    },
    {
      "epoch": 2.401742114293256,
      "grad_norm": 0.2572704553604126,
      "learning_rate": 4.38833321241176e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 8131792,
      "step": 9100
    },
    {
      "epoch": 2.403061897848753,
      "grad_norm": 0.10141805559396744,
      "learning_rate": 4.3876896874517434e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 8136464,
      "step": 9105
    },
    {
      "epoch": 2.40438168140425,
      "grad_norm": 0.18329687416553497,
      "learning_rate": 4.3870458713862554e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 8141424,
      "step": 9110
    },
    {
      "epoch": 2.405701464959747,
      "grad_norm": 0.06782520562410355,
      "learning_rate": 4.386401764314579e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 8145904,
      "step": 9115
    },
    {
      "epoch": 2.4070212485152434,
      "grad_norm": 0.23425708711147308,
      "learning_rate": 4.385757366336045e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 8150480,
      "step": 9120
    },
    {
      "epoch": 2.4083410320707404,
      "grad_norm": 0.25675007700920105,
      "learning_rate": 4.385112677550027e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 8154448,
      "step": 9125
    },
    {
      "epoch": 2.4096608156262374,
      "grad_norm": 0.23761184513568878,
      "learning_rate": 4.384467698055945e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 8158832,
      "step": 9130
    },
    {
      "epoch": 2.4109805991817344,
      "grad_norm": 0.058695238083601,
      "learning_rate": 4.383822427953261e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 8163664,
      "step": 9135
    },
    {
      "epoch": 2.412300382737231,
      "grad_norm": 0.1232694461941719,
      "learning_rate": 4.3831768673414864e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 8168432,
      "step": 9140
    },
    {
      "epoch": 2.413620166292728,
      "grad_norm": 0.159526988863945,
      "learning_rate": 4.382531016320173e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 8173264,
      "step": 9145
    },
    {
      "epoch": 2.414939949848225,
      "grad_norm": 0.1258520632982254,
      "learning_rate": 4.3818848749889184e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 8177808,
      "step": 9150
    },
    {
      "epoch": 2.416259733403722,
      "grad_norm": 0.1542961299419403,
      "learning_rate": 4.381238443447368e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 8182512,
      "step": 9155
    },
    {
      "epoch": 2.4175795169592185,
      "grad_norm": 0.40725672245025635,
      "learning_rate": 4.380591721795208e-05,
      "loss": 0.0777,
      "num_input_tokens_seen": 8187152,
      "step": 9160
    },
    {
      "epoch": 2.4188993005147155,
      "grad_norm": 0.08722355961799622,
      "learning_rate": 4.3799447101321723e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 8191920,
      "step": 9165
    },
    {
      "epoch": 2.4202190840702125,
      "grad_norm": 0.2834886312484741,
      "learning_rate": 4.379297408558036e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 8196592,
      "step": 9170
    },
    {
      "epoch": 2.4215388676257095,
      "grad_norm": 0.035566672682762146,
      "learning_rate": 4.378649817172624e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 8201008,
      "step": 9175
    },
    {
      "epoch": 2.4228586511812065,
      "grad_norm": 0.2594910264015198,
      "learning_rate": 4.378001936075801e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 8205808,
      "step": 9180
    },
    {
      "epoch": 2.424178434736703,
      "grad_norm": 0.4560912251472473,
      "learning_rate": 4.377353765367479e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 8210352,
      "step": 9185
    },
    {
      "epoch": 2.4254982182922,
      "grad_norm": 0.4761858284473419,
      "learning_rate": 4.376705305147614e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 8214672,
      "step": 9190
    },
    {
      "epoch": 2.426818001847697,
      "grad_norm": 0.09230410307645798,
      "learning_rate": 4.376056555516206e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 8219184,
      "step": 9195
    },
    {
      "epoch": 2.428137785403194,
      "grad_norm": 0.14981140196323395,
      "learning_rate": 4.375407516573302e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 8223824,
      "step": 9200
    },
    {
      "epoch": 2.428137785403194,
      "eval_loss": 0.07391242682933807,
      "eval_runtime": 64.7357,
      "eval_samples_per_second": 104.038,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 8223824,
      "step": 9200
    },
    {
      "epoch": 2.4294575689586906,
      "grad_norm": 0.19377146661281586,
      "learning_rate": 4.3747581884189913e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 8228272,
      "step": 9205
    },
    {
      "epoch": 2.4307773525141876,
      "grad_norm": 0.3537294864654541,
      "learning_rate": 4.374108571153408e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 8233040,
      "step": 9210
    },
    {
      "epoch": 2.4320971360696846,
      "grad_norm": 0.26429229974746704,
      "learning_rate": 4.3734586648767316e-05,
      "loss": 0.1141,
      "num_input_tokens_seen": 8237424,
      "step": 9215
    },
    {
      "epoch": 2.4334169196251816,
      "grad_norm": 0.03198317810893059,
      "learning_rate": 4.372808469689186e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 8241968,
      "step": 9220
    },
    {
      "epoch": 2.434736703180678,
      "grad_norm": 0.22830308973789215,
      "learning_rate": 4.372157985691039e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 8246288,
      "step": 9225
    },
    {
      "epoch": 2.436056486736175,
      "grad_norm": 0.36861100792884827,
      "learning_rate": 4.371507212982603e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 8250832,
      "step": 9230
    },
    {
      "epoch": 2.437376270291672,
      "grad_norm": 0.17639894783496857,
      "learning_rate": 4.370856151664236e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 8255408,
      "step": 9235
    },
    {
      "epoch": 2.438696053847169,
      "grad_norm": 0.04036502167582512,
      "learning_rate": 4.3702048018363404e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 8259504,
      "step": 9240
    },
    {
      "epoch": 2.440015837402666,
      "grad_norm": 0.17586800456047058,
      "learning_rate": 4.369553163599362e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 8263984,
      "step": 9245
    },
    {
      "epoch": 2.4413356209581627,
      "grad_norm": 0.22516407072544098,
      "learning_rate": 4.3689012370537904e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 8268240,
      "step": 9250
    },
    {
      "epoch": 2.4426554045136597,
      "grad_norm": 0.15411430597305298,
      "learning_rate": 4.368249022300164e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 8272656,
      "step": 9255
    },
    {
      "epoch": 2.4439751880691567,
      "grad_norm": 0.07034123688936234,
      "learning_rate": 4.367596519439059e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 8276752,
      "step": 9260
    },
    {
      "epoch": 2.4452949716246537,
      "grad_norm": 0.27482086420059204,
      "learning_rate": 4.366943728571101e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 8281104,
      "step": 9265
    },
    {
      "epoch": 2.4466147551801507,
      "grad_norm": 0.1325758844614029,
      "learning_rate": 4.366290649796959e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 8285808,
      "step": 9270
    },
    {
      "epoch": 2.4479345387356473,
      "grad_norm": 0.026735510677099228,
      "learning_rate": 4.3656372832173456e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 8290256,
      "step": 9275
    },
    {
      "epoch": 2.4492543222911443,
      "grad_norm": 0.28094732761383057,
      "learning_rate": 4.364983628933017e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 8294832,
      "step": 9280
    },
    {
      "epoch": 2.4505741058466413,
      "grad_norm": 0.43103864789009094,
      "learning_rate": 4.364329687044777e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 8299088,
      "step": 9285
    },
    {
      "epoch": 2.451893889402138,
      "grad_norm": 0.10007283836603165,
      "learning_rate": 4.36367545765347e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 8303856,
      "step": 9290
    },
    {
      "epoch": 2.453213672957635,
      "grad_norm": 0.17797701060771942,
      "learning_rate": 4.363020940859988e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 8308336,
      "step": 9295
    },
    {
      "epoch": 2.454533456513132,
      "grad_norm": 0.2130151242017746,
      "learning_rate": 4.362366136765263e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 8312944,
      "step": 9300
    },
    {
      "epoch": 2.455853240068629,
      "grad_norm": 0.15052422881126404,
      "learning_rate": 4.361711045470278e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 8317424,
      "step": 9305
    },
    {
      "epoch": 2.457173023624126,
      "grad_norm": 0.1745966672897339,
      "learning_rate": 4.3610556670760524e-05,
      "loss": 0.1217,
      "num_input_tokens_seen": 8321840,
      "step": 9310
    },
    {
      "epoch": 2.4584928071796224,
      "grad_norm": 0.08775819838047028,
      "learning_rate": 4.360400001683657e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 8326480,
      "step": 9315
    },
    {
      "epoch": 2.4598125907351194,
      "grad_norm": 0.1912098079919815,
      "learning_rate": 4.3597440493942e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 8331024,
      "step": 9320
    },
    {
      "epoch": 2.4611323742906164,
      "grad_norm": 0.11903802305459976,
      "learning_rate": 4.3590878103088405e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 8335664,
      "step": 9325
    },
    {
      "epoch": 2.4624521578461134,
      "grad_norm": 0.09403789043426514,
      "learning_rate": 4.358431284528779e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 8340016,
      "step": 9330
    },
    {
      "epoch": 2.4637719414016104,
      "grad_norm": 0.0589064322412014,
      "learning_rate": 4.357774472155257e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 8344368,
      "step": 9335
    },
    {
      "epoch": 2.465091724957107,
      "grad_norm": 0.09838778525590897,
      "learning_rate": 4.3571173732895664e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 8348592,
      "step": 9340
    },
    {
      "epoch": 2.466411508512604,
      "grad_norm": 0.11654862016439438,
      "learning_rate": 4.356459988033039e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 8352880,
      "step": 9345
    },
    {
      "epoch": 2.467731292068101,
      "grad_norm": 0.2738012373447418,
      "learning_rate": 4.355802316487051e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 8357360,
      "step": 9350
    },
    {
      "epoch": 2.4690510756235975,
      "grad_norm": 0.6542018055915833,
      "learning_rate": 4.355144358753025e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 8361840,
      "step": 9355
    },
    {
      "epoch": 2.4703708591790945,
      "grad_norm": 0.05664927139878273,
      "learning_rate": 4.354486114932425e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 8366128,
      "step": 9360
    },
    {
      "epoch": 2.4716906427345915,
      "grad_norm": 0.028456710278987885,
      "learning_rate": 4.353827585126762e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 8370864,
      "step": 9365
    },
    {
      "epoch": 2.4730104262900885,
      "grad_norm": 0.12671597301959991,
      "learning_rate": 4.353168769437588e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 8375440,
      "step": 9370
    },
    {
      "epoch": 2.4743302098455855,
      "grad_norm": 0.20276258885860443,
      "learning_rate": 4.3525096679665014e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 8379856,
      "step": 9375
    },
    {
      "epoch": 2.475649993401082,
      "grad_norm": 0.2866382896900177,
      "learning_rate": 4.351850280815144e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 8384144,
      "step": 9380
    },
    {
      "epoch": 2.476969776956579,
      "grad_norm": 0.23208369314670563,
      "learning_rate": 4.3511906080852014e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 8388592,
      "step": 9385
    },
    {
      "epoch": 2.478289560512076,
      "grad_norm": 0.1869475096464157,
      "learning_rate": 4.350530649878404e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 8393136,
      "step": 9390
    },
    {
      "epoch": 2.479609344067573,
      "grad_norm": 0.24110490083694458,
      "learning_rate": 4.3498704062965246e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 8397648,
      "step": 9395
    },
    {
      "epoch": 2.48092912762307,
      "grad_norm": 0.027560560032725334,
      "learning_rate": 4.3492098774413815e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 8402448,
      "step": 9400
    },
    {
      "epoch": 2.48092912762307,
      "eval_loss": 0.07444201409816742,
      "eval_runtime": 64.7265,
      "eval_samples_per_second": 104.053,
      "eval_steps_per_second": 26.017,
      "num_input_tokens_seen": 8402448,
      "step": 9400
    },
    {
      "epoch": 2.4822489111785666,
      "grad_norm": 0.32648807764053345,
      "learning_rate": 4.3485490634148375e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 8406704,
      "step": 9405
    },
    {
      "epoch": 2.4835686947340636,
      "grad_norm": 0.3031865060329437,
      "learning_rate": 4.347887964318797e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 8411312,
      "step": 9410
    },
    {
      "epoch": 2.4848884782895606,
      "grad_norm": 0.16181257367134094,
      "learning_rate": 4.34722658025521e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 8415472,
      "step": 9415
    },
    {
      "epoch": 2.486208261845057,
      "grad_norm": 0.24239273369312286,
      "learning_rate": 4.346564911326071e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 8419952,
      "step": 9420
    },
    {
      "epoch": 2.487528045400554,
      "grad_norm": 0.3218967616558075,
      "learning_rate": 4.345902957633418e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 8424688,
      "step": 9425
    },
    {
      "epoch": 2.488847828956051,
      "grad_norm": 0.21964791417121887,
      "learning_rate": 4.345240719279331e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 8429168,
      "step": 9430
    },
    {
      "epoch": 2.490167612511548,
      "grad_norm": 0.3412536680698395,
      "learning_rate": 4.3445781963659374e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 8434032,
      "step": 9435
    },
    {
      "epoch": 2.491487396067045,
      "grad_norm": 0.3344704806804657,
      "learning_rate": 4.3439153889954045e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 8438256,
      "step": 9440
    },
    {
      "epoch": 2.4928071796225417,
      "grad_norm": 0.09304055571556091,
      "learning_rate": 4.343252297269946e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 8443152,
      "step": 9445
    },
    {
      "epoch": 2.4941269631780387,
      "grad_norm": 0.14669226109981537,
      "learning_rate": 4.342588921291821e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 8447728,
      "step": 9450
    },
    {
      "epoch": 2.4954467467335357,
      "grad_norm": 0.12506449222564697,
      "learning_rate": 4.341925261163328e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 8452208,
      "step": 9455
    },
    {
      "epoch": 2.4967665302890327,
      "grad_norm": 0.26568660140037537,
      "learning_rate": 4.341261316986813e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 8456912,
      "step": 9460
    },
    {
      "epoch": 2.4980863138445297,
      "grad_norm": 0.3317663073539734,
      "learning_rate": 4.340597088864664e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 8461136,
      "step": 9465
    },
    {
      "epoch": 2.4994060974000263,
      "grad_norm": 0.22316816449165344,
      "learning_rate": 4.339932576899313e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 8465296,
      "step": 9470
    },
    {
      "epoch": 2.5007258809555233,
      "grad_norm": 0.3642165958881378,
      "learning_rate": 4.3392677811932375e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 8469584,
      "step": 9475
    },
    {
      "epoch": 2.5020456645110203,
      "grad_norm": 0.22838711738586426,
      "learning_rate": 4.338602701848956e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 8474352,
      "step": 9480
    },
    {
      "epoch": 2.503365448066517,
      "grad_norm": 0.12118610739707947,
      "learning_rate": 4.337937338969033e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 8478768,
      "step": 9485
    },
    {
      "epoch": 2.504685231622014,
      "grad_norm": 0.535918116569519,
      "learning_rate": 4.337271692656075e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 8483120,
      "step": 9490
    },
    {
      "epoch": 2.506005015177511,
      "grad_norm": 0.05672985315322876,
      "learning_rate": 4.336605763012733e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 8487440,
      "step": 9495
    },
    {
      "epoch": 2.507324798733008,
      "grad_norm": 0.4549371302127838,
      "learning_rate": 4.3359395501417026e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 8491696,
      "step": 9500
    },
    {
      "epoch": 2.508644582288505,
      "grad_norm": 0.05192483216524124,
      "learning_rate": 4.335273054145722e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 8496176,
      "step": 9505
    },
    {
      "epoch": 2.5099643658440014,
      "grad_norm": 0.331556111574173,
      "learning_rate": 4.334606275127572e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 8501008,
      "step": 9510
    },
    {
      "epoch": 2.5112841493994984,
      "grad_norm": 0.18953146040439606,
      "learning_rate": 4.33393921319008e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 8505424,
      "step": 9515
    },
    {
      "epoch": 2.5126039329549954,
      "grad_norm": 0.13035956025123596,
      "learning_rate": 4.3332718684361146e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 8510064,
      "step": 9520
    },
    {
      "epoch": 2.5139237165104924,
      "grad_norm": 0.15347784757614136,
      "learning_rate": 4.332604240968588e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 8514384,
      "step": 9525
    },
    {
      "epoch": 2.5152435000659894,
      "grad_norm": 0.2641039788722992,
      "learning_rate": 4.331936330890459e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 8518864,
      "step": 9530
    },
    {
      "epoch": 2.516563283621486,
      "grad_norm": 0.25328975915908813,
      "learning_rate": 4.331268138304725e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 8523088,
      "step": 9535
    },
    {
      "epoch": 2.517883067176983,
      "grad_norm": 0.24716410040855408,
      "learning_rate": 4.330599663314431e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 8527568,
      "step": 9540
    },
    {
      "epoch": 2.51920285073248,
      "grad_norm": 0.4119676351547241,
      "learning_rate": 4.329930906022665e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 8531728,
      "step": 9545
    },
    {
      "epoch": 2.5205226342879765,
      "grad_norm": 0.1101699247956276,
      "learning_rate": 4.3292618665325564e-05,
      "loss": 0.0725,
      "num_input_tokens_seen": 8536688,
      "step": 9550
    },
    {
      "epoch": 2.5218424178434735,
      "grad_norm": 0.08412696421146393,
      "learning_rate": 4.3285925449472796e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 8541296,
      "step": 9555
    },
    {
      "epoch": 2.5231622013989705,
      "grad_norm": 0.31384074687957764,
      "learning_rate": 4.327922941370054e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 8545776,
      "step": 9560
    },
    {
      "epoch": 2.5244819849544675,
      "grad_norm": 0.27171245217323303,
      "learning_rate": 4.3272530559041384e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 8550288,
      "step": 9565
    },
    {
      "epoch": 2.5258017685099645,
      "grad_norm": 0.1897088885307312,
      "learning_rate": 4.32658288865284e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 8554704,
      "step": 9570
    },
    {
      "epoch": 2.527121552065461,
      "grad_norm": 0.49644559621810913,
      "learning_rate": 4.325912439719505e-05,
      "loss": 0.1047,
      "num_input_tokens_seen": 8559120,
      "step": 9575
    },
    {
      "epoch": 2.528441335620958,
      "grad_norm": 0.19385695457458496,
      "learning_rate": 4.3252417092075266e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 8563632,
      "step": 9580
    },
    {
      "epoch": 2.529761119176455,
      "grad_norm": 0.4870379865169525,
      "learning_rate": 4.3245706972203385e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 8568304,
      "step": 9585
    },
    {
      "epoch": 2.531080902731952,
      "grad_norm": 0.5403352975845337,
      "learning_rate": 4.323899403861421e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 8572976,
      "step": 9590
    },
    {
      "epoch": 2.532400686287449,
      "grad_norm": 0.13458994030952454,
      "learning_rate": 4.3232278292342935e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 8577200,
      "step": 9595
    },
    {
      "epoch": 2.5337204698429456,
      "grad_norm": 0.40126243233680725,
      "learning_rate": 4.322555973442524e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 8581936,
      "step": 9600
    },
    {
      "epoch": 2.5337204698429456,
      "eval_loss": 0.07336914539337158,
      "eval_runtime": 64.792,
      "eval_samples_per_second": 103.948,
      "eval_steps_per_second": 25.991,
      "num_input_tokens_seen": 8581936,
      "step": 9600
    },
    {
      "epoch": 2.5350402533984426,
      "grad_norm": 0.3336700201034546,
      "learning_rate": 4.3218838365897184e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 8586352,
      "step": 9605
    },
    {
      "epoch": 2.5363600369539396,
      "grad_norm": 0.23010782897472382,
      "learning_rate": 4.3212114187795306e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 8590640,
      "step": 9610
    },
    {
      "epoch": 2.537679820509436,
      "grad_norm": 0.18952596187591553,
      "learning_rate": 4.320538720115656e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 8595152,
      "step": 9615
    },
    {
      "epoch": 2.538999604064933,
      "grad_norm": 0.07515153288841248,
      "learning_rate": 4.319865740701831e-05,
      "loss": 0.1215,
      "num_input_tokens_seen": 8599344,
      "step": 9620
    },
    {
      "epoch": 2.54031938762043,
      "grad_norm": 0.11823656409978867,
      "learning_rate": 4.3191924806418396e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 8604112,
      "step": 9625
    },
    {
      "epoch": 2.541639171175927,
      "grad_norm": 0.08906187117099762,
      "learning_rate": 4.318518940039507e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 8608944,
      "step": 9630
    },
    {
      "epoch": 2.542958954731424,
      "grad_norm": 0.22747233510017395,
      "learning_rate": 4.3178451189987e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 8613552,
      "step": 9635
    },
    {
      "epoch": 2.5442787382869207,
      "grad_norm": 0.1265621781349182,
      "learning_rate": 4.3171710176233315e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 8618064,
      "step": 9640
    },
    {
      "epoch": 2.5455985218424177,
      "grad_norm": 0.3723251223564148,
      "learning_rate": 4.316496636017355e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 8622256,
      "step": 9645
    },
    {
      "epoch": 2.5469183053979148,
      "grad_norm": 0.11644499003887177,
      "learning_rate": 4.315821974284771e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 8626704,
      "step": 9650
    },
    {
      "epoch": 2.5482380889534118,
      "grad_norm": 0.4974287450313568,
      "learning_rate": 4.315147032529619e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 8631504,
      "step": 9655
    },
    {
      "epoch": 2.5495578725089088,
      "grad_norm": 0.37407687306404114,
      "learning_rate": 4.3144718108559845e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 8636016,
      "step": 9660
    },
    {
      "epoch": 2.5508776560644053,
      "grad_norm": 0.14029084146022797,
      "learning_rate": 4.3137963093679945e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 8640688,
      "step": 9665
    },
    {
      "epoch": 2.5521974396199023,
      "grad_norm": 0.3493426740169525,
      "learning_rate": 4.31312052816982e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 8644976,
      "step": 9670
    },
    {
      "epoch": 2.5535172231753993,
      "grad_norm": 0.20684921741485596,
      "learning_rate": 4.312444467365675e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 8649232,
      "step": 9675
    },
    {
      "epoch": 2.554837006730896,
      "grad_norm": 0.5462194681167603,
      "learning_rate": 4.311768127059816e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 8653744,
      "step": 9680
    },
    {
      "epoch": 2.556156790286393,
      "grad_norm": 0.1233847364783287,
      "learning_rate": 4.3110915073565444e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 8658384,
      "step": 9685
    },
    {
      "epoch": 2.55747657384189,
      "grad_norm": 0.13533592224121094,
      "learning_rate": 4.310414608360203e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 8662704,
      "step": 9690
    },
    {
      "epoch": 2.558796357397387,
      "grad_norm": 0.2691023051738739,
      "learning_rate": 4.309737430175177e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 8667216,
      "step": 9695
    },
    {
      "epoch": 2.560116140952884,
      "grad_norm": 0.20291408896446228,
      "learning_rate": 4.309059972905897e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 8671760,
      "step": 9700
    },
    {
      "epoch": 2.5614359245083804,
      "grad_norm": 0.03975309431552887,
      "learning_rate": 4.308382236656836e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 8676304,
      "step": 9705
    },
    {
      "epoch": 2.5627557080638774,
      "grad_norm": 0.3087990880012512,
      "learning_rate": 4.307704221532507e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 8680880,
      "step": 9710
    },
    {
      "epoch": 2.5640754916193744,
      "grad_norm": 0.22949576377868652,
      "learning_rate": 4.307025927637471e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 8685520,
      "step": 9715
    },
    {
      "epoch": 2.5653952751748714,
      "grad_norm": 0.27851688861846924,
      "learning_rate": 4.306347355076328e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 8690224,
      "step": 9720
    },
    {
      "epoch": 2.5667150587303684,
      "grad_norm": 0.13673822581768036,
      "learning_rate": 4.305668503953724e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 8694736,
      "step": 9725
    },
    {
      "epoch": 2.568034842285865,
      "grad_norm": 0.08895297348499298,
      "learning_rate": 4.3049893743743436e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 8699568,
      "step": 9730
    },
    {
      "epoch": 2.569354625841362,
      "grad_norm": 0.13384582102298737,
      "learning_rate": 4.304309966442919e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 8703920,
      "step": 9735
    },
    {
      "epoch": 2.570674409396859,
      "grad_norm": 0.24064113199710846,
      "learning_rate": 4.303630280264224e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 8708400,
      "step": 9740
    },
    {
      "epoch": 2.5719941929523555,
      "grad_norm": 0.03973736986517906,
      "learning_rate": 4.302950315943074e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 8713136,
      "step": 9745
    },
    {
      "epoch": 2.573313976507853,
      "grad_norm": 0.1885768324136734,
      "learning_rate": 4.3022700735843275e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 8717392,
      "step": 9750
    },
    {
      "epoch": 2.5746337600633495,
      "grad_norm": 0.3195228576660156,
      "learning_rate": 4.301589553292887e-05,
      "loss": 0.1086,
      "num_input_tokens_seen": 8721936,
      "step": 9755
    },
    {
      "epoch": 2.5759535436188465,
      "grad_norm": 0.3497832417488098,
      "learning_rate": 4.300908755173697e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 8726704,
      "step": 9760
    },
    {
      "epoch": 2.5772733271743435,
      "grad_norm": 0.3529016375541687,
      "learning_rate": 4.300227679331745e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 8731024,
      "step": 9765
    },
    {
      "epoch": 2.57859311072984,
      "grad_norm": 0.16406171023845673,
      "learning_rate": 4.299546325872063e-05,
      "loss": 0.0841,
      "num_input_tokens_seen": 8735504,
      "step": 9770
    },
    {
      "epoch": 2.579912894285337,
      "grad_norm": 0.14302510023117065,
      "learning_rate": 4.2988646948997225e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 8739760,
      "step": 9775
    },
    {
      "epoch": 2.581232677840834,
      "grad_norm": 0.198823481798172,
      "learning_rate": 4.29818278651984e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 8744208,
      "step": 9780
    },
    {
      "epoch": 2.582552461396331,
      "grad_norm": 0.32735323905944824,
      "learning_rate": 4.297500600837574e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 8748592,
      "step": 9785
    },
    {
      "epoch": 2.583872244951828,
      "grad_norm": 0.6087290644645691,
      "learning_rate": 4.2968181379581276e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 8753072,
      "step": 9790
    },
    {
      "epoch": 2.5851920285073247,
      "grad_norm": 0.18827207386493683,
      "learning_rate": 4.296135397986743e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 8757648,
      "step": 9795
    },
    {
      "epoch": 2.5865118120628217,
      "grad_norm": 0.24756808578968048,
      "learning_rate": 4.295452381028709e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 8762128,
      "step": 9800
    },
    {
      "epoch": 2.5865118120628217,
      "eval_loss": 0.07314550876617432,
      "eval_runtime": 64.7604,
      "eval_samples_per_second": 103.999,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 8762128,
      "step": 9800
    },
    {
      "epoch": 2.5878315956183187,
      "grad_norm": 0.31770050525665283,
      "learning_rate": 4.294769087189354e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 8766416,
      "step": 9805
    },
    {
      "epoch": 2.5891513791738157,
      "grad_norm": 0.44346433877944946,
      "learning_rate": 4.294085516574052e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 8770928,
      "step": 9810
    },
    {
      "epoch": 2.5904711627293127,
      "grad_norm": 0.1290636658668518,
      "learning_rate": 4.2934016692882176e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 8775248,
      "step": 9815
    },
    {
      "epoch": 2.591790946284809,
      "grad_norm": 0.24320003390312195,
      "learning_rate": 4.292717545437308e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 8779728,
      "step": 9820
    },
    {
      "epoch": 2.593110729840306,
      "grad_norm": 0.4271619915962219,
      "learning_rate": 4.292033145126825e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 8784208,
      "step": 9825
    },
    {
      "epoch": 2.594430513395803,
      "grad_norm": 0.07138428837060928,
      "learning_rate": 4.29134846846231e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 8788688,
      "step": 9830
    },
    {
      "epoch": 2.5957502969512998,
      "grad_norm": 0.19680219888687134,
      "learning_rate": 4.29066351554935e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 8793008,
      "step": 9835
    },
    {
      "epoch": 2.5970700805067968,
      "grad_norm": 0.1529364138841629,
      "learning_rate": 4.289978286493574e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 8797424,
      "step": 9840
    },
    {
      "epoch": 2.5983898640622938,
      "grad_norm": 0.334871381521225,
      "learning_rate": 4.28929278140065e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 8802192,
      "step": 9845
    },
    {
      "epoch": 2.5997096476177908,
      "grad_norm": 0.3508867621421814,
      "learning_rate": 4.288607000376295e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 8806576,
      "step": 9850
    },
    {
      "epoch": 2.6010294311732878,
      "grad_norm": 0.38575175404548645,
      "learning_rate": 4.2879209435262624e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 8810896,
      "step": 9855
    },
    {
      "epoch": 2.6023492147287843,
      "grad_norm": 0.20000772178173065,
      "learning_rate": 4.287234610956353e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 8815152,
      "step": 9860
    },
    {
      "epoch": 2.6036689982842813,
      "grad_norm": 0.15364022552967072,
      "learning_rate": 4.2865480027724056e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 8819440,
      "step": 9865
    },
    {
      "epoch": 2.6049887818397783,
      "grad_norm": 0.20096606016159058,
      "learning_rate": 4.285861119080306e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 8824048,
      "step": 9870
    },
    {
      "epoch": 2.6063085653952753,
      "grad_norm": 0.09594684094190598,
      "learning_rate": 4.2851739599859784e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 8828688,
      "step": 9875
    },
    {
      "epoch": 2.6076283489507723,
      "grad_norm": 0.24546343088150024,
      "learning_rate": 4.2844865255953934e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 8833360,
      "step": 9880
    },
    {
      "epoch": 2.608948132506269,
      "grad_norm": 0.3628990054130554,
      "learning_rate": 4.2837988160145605e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 8837744,
      "step": 9885
    },
    {
      "epoch": 2.610267916061766,
      "grad_norm": 0.2777470350265503,
      "learning_rate": 4.2831108313495336e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 8842256,
      "step": 9890
    },
    {
      "epoch": 2.611587699617263,
      "grad_norm": 0.36594706773757935,
      "learning_rate": 4.282422571706408e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 8846672,
      "step": 9895
    },
    {
      "epoch": 2.6129074831727594,
      "grad_norm": 0.10054849088191986,
      "learning_rate": 4.281734037191323e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 8851216,
      "step": 9900
    },
    {
      "epoch": 2.6142272667282564,
      "grad_norm": 0.3801535964012146,
      "learning_rate": 4.281045227910459e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 8855408,
      "step": 9905
    },
    {
      "epoch": 2.6155470502837534,
      "grad_norm": 0.08858103305101395,
      "learning_rate": 4.280356143970038e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 8859696,
      "step": 9910
    },
    {
      "epoch": 2.6168668338392505,
      "grad_norm": 0.188070610165596,
      "learning_rate": 4.279666785476327e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 8864176,
      "step": 9915
    },
    {
      "epoch": 2.6181866173947475,
      "grad_norm": 0.27322542667388916,
      "learning_rate": 4.2789771525356325e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 8868560,
      "step": 9920
    },
    {
      "epoch": 2.619506400950244,
      "grad_norm": 0.20771068334579468,
      "learning_rate": 4.2782872452543056e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 8873008,
      "step": 9925
    },
    {
      "epoch": 2.620826184505741,
      "grad_norm": 0.40944206714630127,
      "learning_rate": 4.2775970637387376e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 8877488,
      "step": 9930
    },
    {
      "epoch": 2.622145968061238,
      "grad_norm": 0.14011326432228088,
      "learning_rate": 4.276906608095363e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 8881680,
      "step": 9935
    },
    {
      "epoch": 2.623465751616735,
      "grad_norm": 0.0863652303814888,
      "learning_rate": 4.276215878430661e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 8886000,
      "step": 9940
    },
    {
      "epoch": 2.624785535172232,
      "grad_norm": 0.07602347433567047,
      "learning_rate": 4.275524874851149e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 8890480,
      "step": 9945
    },
    {
      "epoch": 2.6261053187277286,
      "grad_norm": 0.3011094033718109,
      "learning_rate": 4.274833597463388e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 8895088,
      "step": 9950
    },
    {
      "epoch": 2.6274251022832256,
      "grad_norm": 0.2365679293870926,
      "learning_rate": 4.2741420463739824e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 8899600,
      "step": 9955
    },
    {
      "epoch": 2.6287448858387226,
      "grad_norm": 0.2474764585494995,
      "learning_rate": 4.273450221689578e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 8903920,
      "step": 9960
    },
    {
      "epoch": 2.630064669394219,
      "grad_norm": 0.3711318373680115,
      "learning_rate": 4.272758123516863e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 8908464,
      "step": 9965
    },
    {
      "epoch": 2.631384452949716,
      "grad_norm": 0.22012832760810852,
      "learning_rate": 4.272065751962567e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 8912752,
      "step": 9970
    },
    {
      "epoch": 2.632704236505213,
      "grad_norm": 0.1438518613576889,
      "learning_rate": 4.271373107133464e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 8917360,
      "step": 9975
    },
    {
      "epoch": 2.63402402006071,
      "grad_norm": 0.07493764162063599,
      "learning_rate": 4.270680189136366e-05,
      "loss": 0.0982,
      "num_input_tokens_seen": 8921808,
      "step": 9980
    },
    {
      "epoch": 2.635343803616207,
      "grad_norm": 0.297380656003952,
      "learning_rate": 4.269986998078132e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 8926096,
      "step": 9985
    },
    {
      "epoch": 2.6366635871717037,
      "grad_norm": 0.19269724190235138,
      "learning_rate": 4.2692935340656595e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 8930800,
      "step": 9990
    },
    {
      "epoch": 2.6379833707272007,
      "grad_norm": 0.19477353990077972,
      "learning_rate": 4.26859979720589e-05,
      "loss": 0.1146,
      "num_input_tokens_seen": 8935152,
      "step": 9995
    },
    {
      "epoch": 2.6393031542826977,
      "grad_norm": 0.36513808369636536,
      "learning_rate": 4.267905787605806e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 8939600,
      "step": 10000
    },
    {
      "epoch": 2.6393031542826977,
      "eval_loss": 0.0731375515460968,
      "eval_runtime": 64.7622,
      "eval_samples_per_second": 103.996,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 8939600,
      "step": 10000
    },
    {
      "epoch": 2.6406229378381947,
      "grad_norm": 0.09440498799085617,
      "learning_rate": 4.267211505372433e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 8944016,
      "step": 10005
    },
    {
      "epoch": 2.6419427213936917,
      "grad_norm": 0.19922661781311035,
      "learning_rate": 4.266516950612837e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 8948080,
      "step": 10010
    },
    {
      "epoch": 2.6432625049491882,
      "grad_norm": 0.39145827293395996,
      "learning_rate": 4.265822123434128e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 8952400,
      "step": 10015
    },
    {
      "epoch": 2.6445822885046852,
      "grad_norm": 0.4069858491420746,
      "learning_rate": 4.265127023943457e-05,
      "loss": 0.1017,
      "num_input_tokens_seen": 8956912,
      "step": 10020
    },
    {
      "epoch": 2.6459020720601822,
      "grad_norm": 0.1430160403251648,
      "learning_rate": 4.2644316522480176e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 8961520,
      "step": 10025
    },
    {
      "epoch": 2.647221855615679,
      "grad_norm": 0.039314936846494675,
      "learning_rate": 4.263736008455044e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 8965680,
      "step": 10030
    },
    {
      "epoch": 2.648541639171176,
      "grad_norm": 0.2309594452381134,
      "learning_rate": 4.2630400926718125e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 8969776,
      "step": 10035
    },
    {
      "epoch": 2.649861422726673,
      "grad_norm": 0.47212281823158264,
      "learning_rate": 4.262343905005644e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 8974128,
      "step": 10040
    },
    {
      "epoch": 2.65118120628217,
      "grad_norm": 0.3808099031448364,
      "learning_rate": 4.261647445563897e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 8978320,
      "step": 10045
    },
    {
      "epoch": 2.652500989837667,
      "grad_norm": 0.33051803708076477,
      "learning_rate": 4.260950714453976e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 8982768,
      "step": 10050
    },
    {
      "epoch": 2.6538207733931634,
      "grad_norm": 0.16358062624931335,
      "learning_rate": 4.2602537117833266e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 8987536,
      "step": 10055
    },
    {
      "epoch": 2.6551405569486604,
      "grad_norm": 0.14569436013698578,
      "learning_rate": 4.259556437659433e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 8992144,
      "step": 10060
    },
    {
      "epoch": 2.6564603405041574,
      "grad_norm": 0.3683777451515198,
      "learning_rate": 4.258858892189825e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 8996752,
      "step": 10065
    },
    {
      "epoch": 2.6577801240596544,
      "grad_norm": 0.19314073026180267,
      "learning_rate": 4.2581610754820725e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 9001392,
      "step": 10070
    },
    {
      "epoch": 2.6590999076151514,
      "grad_norm": 0.2471987009048462,
      "learning_rate": 4.2574629876437876e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 9005936,
      "step": 10075
    },
    {
      "epoch": 2.660419691170648,
      "grad_norm": 0.3301261067390442,
      "learning_rate": 4.256764628782625e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 9010640,
      "step": 10080
    },
    {
      "epoch": 2.661739474726145,
      "grad_norm": 0.18878726661205292,
      "learning_rate": 4.256065999006279e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 9015120,
      "step": 10085
    },
    {
      "epoch": 2.663059258281642,
      "grad_norm": 0.1538233757019043,
      "learning_rate": 4.2553670984224885e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 9019440,
      "step": 10090
    },
    {
      "epoch": 2.6643790418371385,
      "grad_norm": 0.13958032429218292,
      "learning_rate": 4.254667927139032e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 9023696,
      "step": 10095
    },
    {
      "epoch": 2.6656988253926355,
      "grad_norm": 0.10961255431175232,
      "learning_rate": 4.2539684852637295e-05,
      "loss": 0.027,
      "num_input_tokens_seen": 9027888,
      "step": 10100
    },
    {
      "epoch": 2.6670186089481325,
      "grad_norm": 0.10923238098621368,
      "learning_rate": 4.253268772904446e-05,
      "loss": 0.0326,
      "num_input_tokens_seen": 9032368,
      "step": 10105
    },
    {
      "epoch": 2.6683383925036295,
      "grad_norm": 0.34692806005477905,
      "learning_rate": 4.252568790169085e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 9036784,
      "step": 10110
    },
    {
      "epoch": 2.6696581760591265,
      "grad_norm": 0.13993752002716064,
      "learning_rate": 4.251868537165592e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 9041264,
      "step": 10115
    },
    {
      "epoch": 2.670977959614623,
      "grad_norm": 0.3688141703605652,
      "learning_rate": 4.251168014001955e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 9045872,
      "step": 10120
    },
    {
      "epoch": 2.67229774317012,
      "grad_norm": 0.09629227221012115,
      "learning_rate": 4.250467220786204e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 9050320,
      "step": 10125
    },
    {
      "epoch": 2.673617526725617,
      "grad_norm": 0.17815323173999786,
      "learning_rate": 4.249766157626409e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 9054960,
      "step": 10130
    },
    {
      "epoch": 2.674937310281114,
      "grad_norm": 0.1155419647693634,
      "learning_rate": 4.249064824630684e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 9059632,
      "step": 10135
    },
    {
      "epoch": 2.676257093836611,
      "grad_norm": 0.31764718890190125,
      "learning_rate": 4.248363221907183e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 9064368,
      "step": 10140
    },
    {
      "epoch": 2.6775768773921076,
      "grad_norm": 0.13647539913654327,
      "learning_rate": 4.2476613495641026e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 9068912,
      "step": 10145
    },
    {
      "epoch": 2.6788966609476046,
      "grad_norm": 0.37233930826187134,
      "learning_rate": 4.246959207709679e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 9073424,
      "step": 10150
    },
    {
      "epoch": 2.6802164445031016,
      "grad_norm": 0.056413568556308746,
      "learning_rate": 4.246256796452192e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 9077680,
      "step": 10155
    },
    {
      "epoch": 2.681536228058598,
      "grad_norm": 0.2447153776884079,
      "learning_rate": 4.245554115899962e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 9081968,
      "step": 10160
    },
    {
      "epoch": 2.682856011614095,
      "grad_norm": 0.1946895271539688,
      "learning_rate": 4.2448511661613514e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 9086864,
      "step": 10165
    },
    {
      "epoch": 2.684175795169592,
      "grad_norm": 0.10948691517114639,
      "learning_rate": 4.2441479473447635e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 9091056,
      "step": 10170
    },
    {
      "epoch": 2.685495578725089,
      "grad_norm": 0.097696952521801,
      "learning_rate": 4.243444459558644e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 9095440,
      "step": 10175
    },
    {
      "epoch": 2.686815362280586,
      "grad_norm": 0.2656266987323761,
      "learning_rate": 4.24274070291148e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 9099856,
      "step": 10180
    },
    {
      "epoch": 2.6881351458360827,
      "grad_norm": 0.0735374167561531,
      "learning_rate": 4.242036677511798e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 9103888,
      "step": 10185
    },
    {
      "epoch": 2.6894549293915797,
      "grad_norm": 0.10560992360115051,
      "learning_rate": 4.241332383468169e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 9108496,
      "step": 10190
    },
    {
      "epoch": 2.6907747129470767,
      "grad_norm": 0.04510555788874626,
      "learning_rate": 4.2406278208892034e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 9112752,
      "step": 10195
    },
    {
      "epoch": 2.6920944965025737,
      "grad_norm": 0.345093697309494,
      "learning_rate": 4.2399229898835536e-05,
      "loss": 0.0992,
      "num_input_tokens_seen": 9117424,
      "step": 10200
    },
    {
      "epoch": 2.6920944965025737,
      "eval_loss": 0.07295124977827072,
      "eval_runtime": 64.7565,
      "eval_samples_per_second": 104.005,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 9117424,
      "step": 10200
    },
    {
      "epoch": 2.6934142800580707,
      "grad_norm": 0.23191717267036438,
      "learning_rate": 4.239217890559914e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 9121488,
      "step": 10205
    },
    {
      "epoch": 2.6947340636135673,
      "grad_norm": 0.09339532256126404,
      "learning_rate": 4.238512523027019e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 9126288,
      "step": 10210
    },
    {
      "epoch": 2.6960538471690643,
      "grad_norm": 0.029894517734646797,
      "learning_rate": 4.237806887393645e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 9130768,
      "step": 10215
    },
    {
      "epoch": 2.6973736307245613,
      "grad_norm": 0.043799567967653275,
      "learning_rate": 4.237100983768611e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 9135440,
      "step": 10220
    },
    {
      "epoch": 2.698693414280058,
      "grad_norm": 0.14154568314552307,
      "learning_rate": 4.2363948122607756e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 9139984,
      "step": 10225
    },
    {
      "epoch": 2.700013197835555,
      "grad_norm": 0.1434115171432495,
      "learning_rate": 4.235688372979039e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 9144688,
      "step": 10230
    },
    {
      "epoch": 2.701332981391052,
      "grad_norm": 0.022823425009846687,
      "learning_rate": 4.234981666032343e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 9149488,
      "step": 10235
    },
    {
      "epoch": 2.702652764946549,
      "grad_norm": 0.29769209027290344,
      "learning_rate": 4.2342746915296704e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 9154288,
      "step": 10240
    },
    {
      "epoch": 2.703972548502046,
      "grad_norm": 0.5064764022827148,
      "learning_rate": 4.233567449580047e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 9158960,
      "step": 10245
    },
    {
      "epoch": 2.7052923320575424,
      "grad_norm": 0.0794043019413948,
      "learning_rate": 4.232859940292537e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 9163536,
      "step": 10250
    },
    {
      "epoch": 2.7066121156130394,
      "grad_norm": 0.53033047914505,
      "learning_rate": 4.232152163776248e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 9168144,
      "step": 10255
    },
    {
      "epoch": 2.7079318991685364,
      "grad_norm": 0.1327446848154068,
      "learning_rate": 4.231444120140328e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 9172464,
      "step": 10260
    },
    {
      "epoch": 2.7092516827240334,
      "grad_norm": 0.23512215912342072,
      "learning_rate": 4.230735809493967e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 9177168,
      "step": 10265
    },
    {
      "epoch": 2.7105714662795304,
      "grad_norm": 0.3205755352973938,
      "learning_rate": 4.2300272319463926e-05,
      "loss": 0.1353,
      "num_input_tokens_seen": 9181616,
      "step": 10270
    },
    {
      "epoch": 2.711891249835027,
      "grad_norm": 0.2373356968164444,
      "learning_rate": 4.2293183876068786e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 9186160,
      "step": 10275
    },
    {
      "epoch": 2.713211033390524,
      "grad_norm": 0.25926467776298523,
      "learning_rate": 4.228609276584737e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 9190864,
      "step": 10280
    },
    {
      "epoch": 2.714530816946021,
      "grad_norm": 0.07786808907985687,
      "learning_rate": 4.227899898989323e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 9195600,
      "step": 10285
    },
    {
      "epoch": 2.7158506005015175,
      "grad_norm": 0.0353267639875412,
      "learning_rate": 4.2271902549300293e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 9199984,
      "step": 10290
    },
    {
      "epoch": 2.7171703840570145,
      "grad_norm": 0.16221003234386444,
      "learning_rate": 4.226480344516294e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 9204624,
      "step": 10295
    },
    {
      "epoch": 2.7184901676125115,
      "grad_norm": 0.07511204481124878,
      "learning_rate": 4.2257701678575925e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 9209072,
      "step": 10300
    },
    {
      "epoch": 2.7198099511680085,
      "grad_norm": 0.05817507579922676,
      "learning_rate": 4.225059725063444e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 9213776,
      "step": 10305
    },
    {
      "epoch": 2.7211297347235055,
      "grad_norm": 0.44674259424209595,
      "learning_rate": 4.2243490162434074e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 9218352,
      "step": 10310
    },
    {
      "epoch": 2.722449518279002,
      "grad_norm": 0.34451398253440857,
      "learning_rate": 4.223638041507083e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 9222768,
      "step": 10315
    },
    {
      "epoch": 2.723769301834499,
      "grad_norm": 0.47843116521835327,
      "learning_rate": 4.2229268009641124e-05,
      "loss": 0.1112,
      "num_input_tokens_seen": 9227344,
      "step": 10320
    },
    {
      "epoch": 2.725089085389996,
      "grad_norm": 0.20910105109214783,
      "learning_rate": 4.222215294724177e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 9231440,
      "step": 10325
    },
    {
      "epoch": 2.726408868945493,
      "grad_norm": 0.09851108491420746,
      "learning_rate": 4.2215035228970005e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 9235632,
      "step": 10330
    },
    {
      "epoch": 2.72772865250099,
      "grad_norm": 0.21298708021640778,
      "learning_rate": 4.2207914855923464e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 9240016,
      "step": 10335
    },
    {
      "epoch": 2.7290484360564866,
      "grad_norm": 0.19436372816562653,
      "learning_rate": 4.220079182920021e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 9244560,
      "step": 10340
    },
    {
      "epoch": 2.7303682196119836,
      "grad_norm": 0.3786255121231079,
      "learning_rate": 4.2193666149898705e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 9248912,
      "step": 10345
    },
    {
      "epoch": 2.7316880031674806,
      "grad_norm": 0.13327337801456451,
      "learning_rate": 4.21865378191178e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 9253648,
      "step": 10350
    },
    {
      "epoch": 2.733007786722977,
      "grad_norm": 0.24992167949676514,
      "learning_rate": 4.217940683795678e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 9258416,
      "step": 10355
    },
    {
      "epoch": 2.734327570278474,
      "grad_norm": 0.11601991206407547,
      "learning_rate": 4.217227320751534e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 9262736,
      "step": 10360
    },
    {
      "epoch": 2.735647353833971,
      "grad_norm": 0.3245983123779297,
      "learning_rate": 4.216513692889358e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 9267760,
      "step": 10365
    },
    {
      "epoch": 2.736967137389468,
      "grad_norm": 0.5606169700622559,
      "learning_rate": 4.215799800319199e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 9272240,
      "step": 10370
    },
    {
      "epoch": 2.738286920944965,
      "grad_norm": 0.10975401103496552,
      "learning_rate": 4.2150856431511485e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 9276720,
      "step": 10375
    },
    {
      "epoch": 2.7396067045004617,
      "grad_norm": 0.21813541650772095,
      "learning_rate": 4.214371221495339e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 9281424,
      "step": 10380
    },
    {
      "epoch": 2.7409264880559587,
      "grad_norm": 0.06182369589805603,
      "learning_rate": 4.213656535461942e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 9285840,
      "step": 10385
    },
    {
      "epoch": 2.7422462716114557,
      "grad_norm": 0.2782314419746399,
      "learning_rate": 4.2129415851611734e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 9290160,
      "step": 10390
    },
    {
      "epoch": 2.7435660551669527,
      "grad_norm": 0.3484974503517151,
      "learning_rate": 4.2122263707032855e-05,
      "loss": 0.1228,
      "num_input_tokens_seen": 9294576,
      "step": 10395
    },
    {
      "epoch": 2.7448858387224497,
      "grad_norm": 0.1426055133342743,
      "learning_rate": 4.211510892198574e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 9299120,
      "step": 10400
    },
    {
      "epoch": 2.7448858387224497,
      "eval_loss": 0.07247276604175568,
      "eval_runtime": 64.8426,
      "eval_samples_per_second": 103.867,
      "eval_steps_per_second": 25.971,
      "num_input_tokens_seen": 9299120,
      "step": 10400
    },
    {
      "epoch": 2.7462056222779463,
      "grad_norm": 0.27401527762413025,
      "learning_rate": 4.210795149757375e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 9303952,
      "step": 10405
    },
    {
      "epoch": 2.7475254058334433,
      "grad_norm": 0.0694328099489212,
      "learning_rate": 4.210079143490065e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 9308368,
      "step": 10410
    },
    {
      "epoch": 2.7488451893889403,
      "grad_norm": 0.21915288269519806,
      "learning_rate": 4.2093628735070604e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 9312432,
      "step": 10415
    },
    {
      "epoch": 2.750164972944437,
      "grad_norm": 0.33886176347732544,
      "learning_rate": 4.208646339918819e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 9316784,
      "step": 10420
    },
    {
      "epoch": 2.751484756499934,
      "grad_norm": 0.28061583638191223,
      "learning_rate": 4.2079295428358414e-05,
      "loss": 0.077,
      "num_input_tokens_seen": 9321488,
      "step": 10425
    },
    {
      "epoch": 2.752804540055431,
      "grad_norm": 0.23319807648658752,
      "learning_rate": 4.207212482368664e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 9325936,
      "step": 10430
    },
    {
      "epoch": 2.754124323610928,
      "grad_norm": 0.18096014857292175,
      "learning_rate": 4.206495158627867e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 9330256,
      "step": 10435
    },
    {
      "epoch": 2.755444107166425,
      "grad_norm": 0.2003723680973053,
      "learning_rate": 4.205777571724073e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 9334832,
      "step": 10440
    },
    {
      "epoch": 2.7567638907219214,
      "grad_norm": 0.04375162720680237,
      "learning_rate": 4.20505972176794e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 9339504,
      "step": 10445
    },
    {
      "epoch": 2.7580836742774184,
      "grad_norm": 0.21019849181175232,
      "learning_rate": 4.204341608870171e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 9343920,
      "step": 10450
    },
    {
      "epoch": 2.7594034578329154,
      "grad_norm": 0.12451523542404175,
      "learning_rate": 4.203623233141508e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 9348528,
      "step": 10455
    },
    {
      "epoch": 2.7607232413884124,
      "grad_norm": 0.12882545590400696,
      "learning_rate": 4.2029045946927334e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 9353456,
      "step": 10460
    },
    {
      "epoch": 2.7620430249439094,
      "grad_norm": 0.2153378427028656,
      "learning_rate": 4.20218569363467e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 9357936,
      "step": 10465
    },
    {
      "epoch": 2.763362808499406,
      "grad_norm": 0.21429449319839478,
      "learning_rate": 4.2014665300781834e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 9362416,
      "step": 10470
    },
    {
      "epoch": 2.764682592054903,
      "grad_norm": 0.15040379762649536,
      "learning_rate": 4.200747104134174e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 9366736,
      "step": 10475
    },
    {
      "epoch": 2.7660023756104,
      "grad_norm": 0.3279176354408264,
      "learning_rate": 4.200027415913588e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 9371024,
      "step": 10480
    },
    {
      "epoch": 2.7673221591658965,
      "grad_norm": 0.08205234259366989,
      "learning_rate": 4.1993074655274126e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 9375728,
      "step": 10485
    },
    {
      "epoch": 2.768641942721394,
      "grad_norm": 0.10314976423978806,
      "learning_rate": 4.198587253086669e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 9379920,
      "step": 10490
    },
    {
      "epoch": 2.7699617262768905,
      "grad_norm": 0.3758162558078766,
      "learning_rate": 4.197866778702426e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 9384560,
      "step": 10495
    },
    {
      "epoch": 2.7712815098323875,
      "grad_norm": 0.2838219106197357,
      "learning_rate": 4.197146042485789e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 9388816,
      "step": 10500
    },
    {
      "epoch": 2.7726012933878845,
      "grad_norm": 0.15853995084762573,
      "learning_rate": 4.1964250445479046e-05,
      "loss": 0.1027,
      "num_input_tokens_seen": 9393584,
      "step": 10505
    },
    {
      "epoch": 2.773921076943381,
      "grad_norm": 0.21835008263587952,
      "learning_rate": 4.19570378499996e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 9397968,
      "step": 10510
    },
    {
      "epoch": 2.775240860498878,
      "grad_norm": 0.14547650516033173,
      "learning_rate": 4.194982263953182e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 9402768,
      "step": 10515
    },
    {
      "epoch": 2.776560644054375,
      "grad_norm": 0.2022438496351242,
      "learning_rate": 4.194260481518838e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 9407248,
      "step": 10520
    },
    {
      "epoch": 2.777880427609872,
      "grad_norm": 0.1448986828327179,
      "learning_rate": 4.1935384378082366e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 9411600,
      "step": 10525
    },
    {
      "epoch": 2.779200211165369,
      "grad_norm": 0.4270556569099426,
      "learning_rate": 4.1928161329327267e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 9415952,
      "step": 10530
    },
    {
      "epoch": 2.7805199947208656,
      "grad_norm": 0.4278147518634796,
      "learning_rate": 4.1920935670036945e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 9420848,
      "step": 10535
    },
    {
      "epoch": 2.7818397782763626,
      "grad_norm": 0.23939765989780426,
      "learning_rate": 4.1913707401325705e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 9425072,
      "step": 10540
    },
    {
      "epoch": 2.7831595618318596,
      "grad_norm": 0.18642324209213257,
      "learning_rate": 4.1906476524308235e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 9429008,
      "step": 10545
    },
    {
      "epoch": 2.784479345387356,
      "grad_norm": 0.26902902126312256,
      "learning_rate": 4.189924304009962e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 9433456,
      "step": 10550
    },
    {
      "epoch": 2.7857991289428536,
      "grad_norm": 0.33873802423477173,
      "learning_rate": 4.189200694981537e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 9437680,
      "step": 10555
    },
    {
      "epoch": 2.78711891249835,
      "grad_norm": 0.13564440608024597,
      "learning_rate": 4.188476825457136e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 9442448,
      "step": 10560
    },
    {
      "epoch": 2.788438696053847,
      "grad_norm": 0.11616267263889313,
      "learning_rate": 4.18775269554839e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 9446704,
      "step": 10565
    },
    {
      "epoch": 2.789758479609344,
      "grad_norm": 0.39329490065574646,
      "learning_rate": 4.187028305366969e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 9451312,
      "step": 10570
    },
    {
      "epoch": 2.7910782631648408,
      "grad_norm": 0.2904492914676666,
      "learning_rate": 4.1863036550245824e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 9456016,
      "step": 10575
    },
    {
      "epoch": 2.7923980467203378,
      "grad_norm": 0.26169899106025696,
      "learning_rate": 4.1855787446329806e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 9460368,
      "step": 10580
    },
    {
      "epoch": 2.7937178302758348,
      "grad_norm": 0.20216087996959686,
      "learning_rate": 4.184853574303955e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 9464816,
      "step": 10585
    },
    {
      "epoch": 2.7950376138313318,
      "grad_norm": 0.1291237324476242,
      "learning_rate": 4.184128144149334e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 9469200,
      "step": 10590
    },
    {
      "epoch": 2.7963573973868288,
      "grad_norm": 0.5450581908226013,
      "learning_rate": 4.1834024542809896e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 9473872,
      "step": 10595
    },
    {
      "epoch": 2.7976771809423253,
      "grad_norm": 0.15312978625297546,
      "learning_rate": 4.1826765048108315e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 9477968,
      "step": 10600
    },
    {
      "epoch": 2.7976771809423253,
      "eval_loss": 0.07258456945419312,
      "eval_runtime": 64.7758,
      "eval_samples_per_second": 103.974,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 9477968,
      "step": 10600
    },
    {
      "epoch": 2.7989969644978223,
      "grad_norm": 0.23226916790008545,
      "learning_rate": 4.181950295850811e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 9482288,
      "step": 10605
    },
    {
      "epoch": 2.8003167480533193,
      "grad_norm": 0.2744671404361725,
      "learning_rate": 4.181223827512918e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 9486704,
      "step": 10610
    },
    {
      "epoch": 2.8016365316088163,
      "grad_norm": 0.11475303769111633,
      "learning_rate": 4.180497099909183e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 9491440,
      "step": 10615
    },
    {
      "epoch": 2.8029563151643133,
      "grad_norm": 0.4354880154132843,
      "learning_rate": 4.179770113151677e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 9496304,
      "step": 10620
    },
    {
      "epoch": 2.80427609871981,
      "grad_norm": 0.20914198458194733,
      "learning_rate": 4.179042867352511e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 9500528,
      "step": 10625
    },
    {
      "epoch": 2.805595882275307,
      "grad_norm": 0.22657379508018494,
      "learning_rate": 4.1783153626238334e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 9504944,
      "step": 10630
    },
    {
      "epoch": 2.806915665830804,
      "grad_norm": 0.24687549471855164,
      "learning_rate": 4.177587599077836e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 9509264,
      "step": 10635
    },
    {
      "epoch": 2.8082354493863004,
      "grad_norm": 0.3574235141277313,
      "learning_rate": 4.1768595768267494e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 9513744,
      "step": 10640
    },
    {
      "epoch": 2.8095552329417974,
      "grad_norm": 0.08993898332118988,
      "learning_rate": 4.176131295982843e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 9518352,
      "step": 10645
    },
    {
      "epoch": 2.8108750164972944,
      "grad_norm": 0.10981207340955734,
      "learning_rate": 4.1754027566584276e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 9522896,
      "step": 10650
    },
    {
      "epoch": 2.8121948000527914,
      "grad_norm": 0.40158671140670776,
      "learning_rate": 4.174673958965852e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 9527440,
      "step": 10655
    },
    {
      "epoch": 2.8135145836082884,
      "grad_norm": 0.1160324439406395,
      "learning_rate": 4.173944903017507e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 9531888,
      "step": 10660
    },
    {
      "epoch": 2.814834367163785,
      "grad_norm": 0.09010399878025055,
      "learning_rate": 4.173215588925822e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 9536496,
      "step": 10665
    },
    {
      "epoch": 2.816154150719282,
      "grad_norm": 0.06260544061660767,
      "learning_rate": 4.172486016803266e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 9540880,
      "step": 10670
    },
    {
      "epoch": 2.817473934274779,
      "grad_norm": 0.17468859255313873,
      "learning_rate": 4.171756186762349e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 9545520,
      "step": 10675
    },
    {
      "epoch": 2.818793717830276,
      "grad_norm": 0.1733817756175995,
      "learning_rate": 4.171026098915619e-05,
      "loss": 0.0201,
      "num_input_tokens_seen": 9549872,
      "step": 10680
    },
    {
      "epoch": 2.820113501385773,
      "grad_norm": 0.10742655396461487,
      "learning_rate": 4.170295753375665e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 9554096,
      "step": 10685
    },
    {
      "epoch": 2.8214332849412695,
      "grad_norm": 0.14247892796993256,
      "learning_rate": 4.169565150255117e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 9558448,
      "step": 10690
    },
    {
      "epoch": 2.8227530684967665,
      "grad_norm": 0.04745020344853401,
      "learning_rate": 4.16883428966664e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 9563216,
      "step": 10695
    },
    {
      "epoch": 2.8240728520522635,
      "grad_norm": 0.26827290654182434,
      "learning_rate": 4.168103171722944e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 9567696,
      "step": 10700
    },
    {
      "epoch": 2.82539263560776,
      "grad_norm": 0.3079434037208557,
      "learning_rate": 4.167371796536777e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 9572016,
      "step": 10705
    },
    {
      "epoch": 2.826712419163257,
      "grad_norm": 0.40891391038894653,
      "learning_rate": 4.166640164220924e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 9576304,
      "step": 10710
    },
    {
      "epoch": 2.828032202718754,
      "grad_norm": 0.3849615752696991,
      "learning_rate": 4.1659082748882144e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 9581136,
      "step": 10715
    },
    {
      "epoch": 2.829351986274251,
      "grad_norm": 0.36571359634399414,
      "learning_rate": 4.1651761286515135e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 9585744,
      "step": 10720
    },
    {
      "epoch": 2.830671769829748,
      "grad_norm": 0.25078001618385315,
      "learning_rate": 4.164443725623728e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 9590064,
      "step": 10725
    },
    {
      "epoch": 2.8319915533852447,
      "grad_norm": 0.09294269233942032,
      "learning_rate": 4.163711065917802e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 9594640,
      "step": 10730
    },
    {
      "epoch": 2.8333113369407417,
      "grad_norm": 0.06363637000322342,
      "learning_rate": 4.1629781496467234e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 9599280,
      "step": 10735
    },
    {
      "epoch": 2.8346311204962387,
      "grad_norm": 0.08495685458183289,
      "learning_rate": 4.1622449769235164e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 9603952,
      "step": 10740
    },
    {
      "epoch": 2.8359509040517357,
      "grad_norm": 0.09065207093954086,
      "learning_rate": 4.161511547861243e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 9608752,
      "step": 10745
    },
    {
      "epoch": 2.8372706876072327,
      "grad_norm": 0.16980881989002228,
      "learning_rate": 4.1607778625730104e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 9613520,
      "step": 10750
    },
    {
      "epoch": 2.8385904711627292,
      "grad_norm": 0.28378695249557495,
      "learning_rate": 4.160043921171961e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 9617872,
      "step": 10755
    },
    {
      "epoch": 2.8399102547182262,
      "grad_norm": 0.45464298129081726,
      "learning_rate": 4.159309723771276e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 9622224,
      "step": 10760
    },
    {
      "epoch": 2.8412300382737232,
      "grad_norm": 0.34361666440963745,
      "learning_rate": 4.158575270484181e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 9626480,
      "step": 10765
    },
    {
      "epoch": 2.84254982182922,
      "grad_norm": 0.17745482921600342,
      "learning_rate": 4.157840561423936e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 9631312,
      "step": 10770
    },
    {
      "epoch": 2.843869605384717,
      "grad_norm": 0.2776324450969696,
      "learning_rate": 4.1571055967038416e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 9636016,
      "step": 10775
    },
    {
      "epoch": 2.845189388940214,
      "grad_norm": 0.19971998035907745,
      "learning_rate": 4.156370376437241e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 9640368,
      "step": 10780
    },
    {
      "epoch": 2.846509172495711,
      "grad_norm": 0.23342804610729218,
      "learning_rate": 4.155634900737513e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 9644912,
      "step": 10785
    },
    {
      "epoch": 2.847828956051208,
      "grad_norm": 0.08116573840379715,
      "learning_rate": 4.1548991697180764e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 9649264,
      "step": 10790
    },
    {
      "epoch": 2.8491487396067043,
      "grad_norm": 0.05584760755300522,
      "learning_rate": 4.1541631834923914e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 9653712,
      "step": 10795
    },
    {
      "epoch": 2.8504685231622013,
      "grad_norm": 0.18666653335094452,
      "learning_rate": 4.153426942173956e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 9658192,
      "step": 10800
    },
    {
      "epoch": 2.8504685231622013,
      "eval_loss": 0.07364362478256226,
      "eval_runtime": 64.767,
      "eval_samples_per_second": 103.988,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 9658192,
      "step": 10800
    },
    {
      "epoch": 2.8517883067176983,
      "grad_norm": 0.28547418117523193,
      "learning_rate": 4.152690445876308e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 9662672,
      "step": 10805
    },
    {
      "epoch": 2.8531080902731953,
      "grad_norm": 0.45269665122032166,
      "learning_rate": 4.1519536947130245e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 9667312,
      "step": 10810
    },
    {
      "epoch": 2.8544278738286923,
      "grad_norm": 0.1670391857624054,
      "learning_rate": 4.151216688797722e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 9671920,
      "step": 10815
    },
    {
      "epoch": 2.855747657384189,
      "grad_norm": 0.14056283235549927,
      "learning_rate": 4.150479428244054e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 9676368,
      "step": 10820
    },
    {
      "epoch": 2.857067440939686,
      "grad_norm": 0.5301304459571838,
      "learning_rate": 4.1497419131657176e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 9680944,
      "step": 10825
    },
    {
      "epoch": 2.858387224495183,
      "grad_norm": 0.606558084487915,
      "learning_rate": 4.149004143676447e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 9685296,
      "step": 10830
    },
    {
      "epoch": 2.8597070080506795,
      "grad_norm": 0.13543011248111725,
      "learning_rate": 4.148266119890015e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 9689904,
      "step": 10835
    },
    {
      "epoch": 2.8610267916061765,
      "grad_norm": 0.10696174949407578,
      "learning_rate": 4.1475278419202324e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 9694288,
      "step": 10840
    },
    {
      "epoch": 2.8623465751616735,
      "grad_norm": 0.3865920901298523,
      "learning_rate": 4.146789309880953e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 9699152,
      "step": 10845
    },
    {
      "epoch": 2.8636663587171705,
      "grad_norm": 0.21165058016777039,
      "learning_rate": 4.146050523886068e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 9703440,
      "step": 10850
    },
    {
      "epoch": 2.8649861422726675,
      "grad_norm": 0.4811539947986603,
      "learning_rate": 4.1453114840495055e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 9707984,
      "step": 10855
    },
    {
      "epoch": 2.866305925828164,
      "grad_norm": 0.08387497812509537,
      "learning_rate": 4.1445721904852364e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 9712496,
      "step": 10860
    },
    {
      "epoch": 2.867625709383661,
      "grad_norm": 0.22065885365009308,
      "learning_rate": 4.143832643307269e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 9716912,
      "step": 10865
    },
    {
      "epoch": 2.868945492939158,
      "grad_norm": 0.10948526114225388,
      "learning_rate": 4.1430928426296503e-05,
      "loss": 0.0156,
      "num_input_tokens_seen": 9721520,
      "step": 10870
    },
    {
      "epoch": 2.870265276494655,
      "grad_norm": 0.1257033497095108,
      "learning_rate": 4.142352788566466e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 9725904,
      "step": 10875
    },
    {
      "epoch": 2.871585060050152,
      "grad_norm": 0.4825247824192047,
      "learning_rate": 4.1416124812318424e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 9730480,
      "step": 10880
    },
    {
      "epoch": 2.8729048436056486,
      "grad_norm": 0.2748614549636841,
      "learning_rate": 4.1408719207399453e-05,
      "loss": 0.1398,
      "num_input_tokens_seen": 9735088,
      "step": 10885
    },
    {
      "epoch": 2.8742246271611456,
      "grad_norm": 0.155390664935112,
      "learning_rate": 4.140131107204978e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 9739536,
      "step": 10890
    },
    {
      "epoch": 2.8755444107166426,
      "grad_norm": 0.5951839685440063,
      "learning_rate": 4.139390040741182e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 9743792,
      "step": 10895
    },
    {
      "epoch": 2.876864194272139,
      "grad_norm": 0.06289207190275192,
      "learning_rate": 4.1386487214628396e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 9748304,
      "step": 10900
    },
    {
      "epoch": 2.878183977827636,
      "grad_norm": 0.17405596375465393,
      "learning_rate": 4.137907149484272e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 9752528,
      "step": 10905
    },
    {
      "epoch": 2.879503761383133,
      "grad_norm": 0.043563876301050186,
      "learning_rate": 4.137165324919839e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 9757232,
      "step": 10910
    },
    {
      "epoch": 2.88082354493863,
      "grad_norm": 0.4010505676269531,
      "learning_rate": 4.136423247883939e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 9761680,
      "step": 10915
    },
    {
      "epoch": 2.882143328494127,
      "grad_norm": 0.1429624706506729,
      "learning_rate": 4.135680918491009e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 9766064,
      "step": 10920
    },
    {
      "epoch": 2.8834631120496237,
      "grad_norm": 0.26946377754211426,
      "learning_rate": 4.1349383368555265e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 9770768,
      "step": 10925
    },
    {
      "epoch": 2.8847828956051207,
      "grad_norm": 0.1430450975894928,
      "learning_rate": 4.1341955030920065e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 9775216,
      "step": 10930
    },
    {
      "epoch": 2.8861026791606177,
      "grad_norm": 0.15326353907585144,
      "learning_rate": 4.1334524173150036e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 9779504,
      "step": 10935
    },
    {
      "epoch": 2.8874224627161147,
      "grad_norm": 0.2492009699344635,
      "learning_rate": 4.13270907963911e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 9783696,
      "step": 10940
    },
    {
      "epoch": 2.8887422462716117,
      "grad_norm": 0.42271658778190613,
      "learning_rate": 4.131965490178959e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 9788368,
      "step": 10945
    },
    {
      "epoch": 2.8900620298271082,
      "grad_norm": 0.08454377204179764,
      "learning_rate": 4.131221649049222e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 9792848,
      "step": 10950
    },
    {
      "epoch": 2.8913818133826052,
      "grad_norm": 0.36784616112709045,
      "learning_rate": 4.130477556364606e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 9797136,
      "step": 10955
    },
    {
      "epoch": 2.8927015969381022,
      "grad_norm": 0.19283534586429596,
      "learning_rate": 4.129733212239861e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 9801648,
      "step": 10960
    },
    {
      "epoch": 2.894021380493599,
      "grad_norm": 0.42087846994400024,
      "learning_rate": 4.128988616789774e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 9806224,
      "step": 10965
    },
    {
      "epoch": 2.895341164049096,
      "grad_norm": 0.29365837574005127,
      "learning_rate": 4.1282437701291724e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 9810768,
      "step": 10970
    },
    {
      "epoch": 2.896660947604593,
      "grad_norm": 0.1813187450170517,
      "learning_rate": 4.1274986723729184e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 9815568,
      "step": 10975
    },
    {
      "epoch": 2.89798073116009,
      "grad_norm": 0.12784017622470856,
      "learning_rate": 4.126753323635917e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 9820112,
      "step": 10980
    },
    {
      "epoch": 2.899300514715587,
      "grad_norm": 0.4426461160182953,
      "learning_rate": 4.12600772403311e-05,
      "loss": 0.0919,
      "num_input_tokens_seen": 9824272,
      "step": 10985
    },
    {
      "epoch": 2.9006202982710834,
      "grad_norm": 0.03414088487625122,
      "learning_rate": 4.125261873679479e-05,
      "loss": 0.1085,
      "num_input_tokens_seen": 9828368,
      "step": 10990
    },
    {
      "epoch": 2.9019400818265804,
      "grad_norm": 0.36862286925315857,
      "learning_rate": 4.124515772690042e-05,
      "loss": 0.1163,
      "num_input_tokens_seen": 9833008,
      "step": 10995
    },
    {
      "epoch": 2.9032598653820774,
      "grad_norm": 0.31738120317459106,
      "learning_rate": 4.123769421179858e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 9837392,
      "step": 11000
    },
    {
      "epoch": 2.9032598653820774,
      "eval_loss": 0.0721987709403038,
      "eval_runtime": 64.7228,
      "eval_samples_per_second": 104.059,
      "eval_steps_per_second": 26.019,
      "num_input_tokens_seen": 9837392,
      "step": 11000
    },
    {
      "epoch": 2.9045796489375744,
      "grad_norm": 0.08383646607398987,
      "learning_rate": 4.1230228192640236e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 9841744,
      "step": 11005
    },
    {
      "epoch": 2.9058994324930714,
      "grad_norm": 0.1444535255432129,
      "learning_rate": 4.122275967057675e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 9846224,
      "step": 11010
    },
    {
      "epoch": 2.907219216048568,
      "grad_norm": 0.3993277847766876,
      "learning_rate": 4.1215288646759846e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 9850416,
      "step": 11015
    },
    {
      "epoch": 2.908538999604065,
      "grad_norm": 0.11769595742225647,
      "learning_rate": 4.120781512234166e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 9854800,
      "step": 11020
    },
    {
      "epoch": 2.909858783159562,
      "grad_norm": 0.2832094132900238,
      "learning_rate": 4.120033909847471e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 9859440,
      "step": 11025
    },
    {
      "epoch": 2.9111785667150585,
      "grad_norm": 0.34016942977905273,
      "learning_rate": 4.119286057631187e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 9863792,
      "step": 11030
    },
    {
      "epoch": 2.9124983502705555,
      "grad_norm": 0.09394952654838562,
      "learning_rate": 4.118537955700646e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 9868272,
      "step": 11035
    },
    {
      "epoch": 2.9138181338260525,
      "grad_norm": 0.12728965282440186,
      "learning_rate": 4.11778960417121e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 9872624,
      "step": 11040
    },
    {
      "epoch": 2.9151379173815495,
      "grad_norm": 0.3959643840789795,
      "learning_rate": 4.117041003158288e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 9877232,
      "step": 11045
    },
    {
      "epoch": 2.9164577009370465,
      "grad_norm": 0.19259177148342133,
      "learning_rate": 4.1162921527773215e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 9881616,
      "step": 11050
    },
    {
      "epoch": 2.917777484492543,
      "grad_norm": 0.11183683574199677,
      "learning_rate": 4.115543053143794e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 9886096,
      "step": 11055
    },
    {
      "epoch": 2.91909726804804,
      "grad_norm": 0.5395477414131165,
      "learning_rate": 4.114793704373226e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 9890512,
      "step": 11060
    },
    {
      "epoch": 2.920417051603537,
      "grad_norm": 0.320205420255661,
      "learning_rate": 4.114044106581175e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 9895088,
      "step": 11065
    },
    {
      "epoch": 2.921736835159034,
      "grad_norm": 0.10865378379821777,
      "learning_rate": 4.11329425988324e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 9899472,
      "step": 11070
    },
    {
      "epoch": 2.923056618714531,
      "grad_norm": 0.09640852361917496,
      "learning_rate": 4.112544164395056e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 9904240,
      "step": 11075
    },
    {
      "epoch": 2.9243764022700276,
      "grad_norm": 0.49801191687583923,
      "learning_rate": 4.111793820232297e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 9908720,
      "step": 11080
    },
    {
      "epoch": 2.9256961858255246,
      "grad_norm": 0.2555314600467682,
      "learning_rate": 4.1110432275106767e-05,
      "loss": 0.0918,
      "num_input_tokens_seen": 9913168,
      "step": 11085
    },
    {
      "epoch": 2.9270159693810216,
      "grad_norm": 0.07466030865907669,
      "learning_rate": 4.110292386345944e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 9917712,
      "step": 11090
    },
    {
      "epoch": 2.928335752936518,
      "grad_norm": 0.4928416609764099,
      "learning_rate": 4.109541296853891e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 9921872,
      "step": 11095
    },
    {
      "epoch": 2.929655536492015,
      "grad_norm": 0.07516459375619888,
      "learning_rate": 4.108789959150341e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 9926224,
      "step": 11100
    },
    {
      "epoch": 2.930975320047512,
      "grad_norm": 0.04099766165018082,
      "learning_rate": 4.108038373351163e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 9930864,
      "step": 11105
    },
    {
      "epoch": 2.932295103603009,
      "grad_norm": 0.11317934095859528,
      "learning_rate": 4.10728653957226e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 9935536,
      "step": 11110
    },
    {
      "epoch": 2.933614887158506,
      "grad_norm": 0.1413199007511139,
      "learning_rate": 4.106534457929575e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 9939824,
      "step": 11115
    },
    {
      "epoch": 2.9349346707140027,
      "grad_norm": 0.4590227007865906,
      "learning_rate": 4.105782128539086e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 9943792,
      "step": 11120
    },
    {
      "epoch": 2.9362544542694997,
      "grad_norm": 0.433015912771225,
      "learning_rate": 4.1050295515168144e-05,
      "loss": 0.099,
      "num_input_tokens_seen": 9948240,
      "step": 11125
    },
    {
      "epoch": 2.9375742378249967,
      "grad_norm": 0.20044876635074615,
      "learning_rate": 4.1042767269788155e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 9952656,
      "step": 11130
    },
    {
      "epoch": 2.9388940213804937,
      "grad_norm": 0.13858526945114136,
      "learning_rate": 4.103523655041185e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 9956880,
      "step": 11135
    },
    {
      "epoch": 2.9402138049359907,
      "grad_norm": 0.2685445547103882,
      "learning_rate": 4.102770335820055e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 9961296,
      "step": 11140
    },
    {
      "epoch": 2.9415335884914873,
      "grad_norm": 0.21399009227752686,
      "learning_rate": 4.1020167694315984e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 9965520,
      "step": 11145
    },
    {
      "epoch": 2.9428533720469843,
      "grad_norm": 0.24000246822834015,
      "learning_rate": 4.101262955992023e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 9970128,
      "step": 11150
    },
    {
      "epoch": 2.9441731556024813,
      "grad_norm": 0.26567238569259644,
      "learning_rate": 4.100508895617578e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 9974608,
      "step": 11155
    },
    {
      "epoch": 2.945492939157978,
      "grad_norm": 0.06241559982299805,
      "learning_rate": 4.099754588424547e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 9979056,
      "step": 11160
    },
    {
      "epoch": 2.946812722713475,
      "grad_norm": 0.03150666877627373,
      "learning_rate": 4.0990000345292546e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 9983568,
      "step": 11165
    },
    {
      "epoch": 2.948132506268972,
      "grad_norm": 0.2828163802623749,
      "learning_rate": 4.098245234048064e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 9987792,
      "step": 11170
    },
    {
      "epoch": 2.949452289824469,
      "grad_norm": 0.295186311006546,
      "learning_rate": 4.0974901870973726e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 9992272,
      "step": 11175
    },
    {
      "epoch": 2.950772073379966,
      "grad_norm": 0.17542459070682526,
      "learning_rate": 4.096734893793619e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 9996528,
      "step": 11180
    },
    {
      "epoch": 2.9520918569354624,
      "grad_norm": 0.3067108690738678,
      "learning_rate": 4.095979354253279e-05,
      "loss": 0.1364,
      "num_input_tokens_seen": 10001104,
      "step": 11185
    },
    {
      "epoch": 2.9534116404909594,
      "grad_norm": 0.08225727826356888,
      "learning_rate": 4.0952235685928656e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 10005520,
      "step": 11190
    },
    {
      "epoch": 2.9547314240464564,
      "grad_norm": 0.3279702961444855,
      "learning_rate": 4.094467536928932e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 10009936,
      "step": 11195
    },
    {
      "epoch": 2.9560512076019534,
      "grad_norm": 0.22426649928092957,
      "learning_rate": 4.093711259378067e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 10014320,
      "step": 11200
    },
    {
      "epoch": 2.9560512076019534,
      "eval_loss": 0.07229285687208176,
      "eval_runtime": 64.829,
      "eval_samples_per_second": 103.889,
      "eval_steps_per_second": 25.976,
      "num_input_tokens_seen": 10014320,
      "step": 11200
    },
    {
      "epoch": 2.9573709911574504,
      "grad_norm": 0.23876067996025085,
      "learning_rate": 4.092954736056897e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 10018640,
      "step": 11205
    },
    {
      "epoch": 2.958690774712947,
      "grad_norm": 0.04636706784367561,
      "learning_rate": 4.09219796708209e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 10023248,
      "step": 11210
    },
    {
      "epoch": 2.960010558268444,
      "grad_norm": 0.1479640007019043,
      "learning_rate": 4.0914409525703464e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 10027408,
      "step": 11215
    },
    {
      "epoch": 2.961330341823941,
      "grad_norm": 0.2845725119113922,
      "learning_rate": 4.090683692638408e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 10031632,
      "step": 11220
    },
    {
      "epoch": 2.9626501253794375,
      "grad_norm": 0.19546422362327576,
      "learning_rate": 4.089926187403056e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 10036464,
      "step": 11225
    },
    {
      "epoch": 2.9639699089349345,
      "grad_norm": 0.2750297486782074,
      "learning_rate": 4.0891684369811044e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 10041040,
      "step": 11230
    },
    {
      "epoch": 2.9652896924904315,
      "grad_norm": 0.1769384741783142,
      "learning_rate": 4.0884104414894107e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 10045392,
      "step": 11235
    },
    {
      "epoch": 2.9666094760459285,
      "grad_norm": 0.41263535618782043,
      "learning_rate": 4.087652201044864e-05,
      "loss": 0.1031,
      "num_input_tokens_seen": 10049744,
      "step": 11240
    },
    {
      "epoch": 2.9679292596014255,
      "grad_norm": 0.26773661375045776,
      "learning_rate": 4.086893715764397e-05,
      "loss": 0.0698,
      "num_input_tokens_seen": 10054000,
      "step": 11245
    },
    {
      "epoch": 2.969249043156922,
      "grad_norm": 0.3321188986301422,
      "learning_rate": 4.086134985764977e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 10058448,
      "step": 11250
    },
    {
      "epoch": 2.970568826712419,
      "grad_norm": 0.3590265214443207,
      "learning_rate": 4.0853760111636085e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 10062960,
      "step": 11255
    },
    {
      "epoch": 2.971888610267916,
      "grad_norm": 0.28318166732788086,
      "learning_rate": 4.084616792077337e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 10067376,
      "step": 11260
    },
    {
      "epoch": 2.973208393823413,
      "grad_norm": 0.13597536087036133,
      "learning_rate": 4.083857328623243e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 10071888,
      "step": 11265
    },
    {
      "epoch": 2.97452817737891,
      "grad_norm": 0.056540802121162415,
      "learning_rate": 4.083097620918444e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 10076368,
      "step": 11270
    },
    {
      "epoch": 2.9758479609344066,
      "grad_norm": 0.15361592173576355,
      "learning_rate": 4.082337669080097e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 10080656,
      "step": 11275
    },
    {
      "epoch": 2.9771677444899036,
      "grad_norm": 0.112300805747509,
      "learning_rate": 4.081577473225398e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 10085360,
      "step": 11280
    },
    {
      "epoch": 2.9784875280454006,
      "grad_norm": 0.443852037191391,
      "learning_rate": 4.080817033471577e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 10089424,
      "step": 11285
    },
    {
      "epoch": 2.979807311600897,
      "grad_norm": 0.035306625068187714,
      "learning_rate": 4.080056349935903e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 10093808,
      "step": 11290
    },
    {
      "epoch": 2.9811270951563946,
      "grad_norm": 0.06907152384519577,
      "learning_rate": 4.079295422735684e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 10098320,
      "step": 11295
    },
    {
      "epoch": 2.982446878711891,
      "grad_norm": 0.027408909052610397,
      "learning_rate": 4.078534251988264e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 10102672,
      "step": 11300
    },
    {
      "epoch": 2.983766662267388,
      "grad_norm": 0.15732643008232117,
      "learning_rate": 4.077772837811025e-05,
      "loss": 0.1244,
      "num_input_tokens_seen": 10107184,
      "step": 11305
    },
    {
      "epoch": 2.985086445822885,
      "grad_norm": 0.2766321301460266,
      "learning_rate": 4.0770111803213874e-05,
      "loss": 0.1018,
      "num_input_tokens_seen": 10111696,
      "step": 11310
    },
    {
      "epoch": 2.9864062293783817,
      "grad_norm": 0.05397824943065643,
      "learning_rate": 4.076249279636807e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 10116048,
      "step": 11315
    },
    {
      "epoch": 2.9877260129338787,
      "grad_norm": 0.1666417419910431,
      "learning_rate": 4.075487135874781e-05,
      "loss": 0.0945,
      "num_input_tokens_seen": 10120496,
      "step": 11320
    },
    {
      "epoch": 2.9890457964893757,
      "grad_norm": 0.15698044002056122,
      "learning_rate": 4.074724749152837e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 10124720,
      "step": 11325
    },
    {
      "epoch": 2.9903655800448727,
      "grad_norm": 0.2472420632839203,
      "learning_rate": 4.07396211958855e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 10128816,
      "step": 11330
    },
    {
      "epoch": 2.9916853636003697,
      "grad_norm": 0.38140153884887695,
      "learning_rate": 4.073199247299523e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 10133168,
      "step": 11335
    },
    {
      "epoch": 2.9930051471558663,
      "grad_norm": 0.27673792839050293,
      "learning_rate": 4.072436132403403e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 10137680,
      "step": 11340
    },
    {
      "epoch": 2.9943249307113633,
      "grad_norm": 0.22438573837280273,
      "learning_rate": 4.0716727750178704e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 10142480,
      "step": 11345
    },
    {
      "epoch": 2.9956447142668603,
      "grad_norm": 0.043681684881448746,
      "learning_rate": 4.0709091752606455e-05,
      "loss": 0.1434,
      "num_input_tokens_seen": 10147120,
      "step": 11350
    },
    {
      "epoch": 2.9969644978223573,
      "grad_norm": 0.37502187490463257,
      "learning_rate": 4.070145333249484e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 10151792,
      "step": 11355
    },
    {
      "epoch": 2.9982842813778543,
      "grad_norm": 0.1907288134098053,
      "learning_rate": 4.069381249102181e-05,
      "loss": 0.0263,
      "num_input_tokens_seen": 10156208,
      "step": 11360
    },
    {
      "epoch": 2.999604064933351,
      "grad_norm": 0.08892754465341568,
      "learning_rate": 4.0686169229365665e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 10160688,
      "step": 11365
    },
    {
      "epoch": 3.0007918701332983,
      "grad_norm": 0.2761468291282654,
      "learning_rate": 4.067852354870511e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 10164384,
      "step": 11370
    },
    {
      "epoch": 3.002111653688795,
      "grad_norm": 0.320089727640152,
      "learning_rate": 4.067087545021919e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 10169312,
      "step": 11375
    },
    {
      "epoch": 3.003431437244292,
      "grad_norm": 0.035226814448833466,
      "learning_rate": 4.066322493508734e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 10173664,
      "step": 11380
    },
    {
      "epoch": 3.004751220799789,
      "grad_norm": 0.22582793235778809,
      "learning_rate": 4.065557200448937e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 10178432,
      "step": 11385
    },
    {
      "epoch": 3.006071004355286,
      "grad_norm": 0.31661298871040344,
      "learning_rate": 4.064791665960546e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 10182752,
      "step": 11390
    },
    {
      "epoch": 3.007390787910783,
      "grad_norm": 0.38629651069641113,
      "learning_rate": 4.064025890161615e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 10187488,
      "step": 11395
    },
    {
      "epoch": 3.0087105714662794,
      "grad_norm": 0.09677337855100632,
      "learning_rate": 4.0632598731702373e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 10191904,
      "step": 11400
    },
    {
      "epoch": 3.0087105714662794,
      "eval_loss": 0.07147658616304398,
      "eval_runtime": 64.7462,
      "eval_samples_per_second": 104.022,
      "eval_steps_per_second": 26.009,
      "num_input_tokens_seen": 10191904,
      "step": 11400
    },
    {
      "epoch": 3.0100303550217764,
      "grad_norm": 0.11812586337327957,
      "learning_rate": 4.0624936151045426e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 10196096,
      "step": 11405
    },
    {
      "epoch": 3.0113501385772734,
      "grad_norm": 0.06322041898965836,
      "learning_rate": 4.061727116082696e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 10200384,
      "step": 11410
    },
    {
      "epoch": 3.0126699221327704,
      "grad_norm": 0.09015189111232758,
      "learning_rate": 4.060960376222903e-05,
      "loss": 0.1108,
      "num_input_tokens_seen": 10204736,
      "step": 11415
    },
    {
      "epoch": 3.013989705688267,
      "grad_norm": 0.05702092498540878,
      "learning_rate": 4.0601933956434034e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 10209120,
      "step": 11420
    },
    {
      "epoch": 3.015309489243764,
      "grad_norm": 0.20381246507167816,
      "learning_rate": 4.059426174462476e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 10213856,
      "step": 11425
    },
    {
      "epoch": 3.016629272799261,
      "grad_norm": 0.10250519961118698,
      "learning_rate": 4.058658712798435e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 10217984,
      "step": 11430
    },
    {
      "epoch": 3.017949056354758,
      "grad_norm": 0.2268906682729721,
      "learning_rate": 4.0578910107696336e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 10222592,
      "step": 11435
    },
    {
      "epoch": 3.0192688399102545,
      "grad_norm": 0.08170004934072495,
      "learning_rate": 4.05712306849446e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 10227072,
      "step": 11440
    },
    {
      "epoch": 3.0205886234657515,
      "grad_norm": 0.3212871253490448,
      "learning_rate": 4.0563548860913415e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 10231424,
      "step": 11445
    },
    {
      "epoch": 3.0219084070212485,
      "grad_norm": 0.47481027245521545,
      "learning_rate": 4.0555864636787414e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 10235936,
      "step": 11450
    },
    {
      "epoch": 3.0232281905767455,
      "grad_norm": 0.5487993359565735,
      "learning_rate": 4.054817801375159e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 10240352,
      "step": 11455
    },
    {
      "epoch": 3.0245479741322425,
      "grad_norm": 0.028782833367586136,
      "learning_rate": 4.054048899299134e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 10244768,
      "step": 11460
    },
    {
      "epoch": 3.025867757687739,
      "grad_norm": 0.238408625125885,
      "learning_rate": 4.0532797575692385e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 10249280,
      "step": 11465
    },
    {
      "epoch": 3.027187541243236,
      "grad_norm": 0.08297261595726013,
      "learning_rate": 4.052510376304085e-05,
      "loss": 0.0441,
      "num_input_tokens_seen": 10253632,
      "step": 11470
    },
    {
      "epoch": 3.028507324798733,
      "grad_norm": 0.21018020808696747,
      "learning_rate": 4.051740755622321e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 10257856,
      "step": 11475
    },
    {
      "epoch": 3.02982710835423,
      "grad_norm": 0.11486421525478363,
      "learning_rate": 4.050970895642632e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 10262624,
      "step": 11480
    },
    {
      "epoch": 3.0311468919097266,
      "grad_norm": 0.0838986337184906,
      "learning_rate": 4.050200796483741e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 10267072,
      "step": 11485
    },
    {
      "epoch": 3.0324666754652236,
      "grad_norm": 0.14056506752967834,
      "learning_rate": 4.049430458264405e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 10271744,
      "step": 11490
    },
    {
      "epoch": 3.0337864590207206,
      "grad_norm": 0.04700998216867447,
      "learning_rate": 4.048659881103422e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 10275936,
      "step": 11495
    },
    {
      "epoch": 3.0351062425762176,
      "grad_norm": 0.2795446217060089,
      "learning_rate": 4.0478890651196235e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 10280384,
      "step": 11500
    },
    {
      "epoch": 3.036426026131714,
      "grad_norm": 0.05636051297187805,
      "learning_rate": 4.047118010431879e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 10284736,
      "step": 11505
    },
    {
      "epoch": 3.037745809687211,
      "grad_norm": 0.2649874985218048,
      "learning_rate": 4.046346717159094e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 10289184,
      "step": 11510
    },
    {
      "epoch": 3.039065593242708,
      "grad_norm": 0.08017101883888245,
      "learning_rate": 4.045575185420214e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 10293728,
      "step": 11515
    },
    {
      "epoch": 3.040385376798205,
      "grad_norm": 0.197988823056221,
      "learning_rate": 4.0448034153342165e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 10298272,
      "step": 11520
    },
    {
      "epoch": 3.041705160353702,
      "grad_norm": 0.2966569662094116,
      "learning_rate": 4.0440314070201194e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 10302688,
      "step": 11525
    },
    {
      "epoch": 3.0430249439091988,
      "grad_norm": 0.1451355367898941,
      "learning_rate": 4.043259160596976e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 10307008,
      "step": 11530
    },
    {
      "epoch": 3.0443447274646958,
      "grad_norm": 0.43196558952331543,
      "learning_rate": 4.0424866761838767e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 10311712,
      "step": 11535
    },
    {
      "epoch": 3.0456645110201928,
      "grad_norm": 0.06707499176263809,
      "learning_rate": 4.041713953899948e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 10316288,
      "step": 11540
    },
    {
      "epoch": 3.0469842945756898,
      "grad_norm": 0.16918694972991943,
      "learning_rate": 4.0409409938643515e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 10320832,
      "step": 11545
    },
    {
      "epoch": 3.0483040781311863,
      "grad_norm": 0.4091033637523651,
      "learning_rate": 4.0401677961962904e-05,
      "loss": 0.1196,
      "num_input_tokens_seen": 10324896,
      "step": 11550
    },
    {
      "epoch": 3.0496238616866833,
      "grad_norm": 0.1273830533027649,
      "learning_rate": 4.039394361015001e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 10329440,
      "step": 11555
    },
    {
      "epoch": 3.0509436452421803,
      "grad_norm": 0.08937449753284454,
      "learning_rate": 4.038620688439755e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 10334272,
      "step": 11560
    },
    {
      "epoch": 3.0522634287976773,
      "grad_norm": 0.42565733194351196,
      "learning_rate": 4.037846778589862e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 10338912,
      "step": 11565
    },
    {
      "epoch": 3.053583212353174,
      "grad_norm": 0.18027301132678986,
      "learning_rate": 4.0370726315846715e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 10343296,
      "step": 11570
    },
    {
      "epoch": 3.054902995908671,
      "grad_norm": 0.140743687748909,
      "learning_rate": 4.036298247543565e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 10347840,
      "step": 11575
    },
    {
      "epoch": 3.056222779464168,
      "grad_norm": 0.26292720437049866,
      "learning_rate": 4.035523626585962e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 10352448,
      "step": 11580
    },
    {
      "epoch": 3.057542563019665,
      "grad_norm": 0.11770360916852951,
      "learning_rate": 4.0347487688313194e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 10356992,
      "step": 11585
    },
    {
      "epoch": 3.058862346575162,
      "grad_norm": 0.21183274686336517,
      "learning_rate": 4.0339736743991296e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 10361152,
      "step": 11590
    },
    {
      "epoch": 3.0601821301306584,
      "grad_norm": 0.05844496563076973,
      "learning_rate": 4.0331983434089227e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 10365472,
      "step": 11595
    },
    {
      "epoch": 3.0615019136861554,
      "grad_norm": 0.09435346722602844,
      "learning_rate": 4.032422775980264e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 10369824,
      "step": 11600
    },
    {
      "epoch": 3.0615019136861554,
      "eval_loss": 0.0714099109172821,
      "eval_runtime": 64.7291,
      "eval_samples_per_second": 104.049,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 10369824,
      "step": 11600
    },
    {
      "epoch": 3.0628216972416524,
      "grad_norm": 0.14890049397945404,
      "learning_rate": 4.031646972232754e-05,
      "loss": 0.1142,
      "num_input_tokens_seen": 10374400,
      "step": 11605
    },
    {
      "epoch": 3.0641414807971494,
      "grad_norm": 0.3251514732837677,
      "learning_rate": 4.0308709322860344e-05,
      "loss": 0.0992,
      "num_input_tokens_seen": 10378976,
      "step": 11610
    },
    {
      "epoch": 3.065461264352646,
      "grad_norm": 0.1879168599843979,
      "learning_rate": 4.0300946562597784e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 10383456,
      "step": 11615
    },
    {
      "epoch": 3.066781047908143,
      "grad_norm": 0.2131122648715973,
      "learning_rate": 4.029318144273698e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 10388000,
      "step": 11620
    },
    {
      "epoch": 3.06810083146364,
      "grad_norm": 0.25327858328819275,
      "learning_rate": 4.0285413964475415e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 10392576,
      "step": 11625
    },
    {
      "epoch": 3.069420615019137,
      "grad_norm": 0.15716034173965454,
      "learning_rate": 4.0277644129010927e-05,
      "loss": 0.1017,
      "num_input_tokens_seen": 10396992,
      "step": 11630
    },
    {
      "epoch": 3.0707403985746335,
      "grad_norm": 0.22831644117832184,
      "learning_rate": 4.0269871937541724e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 10401184,
      "step": 11635
    },
    {
      "epoch": 3.0720601821301305,
      "grad_norm": 0.30151504278182983,
      "learning_rate": 4.026209739126637e-05,
      "loss": 0.1281,
      "num_input_tokens_seen": 10405280,
      "step": 11640
    },
    {
      "epoch": 3.0733799656856275,
      "grad_norm": 0.35103875398635864,
      "learning_rate": 4.025432049138381e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 10410112,
      "step": 11645
    },
    {
      "epoch": 3.0746997492411245,
      "grad_norm": 0.17283488810062408,
      "learning_rate": 4.0246541239093325e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 10414528,
      "step": 11650
    },
    {
      "epoch": 3.0760195327966215,
      "grad_norm": 0.14907215535640717,
      "learning_rate": 4.023875963559459e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 10418976,
      "step": 11655
    },
    {
      "epoch": 3.077339316352118,
      "grad_norm": 0.09874022752046585,
      "learning_rate": 4.023097568208761e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 10423648,
      "step": 11660
    },
    {
      "epoch": 3.078659099907615,
      "grad_norm": 0.10067681968212128,
      "learning_rate": 4.022318937977277e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 10427968,
      "step": 11665
    },
    {
      "epoch": 3.079978883463112,
      "grad_norm": 0.057094722986221313,
      "learning_rate": 4.021540072985084e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 10432384,
      "step": 11670
    },
    {
      "epoch": 3.081298667018609,
      "grad_norm": 0.24319683015346527,
      "learning_rate": 4.020760973352289e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 10436768,
      "step": 11675
    },
    {
      "epoch": 3.0826184505741057,
      "grad_norm": 0.10271193832159042,
      "learning_rate": 4.019981639199042e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 10441152,
      "step": 11680
    },
    {
      "epoch": 3.0839382341296027,
      "grad_norm": 0.11218240112066269,
      "learning_rate": 4.0192020706455245e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 10445408,
      "step": 11685
    },
    {
      "epoch": 3.0852580176850997,
      "grad_norm": 0.2968849837779999,
      "learning_rate": 4.018422267811956e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 10449920,
      "step": 11690
    },
    {
      "epoch": 3.0865778012405967,
      "grad_norm": 0.060434628278017044,
      "learning_rate": 4.017642230818592e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 10454560,
      "step": 11695
    },
    {
      "epoch": 3.087897584796093,
      "grad_norm": 0.38852328062057495,
      "learning_rate": 4.0168619597857246e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 10458752,
      "step": 11700
    },
    {
      "epoch": 3.08921736835159,
      "grad_norm": 0.10935933142900467,
      "learning_rate": 4.016081454833681e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 10463200,
      "step": 11705
    },
    {
      "epoch": 3.090537151907087,
      "grad_norm": 0.12661324441432953,
      "learning_rate": 4.0153007160828245e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 10467552,
      "step": 11710
    },
    {
      "epoch": 3.091856935462584,
      "grad_norm": 0.05486902967095375,
      "learning_rate": 4.0145197436535555e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 10471904,
      "step": 11715
    },
    {
      "epoch": 3.0931767190180812,
      "grad_norm": 0.6579387187957764,
      "learning_rate": 4.0137385376663095e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 10476672,
      "step": 11720
    },
    {
      "epoch": 3.094496502573578,
      "grad_norm": 0.13793013989925385,
      "learning_rate": 4.012957098241558e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 10481152,
      "step": 11725
    },
    {
      "epoch": 3.095816286129075,
      "grad_norm": 0.23525623977184296,
      "learning_rate": 4.0121754254998076e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 10485632,
      "step": 11730
    },
    {
      "epoch": 3.097136069684572,
      "grad_norm": 0.29484468698501587,
      "learning_rate": 4.011393519561606e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 10490208,
      "step": 11735
    },
    {
      "epoch": 3.098455853240069,
      "grad_norm": 0.08719450980424881,
      "learning_rate": 4.010611380547529e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 10494496,
      "step": 11740
    },
    {
      "epoch": 3.0997756367955653,
      "grad_norm": 0.4006711542606354,
      "learning_rate": 4.009829008578192e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 10499008,
      "step": 11745
    },
    {
      "epoch": 3.1010954203510623,
      "grad_norm": 0.06774111837148666,
      "learning_rate": 4.00904640377425e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 10503296,
      "step": 11750
    },
    {
      "epoch": 3.1024152039065593,
      "grad_norm": 0.4838588535785675,
      "learning_rate": 4.0082635662563886e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 10507616,
      "step": 11755
    },
    {
      "epoch": 3.1037349874620563,
      "grad_norm": 0.20226526260375977,
      "learning_rate": 4.007480496145331e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 10512032,
      "step": 11760
    },
    {
      "epoch": 3.105054771017553,
      "grad_norm": 0.09381644427776337,
      "learning_rate": 4.006697193561837e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 10516544,
      "step": 11765
    },
    {
      "epoch": 3.10637455457305,
      "grad_norm": 0.2197590470314026,
      "learning_rate": 4.005913658626701e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 10520928,
      "step": 11770
    },
    {
      "epoch": 3.107694338128547,
      "grad_norm": 0.15821683406829834,
      "learning_rate": 4.005129891460754e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 10525472,
      "step": 11775
    },
    {
      "epoch": 3.109014121684044,
      "grad_norm": 0.1322133094072342,
      "learning_rate": 4.004345892184864e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 10529920,
      "step": 11780
    },
    {
      "epoch": 3.110333905239541,
      "grad_norm": 0.22874826192855835,
      "learning_rate": 4.003561660919932e-05,
      "loss": 0.0745,
      "num_input_tokens_seen": 10534048,
      "step": 11785
    },
    {
      "epoch": 3.1116536887950375,
      "grad_norm": 0.07344454526901245,
      "learning_rate": 4.002777197786897e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 10538784,
      "step": 11790
    },
    {
      "epoch": 3.1129734723505345,
      "grad_norm": 0.2665601074695587,
      "learning_rate": 4.0019925029067326e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 10543104,
      "step": 11795
    },
    {
      "epoch": 3.1142932559060315,
      "grad_norm": 0.39506542682647705,
      "learning_rate": 4.0012075764004495e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 10547296,
      "step": 11800
    },
    {
      "epoch": 3.1142932559060315,
      "eval_loss": 0.07149194180965424,
      "eval_runtime": 64.7435,
      "eval_samples_per_second": 104.026,
      "eval_steps_per_second": 26.01,
      "num_input_tokens_seen": 10547296,
      "step": 11800
    },
    {
      "epoch": 3.1156130394615285,
      "grad_norm": 0.1312103420495987,
      "learning_rate": 4.000422418389094e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 10551968,
      "step": 11805
    },
    {
      "epoch": 3.116932823017025,
      "grad_norm": 0.1038178876042366,
      "learning_rate": 3.999637028993744e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 10556512,
      "step": 11810
    },
    {
      "epoch": 3.118252606572522,
      "grad_norm": 0.12283166497945786,
      "learning_rate": 3.99885140833552e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 10560832,
      "step": 11815
    },
    {
      "epoch": 3.119572390128019,
      "grad_norm": 0.0407571904361248,
      "learning_rate": 3.998065556535572e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 10565440,
      "step": 11820
    },
    {
      "epoch": 3.120892173683516,
      "grad_norm": 0.40160784125328064,
      "learning_rate": 3.9972794737150895e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 10569824,
      "step": 11825
    },
    {
      "epoch": 3.122211957239013,
      "grad_norm": 0.2049623280763626,
      "learning_rate": 3.996493159995297e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 10573984,
      "step": 11830
    },
    {
      "epoch": 3.1235317407945096,
      "grad_norm": 0.20785745978355408,
      "learning_rate": 3.995706615497453e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 10578272,
      "step": 11835
    },
    {
      "epoch": 3.1248515243500066,
      "grad_norm": 0.3975113034248352,
      "learning_rate": 3.994919840342852e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 10582752,
      "step": 11840
    },
    {
      "epoch": 3.1261713079055036,
      "grad_norm": 0.7879478335380554,
      "learning_rate": 3.994132834652825e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 10586944,
      "step": 11845
    },
    {
      "epoch": 3.1274910914610006,
      "grad_norm": 0.20513638854026794,
      "learning_rate": 3.99334559854874e-05,
      "loss": 0.1151,
      "num_input_tokens_seen": 10591424,
      "step": 11850
    },
    {
      "epoch": 3.128810875016497,
      "grad_norm": 0.3745201826095581,
      "learning_rate": 3.9925581321519955e-05,
      "loss": 0.0983,
      "num_input_tokens_seen": 10596320,
      "step": 11855
    },
    {
      "epoch": 3.130130658571994,
      "grad_norm": 0.024210810661315918,
      "learning_rate": 3.991770435584031e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 10600896,
      "step": 11860
    },
    {
      "epoch": 3.131450442127491,
      "grad_norm": 0.3471834063529968,
      "learning_rate": 3.990982508966319e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 10605600,
      "step": 11865
    },
    {
      "epoch": 3.132770225682988,
      "grad_norm": 0.10622302442789078,
      "learning_rate": 3.990194352420367e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 10609952,
      "step": 11870
    },
    {
      "epoch": 3.1340900092384847,
      "grad_norm": 0.23643702268600464,
      "learning_rate": 3.9894059660677184e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 10614784,
      "step": 11875
    },
    {
      "epoch": 3.1354097927939817,
      "grad_norm": 0.10862113535404205,
      "learning_rate": 3.9886173500299526e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 10619264,
      "step": 11880
    },
    {
      "epoch": 3.1367295763494787,
      "grad_norm": 0.10649561882019043,
      "learning_rate": 3.987828504428685e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 10623808,
      "step": 11885
    },
    {
      "epoch": 3.1380493599049757,
      "grad_norm": 0.4502353370189667,
      "learning_rate": 3.987039429385565e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 10628352,
      "step": 11890
    },
    {
      "epoch": 3.1393691434604727,
      "grad_norm": 0.24622097611427307,
      "learning_rate": 3.986250125022277e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 10632800,
      "step": 11895
    },
    {
      "epoch": 3.1406889270159692,
      "grad_norm": 0.13194748759269714,
      "learning_rate": 3.985460591460544e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 10637248,
      "step": 11900
    },
    {
      "epoch": 3.1420087105714662,
      "grad_norm": 0.24656416475772858,
      "learning_rate": 3.984670828822118e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 10641632,
      "step": 11905
    },
    {
      "epoch": 3.1433284941269632,
      "grad_norm": 0.23687931895256042,
      "learning_rate": 3.983880837228794e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 10645664,
      "step": 11910
    },
    {
      "epoch": 3.1446482776824602,
      "grad_norm": 0.11159971356391907,
      "learning_rate": 3.983090616802396e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 10649792,
      "step": 11915
    },
    {
      "epoch": 3.145968061237957,
      "grad_norm": 0.02664664015173912,
      "learning_rate": 3.982300167664788e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 10654528,
      "step": 11920
    },
    {
      "epoch": 3.147287844793454,
      "grad_norm": 0.12186938524246216,
      "learning_rate": 3.981509489937868e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 10659168,
      "step": 11925
    },
    {
      "epoch": 3.148607628348951,
      "grad_norm": 0.10504790395498276,
      "learning_rate": 3.9807185837435643e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 10663776,
      "step": 11930
    },
    {
      "epoch": 3.149927411904448,
      "grad_norm": 0.3318347632884979,
      "learning_rate": 3.9799274492038484e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 10668320,
      "step": 11935
    },
    {
      "epoch": 3.1512471954599444,
      "grad_norm": 0.2635408937931061,
      "learning_rate": 3.979136086440722e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 10672640,
      "step": 11940
    },
    {
      "epoch": 3.1525669790154414,
      "grad_norm": 0.2946043014526367,
      "learning_rate": 3.9783444955762226e-05,
      "loss": 0.0895,
      "num_input_tokens_seen": 10676704,
      "step": 11945
    },
    {
      "epoch": 3.1538867625709384,
      "grad_norm": 0.28410494327545166,
      "learning_rate": 3.977552676732424e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 10681056,
      "step": 11950
    },
    {
      "epoch": 3.1552065461264354,
      "grad_norm": 0.054054390639066696,
      "learning_rate": 3.976760630031435e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 10685472,
      "step": 11955
    },
    {
      "epoch": 3.1565263296819324,
      "grad_norm": 0.23285432159900665,
      "learning_rate": 3.975968355595398e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 10689856,
      "step": 11960
    },
    {
      "epoch": 3.157846113237429,
      "grad_norm": 0.169199138879776,
      "learning_rate": 3.9751758535464935e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 10694560,
      "step": 11965
    },
    {
      "epoch": 3.159165896792926,
      "grad_norm": 0.26040729880332947,
      "learning_rate": 3.9743831240069326e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 10699232,
      "step": 11970
    },
    {
      "epoch": 3.160485680348423,
      "grad_norm": 0.058169495314359665,
      "learning_rate": 3.9735901670989675e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 10703808,
      "step": 11975
    },
    {
      "epoch": 3.16180546390392,
      "grad_norm": 0.24905069172382355,
      "learning_rate": 3.97279698294488e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 10708160,
      "step": 11980
    },
    {
      "epoch": 3.1631252474594165,
      "grad_norm": 0.06303642690181732,
      "learning_rate": 3.9720035716669876e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 10712736,
      "step": 11985
    },
    {
      "epoch": 3.1644450310149135,
      "grad_norm": 0.26368066668510437,
      "learning_rate": 3.9712099333876474e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 10717440,
      "step": 11990
    },
    {
      "epoch": 3.1657648145704105,
      "grad_norm": 0.40200507640838623,
      "learning_rate": 3.9704160682292475e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 10722016,
      "step": 11995
    },
    {
      "epoch": 3.1670845981259075,
      "grad_norm": 0.175954669713974,
      "learning_rate": 3.9696219763142106e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 10726592,
      "step": 12000
    },
    {
      "epoch": 3.1670845981259075,
      "eval_loss": 0.07133092731237411,
      "eval_runtime": 64.7285,
      "eval_samples_per_second": 104.05,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 10726592,
      "step": 12000
    },
    {
      "epoch": 3.1684043816814045,
      "grad_norm": 0.06245870888233185,
      "learning_rate": 3.968827657764997e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 10731200,
      "step": 12005
    },
    {
      "epoch": 3.169724165236901,
      "grad_norm": 0.3634347915649414,
      "learning_rate": 3.9680331127041e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 10735552,
      "step": 12010
    },
    {
      "epoch": 3.171043948792398,
      "grad_norm": 0.2791411280632019,
      "learning_rate": 3.9672383412540495e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 10739872,
      "step": 12015
    },
    {
      "epoch": 3.172363732347895,
      "grad_norm": 0.5884649157524109,
      "learning_rate": 3.966443343537407e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 10744448,
      "step": 12020
    },
    {
      "epoch": 3.173683515903392,
      "grad_norm": 0.14738459885120392,
      "learning_rate": 3.965648119676772e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 10748800,
      "step": 12025
    },
    {
      "epoch": 3.1750032994588886,
      "grad_norm": 0.12158913910388947,
      "learning_rate": 3.96485266979478e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 10753056,
      "step": 12030
    },
    {
      "epoch": 3.1763230830143856,
      "grad_norm": 0.36428841948509216,
      "learning_rate": 3.9640569940140974e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 10757728,
      "step": 12035
    },
    {
      "epoch": 3.1776428665698826,
      "grad_norm": 0.29071515798568726,
      "learning_rate": 3.963261092457428e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 10762400,
      "step": 12040
    },
    {
      "epoch": 3.1789626501253796,
      "grad_norm": 0.23073220252990723,
      "learning_rate": 3.962464965247509e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 10767072,
      "step": 12045
    },
    {
      "epoch": 3.180282433680876,
      "grad_norm": 0.0712454691529274,
      "learning_rate": 3.9616686125071135e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 10771424,
      "step": 12050
    },
    {
      "epoch": 3.181602217236373,
      "grad_norm": 0.11387949436903,
      "learning_rate": 3.9608720343590506e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 10775840,
      "step": 12055
    },
    {
      "epoch": 3.18292200079187,
      "grad_norm": 0.2871323525905609,
      "learning_rate": 3.960075230926161e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 10780384,
      "step": 12060
    },
    {
      "epoch": 3.184241784347367,
      "grad_norm": 0.2290744036436081,
      "learning_rate": 3.959278202331322e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 10784800,
      "step": 12065
    },
    {
      "epoch": 3.185561567902864,
      "grad_norm": 0.1367940753698349,
      "learning_rate": 3.958480948697446e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 10789120,
      "step": 12070
    },
    {
      "epoch": 3.1868813514583607,
      "grad_norm": 0.37057942152023315,
      "learning_rate": 3.95768347014748e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 10793408,
      "step": 12075
    },
    {
      "epoch": 3.1882011350138577,
      "grad_norm": 0.42554354667663574,
      "learning_rate": 3.956885766804404e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 10797600,
      "step": 12080
    },
    {
      "epoch": 3.1895209185693547,
      "grad_norm": 0.19073957204818726,
      "learning_rate": 3.956087838791235e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 10802112,
      "step": 12085
    },
    {
      "epoch": 3.1908407021248517,
      "grad_norm": 0.3140142560005188,
      "learning_rate": 3.955289686231022e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 10806624,
      "step": 12090
    },
    {
      "epoch": 3.1921604856803483,
      "grad_norm": 0.4150667190551758,
      "learning_rate": 3.9544913092468504e-05,
      "loss": 0.1311,
      "num_input_tokens_seen": 10811232,
      "step": 12095
    },
    {
      "epoch": 3.1934802692358453,
      "grad_norm": 0.18257953226566315,
      "learning_rate": 3.9536927079618425e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 10815648,
      "step": 12100
    },
    {
      "epoch": 3.1948000527913423,
      "grad_norm": 0.2698443531990051,
      "learning_rate": 3.9528938824991494e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 10820000,
      "step": 12105
    },
    {
      "epoch": 3.1961198363468393,
      "grad_norm": 0.17314468324184418,
      "learning_rate": 3.952094832981962e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 10824416,
      "step": 12110
    },
    {
      "epoch": 3.197439619902336,
      "grad_norm": 0.4345317780971527,
      "learning_rate": 3.951295559533503e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 10828992,
      "step": 12115
    },
    {
      "epoch": 3.198759403457833,
      "grad_norm": 0.27941009402275085,
      "learning_rate": 3.95049606227703e-05,
      "loss": 0.102,
      "num_input_tokens_seen": 10833856,
      "step": 12120
    },
    {
      "epoch": 3.20007918701333,
      "grad_norm": 0.11721979826688766,
      "learning_rate": 3.949696341335838e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 10838368,
      "step": 12125
    },
    {
      "epoch": 3.201398970568827,
      "grad_norm": 0.2687836289405823,
      "learning_rate": 3.9488963968332503e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 10842880,
      "step": 12130
    },
    {
      "epoch": 3.202718754124324,
      "grad_norm": 0.0599936805665493,
      "learning_rate": 3.948096228892631e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 10847552,
      "step": 12135
    },
    {
      "epoch": 3.2040385376798204,
      "grad_norm": 0.3186212182044983,
      "learning_rate": 3.947295837637375e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 10851936,
      "step": 12140
    },
    {
      "epoch": 3.2053583212353174,
      "grad_norm": 0.25621819496154785,
      "learning_rate": 3.9464952231909135e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 10856640,
      "step": 12145
    },
    {
      "epoch": 3.2066781047908144,
      "grad_norm": 0.19948351383209229,
      "learning_rate": 3.945694385676711e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 10860960,
      "step": 12150
    },
    {
      "epoch": 3.2079978883463114,
      "grad_norm": 0.25004154443740845,
      "learning_rate": 3.944893325218265e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 10865504,
      "step": 12155
    },
    {
      "epoch": 3.209317671901808,
      "grad_norm": 0.3196832537651062,
      "learning_rate": 3.944092041939112e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 10870016,
      "step": 12160
    },
    {
      "epoch": 3.210637455457305,
      "grad_norm": 0.22929449379444122,
      "learning_rate": 3.943290535962818e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 10874240,
      "step": 12165
    },
    {
      "epoch": 3.211957239012802,
      "grad_norm": 0.038272034376859665,
      "learning_rate": 3.942488807412985e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 10878752,
      "step": 12170
    },
    {
      "epoch": 3.213277022568299,
      "grad_norm": 0.2779572010040283,
      "learning_rate": 3.941686856413251e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 10883104,
      "step": 12175
    },
    {
      "epoch": 3.2145968061237955,
      "grad_norm": 0.41146060824394226,
      "learning_rate": 3.9408846830872874e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 10887904,
      "step": 12180
    },
    {
      "epoch": 3.2159165896792925,
      "grad_norm": 0.11652016639709473,
      "learning_rate": 3.940082287558798e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 10892448,
      "step": 12185
    },
    {
      "epoch": 3.2172363732347895,
      "grad_norm": 0.09325562417507172,
      "learning_rate": 3.939279669951522e-05,
      "loss": 0.1289,
      "num_input_tokens_seen": 10896896,
      "step": 12190
    },
    {
      "epoch": 3.2185561567902865,
      "grad_norm": 0.052444543689489365,
      "learning_rate": 3.938476830389234e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 10901536,
      "step": 12195
    },
    {
      "epoch": 3.2198759403457835,
      "grad_norm": 0.08701413869857788,
      "learning_rate": 3.937673768995742e-05,
      "loss": 0.0118,
      "num_input_tokens_seen": 10905760,
      "step": 12200
    },
    {
      "epoch": 3.2198759403457835,
      "eval_loss": 0.07127954065799713,
      "eval_runtime": 64.8005,
      "eval_samples_per_second": 103.934,
      "eval_steps_per_second": 25.987,
      "num_input_tokens_seen": 10905760,
      "step": 12200
    },
    {
      "epoch": 3.22119572390128,
      "grad_norm": 0.24336403608322144,
      "learning_rate": 3.936870485894888e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 10909856,
      "step": 12205
    },
    {
      "epoch": 3.222515507456777,
      "grad_norm": 0.10953997820615768,
      "learning_rate": 3.9360669812105475e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 10914624,
      "step": 12210
    },
    {
      "epoch": 3.223835291012274,
      "grad_norm": 0.07805051654577255,
      "learning_rate": 3.9352632550666325e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 10919392,
      "step": 12215
    },
    {
      "epoch": 3.225155074567771,
      "grad_norm": 0.16056789457798004,
      "learning_rate": 3.9344593075870866e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 10923808,
      "step": 12220
    },
    {
      "epoch": 3.2264748581232676,
      "grad_norm": 0.3008278012275696,
      "learning_rate": 3.933655138895889e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 10928256,
      "step": 12225
    },
    {
      "epoch": 3.2277946416787646,
      "grad_norm": 0.12557865679264069,
      "learning_rate": 3.932850749117053e-05,
      "loss": 0.1159,
      "num_input_tokens_seen": 10932768,
      "step": 12230
    },
    {
      "epoch": 3.2291144252342616,
      "grad_norm": 0.1652558594942093,
      "learning_rate": 3.932046138374624e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 10937024,
      "step": 12235
    },
    {
      "epoch": 3.2304342087897586,
      "grad_norm": 0.13291266560554504,
      "learning_rate": 3.9312413067926854e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 10941632,
      "step": 12240
    },
    {
      "epoch": 3.231753992345255,
      "grad_norm": 0.39460572600364685,
      "learning_rate": 3.9304362544953506e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 10946016,
      "step": 12245
    },
    {
      "epoch": 3.233073775900752,
      "grad_norm": 0.10827259719371796,
      "learning_rate": 3.929630981606769e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 10950464,
      "step": 12250
    },
    {
      "epoch": 3.234393559456249,
      "grad_norm": 0.25769445300102234,
      "learning_rate": 3.928825488251124e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 10954912,
      "step": 12255
    },
    {
      "epoch": 3.235713343011746,
      "grad_norm": 0.35614487528800964,
      "learning_rate": 3.9280197745526344e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 10959392,
      "step": 12260
    },
    {
      "epoch": 3.237033126567243,
      "grad_norm": 0.28672635555267334,
      "learning_rate": 3.9272138406355495e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 10963584,
      "step": 12265
    },
    {
      "epoch": 3.2383529101227397,
      "grad_norm": 0.16827726364135742,
      "learning_rate": 3.926407686624154e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 10968064,
      "step": 12270
    },
    {
      "epoch": 3.2396726936782367,
      "grad_norm": 0.1468430608510971,
      "learning_rate": 3.9256013126427684e-05,
      "loss": 0.0204,
      "num_input_tokens_seen": 10972800,
      "step": 12275
    },
    {
      "epoch": 3.2409924772337337,
      "grad_norm": 0.8434406518936157,
      "learning_rate": 3.9247947188157455e-05,
      "loss": 0.147,
      "num_input_tokens_seen": 10977696,
      "step": 12280
    },
    {
      "epoch": 3.2423122607892307,
      "grad_norm": 0.07937264442443848,
      "learning_rate": 3.9239879052674715e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 10982496,
      "step": 12285
    },
    {
      "epoch": 3.2436320443447273,
      "grad_norm": 0.6772339344024658,
      "learning_rate": 3.9231808721223673e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 10986944,
      "step": 12290
    },
    {
      "epoch": 3.2449518279002243,
      "grad_norm": 0.3079594075679779,
      "learning_rate": 3.9223736195048886e-05,
      "loss": 0.0261,
      "num_input_tokens_seen": 10991232,
      "step": 12295
    },
    {
      "epoch": 3.2462716114557213,
      "grad_norm": 0.12396866083145142,
      "learning_rate": 3.921566147539523e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 10995520,
      "step": 12300
    },
    {
      "epoch": 3.2475913950112183,
      "grad_norm": 0.5206104516983032,
      "learning_rate": 3.920758456350792e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 11000192,
      "step": 12305
    },
    {
      "epoch": 3.248911178566715,
      "grad_norm": 0.2579779624938965,
      "learning_rate": 3.919950546063253e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 11004480,
      "step": 12310
    },
    {
      "epoch": 3.250230962122212,
      "grad_norm": 0.20088797807693481,
      "learning_rate": 3.919142416801496e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 11009184,
      "step": 12315
    },
    {
      "epoch": 3.251550745677709,
      "grad_norm": 0.2147436887025833,
      "learning_rate": 3.918334068690144e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 11013664,
      "step": 12320
    },
    {
      "epoch": 3.252870529233206,
      "grad_norm": 0.07938262820243835,
      "learning_rate": 3.917525501853855e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 11018336,
      "step": 12325
    },
    {
      "epoch": 3.254190312788703,
      "grad_norm": 0.13699668645858765,
      "learning_rate": 3.916716716417319e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 11022848,
      "step": 12330
    },
    {
      "epoch": 3.2555100963441994,
      "grad_norm": 0.11194593459367752,
      "learning_rate": 3.915907712505263e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 11027648,
      "step": 12335
    },
    {
      "epoch": 3.2568298798996964,
      "grad_norm": 0.2792606055736542,
      "learning_rate": 3.915098490242444e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 11031968,
      "step": 12340
    },
    {
      "epoch": 3.2581496634551934,
      "grad_norm": 0.42174407839775085,
      "learning_rate": 3.914289049753654e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 11036448,
      "step": 12345
    },
    {
      "epoch": 3.2594694470106904,
      "grad_norm": 0.2800735831260681,
      "learning_rate": 3.913479391163719e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 11040704,
      "step": 12350
    },
    {
      "epoch": 3.260789230566187,
      "grad_norm": 0.16416536271572113,
      "learning_rate": 3.9126695145975e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 11045120,
      "step": 12355
    },
    {
      "epoch": 3.262109014121684,
      "grad_norm": 0.18411599099636078,
      "learning_rate": 3.911859420179889e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 11049568,
      "step": 12360
    },
    {
      "epoch": 3.263428797677181,
      "grad_norm": 0.0886845737695694,
      "learning_rate": 3.911049108035813e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 11054176,
      "step": 12365
    },
    {
      "epoch": 3.264748581232678,
      "grad_norm": 0.07868367433547974,
      "learning_rate": 3.910238578290232e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 11058656,
      "step": 12370
    },
    {
      "epoch": 3.2660683647881745,
      "grad_norm": 0.3242507874965668,
      "learning_rate": 3.90942783106814e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 11063328,
      "step": 12375
    },
    {
      "epoch": 3.2673881483436715,
      "grad_norm": 0.19056358933448792,
      "learning_rate": 3.908616866494564e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 11067776,
      "step": 12380
    },
    {
      "epoch": 3.2687079318991685,
      "grad_norm": 0.06202366575598717,
      "learning_rate": 3.907805684694566e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 11072896,
      "step": 12385
    },
    {
      "epoch": 3.2700277154546655,
      "grad_norm": 0.2809184789657593,
      "learning_rate": 3.90699428579324e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 11077248,
      "step": 12390
    },
    {
      "epoch": 3.2713474990101625,
      "grad_norm": 0.27314749360084534,
      "learning_rate": 3.906182669915713e-05,
      "loss": 0.0326,
      "num_input_tokens_seen": 11081856,
      "step": 12395
    },
    {
      "epoch": 3.272667282565659,
      "grad_norm": 0.3536410927772522,
      "learning_rate": 3.9053708371871476e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 11086528,
      "step": 12400
    },
    {
      "epoch": 3.272667282565659,
      "eval_loss": 0.07118018716573715,
      "eval_runtime": 64.7659,
      "eval_samples_per_second": 103.99,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 11086528,
      "step": 12400
    },
    {
      "epoch": 3.273987066121156,
      "grad_norm": 0.07660038769245148,
      "learning_rate": 3.904558787732738e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 11091360,
      "step": 12405
    },
    {
      "epoch": 3.275306849676653,
      "grad_norm": 0.3222607672214508,
      "learning_rate": 3.9037465216777135e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 11095488,
      "step": 12410
    },
    {
      "epoch": 3.27662663323215,
      "grad_norm": 0.537509024143219,
      "learning_rate": 3.902934039147334e-05,
      "loss": 0.1332,
      "num_input_tokens_seen": 11100000,
      "step": 12415
    },
    {
      "epoch": 3.2779464167876466,
      "grad_norm": 0.1443837285041809,
      "learning_rate": 3.902121340266894e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 11104256,
      "step": 12420
    },
    {
      "epoch": 3.2792662003431436,
      "grad_norm": 0.156450554728508,
      "learning_rate": 3.9013084251617246e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 11108800,
      "step": 12425
    },
    {
      "epoch": 3.2805859838986406,
      "grad_norm": 0.03396226093173027,
      "learning_rate": 3.9004952939571865e-05,
      "loss": 0.0185,
      "num_input_tokens_seen": 11113120,
      "step": 12430
    },
    {
      "epoch": 3.2819057674541376,
      "grad_norm": 0.3223204016685486,
      "learning_rate": 3.899681946778673e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 11117664,
      "step": 12435
    },
    {
      "epoch": 3.283225551009634,
      "grad_norm": 0.08191344887018204,
      "learning_rate": 3.898868383751615e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 11122336,
      "step": 12440
    },
    {
      "epoch": 3.284545334565131,
      "grad_norm": 0.060574185103178024,
      "learning_rate": 3.8980546050014724e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 11126432,
      "step": 12445
    },
    {
      "epoch": 3.285865118120628,
      "grad_norm": 0.4308983087539673,
      "learning_rate": 3.897240610653741e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 11130784,
      "step": 12450
    },
    {
      "epoch": 3.287184901676125,
      "grad_norm": 0.027273230254650116,
      "learning_rate": 3.896426400833948e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 11135168,
      "step": 12455
    },
    {
      "epoch": 3.288504685231622,
      "grad_norm": 0.3020021915435791,
      "learning_rate": 3.895611975667656e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 11139648,
      "step": 12460
    },
    {
      "epoch": 3.2898244687871188,
      "grad_norm": 0.11137300729751587,
      "learning_rate": 3.8947973352804584e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 11144032,
      "step": 12465
    },
    {
      "epoch": 3.2911442523426158,
      "grad_norm": 0.1464376002550125,
      "learning_rate": 3.893982479797984e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 11148416,
      "step": 12470
    },
    {
      "epoch": 3.2924640358981128,
      "grad_norm": 0.0498352125287056,
      "learning_rate": 3.8931674093458926e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 11152736,
      "step": 12475
    },
    {
      "epoch": 3.2937838194536098,
      "grad_norm": 0.4102264940738678,
      "learning_rate": 3.89235212404988e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 11157472,
      "step": 12480
    },
    {
      "epoch": 3.2951036030091063,
      "grad_norm": 0.3158841133117676,
      "learning_rate": 3.891536624035672e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 11161760,
      "step": 12485
    },
    {
      "epoch": 3.2964233865646033,
      "grad_norm": 0.2993851900100708,
      "learning_rate": 3.8907209094290295e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 11166176,
      "step": 12490
    },
    {
      "epoch": 3.2977431701201003,
      "grad_norm": 0.18618209660053253,
      "learning_rate": 3.8899049803557466e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 11170560,
      "step": 12495
    },
    {
      "epoch": 3.2990629536755973,
      "grad_norm": 0.30372899770736694,
      "learning_rate": 3.889088836941648e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 11175136,
      "step": 12500
    },
    {
      "epoch": 3.300382737231094,
      "grad_norm": 0.11637644469738007,
      "learning_rate": 3.8882724793125946e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 11179904,
      "step": 12505
    },
    {
      "epoch": 3.301702520786591,
      "grad_norm": 0.132761150598526,
      "learning_rate": 3.8874559075944794e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 11184032,
      "step": 12510
    },
    {
      "epoch": 3.303022304342088,
      "grad_norm": 0.08603093028068542,
      "learning_rate": 3.886639121913227e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 11188416,
      "step": 12515
    },
    {
      "epoch": 3.304342087897585,
      "grad_norm": 0.44068774580955505,
      "learning_rate": 3.885822122394797e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 11192832,
      "step": 12520
    },
    {
      "epoch": 3.305661871453082,
      "grad_norm": 0.22403687238693237,
      "learning_rate": 3.8850049091651794e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 11197248,
      "step": 12525
    },
    {
      "epoch": 3.3069816550085784,
      "grad_norm": 0.10143960267305374,
      "learning_rate": 3.8841874823504e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 11201888,
      "step": 12530
    },
    {
      "epoch": 3.3083014385640754,
      "grad_norm": 0.18605142831802368,
      "learning_rate": 3.8833698420765157e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 11206400,
      "step": 12535
    },
    {
      "epoch": 3.3096212221195724,
      "grad_norm": 0.1609012931585312,
      "learning_rate": 3.882551988469618e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 11210976,
      "step": 12540
    },
    {
      "epoch": 3.3109410056750694,
      "grad_norm": 0.32810917496681213,
      "learning_rate": 3.881733921655829e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 11215776,
      "step": 12545
    },
    {
      "epoch": 3.312260789230566,
      "grad_norm": 0.1068665161728859,
      "learning_rate": 3.8809156417613054e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 11220128,
      "step": 12550
    },
    {
      "epoch": 3.313580572786063,
      "grad_norm": 0.11308085173368454,
      "learning_rate": 3.8800971489122364e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 11224480,
      "step": 12555
    },
    {
      "epoch": 3.31490035634156,
      "grad_norm": 0.28777286410331726,
      "learning_rate": 3.8792784432348434e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 11229088,
      "step": 12560
    },
    {
      "epoch": 3.316220139897057,
      "grad_norm": 0.14732003211975098,
      "learning_rate": 3.878459524855381e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 11233664,
      "step": 12565
    },
    {
      "epoch": 3.3175399234525536,
      "grad_norm": 0.10012771934270859,
      "learning_rate": 3.8776403939001384e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 11238176,
      "step": 12570
    },
    {
      "epoch": 3.3188597070080506,
      "grad_norm": 0.02408447675406933,
      "learning_rate": 3.876821050495433e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 11242560,
      "step": 12575
    },
    {
      "epoch": 3.3201794905635476,
      "grad_norm": 0.24276204407215118,
      "learning_rate": 3.87600149476762e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 11247456,
      "step": 12580
    },
    {
      "epoch": 3.3214992741190446,
      "grad_norm": 0.23817916214466095,
      "learning_rate": 3.8751817268430843e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 11252224,
      "step": 12585
    },
    {
      "epoch": 3.3228190576745416,
      "grad_norm": 0.30738565325737,
      "learning_rate": 3.8743617468482464e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 11256608,
      "step": 12590
    },
    {
      "epoch": 3.324138841230038,
      "grad_norm": 0.29069724678993225,
      "learning_rate": 3.8735415549095535e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 11261376,
      "step": 12595
    },
    {
      "epoch": 3.325458624785535,
      "grad_norm": 0.312551885843277,
      "learning_rate": 3.8727211511534934e-05,
      "loss": 0.0981,
      "num_input_tokens_seen": 11266208,
      "step": 12600
    },
    {
      "epoch": 3.325458624785535,
      "eval_loss": 0.07108491659164429,
      "eval_runtime": 64.7375,
      "eval_samples_per_second": 104.036,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 11266208,
      "step": 12600
    },
    {
      "epoch": 3.326778408341032,
      "grad_norm": 0.606665849685669,
      "learning_rate": 3.8719005357065804e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 11270784,
      "step": 12605
    },
    {
      "epoch": 3.328098191896529,
      "grad_norm": 0.3690475523471832,
      "learning_rate": 3.8710797086953645e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 11275200,
      "step": 12610
    },
    {
      "epoch": 3.329417975452026,
      "grad_norm": 0.08787667006254196,
      "learning_rate": 3.870258670246427e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 11279744,
      "step": 12615
    },
    {
      "epoch": 3.3307377590075227,
      "grad_norm": 0.17705772817134857,
      "learning_rate": 3.869437420486384e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 11284000,
      "step": 12620
    },
    {
      "epoch": 3.3320575425630197,
      "grad_norm": 0.35230889916419983,
      "learning_rate": 3.8686159595418805e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 11288576,
      "step": 12625
    },
    {
      "epoch": 3.3333773261185167,
      "grad_norm": 0.46084660291671753,
      "learning_rate": 3.867794287539597e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 11292928,
      "step": 12630
    },
    {
      "epoch": 3.3346971096740132,
      "grad_norm": 0.42620107531547546,
      "learning_rate": 3.866972404606245e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 11297280,
      "step": 12635
    },
    {
      "epoch": 3.3360168932295102,
      "grad_norm": 0.059471841901540756,
      "learning_rate": 3.866150310868571e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 11302272,
      "step": 12640
    },
    {
      "epoch": 3.3373366767850072,
      "grad_norm": 0.2551204264163971,
      "learning_rate": 3.8653280064533506e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 11306496,
      "step": 12645
    },
    {
      "epoch": 3.3386564603405042,
      "grad_norm": 0.235077366232872,
      "learning_rate": 3.864505491487394e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 11311040,
      "step": 12650
    },
    {
      "epoch": 3.3399762438960012,
      "grad_norm": 0.23913665115833282,
      "learning_rate": 3.8636827660975414e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 11315616,
      "step": 12655
    },
    {
      "epoch": 3.341296027451498,
      "grad_norm": 0.24818812310695648,
      "learning_rate": 3.862859830410671e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 11319872,
      "step": 12660
    },
    {
      "epoch": 3.342615811006995,
      "grad_norm": 0.0719914510846138,
      "learning_rate": 3.862036684553688e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 11324160,
      "step": 12665
    },
    {
      "epoch": 3.343935594562492,
      "grad_norm": 0.28848952054977417,
      "learning_rate": 3.8612133286535314e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 11328352,
      "step": 12670
    },
    {
      "epoch": 3.345255378117989,
      "grad_norm": 0.15879443287849426,
      "learning_rate": 3.860389762837173e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 11332768,
      "step": 12675
    },
    {
      "epoch": 3.346575161673486,
      "grad_norm": 0.5777024626731873,
      "learning_rate": 3.859565987231618e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 11336896,
      "step": 12680
    },
    {
      "epoch": 3.3478949452289823,
      "grad_norm": 0.2672886550426483,
      "learning_rate": 3.858742001963902e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 11341376,
      "step": 12685
    },
    {
      "epoch": 3.3492147287844793,
      "grad_norm": 0.1624869853258133,
      "learning_rate": 3.857917807161094e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 11345888,
      "step": 12690
    },
    {
      "epoch": 3.3505345123399763,
      "grad_norm": 0.25792065262794495,
      "learning_rate": 3.857093402950296e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 11350592,
      "step": 12695
    },
    {
      "epoch": 3.351854295895473,
      "grad_norm": 0.08953763544559479,
      "learning_rate": 3.8562687894586414e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 11354816,
      "step": 12700
    },
    {
      "epoch": 3.35317407945097,
      "grad_norm": 0.21978148818016052,
      "learning_rate": 3.8554439668132946e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 11359552,
      "step": 12705
    },
    {
      "epoch": 3.354493863006467,
      "grad_norm": 0.19123108685016632,
      "learning_rate": 3.854618935141455e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 11364352,
      "step": 12710
    },
    {
      "epoch": 3.355813646561964,
      "grad_norm": 0.2923915386199951,
      "learning_rate": 3.8537936945703525e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 11368800,
      "step": 12715
    },
    {
      "epoch": 3.357133430117461,
      "grad_norm": 0.05438604950904846,
      "learning_rate": 3.852968245227249e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 11373408,
      "step": 12720
    },
    {
      "epoch": 3.3584532136729575,
      "grad_norm": 0.4267398715019226,
      "learning_rate": 3.85214258723944e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 11378208,
      "step": 12725
    },
    {
      "epoch": 3.3597729972284545,
      "grad_norm": 0.06874749809503555,
      "learning_rate": 3.8513167207342524e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 11382688,
      "step": 12730
    },
    {
      "epoch": 3.3610927807839515,
      "grad_norm": 0.5170417428016663,
      "learning_rate": 3.850490645839044e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 11386944,
      "step": 12735
    },
    {
      "epoch": 3.3624125643394485,
      "grad_norm": 0.09361166507005692,
      "learning_rate": 3.849664362681207e-05,
      "loss": 0.0698,
      "num_input_tokens_seen": 11391424,
      "step": 12740
    },
    {
      "epoch": 3.3637323478949455,
      "grad_norm": 0.1086655855178833,
      "learning_rate": 3.848837871388165e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 11396032,
      "step": 12745
    },
    {
      "epoch": 3.365052131450442,
      "grad_norm": 0.030960671603679657,
      "learning_rate": 3.848011172087371e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 11400864,
      "step": 12750
    },
    {
      "epoch": 3.366371915005939,
      "grad_norm": 0.40509742498397827,
      "learning_rate": 3.847184264906315e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 11405312,
      "step": 12755
    },
    {
      "epoch": 3.367691698561436,
      "grad_norm": 0.14501312375068665,
      "learning_rate": 3.846357149972516e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 11409824,
      "step": 12760
    },
    {
      "epoch": 3.3690114821169326,
      "grad_norm": 0.13309578597545624,
      "learning_rate": 3.8455298274135246e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 11414496,
      "step": 12765
    },
    {
      "epoch": 3.3703312656724296,
      "grad_norm": 0.32120785117149353,
      "learning_rate": 3.8447022973569254e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 11418784,
      "step": 12770
    },
    {
      "epoch": 3.3716510492279266,
      "grad_norm": 0.11422805488109589,
      "learning_rate": 3.843874559930332e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 11423520,
      "step": 12775
    },
    {
      "epoch": 3.3729708327834236,
      "grad_norm": 0.7025569081306458,
      "learning_rate": 3.843046615261394e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 11427552,
      "step": 12780
    },
    {
      "epoch": 3.3742906163389206,
      "grad_norm": 0.10289428383111954,
      "learning_rate": 3.842218463477791e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 11432000,
      "step": 12785
    },
    {
      "epoch": 3.375610399894417,
      "grad_norm": 0.22900713980197906,
      "learning_rate": 3.841390104707233e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 11436160,
      "step": 12790
    },
    {
      "epoch": 3.376930183449914,
      "grad_norm": 0.18482773005962372,
      "learning_rate": 3.8405615390774643e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 11440416,
      "step": 12795
    },
    {
      "epoch": 3.378249967005411,
      "grad_norm": 0.3416880965232849,
      "learning_rate": 3.839732766716259e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 11445184,
      "step": 12800
    },
    {
      "epoch": 3.378249967005411,
      "eval_loss": 0.07112780958414078,
      "eval_runtime": 64.7412,
      "eval_samples_per_second": 104.03,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 11445184,
      "step": 12800
    },
    {
      "epoch": 3.379569750560908,
      "grad_norm": 0.2003156542778015,
      "learning_rate": 3.838903787751425e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 11449728,
      "step": 12805
    },
    {
      "epoch": 3.380889534116405,
      "grad_norm": 0.07657336443662643,
      "learning_rate": 3.838074602310802e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 11454144,
      "step": 12810
    },
    {
      "epoch": 3.3822093176719017,
      "grad_norm": 0.3678944408893585,
      "learning_rate": 3.837245210522258e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 11458496,
      "step": 12815
    },
    {
      "epoch": 3.3835291012273987,
      "grad_norm": 0.05436988174915314,
      "learning_rate": 3.8364156125136996e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 11463200,
      "step": 12820
    },
    {
      "epoch": 3.3848488847828957,
      "grad_norm": 0.7360002398490906,
      "learning_rate": 3.835585808413059e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 11467968,
      "step": 12825
    },
    {
      "epoch": 3.3861686683383927,
      "grad_norm": 0.1318444311618805,
      "learning_rate": 3.8347557983483024e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 11472608,
      "step": 12830
    },
    {
      "epoch": 3.3874884518938893,
      "grad_norm": 0.17243808507919312,
      "learning_rate": 3.833925582447428e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 11477472,
      "step": 12835
    },
    {
      "epoch": 3.3888082354493863,
      "grad_norm": 0.18860001862049103,
      "learning_rate": 3.8330951608384656e-05,
      "loss": 0.0181,
      "num_input_tokens_seen": 11481952,
      "step": 12840
    },
    {
      "epoch": 3.3901280190048833,
      "grad_norm": 0.17510107159614563,
      "learning_rate": 3.832264533649477e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 11485920,
      "step": 12845
    },
    {
      "epoch": 3.3914478025603803,
      "grad_norm": 0.24049000442028046,
      "learning_rate": 3.8314337010085555e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 11490272,
      "step": 12850
    },
    {
      "epoch": 3.392767586115877,
      "grad_norm": 0.030918072909116745,
      "learning_rate": 3.830602663043824e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 11494592,
      "step": 12855
    },
    {
      "epoch": 3.394087369671374,
      "grad_norm": 0.33452993631362915,
      "learning_rate": 3.8297714198834414e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 11499136,
      "step": 12860
    },
    {
      "epoch": 3.395407153226871,
      "grad_norm": 0.25054389238357544,
      "learning_rate": 3.828939971655595e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 11503456,
      "step": 12865
    },
    {
      "epoch": 3.396726936782368,
      "grad_norm": 0.06659526377916336,
      "learning_rate": 3.828108318488505e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 11507712,
      "step": 12870
    },
    {
      "epoch": 3.398046720337865,
      "grad_norm": 0.28113067150115967,
      "learning_rate": 3.8272764605104216e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 11511936,
      "step": 12875
    },
    {
      "epoch": 3.3993665038933614,
      "grad_norm": 0.35489577054977417,
      "learning_rate": 3.826444397849628e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 11516576,
      "step": 12880
    },
    {
      "epoch": 3.4006862874488584,
      "grad_norm": 0.03772253170609474,
      "learning_rate": 3.825612130634439e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 11520928,
      "step": 12885
    },
    {
      "epoch": 3.4020060710043554,
      "grad_norm": 0.05010586604475975,
      "learning_rate": 3.824779658993202e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 11525632,
      "step": 12890
    },
    {
      "epoch": 3.4033258545598524,
      "grad_norm": 0.3769335150718689,
      "learning_rate": 3.823946983054292e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 11530368,
      "step": 12895
    },
    {
      "epoch": 3.404645638115349,
      "grad_norm": 0.2078799158334732,
      "learning_rate": 3.82311410294612e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 11535072,
      "step": 12900
    },
    {
      "epoch": 3.405965421670846,
      "grad_norm": 0.2232269048690796,
      "learning_rate": 3.822281018797127e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 11539392,
      "step": 12905
    },
    {
      "epoch": 3.407285205226343,
      "grad_norm": 0.10016993433237076,
      "learning_rate": 3.821447730735783e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 11543904,
      "step": 12910
    },
    {
      "epoch": 3.40860498878184,
      "grad_norm": 0.2646557688713074,
      "learning_rate": 3.820614238890592e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 11548416,
      "step": 12915
    },
    {
      "epoch": 3.4099247723373365,
      "grad_norm": 0.05379759520292282,
      "learning_rate": 3.819780543390091e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 11552832,
      "step": 12920
    },
    {
      "epoch": 3.4112445558928335,
      "grad_norm": 0.20758458971977234,
      "learning_rate": 3.818946644362844e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 11557696,
      "step": 12925
    },
    {
      "epoch": 3.4125643394483305,
      "grad_norm": 0.09507401287555695,
      "learning_rate": 3.81811254193745e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 11562176,
      "step": 12930
    },
    {
      "epoch": 3.4138841230038275,
      "grad_norm": 0.2030310183763504,
      "learning_rate": 3.8172782362425366e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 11566464,
      "step": 12935
    },
    {
      "epoch": 3.4152039065593245,
      "grad_norm": 0.09187600761651993,
      "learning_rate": 3.816443727406765e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 11570976,
      "step": 12940
    },
    {
      "epoch": 3.416523690114821,
      "grad_norm": 0.4140663146972656,
      "learning_rate": 3.815609015558829e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 11575552,
      "step": 12945
    },
    {
      "epoch": 3.417843473670318,
      "grad_norm": 0.17092835903167725,
      "learning_rate": 3.814774100827448e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 11579808,
      "step": 12950
    },
    {
      "epoch": 3.419163257225815,
      "grad_norm": 0.6382081508636475,
      "learning_rate": 3.813938983341379e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 11584224,
      "step": 12955
    },
    {
      "epoch": 3.420483040781312,
      "grad_norm": 0.05937276408076286,
      "learning_rate": 3.813103663229407e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 11588704,
      "step": 12960
    },
    {
      "epoch": 3.4218028243368086,
      "grad_norm": 0.239060640335083,
      "learning_rate": 3.812268140620349e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 11593216,
      "step": 12965
    },
    {
      "epoch": 3.4231226078923056,
      "grad_norm": 0.07066410034894943,
      "learning_rate": 3.811432415643051e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 11597568,
      "step": 12970
    },
    {
      "epoch": 3.4244423914478026,
      "grad_norm": 0.07205714285373688,
      "learning_rate": 3.8105964884263954e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 11602080,
      "step": 12975
    },
    {
      "epoch": 3.4257621750032996,
      "grad_norm": 0.41819435358047485,
      "learning_rate": 3.809760359099291e-05,
      "loss": 0.0969,
      "num_input_tokens_seen": 11606496,
      "step": 12980
    },
    {
      "epoch": 3.427081958558796,
      "grad_norm": 0.4587157368659973,
      "learning_rate": 3.8089240277906804e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 11610592,
      "step": 12985
    },
    {
      "epoch": 3.428401742114293,
      "grad_norm": 0.2852168679237366,
      "learning_rate": 3.808087494629535e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 11614976,
      "step": 12990
    },
    {
      "epoch": 3.42972152566979,
      "grad_norm": 0.4607377350330353,
      "learning_rate": 3.8072507597448595e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 11619488,
      "step": 12995
    },
    {
      "epoch": 3.431041309225287,
      "grad_norm": 0.160573810338974,
      "learning_rate": 3.806413823265689e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 11623936,
      "step": 13000
    },
    {
      "epoch": 3.431041309225287,
      "eval_loss": 0.07065212726593018,
      "eval_runtime": 64.8025,
      "eval_samples_per_second": 103.931,
      "eval_steps_per_second": 25.987,
      "num_input_tokens_seen": 11623936,
      "step": 13000
    },
    {
      "epoch": 3.432361092780784,
      "grad_norm": 0.12693430483341217,
      "learning_rate": 3.805576685321089e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 11628192,
      "step": 13005
    },
    {
      "epoch": 3.4336808763362807,
      "grad_norm": 0.15356528759002686,
      "learning_rate": 3.804739346040158e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 11632512,
      "step": 13010
    },
    {
      "epoch": 3.4350006598917777,
      "grad_norm": 0.2569732666015625,
      "learning_rate": 3.8039018055520234e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 11636832,
      "step": 13015
    },
    {
      "epoch": 3.4363204434472747,
      "grad_norm": 0.4036039710044861,
      "learning_rate": 3.803064063985844e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 11641344,
      "step": 13020
    },
    {
      "epoch": 3.4376402270027717,
      "grad_norm": 0.05075610801577568,
      "learning_rate": 3.802226121470811e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 11645920,
      "step": 13025
    },
    {
      "epoch": 3.4389600105582683,
      "grad_norm": 0.39474213123321533,
      "learning_rate": 3.801387978136145e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 11650656,
      "step": 13030
    },
    {
      "epoch": 3.4402797941137653,
      "grad_norm": 0.23534747958183289,
      "learning_rate": 3.800549634111099e-05,
      "loss": 0.0992,
      "num_input_tokens_seen": 11655104,
      "step": 13035
    },
    {
      "epoch": 3.4415995776692623,
      "grad_norm": 0.13698719441890717,
      "learning_rate": 3.799711089524955e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 11659680,
      "step": 13040
    },
    {
      "epoch": 3.4429193612247593,
      "grad_norm": 0.44572871923446655,
      "learning_rate": 3.7988723445070285e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 11664256,
      "step": 13045
    },
    {
      "epoch": 3.444239144780256,
      "grad_norm": 0.17778794467449188,
      "learning_rate": 3.798033399186663e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 11668544,
      "step": 13050
    },
    {
      "epoch": 3.445558928335753,
      "grad_norm": 0.3145613968372345,
      "learning_rate": 3.797194253693237e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 11672832,
      "step": 13055
    },
    {
      "epoch": 3.44687871189125,
      "grad_norm": 0.1257438212633133,
      "learning_rate": 3.796354908156153e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 11677344,
      "step": 13060
    },
    {
      "epoch": 3.448198495446747,
      "grad_norm": 0.07033568620681763,
      "learning_rate": 3.795515362704853e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 11681760,
      "step": 13065
    },
    {
      "epoch": 3.449518279002244,
      "grad_norm": 0.5977985858917236,
      "learning_rate": 3.794675617468803e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 11686048,
      "step": 13070
    },
    {
      "epoch": 3.4508380625577404,
      "grad_norm": 0.16040270030498505,
      "learning_rate": 3.793835672577503e-05,
      "loss": 0.0161,
      "num_input_tokens_seen": 11690496,
      "step": 13075
    },
    {
      "epoch": 3.4521578461132374,
      "grad_norm": 0.10661104321479797,
      "learning_rate": 3.7929955281604826e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 11694688,
      "step": 13080
    },
    {
      "epoch": 3.4534776296687344,
      "grad_norm": 0.1306701898574829,
      "learning_rate": 3.7921551843473036e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 11699264,
      "step": 13085
    },
    {
      "epoch": 3.4547974132242314,
      "grad_norm": 0.08522362262010574,
      "learning_rate": 3.791314641267557e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 11703488,
      "step": 13090
    },
    {
      "epoch": 3.456117196779728,
      "grad_norm": 0.3868141770362854,
      "learning_rate": 3.790473899050864e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 11707648,
      "step": 13095
    },
    {
      "epoch": 3.457436980335225,
      "grad_norm": 0.03684745728969574,
      "learning_rate": 3.7896329578268794e-05,
      "loss": 0.0266,
      "num_input_tokens_seen": 11712192,
      "step": 13100
    },
    {
      "epoch": 3.458756763890722,
      "grad_norm": 0.34080183506011963,
      "learning_rate": 3.7887918177252855e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 11716576,
      "step": 13105
    },
    {
      "epoch": 3.460076547446219,
      "grad_norm": 0.2844870090484619,
      "learning_rate": 3.787950478875798e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 11721088,
      "step": 13110
    },
    {
      "epoch": 3.4613963310017155,
      "grad_norm": 0.10446255654096603,
      "learning_rate": 3.787108941408162e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 11725664,
      "step": 13115
    },
    {
      "epoch": 3.4627161145572125,
      "grad_norm": 0.15243922173976898,
      "learning_rate": 3.786267205452151e-05,
      "loss": 0.0241,
      "num_input_tokens_seen": 11729856,
      "step": 13120
    },
    {
      "epoch": 3.4640358981127095,
      "grad_norm": 0.18912728130817413,
      "learning_rate": 3.785425271137573e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 11734400,
      "step": 13125
    },
    {
      "epoch": 3.4653556816682065,
      "grad_norm": 0.1190740317106247,
      "learning_rate": 3.7845831385942655e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 11738944,
      "step": 13130
    },
    {
      "epoch": 3.4666754652237035,
      "grad_norm": 0.19124087691307068,
      "learning_rate": 3.7837408079520944e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 11743456,
      "step": 13135
    },
    {
      "epoch": 3.4679952487792,
      "grad_norm": 0.2830105125904083,
      "learning_rate": 3.782898279340957e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 11747744,
      "step": 13140
    },
    {
      "epoch": 3.469315032334697,
      "grad_norm": 0.2466481775045395,
      "learning_rate": 3.782055552890784e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 11752256,
      "step": 13145
    },
    {
      "epoch": 3.470634815890194,
      "grad_norm": 0.17591819167137146,
      "learning_rate": 3.781212628731534e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 11756832,
      "step": 13150
    },
    {
      "epoch": 3.471954599445691,
      "grad_norm": 0.08080969750881195,
      "learning_rate": 3.7803695069931946e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 11760960,
      "step": 13155
    },
    {
      "epoch": 3.4732743830011876,
      "grad_norm": 0.28955063223838806,
      "learning_rate": 3.779526187805789e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 11765408,
      "step": 13160
    },
    {
      "epoch": 3.4745941665566846,
      "grad_norm": 0.15552116930484772,
      "learning_rate": 3.778682671299364e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 11769600,
      "step": 13165
    },
    {
      "epoch": 3.4759139501121816,
      "grad_norm": 0.3238231837749481,
      "learning_rate": 3.777838957604003e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 11774176,
      "step": 13170
    },
    {
      "epoch": 3.4772337336676786,
      "grad_norm": 0.4140813648700714,
      "learning_rate": 3.776995046849816e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 11778496,
      "step": 13175
    },
    {
      "epoch": 3.478553517223175,
      "grad_norm": 0.5008772611618042,
      "learning_rate": 3.776150939166945e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 11783040,
      "step": 13180
    },
    {
      "epoch": 3.479873300778672,
      "grad_norm": 0.3332741856575012,
      "learning_rate": 3.775306634685562e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 11787520,
      "step": 13185
    },
    {
      "epoch": 3.481193084334169,
      "grad_norm": 0.24662064015865326,
      "learning_rate": 3.7744621335358696e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 11792288,
      "step": 13190
    },
    {
      "epoch": 3.482512867889666,
      "grad_norm": 0.13576284050941467,
      "learning_rate": 3.7736174358481e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 11796800,
      "step": 13195
    },
    {
      "epoch": 3.483832651445163,
      "grad_norm": 0.13013531267642975,
      "learning_rate": 3.7727725417525175e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 11801312,
      "step": 13200
    },
    {
      "epoch": 3.483832651445163,
      "eval_loss": 0.07120130211114883,
      "eval_runtime": 64.7201,
      "eval_samples_per_second": 104.063,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 11801312,
      "step": 13200
    },
    {
      "epoch": 3.4851524350006597,
      "grad_norm": 0.2128872573375702,
      "learning_rate": 3.771927451379414e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 11805760,
      "step": 13205
    },
    {
      "epoch": 3.4864722185561567,
      "grad_norm": 0.3668399751186371,
      "learning_rate": 3.7710821648591135e-05,
      "loss": 0.1133,
      "num_input_tokens_seen": 11810464,
      "step": 13210
    },
    {
      "epoch": 3.4877920021116537,
      "grad_norm": 0.3235186040401459,
      "learning_rate": 3.7702366823219694e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 11814656,
      "step": 13215
    },
    {
      "epoch": 3.4891117856671507,
      "grad_norm": 0.24456529319286346,
      "learning_rate": 3.769391003898366e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 11819424,
      "step": 13220
    },
    {
      "epoch": 3.4904315692226473,
      "grad_norm": 0.10153931379318237,
      "learning_rate": 3.768545129718718e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 11823680,
      "step": 13225
    },
    {
      "epoch": 3.4917513527781443,
      "grad_norm": 0.7076449394226074,
      "learning_rate": 3.7676990599134686e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 11828096,
      "step": 13230
    },
    {
      "epoch": 3.4930711363336413,
      "grad_norm": 0.24989350140094757,
      "learning_rate": 3.766852794613095e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 11832576,
      "step": 13235
    },
    {
      "epoch": 3.4943909198891383,
      "grad_norm": 0.25572073459625244,
      "learning_rate": 3.766006333948099e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 11836896,
      "step": 13240
    },
    {
      "epoch": 3.495710703444635,
      "grad_norm": 0.125401109457016,
      "learning_rate": 3.765159678049017e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 11841280,
      "step": 13245
    },
    {
      "epoch": 3.497030487000132,
      "grad_norm": 0.18690870702266693,
      "learning_rate": 3.7643128270464134e-05,
      "loss": 0.032,
      "num_input_tokens_seen": 11845760,
      "step": 13250
    },
    {
      "epoch": 3.498350270555629,
      "grad_norm": 0.05348171666264534,
      "learning_rate": 3.763465781070884e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 11850240,
      "step": 13255
    },
    {
      "epoch": 3.499670054111126,
      "grad_norm": 0.1838684231042862,
      "learning_rate": 3.762618540253052e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 11855104,
      "step": 13260
    },
    {
      "epoch": 3.500989837666623,
      "grad_norm": 0.486880898475647,
      "learning_rate": 3.761771104723576e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 11859616,
      "step": 13265
    },
    {
      "epoch": 3.5023096212221194,
      "grad_norm": 0.33088698983192444,
      "learning_rate": 3.7609234746131386e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 11863872,
      "step": 13270
    },
    {
      "epoch": 3.5036294047776164,
      "grad_norm": 0.33012714982032776,
      "learning_rate": 3.7600756500524556e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 11867968,
      "step": 13275
    },
    {
      "epoch": 3.5049491883331134,
      "grad_norm": 0.10354222357273102,
      "learning_rate": 3.759227631172271e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 11872064,
      "step": 13280
    },
    {
      "epoch": 3.5062689718886104,
      "grad_norm": 0.08376798778772354,
      "learning_rate": 3.758379418103363e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 11876544,
      "step": 13285
    },
    {
      "epoch": 3.5075887554441074,
      "grad_norm": 0.05344199761748314,
      "learning_rate": 3.757531010976534e-05,
      "loss": 0.0282,
      "num_input_tokens_seen": 11881088,
      "step": 13290
    },
    {
      "epoch": 3.508908538999604,
      "grad_norm": 0.0856684222817421,
      "learning_rate": 3.75668240992262e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 11885408,
      "step": 13295
    },
    {
      "epoch": 3.510228322555101,
      "grad_norm": 0.09118033200502396,
      "learning_rate": 3.7558336150724865e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 11889728,
      "step": 13300
    },
    {
      "epoch": 3.511548106110598,
      "grad_norm": 0.06639908999204636,
      "learning_rate": 3.754984626557028e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 11894016,
      "step": 13305
    },
    {
      "epoch": 3.5128678896660945,
      "grad_norm": 0.06480544060468674,
      "learning_rate": 3.754135444507168e-05,
      "loss": 0.0859,
      "num_input_tokens_seen": 11898720,
      "step": 13310
    },
    {
      "epoch": 3.5141876732215915,
      "grad_norm": 0.03230544924736023,
      "learning_rate": 3.753286069053863e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 11903232,
      "step": 13315
    },
    {
      "epoch": 3.5155074567770885,
      "grad_norm": 0.05360060930252075,
      "learning_rate": 3.7524365003280945e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 11907904,
      "step": 13320
    },
    {
      "epoch": 3.5168272403325855,
      "grad_norm": 0.09171462804079056,
      "learning_rate": 3.75158673846088e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 11912224,
      "step": 13325
    },
    {
      "epoch": 3.5181470238880825,
      "grad_norm": 0.1222505271434784,
      "learning_rate": 3.750736783583262e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 11916608,
      "step": 13330
    },
    {
      "epoch": 3.519466807443579,
      "grad_norm": 0.5553958415985107,
      "learning_rate": 3.7498866358263144e-05,
      "loss": 0.155,
      "num_input_tokens_seen": 11921408,
      "step": 13335
    },
    {
      "epoch": 3.520786590999076,
      "grad_norm": 0.0767131894826889,
      "learning_rate": 3.74903629532114e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 11926016,
      "step": 13340
    },
    {
      "epoch": 3.522106374554573,
      "grad_norm": 0.19707342982292175,
      "learning_rate": 3.748185762198873e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 11930016,
      "step": 13345
    },
    {
      "epoch": 3.52342615811007,
      "grad_norm": 0.3796919882297516,
      "learning_rate": 3.747335036590676e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 11934400,
      "step": 13350
    },
    {
      "epoch": 3.524745941665567,
      "grad_norm": 0.13117749989032745,
      "learning_rate": 3.7464841186277405e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 11938816,
      "step": 13355
    },
    {
      "epoch": 3.5260657252210637,
      "grad_norm": 0.11444339156150818,
      "learning_rate": 3.7456330084412896e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 11943552,
      "step": 13360
    },
    {
      "epoch": 3.5273855087765607,
      "grad_norm": 0.21108338236808777,
      "learning_rate": 3.744781706162576e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 11948032,
      "step": 13365
    },
    {
      "epoch": 3.5287052923320577,
      "grad_norm": 0.28620192408561707,
      "learning_rate": 3.743930211922879e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 11952768,
      "step": 13370
    },
    {
      "epoch": 3.530025075887554,
      "grad_norm": 0.20223495364189148,
      "learning_rate": 3.743078525853513e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 11956992,
      "step": 13375
    },
    {
      "epoch": 3.531344859443051,
      "grad_norm": 0.2729811668395996,
      "learning_rate": 3.7422266480858154e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 11961344,
      "step": 13380
    },
    {
      "epoch": 3.532664642998548,
      "grad_norm": 0.343795508146286,
      "learning_rate": 3.741374578751158e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 11965888,
      "step": 13385
    },
    {
      "epoch": 3.533984426554045,
      "grad_norm": 0.2854900360107422,
      "learning_rate": 3.740522317980941e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 11970368,
      "step": 13390
    },
    {
      "epoch": 3.535304210109542,
      "grad_norm": 0.09296902269124985,
      "learning_rate": 3.739669865906593e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 11974880,
      "step": 13395
    },
    {
      "epoch": 3.5366239936650388,
      "grad_norm": 0.2642110586166382,
      "learning_rate": 3.738817222659573e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 11979552,
      "step": 13400
    },
    {
      "epoch": 3.5366239936650388,
      "eval_loss": 0.0715508833527565,
      "eval_runtime": 64.7286,
      "eval_samples_per_second": 104.05,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 11979552,
      "step": 13400
    },
    {
      "epoch": 3.5379437772205358,
      "grad_norm": 0.2166488617658615,
      "learning_rate": 3.73796438837137e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 11984384,
      "step": 13405
    },
    {
      "epoch": 3.5392635607760328,
      "grad_norm": 0.10222763568162918,
      "learning_rate": 3.7371113631735e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 11988864,
      "step": 13410
    },
    {
      "epoch": 3.5405833443315298,
      "grad_norm": 0.40142932534217834,
      "learning_rate": 3.736258147197512e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 11993088,
      "step": 13415
    },
    {
      "epoch": 3.5419031278870268,
      "grad_norm": 0.1158529669046402,
      "learning_rate": 3.735404740574981e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 11997600,
      "step": 13420
    },
    {
      "epoch": 3.5432229114425233,
      "grad_norm": 0.19183537364006042,
      "learning_rate": 3.7345511434375145e-05,
      "loss": 0.127,
      "num_input_tokens_seen": 12002048,
      "step": 13425
    },
    {
      "epoch": 3.5445426949980203,
      "grad_norm": 0.06798561662435532,
      "learning_rate": 3.733697355916748e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 12006752,
      "step": 13430
    },
    {
      "epoch": 3.5458624785535173,
      "grad_norm": 0.2004760503768921,
      "learning_rate": 3.732843378144345e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 12011360,
      "step": 13435
    },
    {
      "epoch": 3.547182262109014,
      "grad_norm": 0.19981902837753296,
      "learning_rate": 3.7319892102519995e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 12015616,
      "step": 13440
    },
    {
      "epoch": 3.548502045664511,
      "grad_norm": 0.23517592251300812,
      "learning_rate": 3.731134852371436e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 12019904,
      "step": 13445
    },
    {
      "epoch": 3.549821829220008,
      "grad_norm": 0.12135408073663712,
      "learning_rate": 3.730280304634408e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 12024096,
      "step": 13450
    },
    {
      "epoch": 3.551141612775505,
      "grad_norm": 0.31607285141944885,
      "learning_rate": 3.729425567172696e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 12028608,
      "step": 13455
    },
    {
      "epoch": 3.552461396331002,
      "grad_norm": 0.13649743795394897,
      "learning_rate": 3.728570640118111e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 12032896,
      "step": 13460
    },
    {
      "epoch": 3.5537811798864984,
      "grad_norm": 0.2105242758989334,
      "learning_rate": 3.727715523602494e-05,
      "loss": 0.1471,
      "num_input_tokens_seen": 12037088,
      "step": 13465
    },
    {
      "epoch": 3.5551009634419954,
      "grad_norm": 0.030522378161549568,
      "learning_rate": 3.726860217757715e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 12041664,
      "step": 13470
    },
    {
      "epoch": 3.5564207469974924,
      "grad_norm": 0.2716744542121887,
      "learning_rate": 3.726004722715673e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 12046336,
      "step": 13475
    },
    {
      "epoch": 3.5577405305529894,
      "grad_norm": 0.08288174122571945,
      "learning_rate": 3.725149038608296e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 12050656,
      "step": 13480
    },
    {
      "epoch": 3.5590603141084864,
      "grad_norm": 0.3584168553352356,
      "learning_rate": 3.7242931655675404e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 12054944,
      "step": 13485
    },
    {
      "epoch": 3.560380097663983,
      "grad_norm": 0.4059719145298004,
      "learning_rate": 3.7234371037253937e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 12059392,
      "step": 13490
    },
    {
      "epoch": 3.56169988121948,
      "grad_norm": 0.2573276460170746,
      "learning_rate": 3.7225808532138705e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 12064096,
      "step": 13495
    },
    {
      "epoch": 3.563019664774977,
      "grad_norm": 0.16586817800998688,
      "learning_rate": 3.721724414165016e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 12068640,
      "step": 13500
    },
    {
      "epoch": 3.5643394483304736,
      "grad_norm": 0.31450891494750977,
      "learning_rate": 3.720867786710904e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 12073536,
      "step": 13505
    },
    {
      "epoch": 3.5656592318859706,
      "grad_norm": 0.11625517159700394,
      "learning_rate": 3.7200109709836366e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 12078048,
      "step": 13510
    },
    {
      "epoch": 3.5669790154414676,
      "grad_norm": 0.04604598879814148,
      "learning_rate": 3.7191539671153465e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 12082688,
      "step": 13515
    },
    {
      "epoch": 3.5682987989969646,
      "grad_norm": 0.21834248304367065,
      "learning_rate": 3.718296775238193e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 12086848,
      "step": 13520
    },
    {
      "epoch": 3.5696185825524616,
      "grad_norm": 0.30584001541137695,
      "learning_rate": 3.7174393954843675e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 12091552,
      "step": 13525
    },
    {
      "epoch": 3.570938366107958,
      "grad_norm": 0.281668484210968,
      "learning_rate": 3.716581827986087e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 12096064,
      "step": 13530
    },
    {
      "epoch": 3.572258149663455,
      "grad_norm": 0.09635057300329208,
      "learning_rate": 3.7157240728756004e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 12100544,
      "step": 13535
    },
    {
      "epoch": 3.573577933218952,
      "grad_norm": 0.4518418610095978,
      "learning_rate": 3.714866130285184e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 12105056,
      "step": 13540
    },
    {
      "epoch": 3.574897716774449,
      "grad_norm": 0.32803046703338623,
      "learning_rate": 3.714008000347143e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 12109696,
      "step": 13545
    },
    {
      "epoch": 3.576217500329946,
      "grad_norm": 0.027123291045427322,
      "learning_rate": 3.7131496831938126e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 12114528,
      "step": 13550
    },
    {
      "epoch": 3.5775372838854427,
      "grad_norm": 0.3130221664905548,
      "learning_rate": 3.7122911789575565e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 12118976,
      "step": 13555
    },
    {
      "epoch": 3.5788570674409397,
      "grad_norm": 0.1424226611852646,
      "learning_rate": 3.711432487770765e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 12123616,
      "step": 13560
    },
    {
      "epoch": 3.5801768509964367,
      "grad_norm": 0.14411087334156036,
      "learning_rate": 3.710573609765861e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 12128160,
      "step": 13565
    },
    {
      "epoch": 3.5814966345519332,
      "grad_norm": 0.19305528700351715,
      "learning_rate": 3.709714545075292e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 12132352,
      "step": 13570
    },
    {
      "epoch": 3.5828164181074302,
      "grad_norm": 0.1429882049560547,
      "learning_rate": 3.708855293831538e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 12136832,
      "step": 13575
    },
    {
      "epoch": 3.5841362016629272,
      "grad_norm": 0.07496651262044907,
      "learning_rate": 3.707995856167107e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 12141152,
      "step": 13580
    },
    {
      "epoch": 3.5854559852184242,
      "grad_norm": 0.2872995436191559,
      "learning_rate": 3.707136232214534e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 12145408,
      "step": 13585
    },
    {
      "epoch": 3.5867757687739212,
      "grad_norm": 0.08361658453941345,
      "learning_rate": 3.7062764221063844e-05,
      "loss": 0.1092,
      "num_input_tokens_seen": 12149632,
      "step": 13590
    },
    {
      "epoch": 3.588095552329418,
      "grad_norm": 0.22108393907546997,
      "learning_rate": 3.705416425975252e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 12153856,
      "step": 13595
    },
    {
      "epoch": 3.589415335884915,
      "grad_norm": 0.08470819890499115,
      "learning_rate": 3.704556243953758e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 12158496,
      "step": 13600
    },
    {
      "epoch": 3.589415335884915,
      "eval_loss": 0.07047203183174133,
      "eval_runtime": 64.7974,
      "eval_samples_per_second": 103.939,
      "eval_steps_per_second": 25.989,
      "num_input_tokens_seen": 12158496,
      "step": 13600
    },
    {
      "epoch": 3.590735119440412,
      "grad_norm": 0.059494469314813614,
      "learning_rate": 3.7036958761745535e-05,
      "loss": 0.1091,
      "num_input_tokens_seen": 12162976,
      "step": 13605
    },
    {
      "epoch": 3.592054902995909,
      "grad_norm": 0.18056075274944305,
      "learning_rate": 3.702835322770318e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 12167424,
      "step": 13610
    },
    {
      "epoch": 3.593374686551406,
      "grad_norm": 0.13806694746017456,
      "learning_rate": 3.701974583873761e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 12172064,
      "step": 13615
    },
    {
      "epoch": 3.5946944701069024,
      "grad_norm": 0.2815007269382477,
      "learning_rate": 3.701113659617618e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 12176608,
      "step": 13620
    },
    {
      "epoch": 3.5960142536623994,
      "grad_norm": 0.08173112571239471,
      "learning_rate": 3.7002525501346535e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 12180608,
      "step": 13625
    },
    {
      "epoch": 3.5973340372178964,
      "grad_norm": 0.27852824330329895,
      "learning_rate": 3.699391255557664e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 12185088,
      "step": 13630
    },
    {
      "epoch": 3.598653820773393,
      "grad_norm": 0.13107930123806,
      "learning_rate": 3.69852977601947e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 12189504,
      "step": 13635
    },
    {
      "epoch": 3.59997360432889,
      "grad_norm": 0.5061914920806885,
      "learning_rate": 3.697668111652922e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 12193760,
      "step": 13640
    },
    {
      "epoch": 3.601293387884387,
      "grad_norm": 0.1426459699869156,
      "learning_rate": 3.6968062625909005e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 12198208,
      "step": 13645
    },
    {
      "epoch": 3.602613171439884,
      "grad_norm": 0.13407565653324127,
      "learning_rate": 3.6959442289663135e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 12202624,
      "step": 13650
    },
    {
      "epoch": 3.603932954995381,
      "grad_norm": 0.11071648448705673,
      "learning_rate": 3.695082010912098e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 12207200,
      "step": 13655
    },
    {
      "epoch": 3.6052527385508775,
      "grad_norm": 0.3224470913410187,
      "learning_rate": 3.694219608561217e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 12211776,
      "step": 13660
    },
    {
      "epoch": 3.6065725221063745,
      "grad_norm": 0.21437466144561768,
      "learning_rate": 3.693357022046665e-05,
      "loss": 0.1175,
      "num_input_tokens_seen": 12216352,
      "step": 13665
    },
    {
      "epoch": 3.6078923056618715,
      "grad_norm": 0.1642499417066574,
      "learning_rate": 3.6924942515014644e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 12220832,
      "step": 13670
    },
    {
      "epoch": 3.6092120892173685,
      "grad_norm": 0.35209134221076965,
      "learning_rate": 3.691631297058664e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 12225344,
      "step": 13675
    },
    {
      "epoch": 3.6105318727728655,
      "grad_norm": 0.3754257261753082,
      "learning_rate": 3.6907681588513424e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 12229696,
      "step": 13680
    },
    {
      "epoch": 3.611851656328362,
      "grad_norm": 0.20746377110481262,
      "learning_rate": 3.689904837012606e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 12234176,
      "step": 13685
    },
    {
      "epoch": 3.613171439883859,
      "grad_norm": 0.0922841876745224,
      "learning_rate": 3.689041331675591e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 12238688,
      "step": 13690
    },
    {
      "epoch": 3.614491223439356,
      "grad_norm": 0.2862519919872284,
      "learning_rate": 3.688177642973461e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 12243200,
      "step": 13695
    },
    {
      "epoch": 3.6158110069948526,
      "grad_norm": 0.24763265252113342,
      "learning_rate": 3.687313771039406e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 12247712,
      "step": 13700
    },
    {
      "epoch": 3.6171307905503496,
      "grad_norm": 0.3014957904815674,
      "learning_rate": 3.686449716006647e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 12252256,
      "step": 13705
    },
    {
      "epoch": 3.6184505741058466,
      "grad_norm": 0.11168388277292252,
      "learning_rate": 3.685585478008432e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 12256960,
      "step": 13710
    },
    {
      "epoch": 3.6197703576613436,
      "grad_norm": 0.06092044711112976,
      "learning_rate": 3.6847210571780364e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 12261376,
      "step": 13715
    },
    {
      "epoch": 3.6210901412168406,
      "grad_norm": 0.32533618807792664,
      "learning_rate": 3.683856453648767e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 12265952,
      "step": 13720
    },
    {
      "epoch": 3.622409924772337,
      "grad_norm": 0.3705858588218689,
      "learning_rate": 3.682991667553954e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 12270528,
      "step": 13725
    },
    {
      "epoch": 3.623729708327834,
      "grad_norm": 0.23150433599948883,
      "learning_rate": 3.6821266990269606e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 12274912,
      "step": 13730
    },
    {
      "epoch": 3.625049491883331,
      "grad_norm": 0.17243900895118713,
      "learning_rate": 3.681261548201174e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 12279360,
      "step": 13735
    },
    {
      "epoch": 3.626369275438828,
      "grad_norm": 0.17127864062786102,
      "learning_rate": 3.6803962152100125e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 12283840,
      "step": 13740
    },
    {
      "epoch": 3.627689058994325,
      "grad_norm": 0.4241301715373993,
      "learning_rate": 3.67953070018692e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 12288352,
      "step": 13745
    },
    {
      "epoch": 3.6290088425498217,
      "grad_norm": 0.21278217434883118,
      "learning_rate": 3.678665003265371e-05,
      "loss": 0.0906,
      "num_input_tokens_seen": 12292576,
      "step": 13750
    },
    {
      "epoch": 3.6303286261053187,
      "grad_norm": 0.04317848011851311,
      "learning_rate": 3.677799124578867e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 12296896,
      "step": 13755
    },
    {
      "epoch": 3.6316484096608157,
      "grad_norm": 0.09849812835454941,
      "learning_rate": 3.676933064260937e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 12301184,
      "step": 13760
    },
    {
      "epoch": 3.6329681932163123,
      "grad_norm": 0.18287646770477295,
      "learning_rate": 3.6760668224451365e-05,
      "loss": 0.1015,
      "num_input_tokens_seen": 12305760,
      "step": 13765
    },
    {
      "epoch": 3.6342879767718093,
      "grad_norm": 0.09911791980266571,
      "learning_rate": 3.675200399265054e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 12310240,
      "step": 13770
    },
    {
      "epoch": 3.6356077603273063,
      "grad_norm": 0.2794325649738312,
      "learning_rate": 3.6743337948543014e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 12314624,
      "step": 13775
    },
    {
      "epoch": 3.6369275438828033,
      "grad_norm": 0.08199980854988098,
      "learning_rate": 3.6734670093465204e-05,
      "loss": 0.0741,
      "num_input_tokens_seen": 12319072,
      "step": 13780
    },
    {
      "epoch": 3.6382473274383003,
      "grad_norm": 0.22547121345996857,
      "learning_rate": 3.672600042875379e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 12323392,
      "step": 13785
    },
    {
      "epoch": 3.639567110993797,
      "grad_norm": 0.549271285533905,
      "learning_rate": 3.671732895574575e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 12327776,
      "step": 13790
    },
    {
      "epoch": 3.640886894549294,
      "grad_norm": 0.2668686807155609,
      "learning_rate": 3.670865567577834e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 12332512,
      "step": 13795
    },
    {
      "epoch": 3.642206678104791,
      "grad_norm": 0.27998656034469604,
      "learning_rate": 3.669998059018909e-05,
      "loss": 0.0441,
      "num_input_tokens_seen": 12336928,
      "step": 13800
    },
    {
      "epoch": 3.642206678104791,
      "eval_loss": 0.07020489126443863,
      "eval_runtime": 64.7322,
      "eval_samples_per_second": 104.044,
      "eval_steps_per_second": 26.015,
      "num_input_tokens_seen": 12336928,
      "step": 13800
    },
    {
      "epoch": 3.643526461660288,
      "grad_norm": 0.20211559534072876,
      "learning_rate": 3.6691303700315796e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 12341632,
      "step": 13805
    },
    {
      "epoch": 3.644846245215785,
      "grad_norm": 0.46745795011520386,
      "learning_rate": 3.668262500749655e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 12346272,
      "step": 13810
    },
    {
      "epoch": 3.6461660287712814,
      "grad_norm": 0.08376599103212357,
      "learning_rate": 3.667394451306971e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 12350816,
      "step": 13815
    },
    {
      "epoch": 3.6474858123267784,
      "grad_norm": 0.27907389402389526,
      "learning_rate": 3.666526221837393e-05,
      "loss": 0.115,
      "num_input_tokens_seen": 12355136,
      "step": 13820
    },
    {
      "epoch": 3.6488055958822754,
      "grad_norm": 0.24345633387565613,
      "learning_rate": 3.665657812474812e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 12360064,
      "step": 13825
    },
    {
      "epoch": 3.650125379437772,
      "grad_norm": 0.03641634061932564,
      "learning_rate": 3.664789223353147e-05,
      "loss": 0.0144,
      "num_input_tokens_seen": 12364704,
      "step": 13830
    },
    {
      "epoch": 3.6514451629932694,
      "grad_norm": 0.04545697197318077,
      "learning_rate": 3.663920454606347e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 12369248,
      "step": 13835
    },
    {
      "epoch": 3.652764946548766,
      "grad_norm": 0.09697481989860535,
      "learning_rate": 3.6630515063683856e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 12373632,
      "step": 13840
    },
    {
      "epoch": 3.654084730104263,
      "grad_norm": 0.07207610458135605,
      "learning_rate": 3.662182378773267e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 12377952,
      "step": 13845
    },
    {
      "epoch": 3.65540451365976,
      "grad_norm": 0.24276262521743774,
      "learning_rate": 3.66131307195502e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 12382688,
      "step": 13850
    },
    {
      "epoch": 3.6567242972152565,
      "grad_norm": 0.28794342279434204,
      "learning_rate": 3.6604435860477034e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 12386848,
      "step": 13855
    },
    {
      "epoch": 3.6580440807707535,
      "grad_norm": 0.07739073783159256,
      "learning_rate": 3.6595739211854025e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 12391264,
      "step": 13860
    },
    {
      "epoch": 3.6593638643262505,
      "grad_norm": 0.38529080152511597,
      "learning_rate": 3.658704077502231e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 12395616,
      "step": 13865
    },
    {
      "epoch": 3.6606836478817475,
      "grad_norm": 0.5888718366622925,
      "learning_rate": 3.65783405513233e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 12400224,
      "step": 13870
    },
    {
      "epoch": 3.6620034314372445,
      "grad_norm": 0.5400272011756897,
      "learning_rate": 3.656963854209867e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 12404704,
      "step": 13875
    },
    {
      "epoch": 3.663323214992741,
      "grad_norm": 0.10038716346025467,
      "learning_rate": 3.656093474869038e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 12409280,
      "step": 13880
    },
    {
      "epoch": 3.664642998548238,
      "grad_norm": 0.34208759665489197,
      "learning_rate": 3.655222917244068e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 12413664,
      "step": 13885
    },
    {
      "epoch": 3.665962782103735,
      "grad_norm": 0.3931296765804291,
      "learning_rate": 3.6543521814692054e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 12418336,
      "step": 13890
    },
    {
      "epoch": 3.667282565659232,
      "grad_norm": 0.3020230829715729,
      "learning_rate": 3.653481267678731e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 12422720,
      "step": 13895
    },
    {
      "epoch": 3.668602349214729,
      "grad_norm": 0.1967891901731491,
      "learning_rate": 3.652610176006949e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 12427104,
      "step": 13900
    },
    {
      "epoch": 3.6699221327702256,
      "grad_norm": 0.08594027906656265,
      "learning_rate": 3.6517389065881925e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 12431616,
      "step": 13905
    },
    {
      "epoch": 3.6712419163257226,
      "grad_norm": 0.2151394784450531,
      "learning_rate": 3.650867459556824e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 12436064,
      "step": 13910
    },
    {
      "epoch": 3.6725616998812196,
      "grad_norm": 0.05408387631177902,
      "learning_rate": 3.64999583504723e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 12440320,
      "step": 13915
    },
    {
      "epoch": 3.673881483436716,
      "grad_norm": 0.038490179926157,
      "learning_rate": 3.649124033193827e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 12444960,
      "step": 13920
    },
    {
      "epoch": 3.675201266992213,
      "grad_norm": 0.24906596541404724,
      "learning_rate": 3.648252054131057e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 12449216,
      "step": 13925
    },
    {
      "epoch": 3.67652105054771,
      "grad_norm": 0.3120221197605133,
      "learning_rate": 3.647379897993391e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 12454176,
      "step": 13930
    },
    {
      "epoch": 3.677840834103207,
      "grad_norm": 0.08492430299520493,
      "learning_rate": 3.646507564915325e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 12458592,
      "step": 13935
    },
    {
      "epoch": 3.679160617658704,
      "grad_norm": 0.28929901123046875,
      "learning_rate": 3.645635055031385e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 12463296,
      "step": 13940
    },
    {
      "epoch": 3.6804804012142007,
      "grad_norm": 0.05457812175154686,
      "learning_rate": 3.6447623684761224e-05,
      "loss": 0.1091,
      "num_input_tokens_seen": 12467744,
      "step": 13945
    },
    {
      "epoch": 3.6818001847696977,
      "grad_norm": 0.09262394160032272,
      "learning_rate": 3.643889505384117e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 12471936,
      "step": 13950
    },
    {
      "epoch": 3.6831199683251947,
      "grad_norm": 0.31971409916877747,
      "learning_rate": 3.6430164658899744e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 12476064,
      "step": 13955
    },
    {
      "epoch": 3.6844397518806917,
      "grad_norm": 0.34457284212112427,
      "learning_rate": 3.642143250128329e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 12480512,
      "step": 13960
    },
    {
      "epoch": 3.6857595354361887,
      "grad_norm": 0.2229415327310562,
      "learning_rate": 3.641269858233841e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 12485088,
      "step": 13965
    },
    {
      "epoch": 3.6870793189916853,
      "grad_norm": 0.22079087793827057,
      "learning_rate": 3.640396290341199e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 12489696,
      "step": 13970
    },
    {
      "epoch": 3.6883991025471823,
      "grad_norm": 0.2719988226890564,
      "learning_rate": 3.639522546585118e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 12494080,
      "step": 13975
    },
    {
      "epoch": 3.6897188861026793,
      "grad_norm": 0.24038533866405487,
      "learning_rate": 3.6386486271003404e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 12498688,
      "step": 13980
    },
    {
      "epoch": 3.691038669658176,
      "grad_norm": 0.19189786911010742,
      "learning_rate": 3.6377745320216346e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 12502976,
      "step": 13985
    },
    {
      "epoch": 3.692358453213673,
      "grad_norm": 0.29798004031181335,
      "learning_rate": 3.636900261483798e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 12507648,
      "step": 13990
    },
    {
      "epoch": 3.69367823676917,
      "grad_norm": 0.24173560738563538,
      "learning_rate": 3.636025815621654e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 12512352,
      "step": 13995
    },
    {
      "epoch": 3.694998020324667,
      "grad_norm": 0.23445792496204376,
      "learning_rate": 3.635151194570054e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 12516800,
      "step": 14000
    },
    {
      "epoch": 3.694998020324667,
      "eval_loss": 0.07012901455163956,
      "eval_runtime": 64.7621,
      "eval_samples_per_second": 103.996,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 12516800,
      "step": 14000
    },
    {
      "epoch": 3.696317803880164,
      "grad_norm": 0.3708011507987976,
      "learning_rate": 3.634276398463873e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 12521632,
      "step": 14005
    },
    {
      "epoch": 3.6976375874356604,
      "grad_norm": 0.06707891821861267,
      "learning_rate": 3.633401427438018e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 12525920,
      "step": 14010
    },
    {
      "epoch": 3.6989573709911574,
      "grad_norm": 0.5043233036994934,
      "learning_rate": 3.63252628162742e-05,
      "loss": 0.1065,
      "num_input_tokens_seen": 12530368,
      "step": 14015
    },
    {
      "epoch": 3.7002771545466544,
      "grad_norm": 0.15913519263267517,
      "learning_rate": 3.6316509611670364e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 12534848,
      "step": 14020
    },
    {
      "epoch": 3.7015969381021514,
      "grad_norm": 0.05630084499716759,
      "learning_rate": 3.630775466191854e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 12539296,
      "step": 14025
    },
    {
      "epoch": 3.7029167216576484,
      "grad_norm": 0.0715685710310936,
      "learning_rate": 3.629899796836884e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 12544000,
      "step": 14030
    },
    {
      "epoch": 3.704236505213145,
      "grad_norm": 0.10906235873699188,
      "learning_rate": 3.6290239532371666e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 12548608,
      "step": 14035
    },
    {
      "epoch": 3.705556288768642,
      "grad_norm": 0.2301749587059021,
      "learning_rate": 3.628147935527767e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 12553536,
      "step": 14040
    },
    {
      "epoch": 3.706876072324139,
      "grad_norm": 0.03506452217698097,
      "learning_rate": 3.627271743843779e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 12557920,
      "step": 14045
    },
    {
      "epoch": 3.7081958558796355,
      "grad_norm": 0.28077152371406555,
      "learning_rate": 3.626395378320321e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 12562560,
      "step": 14050
    },
    {
      "epoch": 3.7095156394351325,
      "grad_norm": 0.047922611236572266,
      "learning_rate": 3.625518839092541e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 12566816,
      "step": 14055
    },
    {
      "epoch": 3.7108354229906295,
      "grad_norm": 0.19901736080646515,
      "learning_rate": 3.624642126295612e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 12571456,
      "step": 14060
    },
    {
      "epoch": 3.7121552065461265,
      "grad_norm": 0.12431244552135468,
      "learning_rate": 3.6237652400647345e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 12575680,
      "step": 14065
    },
    {
      "epoch": 3.7134749901016235,
      "grad_norm": 0.05594848096370697,
      "learning_rate": 3.622888180535134e-05,
      "loss": 0.0282,
      "num_input_tokens_seen": 12580160,
      "step": 14070
    },
    {
      "epoch": 3.71479477365712,
      "grad_norm": 0.1305118203163147,
      "learning_rate": 3.6220109478420655e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 12584608,
      "step": 14075
    },
    {
      "epoch": 3.716114557212617,
      "grad_norm": 0.25873133540153503,
      "learning_rate": 3.6211335421208084e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 12589120,
      "step": 14080
    },
    {
      "epoch": 3.717434340768114,
      "grad_norm": 0.443569540977478,
      "learning_rate": 3.62025596350667e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 12593440,
      "step": 14085
    },
    {
      "epoch": 3.718754124323611,
      "grad_norm": 0.11574344336986542,
      "learning_rate": 3.619378212134984e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 12597856,
      "step": 14090
    },
    {
      "epoch": 3.720073907879108,
      "grad_norm": 0.11435899883508682,
      "learning_rate": 3.618500288141111e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 12602400,
      "step": 14095
    },
    {
      "epoch": 3.7213936914346046,
      "grad_norm": 0.4830787777900696,
      "learning_rate": 3.617622191660438e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 12606912,
      "step": 14100
    },
    {
      "epoch": 3.7227134749901016,
      "grad_norm": 0.25740718841552734,
      "learning_rate": 3.616743922828377e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 12611072,
      "step": 14105
    },
    {
      "epoch": 3.7240332585455986,
      "grad_norm": 0.05857228860259056,
      "learning_rate": 3.615865481780371e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 12615488,
      "step": 14110
    },
    {
      "epoch": 3.725353042101095,
      "grad_norm": 0.1826494336128235,
      "learning_rate": 3.614986868651883e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 12619936,
      "step": 14115
    },
    {
      "epoch": 3.726672825656592,
      "grad_norm": 0.2047070562839508,
      "learning_rate": 3.614108083578409e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 12624512,
      "step": 14120
    },
    {
      "epoch": 3.727992609212089,
      "grad_norm": 0.2482721507549286,
      "learning_rate": 3.613229126695467e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 12628768,
      "step": 14125
    },
    {
      "epoch": 3.729312392767586,
      "grad_norm": 0.12914778292179108,
      "learning_rate": 3.612349998138605e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 12633344,
      "step": 14130
    },
    {
      "epoch": 3.730632176323083,
      "grad_norm": 0.11181078106164932,
      "learning_rate": 3.6114706980433946e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 12638048,
      "step": 14135
    },
    {
      "epoch": 3.7319519598785797,
      "grad_norm": 0.2526971101760864,
      "learning_rate": 3.610591226545435e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 12642464,
      "step": 14140
    },
    {
      "epoch": 3.7332717434340768,
      "grad_norm": 0.06576159596443176,
      "learning_rate": 3.6097115837803505e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 12646880,
      "step": 14145
    },
    {
      "epoch": 3.7345915269895738,
      "grad_norm": 0.3647424578666687,
      "learning_rate": 3.608831769883795e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 12651072,
      "step": 14150
    },
    {
      "epoch": 3.7359113105450708,
      "grad_norm": 0.35942384600639343,
      "learning_rate": 3.607951784991446e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 12655776,
      "step": 14155
    },
    {
      "epoch": 3.7372310941005678,
      "grad_norm": 0.1862003356218338,
      "learning_rate": 3.6070716292390085e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 12660000,
      "step": 14160
    },
    {
      "epoch": 3.7385508776560643,
      "grad_norm": 0.5439689755439758,
      "learning_rate": 3.606191302762213e-05,
      "loss": 0.1412,
      "num_input_tokens_seen": 12664864,
      "step": 14165
    },
    {
      "epoch": 3.7398706612115613,
      "grad_norm": 0.2243514358997345,
      "learning_rate": 3.605310805696818e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 12669376,
      "step": 14170
    },
    {
      "epoch": 3.7411904447670583,
      "grad_norm": 0.24459627270698547,
      "learning_rate": 3.6044301381786067e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 12673600,
      "step": 14175
    },
    {
      "epoch": 3.742510228322555,
      "grad_norm": 0.2941947281360626,
      "learning_rate": 3.6035493003433883e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 12677760,
      "step": 14180
    },
    {
      "epoch": 3.743830011878052,
      "grad_norm": 0.250480055809021,
      "learning_rate": 3.6026682923269994e-05,
      "loss": 0.1254,
      "num_input_tokens_seen": 12682240,
      "step": 14185
    },
    {
      "epoch": 3.745149795433549,
      "grad_norm": 0.2735949158668518,
      "learning_rate": 3.6017871142653034e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 12686784,
      "step": 14190
    },
    {
      "epoch": 3.746469578989046,
      "grad_norm": 0.21321110427379608,
      "learning_rate": 3.600905766294189e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 12691360,
      "step": 14195
    },
    {
      "epoch": 3.747789362544543,
      "grad_norm": 0.04726054519414902,
      "learning_rate": 3.60002424854957e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 12695648,
      "step": 14200
    },
    {
      "epoch": 3.747789362544543,
      "eval_loss": 0.07010029256343842,
      "eval_runtime": 65.0734,
      "eval_samples_per_second": 103.499,
      "eval_steps_per_second": 25.878,
      "num_input_tokens_seen": 12695648,
      "step": 14200
    },
    {
      "epoch": 3.7491091461000394,
      "grad_norm": 0.19516563415527344,
      "learning_rate": 3.5991425611673876e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 12700288,
      "step": 14205
    },
    {
      "epoch": 3.7504289296555364,
      "grad_norm": 0.40891969203948975,
      "learning_rate": 3.5982607042836105e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 12704544,
      "step": 14210
    },
    {
      "epoch": 3.7517487132110334,
      "grad_norm": 0.7194133400917053,
      "learning_rate": 3.597378678034231e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 12709184,
      "step": 14215
    },
    {
      "epoch": 3.7530684967665304,
      "grad_norm": 0.24140232801437378,
      "learning_rate": 3.596496482555269e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 12713920,
      "step": 14220
    },
    {
      "epoch": 3.7543882803220274,
      "grad_norm": 0.1821589171886444,
      "learning_rate": 3.595614117982769e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 12718336,
      "step": 14225
    },
    {
      "epoch": 3.755708063877524,
      "grad_norm": 0.09190386533737183,
      "learning_rate": 3.594731584452805e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 12722848,
      "step": 14230
    },
    {
      "epoch": 3.757027847433021,
      "grad_norm": 0.5025379061698914,
      "learning_rate": 3.593848882101472e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 12727200,
      "step": 14235
    },
    {
      "epoch": 3.758347630988518,
      "grad_norm": 0.1722545623779297,
      "learning_rate": 3.592966011064896e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 12731872,
      "step": 14240
    },
    {
      "epoch": 3.7596674145440145,
      "grad_norm": 0.2402086704969406,
      "learning_rate": 3.592082971479226e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 12736000,
      "step": 14245
    },
    {
      "epoch": 3.7609871980995115,
      "grad_norm": 0.17915412783622742,
      "learning_rate": 3.5911997634806385e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 12740512,
      "step": 14250
    },
    {
      "epoch": 3.7623069816550085,
      "grad_norm": 0.18790321052074432,
      "learning_rate": 3.5903163872053336e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 12745056,
      "step": 14255
    },
    {
      "epoch": 3.7636267652105055,
      "grad_norm": 0.06330185383558273,
      "learning_rate": 3.58943284278954e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 12749376,
      "step": 14260
    },
    {
      "epoch": 3.7649465487660025,
      "grad_norm": 0.072992704808712,
      "learning_rate": 3.588549130369512e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 12753888,
      "step": 14265
    },
    {
      "epoch": 3.766266332321499,
      "grad_norm": 0.10988523811101913,
      "learning_rate": 3.5876652500815274e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 12758432,
      "step": 14270
    },
    {
      "epoch": 3.767586115876996,
      "grad_norm": 0.20684972405433655,
      "learning_rate": 3.586781202061894e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 12763200,
      "step": 14275
    },
    {
      "epoch": 3.768905899432493,
      "grad_norm": 0.21513719856739044,
      "learning_rate": 3.585896986446942e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 12767680,
      "step": 14280
    },
    {
      "epoch": 3.77022568298799,
      "grad_norm": 0.15895283222198486,
      "learning_rate": 3.585012603373028e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 12771872,
      "step": 14285
    },
    {
      "epoch": 3.771545466543487,
      "grad_norm": 0.22698146104812622,
      "learning_rate": 3.584128052976535e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 12776224,
      "step": 14290
    },
    {
      "epoch": 3.7728652500989837,
      "grad_norm": 0.34538212418556213,
      "learning_rate": 3.5832433353938724e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 12780640,
      "step": 14295
    },
    {
      "epoch": 3.7741850336544807,
      "grad_norm": 0.11162048578262329,
      "learning_rate": 3.5823584507614746e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 12785344,
      "step": 14300
    },
    {
      "epoch": 3.7755048172099777,
      "grad_norm": 0.2139614224433899,
      "learning_rate": 3.581473399215802e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 12789600,
      "step": 14305
    },
    {
      "epoch": 3.776824600765474,
      "grad_norm": 0.08781816065311432,
      "learning_rate": 3.580588180893341e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 12794272,
      "step": 14310
    },
    {
      "epoch": 3.778144384320971,
      "grad_norm": 0.2113962024450302,
      "learning_rate": 3.579702795930602e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 12798720,
      "step": 14315
    },
    {
      "epoch": 3.779464167876468,
      "grad_norm": 0.2857036590576172,
      "learning_rate": 3.578817244464125e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 12803392,
      "step": 14320
    },
    {
      "epoch": 3.780783951431965,
      "grad_norm": 0.33682477474212646,
      "learning_rate": 3.577931526630471e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 12807968,
      "step": 14325
    },
    {
      "epoch": 3.782103734987462,
      "grad_norm": 0.07584571093320847,
      "learning_rate": 3.577045642566229e-05,
      "loss": 0.0263,
      "num_input_tokens_seen": 12812224,
      "step": 14330
    },
    {
      "epoch": 3.7834235185429588,
      "grad_norm": 0.0941348597407341,
      "learning_rate": 3.576159592408014e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 12816576,
      "step": 14335
    },
    {
      "epoch": 3.7847433020984558,
      "grad_norm": 0.07598311454057693,
      "learning_rate": 3.575273376292466e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 12820672,
      "step": 14340
    },
    {
      "epoch": 3.7860630856539528,
      "grad_norm": 0.5496885180473328,
      "learning_rate": 3.574386994356251e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 12824960,
      "step": 14345
    },
    {
      "epoch": 3.7873828692094498,
      "grad_norm": 0.42316362261772156,
      "learning_rate": 3.573500446736059e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 12829504,
      "step": 14350
    },
    {
      "epoch": 3.7887026527649468,
      "grad_norm": 0.3387746214866638,
      "learning_rate": 3.5726137335686094e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 12833888,
      "step": 14355
    },
    {
      "epoch": 3.7900224363204433,
      "grad_norm": 0.2717480957508087,
      "learning_rate": 3.571726854990642e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 12838624,
      "step": 14360
    },
    {
      "epoch": 3.7913422198759403,
      "grad_norm": 0.19734783470630646,
      "learning_rate": 3.570839811138925e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 12843008,
      "step": 14365
    },
    {
      "epoch": 3.7926620034314373,
      "grad_norm": 0.22105637192726135,
      "learning_rate": 3.569952602150252e-05,
      "loss": 0.0725,
      "num_input_tokens_seen": 12847264,
      "step": 14370
    },
    {
      "epoch": 3.793981786986934,
      "grad_norm": 0.13598424196243286,
      "learning_rate": 3.569065228161442e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 12852000,
      "step": 14375
    },
    {
      "epoch": 3.795301570542431,
      "grad_norm": 0.23259256780147552,
      "learning_rate": 3.5681776893093395e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 12856608,
      "step": 14380
    },
    {
      "epoch": 3.796621354097928,
      "grad_norm": 0.13966037333011627,
      "learning_rate": 3.5672899857308134e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 12861024,
      "step": 14385
    },
    {
      "epoch": 3.797941137653425,
      "grad_norm": 0.13671354949474335,
      "learning_rate": 3.566402117562759e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 12865600,
      "step": 14390
    },
    {
      "epoch": 3.799260921208922,
      "grad_norm": 0.1927642673254013,
      "learning_rate": 3.565514084942097e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 12870080,
      "step": 14395
    },
    {
      "epoch": 3.8005807047644184,
      "grad_norm": 0.3473242521286011,
      "learning_rate": 3.564625888005773e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 12874656,
      "step": 14400
    },
    {
      "epoch": 3.8005807047644184,
      "eval_loss": 0.07020100206136703,
      "eval_runtime": 64.7445,
      "eval_samples_per_second": 104.024,
      "eval_steps_per_second": 26.01,
      "num_input_tokens_seen": 12874656,
      "step": 14400
    },
    {
      "epoch": 3.8019004883199154,
      "grad_norm": 0.0446460098028183,
      "learning_rate": 3.563737526890759e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 12879392,
      "step": 14405
    },
    {
      "epoch": 3.8032202718754125,
      "grad_norm": 0.21624614298343658,
      "learning_rate": 3.562849001734049e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 12883744,
      "step": 14410
    },
    {
      "epoch": 3.8045400554309095,
      "grad_norm": 0.12420845031738281,
      "learning_rate": 3.561960312672667e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 12887936,
      "step": 14415
    },
    {
      "epoch": 3.8058598389864065,
      "grad_norm": 0.223580002784729,
      "learning_rate": 3.5610714598436596e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 12892256,
      "step": 14420
    },
    {
      "epoch": 3.807179622541903,
      "grad_norm": 0.11566367745399475,
      "learning_rate": 3.5601824433840986e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 12897152,
      "step": 14425
    },
    {
      "epoch": 3.8084994060974,
      "grad_norm": 0.06418372690677643,
      "learning_rate": 3.559293263431082e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 12901504,
      "step": 14430
    },
    {
      "epoch": 3.809819189652897,
      "grad_norm": 0.04414607957005501,
      "learning_rate": 3.558403920121732e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 12905920,
      "step": 14435
    },
    {
      "epoch": 3.8111389732083936,
      "grad_norm": 0.12241875380277634,
      "learning_rate": 3.557514413593197e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 12910432,
      "step": 14440
    },
    {
      "epoch": 3.8124587567638906,
      "grad_norm": 0.06376060098409653,
      "learning_rate": 3.55662474398265e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 12915040,
      "step": 14445
    },
    {
      "epoch": 3.8137785403193876,
      "grad_norm": 0.1842741072177887,
      "learning_rate": 3.555734911427288e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 12919328,
      "step": 14450
    },
    {
      "epoch": 3.8150983238748846,
      "grad_norm": 0.06489312648773193,
      "learning_rate": 3.5548449160643363e-05,
      "loss": 0.0994,
      "num_input_tokens_seen": 12923936,
      "step": 14455
    },
    {
      "epoch": 3.8164181074303816,
      "grad_norm": 0.1813816875219345,
      "learning_rate": 3.553954758031043e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 12928672,
      "step": 14460
    },
    {
      "epoch": 3.817737890985878,
      "grad_norm": 0.3094000220298767,
      "learning_rate": 3.5530644374646815e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 12932960,
      "step": 14465
    },
    {
      "epoch": 3.819057674541375,
      "grad_norm": 0.3110158145427704,
      "learning_rate": 3.552173954502549e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 12937568,
      "step": 14470
    },
    {
      "epoch": 3.820377458096872,
      "grad_norm": 0.08041540533304214,
      "learning_rate": 3.55128330928197e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 12941856,
      "step": 14475
    },
    {
      "epoch": 3.821697241652369,
      "grad_norm": 0.22771687805652618,
      "learning_rate": 3.550392501940294e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 12946560,
      "step": 14480
    },
    {
      "epoch": 3.823017025207866,
      "grad_norm": 0.18116185069084167,
      "learning_rate": 3.5495015326148945e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 12951072,
      "step": 14485
    },
    {
      "epoch": 3.8243368087633627,
      "grad_norm": 0.1734246015548706,
      "learning_rate": 3.548610401443169e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 12955744,
      "step": 14490
    },
    {
      "epoch": 3.8256565923188597,
      "grad_norm": 0.21937955915927887,
      "learning_rate": 3.547719108562543e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 12960448,
      "step": 14495
    },
    {
      "epoch": 3.8269763758743567,
      "grad_norm": 0.2469603419303894,
      "learning_rate": 3.546827654110464e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 12965184,
      "step": 14500
    },
    {
      "epoch": 3.8282961594298532,
      "grad_norm": 0.33697938919067383,
      "learning_rate": 3.545936038224405e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 12969568,
      "step": 14505
    },
    {
      "epoch": 3.8296159429853502,
      "grad_norm": 0.23930227756500244,
      "learning_rate": 3.545044261041864e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 12974112,
      "step": 14510
    },
    {
      "epoch": 3.8309357265408472,
      "grad_norm": 0.05051136389374733,
      "learning_rate": 3.5441523227003657e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 12978432,
      "step": 14515
    },
    {
      "epoch": 3.8322555100963442,
      "grad_norm": 0.15185415744781494,
      "learning_rate": 3.543260223337459e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 12982816,
      "step": 14520
    },
    {
      "epoch": 3.8335752936518412,
      "grad_norm": 0.0468449629843235,
      "learning_rate": 3.542367963090714e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 12987488,
      "step": 14525
    },
    {
      "epoch": 3.834895077207338,
      "grad_norm": 0.1370970457792282,
      "learning_rate": 3.5414755420977295e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 12992064,
      "step": 14530
    },
    {
      "epoch": 3.836214860762835,
      "grad_norm": 0.2558385729789734,
      "learning_rate": 3.54058296049613e-05,
      "loss": 0.0926,
      "num_input_tokens_seen": 12996128,
      "step": 14535
    },
    {
      "epoch": 3.837534644318332,
      "grad_norm": 0.1799997091293335,
      "learning_rate": 3.53969021842356e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 13000352,
      "step": 14540
    },
    {
      "epoch": 3.838854427873829,
      "grad_norm": 0.12324324250221252,
      "learning_rate": 3.5387973160176926e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 13005056,
      "step": 14545
    },
    {
      "epoch": 3.840174211429326,
      "grad_norm": 0.28142648935317993,
      "learning_rate": 3.537904253416224e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 13009472,
      "step": 14550
    },
    {
      "epoch": 3.8414939949848224,
      "grad_norm": 0.31049129366874695,
      "learning_rate": 3.537011030756878e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 13013728,
      "step": 14555
    },
    {
      "epoch": 3.8428137785403194,
      "grad_norm": 0.04380623623728752,
      "learning_rate": 3.536117648177399e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 13018144,
      "step": 14560
    },
    {
      "epoch": 3.8441335620958164,
      "grad_norm": 0.3408360779285431,
      "learning_rate": 3.535224105815558e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 13022528,
      "step": 14565
    },
    {
      "epoch": 3.845453345651313,
      "grad_norm": 0.30420738458633423,
      "learning_rate": 3.5343304038091494e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 13026816,
      "step": 14570
    },
    {
      "epoch": 3.8467731292068104,
      "grad_norm": 0.0713987797498703,
      "learning_rate": 3.5334365422959955e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 13031264,
      "step": 14575
    },
    {
      "epoch": 3.848092912762307,
      "grad_norm": 0.232406347990036,
      "learning_rate": 3.5325425214139396e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 13035488,
      "step": 14580
    },
    {
      "epoch": 3.849412696317804,
      "grad_norm": 0.2916276156902313,
      "learning_rate": 3.531648341300851e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 13040032,
      "step": 14585
    },
    {
      "epoch": 3.850732479873301,
      "grad_norm": 0.14552929997444153,
      "learning_rate": 3.530754002094623e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 13044448,
      "step": 14590
    },
    {
      "epoch": 3.8520522634287975,
      "grad_norm": 0.14198054373264313,
      "learning_rate": 3.529859503933175e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 13048736,
      "step": 14595
    },
    {
      "epoch": 3.8533720469842945,
      "grad_norm": 0.3221355378627777,
      "learning_rate": 3.52896484695445e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 13053184,
      "step": 14600
    },
    {
      "epoch": 3.8533720469842945,
      "eval_loss": 0.06999914348125458,
      "eval_runtime": 64.7764,
      "eval_samples_per_second": 103.973,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 13053184,
      "step": 14600
    },
    {
      "epoch": 3.8546918305397915,
      "grad_norm": 0.04182983189821243,
      "learning_rate": 3.528070031296414e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 13058240,
      "step": 14605
    },
    {
      "epoch": 3.8560116140952885,
      "grad_norm": 0.48399922251701355,
      "learning_rate": 3.5271750570970605e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 13063008,
      "step": 14610
    },
    {
      "epoch": 3.8573313976507855,
      "grad_norm": 0.24412967264652252,
      "learning_rate": 3.526279924494405e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 13067296,
      "step": 14615
    },
    {
      "epoch": 3.858651181206282,
      "grad_norm": 0.21276412904262543,
      "learning_rate": 3.5253846336264874e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 13071840,
      "step": 14620
    },
    {
      "epoch": 3.859970964761779,
      "grad_norm": 0.08351841568946838,
      "learning_rate": 3.5244891846313736e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 13076192,
      "step": 14625
    },
    {
      "epoch": 3.861290748317276,
      "grad_norm": 0.22587162256240845,
      "learning_rate": 3.5235935776471527e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 13080768,
      "step": 14630
    },
    {
      "epoch": 3.8626105318727726,
      "grad_norm": 0.10556302964687347,
      "learning_rate": 3.522697812811939e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 13085440,
      "step": 14635
    },
    {
      "epoch": 3.86393031542827,
      "grad_norm": 0.1070423275232315,
      "learning_rate": 3.521801890263871e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 13089952,
      "step": 14640
    },
    {
      "epoch": 3.8652500989837666,
      "grad_norm": 0.13928936421871185,
      "learning_rate": 3.5209058101411114e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 13094336,
      "step": 14645
    },
    {
      "epoch": 3.8665698825392636,
      "grad_norm": 0.07713605463504791,
      "learning_rate": 3.520009572581845e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 13099136,
      "step": 14650
    },
    {
      "epoch": 3.8678896660947606,
      "grad_norm": 0.044993143528699875,
      "learning_rate": 3.519113177724285e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 13103520,
      "step": 14655
    },
    {
      "epoch": 3.869209449650257,
      "grad_norm": 0.2854921519756317,
      "learning_rate": 3.5182166257066656e-05,
      "loss": 0.1164,
      "num_input_tokens_seen": 13107648,
      "step": 14660
    },
    {
      "epoch": 3.870529233205754,
      "grad_norm": 0.11618240177631378,
      "learning_rate": 3.517319916667247e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 13112256,
      "step": 14665
    },
    {
      "epoch": 3.871849016761251,
      "grad_norm": 0.07345984876155853,
      "learning_rate": 3.516423050744313e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 13117088,
      "step": 14670
    },
    {
      "epoch": 3.873168800316748,
      "grad_norm": 0.13963858783245087,
      "learning_rate": 3.5155260280761704e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 13121472,
      "step": 14675
    },
    {
      "epoch": 3.874488583872245,
      "grad_norm": 0.0901065468788147,
      "learning_rate": 3.514628848801154e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 13125856,
      "step": 14680
    },
    {
      "epoch": 3.8758083674277417,
      "grad_norm": 0.14650169014930725,
      "learning_rate": 3.5137315130576174e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 13130240,
      "step": 14685
    },
    {
      "epoch": 3.8771281509832387,
      "grad_norm": 0.4479154646396637,
      "learning_rate": 3.512834020983942e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 13134592,
      "step": 14690
    },
    {
      "epoch": 3.8784479345387357,
      "grad_norm": 0.2864721417427063,
      "learning_rate": 3.5119363727185334e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 13139040,
      "step": 14695
    },
    {
      "epoch": 3.8797677180942327,
      "grad_norm": 0.045504871755838394,
      "learning_rate": 3.511038568399819e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 13143360,
      "step": 14700
    },
    {
      "epoch": 3.8810875016497297,
      "grad_norm": 0.17049123346805573,
      "learning_rate": 3.510140608166251e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 13147552,
      "step": 14705
    },
    {
      "epoch": 3.8824072852052263,
      "grad_norm": 0.059922803193330765,
      "learning_rate": 3.509242492156308e-05,
      "loss": 0.0119,
      "num_input_tokens_seen": 13152096,
      "step": 14710
    },
    {
      "epoch": 3.8837270687607233,
      "grad_norm": 0.11769622564315796,
      "learning_rate": 3.5083442205084896e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 13156480,
      "step": 14715
    },
    {
      "epoch": 3.8850468523162203,
      "grad_norm": 0.23994484543800354,
      "learning_rate": 3.507445793361321e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 13160864,
      "step": 14720
    },
    {
      "epoch": 3.886366635871717,
      "grad_norm": 0.32888466119766235,
      "learning_rate": 3.5065472108533505e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 13165152,
      "step": 14725
    },
    {
      "epoch": 3.887686419427214,
      "grad_norm": 0.4879530370235443,
      "learning_rate": 3.5056484731231504e-05,
      "loss": 0.1403,
      "num_input_tokens_seen": 13169664,
      "step": 14730
    },
    {
      "epoch": 3.889006202982711,
      "grad_norm": 0.13911044597625732,
      "learning_rate": 3.504749580309319e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 13174144,
      "step": 14735
    },
    {
      "epoch": 3.890325986538208,
      "grad_norm": 0.26912108063697815,
      "learning_rate": 3.5038505325504753e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 13178720,
      "step": 14740
    },
    {
      "epoch": 3.891645770093705,
      "grad_norm": 0.05984117090702057,
      "learning_rate": 3.502951329985264e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 13183072,
      "step": 14745
    },
    {
      "epoch": 3.8929655536492014,
      "grad_norm": 0.39965179562568665,
      "learning_rate": 3.502051972752354e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 13187424,
      "step": 14750
    },
    {
      "epoch": 3.8942853372046984,
      "grad_norm": 0.03279373422265053,
      "learning_rate": 3.5011524609904374e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 13191936,
      "step": 14755
    },
    {
      "epoch": 3.8956051207601954,
      "grad_norm": 0.10416101664304733,
      "learning_rate": 3.50025279483823e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 13196160,
      "step": 14760
    },
    {
      "epoch": 3.8969249043156924,
      "grad_norm": 0.29175421595573425,
      "learning_rate": 3.499352974434472e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 13200608,
      "step": 14765
    },
    {
      "epoch": 3.8982446878711894,
      "grad_norm": 0.15377305448055267,
      "learning_rate": 3.498452999917926e-05,
      "loss": 0.1065,
      "num_input_tokens_seen": 13205248,
      "step": 14770
    },
    {
      "epoch": 3.899564471426686,
      "grad_norm": 0.793651819229126,
      "learning_rate": 3.4975528714273795e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 13209984,
      "step": 14775
    },
    {
      "epoch": 3.900884254982183,
      "grad_norm": 0.06444971263408661,
      "learning_rate": 3.4966525891016454e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 13214592,
      "step": 14780
    },
    {
      "epoch": 3.90220403853768,
      "grad_norm": 0.1338070034980774,
      "learning_rate": 3.495752153079557e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 13218912,
      "step": 14785
    },
    {
      "epoch": 3.9035238220931765,
      "grad_norm": 0.037054501473903656,
      "learning_rate": 3.494851563499974e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 13223616,
      "step": 14790
    },
    {
      "epoch": 3.9048436056486735,
      "grad_norm": 0.219995379447937,
      "learning_rate": 3.493950820501777e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 13228064,
      "step": 14795
    },
    {
      "epoch": 3.9061633892041705,
      "grad_norm": 0.10364536195993423,
      "learning_rate": 3.493049924223872e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 13232576,
      "step": 14800
    },
    {
      "epoch": 3.9061633892041705,
      "eval_loss": 0.07008786499500275,
      "eval_runtime": 64.8015,
      "eval_samples_per_second": 103.933,
      "eval_steps_per_second": 25.987,
      "num_input_tokens_seen": 13232576,
      "step": 14800
    },
    {
      "epoch": 3.9074831727596675,
      "grad_norm": 0.10320975631475449,
      "learning_rate": 3.49214887480519e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 13236992,
      "step": 14805
    },
    {
      "epoch": 3.9088029563151645,
      "grad_norm": 0.13766777515411377,
      "learning_rate": 3.4912476723846834e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 13241216,
      "step": 14810
    },
    {
      "epoch": 3.910122739870661,
      "grad_norm": 0.2176133692264557,
      "learning_rate": 3.490346317101328e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 13245280,
      "step": 14815
    },
    {
      "epoch": 3.911442523426158,
      "grad_norm": 0.14945746958255768,
      "learning_rate": 3.4894448090941266e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 13249632,
      "step": 14820
    },
    {
      "epoch": 3.912762306981655,
      "grad_norm": 0.11111060529947281,
      "learning_rate": 3.488543148502101e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 13253952,
      "step": 14825
    },
    {
      "epoch": 3.914082090537152,
      "grad_norm": 0.21643146872520447,
      "learning_rate": 3.487641335464299e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 13258464,
      "step": 14830
    },
    {
      "epoch": 3.915401874092649,
      "grad_norm": 0.2436448633670807,
      "learning_rate": 3.4867393701197914e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 13262880,
      "step": 14835
    },
    {
      "epoch": 3.9167216576481456,
      "grad_norm": 0.5105699300765991,
      "learning_rate": 3.485837252607673e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 13267520,
      "step": 14840
    },
    {
      "epoch": 3.9180414412036426,
      "grad_norm": 0.1857278198003769,
      "learning_rate": 3.4849349830670615e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 13272256,
      "step": 14845
    },
    {
      "epoch": 3.9193612247591396,
      "grad_norm": 0.022915441542863846,
      "learning_rate": 3.4840325616370976e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 13276320,
      "step": 14850
    },
    {
      "epoch": 3.920681008314636,
      "grad_norm": 0.032792024314403534,
      "learning_rate": 3.483129988456947e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 13280704,
      "step": 14855
    },
    {
      "epoch": 3.922000791870133,
      "grad_norm": 0.14063799381256104,
      "learning_rate": 3.482227263665797e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 13284992,
      "step": 14860
    },
    {
      "epoch": 3.92332057542563,
      "grad_norm": 0.12707214057445526,
      "learning_rate": 3.48132438740286e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 13289664,
      "step": 14865
    },
    {
      "epoch": 3.924640358981127,
      "grad_norm": 0.5708481073379517,
      "learning_rate": 3.48042135980737e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 13294304,
      "step": 14870
    },
    {
      "epoch": 3.925960142536624,
      "grad_norm": 0.05529450997710228,
      "learning_rate": 3.479518181018586e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 13298848,
      "step": 14875
    },
    {
      "epoch": 3.9272799260921207,
      "grad_norm": 0.266022264957428,
      "learning_rate": 3.4786148511757886e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 13303200,
      "step": 14880
    },
    {
      "epoch": 3.9285997096476177,
      "grad_norm": 0.19059279561042786,
      "learning_rate": 3.477711370418284e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 13307520,
      "step": 14885
    },
    {
      "epoch": 3.9299194932031147,
      "grad_norm": 0.17849835753440857,
      "learning_rate": 3.476807738885399e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 13312096,
      "step": 14890
    },
    {
      "epoch": 3.9312392767586117,
      "grad_norm": 0.22950884699821472,
      "learning_rate": 3.475903956716485e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 13316608,
      "step": 14895
    },
    {
      "epoch": 3.9325590603141087,
      "grad_norm": 0.5360502004623413,
      "learning_rate": 3.475000024050917e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 13320800,
      "step": 14900
    },
    {
      "epoch": 3.9338788438696053,
      "grad_norm": 0.2870543897151947,
      "learning_rate": 3.4740959410280926e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 13325152,
      "step": 14905
    },
    {
      "epoch": 3.9351986274251023,
      "grad_norm": 0.171759694814682,
      "learning_rate": 3.4731917077874324e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 13329696,
      "step": 14910
    },
    {
      "epoch": 3.9365184109805993,
      "grad_norm": 0.3141838014125824,
      "learning_rate": 3.4722873244683816e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 13334336,
      "step": 14915
    },
    {
      "epoch": 3.937838194536096,
      "grad_norm": 0.22930733859539032,
      "learning_rate": 3.4713827912104065e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 13338560,
      "step": 14920
    },
    {
      "epoch": 3.939157978091593,
      "grad_norm": 0.030559547245502472,
      "learning_rate": 3.470478108152998e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 13343456,
      "step": 14925
    },
    {
      "epoch": 3.94047776164709,
      "grad_norm": 0.0794212594628334,
      "learning_rate": 3.4695732754356695e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 13348064,
      "step": 14930
    },
    {
      "epoch": 3.941797545202587,
      "grad_norm": 0.24997788667678833,
      "learning_rate": 3.4686682931979576e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 13352352,
      "step": 14935
    },
    {
      "epoch": 3.943117328758084,
      "grad_norm": 0.04639951139688492,
      "learning_rate": 3.467763161579422e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 13356928,
      "step": 14940
    },
    {
      "epoch": 3.9444371123135804,
      "grad_norm": 0.113305002450943,
      "learning_rate": 3.466857880719645e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 13361184,
      "step": 14945
    },
    {
      "epoch": 3.9457568958690774,
      "grad_norm": 0.44650429487228394,
      "learning_rate": 3.465952450758233e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 13365696,
      "step": 14950
    },
    {
      "epoch": 3.9470766794245744,
      "grad_norm": 0.12692926824092865,
      "learning_rate": 3.4650468718348126e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 13370080,
      "step": 14955
    },
    {
      "epoch": 3.9483964629800714,
      "grad_norm": 0.15140819549560547,
      "learning_rate": 3.464141144089038e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 13374560,
      "step": 14960
    },
    {
      "epoch": 3.9497162465355684,
      "grad_norm": 0.03895285353064537,
      "learning_rate": 3.463235267660583e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 13378880,
      "step": 14965
    },
    {
      "epoch": 3.951036030091065,
      "grad_norm": 0.1654711216688156,
      "learning_rate": 3.462329242689145e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 13383232,
      "step": 14970
    },
    {
      "epoch": 3.952355813646562,
      "grad_norm": 0.19105668365955353,
      "learning_rate": 3.461423069314444e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 13387872,
      "step": 14975
    },
    {
      "epoch": 3.953675597202059,
      "grad_norm": 0.05818028748035431,
      "learning_rate": 3.460516747676224e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 13392320,
      "step": 14980
    },
    {
      "epoch": 3.9549953807575555,
      "grad_norm": 0.14191877841949463,
      "learning_rate": 3.459610277914251e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 13396448,
      "step": 14985
    },
    {
      "epoch": 3.9563151643130525,
      "grad_norm": 0.31594961881637573,
      "learning_rate": 3.458703660168314e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 13401056,
      "step": 14990
    },
    {
      "epoch": 3.9576349478685495,
      "grad_norm": 0.3360626995563507,
      "learning_rate": 3.457796894578224e-05,
      "loss": 0.1145,
      "num_input_tokens_seen": 13405632,
      "step": 14995
    },
    {
      "epoch": 3.9589547314240465,
      "grad_norm": 0.13125957548618317,
      "learning_rate": 3.456889981283817e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 13410176,
      "step": 15000
    },
    {
      "epoch": 3.9589547314240465,
      "eval_loss": 0.0699625238776207,
      "eval_runtime": 64.7956,
      "eval_samples_per_second": 103.942,
      "eval_steps_per_second": 25.989,
      "num_input_tokens_seen": 13410176,
      "step": 15000
    },
    {
      "epoch": 3.9602745149795435,
      "grad_norm": 0.278329074382782,
      "learning_rate": 3.45598292042495e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 13414656,
      "step": 15005
    },
    {
      "epoch": 3.96159429853504,
      "grad_norm": 0.21147602796554565,
      "learning_rate": 3.4550757121415035e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 13418976,
      "step": 15010
    },
    {
      "epoch": 3.962914082090537,
      "grad_norm": 0.2862236797809601,
      "learning_rate": 3.454168356573378e-05,
      "loss": 0.0874,
      "num_input_tokens_seen": 13423136,
      "step": 15015
    },
    {
      "epoch": 3.964233865646034,
      "grad_norm": 0.11077149957418442,
      "learning_rate": 3.453260853860503e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 13427552,
      "step": 15020
    },
    {
      "epoch": 3.965553649201531,
      "grad_norm": 0.04835366830229759,
      "learning_rate": 3.452353204142824e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 13432384,
      "step": 15025
    },
    {
      "epoch": 3.966873432757028,
      "grad_norm": 0.07445228844881058,
      "learning_rate": 3.4514454075603136e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 13436864,
      "step": 15030
    },
    {
      "epoch": 3.9681932163125246,
      "grad_norm": 0.22780664265155792,
      "learning_rate": 3.450537464252964e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 13441216,
      "step": 15035
    },
    {
      "epoch": 3.9695129998680216,
      "grad_norm": 0.2804659605026245,
      "learning_rate": 3.4496293743607925e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 13445792,
      "step": 15040
    },
    {
      "epoch": 3.9708327834235186,
      "grad_norm": 0.1564619541168213,
      "learning_rate": 3.448721138023838e-05,
      "loss": 0.0783,
      "num_input_tokens_seen": 13450016,
      "step": 15045
    },
    {
      "epoch": 3.972152566979015,
      "grad_norm": 0.44536006450653076,
      "learning_rate": 3.447812755382162e-05,
      "loss": 0.1,
      "num_input_tokens_seen": 13454176,
      "step": 15050
    },
    {
      "epoch": 3.973472350534512,
      "grad_norm": 0.3413216769695282,
      "learning_rate": 3.446904226575847e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 13458880,
      "step": 15055
    },
    {
      "epoch": 3.974792134090009,
      "grad_norm": 0.33574607968330383,
      "learning_rate": 3.445995551745002e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 13463232,
      "step": 15060
    },
    {
      "epoch": 3.976111917645506,
      "grad_norm": 0.07973798364400864,
      "learning_rate": 3.445086731029753e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 13467904,
      "step": 15065
    },
    {
      "epoch": 3.977431701201003,
      "grad_norm": 0.21384595334529877,
      "learning_rate": 3.444177764570255e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 13472512,
      "step": 15070
    },
    {
      "epoch": 3.9787514847564998,
      "grad_norm": 0.37786194682121277,
      "learning_rate": 3.44326865250668e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 13476960,
      "step": 15075
    },
    {
      "epoch": 3.9800712683119968,
      "grad_norm": 0.3112943768501282,
      "learning_rate": 3.442359394979225e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 13481312,
      "step": 15080
    },
    {
      "epoch": 3.9813910518674938,
      "grad_norm": 0.1531369686126709,
      "learning_rate": 3.441449992128108e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 13485984,
      "step": 15085
    },
    {
      "epoch": 3.9827108354229908,
      "grad_norm": 0.12034045159816742,
      "learning_rate": 3.440540444093573e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 13490496,
      "step": 15090
    },
    {
      "epoch": 3.9840306189784878,
      "grad_norm": 0.2524610161781311,
      "learning_rate": 3.43963075101588e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 13494720,
      "step": 15095
    },
    {
      "epoch": 3.9853504025339843,
      "grad_norm": 0.2634822726249695,
      "learning_rate": 3.438720913035318e-05,
      "loss": 0.1152,
      "num_input_tokens_seen": 13499136,
      "step": 15100
    },
    {
      "epoch": 3.9866701860894813,
      "grad_norm": 0.12332644313573837,
      "learning_rate": 3.437810930292195e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 13503584,
      "step": 15105
    },
    {
      "epoch": 3.9879899696449783,
      "grad_norm": 0.06713055074214935,
      "learning_rate": 3.43690080292684e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 13508160,
      "step": 15110
    },
    {
      "epoch": 3.989309753200475,
      "grad_norm": 0.03950841724872589,
      "learning_rate": 3.435990531079608e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 13512960,
      "step": 15115
    },
    {
      "epoch": 3.990629536755972,
      "grad_norm": 0.21860715746879578,
      "learning_rate": 3.435080114890874e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 13516992,
      "step": 15120
    },
    {
      "epoch": 3.991949320311469,
      "grad_norm": 0.11835494637489319,
      "learning_rate": 3.434169554501035e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 13521472,
      "step": 15125
    },
    {
      "epoch": 3.993269103866966,
      "grad_norm": 0.11930891126394272,
      "learning_rate": 3.433258850050511e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 13525920,
      "step": 15130
    },
    {
      "epoch": 3.994588887422463,
      "grad_norm": 0.16619794070720673,
      "learning_rate": 3.4323480016797446e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 13530976,
      "step": 15135
    },
    {
      "epoch": 3.9959086709779594,
      "grad_norm": 0.0904647633433342,
      "learning_rate": 3.4314370095291995e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 13535424,
      "step": 15140
    },
    {
      "epoch": 3.9972284545334564,
      "grad_norm": 0.35754939913749695,
      "learning_rate": 3.430525873739363e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 13539808,
      "step": 15145
    },
    {
      "epoch": 3.9985482380889534,
      "grad_norm": 0.1245490312576294,
      "learning_rate": 3.429614594450743e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 13544416,
      "step": 15150
    },
    {
      "epoch": 3.9998680216444504,
      "grad_norm": 0.2929128408432007,
      "learning_rate": 3.428703171803869e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 13548832,
      "step": 15155
    },
    {
      "epoch": 4.001055826844397,
      "grad_norm": 0.24679122865200043,
      "learning_rate": 3.4277916059392964e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 13552688,
      "step": 15160
    },
    {
      "epoch": 4.002375610399894,
      "grad_norm": 0.44888702034950256,
      "learning_rate": 3.426879896997598e-05,
      "loss": 0.0918,
      "num_input_tokens_seen": 13557168,
      "step": 15165
    },
    {
      "epoch": 4.003695393955391,
      "grad_norm": 0.18724696338176727,
      "learning_rate": 3.425968045119372e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 13561424,
      "step": 15170
    },
    {
      "epoch": 4.005015177510888,
      "grad_norm": 0.18873801827430725,
      "learning_rate": 3.425056050445237e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 13565968,
      "step": 15175
    },
    {
      "epoch": 4.006334961066385,
      "grad_norm": 0.10356069356203079,
      "learning_rate": 3.4241439131158336e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 13570352,
      "step": 15180
    },
    {
      "epoch": 4.007654744621882,
      "grad_norm": 0.2512379288673401,
      "learning_rate": 3.423231633271825e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 13574736,
      "step": 15185
    },
    {
      "epoch": 4.0089745281773785,
      "grad_norm": 0.14024873077869415,
      "learning_rate": 3.4223192110538985e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 13579376,
      "step": 15190
    },
    {
      "epoch": 4.010294311732876,
      "grad_norm": 0.0481589175760746,
      "learning_rate": 3.4214066466027575e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 13583856,
      "step": 15195
    },
    {
      "epoch": 4.0116140952883725,
      "grad_norm": 0.26079243421554565,
      "learning_rate": 3.4204939400591325e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 13588176,
      "step": 15200
    },
    {
      "epoch": 4.0116140952883725,
      "eval_loss": 0.06972558051347733,
      "eval_runtime": 64.7979,
      "eval_samples_per_second": 103.939,
      "eval_steps_per_second": 25.989,
      "num_input_tokens_seen": 13588176,
      "step": 15200
    },
    {
      "epoch": 4.01293387884387,
      "grad_norm": 0.370058536529541,
      "learning_rate": 3.419581091563775e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 13592560,
      "step": 15205
    },
    {
      "epoch": 4.0142536623993665,
      "grad_norm": 0.11809702962636948,
      "learning_rate": 3.418668101257456e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 13596816,
      "step": 15210
    },
    {
      "epoch": 4.015573445954863,
      "grad_norm": 0.09288965910673141,
      "learning_rate": 3.417754969280971e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 13601424,
      "step": 15215
    },
    {
      "epoch": 4.0168932295103605,
      "grad_norm": 0.32196182012557983,
      "learning_rate": 3.416841695775137e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 13606000,
      "step": 15220
    },
    {
      "epoch": 4.018213013065857,
      "grad_norm": 0.2525179386138916,
      "learning_rate": 3.415928280880792e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 13610640,
      "step": 15225
    },
    {
      "epoch": 4.019532796621354,
      "grad_norm": 0.18314425647258759,
      "learning_rate": 3.4150147247387965e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 13615440,
      "step": 15230
    },
    {
      "epoch": 4.020852580176851,
      "grad_norm": 0.48517295718193054,
      "learning_rate": 3.4141010274900306e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 13619920,
      "step": 15235
    },
    {
      "epoch": 4.022172363732348,
      "grad_norm": 0.444707989692688,
      "learning_rate": 3.413187189275399e-05,
      "loss": 0.0837,
      "num_input_tokens_seen": 13624400,
      "step": 15240
    },
    {
      "epoch": 4.023492147287845,
      "grad_norm": 0.3515114188194275,
      "learning_rate": 3.4122732102358265e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 13629072,
      "step": 15245
    },
    {
      "epoch": 4.024811930843342,
      "grad_norm": 0.39779072999954224,
      "learning_rate": 3.411359090512261e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 13633616,
      "step": 15250
    },
    {
      "epoch": 4.026131714398838,
      "grad_norm": 0.3236246407032013,
      "learning_rate": 3.410444830245672e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 13638224,
      "step": 15255
    },
    {
      "epoch": 4.027451497954336,
      "grad_norm": 0.020479105412960052,
      "learning_rate": 3.409530429577048e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 13642768,
      "step": 15260
    },
    {
      "epoch": 4.028771281509832,
      "grad_norm": 0.25240564346313477,
      "learning_rate": 3.408615888647402e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 13647056,
      "step": 15265
    },
    {
      "epoch": 4.03009106506533,
      "grad_norm": 0.0888763815164566,
      "learning_rate": 3.4077012075977675e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 13651408,
      "step": 15270
    },
    {
      "epoch": 4.031410848620826,
      "grad_norm": 0.19299018383026123,
      "learning_rate": 3.4067863865692e-05,
      "loss": 0.1245,
      "num_input_tokens_seen": 13655824,
      "step": 15275
    },
    {
      "epoch": 4.032730632176323,
      "grad_norm": 0.06421855837106705,
      "learning_rate": 3.4058714257027755e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 13660304,
      "step": 15280
    },
    {
      "epoch": 4.03405041573182,
      "grad_norm": 0.115789033472538,
      "learning_rate": 3.404956325139594e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 13664624,
      "step": 15285
    },
    {
      "epoch": 4.035370199287317,
      "grad_norm": 0.3456934690475464,
      "learning_rate": 3.404041085020775e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 13669104,
      "step": 15290
    },
    {
      "epoch": 4.036689982842814,
      "grad_norm": 0.3957151472568512,
      "learning_rate": 3.403125705487459e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 13673520,
      "step": 15295
    },
    {
      "epoch": 4.038009766398311,
      "grad_norm": 0.1952756643295288,
      "learning_rate": 3.402210186680811e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 13677744,
      "step": 15300
    },
    {
      "epoch": 4.039329549953807,
      "grad_norm": 0.1507146805524826,
      "learning_rate": 3.4012945287420137e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 13681872,
      "step": 15305
    },
    {
      "epoch": 4.040649333509305,
      "grad_norm": 0.09474373608827591,
      "learning_rate": 3.400378731812274e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 13686384,
      "step": 15310
    },
    {
      "epoch": 4.041969117064801,
      "grad_norm": 0.18544119596481323,
      "learning_rate": 3.399462796032817e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 13690704,
      "step": 15315
    },
    {
      "epoch": 4.043288900620298,
      "grad_norm": 0.2785351276397705,
      "learning_rate": 3.3985467215448954e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 13694992,
      "step": 15320
    },
    {
      "epoch": 4.044608684175795,
      "grad_norm": 0.4678272604942322,
      "learning_rate": 3.3976305084897776e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 13699056,
      "step": 15325
    },
    {
      "epoch": 4.045928467731292,
      "grad_norm": 0.2782832384109497,
      "learning_rate": 3.3967141570087544e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 13703824,
      "step": 15330
    },
    {
      "epoch": 4.047248251286789,
      "grad_norm": 0.27288955450057983,
      "learning_rate": 3.39579766724314e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 13708304,
      "step": 15335
    },
    {
      "epoch": 4.048568034842286,
      "grad_norm": 0.22154158353805542,
      "learning_rate": 3.3948810393342677e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 13712720,
      "step": 15340
    },
    {
      "epoch": 4.0498878183977824,
      "grad_norm": 0.16392801702022552,
      "learning_rate": 3.3939642734234936e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 13717040,
      "step": 15345
    },
    {
      "epoch": 4.05120760195328,
      "grad_norm": 0.2467198520898819,
      "learning_rate": 3.393047369652194e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 13721584,
      "step": 15350
    },
    {
      "epoch": 4.0525273855087764,
      "grad_norm": 0.07231663167476654,
      "learning_rate": 3.3921303281617664e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 13726128,
      "step": 15355
    },
    {
      "epoch": 4.053847169064274,
      "grad_norm": 0.07332615554332733,
      "learning_rate": 3.391213149093632e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 13730640,
      "step": 15360
    },
    {
      "epoch": 4.0551669526197704,
      "grad_norm": 0.16019225120544434,
      "learning_rate": 3.3902958325892303e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 13734896,
      "step": 15365
    },
    {
      "epoch": 4.056486736175267,
      "grad_norm": 0.2087184190750122,
      "learning_rate": 3.389378378790023e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 13739504,
      "step": 15370
    },
    {
      "epoch": 4.0578065197307644,
      "grad_norm": 0.35674214363098145,
      "learning_rate": 3.388460787837493e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 13744112,
      "step": 15375
    },
    {
      "epoch": 4.059126303286261,
      "grad_norm": 0.23047539591789246,
      "learning_rate": 3.387543059873145e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 13748528,
      "step": 15380
    },
    {
      "epoch": 4.060446086841758,
      "grad_norm": 0.17903640866279602,
      "learning_rate": 3.386625195038503e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 13752944,
      "step": 15385
    },
    {
      "epoch": 4.061765870397255,
      "grad_norm": 0.17652344703674316,
      "learning_rate": 3.3857071934751136e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 13757392,
      "step": 15390
    },
    {
      "epoch": 4.063085653952752,
      "grad_norm": 0.1955309361219406,
      "learning_rate": 3.384789055324544e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 13761904,
      "step": 15395
    },
    {
      "epoch": 4.064405437508249,
      "grad_norm": 0.13614407181739807,
      "learning_rate": 3.3838707807283843e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 13766160,
      "step": 15400
    },
    {
      "epoch": 4.064405437508249,
      "eval_loss": 0.06941487640142441,
      "eval_runtime": 64.7667,
      "eval_samples_per_second": 103.989,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 13766160,
      "step": 15400
    },
    {
      "epoch": 4.065725221063746,
      "grad_norm": 0.052259884774684906,
      "learning_rate": 3.382952369828243e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 13770672,
      "step": 15405
    },
    {
      "epoch": 4.067045004619242,
      "grad_norm": 0.34939345717430115,
      "learning_rate": 3.38203382276575e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 13775088,
      "step": 15410
    },
    {
      "epoch": 4.06836478817474,
      "grad_norm": 0.09233438968658447,
      "learning_rate": 3.381115139682557e-05,
      "loss": 0.0764,
      "num_input_tokens_seen": 13779376,
      "step": 15415
    },
    {
      "epoch": 4.069684571730236,
      "grad_norm": 0.046208497136831284,
      "learning_rate": 3.3801963207203366e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 13783984,
      "step": 15420
    },
    {
      "epoch": 4.071004355285734,
      "grad_norm": 0.23162056505680084,
      "learning_rate": 3.379277366020782e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 13788432,
      "step": 15425
    },
    {
      "epoch": 4.07232413884123,
      "grad_norm": 0.31733816862106323,
      "learning_rate": 3.3783582757256085e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 13793104,
      "step": 15430
    },
    {
      "epoch": 4.073643922396727,
      "grad_norm": 0.3538193702697754,
      "learning_rate": 3.3774390499765504e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 13797200,
      "step": 15435
    },
    {
      "epoch": 4.074963705952224,
      "grad_norm": 0.19041797518730164,
      "learning_rate": 3.376519688915364e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 13801680,
      "step": 15440
    },
    {
      "epoch": 4.076283489507721,
      "grad_norm": 0.23964625597000122,
      "learning_rate": 3.3756001926838273e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 13806544,
      "step": 15445
    },
    {
      "epoch": 4.077603273063217,
      "grad_norm": 0.19917507469654083,
      "learning_rate": 3.374680561423737e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 13810896,
      "step": 15450
    },
    {
      "epoch": 4.078923056618715,
      "grad_norm": 0.10634046047925949,
      "learning_rate": 3.373760795276912e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 13815536,
      "step": 15455
    },
    {
      "epoch": 4.080242840174211,
      "grad_norm": 0.1096583753824234,
      "learning_rate": 3.372840894385192e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 13820336,
      "step": 15460
    },
    {
      "epoch": 4.081562623729709,
      "grad_norm": 0.09835968166589737,
      "learning_rate": 3.3719208588904375e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 13825104,
      "step": 15465
    },
    {
      "epoch": 4.082882407285205,
      "grad_norm": 0.5322155952453613,
      "learning_rate": 3.371000688934529e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 13829200,
      "step": 15470
    },
    {
      "epoch": 4.084202190840702,
      "grad_norm": 0.2679766118526459,
      "learning_rate": 3.370080384659369e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 13833584,
      "step": 15475
    },
    {
      "epoch": 4.085521974396199,
      "grad_norm": 0.1547887772321701,
      "learning_rate": 3.36915994620688e-05,
      "loss": 0.0248,
      "num_input_tokens_seen": 13837936,
      "step": 15480
    },
    {
      "epoch": 4.086841757951696,
      "grad_norm": 0.088457390666008,
      "learning_rate": 3.3682393737190035e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 13842320,
      "step": 15485
    },
    {
      "epoch": 4.088161541507193,
      "grad_norm": 0.21799533069133759,
      "learning_rate": 3.3673186673377054e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 13847120,
      "step": 15490
    },
    {
      "epoch": 4.08948132506269,
      "grad_norm": 0.18188966810703278,
      "learning_rate": 3.366397827204969e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 13851696,
      "step": 15495
    },
    {
      "epoch": 4.090801108618186,
      "grad_norm": 0.08275601267814636,
      "learning_rate": 3.3654768534628e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 13856240,
      "step": 15500
    },
    {
      "epoch": 4.092120892173684,
      "grad_norm": 0.27553048729896545,
      "learning_rate": 3.3645557462532245e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 13860880,
      "step": 15505
    },
    {
      "epoch": 4.09344067572918,
      "grad_norm": 0.31351494789123535,
      "learning_rate": 3.363634505718288e-05,
      "loss": 0.0955,
      "num_input_tokens_seen": 13865264,
      "step": 15510
    },
    {
      "epoch": 4.094760459284677,
      "grad_norm": 0.3622507154941559,
      "learning_rate": 3.362713132000057e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 13869552,
      "step": 15515
    },
    {
      "epoch": 4.096080242840174,
      "grad_norm": 0.2192094773054123,
      "learning_rate": 3.36179162524062e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 13873904,
      "step": 15520
    },
    {
      "epoch": 4.097400026395671,
      "grad_norm": 0.2093241959810257,
      "learning_rate": 3.3608699855820846e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 13878352,
      "step": 15525
    },
    {
      "epoch": 4.098719809951168,
      "grad_norm": 0.1291126012802124,
      "learning_rate": 3.359948213166578e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 13882608,
      "step": 15530
    },
    {
      "epoch": 4.100039593506665,
      "grad_norm": 0.38856059312820435,
      "learning_rate": 3.359026308136252e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 13887088,
      "step": 15535
    },
    {
      "epoch": 4.1013593770621615,
      "grad_norm": 0.20325766503810883,
      "learning_rate": 3.358104270633272e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 13891600,
      "step": 15540
    },
    {
      "epoch": 4.102679160617659,
      "grad_norm": 0.32808127999305725,
      "learning_rate": 3.357182100799831e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 13896176,
      "step": 15545
    },
    {
      "epoch": 4.1039989441731555,
      "grad_norm": 0.3997575044631958,
      "learning_rate": 3.3562597987781384e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 13900848,
      "step": 15550
    },
    {
      "epoch": 4.105318727728653,
      "grad_norm": 0.38658228516578674,
      "learning_rate": 3.355337364710424e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 13905616,
      "step": 15555
    },
    {
      "epoch": 4.1066385112841495,
      "grad_norm": 0.1041800007224083,
      "learning_rate": 3.354414798738939e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 13909680,
      "step": 15560
    },
    {
      "epoch": 4.107958294839646,
      "grad_norm": 0.05422928184270859,
      "learning_rate": 3.353492101005955e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 13914352,
      "step": 15565
    },
    {
      "epoch": 4.1092780783951435,
      "grad_norm": 0.07094903290271759,
      "learning_rate": 3.352569271653763e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 13918896,
      "step": 15570
    },
    {
      "epoch": 4.11059786195064,
      "grad_norm": 0.44785088300704956,
      "learning_rate": 3.351646310824675e-05,
      "loss": 0.0764,
      "num_input_tokens_seen": 13923376,
      "step": 15575
    },
    {
      "epoch": 4.111917645506137,
      "grad_norm": 0.19541163742542267,
      "learning_rate": 3.350723218661023e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 13927696,
      "step": 15580
    },
    {
      "epoch": 4.113237429061634,
      "grad_norm": 0.1924787312746048,
      "learning_rate": 3.349799995305162e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 13932304,
      "step": 15585
    },
    {
      "epoch": 4.114557212617131,
      "grad_norm": 0.04075240343809128,
      "learning_rate": 3.348876640899461e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 13936944,
      "step": 15590
    },
    {
      "epoch": 4.115876996172628,
      "grad_norm": 0.08251579850912094,
      "learning_rate": 3.3479531555863144e-05,
      "loss": 0.0244,
      "num_input_tokens_seen": 13941424,
      "step": 15595
    },
    {
      "epoch": 4.117196779728125,
      "grad_norm": 0.12373322248458862,
      "learning_rate": 3.3470295395081344e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 13945776,
      "step": 15600
    },
    {
      "epoch": 4.117196779728125,
      "eval_loss": 0.06979098916053772,
      "eval_runtime": 64.7782,
      "eval_samples_per_second": 103.97,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 13945776,
      "step": 15600
    },
    {
      "epoch": 4.118516563283621,
      "grad_norm": 0.17074279487133026,
      "learning_rate": 3.3461057928073556e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 13950128,
      "step": 15605
    },
    {
      "epoch": 4.119836346839119,
      "grad_norm": 0.3415358066558838,
      "learning_rate": 3.345181915626431e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 13954416,
      "step": 15610
    },
    {
      "epoch": 4.121156130394615,
      "grad_norm": 0.15275879204273224,
      "learning_rate": 3.344257908107834e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 13958544,
      "step": 15615
    },
    {
      "epoch": 4.122475913950113,
      "grad_norm": 0.23040013015270233,
      "learning_rate": 3.343333770394058e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 13962928,
      "step": 15620
    },
    {
      "epoch": 4.123795697505609,
      "grad_norm": 0.04406331852078438,
      "learning_rate": 3.342409502627616e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 13967152,
      "step": 15625
    },
    {
      "epoch": 4.125115481061106,
      "grad_norm": 0.3055570423603058,
      "learning_rate": 3.341485104951043e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 13971408,
      "step": 15630
    },
    {
      "epoch": 4.126435264616603,
      "grad_norm": 0.4481905698776245,
      "learning_rate": 3.340560577506892e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 13975856,
      "step": 15635
    },
    {
      "epoch": 4.1277550481721,
      "grad_norm": 0.27619317173957825,
      "learning_rate": 3.339635920437735e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 13980560,
      "step": 15640
    },
    {
      "epoch": 4.129074831727596,
      "grad_norm": 0.24358342587947845,
      "learning_rate": 3.338711133886169e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 13985040,
      "step": 15645
    },
    {
      "epoch": 4.130394615283094,
      "grad_norm": 0.14848750829696655,
      "learning_rate": 3.3377862179948064e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 13989712,
      "step": 15650
    },
    {
      "epoch": 4.13171439883859,
      "grad_norm": 0.12414304912090302,
      "learning_rate": 3.336861172906281e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 13994160,
      "step": 15655
    },
    {
      "epoch": 4.133034182394088,
      "grad_norm": 0.2855603098869324,
      "learning_rate": 3.335935998763245e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 13998736,
      "step": 15660
    },
    {
      "epoch": 4.134353965949584,
      "grad_norm": 0.14758923649787903,
      "learning_rate": 3.3350106957083744e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 14003312,
      "step": 15665
    },
    {
      "epoch": 4.135673749505081,
      "grad_norm": 0.32446083426475525,
      "learning_rate": 3.33408526388436e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 14007760,
      "step": 15670
    },
    {
      "epoch": 4.136993533060578,
      "grad_norm": 0.5859151482582092,
      "learning_rate": 3.3331597034339166e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 14012560,
      "step": 15675
    },
    {
      "epoch": 4.138313316616075,
      "grad_norm": 0.22588935494422913,
      "learning_rate": 3.3322340144997764e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 14016816,
      "step": 15680
    },
    {
      "epoch": 4.139633100171572,
      "grad_norm": 0.274039626121521,
      "learning_rate": 3.331308197224693e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 14020880,
      "step": 15685
    },
    {
      "epoch": 4.140952883727069,
      "grad_norm": 0.05712407827377319,
      "learning_rate": 3.330382251751438e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 14025584,
      "step": 15690
    },
    {
      "epoch": 4.142272667282565,
      "grad_norm": 0.20152127742767334,
      "learning_rate": 3.3294561782228054e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 14029904,
      "step": 15695
    },
    {
      "epoch": 4.143592450838063,
      "grad_norm": 0.16317099332809448,
      "learning_rate": 3.328529976781607e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 14034544,
      "step": 15700
    },
    {
      "epoch": 4.144912234393559,
      "grad_norm": 0.30084744095802307,
      "learning_rate": 3.327603647570673e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 14039152,
      "step": 15705
    },
    {
      "epoch": 4.146232017949056,
      "grad_norm": 0.2822813093662262,
      "learning_rate": 3.326677190732857e-05,
      "loss": 0.0948,
      "num_input_tokens_seen": 14043504,
      "step": 15710
    },
    {
      "epoch": 4.147551801504553,
      "grad_norm": 0.058278538286685944,
      "learning_rate": 3.325750606411029e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 14048016,
      "step": 15715
    },
    {
      "epoch": 4.14887158506005,
      "grad_norm": 0.19492752850055695,
      "learning_rate": 3.3248238947480804e-05,
      "loss": 0.0819,
      "num_input_tokens_seen": 14052400,
      "step": 15720
    },
    {
      "epoch": 4.150191368615547,
      "grad_norm": 0.14393779635429382,
      "learning_rate": 3.323897055886922e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 14057104,
      "step": 15725
    },
    {
      "epoch": 4.151511152171044,
      "grad_norm": 0.4962517023086548,
      "learning_rate": 3.322970089970484e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 14061648,
      "step": 15730
    },
    {
      "epoch": 4.1528309357265405,
      "grad_norm": 0.19396886229515076,
      "learning_rate": 3.3220429971417165e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 14066192,
      "step": 15735
    },
    {
      "epoch": 4.154150719282038,
      "grad_norm": 0.5927832722663879,
      "learning_rate": 3.321115777543588e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 14070704,
      "step": 15740
    },
    {
      "epoch": 4.1554705028375345,
      "grad_norm": 0.11209552735090256,
      "learning_rate": 3.320188431319088e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 14075152,
      "step": 15745
    },
    {
      "epoch": 4.156790286393032,
      "grad_norm": 0.24031676352024078,
      "learning_rate": 3.319260958611224e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 14079376,
      "step": 15750
    },
    {
      "epoch": 4.1581100699485285,
      "grad_norm": 0.36269837617874146,
      "learning_rate": 3.3183333595630256e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 14083728,
      "step": 15755
    },
    {
      "epoch": 4.159429853504025,
      "grad_norm": 0.23398347198963165,
      "learning_rate": 3.317405634317538e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 14088400,
      "step": 15760
    },
    {
      "epoch": 4.1607496370595225,
      "grad_norm": 0.17716103792190552,
      "learning_rate": 3.3164777830178315e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 14092688,
      "step": 15765
    },
    {
      "epoch": 4.162069420615019,
      "grad_norm": 0.10484951734542847,
      "learning_rate": 3.315549805806989e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 14097296,
      "step": 15770
    },
    {
      "epoch": 4.163389204170516,
      "grad_norm": 0.10782454162836075,
      "learning_rate": 3.314621702828118e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 14101744,
      "step": 15775
    },
    {
      "epoch": 4.164708987726013,
      "grad_norm": 0.0995270311832428,
      "learning_rate": 3.313693474224342e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 14105872,
      "step": 15780
    },
    {
      "epoch": 4.16602877128151,
      "grad_norm": 0.29905152320861816,
      "learning_rate": 3.312765120138809e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 14110000,
      "step": 15785
    },
    {
      "epoch": 4.167348554837007,
      "grad_norm": 0.3319566249847412,
      "learning_rate": 3.311836640714679e-05,
      "loss": 0.0768,
      "num_input_tokens_seen": 14114192,
      "step": 15790
    },
    {
      "epoch": 4.168668338392504,
      "grad_norm": 0.049213580787181854,
      "learning_rate": 3.310908036095137e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 14118640,
      "step": 15795
    },
    {
      "epoch": 4.169988121948,
      "grad_norm": 0.2921159267425537,
      "learning_rate": 3.309979306423386e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 14123120,
      "step": 15800
    },
    {
      "epoch": 4.169988121948,
      "eval_loss": 0.06947094947099686,
      "eval_runtime": 64.7206,
      "eval_samples_per_second": 104.063,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 14123120,
      "step": 15800
    },
    {
      "epoch": 4.171307905503498,
      "grad_norm": 0.5583330988883972,
      "learning_rate": 3.309050451842647e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 14127664,
      "step": 15805
    },
    {
      "epoch": 4.172627689058994,
      "grad_norm": 0.09117178618907928,
      "learning_rate": 3.3081214724961604e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 14132016,
      "step": 15810
    },
    {
      "epoch": 4.173947472614492,
      "grad_norm": 0.36706581711769104,
      "learning_rate": 3.307192368527188e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 14136336,
      "step": 15815
    },
    {
      "epoch": 4.175267256169988,
      "grad_norm": 0.369368314743042,
      "learning_rate": 3.306263140079008e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 14140784,
      "step": 15820
    },
    {
      "epoch": 4.176587039725485,
      "grad_norm": 0.22923415899276733,
      "learning_rate": 3.30533378729492e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 14145360,
      "step": 15825
    },
    {
      "epoch": 4.177906823280982,
      "grad_norm": 0.2262612134218216,
      "learning_rate": 3.304404310318242e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 14149904,
      "step": 15830
    },
    {
      "epoch": 4.179226606836479,
      "grad_norm": 0.07125329971313477,
      "learning_rate": 3.3034747092923105e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 14154320,
      "step": 15835
    },
    {
      "epoch": 4.180546390391975,
      "grad_norm": 0.2851259410381317,
      "learning_rate": 3.3025449843604806e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 14158832,
      "step": 15840
    },
    {
      "epoch": 4.181866173947473,
      "grad_norm": 0.16051997244358063,
      "learning_rate": 3.30161513566613e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 14163536,
      "step": 15845
    },
    {
      "epoch": 4.183185957502969,
      "grad_norm": 0.12674543261528015,
      "learning_rate": 3.3006851633526506e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 14167312,
      "step": 15850
    },
    {
      "epoch": 4.184505741058467,
      "grad_norm": 0.31514954566955566,
      "learning_rate": 3.2997550675634584e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 14171824,
      "step": 15855
    },
    {
      "epoch": 4.185825524613963,
      "grad_norm": 0.28195643424987793,
      "learning_rate": 3.2988248484419825e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 14176368,
      "step": 15860
    },
    {
      "epoch": 4.18714530816946,
      "grad_norm": 0.27311739325523376,
      "learning_rate": 3.2978945061316776e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 14180848,
      "step": 15865
    },
    {
      "epoch": 4.188465091724957,
      "grad_norm": 0.16340073943138123,
      "learning_rate": 3.296964040776013e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 14185360,
      "step": 15870
    },
    {
      "epoch": 4.189784875280454,
      "grad_norm": 0.24772635102272034,
      "learning_rate": 3.296033452518478e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 14189808,
      "step": 15875
    },
    {
      "epoch": 4.191104658835951,
      "grad_norm": 0.16504834592342377,
      "learning_rate": 3.2951027415025806e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 14194128,
      "step": 15880
    },
    {
      "epoch": 4.192424442391448,
      "grad_norm": 0.2937238812446594,
      "learning_rate": 3.294171907871849e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 14198640,
      "step": 15885
    },
    {
      "epoch": 4.193744225946944,
      "grad_norm": 0.04853179305791855,
      "learning_rate": 3.293240951769828e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 14202736,
      "step": 15890
    },
    {
      "epoch": 4.195064009502442,
      "grad_norm": 0.325665682554245,
      "learning_rate": 3.2923098733400846e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 14207376,
      "step": 15895
    },
    {
      "epoch": 4.196383793057938,
      "grad_norm": 0.14807242155075073,
      "learning_rate": 3.291378672726202e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 14211856,
      "step": 15900
    },
    {
      "epoch": 4.197703576613435,
      "grad_norm": 0.13883376121520996,
      "learning_rate": 3.2904473500717824e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 14216368,
      "step": 15905
    },
    {
      "epoch": 4.199023360168932,
      "grad_norm": 0.08326022326946259,
      "learning_rate": 3.289515905520449e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 14220848,
      "step": 15910
    },
    {
      "epoch": 4.200343143724429,
      "grad_norm": 0.0963999554514885,
      "learning_rate": 3.288584339215841e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 14225456,
      "step": 15915
    },
    {
      "epoch": 4.201662927279926,
      "grad_norm": 0.3117450177669525,
      "learning_rate": 3.287652651301617e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 14230384,
      "step": 15920
    },
    {
      "epoch": 4.202982710835423,
      "grad_norm": 0.059780437499284744,
      "learning_rate": 3.286720841921457e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 14234704,
      "step": 15925
    },
    {
      "epoch": 4.2043024943909195,
      "grad_norm": 0.08157684653997421,
      "learning_rate": 3.285788911219056e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 14238928,
      "step": 15930
    },
    {
      "epoch": 4.205622277946417,
      "grad_norm": 0.19843405485153198,
      "learning_rate": 3.284856859338131e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 14243216,
      "step": 15935
    },
    {
      "epoch": 4.2069420615019135,
      "grad_norm": 0.3667726516723633,
      "learning_rate": 3.283924686422414e-05,
      "loss": 0.1147,
      "num_input_tokens_seen": 14247536,
      "step": 15940
    },
    {
      "epoch": 4.208261845057411,
      "grad_norm": 0.5752120018005371,
      "learning_rate": 3.282992392615659e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 14252432,
      "step": 15945
    },
    {
      "epoch": 4.2095816286129075,
      "grad_norm": 0.15009431540966034,
      "learning_rate": 3.282059978061638e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 14257040,
      "step": 15950
    },
    {
      "epoch": 4.210901412168404,
      "grad_norm": 0.38392356038093567,
      "learning_rate": 3.28112744290414e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 14260816,
      "step": 15955
    },
    {
      "epoch": 4.2122211957239015,
      "grad_norm": 0.22872377932071686,
      "learning_rate": 3.280194787286974e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 14265392,
      "step": 15960
    },
    {
      "epoch": 4.213540979279398,
      "grad_norm": 0.10152777284383774,
      "learning_rate": 3.2792620113539674e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 14269872,
      "step": 15965
    },
    {
      "epoch": 4.2148607628348955,
      "grad_norm": 0.530021607875824,
      "learning_rate": 3.278329115248966e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 14274416,
      "step": 15970
    },
    {
      "epoch": 4.216180546390392,
      "grad_norm": 0.11004350334405899,
      "learning_rate": 3.277396099115834e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 14278768,
      "step": 15975
    },
    {
      "epoch": 4.217500329945889,
      "grad_norm": 0.09730465710163116,
      "learning_rate": 3.276462963098454e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 14283280,
      "step": 15980
    },
    {
      "epoch": 4.218820113501386,
      "grad_norm": 0.40982648730278015,
      "learning_rate": 3.275529707340728e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 14287728,
      "step": 15985
    },
    {
      "epoch": 4.220139897056883,
      "grad_norm": 0.20734404027462006,
      "learning_rate": 3.274596331986574e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 14292080,
      "step": 15990
    },
    {
      "epoch": 4.221459680612379,
      "grad_norm": 0.11304210871458054,
      "learning_rate": 3.273662837179932e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 14296272,
      "step": 15995
    },
    {
      "epoch": 4.222779464167877,
      "grad_norm": 0.14152880012989044,
      "learning_rate": 3.272729223064758e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 14300816,
      "step": 16000
    },
    {
      "epoch": 4.222779464167877,
      "eval_loss": 0.06940384209156036,
      "eval_runtime": 64.7909,
      "eval_samples_per_second": 103.95,
      "eval_steps_per_second": 25.991,
      "num_input_tokens_seen": 14300816,
      "step": 16000
    },
    {
      "epoch": 4.224099247723373,
      "grad_norm": 0.21146002411842346,
      "learning_rate": 3.2717954897850264e-05,
      "loss": 0.0211,
      "num_input_tokens_seen": 14305296,
      "step": 16005
    },
    {
      "epoch": 4.225419031278871,
      "grad_norm": 0.085642009973526,
      "learning_rate": 3.270861637484733e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 14309616,
      "step": 16010
    },
    {
      "epoch": 4.226738814834367,
      "grad_norm": 0.09676684439182281,
      "learning_rate": 3.2699276663078867e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 14314128,
      "step": 16015
    },
    {
      "epoch": 4.228058598389864,
      "grad_norm": 0.2536756694316864,
      "learning_rate": 3.268993576398519e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 14318256,
      "step": 16020
    },
    {
      "epoch": 4.229378381945361,
      "grad_norm": 0.09631747752428055,
      "learning_rate": 3.268059367900678e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 14322480,
      "step": 16025
    },
    {
      "epoch": 4.230698165500858,
      "grad_norm": 0.14271752536296844,
      "learning_rate": 3.26712504095843e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 14327024,
      "step": 16030
    },
    {
      "epoch": 4.232017949056354,
      "grad_norm": 0.11234790831804276,
      "learning_rate": 3.2661905957158615e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 14331504,
      "step": 16035
    },
    {
      "epoch": 4.233337732611852,
      "grad_norm": 0.19353853166103363,
      "learning_rate": 3.2652560323170734e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 14336016,
      "step": 16040
    },
    {
      "epoch": 4.234657516167348,
      "grad_norm": 0.05595673993229866,
      "learning_rate": 3.264321350906189e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 14340752,
      "step": 16045
    },
    {
      "epoch": 4.235977299722846,
      "grad_norm": 0.15499264001846313,
      "learning_rate": 3.263386551627346e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 14345168,
      "step": 16050
    },
    {
      "epoch": 4.237297083278342,
      "grad_norm": 0.25461477041244507,
      "learning_rate": 3.2624516346247055e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 14349552,
      "step": 16055
    },
    {
      "epoch": 4.238616866833839,
      "grad_norm": 0.05314868688583374,
      "learning_rate": 3.2615166000424404e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 14354288,
      "step": 16060
    },
    {
      "epoch": 4.239936650389336,
      "grad_norm": 0.11047704517841339,
      "learning_rate": 3.260581448024745e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 14358800,
      "step": 16065
    },
    {
      "epoch": 4.241256433944833,
      "grad_norm": 0.36406680941581726,
      "learning_rate": 3.2596461787158335e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 14363248,
      "step": 16070
    },
    {
      "epoch": 4.24257621750033,
      "grad_norm": 0.14550025761127472,
      "learning_rate": 3.258710792259934e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 14367632,
      "step": 16075
    },
    {
      "epoch": 4.243896001055827,
      "grad_norm": 0.0645856112241745,
      "learning_rate": 3.257775288801296e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 14372272,
      "step": 16080
    },
    {
      "epoch": 4.245215784611323,
      "grad_norm": 0.17523810267448425,
      "learning_rate": 3.256839668484186e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 14376432,
      "step": 16085
    },
    {
      "epoch": 4.246535568166821,
      "grad_norm": 0.09288103878498077,
      "learning_rate": 3.255903931452888e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 14381008,
      "step": 16090
    },
    {
      "epoch": 4.247855351722317,
      "grad_norm": 0.16668589413166046,
      "learning_rate": 3.2549680778517045e-05,
      "loss": 0.1116,
      "num_input_tokens_seen": 14385488,
      "step": 16095
    },
    {
      "epoch": 4.249175135277815,
      "grad_norm": 0.08292868733406067,
      "learning_rate": 3.2540321078249556e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 14389680,
      "step": 16100
    },
    {
      "epoch": 4.250494918833311,
      "grad_norm": 0.3844326436519623,
      "learning_rate": 3.2530960215169795e-05,
      "loss": 0.0698,
      "num_input_tokens_seen": 14394160,
      "step": 16105
    },
    {
      "epoch": 4.251814702388808,
      "grad_norm": 0.135445699095726,
      "learning_rate": 3.2521598190721345e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 14398832,
      "step": 16110
    },
    {
      "epoch": 4.253134485944305,
      "grad_norm": 0.05857017636299133,
      "learning_rate": 3.251223500634792e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 14403056,
      "step": 16115
    },
    {
      "epoch": 4.254454269499802,
      "grad_norm": 0.16728360950946808,
      "learning_rate": 3.2502870663493445e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 14407248,
      "step": 16120
    },
    {
      "epoch": 4.2557740530552985,
      "grad_norm": 0.0856282189488411,
      "learning_rate": 3.249350516360203e-05,
      "loss": 0.0783,
      "num_input_tokens_seen": 14411408,
      "step": 16125
    },
    {
      "epoch": 4.257093836610796,
      "grad_norm": 0.024195149540901184,
      "learning_rate": 3.248413850811797e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 14415824,
      "step": 16130
    },
    {
      "epoch": 4.2584136201662925,
      "grad_norm": 0.23630072176456451,
      "learning_rate": 3.2474770698485677e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 14420368,
      "step": 16135
    },
    {
      "epoch": 4.25973340372179,
      "grad_norm": 0.22448642551898956,
      "learning_rate": 3.246540173614983e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 14425104,
      "step": 16140
    },
    {
      "epoch": 4.2610531872772865,
      "grad_norm": 0.20074792206287384,
      "learning_rate": 3.2456031622555197e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 14429360,
      "step": 16145
    },
    {
      "epoch": 4.262372970832783,
      "grad_norm": 0.10946731269359589,
      "learning_rate": 3.2446660359146794e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 14433680,
      "step": 16150
    },
    {
      "epoch": 4.2636927543882805,
      "grad_norm": 0.35188889503479004,
      "learning_rate": 3.2437287947369786e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 14438352,
      "step": 16155
    },
    {
      "epoch": 4.265012537943777,
      "grad_norm": 0.46400460600852966,
      "learning_rate": 3.2427914388669525e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 14442800,
      "step": 16160
    },
    {
      "epoch": 4.266332321499274,
      "grad_norm": 0.21883760392665863,
      "learning_rate": 3.241853968449151e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 14447568,
      "step": 16165
    },
    {
      "epoch": 4.267652105054771,
      "grad_norm": 0.3653423488140106,
      "learning_rate": 3.240916383628144e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 14451920,
      "step": 16170
    },
    {
      "epoch": 4.268971888610268,
      "grad_norm": 0.48588311672210693,
      "learning_rate": 3.239978684548521e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 14456464,
      "step": 16175
    },
    {
      "epoch": 4.270291672165765,
      "grad_norm": 0.3170522451400757,
      "learning_rate": 3.239040871354885e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 14460944,
      "step": 16180
    },
    {
      "epoch": 4.271611455721262,
      "grad_norm": 0.330966979265213,
      "learning_rate": 3.2381029441918596e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 14465648,
      "step": 16185
    },
    {
      "epoch": 4.272931239276758,
      "grad_norm": 0.187547966837883,
      "learning_rate": 3.2371649032040845e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 14470128,
      "step": 16190
    },
    {
      "epoch": 4.274251022832256,
      "grad_norm": 0.3601433336734772,
      "learning_rate": 3.2362267485362174e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 14474768,
      "step": 16195
    },
    {
      "epoch": 4.275570806387752,
      "grad_norm": 0.06938692927360535,
      "learning_rate": 3.235288480332934e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 14479248,
      "step": 16200
    },
    {
      "epoch": 4.275570806387752,
      "eval_loss": 0.06928060948848724,
      "eval_runtime": 64.7828,
      "eval_samples_per_second": 103.963,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 14479248,
      "step": 16200
    },
    {
      "epoch": 4.27689058994325,
      "grad_norm": 0.08146052062511444,
      "learning_rate": 3.234350098738927e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 14483504,
      "step": 16205
    },
    {
      "epoch": 4.278210373498746,
      "grad_norm": 0.18147023022174835,
      "learning_rate": 3.233411603898906e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 14487888,
      "step": 16210
    },
    {
      "epoch": 4.279530157054243,
      "grad_norm": 0.2658587396144867,
      "learning_rate": 3.232472995957599e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 14492912,
      "step": 16215
    },
    {
      "epoch": 4.28084994060974,
      "grad_norm": 0.16622881591320038,
      "learning_rate": 3.231534275059751e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 14497520,
      "step": 16220
    },
    {
      "epoch": 4.282169724165237,
      "grad_norm": 0.05868997052311897,
      "learning_rate": 3.230595441350125e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 14502128,
      "step": 16225
    },
    {
      "epoch": 4.283489507720734,
      "grad_norm": 0.15182790160179138,
      "learning_rate": 3.2296564949735e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 14506640,
      "step": 16230
    },
    {
      "epoch": 4.284809291276231,
      "grad_norm": 0.17041447758674622,
      "learning_rate": 3.228717436074675e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 14511376,
      "step": 16235
    },
    {
      "epoch": 4.286129074831727,
      "grad_norm": 0.09544391930103302,
      "learning_rate": 3.227778264798463e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 14515696,
      "step": 16240
    },
    {
      "epoch": 4.287448858387225,
      "grad_norm": 0.08945830166339874,
      "learning_rate": 3.226838981289698e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 14520336,
      "step": 16245
    },
    {
      "epoch": 4.288768641942721,
      "grad_norm": 0.13536033034324646,
      "learning_rate": 3.225899585693227e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 14524592,
      "step": 16250
    },
    {
      "epoch": 4.290088425498218,
      "grad_norm": 0.03366810828447342,
      "learning_rate": 3.224960078153918e-05,
      "loss": 0.1031,
      "num_input_tokens_seen": 14529200,
      "step": 16255
    },
    {
      "epoch": 4.291408209053715,
      "grad_norm": 0.20745255053043365,
      "learning_rate": 3.224020458816655e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 14534000,
      "step": 16260
    },
    {
      "epoch": 4.292727992609212,
      "grad_norm": 0.4370614290237427,
      "learning_rate": 3.223080727826337e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 14538480,
      "step": 16265
    },
    {
      "epoch": 4.294047776164709,
      "grad_norm": 0.5274820923805237,
      "learning_rate": 3.222140885327885e-05,
      "loss": 0.1127,
      "num_input_tokens_seen": 14543152,
      "step": 16270
    },
    {
      "epoch": 4.295367559720206,
      "grad_norm": 0.31907156109809875,
      "learning_rate": 3.221200931466234e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 14547696,
      "step": 16275
    },
    {
      "epoch": 4.2966873432757025,
      "grad_norm": 0.24983958899974823,
      "learning_rate": 3.220260866386336e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 14552176,
      "step": 16280
    },
    {
      "epoch": 4.2980071268312,
      "grad_norm": 0.2035602480173111,
      "learning_rate": 3.21932069023316e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 14556912,
      "step": 16285
    },
    {
      "epoch": 4.2993269103866965,
      "grad_norm": 0.33681946992874146,
      "learning_rate": 3.218380403151695e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 14561168,
      "step": 16290
    },
    {
      "epoch": 4.300646693942193,
      "grad_norm": 0.1661912202835083,
      "learning_rate": 3.217440005286943e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 14565584,
      "step": 16295
    },
    {
      "epoch": 4.3019664774976905,
      "grad_norm": 0.19048462808132172,
      "learning_rate": 3.216499496783928e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 14570096,
      "step": 16300
    },
    {
      "epoch": 4.303286261053187,
      "grad_norm": 0.38355356454849243,
      "learning_rate": 3.2155588777876856e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 14574576,
      "step": 16305
    },
    {
      "epoch": 4.3046060446086845,
      "grad_norm": 0.26800528168678284,
      "learning_rate": 3.214618148443273e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 14579184,
      "step": 16310
    },
    {
      "epoch": 4.305925828164181,
      "grad_norm": 0.3803464472293854,
      "learning_rate": 3.2136773088957595e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 14583792,
      "step": 16315
    },
    {
      "epoch": 4.307245611719678,
      "grad_norm": 0.1807071566581726,
      "learning_rate": 3.2127363592902374e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 14588304,
      "step": 16320
    },
    {
      "epoch": 4.308565395275175,
      "grad_norm": 0.20470990240573883,
      "learning_rate": 3.211795299771812e-05,
      "loss": 0.105,
      "num_input_tokens_seen": 14592912,
      "step": 16325
    },
    {
      "epoch": 4.309885178830672,
      "grad_norm": 0.18536344170570374,
      "learning_rate": 3.210854130485605e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 14597424,
      "step": 16330
    },
    {
      "epoch": 4.311204962386169,
      "grad_norm": 0.12596353888511658,
      "learning_rate": 3.209912851576759e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 14601808,
      "step": 16335
    },
    {
      "epoch": 4.312524745941666,
      "grad_norm": 0.21420417726039886,
      "learning_rate": 3.208971463190431e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 14606448,
      "step": 16340
    },
    {
      "epoch": 4.313844529497162,
      "grad_norm": 0.15049614012241364,
      "learning_rate": 3.208029965471793e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 14611056,
      "step": 16345
    },
    {
      "epoch": 4.31516431305266,
      "grad_norm": 0.46172210574150085,
      "learning_rate": 3.2070883585660364e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 14615568,
      "step": 16350
    },
    {
      "epoch": 4.316484096608156,
      "grad_norm": 0.1275540590286255,
      "learning_rate": 3.20614664261837e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 14620688,
      "step": 16355
    },
    {
      "epoch": 4.317803880163654,
      "grad_norm": 0.25704100728034973,
      "learning_rate": 3.205204817774016e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 14625424,
      "step": 16360
    },
    {
      "epoch": 4.31912366371915,
      "grad_norm": 0.20196960866451263,
      "learning_rate": 3.204262884178218e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 14629840,
      "step": 16365
    },
    {
      "epoch": 4.320443447274647,
      "grad_norm": 0.374894380569458,
      "learning_rate": 3.2033208419762314e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 14634224,
      "step": 16370
    },
    {
      "epoch": 4.321763230830144,
      "grad_norm": 0.047620516270399094,
      "learning_rate": 3.2023786913133344e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 14638672,
      "step": 16375
    },
    {
      "epoch": 4.323083014385641,
      "grad_norm": 0.06040734425187111,
      "learning_rate": 3.201436432334816e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 14642896,
      "step": 16380
    },
    {
      "epoch": 4.324402797941137,
      "grad_norm": 0.1091158390045166,
      "learning_rate": 3.2004940651859844e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 14647120,
      "step": 16385
    },
    {
      "epoch": 4.325722581496635,
      "grad_norm": 0.14791551232337952,
      "learning_rate": 3.1995515900121655e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 14651824,
      "step": 16390
    },
    {
      "epoch": 4.327042365052131,
      "grad_norm": 0.06846854835748672,
      "learning_rate": 3.1986090069587e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 14656464,
      "step": 16395
    },
    {
      "epoch": 4.328362148607629,
      "grad_norm": 0.2865278720855713,
      "learning_rate": 3.1976663161709466e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 14660976,
      "step": 16400
    },
    {
      "epoch": 4.328362148607629,
      "eval_loss": 0.0692586898803711,
      "eval_runtime": 64.7386,
      "eval_samples_per_second": 104.034,
      "eval_steps_per_second": 26.012,
      "num_input_tokens_seen": 14660976,
      "step": 16400
    },
    {
      "epoch": 4.329681932163125,
      "grad_norm": 0.279731422662735,
      "learning_rate": 3.196723517794279e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 14665392,
      "step": 16405
    },
    {
      "epoch": 4.331001715718622,
      "grad_norm": 0.17093487083911896,
      "learning_rate": 3.19578061197409e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 14669712,
      "step": 16410
    },
    {
      "epoch": 4.332321499274119,
      "grad_norm": 0.26023542881011963,
      "learning_rate": 3.194837598855787e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 14673808,
      "step": 16415
    },
    {
      "epoch": 4.333641282829616,
      "grad_norm": 0.25157028436660767,
      "learning_rate": 3.193894478584794e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 14678256,
      "step": 16420
    },
    {
      "epoch": 4.334961066385113,
      "grad_norm": 0.31848573684692383,
      "learning_rate": 3.192951251306553e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 14682640,
      "step": 16425
    },
    {
      "epoch": 4.33628084994061,
      "grad_norm": 0.10602187365293503,
      "learning_rate": 3.192007917166521e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 14687376,
      "step": 16430
    },
    {
      "epoch": 4.337600633496106,
      "grad_norm": 0.19057434797286987,
      "learning_rate": 3.191064476310171e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 14692048,
      "step": 16435
    },
    {
      "epoch": 4.338920417051604,
      "grad_norm": 0.2442363053560257,
      "learning_rate": 3.1901209288829944e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 14696720,
      "step": 16440
    },
    {
      "epoch": 4.3402402006071,
      "grad_norm": 0.24493849277496338,
      "learning_rate": 3.1891772750304985e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 14701072,
      "step": 16445
    },
    {
      "epoch": 4.341559984162597,
      "grad_norm": 0.2435508817434311,
      "learning_rate": 3.188233514898206e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 14705744,
      "step": 16450
    },
    {
      "epoch": 4.342879767718094,
      "grad_norm": 0.287352055311203,
      "learning_rate": 3.187289648631657e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 14710352,
      "step": 16455
    },
    {
      "epoch": 4.344199551273591,
      "grad_norm": 0.5548743605613708,
      "learning_rate": 3.186345676376406e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 14714544,
      "step": 16460
    },
    {
      "epoch": 4.345519334829088,
      "grad_norm": 0.08118835091590881,
      "learning_rate": 3.1854015982780275e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 14718960,
      "step": 16465
    },
    {
      "epoch": 4.346839118384585,
      "grad_norm": 0.5737006664276123,
      "learning_rate": 3.1844574144821084e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 14723504,
      "step": 16470
    },
    {
      "epoch": 4.3481589019400815,
      "grad_norm": 0.3589685559272766,
      "learning_rate": 3.1835131251342554e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 14727696,
      "step": 16475
    },
    {
      "epoch": 4.349478685495579,
      "grad_norm": 0.3090875446796417,
      "learning_rate": 3.182568730380089e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 14731792,
      "step": 16480
    },
    {
      "epoch": 4.3507984690510755,
      "grad_norm": 0.6478426456451416,
      "learning_rate": 3.181624230365245e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 14736496,
      "step": 16485
    },
    {
      "epoch": 4.352118252606573,
      "grad_norm": 0.3426189422607422,
      "learning_rate": 3.180679625235381e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 14740784,
      "step": 16490
    },
    {
      "epoch": 4.3534380361620695,
      "grad_norm": 0.5377569198608398,
      "learning_rate": 3.1797349151361646e-05,
      "loss": 0.1396,
      "num_input_tokens_seen": 14745200,
      "step": 16495
    },
    {
      "epoch": 4.354757819717566,
      "grad_norm": 0.272971510887146,
      "learning_rate": 3.178790100213281e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 14749744,
      "step": 16500
    },
    {
      "epoch": 4.3560776032730635,
      "grad_norm": 0.2793237864971161,
      "learning_rate": 3.1778451806124346e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 14754096,
      "step": 16505
    },
    {
      "epoch": 4.35739738682856,
      "grad_norm": 0.35539141297340393,
      "learning_rate": 3.176900156479342e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 14758640,
      "step": 16510
    },
    {
      "epoch": 4.3587171703840575,
      "grad_norm": 0.19726309180259705,
      "learning_rate": 3.17595502795974e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 14762928,
      "step": 16515
    },
    {
      "epoch": 4.360036953939554,
      "grad_norm": 0.1303650289773941,
      "learning_rate": 3.175009795199377e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 14767440,
      "step": 16520
    },
    {
      "epoch": 4.361356737495051,
      "grad_norm": 0.21357670426368713,
      "learning_rate": 3.1740644583440224e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 14772144,
      "step": 16525
    },
    {
      "epoch": 4.362676521050548,
      "grad_norm": 0.4475732147693634,
      "learning_rate": 3.173119017539457e-05,
      "loss": 0.1034,
      "num_input_tokens_seen": 14776624,
      "step": 16530
    },
    {
      "epoch": 4.363996304606045,
      "grad_norm": 0.20076723396778107,
      "learning_rate": 3.172173472931479e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 14781232,
      "step": 16535
    },
    {
      "epoch": 4.365316088161541,
      "grad_norm": 0.11919048428535461,
      "learning_rate": 3.1712278246659055e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 14785648,
      "step": 16540
    },
    {
      "epoch": 4.366635871717039,
      "grad_norm": 0.18762211501598358,
      "learning_rate": 3.170282072888566e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 14790256,
      "step": 16545
    },
    {
      "epoch": 4.367955655272535,
      "grad_norm": 0.06658461689949036,
      "learning_rate": 3.169336217745307e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 14794736,
      "step": 16550
    },
    {
      "epoch": 4.369275438828033,
      "grad_norm": 0.1254887729883194,
      "learning_rate": 3.1683902593819924e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 14799408,
      "step": 16555
    },
    {
      "epoch": 4.370595222383529,
      "grad_norm": 0.2671094536781311,
      "learning_rate": 3.1674441979445e-05,
      "loss": 0.11,
      "num_input_tokens_seen": 14803760,
      "step": 16560
    },
    {
      "epoch": 4.371915005939026,
      "grad_norm": 0.3242781162261963,
      "learning_rate": 3.166498033578725e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 14808112,
      "step": 16565
    },
    {
      "epoch": 4.373234789494523,
      "grad_norm": 0.2077052742242813,
      "learning_rate": 3.165551766430578e-05,
      "loss": 0.0764,
      "num_input_tokens_seen": 14812528,
      "step": 16570
    },
    {
      "epoch": 4.37455457305002,
      "grad_norm": 0.03699875250458717,
      "learning_rate": 3.164605396645984e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 14816816,
      "step": 16575
    },
    {
      "epoch": 4.375874356605516,
      "grad_norm": 0.10970988869667053,
      "learning_rate": 3.163658924370886e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 14821424,
      "step": 16580
    },
    {
      "epoch": 4.377194140161014,
      "grad_norm": 0.19884027540683746,
      "learning_rate": 3.1627123497512415e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 14826096,
      "step": 16585
    },
    {
      "epoch": 4.37851392371651,
      "grad_norm": 0.26470932364463806,
      "learning_rate": 3.1617656729330245e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 14830352,
      "step": 16590
    },
    {
      "epoch": 4.379833707272008,
      "grad_norm": 0.2684254050254822,
      "learning_rate": 3.1608188940622255e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 14834736,
      "step": 16595
    },
    {
      "epoch": 4.381153490827504,
      "grad_norm": 0.30212345719337463,
      "learning_rate": 3.159872013284847e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 14839056,
      "step": 16600
    },
    {
      "epoch": 4.381153490827504,
      "eval_loss": 0.06891518831253052,
      "eval_runtime": 64.8211,
      "eval_samples_per_second": 103.901,
      "eval_steps_per_second": 25.979,
      "num_input_tokens_seen": 14839056,
      "step": 16600
    },
    {
      "epoch": 4.382473274383001,
      "grad_norm": 0.4185463786125183,
      "learning_rate": 3.1589250307469134e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 14843440,
      "step": 16605
    },
    {
      "epoch": 4.383793057938498,
      "grad_norm": 0.1523248702287674,
      "learning_rate": 3.1579779465944586e-05,
      "loss": 0.0571,
      "num_input_tokens_seen": 14847664,
      "step": 16610
    },
    {
      "epoch": 4.385112841493995,
      "grad_norm": 0.08444304019212723,
      "learning_rate": 3.1570307609735363e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 14852080,
      "step": 16615
    },
    {
      "epoch": 4.386432625049492,
      "grad_norm": 0.32412415742874146,
      "learning_rate": 3.156083474030213e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 14856336,
      "step": 16620
    },
    {
      "epoch": 4.387752408604989,
      "grad_norm": 0.39373818039894104,
      "learning_rate": 3.155136085910573e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 14861040,
      "step": 16625
    },
    {
      "epoch": 4.389072192160485,
      "grad_norm": 0.10800865292549133,
      "learning_rate": 3.154188596760717e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 14865680,
      "step": 16630
    },
    {
      "epoch": 4.390391975715983,
      "grad_norm": 0.04714430496096611,
      "learning_rate": 3.153241006726757e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 14870256,
      "step": 16635
    },
    {
      "epoch": 4.391711759271479,
      "grad_norm": 0.16838154196739197,
      "learning_rate": 3.152293315954825e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 14874544,
      "step": 16640
    },
    {
      "epoch": 4.393031542826977,
      "grad_norm": 0.16970589756965637,
      "learning_rate": 3.1513455245910666e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 14879088,
      "step": 16645
    },
    {
      "epoch": 4.394351326382473,
      "grad_norm": 0.06012123450636864,
      "learning_rate": 3.150397632781643e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 14883696,
      "step": 16650
    },
    {
      "epoch": 4.39567110993797,
      "grad_norm": 0.43462684750556946,
      "learning_rate": 3.149449640672731e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 14887952,
      "step": 16655
    },
    {
      "epoch": 4.396990893493467,
      "grad_norm": 0.14801432192325592,
      "learning_rate": 3.148501548410523e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 14892304,
      "step": 16660
    },
    {
      "epoch": 4.398310677048964,
      "grad_norm": 0.20283685624599457,
      "learning_rate": 3.1475533561412256e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 14896880,
      "step": 16665
    },
    {
      "epoch": 4.3996304606044605,
      "grad_norm": 0.3539446294307709,
      "learning_rate": 3.146605064011065e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 14901488,
      "step": 16670
    },
    {
      "epoch": 4.400950244159958,
      "grad_norm": 0.4137662947177887,
      "learning_rate": 3.145656672166277e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 14905872,
      "step": 16675
    },
    {
      "epoch": 4.4022700277154545,
      "grad_norm": 0.13412520289421082,
      "learning_rate": 3.144708180753116e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 14909872,
      "step": 16680
    },
    {
      "epoch": 4.403589811270952,
      "grad_norm": 0.07847852259874344,
      "learning_rate": 3.143759589917851e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 14914096,
      "step": 16685
    },
    {
      "epoch": 4.4049095948264485,
      "grad_norm": 0.06109490245580673,
      "learning_rate": 3.142810899806768e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 14918320,
      "step": 16690
    },
    {
      "epoch": 4.406229378381945,
      "grad_norm": 0.25303465127944946,
      "learning_rate": 3.141862110566166e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 14922608,
      "step": 16695
    },
    {
      "epoch": 4.4075491619374425,
      "grad_norm": 0.18817973136901855,
      "learning_rate": 3.1409132223423606e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 14927248,
      "step": 16700
    },
    {
      "epoch": 4.408868945492939,
      "grad_norm": 0.3521624505519867,
      "learning_rate": 3.139964235281682e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 14931664,
      "step": 16705
    },
    {
      "epoch": 4.410188729048436,
      "grad_norm": 0.20132042467594147,
      "learning_rate": 3.139015149530476e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 14935792,
      "step": 16710
    },
    {
      "epoch": 4.411508512603933,
      "grad_norm": 0.1608770489692688,
      "learning_rate": 3.1380659652351034e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 14940208,
      "step": 16715
    },
    {
      "epoch": 4.41282829615943,
      "grad_norm": 0.11961697787046432,
      "learning_rate": 3.137116682541941e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 14944752,
      "step": 16720
    },
    {
      "epoch": 4.414148079714927,
      "grad_norm": 0.38208794593811035,
      "learning_rate": 3.136167301597379e-05,
      "loss": 0.1234,
      "num_input_tokens_seen": 14949136,
      "step": 16725
    },
    {
      "epoch": 4.415467863270424,
      "grad_norm": 0.2332635074853897,
      "learning_rate": 3.1352178225478254e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 14953488,
      "step": 16730
    },
    {
      "epoch": 4.41678764682592,
      "grad_norm": 0.11598386615514755,
      "learning_rate": 3.1342682455396996e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 14958096,
      "step": 16735
    },
    {
      "epoch": 4.418107430381418,
      "grad_norm": 0.10811329632997513,
      "learning_rate": 3.133318570719441e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 14962640,
      "step": 16740
    },
    {
      "epoch": 4.419427213936914,
      "grad_norm": 0.12695413827896118,
      "learning_rate": 3.132368798233499e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 14967088,
      "step": 16745
    },
    {
      "epoch": 4.420746997492412,
      "grad_norm": 0.3933844566345215,
      "learning_rate": 3.131418928228342e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 14971696,
      "step": 16750
    },
    {
      "epoch": 4.422066781047908,
      "grad_norm": 0.22276704013347626,
      "learning_rate": 3.1304689608504514e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 14976272,
      "step": 16755
    },
    {
      "epoch": 4.423386564603405,
      "grad_norm": 0.2870182991027832,
      "learning_rate": 3.129518896246324e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 14980752,
      "step": 16760
    },
    {
      "epoch": 4.424706348158902,
      "grad_norm": 0.10488896816968918,
      "learning_rate": 3.128568734562472e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 14985488,
      "step": 16765
    },
    {
      "epoch": 4.426026131714399,
      "grad_norm": 0.07172037661075592,
      "learning_rate": 3.127618475945421e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 14989616,
      "step": 16770
    },
    {
      "epoch": 4.427345915269896,
      "grad_norm": 0.24923862516880035,
      "learning_rate": 3.126668120541715e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 14993936,
      "step": 16775
    },
    {
      "epoch": 4.428665698825393,
      "grad_norm": 0.0857858955860138,
      "learning_rate": 3.1257176684979096e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 14998768,
      "step": 16780
    },
    {
      "epoch": 4.429985482380889,
      "grad_norm": 0.05052291229367256,
      "learning_rate": 3.124767119960576e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 15003184,
      "step": 16785
    },
    {
      "epoch": 4.431305265936387,
      "grad_norm": 0.2406575083732605,
      "learning_rate": 3.123816475076301e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 15007376,
      "step": 16790
    },
    {
      "epoch": 4.432625049491883,
      "grad_norm": 0.42316174507141113,
      "learning_rate": 3.122865733991687e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 15011536,
      "step": 16795
    },
    {
      "epoch": 4.43394483304738,
      "grad_norm": 0.357875257730484,
      "learning_rate": 3.1219148968533486e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 15016048,
      "step": 16800
    },
    {
      "epoch": 4.43394483304738,
      "eval_loss": 0.06896334141492844,
      "eval_runtime": 64.7731,
      "eval_samples_per_second": 103.978,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 15016048,
      "step": 16800
    },
    {
      "epoch": 4.435264616602877,
      "grad_norm": 0.1395978480577469,
      "learning_rate": 3.120963963807918e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 15020720,
      "step": 16805
    },
    {
      "epoch": 4.436584400158374,
      "grad_norm": 0.15530188381671906,
      "learning_rate": 3.12001293500204e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 15025168,
      "step": 16810
    },
    {
      "epoch": 4.437904183713871,
      "grad_norm": 0.10831574350595474,
      "learning_rate": 3.1190618105823765e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 15029360,
      "step": 16815
    },
    {
      "epoch": 4.439223967269368,
      "grad_norm": 0.2043435126543045,
      "learning_rate": 3.118110590695603e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 15034064,
      "step": 16820
    },
    {
      "epoch": 4.440543750824864,
      "grad_norm": 0.12220444530248642,
      "learning_rate": 3.117159275488407e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 15038672,
      "step": 16825
    },
    {
      "epoch": 4.441863534380362,
      "grad_norm": 0.15089628100395203,
      "learning_rate": 3.1162078651074956e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 15043408,
      "step": 16830
    },
    {
      "epoch": 4.443183317935858,
      "grad_norm": 0.04405693709850311,
      "learning_rate": 3.1152563596995885e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 15048112,
      "step": 16835
    },
    {
      "epoch": 4.444503101491355,
      "grad_norm": 0.13576270639896393,
      "learning_rate": 3.1143047594114186e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 15052912,
      "step": 16840
    },
    {
      "epoch": 4.445822885046852,
      "grad_norm": 0.08532340824604034,
      "learning_rate": 3.113353064389734e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 15057744,
      "step": 16845
    },
    {
      "epoch": 4.447142668602349,
      "grad_norm": 0.1438058614730835,
      "learning_rate": 3.1124012747812993e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 15062128,
      "step": 16850
    },
    {
      "epoch": 4.448462452157846,
      "grad_norm": 0.13937461376190186,
      "learning_rate": 3.1114493907328936e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 15066640,
      "step": 16855
    },
    {
      "epoch": 4.449782235713343,
      "grad_norm": 0.1294962465763092,
      "learning_rate": 3.110497412391306e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 15071216,
      "step": 16860
    },
    {
      "epoch": 4.4511020192688395,
      "grad_norm": 0.11827854067087173,
      "learning_rate": 3.1095453399033466e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 15075984,
      "step": 16865
    },
    {
      "epoch": 4.452421802824337,
      "grad_norm": 0.17024582624435425,
      "learning_rate": 3.108593173415835e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 15080592,
      "step": 16870
    },
    {
      "epoch": 4.4537415863798335,
      "grad_norm": 0.4058993458747864,
      "learning_rate": 3.107640913075609e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 15085136,
      "step": 16875
    },
    {
      "epoch": 4.455061369935331,
      "grad_norm": 0.11020553112030029,
      "learning_rate": 3.106688559029517e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 15089776,
      "step": 16880
    },
    {
      "epoch": 4.4563811534908275,
      "grad_norm": 0.4051533341407776,
      "learning_rate": 3.105736111424425e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 15094288,
      "step": 16885
    },
    {
      "epoch": 4.457700937046324,
      "grad_norm": 0.22695191204547882,
      "learning_rate": 3.1047835704072136e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 15098512,
      "step": 16890
    },
    {
      "epoch": 4.4590207206018215,
      "grad_norm": 0.27577629685401917,
      "learning_rate": 3.103830936124775e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 15102896,
      "step": 16895
    },
    {
      "epoch": 4.460340504157318,
      "grad_norm": 0.22035843133926392,
      "learning_rate": 3.102878208724018e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 15107568,
      "step": 16900
    },
    {
      "epoch": 4.4616602877128155,
      "grad_norm": 0.279906690120697,
      "learning_rate": 3.101925388351865e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 15111920,
      "step": 16905
    },
    {
      "epoch": 4.462980071268312,
      "grad_norm": 0.40887123346328735,
      "learning_rate": 3.1009724751552515e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 15116528,
      "step": 16910
    },
    {
      "epoch": 4.464299854823809,
      "grad_norm": 0.19499571621418,
      "learning_rate": 3.100019469281131e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 15120944,
      "step": 16915
    },
    {
      "epoch": 4.465619638379306,
      "grad_norm": 0.16979607939720154,
      "learning_rate": 3.0990663708764685e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 15125328,
      "step": 16920
    },
    {
      "epoch": 4.466939421934803,
      "grad_norm": 0.14564411342144012,
      "learning_rate": 3.098113180088243e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 15129584,
      "step": 16925
    },
    {
      "epoch": 4.468259205490299,
      "grad_norm": 0.2107173055410385,
      "learning_rate": 3.097159897063448e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 15133808,
      "step": 16930
    },
    {
      "epoch": 4.469578989045797,
      "grad_norm": 0.22117169201374054,
      "learning_rate": 3.096206521949094e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 15138672,
      "step": 16935
    },
    {
      "epoch": 4.470898772601293,
      "grad_norm": 0.1824360489845276,
      "learning_rate": 3.0952530548922006e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 15143472,
      "step": 16940
    },
    {
      "epoch": 4.472218556156791,
      "grad_norm": 0.20703795552253723,
      "learning_rate": 3.0942994960398064e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 15147984,
      "step": 16945
    },
    {
      "epoch": 4.473538339712287,
      "grad_norm": 0.5699824690818787,
      "learning_rate": 3.093345845538961e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 15152336,
      "step": 16950
    },
    {
      "epoch": 4.474858123267784,
      "grad_norm": 0.1892164796590805,
      "learning_rate": 3.09239210353673e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 15156720,
      "step": 16955
    },
    {
      "epoch": 4.476177906823281,
      "grad_norm": 0.12792544066905975,
      "learning_rate": 3.0914382701801926e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 15161104,
      "step": 16960
    },
    {
      "epoch": 4.477497690378778,
      "grad_norm": 0.06650001555681229,
      "learning_rate": 3.090484345616441e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 15165744,
      "step": 16965
    },
    {
      "epoch": 4.478817473934274,
      "grad_norm": 0.6658869981765747,
      "learning_rate": 3.0895303299925825e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 15170384,
      "step": 16970
    },
    {
      "epoch": 4.480137257489772,
      "grad_norm": 0.24367541074752808,
      "learning_rate": 3.0885762234557393e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 15174544,
      "step": 16975
    },
    {
      "epoch": 4.481457041045268,
      "grad_norm": 0.2661319077014923,
      "learning_rate": 3.087622026153045e-05,
      "loss": 0.1055,
      "num_input_tokens_seen": 15178736,
      "step": 16980
    },
    {
      "epoch": 4.482776824600766,
      "grad_norm": 0.12424825876951218,
      "learning_rate": 3.086667738231651e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 15183024,
      "step": 16985
    },
    {
      "epoch": 4.484096608156262,
      "grad_norm": 0.14623700082302094,
      "learning_rate": 3.085713359838718e-05,
      "loss": 0.1048,
      "num_input_tokens_seen": 15187664,
      "step": 16990
    },
    {
      "epoch": 4.485416391711759,
      "grad_norm": 0.11514277756214142,
      "learning_rate": 3.084758891121425e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 15192048,
      "step": 16995
    },
    {
      "epoch": 4.486736175267256,
      "grad_norm": 0.032668016850948334,
      "learning_rate": 3.083804332226963e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 15196432,
      "step": 17000
    },
    {
      "epoch": 4.486736175267256,
      "eval_loss": 0.06966648995876312,
      "eval_runtime": 64.7628,
      "eval_samples_per_second": 103.995,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 15196432,
      "step": 17000
    },
    {
      "epoch": 4.488055958822753,
      "grad_norm": 0.2274118810892105,
      "learning_rate": 3.082849683302536e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 15200880,
      "step": 17005
    },
    {
      "epoch": 4.48937574237825,
      "grad_norm": 0.03433378413319588,
      "learning_rate": 3.081894944495363e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 15205296,
      "step": 17010
    },
    {
      "epoch": 4.490695525933747,
      "grad_norm": 0.08687478303909302,
      "learning_rate": 3.080940115952677e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 15210032,
      "step": 17015
    },
    {
      "epoch": 4.492015309489243,
      "grad_norm": 0.14901818335056305,
      "learning_rate": 3.0799851978217245e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 15214384,
      "step": 17020
    },
    {
      "epoch": 4.493335093044741,
      "grad_norm": 0.25301143527030945,
      "learning_rate": 3.0790301902497666e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 15218992,
      "step": 17025
    },
    {
      "epoch": 4.494654876600237,
      "grad_norm": 0.09659424424171448,
      "learning_rate": 3.078075093384076e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 15223408,
      "step": 17030
    },
    {
      "epoch": 4.495974660155735,
      "grad_norm": 0.24253039062023163,
      "learning_rate": 3.077119907371942e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 15227664,
      "step": 17035
    },
    {
      "epoch": 4.497294443711231,
      "grad_norm": 0.10424157232046127,
      "learning_rate": 3.076164632360666e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 15232368,
      "step": 17040
    },
    {
      "epoch": 4.498614227266728,
      "grad_norm": 0.1829790621995926,
      "learning_rate": 3.075209268497563e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 15236496,
      "step": 17045
    },
    {
      "epoch": 4.499934010822225,
      "grad_norm": 0.13616782426834106,
      "learning_rate": 3.074253815929961e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 15240976,
      "step": 17050
    },
    {
      "epoch": 4.501253794377722,
      "grad_norm": 0.05685742199420929,
      "learning_rate": 3.0732982748052054e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 15245456,
      "step": 17055
    },
    {
      "epoch": 4.502573577933219,
      "grad_norm": 0.12395831942558289,
      "learning_rate": 3.072342645270651e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 15249872,
      "step": 17060
    },
    {
      "epoch": 4.503893361488716,
      "grad_norm": 0.2523971199989319,
      "learning_rate": 3.071386927473668e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 15254480,
      "step": 17065
    },
    {
      "epoch": 4.5052131450442126,
      "grad_norm": 0.12605805695056915,
      "learning_rate": 3.0704311215616404e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 15258704,
      "step": 17070
    },
    {
      "epoch": 4.50653292859971,
      "grad_norm": 0.21279558539390564,
      "learning_rate": 3.0694752276819656e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 15263120,
      "step": 17075
    },
    {
      "epoch": 4.5078527121552066,
      "grad_norm": 0.09739082306623459,
      "learning_rate": 3.068519245982054e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 15267664,
      "step": 17080
    },
    {
      "epoch": 4.509172495710703,
      "grad_norm": 0.21431399881839752,
      "learning_rate": 3.0675631766093304e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 15272144,
      "step": 17085
    },
    {
      "epoch": 4.5104922792662006,
      "grad_norm": 0.06275000423192978,
      "learning_rate": 3.066607019711232e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 15276816,
      "step": 17090
    },
    {
      "epoch": 4.511812062821697,
      "grad_norm": 0.2777114510536194,
      "learning_rate": 3.065650775435211e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 15281424,
      "step": 17095
    },
    {
      "epoch": 4.513131846377194,
      "grad_norm": 0.06710651516914368,
      "learning_rate": 3.0646944439287326e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 15286064,
      "step": 17100
    },
    {
      "epoch": 4.514451629932691,
      "grad_norm": 0.23893192410469055,
      "learning_rate": 3.0637380253392736e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 15290736,
      "step": 17105
    },
    {
      "epoch": 4.515771413488188,
      "grad_norm": 0.1470186412334442,
      "learning_rate": 3.062781519814327e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 15294960,
      "step": 17110
    },
    {
      "epoch": 4.517091197043685,
      "grad_norm": 0.1950264424085617,
      "learning_rate": 3.0618249275013985e-05,
      "loss": 0.0896,
      "num_input_tokens_seen": 15299088,
      "step": 17115
    },
    {
      "epoch": 4.518410980599182,
      "grad_norm": 0.3128533363342285,
      "learning_rate": 3.060868248548005e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 15303760,
      "step": 17120
    },
    {
      "epoch": 4.519730764154678,
      "grad_norm": 0.20763787627220154,
      "learning_rate": 3.0599114831016796e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 15308208,
      "step": 17125
    },
    {
      "epoch": 4.521050547710176,
      "grad_norm": 0.10270621627569199,
      "learning_rate": 3.0589546313099666e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 15312496,
      "step": 17130
    },
    {
      "epoch": 4.522370331265672,
      "grad_norm": 0.10230451822280884,
      "learning_rate": 3.0579976933204255e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 15317040,
      "step": 17135
    },
    {
      "epoch": 4.52369011482117,
      "grad_norm": 0.06859953701496124,
      "learning_rate": 3.0570406692806284e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 15321424,
      "step": 17140
    },
    {
      "epoch": 4.525009898376666,
      "grad_norm": 0.5696061849594116,
      "learning_rate": 3.05608355933816e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 15325808,
      "step": 17145
    },
    {
      "epoch": 4.526329681932163,
      "grad_norm": 0.07461816072463989,
      "learning_rate": 3.055126363640618e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 15329936,
      "step": 17150
    },
    {
      "epoch": 4.52764946548766,
      "grad_norm": 0.23385633528232574,
      "learning_rate": 3.0541690823356146e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 15334576,
      "step": 17155
    },
    {
      "epoch": 4.528969249043157,
      "grad_norm": 0.24719227850437164,
      "learning_rate": 3.053211715570775e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 15339024,
      "step": 17160
    },
    {
      "epoch": 4.530289032598654,
      "grad_norm": 0.15709394216537476,
      "learning_rate": 3.052254263493736e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 15343280,
      "step": 17165
    },
    {
      "epoch": 4.531608816154151,
      "grad_norm": 0.26371175050735474,
      "learning_rate": 3.0512967262521498e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 15347792,
      "step": 17170
    },
    {
      "epoch": 4.532928599709647,
      "grad_norm": 0.08827108144760132,
      "learning_rate": 3.0503391039936803e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 15352176,
      "step": 17175
    },
    {
      "epoch": 4.534248383265145,
      "grad_norm": 0.14745686948299408,
      "learning_rate": 3.0493813968660056e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 15356528,
      "step": 17180
    },
    {
      "epoch": 4.535568166820641,
      "grad_norm": 0.13035918772220612,
      "learning_rate": 3.0484236050168153e-05,
      "loss": 0.0203,
      "num_input_tokens_seen": 15360880,
      "step": 17185
    },
    {
      "epoch": 4.536887950376139,
      "grad_norm": 0.23638279736042023,
      "learning_rate": 3.0474657285938123e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 15365104,
      "step": 17190
    },
    {
      "epoch": 4.538207733931635,
      "grad_norm": 0.24518848955631256,
      "learning_rate": 3.046507767744715e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 15369680,
      "step": 17195
    },
    {
      "epoch": 4.539527517487132,
      "grad_norm": 0.129789799451828,
      "learning_rate": 3.045549722617252e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 15374128,
      "step": 17200
    },
    {
      "epoch": 4.539527517487132,
      "eval_loss": 0.06896604597568512,
      "eval_runtime": 64.7282,
      "eval_samples_per_second": 104.05,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 15374128,
      "step": 17200
    },
    {
      "epoch": 4.540847301042629,
      "grad_norm": 0.21089038252830505,
      "learning_rate": 3.0445915933591658e-05,
      "loss": 0.0896,
      "num_input_tokens_seen": 15378640,
      "step": 17205
    },
    {
      "epoch": 4.542167084598126,
      "grad_norm": 0.2550269663333893,
      "learning_rate": 3.0436333801182114e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 15383504,
      "step": 17210
    },
    {
      "epoch": 4.5434868681536225,
      "grad_norm": 0.27197185158729553,
      "learning_rate": 3.0426750830421596e-05,
      "loss": 0.1017,
      "num_input_tokens_seen": 15388272,
      "step": 17215
    },
    {
      "epoch": 4.54480665170912,
      "grad_norm": 0.23533427715301514,
      "learning_rate": 3.0417167022787897e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 15392816,
      "step": 17220
    },
    {
      "epoch": 4.5461264352646165,
      "grad_norm": 0.15790709853172302,
      "learning_rate": 3.0407582379758966e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 15397296,
      "step": 17225
    },
    {
      "epoch": 4.547446218820113,
      "grad_norm": 0.25732430815696716,
      "learning_rate": 3.039799690281287e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 15402032,
      "step": 17230
    },
    {
      "epoch": 4.5487660023756105,
      "grad_norm": 0.6158699989318848,
      "learning_rate": 3.0388410593427823e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 15406480,
      "step": 17235
    },
    {
      "epoch": 4.550085785931107,
      "grad_norm": 0.07004453986883163,
      "learning_rate": 3.0378823453082146e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 15410768,
      "step": 17240
    },
    {
      "epoch": 4.5514055694866045,
      "grad_norm": 0.4746168255805969,
      "learning_rate": 3.03692354832543e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 15415248,
      "step": 17245
    },
    {
      "epoch": 4.552725353042101,
      "grad_norm": 0.23729097843170166,
      "learning_rate": 3.0359646685422865e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 15419984,
      "step": 17250
    },
    {
      "epoch": 4.554045136597598,
      "grad_norm": 0.02623366378247738,
      "learning_rate": 3.035005706106656e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 15424208,
      "step": 17255
    },
    {
      "epoch": 4.555364920153095,
      "grad_norm": 0.25380733609199524,
      "learning_rate": 3.034046661166422e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 15428368,
      "step": 17260
    },
    {
      "epoch": 4.556684703708592,
      "grad_norm": 0.16102853417396545,
      "learning_rate": 3.033087533869482e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 15433040,
      "step": 17265
    },
    {
      "epoch": 4.558004487264089,
      "grad_norm": 0.1977294385433197,
      "learning_rate": 3.0321283243637444e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 15437648,
      "step": 17270
    },
    {
      "epoch": 4.559324270819586,
      "grad_norm": 0.3743496537208557,
      "learning_rate": 3.0311690327971326e-05,
      "loss": 0.1182,
      "num_input_tokens_seen": 15442448,
      "step": 17275
    },
    {
      "epoch": 4.560644054375082,
      "grad_norm": 0.28324317932128906,
      "learning_rate": 3.030209659317581e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 15447024,
      "step": 17280
    },
    {
      "epoch": 4.56196383793058,
      "grad_norm": 0.08828887343406677,
      "learning_rate": 3.0292502040730362e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 15451472,
      "step": 17285
    },
    {
      "epoch": 4.563283621486076,
      "grad_norm": 0.3007761240005493,
      "learning_rate": 3.0282906672114597e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 15455824,
      "step": 17290
    },
    {
      "epoch": 4.564603405041574,
      "grad_norm": 0.2322336882352829,
      "learning_rate": 3.027331048880823e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 15460336,
      "step": 17295
    },
    {
      "epoch": 4.56592318859707,
      "grad_norm": 0.11608325690031052,
      "learning_rate": 3.0263713492291123e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 15464784,
      "step": 17300
    },
    {
      "epoch": 4.567242972152567,
      "grad_norm": 0.315477579832077,
      "learning_rate": 3.0254115684043242e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 15469200,
      "step": 17305
    },
    {
      "epoch": 4.568562755708064,
      "grad_norm": 0.4073043465614319,
      "learning_rate": 3.024451706554469e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 15473552,
      "step": 17310
    },
    {
      "epoch": 4.569882539263561,
      "grad_norm": 0.032944973558187485,
      "learning_rate": 3.0234917638275705e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 15478256,
      "step": 17315
    },
    {
      "epoch": 4.571202322819058,
      "grad_norm": 0.1115405410528183,
      "learning_rate": 3.0225317403716635e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 15482448,
      "step": 17320
    },
    {
      "epoch": 4.572522106374555,
      "grad_norm": 0.2591233551502228,
      "learning_rate": 3.0215716363347956e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 15486896,
      "step": 17325
    },
    {
      "epoch": 4.573841889930051,
      "grad_norm": 0.15288959443569183,
      "learning_rate": 3.0206114518650275e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 15491472,
      "step": 17330
    },
    {
      "epoch": 4.575161673485549,
      "grad_norm": 0.18196724355220795,
      "learning_rate": 3.0196511871104304e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 15495696,
      "step": 17335
    },
    {
      "epoch": 4.576481457041045,
      "grad_norm": 0.3367064893245697,
      "learning_rate": 3.01869084221909e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 15499920,
      "step": 17340
    },
    {
      "epoch": 4.577801240596542,
      "grad_norm": 0.2511852979660034,
      "learning_rate": 3.0177304173391037e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 15504272,
      "step": 17345
    },
    {
      "epoch": 4.579121024152039,
      "grad_norm": 0.374184787273407,
      "learning_rate": 3.01676991261858e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 15508624,
      "step": 17350
    },
    {
      "epoch": 4.580440807707536,
      "grad_norm": 0.20775195956230164,
      "learning_rate": 3.015809328205642e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 15513040,
      "step": 17355
    },
    {
      "epoch": 4.581760591263032,
      "grad_norm": 0.23273865878582,
      "learning_rate": 3.0148486642484248e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 15517360,
      "step": 17360
    },
    {
      "epoch": 4.58308037481853,
      "grad_norm": 0.09257908910512924,
      "learning_rate": 3.0138879208950722e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 15522128,
      "step": 17365
    },
    {
      "epoch": 4.584400158374026,
      "grad_norm": 0.11899735033512115,
      "learning_rate": 3.012927098293744e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 15526608,
      "step": 17370
    },
    {
      "epoch": 4.585719941929524,
      "grad_norm": 0.23780351877212524,
      "learning_rate": 3.0119661965926123e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 15530928,
      "step": 17375
    },
    {
      "epoch": 4.58703972548502,
      "grad_norm": 0.156488835811615,
      "learning_rate": 3.0110052159398587e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 15535792,
      "step": 17380
    },
    {
      "epoch": 4.588359509040517,
      "grad_norm": 0.04205513000488281,
      "learning_rate": 3.0100441564836802e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 15540080,
      "step": 17385
    },
    {
      "epoch": 4.589679292596014,
      "grad_norm": 0.3483344614505768,
      "learning_rate": 3.0090830183722817e-05,
      "loss": 0.0981,
      "num_input_tokens_seen": 15544528,
      "step": 17390
    },
    {
      "epoch": 4.590999076151511,
      "grad_norm": 0.30300524830818176,
      "learning_rate": 3.0081218017538852e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 15549328,
      "step": 17395
    },
    {
      "epoch": 4.592318859707008,
      "grad_norm": 0.22110016644001007,
      "learning_rate": 3.0071605067767212e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 15553776,
      "step": 17400
    },
    {
      "epoch": 4.592318859707008,
      "eval_loss": 0.06884676963090897,
      "eval_runtime": 64.8367,
      "eval_samples_per_second": 103.876,
      "eval_steps_per_second": 25.973,
      "num_input_tokens_seen": 15553776,
      "step": 17400
    },
    {
      "epoch": 4.593638643262505,
      "grad_norm": 0.1262214630842209,
      "learning_rate": 3.006199133589034e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 15558384,
      "step": 17405
    },
    {
      "epoch": 4.5949584268180015,
      "grad_norm": 0.1509024202823639,
      "learning_rate": 3.005237682339079e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 15562736,
      "step": 17410
    },
    {
      "epoch": 4.596278210373499,
      "grad_norm": 0.33995139598846436,
      "learning_rate": 3.0042761531751228e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 15567312,
      "step": 17415
    },
    {
      "epoch": 4.5975979939289955,
      "grad_norm": 0.40181964635849,
      "learning_rate": 3.0033145462454482e-05,
      "loss": 0.1055,
      "num_input_tokens_seen": 15571632,
      "step": 17420
    },
    {
      "epoch": 4.598917777484493,
      "grad_norm": 0.2430763840675354,
      "learning_rate": 3.002352861698345e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 15576048,
      "step": 17425
    },
    {
      "epoch": 4.6002375610399895,
      "grad_norm": 0.2577477693557739,
      "learning_rate": 3.0013910996821178e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 15581104,
      "step": 17430
    },
    {
      "epoch": 4.601557344595486,
      "grad_norm": 0.0804043859243393,
      "learning_rate": 3.0004292603450817e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 15585456,
      "step": 17435
    },
    {
      "epoch": 4.6028771281509835,
      "grad_norm": 0.26149576902389526,
      "learning_rate": 2.9994673438355653e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 15589776,
      "step": 17440
    },
    {
      "epoch": 4.60419691170648,
      "grad_norm": 0.33292320370674133,
      "learning_rate": 2.9985053503019078e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 15594320,
      "step": 17445
    },
    {
      "epoch": 4.6055166952619775,
      "grad_norm": 0.06740719825029373,
      "learning_rate": 2.99754327989246e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 15598576,
      "step": 17450
    },
    {
      "epoch": 4.606836478817474,
      "grad_norm": 0.3843245804309845,
      "learning_rate": 2.9965811327555864e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 15602928,
      "step": 17455
    },
    {
      "epoch": 4.608156262372971,
      "grad_norm": 0.42473652958869934,
      "learning_rate": 2.995618909039662e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 15607504,
      "step": 17460
    },
    {
      "epoch": 4.609476045928468,
      "grad_norm": 0.23682084679603577,
      "learning_rate": 2.9946566088930727e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 15612080,
      "step": 17465
    },
    {
      "epoch": 4.610795829483965,
      "grad_norm": 0.062377359718084335,
      "learning_rate": 2.9936942324642192e-05,
      "loss": 0.0186,
      "num_input_tokens_seen": 15616496,
      "step": 17470
    },
    {
      "epoch": 4.612115613039461,
      "grad_norm": 0.2805708348751068,
      "learning_rate": 2.9927317799015097e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 15620848,
      "step": 17475
    },
    {
      "epoch": 4.613435396594959,
      "grad_norm": 0.3321366608142853,
      "learning_rate": 2.9917692513533685e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 15625168,
      "step": 17480
    },
    {
      "epoch": 4.614755180150455,
      "grad_norm": 0.1830660104751587,
      "learning_rate": 2.990806646968229e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 15629680,
      "step": 17485
    },
    {
      "epoch": 4.616074963705953,
      "grad_norm": 0.4089912176132202,
      "learning_rate": 2.989843966894536e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 15634224,
      "step": 17490
    },
    {
      "epoch": 4.617394747261449,
      "grad_norm": 0.23219004273414612,
      "learning_rate": 2.9888812112807472e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 15638608,
      "step": 17495
    },
    {
      "epoch": 4.618714530816946,
      "grad_norm": 0.17908340692520142,
      "learning_rate": 2.987918380275333e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 15643280,
      "step": 17500
    },
    {
      "epoch": 4.620034314372443,
      "grad_norm": 0.13531553745269775,
      "learning_rate": 2.9869554740267724e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 15647536,
      "step": 17505
    },
    {
      "epoch": 4.62135409792794,
      "grad_norm": 0.08422496914863586,
      "learning_rate": 2.9859924926835585e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 15652080,
      "step": 17510
    },
    {
      "epoch": 4.622673881483436,
      "grad_norm": 0.08091544359922409,
      "learning_rate": 2.9850294363941944e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 15656848,
      "step": 17515
    },
    {
      "epoch": 4.623993665038934,
      "grad_norm": 0.10226783901453018,
      "learning_rate": 2.9840663053071967e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 15661584,
      "step": 17520
    },
    {
      "epoch": 4.62531344859443,
      "grad_norm": 0.24376888573169708,
      "learning_rate": 2.983103099571091e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 15665872,
      "step": 17525
    },
    {
      "epoch": 4.626633232149928,
      "grad_norm": 0.1510409116744995,
      "learning_rate": 2.9821398193344164e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 15670544,
      "step": 17530
    },
    {
      "epoch": 4.627953015705424,
      "grad_norm": 0.10583268851041794,
      "learning_rate": 2.9811764647457226e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 15675088,
      "step": 17535
    },
    {
      "epoch": 4.629272799260921,
      "grad_norm": 0.1281079202890396,
      "learning_rate": 2.9802130359535714e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 15679504,
      "step": 17540
    },
    {
      "epoch": 4.630592582816418,
      "grad_norm": 0.22123976051807404,
      "learning_rate": 2.979249533106535e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 15684112,
      "step": 17545
    },
    {
      "epoch": 4.631912366371915,
      "grad_norm": 0.19087405502796173,
      "learning_rate": 2.9782859563531986e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 15688816,
      "step": 17550
    },
    {
      "epoch": 4.633232149927412,
      "grad_norm": 0.09876422584056854,
      "learning_rate": 2.977322305842156e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 15693264,
      "step": 17555
    },
    {
      "epoch": 4.634551933482909,
      "grad_norm": 0.10813527554273605,
      "learning_rate": 2.9763585817220162e-05,
      "loss": 0.0452,
      "num_input_tokens_seen": 15697392,
      "step": 17560
    },
    {
      "epoch": 4.635871717038405,
      "grad_norm": 0.15994390845298767,
      "learning_rate": 2.975394784141397e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 15701840,
      "step": 17565
    },
    {
      "epoch": 4.637191500593903,
      "grad_norm": 0.048398297280073166,
      "learning_rate": 2.974430913248928e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 15706384,
      "step": 17570
    },
    {
      "epoch": 4.638511284149399,
      "grad_norm": 0.3100918233394623,
      "learning_rate": 2.9734669691932497e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 15710864,
      "step": 17575
    },
    {
      "epoch": 4.639831067704897,
      "grad_norm": 0.4119435250759125,
      "learning_rate": 2.9725029521230147e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 15715472,
      "step": 17580
    },
    {
      "epoch": 4.641150851260393,
      "grad_norm": 0.590843141078949,
      "learning_rate": 2.9715388621868873e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 15720016,
      "step": 17585
    },
    {
      "epoch": 4.64247063481589,
      "grad_norm": 0.34467923641204834,
      "learning_rate": 2.970574699533541e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 15724560,
      "step": 17590
    },
    {
      "epoch": 4.643790418371387,
      "grad_norm": 0.12566322088241577,
      "learning_rate": 2.969610464311662e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 15729008,
      "step": 17595
    },
    {
      "epoch": 4.645110201926884,
      "grad_norm": 0.23434963822364807,
      "learning_rate": 2.9686461566699487e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 15733520,
      "step": 17600
    },
    {
      "epoch": 4.645110201926884,
      "eval_loss": 0.06890065968036652,
      "eval_runtime": 64.7883,
      "eval_samples_per_second": 103.954,
      "eval_steps_per_second": 25.992,
      "num_input_tokens_seen": 15733520,
      "step": 17600
    },
    {
      "epoch": 4.6464299854823805,
      "grad_norm": 0.47985199093818665,
      "learning_rate": 2.9676817767571086e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 15737808,
      "step": 17605
    },
    {
      "epoch": 4.647749769037878,
      "grad_norm": 0.3714390695095062,
      "learning_rate": 2.966717324721861e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 15742320,
      "step": 17610
    },
    {
      "epoch": 4.6490695525933745,
      "grad_norm": 0.05435813218355179,
      "learning_rate": 2.9657528007129366e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 15746928,
      "step": 17615
    },
    {
      "epoch": 4.650389336148872,
      "grad_norm": 0.29634612798690796,
      "learning_rate": 2.9647882048790777e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 15751792,
      "step": 17620
    },
    {
      "epoch": 4.6517091197043685,
      "grad_norm": 0.3147405982017517,
      "learning_rate": 2.963823537369037e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 15756240,
      "step": 17625
    },
    {
      "epoch": 4.653028903259865,
      "grad_norm": 0.11478442698717117,
      "learning_rate": 2.9628587983315775e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 15760656,
      "step": 17630
    },
    {
      "epoch": 4.6543486868153625,
      "grad_norm": 0.13129062950611115,
      "learning_rate": 2.9618939879154746e-05,
      "loss": 0.0189,
      "num_input_tokens_seen": 15765040,
      "step": 17635
    },
    {
      "epoch": 4.655668470370859,
      "grad_norm": 0.3957226276397705,
      "learning_rate": 2.9609291062695143e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 15769264,
      "step": 17640
    },
    {
      "epoch": 4.656988253926356,
      "grad_norm": 0.13761353492736816,
      "learning_rate": 2.9599641535424938e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 15773520,
      "step": 17645
    },
    {
      "epoch": 4.658308037481853,
      "grad_norm": 0.05479968339204788,
      "learning_rate": 2.9589991298832202e-05,
      "loss": 0.0452,
      "num_input_tokens_seen": 15777872,
      "step": 17650
    },
    {
      "epoch": 4.65962782103735,
      "grad_norm": 0.2791734039783478,
      "learning_rate": 2.958034035440513e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 15782608,
      "step": 17655
    },
    {
      "epoch": 4.660947604592847,
      "grad_norm": 0.08973439782857895,
      "learning_rate": 2.957068870363201e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 15786896,
      "step": 17660
    },
    {
      "epoch": 4.662267388148344,
      "grad_norm": 0.05793585255742073,
      "learning_rate": 2.956103634800126e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 15791184,
      "step": 17665
    },
    {
      "epoch": 4.66358717170384,
      "grad_norm": 0.4634318947792053,
      "learning_rate": 2.9551383289001384e-05,
      "loss": 0.1204,
      "num_input_tokens_seen": 15795440,
      "step": 17670
    },
    {
      "epoch": 4.664906955259338,
      "grad_norm": 0.2288123518228531,
      "learning_rate": 2.9541729528121005e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 15799824,
      "step": 17675
    },
    {
      "epoch": 4.666226738814834,
      "grad_norm": 0.17236408591270447,
      "learning_rate": 2.9532075066848856e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 15804240,
      "step": 17680
    },
    {
      "epoch": 4.667546522370332,
      "grad_norm": 0.31086984276771545,
      "learning_rate": 2.9522419906673786e-05,
      "loss": 0.1438,
      "num_input_tokens_seen": 15808720,
      "step": 17685
    },
    {
      "epoch": 4.668866305925828,
      "grad_norm": 0.08635164797306061,
      "learning_rate": 2.951276404908474e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 15813200,
      "step": 17690
    },
    {
      "epoch": 4.670186089481325,
      "grad_norm": 0.10308081656694412,
      "learning_rate": 2.9503107495570752e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 15817648,
      "step": 17695
    },
    {
      "epoch": 4.671505873036822,
      "grad_norm": 0.06466356664896011,
      "learning_rate": 2.9493450247621003e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 15822032,
      "step": 17700
    },
    {
      "epoch": 4.672825656592319,
      "grad_norm": 0.2598475515842438,
      "learning_rate": 2.948379230672476e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 15826544,
      "step": 17705
    },
    {
      "epoch": 4.674145440147816,
      "grad_norm": 0.07291156053543091,
      "learning_rate": 2.9474133674371396e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 15831280,
      "step": 17710
    },
    {
      "epoch": 4.675465223703313,
      "grad_norm": 0.27887532114982605,
      "learning_rate": 2.9464474352050387e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 15835824,
      "step": 17715
    },
    {
      "epoch": 4.676785007258809,
      "grad_norm": 0.0346548855304718,
      "learning_rate": 2.9454814341251336e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 15840496,
      "step": 17720
    },
    {
      "epoch": 4.678104790814307,
      "grad_norm": 0.42805132269859314,
      "learning_rate": 2.9445153643463942e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 15845104,
      "step": 17725
    },
    {
      "epoch": 4.679424574369803,
      "grad_norm": 0.1374506801366806,
      "learning_rate": 2.943549226017798e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 15849456,
      "step": 17730
    },
    {
      "epoch": 4.680744357925301,
      "grad_norm": 0.027222875505685806,
      "learning_rate": 2.942583019288337e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 15853744,
      "step": 17735
    },
    {
      "epoch": 4.682064141480797,
      "grad_norm": 0.19105306267738342,
      "learning_rate": 2.9416167443070132e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 15858128,
      "step": 17740
    },
    {
      "epoch": 4.683383925036294,
      "grad_norm": 0.0392376072704792,
      "learning_rate": 2.9406504012228375e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 15862608,
      "step": 17745
    },
    {
      "epoch": 4.684703708591791,
      "grad_norm": 0.40456998348236084,
      "learning_rate": 2.939683990184832e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 15866928,
      "step": 17750
    },
    {
      "epoch": 4.686023492147288,
      "grad_norm": 0.1575268805027008,
      "learning_rate": 2.93871751134203e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 15871184,
      "step": 17755
    },
    {
      "epoch": 4.687343275702784,
      "grad_norm": 0.14288605749607086,
      "learning_rate": 2.9377509648434752e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 15875408,
      "step": 17760
    },
    {
      "epoch": 4.688663059258282,
      "grad_norm": 0.08281012624502182,
      "learning_rate": 2.9367843508382203e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 15879728,
      "step": 17765
    },
    {
      "epoch": 4.689982842813778,
      "grad_norm": 0.024431154131889343,
      "learning_rate": 2.9358176694753293e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 15884528,
      "step": 17770
    },
    {
      "epoch": 4.691302626369275,
      "grad_norm": 0.23028463125228882,
      "learning_rate": 2.9348509209038766e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 15888944,
      "step": 17775
    },
    {
      "epoch": 4.692622409924772,
      "grad_norm": 0.08968193829059601,
      "learning_rate": 2.933884105272947e-05,
      "loss": 0.0133,
      "num_input_tokens_seen": 15893840,
      "step": 17780
    },
    {
      "epoch": 4.693942193480269,
      "grad_norm": 0.4847165644168854,
      "learning_rate": 2.9329172227316366e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 15898224,
      "step": 17785
    },
    {
      "epoch": 4.695261977035766,
      "grad_norm": 0.3025650680065155,
      "learning_rate": 2.93195027342905e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 15902832,
      "step": 17790
    },
    {
      "epoch": 4.696581760591263,
      "grad_norm": 0.22033993899822235,
      "learning_rate": 2.9309832575143024e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 15907280,
      "step": 17795
    },
    {
      "epoch": 4.6979015441467595,
      "grad_norm": 0.28822341561317444,
      "learning_rate": 2.930016175136521e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 15911728,
      "step": 17800
    },
    {
      "epoch": 4.6979015441467595,
      "eval_loss": 0.06856284290552139,
      "eval_runtime": 64.781,
      "eval_samples_per_second": 103.966,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 15911728,
      "step": 17800
    },
    {
      "epoch": 4.699221327702257,
      "grad_norm": 0.30911681056022644,
      "learning_rate": 2.9290490264448412e-05,
      "loss": 0.1562,
      "num_input_tokens_seen": 15916208,
      "step": 17805
    },
    {
      "epoch": 4.7005411112577535,
      "grad_norm": 0.1361420899629593,
      "learning_rate": 2.9280818115884094e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 15920496,
      "step": 17810
    },
    {
      "epoch": 4.701860894813251,
      "grad_norm": 0.04205937311053276,
      "learning_rate": 2.9271145307163828e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 15924880,
      "step": 17815
    },
    {
      "epoch": 4.7031806783687475,
      "grad_norm": 0.0512765608727932,
      "learning_rate": 2.9261471839779287e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 15929744,
      "step": 17820
    },
    {
      "epoch": 4.704500461924244,
      "grad_norm": 0.18293075263500214,
      "learning_rate": 2.925179771522223e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 15934352,
      "step": 17825
    },
    {
      "epoch": 4.7058202454797415,
      "grad_norm": 0.23262397944927216,
      "learning_rate": 2.9242122934984535e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 15938640,
      "step": 17830
    },
    {
      "epoch": 4.707140029035238,
      "grad_norm": 0.2571221590042114,
      "learning_rate": 2.9232447500558176e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 15943152,
      "step": 17835
    },
    {
      "epoch": 4.7084598125907355,
      "grad_norm": 0.05716380104422569,
      "learning_rate": 2.9222771413435225e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 15947408,
      "step": 17840
    },
    {
      "epoch": 4.709779596146232,
      "grad_norm": 0.09259232133626938,
      "learning_rate": 2.9213094675107848e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 15951664,
      "step": 17845
    },
    {
      "epoch": 4.711099379701729,
      "grad_norm": 0.03979388624429703,
      "learning_rate": 2.9203417287068335e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 15955824,
      "step": 17850
    },
    {
      "epoch": 4.712419163257226,
      "grad_norm": 0.061700619757175446,
      "learning_rate": 2.9193739250809042e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 15959856,
      "step": 17855
    },
    {
      "epoch": 4.713738946812723,
      "grad_norm": 0.16454225778579712,
      "learning_rate": 2.9184060567822463e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 15964560,
      "step": 17860
    },
    {
      "epoch": 4.71505873036822,
      "grad_norm": 0.25724831223487854,
      "learning_rate": 2.9174381239601166e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 15969040,
      "step": 17865
    },
    {
      "epoch": 4.716378513923717,
      "grad_norm": 0.24976512789726257,
      "learning_rate": 2.916470126763783e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 15973328,
      "step": 17870
    },
    {
      "epoch": 4.717698297479213,
      "grad_norm": 0.26603972911834717,
      "learning_rate": 2.9155020653425203e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 15978000,
      "step": 17875
    },
    {
      "epoch": 4.719018081034711,
      "grad_norm": 0.23804716765880585,
      "learning_rate": 2.9145339398456184e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 15982320,
      "step": 17880
    },
    {
      "epoch": 4.720337864590207,
      "grad_norm": 0.08322809636592865,
      "learning_rate": 2.913565750422374e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 15987216,
      "step": 17885
    },
    {
      "epoch": 4.721657648145704,
      "grad_norm": 0.21547013521194458,
      "learning_rate": 2.9125974972220938e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 15991600,
      "step": 17890
    },
    {
      "epoch": 4.722977431701201,
      "grad_norm": 0.2432098239660263,
      "learning_rate": 2.9116291803940932e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 15995984,
      "step": 17895
    },
    {
      "epoch": 4.724297215256698,
      "grad_norm": 0.03919065743684769,
      "learning_rate": 2.910660800087701e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 16000688,
      "step": 17900
    },
    {
      "epoch": 4.725616998812194,
      "grad_norm": 0.11776836961507797,
      "learning_rate": 2.909692356452254e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 16005136,
      "step": 17905
    },
    {
      "epoch": 4.726936782367692,
      "grad_norm": 0.18365944921970367,
      "learning_rate": 2.9087238496370962e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 16009392,
      "step": 17910
    },
    {
      "epoch": 4.728256565923188,
      "grad_norm": 0.5774285197257996,
      "learning_rate": 2.907755279791583e-05,
      "loss": 0.0994,
      "num_input_tokens_seen": 16013648,
      "step": 17915
    },
    {
      "epoch": 4.729576349478686,
      "grad_norm": 0.09283986687660217,
      "learning_rate": 2.906786647065083e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 16018192,
      "step": 17920
    },
    {
      "epoch": 4.730896133034182,
      "grad_norm": 0.01985105499625206,
      "learning_rate": 2.9058179516069695e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 16023088,
      "step": 17925
    },
    {
      "epoch": 4.732215916589679,
      "grad_norm": 0.6372429728507996,
      "learning_rate": 2.9048491935666282e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 16027536,
      "step": 17930
    },
    {
      "epoch": 4.733535700145176,
      "grad_norm": 0.26704344153404236,
      "learning_rate": 2.9038803730934534e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 16032208,
      "step": 17935
    },
    {
      "epoch": 4.734855483700673,
      "grad_norm": 0.20905055105686188,
      "learning_rate": 2.9029114903368503e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 16036976,
      "step": 17940
    },
    {
      "epoch": 4.73617526725617,
      "grad_norm": 0.27750009298324585,
      "learning_rate": 2.9019425454462318e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 16041392,
      "step": 17945
    },
    {
      "epoch": 4.737495050811667,
      "grad_norm": 0.18839259445667267,
      "learning_rate": 2.9009735385710212e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 16046032,
      "step": 17950
    },
    {
      "epoch": 4.738814834367163,
      "grad_norm": 0.26710817217826843,
      "learning_rate": 2.900004469860652e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 16050672,
      "step": 17955
    },
    {
      "epoch": 4.740134617922661,
      "grad_norm": 0.20179186761379242,
      "learning_rate": 2.8990353394645668e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 16055216,
      "step": 17960
    },
    {
      "epoch": 4.741454401478157,
      "grad_norm": 0.2130323350429535,
      "learning_rate": 2.8980661475322186e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 16059664,
      "step": 17965
    },
    {
      "epoch": 4.742774185033655,
      "grad_norm": 0.15826711058616638,
      "learning_rate": 2.897096894213067e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 16064464,
      "step": 17970
    },
    {
      "epoch": 4.7440939685891514,
      "grad_norm": 0.34615734219551086,
      "learning_rate": 2.8961275796565845e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 16069072,
      "step": 17975
    },
    {
      "epoch": 4.745413752144648,
      "grad_norm": 0.03634057939052582,
      "learning_rate": 2.8951582040122517e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 16073744,
      "step": 17980
    },
    {
      "epoch": 4.7467335357001454,
      "grad_norm": 0.2439921349287033,
      "learning_rate": 2.894188767429557e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 16078384,
      "step": 17985
    },
    {
      "epoch": 4.748053319255642,
      "grad_norm": 0.030153362080454826,
      "learning_rate": 2.8932192700580014e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 16082832,
      "step": 17990
    },
    {
      "epoch": 4.7493731028111394,
      "grad_norm": 0.1239658072590828,
      "learning_rate": 2.8922497120470916e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 16087280,
      "step": 17995
    },
    {
      "epoch": 4.750692886366636,
      "grad_norm": 0.13337016105651855,
      "learning_rate": 2.891280093546348e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 16091728,
      "step": 18000
    },
    {
      "epoch": 4.750692886366636,
      "eval_loss": 0.06902699917554855,
      "eval_runtime": 64.7828,
      "eval_samples_per_second": 103.963,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 16091728,
      "step": 18000
    },
    {
      "epoch": 4.752012669922133,
      "grad_norm": 0.08740201592445374,
      "learning_rate": 2.890310414705297e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 16096112,
      "step": 18005
    },
    {
      "epoch": 4.75333245347763,
      "grad_norm": 0.10230229794979095,
      "learning_rate": 2.8893406756734742e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 16100272,
      "step": 18010
    },
    {
      "epoch": 4.754652237033127,
      "grad_norm": 0.23941786587238312,
      "learning_rate": 2.888370876600427e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 16104592,
      "step": 18015
    },
    {
      "epoch": 4.755972020588623,
      "grad_norm": 0.09854938089847565,
      "learning_rate": 2.8874010176357104e-05,
      "loss": 0.0967,
      "num_input_tokens_seen": 16108688,
      "step": 18020
    },
    {
      "epoch": 4.757291804144121,
      "grad_norm": 0.12685872614383698,
      "learning_rate": 2.886431098928888e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 16113584,
      "step": 18025
    },
    {
      "epoch": 4.758611587699617,
      "grad_norm": 0.16588932275772095,
      "learning_rate": 2.885461120629534e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 16118000,
      "step": 18030
    },
    {
      "epoch": 4.759931371255114,
      "grad_norm": 0.13890865445137024,
      "learning_rate": 2.8844910828872317e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 16122608,
      "step": 18035
    },
    {
      "epoch": 4.761251154810611,
      "grad_norm": 0.042833294719457626,
      "learning_rate": 2.8835209858515715e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 16127088,
      "step": 18040
    },
    {
      "epoch": 4.762570938366108,
      "grad_norm": 0.04848520830273628,
      "learning_rate": 2.8825508296721566e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 16131440,
      "step": 18045
    },
    {
      "epoch": 4.763890721921605,
      "grad_norm": 0.07415361702442169,
      "learning_rate": 2.881580614498596e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 16135824,
      "step": 18050
    },
    {
      "epoch": 4.765210505477102,
      "grad_norm": 0.3715346157550812,
      "learning_rate": 2.8806103404805103e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 16140240,
      "step": 18055
    },
    {
      "epoch": 4.766530289032598,
      "grad_norm": 0.15694566071033478,
      "learning_rate": 2.8796400077675257e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 16144336,
      "step": 18060
    },
    {
      "epoch": 4.767850072588096,
      "grad_norm": 0.1628815233707428,
      "learning_rate": 2.8786696165092812e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 16149072,
      "step": 18065
    },
    {
      "epoch": 4.769169856143592,
      "grad_norm": 0.46819257736206055,
      "learning_rate": 2.8776991668554236e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 16153616,
      "step": 18070
    },
    {
      "epoch": 4.77048963969909,
      "grad_norm": 0.2627697288990021,
      "learning_rate": 2.876728658955608e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 16157744,
      "step": 18075
    },
    {
      "epoch": 4.771809423254586,
      "grad_norm": 0.25993549823760986,
      "learning_rate": 2.8757580929594986e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 16162320,
      "step": 18080
    },
    {
      "epoch": 4.773129206810083,
      "grad_norm": 0.1118636205792427,
      "learning_rate": 2.87478746901677e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 16166480,
      "step": 18085
    },
    {
      "epoch": 4.77444899036558,
      "grad_norm": 0.16295285522937775,
      "learning_rate": 2.873816787277103e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 16170544,
      "step": 18090
    },
    {
      "epoch": 4.775768773921077,
      "grad_norm": 0.3596700131893158,
      "learning_rate": 2.8728460478901903e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 16174672,
      "step": 18095
    },
    {
      "epoch": 4.777088557476574,
      "grad_norm": 0.32202497124671936,
      "learning_rate": 2.8718752510057307e-05,
      "loss": 0.0949,
      "num_input_tokens_seen": 16179216,
      "step": 18100
    },
    {
      "epoch": 4.778408341032071,
      "grad_norm": 0.1442602425813675,
      "learning_rate": 2.870904396773435e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 16183632,
      "step": 18105
    },
    {
      "epoch": 4.779728124587567,
      "grad_norm": 0.2890265882015228,
      "learning_rate": 2.86993348534302e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 16188304,
      "step": 18110
    },
    {
      "epoch": 4.781047908143065,
      "grad_norm": 0.034286852926015854,
      "learning_rate": 2.868962516864212e-05,
      "loss": 0.0135,
      "num_input_tokens_seen": 16192848,
      "step": 18115
    },
    {
      "epoch": 4.782367691698561,
      "grad_norm": 0.21903228759765625,
      "learning_rate": 2.8679914914867477e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 16197264,
      "step": 18120
    },
    {
      "epoch": 4.783687475254059,
      "grad_norm": 0.1597234606742859,
      "learning_rate": 2.8670204093603713e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 16201968,
      "step": 18125
    },
    {
      "epoch": 4.785007258809555,
      "grad_norm": 0.44640466570854187,
      "learning_rate": 2.8660492706348357e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 16206320,
      "step": 18130
    },
    {
      "epoch": 4.786327042365052,
      "grad_norm": 0.5237805843353271,
      "learning_rate": 2.8650780754599022e-05,
      "loss": 0.0775,
      "num_input_tokens_seen": 16210736,
      "step": 18135
    },
    {
      "epoch": 4.787646825920549,
      "grad_norm": 0.28884586691856384,
      "learning_rate": 2.8641068239853407e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 16215280,
      "step": 18140
    },
    {
      "epoch": 4.788966609476046,
      "grad_norm": 0.2676484286785126,
      "learning_rate": 2.863135516360932e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 16219792,
      "step": 18145
    },
    {
      "epoch": 4.7902863930315425,
      "grad_norm": 0.21679319441318512,
      "learning_rate": 2.8621641527364633e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 16224304,
      "step": 18150
    },
    {
      "epoch": 4.79160617658704,
      "grad_norm": 0.08065123111009598,
      "learning_rate": 2.8611927332617313e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 16228720,
      "step": 18155
    },
    {
      "epoch": 4.7929259601425365,
      "grad_norm": 0.03950365632772446,
      "learning_rate": 2.8602212580865405e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 16233200,
      "step": 18160
    },
    {
      "epoch": 4.794245743698033,
      "grad_norm": 0.10011640191078186,
      "learning_rate": 2.859249727360705e-05,
      "loss": 0.0282,
      "num_input_tokens_seen": 16237648,
      "step": 18165
    },
    {
      "epoch": 4.7955655272535305,
      "grad_norm": 0.37731754779815674,
      "learning_rate": 2.8582781412340465e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 16242064,
      "step": 18170
    },
    {
      "epoch": 4.796885310809027,
      "grad_norm": 0.28352510929107666,
      "learning_rate": 2.857306499856397e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 16246256,
      "step": 18175
    },
    {
      "epoch": 4.7982050943645245,
      "grad_norm": 0.2848483622074127,
      "learning_rate": 2.856334803377594e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 16250544,
      "step": 18180
    },
    {
      "epoch": 4.799524877920021,
      "grad_norm": 0.2371665984392166,
      "learning_rate": 2.8553630519474867e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 16254864,
      "step": 18185
    },
    {
      "epoch": 4.800844661475518,
      "grad_norm": 0.06103622540831566,
      "learning_rate": 2.8543912457159317e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 16259376,
      "step": 18190
    },
    {
      "epoch": 4.802164445031015,
      "grad_norm": 0.3026491701602936,
      "learning_rate": 2.853419384832792e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 16263792,
      "step": 18195
    },
    {
      "epoch": 4.803484228586512,
      "grad_norm": 0.7308284044265747,
      "learning_rate": 2.8524474694479423e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 16268208,
      "step": 18200
    },
    {
      "epoch": 4.803484228586512,
      "eval_loss": 0.06847847253084183,
      "eval_runtime": 64.7227,
      "eval_samples_per_second": 104.059,
      "eval_steps_per_second": 26.019,
      "num_input_tokens_seen": 16268208,
      "step": 18200
    },
    {
      "epoch": 4.804804012142009,
      "grad_norm": 0.23003126680850983,
      "learning_rate": 2.851475499711264e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 16272624,
      "step": 18205
    },
    {
      "epoch": 4.806123795697506,
      "grad_norm": 0.19622860848903656,
      "learning_rate": 2.8505034757726468e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 16277104,
      "step": 18210
    },
    {
      "epoch": 4.807443579253002,
      "grad_norm": 0.15494847297668457,
      "learning_rate": 2.8495313977819886e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 16281424,
      "step": 18215
    },
    {
      "epoch": 4.8087633628085,
      "grad_norm": 0.29320693016052246,
      "learning_rate": 2.8485592658891956e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 16285968,
      "step": 18220
    },
    {
      "epoch": 4.810083146363996,
      "grad_norm": 0.2857033312320709,
      "learning_rate": 2.8475870802441844e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 16290512,
      "step": 18225
    },
    {
      "epoch": 4.811402929919494,
      "grad_norm": 0.2897636890411377,
      "learning_rate": 2.8466148409968774e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 16294928,
      "step": 18230
    },
    {
      "epoch": 4.81272271347499,
      "grad_norm": 0.27795130014419556,
      "learning_rate": 2.8456425482972067e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 16299088,
      "step": 18235
    },
    {
      "epoch": 4.814042497030487,
      "grad_norm": 0.44196945428848267,
      "learning_rate": 2.84467020229511e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 16303920,
      "step": 18240
    },
    {
      "epoch": 4.815362280585984,
      "grad_norm": 0.14868836104869843,
      "learning_rate": 2.8436978031405375e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 16308432,
      "step": 18245
    },
    {
      "epoch": 4.816682064141481,
      "grad_norm": 0.2528722584247589,
      "learning_rate": 2.842725350983445e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 16313168,
      "step": 18250
    },
    {
      "epoch": 4.818001847696978,
      "grad_norm": 0.2338588982820511,
      "learning_rate": 2.8417528459737957e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 16317552,
      "step": 18255
    },
    {
      "epoch": 4.819321631252475,
      "grad_norm": 0.37647566199302673,
      "learning_rate": 2.8407802882615624e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 16321904,
      "step": 18260
    },
    {
      "epoch": 4.820641414807971,
      "grad_norm": 0.20688609778881073,
      "learning_rate": 2.8398076779967277e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 16326320,
      "step": 18265
    },
    {
      "epoch": 4.821961198363469,
      "grad_norm": 0.19370673596858978,
      "learning_rate": 2.8388350153292774e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 16331024,
      "step": 18270
    },
    {
      "epoch": 4.823280981918965,
      "grad_norm": 0.08151532709598541,
      "learning_rate": 2.8378623004092103e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 16335280,
      "step": 18275
    },
    {
      "epoch": 4.824600765474462,
      "grad_norm": 0.15876011550426483,
      "learning_rate": 2.8368895333865302e-05,
      "loss": 0.0271,
      "num_input_tokens_seen": 16339472,
      "step": 18280
    },
    {
      "epoch": 4.825920549029959,
      "grad_norm": 0.4037726819515228,
      "learning_rate": 2.835916714411251e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 16344176,
      "step": 18285
    },
    {
      "epoch": 4.827240332585456,
      "grad_norm": 0.1780112236738205,
      "learning_rate": 2.8349438436333926e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 16348656,
      "step": 18290
    },
    {
      "epoch": 4.828560116140953,
      "grad_norm": 0.5652247071266174,
      "learning_rate": 2.833970921202984e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 16352912,
      "step": 18295
    },
    {
      "epoch": 4.82987989969645,
      "grad_norm": 0.19574569165706635,
      "learning_rate": 2.8329979472700628e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 16357136,
      "step": 18300
    },
    {
      "epoch": 4.831199683251946,
      "grad_norm": 0.2805270850658417,
      "learning_rate": 2.832024921984674e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 16361648,
      "step": 18305
    },
    {
      "epoch": 4.832519466807444,
      "grad_norm": 0.20967717468738556,
      "learning_rate": 2.8310518454968693e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 16366352,
      "step": 18310
    },
    {
      "epoch": 4.83383925036294,
      "grad_norm": 0.13764281570911407,
      "learning_rate": 2.8300787179567095e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 16370640,
      "step": 18315
    },
    {
      "epoch": 4.835159033918437,
      "grad_norm": 0.4281664490699768,
      "learning_rate": 2.8291055395142636e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 16375280,
      "step": 18320
    },
    {
      "epoch": 4.836478817473934,
      "grad_norm": 0.30913692712783813,
      "learning_rate": 2.8281323103196073e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 16380048,
      "step": 18325
    },
    {
      "epoch": 4.837798601029431,
      "grad_norm": 0.43322595953941345,
      "learning_rate": 2.8271590305228256e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 16384560,
      "step": 18330
    },
    {
      "epoch": 4.839118384584928,
      "grad_norm": 0.4322415888309479,
      "learning_rate": 2.82618570027401e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 16388816,
      "step": 18335
    },
    {
      "epoch": 4.840438168140425,
      "grad_norm": 0.20907701551914215,
      "learning_rate": 2.8252123197232604e-05,
      "loss": 0.0249,
      "num_input_tokens_seen": 16393424,
      "step": 18340
    },
    {
      "epoch": 4.8417579516959215,
      "grad_norm": 0.40784138441085815,
      "learning_rate": 2.8242388890206843e-05,
      "loss": 0.1337,
      "num_input_tokens_seen": 16397872,
      "step": 18345
    },
    {
      "epoch": 4.843077735251419,
      "grad_norm": 0.3420793116092682,
      "learning_rate": 2.8232654083163967e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 16402224,
      "step": 18350
    },
    {
      "epoch": 4.8443975188069155,
      "grad_norm": 0.5291674137115479,
      "learning_rate": 2.822291877760521e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 16406672,
      "step": 18355
    },
    {
      "epoch": 4.845717302362413,
      "grad_norm": 0.2153484970331192,
      "learning_rate": 2.8213182975031864e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 16411024,
      "step": 18360
    },
    {
      "epoch": 4.8470370859179095,
      "grad_norm": 0.19715672731399536,
      "learning_rate": 2.8203446676945337e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 16415600,
      "step": 18365
    },
    {
      "epoch": 4.848356869473406,
      "grad_norm": 0.5209136605262756,
      "learning_rate": 2.8193709884847075e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 16420048,
      "step": 18370
    },
    {
      "epoch": 4.8496766530289035,
      "grad_norm": 0.23625631630420685,
      "learning_rate": 2.8183972600238605e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 16424656,
      "step": 18375
    },
    {
      "epoch": 4.8509964365844,
      "grad_norm": 0.34111475944519043,
      "learning_rate": 2.817423482462156e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 16428880,
      "step": 18380
    },
    {
      "epoch": 4.8523162201398975,
      "grad_norm": 0.16406036913394928,
      "learning_rate": 2.8164496559497605e-05,
      "loss": 0.0745,
      "num_input_tokens_seen": 16433360,
      "step": 18385
    },
    {
      "epoch": 4.853636003695394,
      "grad_norm": 0.11744712293148041,
      "learning_rate": 2.815475780636852e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 16437680,
      "step": 18390
    },
    {
      "epoch": 4.854955787250891,
      "grad_norm": 0.3759215176105499,
      "learning_rate": 2.814501856673613e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 16442224,
      "step": 18395
    },
    {
      "epoch": 4.856275570806388,
      "grad_norm": 0.22512462735176086,
      "learning_rate": 2.8135278842102353e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 16446704,
      "step": 18400
    },
    {
      "epoch": 4.856275570806388,
      "eval_loss": 0.06849317997694016,
      "eval_runtime": 64.7706,
      "eval_samples_per_second": 103.982,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 16446704,
      "step": 18400
    },
    {
      "epoch": 4.857595354361885,
      "grad_norm": 0.0899285227060318,
      "learning_rate": 2.8125538633969183e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 16451120,
      "step": 18405
    },
    {
      "epoch": 4.858915137917381,
      "grad_norm": 0.30360764265060425,
      "learning_rate": 2.8115797943838677e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 16455792,
      "step": 18410
    },
    {
      "epoch": 4.860234921472879,
      "grad_norm": 0.3270409107208252,
      "learning_rate": 2.810605677321298e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 16460080,
      "step": 18415
    },
    {
      "epoch": 4.861554705028375,
      "grad_norm": 0.11539432406425476,
      "learning_rate": 2.809631512359428e-05,
      "loss": 0.0261,
      "num_input_tokens_seen": 16464432,
      "step": 18420
    },
    {
      "epoch": 4.862874488583873,
      "grad_norm": 0.1628473997116089,
      "learning_rate": 2.8086572996484884e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 16468848,
      "step": 18425
    },
    {
      "epoch": 4.864194272139369,
      "grad_norm": 0.579860270023346,
      "learning_rate": 2.8076830393387143e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 16473232,
      "step": 18430
    },
    {
      "epoch": 4.865514055694866,
      "grad_norm": 0.2550118565559387,
      "learning_rate": 2.8067087315803497e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 16477840,
      "step": 18435
    },
    {
      "epoch": 4.866833839250363,
      "grad_norm": 0.11569822579622269,
      "learning_rate": 2.8057343765236433e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 16482256,
      "step": 18440
    },
    {
      "epoch": 4.86815362280586,
      "grad_norm": 0.11786375939846039,
      "learning_rate": 2.804759974318854e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 16486704,
      "step": 18445
    },
    {
      "epoch": 4.869473406361356,
      "grad_norm": 0.2913278341293335,
      "learning_rate": 2.8037855251162482e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 16491184,
      "step": 18450
    },
    {
      "epoch": 4.870793189916854,
      "grad_norm": 0.16652972996234894,
      "learning_rate": 2.802811029066096e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 16495888,
      "step": 18455
    },
    {
      "epoch": 4.87211297347235,
      "grad_norm": 0.10203798115253448,
      "learning_rate": 2.8018364863186764e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 16500368,
      "step": 18460
    },
    {
      "epoch": 4.873432757027848,
      "grad_norm": 0.1026599109172821,
      "learning_rate": 2.800861897024279e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 16504912,
      "step": 18465
    },
    {
      "epoch": 4.874752540583344,
      "grad_norm": 0.01670807972550392,
      "learning_rate": 2.799887261333196e-05,
      "loss": 0.0441,
      "num_input_tokens_seen": 16509584,
      "step": 18470
    },
    {
      "epoch": 4.876072324138841,
      "grad_norm": 0.15002210438251495,
      "learning_rate": 2.798912579395728e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 16514064,
      "step": 18475
    },
    {
      "epoch": 4.877392107694338,
      "grad_norm": 0.14212659001350403,
      "learning_rate": 2.797937851362185e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 16518512,
      "step": 18480
    },
    {
      "epoch": 4.878711891249835,
      "grad_norm": 0.15426969528198242,
      "learning_rate": 2.7969630773828802e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 16523056,
      "step": 18485
    },
    {
      "epoch": 4.880031674805332,
      "grad_norm": 0.18571124970912933,
      "learning_rate": 2.7959882576081382e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 16527632,
      "step": 18490
    },
    {
      "epoch": 4.881351458360829,
      "grad_norm": 0.553921103477478,
      "learning_rate": 2.795013392188286e-05,
      "loss": 0.104,
      "num_input_tokens_seen": 16531920,
      "step": 18495
    },
    {
      "epoch": 4.882671241916325,
      "grad_norm": 0.28923413157463074,
      "learning_rate": 2.7940384812736614e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 16536656,
      "step": 18500
    },
    {
      "epoch": 4.883991025471823,
      "grad_norm": 0.11596778780221939,
      "learning_rate": 2.7930635250146087e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 16540944,
      "step": 18505
    },
    {
      "epoch": 4.885310809027319,
      "grad_norm": 0.0761212557554245,
      "learning_rate": 2.792088523561477e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 16545616,
      "step": 18510
    },
    {
      "epoch": 4.886630592582817,
      "grad_norm": 0.09057457745075226,
      "learning_rate": 2.7911134770646246e-05,
      "loss": 0.0131,
      "num_input_tokens_seen": 16550160,
      "step": 18515
    },
    {
      "epoch": 4.887950376138313,
      "grad_norm": 0.15105414390563965,
      "learning_rate": 2.7901383856744157e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 16554768,
      "step": 18520
    },
    {
      "epoch": 4.88927015969381,
      "grad_norm": 0.2015790492296219,
      "learning_rate": 2.7891632495412217e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 16559248,
      "step": 18525
    },
    {
      "epoch": 4.890589943249307,
      "grad_norm": 0.10020840167999268,
      "learning_rate": 2.7881880688154205e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 16563568,
      "step": 18530
    },
    {
      "epoch": 4.891909726804804,
      "grad_norm": 0.017234444618225098,
      "learning_rate": 2.7872128436473977e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 16568016,
      "step": 18535
    },
    {
      "epoch": 4.893229510360301,
      "grad_norm": 0.3275602459907532,
      "learning_rate": 2.7862375741875448e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 16572688,
      "step": 18540
    },
    {
      "epoch": 4.894549293915798,
      "grad_norm": 0.217233806848526,
      "learning_rate": 2.785262260586261e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 16577296,
      "step": 18545
    },
    {
      "epoch": 4.8958690774712945,
      "grad_norm": 0.24206528067588806,
      "learning_rate": 2.7842869029939517e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 16581872,
      "step": 18550
    },
    {
      "epoch": 4.897188861026792,
      "grad_norm": 0.16623778641223907,
      "learning_rate": 2.7833115015610296e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 16586544,
      "step": 18555
    },
    {
      "epoch": 4.8985086445822885,
      "grad_norm": 0.410235196352005,
      "learning_rate": 2.7823360564379136e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 16590960,
      "step": 18560
    },
    {
      "epoch": 4.899828428137785,
      "grad_norm": 0.1687140166759491,
      "learning_rate": 2.7813605677750297e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 16595440,
      "step": 18565
    },
    {
      "epoch": 4.9011482116932825,
      "grad_norm": 0.1012502908706665,
      "learning_rate": 2.7803850357228102e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 16600336,
      "step": 18570
    },
    {
      "epoch": 4.902467995248779,
      "grad_norm": 0.16951428353786469,
      "learning_rate": 2.779409460431695e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 16604752,
      "step": 18575
    },
    {
      "epoch": 4.903787778804276,
      "grad_norm": 0.2252022624015808,
      "learning_rate": 2.778433842052129e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 16609360,
      "step": 18580
    },
    {
      "epoch": 4.905107562359773,
      "grad_norm": 0.15820682048797607,
      "learning_rate": 2.7774581807345664e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 16614000,
      "step": 18585
    },
    {
      "epoch": 4.90642734591527,
      "grad_norm": 0.050562042742967606,
      "learning_rate": 2.776482476629465e-05,
      "loss": 0.0241,
      "num_input_tokens_seen": 16618032,
      "step": 18590
    },
    {
      "epoch": 4.907747129470767,
      "grad_norm": 0.4059712886810303,
      "learning_rate": 2.7755067298872924e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 16622576,
      "step": 18595
    },
    {
      "epoch": 4.909066913026264,
      "grad_norm": 0.14000160992145538,
      "learning_rate": 2.774530940658518e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 16627152,
      "step": 18600
    },
    {
      "epoch": 4.909066913026264,
      "eval_loss": 0.06863165646791458,
      "eval_runtime": 64.7735,
      "eval_samples_per_second": 103.978,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 16627152,
      "step": 18600
    },
    {
      "epoch": 4.91038669658176,
      "grad_norm": 0.5590384602546692,
      "learning_rate": 2.7735551090936236e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 16631568,
      "step": 18605
    },
    {
      "epoch": 4.911706480137258,
      "grad_norm": 0.15937602519989014,
      "learning_rate": 2.7725792353430934e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 16635888,
      "step": 18610
    },
    {
      "epoch": 4.913026263692754,
      "grad_norm": 0.20967191457748413,
      "learning_rate": 2.77160331955742e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 16640560,
      "step": 18615
    },
    {
      "epoch": 4.914346047248252,
      "grad_norm": 0.32316821813583374,
      "learning_rate": 2.7706273618871008e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 16644848,
      "step": 18620
    },
    {
      "epoch": 4.915665830803748,
      "grad_norm": 0.11100181192159653,
      "learning_rate": 2.769651362482642e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 16649424,
      "step": 18625
    },
    {
      "epoch": 4.916985614359245,
      "grad_norm": 0.04020622745156288,
      "learning_rate": 2.768675321494555e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 16653808,
      "step": 18630
    },
    {
      "epoch": 4.918305397914742,
      "grad_norm": 0.2699630558490753,
      "learning_rate": 2.7676992390733565e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 16658480,
      "step": 18635
    },
    {
      "epoch": 4.919625181470239,
      "grad_norm": 0.23626990616321564,
      "learning_rate": 2.766723115369571e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 16662928,
      "step": 18640
    },
    {
      "epoch": 4.920944965025736,
      "grad_norm": 0.07744468748569489,
      "learning_rate": 2.765746950533729e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 16667440,
      "step": 18645
    },
    {
      "epoch": 4.922264748581233,
      "grad_norm": 0.15091878175735474,
      "learning_rate": 2.7647707447163684e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 16672048,
      "step": 18650
    },
    {
      "epoch": 4.923584532136729,
      "grad_norm": 0.04735223576426506,
      "learning_rate": 2.7637944980680315e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 16676336,
      "step": 18655
    },
    {
      "epoch": 4.924904315692227,
      "grad_norm": 0.18337567150592804,
      "learning_rate": 2.762818210739268e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 16680816,
      "step": 18660
    },
    {
      "epoch": 4.926224099247723,
      "grad_norm": 0.134157195687294,
      "learning_rate": 2.7618418828806332e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 16685200,
      "step": 18665
    },
    {
      "epoch": 4.927543882803221,
      "grad_norm": 0.20517902076244354,
      "learning_rate": 2.76086551464269e-05,
      "loss": 0.0211,
      "num_input_tokens_seen": 16689360,
      "step": 18670
    },
    {
      "epoch": 4.928863666358717,
      "grad_norm": 0.13440768420696259,
      "learning_rate": 2.759889106176006e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 16693776,
      "step": 18675
    },
    {
      "epoch": 4.930183449914214,
      "grad_norm": 0.037327684462070465,
      "learning_rate": 2.758912657631156e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 16698256,
      "step": 18680
    },
    {
      "epoch": 4.931503233469711,
      "grad_norm": 0.05003780126571655,
      "learning_rate": 2.7579361691587198e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 16702736,
      "step": 18685
    },
    {
      "epoch": 4.932823017025208,
      "grad_norm": 0.18410705029964447,
      "learning_rate": 2.756959640909285e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 16707312,
      "step": 18690
    },
    {
      "epoch": 4.934142800580704,
      "grad_norm": 0.09078793227672577,
      "learning_rate": 2.7559830730334452e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 16711664,
      "step": 18695
    },
    {
      "epoch": 4.935462584136202,
      "grad_norm": 0.24783697724342346,
      "learning_rate": 2.7550064656817988e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 16715952,
      "step": 18700
    },
    {
      "epoch": 4.936782367691698,
      "grad_norm": 0.17165496945381165,
      "learning_rate": 2.7540298190049503e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 16720528,
      "step": 18705
    },
    {
      "epoch": 4.938102151247195,
      "grad_norm": 0.36842474341392517,
      "learning_rate": 2.7530531331535107e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 16724944,
      "step": 18710
    },
    {
      "epoch": 4.939421934802692,
      "grad_norm": 0.1413727104663849,
      "learning_rate": 2.752076408278099e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 16729648,
      "step": 18715
    },
    {
      "epoch": 4.940741718358189,
      "grad_norm": 0.046087827533483505,
      "learning_rate": 2.751099644529337e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 16734288,
      "step": 18720
    },
    {
      "epoch": 4.942061501913686,
      "grad_norm": 0.06575822830200195,
      "learning_rate": 2.7501228420578533e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 16738768,
      "step": 18725
    },
    {
      "epoch": 4.943381285469183,
      "grad_norm": 0.2496006339788437,
      "learning_rate": 2.7491460010142857e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 16743632,
      "step": 18730
    },
    {
      "epoch": 4.9447010690246795,
      "grad_norm": 0.3014904260635376,
      "learning_rate": 2.7481691215492727e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 16747728,
      "step": 18735
    },
    {
      "epoch": 4.946020852580177,
      "grad_norm": 0.29435354471206665,
      "learning_rate": 2.747192203813463e-05,
      "loss": 0.1188,
      "num_input_tokens_seen": 16752368,
      "step": 18740
    },
    {
      "epoch": 4.9473406361356735,
      "grad_norm": 0.2004237174987793,
      "learning_rate": 2.7462152479575087e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 16757104,
      "step": 18745
    },
    {
      "epoch": 4.948660419691171,
      "grad_norm": 0.41868647933006287,
      "learning_rate": 2.7452382541320697e-05,
      "loss": 0.0548,
      "num_input_tokens_seen": 16761520,
      "step": 18750
    },
    {
      "epoch": 4.9499802032466675,
      "grad_norm": 0.0709867998957634,
      "learning_rate": 2.7442612224878096e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 16766128,
      "step": 18755
    },
    {
      "epoch": 4.951299986802164,
      "grad_norm": 0.04019893333315849,
      "learning_rate": 2.7432841531753994e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 16770512,
      "step": 18760
    },
    {
      "epoch": 4.9526197703576615,
      "grad_norm": 0.2900136411190033,
      "learning_rate": 2.7423070463455147e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 16774960,
      "step": 18765
    },
    {
      "epoch": 4.953939553913158,
      "grad_norm": 0.07563133537769318,
      "learning_rate": 2.7413299021488397e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 16779440,
      "step": 18770
    },
    {
      "epoch": 4.9552593374686555,
      "grad_norm": 0.04926672577857971,
      "learning_rate": 2.7403527207360615e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 16783920,
      "step": 18775
    },
    {
      "epoch": 4.956579121024152,
      "grad_norm": 0.11770081520080566,
      "learning_rate": 2.7393755022578722e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 16788496,
      "step": 18780
    },
    {
      "epoch": 4.957898904579649,
      "grad_norm": 0.24784019589424133,
      "learning_rate": 2.7383982468649714e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 16792880,
      "step": 18785
    },
    {
      "epoch": 4.959218688135146,
      "grad_norm": 0.10999373346567154,
      "learning_rate": 2.7374209547080665e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 16797168,
      "step": 18790
    },
    {
      "epoch": 4.960538471690643,
      "grad_norm": 0.21823538839817047,
      "learning_rate": 2.7364436259378663e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 16801680,
      "step": 18795
    },
    {
      "epoch": 4.96185825524614,
      "grad_norm": 0.6799261569976807,
      "learning_rate": 2.735466260705088e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 16806032,
      "step": 18800
    },
    {
      "epoch": 4.96185825524614,
      "eval_loss": 0.06862161308526993,
      "eval_runtime": 64.7769,
      "eval_samples_per_second": 103.972,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 16806032,
      "step": 18800
    },
    {
      "epoch": 4.963178038801637,
      "grad_norm": 0.09917201846837997,
      "learning_rate": 2.7344888591604524e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 16810384,
      "step": 18805
    },
    {
      "epoch": 4.964497822357133,
      "grad_norm": 0.1404743492603302,
      "learning_rate": 2.7335114214546893e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 16815152,
      "step": 18810
    },
    {
      "epoch": 4.965817605912631,
      "grad_norm": 0.198434516787529,
      "learning_rate": 2.7325339477385293e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 16819888,
      "step": 18815
    },
    {
      "epoch": 4.967137389468127,
      "grad_norm": 0.10376154631376266,
      "learning_rate": 2.7315564381627128e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 16824432,
      "step": 18820
    },
    {
      "epoch": 4.968457173023624,
      "grad_norm": 0.060383908450603485,
      "learning_rate": 2.7305788928779835e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 16829008,
      "step": 18825
    },
    {
      "epoch": 4.969776956579121,
      "grad_norm": 0.13487784564495087,
      "learning_rate": 2.729601312035091e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 16833584,
      "step": 18830
    },
    {
      "epoch": 4.971096740134618,
      "grad_norm": 0.20110732316970825,
      "learning_rate": 2.7286236957847915e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 16838032,
      "step": 18835
    },
    {
      "epoch": 4.972416523690114,
      "grad_norm": 0.1296185702085495,
      "learning_rate": 2.7276460442778446e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 16842448,
      "step": 18840
    },
    {
      "epoch": 4.973736307245612,
      "grad_norm": 0.27330026030540466,
      "learning_rate": 2.726668357665017e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 16846832,
      "step": 18845
    },
    {
      "epoch": 4.975056090801108,
      "grad_norm": 0.07113484293222427,
      "learning_rate": 2.7256906360970808e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 16851472,
      "step": 18850
    },
    {
      "epoch": 4.976375874356606,
      "grad_norm": 0.4439070224761963,
      "learning_rate": 2.7247128797248117e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 16855920,
      "step": 18855
    },
    {
      "epoch": 4.977695657912102,
      "grad_norm": 0.07135985791683197,
      "learning_rate": 2.7237350886989925e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 16860176,
      "step": 18860
    },
    {
      "epoch": 4.979015441467599,
      "grad_norm": 0.17006325721740723,
      "learning_rate": 2.7227572631704107e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 16864560,
      "step": 18865
    },
    {
      "epoch": 4.980335225023096,
      "grad_norm": 0.22113637626171112,
      "learning_rate": 2.7217794032898596e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 16869008,
      "step": 18870
    },
    {
      "epoch": 4.981655008578593,
      "grad_norm": 0.27087342739105225,
      "learning_rate": 2.7208015092081384e-05,
      "loss": 0.1003,
      "num_input_tokens_seen": 16873232,
      "step": 18875
    },
    {
      "epoch": 4.98297479213409,
      "grad_norm": 0.1871531456708908,
      "learning_rate": 2.719823581076049e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 16877680,
      "step": 18880
    },
    {
      "epoch": 4.984294575689587,
      "grad_norm": 0.5451454520225525,
      "learning_rate": 2.718845619044401e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 16882256,
      "step": 18885
    },
    {
      "epoch": 4.9856143592450834,
      "grad_norm": 0.35073158144950867,
      "learning_rate": 2.7178676232640088e-05,
      "loss": 0.122,
      "num_input_tokens_seen": 16886512,
      "step": 18890
    },
    {
      "epoch": 4.986934142800581,
      "grad_norm": 0.23698747158050537,
      "learning_rate": 2.716889593885691e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 16890928,
      "step": 18895
    },
    {
      "epoch": 4.9882539263560775,
      "grad_norm": 0.3007442355155945,
      "learning_rate": 2.7159115310602716e-05,
      "loss": 0.1175,
      "num_input_tokens_seen": 16895248,
      "step": 18900
    },
    {
      "epoch": 4.989573709911575,
      "grad_norm": 0.059046510607004166,
      "learning_rate": 2.7149334349385814e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 16899952,
      "step": 18905
    },
    {
      "epoch": 4.9908934934670715,
      "grad_norm": 0.3999674618244171,
      "learning_rate": 2.713955305671454e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 16904592,
      "step": 18910
    },
    {
      "epoch": 4.992213277022568,
      "grad_norm": 0.42112573981285095,
      "learning_rate": 2.71297714340973e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 16909040,
      "step": 18915
    },
    {
      "epoch": 4.9935330605780655,
      "grad_norm": 0.2466246485710144,
      "learning_rate": 2.7119989483042545e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 16913808,
      "step": 18920
    },
    {
      "epoch": 4.994852844133562,
      "grad_norm": 0.0232480950653553,
      "learning_rate": 2.7110207205058768e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 16918640,
      "step": 18925
    },
    {
      "epoch": 4.9961726276890595,
      "grad_norm": 0.1594805270433426,
      "learning_rate": 2.7100424601654517e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 16922992,
      "step": 18930
    },
    {
      "epoch": 4.997492411244556,
      "grad_norm": 0.2975085973739624,
      "learning_rate": 2.7090641674338403e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 16927152,
      "step": 18935
    },
    {
      "epoch": 4.998812194800053,
      "grad_norm": 0.19951516389846802,
      "learning_rate": 2.7080858424619072e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 16931664,
      "step": 18940
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.22755782306194305,
      "learning_rate": 2.707107485400521e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 16935568,
      "step": 18945
    },
    {
      "epoch": 5.001319783555497,
      "grad_norm": 0.48928946256637573,
      "learning_rate": 2.7061290964005586e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 16940176,
      "step": 18950
    },
    {
      "epoch": 5.002639567110994,
      "grad_norm": 0.15697601437568665,
      "learning_rate": 2.7051506756129e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 16944720,
      "step": 18955
    },
    {
      "epoch": 5.003959350666491,
      "grad_norm": 0.3455692529678345,
      "learning_rate": 2.704172223188428e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 16949200,
      "step": 18960
    },
    {
      "epoch": 5.005279134221988,
      "grad_norm": 0.22429300844669342,
      "learning_rate": 2.7031937392780334e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 16953968,
      "step": 18965
    },
    {
      "epoch": 5.006598917777485,
      "grad_norm": 0.27705511450767517,
      "learning_rate": 2.702215224032611e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 16958448,
      "step": 18970
    },
    {
      "epoch": 5.007918701332981,
      "grad_norm": 0.21976374089717865,
      "learning_rate": 2.70123667760306e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 16963024,
      "step": 18975
    },
    {
      "epoch": 5.009238484888479,
      "grad_norm": 0.1555311381816864,
      "learning_rate": 2.7002581001402845e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 16967376,
      "step": 18980
    },
    {
      "epoch": 5.010558268443975,
      "grad_norm": 0.06559067964553833,
      "learning_rate": 2.6992794917951923e-05,
      "loss": 0.0981,
      "num_input_tokens_seen": 16971952,
      "step": 18985
    },
    {
      "epoch": 5.011878051999472,
      "grad_norm": 0.09878981113433838,
      "learning_rate": 2.6983008527187e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 16976720,
      "step": 18990
    },
    {
      "epoch": 5.013197835554969,
      "grad_norm": 0.027351541444659233,
      "learning_rate": 2.697322183061723e-05,
      "loss": 0.0197,
      "num_input_tokens_seen": 16981328,
      "step": 18995
    },
    {
      "epoch": 5.014517619110466,
      "grad_norm": 0.28916311264038086,
      "learning_rate": 2.696343482975186e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 16986160,
      "step": 19000
    },
    {
      "epoch": 5.014517619110466,
      "eval_loss": 0.06854243576526642,
      "eval_runtime": 64.7706,
      "eval_samples_per_second": 103.982,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 16986160,
      "step": 19000
    },
    {
      "epoch": 5.015837402665963,
      "grad_norm": 0.07987211644649506,
      "learning_rate": 2.695364752610016e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 16990352,
      "step": 19005
    },
    {
      "epoch": 5.01715718622146,
      "grad_norm": 0.3318217992782593,
      "learning_rate": 2.6943859921171467e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 16994960,
      "step": 19010
    },
    {
      "epoch": 5.018476969776956,
      "grad_norm": 0.11486142873764038,
      "learning_rate": 2.6934072016475143e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 16999376,
      "step": 19015
    },
    {
      "epoch": 5.019796753332454,
      "grad_norm": 0.4915933907032013,
      "learning_rate": 2.6924283813520606e-05,
      "loss": 0.1208,
      "num_input_tokens_seen": 17004048,
      "step": 19020
    },
    {
      "epoch": 5.02111653688795,
      "grad_norm": 0.17441192269325256,
      "learning_rate": 2.691449531381733e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 17008464,
      "step": 19025
    },
    {
      "epoch": 5.022436320443448,
      "grad_norm": 0.16601507365703583,
      "learning_rate": 2.6904706518874816e-05,
      "loss": 0.0186,
      "num_input_tokens_seen": 17013072,
      "step": 19030
    },
    {
      "epoch": 5.023756103998944,
      "grad_norm": 0.26146870851516724,
      "learning_rate": 2.6894917430202615e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 17017424,
      "step": 19035
    },
    {
      "epoch": 5.025075887554441,
      "grad_norm": 0.2478986233472824,
      "learning_rate": 2.6885128049310343e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 17021744,
      "step": 19040
    },
    {
      "epoch": 5.026395671109938,
      "grad_norm": 0.19030870497226715,
      "learning_rate": 2.687533837770762e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 17026128,
      "step": 19045
    },
    {
      "epoch": 5.027715454665435,
      "grad_norm": 0.22293902933597565,
      "learning_rate": 2.6865548416904162e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 17030576,
      "step": 19050
    },
    {
      "epoch": 5.029035238220931,
      "grad_norm": 0.17587222158908844,
      "learning_rate": 2.68557581684097e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 17035120,
      "step": 19055
    },
    {
      "epoch": 5.030355021776429,
      "grad_norm": 0.05642015114426613,
      "learning_rate": 2.6845967633733998e-05,
      "loss": 0.0895,
      "num_input_tokens_seen": 17039568,
      "step": 19060
    },
    {
      "epoch": 5.031674805331925,
      "grad_norm": 0.43870484828948975,
      "learning_rate": 2.683617681438689e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 17043984,
      "step": 19065
    },
    {
      "epoch": 5.032994588887423,
      "grad_norm": 0.11381293088197708,
      "learning_rate": 2.682638571187825e-05,
      "loss": 0.0725,
      "num_input_tokens_seen": 17048432,
      "step": 19070
    },
    {
      "epoch": 5.034314372442919,
      "grad_norm": 0.4731496572494507,
      "learning_rate": 2.6816594327717976e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 17053104,
      "step": 19075
    },
    {
      "epoch": 5.035634155998416,
      "grad_norm": 0.34311068058013916,
      "learning_rate": 2.680680266341603e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 17057424,
      "step": 19080
    },
    {
      "epoch": 5.036953939553913,
      "grad_norm": 0.20074953138828278,
      "learning_rate": 2.67970107204824e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 17061776,
      "step": 19085
    },
    {
      "epoch": 5.03827372310941,
      "grad_norm": 0.232033371925354,
      "learning_rate": 2.6787218500427142e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 17066640,
      "step": 19090
    },
    {
      "epoch": 5.039593506664907,
      "grad_norm": 0.16428549587726593,
      "learning_rate": 2.6777426004760332e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 17071280,
      "step": 19095
    },
    {
      "epoch": 5.040913290220404,
      "grad_norm": 0.1406024694442749,
      "learning_rate": 2.6767633234992094e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 17075696,
      "step": 19100
    },
    {
      "epoch": 5.0422330737759005,
      "grad_norm": 0.34932538866996765,
      "learning_rate": 2.6757840192632598e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 17080048,
      "step": 19105
    },
    {
      "epoch": 5.043552857331398,
      "grad_norm": 0.1890619695186615,
      "learning_rate": 2.6748046879192052e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 17084432,
      "step": 19110
    },
    {
      "epoch": 5.0448726408868945,
      "grad_norm": 0.14648717641830444,
      "learning_rate": 2.673825329618071e-05,
      "loss": 0.0644,
      "num_input_tokens_seen": 17088464,
      "step": 19115
    },
    {
      "epoch": 5.046192424442391,
      "grad_norm": 0.29285961389541626,
      "learning_rate": 2.6728459445108866e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 17093168,
      "step": 19120
    },
    {
      "epoch": 5.0475122079978885,
      "grad_norm": 0.11911913752555847,
      "learning_rate": 2.6718665327486854e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 17097712,
      "step": 19125
    },
    {
      "epoch": 5.048831991553385,
      "grad_norm": 0.1464732140302658,
      "learning_rate": 2.6708870944825048e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 17102096,
      "step": 19130
    },
    {
      "epoch": 5.0501517751088825,
      "grad_norm": 0.3286367356777191,
      "learning_rate": 2.6699076298633874e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 17106992,
      "step": 19135
    },
    {
      "epoch": 5.051471558664379,
      "grad_norm": 0.08369673788547516,
      "learning_rate": 2.6689281390423788e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 17111344,
      "step": 19140
    },
    {
      "epoch": 5.052791342219876,
      "grad_norm": 0.13633674383163452,
      "learning_rate": 2.667948622170527e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 17115504,
      "step": 19145
    },
    {
      "epoch": 5.054111125775373,
      "grad_norm": 0.18322749435901642,
      "learning_rate": 2.6669690793988873e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 17120368,
      "step": 19150
    },
    {
      "epoch": 5.05543090933087,
      "grad_norm": 0.20799368619918823,
      "learning_rate": 2.665989510878518e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 17124720,
      "step": 19155
    },
    {
      "epoch": 5.056750692886367,
      "grad_norm": 0.3070172369480133,
      "learning_rate": 2.6650099167604793e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 17129296,
      "step": 19160
    },
    {
      "epoch": 5.058070476441864,
      "grad_norm": 0.32486045360565186,
      "learning_rate": 2.6640302971958376e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 17133936,
      "step": 19165
    },
    {
      "epoch": 5.05939025999736,
      "grad_norm": 0.48955875635147095,
      "learning_rate": 2.6630506523356635e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 17138352,
      "step": 19170
    },
    {
      "epoch": 5.060710043552858,
      "grad_norm": 0.3325125575065613,
      "learning_rate": 2.6620709823310297e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 17142832,
      "step": 19175
    },
    {
      "epoch": 5.062029827108354,
      "grad_norm": 0.03684444725513458,
      "learning_rate": 2.661091287333014e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 17147440,
      "step": 19180
    },
    {
      "epoch": 5.063349610663851,
      "grad_norm": 0.09186553955078125,
      "learning_rate": 2.660111567492696e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 17151984,
      "step": 19185
    },
    {
      "epoch": 5.064669394219348,
      "grad_norm": 0.2506871521472931,
      "learning_rate": 2.6591318229611635e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 17156240,
      "step": 19190
    },
    {
      "epoch": 5.065989177774845,
      "grad_norm": 0.3182862401008606,
      "learning_rate": 2.6581520538895037e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 17160432,
      "step": 19195
    },
    {
      "epoch": 5.067308961330342,
      "grad_norm": 0.25975215435028076,
      "learning_rate": 2.6571722604288102e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 17164848,
      "step": 19200
    },
    {
      "epoch": 5.067308961330342,
      "eval_loss": 0.06870897114276886,
      "eval_runtime": 64.7963,
      "eval_samples_per_second": 103.941,
      "eval_steps_per_second": 25.989,
      "num_input_tokens_seen": 17164848,
      "step": 19200
    },
    {
      "epoch": 5.068628744885839,
      "grad_norm": 0.10210270434617996,
      "learning_rate": 2.656192442730179e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 17169168,
      "step": 19205
    },
    {
      "epoch": 5.069948528441335,
      "grad_norm": 0.033850397914648056,
      "learning_rate": 2.6552126009447098e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 17173712,
      "step": 19210
    },
    {
      "epoch": 5.071268311996833,
      "grad_norm": 0.44324642419815063,
      "learning_rate": 2.654232735223507e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 17178192,
      "step": 19215
    },
    {
      "epoch": 5.072588095552329,
      "grad_norm": 0.05681411549448967,
      "learning_rate": 2.6532528457176787e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 17182896,
      "step": 19220
    },
    {
      "epoch": 5.073907879107827,
      "grad_norm": 0.26370126008987427,
      "learning_rate": 2.6522729325783348e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 17187280,
      "step": 19225
    },
    {
      "epoch": 5.075227662663323,
      "grad_norm": 0.2961690127849579,
      "learning_rate": 2.6512929959565914e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 17191888,
      "step": 19230
    },
    {
      "epoch": 5.07654744621882,
      "grad_norm": 0.20757268369197845,
      "learning_rate": 2.6503130360035673e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 17196368,
      "step": 19235
    },
    {
      "epoch": 5.077867229774317,
      "grad_norm": 0.07291275262832642,
      "learning_rate": 2.6493330528703835e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 17200656,
      "step": 19240
    },
    {
      "epoch": 5.079187013329814,
      "grad_norm": 0.0547778382897377,
      "learning_rate": 2.648353046708167e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 17204976,
      "step": 19245
    },
    {
      "epoch": 5.08050679688531,
      "grad_norm": 0.21507373452186584,
      "learning_rate": 2.647373017668046e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 17209136,
      "step": 19250
    },
    {
      "epoch": 5.081826580440808,
      "grad_norm": 0.3954898715019226,
      "learning_rate": 2.6463929659011537e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 17213552,
      "step": 19255
    },
    {
      "epoch": 5.083146363996304,
      "grad_norm": 0.22904185950756073,
      "learning_rate": 2.6454128915586262e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 17217872,
      "step": 19260
    },
    {
      "epoch": 5.084466147551802,
      "grad_norm": 0.20396392047405243,
      "learning_rate": 2.6444327947916036e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 17222096,
      "step": 19265
    },
    {
      "epoch": 5.085785931107298,
      "grad_norm": 0.14217182993888855,
      "learning_rate": 2.6434526757512292e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 17226608,
      "step": 19270
    },
    {
      "epoch": 5.087105714662795,
      "grad_norm": 0.07806000113487244,
      "learning_rate": 2.6424725345886486e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 17231440,
      "step": 19275
    },
    {
      "epoch": 5.088425498218292,
      "grad_norm": 0.12203270196914673,
      "learning_rate": 2.641492371455014e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 17235760,
      "step": 19280
    },
    {
      "epoch": 5.089745281773789,
      "grad_norm": 0.03367949649691582,
      "learning_rate": 2.640512186501477e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 17240240,
      "step": 19285
    },
    {
      "epoch": 5.091065065329286,
      "grad_norm": 0.16777314245700836,
      "learning_rate": 2.639531979879195e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 17245104,
      "step": 19290
    },
    {
      "epoch": 5.092384848884783,
      "grad_norm": 0.25858274102211,
      "learning_rate": 2.638551751739328e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 17249680,
      "step": 19295
    },
    {
      "epoch": 5.0937046324402795,
      "grad_norm": 0.14053095877170563,
      "learning_rate": 2.6375715022330404e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 17254160,
      "step": 19300
    },
    {
      "epoch": 5.095024415995777,
      "grad_norm": 0.19657224416732788,
      "learning_rate": 2.6365912315114976e-05,
      "loss": 0.0721,
      "num_input_tokens_seen": 17258832,
      "step": 19305
    },
    {
      "epoch": 5.0963441995512735,
      "grad_norm": 0.14245790243148804,
      "learning_rate": 2.6356109397258704e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 17263120,
      "step": 19310
    },
    {
      "epoch": 5.09766398310677,
      "grad_norm": 0.3223910331726074,
      "learning_rate": 2.6346306270273325e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 17267472,
      "step": 19315
    },
    {
      "epoch": 5.0989837666622675,
      "grad_norm": 0.4378238618373871,
      "learning_rate": 2.6336502935670608e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 17271632,
      "step": 19320
    },
    {
      "epoch": 5.100303550217764,
      "grad_norm": 0.3300918936729431,
      "learning_rate": 2.6326699394962333e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 17275760,
      "step": 19325
    },
    {
      "epoch": 5.1016233337732615,
      "grad_norm": 0.3612113893032074,
      "learning_rate": 2.6316895649660334e-05,
      "loss": 0.1097,
      "num_input_tokens_seen": 17280368,
      "step": 19330
    },
    {
      "epoch": 5.102943117328758,
      "grad_norm": 0.360798180103302,
      "learning_rate": 2.6307091701276486e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 17284688,
      "step": 19335
    },
    {
      "epoch": 5.104262900884255,
      "grad_norm": 0.09335128217935562,
      "learning_rate": 2.629728755132267e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 17289168,
      "step": 19340
    },
    {
      "epoch": 5.105582684439752,
      "grad_norm": 0.08992238342761993,
      "learning_rate": 2.628748320131081e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 17293680,
      "step": 19345
    },
    {
      "epoch": 5.106902467995249,
      "grad_norm": 0.21821841597557068,
      "learning_rate": 2.6277678652752856e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 17298096,
      "step": 19350
    },
    {
      "epoch": 5.108222251550746,
      "grad_norm": 0.3959363102912903,
      "learning_rate": 2.6267873907160807e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 17302672,
      "step": 19355
    },
    {
      "epoch": 5.109542035106243,
      "grad_norm": 0.2679448425769806,
      "learning_rate": 2.6258068966046668e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 17307568,
      "step": 19360
    },
    {
      "epoch": 5.110861818661739,
      "grad_norm": 0.08793846517801285,
      "learning_rate": 2.6248263830922475e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 17312176,
      "step": 19365
    },
    {
      "epoch": 5.112181602217237,
      "grad_norm": 0.10815593600273132,
      "learning_rate": 2.6238458503300318e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 17316688,
      "step": 19370
    },
    {
      "epoch": 5.113501385772733,
      "grad_norm": 0.08007234334945679,
      "learning_rate": 2.6228652984692292e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 17320880,
      "step": 19375
    },
    {
      "epoch": 5.114821169328231,
      "grad_norm": 0.34647461771965027,
      "learning_rate": 2.621884727661054e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 17325104,
      "step": 19380
    },
    {
      "epoch": 5.116140952883727,
      "grad_norm": 0.23384162783622742,
      "learning_rate": 2.6209041380567222e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 17329360,
      "step": 19385
    },
    {
      "epoch": 5.117460736439224,
      "grad_norm": 0.2360575944185257,
      "learning_rate": 2.6199235298074527e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 17333776,
      "step": 19390
    },
    {
      "epoch": 5.118780519994721,
      "grad_norm": 0.09644157439470291,
      "learning_rate": 2.618942903064468e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 17338544,
      "step": 19395
    },
    {
      "epoch": 5.120100303550218,
      "grad_norm": 0.1589479297399521,
      "learning_rate": 2.6179622579789932e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 17342800,
      "step": 19400
    },
    {
      "epoch": 5.120100303550218,
      "eval_loss": 0.06855374574661255,
      "eval_runtime": 64.7426,
      "eval_samples_per_second": 104.027,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 17342800,
      "step": 19400
    },
    {
      "epoch": 5.121420087105714,
      "grad_norm": 0.31678342819213867,
      "learning_rate": 2.6169815947022553e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 17347152,
      "step": 19405
    },
    {
      "epoch": 5.122739870661212,
      "grad_norm": 0.14495675265789032,
      "learning_rate": 2.6160009133854853e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 17351984,
      "step": 19410
    },
    {
      "epoch": 5.124059654216708,
      "grad_norm": 0.24549011886119843,
      "learning_rate": 2.6150202141799168e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 17356432,
      "step": 19415
    },
    {
      "epoch": 5.125379437772206,
      "grad_norm": 0.1788676381111145,
      "learning_rate": 2.614039497236786e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 17361104,
      "step": 19420
    },
    {
      "epoch": 5.126699221327702,
      "grad_norm": 0.2293304204940796,
      "learning_rate": 2.6130587627073315e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 17365488,
      "step": 19425
    },
    {
      "epoch": 5.128019004883199,
      "grad_norm": 0.10142004489898682,
      "learning_rate": 2.6120780107427956e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 17369808,
      "step": 19430
    },
    {
      "epoch": 5.129338788438696,
      "grad_norm": 0.41322341561317444,
      "learning_rate": 2.6110972414944214e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 17374256,
      "step": 19435
    },
    {
      "epoch": 5.130658571994193,
      "grad_norm": 0.026502422988414764,
      "learning_rate": 2.6101164551134565e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 17378800,
      "step": 19440
    },
    {
      "epoch": 5.131978355549689,
      "grad_norm": 0.26221224665641785,
      "learning_rate": 2.6091356517511505e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 17383248,
      "step": 19445
    },
    {
      "epoch": 5.133298139105187,
      "grad_norm": 0.23653709888458252,
      "learning_rate": 2.608154831558755e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 17387536,
      "step": 19450
    },
    {
      "epoch": 5.134617922660683,
      "grad_norm": 0.10292716324329376,
      "learning_rate": 2.607173994687526e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 17392208,
      "step": 19455
    },
    {
      "epoch": 5.135937706216181,
      "grad_norm": 0.22351998090744019,
      "learning_rate": 2.6061931412887196e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 17396816,
      "step": 19460
    },
    {
      "epoch": 5.137257489771677,
      "grad_norm": 0.2204088568687439,
      "learning_rate": 2.6052122715135973e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 17401232,
      "step": 19465
    },
    {
      "epoch": 5.138577273327174,
      "grad_norm": 0.2688142657279968,
      "learning_rate": 2.60423138551342e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 17405648,
      "step": 19470
    },
    {
      "epoch": 5.139897056882671,
      "grad_norm": 0.08223710209131241,
      "learning_rate": 2.6032504834394527e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 17410064,
      "step": 19475
    },
    {
      "epoch": 5.141216840438168,
      "grad_norm": 0.31379571557044983,
      "learning_rate": 2.602269565442964e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 17414448,
      "step": 19480
    },
    {
      "epoch": 5.142536623993665,
      "grad_norm": 0.06925041973590851,
      "learning_rate": 2.6012886316752227e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 17418864,
      "step": 19485
    },
    {
      "epoch": 5.143856407549162,
      "grad_norm": 0.4157066345214844,
      "learning_rate": 2.6003076822875018e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 17422960,
      "step": 19490
    },
    {
      "epoch": 5.1451761911046585,
      "grad_norm": 0.08373609185218811,
      "learning_rate": 2.5993267174310755e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 17427568,
      "step": 19495
    },
    {
      "epoch": 5.146495974660156,
      "grad_norm": 0.3729817271232605,
      "learning_rate": 2.5983457372572218e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 17432400,
      "step": 19500
    },
    {
      "epoch": 5.1478157582156525,
      "grad_norm": 0.10966972261667252,
      "learning_rate": 2.597364741917219e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 17436848,
      "step": 19505
    },
    {
      "epoch": 5.14913554177115,
      "grad_norm": 0.44847801327705383,
      "learning_rate": 2.5963837315623492e-05,
      "loss": 0.1241,
      "num_input_tokens_seen": 17441168,
      "step": 19510
    },
    {
      "epoch": 5.1504553253266465,
      "grad_norm": 0.186454176902771,
      "learning_rate": 2.595402706343897e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 17445392,
      "step": 19515
    },
    {
      "epoch": 5.151775108882143,
      "grad_norm": 0.20327343046665192,
      "learning_rate": 2.594421666413148e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 17449968,
      "step": 19520
    },
    {
      "epoch": 5.1530948924376405,
      "grad_norm": 0.1192198246717453,
      "learning_rate": 2.5934406119213928e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 17454192,
      "step": 19525
    },
    {
      "epoch": 5.154414675993137,
      "grad_norm": 0.13682861626148224,
      "learning_rate": 2.5924595430199193e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 17458416,
      "step": 19530
    },
    {
      "epoch": 5.155734459548634,
      "grad_norm": 0.1316082924604416,
      "learning_rate": 2.5914784598600238e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 17462672,
      "step": 19535
    },
    {
      "epoch": 5.157054243104131,
      "grad_norm": 0.2377040833234787,
      "learning_rate": 2.5904973625930002e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 17467376,
      "step": 19540
    },
    {
      "epoch": 5.158374026659628,
      "grad_norm": 0.14712467789649963,
      "learning_rate": 2.5895162513701456e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 17471792,
      "step": 19545
    },
    {
      "epoch": 5.159693810215125,
      "grad_norm": 0.25829941034317017,
      "learning_rate": 2.5885351263427593e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 17476304,
      "step": 19550
    },
    {
      "epoch": 5.161013593770622,
      "grad_norm": 0.1244855746626854,
      "learning_rate": 2.5875539876621448e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 17480592,
      "step": 19555
    },
    {
      "epoch": 5.162333377326118,
      "grad_norm": 0.2959396541118622,
      "learning_rate": 2.586572835479605e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 17485168,
      "step": 19560
    },
    {
      "epoch": 5.163653160881616,
      "grad_norm": 0.2747514247894287,
      "learning_rate": 2.585591669946446e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 17489456,
      "step": 19565
    },
    {
      "epoch": 5.164972944437112,
      "grad_norm": 0.4393116235733032,
      "learning_rate": 2.5846104912139756e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 17493584,
      "step": 19570
    },
    {
      "epoch": 5.16629272799261,
      "grad_norm": 0.3752378821372986,
      "learning_rate": 2.583629299433505e-05,
      "loss": 0.0802,
      "num_input_tokens_seen": 17497968,
      "step": 19575
    },
    {
      "epoch": 5.167612511548106,
      "grad_norm": 0.11787020415067673,
      "learning_rate": 2.582648094756345e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 17502224,
      "step": 19580
    },
    {
      "epoch": 5.168932295103603,
      "grad_norm": 0.16222721338272095,
      "learning_rate": 2.5816668773338098e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 17506768,
      "step": 19585
    },
    {
      "epoch": 5.1702520786591,
      "grad_norm": 0.06564243882894516,
      "learning_rate": 2.580685647317216e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 17511056,
      "step": 19590
    },
    {
      "epoch": 5.171571862214597,
      "grad_norm": 0.594566822052002,
      "learning_rate": 2.5797044048578818e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 17515568,
      "step": 19595
    },
    {
      "epoch": 5.172891645770093,
      "grad_norm": 0.18579146265983582,
      "learning_rate": 2.5787231501071262e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 17520144,
      "step": 19600
    },
    {
      "epoch": 5.172891645770093,
      "eval_loss": 0.06836747378110886,
      "eval_runtime": 64.7298,
      "eval_samples_per_second": 104.048,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 17520144,
      "step": 19600
    },
    {
      "epoch": 5.174211429325591,
      "grad_norm": 0.2781524658203125,
      "learning_rate": 2.577741883216272e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 17524592,
      "step": 19605
    },
    {
      "epoch": 5.175531212881087,
      "grad_norm": 0.08001847565174103,
      "learning_rate": 2.576760604336642e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 17529072,
      "step": 19610
    },
    {
      "epoch": 5.176850996436585,
      "grad_norm": 0.21498465538024902,
      "learning_rate": 2.575779313619563e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 17533776,
      "step": 19615
    },
    {
      "epoch": 5.178170779992081,
      "grad_norm": 0.10978720337152481,
      "learning_rate": 2.5747980112163605e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 17537872,
      "step": 19620
    },
    {
      "epoch": 5.179490563547578,
      "grad_norm": 0.18835125863552094,
      "learning_rate": 2.5738166972783656e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 17542640,
      "step": 19625
    },
    {
      "epoch": 5.180810347103075,
      "grad_norm": 0.12763911485671997,
      "learning_rate": 2.5728353719569075e-05,
      "loss": 0.0842,
      "num_input_tokens_seen": 17547088,
      "step": 19630
    },
    {
      "epoch": 5.182130130658572,
      "grad_norm": 0.15005506575107574,
      "learning_rate": 2.57185403540332e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 17551600,
      "step": 19635
    },
    {
      "epoch": 5.183449914214069,
      "grad_norm": 0.18843179941177368,
      "learning_rate": 2.5708726877689375e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 17555888,
      "step": 19640
    },
    {
      "epoch": 5.184769697769566,
      "grad_norm": 0.027912145480513573,
      "learning_rate": 2.5698913292050964e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 17560464,
      "step": 19645
    },
    {
      "epoch": 5.186089481325062,
      "grad_norm": 0.334033340215683,
      "learning_rate": 2.568909959863133e-05,
      "loss": 0.1307,
      "num_input_tokens_seen": 17564784,
      "step": 19650
    },
    {
      "epoch": 5.18740926488056,
      "grad_norm": 0.1925666481256485,
      "learning_rate": 2.5679285798943887e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 17569296,
      "step": 19655
    },
    {
      "epoch": 5.188729048436056,
      "grad_norm": 0.30758169293403625,
      "learning_rate": 2.5669471894502035e-05,
      "loss": 0.098,
      "num_input_tokens_seen": 17573552,
      "step": 19660
    },
    {
      "epoch": 5.190048831991553,
      "grad_norm": 0.2587870657444,
      "learning_rate": 2.56596578868192e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 17578448,
      "step": 19665
    },
    {
      "epoch": 5.19136861554705,
      "grad_norm": 0.049315791577100754,
      "learning_rate": 2.564984377740883e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 17583024,
      "step": 19670
    },
    {
      "epoch": 5.192688399102547,
      "grad_norm": 0.6486089825630188,
      "learning_rate": 2.564002956778438e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 17587376,
      "step": 19675
    },
    {
      "epoch": 5.194008182658044,
      "grad_norm": 0.3426949679851532,
      "learning_rate": 2.563021525945934e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 17591888,
      "step": 19680
    },
    {
      "epoch": 5.195327966213541,
      "grad_norm": 0.21619224548339844,
      "learning_rate": 2.562040085394718e-05,
      "loss": 0.1097,
      "num_input_tokens_seen": 17596272,
      "step": 19685
    },
    {
      "epoch": 5.1966477497690375,
      "grad_norm": 0.04153769835829735,
      "learning_rate": 2.56105863527614e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 17600720,
      "step": 19690
    },
    {
      "epoch": 5.197967533324535,
      "grad_norm": 0.2781229615211487,
      "learning_rate": 2.5600771757415548e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 17605008,
      "step": 19695
    },
    {
      "epoch": 5.1992873168800315,
      "grad_norm": 0.30697721242904663,
      "learning_rate": 2.5590957069423134e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 17609392,
      "step": 19700
    },
    {
      "epoch": 5.200607100435529,
      "grad_norm": 0.34575575590133667,
      "learning_rate": 2.5581142290297716e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 17613904,
      "step": 19705
    },
    {
      "epoch": 5.2019268839910255,
      "grad_norm": 0.4239283800125122,
      "learning_rate": 2.557132742155285e-05,
      "loss": 0.0874,
      "num_input_tokens_seen": 17618160,
      "step": 19710
    },
    {
      "epoch": 5.203246667546522,
      "grad_norm": 0.3491196036338806,
      "learning_rate": 2.556151246470212e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 17622576,
      "step": 19715
    },
    {
      "epoch": 5.2045664511020195,
      "grad_norm": 0.052440203726291656,
      "learning_rate": 2.5551697421259114e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 17627248,
      "step": 19720
    },
    {
      "epoch": 5.205886234657516,
      "grad_norm": 0.24077579379081726,
      "learning_rate": 2.554188229273743e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 17631664,
      "step": 19725
    },
    {
      "epoch": 5.207206018213013,
      "grad_norm": 0.0738476887345314,
      "learning_rate": 2.5532067080650678e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 17635920,
      "step": 19730
    },
    {
      "epoch": 5.20852580176851,
      "grad_norm": 0.10256332159042358,
      "learning_rate": 2.55222517865125e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 17640272,
      "step": 19735
    },
    {
      "epoch": 5.209845585324007,
      "grad_norm": 0.09147744625806808,
      "learning_rate": 2.5512436411836538e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 17644784,
      "step": 19740
    },
    {
      "epoch": 5.211165368879504,
      "grad_norm": 0.11400903016328812,
      "learning_rate": 2.5502620958136443e-05,
      "loss": 0.0187,
      "num_input_tokens_seen": 17649456,
      "step": 19745
    },
    {
      "epoch": 5.212485152435001,
      "grad_norm": 0.41656580567359924,
      "learning_rate": 2.5492805426925874e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 17653520,
      "step": 19750
    },
    {
      "epoch": 5.213804935990497,
      "grad_norm": 0.24157501757144928,
      "learning_rate": 2.5482989819718523e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 17657712,
      "step": 19755
    },
    {
      "epoch": 5.215124719545995,
      "grad_norm": 0.0967484563589096,
      "learning_rate": 2.5473174138028065e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 17662224,
      "step": 19760
    },
    {
      "epoch": 5.216444503101491,
      "grad_norm": 0.261910617351532,
      "learning_rate": 2.5463358383368212e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 17666896,
      "step": 19765
    },
    {
      "epoch": 5.217764286656989,
      "grad_norm": 0.10458493232727051,
      "learning_rate": 2.545354255725267e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 17671536,
      "step": 19770
    },
    {
      "epoch": 5.219084070212485,
      "grad_norm": 0.11341068893671036,
      "learning_rate": 2.5443726661195165e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 17675888,
      "step": 19775
    },
    {
      "epoch": 5.220403853767982,
      "grad_norm": 0.26071885228157043,
      "learning_rate": 2.543391069670944e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 17680848,
      "step": 19780
    },
    {
      "epoch": 5.221723637323479,
      "grad_norm": 0.17147307097911835,
      "learning_rate": 2.5424094665309228e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 17685200,
      "step": 19785
    },
    {
      "epoch": 5.223043420878976,
      "grad_norm": 0.22167818248271942,
      "learning_rate": 2.5414278568508292e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 17689488,
      "step": 19790
    },
    {
      "epoch": 5.224363204434472,
      "grad_norm": 0.278536856174469,
      "learning_rate": 2.540446240782039e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 17693744,
      "step": 19795
    },
    {
      "epoch": 5.22568298798997,
      "grad_norm": 0.21481439471244812,
      "learning_rate": 2.5394646184759307e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 17697936,
      "step": 19800
    },
    {
      "epoch": 5.22568298798997,
      "eval_loss": 0.06815442442893982,
      "eval_runtime": 64.7625,
      "eval_samples_per_second": 103.995,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 17697936,
      "step": 19800
    },
    {
      "epoch": 5.227002771545466,
      "grad_norm": 0.04391780495643616,
      "learning_rate": 2.538482990083882e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 17702416,
      "step": 19805
    },
    {
      "epoch": 5.228322555100964,
      "grad_norm": 0.5930991172790527,
      "learning_rate": 2.5375013557572725e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 17707216,
      "step": 19810
    },
    {
      "epoch": 5.22964233865646,
      "grad_norm": 0.10810256004333496,
      "learning_rate": 2.536519715647483e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 17712048,
      "step": 19815
    },
    {
      "epoch": 5.230962122211957,
      "grad_norm": 0.14258959889411926,
      "learning_rate": 2.535538069905894e-05,
      "loss": 0.1121,
      "num_input_tokens_seen": 17716528,
      "step": 19820
    },
    {
      "epoch": 5.232281905767454,
      "grad_norm": 0.0782979428768158,
      "learning_rate": 2.534556418683888e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 17720784,
      "step": 19825
    },
    {
      "epoch": 5.233601689322951,
      "grad_norm": 0.19077028334140778,
      "learning_rate": 2.5335747621328486e-05,
      "loss": 0.0215,
      "num_input_tokens_seen": 17725072,
      "step": 19830
    },
    {
      "epoch": 5.234921472878448,
      "grad_norm": 0.06264678388834,
      "learning_rate": 2.5325931004041586e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 17729552,
      "step": 19835
    },
    {
      "epoch": 5.236241256433945,
      "grad_norm": 0.1429455280303955,
      "learning_rate": 2.5316114336492032e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 17734000,
      "step": 19840
    },
    {
      "epoch": 5.2375610399894414,
      "grad_norm": 0.11153917759656906,
      "learning_rate": 2.530629762019367e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 17738576,
      "step": 19845
    },
    {
      "epoch": 5.238880823544939,
      "grad_norm": 0.0435015968978405,
      "learning_rate": 2.5296480856660364e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 17742992,
      "step": 19850
    },
    {
      "epoch": 5.2402006071004354,
      "grad_norm": 0.19637437164783478,
      "learning_rate": 2.528666404740599e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 17747440,
      "step": 19855
    },
    {
      "epoch": 5.241520390655932,
      "grad_norm": 0.2633301615715027,
      "learning_rate": 2.527684719394442e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 17751984,
      "step": 19860
    },
    {
      "epoch": 5.2428401742114294,
      "grad_norm": 0.21777556836605072,
      "learning_rate": 2.526703029778953e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 17756368,
      "step": 19865
    },
    {
      "epoch": 5.244159957766926,
      "grad_norm": 0.16902273893356323,
      "learning_rate": 2.5257213360455208e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 17761008,
      "step": 19870
    },
    {
      "epoch": 5.2454797413224235,
      "grad_norm": 0.09370040148496628,
      "learning_rate": 2.5247396383455353e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 17765360,
      "step": 19875
    },
    {
      "epoch": 5.24679952487792,
      "grad_norm": 0.2592080235481262,
      "learning_rate": 2.523757936830387e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 17769616,
      "step": 19880
    },
    {
      "epoch": 5.248119308433417,
      "grad_norm": 0.060568470507860184,
      "learning_rate": 2.5227762316514662e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 17774096,
      "step": 19885
    },
    {
      "epoch": 5.249439091988914,
      "grad_norm": 0.20296640694141388,
      "learning_rate": 2.5217945229601648e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 17778512,
      "step": 19890
    },
    {
      "epoch": 5.250758875544411,
      "grad_norm": 0.057543933391571045,
      "learning_rate": 2.5208128109078738e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 17782832,
      "step": 19895
    },
    {
      "epoch": 5.252078659099908,
      "grad_norm": 0.2966860234737396,
      "learning_rate": 2.5198310956459853e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 17787504,
      "step": 19900
    },
    {
      "epoch": 5.253398442655405,
      "grad_norm": 0.10221467912197113,
      "learning_rate": 2.518849377325893e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 17791920,
      "step": 19905
    },
    {
      "epoch": 5.254718226210901,
      "grad_norm": 0.15956884622573853,
      "learning_rate": 2.51786765609899e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 17796688,
      "step": 19910
    },
    {
      "epoch": 5.256038009766399,
      "grad_norm": 0.044606950134038925,
      "learning_rate": 2.5168859321166694e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 17801072,
      "step": 19915
    },
    {
      "epoch": 5.257357793321895,
      "grad_norm": 0.21111707389354706,
      "learning_rate": 2.515904205530326e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 17805200,
      "step": 19920
    },
    {
      "epoch": 5.258677576877393,
      "grad_norm": 0.1135886162519455,
      "learning_rate": 2.514922476491355e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 17809552,
      "step": 19925
    },
    {
      "epoch": 5.259997360432889,
      "grad_norm": 0.2286888062953949,
      "learning_rate": 2.51394074515115e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 17814064,
      "step": 19930
    },
    {
      "epoch": 5.261317143988386,
      "grad_norm": 0.41446393728256226,
      "learning_rate": 2.5129590116611067e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 17818256,
      "step": 19935
    },
    {
      "epoch": 5.262636927543883,
      "grad_norm": 0.27736788988113403,
      "learning_rate": 2.5119772761726212e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 17822768,
      "step": 19940
    },
    {
      "epoch": 5.26395671109938,
      "grad_norm": 0.03485806658864021,
      "learning_rate": 2.5109955388370893e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 17827312,
      "step": 19945
    },
    {
      "epoch": 5.265276494654876,
      "grad_norm": 0.20494556427001953,
      "learning_rate": 2.510013799805907e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 17831760,
      "step": 19950
    },
    {
      "epoch": 5.266596278210374,
      "grad_norm": 0.27125608921051025,
      "learning_rate": 2.5090320592304706e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 17836336,
      "step": 19955
    },
    {
      "epoch": 5.26791606176587,
      "grad_norm": 0.18720531463623047,
      "learning_rate": 2.5080503172621777e-05,
      "loss": 0.0193,
      "num_input_tokens_seen": 17840720,
      "step": 19960
    },
    {
      "epoch": 5.269235845321368,
      "grad_norm": 0.11960024386644363,
      "learning_rate": 2.5070685740524246e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 17845008,
      "step": 19965
    },
    {
      "epoch": 5.270555628876864,
      "grad_norm": 0.07140699774026871,
      "learning_rate": 2.5060868297526084e-05,
      "loss": 0.0706,
      "num_input_tokens_seen": 17849584,
      "step": 19970
    },
    {
      "epoch": 5.271875412432361,
      "grad_norm": 0.5706459283828735,
      "learning_rate": 2.5051050845141267e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 17853776,
      "step": 19975
    },
    {
      "epoch": 5.273195195987858,
      "grad_norm": 0.22265766561031342,
      "learning_rate": 2.5041233384883765e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 17858512,
      "step": 19980
    },
    {
      "epoch": 5.274514979543355,
      "grad_norm": 0.1970408856868744,
      "learning_rate": 2.5031415918267564e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 17863056,
      "step": 19985
    },
    {
      "epoch": 5.275834763098851,
      "grad_norm": 0.21469996869564056,
      "learning_rate": 2.5021598446806626e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 17867440,
      "step": 19990
    },
    {
      "epoch": 5.277154546654349,
      "grad_norm": 0.1320730745792389,
      "learning_rate": 2.5011780972014937e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 17871920,
      "step": 19995
    },
    {
      "epoch": 5.278474330209845,
      "grad_norm": 0.28926241397857666,
      "learning_rate": 2.5001963495406478e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 17876496,
      "step": 20000
    },
    {
      "epoch": 5.278474330209845,
      "eval_loss": 0.06812906265258789,
      "eval_runtime": 64.813,
      "eval_samples_per_second": 103.914,
      "eval_steps_per_second": 25.982,
      "num_input_tokens_seen": 17876496,
      "step": 20000
    },
    {
      "epoch": 5.279794113765343,
      "grad_norm": 0.12972353398799896,
      "learning_rate": 2.499214601849522e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 17881040,
      "step": 20005
    },
    {
      "epoch": 5.281113897320839,
      "grad_norm": 0.12967248260974884,
      "learning_rate": 2.4982328542795148e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 17885680,
      "step": 20010
    },
    {
      "epoch": 5.282433680876336,
      "grad_norm": 0.2909599840641022,
      "learning_rate": 2.497251106982024e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 17890416,
      "step": 20015
    },
    {
      "epoch": 5.283753464431833,
      "grad_norm": 0.06966737657785416,
      "learning_rate": 2.4962693601084458e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 17894928,
      "step": 20020
    },
    {
      "epoch": 5.28507324798733,
      "grad_norm": 0.15251995623111725,
      "learning_rate": 2.4952876138101794e-05,
      "loss": 0.0153,
      "num_input_tokens_seen": 17899632,
      "step": 20025
    },
    {
      "epoch": 5.286393031542827,
      "grad_norm": 0.49357345700263977,
      "learning_rate": 2.4943058682386233e-05,
      "loss": 0.0975,
      "num_input_tokens_seen": 17904048,
      "step": 20030
    },
    {
      "epoch": 5.287712815098324,
      "grad_norm": 0.2558630108833313,
      "learning_rate": 2.493324123545173e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 17908464,
      "step": 20035
    },
    {
      "epoch": 5.2890325986538205,
      "grad_norm": 0.38097602128982544,
      "learning_rate": 2.4923423798812272e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 17912944,
      "step": 20040
    },
    {
      "epoch": 5.290352382209318,
      "grad_norm": 0.28355035185813904,
      "learning_rate": 2.4913606373981825e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 17917360,
      "step": 20045
    },
    {
      "epoch": 5.2916721657648145,
      "grad_norm": 0.08607621490955353,
      "learning_rate": 2.4903788962474357e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 17921840,
      "step": 20050
    },
    {
      "epoch": 5.292991949320312,
      "grad_norm": 0.0862642303109169,
      "learning_rate": 2.489397156580385e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 17926192,
      "step": 20055
    },
    {
      "epoch": 5.2943117328758085,
      "grad_norm": 0.09222869575023651,
      "learning_rate": 2.4884154185484246e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 17931024,
      "step": 20060
    },
    {
      "epoch": 5.295631516431305,
      "grad_norm": 0.08120808750391006,
      "learning_rate": 2.4874336823029526e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 17935344,
      "step": 20065
    },
    {
      "epoch": 5.2969512999868025,
      "grad_norm": 0.1858755201101303,
      "learning_rate": 2.4864519479953656e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 17939728,
      "step": 20070
    },
    {
      "epoch": 5.298271083542299,
      "grad_norm": 0.06332714110612869,
      "learning_rate": 2.485470215777058e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 17944368,
      "step": 20075
    },
    {
      "epoch": 5.299590867097796,
      "grad_norm": 0.17160849273204803,
      "learning_rate": 2.4844884857994258e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 17948336,
      "step": 20080
    },
    {
      "epoch": 5.300910650653293,
      "grad_norm": 0.20850728452205658,
      "learning_rate": 2.4835067582138638e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 17952496,
      "step": 20085
    },
    {
      "epoch": 5.30223043420879,
      "grad_norm": 0.10861066728830338,
      "learning_rate": 2.4825250331717666e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 17957168,
      "step": 20090
    },
    {
      "epoch": 5.303550217764287,
      "grad_norm": 0.07189831882715225,
      "learning_rate": 2.4815433108245298e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 17962064,
      "step": 20095
    },
    {
      "epoch": 5.304870001319784,
      "grad_norm": 0.21253249049186707,
      "learning_rate": 2.4805615913235456e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 17966480,
      "step": 20100
    },
    {
      "epoch": 5.30618978487528,
      "grad_norm": 0.049017295241355896,
      "learning_rate": 2.479579874820208e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 17970960,
      "step": 20105
    },
    {
      "epoch": 5.307509568430778,
      "grad_norm": 0.04952039197087288,
      "learning_rate": 2.4785981614659115e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 17975440,
      "step": 20110
    },
    {
      "epoch": 5.308829351986274,
      "grad_norm": 0.23236462473869324,
      "learning_rate": 2.477616451412047e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 17979824,
      "step": 20115
    },
    {
      "epoch": 5.310149135541771,
      "grad_norm": 0.2543785274028778,
      "learning_rate": 2.476634744810007e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 17984336,
      "step": 20120
    },
    {
      "epoch": 5.311468919097268,
      "grad_norm": 0.03549548238515854,
      "learning_rate": 2.475653041811183e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 17988816,
      "step": 20125
    },
    {
      "epoch": 5.312788702652765,
      "grad_norm": 0.07516444474458694,
      "learning_rate": 2.4746713425669652e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 17993232,
      "step": 20130
    },
    {
      "epoch": 5.314108486208262,
      "grad_norm": 0.3256787359714508,
      "learning_rate": 2.4736896472287458e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 17997584,
      "step": 20135
    },
    {
      "epoch": 5.315428269763759,
      "grad_norm": 0.3003339469432831,
      "learning_rate": 2.4727079559479124e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 18001936,
      "step": 20140
    },
    {
      "epoch": 5.316748053319255,
      "grad_norm": 0.2503361105918884,
      "learning_rate": 2.4717262688758557e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 18006128,
      "step": 20145
    },
    {
      "epoch": 5.318067836874753,
      "grad_norm": 0.49126237630844116,
      "learning_rate": 2.4707445861639637e-05,
      "loss": 0.0738,
      "num_input_tokens_seen": 18010320,
      "step": 20150
    },
    {
      "epoch": 5.319387620430249,
      "grad_norm": 0.04810823127627373,
      "learning_rate": 2.4697629079636244e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 18014704,
      "step": 20155
    },
    {
      "epoch": 5.320707403985747,
      "grad_norm": 0.06196076050400734,
      "learning_rate": 2.4687812344262244e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 18019152,
      "step": 20160
    },
    {
      "epoch": 5.322027187541243,
      "grad_norm": 0.26994767785072327,
      "learning_rate": 2.46779956570315e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 18023600,
      "step": 20165
    },
    {
      "epoch": 5.32334697109674,
      "grad_norm": 0.170233353972435,
      "learning_rate": 2.466817901945787e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 18028016,
      "step": 20170
    },
    {
      "epoch": 5.324666754652237,
      "grad_norm": 0.03940499201416969,
      "learning_rate": 2.4658362433055217e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 18032592,
      "step": 20175
    },
    {
      "epoch": 5.325986538207734,
      "grad_norm": 0.28275907039642334,
      "learning_rate": 2.4648545899337356e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 18036912,
      "step": 20180
    },
    {
      "epoch": 5.327306321763231,
      "grad_norm": 0.059048958122730255,
      "learning_rate": 2.4638729419818143e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 18041328,
      "step": 20185
    },
    {
      "epoch": 5.328626105318728,
      "grad_norm": 0.3982534110546112,
      "learning_rate": 2.46289129960114e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 18046032,
      "step": 20190
    },
    {
      "epoch": 5.329945888874224,
      "grad_norm": 0.24517352879047394,
      "learning_rate": 2.4619096629430924e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 18050256,
      "step": 20195
    },
    {
      "epoch": 5.331265672429722,
      "grad_norm": 0.2974519431591034,
      "learning_rate": 2.4609280321590543e-05,
      "loss": 0.0185,
      "num_input_tokens_seen": 18054800,
      "step": 20200
    },
    {
      "epoch": 5.331265672429722,
      "eval_loss": 0.06868622452020645,
      "eval_runtime": 64.778,
      "eval_samples_per_second": 103.971,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 18054800,
      "step": 20200
    },
    {
      "epoch": 5.332585455985218,
      "grad_norm": 0.30404600501060486,
      "learning_rate": 2.4599464074004037e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 18059056,
      "step": 20205
    },
    {
      "epoch": 5.333905239540715,
      "grad_norm": 0.39369818568229675,
      "learning_rate": 2.4589647888185204e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 18063504,
      "step": 20210
    },
    {
      "epoch": 5.335225023096212,
      "grad_norm": 0.2401280701160431,
      "learning_rate": 2.4579831765647836e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 18067856,
      "step": 20215
    },
    {
      "epoch": 5.336544806651709,
      "grad_norm": 0.15014079213142395,
      "learning_rate": 2.4570015707905676e-05,
      "loss": 0.1261,
      "num_input_tokens_seen": 18072144,
      "step": 20220
    },
    {
      "epoch": 5.337864590207206,
      "grad_norm": 0.30884531140327454,
      "learning_rate": 2.4560199716472508e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 18076688,
      "step": 20225
    },
    {
      "epoch": 5.339184373762703,
      "grad_norm": 0.046297959983348846,
      "learning_rate": 2.455038379286207e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 18081104,
      "step": 20230
    },
    {
      "epoch": 5.3405041573181995,
      "grad_norm": 0.2118537724018097,
      "learning_rate": 2.4540567938588095e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 18085680,
      "step": 20235
    },
    {
      "epoch": 5.341823940873697,
      "grad_norm": 0.08932926505804062,
      "learning_rate": 2.4530752155164328e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 18089808,
      "step": 20240
    },
    {
      "epoch": 5.3431437244291935,
      "grad_norm": 0.24854230880737305,
      "learning_rate": 2.4520936444104463e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 18094224,
      "step": 20245
    },
    {
      "epoch": 5.34446350798469,
      "grad_norm": 0.23663578927516937,
      "learning_rate": 2.4511120806922218e-05,
      "loss": 0.0388,
      "num_input_tokens_seen": 18098640,
      "step": 20250
    },
    {
      "epoch": 5.3457832915401875,
      "grad_norm": 0.28054967522621155,
      "learning_rate": 2.45013052451313e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 18103088,
      "step": 20255
    },
    {
      "epoch": 5.347103075095684,
      "grad_norm": 0.14086765050888062,
      "learning_rate": 2.4491489760245376e-05,
      "loss": 0.0706,
      "num_input_tokens_seen": 18107536,
      "step": 20260
    },
    {
      "epoch": 5.3484228586511815,
      "grad_norm": 0.3821715712547302,
      "learning_rate": 2.4481674353778115e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 18111824,
      "step": 20265
    },
    {
      "epoch": 5.349742642206678,
      "grad_norm": 0.3472025394439697,
      "learning_rate": 2.447185902724319e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 18116048,
      "step": 20270
    },
    {
      "epoch": 5.351062425762175,
      "grad_norm": 0.4772235155105591,
      "learning_rate": 2.4462043782154233e-05,
      "loss": 0.0801,
      "num_input_tokens_seen": 18120368,
      "step": 20275
    },
    {
      "epoch": 5.352382209317672,
      "grad_norm": 0.2287965565919876,
      "learning_rate": 2.4452228620024895e-05,
      "loss": 0.1044,
      "num_input_tokens_seen": 18124880,
      "step": 20280
    },
    {
      "epoch": 5.353701992873169,
      "grad_norm": 0.052422914654016495,
      "learning_rate": 2.4442413542368776e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 18129456,
      "step": 20285
    },
    {
      "epoch": 5.355021776428666,
      "grad_norm": 0.34070533514022827,
      "learning_rate": 2.4432598550699502e-05,
      "loss": 0.0809,
      "num_input_tokens_seen": 18133616,
      "step": 20290
    },
    {
      "epoch": 5.356341559984163,
      "grad_norm": 0.3570539355278015,
      "learning_rate": 2.4422783646530663e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 18138032,
      "step": 20295
    },
    {
      "epoch": 5.357661343539659,
      "grad_norm": 0.20943960547447205,
      "learning_rate": 2.441296883137584e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 18142480,
      "step": 20300
    },
    {
      "epoch": 5.358981127095157,
      "grad_norm": 0.1433006376028061,
      "learning_rate": 2.4403154106748592e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 18147184,
      "step": 20305
    },
    {
      "epoch": 5.360300910650653,
      "grad_norm": 0.20047394931316376,
      "learning_rate": 2.4393339474162494e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 18151600,
      "step": 20310
    },
    {
      "epoch": 5.361620694206151,
      "grad_norm": 0.154203400015831,
      "learning_rate": 2.4383524935131062e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 18155792,
      "step": 20315
    },
    {
      "epoch": 5.362940477761647,
      "grad_norm": 0.1663346141576767,
      "learning_rate": 2.437371049116784e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 18159952,
      "step": 20320
    },
    {
      "epoch": 5.364260261317144,
      "grad_norm": 0.24978007376194,
      "learning_rate": 2.436389614378632e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 18164272,
      "step": 20325
    },
    {
      "epoch": 5.365580044872641,
      "grad_norm": 0.23309314250946045,
      "learning_rate": 2.435408189450002e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 18168976,
      "step": 20330
    },
    {
      "epoch": 5.366899828428138,
      "grad_norm": 0.24592380225658417,
      "learning_rate": 2.4344267744822406e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 18173456,
      "step": 20335
    },
    {
      "epoch": 5.368219611983634,
      "grad_norm": 0.26433998346328735,
      "learning_rate": 2.4334453696266944e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 18177936,
      "step": 20340
    },
    {
      "epoch": 5.369539395539132,
      "grad_norm": 0.2110581248998642,
      "learning_rate": 2.432463975034708e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 18182320,
      "step": 20345
    },
    {
      "epoch": 5.370859179094628,
      "grad_norm": 0.08799154311418533,
      "learning_rate": 2.4314825908576265e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 18186768,
      "step": 20350
    },
    {
      "epoch": 5.372178962650126,
      "grad_norm": 0.4860915243625641,
      "learning_rate": 2.4305012172467897e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 18191504,
      "step": 20355
    },
    {
      "epoch": 5.373498746205622,
      "grad_norm": 0.04163866490125656,
      "learning_rate": 2.4295198543535393e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 18195824,
      "step": 20360
    },
    {
      "epoch": 5.374818529761119,
      "grad_norm": 0.05976337939500809,
      "learning_rate": 2.4285385023292124e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 18200176,
      "step": 20365
    },
    {
      "epoch": 5.376138313316616,
      "grad_norm": 0.14654403924942017,
      "learning_rate": 2.427557161325147e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 18204624,
      "step": 20370
    },
    {
      "epoch": 5.377458096872113,
      "grad_norm": 0.21097059547901154,
      "learning_rate": 2.4265758314926778e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 18209136,
      "step": 20375
    },
    {
      "epoch": 5.378777880427609,
      "grad_norm": 0.2892642617225647,
      "learning_rate": 2.4255945129831373e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 18213712,
      "step": 20380
    },
    {
      "epoch": 5.380097663983107,
      "grad_norm": 0.3796842396259308,
      "learning_rate": 2.4246132059478578e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 18218352,
      "step": 20385
    },
    {
      "epoch": 5.381417447538603,
      "grad_norm": 0.15289348363876343,
      "learning_rate": 2.4236319105381706e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 18223088,
      "step": 20390
    },
    {
      "epoch": 5.382737231094101,
      "grad_norm": 0.5039880871772766,
      "learning_rate": 2.422650626905401e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 18227152,
      "step": 20395
    },
    {
      "epoch": 5.384057014649597,
      "grad_norm": 0.13780230283737183,
      "learning_rate": 2.4216693552008785e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 18232176,
      "step": 20400
    },
    {
      "epoch": 5.384057014649597,
      "eval_loss": 0.06840289384126663,
      "eval_runtime": 64.7619,
      "eval_samples_per_second": 103.996,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 18232176,
      "step": 20400
    },
    {
      "epoch": 5.385376798205094,
      "grad_norm": 0.13231101632118225,
      "learning_rate": 2.4206880955759247e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 18236432,
      "step": 20405
    },
    {
      "epoch": 5.386696581760591,
      "grad_norm": 0.1054055392742157,
      "learning_rate": 2.419706848181863e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 18241136,
      "step": 20410
    },
    {
      "epoch": 5.388016365316088,
      "grad_norm": 0.2758007049560547,
      "learning_rate": 2.4187256131700153e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 18245744,
      "step": 20415
    },
    {
      "epoch": 5.389336148871585,
      "grad_norm": 0.08499789983034134,
      "learning_rate": 2.4177443906916985e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 18250160,
      "step": 20420
    },
    {
      "epoch": 5.390655932427082,
      "grad_norm": 0.14825163781642914,
      "learning_rate": 2.4167631808982303e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 18254416,
      "step": 20425
    },
    {
      "epoch": 5.3919757159825785,
      "grad_norm": 0.19929571449756622,
      "learning_rate": 2.4157819839409264e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 18258928,
      "step": 20430
    },
    {
      "epoch": 5.393295499538076,
      "grad_norm": 0.10604074597358704,
      "learning_rate": 2.414800799971098e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 18263344,
      "step": 20435
    },
    {
      "epoch": 5.3946152830935725,
      "grad_norm": 0.17447803914546967,
      "learning_rate": 2.4138196291400582e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 18267696,
      "step": 20440
    },
    {
      "epoch": 5.39593506664907,
      "grad_norm": 0.19025644659996033,
      "learning_rate": 2.412838471599114e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 18272240,
      "step": 20445
    },
    {
      "epoch": 5.3972548502045665,
      "grad_norm": 0.7038541436195374,
      "learning_rate": 2.411857327499572e-05,
      "loss": 0.0853,
      "num_input_tokens_seen": 18276880,
      "step": 20450
    },
    {
      "epoch": 5.398574633760063,
      "grad_norm": 0.2678239345550537,
      "learning_rate": 2.410876196992739e-05,
      "loss": 0.1093,
      "num_input_tokens_seen": 18281456,
      "step": 20455
    },
    {
      "epoch": 5.3998944173155605,
      "grad_norm": 0.10129847377538681,
      "learning_rate": 2.4098950802299156e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 18285776,
      "step": 20460
    },
    {
      "epoch": 5.401214200871057,
      "grad_norm": 0.47548508644104004,
      "learning_rate": 2.4089139773624027e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 18290320,
      "step": 20465
    },
    {
      "epoch": 5.402533984426554,
      "grad_norm": 0.10117387026548386,
      "learning_rate": 2.4079328885415007e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 18294608,
      "step": 20470
    },
    {
      "epoch": 5.403853767982051,
      "grad_norm": 0.05529160052537918,
      "learning_rate": 2.4069518139185036e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 18298928,
      "step": 20475
    },
    {
      "epoch": 5.405173551537548,
      "grad_norm": 0.02918175421655178,
      "learning_rate": 2.405970753644706e-05,
      "loss": 0.099,
      "num_input_tokens_seen": 18303600,
      "step": 20480
    },
    {
      "epoch": 5.406493335093045,
      "grad_norm": 0.34676364064216614,
      "learning_rate": 2.4049897078714e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 18308304,
      "step": 20485
    },
    {
      "epoch": 5.407813118648542,
      "grad_norm": 0.492438405752182,
      "learning_rate": 2.404008676749874e-05,
      "loss": 0.1382,
      "num_input_tokens_seen": 18312912,
      "step": 20490
    },
    {
      "epoch": 5.409132902204038,
      "grad_norm": 0.36976322531700134,
      "learning_rate": 2.403027660431418e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 18317296,
      "step": 20495
    },
    {
      "epoch": 5.410452685759536,
      "grad_norm": 0.11185924708843231,
      "learning_rate": 2.402046659067314e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 18321968,
      "step": 20500
    },
    {
      "epoch": 5.411772469315032,
      "grad_norm": 0.03092212975025177,
      "learning_rate": 2.401065672808847e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 18326448,
      "step": 20505
    },
    {
      "epoch": 5.41309225287053,
      "grad_norm": 0.04172932729125023,
      "learning_rate": 2.400084701807296e-05,
      "loss": 0.0259,
      "num_input_tokens_seen": 18331216,
      "step": 20510
    },
    {
      "epoch": 5.414412036426026,
      "grad_norm": 0.13894903659820557,
      "learning_rate": 2.39910374621394e-05,
      "loss": 0.1127,
      "num_input_tokens_seen": 18336016,
      "step": 20515
    },
    {
      "epoch": 5.415731819981523,
      "grad_norm": 0.35350263118743896,
      "learning_rate": 2.3981228061800544e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 18340368,
      "step": 20520
    },
    {
      "epoch": 5.41705160353702,
      "grad_norm": 0.0676608756184578,
      "learning_rate": 2.3971418818569115e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 18344784,
      "step": 20525
    },
    {
      "epoch": 5.418371387092517,
      "grad_norm": 0.40052247047424316,
      "learning_rate": 2.3961609733957832e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 18349136,
      "step": 20530
    },
    {
      "epoch": 5.419691170648013,
      "grad_norm": 0.44407305121421814,
      "learning_rate": 2.395180080947939e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 18353872,
      "step": 20535
    },
    {
      "epoch": 5.421010954203511,
      "grad_norm": 0.12540219724178314,
      "learning_rate": 2.394199204664642e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 18358256,
      "step": 20540
    },
    {
      "epoch": 5.422330737759007,
      "grad_norm": 0.0647035464644432,
      "learning_rate": 2.3932183446971583e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 18362512,
      "step": 20545
    },
    {
      "epoch": 5.423650521314505,
      "grad_norm": 0.34237101674079895,
      "learning_rate": 2.3922375011967473e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 18367024,
      "step": 20550
    },
    {
      "epoch": 5.424970304870001,
      "grad_norm": 0.057282138615846634,
      "learning_rate": 2.3912566743146676e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 18371568,
      "step": 20555
    },
    {
      "epoch": 5.426290088425498,
      "grad_norm": 0.15275301039218903,
      "learning_rate": 2.390275864202176e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 18375952,
      "step": 20560
    },
    {
      "epoch": 5.427609871980995,
      "grad_norm": 0.15629975497722626,
      "learning_rate": 2.3892950710105243e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 18380528,
      "step": 20565
    },
    {
      "epoch": 5.428929655536492,
      "grad_norm": 0.5885166525840759,
      "learning_rate": 2.3883142948909635e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 18384880,
      "step": 20570
    },
    {
      "epoch": 5.430249439091989,
      "grad_norm": 0.34113261103630066,
      "learning_rate": 2.3873335359947433e-05,
      "loss": 0.0484,
      "num_input_tokens_seen": 18389328,
      "step": 20575
    },
    {
      "epoch": 5.431569222647486,
      "grad_norm": 0.40408459305763245,
      "learning_rate": 2.3863527944731066e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 18393840,
      "step": 20580
    },
    {
      "epoch": 5.432889006202982,
      "grad_norm": 0.09386181086301804,
      "learning_rate": 2.385372070477298e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 18398256,
      "step": 20585
    },
    {
      "epoch": 5.43420878975848,
      "grad_norm": 0.7162038683891296,
      "learning_rate": 2.384391364158556e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 18402608,
      "step": 20590
    },
    {
      "epoch": 5.435528573313976,
      "grad_norm": 0.10104630887508392,
      "learning_rate": 2.3834106756681185e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 18407120,
      "step": 20595
    },
    {
      "epoch": 5.436848356869474,
      "grad_norm": 0.09121514856815338,
      "learning_rate": 2.3824300051572206e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 18411760,
      "step": 20600
    },
    {
      "epoch": 5.436848356869474,
      "eval_loss": 0.06782509386539459,
      "eval_runtime": 64.773,
      "eval_samples_per_second": 103.979,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 18411760,
      "step": 20600
    },
    {
      "epoch": 5.43816814042497,
      "grad_norm": 0.25113779306411743,
      "learning_rate": 2.3814493527770923e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 18416464,
      "step": 20605
    },
    {
      "epoch": 5.439487923980467,
      "grad_norm": 0.0962737500667572,
      "learning_rate": 2.3804687186789637e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 18421136,
      "step": 20610
    },
    {
      "epoch": 5.440807707535964,
      "grad_norm": 0.22451110184192657,
      "learning_rate": 2.379488103014062e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 18425712,
      "step": 20615
    },
    {
      "epoch": 5.442127491091461,
      "grad_norm": 0.26961997151374817,
      "learning_rate": 2.3785075059336086e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 18430480,
      "step": 20620
    },
    {
      "epoch": 5.4434472746469575,
      "grad_norm": 0.15414807200431824,
      "learning_rate": 2.3775269275888248e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 18435056,
      "step": 20625
    },
    {
      "epoch": 5.444767058202455,
      "grad_norm": 0.3021535873413086,
      "learning_rate": 2.3765463681309274e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 18439440,
      "step": 20630
    },
    {
      "epoch": 5.4460868417579515,
      "grad_norm": 0.27356576919555664,
      "learning_rate": 2.3755658277111313e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 18444144,
      "step": 20635
    },
    {
      "epoch": 5.447406625313449,
      "grad_norm": 0.29202792048454285,
      "learning_rate": 2.374585306480649e-05,
      "loss": 0.1203,
      "num_input_tokens_seen": 18448688,
      "step": 20640
    },
    {
      "epoch": 5.4487264088689455,
      "grad_norm": 0.12554433941841125,
      "learning_rate": 2.3736048045906877e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 18453264,
      "step": 20645
    },
    {
      "epoch": 5.450046192424442,
      "grad_norm": 0.28185394406318665,
      "learning_rate": 2.372624322192454e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 18457584,
      "step": 20650
    },
    {
      "epoch": 5.4513659759799395,
      "grad_norm": 0.18556854128837585,
      "learning_rate": 2.3716438594371516e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 18462032,
      "step": 20655
    },
    {
      "epoch": 5.452685759535436,
      "grad_norm": 0.37454450130462646,
      "learning_rate": 2.3706634164759784e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 18466576,
      "step": 20660
    },
    {
      "epoch": 5.454005543090933,
      "grad_norm": 0.1118970513343811,
      "learning_rate": 2.3696829934601323e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 18470928,
      "step": 20665
    },
    {
      "epoch": 5.45532532664643,
      "grad_norm": 0.18228600919246674,
      "learning_rate": 2.3687025905408053e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 18475056,
      "step": 20670
    },
    {
      "epoch": 5.456645110201927,
      "grad_norm": 0.4120979607105255,
      "learning_rate": 2.3677222078691886e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 18479664,
      "step": 20675
    },
    {
      "epoch": 5.457964893757424,
      "grad_norm": 0.2520212233066559,
      "learning_rate": 2.366741845596471e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 18484016,
      "step": 20680
    },
    {
      "epoch": 5.459284677312921,
      "grad_norm": 0.15789498388767242,
      "learning_rate": 2.3657615038738343e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 18488496,
      "step": 20685
    },
    {
      "epoch": 5.460604460868417,
      "grad_norm": 0.2324552834033966,
      "learning_rate": 2.3647811828524614e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 18492880,
      "step": 20690
    },
    {
      "epoch": 5.461924244423915,
      "grad_norm": 0.08744845539331436,
      "learning_rate": 2.363800882683529e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 18497360,
      "step": 20695
    },
    {
      "epoch": 5.463244027979411,
      "grad_norm": 0.2076427936553955,
      "learning_rate": 2.3628206035182125e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 18501904,
      "step": 20700
    },
    {
      "epoch": 5.464563811534909,
      "grad_norm": 0.33741968870162964,
      "learning_rate": 2.361840345507683e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 18506320,
      "step": 20705
    },
    {
      "epoch": 5.465883595090405,
      "grad_norm": 0.1196742057800293,
      "learning_rate": 2.3608601088031073e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 18510640,
      "step": 20710
    },
    {
      "epoch": 5.467203378645902,
      "grad_norm": 0.21732641756534576,
      "learning_rate": 2.3598798935556516e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 18515280,
      "step": 20715
    },
    {
      "epoch": 5.468523162201399,
      "grad_norm": 0.20192818343639374,
      "learning_rate": 2.3588996999164784e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 18519728,
      "step": 20720
    },
    {
      "epoch": 5.469842945756896,
      "grad_norm": 0.226049542427063,
      "learning_rate": 2.3579195280367434e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 18523984,
      "step": 20725
    },
    {
      "epoch": 5.471162729312393,
      "grad_norm": 0.19992336630821228,
      "learning_rate": 2.356939378067603e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 18528560,
      "step": 20730
    },
    {
      "epoch": 5.47248251286789,
      "grad_norm": 0.38813725113868713,
      "learning_rate": 2.3559592501602092e-05,
      "loss": 0.077,
      "num_input_tokens_seen": 18533040,
      "step": 20735
    },
    {
      "epoch": 5.473802296423386,
      "grad_norm": 0.23195841908454895,
      "learning_rate": 2.3549791444657076e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 18537744,
      "step": 20740
    },
    {
      "epoch": 5.475122079978884,
      "grad_norm": 0.11163752526044846,
      "learning_rate": 2.353999061135246e-05,
      "loss": 0.0273,
      "num_input_tokens_seen": 18542000,
      "step": 20745
    },
    {
      "epoch": 5.47644186353438,
      "grad_norm": 0.2662763297557831,
      "learning_rate": 2.3530190003199626e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 18546576,
      "step": 20750
    },
    {
      "epoch": 5.477761647089877,
      "grad_norm": 0.06427902728319168,
      "learning_rate": 2.3520389621709965e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 18551152,
      "step": 20755
    },
    {
      "epoch": 5.479081430645374,
      "grad_norm": 0.17020420730113983,
      "learning_rate": 2.351058946839483e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 18555632,
      "step": 20760
    },
    {
      "epoch": 5.480401214200871,
      "grad_norm": 0.17938990890979767,
      "learning_rate": 2.350078954476551e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 18559920,
      "step": 20765
    },
    {
      "epoch": 5.481720997756368,
      "grad_norm": 0.0953431949019432,
      "learning_rate": 2.3490989852333272e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 18564080,
      "step": 20770
    },
    {
      "epoch": 5.483040781311865,
      "grad_norm": 0.18800514936447144,
      "learning_rate": 2.3481190392609377e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 18568272,
      "step": 20775
    },
    {
      "epoch": 5.4843605648673615,
      "grad_norm": 0.35476887226104736,
      "learning_rate": 2.3471391167105e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 18572720,
      "step": 20780
    },
    {
      "epoch": 5.485680348422859,
      "grad_norm": 0.056558284908533096,
      "learning_rate": 2.3461592177331325e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 18577008,
      "step": 20785
    },
    {
      "epoch": 5.4870001319783555,
      "grad_norm": 0.0970536395907402,
      "learning_rate": 2.345179342479946e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 18581424,
      "step": 20790
    },
    {
      "epoch": 5.488319915533852,
      "grad_norm": 0.25388211011886597,
      "learning_rate": 2.3441994911020503e-05,
      "loss": 0.0281,
      "num_input_tokens_seen": 18586192,
      "step": 20795
    },
    {
      "epoch": 5.4896396990893495,
      "grad_norm": 0.03132476657629013,
      "learning_rate": 2.3432196637505522e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 18590672,
      "step": 20800
    },
    {
      "epoch": 5.4896396990893495,
      "eval_loss": 0.06822258979082108,
      "eval_runtime": 64.7514,
      "eval_samples_per_second": 104.013,
      "eval_steps_per_second": 26.007,
      "num_input_tokens_seen": 18590672,
      "step": 20800
    },
    {
      "epoch": 5.490959482644846,
      "grad_norm": 0.051991067826747894,
      "learning_rate": 2.3422398605765515e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 18595216,
      "step": 20805
    },
    {
      "epoch": 5.4922792662003435,
      "grad_norm": 0.12102445214986801,
      "learning_rate": 2.3412600817311462e-05,
      "loss": 0.1101,
      "num_input_tokens_seen": 18599728,
      "step": 20810
    },
    {
      "epoch": 5.49359904975584,
      "grad_norm": 0.1357477307319641,
      "learning_rate": 2.3402803273654326e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 18604304,
      "step": 20815
    },
    {
      "epoch": 5.494918833311337,
      "grad_norm": 0.2408204823732376,
      "learning_rate": 2.3393005976304983e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 18608944,
      "step": 20820
    },
    {
      "epoch": 5.496238616866834,
      "grad_norm": 0.249470517039299,
      "learning_rate": 2.338320892677432e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 18613552,
      "step": 20825
    },
    {
      "epoch": 5.497558400422331,
      "grad_norm": 0.06542292982339859,
      "learning_rate": 2.3373412126573155e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 18618160,
      "step": 20830
    },
    {
      "epoch": 5.498878183977828,
      "grad_norm": 0.44577333331108093,
      "learning_rate": 2.3363615577212285e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 18622320,
      "step": 20835
    },
    {
      "epoch": 5.500197967533325,
      "grad_norm": 0.46828821301460266,
      "learning_rate": 2.3353819280202455e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 18626896,
      "step": 20840
    },
    {
      "epoch": 5.501517751088821,
      "grad_norm": 0.32352742552757263,
      "learning_rate": 2.334402323705438e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 18631088,
      "step": 20845
    },
    {
      "epoch": 5.502837534644319,
      "grad_norm": 0.21690423786640167,
      "learning_rate": 2.3334227449278725e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 18635504,
      "step": 20850
    },
    {
      "epoch": 5.504157318199815,
      "grad_norm": 0.05946047604084015,
      "learning_rate": 2.3324431918386143e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 18639984,
      "step": 20855
    },
    {
      "epoch": 5.505477101755313,
      "grad_norm": 0.2893528342247009,
      "learning_rate": 2.3314636645887207e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 18644656,
      "step": 20860
    },
    {
      "epoch": 5.506796885310809,
      "grad_norm": 0.08694219589233398,
      "learning_rate": 2.3304841633292487e-05,
      "loss": 0.1027,
      "num_input_tokens_seen": 18648944,
      "step": 20865
    },
    {
      "epoch": 5.508116668866306,
      "grad_norm": 0.42229074239730835,
      "learning_rate": 2.329504688211248e-05,
      "loss": 0.1129,
      "num_input_tokens_seen": 18653168,
      "step": 20870
    },
    {
      "epoch": 5.509436452421803,
      "grad_norm": 0.19916792213916779,
      "learning_rate": 2.3285252393857677e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 18657392,
      "step": 20875
    },
    {
      "epoch": 5.5107562359773,
      "grad_norm": 0.09439694136381149,
      "learning_rate": 2.327545817003851e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 18661840,
      "step": 20880
    },
    {
      "epoch": 5.512076019532796,
      "grad_norm": 0.10494999587535858,
      "learning_rate": 2.326566421216535e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 18666320,
      "step": 20885
    },
    {
      "epoch": 5.513395803088294,
      "grad_norm": 0.03207461163401604,
      "learning_rate": 2.3255870521748565e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 18670800,
      "step": 20890
    },
    {
      "epoch": 5.51471558664379,
      "grad_norm": 0.1524760127067566,
      "learning_rate": 2.3246077100298474e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 18675472,
      "step": 20895
    },
    {
      "epoch": 5.516035370199288,
      "grad_norm": 0.23691290616989136,
      "learning_rate": 2.3236283949325328e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 18679856,
      "step": 20900
    },
    {
      "epoch": 5.517355153754784,
      "grad_norm": 0.2641846239566803,
      "learning_rate": 2.3226491070339368e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 18684304,
      "step": 20905
    },
    {
      "epoch": 5.518674937310281,
      "grad_norm": 0.09911259263753891,
      "learning_rate": 2.3216698464850762e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 18688656,
      "step": 20910
    },
    {
      "epoch": 5.519994720865778,
      "grad_norm": 0.18849235773086548,
      "learning_rate": 2.320690613436967e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 18693008,
      "step": 20915
    },
    {
      "epoch": 5.521314504421275,
      "grad_norm": 0.0780048817396164,
      "learning_rate": 2.3197114080406192e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 18697584,
      "step": 20920
    },
    {
      "epoch": 5.522634287976771,
      "grad_norm": 0.45181116461753845,
      "learning_rate": 2.3187322304470365e-05,
      "loss": 0.0994,
      "num_input_tokens_seen": 18702160,
      "step": 20925
    },
    {
      "epoch": 5.523954071532269,
      "grad_norm": 0.11812064796686172,
      "learning_rate": 2.3177530808072222e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 18706224,
      "step": 20930
    },
    {
      "epoch": 5.525273855087765,
      "grad_norm": 0.33413559198379517,
      "learning_rate": 2.316773959272174e-05,
      "loss": 0.11,
      "num_input_tokens_seen": 18710800,
      "step": 20935
    },
    {
      "epoch": 5.526593638643263,
      "grad_norm": 0.06064317375421524,
      "learning_rate": 2.3157948659928823e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 18715376,
      "step": 20940
    },
    {
      "epoch": 5.527913422198759,
      "grad_norm": 0.11536934971809387,
      "learning_rate": 2.3148158011203388e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 18719792,
      "step": 20945
    },
    {
      "epoch": 5.529233205754256,
      "grad_norm": 0.12524516880512238,
      "learning_rate": 2.3138367648055253e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 18724336,
      "step": 20950
    },
    {
      "epoch": 5.530552989309753,
      "grad_norm": 0.20421750843524933,
      "learning_rate": 2.312857757199422e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 18729072,
      "step": 20955
    },
    {
      "epoch": 5.53187277286525,
      "grad_norm": 0.35203099250793457,
      "learning_rate": 2.3118787784530048e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 18733936,
      "step": 20960
    },
    {
      "epoch": 5.533192556420747,
      "grad_norm": 0.04522039741277695,
      "learning_rate": 2.310899828717243e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 18738416,
      "step": 20965
    },
    {
      "epoch": 5.534512339976244,
      "grad_norm": 0.3366948962211609,
      "learning_rate": 2.309920908143104e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 18742864,
      "step": 20970
    },
    {
      "epoch": 5.5358321235317405,
      "grad_norm": 0.019371913745999336,
      "learning_rate": 2.308942016881551e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 18747312,
      "step": 20975
    },
    {
      "epoch": 5.537151907087238,
      "grad_norm": 0.17780911922454834,
      "learning_rate": 2.307963155083539e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 18751824,
      "step": 20980
    },
    {
      "epoch": 5.5384716906427345,
      "grad_norm": 0.05368199571967125,
      "learning_rate": 2.306984322900022e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 18756208,
      "step": 20985
    },
    {
      "epoch": 5.539791474198232,
      "grad_norm": 0.03777335211634636,
      "learning_rate": 2.3060055204819482e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 18760720,
      "step": 20990
    },
    {
      "epoch": 5.5411112577537285,
      "grad_norm": 0.10467682033777237,
      "learning_rate": 2.3050267479802604e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 18765360,
      "step": 20995
    },
    {
      "epoch": 5.542431041309225,
      "grad_norm": 0.19103550910949707,
      "learning_rate": 2.304048005545899e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 18770000,
      "step": 21000
    },
    {
      "epoch": 5.542431041309225,
      "eval_loss": 0.06886310875415802,
      "eval_runtime": 64.8064,
      "eval_samples_per_second": 103.925,
      "eval_steps_per_second": 25.985,
      "num_input_tokens_seen": 18770000,
      "step": 21000
    },
    {
      "epoch": 5.5437508248647225,
      "grad_norm": 0.045127011835575104,
      "learning_rate": 2.3030692933297972e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 18774704,
      "step": 21005
    },
    {
      "epoch": 5.545070608420219,
      "grad_norm": 0.04276011884212494,
      "learning_rate": 2.3020906114828843e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 18779184,
      "step": 21010
    },
    {
      "epoch": 5.546390391975716,
      "grad_norm": 0.17357252538204193,
      "learning_rate": 2.301111960156088e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 18783536,
      "step": 21015
    },
    {
      "epoch": 5.547710175531213,
      "grad_norm": 0.11120226979255676,
      "learning_rate": 2.300133339500326e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 18788048,
      "step": 21020
    },
    {
      "epoch": 5.54902995908671,
      "grad_norm": 0.1699865758419037,
      "learning_rate": 2.2991547496665148e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 18792624,
      "step": 21025
    },
    {
      "epoch": 5.550349742642207,
      "grad_norm": 0.2163824588060379,
      "learning_rate": 2.298176190805565e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 18797264,
      "step": 21030
    },
    {
      "epoch": 5.551669526197704,
      "grad_norm": 0.4397863447666168,
      "learning_rate": 2.2971976630683826e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 18801584,
      "step": 21035
    },
    {
      "epoch": 5.5529893097532,
      "grad_norm": 0.37010493874549866,
      "learning_rate": 2.29621916660587e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 18806192,
      "step": 21040
    },
    {
      "epoch": 5.554309093308698,
      "grad_norm": 0.15090349316596985,
      "learning_rate": 2.295240701568922e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 18810576,
      "step": 21045
    },
    {
      "epoch": 5.555628876864194,
      "grad_norm": 0.44576624035835266,
      "learning_rate": 2.2942622681084312e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 18814896,
      "step": 21050
    },
    {
      "epoch": 5.556948660419691,
      "grad_norm": 0.3156450390815735,
      "learning_rate": 2.293283866375284e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 18819152,
      "step": 21055
    },
    {
      "epoch": 5.558268443975188,
      "grad_norm": 0.13547496497631073,
      "learning_rate": 2.2923054965203627e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 18823472,
      "step": 21060
    },
    {
      "epoch": 5.559588227530685,
      "grad_norm": 0.2864745855331421,
      "learning_rate": 2.2913271586945443e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 18828112,
      "step": 21065
    },
    {
      "epoch": 5.560908011086182,
      "grad_norm": 0.08022721111774445,
      "learning_rate": 2.290348853048699e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 18832592,
      "step": 21070
    },
    {
      "epoch": 5.562227794641679,
      "grad_norm": 0.1406233161687851,
      "learning_rate": 2.2893705797336956e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 18837328,
      "step": 21075
    },
    {
      "epoch": 5.563547578197175,
      "grad_norm": 0.29012757539749146,
      "learning_rate": 2.288392338900397e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 18841936,
      "step": 21080
    },
    {
      "epoch": 5.564867361752673,
      "grad_norm": 0.15457555651664734,
      "learning_rate": 2.2874141306996576e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 18846096,
      "step": 21085
    },
    {
      "epoch": 5.566187145308169,
      "grad_norm": 0.09114647656679153,
      "learning_rate": 2.2864359552823312e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 18850352,
      "step": 21090
    },
    {
      "epoch": 5.567506928863667,
      "grad_norm": 0.28845515847206116,
      "learning_rate": 2.2854578127992648e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 18854544,
      "step": 21095
    },
    {
      "epoch": 5.568826712419163,
      "grad_norm": 0.014079189859330654,
      "learning_rate": 2.2844797034012988e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 18858928,
      "step": 21100
    },
    {
      "epoch": 5.57014649597466,
      "grad_norm": 0.191579207777977,
      "learning_rate": 2.2835016272392722e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 18863440,
      "step": 21105
    },
    {
      "epoch": 5.571466279530157,
      "grad_norm": 0.24062155187129974,
      "learning_rate": 2.2825235844640142e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 18867728,
      "step": 21110
    },
    {
      "epoch": 5.572786063085654,
      "grad_norm": 0.0639524832367897,
      "learning_rate": 2.2815455752263522e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 18872400,
      "step": 21115
    },
    {
      "epoch": 5.574105846641151,
      "grad_norm": 0.10769195854663849,
      "learning_rate": 2.2805675996771092e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 18876784,
      "step": 21120
    },
    {
      "epoch": 5.575425630196648,
      "grad_norm": 0.1816895604133606,
      "learning_rate": 2.2795896579670987e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 18881008,
      "step": 21125
    },
    {
      "epoch": 5.576745413752144,
      "grad_norm": 0.2661650776863098,
      "learning_rate": 2.2786117502471337e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 18885808,
      "step": 21130
    },
    {
      "epoch": 5.578065197307642,
      "grad_norm": 0.4620969593524933,
      "learning_rate": 2.2776338766680185e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 18890192,
      "step": 21135
    },
    {
      "epoch": 5.579384980863138,
      "grad_norm": 0.17995209991931915,
      "learning_rate": 2.2766560373805533e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 18894512,
      "step": 21140
    },
    {
      "epoch": 5.580704764418636,
      "grad_norm": 0.1734919399023056,
      "learning_rate": 2.2756782325355353e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 18898768,
      "step": 21145
    },
    {
      "epoch": 5.582024547974132,
      "grad_norm": 0.11131265014410019,
      "learning_rate": 2.2747004622837514e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 18903344,
      "step": 21150
    },
    {
      "epoch": 5.583344331529629,
      "grad_norm": 0.44984909892082214,
      "learning_rate": 2.2737227267759878e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 18907920,
      "step": 21155
    },
    {
      "epoch": 5.584664115085126,
      "grad_norm": 0.04744124785065651,
      "learning_rate": 2.272745026163024e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 18912432,
      "step": 21160
    },
    {
      "epoch": 5.585983898640623,
      "grad_norm": 0.24574008584022522,
      "learning_rate": 2.271767360595633e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 18916688,
      "step": 21165
    },
    {
      "epoch": 5.5873036821961195,
      "grad_norm": 0.15691885352134705,
      "learning_rate": 2.270789730224583e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 18921072,
      "step": 21170
    },
    {
      "epoch": 5.588623465751617,
      "grad_norm": 0.043993137776851654,
      "learning_rate": 2.2698121352006367e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 18925360,
      "step": 21175
    },
    {
      "epoch": 5.5899432493071135,
      "grad_norm": 0.0546710267663002,
      "learning_rate": 2.2688345756745517e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 18929840,
      "step": 21180
    },
    {
      "epoch": 5.59126303286261,
      "grad_norm": 0.1551498919725418,
      "learning_rate": 2.267857051797081e-05,
      "loss": 0.0873,
      "num_input_tokens_seen": 18934128,
      "step": 21185
    },
    {
      "epoch": 5.5925828164181075,
      "grad_norm": 0.26744258403778076,
      "learning_rate": 2.2668795637189695e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 18938544,
      "step": 21190
    },
    {
      "epoch": 5.593902599973604,
      "grad_norm": 0.2521877586841583,
      "learning_rate": 2.2659021115909586e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 18943344,
      "step": 21195
    },
    {
      "epoch": 5.5952223835291015,
      "grad_norm": 0.34809041023254395,
      "learning_rate": 2.2649246955637847e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 18947664,
      "step": 21200
    },
    {
      "epoch": 5.5952223835291015,
      "eval_loss": 0.06824265420436859,
      "eval_runtime": 64.7426,
      "eval_samples_per_second": 104.027,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 18947664,
      "step": 21200
    },
    {
      "epoch": 5.596542167084598,
      "grad_norm": 0.08779877424240112,
      "learning_rate": 2.2639473157881766e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 18952176,
      "step": 21205
    },
    {
      "epoch": 5.597861950640095,
      "grad_norm": 0.28717777132987976,
      "learning_rate": 2.2629699724148594e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 18956720,
      "step": 21210
    },
    {
      "epoch": 5.599181734195592,
      "grad_norm": 0.15190094709396362,
      "learning_rate": 2.26199266559455e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 18961200,
      "step": 21215
    },
    {
      "epoch": 5.600501517751089,
      "grad_norm": 0.13326530158519745,
      "learning_rate": 2.2610153954779625e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 18965616,
      "step": 21220
    },
    {
      "epoch": 5.601821301306586,
      "grad_norm": 0.0784735456109047,
      "learning_rate": 2.2600381622158056e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 18970448,
      "step": 21225
    },
    {
      "epoch": 5.603141084862083,
      "grad_norm": 0.35813286900520325,
      "learning_rate": 2.2590609659587783e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 18975024,
      "step": 21230
    },
    {
      "epoch": 5.604460868417579,
      "grad_norm": 0.2475057691335678,
      "learning_rate": 2.2580838068575787e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 18979504,
      "step": 21235
    },
    {
      "epoch": 5.605780651973077,
      "grad_norm": 0.06599078327417374,
      "learning_rate": 2.257106685062896e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 18983856,
      "step": 21240
    },
    {
      "epoch": 5.607100435528573,
      "grad_norm": 0.24672354757785797,
      "learning_rate": 2.256129600725415e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 18988400,
      "step": 21245
    },
    {
      "epoch": 5.608420219084071,
      "grad_norm": 0.20182178914546967,
      "learning_rate": 2.2551525539958145e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 18993136,
      "step": 21250
    },
    {
      "epoch": 5.609740002639567,
      "grad_norm": 0.4448821246623993,
      "learning_rate": 2.2541755450247663e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 18997552,
      "step": 21255
    },
    {
      "epoch": 5.611059786195064,
      "grad_norm": 0.22393456101417542,
      "learning_rate": 2.2531985739629382e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 19001936,
      "step": 21260
    },
    {
      "epoch": 5.612379569750561,
      "grad_norm": 0.2718956470489502,
      "learning_rate": 2.2522216409609924e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 19006320,
      "step": 21265
    },
    {
      "epoch": 5.613699353306058,
      "grad_norm": 0.21783769130706787,
      "learning_rate": 2.2512447461695826e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 19010544,
      "step": 21270
    },
    {
      "epoch": 5.615019136861555,
      "grad_norm": 0.23383860290050507,
      "learning_rate": 2.2502678897393593e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 19014896,
      "step": 21275
    },
    {
      "epoch": 5.616338920417052,
      "grad_norm": 0.06760753691196442,
      "learning_rate": 2.2492910718209665e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 19019632,
      "step": 21280
    },
    {
      "epoch": 5.617658703972548,
      "grad_norm": 0.17811067402362823,
      "learning_rate": 2.2483142925650398e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 19024240,
      "step": 21285
    },
    {
      "epoch": 5.618978487528046,
      "grad_norm": 0.11221028119325638,
      "learning_rate": 2.247337552122213e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 19029008,
      "step": 21290
    },
    {
      "epoch": 5.620298271083542,
      "grad_norm": 0.4110196530818939,
      "learning_rate": 2.24636085064311e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 19033584,
      "step": 21295
    },
    {
      "epoch": 5.621618054639039,
      "grad_norm": 0.18568487465381622,
      "learning_rate": 2.245384188278351e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 19038224,
      "step": 21300
    },
    {
      "epoch": 5.622937838194536,
      "grad_norm": 0.20742888748645782,
      "learning_rate": 2.2444075651785513e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 19042384,
      "step": 21305
    },
    {
      "epoch": 5.624257621750033,
      "grad_norm": 0.6016423106193542,
      "learning_rate": 2.243430981494316e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 19046896,
      "step": 21310
    },
    {
      "epoch": 5.625577405305529,
      "grad_norm": 0.6026350259780884,
      "learning_rate": 2.2424544373762475e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 19051440,
      "step": 21315
    },
    {
      "epoch": 5.626897188861027,
      "grad_norm": 0.04013843461871147,
      "learning_rate": 2.2414779329749418e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 19055760,
      "step": 21320
    },
    {
      "epoch": 5.628216972416523,
      "grad_norm": 0.3939080238342285,
      "learning_rate": 2.2405014684409873e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 19060240,
      "step": 21325
    },
    {
      "epoch": 5.629536755972021,
      "grad_norm": 0.18597635626792908,
      "learning_rate": 2.239525043924968e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 19064336,
      "step": 21330
    },
    {
      "epoch": 5.630856539527517,
      "grad_norm": 0.30208033323287964,
      "learning_rate": 2.2385486595774592e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 19068720,
      "step": 21335
    },
    {
      "epoch": 5.632176323083014,
      "grad_norm": 0.09490006417036057,
      "learning_rate": 2.237572315549033e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 19073360,
      "step": 21340
    },
    {
      "epoch": 5.633496106638511,
      "grad_norm": 0.5111339688301086,
      "learning_rate": 2.2365960119902545e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 19077776,
      "step": 21345
    },
    {
      "epoch": 5.634815890194008,
      "grad_norm": 0.17606085538864136,
      "learning_rate": 2.2356197490516806e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 19082064,
      "step": 21350
    },
    {
      "epoch": 5.636135673749505,
      "grad_norm": 0.24307148158550262,
      "learning_rate": 2.234643526883863e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 19086576,
      "step": 21355
    },
    {
      "epoch": 5.637455457305002,
      "grad_norm": 0.4026353657245636,
      "learning_rate": 2.2336673456373497e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 19091056,
      "step": 21360
    },
    {
      "epoch": 5.6387752408604985,
      "grad_norm": 0.2667510509490967,
      "learning_rate": 2.2326912054626772e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 19095568,
      "step": 21365
    },
    {
      "epoch": 5.640095024415996,
      "grad_norm": 0.045207541435956955,
      "learning_rate": 2.2317151065103813e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 19100304,
      "step": 21370
    },
    {
      "epoch": 5.6414148079714925,
      "grad_norm": 0.2859058082103729,
      "learning_rate": 2.2307390489309865e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 19105072,
      "step": 21375
    },
    {
      "epoch": 5.64273459152699,
      "grad_norm": 0.08890268206596375,
      "learning_rate": 2.2297630328750146e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 19109488,
      "step": 21380
    },
    {
      "epoch": 5.6440543750824865,
      "grad_norm": 0.08340633660554886,
      "learning_rate": 2.228787058492979e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 19114128,
      "step": 21385
    },
    {
      "epoch": 5.645374158637983,
      "grad_norm": 0.20061054825782776,
      "learning_rate": 2.2278111259353875e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 19118704,
      "step": 21390
    },
    {
      "epoch": 5.6466939421934805,
      "grad_norm": 0.047674551606178284,
      "learning_rate": 2.2268352353527395e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 19123088,
      "step": 21395
    },
    {
      "epoch": 5.648013725748977,
      "grad_norm": 0.13338921964168549,
      "learning_rate": 2.225859386895533e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 19127344,
      "step": 21400
    },
    {
      "epoch": 5.648013725748977,
      "eval_loss": 0.06848660111427307,
      "eval_runtime": 64.7802,
      "eval_samples_per_second": 103.967,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 19127344,
      "step": 21400
    },
    {
      "epoch": 5.6493335093044745,
      "grad_norm": 0.4650416672229767,
      "learning_rate": 2.2248835807142525e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 19131856,
      "step": 21405
    },
    {
      "epoch": 5.650653292859971,
      "grad_norm": 0.06632301211357117,
      "learning_rate": 2.2239078169593826e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 19136400,
      "step": 21410
    },
    {
      "epoch": 5.651973076415468,
      "grad_norm": 0.15004661679267883,
      "learning_rate": 2.222932095781396e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 19140944,
      "step": 21415
    },
    {
      "epoch": 5.653292859970965,
      "grad_norm": 0.1269366443157196,
      "learning_rate": 2.221956417330762e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 19145520,
      "step": 21420
    },
    {
      "epoch": 5.654612643526462,
      "grad_norm": 0.11003465950489044,
      "learning_rate": 2.2209807817579438e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 19149680,
      "step": 21425
    },
    {
      "epoch": 5.655932427081958,
      "grad_norm": 0.5999953150749207,
      "learning_rate": 2.220005189213394e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 19154032,
      "step": 21430
    },
    {
      "epoch": 5.657252210637456,
      "grad_norm": 0.07152295857667923,
      "learning_rate": 2.2190296398475624e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 19158704,
      "step": 21435
    },
    {
      "epoch": 5.658571994192952,
      "grad_norm": 0.2252178192138672,
      "learning_rate": 2.2180541338108926e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 19163280,
      "step": 21440
    },
    {
      "epoch": 5.659891777748449,
      "grad_norm": 0.3121447265148163,
      "learning_rate": 2.2170786712538176e-05,
      "loss": 0.1096,
      "num_input_tokens_seen": 19167632,
      "step": 21445
    },
    {
      "epoch": 5.661211561303946,
      "grad_norm": 0.5204066634178162,
      "learning_rate": 2.216103252326768e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 19171952,
      "step": 21450
    },
    {
      "epoch": 5.662531344859443,
      "grad_norm": 0.19713424146175385,
      "learning_rate": 2.2151278771801635e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 19176336,
      "step": 21455
    },
    {
      "epoch": 5.66385112841494,
      "grad_norm": 0.17127105593681335,
      "learning_rate": 2.21415254596442e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 19180880,
      "step": 21460
    },
    {
      "epoch": 5.665170911970437,
      "grad_norm": 0.23283417522907257,
      "learning_rate": 2.213177258829947e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 19185584,
      "step": 21465
    },
    {
      "epoch": 5.666490695525933,
      "grad_norm": 0.21299847960472107,
      "learning_rate": 2.2122020159271445e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 19189936,
      "step": 21470
    },
    {
      "epoch": 5.667810479081431,
      "grad_norm": 0.21029940247535706,
      "learning_rate": 2.2112268174064075e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 19194448,
      "step": 21475
    },
    {
      "epoch": 5.669130262636927,
      "grad_norm": 0.14166143536567688,
      "learning_rate": 2.2102516634181253e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 19198864,
      "step": 21480
    },
    {
      "epoch": 5.670450046192425,
      "grad_norm": 0.17571035027503967,
      "learning_rate": 2.209276554112677e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 19203280,
      "step": 21485
    },
    {
      "epoch": 5.671769829747921,
      "grad_norm": 0.13505153357982635,
      "learning_rate": 2.2083014896404384e-05,
      "loss": 0.0259,
      "num_input_tokens_seen": 19208176,
      "step": 21490
    },
    {
      "epoch": 5.673089613303418,
      "grad_norm": 0.13763467967510223,
      "learning_rate": 2.207326470151775e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 19213040,
      "step": 21495
    },
    {
      "epoch": 5.674409396858915,
      "grad_norm": 0.24686755239963531,
      "learning_rate": 2.2063514957970477e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 19217744,
      "step": 21500
    },
    {
      "epoch": 5.675729180414412,
      "grad_norm": 0.04710644856095314,
      "learning_rate": 2.205376566726611e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 19222192,
      "step": 21505
    },
    {
      "epoch": 5.677048963969909,
      "grad_norm": 0.16848784685134888,
      "learning_rate": 2.204401683090809e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 19227152,
      "step": 21510
    },
    {
      "epoch": 5.678368747525406,
      "grad_norm": 0.22841808199882507,
      "learning_rate": 2.203426845039982e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 19231472,
      "step": 21515
    },
    {
      "epoch": 5.679688531080902,
      "grad_norm": 0.23605351150035858,
      "learning_rate": 2.202452052724464e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 19235888,
      "step": 21520
    },
    {
      "epoch": 5.6810083146364,
      "grad_norm": 0.3673918545246124,
      "learning_rate": 2.2014773062945777e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 19240464,
      "step": 21525
    },
    {
      "epoch": 5.682328098191896,
      "grad_norm": 0.3230617344379425,
      "learning_rate": 2.2005026059006427e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 19244880,
      "step": 21530
    },
    {
      "epoch": 5.683647881747394,
      "grad_norm": 0.13052794337272644,
      "learning_rate": 2.1995279516929695e-05,
      "loss": 0.0399,
      "num_input_tokens_seen": 19249008,
      "step": 21535
    },
    {
      "epoch": 5.68496766530289,
      "grad_norm": 0.7355924248695374,
      "learning_rate": 2.1985533438218613e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 19253328,
      "step": 21540
    },
    {
      "epoch": 5.686287448858387,
      "grad_norm": 0.29957613348960876,
      "learning_rate": 2.197578782437617e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 19258064,
      "step": 21545
    },
    {
      "epoch": 5.687607232413884,
      "grad_norm": 0.18086326122283936,
      "learning_rate": 2.196604267690524e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 19262320,
      "step": 21550
    },
    {
      "epoch": 5.688927015969381,
      "grad_norm": 0.16441041231155396,
      "learning_rate": 2.195629799730865e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 19266704,
      "step": 21555
    },
    {
      "epoch": 5.6902467995248776,
      "grad_norm": 0.4081951975822449,
      "learning_rate": 2.1946553787089173e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 19271504,
      "step": 21560
    },
    {
      "epoch": 5.691566583080375,
      "grad_norm": 0.05400290712714195,
      "learning_rate": 2.193681004774947e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 19275696,
      "step": 21565
    },
    {
      "epoch": 5.6928863666358716,
      "grad_norm": 0.16085407137870789,
      "learning_rate": 2.1927066780792154e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 19280272,
      "step": 21570
    },
    {
      "epoch": 5.694206150191369,
      "grad_norm": 0.1212693378329277,
      "learning_rate": 2.191732398771975e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 19284720,
      "step": 21575
    },
    {
      "epoch": 5.6955259337468656,
      "grad_norm": 0.1213379055261612,
      "learning_rate": 2.1907581670034725e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 19289040,
      "step": 21580
    },
    {
      "epoch": 5.696845717302362,
      "grad_norm": 0.040082138031721115,
      "learning_rate": 2.189783982923948e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 19293552,
      "step": 21585
    },
    {
      "epoch": 5.6981655008578596,
      "grad_norm": 0.0658331885933876,
      "learning_rate": 2.1888098466836303e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 19297808,
      "step": 21590
    },
    {
      "epoch": 5.699485284413356,
      "grad_norm": 0.18205708265304565,
      "learning_rate": 2.1878357584327457e-05,
      "loss": 0.0241,
      "num_input_tokens_seen": 19302352,
      "step": 21595
    },
    {
      "epoch": 5.700805067968853,
      "grad_norm": 0.1503007858991623,
      "learning_rate": 2.1868617183215103e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 19306864,
      "step": 21600
    },
    {
      "epoch": 5.700805067968853,
      "eval_loss": 0.06780818849802017,
      "eval_runtime": 64.7836,
      "eval_samples_per_second": 103.962,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 19306864,
      "step": 21600
    },
    {
      "epoch": 5.70212485152435,
      "grad_norm": 0.09879620373249054,
      "learning_rate": 2.1858877265001327e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 19310992,
      "step": 21605
    },
    {
      "epoch": 5.703444635079847,
      "grad_norm": 0.4708520472049713,
      "learning_rate": 2.184913783118816e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 19315632,
      "step": 21610
    },
    {
      "epoch": 5.704764418635344,
      "grad_norm": 0.22414879500865936,
      "learning_rate": 2.1839398883277522e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 19320080,
      "step": 21615
    },
    {
      "epoch": 5.706084202190841,
      "grad_norm": 0.27883031964302063,
      "learning_rate": 2.182966042277129e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 19324176,
      "step": 21620
    },
    {
      "epoch": 5.707403985746337,
      "grad_norm": 0.19263526797294617,
      "learning_rate": 2.181992245117128e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 19328784,
      "step": 21625
    },
    {
      "epoch": 5.708723769301835,
      "grad_norm": 0.1222437396645546,
      "learning_rate": 2.181018496997918e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 19332912,
      "step": 21630
    },
    {
      "epoch": 5.710043552857331,
      "grad_norm": 0.20359845459461212,
      "learning_rate": 2.1800447980696648e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 19337360,
      "step": 21635
    },
    {
      "epoch": 5.711363336412829,
      "grad_norm": 0.4364054501056671,
      "learning_rate": 2.1790711484825248e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 19341648,
      "step": 21640
    },
    {
      "epoch": 5.712683119968325,
      "grad_norm": 0.24508921802043915,
      "learning_rate": 2.178097548386646e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 19346448,
      "step": 21645
    },
    {
      "epoch": 5.714002903523822,
      "grad_norm": 0.0940992683172226,
      "learning_rate": 2.1771239979321712e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 19350800,
      "step": 21650
    },
    {
      "epoch": 5.715322687079319,
      "grad_norm": 0.09137574583292007,
      "learning_rate": 2.1761504972692327e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 19355152,
      "step": 21655
    },
    {
      "epoch": 5.716642470634816,
      "grad_norm": 0.06133875995874405,
      "learning_rate": 2.1751770465479572e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 19359856,
      "step": 21660
    },
    {
      "epoch": 5.717962254190313,
      "grad_norm": 0.2081190049648285,
      "learning_rate": 2.174203645918464e-05,
      "loss": 0.1054,
      "num_input_tokens_seen": 19364144,
      "step": 21665
    },
    {
      "epoch": 5.71928203774581,
      "grad_norm": 0.3296848237514496,
      "learning_rate": 2.1732302955308624e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 19368656,
      "step": 21670
    },
    {
      "epoch": 5.720601821301306,
      "grad_norm": 0.14269819855690002,
      "learning_rate": 2.172256995535255e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 19373264,
      "step": 21675
    },
    {
      "epoch": 5.721921604856804,
      "grad_norm": 0.16286475956439972,
      "learning_rate": 2.171283746081739e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 19377680,
      "step": 21680
    },
    {
      "epoch": 5.7232413884123,
      "grad_norm": 0.05276135355234146,
      "learning_rate": 2.1703105473203988e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 19382064,
      "step": 21685
    },
    {
      "epoch": 5.724561171967797,
      "grad_norm": 0.24705389142036438,
      "learning_rate": 2.1693373994013168e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 19386384,
      "step": 21690
    },
    {
      "epoch": 5.725880955523294,
      "grad_norm": 0.10040836781263351,
      "learning_rate": 2.168364302474562e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 19391088,
      "step": 21695
    },
    {
      "epoch": 5.727200739078791,
      "grad_norm": 0.1523878574371338,
      "learning_rate": 2.167391256690199e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 19395504,
      "step": 21700
    },
    {
      "epoch": 5.728520522634288,
      "grad_norm": 0.15728230774402618,
      "learning_rate": 2.1664182621982855e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 19400048,
      "step": 21705
    },
    {
      "epoch": 5.729840306189785,
      "grad_norm": 0.1888779103755951,
      "learning_rate": 2.1654453191488673e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 19404624,
      "step": 21710
    },
    {
      "epoch": 5.7311600897452815,
      "grad_norm": 0.0433732308447361,
      "learning_rate": 2.1644724276919846e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 19409008,
      "step": 21715
    },
    {
      "epoch": 5.732479873300779,
      "grad_norm": 0.1991540491580963,
      "learning_rate": 2.1634995879776715e-05,
      "loss": 0.0281,
      "num_input_tokens_seen": 19413872,
      "step": 21720
    },
    {
      "epoch": 5.7337996568562755,
      "grad_norm": 0.3610057830810547,
      "learning_rate": 2.162526800155949e-05,
      "loss": 0.1048,
      "num_input_tokens_seen": 19417904,
      "step": 21725
    },
    {
      "epoch": 5.735119440411772,
      "grad_norm": 0.13592877984046936,
      "learning_rate": 2.1615540643768363e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 19422576,
      "step": 21730
    },
    {
      "epoch": 5.7364392239672695,
      "grad_norm": 0.19473311305046082,
      "learning_rate": 2.160581380790339e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 19426864,
      "step": 21735
    },
    {
      "epoch": 5.737759007522766,
      "grad_norm": 0.6999038457870483,
      "learning_rate": 2.1596087495464586e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 19431408,
      "step": 21740
    },
    {
      "epoch": 5.7390787910782635,
      "grad_norm": 0.06212056055665016,
      "learning_rate": 2.1586361707951866e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 19435760,
      "step": 21745
    },
    {
      "epoch": 5.74039857463376,
      "grad_norm": 0.21014398336410522,
      "learning_rate": 2.157663644686507e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 19440336,
      "step": 21750
    },
    {
      "epoch": 5.741718358189257,
      "grad_norm": 0.285422682762146,
      "learning_rate": 2.156691171370396e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 19444624,
      "step": 21755
    },
    {
      "epoch": 5.743038141744754,
      "grad_norm": 0.14195486903190613,
      "learning_rate": 2.1557187509968195e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 19449168,
      "step": 21760
    },
    {
      "epoch": 5.744357925300251,
      "grad_norm": 0.19457361102104187,
      "learning_rate": 2.1547463837157382e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 19453744,
      "step": 21765
    },
    {
      "epoch": 5.745677708855748,
      "grad_norm": 0.27239757776260376,
      "learning_rate": 2.1537740696771045e-05,
      "loss": 0.1174,
      "num_input_tokens_seen": 19458192,
      "step": 21770
    },
    {
      "epoch": 5.746997492411245,
      "grad_norm": 0.053028497844934464,
      "learning_rate": 2.1528018090308587e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 19462416,
      "step": 21775
    },
    {
      "epoch": 5.748317275966741,
      "grad_norm": 0.11894752085208893,
      "learning_rate": 2.151829601926938e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 19467024,
      "step": 21780
    },
    {
      "epoch": 5.749637059522239,
      "grad_norm": 0.2708461582660675,
      "learning_rate": 2.1508574485152684e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 19471856,
      "step": 21785
    },
    {
      "epoch": 5.750956843077735,
      "grad_norm": 0.40038877725601196,
      "learning_rate": 2.1498853489457667e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 19476336,
      "step": 21790
    },
    {
      "epoch": 5.752276626633233,
      "grad_norm": 0.239413782954216,
      "learning_rate": 2.1489133033683455e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 19480816,
      "step": 21795
    },
    {
      "epoch": 5.753596410188729,
      "grad_norm": 0.13613425195217133,
      "learning_rate": 2.1479413119329038e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 19485200,
      "step": 21800
    },
    {
      "epoch": 5.753596410188729,
      "eval_loss": 0.06795632094144821,
      "eval_runtime": 64.8305,
      "eval_samples_per_second": 103.886,
      "eval_steps_per_second": 25.975,
      "num_input_tokens_seen": 19485200,
      "step": 21800
    },
    {
      "epoch": 5.754916193744226,
      "grad_norm": 0.09529406577348709,
      "learning_rate": 2.1469693747893355e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 19489616,
      "step": 21805
    },
    {
      "epoch": 5.756235977299723,
      "grad_norm": 0.048222169280052185,
      "learning_rate": 2.1459974920875274e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 19494192,
      "step": 21810
    },
    {
      "epoch": 5.75755576085522,
      "grad_norm": 0.10930148512125015,
      "learning_rate": 2.145025663977354e-05,
      "loss": 0.1443,
      "num_input_tokens_seen": 19498640,
      "step": 21815
    },
    {
      "epoch": 5.758875544410717,
      "grad_norm": 0.22699697315692902,
      "learning_rate": 2.1440538906086844e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 19503152,
      "step": 21820
    },
    {
      "epoch": 5.760195327966214,
      "grad_norm": 0.06325110048055649,
      "learning_rate": 2.1430821721313782e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 19507664,
      "step": 21825
    },
    {
      "epoch": 5.76151511152171,
      "grad_norm": 0.09122063219547272,
      "learning_rate": 2.142110508695286e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 19511888,
      "step": 21830
    },
    {
      "epoch": 5.762834895077208,
      "grad_norm": 0.25243082642555237,
      "learning_rate": 2.1411389004502515e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 19516400,
      "step": 21835
    },
    {
      "epoch": 5.764154678632704,
      "grad_norm": 0.1664630025625229,
      "learning_rate": 2.140167347546107e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 19521008,
      "step": 21840
    },
    {
      "epoch": 5.765474462188201,
      "grad_norm": 0.08837154507637024,
      "learning_rate": 2.1391958501326793e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 19525488,
      "step": 21845
    },
    {
      "epoch": 5.766794245743698,
      "grad_norm": 0.3557136654853821,
      "learning_rate": 2.1382244083597873e-05,
      "loss": 0.1501,
      "num_input_tokens_seen": 19529744,
      "step": 21850
    },
    {
      "epoch": 5.768114029299195,
      "grad_norm": 0.06833229959011078,
      "learning_rate": 2.137253022377237e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 19533840,
      "step": 21855
    },
    {
      "epoch": 5.769433812854691,
      "grad_norm": 0.1684255301952362,
      "learning_rate": 2.136281692334829e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 19538192,
      "step": 21860
    },
    {
      "epoch": 5.770753596410189,
      "grad_norm": 0.2881527245044708,
      "learning_rate": 2.135310418382356e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 19543024,
      "step": 21865
    },
    {
      "epoch": 5.772073379965685,
      "grad_norm": 0.029740381985902786,
      "learning_rate": 2.134339200669598e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 19547504,
      "step": 21870
    },
    {
      "epoch": 5.773393163521183,
      "grad_norm": 0.26382920145988464,
      "learning_rate": 2.133368039346331e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 19552080,
      "step": 21875
    },
    {
      "epoch": 5.774712947076679,
      "grad_norm": 0.07745521515607834,
      "learning_rate": 2.1323969345623195e-05,
      "loss": 0.0842,
      "num_input_tokens_seen": 19556496,
      "step": 21880
    },
    {
      "epoch": 5.776032730632176,
      "grad_norm": 0.28076961636543274,
      "learning_rate": 2.1314258864673207e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 19560848,
      "step": 21885
    },
    {
      "epoch": 5.777352514187673,
      "grad_norm": 0.07795155793428421,
      "learning_rate": 2.130454895211082e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 19565296,
      "step": 21890
    },
    {
      "epoch": 5.77867229774317,
      "grad_norm": 0.12019369751214981,
      "learning_rate": 2.129483960943342e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 19569712,
      "step": 21895
    },
    {
      "epoch": 5.779992081298667,
      "grad_norm": 0.43047478795051575,
      "learning_rate": 2.128513083813831e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 19574000,
      "step": 21900
    },
    {
      "epoch": 5.781311864854164,
      "grad_norm": 0.1759934425354004,
      "learning_rate": 2.1275422639722724e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 19578384,
      "step": 21905
    },
    {
      "epoch": 5.7826316484096605,
      "grad_norm": 0.3217279613018036,
      "learning_rate": 2.126571501568376e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 19583280,
      "step": 21910
    },
    {
      "epoch": 5.783951431965158,
      "grad_norm": 0.27746906876564026,
      "learning_rate": 2.1256007967518478e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 19587632,
      "step": 21915
    },
    {
      "epoch": 5.7852712155206545,
      "grad_norm": 0.1901746541261673,
      "learning_rate": 2.124630149672381e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 19592368,
      "step": 21920
    },
    {
      "epoch": 5.786590999076152,
      "grad_norm": 0.3257935345172882,
      "learning_rate": 2.1236595604796624e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 19596848,
      "step": 21925
    },
    {
      "epoch": 5.7879107826316485,
      "grad_norm": 0.27395591139793396,
      "learning_rate": 2.1226890293233693e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 19601296,
      "step": 21930
    },
    {
      "epoch": 5.789230566187145,
      "grad_norm": 0.17663167417049408,
      "learning_rate": 2.1217185563531694e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 19605872,
      "step": 21935
    },
    {
      "epoch": 5.7905503497426425,
      "grad_norm": 0.2962487041950226,
      "learning_rate": 2.120748141718721e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 19610352,
      "step": 21940
    },
    {
      "epoch": 5.791870133298139,
      "grad_norm": 0.508523166179657,
      "learning_rate": 2.1197777855696765e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 19614800,
      "step": 21945
    },
    {
      "epoch": 5.7931899168536365,
      "grad_norm": 0.038787566125392914,
      "learning_rate": 2.1188074880556746e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 19619216,
      "step": 21950
    },
    {
      "epoch": 5.794509700409133,
      "grad_norm": 0.022323528304696083,
      "learning_rate": 2.1178372493263495e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 19623792,
      "step": 21955
    },
    {
      "epoch": 5.79582948396463,
      "grad_norm": 0.31188657879829407,
      "learning_rate": 2.116867069531322e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 19628400,
      "step": 21960
    },
    {
      "epoch": 5.797149267520127,
      "grad_norm": 0.05209731310606003,
      "learning_rate": 2.1158969488202073e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 19632784,
      "step": 21965
    },
    {
      "epoch": 5.798469051075624,
      "grad_norm": 0.09773904085159302,
      "learning_rate": 2.114926887342611e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 19637488,
      "step": 21970
    },
    {
      "epoch": 5.79978883463112,
      "grad_norm": 0.1102888211607933,
      "learning_rate": 2.113956885248127e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 19641680,
      "step": 21975
    },
    {
      "epoch": 5.801108618186618,
      "grad_norm": 0.15374262630939484,
      "learning_rate": 2.112986942686342e-05,
      "loss": 0.0217,
      "num_input_tokens_seen": 19646160,
      "step": 21980
    },
    {
      "epoch": 5.802428401742114,
      "grad_norm": 0.024703998118638992,
      "learning_rate": 2.112017059806835e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 19650480,
      "step": 21985
    },
    {
      "epoch": 5.803748185297611,
      "grad_norm": 0.5820759534835815,
      "learning_rate": 2.1110472367591724e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 19654992,
      "step": 21990
    },
    {
      "epoch": 5.805067968853108,
      "grad_norm": 0.15050703287124634,
      "learning_rate": 2.1100774736929145e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 19659472,
      "step": 21995
    },
    {
      "epoch": 5.806387752408605,
      "grad_norm": 0.34700727462768555,
      "learning_rate": 2.10910777075761e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 19664112,
      "step": 22000
    },
    {
      "epoch": 5.806387752408605,
      "eval_loss": 0.06791354715824127,
      "eval_runtime": 64.7415,
      "eval_samples_per_second": 104.029,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 19664112,
      "step": 22000
    },
    {
      "epoch": 5.807707535964102,
      "grad_norm": 0.29714256525039673,
      "learning_rate": 2.108138128102799e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 19668784,
      "step": 22005
    },
    {
      "epoch": 5.809027319519599,
      "grad_norm": 0.2102889120578766,
      "learning_rate": 2.107168545878014e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 19673296,
      "step": 22010
    },
    {
      "epoch": 5.810347103075095,
      "grad_norm": 0.48718398809432983,
      "learning_rate": 2.106199024232775e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 19677680,
      "step": 22015
    },
    {
      "epoch": 5.811666886630593,
      "grad_norm": 0.07027366012334824,
      "learning_rate": 2.105229563316595e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 19682256,
      "step": 22020
    },
    {
      "epoch": 5.812986670186089,
      "grad_norm": 0.2676008343696594,
      "learning_rate": 2.1042601632789784e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 19686640,
      "step": 22025
    },
    {
      "epoch": 5.814306453741587,
      "grad_norm": 0.05093734711408615,
      "learning_rate": 2.103290824269417e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 19691088,
      "step": 22030
    },
    {
      "epoch": 5.815626237297083,
      "grad_norm": 0.3033090829849243,
      "learning_rate": 2.1023215464373965e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 19695760,
      "step": 22035
    },
    {
      "epoch": 5.81694602085258,
      "grad_norm": 0.18333011865615845,
      "learning_rate": 2.1013523299323908e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 19700144,
      "step": 22040
    },
    {
      "epoch": 5.818265804408077,
      "grad_norm": 0.18397235870361328,
      "learning_rate": 2.1003831749038654e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 19704560,
      "step": 22045
    },
    {
      "epoch": 5.819585587963574,
      "grad_norm": 0.163334459066391,
      "learning_rate": 2.099414081501277e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 19709040,
      "step": 22050
    },
    {
      "epoch": 5.820905371519071,
      "grad_norm": 0.5005264282226562,
      "learning_rate": 2.09844504987407e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 19713296,
      "step": 22055
    },
    {
      "epoch": 5.822225155074568,
      "grad_norm": 0.3711463212966919,
      "learning_rate": 2.097476080171683e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 19717904,
      "step": 22060
    },
    {
      "epoch": 5.823544938630064,
      "grad_norm": 0.09006980061531067,
      "learning_rate": 2.0965071725435436e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 19722224,
      "step": 22065
    },
    {
      "epoch": 5.824864722185562,
      "grad_norm": 0.20882688462734222,
      "learning_rate": 2.0955383271390684e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 19726544,
      "step": 22070
    },
    {
      "epoch": 5.826184505741058,
      "grad_norm": 0.27874624729156494,
      "learning_rate": 2.094569544107666e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 19731056,
      "step": 22075
    },
    {
      "epoch": 5.827504289296556,
      "grad_norm": 0.11233468353748322,
      "learning_rate": 2.093600823598735e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 19735440,
      "step": 22080
    },
    {
      "epoch": 5.828824072852052,
      "grad_norm": 0.05246885120868683,
      "learning_rate": 2.092632165761663e-05,
      "loss": 0.1071,
      "num_input_tokens_seen": 19739632,
      "step": 22085
    },
    {
      "epoch": 5.830143856407549,
      "grad_norm": 0.27770912647247314,
      "learning_rate": 2.091663570745832e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 19744304,
      "step": 22090
    },
    {
      "epoch": 5.831463639963046,
      "grad_norm": 0.152957946062088,
      "learning_rate": 2.0906950387006086e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 19748656,
      "step": 22095
    },
    {
      "epoch": 5.832783423518543,
      "grad_norm": 0.6110996603965759,
      "learning_rate": 2.0897265697753543e-05,
      "loss": 0.0908,
      "num_input_tokens_seen": 19753168,
      "step": 22100
    },
    {
      "epoch": 5.8341032070740395,
      "grad_norm": 0.08941664546728134,
      "learning_rate": 2.088758164119419e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 19757552,
      "step": 22105
    },
    {
      "epoch": 5.835422990629537,
      "grad_norm": 0.12796351313591003,
      "learning_rate": 2.0877898218821428e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 19762064,
      "step": 22110
    },
    {
      "epoch": 5.8367427741850335,
      "grad_norm": 0.059940509498119354,
      "learning_rate": 2.0868215432128565e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 19766320,
      "step": 22115
    },
    {
      "epoch": 5.83806255774053,
      "grad_norm": 0.15658944845199585,
      "learning_rate": 2.0858533282608796e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 19770544,
      "step": 22120
    },
    {
      "epoch": 5.8393823412960275,
      "grad_norm": 0.08887448161840439,
      "learning_rate": 2.084885177175524e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 19775408,
      "step": 22125
    },
    {
      "epoch": 5.840702124851524,
      "grad_norm": 0.19172848761081696,
      "learning_rate": 2.0839170901060917e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 19779760,
      "step": 22130
    },
    {
      "epoch": 5.8420219084070215,
      "grad_norm": 0.22904475033283234,
      "learning_rate": 2.082949067201872e-05,
      "loss": 0.1024,
      "num_input_tokens_seen": 19784464,
      "step": 22135
    },
    {
      "epoch": 5.843341691962518,
      "grad_norm": 0.17465868592262268,
      "learning_rate": 2.0819811086121475e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 19789072,
      "step": 22140
    },
    {
      "epoch": 5.844661475518015,
      "grad_norm": 0.31478258967399597,
      "learning_rate": 2.08101321448619e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 19793456,
      "step": 22145
    },
    {
      "epoch": 5.845981259073512,
      "grad_norm": 0.07886745035648346,
      "learning_rate": 2.080045384973259e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 19798032,
      "step": 22150
    },
    {
      "epoch": 5.847301042629009,
      "grad_norm": 0.10830595344305038,
      "learning_rate": 2.0790776202226082e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 19802640,
      "step": 22155
    },
    {
      "epoch": 5.848620826184506,
      "grad_norm": 0.47462987899780273,
      "learning_rate": 2.078109920383477e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 19807152,
      "step": 22160
    },
    {
      "epoch": 5.849940609740003,
      "grad_norm": 0.2559426724910736,
      "learning_rate": 2.0771422856050978e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 19811472,
      "step": 22165
    },
    {
      "epoch": 5.851260393295499,
      "grad_norm": 0.06618694961071014,
      "learning_rate": 2.076174716036693e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 19816048,
      "step": 22170
    },
    {
      "epoch": 5.852580176850997,
      "grad_norm": 0.28502368927001953,
      "learning_rate": 2.075207211827472e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 19820432,
      "step": 22175
    },
    {
      "epoch": 5.853899960406493,
      "grad_norm": 0.1422804445028305,
      "learning_rate": 2.074239773126638e-05,
      "loss": 0.0268,
      "num_input_tokens_seen": 19825232,
      "step": 22180
    },
    {
      "epoch": 5.855219743961991,
      "grad_norm": 0.1599172204732895,
      "learning_rate": 2.073272400083382e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 19829488,
      "step": 22185
    },
    {
      "epoch": 5.856539527517487,
      "grad_norm": 0.17204247415065765,
      "learning_rate": 2.072305092846883e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 19834224,
      "step": 22190
    },
    {
      "epoch": 5.857859311072984,
      "grad_norm": 0.3195152282714844,
      "learning_rate": 2.0713378515663152e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 19838896,
      "step": 22195
    },
    {
      "epoch": 5.859179094628481,
      "grad_norm": 0.3729403614997864,
      "learning_rate": 2.070370676390836e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 19843216,
      "step": 22200
    },
    {
      "epoch": 5.859179094628481,
      "eval_loss": 0.06794528663158417,
      "eval_runtime": 64.7917,
      "eval_samples_per_second": 103.948,
      "eval_steps_per_second": 25.991,
      "num_input_tokens_seen": 19843216,
      "step": 22200
    },
    {
      "epoch": 5.860498878183978,
      "grad_norm": 0.21294929087162018,
      "learning_rate": 2.0694035674695974e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 19847632,
      "step": 22205
    },
    {
      "epoch": 5.861818661739475,
      "grad_norm": 0.20426017045974731,
      "learning_rate": 2.0684365249517416e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 19852048,
      "step": 22210
    },
    {
      "epoch": 5.863138445294972,
      "grad_norm": 0.22793236374855042,
      "learning_rate": 2.067469548986396e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 19856400,
      "step": 22215
    },
    {
      "epoch": 5.864458228850468,
      "grad_norm": 0.11497310549020767,
      "learning_rate": 2.066502639722681e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 19860656,
      "step": 22220
    },
    {
      "epoch": 5.865778012405966,
      "grad_norm": 0.20773261785507202,
      "learning_rate": 2.065535797309708e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 19865200,
      "step": 22225
    },
    {
      "epoch": 5.867097795961462,
      "grad_norm": 0.09034489840269089,
      "learning_rate": 2.0645690218965736e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 19869456,
      "step": 22230
    },
    {
      "epoch": 5.868417579516959,
      "grad_norm": 0.0734708234667778,
      "learning_rate": 2.063602313632369e-05,
      "loss": 0.0204,
      "num_input_tokens_seen": 19873872,
      "step": 22235
    },
    {
      "epoch": 5.869737363072456,
      "grad_norm": 0.0815160870552063,
      "learning_rate": 2.0626356726661704e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 19878320,
      "step": 22240
    },
    {
      "epoch": 5.871057146627953,
      "grad_norm": 0.04125366359949112,
      "learning_rate": 2.0616690991470477e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 19883056,
      "step": 22245
    },
    {
      "epoch": 5.872376930183449,
      "grad_norm": 0.15904560685157776,
      "learning_rate": 2.0607025932240595e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 19887536,
      "step": 22250
    },
    {
      "epoch": 5.873696713738947,
      "grad_norm": 0.0795137956738472,
      "learning_rate": 2.059736155046251e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 19892112,
      "step": 22255
    },
    {
      "epoch": 5.875016497294443,
      "grad_norm": 0.06374108046293259,
      "learning_rate": 2.0587697847626603e-05,
      "loss": 0.0305,
      "num_input_tokens_seen": 19896336,
      "step": 22260
    },
    {
      "epoch": 5.876336280849941,
      "grad_norm": 0.2981642782688141,
      "learning_rate": 2.057803482522314e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 19900592,
      "step": 22265
    },
    {
      "epoch": 5.877656064405437,
      "grad_norm": 0.206697478890419,
      "learning_rate": 2.056837248474227e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 19905456,
      "step": 22270
    },
    {
      "epoch": 5.878975847960934,
      "grad_norm": 0.40407702326774597,
      "learning_rate": 2.0558710827674064e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 19909968,
      "step": 22275
    },
    {
      "epoch": 5.880295631516431,
      "grad_norm": 0.643403172492981,
      "learning_rate": 2.054904985550845e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 19914544,
      "step": 22280
    },
    {
      "epoch": 5.881615415071928,
      "grad_norm": 0.04485026001930237,
      "learning_rate": 2.0539389569735287e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 19919088,
      "step": 22285
    },
    {
      "epoch": 5.882935198627425,
      "grad_norm": 0.37334638833999634,
      "learning_rate": 2.052972997184431e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 19923280,
      "step": 22290
    },
    {
      "epoch": 5.884254982182922,
      "grad_norm": 0.3444516658782959,
      "learning_rate": 2.0520071063325146e-05,
      "loss": 0.0841,
      "num_input_tokens_seen": 19928080,
      "step": 22295
    },
    {
      "epoch": 5.8855747657384185,
      "grad_norm": 0.12093238532543182,
      "learning_rate": 2.051041284566732e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 19932592,
      "step": 22300
    },
    {
      "epoch": 5.886894549293916,
      "grad_norm": 0.11455899477005005,
      "learning_rate": 2.050075532036026e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 19936944,
      "step": 22305
    },
    {
      "epoch": 5.8882143328494125,
      "grad_norm": 0.36267268657684326,
      "learning_rate": 2.0491098488893264e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 19941616,
      "step": 22310
    },
    {
      "epoch": 5.88953411640491,
      "grad_norm": 0.17083221673965454,
      "learning_rate": 2.0481442352755546e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 19946160,
      "step": 22315
    },
    {
      "epoch": 5.8908538999604065,
      "grad_norm": 0.13374413549900055,
      "learning_rate": 2.0471786913436198e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 19951024,
      "step": 22320
    },
    {
      "epoch": 5.892173683515903,
      "grad_norm": 0.637289822101593,
      "learning_rate": 2.0462132172424218e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 19955632,
      "step": 22325
    },
    {
      "epoch": 5.8934934670714005,
      "grad_norm": 0.19507372379302979,
      "learning_rate": 2.0452478131208484e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 19960048,
      "step": 22330
    },
    {
      "epoch": 5.894813250626897,
      "grad_norm": 0.2867993712425232,
      "learning_rate": 2.0442824791277765e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 19964496,
      "step": 22335
    },
    {
      "epoch": 5.8961330341823945,
      "grad_norm": 0.18662673234939575,
      "learning_rate": 2.0433172154120727e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 19969008,
      "step": 22340
    },
    {
      "epoch": 5.897452817737891,
      "grad_norm": 0.22194775938987732,
      "learning_rate": 2.0423520221225947e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 19973616,
      "step": 22345
    },
    {
      "epoch": 5.898772601293388,
      "grad_norm": 0.42935243248939514,
      "learning_rate": 2.0413868994081848e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 19977872,
      "step": 22350
    },
    {
      "epoch": 5.900092384848885,
      "grad_norm": 0.04585396498441696,
      "learning_rate": 2.0404218474176795e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 19982704,
      "step": 22355
    },
    {
      "epoch": 5.901412168404382,
      "grad_norm": 0.07347700744867325,
      "learning_rate": 2.0394568662999002e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 19987088,
      "step": 22360
    },
    {
      "epoch": 5.902731951959878,
      "grad_norm": 0.10140080749988556,
      "learning_rate": 2.0384919562036593e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 19991664,
      "step": 22365
    },
    {
      "epoch": 5.904051735515376,
      "grad_norm": 0.2450418621301651,
      "learning_rate": 2.0375271172777593e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 19995888,
      "step": 22370
    },
    {
      "epoch": 5.905371519070872,
      "grad_norm": 0.08728967607021332,
      "learning_rate": 2.0365623496709885e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 20000240,
      "step": 22375
    },
    {
      "epoch": 5.90669130262637,
      "grad_norm": 0.07881025224924088,
      "learning_rate": 2.0355976535321283e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 20004688,
      "step": 22380
    },
    {
      "epoch": 5.908011086181866,
      "grad_norm": 0.3610896170139313,
      "learning_rate": 2.034633029009945e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 20009264,
      "step": 22385
    },
    {
      "epoch": 5.909330869737363,
      "grad_norm": 0.26583799719810486,
      "learning_rate": 2.0336684762531972e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 20013712,
      "step": 22390
    },
    {
      "epoch": 5.91065065329286,
      "grad_norm": 0.1178184300661087,
      "learning_rate": 2.032703995410631e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 20018288,
      "step": 22395
    },
    {
      "epoch": 5.911970436848357,
      "grad_norm": 0.19758890569210052,
      "learning_rate": 2.031739586630981e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 20022672,
      "step": 22400
    },
    {
      "epoch": 5.911970436848357,
      "eval_loss": 0.06776358187198639,
      "eval_runtime": 64.7323,
      "eval_samples_per_second": 104.044,
      "eval_steps_per_second": 26.015,
      "num_input_tokens_seen": 20022672,
      "step": 22400
    },
    {
      "epoch": 5.913290220403853,
      "grad_norm": 0.34361952543258667,
      "learning_rate": 2.0307752500629707e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 20027120,
      "step": 22405
    },
    {
      "epoch": 5.914610003959351,
      "grad_norm": 0.2447052150964737,
      "learning_rate": 2.0298109858553144e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 20031504,
      "step": 22410
    },
    {
      "epoch": 5.915929787514847,
      "grad_norm": 0.17563025653362274,
      "learning_rate": 2.028846794156712e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 20035952,
      "step": 22415
    },
    {
      "epoch": 5.917249571070345,
      "grad_norm": 0.06616902351379395,
      "learning_rate": 2.027882675115856e-05,
      "loss": 0.1116,
      "num_input_tokens_seen": 20040464,
      "step": 22420
    },
    {
      "epoch": 5.918569354625841,
      "grad_norm": 0.12638072669506073,
      "learning_rate": 2.026918628881423e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 20045136,
      "step": 22425
    },
    {
      "epoch": 5.919889138181338,
      "grad_norm": 0.11560352146625519,
      "learning_rate": 2.0259546556020833e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 20049680,
      "step": 22430
    },
    {
      "epoch": 5.921208921736835,
      "grad_norm": 0.11841684579849243,
      "learning_rate": 2.024990755426493e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 20053840,
      "step": 22435
    },
    {
      "epoch": 5.922528705292332,
      "grad_norm": 0.39667266607284546,
      "learning_rate": 2.0240269285032975e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 20058160,
      "step": 22440
    },
    {
      "epoch": 5.923848488847829,
      "grad_norm": 0.06110318377614021,
      "learning_rate": 2.0230631749811306e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 20062640,
      "step": 22445
    },
    {
      "epoch": 5.925168272403326,
      "grad_norm": 0.07051922380924225,
      "learning_rate": 2.0220994950086162e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 20066800,
      "step": 22450
    },
    {
      "epoch": 5.926488055958822,
      "grad_norm": 0.07160522043704987,
      "learning_rate": 2.021135888734365e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 20071152,
      "step": 22455
    },
    {
      "epoch": 5.92780783951432,
      "grad_norm": 0.19556304812431335,
      "learning_rate": 2.0201723563069783e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 20075376,
      "step": 22460
    },
    {
      "epoch": 5.929127623069816,
      "grad_norm": 0.09849676489830017,
      "learning_rate": 2.0192088978750433e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 20080048,
      "step": 22465
    },
    {
      "epoch": 5.930447406625314,
      "grad_norm": 0.352900892496109,
      "learning_rate": 2.0182455135871385e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 20084400,
      "step": 22470
    },
    {
      "epoch": 5.9317671901808104,
      "grad_norm": 0.07026854157447815,
      "learning_rate": 2.0172822035918305e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 20089264,
      "step": 22475
    },
    {
      "epoch": 5.933086973736307,
      "grad_norm": 0.4231656491756439,
      "learning_rate": 2.016318968037671e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 20093808,
      "step": 22480
    },
    {
      "epoch": 5.9344067572918044,
      "grad_norm": 0.07129927724599838,
      "learning_rate": 2.015355807073206e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 20098544,
      "step": 22485
    },
    {
      "epoch": 5.935726540847301,
      "grad_norm": 0.07132765650749207,
      "learning_rate": 2.0143927208469664e-05,
      "loss": 0.0801,
      "num_input_tokens_seen": 20102992,
      "step": 22490
    },
    {
      "epoch": 5.937046324402798,
      "grad_norm": 0.225108340382576,
      "learning_rate": 2.0134297095074708e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 20107696,
      "step": 22495
    },
    {
      "epoch": 5.938366107958295,
      "grad_norm": 0.12584401667118073,
      "learning_rate": 2.0124667732032297e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 20112336,
      "step": 22500
    },
    {
      "epoch": 5.939685891513792,
      "grad_norm": 0.2663365602493286,
      "learning_rate": 2.011503912082738e-05,
      "loss": 0.1231,
      "num_input_tokens_seen": 20116368,
      "step": 22505
    },
    {
      "epoch": 5.941005675069289,
      "grad_norm": 0.18252389132976532,
      "learning_rate": 2.0105411262944823e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 20120976,
      "step": 22510
    },
    {
      "epoch": 5.942325458624786,
      "grad_norm": 0.24875110387802124,
      "learning_rate": 2.0095784159869366e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 20125392,
      "step": 22515
    },
    {
      "epoch": 5.943645242180282,
      "grad_norm": 0.07982087880373001,
      "learning_rate": 2.0086157813085608e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 20129776,
      "step": 22520
    },
    {
      "epoch": 5.94496502573578,
      "grad_norm": 0.17628490924835205,
      "learning_rate": 2.0076532224078068e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 20134320,
      "step": 22525
    },
    {
      "epoch": 5.946284809291276,
      "grad_norm": 0.17296676337718964,
      "learning_rate": 2.0066907394331142e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 20138992,
      "step": 22530
    },
    {
      "epoch": 5.947604592846773,
      "grad_norm": 0.3404502272605896,
      "learning_rate": 2.0057283325329077e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 20143504,
      "step": 22535
    },
    {
      "epoch": 5.94892437640227,
      "grad_norm": 0.07860489934682846,
      "learning_rate": 2.0047660018556047e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 20147472,
      "step": 22540
    },
    {
      "epoch": 5.950244159957767,
      "grad_norm": 0.13407465815544128,
      "learning_rate": 2.0038037475496075e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 20151824,
      "step": 22545
    },
    {
      "epoch": 5.951563943513264,
      "grad_norm": 0.2956903576850891,
      "learning_rate": 2.0028415697633073e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 20156464,
      "step": 22550
    },
    {
      "epoch": 5.952883727068761,
      "grad_norm": 0.07711460441350937,
      "learning_rate": 2.0018794686450858e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 20161072,
      "step": 22555
    },
    {
      "epoch": 5.954203510624257,
      "grad_norm": 0.04942665994167328,
      "learning_rate": 2.0009174443433088e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 20165808,
      "step": 22560
    },
    {
      "epoch": 5.955523294179755,
      "grad_norm": 0.20319174230098724,
      "learning_rate": 1.999955497006334e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 20170160,
      "step": 22565
    },
    {
      "epoch": 5.956843077735251,
      "grad_norm": 0.17572377622127533,
      "learning_rate": 1.9989936267825067e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 20174768,
      "step": 22570
    },
    {
      "epoch": 5.958162861290749,
      "grad_norm": 0.027589555829763412,
      "learning_rate": 1.9980318338201572e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 20179344,
      "step": 22575
    },
    {
      "epoch": 5.959482644846245,
      "grad_norm": 0.1353382170200348,
      "learning_rate": 1.997070118267607e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 20183952,
      "step": 22580
    },
    {
      "epoch": 5.960802428401742,
      "grad_norm": 0.27334097027778625,
      "learning_rate": 1.9961084802731654e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 20188336,
      "step": 22585
    },
    {
      "epoch": 5.962122211957239,
      "grad_norm": 0.11162400990724564,
      "learning_rate": 1.9951469199851273e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 20192688,
      "step": 22590
    },
    {
      "epoch": 5.963441995512736,
      "grad_norm": 0.08960824459791183,
      "learning_rate": 1.99418543755178e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 20196848,
      "step": 22595
    },
    {
      "epoch": 5.964761779068233,
      "grad_norm": 0.2577708959579468,
      "learning_rate": 1.9932240331213936e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 20201808,
      "step": 22600
    },
    {
      "epoch": 5.964761779068233,
      "eval_loss": 0.06784471869468689,
      "eval_runtime": 64.8003,
      "eval_samples_per_second": 103.935,
      "eval_steps_per_second": 25.988,
      "num_input_tokens_seen": 20201808,
      "step": 22600
    },
    {
      "epoch": 5.96608156262373,
      "grad_norm": 0.11876357346773148,
      "learning_rate": 1.9922627068422297e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 20206064,
      "step": 22605
    },
    {
      "epoch": 5.967401346179226,
      "grad_norm": 0.09185434877872467,
      "learning_rate": 1.991301458862538e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 20210480,
      "step": 22610
    },
    {
      "epoch": 5.968721129734724,
      "grad_norm": 0.21985825896263123,
      "learning_rate": 1.9903402893305536e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 20215024,
      "step": 22615
    },
    {
      "epoch": 5.97004091329022,
      "grad_norm": 0.15698637068271637,
      "learning_rate": 1.9893791983945016e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 20219312,
      "step": 22620
    },
    {
      "epoch": 5.971360696845718,
      "grad_norm": 0.14252899587154388,
      "learning_rate": 1.988418186202594e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 20223920,
      "step": 22625
    },
    {
      "epoch": 5.972680480401214,
      "grad_norm": 0.17126336693763733,
      "learning_rate": 1.98745725290303e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 20228144,
      "step": 22630
    },
    {
      "epoch": 5.974000263956711,
      "grad_norm": 0.1142592802643776,
      "learning_rate": 1.986496398644e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 20232752,
      "step": 22635
    },
    {
      "epoch": 5.975320047512208,
      "grad_norm": 0.07449564337730408,
      "learning_rate": 1.9855356235736777e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 20237648,
      "step": 22640
    },
    {
      "epoch": 5.976639831067705,
      "grad_norm": 0.596913754940033,
      "learning_rate": 1.9845749278402277e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 20242064,
      "step": 22645
    },
    {
      "epoch": 5.9779596146232015,
      "grad_norm": 0.22089427709579468,
      "learning_rate": 1.9836143115918006e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 20246608,
      "step": 22650
    },
    {
      "epoch": 5.979279398178699,
      "grad_norm": 0.06661208719015121,
      "learning_rate": 1.9826537749765367e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 20251120,
      "step": 22655
    },
    {
      "epoch": 5.9805991817341955,
      "grad_norm": 0.1210612803697586,
      "learning_rate": 1.9816933181425625e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 20255440,
      "step": 22660
    },
    {
      "epoch": 5.981918965289692,
      "grad_norm": 0.23772338032722473,
      "learning_rate": 1.9807329412379903e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 20259760,
      "step": 22665
    },
    {
      "epoch": 5.9832387488451895,
      "grad_norm": 0.05137418210506439,
      "learning_rate": 1.9797726444109247e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 20264272,
      "step": 22670
    },
    {
      "epoch": 5.984558532400686,
      "grad_norm": 0.4696442782878876,
      "learning_rate": 1.9788124278094557e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 20268592,
      "step": 22675
    },
    {
      "epoch": 5.9858783159561835,
      "grad_norm": 0.03975774347782135,
      "learning_rate": 1.9778522915816594e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 20272944,
      "step": 22680
    },
    {
      "epoch": 5.98719809951168,
      "grad_norm": 0.2779412269592285,
      "learning_rate": 1.9768922358756014e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 20277584,
      "step": 22685
    },
    {
      "epoch": 5.988517883067177,
      "grad_norm": 0.10319958627223969,
      "learning_rate": 1.9759322608393353e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 20282256,
      "step": 22690
    },
    {
      "epoch": 5.989837666622674,
      "grad_norm": 0.08342566341161728,
      "learning_rate": 1.9749723666208992e-05,
      "loss": 0.096,
      "num_input_tokens_seen": 20286416,
      "step": 22695
    },
    {
      "epoch": 5.991157450178171,
      "grad_norm": 0.22150297462940216,
      "learning_rate": 1.9740125533683235e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 20290864,
      "step": 22700
    },
    {
      "epoch": 5.992477233733668,
      "grad_norm": 0.2464013397693634,
      "learning_rate": 1.9730528212296208e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 20295440,
      "step": 22705
    },
    {
      "epoch": 5.993797017289165,
      "grad_norm": 0.024801168590784073,
      "learning_rate": 1.9720931703527945e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 20299984,
      "step": 22710
    },
    {
      "epoch": 5.995116800844661,
      "grad_norm": 0.10164182633161545,
      "learning_rate": 1.9711336008858373e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 20304560,
      "step": 22715
    },
    {
      "epoch": 5.996436584400159,
      "grad_norm": 0.2264719307422638,
      "learning_rate": 1.9701741129767233e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 20309200,
      "step": 22720
    },
    {
      "epoch": 5.997756367955655,
      "grad_norm": 0.047787144780159,
      "learning_rate": 1.9692147067734202e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 20313744,
      "step": 22725
    },
    {
      "epoch": 5.999076151511153,
      "grad_norm": 0.24971498548984528,
      "learning_rate": 1.96825538242388e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 20318000,
      "step": 22730
    },
    {
      "epoch": 6.000263956711099,
      "grad_norm": 0.1991625726222992,
      "learning_rate": 1.967296140076041e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 20321824,
      "step": 22735
    },
    {
      "epoch": 6.001583740266597,
      "grad_norm": 0.4505075216293335,
      "learning_rate": 1.966336979877833e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 20326432,
      "step": 22740
    },
    {
      "epoch": 6.002903523822093,
      "grad_norm": 0.05449704825878143,
      "learning_rate": 1.9653779019771678e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 20331008,
      "step": 22745
    },
    {
      "epoch": 6.00422330737759,
      "grad_norm": 0.18573202192783356,
      "learning_rate": 1.9644189065219488e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 20335520,
      "step": 22750
    },
    {
      "epoch": 6.005543090933087,
      "grad_norm": 0.2909197211265564,
      "learning_rate": 1.9634599936600655e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 20340064,
      "step": 22755
    },
    {
      "epoch": 6.006862874488584,
      "grad_norm": 0.3009883165359497,
      "learning_rate": 1.9625011635393935e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 20344736,
      "step": 22760
    },
    {
      "epoch": 6.008182658044081,
      "grad_norm": 0.22672471404075623,
      "learning_rate": 1.9615424163077963e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 20349536,
      "step": 22765
    },
    {
      "epoch": 6.009502441599578,
      "grad_norm": 0.03663812577724457,
      "learning_rate": 1.9605837521131263e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 20354048,
      "step": 22770
    },
    {
      "epoch": 6.010822225155074,
      "grad_norm": 0.37975892424583435,
      "learning_rate": 1.9596251711032192e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 20358592,
      "step": 22775
    },
    {
      "epoch": 6.012142008710572,
      "grad_norm": 0.37234705686569214,
      "learning_rate": 1.958666673425903e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 20362752,
      "step": 22780
    },
    {
      "epoch": 6.013461792266068,
      "grad_norm": 0.0928589329123497,
      "learning_rate": 1.957708259228987e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 20367232,
      "step": 22785
    },
    {
      "epoch": 6.014781575821566,
      "grad_norm": 0.3026881814002991,
      "learning_rate": 1.956749928660273e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 20371712,
      "step": 22790
    },
    {
      "epoch": 6.016101359377062,
      "grad_norm": 0.34005528688430786,
      "learning_rate": 1.955791681867547e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 20376096,
      "step": 22795
    },
    {
      "epoch": 6.017421142932559,
      "grad_norm": 0.47880837321281433,
      "learning_rate": 1.9548335189985824e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 20380512,
      "step": 22800
    },
    {
      "epoch": 6.017421142932559,
      "eval_loss": 0.06773233413696289,
      "eval_runtime": 64.7924,
      "eval_samples_per_second": 103.947,
      "eval_steps_per_second": 25.991,
      "num_input_tokens_seen": 20380512,
      "step": 22800
    },
    {
      "epoch": 6.018740926488056,
      "grad_norm": 0.2146192044019699,
      "learning_rate": 1.9538754402011396e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 20384864,
      "step": 22805
    },
    {
      "epoch": 6.020060710043553,
      "grad_norm": 0.26803654432296753,
      "learning_rate": 1.952917445622968e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 20389280,
      "step": 22810
    },
    {
      "epoch": 6.021380493599049,
      "grad_norm": 0.08758390694856644,
      "learning_rate": 1.9519595354118005e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 20393632,
      "step": 22815
    },
    {
      "epoch": 6.022700277154547,
      "grad_norm": 0.07868046313524246,
      "learning_rate": 1.951001709715361e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 20398016,
      "step": 22820
    },
    {
      "epoch": 6.024020060710043,
      "grad_norm": 0.04365616664290428,
      "learning_rate": 1.9500439686813556e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 20402432,
      "step": 22825
    },
    {
      "epoch": 6.025339844265541,
      "grad_norm": 0.09426572918891907,
      "learning_rate": 1.949086312457482e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 20406656,
      "step": 22830
    },
    {
      "epoch": 6.026659627821037,
      "grad_norm": 0.034761928021907806,
      "learning_rate": 1.9481287411914223e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 20411136,
      "step": 22835
    },
    {
      "epoch": 6.027979411376534,
      "grad_norm": 0.15311692655086517,
      "learning_rate": 1.9471712550308457e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 20415584,
      "step": 22840
    },
    {
      "epoch": 6.029299194932031,
      "grad_norm": 0.4278508126735687,
      "learning_rate": 1.946213854123409e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 20420096,
      "step": 22845
    },
    {
      "epoch": 6.030618978487528,
      "grad_norm": 0.06508508324623108,
      "learning_rate": 1.9452565386167554e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 20424640,
      "step": 22850
    },
    {
      "epoch": 6.031938762043025,
      "grad_norm": 0.31825318932533264,
      "learning_rate": 1.9442993086585142e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 20429312,
      "step": 22855
    },
    {
      "epoch": 6.033258545598522,
      "grad_norm": 0.028642315417528152,
      "learning_rate": 1.9433421643963043e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 20433792,
      "step": 22860
    },
    {
      "epoch": 6.0345783291540185,
      "grad_norm": 0.25238311290740967,
      "learning_rate": 1.942385105977727e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 20438720,
      "step": 22865
    },
    {
      "epoch": 6.035898112709516,
      "grad_norm": 0.15502043068408966,
      "learning_rate": 1.9414281335503743e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 20443232,
      "step": 22870
    },
    {
      "epoch": 6.0372178962650125,
      "grad_norm": 0.4300287663936615,
      "learning_rate": 1.9404712472618232e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 20447680,
      "step": 22875
    },
    {
      "epoch": 6.038537679820509,
      "grad_norm": 0.24386338889598846,
      "learning_rate": 1.939514447259636e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 20452160,
      "step": 22880
    },
    {
      "epoch": 6.0398574633760065,
      "grad_norm": 0.16614852845668793,
      "learning_rate": 1.938557733691365e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 20456544,
      "step": 22885
    },
    {
      "epoch": 6.041177246931503,
      "grad_norm": 0.08854377269744873,
      "learning_rate": 1.9376011067045476e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 20461248,
      "step": 22890
    },
    {
      "epoch": 6.0424970304870005,
      "grad_norm": 0.11368018388748169,
      "learning_rate": 1.9366445664467065e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 20465632,
      "step": 22895
    },
    {
      "epoch": 6.043816814042497,
      "grad_norm": 0.05930309742689133,
      "learning_rate": 1.9356881130653533e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 20470304,
      "step": 22900
    },
    {
      "epoch": 6.045136597597994,
      "grad_norm": 0.340002179145813,
      "learning_rate": 1.9347317467079846e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 20474784,
      "step": 22905
    },
    {
      "epoch": 6.046456381153491,
      "grad_norm": 0.11450517177581787,
      "learning_rate": 1.9337754675220836e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 20479296,
      "step": 22910
    },
    {
      "epoch": 6.047776164708988,
      "grad_norm": 0.11731050908565521,
      "learning_rate": 1.9328192756551218e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 20483584,
      "step": 22915
    },
    {
      "epoch": 6.049095948264485,
      "grad_norm": 0.33171775937080383,
      "learning_rate": 1.931863171254555e-05,
      "loss": 0.106,
      "num_input_tokens_seen": 20488320,
      "step": 22920
    },
    {
      "epoch": 6.050415731819982,
      "grad_norm": 0.2961711287498474,
      "learning_rate": 1.930907154467826e-05,
      "loss": 0.0737,
      "num_input_tokens_seen": 20492736,
      "step": 22925
    },
    {
      "epoch": 6.051735515375478,
      "grad_norm": 0.2513962984085083,
      "learning_rate": 1.9299512254423673e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 20497216,
      "step": 22930
    },
    {
      "epoch": 6.053055298930976,
      "grad_norm": 0.03956704959273338,
      "learning_rate": 1.9289953843255914e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 20501696,
      "step": 22935
    },
    {
      "epoch": 6.054375082486472,
      "grad_norm": 0.2110985964536667,
      "learning_rate": 1.9280396312649048e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 20506368,
      "step": 22940
    },
    {
      "epoch": 6.055694866041969,
      "grad_norm": 0.12699942290782928,
      "learning_rate": 1.9270839664076936e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 20510688,
      "step": 22945
    },
    {
      "epoch": 6.057014649597466,
      "grad_norm": 0.08285914361476898,
      "learning_rate": 1.9261283899013345e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 20515424,
      "step": 22950
    },
    {
      "epoch": 6.058334433152963,
      "grad_norm": 0.10564824938774109,
      "learning_rate": 1.92517290189319e-05,
      "loss": 0.0468,
      "num_input_tokens_seen": 20520000,
      "step": 22955
    },
    {
      "epoch": 6.05965421670846,
      "grad_norm": 0.11849465221166611,
      "learning_rate": 1.924217502530607e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 20524480,
      "step": 22960
    },
    {
      "epoch": 6.060974000263957,
      "grad_norm": 0.21841368079185486,
      "learning_rate": 1.9232621919609207e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 20529120,
      "step": 22965
    },
    {
      "epoch": 6.062293783819453,
      "grad_norm": 0.039363421499729156,
      "learning_rate": 1.9223069703314534e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 20533824,
      "step": 22970
    },
    {
      "epoch": 6.063613567374951,
      "grad_norm": 0.17661263048648834,
      "learning_rate": 1.92135183778951e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 20538240,
      "step": 22975
    },
    {
      "epoch": 6.064933350930447,
      "grad_norm": 0.13207556307315826,
      "learning_rate": 1.9203967944823857e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 20542624,
      "step": 22980
    },
    {
      "epoch": 6.066253134485945,
      "grad_norm": 0.33705660700798035,
      "learning_rate": 1.9194418405573588e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 20547040,
      "step": 22985
    },
    {
      "epoch": 6.067572918041441,
      "grad_norm": 0.12375576794147491,
      "learning_rate": 1.9184869761616954e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 20551616,
      "step": 22990
    },
    {
      "epoch": 6.068892701596938,
      "grad_norm": 0.08161190897226334,
      "learning_rate": 1.9175322014426495e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 20555712,
      "step": 22995
    },
    {
      "epoch": 6.070212485152435,
      "grad_norm": 0.5069274306297302,
      "learning_rate": 1.9165775165474565e-05,
      "loss": 0.0986,
      "num_input_tokens_seen": 20560608,
      "step": 23000
    },
    {
      "epoch": 6.070212485152435,
      "eval_loss": 0.06800545006990433,
      "eval_runtime": 64.7414,
      "eval_samples_per_second": 104.029,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 20560608,
      "step": 23000
    },
    {
      "epoch": 6.071532268707932,
      "grad_norm": 0.1637074500322342,
      "learning_rate": 1.9156229216233434e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 20565088,
      "step": 23005
    },
    {
      "epoch": 6.072852052263428,
      "grad_norm": 0.354707807302475,
      "learning_rate": 1.9146684168175184e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 20569888,
      "step": 23010
    },
    {
      "epoch": 6.074171835818926,
      "grad_norm": 0.1474297046661377,
      "learning_rate": 1.9137140022771796e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 20574080,
      "step": 23015
    },
    {
      "epoch": 6.075491619374422,
      "grad_norm": 0.2152809053659439,
      "learning_rate": 1.9127596781495103e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 20578528,
      "step": 23020
    },
    {
      "epoch": 6.07681140292992,
      "grad_norm": 0.04411572590470314,
      "learning_rate": 1.9118054445816767e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 20583104,
      "step": 23025
    },
    {
      "epoch": 6.078131186485416,
      "grad_norm": 0.22132724523544312,
      "learning_rate": 1.9108513017208356e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 20587744,
      "step": 23030
    },
    {
      "epoch": 6.079450970040913,
      "grad_norm": 0.11532240360975266,
      "learning_rate": 1.9098972497141287e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 20591872,
      "step": 23035
    },
    {
      "epoch": 6.08077075359641,
      "grad_norm": 0.04132236912846565,
      "learning_rate": 1.9089432887086806e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 20596032,
      "step": 23040
    },
    {
      "epoch": 6.082090537151907,
      "grad_norm": 0.20819427073001862,
      "learning_rate": 1.9079894188516056e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 20600448,
      "step": 23045
    },
    {
      "epoch": 6.083410320707404,
      "grad_norm": 0.18943600356578827,
      "learning_rate": 1.907035640290002e-05,
      "loss": 0.1047,
      "num_input_tokens_seen": 20604832,
      "step": 23050
    },
    {
      "epoch": 6.084730104262901,
      "grad_norm": 0.055999405682086945,
      "learning_rate": 1.9060819531709534e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 20609152,
      "step": 23055
    },
    {
      "epoch": 6.0860498878183975,
      "grad_norm": 0.13046874105930328,
      "learning_rate": 1.9051283576415325e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 20613696,
      "step": 23060
    },
    {
      "epoch": 6.087369671373895,
      "grad_norm": 0.22738541662693024,
      "learning_rate": 1.904174853848793e-05,
      "loss": 0.077,
      "num_input_tokens_seen": 20618496,
      "step": 23065
    },
    {
      "epoch": 6.0886894549293915,
      "grad_norm": 0.14440608024597168,
      "learning_rate": 1.903221441939779e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 20622976,
      "step": 23070
    },
    {
      "epoch": 6.090009238484888,
      "grad_norm": 0.39488643407821655,
      "learning_rate": 1.9022681220615194e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 20627648,
      "step": 23075
    },
    {
      "epoch": 6.0913290220403855,
      "grad_norm": 0.23189353942871094,
      "learning_rate": 1.9013148943610255e-05,
      "loss": 0.1224,
      "num_input_tokens_seen": 20632128,
      "step": 23080
    },
    {
      "epoch": 6.092648805595882,
      "grad_norm": 0.24145640432834625,
      "learning_rate": 1.9003617589852998e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 20636672,
      "step": 23085
    },
    {
      "epoch": 6.0939685891513795,
      "grad_norm": 0.1598062366247177,
      "learning_rate": 1.899408716081326e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 20640992,
      "step": 23090
    },
    {
      "epoch": 6.095288372706876,
      "grad_norm": 0.1900138109922409,
      "learning_rate": 1.898455765796075e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 20645440,
      "step": 23095
    },
    {
      "epoch": 6.096608156262373,
      "grad_norm": 0.0436834841966629,
      "learning_rate": 1.8975029082765053e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 20650048,
      "step": 23100
    },
    {
      "epoch": 6.09792793981787,
      "grad_norm": 0.15502937138080597,
      "learning_rate": 1.8965501436695577e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 20654688,
      "step": 23105
    },
    {
      "epoch": 6.099247723373367,
      "grad_norm": 0.14915795624256134,
      "learning_rate": 1.895597472122161e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 20659488,
      "step": 23110
    },
    {
      "epoch": 6.100567506928864,
      "grad_norm": 0.27350491285324097,
      "learning_rate": 1.894644893781231e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 20663968,
      "step": 23115
    },
    {
      "epoch": 6.101887290484361,
      "grad_norm": 0.08406408131122589,
      "learning_rate": 1.893692408793665e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 20668576,
      "step": 23120
    },
    {
      "epoch": 6.103207074039857,
      "grad_norm": 0.24237757921218872,
      "learning_rate": 1.8927400173063493e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 20673248,
      "step": 23125
    },
    {
      "epoch": 6.104526857595355,
      "grad_norm": 0.07170090824365616,
      "learning_rate": 1.891787719466154e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 20677728,
      "step": 23130
    },
    {
      "epoch": 6.105846641150851,
      "grad_norm": 0.14027707278728485,
      "learning_rate": 1.8908355154199346e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 20682112,
      "step": 23135
    },
    {
      "epoch": 6.107166424706348,
      "grad_norm": 0.06461618095636368,
      "learning_rate": 1.8898834053145357e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 20686304,
      "step": 23140
    },
    {
      "epoch": 6.108486208261845,
      "grad_norm": 0.14769235253334045,
      "learning_rate": 1.8889313892967813e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 20690752,
      "step": 23145
    },
    {
      "epoch": 6.109805991817342,
      "grad_norm": 0.2689723074436188,
      "learning_rate": 1.8879794675134863e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 20695392,
      "step": 23150
    },
    {
      "epoch": 6.111125775372839,
      "grad_norm": 0.3610045909881592,
      "learning_rate": 1.8870276401114494e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 20699712,
      "step": 23155
    },
    {
      "epoch": 6.112445558928336,
      "grad_norm": 0.3233611583709717,
      "learning_rate": 1.886075907237453e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 20704160,
      "step": 23160
    },
    {
      "epoch": 6.113765342483832,
      "grad_norm": 0.15273818373680115,
      "learning_rate": 1.8851242690382672e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 20708672,
      "step": 23165
    },
    {
      "epoch": 6.11508512603933,
      "grad_norm": 0.25217145681381226,
      "learning_rate": 1.884172725660645e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 20713056,
      "step": 23170
    },
    {
      "epoch": 6.116404909594826,
      "grad_norm": 0.175252765417099,
      "learning_rate": 1.8832212772513277e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 20717344,
      "step": 23175
    },
    {
      "epoch": 6.117724693150324,
      "grad_norm": 0.05805163457989693,
      "learning_rate": 1.8822699239570414e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 20721472,
      "step": 23180
    },
    {
      "epoch": 6.11904447670582,
      "grad_norm": 0.06072873994708061,
      "learning_rate": 1.8813186659244943e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 20725792,
      "step": 23185
    },
    {
      "epoch": 6.120364260261317,
      "grad_norm": 0.16884228587150574,
      "learning_rate": 1.880367503300385e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 20730144,
      "step": 23190
    },
    {
      "epoch": 6.121684043816814,
      "grad_norm": 0.47738370299339294,
      "learning_rate": 1.8794164362313927e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 20734656,
      "step": 23195
    },
    {
      "epoch": 6.123003827372311,
      "grad_norm": 0.3757287859916687,
      "learning_rate": 1.878465464864185e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 20739200,
      "step": 23200
    },
    {
      "epoch": 6.123003827372311,
      "eval_loss": 0.06785519421100616,
      "eval_runtime": 64.7514,
      "eval_samples_per_second": 104.013,
      "eval_steps_per_second": 26.007,
      "num_input_tokens_seen": 20739200,
      "step": 23200
    },
    {
      "epoch": 6.124323610927807,
      "grad_norm": 0.47358763217926025,
      "learning_rate": 1.877514589345414e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 20743648,
      "step": 23205
    },
    {
      "epoch": 6.125643394483305,
      "grad_norm": 0.16667725145816803,
      "learning_rate": 1.876563809821715e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 20748000,
      "step": 23210
    },
    {
      "epoch": 6.126963178038801,
      "grad_norm": 0.5252594351768494,
      "learning_rate": 1.8756131264397106e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 20752576,
      "step": 23215
    },
    {
      "epoch": 6.128282961594299,
      "grad_norm": 0.08749470859766006,
      "learning_rate": 1.87466253934601e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 20756832,
      "step": 23220
    },
    {
      "epoch": 6.129602745149795,
      "grad_norm": 0.20065639913082123,
      "learning_rate": 1.8737120486872033e-05,
      "loss": 0.0178,
      "num_input_tokens_seen": 20761472,
      "step": 23225
    },
    {
      "epoch": 6.130922528705292,
      "grad_norm": 0.10488150268793106,
      "learning_rate": 1.8727616546098696e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 20765984,
      "step": 23230
    },
    {
      "epoch": 6.132242312260789,
      "grad_norm": 0.4782564043998718,
      "learning_rate": 1.8718113572605716e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 20770496,
      "step": 23235
    },
    {
      "epoch": 6.133562095816286,
      "grad_norm": 0.0979752317070961,
      "learning_rate": 1.8708611567858554e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 20775168,
      "step": 23240
    },
    {
      "epoch": 6.134881879371783,
      "grad_norm": 0.07275248318910599,
      "learning_rate": 1.8699110533322565e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 20779840,
      "step": 23245
    },
    {
      "epoch": 6.13620166292728,
      "grad_norm": 0.13235941529273987,
      "learning_rate": 1.8689610470462897e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 20784192,
      "step": 23250
    },
    {
      "epoch": 6.1375214464827765,
      "grad_norm": 0.31818288564682007,
      "learning_rate": 1.8680111380744604e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 20788736,
      "step": 23255
    },
    {
      "epoch": 6.138841230038274,
      "grad_norm": 0.294825941324234,
      "learning_rate": 1.8670613265632564e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 20793248,
      "step": 23260
    },
    {
      "epoch": 6.1401610135937705,
      "grad_norm": 0.0656796470284462,
      "learning_rate": 1.866111612659149e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 20797728,
      "step": 23265
    },
    {
      "epoch": 6.141480797149267,
      "grad_norm": 0.06125950813293457,
      "learning_rate": 1.8651619965085967e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 20802304,
      "step": 23270
    },
    {
      "epoch": 6.1428005807047645,
      "grad_norm": 0.18593533337116241,
      "learning_rate": 1.8642124782580433e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 20806432,
      "step": 23275
    },
    {
      "epoch": 6.144120364260261,
      "grad_norm": 0.05149536952376366,
      "learning_rate": 1.8632630580539144e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 20810976,
      "step": 23280
    },
    {
      "epoch": 6.1454401478157585,
      "grad_norm": 0.6380911469459534,
      "learning_rate": 1.862313736042625e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 20815584,
      "step": 23285
    },
    {
      "epoch": 6.146759931371255,
      "grad_norm": 0.24208343029022217,
      "learning_rate": 1.8613645123705703e-05,
      "loss": 0.0357,
      "num_input_tokens_seen": 20819904,
      "step": 23290
    },
    {
      "epoch": 6.148079714926752,
      "grad_norm": 0.2303633689880371,
      "learning_rate": 1.8604153871841328e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 20824608,
      "step": 23295
    },
    {
      "epoch": 6.149399498482249,
      "grad_norm": 0.34356626868247986,
      "learning_rate": 1.859466360629682e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 20828896,
      "step": 23300
    },
    {
      "epoch": 6.150719282037746,
      "grad_norm": 0.21986468136310577,
      "learning_rate": 1.8585174328535666e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 20833024,
      "step": 23305
    },
    {
      "epoch": 6.152039065593243,
      "grad_norm": 0.16313406825065613,
      "learning_rate": 1.857568604002124e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 20837632,
      "step": 23310
    },
    {
      "epoch": 6.15335884914874,
      "grad_norm": 0.2151777744293213,
      "learning_rate": 1.8566198742216774e-05,
      "loss": 0.1034,
      "num_input_tokens_seen": 20841728,
      "step": 23315
    },
    {
      "epoch": 6.154678632704236,
      "grad_norm": 0.03583676740527153,
      "learning_rate": 1.85567124365853e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 20846400,
      "step": 23320
    },
    {
      "epoch": 6.155998416259734,
      "grad_norm": 0.17931176722049713,
      "learning_rate": 1.854722712458975e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 20850912,
      "step": 23325
    },
    {
      "epoch": 6.15731819981523,
      "grad_norm": 0.04011833667755127,
      "learning_rate": 1.853774280769286e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 20855488,
      "step": 23330
    },
    {
      "epoch": 6.158637983370728,
      "grad_norm": 0.3229314088821411,
      "learning_rate": 1.852825948735724e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 20860288,
      "step": 23335
    },
    {
      "epoch": 6.159957766926224,
      "grad_norm": 0.17152704298496246,
      "learning_rate": 1.851877716504534e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 20864416,
      "step": 23340
    },
    {
      "epoch": 6.161277550481721,
      "grad_norm": 0.21125110983848572,
      "learning_rate": 1.8509295842219448e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 20868864,
      "step": 23345
    },
    {
      "epoch": 6.162597334037218,
      "grad_norm": 0.5406584143638611,
      "learning_rate": 1.8499815520341697e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 20873472,
      "step": 23350
    },
    {
      "epoch": 6.163917117592715,
      "grad_norm": 0.14122548699378967,
      "learning_rate": 1.8490336200874094e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 20877568,
      "step": 23355
    },
    {
      "epoch": 6.165236901148211,
      "grad_norm": 0.032147087156772614,
      "learning_rate": 1.848085788527844e-05,
      "loss": 0.0379,
      "num_input_tokens_seen": 20882336,
      "step": 23360
    },
    {
      "epoch": 6.166556684703709,
      "grad_norm": 0.2314123958349228,
      "learning_rate": 1.847138057501644e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 20886912,
      "step": 23365
    },
    {
      "epoch": 6.167876468259205,
      "grad_norm": 0.3218337893486023,
      "learning_rate": 1.8461904271549582e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 20891200,
      "step": 23370
    },
    {
      "epoch": 6.169196251814703,
      "grad_norm": 0.2519700825214386,
      "learning_rate": 1.845242897633926e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 20895840,
      "step": 23375
    },
    {
      "epoch": 6.170516035370199,
      "grad_norm": 0.10085026919841766,
      "learning_rate": 1.844295469084667e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 20900384,
      "step": 23380
    },
    {
      "epoch": 6.171835818925696,
      "grad_norm": 0.08631719648838043,
      "learning_rate": 1.843348141653286e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 20904640,
      "step": 23385
    },
    {
      "epoch": 6.173155602481193,
      "grad_norm": 0.32672932744026184,
      "learning_rate": 1.842400915485874e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 20909120,
      "step": 23390
    },
    {
      "epoch": 6.17447538603669,
      "grad_norm": 0.41695156693458557,
      "learning_rate": 1.8414537907285053e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 20913376,
      "step": 23395
    },
    {
      "epoch": 6.175795169592186,
      "grad_norm": 0.0174611397087574,
      "learning_rate": 1.840506767527237e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 20917728,
      "step": 23400
    },
    {
      "epoch": 6.175795169592186,
      "eval_loss": 0.06792458891868591,
      "eval_runtime": 64.7372,
      "eval_samples_per_second": 104.036,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 20917728,
      "step": 23400
    },
    {
      "epoch": 6.177114953147684,
      "grad_norm": 0.09049680083990097,
      "learning_rate": 1.8395598460281137e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 20922176,
      "step": 23405
    },
    {
      "epoch": 6.17843473670318,
      "grad_norm": 0.06453737616539001,
      "learning_rate": 1.838613026377161e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 20926496,
      "step": 23410
    },
    {
      "epoch": 6.179754520258678,
      "grad_norm": 0.3907078206539154,
      "learning_rate": 1.8376663087203917e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 20931296,
      "step": 23415
    },
    {
      "epoch": 6.181074303814174,
      "grad_norm": 0.3931455612182617,
      "learning_rate": 1.8367196932038014e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 20935424,
      "step": 23420
    },
    {
      "epoch": 6.182394087369671,
      "grad_norm": 0.09347866475582123,
      "learning_rate": 1.8357731799733686e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 20939872,
      "step": 23425
    },
    {
      "epoch": 6.183713870925168,
      "grad_norm": 0.14979027211666107,
      "learning_rate": 1.8348267691750586e-05,
      "loss": 0.0983,
      "num_input_tokens_seen": 20944576,
      "step": 23430
    },
    {
      "epoch": 6.185033654480665,
      "grad_norm": 0.09732377529144287,
      "learning_rate": 1.833880460954821e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 20949472,
      "step": 23435
    },
    {
      "epoch": 6.1863534380361624,
      "grad_norm": 0.10456091910600662,
      "learning_rate": 1.8329342554585866e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 20953856,
      "step": 23440
    },
    {
      "epoch": 6.187673221591659,
      "grad_norm": 0.5026780366897583,
      "learning_rate": 1.8319881528322735e-05,
      "loss": 0.1131,
      "num_input_tokens_seen": 20958496,
      "step": 23445
    },
    {
      "epoch": 6.188993005147156,
      "grad_norm": 0.2576005756855011,
      "learning_rate": 1.8310421532217815e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 20962784,
      "step": 23450
    },
    {
      "epoch": 6.190312788702653,
      "grad_norm": 0.21271757781505585,
      "learning_rate": 1.8300962567729958e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 20967360,
      "step": 23455
    },
    {
      "epoch": 6.19163257225815,
      "grad_norm": 0.13919296860694885,
      "learning_rate": 1.8291504636317866e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 20971936,
      "step": 23460
    },
    {
      "epoch": 6.192952355813647,
      "grad_norm": 0.21724295616149902,
      "learning_rate": 1.8282047739440055e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 20976576,
      "step": 23465
    },
    {
      "epoch": 6.194272139369144,
      "grad_norm": 0.22113631665706635,
      "learning_rate": 1.8272591878554903e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 20980960,
      "step": 23470
    },
    {
      "epoch": 6.19559192292464,
      "grad_norm": 0.2735648453235626,
      "learning_rate": 1.8263137055120638e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 20985472,
      "step": 23475
    },
    {
      "epoch": 6.196911706480138,
      "grad_norm": 0.0804530531167984,
      "learning_rate": 1.8253683270595295e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 20989760,
      "step": 23480
    },
    {
      "epoch": 6.198231490035634,
      "grad_norm": 0.03825506567955017,
      "learning_rate": 1.824423052643677e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 20994112,
      "step": 23485
    },
    {
      "epoch": 6.199551273591131,
      "grad_norm": 0.18740999698638916,
      "learning_rate": 1.82347788241028e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 20998720,
      "step": 23490
    },
    {
      "epoch": 6.200871057146628,
      "grad_norm": 0.16862717270851135,
      "learning_rate": 1.8225328165050942e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 21002976,
      "step": 23495
    },
    {
      "epoch": 6.202190840702125,
      "grad_norm": 0.04621472954750061,
      "learning_rate": 1.821587855073863e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 21007424,
      "step": 23500
    },
    {
      "epoch": 6.203510624257622,
      "grad_norm": 0.06968823820352554,
      "learning_rate": 1.8206429982623086e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 21011936,
      "step": 23505
    },
    {
      "epoch": 6.204830407813119,
      "grad_norm": 0.05870481953024864,
      "learning_rate": 1.8196982462161416e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 21016352,
      "step": 23510
    },
    {
      "epoch": 6.206150191368615,
      "grad_norm": 0.283359169960022,
      "learning_rate": 1.818753599081055e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 21021120,
      "step": 23515
    },
    {
      "epoch": 6.207469974924113,
      "grad_norm": 0.42665475606918335,
      "learning_rate": 1.817809057002724e-05,
      "loss": 0.0801,
      "num_input_tokens_seen": 21025280,
      "step": 23520
    },
    {
      "epoch": 6.208789758479609,
      "grad_norm": 0.17830058932304382,
      "learning_rate": 1.8168646201268096e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 21029600,
      "step": 23525
    },
    {
      "epoch": 6.210109542035106,
      "grad_norm": 0.19644585251808167,
      "learning_rate": 1.8159202885989557e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 21034112,
      "step": 23530
    },
    {
      "epoch": 6.211429325590603,
      "grad_norm": 0.3622327148914337,
      "learning_rate": 1.814976062564789e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 21038368,
      "step": 23535
    },
    {
      "epoch": 6.2127491091461,
      "grad_norm": 0.030862197279930115,
      "learning_rate": 1.8140319421699234e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 21043232,
      "step": 23540
    },
    {
      "epoch": 6.214068892701597,
      "grad_norm": 0.28243887424468994,
      "learning_rate": 1.8130879275599515e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 21047680,
      "step": 23545
    },
    {
      "epoch": 6.215388676257094,
      "grad_norm": 0.11172232031822205,
      "learning_rate": 1.8121440188804544e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 21052128,
      "step": 23550
    },
    {
      "epoch": 6.21670845981259,
      "grad_norm": 0.0934629812836647,
      "learning_rate": 1.811200216276993e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 21056832,
      "step": 23555
    },
    {
      "epoch": 6.218028243368088,
      "grad_norm": 0.39137664437294006,
      "learning_rate": 1.810256519895115e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 21061024,
      "step": 23560
    },
    {
      "epoch": 6.219348026923584,
      "grad_norm": 0.14804111421108246,
      "learning_rate": 1.8093129298803494e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 21065568,
      "step": 23565
    },
    {
      "epoch": 6.220667810479082,
      "grad_norm": 0.272752970457077,
      "learning_rate": 1.808369446378209e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 21069952,
      "step": 23570
    },
    {
      "epoch": 6.221987594034578,
      "grad_norm": 0.09275731444358826,
      "learning_rate": 1.8074260695341914e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 21074496,
      "step": 23575
    },
    {
      "epoch": 6.223307377590075,
      "grad_norm": 0.1410493552684784,
      "learning_rate": 1.8064827994937782e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 21078784,
      "step": 23580
    },
    {
      "epoch": 6.224627161145572,
      "grad_norm": 0.2554750144481659,
      "learning_rate": 1.8055396364024317e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 21083328,
      "step": 23585
    },
    {
      "epoch": 6.225946944701069,
      "grad_norm": 0.05474372208118439,
      "learning_rate": 1.804596580405601e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 21087936,
      "step": 23590
    },
    {
      "epoch": 6.227266728256566,
      "grad_norm": 0.24490174651145935,
      "learning_rate": 1.8036536316487174e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 21092576,
      "step": 23595
    },
    {
      "epoch": 6.228586511812063,
      "grad_norm": 0.21805471181869507,
      "learning_rate": 1.802710790277193e-05,
      "loss": 0.0304,
      "num_input_tokens_seen": 21097088,
      "step": 23600
    },
    {
      "epoch": 6.228586511812063,
      "eval_loss": 0.06765655428171158,
      "eval_runtime": 64.8117,
      "eval_samples_per_second": 103.916,
      "eval_steps_per_second": 25.983,
      "num_input_tokens_seen": 21097088,
      "step": 23600
    },
    {
      "epoch": 6.2299062953675595,
      "grad_norm": 0.10086227208375931,
      "learning_rate": 1.801768056436429e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 21101568,
      "step": 23605
    },
    {
      "epoch": 6.231226078923057,
      "grad_norm": 0.3419232964515686,
      "learning_rate": 1.8008254302718035e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 21106272,
      "step": 23610
    },
    {
      "epoch": 6.2325458624785535,
      "grad_norm": 0.40463292598724365,
      "learning_rate": 1.7998829119286837e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 21110688,
      "step": 23615
    },
    {
      "epoch": 6.23386564603405,
      "grad_norm": 0.2521134912967682,
      "learning_rate": 1.798940501552418e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 21115520,
      "step": 23620
    },
    {
      "epoch": 6.2351854295895475,
      "grad_norm": 0.08768836408853531,
      "learning_rate": 1.797998199288336e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 21119904,
      "step": 23625
    },
    {
      "epoch": 6.236505213145044,
      "grad_norm": 0.51609206199646,
      "learning_rate": 1.7970560052817543e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 21124224,
      "step": 23630
    },
    {
      "epoch": 6.2378249967005415,
      "grad_norm": 0.35435980558395386,
      "learning_rate": 1.7961139196779702e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 21128800,
      "step": 23635
    },
    {
      "epoch": 6.239144780256038,
      "grad_norm": 0.2641701400279999,
      "learning_rate": 1.7951719426222647e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 21133088,
      "step": 23640
    },
    {
      "epoch": 6.240464563811535,
      "grad_norm": 0.040627576410770416,
      "learning_rate": 1.794230074259904e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 21137696,
      "step": 23645
    },
    {
      "epoch": 6.241784347367032,
      "grad_norm": 0.06558166444301605,
      "learning_rate": 1.7932883147361336e-05,
      "loss": 0.1099,
      "num_input_tokens_seen": 21142080,
      "step": 23650
    },
    {
      "epoch": 6.243104130922529,
      "grad_norm": 0.234835684299469,
      "learning_rate": 1.7923466641961865e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 21146688,
      "step": 23655
    },
    {
      "epoch": 6.244423914478026,
      "grad_norm": 0.3145919740200043,
      "learning_rate": 1.791405122785278e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 21151424,
      "step": 23660
    },
    {
      "epoch": 6.245743698033523,
      "grad_norm": 0.06777172535657883,
      "learning_rate": 1.7904636906486037e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 21155808,
      "step": 23665
    },
    {
      "epoch": 6.247063481589019,
      "grad_norm": 0.2834625840187073,
      "learning_rate": 1.7895223679313448e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 21160160,
      "step": 23670
    },
    {
      "epoch": 6.248383265144517,
      "grad_norm": 0.3051503300666809,
      "learning_rate": 1.7885811547786653e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 21164256,
      "step": 23675
    },
    {
      "epoch": 6.249703048700013,
      "grad_norm": 0.43011030554771423,
      "learning_rate": 1.7876400513357115e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 21168448,
      "step": 23680
    },
    {
      "epoch": 6.25102283225551,
      "grad_norm": 0.12648195028305054,
      "learning_rate": 1.7866990577476146e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 21172704,
      "step": 23685
    },
    {
      "epoch": 6.252342615811007,
      "grad_norm": 0.1834063082933426,
      "learning_rate": 1.7857581741594863e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 21177344,
      "step": 23690
    },
    {
      "epoch": 6.253662399366504,
      "grad_norm": 0.4432027339935303,
      "learning_rate": 1.7848174007164237e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 21181568,
      "step": 23695
    },
    {
      "epoch": 6.254982182922001,
      "grad_norm": 0.3047706186771393,
      "learning_rate": 1.7838767375635052e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 21185920,
      "step": 23700
    },
    {
      "epoch": 6.256301966477498,
      "grad_norm": 0.032131265848875046,
      "learning_rate": 1.782936184845793e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 21190432,
      "step": 23705
    },
    {
      "epoch": 6.257621750032994,
      "grad_norm": 0.33897316455841064,
      "learning_rate": 1.7819957427083334e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 21194976,
      "step": 23710
    },
    {
      "epoch": 6.258941533588492,
      "grad_norm": 0.2434198260307312,
      "learning_rate": 1.7810554112961516e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 21199392,
      "step": 23715
    },
    {
      "epoch": 6.260261317143988,
      "grad_norm": 0.26061710715293884,
      "learning_rate": 1.7801151907542607e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 21203680,
      "step": 23720
    },
    {
      "epoch": 6.261581100699486,
      "grad_norm": 0.278072327375412,
      "learning_rate": 1.7791750812276547e-05,
      "loss": 0.0636,
      "num_input_tokens_seen": 21207744,
      "step": 23725
    },
    {
      "epoch": 6.262900884254982,
      "grad_norm": 0.3654395043849945,
      "learning_rate": 1.778235082861309e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 21211968,
      "step": 23730
    },
    {
      "epoch": 6.264220667810479,
      "grad_norm": 0.14277417957782745,
      "learning_rate": 1.777295195800184e-05,
      "loss": 0.0857,
      "num_input_tokens_seen": 21216320,
      "step": 23735
    },
    {
      "epoch": 6.265540451365976,
      "grad_norm": 0.31520983576774597,
      "learning_rate": 1.7763554201892215e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 21221024,
      "step": 23740
    },
    {
      "epoch": 6.266860234921473,
      "grad_norm": 0.047309454530477524,
      "learning_rate": 1.7754157561733476e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 21225568,
      "step": 23745
    },
    {
      "epoch": 6.268180018476969,
      "grad_norm": 0.3717719316482544,
      "learning_rate": 1.7744762038974702e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 21230016,
      "step": 23750
    },
    {
      "epoch": 6.269499802032467,
      "grad_norm": 0.20820075273513794,
      "learning_rate": 1.7735367635064788e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 21234400,
      "step": 23755
    },
    {
      "epoch": 6.270819585587963,
      "grad_norm": 0.16844941675662994,
      "learning_rate": 1.7725974351452474e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 21239136,
      "step": 23760
    },
    {
      "epoch": 6.272139369143461,
      "grad_norm": 0.12091757357120514,
      "learning_rate": 1.771658218958634e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 21243904,
      "step": 23765
    },
    {
      "epoch": 6.273459152698957,
      "grad_norm": 0.1907871514558792,
      "learning_rate": 1.770719115091475e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 21248576,
      "step": 23770
    },
    {
      "epoch": 6.274778936254454,
      "grad_norm": 0.21142634749412537,
      "learning_rate": 1.7697801236885935e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 21253120,
      "step": 23775
    },
    {
      "epoch": 6.276098719809951,
      "grad_norm": 0.468583881855011,
      "learning_rate": 1.7688412448947944e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 21257568,
      "step": 23780
    },
    {
      "epoch": 6.277418503365448,
      "grad_norm": 0.20885169506072998,
      "learning_rate": 1.767902478854862e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 21261984,
      "step": 23785
    },
    {
      "epoch": 6.278738286920945,
      "grad_norm": 0.16629639267921448,
      "learning_rate": 1.766963825713569e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 21266336,
      "step": 23790
    },
    {
      "epoch": 6.280058070476442,
      "grad_norm": 0.2886351943016052,
      "learning_rate": 1.766025285615665e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 21270816,
      "step": 23795
    },
    {
      "epoch": 6.2813778540319385,
      "grad_norm": 0.39841577410697937,
      "learning_rate": 1.7650868587058854e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 21275360,
      "step": 23800
    },
    {
      "epoch": 6.2813778540319385,
      "eval_loss": 0.06739436835050583,
      "eval_runtime": 64.7756,
      "eval_samples_per_second": 103.974,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 21275360,
      "step": 23800
    },
    {
      "epoch": 6.282697637587436,
      "grad_norm": 0.04047234356403351,
      "learning_rate": 1.7641485451289484e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 21279808,
      "step": 23805
    },
    {
      "epoch": 6.2840174211429325,
      "grad_norm": 0.04487486183643341,
      "learning_rate": 1.7632103450295534e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 21284224,
      "step": 23810
    },
    {
      "epoch": 6.285337204698429,
      "grad_norm": 0.2171248495578766,
      "learning_rate": 1.762272258552381e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 21288768,
      "step": 23815
    },
    {
      "epoch": 6.2866569882539265,
      "grad_norm": 0.034815799444913864,
      "learning_rate": 1.7613342858420988e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 21293536,
      "step": 23820
    },
    {
      "epoch": 6.287976771809423,
      "grad_norm": 0.2077503651380539,
      "learning_rate": 1.760396427043351e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 21298016,
      "step": 23825
    },
    {
      "epoch": 6.2892965553649205,
      "grad_norm": 0.06525364518165588,
      "learning_rate": 1.7594586823007696e-05,
      "loss": 0.0261,
      "num_input_tokens_seen": 21302528,
      "step": 23830
    },
    {
      "epoch": 6.290616338920417,
      "grad_norm": 0.295021116733551,
      "learning_rate": 1.7585210517589646e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 21307168,
      "step": 23835
    },
    {
      "epoch": 6.291936122475914,
      "grad_norm": 0.302420973777771,
      "learning_rate": 1.7575835355625314e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 21311712,
      "step": 23840
    },
    {
      "epoch": 6.293255906031411,
      "grad_norm": 0.1583237200975418,
      "learning_rate": 1.756646133856048e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 21316160,
      "step": 23845
    },
    {
      "epoch": 6.294575689586908,
      "grad_norm": 0.07820507138967514,
      "learning_rate": 1.7557088467840714e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 21320512,
      "step": 23850
    },
    {
      "epoch": 6.295895473142405,
      "grad_norm": 0.05903852358460426,
      "learning_rate": 1.7547716744911438e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 21325056,
      "step": 23855
    },
    {
      "epoch": 6.297215256697902,
      "grad_norm": 0.163959339261055,
      "learning_rate": 1.7538346171217902e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 21329824,
      "step": 23860
    },
    {
      "epoch": 6.298535040253398,
      "grad_norm": 0.2888582944869995,
      "learning_rate": 1.7528976748205146e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 21334272,
      "step": 23865
    },
    {
      "epoch": 6.299854823808896,
      "grad_norm": 0.13921359181404114,
      "learning_rate": 1.751960847731807e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 21338720,
      "step": 23870
    },
    {
      "epoch": 6.301174607364392,
      "grad_norm": 0.41786932945251465,
      "learning_rate": 1.7510241360001362e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 21343232,
      "step": 23875
    },
    {
      "epoch": 6.302494390919889,
      "grad_norm": 0.04720594361424446,
      "learning_rate": 1.7500875397699562e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 21347744,
      "step": 23880
    },
    {
      "epoch": 6.303814174475386,
      "grad_norm": 0.12537865340709686,
      "learning_rate": 1.7491510591857015e-05,
      "loss": 0.0317,
      "num_input_tokens_seen": 21352352,
      "step": 23885
    },
    {
      "epoch": 6.305133958030883,
      "grad_norm": 0.21012453734874725,
      "learning_rate": 1.7482146943917896e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 21357024,
      "step": 23890
    },
    {
      "epoch": 6.30645374158638,
      "grad_norm": 0.3014776408672333,
      "learning_rate": 1.7472784455326185e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 21361376,
      "step": 23895
    },
    {
      "epoch": 6.307773525141877,
      "grad_norm": 0.20395205914974213,
      "learning_rate": 1.746342312752572e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 21365696,
      "step": 23900
    },
    {
      "epoch": 6.309093308697373,
      "grad_norm": 0.05851678177714348,
      "learning_rate": 1.74540629619601e-05,
      "loss": 0.0201,
      "num_input_tokens_seen": 21370304,
      "step": 23905
    },
    {
      "epoch": 6.310413092252871,
      "grad_norm": 0.510770857334137,
      "learning_rate": 1.7444703960072815e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 21374400,
      "step": 23910
    },
    {
      "epoch": 6.311732875808367,
      "grad_norm": 0.30311185121536255,
      "learning_rate": 1.7435346123307118e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 21378816,
      "step": 23915
    },
    {
      "epoch": 6.313052659363865,
      "grad_norm": 0.08239750564098358,
      "learning_rate": 1.742598945310611e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 21383200,
      "step": 23920
    },
    {
      "epoch": 6.314372442919361,
      "grad_norm": 0.3715181350708008,
      "learning_rate": 1.741663395091272e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 21387584,
      "step": 23925
    },
    {
      "epoch": 6.315692226474858,
      "grad_norm": 0.447177529335022,
      "learning_rate": 1.7407279618169657e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 21392384,
      "step": 23930
    },
    {
      "epoch": 6.317012010030355,
      "grad_norm": 0.11280747503042221,
      "learning_rate": 1.73979264563195e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 21396704,
      "step": 23935
    },
    {
      "epoch": 6.318331793585852,
      "grad_norm": 0.3707388937473297,
      "learning_rate": 1.7388574466804625e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 21401184,
      "step": 23940
    },
    {
      "epoch": 6.319651577141348,
      "grad_norm": 0.12804940342903137,
      "learning_rate": 1.7379223651067207e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 21405888,
      "step": 23945
    },
    {
      "epoch": 6.320971360696846,
      "grad_norm": 0.18792009353637695,
      "learning_rate": 1.736987401054928e-05,
      "loss": 0.0911,
      "num_input_tokens_seen": 21410112,
      "step": 23950
    },
    {
      "epoch": 6.322291144252342,
      "grad_norm": 0.22138424217700958,
      "learning_rate": 1.736052554669266e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 21414240,
      "step": 23955
    },
    {
      "epoch": 6.32361092780784,
      "grad_norm": 0.1386704444885254,
      "learning_rate": 1.7351178260939007e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 21418912,
      "step": 23960
    },
    {
      "epoch": 6.324930711363336,
      "grad_norm": 0.2715069651603699,
      "learning_rate": 1.7341832154729794e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 21423136,
      "step": 23965
    },
    {
      "epoch": 6.326250494918833,
      "grad_norm": 0.1740335375070572,
      "learning_rate": 1.7332487229506286e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 21427520,
      "step": 23970
    },
    {
      "epoch": 6.32757027847433,
      "grad_norm": 0.18758241832256317,
      "learning_rate": 1.732314348670961e-05,
      "loss": 0.0646,
      "num_input_tokens_seen": 21432352,
      "step": 23975
    },
    {
      "epoch": 6.328890062029827,
      "grad_norm": 0.3596736788749695,
      "learning_rate": 1.7313800927780686e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 21436640,
      "step": 23980
    },
    {
      "epoch": 6.330209845585324,
      "grad_norm": 0.3624059557914734,
      "learning_rate": 1.7304459554160245e-05,
      "loss": 0.052,
      "num_input_tokens_seen": 21441056,
      "step": 23985
    },
    {
      "epoch": 6.331529629140821,
      "grad_norm": 0.20522797107696533,
      "learning_rate": 1.7295119367288853e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 21445440,
      "step": 23990
    },
    {
      "epoch": 6.3328494126963175,
      "grad_norm": 0.16960488259792328,
      "learning_rate": 1.728578036860688e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 21449728,
      "step": 23995
    },
    {
      "epoch": 6.334169196251815,
      "grad_norm": 0.3596837818622589,
      "learning_rate": 1.7276442559554513e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 21454048,
      "step": 24000
    },
    {
      "epoch": 6.334169196251815,
      "eval_loss": 0.06759290397167206,
      "eval_runtime": 64.7681,
      "eval_samples_per_second": 103.986,
      "eval_steps_per_second": 26.0,
      "num_input_tokens_seen": 21454048,
      "step": 24000
    },
    {
      "epoch": 6.3354889798073115,
      "grad_norm": 0.11799906194210052,
      "learning_rate": 1.726710594157177e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 21458976,
      "step": 24005
    },
    {
      "epoch": 6.336808763362809,
      "grad_norm": 0.1542358547449112,
      "learning_rate": 1.725777051609846e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 21463488,
      "step": 24010
    },
    {
      "epoch": 6.3381285469183055,
      "grad_norm": 0.13928136229515076,
      "learning_rate": 1.7248436284574228e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 21468064,
      "step": 24015
    },
    {
      "epoch": 6.339448330473802,
      "grad_norm": 0.17160825431346893,
      "learning_rate": 1.723910324843855e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 21472512,
      "step": 24020
    },
    {
      "epoch": 6.3407681140292995,
      "grad_norm": 0.10874205082654953,
      "learning_rate": 1.722977140913067e-05,
      "loss": 0.0483,
      "num_input_tokens_seen": 21476832,
      "step": 24025
    },
    {
      "epoch": 6.342087897584796,
      "grad_norm": 0.22802071273326874,
      "learning_rate": 1.7220440768089688e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 21481696,
      "step": 24030
    },
    {
      "epoch": 6.343407681140293,
      "grad_norm": 0.09431540220975876,
      "learning_rate": 1.7211111326754505e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 21486048,
      "step": 24035
    },
    {
      "epoch": 6.34472746469579,
      "grad_norm": 0.1326177418231964,
      "learning_rate": 1.720178308656383e-05,
      "loss": 0.1008,
      "num_input_tokens_seen": 21490656,
      "step": 24040
    },
    {
      "epoch": 6.346047248251287,
      "grad_norm": 0.0991852805018425,
      "learning_rate": 1.719245604895621e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 21495168,
      "step": 24045
    },
    {
      "epoch": 6.347367031806784,
      "grad_norm": 0.5619545578956604,
      "learning_rate": 1.7183130215369972e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 21499584,
      "step": 24050
    },
    {
      "epoch": 6.348686815362281,
      "grad_norm": 0.26978424191474915,
      "learning_rate": 1.7173805587243292e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 21503808,
      "step": 24055
    },
    {
      "epoch": 6.350006598917777,
      "grad_norm": 0.13514144718647003,
      "learning_rate": 1.7164482166014147e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 21508320,
      "step": 24060
    },
    {
      "epoch": 6.351326382473275,
      "grad_norm": 0.3354870080947876,
      "learning_rate": 1.7155159953120313e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 21512672,
      "step": 24065
    },
    {
      "epoch": 6.352646166028771,
      "grad_norm": 0.3133455216884613,
      "learning_rate": 1.714583894999941e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 21517152,
      "step": 24070
    },
    {
      "epoch": 6.353965949584268,
      "grad_norm": 0.06239788606762886,
      "learning_rate": 1.7136519158088826e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 21521216,
      "step": 24075
    },
    {
      "epoch": 6.355285733139765,
      "grad_norm": 0.47262364625930786,
      "learning_rate": 1.712720057882581e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 21525632,
      "step": 24080
    },
    {
      "epoch": 6.356605516695262,
      "grad_norm": 0.3394815921783447,
      "learning_rate": 1.7117883213647413e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 21530400,
      "step": 24085
    },
    {
      "epoch": 6.357925300250759,
      "grad_norm": 0.22428442537784576,
      "learning_rate": 1.710856706399046e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 21534944,
      "step": 24090
    },
    {
      "epoch": 6.359245083806256,
      "grad_norm": 0.27164584398269653,
      "learning_rate": 1.7099252131291648e-05,
      "loss": 0.0978,
      "num_input_tokens_seen": 21539520,
      "step": 24095
    },
    {
      "epoch": 6.360564867361752,
      "grad_norm": 0.14305263757705688,
      "learning_rate": 1.708993841698744e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 21543936,
      "step": 24100
    },
    {
      "epoch": 6.36188465091725,
      "grad_norm": 0.0484158955514431,
      "learning_rate": 1.7080625922514132e-05,
      "loss": 0.012,
      "num_input_tokens_seen": 21548448,
      "step": 24105
    },
    {
      "epoch": 6.363204434472746,
      "grad_norm": 0.30051058530807495,
      "learning_rate": 1.7071314649307836e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 21552576,
      "step": 24110
    },
    {
      "epoch": 6.364524218028244,
      "grad_norm": 0.1470867246389389,
      "learning_rate": 1.7062004598804448e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 21556896,
      "step": 24115
    },
    {
      "epoch": 6.36584400158374,
      "grad_norm": 0.22805409133434296,
      "learning_rate": 1.7052695772439702e-05,
      "loss": 0.0799,
      "num_input_tokens_seen": 21561504,
      "step": 24120
    },
    {
      "epoch": 6.367163785139237,
      "grad_norm": 0.2706526815891266,
      "learning_rate": 1.7043388171649154e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 21565952,
      "step": 24125
    },
    {
      "epoch": 6.368483568694734,
      "grad_norm": 0.11722206324338913,
      "learning_rate": 1.7034081797868127e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 21570496,
      "step": 24130
    },
    {
      "epoch": 6.369803352250231,
      "grad_norm": 0.07856651395559311,
      "learning_rate": 1.70247766525318e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 21575232,
      "step": 24135
    },
    {
      "epoch": 6.371123135805728,
      "grad_norm": 0.2467634230852127,
      "learning_rate": 1.701547273707514e-05,
      "loss": 0.1884,
      "num_input_tokens_seen": 21579456,
      "step": 24140
    },
    {
      "epoch": 6.372442919361225,
      "grad_norm": 0.08213524520397186,
      "learning_rate": 1.7006170052932916e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 21583840,
      "step": 24145
    },
    {
      "epoch": 6.373762702916721,
      "grad_norm": 0.1368740350008011,
      "learning_rate": 1.6996868601539735e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 21588064,
      "step": 24150
    },
    {
      "epoch": 6.375082486472219,
      "grad_norm": 0.24248552322387695,
      "learning_rate": 1.6987568384329977e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 21592224,
      "step": 24155
    },
    {
      "epoch": 6.376402270027715,
      "grad_norm": 0.41842830181121826,
      "learning_rate": 1.6978269402737866e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 21596608,
      "step": 24160
    },
    {
      "epoch": 6.377722053583212,
      "grad_norm": 0.3097465932369232,
      "learning_rate": 1.696897165819743e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 21600832,
      "step": 24165
    },
    {
      "epoch": 6.379041837138709,
      "grad_norm": 0.11658819019794464,
      "learning_rate": 1.6959675152142487e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 21605312,
      "step": 24170
    },
    {
      "epoch": 6.380361620694206,
      "grad_norm": 0.3191256523132324,
      "learning_rate": 1.6950379886006667e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 21609472,
      "step": 24175
    },
    {
      "epoch": 6.381681404249703,
      "grad_norm": 0.21096785366535187,
      "learning_rate": 1.6941085861223438e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 21613888,
      "step": 24180
    },
    {
      "epoch": 6.3830011878052,
      "grad_norm": 0.19102483987808228,
      "learning_rate": 1.6931793079226034e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 21618496,
      "step": 24185
    },
    {
      "epoch": 6.3843209713606965,
      "grad_norm": 0.06051099672913551,
      "learning_rate": 1.692250154144754e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 21622656,
      "step": 24190
    },
    {
      "epoch": 6.385640754916194,
      "grad_norm": 0.0768212378025055,
      "learning_rate": 1.6913211249320807e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 21626720,
      "step": 24195
    },
    {
      "epoch": 6.3869605384716905,
      "grad_norm": 0.30191799998283386,
      "learning_rate": 1.6903922204278522e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 21631232,
      "step": 24200
    },
    {
      "epoch": 6.3869605384716905,
      "eval_loss": 0.0673343613743782,
      "eval_runtime": 64.7615,
      "eval_samples_per_second": 103.997,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 21631232,
      "step": 24200
    },
    {
      "epoch": 6.388280322027187,
      "grad_norm": 0.1655900776386261,
      "learning_rate": 1.6894634407753186e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 21636000,
      "step": 24205
    },
    {
      "epoch": 6.3896001055826845,
      "grad_norm": 0.7960726022720337,
      "learning_rate": 1.6885347861177077e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 21640352,
      "step": 24210
    },
    {
      "epoch": 6.390919889138181,
      "grad_norm": 0.06886322051286697,
      "learning_rate": 1.6876062565982298e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 21645120,
      "step": 24215
    },
    {
      "epoch": 6.3922396726936785,
      "grad_norm": 0.17605438828468323,
      "learning_rate": 1.6866778523600774e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 21649312,
      "step": 24220
    },
    {
      "epoch": 6.393559456249175,
      "grad_norm": 0.45297589898109436,
      "learning_rate": 1.6857495735464195e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 21653664,
      "step": 24225
    },
    {
      "epoch": 6.394879239804672,
      "grad_norm": 0.27499887347221375,
      "learning_rate": 1.6848214203004115e-05,
      "loss": 0.1189,
      "num_input_tokens_seen": 21658208,
      "step": 24230
    },
    {
      "epoch": 6.396199023360169,
      "grad_norm": 0.30921897292137146,
      "learning_rate": 1.6838933927651835e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 21662624,
      "step": 24235
    },
    {
      "epoch": 6.397518806915666,
      "grad_norm": 0.5695245862007141,
      "learning_rate": 1.6829654910838506e-05,
      "loss": 0.1108,
      "num_input_tokens_seen": 21667072,
      "step": 24240
    },
    {
      "epoch": 6.398838590471163,
      "grad_norm": 0.6028729677200317,
      "learning_rate": 1.6820377153995065e-05,
      "loss": 0.1081,
      "num_input_tokens_seen": 21671776,
      "step": 24245
    },
    {
      "epoch": 6.40015837402666,
      "grad_norm": 0.18353049457073212,
      "learning_rate": 1.681110065855226e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 21676352,
      "step": 24250
    },
    {
      "epoch": 6.401478157582156,
      "grad_norm": 0.1644231528043747,
      "learning_rate": 1.6801825425940642e-05,
      "loss": 0.027,
      "num_input_tokens_seen": 21680640,
      "step": 24255
    },
    {
      "epoch": 6.402797941137654,
      "grad_norm": 0.11812327057123184,
      "learning_rate": 1.679255145759056e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 21685056,
      "step": 24260
    },
    {
      "epoch": 6.40411772469315,
      "grad_norm": 0.13153013586997986,
      "learning_rate": 1.6783278754932187e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 21689568,
      "step": 24265
    },
    {
      "epoch": 6.405437508248648,
      "grad_norm": 0.061284635215997696,
      "learning_rate": 1.6774007319395496e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 21693728,
      "step": 24270
    },
    {
      "epoch": 6.406757291804144,
      "grad_norm": 0.14705680310726166,
      "learning_rate": 1.6764737152410243e-05,
      "loss": 0.0304,
      "num_input_tokens_seen": 21697792,
      "step": 24275
    },
    {
      "epoch": 6.408077075359641,
      "grad_norm": 0.06740880757570267,
      "learning_rate": 1.6755468255406016e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 21702240,
      "step": 24280
    },
    {
      "epoch": 6.409396858915138,
      "grad_norm": 0.09752621501684189,
      "learning_rate": 1.674620062981219e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 21706400,
      "step": 24285
    },
    {
      "epoch": 6.410716642470635,
      "grad_norm": 0.07922883331775665,
      "learning_rate": 1.6736934277057947e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 21710624,
      "step": 24290
    },
    {
      "epoch": 6.412036426026131,
      "grad_norm": 0.25735238194465637,
      "learning_rate": 1.6727669198572286e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 21714688,
      "step": 24295
    },
    {
      "epoch": 6.413356209581629,
      "grad_norm": 0.31563568115234375,
      "learning_rate": 1.6718405395783984e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 21719136,
      "step": 24300
    },
    {
      "epoch": 6.414675993137125,
      "grad_norm": 0.22938284277915955,
      "learning_rate": 1.6709142870121643e-05,
      "loss": 0.1273,
      "num_input_tokens_seen": 21723552,
      "step": 24305
    },
    {
      "epoch": 6.415995776692623,
      "grad_norm": 0.1776706576347351,
      "learning_rate": 1.669988162301367e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 21728512,
      "step": 24310
    },
    {
      "epoch": 6.417315560248119,
      "grad_norm": 0.17078739404678345,
      "learning_rate": 1.6690621655888243e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 21733024,
      "step": 24315
    },
    {
      "epoch": 6.418635343803616,
      "grad_norm": 0.4248080253601074,
      "learning_rate": 1.6681362970173386e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 21737504,
      "step": 24320
    },
    {
      "epoch": 6.419955127359113,
      "grad_norm": 0.22397294640541077,
      "learning_rate": 1.6672105567296904e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 21741920,
      "step": 24325
    },
    {
      "epoch": 6.42127491091461,
      "grad_norm": 0.13692547380924225,
      "learning_rate": 1.666284944868639e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 21746272,
      "step": 24330
    },
    {
      "epoch": 6.4225946944701064,
      "grad_norm": 0.16578029096126556,
      "learning_rate": 1.665359461576927e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 21750816,
      "step": 24335
    },
    {
      "epoch": 6.423914478025604,
      "grad_norm": 0.3957580029964447,
      "learning_rate": 1.6644341069972736e-05,
      "loss": 0.1244,
      "num_input_tokens_seen": 21755168,
      "step": 24340
    },
    {
      "epoch": 6.4252342615811004,
      "grad_norm": 0.12295303493738174,
      "learning_rate": 1.6635088812723813e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 21759968,
      "step": 24345
    },
    {
      "epoch": 6.426554045136598,
      "grad_norm": 0.17227862775325775,
      "learning_rate": 1.6625837845449328e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 21764608,
      "step": 24350
    },
    {
      "epoch": 6.4278738286920944,
      "grad_norm": 0.16177645325660706,
      "learning_rate": 1.6616588169575874e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 21769344,
      "step": 24355
    },
    {
      "epoch": 6.429193612247591,
      "grad_norm": 0.3222702443599701,
      "learning_rate": 1.6607339786529878e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 21773760,
      "step": 24360
    },
    {
      "epoch": 6.4305133958030885,
      "grad_norm": 0.151154026389122,
      "learning_rate": 1.659809269773756e-05,
      "loss": 0.0223,
      "num_input_tokens_seen": 21778272,
      "step": 24365
    },
    {
      "epoch": 6.431833179358585,
      "grad_norm": 0.1167394146323204,
      "learning_rate": 1.658884690462493e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 21782752,
      "step": 24370
    },
    {
      "epoch": 6.4331529629140825,
      "grad_norm": 0.1073404848575592,
      "learning_rate": 1.6579602408617813e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 21787136,
      "step": 24375
    },
    {
      "epoch": 6.434472746469579,
      "grad_norm": 0.37267884612083435,
      "learning_rate": 1.657035921114181e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 21791712,
      "step": 24380
    },
    {
      "epoch": 6.435792530025076,
      "grad_norm": 0.196873277425766,
      "learning_rate": 1.656111731362236e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 21796128,
      "step": 24385
    },
    {
      "epoch": 6.437112313580573,
      "grad_norm": 0.11980720609426498,
      "learning_rate": 1.6551876717484666e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 21800640,
      "step": 24390
    },
    {
      "epoch": 6.43843209713607,
      "grad_norm": 0.2921040952205658,
      "learning_rate": 1.6542637424153752e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 21804960,
      "step": 24395
    },
    {
      "epoch": 6.439751880691567,
      "grad_norm": 0.24739579856395721,
      "learning_rate": 1.6533399435054418e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 21809632,
      "step": 24400
    },
    {
      "epoch": 6.439751880691567,
      "eval_loss": 0.06740210950374603,
      "eval_runtime": 64.7862,
      "eval_samples_per_second": 103.957,
      "eval_steps_per_second": 25.993,
      "num_input_tokens_seen": 21809632,
      "step": 24400
    },
    {
      "epoch": 6.441071664247064,
      "grad_norm": 0.3582490384578705,
      "learning_rate": 1.6524162751611304e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 21813632,
      "step": 24405
    },
    {
      "epoch": 6.44239144780256,
      "grad_norm": 0.2821758985519409,
      "learning_rate": 1.6514927375248796e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 21817920,
      "step": 24410
    },
    {
      "epoch": 6.443711231358058,
      "grad_norm": 0.3133186399936676,
      "learning_rate": 1.6505693307391127e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 21822048,
      "step": 24415
    },
    {
      "epoch": 6.445031014913554,
      "grad_norm": 0.0899176225066185,
      "learning_rate": 1.6496460549462288e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 21826656,
      "step": 24420
    },
    {
      "epoch": 6.446350798469051,
      "grad_norm": 0.09695081412792206,
      "learning_rate": 1.6487229102886097e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 21831104,
      "step": 24425
    },
    {
      "epoch": 6.447670582024548,
      "grad_norm": 0.09089024364948273,
      "learning_rate": 1.6477998969086155e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 21835264,
      "step": 24430
    },
    {
      "epoch": 6.448990365580045,
      "grad_norm": 0.10365107655525208,
      "learning_rate": 1.646877014948587e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 21839744,
      "step": 24435
    },
    {
      "epoch": 6.450310149135542,
      "grad_norm": 0.0416833758354187,
      "learning_rate": 1.6459542645508433e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 21844512,
      "step": 24440
    },
    {
      "epoch": 6.451629932691039,
      "grad_norm": 0.0596088208258152,
      "learning_rate": 1.6450316458576852e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 21849152,
      "step": 24445
    },
    {
      "epoch": 6.452949716246535,
      "grad_norm": 0.07115906476974487,
      "learning_rate": 1.6441091590113912e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 21853696,
      "step": 24450
    },
    {
      "epoch": 6.454269499802033,
      "grad_norm": 0.4276336133480072,
      "learning_rate": 1.6431868041542213e-05,
      "loss": 0.099,
      "num_input_tokens_seen": 21857952,
      "step": 24455
    },
    {
      "epoch": 6.455589283357529,
      "grad_norm": 0.10453805327415466,
      "learning_rate": 1.6422645814284123e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 21862528,
      "step": 24460
    },
    {
      "epoch": 6.456909066913026,
      "grad_norm": 0.2848013639450073,
      "learning_rate": 1.6413424909761846e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 21866912,
      "step": 24465
    },
    {
      "epoch": 6.458228850468523,
      "grad_norm": 0.04527978226542473,
      "learning_rate": 1.640420532939736e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 21871296,
      "step": 24470
    },
    {
      "epoch": 6.45954863402402,
      "grad_norm": 0.13420382142066956,
      "learning_rate": 1.639498707461242e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 21875488,
      "step": 24475
    },
    {
      "epoch": 6.460868417579517,
      "grad_norm": 0.10216771811246872,
      "learning_rate": 1.6385770146828614e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 21879712,
      "step": 24480
    },
    {
      "epoch": 6.462188201135014,
      "grad_norm": 0.08184622973203659,
      "learning_rate": 1.637655454746731e-05,
      "loss": 0.0379,
      "num_input_tokens_seen": 21884384,
      "step": 24485
    },
    {
      "epoch": 6.46350798469051,
      "grad_norm": 0.02707076072692871,
      "learning_rate": 1.6367340277949658e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 21888800,
      "step": 24490
    },
    {
      "epoch": 6.464827768246008,
      "grad_norm": 0.034223467111587524,
      "learning_rate": 1.635812733969663e-05,
      "loss": 0.0205,
      "num_input_tokens_seen": 21893344,
      "step": 24495
    },
    {
      "epoch": 6.466147551801504,
      "grad_norm": 0.025290852412581444,
      "learning_rate": 1.634891573412896e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 21897664,
      "step": 24500
    },
    {
      "epoch": 6.467467335357002,
      "grad_norm": 0.1661483496427536,
      "learning_rate": 1.6339705462667196e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 21902048,
      "step": 24505
    },
    {
      "epoch": 6.468787118912498,
      "grad_norm": 0.2683849036693573,
      "learning_rate": 1.633049652673169e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 21906752,
      "step": 24510
    },
    {
      "epoch": 6.470106902467995,
      "grad_norm": 0.07957678288221359,
      "learning_rate": 1.632128892774256e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 21911360,
      "step": 24515
    },
    {
      "epoch": 6.471426686023492,
      "grad_norm": 0.28816160559654236,
      "learning_rate": 1.6312082667119737e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 21915744,
      "step": 24520
    },
    {
      "epoch": 6.472746469578989,
      "grad_norm": 0.36751312017440796,
      "learning_rate": 1.630287774628296e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 21920480,
      "step": 24525
    },
    {
      "epoch": 6.474066253134486,
      "grad_norm": 0.04958411678671837,
      "learning_rate": 1.6293674166651718e-05,
      "loss": 0.0111,
      "num_input_tokens_seen": 21924960,
      "step": 24530
    },
    {
      "epoch": 6.475386036689983,
      "grad_norm": 0.3844093978404999,
      "learning_rate": 1.6284471929645338e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 21929472,
      "step": 24535
    },
    {
      "epoch": 6.4767058202454795,
      "grad_norm": 0.24479833245277405,
      "learning_rate": 1.627527103668291e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 21934208,
      "step": 24540
    },
    {
      "epoch": 6.478025603800977,
      "grad_norm": 0.19606706500053406,
      "learning_rate": 1.6266071489183327e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 21938880,
      "step": 24545
    },
    {
      "epoch": 6.4793453873564735,
      "grad_norm": 0.31466829776763916,
      "learning_rate": 1.6256873288565283e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 21942912,
      "step": 24550
    },
    {
      "epoch": 6.48066517091197,
      "grad_norm": 0.23563647270202637,
      "learning_rate": 1.6247676436247245e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 21947520,
      "step": 24555
    },
    {
      "epoch": 6.4819849544674675,
      "grad_norm": 0.2779657244682312,
      "learning_rate": 1.6238480933647486e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 21952192,
      "step": 24560
    },
    {
      "epoch": 6.483304738022964,
      "grad_norm": 0.04489078000187874,
      "learning_rate": 1.6229286782184083e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 21956640,
      "step": 24565
    },
    {
      "epoch": 6.4846245215784615,
      "grad_norm": 0.03793451935052872,
      "learning_rate": 1.622009398327487e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 21961184,
      "step": 24570
    },
    {
      "epoch": 6.485944305133958,
      "grad_norm": 0.21921348571777344,
      "learning_rate": 1.6210902538337502e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 21965568,
      "step": 24575
    },
    {
      "epoch": 6.487264088689455,
      "grad_norm": 0.14004310965538025,
      "learning_rate": 1.6201712448789413e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 21970432,
      "step": 24580
    },
    {
      "epoch": 6.488583872244952,
      "grad_norm": 0.08286130428314209,
      "learning_rate": 1.6192523716047827e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 21974784,
      "step": 24585
    },
    {
      "epoch": 6.489903655800449,
      "grad_norm": 0.19946348667144775,
      "learning_rate": 1.6183336341529776e-05,
      "loss": 0.0169,
      "num_input_tokens_seen": 21979456,
      "step": 24590
    },
    {
      "epoch": 6.491223439355946,
      "grad_norm": 0.041830938309431076,
      "learning_rate": 1.6174150326652047e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 21983872,
      "step": 24595
    },
    {
      "epoch": 6.492543222911443,
      "grad_norm": 0.06078316271305084,
      "learning_rate": 1.6164965672831256e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 21988192,
      "step": 24600
    },
    {
      "epoch": 6.492543222911443,
      "eval_loss": 0.06740046292543411,
      "eval_runtime": 64.7464,
      "eval_samples_per_second": 104.021,
      "eval_steps_per_second": 26.009,
      "num_input_tokens_seen": 21988192,
      "step": 24600
    },
    {
      "epoch": 6.493863006466939,
      "grad_norm": 0.09459691494703293,
      "learning_rate": 1.6155782381483784e-05,
      "loss": 0.0326,
      "num_input_tokens_seen": 21992928,
      "step": 24605
    },
    {
      "epoch": 6.495182790022437,
      "grad_norm": 0.373667448759079,
      "learning_rate": 1.6146600454025813e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 21997376,
      "step": 24610
    },
    {
      "epoch": 6.496502573577933,
      "grad_norm": 0.15895818173885345,
      "learning_rate": 1.6137419891873317e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 22001952,
      "step": 24615
    },
    {
      "epoch": 6.49782235713343,
      "grad_norm": 0.26278477907180786,
      "learning_rate": 1.6128240696442038e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 22006432,
      "step": 24620
    },
    {
      "epoch": 6.499142140688927,
      "grad_norm": 0.11717765033245087,
      "learning_rate": 1.611906286914753e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 22010784,
      "step": 24625
    },
    {
      "epoch": 6.500461924244424,
      "grad_norm": 0.1504027098417282,
      "learning_rate": 1.6109886411405144e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 22015072,
      "step": 24630
    },
    {
      "epoch": 6.501781707799921,
      "grad_norm": 0.4211582541465759,
      "learning_rate": 1.6100711324629985e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 22019360,
      "step": 24635
    },
    {
      "epoch": 6.503101491355418,
      "grad_norm": 0.21142257750034332,
      "learning_rate": 1.609153761023698e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 22023648,
      "step": 24640
    },
    {
      "epoch": 6.504421274910914,
      "grad_norm": 0.11218905448913574,
      "learning_rate": 1.608236526964083e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 22028576,
      "step": 24645
    },
    {
      "epoch": 6.505741058466412,
      "grad_norm": 0.3195367753505707,
      "learning_rate": 1.607319430425601e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 22033216,
      "step": 24650
    },
    {
      "epoch": 6.507060842021908,
      "grad_norm": 0.12473928183317184,
      "learning_rate": 1.606402471549682e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 22038176,
      "step": 24655
    },
    {
      "epoch": 6.508380625577406,
      "grad_norm": 0.2630380392074585,
      "learning_rate": 1.6054856504777312e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 22042944,
      "step": 24660
    },
    {
      "epoch": 6.509700409132902,
      "grad_norm": 0.3162872791290283,
      "learning_rate": 1.6045689673511334e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 22047360,
      "step": 24665
    },
    {
      "epoch": 6.511020192688399,
      "grad_norm": 0.0589139498770237,
      "learning_rate": 1.6036524223112548e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 22051680,
      "step": 24670
    },
    {
      "epoch": 6.512339976243896,
      "grad_norm": 0.1132378876209259,
      "learning_rate": 1.602736015499436e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 22056064,
      "step": 24675
    },
    {
      "epoch": 6.513659759799393,
      "grad_norm": 0.41450002789497375,
      "learning_rate": 1.601819747057e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 22060512,
      "step": 24680
    },
    {
      "epoch": 6.51497954335489,
      "grad_norm": 0.3552088439464569,
      "learning_rate": 1.6009036171252465e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 22065088,
      "step": 24685
    },
    {
      "epoch": 6.516299326910387,
      "grad_norm": 0.1403990238904953,
      "learning_rate": 1.599987625845453e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 22069696,
      "step": 24690
    },
    {
      "epoch": 6.517619110465883,
      "grad_norm": 0.20467954874038696,
      "learning_rate": 1.599071773358879e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 22074272,
      "step": 24695
    },
    {
      "epoch": 6.518938894021381,
      "grad_norm": 0.38528159260749817,
      "learning_rate": 1.598156059806758e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 22078880,
      "step": 24700
    },
    {
      "epoch": 6.520258677576877,
      "grad_norm": 0.0723615288734436,
      "learning_rate": 1.5972404853303062e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 22083328,
      "step": 24705
    },
    {
      "epoch": 6.521578461132374,
      "grad_norm": 0.1609569489955902,
      "learning_rate": 1.5963250500707172e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 22087776,
      "step": 24710
    },
    {
      "epoch": 6.522898244687871,
      "grad_norm": 0.0764029324054718,
      "learning_rate": 1.5954097541691612e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 22092192,
      "step": 24715
    },
    {
      "epoch": 6.524218028243368,
      "grad_norm": 0.4174157977104187,
      "learning_rate": 1.5944945977667884e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 22096768,
      "step": 24720
    },
    {
      "epoch": 6.5255378117988645,
      "grad_norm": 0.1853865534067154,
      "learning_rate": 1.593579581004729e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 22101376,
      "step": 24725
    },
    {
      "epoch": 6.526857595354362,
      "grad_norm": 0.38852012157440186,
      "learning_rate": 1.592664704024088e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 22105696,
      "step": 24730
    },
    {
      "epoch": 6.5281773789098585,
      "grad_norm": 0.24551379680633545,
      "learning_rate": 1.591749966965953e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 22109888,
      "step": 24735
    },
    {
      "epoch": 6.529497162465356,
      "grad_norm": 0.12772448360919952,
      "learning_rate": 1.5908353699713856e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 22114240,
      "step": 24740
    },
    {
      "epoch": 6.5308169460208525,
      "grad_norm": 0.2763034999370575,
      "learning_rate": 1.5899209131814298e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 22118464,
      "step": 24745
    },
    {
      "epoch": 6.532136729576349,
      "grad_norm": 0.26864516735076904,
      "learning_rate": 1.5890065967371067e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 22123008,
      "step": 24750
    },
    {
      "epoch": 6.5334565131318465,
      "grad_norm": 0.45077452063560486,
      "learning_rate": 1.5880924207794144e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 22127616,
      "step": 24755
    },
    {
      "epoch": 6.534776296687343,
      "grad_norm": 0.06777957826852798,
      "learning_rate": 1.5871783854493298e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 22132512,
      "step": 24760
    },
    {
      "epoch": 6.5360960802428405,
      "grad_norm": 0.05977372080087662,
      "learning_rate": 1.5862644908878106e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 22137120,
      "step": 24765
    },
    {
      "epoch": 6.537415863798337,
      "grad_norm": 0.1738738864660263,
      "learning_rate": 1.5853507372357885e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 22141600,
      "step": 24770
    },
    {
      "epoch": 6.538735647353834,
      "grad_norm": 0.07651665806770325,
      "learning_rate": 1.5844371246341776e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 22146240,
      "step": 24775
    },
    {
      "epoch": 6.540055430909331,
      "grad_norm": 0.05778427794575691,
      "learning_rate": 1.5835236532238674e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 22150528,
      "step": 24780
    },
    {
      "epoch": 6.541375214464828,
      "grad_norm": 0.2825198769569397,
      "learning_rate": 1.582610323145727e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 22155168,
      "step": 24785
    },
    {
      "epoch": 6.542694998020325,
      "grad_norm": 0.3366435766220093,
      "learning_rate": 1.5816971345406035e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 22159840,
      "step": 24790
    },
    {
      "epoch": 6.544014781575822,
      "grad_norm": 0.21497344970703125,
      "learning_rate": 1.5807840875493225e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 22164608,
      "step": 24795
    },
    {
      "epoch": 6.545334565131318,
      "grad_norm": 0.23767434060573578,
      "learning_rate": 1.5798711823126854e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 22168864,
      "step": 24800
    },
    {
      "epoch": 6.545334565131318,
      "eval_loss": 0.06724140048027039,
      "eval_runtime": 64.737,
      "eval_samples_per_second": 104.036,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 22168864,
      "step": 24800
    },
    {
      "epoch": 6.546654348686816,
      "grad_norm": 0.3482906222343445,
      "learning_rate": 1.578958418971477e-05,
      "loss": 0.0916,
      "num_input_tokens_seen": 22173344,
      "step": 24805
    },
    {
      "epoch": 6.547974132242312,
      "grad_norm": 0.1272256225347519,
      "learning_rate": 1.578045797666453e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 22177856,
      "step": 24810
    },
    {
      "epoch": 6.54929391579781,
      "grad_norm": 0.16916589438915253,
      "learning_rate": 1.5771333185383548e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 22182304,
      "step": 24815
    },
    {
      "epoch": 6.550613699353306,
      "grad_norm": 0.25160858035087585,
      "learning_rate": 1.576220981727895e-05,
      "loss": 0.1231,
      "num_input_tokens_seen": 22186528,
      "step": 24820
    },
    {
      "epoch": 6.551933482908803,
      "grad_norm": 0.20918305218219757,
      "learning_rate": 1.575308787375769e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 22191040,
      "step": 24825
    },
    {
      "epoch": 6.5532532664643,
      "grad_norm": 0.3416500985622406,
      "learning_rate": 1.5743967356226492e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 22195360,
      "step": 24830
    },
    {
      "epoch": 6.554573050019797,
      "grad_norm": 0.2278144508600235,
      "learning_rate": 1.5734848266091835e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 22199904,
      "step": 24835
    },
    {
      "epoch": 6.555892833575293,
      "grad_norm": 0.15138262510299683,
      "learning_rate": 1.572573060476001e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 22204608,
      "step": 24840
    },
    {
      "epoch": 6.557212617130791,
      "grad_norm": 0.24951870739459991,
      "learning_rate": 1.5716614373637085e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 22208736,
      "step": 24845
    },
    {
      "epoch": 6.558532400686287,
      "grad_norm": 0.08371890336275101,
      "learning_rate": 1.570749957412887e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 22213152,
      "step": 24850
    },
    {
      "epoch": 6.559852184241785,
      "grad_norm": 0.16481614112854004,
      "learning_rate": 1.5698386207641013e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 22217664,
      "step": 24855
    },
    {
      "epoch": 6.561171967797281,
      "grad_norm": 0.19218680262565613,
      "learning_rate": 1.5689274275578884e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 22222016,
      "step": 24860
    },
    {
      "epoch": 6.562491751352778,
      "grad_norm": 0.2483195662498474,
      "learning_rate": 1.5680163779347667e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 22226400,
      "step": 24865
    },
    {
      "epoch": 6.563811534908275,
      "grad_norm": 0.13551373779773712,
      "learning_rate": 1.5671054720352327e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 22230912,
      "step": 24870
    },
    {
      "epoch": 6.565131318463772,
      "grad_norm": 0.3919602334499359,
      "learning_rate": 1.566194709999757e-05,
      "loss": 0.1095,
      "num_input_tokens_seen": 22235264,
      "step": 24875
    },
    {
      "epoch": 6.566451102019268,
      "grad_norm": 0.10560186952352524,
      "learning_rate": 1.5652840919687933e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 22239840,
      "step": 24880
    },
    {
      "epoch": 6.567770885574766,
      "grad_norm": 0.20058444142341614,
      "learning_rate": 1.5643736180827676e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 22244288,
      "step": 24885
    },
    {
      "epoch": 6.569090669130262,
      "grad_norm": 0.5639130473136902,
      "learning_rate": 1.5634632884820878e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 22248896,
      "step": 24890
    },
    {
      "epoch": 6.57041045268576,
      "grad_norm": 0.20073877274990082,
      "learning_rate": 1.5625531033071395e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 22253408,
      "step": 24895
    },
    {
      "epoch": 6.571730236241256,
      "grad_norm": 0.03171442076563835,
      "learning_rate": 1.5616430626982828e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 22257792,
      "step": 24900
    },
    {
      "epoch": 6.573050019796753,
      "grad_norm": 0.5195639729499817,
      "learning_rate": 1.5607331667958575e-05,
      "loss": 0.0518,
      "num_input_tokens_seen": 22262080,
      "step": 24905
    },
    {
      "epoch": 6.57436980335225,
      "grad_norm": 0.07130123674869537,
      "learning_rate": 1.5598234157401824e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 22266656,
      "step": 24910
    },
    {
      "epoch": 6.575689586907747,
      "grad_norm": 0.245599627494812,
      "learning_rate": 1.5589138096715503e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 22271296,
      "step": 24915
    },
    {
      "epoch": 6.577009370463244,
      "grad_norm": 0.07163049280643463,
      "learning_rate": 1.5580043487302365e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 22275744,
      "step": 24920
    },
    {
      "epoch": 6.578329154018741,
      "grad_norm": 0.3264152407646179,
      "learning_rate": 1.5570950330564888e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 22280320,
      "step": 24925
    },
    {
      "epoch": 6.5796489375742375,
      "grad_norm": 0.09860634803771973,
      "learning_rate": 1.5561858627905367e-05,
      "loss": 0.0824,
      "num_input_tokens_seen": 22284512,
      "step": 24930
    },
    {
      "epoch": 6.580968721129735,
      "grad_norm": 0.16872276365756989,
      "learning_rate": 1.5552768380725857e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 22289376,
      "step": 24935
    },
    {
      "epoch": 6.5822885046852315,
      "grad_norm": 0.36777904629707336,
      "learning_rate": 1.5543679590428183e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 22293696,
      "step": 24940
    },
    {
      "epoch": 6.583608288240729,
      "grad_norm": 0.24464012682437897,
      "learning_rate": 1.5534592258413943e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 22298144,
      "step": 24945
    },
    {
      "epoch": 6.5849280717962255,
      "grad_norm": 0.3762947916984558,
      "learning_rate": 1.5525506386084538e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 22302784,
      "step": 24950
    },
    {
      "epoch": 6.586247855351722,
      "grad_norm": 0.15514510869979858,
      "learning_rate": 1.55164219748411e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 22307232,
      "step": 24955
    },
    {
      "epoch": 6.5875676389072195,
      "grad_norm": 0.24981392920017242,
      "learning_rate": 1.550733902608459e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 22311712,
      "step": 24960
    },
    {
      "epoch": 6.588887422462716,
      "grad_norm": 0.13734064996242523,
      "learning_rate": 1.549825754121568e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 22316160,
      "step": 24965
    },
    {
      "epoch": 6.590207206018213,
      "grad_norm": 0.028818149119615555,
      "learning_rate": 1.5489177521634864e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 22320704,
      "step": 24970
    },
    {
      "epoch": 6.59152698957371,
      "grad_norm": 0.20127154886722565,
      "learning_rate": 1.5480098968742402e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 22325024,
      "step": 24975
    },
    {
      "epoch": 6.592846773129207,
      "grad_norm": 0.23138581216335297,
      "learning_rate": 1.5471021883938304e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 22329568,
      "step": 24980
    },
    {
      "epoch": 6.594166556684704,
      "grad_norm": 0.3950396776199341,
      "learning_rate": 1.546194626862238e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 22334016,
      "step": 24985
    },
    {
      "epoch": 6.595486340240201,
      "grad_norm": 0.3257615268230438,
      "learning_rate": 1.5452872124194216e-05,
      "loss": 0.0926,
      "num_input_tokens_seen": 22338336,
      "step": 24990
    },
    {
      "epoch": 6.596806123795697,
      "grad_norm": 0.2152331918478012,
      "learning_rate": 1.5443799452053136e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 22342688,
      "step": 24995
    },
    {
      "epoch": 6.598125907351195,
      "grad_norm": 0.10007510334253311,
      "learning_rate": 1.543472825359828e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 22347392,
      "step": 25000
    },
    {
      "epoch": 6.598125907351195,
      "eval_loss": 0.06756851077079773,
      "eval_runtime": 64.7562,
      "eval_samples_per_second": 104.005,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 22347392,
      "step": 25000
    },
    {
      "epoch": 6.599445690906691,
      "grad_norm": 0.3414227068424225,
      "learning_rate": 1.5425658530228522e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 22351712,
      "step": 25005
    },
    {
      "epoch": 6.600765474462188,
      "grad_norm": 0.024705443531274796,
      "learning_rate": 1.5416590283342546e-05,
      "loss": 0.0137,
      "num_input_tokens_seen": 22356192,
      "step": 25010
    },
    {
      "epoch": 6.602085258017685,
      "grad_norm": 0.3722876310348511,
      "learning_rate": 1.5407523514338783e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 22360960,
      "step": 25015
    },
    {
      "epoch": 6.603405041573182,
      "grad_norm": 0.3392707109451294,
      "learning_rate": 1.539845822461543e-05,
      "loss": 0.0719,
      "num_input_tokens_seen": 22365888,
      "step": 25020
    },
    {
      "epoch": 6.604724825128679,
      "grad_norm": 0.022637538611888885,
      "learning_rate": 1.538939441557048e-05,
      "loss": 0.0275,
      "num_input_tokens_seen": 22370176,
      "step": 25025
    },
    {
      "epoch": 6.606044608684176,
      "grad_norm": 0.15895098447799683,
      "learning_rate": 1.5380332088601696e-05,
      "loss": 0.0203,
      "num_input_tokens_seen": 22374368,
      "step": 25030
    },
    {
      "epoch": 6.607364392239672,
      "grad_norm": 0.1940978318452835,
      "learning_rate": 1.537127124510658e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 22379008,
      "step": 25035
    },
    {
      "epoch": 6.60868417579517,
      "grad_norm": 0.10255209356546402,
      "learning_rate": 1.5362211886482457e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 22383296,
      "step": 25040
    },
    {
      "epoch": 6.610003959350666,
      "grad_norm": 0.9258224368095398,
      "learning_rate": 1.5353154014126363e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 22387776,
      "step": 25045
    },
    {
      "epoch": 6.611323742906164,
      "grad_norm": 0.39313894510269165,
      "learning_rate": 1.534409762943515e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 22392480,
      "step": 25050
    },
    {
      "epoch": 6.61264352646166,
      "grad_norm": 0.2638201415538788,
      "learning_rate": 1.5335042733805438e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 22396896,
      "step": 25055
    },
    {
      "epoch": 6.613963310017157,
      "grad_norm": 0.2537665069103241,
      "learning_rate": 1.532598932863358e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 22401312,
      "step": 25060
    },
    {
      "epoch": 6.615283093572654,
      "grad_norm": 0.37750327587127686,
      "learning_rate": 1.531693741531574e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 22405888,
      "step": 25065
    },
    {
      "epoch": 6.616602877128151,
      "grad_norm": 0.3170725405216217,
      "learning_rate": 1.5307886995247844e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 22410560,
      "step": 25070
    },
    {
      "epoch": 6.617922660683648,
      "grad_norm": 0.07566087692975998,
      "learning_rate": 1.529883806982557e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 22415072,
      "step": 25075
    },
    {
      "epoch": 6.619242444239145,
      "grad_norm": 0.040788568556308746,
      "learning_rate": 1.5289790640444376e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 22419264,
      "step": 25080
    },
    {
      "epoch": 6.620562227794641,
      "grad_norm": 0.08925555646419525,
      "learning_rate": 1.5280744708499494e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 22423424,
      "step": 25085
    },
    {
      "epoch": 6.621882011350139,
      "grad_norm": 0.0710161030292511,
      "learning_rate": 1.527170027538591e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 22427680,
      "step": 25090
    },
    {
      "epoch": 6.623201794905635,
      "grad_norm": 0.14255192875862122,
      "learning_rate": 1.5262657342498407e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 22432096,
      "step": 25095
    },
    {
      "epoch": 6.624521578461132,
      "grad_norm": 0.3095634877681732,
      "learning_rate": 1.52536159112315e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 22436608,
      "step": 25100
    },
    {
      "epoch": 6.625841362016629,
      "grad_norm": 0.08310741186141968,
      "learning_rate": 1.5244575982979497e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 22440992,
      "step": 25105
    },
    {
      "epoch": 6.627161145572126,
      "grad_norm": 0.1763300746679306,
      "learning_rate": 1.5235537559136487e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 22445120,
      "step": 25110
    },
    {
      "epoch": 6.628480929127623,
      "grad_norm": 0.31509190797805786,
      "learning_rate": 1.5226500641096286e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 22449696,
      "step": 25115
    },
    {
      "epoch": 6.62980071268312,
      "grad_norm": 0.35046854615211487,
      "learning_rate": 1.5217465230252509e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 22454080,
      "step": 25120
    },
    {
      "epoch": 6.6311204962386165,
      "grad_norm": 0.28767433762550354,
      "learning_rate": 1.5208431327998523e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 22458528,
      "step": 25125
    },
    {
      "epoch": 6.632440279794114,
      "grad_norm": 0.3922910988330841,
      "learning_rate": 1.5199398935727477e-05,
      "loss": 0.0745,
      "num_input_tokens_seen": 22462720,
      "step": 25130
    },
    {
      "epoch": 6.6337600633496105,
      "grad_norm": 0.19528095424175262,
      "learning_rate": 1.5190368054832282e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 22467264,
      "step": 25135
    },
    {
      "epoch": 6.635079846905107,
      "grad_norm": 0.36180612444877625,
      "learning_rate": 1.5181338686705601e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 22471648,
      "step": 25140
    },
    {
      "epoch": 6.6363996304606045,
      "grad_norm": 0.4259493350982666,
      "learning_rate": 1.5172310832739889e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 22476320,
      "step": 25145
    },
    {
      "epoch": 6.637719414016101,
      "grad_norm": 0.15420609712600708,
      "learning_rate": 1.5163284494327346e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 22480672,
      "step": 25150
    },
    {
      "epoch": 6.6390391975715985,
      "grad_norm": 0.06095863878726959,
      "learning_rate": 1.5154259672859952e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 22485088,
      "step": 25155
    },
    {
      "epoch": 6.640358981127095,
      "grad_norm": 0.09854001551866531,
      "learning_rate": 1.5145236369729452e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 22490112,
      "step": 25160
    },
    {
      "epoch": 6.641678764682592,
      "grad_norm": 0.23597709834575653,
      "learning_rate": 1.5136214586327335e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 22494208,
      "step": 25165
    },
    {
      "epoch": 6.642998548238089,
      "grad_norm": 0.17460519075393677,
      "learning_rate": 1.5127194324044885e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 22498624,
      "step": 25170
    },
    {
      "epoch": 6.644318331793586,
      "grad_norm": 0.21749743819236755,
      "learning_rate": 1.5118175584273148e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 22503360,
      "step": 25175
    },
    {
      "epoch": 6.645638115349083,
      "grad_norm": 0.3704953193664551,
      "learning_rate": 1.5109158368402909e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 22507840,
      "step": 25180
    },
    {
      "epoch": 6.64695789890458,
      "grad_norm": 0.0787973701953888,
      "learning_rate": 1.5100142677824753e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 22512512,
      "step": 25185
    },
    {
      "epoch": 6.648277682460076,
      "grad_norm": 0.21899017691612244,
      "learning_rate": 1.509112851392901e-05,
      "loss": 0.0755,
      "num_input_tokens_seen": 22517248,
      "step": 25190
    },
    {
      "epoch": 6.649597466015574,
      "grad_norm": 0.08616723865270615,
      "learning_rate": 1.5082115878105763e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 22521760,
      "step": 25195
    },
    {
      "epoch": 6.65091724957107,
      "grad_norm": 0.23156920075416565,
      "learning_rate": 1.5073104771744892e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 22526048,
      "step": 25200
    },
    {
      "epoch": 6.65091724957107,
      "eval_loss": 0.0673813447356224,
      "eval_runtime": 64.7733,
      "eval_samples_per_second": 103.978,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 22526048,
      "step": 25200
    },
    {
      "epoch": 6.652237033126568,
      "grad_norm": 0.21575644612312317,
      "learning_rate": 1.5064095196236006e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 22530560,
      "step": 25205
    },
    {
      "epoch": 6.653556816682064,
      "grad_norm": 0.2730697989463806,
      "learning_rate": 1.50550871529685e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 22535296,
      "step": 25210
    },
    {
      "epoch": 6.654876600237561,
      "grad_norm": 0.23770639300346375,
      "learning_rate": 1.5046080643331546e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 22539392,
      "step": 25215
    },
    {
      "epoch": 6.656196383793058,
      "grad_norm": 0.14808988571166992,
      "learning_rate": 1.5037075668714028e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 22543872,
      "step": 25220
    },
    {
      "epoch": 6.657516167348555,
      "grad_norm": 0.09054088592529297,
      "learning_rate": 1.5028072230504656e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 22548512,
      "step": 25225
    },
    {
      "epoch": 6.658835950904052,
      "grad_norm": 0.023256955668330193,
      "learning_rate": 1.5019070330091861e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 22552864,
      "step": 25230
    },
    {
      "epoch": 6.660155734459549,
      "grad_norm": 0.15788036584854126,
      "learning_rate": 1.5010069968863843e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 22557120,
      "step": 25235
    },
    {
      "epoch": 6.661475518015045,
      "grad_norm": 0.3684157133102417,
      "learning_rate": 1.5001071148208584e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 22561696,
      "step": 25240
    },
    {
      "epoch": 6.662795301570543,
      "grad_norm": 0.041230540722608566,
      "learning_rate": 1.49920738695138e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 22566240,
      "step": 25245
    },
    {
      "epoch": 6.664115085126039,
      "grad_norm": 0.1836252361536026,
      "learning_rate": 1.4983078134166995e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 22570720,
      "step": 25250
    },
    {
      "epoch": 6.665434868681536,
      "grad_norm": 0.17589899897575378,
      "learning_rate": 1.4974083943555428e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 22575136,
      "step": 25255
    },
    {
      "epoch": 6.666754652237033,
      "grad_norm": 0.16102515161037445,
      "learning_rate": 1.496509129906611e-05,
      "loss": 0.1505,
      "num_input_tokens_seen": 22579424,
      "step": 25260
    },
    {
      "epoch": 6.66807443579253,
      "grad_norm": 0.5027807354927063,
      "learning_rate": 1.4956100202085809e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 22584352,
      "step": 25265
    },
    {
      "epoch": 6.6693942193480265,
      "grad_norm": 0.25196483731269836,
      "learning_rate": 1.4947110654001093e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 22588896,
      "step": 25270
    },
    {
      "epoch": 6.670714002903524,
      "grad_norm": 0.11714130640029907,
      "learning_rate": 1.4938122656198234e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 22593280,
      "step": 25275
    },
    {
      "epoch": 6.6720337864590205,
      "grad_norm": 0.25152313709259033,
      "learning_rate": 1.4929136210063316e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 22597536,
      "step": 25280
    },
    {
      "epoch": 6.673353570014518,
      "grad_norm": 0.13583233952522278,
      "learning_rate": 1.4920151316982146e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 22602048,
      "step": 25285
    },
    {
      "epoch": 6.6746733535700145,
      "grad_norm": 0.1612614542245865,
      "learning_rate": 1.4911167978340312e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 22606528,
      "step": 25290
    },
    {
      "epoch": 6.675993137125511,
      "grad_norm": 0.306769996881485,
      "learning_rate": 1.4902186195523166e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 22610816,
      "step": 25295
    },
    {
      "epoch": 6.6773129206810085,
      "grad_norm": 0.3543424606323242,
      "learning_rate": 1.4893205969915805e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 22615296,
      "step": 25300
    },
    {
      "epoch": 6.678632704236505,
      "grad_norm": 0.41265955567359924,
      "learning_rate": 1.4884227302903086e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 22619712,
      "step": 25305
    },
    {
      "epoch": 6.6799524877920025,
      "grad_norm": 0.12850342690944672,
      "learning_rate": 1.4875250195869653e-05,
      "loss": 0.0549,
      "num_input_tokens_seen": 22624352,
      "step": 25310
    },
    {
      "epoch": 6.681272271347499,
      "grad_norm": 0.15705634653568268,
      "learning_rate": 1.4866274650199862e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 22628608,
      "step": 25315
    },
    {
      "epoch": 6.682592054902996,
      "grad_norm": 0.13536995649337769,
      "learning_rate": 1.485730066727788e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 22633088,
      "step": 25320
    },
    {
      "epoch": 6.683911838458493,
      "grad_norm": 0.27390095591545105,
      "learning_rate": 1.4848328248487586e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 22637440,
      "step": 25325
    },
    {
      "epoch": 6.68523162201399,
      "grad_norm": 0.2808273434638977,
      "learning_rate": 1.4839357395212656e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 22641984,
      "step": 25330
    },
    {
      "epoch": 6.686551405569487,
      "grad_norm": 0.15451383590698242,
      "learning_rate": 1.4830388108836502e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 22646464,
      "step": 25335
    },
    {
      "epoch": 6.687871189124984,
      "grad_norm": 0.0476226881146431,
      "learning_rate": 1.4821420390742299e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 22651424,
      "step": 25340
    },
    {
      "epoch": 6.68919097268048,
      "grad_norm": 0.06523685157299042,
      "learning_rate": 1.4812454242312979e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 22655968,
      "step": 25345
    },
    {
      "epoch": 6.690510756235978,
      "grad_norm": 0.09467204660177231,
      "learning_rate": 1.4803489664931253e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 22660384,
      "step": 25350
    },
    {
      "epoch": 6.691830539791474,
      "grad_norm": 0.16191187500953674,
      "learning_rate": 1.4794526659979544e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 22664896,
      "step": 25355
    },
    {
      "epoch": 6.693150323346972,
      "grad_norm": 0.2829194962978363,
      "learning_rate": 1.4785565228840086e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 22669312,
      "step": 25360
    },
    {
      "epoch": 6.694470106902468,
      "grad_norm": 0.15123167634010315,
      "learning_rate": 1.4776605372894819e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 22673760,
      "step": 25365
    },
    {
      "epoch": 6.695789890457965,
      "grad_norm": 0.10001382231712341,
      "learning_rate": 1.4767647093525488e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 22678464,
      "step": 25370
    },
    {
      "epoch": 6.697109674013462,
      "grad_norm": 0.4783877432346344,
      "learning_rate": 1.4758690392113566e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 22682976,
      "step": 25375
    },
    {
      "epoch": 6.698429457568959,
      "grad_norm": 0.20336632430553436,
      "learning_rate": 1.4749735270040276e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 22687616,
      "step": 25380
    },
    {
      "epoch": 6.699749241124455,
      "grad_norm": 0.1614319086074829,
      "learning_rate": 1.4740781728686623e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 22692128,
      "step": 25385
    },
    {
      "epoch": 6.701069024679953,
      "grad_norm": 0.0773206576704979,
      "learning_rate": 1.4731829769433358e-05,
      "loss": 0.0275,
      "num_input_tokens_seen": 22696416,
      "step": 25390
    },
    {
      "epoch": 6.702388808235449,
      "grad_norm": 0.08341587334871292,
      "learning_rate": 1.4722879393660976e-05,
      "loss": 0.0266,
      "num_input_tokens_seen": 22700512,
      "step": 25395
    },
    {
      "epoch": 6.703708591790946,
      "grad_norm": 0.2792700529098511,
      "learning_rate": 1.4713930602749748e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 22704800,
      "step": 25400
    },
    {
      "epoch": 6.703708591790946,
      "eval_loss": 0.06725440174341202,
      "eval_runtime": 64.8237,
      "eval_samples_per_second": 103.897,
      "eval_steps_per_second": 25.978,
      "num_input_tokens_seen": 22704800,
      "step": 25400
    },
    {
      "epoch": 6.705028375346443,
      "grad_norm": 0.17292864620685577,
      "learning_rate": 1.470498339807968e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 22709152,
      "step": 25405
    },
    {
      "epoch": 6.70634815890194,
      "grad_norm": 0.21330343186855316,
      "learning_rate": 1.4696037781030542e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 22713408,
      "step": 25410
    },
    {
      "epoch": 6.707667942457437,
      "grad_norm": 0.23509669303894043,
      "learning_rate": 1.4687093752981876e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 22717696,
      "step": 25415
    },
    {
      "epoch": 6.708987726012934,
      "grad_norm": 0.18300104141235352,
      "learning_rate": 1.4678151315312943e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 22722560,
      "step": 25420
    },
    {
      "epoch": 6.71030750956843,
      "grad_norm": 0.295907199382782,
      "learning_rate": 1.4669210469402789e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 22727104,
      "step": 25425
    },
    {
      "epoch": 6.711627293123928,
      "grad_norm": 0.2507394254207611,
      "learning_rate": 1.4660271216630218e-05,
      "loss": 0.0271,
      "num_input_tokens_seen": 22731712,
      "step": 25430
    },
    {
      "epoch": 6.712947076679424,
      "grad_norm": 0.40745922923088074,
      "learning_rate": 1.4651333558373748e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 22736224,
      "step": 25435
    },
    {
      "epoch": 6.714266860234922,
      "grad_norm": 0.03623587638139725,
      "learning_rate": 1.4642397496011707e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 22740736,
      "step": 25440
    },
    {
      "epoch": 6.715586643790418,
      "grad_norm": 0.08631671965122223,
      "learning_rate": 1.4633463030922129e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 22744992,
      "step": 25445
    },
    {
      "epoch": 6.716906427345915,
      "grad_norm": 0.2634826898574829,
      "learning_rate": 1.462453016448282e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 22749184,
      "step": 25450
    },
    {
      "epoch": 6.718226210901412,
      "grad_norm": 0.19099149107933044,
      "learning_rate": 1.4615598898071354e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 22753472,
      "step": 25455
    },
    {
      "epoch": 6.719545994456909,
      "grad_norm": 0.33130350708961487,
      "learning_rate": 1.4606669233065026e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 22757984,
      "step": 25460
    },
    {
      "epoch": 6.720865778012406,
      "grad_norm": 0.25199878215789795,
      "learning_rate": 1.4597741170840914e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 22762240,
      "step": 25465
    },
    {
      "epoch": 6.722185561567903,
      "grad_norm": 0.1882053017616272,
      "learning_rate": 1.4588814712775853e-05,
      "loss": 0.0575,
      "num_input_tokens_seen": 22767008,
      "step": 25470
    },
    {
      "epoch": 6.7235053451233995,
      "grad_norm": 0.18165087699890137,
      "learning_rate": 1.4579889860246382e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 22771648,
      "step": 25475
    },
    {
      "epoch": 6.724825128678897,
      "grad_norm": 0.15447381138801575,
      "learning_rate": 1.457096661462885e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 22776000,
      "step": 25480
    },
    {
      "epoch": 6.7261449122343935,
      "grad_norm": 0.2860298156738281,
      "learning_rate": 1.4562044977299322e-05,
      "loss": 0.1682,
      "num_input_tokens_seen": 22780416,
      "step": 25485
    },
    {
      "epoch": 6.727464695789891,
      "grad_norm": 0.0845583975315094,
      "learning_rate": 1.4553124949633623e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 22785088,
      "step": 25490
    },
    {
      "epoch": 6.7287844793453875,
      "grad_norm": 0.3871162235736847,
      "learning_rate": 1.4544206533007354e-05,
      "loss": 0.1165,
      "num_input_tokens_seen": 22789760,
      "step": 25495
    },
    {
      "epoch": 6.730104262900884,
      "grad_norm": 0.0702676922082901,
      "learning_rate": 1.4535289728795821e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 22794464,
      "step": 25500
    },
    {
      "epoch": 6.7314240464563815,
      "grad_norm": 0.07201121747493744,
      "learning_rate": 1.4526374538374132e-05,
      "loss": 0.0796,
      "num_input_tokens_seen": 22799072,
      "step": 25505
    },
    {
      "epoch": 6.732743830011878,
      "grad_norm": 0.14322853088378906,
      "learning_rate": 1.4517460963117097e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 22803424,
      "step": 25510
    },
    {
      "epoch": 6.734063613567375,
      "grad_norm": 0.11913283169269562,
      "learning_rate": 1.4508549004399314e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 22808064,
      "step": 25515
    },
    {
      "epoch": 6.735383397122872,
      "grad_norm": 0.2387559562921524,
      "learning_rate": 1.449963866359513e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 22812608,
      "step": 25520
    },
    {
      "epoch": 6.736703180678369,
      "grad_norm": 0.05257706344127655,
      "learning_rate": 1.4490729942078607e-05,
      "loss": 0.0273,
      "num_input_tokens_seen": 22816960,
      "step": 25525
    },
    {
      "epoch": 6.738022964233865,
      "grad_norm": 0.03567688539624214,
      "learning_rate": 1.4481822841223608e-05,
      "loss": 0.0138,
      "num_input_tokens_seen": 22821120,
      "step": 25530
    },
    {
      "epoch": 6.739342747789363,
      "grad_norm": 0.1068384200334549,
      "learning_rate": 1.4472917362403704e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 22825216,
      "step": 25535
    },
    {
      "epoch": 6.740662531344859,
      "grad_norm": 0.04199467599391937,
      "learning_rate": 1.4464013506992224e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 22829952,
      "step": 25540
    },
    {
      "epoch": 6.741982314900357,
      "grad_norm": 0.21400941908359528,
      "learning_rate": 1.4455111276362277e-05,
      "loss": 0.1039,
      "num_input_tokens_seen": 22834368,
      "step": 25545
    },
    {
      "epoch": 6.743302098455853,
      "grad_norm": 0.4346313774585724,
      "learning_rate": 1.4446210671886676e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 22838784,
      "step": 25550
    },
    {
      "epoch": 6.74462188201135,
      "grad_norm": 0.11817231774330139,
      "learning_rate": 1.4437311694938015e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 22843232,
      "step": 25555
    },
    {
      "epoch": 6.745941665566847,
      "grad_norm": 0.060409996658563614,
      "learning_rate": 1.442841434688864e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 22847808,
      "step": 25560
    },
    {
      "epoch": 6.747261449122344,
      "grad_norm": 0.33465641736984253,
      "learning_rate": 1.4419518629110615e-05,
      "loss": 0.1271,
      "num_input_tokens_seen": 22852224,
      "step": 25565
    },
    {
      "epoch": 6.748581232677841,
      "grad_norm": 0.18665072321891785,
      "learning_rate": 1.4410624542975778e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 22856736,
      "step": 25570
    },
    {
      "epoch": 6.749901016233338,
      "grad_norm": 0.11938192695379257,
      "learning_rate": 1.4401732089855724e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 22861088,
      "step": 25575
    },
    {
      "epoch": 6.751220799788834,
      "grad_norm": 0.24802088737487793,
      "learning_rate": 1.4392841271121754e-05,
      "loss": 0.0664,
      "num_input_tokens_seen": 22865440,
      "step": 25580
    },
    {
      "epoch": 6.752540583344332,
      "grad_norm": 0.09298299252986908,
      "learning_rate": 1.438395208814497e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 22869728,
      "step": 25585
    },
    {
      "epoch": 6.753860366899828,
      "grad_norm": 0.08434057980775833,
      "learning_rate": 1.4375064542296174e-05,
      "loss": 0.1036,
      "num_input_tokens_seen": 22874400,
      "step": 25590
    },
    {
      "epoch": 6.755180150455326,
      "grad_norm": 0.43707507848739624,
      "learning_rate": 1.4366178634945946e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 22878784,
      "step": 25595
    },
    {
      "epoch": 6.756499934010822,
      "grad_norm": 0.4586452543735504,
      "learning_rate": 1.4357294367464616e-05,
      "loss": 0.0837,
      "num_input_tokens_seen": 22883200,
      "step": 25600
    },
    {
      "epoch": 6.756499934010822,
      "eval_loss": 0.06741667538881302,
      "eval_runtime": 65.0615,
      "eval_samples_per_second": 103.517,
      "eval_steps_per_second": 25.883,
      "num_input_tokens_seen": 22883200,
      "step": 25600
    },
    {
      "epoch": 6.757819717566319,
      "grad_norm": 0.18190863728523254,
      "learning_rate": 1.434841174122224e-05,
      "loss": 0.031,
      "num_input_tokens_seen": 22887648,
      "step": 25605
    },
    {
      "epoch": 6.759139501121816,
      "grad_norm": 0.05596841871738434,
      "learning_rate": 1.4339530757588615e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 22891968,
      "step": 25610
    },
    {
      "epoch": 6.760459284677313,
      "grad_norm": 0.07477214932441711,
      "learning_rate": 1.433065141793333e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 22896512,
      "step": 25615
    },
    {
      "epoch": 6.76177906823281,
      "grad_norm": 0.08816975355148315,
      "learning_rate": 1.4321773723625665e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 22900960,
      "step": 25620
    },
    {
      "epoch": 6.763098851788307,
      "grad_norm": 0.2804015874862671,
      "learning_rate": 1.4312897676034693e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 22905600,
      "step": 25625
    },
    {
      "epoch": 6.764418635343803,
      "grad_norm": 0.10673972964286804,
      "learning_rate": 1.4304023276529188e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 22909728,
      "step": 25630
    },
    {
      "epoch": 6.765738418899301,
      "grad_norm": 0.1238364577293396,
      "learning_rate": 1.4295150526477712e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 22914336,
      "step": 25635
    },
    {
      "epoch": 6.767058202454797,
      "grad_norm": 0.39629462361335754,
      "learning_rate": 1.4286279427248562e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 22919104,
      "step": 25640
    },
    {
      "epoch": 6.768377986010294,
      "grad_norm": 0.3609510660171509,
      "learning_rate": 1.4277409980209747e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 22923680,
      "step": 25645
    },
    {
      "epoch": 6.769697769565791,
      "grad_norm": 0.429675430059433,
      "learning_rate": 1.4268542186729061e-05,
      "loss": 0.0777,
      "num_input_tokens_seen": 22928320,
      "step": 25650
    },
    {
      "epoch": 6.771017553121288,
      "grad_norm": 0.23814165592193604,
      "learning_rate": 1.4259676048174043e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 22932800,
      "step": 25655
    },
    {
      "epoch": 6.772337336676785,
      "grad_norm": 0.034748248755931854,
      "learning_rate": 1.4250811565911937e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 22937152,
      "step": 25660
    },
    {
      "epoch": 6.773657120232282,
      "grad_norm": 0.5204176902770996,
      "learning_rate": 1.4241948741309782e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 22941728,
      "step": 25665
    },
    {
      "epoch": 6.7749769037877785,
      "grad_norm": 0.1144670844078064,
      "learning_rate": 1.4233087575734317e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 22946112,
      "step": 25670
    },
    {
      "epoch": 6.776296687343276,
      "grad_norm": 0.06653739511966705,
      "learning_rate": 1.422422807055206e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 22950784,
      "step": 25675
    },
    {
      "epoch": 6.7776164708987725,
      "grad_norm": 0.16192932426929474,
      "learning_rate": 1.4215370227129243e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 22954880,
      "step": 25680
    },
    {
      "epoch": 6.778936254454269,
      "grad_norm": 0.06337298452854156,
      "learning_rate": 1.4206514046831876e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 22959648,
      "step": 25685
    },
    {
      "epoch": 6.7802560380097665,
      "grad_norm": 0.1323336809873581,
      "learning_rate": 1.419765953102567e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 22964192,
      "step": 25690
    },
    {
      "epoch": 6.781575821565263,
      "grad_norm": 0.039876069873571396,
      "learning_rate": 1.4188806681076125e-05,
      "loss": 0.0273,
      "num_input_tokens_seen": 22968608,
      "step": 25695
    },
    {
      "epoch": 6.7828956051207605,
      "grad_norm": 0.17432242631912231,
      "learning_rate": 1.4179955498348443e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 22973024,
      "step": 25700
    },
    {
      "epoch": 6.784215388676257,
      "grad_norm": 0.15356789529323578,
      "learning_rate": 1.4171105984207605e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 22977504,
      "step": 25705
    },
    {
      "epoch": 6.785535172231754,
      "grad_norm": 0.43902450799942017,
      "learning_rate": 1.4162258140018304e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 22981888,
      "step": 25710
    },
    {
      "epoch": 6.786854955787251,
      "grad_norm": 0.08482851833105087,
      "learning_rate": 1.4153411967144986e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 22986624,
      "step": 25715
    },
    {
      "epoch": 6.788174739342748,
      "grad_norm": 0.35725027322769165,
      "learning_rate": 1.4144567466951864e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 22991008,
      "step": 25720
    },
    {
      "epoch": 6.789494522898245,
      "grad_norm": 0.25625497102737427,
      "learning_rate": 1.4135724640802844e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 22995680,
      "step": 25725
    },
    {
      "epoch": 6.790814306453742,
      "grad_norm": 0.22969114780426025,
      "learning_rate": 1.4126883490061615e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 23000128,
      "step": 25730
    },
    {
      "epoch": 6.792134090009238,
      "grad_norm": 0.24468931555747986,
      "learning_rate": 1.4118044016091603e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 23004832,
      "step": 25735
    },
    {
      "epoch": 6.793453873564736,
      "grad_norm": 0.09296294301748276,
      "learning_rate": 1.410920622025594e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 23009184,
      "step": 25740
    },
    {
      "epoch": 6.794773657120232,
      "grad_norm": 0.1221107766032219,
      "learning_rate": 1.4100370103917554e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 23013920,
      "step": 25745
    },
    {
      "epoch": 6.79609344067573,
      "grad_norm": 0.05854547768831253,
      "learning_rate": 1.409153566843907e-05,
      "loss": 0.1117,
      "num_input_tokens_seen": 23018400,
      "step": 25750
    },
    {
      "epoch": 6.797413224231226,
      "grad_norm": 0.052798379212617874,
      "learning_rate": 1.408270291518286e-05,
      "loss": 0.0168,
      "num_input_tokens_seen": 23022912,
      "step": 25755
    },
    {
      "epoch": 6.798733007786723,
      "grad_norm": 0.12160706520080566,
      "learning_rate": 1.407387184551107e-05,
      "loss": 0.0814,
      "num_input_tokens_seen": 23027296,
      "step": 25760
    },
    {
      "epoch": 6.80005279134222,
      "grad_norm": 0.5155794620513916,
      "learning_rate": 1.4065042460785532e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 23031392,
      "step": 25765
    },
    {
      "epoch": 6.801372574897717,
      "grad_norm": 0.11116791516542435,
      "learning_rate": 1.405621476236787e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 23035584,
      "step": 25770
    },
    {
      "epoch": 6.802692358453213,
      "grad_norm": 0.2727549076080322,
      "learning_rate": 1.4047388751619423e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 23040288,
      "step": 25775
    },
    {
      "epoch": 6.804012142008711,
      "grad_norm": 0.18266388773918152,
      "learning_rate": 1.4038564429901264e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 23044800,
      "step": 25780
    },
    {
      "epoch": 6.805331925564207,
      "grad_norm": 0.0636642724275589,
      "learning_rate": 1.4029741798574227e-05,
      "loss": 0.0189,
      "num_input_tokens_seen": 23049184,
      "step": 25785
    },
    {
      "epoch": 6.806651709119705,
      "grad_norm": 0.1771543323993683,
      "learning_rate": 1.402092085899886e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 23053952,
      "step": 25790
    },
    {
      "epoch": 6.807971492675201,
      "grad_norm": 0.12185461074113846,
      "learning_rate": 1.4012101612535464e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 23058720,
      "step": 25795
    },
    {
      "epoch": 6.809291276230698,
      "grad_norm": 0.20256946980953217,
      "learning_rate": 1.4003284060544092e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 23063104,
      "step": 25800
    },
    {
      "epoch": 6.809291276230698,
      "eval_loss": 0.06739328056573868,
      "eval_runtime": 64.7738,
      "eval_samples_per_second": 103.977,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 23063104,
      "step": 25800
    },
    {
      "epoch": 6.810611059786195,
      "grad_norm": 0.4812100827693939,
      "learning_rate": 1.3994468204384504e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 23067360,
      "step": 25805
    },
    {
      "epoch": 6.811930843341692,
      "grad_norm": 0.23098504543304443,
      "learning_rate": 1.398565404541622e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 23071520,
      "step": 25810
    },
    {
      "epoch": 6.813250626897188,
      "grad_norm": 0.39465436339378357,
      "learning_rate": 1.3976841584998513e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 23076384,
      "step": 25815
    },
    {
      "epoch": 6.814570410452686,
      "grad_norm": 0.23772825300693512,
      "learning_rate": 1.3968030824490352e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 23080896,
      "step": 25820
    },
    {
      "epoch": 6.815890194008182,
      "grad_norm": 0.22949117422103882,
      "learning_rate": 1.3959221765250469e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 23085216,
      "step": 25825
    },
    {
      "epoch": 6.81720997756368,
      "grad_norm": 0.3033384084701538,
      "learning_rate": 1.3950414408637343e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 23089472,
      "step": 25830
    },
    {
      "epoch": 6.818529761119176,
      "grad_norm": 0.4588758051395416,
      "learning_rate": 1.3941608756009166e-05,
      "loss": 0.0624,
      "num_input_tokens_seen": 23093824,
      "step": 25835
    },
    {
      "epoch": 6.819849544674673,
      "grad_norm": 0.18282847106456757,
      "learning_rate": 1.3932804808723898e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 23098304,
      "step": 25840
    },
    {
      "epoch": 6.82116932823017,
      "grad_norm": 0.05114181712269783,
      "learning_rate": 1.3924002568139194e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 23102752,
      "step": 25845
    },
    {
      "epoch": 6.822489111785667,
      "grad_norm": 0.2718784511089325,
      "learning_rate": 1.3915202035612485e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 23107360,
      "step": 25850
    },
    {
      "epoch": 6.823808895341164,
      "grad_norm": 0.09425494074821472,
      "learning_rate": 1.3906403212500935e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 23111968,
      "step": 25855
    },
    {
      "epoch": 6.825128678896661,
      "grad_norm": 0.18809205293655396,
      "learning_rate": 1.3897606100161409e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 23116416,
      "step": 25860
    },
    {
      "epoch": 6.8264484624521575,
      "grad_norm": 0.08508098870515823,
      "learning_rate": 1.388881069995055e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 23121184,
      "step": 25865
    },
    {
      "epoch": 6.827768246007655,
      "grad_norm": 0.3545161783695221,
      "learning_rate": 1.3880017013224708e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 23125536,
      "step": 25870
    },
    {
      "epoch": 6.8290880295631515,
      "grad_norm": 0.3881031572818756,
      "learning_rate": 1.3871225041339984e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 23129824,
      "step": 25875
    },
    {
      "epoch": 6.830407813118649,
      "grad_norm": 0.09362603724002838,
      "learning_rate": 1.386243478565222e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 23134496,
      "step": 25880
    },
    {
      "epoch": 6.8317275966741455,
      "grad_norm": 0.1557009518146515,
      "learning_rate": 1.3853646247516966e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 23138848,
      "step": 25885
    },
    {
      "epoch": 6.833047380229642,
      "grad_norm": 0.2811470031738281,
      "learning_rate": 1.3844859428289545e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 23143488,
      "step": 25890
    },
    {
      "epoch": 6.8343671637851395,
      "grad_norm": 0.15471263229846954,
      "learning_rate": 1.3836074329324984e-05,
      "loss": 0.0275,
      "num_input_tokens_seen": 23147712,
      "step": 25895
    },
    {
      "epoch": 6.835686947340636,
      "grad_norm": 0.3714939057826996,
      "learning_rate": 1.3827290951978044e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 23152192,
      "step": 25900
    },
    {
      "epoch": 6.8370067308961335,
      "grad_norm": 0.19825437664985657,
      "learning_rate": 1.381850929760326e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 23156576,
      "step": 25905
    },
    {
      "epoch": 6.83832651445163,
      "grad_norm": 0.12204756587743759,
      "learning_rate": 1.3809729367554842e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 23161184,
      "step": 25910
    },
    {
      "epoch": 6.839646298007127,
      "grad_norm": 0.06243292614817619,
      "learning_rate": 1.3800951163186784e-05,
      "loss": 0.0158,
      "num_input_tokens_seen": 23165632,
      "step": 25915
    },
    {
      "epoch": 6.840966081562624,
      "grad_norm": 0.16030694544315338,
      "learning_rate": 1.3792174685852801e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 23170240,
      "step": 25920
    },
    {
      "epoch": 6.842285865118121,
      "grad_norm": 0.06883686035871506,
      "learning_rate": 1.378339993690632e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 23174880,
      "step": 25925
    },
    {
      "epoch": 6.843605648673617,
      "grad_norm": 0.19646264612674713,
      "learning_rate": 1.3774626917700523e-05,
      "loss": 0.0542,
      "num_input_tokens_seen": 23179488,
      "step": 25930
    },
    {
      "epoch": 6.844925432229115,
      "grad_norm": 0.34963467717170715,
      "learning_rate": 1.3765855629588334e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 23183744,
      "step": 25935
    },
    {
      "epoch": 6.846245215784611,
      "grad_norm": 0.7438236474990845,
      "learning_rate": 1.3757086073922374e-05,
      "loss": 0.1082,
      "num_input_tokens_seen": 23188416,
      "step": 25940
    },
    {
      "epoch": 6.847564999340108,
      "grad_norm": 0.2681655287742615,
      "learning_rate": 1.3748318252055038e-05,
      "loss": 0.0999,
      "num_input_tokens_seen": 23192992,
      "step": 25945
    },
    {
      "epoch": 6.848884782895605,
      "grad_norm": 0.44821324944496155,
      "learning_rate": 1.3739552165338416e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 23197600,
      "step": 25950
    },
    {
      "epoch": 6.850204566451102,
      "grad_norm": 0.13945230841636658,
      "learning_rate": 1.3730787815124354e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 23202144,
      "step": 25955
    },
    {
      "epoch": 6.851524350006599,
      "grad_norm": 0.1082872673869133,
      "learning_rate": 1.3722025202764443e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 23206624,
      "step": 25960
    },
    {
      "epoch": 6.852844133562096,
      "grad_norm": 0.1621987372636795,
      "learning_rate": 1.371326432960997e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 23211264,
      "step": 25965
    },
    {
      "epoch": 6.854163917117592,
      "grad_norm": 0.2991095781326294,
      "learning_rate": 1.3704505197011969e-05,
      "loss": 0.0364,
      "num_input_tokens_seen": 23215648,
      "step": 25970
    },
    {
      "epoch": 6.85548370067309,
      "grad_norm": 0.3187922239303589,
      "learning_rate": 1.3695747806321224e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 23220032,
      "step": 25975
    },
    {
      "epoch": 6.856803484228586,
      "grad_norm": 0.3495848476886749,
      "learning_rate": 1.3686992158888212e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 23224448,
      "step": 25980
    },
    {
      "epoch": 6.858123267784084,
      "grad_norm": 0.20014162361621857,
      "learning_rate": 1.367823825606319e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 23228864,
      "step": 25985
    },
    {
      "epoch": 6.85944305133958,
      "grad_norm": 0.10503510385751724,
      "learning_rate": 1.36694860991961e-05,
      "loss": 0.1124,
      "num_input_tokens_seen": 23233312,
      "step": 25990
    },
    {
      "epoch": 6.860762834895077,
      "grad_norm": 0.3018341064453125,
      "learning_rate": 1.3660735689636636e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 23237792,
      "step": 25995
    },
    {
      "epoch": 6.862082618450574,
      "grad_norm": 0.36216849088668823,
      "learning_rate": 1.365198702873424e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 23242080,
      "step": 26000
    },
    {
      "epoch": 6.862082618450574,
      "eval_loss": 0.06706790626049042,
      "eval_runtime": 64.7579,
      "eval_samples_per_second": 104.003,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 23242080,
      "step": 26000
    },
    {
      "epoch": 6.863402402006071,
      "grad_norm": 0.18120232224464417,
      "learning_rate": 1.364324011783804e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 23246496,
      "step": 26005
    },
    {
      "epoch": 6.864722185561568,
      "grad_norm": 0.09264125674962997,
      "learning_rate": 1.3634494958296934e-05,
      "loss": 0.048,
      "num_input_tokens_seen": 23250944,
      "step": 26010
    },
    {
      "epoch": 6.866041969117065,
      "grad_norm": 0.04794415459036827,
      "learning_rate": 1.3625751551459542e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 23254912,
      "step": 26015
    },
    {
      "epoch": 6.867361752672561,
      "grad_norm": 0.057788681238889694,
      "learning_rate": 1.3617009898674188e-05,
      "loss": 0.0153,
      "num_input_tokens_seen": 23259552,
      "step": 26020
    },
    {
      "epoch": 6.868681536228059,
      "grad_norm": 0.39926236867904663,
      "learning_rate": 1.3608270001288967e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 23264416,
      "step": 26025
    },
    {
      "epoch": 6.870001319783555,
      "grad_norm": 0.06625698506832123,
      "learning_rate": 1.359953186065166e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 23269056,
      "step": 26030
    },
    {
      "epoch": 6.871321103339053,
      "grad_norm": 0.18107272684574127,
      "learning_rate": 1.3590795478109814e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 23273408,
      "step": 26035
    },
    {
      "epoch": 6.872640886894549,
      "grad_norm": 0.22623808681964874,
      "learning_rate": 1.3582060855010675e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 23278080,
      "step": 26040
    },
    {
      "epoch": 6.873960670450046,
      "grad_norm": 0.49687936902046204,
      "learning_rate": 1.3573327992701245e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 23282784,
      "step": 26045
    },
    {
      "epoch": 6.875280454005543,
      "grad_norm": 0.21760812401771545,
      "learning_rate": 1.356459689252823e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 23287392,
      "step": 26050
    },
    {
      "epoch": 6.87660023756104,
      "grad_norm": 0.08660692721605301,
      "learning_rate": 1.3555867555838087e-05,
      "loss": 0.0135,
      "num_input_tokens_seen": 23291872,
      "step": 26055
    },
    {
      "epoch": 6.8779200211165366,
      "grad_norm": 0.1054881289601326,
      "learning_rate": 1.3547139983976975e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 23296128,
      "step": 26060
    },
    {
      "epoch": 6.879239804672034,
      "grad_norm": 0.3173291087150574,
      "learning_rate": 1.3538414178290815e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 23300352,
      "step": 26065
    },
    {
      "epoch": 6.8805595882275306,
      "grad_norm": 0.20991815626621246,
      "learning_rate": 1.3529690140125209e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 23304864,
      "step": 26070
    },
    {
      "epoch": 6.881879371783027,
      "grad_norm": 0.2373851090669632,
      "learning_rate": 1.352096787082553e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 23309280,
      "step": 26075
    },
    {
      "epoch": 6.8831991553385246,
      "grad_norm": 0.11811939626932144,
      "learning_rate": 1.3512247371736871e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 23313664,
      "step": 26080
    },
    {
      "epoch": 6.884518938894021,
      "grad_norm": 0.22317059338092804,
      "learning_rate": 1.3503528644204022e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 23318688,
      "step": 26085
    },
    {
      "epoch": 6.885838722449519,
      "grad_norm": 0.17305819690227509,
      "learning_rate": 1.349481168957153e-05,
      "loss": 0.0268,
      "num_input_tokens_seen": 23323008,
      "step": 26090
    },
    {
      "epoch": 6.887158506005015,
      "grad_norm": 0.21702373027801514,
      "learning_rate": 1.3486096509183665e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 23327360,
      "step": 26095
    },
    {
      "epoch": 6.888478289560512,
      "grad_norm": 0.24821020662784576,
      "learning_rate": 1.3477383104384406e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 23331968,
      "step": 26100
    },
    {
      "epoch": 6.889798073116009,
      "grad_norm": 0.5861372351646423,
      "learning_rate": 1.3468671476517481e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 23336640,
      "step": 26105
    },
    {
      "epoch": 6.891117856671506,
      "grad_norm": 0.28352785110473633,
      "learning_rate": 1.3459961626926326e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 23340992,
      "step": 26110
    },
    {
      "epoch": 6.892437640227003,
      "grad_norm": 0.03513330593705177,
      "learning_rate": 1.3451253556954101e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 23345632,
      "step": 26115
    },
    {
      "epoch": 6.8937574237825,
      "grad_norm": 0.35132622718811035,
      "learning_rate": 1.3442547267943717e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 23350112,
      "step": 26120
    },
    {
      "epoch": 6.895077207337996,
      "grad_norm": 0.3103977143764496,
      "learning_rate": 1.3433842761237774e-05,
      "loss": 0.0445,
      "num_input_tokens_seen": 23354528,
      "step": 26125
    },
    {
      "epoch": 6.896396990893494,
      "grad_norm": 0.4659053385257721,
      "learning_rate": 1.3425140038178639e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 23358816,
      "step": 26130
    },
    {
      "epoch": 6.89771677444899,
      "grad_norm": 0.24028538167476654,
      "learning_rate": 1.3416439100108358e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 23363392,
      "step": 26135
    },
    {
      "epoch": 6.899036558004488,
      "grad_norm": 0.10888324677944183,
      "learning_rate": 1.3407739948368734e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 23367712,
      "step": 26140
    },
    {
      "epoch": 6.900356341559984,
      "grad_norm": 0.08844508975744247,
      "learning_rate": 1.3399042584301298e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 23372384,
      "step": 26145
    },
    {
      "epoch": 6.901676125115481,
      "grad_norm": 0.2642793655395508,
      "learning_rate": 1.3390347009247272e-05,
      "loss": 0.0543,
      "num_input_tokens_seen": 23376544,
      "step": 26150
    },
    {
      "epoch": 6.902995908670978,
      "grad_norm": 0.0783705785870552,
      "learning_rate": 1.3381653224547635e-05,
      "loss": 0.0741,
      "num_input_tokens_seen": 23380928,
      "step": 26155
    },
    {
      "epoch": 6.904315692226475,
      "grad_norm": 0.3414996862411499,
      "learning_rate": 1.3372961231543086e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 23385568,
      "step": 26160
    },
    {
      "epoch": 6.905635475781972,
      "grad_norm": 0.20869261026382446,
      "learning_rate": 1.3364271031574016e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 23390176,
      "step": 26165
    },
    {
      "epoch": 6.906955259337469,
      "grad_norm": 0.37239232659339905,
      "learning_rate": 1.335558262598059e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 23394656,
      "step": 26170
    },
    {
      "epoch": 6.908275042892965,
      "grad_norm": 0.051340557634830475,
      "learning_rate": 1.3346896016102645e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 23399296,
      "step": 26175
    },
    {
      "epoch": 6.909594826448463,
      "grad_norm": 0.33996284008026123,
      "learning_rate": 1.3338211203279788e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 23403776,
      "step": 26180
    },
    {
      "epoch": 6.910914610003959,
      "grad_norm": 0.050017885863780975,
      "learning_rate": 1.3329528188851303e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 23408192,
      "step": 26185
    },
    {
      "epoch": 6.912234393559456,
      "grad_norm": 0.2997513711452484,
      "learning_rate": 1.3320846974156242e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 23412416,
      "step": 26190
    },
    {
      "epoch": 6.913554177114953,
      "grad_norm": 0.2910624146461487,
      "learning_rate": 1.3312167560533337e-05,
      "loss": 0.1263,
      "num_input_tokens_seen": 23416800,
      "step": 26195
    },
    {
      "epoch": 6.91487396067045,
      "grad_norm": 0.02987534925341606,
      "learning_rate": 1.3303489949321082e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 23421312,
      "step": 26200
    },
    {
      "epoch": 6.91487396067045,
      "eval_loss": 0.06737001985311508,
      "eval_runtime": 64.7831,
      "eval_samples_per_second": 103.962,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 23421312,
      "step": 26200
    },
    {
      "epoch": 6.9161937442259465,
      "grad_norm": 0.10874485969543457,
      "learning_rate": 1.3294814141857653e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 23425888,
      "step": 26205
    },
    {
      "epoch": 6.917513527781444,
      "grad_norm": 0.02958831377327442,
      "learning_rate": 1.3286140139480992e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 23430432,
      "step": 26210
    },
    {
      "epoch": 6.9188333113369405,
      "grad_norm": 0.2245890349149704,
      "learning_rate": 1.3277467943528719e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 23435264,
      "step": 26215
    },
    {
      "epoch": 6.920153094892438,
      "grad_norm": 0.03356247395277023,
      "learning_rate": 1.3268797555338203e-05,
      "loss": 0.083,
      "num_input_tokens_seen": 23439808,
      "step": 26220
    },
    {
      "epoch": 6.9214728784479345,
      "grad_norm": 0.4324404299259186,
      "learning_rate": 1.3260128976246533e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 23444128,
      "step": 26225
    },
    {
      "epoch": 6.922792662003431,
      "grad_norm": 0.25428101420402527,
      "learning_rate": 1.32514622075905e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 23448544,
      "step": 26230
    },
    {
      "epoch": 6.9241124455589285,
      "grad_norm": 0.19580595195293427,
      "learning_rate": 1.3242797250706638e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 23453216,
      "step": 26235
    },
    {
      "epoch": 6.925432229114425,
      "grad_norm": 0.45693421363830566,
      "learning_rate": 1.3234134106931195e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 23457856,
      "step": 26240
    },
    {
      "epoch": 6.9267520126699225,
      "grad_norm": 0.17358365654945374,
      "learning_rate": 1.322547277760013e-05,
      "loss": 0.0417,
      "num_input_tokens_seen": 23462400,
      "step": 26245
    },
    {
      "epoch": 6.928071796225419,
      "grad_norm": 0.1895885169506073,
      "learning_rate": 1.3216813264049132e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 23466848,
      "step": 26250
    },
    {
      "epoch": 6.929391579780916,
      "grad_norm": 0.062415365129709244,
      "learning_rate": 1.32081555676136e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 23471200,
      "step": 26255
    },
    {
      "epoch": 6.930711363336413,
      "grad_norm": 0.12897033989429474,
      "learning_rate": 1.3199499689628674e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 23475488,
      "step": 26260
    },
    {
      "epoch": 6.93203114689191,
      "grad_norm": 0.38613736629486084,
      "learning_rate": 1.3190845631429192e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 23479872,
      "step": 26265
    },
    {
      "epoch": 6.933350930447407,
      "grad_norm": 0.14430344104766846,
      "learning_rate": 1.3182193394349704e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 23484512,
      "step": 26270
    },
    {
      "epoch": 6.934670714002904,
      "grad_norm": 0.285412073135376,
      "learning_rate": 1.3173542979724507e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 23488992,
      "step": 26275
    },
    {
      "epoch": 6.9359904975584,
      "grad_norm": 0.08865345269441605,
      "learning_rate": 1.3164894388887617e-05,
      "loss": 0.0248,
      "num_input_tokens_seen": 23493216,
      "step": 26280
    },
    {
      "epoch": 6.937310281113898,
      "grad_norm": 0.37336641550064087,
      "learning_rate": 1.3156247623172727e-05,
      "loss": 0.0952,
      "num_input_tokens_seen": 23497536,
      "step": 26285
    },
    {
      "epoch": 6.938630064669394,
      "grad_norm": 0.17545442283153534,
      "learning_rate": 1.3147602683913302e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 23501888,
      "step": 26290
    },
    {
      "epoch": 6.939949848224892,
      "grad_norm": 0.06161971762776375,
      "learning_rate": 1.3138959572442481e-05,
      "loss": 0.0259,
      "num_input_tokens_seen": 23506560,
      "step": 26295
    },
    {
      "epoch": 6.941269631780388,
      "grad_norm": 0.2512925863265991,
      "learning_rate": 1.3130318290093146e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 23511104,
      "step": 26300
    },
    {
      "epoch": 6.942589415335885,
      "grad_norm": 0.38723674416542053,
      "learning_rate": 1.3121678838197909e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 23515392,
      "step": 26305
    },
    {
      "epoch": 6.943909198891382,
      "grad_norm": 0.03177139163017273,
      "learning_rate": 1.3113041218089056e-05,
      "loss": 0.1097,
      "num_input_tokens_seen": 23519680,
      "step": 26310
    },
    {
      "epoch": 6.945228982446879,
      "grad_norm": 0.11637531965970993,
      "learning_rate": 1.3104405431098626e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 23524192,
      "step": 26315
    },
    {
      "epoch": 6.946548766002375,
      "grad_norm": 0.34624460339546204,
      "learning_rate": 1.3095771478558377e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 23528576,
      "step": 26320
    },
    {
      "epoch": 6.947868549557873,
      "grad_norm": 0.3507158160209656,
      "learning_rate": 1.3087139361799766e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 23533376,
      "step": 26325
    },
    {
      "epoch": 6.949188333113369,
      "grad_norm": 0.2701948881149292,
      "learning_rate": 1.3078509082153964e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 23538048,
      "step": 26330
    },
    {
      "epoch": 6.950508116668866,
      "grad_norm": 0.11996275931596756,
      "learning_rate": 1.3069880640951885e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 23542336,
      "step": 26335
    },
    {
      "epoch": 6.951827900224363,
      "grad_norm": 0.2698823809623718,
      "learning_rate": 1.3061254039524123e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 23546624,
      "step": 26340
    },
    {
      "epoch": 6.95314768377986,
      "grad_norm": 0.2736412286758423,
      "learning_rate": 1.3052629279201028e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 23551104,
      "step": 26345
    },
    {
      "epoch": 6.954467467335357,
      "grad_norm": 0.10672712326049805,
      "learning_rate": 1.3044006361312633e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 23555104,
      "step": 26350
    },
    {
      "epoch": 6.955787250890854,
      "grad_norm": 0.22339585423469543,
      "learning_rate": 1.30353852871887e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 23559616,
      "step": 26355
    },
    {
      "epoch": 6.95710703444635,
      "grad_norm": 0.4189163148403168,
      "learning_rate": 1.302676605815873e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 23564064,
      "step": 26360
    },
    {
      "epoch": 6.958426818001848,
      "grad_norm": 0.2866857945919037,
      "learning_rate": 1.3018148675551884e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 23568320,
      "step": 26365
    },
    {
      "epoch": 6.959746601557344,
      "grad_norm": 0.24438057839870453,
      "learning_rate": 1.3009533140697094e-05,
      "loss": 0.1338,
      "num_input_tokens_seen": 23572640,
      "step": 26370
    },
    {
      "epoch": 6.961066385112842,
      "grad_norm": 0.5954139828681946,
      "learning_rate": 1.3000919454922966e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 23576992,
      "step": 26375
    },
    {
      "epoch": 6.962386168668338,
      "grad_norm": 0.21665580570697784,
      "learning_rate": 1.299230761955785e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 23581312,
      "step": 26380
    },
    {
      "epoch": 6.963705952223835,
      "grad_norm": 0.02980898879468441,
      "learning_rate": 1.2983697635929807e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 23586240,
      "step": 26385
    },
    {
      "epoch": 6.965025735779332,
      "grad_norm": 0.10365483164787292,
      "learning_rate": 1.2975089505366584e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 23590560,
      "step": 26390
    },
    {
      "epoch": 6.966345519334829,
      "grad_norm": 0.05207838490605354,
      "learning_rate": 1.2966483229195683e-05,
      "loss": 0.0127,
      "num_input_tokens_seen": 23594624,
      "step": 26395
    },
    {
      "epoch": 6.967665302890326,
      "grad_norm": 0.21801352500915527,
      "learning_rate": 1.2957878808744283e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 23599008,
      "step": 26400
    },
    {
      "epoch": 6.967665302890326,
      "eval_loss": 0.06753312051296234,
      "eval_runtime": 64.7619,
      "eval_samples_per_second": 103.996,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 23599008,
      "step": 26400
    },
    {
      "epoch": 6.968985086445823,
      "grad_norm": 0.22268319129943848,
      "learning_rate": 1.294927624533931e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 23603296,
      "step": 26405
    },
    {
      "epoch": 6.9703048700013195,
      "grad_norm": 0.1794978231191635,
      "learning_rate": 1.2940675540307378e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 23607584,
      "step": 26410
    },
    {
      "epoch": 6.971624653556817,
      "grad_norm": 0.14915111660957336,
      "learning_rate": 1.2932076694974814e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 23611840,
      "step": 26415
    },
    {
      "epoch": 6.9729444371123135,
      "grad_norm": 0.05516654998064041,
      "learning_rate": 1.2923479710667682e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 23616448,
      "step": 26420
    },
    {
      "epoch": 6.974264220667811,
      "grad_norm": 0.23909355700016022,
      "learning_rate": 1.2914884588711751e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 23620928,
      "step": 26425
    },
    {
      "epoch": 6.9755840042233075,
      "grad_norm": 0.06487134099006653,
      "learning_rate": 1.2906291330432475e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 23625504,
      "step": 26430
    },
    {
      "epoch": 6.976903787778804,
      "grad_norm": 0.01848084293305874,
      "learning_rate": 1.2897699937155055e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 23630176,
      "step": 26435
    },
    {
      "epoch": 6.9782235713343015,
      "grad_norm": 0.2839110493659973,
      "learning_rate": 1.2889110410204403e-05,
      "loss": 0.1151,
      "num_input_tokens_seen": 23634528,
      "step": 26440
    },
    {
      "epoch": 6.979543354889798,
      "grad_norm": 0.10108304768800735,
      "learning_rate": 1.2880522750905111e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 23638816,
      "step": 26445
    },
    {
      "epoch": 6.980863138445295,
      "grad_norm": 0.2128608524799347,
      "learning_rate": 1.2871936960581523e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 23643360,
      "step": 26450
    },
    {
      "epoch": 6.982182922000792,
      "grad_norm": 0.09511413425207138,
      "learning_rate": 1.2863353040557658e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 23647872,
      "step": 26455
    },
    {
      "epoch": 6.983502705556289,
      "grad_norm": 0.35618311166763306,
      "learning_rate": 1.2854770992157273e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 23652352,
      "step": 26460
    },
    {
      "epoch": 6.984822489111786,
      "grad_norm": 0.288748174905777,
      "learning_rate": 1.2846190816703835e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 23656864,
      "step": 26465
    },
    {
      "epoch": 6.986142272667283,
      "grad_norm": 0.3650299906730652,
      "learning_rate": 1.2837612515520498e-05,
      "loss": 0.105,
      "num_input_tokens_seen": 23661344,
      "step": 26470
    },
    {
      "epoch": 6.987462056222779,
      "grad_norm": 0.19632835686206818,
      "learning_rate": 1.2829036089930163e-05,
      "loss": 0.0326,
      "num_input_tokens_seen": 23665952,
      "step": 26475
    },
    {
      "epoch": 6.988781839778277,
      "grad_norm": 0.0808877944946289,
      "learning_rate": 1.2820461541255412e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 23670752,
      "step": 26480
    },
    {
      "epoch": 6.990101623333773,
      "grad_norm": 0.19864478707313538,
      "learning_rate": 1.2811888870818543e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 23675584,
      "step": 26485
    },
    {
      "epoch": 6.99142140688927,
      "grad_norm": 0.1210024431347847,
      "learning_rate": 1.2803318079941581e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 23679840,
      "step": 26490
    },
    {
      "epoch": 6.992741190444767,
      "grad_norm": 0.0403180867433548,
      "learning_rate": 1.2794749169946235e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 23684192,
      "step": 26495
    },
    {
      "epoch": 6.994060974000264,
      "grad_norm": 0.3332643508911133,
      "learning_rate": 1.2786182142153952e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 23688512,
      "step": 26500
    },
    {
      "epoch": 6.995380757555761,
      "grad_norm": 0.46026432514190674,
      "learning_rate": 1.2777616997885878e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 23693152,
      "step": 26505
    },
    {
      "epoch": 6.996700541111258,
      "grad_norm": 0.3616117537021637,
      "learning_rate": 1.2769053738462847e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 23697600,
      "step": 26510
    },
    {
      "epoch": 6.998020324666754,
      "grad_norm": 0.15544871985912323,
      "learning_rate": 1.2760492365205434e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 23702208,
      "step": 26515
    },
    {
      "epoch": 6.999340108222252,
      "grad_norm": 0.13193447887897491,
      "learning_rate": 1.2751932879433919e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 23706720,
      "step": 26520
    },
    {
      "epoch": 7.000527913422199,
      "grad_norm": 0.09995996206998825,
      "learning_rate": 1.2743375282468267e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 23710928,
      "step": 26525
    },
    {
      "epoch": 7.001847696977696,
      "grad_norm": 0.3785555362701416,
      "learning_rate": 1.2734819575628182e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 23715696,
      "step": 26530
    },
    {
      "epoch": 7.003167480533192,
      "grad_norm": 0.07407134026288986,
      "learning_rate": 1.2726265760233039e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 23720176,
      "step": 26535
    },
    {
      "epoch": 7.00448726408869,
      "grad_norm": 0.040014345198869705,
      "learning_rate": 1.271771383760197e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 23724464,
      "step": 26540
    },
    {
      "epoch": 7.005807047644186,
      "grad_norm": 0.21336564421653748,
      "learning_rate": 1.2709163809053764e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 23728976,
      "step": 26545
    },
    {
      "epoch": 7.007126831199683,
      "grad_norm": 0.3592245578765869,
      "learning_rate": 1.2700615675906963e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 23733072,
      "step": 26550
    },
    {
      "epoch": 7.00844661475518,
      "grad_norm": 0.07405228912830353,
      "learning_rate": 1.269206943947978e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 23737616,
      "step": 26555
    },
    {
      "epoch": 7.009766398310677,
      "grad_norm": 0.1678626984357834,
      "learning_rate": 1.2683525101090177e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 23742064,
      "step": 26560
    },
    {
      "epoch": 7.011086181866174,
      "grad_norm": 0.2535615861415863,
      "learning_rate": 1.2674982662055765e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 23746224,
      "step": 26565
    },
    {
      "epoch": 7.012405965421671,
      "grad_norm": 0.17748519778251648,
      "learning_rate": 1.2666442123693922e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 23750736,
      "step": 26570
    },
    {
      "epoch": 7.013725748977167,
      "grad_norm": 0.22192297875881195,
      "learning_rate": 1.265790348732169e-05,
      "loss": 0.1172,
      "num_input_tokens_seen": 23754864,
      "step": 26575
    },
    {
      "epoch": 7.015045532532665,
      "grad_norm": 0.599966824054718,
      "learning_rate": 1.264936675425584e-05,
      "loss": 0.1043,
      "num_input_tokens_seen": 23759504,
      "step": 26580
    },
    {
      "epoch": 7.016365316088161,
      "grad_norm": 0.2450350522994995,
      "learning_rate": 1.2640831925812852e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 23764176,
      "step": 26585
    },
    {
      "epoch": 7.017685099643659,
      "grad_norm": 0.14555564522743225,
      "learning_rate": 1.263229900330889e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 23768784,
      "step": 26590
    },
    {
      "epoch": 7.019004883199155,
      "grad_norm": 0.1665077954530716,
      "learning_rate": 1.2623767988059843e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 23773008,
      "step": 26595
    },
    {
      "epoch": 7.020324666754652,
      "grad_norm": 0.3017309010028839,
      "learning_rate": 1.2615238881381309e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 23777520,
      "step": 26600
    },
    {
      "epoch": 7.020324666754652,
      "eval_loss": 0.06739455461502075,
      "eval_runtime": 64.778,
      "eval_samples_per_second": 103.97,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 23777520,
      "step": 26600
    },
    {
      "epoch": 7.021644450310149,
      "grad_norm": 0.20437005162239075,
      "learning_rate": 1.2606711684588568e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 23782160,
      "step": 26605
    },
    {
      "epoch": 7.022964233865646,
      "grad_norm": 0.18186326324939728,
      "learning_rate": 1.2598186398996636e-05,
      "loss": 0.0204,
      "num_input_tokens_seen": 23786416,
      "step": 26610
    },
    {
      "epoch": 7.0242840174211425,
      "grad_norm": 0.20386849343776703,
      "learning_rate": 1.2589663025920207e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 23791056,
      "step": 26615
    },
    {
      "epoch": 7.02560380097664,
      "grad_norm": 0.24426443874835968,
      "learning_rate": 1.2581141566673705e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 23795344,
      "step": 26620
    },
    {
      "epoch": 7.0269235845321365,
      "grad_norm": 0.30892109870910645,
      "learning_rate": 1.257262202257124e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 23799792,
      "step": 26625
    },
    {
      "epoch": 7.028243368087634,
      "grad_norm": 0.07429211586713791,
      "learning_rate": 1.2564104394926618e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 23804336,
      "step": 26630
    },
    {
      "epoch": 7.0295631516431305,
      "grad_norm": 0.28406187891960144,
      "learning_rate": 1.2555588685053383e-05,
      "loss": 0.107,
      "num_input_tokens_seen": 23808880,
      "step": 26635
    },
    {
      "epoch": 7.030882935198627,
      "grad_norm": 0.3646852970123291,
      "learning_rate": 1.2547074894264762e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 23813456,
      "step": 26640
    },
    {
      "epoch": 7.0322027187541245,
      "grad_norm": 0.2474670708179474,
      "learning_rate": 1.2538563023873679e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 23817712,
      "step": 26645
    },
    {
      "epoch": 7.033522502309621,
      "grad_norm": 0.19502291083335876,
      "learning_rate": 1.2530053075192789e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 23822160,
      "step": 26650
    },
    {
      "epoch": 7.0348422858651185,
      "grad_norm": 0.34203797578811646,
      "learning_rate": 1.252154504953441e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 23826480,
      "step": 26655
    },
    {
      "epoch": 7.036162069420615,
      "grad_norm": 0.1713544875383377,
      "learning_rate": 1.25130389482106e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 23830960,
      "step": 26660
    },
    {
      "epoch": 7.037481852976112,
      "grad_norm": 0.11495032161474228,
      "learning_rate": 1.2504534772533116e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 23835408,
      "step": 26665
    },
    {
      "epoch": 7.038801636531609,
      "grad_norm": 0.04386752471327782,
      "learning_rate": 1.2496032523813387e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 23839856,
      "step": 26670
    },
    {
      "epoch": 7.040121420087106,
      "grad_norm": 0.16792790591716766,
      "learning_rate": 1.2487532203362576e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 23844528,
      "step": 26675
    },
    {
      "epoch": 7.041441203642602,
      "grad_norm": 0.06166511029005051,
      "learning_rate": 1.247903381249155e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 23849264,
      "step": 26680
    },
    {
      "epoch": 7.0427609871981,
      "grad_norm": 0.23144420981407166,
      "learning_rate": 1.2470537352510853e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 23853712,
      "step": 26685
    },
    {
      "epoch": 7.044080770753596,
      "grad_norm": 0.24693745374679565,
      "learning_rate": 1.2462042824730758e-05,
      "loss": 0.0815,
      "num_input_tokens_seen": 23857968,
      "step": 26690
    },
    {
      "epoch": 7.045400554309094,
      "grad_norm": 0.36573684215545654,
      "learning_rate": 1.245355023046122e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 23862352,
      "step": 26695
    },
    {
      "epoch": 7.04672033786459,
      "grad_norm": 0.057069532573223114,
      "learning_rate": 1.2445059571011896e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 23866640,
      "step": 26700
    },
    {
      "epoch": 7.048040121420087,
      "grad_norm": 0.15891991555690765,
      "learning_rate": 1.2436570847692173e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 23870896,
      "step": 26705
    },
    {
      "epoch": 7.049359904975584,
      "grad_norm": 0.065348781645298,
      "learning_rate": 1.2428084061811096e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 23875152,
      "step": 26710
    },
    {
      "epoch": 7.050679688531081,
      "grad_norm": 0.1477564126253128,
      "learning_rate": 1.2419599214677447e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 23879504,
      "step": 26715
    },
    {
      "epoch": 7.051999472086578,
      "grad_norm": 0.37308457493782043,
      "learning_rate": 1.2411116307599702e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 23883920,
      "step": 26720
    },
    {
      "epoch": 7.053319255642075,
      "grad_norm": 0.3137248158454895,
      "learning_rate": 1.2402635341886016e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 23888400,
      "step": 26725
    },
    {
      "epoch": 7.054639039197571,
      "grad_norm": 0.2803465723991394,
      "learning_rate": 1.2394156318844278e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 23892944,
      "step": 26730
    },
    {
      "epoch": 7.055958822753069,
      "grad_norm": 0.3562689423561096,
      "learning_rate": 1.2385679239782039e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 23897552,
      "step": 26735
    },
    {
      "epoch": 7.057278606308565,
      "grad_norm": 0.21104729175567627,
      "learning_rate": 1.2377204106006585e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 23901872,
      "step": 26740
    },
    {
      "epoch": 7.058598389864063,
      "grad_norm": 0.2214275747537613,
      "learning_rate": 1.2368730918824891e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 23906192,
      "step": 26745
    },
    {
      "epoch": 7.059918173419559,
      "grad_norm": 0.407850056886673,
      "learning_rate": 1.236025967954362e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 23910192,
      "step": 26750
    },
    {
      "epoch": 7.061237956975056,
      "grad_norm": 0.14235419034957886,
      "learning_rate": 1.2351790389469153e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 23914768,
      "step": 26755
    },
    {
      "epoch": 7.062557740530553,
      "grad_norm": 0.10189440101385117,
      "learning_rate": 1.234332304990755e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 23919152,
      "step": 26760
    },
    {
      "epoch": 7.06387752408605,
      "grad_norm": 0.04264217987656593,
      "learning_rate": 1.2334857662164593e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 23923728,
      "step": 26765
    },
    {
      "epoch": 7.065197307641546,
      "grad_norm": 0.21801042556762695,
      "learning_rate": 1.2326394227545743e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 23928272,
      "step": 26770
    },
    {
      "epoch": 7.066517091197044,
      "grad_norm": 0.2024906724691391,
      "learning_rate": 1.2317932747356162e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 23932400,
      "step": 26775
    },
    {
      "epoch": 7.06783687475254,
      "grad_norm": 0.2778661549091339,
      "learning_rate": 1.2309473222900726e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 23936656,
      "step": 26780
    },
    {
      "epoch": 7.069156658308038,
      "grad_norm": 0.3980865478515625,
      "learning_rate": 1.2301015655484006e-05,
      "loss": 0.0854,
      "num_input_tokens_seen": 23941040,
      "step": 26785
    },
    {
      "epoch": 7.070476441863534,
      "grad_norm": 0.23265591263771057,
      "learning_rate": 1.2292560046410245e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 23945328,
      "step": 26790
    },
    {
      "epoch": 7.071796225419031,
      "grad_norm": 0.18941642343997955,
      "learning_rate": 1.228410639698343e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 23949904,
      "step": 26795
    },
    {
      "epoch": 7.073116008974528,
      "grad_norm": 0.47927021980285645,
      "learning_rate": 1.2275654708507195e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 23954320,
      "step": 26800
    },
    {
      "epoch": 7.073116008974528,
      "eval_loss": 0.06721608340740204,
      "eval_runtime": 64.779,
      "eval_samples_per_second": 103.969,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 23954320,
      "step": 26800
    },
    {
      "epoch": 7.074435792530025,
      "grad_norm": 0.030394135043025017,
      "learning_rate": 1.2267204982284908e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 23958704,
      "step": 26805
    },
    {
      "epoch": 7.0757555760855215,
      "grad_norm": 0.05200476571917534,
      "learning_rate": 1.2258757219619635e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 23963376,
      "step": 26810
    },
    {
      "epoch": 7.077075359641019,
      "grad_norm": 0.17522476613521576,
      "learning_rate": 1.2250311421814104e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 23968080,
      "step": 26815
    },
    {
      "epoch": 7.0783951431965155,
      "grad_norm": 0.030085133388638496,
      "learning_rate": 1.2241867590170772e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 23972592,
      "step": 26820
    },
    {
      "epoch": 7.079714926752013,
      "grad_norm": 0.1663040816783905,
      "learning_rate": 1.2233425725991799e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 23977232,
      "step": 26825
    },
    {
      "epoch": 7.0810347103075095,
      "grad_norm": 0.11070884764194489,
      "learning_rate": 1.2224985830579003e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 23981648,
      "step": 26830
    },
    {
      "epoch": 7.082354493863006,
      "grad_norm": 0.1555173695087433,
      "learning_rate": 1.2216547905233944e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 23986096,
      "step": 26835
    },
    {
      "epoch": 7.0836742774185035,
      "grad_norm": 0.17865745723247528,
      "learning_rate": 1.2208111951257842e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 23990800,
      "step": 26840
    },
    {
      "epoch": 7.084994060974,
      "grad_norm": 0.20165173709392548,
      "learning_rate": 1.2199677969951622e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 23995312,
      "step": 26845
    },
    {
      "epoch": 7.0863138445294975,
      "grad_norm": 0.08363146334886551,
      "learning_rate": 1.2191245962615927e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 23999440,
      "step": 26850
    },
    {
      "epoch": 7.087633628084994,
      "grad_norm": 0.03990233317017555,
      "learning_rate": 1.218281593055106e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 24004144,
      "step": 26855
    },
    {
      "epoch": 7.088953411640491,
      "grad_norm": 0.2977573871612549,
      "learning_rate": 1.217438787505705e-05,
      "loss": 0.0874,
      "num_input_tokens_seen": 24008592,
      "step": 26860
    },
    {
      "epoch": 7.090273195195988,
      "grad_norm": 0.1399790197610855,
      "learning_rate": 1.2165961797433615e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 24013136,
      "step": 26865
    },
    {
      "epoch": 7.091592978751485,
      "grad_norm": 0.14798660576343536,
      "learning_rate": 1.215753769898014e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 24017680,
      "step": 26870
    },
    {
      "epoch": 7.092912762306982,
      "grad_norm": 0.1649623066186905,
      "learning_rate": 1.2149115580995755e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 24022384,
      "step": 26875
    },
    {
      "epoch": 7.094232545862479,
      "grad_norm": 0.23406080901622772,
      "learning_rate": 1.2140695444779227e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 24026672,
      "step": 26880
    },
    {
      "epoch": 7.095552329417975,
      "grad_norm": 0.10084214806556702,
      "learning_rate": 1.2132277291629066e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 24031184,
      "step": 26885
    },
    {
      "epoch": 7.096872112973473,
      "grad_norm": 0.03310735151171684,
      "learning_rate": 1.2123861122843458e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 24035728,
      "step": 26890
    },
    {
      "epoch": 7.098191896528969,
      "grad_norm": 0.06398425251245499,
      "learning_rate": 1.2115446939720271e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 24040144,
      "step": 26895
    },
    {
      "epoch": 7.099511680084466,
      "grad_norm": 0.35846656560897827,
      "learning_rate": 1.210703474355708e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 24044464,
      "step": 26900
    },
    {
      "epoch": 7.100831463639963,
      "grad_norm": 0.10406621545553207,
      "learning_rate": 1.2098624535651164e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 24048976,
      "step": 26905
    },
    {
      "epoch": 7.10215124719546,
      "grad_norm": 0.3346163332462311,
      "learning_rate": 1.2090216317299477e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 24053232,
      "step": 26910
    },
    {
      "epoch": 7.103471030750957,
      "grad_norm": 0.07963395863771439,
      "learning_rate": 1.2081810089798668e-05,
      "loss": 0.0275,
      "num_input_tokens_seen": 24057680,
      "step": 26915
    },
    {
      "epoch": 7.104790814306454,
      "grad_norm": 0.06073209270834923,
      "learning_rate": 1.2073405854445072e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 24062384,
      "step": 26920
    },
    {
      "epoch": 7.10611059786195,
      "grad_norm": 0.3106163442134857,
      "learning_rate": 1.206500361253474e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 24066736,
      "step": 26925
    },
    {
      "epoch": 7.107430381417448,
      "grad_norm": 0.1260586529970169,
      "learning_rate": 1.2056603365363409e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 24071152,
      "step": 26930
    },
    {
      "epoch": 7.108750164972944,
      "grad_norm": 0.21250569820404053,
      "learning_rate": 1.2048205114226487e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 24075472,
      "step": 26935
    },
    {
      "epoch": 7.110069948528442,
      "grad_norm": 0.09910744428634644,
      "learning_rate": 1.2039808860419102e-05,
      "loss": 0.0544,
      "num_input_tokens_seen": 24079728,
      "step": 26940
    },
    {
      "epoch": 7.111389732083938,
      "grad_norm": 0.2636645436286926,
      "learning_rate": 1.2031414605236066e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 24084240,
      "step": 26945
    },
    {
      "epoch": 7.112709515639435,
      "grad_norm": 0.08178254961967468,
      "learning_rate": 1.2023022349971862e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 24089136,
      "step": 26950
    },
    {
      "epoch": 7.114029299194932,
      "grad_norm": 0.1292632818222046,
      "learning_rate": 1.20146320959207e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 24093840,
      "step": 26955
    },
    {
      "epoch": 7.115349082750429,
      "grad_norm": 0.2928435504436493,
      "learning_rate": 1.2006243844376445e-05,
      "loss": 0.0723,
      "num_input_tokens_seen": 24098864,
      "step": 26960
    },
    {
      "epoch": 7.116668866305925,
      "grad_norm": 0.041891586035490036,
      "learning_rate": 1.1997857596632678e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 24103216,
      "step": 26965
    },
    {
      "epoch": 7.117988649861423,
      "grad_norm": 0.13848808407783508,
      "learning_rate": 1.1989473353982672e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 24107728,
      "step": 26970
    },
    {
      "epoch": 7.119308433416919,
      "grad_norm": 0.323123961687088,
      "learning_rate": 1.198109111771937e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 24112176,
      "step": 26975
    },
    {
      "epoch": 7.120628216972417,
      "grad_norm": 0.09815607964992523,
      "learning_rate": 1.197271088913543e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 24116848,
      "step": 26980
    },
    {
      "epoch": 7.121948000527913,
      "grad_norm": 0.38988539576530457,
      "learning_rate": 1.1964332669523182e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 24121360,
      "step": 26985
    },
    {
      "epoch": 7.12326778408341,
      "grad_norm": 0.24771583080291748,
      "learning_rate": 1.1955956460174645e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 24125872,
      "step": 26990
    },
    {
      "epoch": 7.124587567638907,
      "grad_norm": 0.25931790471076965,
      "learning_rate": 1.1947582262381552e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 24130288,
      "step": 26995
    },
    {
      "epoch": 7.125907351194404,
      "grad_norm": 0.24531014263629913,
      "learning_rate": 1.1939210077435293e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 24134608,
      "step": 27000
    },
    {
      "epoch": 7.125907351194404,
      "eval_loss": 0.06724131852388382,
      "eval_runtime": 64.7599,
      "eval_samples_per_second": 104.0,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 24134608,
      "step": 27000
    },
    {
      "epoch": 7.127227134749901,
      "grad_norm": 0.31773126125335693,
      "learning_rate": 1.193083990662697e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 24139184,
      "step": 27005
    },
    {
      "epoch": 7.128546918305398,
      "grad_norm": 0.38315433263778687,
      "learning_rate": 1.192247175124738e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 24143440,
      "step": 27010
    },
    {
      "epoch": 7.1298667018608946,
      "grad_norm": 0.1428283452987671,
      "learning_rate": 1.191410561258698e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 24147760,
      "step": 27015
    },
    {
      "epoch": 7.131186485416392,
      "grad_norm": 0.22901950776576996,
      "learning_rate": 1.1905741491935944e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 24152112,
      "step": 27020
    },
    {
      "epoch": 7.1325062689718886,
      "grad_norm": 0.10218995064496994,
      "learning_rate": 1.1897379390584129e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 24156176,
      "step": 27025
    },
    {
      "epoch": 7.133826052527385,
      "grad_norm": 0.19707873463630676,
      "learning_rate": 1.1889019309821062e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 24160720,
      "step": 27030
    },
    {
      "epoch": 7.1351458360828826,
      "grad_norm": 0.06293908506631851,
      "learning_rate": 1.188066125093599e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 24165104,
      "step": 27035
    },
    {
      "epoch": 7.136465619638379,
      "grad_norm": 0.2122800648212433,
      "learning_rate": 1.1872305215217811e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 24169616,
      "step": 27040
    },
    {
      "epoch": 7.1377854031938766,
      "grad_norm": 0.0775170847773552,
      "learning_rate": 1.186395120395514e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 24174096,
      "step": 27045
    },
    {
      "epoch": 7.139105186749373,
      "grad_norm": 0.29120853543281555,
      "learning_rate": 1.1855599218436283e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 24178544,
      "step": 27050
    },
    {
      "epoch": 7.14042497030487,
      "grad_norm": 0.04172292724251747,
      "learning_rate": 1.1847249259949209e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 24183344,
      "step": 27055
    },
    {
      "epoch": 7.141744753860367,
      "grad_norm": 0.224239319562912,
      "learning_rate": 1.1838901329781574e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 24187728,
      "step": 27060
    },
    {
      "epoch": 7.143064537415864,
      "grad_norm": 0.19917020201683044,
      "learning_rate": 1.1830555429220758e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 24192368,
      "step": 27065
    },
    {
      "epoch": 7.144384320971361,
      "grad_norm": 0.2386878877878189,
      "learning_rate": 1.1822211559553784e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 24196688,
      "step": 27070
    },
    {
      "epoch": 7.145704104526858,
      "grad_norm": 0.06777238100767136,
      "learning_rate": 1.18138697220674e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 24201360,
      "step": 27075
    },
    {
      "epoch": 7.147023888082354,
      "grad_norm": 0.1428188532590866,
      "learning_rate": 1.1805529918048e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 24205840,
      "step": 27080
    },
    {
      "epoch": 7.148343671637852,
      "grad_norm": 0.0450635589659214,
      "learning_rate": 1.1797192148781702e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 24210000,
      "step": 27085
    },
    {
      "epoch": 7.149663455193348,
      "grad_norm": 0.5597084760665894,
      "learning_rate": 1.1788856415554297e-05,
      "loss": 0.0969,
      "num_input_tokens_seen": 24214544,
      "step": 27090
    },
    {
      "epoch": 7.150983238748845,
      "grad_norm": 0.31675687432289124,
      "learning_rate": 1.1780522719651249e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 24218736,
      "step": 27095
    },
    {
      "epoch": 7.152303022304342,
      "grad_norm": 0.26600244641304016,
      "learning_rate": 1.1772191062357721e-05,
      "loss": 0.0801,
      "num_input_tokens_seen": 24223376,
      "step": 27100
    },
    {
      "epoch": 7.153622805859839,
      "grad_norm": 0.2309029996395111,
      "learning_rate": 1.1763861444958573e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 24227984,
      "step": 27105
    },
    {
      "epoch": 7.154942589415336,
      "grad_norm": 0.46515825390815735,
      "learning_rate": 1.1755533868738317e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 24232368,
      "step": 27110
    },
    {
      "epoch": 7.156262372970833,
      "grad_norm": 0.04308098554611206,
      "learning_rate": 1.1747208334981185e-05,
      "loss": 0.0224,
      "num_input_tokens_seen": 24236944,
      "step": 27115
    },
    {
      "epoch": 7.157582156526329,
      "grad_norm": 0.3021142780780792,
      "learning_rate": 1.1738884844971067e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 24241488,
      "step": 27120
    },
    {
      "epoch": 7.158901940081827,
      "grad_norm": 0.23209217190742493,
      "learning_rate": 1.1730563399991563e-05,
      "loss": 0.0569,
      "num_input_tokens_seen": 24245904,
      "step": 27125
    },
    {
      "epoch": 7.160221723637323,
      "grad_norm": 0.10624974220991135,
      "learning_rate": 1.1722244001325938e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 24250288,
      "step": 27130
    },
    {
      "epoch": 7.161541507192821,
      "grad_norm": 0.04418506845831871,
      "learning_rate": 1.1713926650257137e-05,
      "loss": 0.012,
      "num_input_tokens_seen": 24255056,
      "step": 27135
    },
    {
      "epoch": 7.162861290748317,
      "grad_norm": 0.28927895426750183,
      "learning_rate": 1.170561134806781e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 24259184,
      "step": 27140
    },
    {
      "epoch": 7.164181074303814,
      "grad_norm": 0.327822744846344,
      "learning_rate": 1.1697298096040287e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 24263664,
      "step": 27145
    },
    {
      "epoch": 7.165500857859311,
      "grad_norm": 0.1626204401254654,
      "learning_rate": 1.1688986895456567e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 24267664,
      "step": 27150
    },
    {
      "epoch": 7.166820641414808,
      "grad_norm": 0.1624266803264618,
      "learning_rate": 1.1680677747598349e-05,
      "loss": 0.0356,
      "num_input_tokens_seen": 24271920,
      "step": 27155
    },
    {
      "epoch": 7.1681404249703045,
      "grad_norm": 0.3008864223957062,
      "learning_rate": 1.1672370653746995e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 24276368,
      "step": 27160
    },
    {
      "epoch": 7.169460208525802,
      "grad_norm": 0.3909413814544678,
      "learning_rate": 1.166406561518357e-05,
      "loss": 0.0764,
      "num_input_tokens_seen": 24280880,
      "step": 27165
    },
    {
      "epoch": 7.1707799920812985,
      "grad_norm": 0.23474417626857758,
      "learning_rate": 1.1655762633188826e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 24285488,
      "step": 27170
    },
    {
      "epoch": 7.172099775636796,
      "grad_norm": 0.2852865755558014,
      "learning_rate": 1.1647461709043172e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 24290128,
      "step": 27175
    },
    {
      "epoch": 7.1734195591922925,
      "grad_norm": 0.07236635684967041,
      "learning_rate": 1.1639162844026722e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 24294480,
      "step": 27180
    },
    {
      "epoch": 7.174739342747789,
      "grad_norm": 0.42925286293029785,
      "learning_rate": 1.163086603941927e-05,
      "loss": 0.1157,
      "num_input_tokens_seen": 24299088,
      "step": 27185
    },
    {
      "epoch": 7.1760591263032865,
      "grad_norm": 0.14288103580474854,
      "learning_rate": 1.1622571296500273e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 24303120,
      "step": 27190
    },
    {
      "epoch": 7.177378909858783,
      "grad_norm": 0.2055729180574417,
      "learning_rate": 1.1614278616548904e-05,
      "loss": 0.0639,
      "num_input_tokens_seen": 24307760,
      "step": 27195
    },
    {
      "epoch": 7.1786986934142805,
      "grad_norm": 0.17746588587760925,
      "learning_rate": 1.1605988000843986e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 24312464,
      "step": 27200
    },
    {
      "epoch": 7.1786986934142805,
      "eval_loss": 0.0671549066901207,
      "eval_runtime": 64.7709,
      "eval_samples_per_second": 103.982,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 24312464,
      "step": 27200
    },
    {
      "epoch": 7.180018476969777,
      "grad_norm": 0.1453004628419876,
      "learning_rate": 1.1597699450664028e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 24316976,
      "step": 27205
    },
    {
      "epoch": 7.181338260525274,
      "grad_norm": 0.0941695049405098,
      "learning_rate": 1.1589412967287252e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 24321104,
      "step": 27210
    },
    {
      "epoch": 7.182658044080771,
      "grad_norm": 0.2976839542388916,
      "learning_rate": 1.1581128551991514e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 24325840,
      "step": 27215
    },
    {
      "epoch": 7.183977827636268,
      "grad_norm": 0.044943131506443024,
      "learning_rate": 1.1572846206054383e-05,
      "loss": 0.0499,
      "num_input_tokens_seen": 24330512,
      "step": 27220
    },
    {
      "epoch": 7.185297611191764,
      "grad_norm": 0.0658232718706131,
      "learning_rate": 1.1564565930753113e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 24335088,
      "step": 27225
    },
    {
      "epoch": 7.186617394747262,
      "grad_norm": 0.06307104974985123,
      "learning_rate": 1.1556287727364606e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 24339728,
      "step": 27230
    },
    {
      "epoch": 7.187937178302758,
      "grad_norm": 0.1420002430677414,
      "learning_rate": 1.1548011597165489e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 24344368,
      "step": 27235
    },
    {
      "epoch": 7.189256961858256,
      "grad_norm": 0.18477608263492584,
      "learning_rate": 1.1539737541432019e-05,
      "loss": 0.0929,
      "num_input_tokens_seen": 24348720,
      "step": 27240
    },
    {
      "epoch": 7.190576745413752,
      "grad_norm": 0.11105629801750183,
      "learning_rate": 1.1531465561440174e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 24353040,
      "step": 27245
    },
    {
      "epoch": 7.191896528969249,
      "grad_norm": 0.20076912641525269,
      "learning_rate": 1.1523195658465605e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 24357680,
      "step": 27250
    },
    {
      "epoch": 7.193216312524746,
      "grad_norm": 0.16600032150745392,
      "learning_rate": 1.1514927833783618e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 24362256,
      "step": 27255
    },
    {
      "epoch": 7.194536096080243,
      "grad_norm": 0.3126922845840454,
      "learning_rate": 1.150666208866922e-05,
      "loss": 0.0288,
      "num_input_tokens_seen": 24366992,
      "step": 27260
    },
    {
      "epoch": 7.19585587963574,
      "grad_norm": 0.13824398815631866,
      "learning_rate": 1.1498398424397106e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 24371280,
      "step": 27265
    },
    {
      "epoch": 7.197175663191237,
      "grad_norm": 0.09463849663734436,
      "learning_rate": 1.1490136842241628e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 24375504,
      "step": 27270
    },
    {
      "epoch": 7.198495446746733,
      "grad_norm": 0.411231130361557,
      "learning_rate": 1.1481877343476813e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 24379824,
      "step": 27275
    },
    {
      "epoch": 7.199815230302231,
      "grad_norm": 0.23730726540088654,
      "learning_rate": 1.14736199293764e-05,
      "loss": 0.0697,
      "num_input_tokens_seen": 24384336,
      "step": 27280
    },
    {
      "epoch": 7.201135013857727,
      "grad_norm": 0.1440921127796173,
      "learning_rate": 1.1465364601213771e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 24388880,
      "step": 27285
    },
    {
      "epoch": 7.202454797413224,
      "grad_norm": 0.17842113971710205,
      "learning_rate": 1.1457111360262012e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 24393072,
      "step": 27290
    },
    {
      "epoch": 7.203774580968721,
      "grad_norm": 0.15534725785255432,
      "learning_rate": 1.1448860207793869e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 24397360,
      "step": 27295
    },
    {
      "epoch": 7.205094364524218,
      "grad_norm": 0.2299571931362152,
      "learning_rate": 1.144061114508177e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 24402160,
      "step": 27300
    },
    {
      "epoch": 7.206414148079715,
      "grad_norm": 0.0235295407474041,
      "learning_rate": 1.1432364173397842e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 24406576,
      "step": 27305
    },
    {
      "epoch": 7.207733931635212,
      "grad_norm": 0.07702226936817169,
      "learning_rate": 1.1424119294013852e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 24411504,
      "step": 27310
    },
    {
      "epoch": 7.209053715190708,
      "grad_norm": 0.11302943527698517,
      "learning_rate": 1.1415876508201279e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 24416048,
      "step": 27315
    },
    {
      "epoch": 7.210373498746206,
      "grad_norm": 0.0733208954334259,
      "learning_rate": 1.140763581723125e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 24420112,
      "step": 27320
    },
    {
      "epoch": 7.211693282301702,
      "grad_norm": 0.062075842171907425,
      "learning_rate": 1.1399397222374588e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 24424432,
      "step": 27325
    },
    {
      "epoch": 7.2130130658572,
      "grad_norm": 0.07009834051132202,
      "learning_rate": 1.1391160724901804e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 24428784,
      "step": 27330
    },
    {
      "epoch": 7.214332849412696,
      "grad_norm": 0.08232706040143967,
      "learning_rate": 1.138292632608304e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 24433136,
      "step": 27335
    },
    {
      "epoch": 7.215652632968193,
      "grad_norm": 0.39854931831359863,
      "learning_rate": 1.1374694027188174e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 24437456,
      "step": 27340
    },
    {
      "epoch": 7.21697241652369,
      "grad_norm": 0.18461249768733978,
      "learning_rate": 1.1366463829486711e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 24441936,
      "step": 27345
    },
    {
      "epoch": 7.218292200079187,
      "grad_norm": 0.14270691573619843,
      "learning_rate": 1.1358235734247849e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 24446640,
      "step": 27350
    },
    {
      "epoch": 7.2196119836346835,
      "grad_norm": 0.3645656704902649,
      "learning_rate": 1.1350009742740478e-05,
      "loss": 0.0782,
      "num_input_tokens_seen": 24450992,
      "step": 27355
    },
    {
      "epoch": 7.220931767190181,
      "grad_norm": 0.255948007106781,
      "learning_rate": 1.134178585623313e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 24455376,
      "step": 27360
    },
    {
      "epoch": 7.2222515507456775,
      "grad_norm": 0.16897998750209808,
      "learning_rate": 1.1333564075994047e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 24459952,
      "step": 27365
    },
    {
      "epoch": 7.223571334301175,
      "grad_norm": 0.18294745683670044,
      "learning_rate": 1.1325344403291133e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 24464336,
      "step": 27370
    },
    {
      "epoch": 7.2248911178566715,
      "grad_norm": 0.403415709733963,
      "learning_rate": 1.1317126839391951e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 24468944,
      "step": 27375
    },
    {
      "epoch": 7.226210901412168,
      "grad_norm": 0.473206102848053,
      "learning_rate": 1.1308911385563766e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 24473744,
      "step": 27380
    },
    {
      "epoch": 7.2275306849676655,
      "grad_norm": 0.048293884843587875,
      "learning_rate": 1.1300698043073494e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 24478256,
      "step": 27385
    },
    {
      "epoch": 7.228850468523162,
      "grad_norm": 0.04211829975247383,
      "learning_rate": 1.1292486813187736e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 24482832,
      "step": 27390
    },
    {
      "epoch": 7.2301702520786595,
      "grad_norm": 0.2053355872631073,
      "learning_rate": 1.1284277697172782e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 24487248,
      "step": 27395
    },
    {
      "epoch": 7.231490035634156,
      "grad_norm": 0.262908399105072,
      "learning_rate": 1.127607069629456e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 24491696,
      "step": 27400
    },
    {
      "epoch": 7.231490035634156,
      "eval_loss": 0.06733673065900803,
      "eval_runtime": 64.7926,
      "eval_samples_per_second": 103.947,
      "eval_steps_per_second": 25.991,
      "num_input_tokens_seen": 24491696,
      "step": 27400
    },
    {
      "epoch": 7.232809819189653,
      "grad_norm": 0.37509220838546753,
      "learning_rate": 1.1267865811818701e-05,
      "loss": 0.0912,
      "num_input_tokens_seen": 24495856,
      "step": 27405
    },
    {
      "epoch": 7.23412960274515,
      "grad_norm": 0.385707288980484,
      "learning_rate": 1.1259663045010513e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 24500496,
      "step": 27410
    },
    {
      "epoch": 7.235449386300647,
      "grad_norm": 0.045231495052576065,
      "learning_rate": 1.1251462397134957e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 24504816,
      "step": 27415
    },
    {
      "epoch": 7.236769169856144,
      "grad_norm": 0.06550495326519012,
      "learning_rate": 1.1243263869456664e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 24509424,
      "step": 27420
    },
    {
      "epoch": 7.238088953411641,
      "grad_norm": 0.07850643992424011,
      "learning_rate": 1.1235067463239967e-05,
      "loss": 0.0795,
      "num_input_tokens_seen": 24513712,
      "step": 27425
    },
    {
      "epoch": 7.239408736967137,
      "grad_norm": 0.338356614112854,
      "learning_rate": 1.122687317974884e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 24517936,
      "step": 27430
    },
    {
      "epoch": 7.240728520522635,
      "grad_norm": 0.5090805292129517,
      "learning_rate": 1.1218681020246963e-05,
      "loss": 0.0909,
      "num_input_tokens_seen": 24522448,
      "step": 27435
    },
    {
      "epoch": 7.242048304078131,
      "grad_norm": 0.10686513036489487,
      "learning_rate": 1.1210490985997652e-05,
      "loss": 0.0977,
      "num_input_tokens_seen": 24526736,
      "step": 27440
    },
    {
      "epoch": 7.243368087633628,
      "grad_norm": 0.038170576095581055,
      "learning_rate": 1.1202303078263917e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 24531088,
      "step": 27445
    },
    {
      "epoch": 7.244687871189125,
      "grad_norm": 0.04953524470329285,
      "learning_rate": 1.1194117298308451e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 24535760,
      "step": 27450
    },
    {
      "epoch": 7.246007654744622,
      "grad_norm": 0.3726954758167267,
      "learning_rate": 1.1185933647393585e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 24540048,
      "step": 27455
    },
    {
      "epoch": 7.247327438300119,
      "grad_norm": 0.31466010212898254,
      "learning_rate": 1.1177752126781354e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 24544464,
      "step": 27460
    },
    {
      "epoch": 7.248647221855616,
      "grad_norm": 0.11652236431837082,
      "learning_rate": 1.1169572737733441e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 24548720,
      "step": 27465
    },
    {
      "epoch": 7.249967005411112,
      "grad_norm": 0.038556769490242004,
      "learning_rate": 1.1161395481511216e-05,
      "loss": 0.1219,
      "num_input_tokens_seen": 24553136,
      "step": 27470
    },
    {
      "epoch": 7.25128678896661,
      "grad_norm": 0.3688308000564575,
      "learning_rate": 1.1153220359375722e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 24557904,
      "step": 27475
    },
    {
      "epoch": 7.252606572522106,
      "grad_norm": 0.1164795458316803,
      "learning_rate": 1.114504737258765e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 24562448,
      "step": 27480
    },
    {
      "epoch": 7.253926356077603,
      "grad_norm": 0.4749835133552551,
      "learning_rate": 1.1136876522407393e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 24566928,
      "step": 27485
    },
    {
      "epoch": 7.2552461396331,
      "grad_norm": 0.18950961530208588,
      "learning_rate": 1.1128707810094985e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 24571216,
      "step": 27490
    },
    {
      "epoch": 7.256565923188597,
      "grad_norm": 0.29345831274986267,
      "learning_rate": 1.1120541236910157e-05,
      "loss": 0.0521,
      "num_input_tokens_seen": 24575728,
      "step": 27495
    },
    {
      "epoch": 7.257885706744094,
      "grad_norm": 0.2807425856590271,
      "learning_rate": 1.111237680411229e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 24580240,
      "step": 27500
    },
    {
      "epoch": 7.259205490299591,
      "grad_norm": 0.558468222618103,
      "learning_rate": 1.1104214512960433e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 24584656,
      "step": 27505
    },
    {
      "epoch": 7.260525273855087,
      "grad_norm": 0.07164488732814789,
      "learning_rate": 1.1096054364713327e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 24589520,
      "step": 27510
    },
    {
      "epoch": 7.261845057410585,
      "grad_norm": 0.4163208603858948,
      "learning_rate": 1.1087896360629371e-05,
      "loss": 0.0591,
      "num_input_tokens_seen": 24593936,
      "step": 27515
    },
    {
      "epoch": 7.263164840966081,
      "grad_norm": 0.06254494935274124,
      "learning_rate": 1.107974050196662e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 24598416,
      "step": 27520
    },
    {
      "epoch": 7.264484624521579,
      "grad_norm": 0.173482283949852,
      "learning_rate": 1.1071586789982816e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 24602896,
      "step": 27525
    },
    {
      "epoch": 7.265804408077075,
      "grad_norm": 0.3245493471622467,
      "learning_rate": 1.1063435225935373e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 24607504,
      "step": 27530
    },
    {
      "epoch": 7.267124191632572,
      "grad_norm": 0.3367374837398529,
      "learning_rate": 1.1055285811081348e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 24611856,
      "step": 27535
    },
    {
      "epoch": 7.268443975188069,
      "grad_norm": 0.11486507207155228,
      "learning_rate": 1.1047138546677499e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 24616816,
      "step": 27540
    },
    {
      "epoch": 7.269763758743566,
      "grad_norm": 0.29035088419914246,
      "learning_rate": 1.1038993433980219e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 24621456,
      "step": 27545
    },
    {
      "epoch": 7.271083542299063,
      "grad_norm": 0.08633492887020111,
      "learning_rate": 1.1030850474245597e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 24626096,
      "step": 27550
    },
    {
      "epoch": 7.27240332585456,
      "grad_norm": 0.3409574627876282,
      "learning_rate": 1.102270966872939e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 24630384,
      "step": 27555
    },
    {
      "epoch": 7.2737231094100565,
      "grad_norm": 0.2998826801776886,
      "learning_rate": 1.1014571018687e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 24634608,
      "step": 27560
    },
    {
      "epoch": 7.275042892965554,
      "grad_norm": 0.06862879544496536,
      "learning_rate": 1.1006434525373502e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 24639024,
      "step": 27565
    },
    {
      "epoch": 7.2763626765210505,
      "grad_norm": 0.11353106796741486,
      "learning_rate": 1.0998300190043664e-05,
      "loss": 0.111,
      "num_input_tokens_seen": 24643632,
      "step": 27570
    },
    {
      "epoch": 7.277682460076547,
      "grad_norm": 0.3312112092971802,
      "learning_rate": 1.0990168013951882e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 24648080,
      "step": 27575
    },
    {
      "epoch": 7.2790022436320445,
      "grad_norm": 0.5687589645385742,
      "learning_rate": 1.0982037998352263e-05,
      "loss": 0.0983,
      "num_input_tokens_seen": 24652624,
      "step": 27580
    },
    {
      "epoch": 7.280322027187541,
      "grad_norm": 0.34436213970184326,
      "learning_rate": 1.0973910144498534e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 24657072,
      "step": 27585
    },
    {
      "epoch": 7.2816418107430385,
      "grad_norm": 0.1591799408197403,
      "learning_rate": 1.0965784453644123e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 24661424,
      "step": 27590
    },
    {
      "epoch": 7.282961594298535,
      "grad_norm": 0.20682141184806824,
      "learning_rate": 1.0957660927042127e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 24665808,
      "step": 27595
    },
    {
      "epoch": 7.284281377854032,
      "grad_norm": 0.1414584368467331,
      "learning_rate": 1.094953956594527e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 24670160,
      "step": 27600
    },
    {
      "epoch": 7.284281377854032,
      "eval_loss": 0.0670321062207222,
      "eval_runtime": 64.7425,
      "eval_samples_per_second": 104.027,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 24670160,
      "step": 27600
    },
    {
      "epoch": 7.285601161409529,
      "grad_norm": 0.10209903120994568,
      "learning_rate": 1.0941420371605981e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 24674672,
      "step": 27605
    },
    {
      "epoch": 7.286920944965026,
      "grad_norm": 0.2846114933490753,
      "learning_rate": 1.0933303345276354e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 24678992,
      "step": 27610
    },
    {
      "epoch": 7.288240728520522,
      "grad_norm": 0.04901299625635147,
      "learning_rate": 1.0925188488208112e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 24683472,
      "step": 27615
    },
    {
      "epoch": 7.28956051207602,
      "grad_norm": 0.3092661499977112,
      "learning_rate": 1.0917075801652694e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 24687536,
      "step": 27620
    },
    {
      "epoch": 7.290880295631516,
      "grad_norm": 0.09395746141672134,
      "learning_rate": 1.0908965286861151e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 24691696,
      "step": 27625
    },
    {
      "epoch": 7.292200079187014,
      "grad_norm": 0.47397541999816895,
      "learning_rate": 1.090085694508425e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 24696080,
      "step": 27630
    },
    {
      "epoch": 7.29351986274251,
      "grad_norm": 0.4713501036167145,
      "learning_rate": 1.089275077757238e-05,
      "loss": 0.1008,
      "num_input_tokens_seen": 24700560,
      "step": 27635
    },
    {
      "epoch": 7.294839646298007,
      "grad_norm": 0.06048964336514473,
      "learning_rate": 1.0884646785575633e-05,
      "loss": 0.032,
      "num_input_tokens_seen": 24705104,
      "step": 27640
    },
    {
      "epoch": 7.296159429853504,
      "grad_norm": 0.04711132124066353,
      "learning_rate": 1.0876544970343728e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 24709584,
      "step": 27645
    },
    {
      "epoch": 7.297479213409001,
      "grad_norm": 0.2942809462547302,
      "learning_rate": 1.0868445333126082e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 24714064,
      "step": 27650
    },
    {
      "epoch": 7.298798996964498,
      "grad_norm": 0.053972892463207245,
      "learning_rate": 1.0860347875171745e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 24718672,
      "step": 27655
    },
    {
      "epoch": 7.300118780519995,
      "grad_norm": 0.08592488616704941,
      "learning_rate": 1.0852252597729465e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 24722800,
      "step": 27660
    },
    {
      "epoch": 7.301438564075491,
      "grad_norm": 0.22755536437034607,
      "learning_rate": 1.0844159502047615e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 24727120,
      "step": 27665
    },
    {
      "epoch": 7.302758347630989,
      "grad_norm": 0.3051374852657318,
      "learning_rate": 1.0836068589374265e-05,
      "loss": 0.0556,
      "num_input_tokens_seen": 24731696,
      "step": 27670
    },
    {
      "epoch": 7.304078131186485,
      "grad_norm": 0.6084445714950562,
      "learning_rate": 1.0827979860957144e-05,
      "loss": 0.0622,
      "num_input_tokens_seen": 24736048,
      "step": 27675
    },
    {
      "epoch": 7.305397914741983,
      "grad_norm": 0.08674095571041107,
      "learning_rate": 1.0819893318043615e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 24740784,
      "step": 27680
    },
    {
      "epoch": 7.306717698297479,
      "grad_norm": 0.2730005085468292,
      "learning_rate": 1.0811808961880734e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 24745520,
      "step": 27685
    },
    {
      "epoch": 7.308037481852976,
      "grad_norm": 0.14386364817619324,
      "learning_rate": 1.080372679371522e-05,
      "loss": 0.0744,
      "num_input_tokens_seen": 24750352,
      "step": 27690
    },
    {
      "epoch": 7.309357265408473,
      "grad_norm": 0.1752229630947113,
      "learning_rate": 1.0795646814793428e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 24754800,
      "step": 27695
    },
    {
      "epoch": 7.31067704896397,
      "grad_norm": 0.40007075667381287,
      "learning_rate": 1.078756902636141e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 24759216,
      "step": 27700
    },
    {
      "epoch": 7.311996832519466,
      "grad_norm": 0.24604114890098572,
      "learning_rate": 1.077949342966485e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 24763504,
      "step": 27705
    },
    {
      "epoch": 7.313316616074964,
      "grad_norm": 0.1704041212797165,
      "learning_rate": 1.0771420025949103e-05,
      "loss": 0.0238,
      "num_input_tokens_seen": 24768112,
      "step": 27710
    },
    {
      "epoch": 7.31463639963046,
      "grad_norm": 0.15443898737430573,
      "learning_rate": 1.0763348816459204e-05,
      "loss": 0.0599,
      "num_input_tokens_seen": 24772464,
      "step": 27715
    },
    {
      "epoch": 7.315956183185958,
      "grad_norm": 0.1411323994398117,
      "learning_rate": 1.0755279802439816e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 24776816,
      "step": 27720
    },
    {
      "epoch": 7.317275966741454,
      "grad_norm": 0.1422702968120575,
      "learning_rate": 1.0747212985135293e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 24781328,
      "step": 27725
    },
    {
      "epoch": 7.318595750296951,
      "grad_norm": 0.27496621012687683,
      "learning_rate": 1.073914836578965e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 24785744,
      "step": 27730
    },
    {
      "epoch": 7.319915533852448,
      "grad_norm": 0.2235908955335617,
      "learning_rate": 1.0731085945646529e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 24790288,
      "step": 27735
    },
    {
      "epoch": 7.321235317407945,
      "grad_norm": 0.17384664714336395,
      "learning_rate": 1.0723025725949285e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 24794768,
      "step": 27740
    },
    {
      "epoch": 7.3225551009634415,
      "grad_norm": 0.09952058643102646,
      "learning_rate": 1.0714967707940875e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 24799504,
      "step": 27745
    },
    {
      "epoch": 7.323874884518939,
      "grad_norm": 0.08403459191322327,
      "learning_rate": 1.0706911892863963e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 24804272,
      "step": 27750
    },
    {
      "epoch": 7.3251946680744355,
      "grad_norm": 0.09434563666582108,
      "learning_rate": 1.0698858281960866e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 24808720,
      "step": 27755
    },
    {
      "epoch": 7.326514451629933,
      "grad_norm": 0.372171550989151,
      "learning_rate": 1.069080687647353e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 24813040,
      "step": 27760
    },
    {
      "epoch": 7.3278342351854295,
      "grad_norm": 0.09060271084308624,
      "learning_rate": 1.0682757677643596e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 24817680,
      "step": 27765
    },
    {
      "epoch": 7.329154018740926,
      "grad_norm": 0.16516268253326416,
      "learning_rate": 1.0674710686712359e-05,
      "loss": 0.0485,
      "num_input_tokens_seen": 24822224,
      "step": 27770
    },
    {
      "epoch": 7.3304738022964235,
      "grad_norm": 0.12392836809158325,
      "learning_rate": 1.0666665904920756e-05,
      "loss": 0.0401,
      "num_input_tokens_seen": 24826704,
      "step": 27775
    },
    {
      "epoch": 7.33179358585192,
      "grad_norm": 0.02740810438990593,
      "learning_rate": 1.0658623333509385e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 24831024,
      "step": 27780
    },
    {
      "epoch": 7.3331133694074175,
      "grad_norm": 0.13148590922355652,
      "learning_rate": 1.0650582973718532e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 24835536,
      "step": 27785
    },
    {
      "epoch": 7.334433152962914,
      "grad_norm": 0.21084579825401306,
      "learning_rate": 1.0642544826788098e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 24839760,
      "step": 27790
    },
    {
      "epoch": 7.335752936518411,
      "grad_norm": 0.13928169012069702,
      "learning_rate": 1.063450889395769e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 24844368,
      "step": 27795
    },
    {
      "epoch": 7.337072720073908,
      "grad_norm": 0.21707572042942047,
      "learning_rate": 1.062647517646653e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 24848976,
      "step": 27800
    },
    {
      "epoch": 7.337072720073908,
      "eval_loss": 0.06726846098899841,
      "eval_runtime": 64.7517,
      "eval_samples_per_second": 104.013,
      "eval_steps_per_second": 26.007,
      "num_input_tokens_seen": 24848976,
      "step": 27800
    },
    {
      "epoch": 7.338392503629405,
      "grad_norm": 0.055873699486255646,
      "learning_rate": 1.0618443675553527e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 24853168,
      "step": 27805
    },
    {
      "epoch": 7.339712287184902,
      "grad_norm": 0.06328243762254715,
      "learning_rate": 1.0610414392457247e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 24857456,
      "step": 27810
    },
    {
      "epoch": 7.341032070740399,
      "grad_norm": 0.24468226730823517,
      "learning_rate": 1.0602387328415888e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 24861712,
      "step": 27815
    },
    {
      "epoch": 7.342351854295895,
      "grad_norm": 0.7732455134391785,
      "learning_rate": 1.0594362484667347e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 24866320,
      "step": 27820
    },
    {
      "epoch": 7.343671637851393,
      "grad_norm": 0.27959728240966797,
      "learning_rate": 1.0586339862449132e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 24870864,
      "step": 27825
    },
    {
      "epoch": 7.344991421406889,
      "grad_norm": 0.2536408603191376,
      "learning_rate": 1.0578319462998445e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 24875312,
      "step": 27830
    },
    {
      "epoch": 7.346311204962386,
      "grad_norm": 0.38055500388145447,
      "learning_rate": 1.057030128755214e-05,
      "loss": 0.0959,
      "num_input_tokens_seen": 24879312,
      "step": 27835
    },
    {
      "epoch": 7.347630988517883,
      "grad_norm": 0.07191559672355652,
      "learning_rate": 1.0562285337346703e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 24884048,
      "step": 27840
    },
    {
      "epoch": 7.34895077207338,
      "grad_norm": 0.15985798835754395,
      "learning_rate": 1.0554271613618308e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 24888720,
      "step": 27845
    },
    {
      "epoch": 7.350270555628877,
      "grad_norm": 0.05079872906208038,
      "learning_rate": 1.054626011760276e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 24893104,
      "step": 27850
    },
    {
      "epoch": 7.351590339184374,
      "grad_norm": 0.16903164982795715,
      "learning_rate": 1.0538250850535549e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 24897552,
      "step": 27855
    },
    {
      "epoch": 7.35291012273987,
      "grad_norm": 0.267366886138916,
      "learning_rate": 1.0530243813651794e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 24901776,
      "step": 27860
    },
    {
      "epoch": 7.354229906295368,
      "grad_norm": 0.0450739823281765,
      "learning_rate": 1.0522239008186271e-05,
      "loss": 0.0628,
      "num_input_tokens_seen": 24906320,
      "step": 27865
    },
    {
      "epoch": 7.355549689850864,
      "grad_norm": 0.09771493822336197,
      "learning_rate": 1.0514236435373434e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 24910736,
      "step": 27870
    },
    {
      "epoch": 7.356869473406362,
      "grad_norm": 0.04377949610352516,
      "learning_rate": 1.0506236096447386e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 24915344,
      "step": 27875
    },
    {
      "epoch": 7.358189256961858,
      "grad_norm": 0.4896205961704254,
      "learning_rate": 1.049823799264186e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 24919920,
      "step": 27880
    },
    {
      "epoch": 7.359509040517355,
      "grad_norm": 0.26843783259391785,
      "learning_rate": 1.049024212519028e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 24924432,
      "step": 27885
    },
    {
      "epoch": 7.360828824072852,
      "grad_norm": 0.10107819736003876,
      "learning_rate": 1.0482248495325713e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 24928880,
      "step": 27890
    },
    {
      "epoch": 7.362148607628349,
      "grad_norm": 0.047026123851537704,
      "learning_rate": 1.047425710428086e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 24933616,
      "step": 27895
    },
    {
      "epoch": 7.363468391183845,
      "grad_norm": 0.5455624461174011,
      "learning_rate": 1.0466267953288114e-05,
      "loss": 0.0882,
      "num_input_tokens_seen": 24938000,
      "step": 27900
    },
    {
      "epoch": 7.364788174739343,
      "grad_norm": 0.1768079400062561,
      "learning_rate": 1.0458281043579482e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 24942352,
      "step": 27905
    },
    {
      "epoch": 7.366107958294839,
      "grad_norm": 0.2302229404449463,
      "learning_rate": 1.0450296376386657e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 24947216,
      "step": 27910
    },
    {
      "epoch": 7.367427741850337,
      "grad_norm": 0.09673712402582169,
      "learning_rate": 1.044231395294098e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 24951888,
      "step": 27915
    },
    {
      "epoch": 7.368747525405833,
      "grad_norm": 0.021349217742681503,
      "learning_rate": 1.0434333774473435e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 24956464,
      "step": 27920
    },
    {
      "epoch": 7.37006730896133,
      "grad_norm": 0.06207912415266037,
      "learning_rate": 1.0426355842214657e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 24960752,
      "step": 27925
    },
    {
      "epoch": 7.371387092516827,
      "grad_norm": 0.36458393931388855,
      "learning_rate": 1.0418380157394963e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 24965072,
      "step": 27930
    },
    {
      "epoch": 7.372706876072324,
      "grad_norm": 0.143574520945549,
      "learning_rate": 1.0410406721244281e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 24969520,
      "step": 27935
    },
    {
      "epoch": 7.3740266596278214,
      "grad_norm": 0.20523373782634735,
      "learning_rate": 1.0402435534992238e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 24973744,
      "step": 27940
    },
    {
      "epoch": 7.375346443183318,
      "grad_norm": 0.25944244861602783,
      "learning_rate": 1.0394466599868071e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 24978256,
      "step": 27945
    },
    {
      "epoch": 7.376666226738815,
      "grad_norm": 0.048216287046670914,
      "learning_rate": 1.0386499917100697e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 24983024,
      "step": 27950
    },
    {
      "epoch": 7.377986010294312,
      "grad_norm": 0.2627435326576233,
      "learning_rate": 1.0378535487918692e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 24987440,
      "step": 27955
    },
    {
      "epoch": 7.379305793849809,
      "grad_norm": 0.2561013996601105,
      "learning_rate": 1.037057331355025e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 24991920,
      "step": 27960
    },
    {
      "epoch": 7.380625577405305,
      "grad_norm": 0.41921135783195496,
      "learning_rate": 1.0362613395223247e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 24996688,
      "step": 27965
    },
    {
      "epoch": 7.381945360960803,
      "grad_norm": 0.16882388293743134,
      "learning_rate": 1.0354655734165212e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 25000976,
      "step": 27970
    },
    {
      "epoch": 7.383265144516299,
      "grad_norm": 0.32885047793388367,
      "learning_rate": 1.03467003316033e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 25005264,
      "step": 27975
    },
    {
      "epoch": 7.384584928071797,
      "grad_norm": 0.2750985324382782,
      "learning_rate": 1.033874718876435e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 25009776,
      "step": 27980
    },
    {
      "epoch": 7.385904711627293,
      "grad_norm": 0.4717799127101898,
      "learning_rate": 1.0330796306874818e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 25014288,
      "step": 27985
    },
    {
      "epoch": 7.38722449518279,
      "grad_norm": 0.2063411921262741,
      "learning_rate": 1.032284768716085e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 25018736,
      "step": 27990
    },
    {
      "epoch": 7.388544278738287,
      "grad_norm": 0.22211600840091705,
      "learning_rate": 1.0314901330848206e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 25023280,
      "step": 27995
    },
    {
      "epoch": 7.389864062293784,
      "grad_norm": 0.0464959554374218,
      "learning_rate": 1.030695723916233e-05,
      "loss": 0.0169,
      "num_input_tokens_seen": 25027536,
      "step": 28000
    },
    {
      "epoch": 7.389864062293784,
      "eval_loss": 0.06733068078756332,
      "eval_runtime": 64.819,
      "eval_samples_per_second": 103.905,
      "eval_steps_per_second": 25.98,
      "num_input_tokens_seen": 25027536,
      "step": 28000
    },
    {
      "epoch": 7.391183845849281,
      "grad_norm": 0.37597766518592834,
      "learning_rate": 1.0299015413328289e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 25031856,
      "step": 28005
    },
    {
      "epoch": 7.392503629404778,
      "grad_norm": 0.25106000900268555,
      "learning_rate": 1.0291075854570809e-05,
      "loss": 0.0536,
      "num_input_tokens_seen": 25036464,
      "step": 28010
    },
    {
      "epoch": 7.393823412960274,
      "grad_norm": 0.289782851934433,
      "learning_rate": 1.0283138564114275e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 25040752,
      "step": 28015
    },
    {
      "epoch": 7.395143196515772,
      "grad_norm": 0.07802927494049072,
      "learning_rate": 1.027520354318273e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 25045392,
      "step": 28020
    },
    {
      "epoch": 7.396462980071268,
      "grad_norm": 0.24733249843120575,
      "learning_rate": 1.0267270792999828e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 25049744,
      "step": 28025
    },
    {
      "epoch": 7.397782763626765,
      "grad_norm": 0.6040419936180115,
      "learning_rate": 1.0259340314788919e-05,
      "loss": 0.0825,
      "num_input_tokens_seen": 25053968,
      "step": 28030
    },
    {
      "epoch": 7.399102547182262,
      "grad_norm": 0.1716860681772232,
      "learning_rate": 1.0251412109772979e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 25058384,
      "step": 28035
    },
    {
      "epoch": 7.400422330737759,
      "grad_norm": 0.2641092538833618,
      "learning_rate": 1.0243486179174627e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 25063152,
      "step": 28040
    },
    {
      "epoch": 7.401742114293256,
      "grad_norm": 0.07201898843050003,
      "learning_rate": 1.0235562524216158e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 25067632,
      "step": 28045
    },
    {
      "epoch": 7.403061897848753,
      "grad_norm": 0.293312668800354,
      "learning_rate": 1.022764114611948e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 25071824,
      "step": 28050
    },
    {
      "epoch": 7.404381681404249,
      "grad_norm": 0.3897400200366974,
      "learning_rate": 1.0219722046106178e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 25076336,
      "step": 28055
    },
    {
      "epoch": 7.405701464959747,
      "grad_norm": 0.16458198428153992,
      "learning_rate": 1.0211805225397486e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 25080816,
      "step": 28060
    },
    {
      "epoch": 7.407021248515243,
      "grad_norm": 0.7404221892356873,
      "learning_rate": 1.020389068521426e-05,
      "loss": 0.0993,
      "num_input_tokens_seen": 25085200,
      "step": 28065
    },
    {
      "epoch": 7.408341032070741,
      "grad_norm": 0.21583284437656403,
      "learning_rate": 1.0195978426777039e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 25089232,
      "step": 28070
    },
    {
      "epoch": 7.409660815626237,
      "grad_norm": 0.2122388333082199,
      "learning_rate": 1.0188068451305982e-05,
      "loss": 0.0725,
      "num_input_tokens_seen": 25093552,
      "step": 28075
    },
    {
      "epoch": 7.410980599181734,
      "grad_norm": 0.07161913067102432,
      "learning_rate": 1.0180160760020902e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 25097968,
      "step": 28080
    },
    {
      "epoch": 7.412300382737231,
      "grad_norm": 0.3532697558403015,
      "learning_rate": 1.0172255354141278e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 25101808,
      "step": 28085
    },
    {
      "epoch": 7.413620166292728,
      "grad_norm": 0.17360340058803558,
      "learning_rate": 1.0164352234886205e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 25106320,
      "step": 28090
    },
    {
      "epoch": 7.414939949848225,
      "grad_norm": 0.13384853303432465,
      "learning_rate": 1.0156451403474454e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 25110800,
      "step": 28095
    },
    {
      "epoch": 7.416259733403722,
      "grad_norm": 0.36028075218200684,
      "learning_rate": 1.0148552861124443e-05,
      "loss": 0.0592,
      "num_input_tokens_seen": 25115376,
      "step": 28100
    },
    {
      "epoch": 7.4175795169592185,
      "grad_norm": 0.24620217084884644,
      "learning_rate": 1.0140656609054205e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 25119952,
      "step": 28105
    },
    {
      "epoch": 7.418899300514716,
      "grad_norm": 0.34116989374160767,
      "learning_rate": 1.0132762648481455e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 25124496,
      "step": 28110
    },
    {
      "epoch": 7.4202190840702125,
      "grad_norm": 0.05062805116176605,
      "learning_rate": 1.0124870980623543e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 25128816,
      "step": 28115
    },
    {
      "epoch": 7.421538867625709,
      "grad_norm": 0.05739118531346321,
      "learning_rate": 1.0116981606697453e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 25133328,
      "step": 28120
    },
    {
      "epoch": 7.4228586511812065,
      "grad_norm": 0.9214297533035278,
      "learning_rate": 1.0109094527919838e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 25137680,
      "step": 28125
    },
    {
      "epoch": 7.424178434736703,
      "grad_norm": 0.2304917573928833,
      "learning_rate": 1.010120974550697e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 25141872,
      "step": 28130
    },
    {
      "epoch": 7.4254982182922005,
      "grad_norm": 0.1777554601430893,
      "learning_rate": 1.0093327260674795e-05,
      "loss": 0.0195,
      "num_input_tokens_seen": 25146192,
      "step": 28135
    },
    {
      "epoch": 7.426818001847697,
      "grad_norm": 0.28297916054725647,
      "learning_rate": 1.0085447074638878e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 25150992,
      "step": 28140
    },
    {
      "epoch": 7.428137785403194,
      "grad_norm": 0.2068546712398529,
      "learning_rate": 1.0077569188614461e-05,
      "loss": 0.0139,
      "num_input_tokens_seen": 25155696,
      "step": 28145
    },
    {
      "epoch": 7.429457568958691,
      "grad_norm": 0.13197863101959229,
      "learning_rate": 1.0069693603816393e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 25160432,
      "step": 28150
    },
    {
      "epoch": 7.430777352514188,
      "grad_norm": 0.5150306224822998,
      "learning_rate": 1.0061820321459204e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 25165040,
      "step": 28155
    },
    {
      "epoch": 7.432097136069684,
      "grad_norm": 0.15374921262264252,
      "learning_rate": 1.0053949342757038e-05,
      "loss": 0.0648,
      "num_input_tokens_seen": 25169584,
      "step": 28160
    },
    {
      "epoch": 7.433416919625182,
      "grad_norm": 0.15092085301876068,
      "learning_rate": 1.0046080668923717e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 25174064,
      "step": 28165
    },
    {
      "epoch": 7.434736703180678,
      "grad_norm": 0.05484118312597275,
      "learning_rate": 1.003821430117267e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 25178672,
      "step": 28170
    },
    {
      "epoch": 7.436056486736176,
      "grad_norm": 0.4082276225090027,
      "learning_rate": 1.0030350240716999e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 25182992,
      "step": 28175
    },
    {
      "epoch": 7.437376270291672,
      "grad_norm": 0.2505885660648346,
      "learning_rate": 1.0022488488769449e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 25187664,
      "step": 28180
    },
    {
      "epoch": 7.438696053847169,
      "grad_norm": 0.23509156703948975,
      "learning_rate": 1.0014629046542387e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 25192304,
      "step": 28185
    },
    {
      "epoch": 7.440015837402666,
      "grad_norm": 0.04883621633052826,
      "learning_rate": 1.0006771915247842e-05,
      "loss": 0.1295,
      "num_input_tokens_seen": 25196880,
      "step": 28190
    },
    {
      "epoch": 7.441335620958163,
      "grad_norm": 0.19336022436618805,
      "learning_rate": 9.998917096097495e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 25201296,
      "step": 28195
    },
    {
      "epoch": 7.44265540451366,
      "grad_norm": 0.23239991068840027,
      "learning_rate": 9.991064590302638e-06,
      "loss": 0.0395,
      "num_input_tokens_seen": 25205648,
      "step": 28200
    },
    {
      "epoch": 7.44265540451366,
      "eval_loss": 0.06724493950605392,
      "eval_runtime": 64.7838,
      "eval_samples_per_second": 103.961,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 25205648,
      "step": 28200
    },
    {
      "epoch": 7.443975188069157,
      "grad_norm": 0.1710798293352127,
      "learning_rate": 9.983214399074241e-06,
      "loss": 0.0303,
      "num_input_tokens_seen": 25209968,
      "step": 28205
    },
    {
      "epoch": 7.445294971624653,
      "grad_norm": 0.4014093279838562,
      "learning_rate": 9.975366523622893e-06,
      "loss": 0.1064,
      "num_input_tokens_seen": 25214992,
      "step": 28210
    },
    {
      "epoch": 7.446614755180151,
      "grad_norm": 0.2548074722290039,
      "learning_rate": 9.967520965158841e-06,
      "loss": 0.0506,
      "num_input_tokens_seen": 25219312,
      "step": 28215
    },
    {
      "epoch": 7.447934538735647,
      "grad_norm": 0.17595715820789337,
      "learning_rate": 9.95967772489197e-06,
      "loss": 0.0481,
      "num_input_tokens_seen": 25223888,
      "step": 28220
    },
    {
      "epoch": 7.449254322291145,
      "grad_norm": 0.19337214529514313,
      "learning_rate": 9.951836804031794e-06,
      "loss": 0.0455,
      "num_input_tokens_seen": 25228208,
      "step": 28225
    },
    {
      "epoch": 7.450574105846641,
      "grad_norm": 0.20132915675640106,
      "learning_rate": 9.943998203787489e-06,
      "loss": 0.1062,
      "num_input_tokens_seen": 25232912,
      "step": 28230
    },
    {
      "epoch": 7.451893889402138,
      "grad_norm": 0.19663147628307343,
      "learning_rate": 9.936161925367874e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 25237488,
      "step": 28235
    },
    {
      "epoch": 7.453213672957635,
      "grad_norm": 0.05027906969189644,
      "learning_rate": 9.928327969981386e-06,
      "loss": 0.05,
      "num_input_tokens_seen": 25241840,
      "step": 28240
    },
    {
      "epoch": 7.454533456513132,
      "grad_norm": 0.036288391798734665,
      "learning_rate": 9.920496338836135e-06,
      "loss": 0.0708,
      "num_input_tokens_seen": 25246384,
      "step": 28245
    },
    {
      "epoch": 7.455853240068628,
      "grad_norm": 0.29450228810310364,
      "learning_rate": 9.912667033139844e-06,
      "loss": 0.0427,
      "num_input_tokens_seen": 25250928,
      "step": 28250
    },
    {
      "epoch": 7.457173023624126,
      "grad_norm": 0.2381192296743393,
      "learning_rate": 9.904840054099893e-06,
      "loss": 0.0492,
      "num_input_tokens_seen": 25255248,
      "step": 28255
    },
    {
      "epoch": 7.458492807179622,
      "grad_norm": 0.19136130809783936,
      "learning_rate": 9.897015402923312e-06,
      "loss": 0.0557,
      "num_input_tokens_seen": 25259280,
      "step": 28260
    },
    {
      "epoch": 7.45981259073512,
      "grad_norm": 0.2605243921279907,
      "learning_rate": 9.889193080816744e-06,
      "loss": 0.065,
      "num_input_tokens_seen": 25263728,
      "step": 28265
    },
    {
      "epoch": 7.461132374290616,
      "grad_norm": 0.4398597776889801,
      "learning_rate": 9.881373088986498e-06,
      "loss": 0.0522,
      "num_input_tokens_seen": 25268208,
      "step": 28270
    },
    {
      "epoch": 7.462452157846113,
      "grad_norm": 0.0907437726855278,
      "learning_rate": 9.873555428638523e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 25272816,
      "step": 28275
    },
    {
      "epoch": 7.46377194140161,
      "grad_norm": 0.1638597697019577,
      "learning_rate": 9.865740100978383e-06,
      "loss": 0.1038,
      "num_input_tokens_seen": 25277392,
      "step": 28280
    },
    {
      "epoch": 7.465091724957107,
      "grad_norm": 0.3589237928390503,
      "learning_rate": 9.857927107211315e-06,
      "loss": 0.0287,
      "num_input_tokens_seen": 25281648,
      "step": 28285
    },
    {
      "epoch": 7.4664115085126035,
      "grad_norm": 0.43596720695495605,
      "learning_rate": 9.850116448542177e-06,
      "loss": 0.0853,
      "num_input_tokens_seen": 25286064,
      "step": 28290
    },
    {
      "epoch": 7.467731292068101,
      "grad_norm": 0.18783985078334808,
      "learning_rate": 9.842308126175457e-06,
      "loss": 0.0349,
      "num_input_tokens_seen": 25290384,
      "step": 28295
    },
    {
      "epoch": 7.4690510756235975,
      "grad_norm": 0.18272094428539276,
      "learning_rate": 9.834502141315315e-06,
      "loss": 0.0515,
      "num_input_tokens_seen": 25294736,
      "step": 28300
    },
    {
      "epoch": 7.470370859179095,
      "grad_norm": 0.06580014526844025,
      "learning_rate": 9.82669849516552e-06,
      "loss": 0.0217,
      "num_input_tokens_seen": 25299280,
      "step": 28305
    },
    {
      "epoch": 7.4716906427345915,
      "grad_norm": 0.07384705543518066,
      "learning_rate": 9.818897188929493e-06,
      "loss": 0.0576,
      "num_input_tokens_seen": 25303728,
      "step": 28310
    },
    {
      "epoch": 7.473010426290088,
      "grad_norm": 0.2425065040588379,
      "learning_rate": 9.811098223810309e-06,
      "loss": 0.0571,
      "num_input_tokens_seen": 25308112,
      "step": 28315
    },
    {
      "epoch": 7.4743302098455855,
      "grad_norm": 0.14675264060497284,
      "learning_rate": 9.803301601010641e-06,
      "loss": 0.0592,
      "num_input_tokens_seen": 25312656,
      "step": 28320
    },
    {
      "epoch": 7.475649993401082,
      "grad_norm": 0.20976220071315765,
      "learning_rate": 9.795507321732853e-06,
      "loss": 0.0526,
      "num_input_tokens_seen": 25317008,
      "step": 28325
    },
    {
      "epoch": 7.4769697769565795,
      "grad_norm": 0.2456665337085724,
      "learning_rate": 9.787715387178898e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 25321520,
      "step": 28330
    },
    {
      "epoch": 7.478289560512076,
      "grad_norm": 0.15069472789764404,
      "learning_rate": 9.779925798550399e-06,
      "loss": 0.0789,
      "num_input_tokens_seen": 25326032,
      "step": 28335
    },
    {
      "epoch": 7.479609344067573,
      "grad_norm": 0.26236090064048767,
      "learning_rate": 9.772138557048619e-06,
      "loss": 0.075,
      "num_input_tokens_seen": 25330320,
      "step": 28340
    },
    {
      "epoch": 7.48092912762307,
      "grad_norm": 0.13382378220558167,
      "learning_rate": 9.764353663874426e-06,
      "loss": 0.0592,
      "num_input_tokens_seen": 25334704,
      "step": 28345
    },
    {
      "epoch": 7.482248911178567,
      "grad_norm": 0.12811772525310516,
      "learning_rate": 9.756571120228375e-06,
      "loss": 0.034,
      "num_input_tokens_seen": 25339088,
      "step": 28350
    },
    {
      "epoch": 7.483568694734064,
      "grad_norm": 0.1899196207523346,
      "learning_rate": 9.748790927310605e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 25343728,
      "step": 28355
    },
    {
      "epoch": 7.484888478289561,
      "grad_norm": 0.7166891694068909,
      "learning_rate": 9.741013086320946e-06,
      "loss": 0.0926,
      "num_input_tokens_seen": 25348304,
      "step": 28360
    },
    {
      "epoch": 7.486208261845057,
      "grad_norm": 0.05646941438317299,
      "learning_rate": 9.733237598458821e-06,
      "loss": 0.0129,
      "num_input_tokens_seen": 25352720,
      "step": 28365
    },
    {
      "epoch": 7.487528045400555,
      "grad_norm": 0.07555802166461945,
      "learning_rate": 9.725464464923308e-06,
      "loss": 0.026,
      "num_input_tokens_seen": 25357232,
      "step": 28370
    },
    {
      "epoch": 7.488847828956051,
      "grad_norm": 0.14126692712306976,
      "learning_rate": 9.717693686913123e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 25361872,
      "step": 28375
    },
    {
      "epoch": 7.490167612511548,
      "grad_norm": 0.269870787858963,
      "learning_rate": 9.709925265626632e-06,
      "loss": 0.0754,
      "num_input_tokens_seen": 25366256,
      "step": 28380
    },
    {
      "epoch": 7.491487396067045,
      "grad_norm": 0.05480840057134628,
      "learning_rate": 9.702159202261801e-06,
      "loss": 0.0924,
      "num_input_tokens_seen": 25371056,
      "step": 28385
    },
    {
      "epoch": 7.492807179622542,
      "grad_norm": 0.27994129061698914,
      "learning_rate": 9.694395498016268e-06,
      "loss": 0.0468,
      "num_input_tokens_seen": 25375696,
      "step": 28390
    },
    {
      "epoch": 7.494126963178039,
      "grad_norm": 0.10177845507860184,
      "learning_rate": 9.686634154087298e-06,
      "loss": 0.0582,
      "num_input_tokens_seen": 25379888,
      "step": 28395
    },
    {
      "epoch": 7.495446746733536,
      "grad_norm": 0.44680944085121155,
      "learning_rate": 9.678875171671776e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 25384496,
      "step": 28400
    },
    {
      "epoch": 7.495446746733536,
      "eval_loss": 0.06719040870666504,
      "eval_runtime": 64.7755,
      "eval_samples_per_second": 103.975,
      "eval_steps_per_second": 25.997,
      "num_input_tokens_seen": 25384496,
      "step": 28400
    },
    {
      "epoch": 7.496766530289032,
      "grad_norm": 0.18420284986495972,
      "learning_rate": 9.671118551966246e-06,
      "loss": 0.0279,
      "num_input_tokens_seen": 25389200,
      "step": 28405
    },
    {
      "epoch": 7.49808631384453,
      "grad_norm": 0.09603328257799149,
      "learning_rate": 9.66336429616686e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 25393744,
      "step": 28410
    },
    {
      "epoch": 7.499406097400026,
      "grad_norm": 0.07899913191795349,
      "learning_rate": 9.655612405469436e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 25398192,
      "step": 28415
    },
    {
      "epoch": 7.500725880955523,
      "grad_norm": 0.12213881313800812,
      "learning_rate": 9.647862881069413e-06,
      "loss": 0.1169,
      "num_input_tokens_seen": 25402960,
      "step": 28420
    },
    {
      "epoch": 7.50204566451102,
      "grad_norm": 0.3203723430633545,
      "learning_rate": 9.640115724161855e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 25407696,
      "step": 28425
    },
    {
      "epoch": 7.503365448066517,
      "grad_norm": 0.30350419878959656,
      "learning_rate": 9.632370935941483e-06,
      "loss": 0.1009,
      "num_input_tokens_seen": 25412272,
      "step": 28430
    },
    {
      "epoch": 7.504685231622014,
      "grad_norm": 0.20776693522930145,
      "learning_rate": 9.624628517602634e-06,
      "loss": 0.0332,
      "num_input_tokens_seen": 25416624,
      "step": 28435
    },
    {
      "epoch": 7.506005015177511,
      "grad_norm": 0.24014948308467865,
      "learning_rate": 9.61688847033928e-06,
      "loss": 0.0268,
      "num_input_tokens_seen": 25421200,
      "step": 28440
    },
    {
      "epoch": 7.507324798733007,
      "grad_norm": 0.07349469512701035,
      "learning_rate": 9.609150795345051e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 25425904,
      "step": 28445
    },
    {
      "epoch": 7.508644582288505,
      "grad_norm": 0.05070003122091293,
      "learning_rate": 9.601415493813171e-06,
      "loss": 0.0319,
      "num_input_tokens_seen": 25430512,
      "step": 28450
    },
    {
      "epoch": 7.509964365844001,
      "grad_norm": 0.06833989918231964,
      "learning_rate": 9.593682566936533e-06,
      "loss": 0.0255,
      "num_input_tokens_seen": 25435216,
      "step": 28455
    },
    {
      "epoch": 7.511284149399499,
      "grad_norm": 0.05357332527637482,
      "learning_rate": 9.58595201590766e-06,
      "loss": 0.0502,
      "num_input_tokens_seen": 25439856,
      "step": 28460
    },
    {
      "epoch": 7.512603932954995,
      "grad_norm": 0.12030693143606186,
      "learning_rate": 9.578223841918681e-06,
      "loss": 0.0202,
      "num_input_tokens_seen": 25444496,
      "step": 28465
    },
    {
      "epoch": 7.513923716510492,
      "grad_norm": 0.23244842886924744,
      "learning_rate": 9.570498046161389e-06,
      "loss": 0.0548,
      "num_input_tokens_seen": 25448976,
      "step": 28470
    },
    {
      "epoch": 7.515243500065989,
      "grad_norm": 0.11157065629959106,
      "learning_rate": 9.562774629827206e-06,
      "loss": 0.0439,
      "num_input_tokens_seen": 25453328,
      "step": 28475
    },
    {
      "epoch": 7.516563283621486,
      "grad_norm": 0.1607305109500885,
      "learning_rate": 9.555053594107163e-06,
      "loss": 0.0295,
      "num_input_tokens_seen": 25457712,
      "step": 28480
    },
    {
      "epoch": 7.517883067176983,
      "grad_norm": 0.08802018314599991,
      "learning_rate": 9.547334940191957e-06,
      "loss": 0.0411,
      "num_input_tokens_seen": 25462512,
      "step": 28485
    },
    {
      "epoch": 7.51920285073248,
      "grad_norm": 0.3966522514820099,
      "learning_rate": 9.539618669271886e-06,
      "loss": 0.1033,
      "num_input_tokens_seen": 25466864,
      "step": 28490
    },
    {
      "epoch": 7.5205226342879765,
      "grad_norm": 0.15384790301322937,
      "learning_rate": 9.531904782536904e-06,
      "loss": 0.0305,
      "num_input_tokens_seen": 25471248,
      "step": 28495
    },
    {
      "epoch": 7.521842417843474,
      "grad_norm": 0.056325383484363556,
      "learning_rate": 9.524193281176597e-06,
      "loss": 0.0997,
      "num_input_tokens_seen": 25475696,
      "step": 28500
    },
    {
      "epoch": 7.5231622013989705,
      "grad_norm": 0.18907099962234497,
      "learning_rate": 9.516484166380165e-06,
      "loss": 0.0446,
      "num_input_tokens_seen": 25480112,
      "step": 28505
    },
    {
      "epoch": 7.524481984954468,
      "grad_norm": 0.08811620622873306,
      "learning_rate": 9.508777439336447e-06,
      "loss": 0.043,
      "num_input_tokens_seen": 25484368,
      "step": 28510
    },
    {
      "epoch": 7.5258017685099645,
      "grad_norm": 0.0339387021958828,
      "learning_rate": 9.50107310123393e-06,
      "loss": 0.0315,
      "num_input_tokens_seen": 25488624,
      "step": 28515
    },
    {
      "epoch": 7.527121552065461,
      "grad_norm": 0.09460266679525375,
      "learning_rate": 9.493371153260702e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 25492912,
      "step": 28520
    },
    {
      "epoch": 7.5284413356209585,
      "grad_norm": 0.20555604994297028,
      "learning_rate": 9.485671596604523e-06,
      "loss": 0.0395,
      "num_input_tokens_seen": 25497424,
      "step": 28525
    },
    {
      "epoch": 7.529761119176455,
      "grad_norm": 0.07615350931882858,
      "learning_rate": 9.477974432452738e-06,
      "loss": 0.052,
      "num_input_tokens_seen": 25502064,
      "step": 28530
    },
    {
      "epoch": 7.531080902731952,
      "grad_norm": 0.08699978142976761,
      "learning_rate": 9.470279661992356e-06,
      "loss": 0.0224,
      "num_input_tokens_seen": 25506288,
      "step": 28535
    },
    {
      "epoch": 7.532400686287449,
      "grad_norm": 0.14378374814987183,
      "learning_rate": 9.462587286410021e-06,
      "loss": 0.0754,
      "num_input_tokens_seen": 25510960,
      "step": 28540
    },
    {
      "epoch": 7.533720469842946,
      "grad_norm": 0.20876257121562958,
      "learning_rate": 9.454897306891972e-06,
      "loss": 0.0441,
      "num_input_tokens_seen": 25515408,
      "step": 28545
    },
    {
      "epoch": 7.535040253398442,
      "grad_norm": 0.4181246757507324,
      "learning_rate": 9.44720972462411e-06,
      "loss": 0.0853,
      "num_input_tokens_seen": 25519696,
      "step": 28550
    },
    {
      "epoch": 7.53636003695394,
      "grad_norm": 0.05389338359236717,
      "learning_rate": 9.439524540791964e-06,
      "loss": 0.0575,
      "num_input_tokens_seen": 25524144,
      "step": 28555
    },
    {
      "epoch": 7.537679820509436,
      "grad_norm": 0.5026947259902954,
      "learning_rate": 9.431841756580673e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 25528560,
      "step": 28560
    },
    {
      "epoch": 7.538999604064934,
      "grad_norm": 0.34509214758872986,
      "learning_rate": 9.42416137317503e-06,
      "loss": 0.091,
      "num_input_tokens_seen": 25532912,
      "step": 28565
    },
    {
      "epoch": 7.54031938762043,
      "grad_norm": 0.30521687865257263,
      "learning_rate": 9.416483391759437e-06,
      "loss": 0.0618,
      "num_input_tokens_seen": 25537296,
      "step": 28570
    },
    {
      "epoch": 7.541639171175927,
      "grad_norm": 0.16312257945537567,
      "learning_rate": 9.408807813517945e-06,
      "loss": 0.0295,
      "num_input_tokens_seen": 25541424,
      "step": 28575
    },
    {
      "epoch": 7.542958954731424,
      "grad_norm": 0.2657415568828583,
      "learning_rate": 9.401134639634221e-06,
      "loss": 0.0453,
      "num_input_tokens_seen": 25545936,
      "step": 28580
    },
    {
      "epoch": 7.544278738286921,
      "grad_norm": 0.1762724369764328,
      "learning_rate": 9.393463871291555e-06,
      "loss": 0.0468,
      "num_input_tokens_seen": 25550448,
      "step": 28585
    },
    {
      "epoch": 7.545598521842418,
      "grad_norm": 0.05827685818076134,
      "learning_rate": 9.385795509672881e-06,
      "loss": 0.0714,
      "num_input_tokens_seen": 25554640,
      "step": 28590
    },
    {
      "epoch": 7.546918305397915,
      "grad_norm": 0.12334321439266205,
      "learning_rate": 9.378129555960771e-06,
      "loss": 0.0349,
      "num_input_tokens_seen": 25559248,
      "step": 28595
    },
    {
      "epoch": 7.548238088953411,
      "grad_norm": 0.06427537649869919,
      "learning_rate": 9.370466011337392e-06,
      "loss": 0.0817,
      "num_input_tokens_seen": 25563856,
      "step": 28600
    },
    {
      "epoch": 7.548238088953411,
      "eval_loss": 0.06737852096557617,
      "eval_runtime": 64.7733,
      "eval_samples_per_second": 103.978,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 25563856,
      "step": 28600
    },
    {
      "epoch": 7.549557872508909,
      "grad_norm": 0.12501057982444763,
      "learning_rate": 9.362804876984573e-06,
      "loss": 0.044,
      "num_input_tokens_seen": 25568464,
      "step": 28605
    },
    {
      "epoch": 7.550877656064405,
      "grad_norm": 0.5499928593635559,
      "learning_rate": 9.355146154083747e-06,
      "loss": 0.0488,
      "num_input_tokens_seen": 25572784,
      "step": 28610
    },
    {
      "epoch": 7.552197439619903,
      "grad_norm": 0.22544871270656586,
      "learning_rate": 9.347489843815987e-06,
      "loss": 0.0492,
      "num_input_tokens_seen": 25577168,
      "step": 28615
    },
    {
      "epoch": 7.553517223175399,
      "grad_norm": 0.2568769156932831,
      "learning_rate": 9.339835947362002e-06,
      "loss": 0.0713,
      "num_input_tokens_seen": 25581680,
      "step": 28620
    },
    {
      "epoch": 7.554837006730896,
      "grad_norm": 0.28132879734039307,
      "learning_rate": 9.332184465902105e-06,
      "loss": 0.0458,
      "num_input_tokens_seen": 25586384,
      "step": 28625
    },
    {
      "epoch": 7.556156790286393,
      "grad_norm": 0.11143406480550766,
      "learning_rate": 9.324535400616266e-06,
      "loss": 0.0603,
      "num_input_tokens_seen": 25590928,
      "step": 28630
    },
    {
      "epoch": 7.55747657384189,
      "grad_norm": 0.3229554295539856,
      "learning_rate": 9.31688875268405e-06,
      "loss": 0.0357,
      "num_input_tokens_seen": 25595600,
      "step": 28635
    },
    {
      "epoch": 7.558796357397387,
      "grad_norm": 0.14877256751060486,
      "learning_rate": 9.309244523284674e-06,
      "loss": 0.0649,
      "num_input_tokens_seen": 25599920,
      "step": 28640
    },
    {
      "epoch": 7.560116140952884,
      "grad_norm": 0.1517946869134903,
      "learning_rate": 9.301602713596982e-06,
      "loss": 0.0166,
      "num_input_tokens_seen": 25604304,
      "step": 28645
    },
    {
      "epoch": 7.56143592450838,
      "grad_norm": 0.07871007174253464,
      "learning_rate": 9.293963324799432e-06,
      "loss": 0.0343,
      "num_input_tokens_seen": 25608912,
      "step": 28650
    },
    {
      "epoch": 7.562755708063878,
      "grad_norm": 0.11559867858886719,
      "learning_rate": 9.286326358070104e-06,
      "loss": 0.034,
      "num_input_tokens_seen": 25613360,
      "step": 28655
    },
    {
      "epoch": 7.564075491619374,
      "grad_norm": 0.262851357460022,
      "learning_rate": 9.278691814586729e-06,
      "loss": 0.0743,
      "num_input_tokens_seen": 25617552,
      "step": 28660
    },
    {
      "epoch": 7.565395275174871,
      "grad_norm": 0.21714268624782562,
      "learning_rate": 9.271059695526635e-06,
      "loss": 0.0658,
      "num_input_tokens_seen": 25621840,
      "step": 28665
    },
    {
      "epoch": 7.566715058730368,
      "grad_norm": 0.24854934215545654,
      "learning_rate": 9.263430002066805e-06,
      "loss": 0.1493,
      "num_input_tokens_seen": 25626288,
      "step": 28670
    },
    {
      "epoch": 7.568034842285865,
      "grad_norm": 0.30293935537338257,
      "learning_rate": 9.25580273538382e-06,
      "loss": 0.0663,
      "num_input_tokens_seen": 25630960,
      "step": 28675
    },
    {
      "epoch": 7.5693546258413615,
      "grad_norm": 0.37781822681427,
      "learning_rate": 9.248177896653907e-06,
      "loss": 0.0591,
      "num_input_tokens_seen": 25635632,
      "step": 28680
    },
    {
      "epoch": 7.570674409396859,
      "grad_norm": 0.3661993741989136,
      "learning_rate": 9.240555487052918e-06,
      "loss": 0.0207,
      "num_input_tokens_seen": 25640336,
      "step": 28685
    },
    {
      "epoch": 7.5719941929523555,
      "grad_norm": 0.2354733943939209,
      "learning_rate": 9.232935507756313e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 25644912,
      "step": 28690
    },
    {
      "epoch": 7.573313976507853,
      "grad_norm": 0.2871819734573364,
      "learning_rate": 9.225317959939193e-06,
      "loss": 0.0704,
      "num_input_tokens_seen": 25649296,
      "step": 28695
    },
    {
      "epoch": 7.5746337600633495,
      "grad_norm": 0.0757763460278511,
      "learning_rate": 9.217702844776287e-06,
      "loss": 0.066,
      "num_input_tokens_seen": 25653552,
      "step": 28700
    },
    {
      "epoch": 7.575953543618846,
      "grad_norm": 0.16725897789001465,
      "learning_rate": 9.210090163441929e-06,
      "loss": 0.0361,
      "num_input_tokens_seen": 25657808,
      "step": 28705
    },
    {
      "epoch": 7.5772733271743435,
      "grad_norm": 0.08254332840442657,
      "learning_rate": 9.202479917110105e-06,
      "loss": 0.0787,
      "num_input_tokens_seen": 25662128,
      "step": 28710
    },
    {
      "epoch": 7.57859311072984,
      "grad_norm": 0.06442095339298248,
      "learning_rate": 9.194872106954392e-06,
      "loss": 0.0762,
      "num_input_tokens_seen": 25666960,
      "step": 28715
    },
    {
      "epoch": 7.5799128942853375,
      "grad_norm": 0.5059494376182556,
      "learning_rate": 9.187266734148029e-06,
      "loss": 0.1057,
      "num_input_tokens_seen": 25671632,
      "step": 28720
    },
    {
      "epoch": 7.581232677840834,
      "grad_norm": 0.32733625173568726,
      "learning_rate": 9.179663799863849e-06,
      "loss": 0.0536,
      "num_input_tokens_seen": 25676016,
      "step": 28725
    },
    {
      "epoch": 7.582552461396331,
      "grad_norm": 0.33237534761428833,
      "learning_rate": 9.172063305274317e-06,
      "loss": 0.0835,
      "num_input_tokens_seen": 25680752,
      "step": 28730
    },
    {
      "epoch": 7.583872244951828,
      "grad_norm": 0.5302887558937073,
      "learning_rate": 9.164465251551527e-06,
      "loss": 0.183,
      "num_input_tokens_seen": 25685232,
      "step": 28735
    },
    {
      "epoch": 7.585192028507325,
      "grad_norm": 0.07520095258951187,
      "learning_rate": 9.156869639867205e-06,
      "loss": 0.0432,
      "num_input_tokens_seen": 25689552,
      "step": 28740
    },
    {
      "epoch": 7.586511812062822,
      "grad_norm": 0.029999621212482452,
      "learning_rate": 9.149276471392677e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 25694224,
      "step": 28745
    },
    {
      "epoch": 7.587831595618319,
      "grad_norm": 0.2688811123371124,
      "learning_rate": 9.141685747298914e-06,
      "loss": 0.0778,
      "num_input_tokens_seen": 25699056,
      "step": 28750
    },
    {
      "epoch": 7.589151379173815,
      "grad_norm": 0.20439714193344116,
      "learning_rate": 9.13409746875649e-06,
      "loss": 0.0251,
      "num_input_tokens_seen": 25703312,
      "step": 28755
    },
    {
      "epoch": 7.590471162729313,
      "grad_norm": 0.045238979160785675,
      "learning_rate": 9.12651163693562e-06,
      "loss": 0.0348,
      "num_input_tokens_seen": 25707824,
      "step": 28760
    },
    {
      "epoch": 7.591790946284809,
      "grad_norm": 0.5429162383079529,
      "learning_rate": 9.11892825300614e-06,
      "loss": 0.0979,
      "num_input_tokens_seen": 25712528,
      "step": 28765
    },
    {
      "epoch": 7.593110729840307,
      "grad_norm": 0.21179956197738647,
      "learning_rate": 9.111347318137491e-06,
      "loss": 0.0966,
      "num_input_tokens_seen": 25717328,
      "step": 28770
    },
    {
      "epoch": 7.594430513395803,
      "grad_norm": 0.03599433973431587,
      "learning_rate": 9.103768833498755e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 25721424,
      "step": 28775
    },
    {
      "epoch": 7.5957502969513,
      "grad_norm": 0.0834731012582779,
      "learning_rate": 9.096192800258639e-06,
      "loss": 0.0437,
      "num_input_tokens_seen": 25725808,
      "step": 28780
    },
    {
      "epoch": 7.597070080506797,
      "grad_norm": 0.1773025542497635,
      "learning_rate": 9.088619219585443e-06,
      "loss": 0.0428,
      "num_input_tokens_seen": 25730192,
      "step": 28785
    },
    {
      "epoch": 7.598389864062294,
      "grad_norm": 0.3067278265953064,
      "learning_rate": 9.081048092647127e-06,
      "loss": 0.0459,
      "num_input_tokens_seen": 25734512,
      "step": 28790
    },
    {
      "epoch": 7.59970964761779,
      "grad_norm": 0.1530575156211853,
      "learning_rate": 9.073479420611245e-06,
      "loss": 0.0679,
      "num_input_tokens_seen": 25738864,
      "step": 28795
    },
    {
      "epoch": 7.601029431173288,
      "grad_norm": 0.20797239243984222,
      "learning_rate": 9.065913204644974e-06,
      "loss": 0.0346,
      "num_input_tokens_seen": 25743536,
      "step": 28800
    },
    {
      "epoch": 7.601029431173288,
      "eval_loss": 0.06708266586065292,
      "eval_runtime": 64.8109,
      "eval_samples_per_second": 103.918,
      "eval_steps_per_second": 25.983,
      "num_input_tokens_seen": 25743536,
      "step": 28800
    },
    {
      "epoch": 7.602349214728784,
      "grad_norm": 0.20771774649620056,
      "learning_rate": 9.058349445915135e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 25747984,
      "step": 28805
    },
    {
      "epoch": 7.603668998284281,
      "grad_norm": 0.14603017270565033,
      "learning_rate": 9.050788145588138e-06,
      "loss": 0.0607,
      "num_input_tokens_seen": 25752400,
      "step": 28810
    },
    {
      "epoch": 7.604988781839778,
      "grad_norm": 0.18702679872512817,
      "learning_rate": 9.043229304830039e-06,
      "loss": 0.039,
      "num_input_tokens_seen": 25756976,
      "step": 28815
    },
    {
      "epoch": 7.606308565395275,
      "grad_norm": 0.46001186966896057,
      "learning_rate": 9.035672924806515e-06,
      "loss": 0.1038,
      "num_input_tokens_seen": 25761040,
      "step": 28820
    },
    {
      "epoch": 7.607628348950772,
      "grad_norm": 0.24940703809261322,
      "learning_rate": 9.028119006682839e-06,
      "loss": 0.0602,
      "num_input_tokens_seen": 25765712,
      "step": 28825
    },
    {
      "epoch": 7.608948132506269,
      "grad_norm": 0.33831915259361267,
      "learning_rate": 9.020567551623935e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 25770352,
      "step": 28830
    },
    {
      "epoch": 7.6102679160617654,
      "grad_norm": 0.14558133482933044,
      "learning_rate": 9.013018560794318e-06,
      "loss": 0.0321,
      "num_input_tokens_seen": 25774864,
      "step": 28835
    },
    {
      "epoch": 7.611587699617263,
      "grad_norm": 0.0739254280924797,
      "learning_rate": 9.005472035358139e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 25779408,
      "step": 28840
    },
    {
      "epoch": 7.6129074831727594,
      "grad_norm": 0.4416583180427551,
      "learning_rate": 8.997927976479185e-06,
      "loss": 0.0717,
      "num_input_tokens_seen": 25783632,
      "step": 28845
    },
    {
      "epoch": 7.614227266728257,
      "grad_norm": 0.3231644034385681,
      "learning_rate": 8.99038638532082e-06,
      "loss": 0.0565,
      "num_input_tokens_seen": 25788368,
      "step": 28850
    },
    {
      "epoch": 7.6155470502837534,
      "grad_norm": 0.053276024758815765,
      "learning_rate": 8.982847263046065e-06,
      "loss": 0.0483,
      "num_input_tokens_seen": 25793104,
      "step": 28855
    },
    {
      "epoch": 7.61686683383925,
      "grad_norm": 0.23859666287899017,
      "learning_rate": 8.975310610817555e-06,
      "loss": 0.0449,
      "num_input_tokens_seen": 25797296,
      "step": 28860
    },
    {
      "epoch": 7.6181866173947475,
      "grad_norm": 0.07422548532485962,
      "learning_rate": 8.967776429797528e-06,
      "loss": 0.0756,
      "num_input_tokens_seen": 25801488,
      "step": 28865
    },
    {
      "epoch": 7.619506400950244,
      "grad_norm": 0.06326866149902344,
      "learning_rate": 8.960244721147842e-06,
      "loss": 0.0353,
      "num_input_tokens_seen": 25805776,
      "step": 28870
    },
    {
      "epoch": 7.6208261845057415,
      "grad_norm": 0.1297316998243332,
      "learning_rate": 8.952715486029995e-06,
      "loss": 0.0215,
      "num_input_tokens_seen": 25810000,
      "step": 28875
    },
    {
      "epoch": 7.622145968061238,
      "grad_norm": 0.03616291657090187,
      "learning_rate": 8.945188725605075e-06,
      "loss": 0.0784,
      "num_input_tokens_seen": 25814512,
      "step": 28880
    },
    {
      "epoch": 7.623465751616735,
      "grad_norm": 0.10423058271408081,
      "learning_rate": 8.937664441033817e-06,
      "loss": 0.0249,
      "num_input_tokens_seen": 25818640,
      "step": 28885
    },
    {
      "epoch": 7.624785535172232,
      "grad_norm": 0.024196263402700424,
      "learning_rate": 8.930142633476549e-06,
      "loss": 0.0465,
      "num_input_tokens_seen": 25822736,
      "step": 28890
    },
    {
      "epoch": 7.626105318727729,
      "grad_norm": 0.26560425758361816,
      "learning_rate": 8.92262330409323e-06,
      "loss": 0.0853,
      "num_input_tokens_seen": 25827024,
      "step": 28895
    },
    {
      "epoch": 7.627425102283226,
      "grad_norm": 0.14553707838058472,
      "learning_rate": 8.915106454043448e-06,
      "loss": 0.0271,
      "num_input_tokens_seen": 25831536,
      "step": 28900
    },
    {
      "epoch": 7.628744885838723,
      "grad_norm": 0.25742220878601074,
      "learning_rate": 8.90759208448638e-06,
      "loss": 0.0823,
      "num_input_tokens_seen": 25836144,
      "step": 28905
    },
    {
      "epoch": 7.630064669394219,
      "grad_norm": 0.23736485838890076,
      "learning_rate": 8.900080196580848e-06,
      "loss": 0.0552,
      "num_input_tokens_seen": 25840496,
      "step": 28910
    },
    {
      "epoch": 7.631384452949717,
      "grad_norm": 0.16533923149108887,
      "learning_rate": 8.892570791485267e-06,
      "loss": 0.0572,
      "num_input_tokens_seen": 25845072,
      "step": 28915
    },
    {
      "epoch": 7.632704236505213,
      "grad_norm": 0.18870149552822113,
      "learning_rate": 8.885063870357688e-06,
      "loss": 0.0344,
      "num_input_tokens_seen": 25849392,
      "step": 28920
    },
    {
      "epoch": 7.63402402006071,
      "grad_norm": 0.2535285949707031,
      "learning_rate": 8.87755943435578e-06,
      "loss": 0.045,
      "num_input_tokens_seen": 25854064,
      "step": 28925
    },
    {
      "epoch": 7.635343803616207,
      "grad_norm": 0.05172240361571312,
      "learning_rate": 8.87005748463681e-06,
      "loss": 0.0509,
      "num_input_tokens_seen": 25858640,
      "step": 28930
    },
    {
      "epoch": 7.636663587171704,
      "grad_norm": 0.1622723489999771,
      "learning_rate": 8.862558022357681e-06,
      "loss": 0.0698,
      "num_input_tokens_seen": 25862800,
      "step": 28935
    },
    {
      "epoch": 7.637983370727201,
      "grad_norm": 0.06722614169120789,
      "learning_rate": 8.855061048674903e-06,
      "loss": 0.024,
      "num_input_tokens_seen": 25866992,
      "step": 28940
    },
    {
      "epoch": 7.639303154282698,
      "grad_norm": 0.5979558229446411,
      "learning_rate": 8.847566564744595e-06,
      "loss": 0.0502,
      "num_input_tokens_seen": 25871600,
      "step": 28945
    },
    {
      "epoch": 7.640622937838194,
      "grad_norm": 0.10905921459197998,
      "learning_rate": 8.840074571722512e-06,
      "loss": 0.1028,
      "num_input_tokens_seen": 25876240,
      "step": 28950
    },
    {
      "epoch": 7.641942721393692,
      "grad_norm": 0.3263806700706482,
      "learning_rate": 8.832585070764002e-06,
      "loss": 0.058,
      "num_input_tokens_seen": 25880912,
      "step": 28955
    },
    {
      "epoch": 7.643262504949188,
      "grad_norm": 0.4214063584804535,
      "learning_rate": 8.825098063024045e-06,
      "loss": 0.0343,
      "num_input_tokens_seen": 25885328,
      "step": 28960
    },
    {
      "epoch": 7.644582288504685,
      "grad_norm": 0.1956067681312561,
      "learning_rate": 8.817613549657244e-06,
      "loss": 0.0391,
      "num_input_tokens_seen": 25889744,
      "step": 28965
    },
    {
      "epoch": 7.645902072060182,
      "grad_norm": 0.19761741161346436,
      "learning_rate": 8.810131531817783e-06,
      "loss": 0.0548,
      "num_input_tokens_seen": 25894352,
      "step": 28970
    },
    {
      "epoch": 7.647221855615679,
      "grad_norm": 0.48980140686035156,
      "learning_rate": 8.802652010659496e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 25898480,
      "step": 28975
    },
    {
      "epoch": 7.648541639171176,
      "grad_norm": 0.3829546272754669,
      "learning_rate": 8.795174987335827e-06,
      "loss": 0.1288,
      "num_input_tokens_seen": 25902864,
      "step": 28980
    },
    {
      "epoch": 7.649861422726673,
      "grad_norm": 0.23275887966156006,
      "learning_rate": 8.787700462999807e-06,
      "loss": 0.063,
      "num_input_tokens_seen": 25907792,
      "step": 28985
    },
    {
      "epoch": 7.651181206282169,
      "grad_norm": 0.25544312596321106,
      "learning_rate": 8.780228438804122e-06,
      "loss": 0.0397,
      "num_input_tokens_seen": 25912432,
      "step": 28990
    },
    {
      "epoch": 7.652500989837667,
      "grad_norm": 0.2271474301815033,
      "learning_rate": 8.772758915901032e-06,
      "loss": 0.0665,
      "num_input_tokens_seen": 25917136,
      "step": 28995
    },
    {
      "epoch": 7.653820773393163,
      "grad_norm": 0.22674846649169922,
      "learning_rate": 8.765291895442443e-06,
      "loss": 0.0321,
      "num_input_tokens_seen": 25921616,
      "step": 29000
    },
    {
      "epoch": 7.653820773393163,
      "eval_loss": 0.06706395745277405,
      "eval_runtime": 64.8256,
      "eval_samples_per_second": 103.894,
      "eval_steps_per_second": 25.977,
      "num_input_tokens_seen": 25921616,
      "step": 29000
    },
    {
      "epoch": 7.655140556948661,
      "grad_norm": 0.15786947309970856,
      "learning_rate": 8.75782737857987e-06,
      "loss": 0.0828,
      "num_input_tokens_seen": 25926160,
      "step": 29005
    },
    {
      "epoch": 7.656460340504157,
      "grad_norm": 0.32487648725509644,
      "learning_rate": 8.750365366464425e-06,
      "loss": 0.1307,
      "num_input_tokens_seen": 25930736,
      "step": 29010
    },
    {
      "epoch": 7.657780124059654,
      "grad_norm": 0.12603232264518738,
      "learning_rate": 8.742905860246838e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 25935184,
      "step": 29015
    },
    {
      "epoch": 7.659099907615151,
      "grad_norm": 0.3401598036289215,
      "learning_rate": 8.735448861077478e-06,
      "loss": 0.1034,
      "num_input_tokens_seen": 25939408,
      "step": 29020
    },
    {
      "epoch": 7.660419691170648,
      "grad_norm": 0.03564178943634033,
      "learning_rate": 8.727994370106288e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 25943760,
      "step": 29025
    },
    {
      "epoch": 7.661739474726145,
      "grad_norm": 0.12660466134548187,
      "learning_rate": 8.720542388482861e-06,
      "loss": 0.0298,
      "num_input_tokens_seen": 25948304,
      "step": 29030
    },
    {
      "epoch": 7.663059258281642,
      "grad_norm": 0.12257567793130875,
      "learning_rate": 8.71309291735637e-06,
      "loss": 0.0539,
      "num_input_tokens_seen": 25952624,
      "step": 29035
    },
    {
      "epoch": 7.6643790418371385,
      "grad_norm": 0.09583305567502975,
      "learning_rate": 8.705645957875621e-06,
      "loss": 0.0418,
      "num_input_tokens_seen": 25957072,
      "step": 29040
    },
    {
      "epoch": 7.665698825392636,
      "grad_norm": 0.0794573649764061,
      "learning_rate": 8.698201511189048e-06,
      "loss": 0.0414,
      "num_input_tokens_seen": 25961712,
      "step": 29045
    },
    {
      "epoch": 7.6670186089481325,
      "grad_norm": 0.045155834406614304,
      "learning_rate": 8.690759578444649e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 25966288,
      "step": 29050
    },
    {
      "epoch": 7.668338392503629,
      "grad_norm": 0.207078754901886,
      "learning_rate": 8.68332016079008e-06,
      "loss": 0.044,
      "num_input_tokens_seen": 25970704,
      "step": 29055
    },
    {
      "epoch": 7.6696581760591265,
      "grad_norm": 0.2654801309108734,
      "learning_rate": 8.6758832593726e-06,
      "loss": 0.0818,
      "num_input_tokens_seen": 25975440,
      "step": 29060
    },
    {
      "epoch": 7.670977959614623,
      "grad_norm": 0.06456955522298813,
      "learning_rate": 8.668448875339053e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 25980016,
      "step": 29065
    },
    {
      "epoch": 7.6722977431701205,
      "grad_norm": 0.08655328303575516,
      "learning_rate": 8.661017009835933e-06,
      "loss": 0.0741,
      "num_input_tokens_seen": 25984624,
      "step": 29070
    },
    {
      "epoch": 7.673617526725617,
      "grad_norm": 0.16649572551250458,
      "learning_rate": 8.653587664009311e-06,
      "loss": 0.0315,
      "num_input_tokens_seen": 25989168,
      "step": 29075
    },
    {
      "epoch": 7.674937310281114,
      "grad_norm": 0.16485533118247986,
      "learning_rate": 8.646160839004902e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 25993552,
      "step": 29080
    },
    {
      "epoch": 7.676257093836611,
      "grad_norm": 0.4182368218898773,
      "learning_rate": 8.638736535967998e-06,
      "loss": 0.0618,
      "num_input_tokens_seen": 25998160,
      "step": 29085
    },
    {
      "epoch": 7.677576877392108,
      "grad_norm": 0.4864540696144104,
      "learning_rate": 8.631314756043535e-06,
      "loss": 0.1188,
      "num_input_tokens_seen": 26003056,
      "step": 29090
    },
    {
      "epoch": 7.678896660947604,
      "grad_norm": 0.3631393313407898,
      "learning_rate": 8.62389550037603e-06,
      "loss": 0.0752,
      "num_input_tokens_seen": 26007568,
      "step": 29095
    },
    {
      "epoch": 7.680216444503102,
      "grad_norm": 0.06107712537050247,
      "learning_rate": 8.616478770109646e-06,
      "loss": 0.0465,
      "num_input_tokens_seen": 26011984,
      "step": 29100
    },
    {
      "epoch": 7.681536228058598,
      "grad_norm": 0.20940859615802765,
      "learning_rate": 8.609064566388111e-06,
      "loss": 0.0452,
      "num_input_tokens_seen": 26016368,
      "step": 29105
    },
    {
      "epoch": 7.682856011614096,
      "grad_norm": 0.26821354031562805,
      "learning_rate": 8.601652890354815e-06,
      "loss": 0.0652,
      "num_input_tokens_seen": 26020848,
      "step": 29110
    },
    {
      "epoch": 7.684175795169592,
      "grad_norm": 0.17707842588424683,
      "learning_rate": 8.594243743152705e-06,
      "loss": 0.0542,
      "num_input_tokens_seen": 26025680,
      "step": 29115
    },
    {
      "epoch": 7.685495578725089,
      "grad_norm": 0.21193699538707733,
      "learning_rate": 8.58683712592438e-06,
      "loss": 0.0383,
      "num_input_tokens_seen": 26030416,
      "step": 29120
    },
    {
      "epoch": 7.686815362280586,
      "grad_norm": 0.45034798979759216,
      "learning_rate": 8.579433039812037e-06,
      "loss": 0.1112,
      "num_input_tokens_seen": 26034864,
      "step": 29125
    },
    {
      "epoch": 7.688135145836083,
      "grad_norm": 0.2201678305864334,
      "learning_rate": 8.572031485957466e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 26039408,
      "step": 29130
    },
    {
      "epoch": 7.68945492939158,
      "grad_norm": 0.16068312525749207,
      "learning_rate": 8.564632465502084e-06,
      "loss": 0.0636,
      "num_input_tokens_seen": 26044112,
      "step": 29135
    },
    {
      "epoch": 7.690774712947077,
      "grad_norm": 0.07296279072761536,
      "learning_rate": 8.557235979586928e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 26048848,
      "step": 29140
    },
    {
      "epoch": 7.692094496502573,
      "grad_norm": 0.5406801104545593,
      "learning_rate": 8.549842029352606e-06,
      "loss": 0.1161,
      "num_input_tokens_seen": 26053328,
      "step": 29145
    },
    {
      "epoch": 7.693414280058071,
      "grad_norm": 0.04569797217845917,
      "learning_rate": 8.542450615939376e-06,
      "loss": 0.0634,
      "num_input_tokens_seen": 26057488,
      "step": 29150
    },
    {
      "epoch": 7.694734063613567,
      "grad_norm": 0.08818359673023224,
      "learning_rate": 8.535061740487082e-06,
      "loss": 0.0728,
      "num_input_tokens_seen": 26061968,
      "step": 29155
    },
    {
      "epoch": 7.696053847169065,
      "grad_norm": 0.2057567685842514,
      "learning_rate": 8.527675404135168e-06,
      "loss": 0.0492,
      "num_input_tokens_seen": 26066288,
      "step": 29160
    },
    {
      "epoch": 7.697373630724561,
      "grad_norm": 0.47638198733329773,
      "learning_rate": 8.520291608022724e-06,
      "loss": 0.0322,
      "num_input_tokens_seen": 26070832,
      "step": 29165
    },
    {
      "epoch": 7.698693414280058,
      "grad_norm": 0.2535324692726135,
      "learning_rate": 8.512910353288398e-06,
      "loss": 0.0435,
      "num_input_tokens_seen": 26075440,
      "step": 29170
    },
    {
      "epoch": 7.700013197835555,
      "grad_norm": 0.36763906478881836,
      "learning_rate": 8.505531641070486e-06,
      "loss": 0.0809,
      "num_input_tokens_seen": 26079984,
      "step": 29175
    },
    {
      "epoch": 7.701332981391052,
      "grad_norm": 0.12091916799545288,
      "learning_rate": 8.498155472506885e-06,
      "loss": 0.0215,
      "num_input_tokens_seen": 26084592,
      "step": 29180
    },
    {
      "epoch": 7.702652764946548,
      "grad_norm": 0.3094196021556854,
      "learning_rate": 8.49078184873508e-06,
      "loss": 0.1353,
      "num_input_tokens_seen": 26089168,
      "step": 29185
    },
    {
      "epoch": 7.703972548502046,
      "grad_norm": 0.3009713888168335,
      "learning_rate": 8.483410770892188e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 26093808,
      "step": 29190
    },
    {
      "epoch": 7.705292332057542,
      "grad_norm": 0.28877606987953186,
      "learning_rate": 8.476042240114909e-06,
      "loss": 0.0667,
      "num_input_tokens_seen": 26098544,
      "step": 29195
    },
    {
      "epoch": 7.70661211561304,
      "grad_norm": 0.42427170276641846,
      "learning_rate": 8.468676257539568e-06,
      "loss": 0.091,
      "num_input_tokens_seen": 26103376,
      "step": 29200
    },
    {
      "epoch": 7.70661211561304,
      "eval_loss": 0.06723228842020035,
      "eval_runtime": 64.8353,
      "eval_samples_per_second": 103.879,
      "eval_steps_per_second": 25.974,
      "num_input_tokens_seen": 26103376,
      "step": 29200
    },
    {
      "epoch": 7.707931899168536,
      "grad_norm": 0.15870288014411926,
      "learning_rate": 8.4613128243021e-06,
      "loss": 0.0553,
      "num_input_tokens_seen": 26107696,
      "step": 29205
    },
    {
      "epoch": 7.709251682724033,
      "grad_norm": 0.10355361551046371,
      "learning_rate": 8.453951941538028e-06,
      "loss": 0.0596,
      "num_input_tokens_seen": 26112240,
      "step": 29210
    },
    {
      "epoch": 7.71057146627953,
      "grad_norm": 0.036306701600551605,
      "learning_rate": 8.446593610382495e-06,
      "loss": 0.084,
      "num_input_tokens_seen": 26116752,
      "step": 29215
    },
    {
      "epoch": 7.711891249835027,
      "grad_norm": 0.4973589777946472,
      "learning_rate": 8.439237831970259e-06,
      "loss": 0.0624,
      "num_input_tokens_seen": 26121072,
      "step": 29220
    },
    {
      "epoch": 7.7132110333905235,
      "grad_norm": 0.40243661403656006,
      "learning_rate": 8.431884607435667e-06,
      "loss": 0.0664,
      "num_input_tokens_seen": 26125840,
      "step": 29225
    },
    {
      "epoch": 7.714530816946021,
      "grad_norm": 0.14700132608413696,
      "learning_rate": 8.424533937912665e-06,
      "loss": 0.0527,
      "num_input_tokens_seen": 26130192,
      "step": 29230
    },
    {
      "epoch": 7.7158506005015175,
      "grad_norm": 0.195792555809021,
      "learning_rate": 8.41718582453484e-06,
      "loss": 0.0539,
      "num_input_tokens_seen": 26134416,
      "step": 29235
    },
    {
      "epoch": 7.717170384057015,
      "grad_norm": 0.17996478080749512,
      "learning_rate": 8.409840268435346e-06,
      "loss": 0.0468,
      "num_input_tokens_seen": 26138928,
      "step": 29240
    },
    {
      "epoch": 7.7184901676125115,
      "grad_norm": 0.21712270379066467,
      "learning_rate": 8.402497270746976e-06,
      "loss": 0.0442,
      "num_input_tokens_seen": 26143632,
      "step": 29245
    },
    {
      "epoch": 7.719809951168008,
      "grad_norm": 0.1893381029367447,
      "learning_rate": 8.395156832602095e-06,
      "loss": 0.0629,
      "num_input_tokens_seen": 26148336,
      "step": 29250
    },
    {
      "epoch": 7.7211297347235055,
      "grad_norm": 0.048457589000463486,
      "learning_rate": 8.387818955132707e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 26153104,
      "step": 29255
    },
    {
      "epoch": 7.722449518279002,
      "grad_norm": 0.058434098958969116,
      "learning_rate": 8.38048363947039e-06,
      "loss": 0.1124,
      "num_input_tokens_seen": 26157712,
      "step": 29260
    },
    {
      "epoch": 7.7237693018344995,
      "grad_norm": 0.09891604632139206,
      "learning_rate": 8.373150886746351e-06,
      "loss": 0.0805,
      "num_input_tokens_seen": 26162384,
      "step": 29265
    },
    {
      "epoch": 7.725089085389996,
      "grad_norm": 0.30577352643013,
      "learning_rate": 8.365820698091397e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 26166960,
      "step": 29270
    },
    {
      "epoch": 7.726408868945493,
      "grad_norm": 0.03424977511167526,
      "learning_rate": 8.358493074635922e-06,
      "loss": 0.0665,
      "num_input_tokens_seen": 26171600,
      "step": 29275
    },
    {
      "epoch": 7.72772865250099,
      "grad_norm": 0.04744609072804451,
      "learning_rate": 8.351168017509948e-06,
      "loss": 0.068,
      "num_input_tokens_seen": 26175888,
      "step": 29280
    },
    {
      "epoch": 7.729048436056487,
      "grad_norm": 0.32994920015335083,
      "learning_rate": 8.343845527843094e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 26180432,
      "step": 29285
    },
    {
      "epoch": 7.730368219611984,
      "grad_norm": 0.2943839430809021,
      "learning_rate": 8.336525606764566e-06,
      "loss": 0.065,
      "num_input_tokens_seen": 26184880,
      "step": 29290
    },
    {
      "epoch": 7.731688003167481,
      "grad_norm": 0.29377907514572144,
      "learning_rate": 8.329208255403204e-06,
      "loss": 0.0589,
      "num_input_tokens_seen": 26189392,
      "step": 29295
    },
    {
      "epoch": 7.733007786722977,
      "grad_norm": 0.015558062121272087,
      "learning_rate": 8.321893474887426e-06,
      "loss": 0.0326,
      "num_input_tokens_seen": 26193904,
      "step": 29300
    },
    {
      "epoch": 7.734327570278475,
      "grad_norm": 0.18897658586502075,
      "learning_rate": 8.31458126634526e-06,
      "loss": 0.0692,
      "num_input_tokens_seen": 26198576,
      "step": 29305
    },
    {
      "epoch": 7.735647353833971,
      "grad_norm": 0.14393861591815948,
      "learning_rate": 8.30727163090435e-06,
      "loss": 0.0711,
      "num_input_tokens_seen": 26203152,
      "step": 29310
    },
    {
      "epoch": 7.736967137389469,
      "grad_norm": 0.39964500069618225,
      "learning_rate": 8.29996456969192e-06,
      "loss": 0.0453,
      "num_input_tokens_seen": 26207664,
      "step": 29315
    },
    {
      "epoch": 7.738286920944965,
      "grad_norm": 0.15900760889053345,
      "learning_rate": 8.292660083834818e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 26211888,
      "step": 29320
    },
    {
      "epoch": 7.739606704500462,
      "grad_norm": 0.06413280963897705,
      "learning_rate": 8.2853581744595e-06,
      "loss": 0.063,
      "num_input_tokens_seen": 26216592,
      "step": 29325
    },
    {
      "epoch": 7.740926488055959,
      "grad_norm": 0.12114482372999191,
      "learning_rate": 8.278058842691991e-06,
      "loss": 0.056,
      "num_input_tokens_seen": 26220976,
      "step": 29330
    },
    {
      "epoch": 7.742246271611456,
      "grad_norm": 0.09418950974941254,
      "learning_rate": 8.27076208965796e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 26225488,
      "step": 29335
    },
    {
      "epoch": 7.743566055166952,
      "grad_norm": 0.25682172179222107,
      "learning_rate": 8.263467916482637e-06,
      "loss": 0.0629,
      "num_input_tokens_seen": 26229904,
      "step": 29340
    },
    {
      "epoch": 7.74488583872245,
      "grad_norm": 0.2081858366727829,
      "learning_rate": 8.256176324290885e-06,
      "loss": 0.0875,
      "num_input_tokens_seen": 26234416,
      "step": 29345
    },
    {
      "epoch": 7.746205622277946,
      "grad_norm": 0.03535403311252594,
      "learning_rate": 8.248887314207168e-06,
      "loss": 0.0654,
      "num_input_tokens_seen": 26238992,
      "step": 29350
    },
    {
      "epoch": 7.747525405833443,
      "grad_norm": 0.09094540774822235,
      "learning_rate": 8.24160088735553e-06,
      "loss": 0.0248,
      "num_input_tokens_seen": 26243344,
      "step": 29355
    },
    {
      "epoch": 7.74884518938894,
      "grad_norm": 0.07507632672786713,
      "learning_rate": 8.234317044859629e-06,
      "loss": 0.0734,
      "num_input_tokens_seen": 26247888,
      "step": 29360
    },
    {
      "epoch": 7.750164972944437,
      "grad_norm": 0.08941983431577682,
      "learning_rate": 8.227035787842744e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 26252176,
      "step": 29365
    },
    {
      "epoch": 7.751484756499934,
      "grad_norm": 0.09300301223993301,
      "learning_rate": 8.219757117427721e-06,
      "loss": 0.0334,
      "num_input_tokens_seen": 26256848,
      "step": 29370
    },
    {
      "epoch": 7.752804540055431,
      "grad_norm": 0.30176451802253723,
      "learning_rate": 8.212481034737014e-06,
      "loss": 0.0725,
      "num_input_tokens_seen": 26261456,
      "step": 29375
    },
    {
      "epoch": 7.754124323610927,
      "grad_norm": 0.2823893129825592,
      "learning_rate": 8.205207540892707e-06,
      "loss": 0.0464,
      "num_input_tokens_seen": 26265840,
      "step": 29380
    },
    {
      "epoch": 7.755444107166425,
      "grad_norm": 0.16713373363018036,
      "learning_rate": 8.197936637016442e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 26270256,
      "step": 29385
    },
    {
      "epoch": 7.756763890721921,
      "grad_norm": 0.21369557082653046,
      "learning_rate": 8.190668324229508e-06,
      "loss": 0.0413,
      "num_input_tokens_seen": 26274672,
      "step": 29390
    },
    {
      "epoch": 7.758083674277419,
      "grad_norm": 0.12407999485731125,
      "learning_rate": 8.183402603652749e-06,
      "loss": 0.0726,
      "num_input_tokens_seen": 26279088,
      "step": 29395
    },
    {
      "epoch": 7.759403457832915,
      "grad_norm": 0.06376142054796219,
      "learning_rate": 8.176139476406635e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 26283664,
      "step": 29400
    },
    {
      "epoch": 7.759403457832915,
      "eval_loss": 0.0671762153506279,
      "eval_runtime": 64.7781,
      "eval_samples_per_second": 103.97,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 26283664,
      "step": 29400
    },
    {
      "epoch": 7.760723241388412,
      "grad_norm": 0.18328458070755005,
      "learning_rate": 8.16887894361125e-06,
      "loss": 0.0426,
      "num_input_tokens_seen": 26288464,
      "step": 29405
    },
    {
      "epoch": 7.762043024943909,
      "grad_norm": 0.07232607901096344,
      "learning_rate": 8.161621006386233e-06,
      "loss": 0.1119,
      "num_input_tokens_seen": 26292912,
      "step": 29410
    },
    {
      "epoch": 7.763362808499406,
      "grad_norm": 0.40565112233161926,
      "learning_rate": 8.154365665850869e-06,
      "loss": 0.0982,
      "num_input_tokens_seen": 26297200,
      "step": 29415
    },
    {
      "epoch": 7.764682592054903,
      "grad_norm": 0.014998036436736584,
      "learning_rate": 8.147112923124005e-06,
      "loss": 0.0935,
      "num_input_tokens_seen": 26301648,
      "step": 29420
    },
    {
      "epoch": 7.7660023756104,
      "grad_norm": 0.08331792056560516,
      "learning_rate": 8.13986277932412e-06,
      "loss": 0.0325,
      "num_input_tokens_seen": 26306192,
      "step": 29425
    },
    {
      "epoch": 7.7673221591658965,
      "grad_norm": 0.33068928122520447,
      "learning_rate": 8.132615235569277e-06,
      "loss": 0.0836,
      "num_input_tokens_seen": 26310608,
      "step": 29430
    },
    {
      "epoch": 7.768641942721394,
      "grad_norm": 0.3390275239944458,
      "learning_rate": 8.125370292977124e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 26315248,
      "step": 29435
    },
    {
      "epoch": 7.7699617262768905,
      "grad_norm": 0.05445443466305733,
      "learning_rate": 8.118127952664944e-06,
      "loss": 0.0289,
      "num_input_tokens_seen": 26319888,
      "step": 29440
    },
    {
      "epoch": 7.771281509832388,
      "grad_norm": 0.11204639077186584,
      "learning_rate": 8.110888215749574e-06,
      "loss": 0.0326,
      "num_input_tokens_seen": 26324336,
      "step": 29445
    },
    {
      "epoch": 7.7726012933878845,
      "grad_norm": 0.036024268716573715,
      "learning_rate": 8.10365108334749e-06,
      "loss": 0.0692,
      "num_input_tokens_seen": 26328688,
      "step": 29450
    },
    {
      "epoch": 7.773921076943381,
      "grad_norm": 0.13047192990779877,
      "learning_rate": 8.096416556574743e-06,
      "loss": 0.0234,
      "num_input_tokens_seen": 26333104,
      "step": 29455
    },
    {
      "epoch": 7.7752408604988785,
      "grad_norm": 0.316723495721817,
      "learning_rate": 8.08918463654698e-06,
      "loss": 0.0621,
      "num_input_tokens_seen": 26337776,
      "step": 29460
    },
    {
      "epoch": 7.776560644054375,
      "grad_norm": 0.2604009211063385,
      "learning_rate": 8.081955324379458e-06,
      "loss": 0.0768,
      "num_input_tokens_seen": 26342224,
      "step": 29465
    },
    {
      "epoch": 7.777880427609872,
      "grad_norm": 0.28379932045936584,
      "learning_rate": 8.074728621187039e-06,
      "loss": 0.0914,
      "num_input_tokens_seen": 26346832,
      "step": 29470
    },
    {
      "epoch": 7.779200211165369,
      "grad_norm": 0.0667523443698883,
      "learning_rate": 8.067504528084158e-06,
      "loss": 0.071,
      "num_input_tokens_seen": 26351312,
      "step": 29475
    },
    {
      "epoch": 7.780519994720866,
      "grad_norm": 0.3414291441440582,
      "learning_rate": 8.060283046184861e-06,
      "loss": 0.108,
      "num_input_tokens_seen": 26355920,
      "step": 29480
    },
    {
      "epoch": 7.781839778276362,
      "grad_norm": 0.2726857364177704,
      "learning_rate": 8.053064176602806e-06,
      "loss": 0.1423,
      "num_input_tokens_seen": 26360400,
      "step": 29485
    },
    {
      "epoch": 7.78315956183186,
      "grad_norm": 0.05482906848192215,
      "learning_rate": 8.045847920451216e-06,
      "loss": 0.0547,
      "num_input_tokens_seen": 26364752,
      "step": 29490
    },
    {
      "epoch": 7.784479345387356,
      "grad_norm": 0.2447459101676941,
      "learning_rate": 8.038634278842944e-06,
      "loss": 0.0461,
      "num_input_tokens_seen": 26369552,
      "step": 29495
    },
    {
      "epoch": 7.785799128942854,
      "grad_norm": 0.14525727927684784,
      "learning_rate": 8.031423252890408e-06,
      "loss": 0.0371,
      "num_input_tokens_seen": 26374064,
      "step": 29500
    },
    {
      "epoch": 7.78711891249835,
      "grad_norm": 0.31675413250923157,
      "learning_rate": 8.024214843705646e-06,
      "loss": 0.079,
      "num_input_tokens_seen": 26378288,
      "step": 29505
    },
    {
      "epoch": 7.788438696053847,
      "grad_norm": 0.23156945407390594,
      "learning_rate": 8.017009052400295e-06,
      "loss": 0.0448,
      "num_input_tokens_seen": 26382704,
      "step": 29510
    },
    {
      "epoch": 7.789758479609344,
      "grad_norm": 0.5436953902244568,
      "learning_rate": 8.00980588008557e-06,
      "loss": 0.0912,
      "num_input_tokens_seen": 26387248,
      "step": 29515
    },
    {
      "epoch": 7.791078263164841,
      "grad_norm": 0.13787122070789337,
      "learning_rate": 8.002605327872282e-06,
      "loss": 0.0483,
      "num_input_tokens_seen": 26391696,
      "step": 29520
    },
    {
      "epoch": 7.792398046720338,
      "grad_norm": 0.22874706983566284,
      "learning_rate": 7.995407396870862e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 26396176,
      "step": 29525
    },
    {
      "epoch": 7.793717830275835,
      "grad_norm": 0.11142507940530777,
      "learning_rate": 7.988212088191307e-06,
      "loss": 0.0301,
      "num_input_tokens_seen": 26400624,
      "step": 29530
    },
    {
      "epoch": 7.795037613831331,
      "grad_norm": 0.026693236082792282,
      "learning_rate": 7.98101940294324e-06,
      "loss": 0.0279,
      "num_input_tokens_seen": 26405392,
      "step": 29535
    },
    {
      "epoch": 7.796357397386829,
      "grad_norm": 0.10701094567775726,
      "learning_rate": 7.973829342235847e-06,
      "loss": 0.1023,
      "num_input_tokens_seen": 26410256,
      "step": 29540
    },
    {
      "epoch": 7.797677180942325,
      "grad_norm": 0.21908967196941376,
      "learning_rate": 7.966641907177936e-06,
      "loss": 0.0809,
      "num_input_tokens_seen": 26414832,
      "step": 29545
    },
    {
      "epoch": 7.798996964497823,
      "grad_norm": 0.038357775658369064,
      "learning_rate": 7.959457098877901e-06,
      "loss": 0.0201,
      "num_input_tokens_seen": 26419504,
      "step": 29550
    },
    {
      "epoch": 7.800316748053319,
      "grad_norm": 0.08825361728668213,
      "learning_rate": 7.952274918443719e-06,
      "loss": 0.063,
      "num_input_tokens_seen": 26424144,
      "step": 29555
    },
    {
      "epoch": 7.801636531608816,
      "grad_norm": 0.5710088014602661,
      "learning_rate": 7.945095366982983e-06,
      "loss": 0.1055,
      "num_input_tokens_seen": 26428496,
      "step": 29560
    },
    {
      "epoch": 7.802956315164313,
      "grad_norm": 0.4130508303642273,
      "learning_rate": 7.937918445602871e-06,
      "loss": 0.0877,
      "num_input_tokens_seen": 26433168,
      "step": 29565
    },
    {
      "epoch": 7.80427609871981,
      "grad_norm": 0.06366489082574844,
      "learning_rate": 7.930744155410145e-06,
      "loss": 0.0725,
      "num_input_tokens_seen": 26437360,
      "step": 29570
    },
    {
      "epoch": 7.805595882275307,
      "grad_norm": 0.11841519176959991,
      "learning_rate": 7.923572497511181e-06,
      "loss": 0.0857,
      "num_input_tokens_seen": 26441552,
      "step": 29575
    },
    {
      "epoch": 7.806915665830804,
      "grad_norm": 0.12524597346782684,
      "learning_rate": 7.916403473011927e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 26445712,
      "step": 29580
    },
    {
      "epoch": 7.8082354493863,
      "grad_norm": 0.388108491897583,
      "learning_rate": 7.909237083017953e-06,
      "loss": 0.0525,
      "num_input_tokens_seen": 26449968,
      "step": 29585
    },
    {
      "epoch": 7.809555232941798,
      "grad_norm": 0.21542654931545258,
      "learning_rate": 7.902073328634389e-06,
      "loss": 0.047,
      "num_input_tokens_seen": 26454640,
      "step": 29590
    },
    {
      "epoch": 7.810875016497294,
      "grad_norm": 0.05684393271803856,
      "learning_rate": 7.894912210965987e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 26459280,
      "step": 29595
    },
    {
      "epoch": 7.812194800052791,
      "grad_norm": 0.2622746229171753,
      "learning_rate": 7.887753731117075e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 26463440,
      "step": 29600
    },
    {
      "epoch": 7.812194800052791,
      "eval_loss": 0.06682220846414566,
      "eval_runtime": 64.7479,
      "eval_samples_per_second": 104.019,
      "eval_steps_per_second": 26.009,
      "num_input_tokens_seen": 26463440,
      "step": 29600
    },
    {
      "epoch": 7.813514583608288,
      "grad_norm": 0.44331231713294983,
      "learning_rate": 7.880597890191587e-06,
      "loss": 0.1185,
      "num_input_tokens_seen": 26467888,
      "step": 29605
    },
    {
      "epoch": 7.814834367163785,
      "grad_norm": 0.17849665880203247,
      "learning_rate": 7.873444689293036e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 26472144,
      "step": 29610
    },
    {
      "epoch": 7.8161541507192815,
      "grad_norm": 0.10475289076566696,
      "learning_rate": 7.866294129524548e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 26476592,
      "step": 29615
    },
    {
      "epoch": 7.817473934274779,
      "grad_norm": 0.44389402866363525,
      "learning_rate": 7.859146211988811e-06,
      "loss": 0.0679,
      "num_input_tokens_seen": 26481040,
      "step": 29620
    },
    {
      "epoch": 7.8187937178302755,
      "grad_norm": 0.21584081649780273,
      "learning_rate": 7.852000937788134e-06,
      "loss": 0.0543,
      "num_input_tokens_seen": 26485488,
      "step": 29625
    },
    {
      "epoch": 7.820113501385773,
      "grad_norm": 0.07895476371049881,
      "learning_rate": 7.844858308024416e-06,
      "loss": 0.0636,
      "num_input_tokens_seen": 26490032,
      "step": 29630
    },
    {
      "epoch": 7.8214332849412695,
      "grad_norm": 0.07522094994783401,
      "learning_rate": 7.837718323799122e-06,
      "loss": 0.0523,
      "num_input_tokens_seen": 26494672,
      "step": 29635
    },
    {
      "epoch": 7.822753068496766,
      "grad_norm": 0.12387192249298096,
      "learning_rate": 7.83058098621334e-06,
      "loss": 0.0413,
      "num_input_tokens_seen": 26498832,
      "step": 29640
    },
    {
      "epoch": 7.8240728520522635,
      "grad_norm": 0.22803911566734314,
      "learning_rate": 7.823446296367739e-06,
      "loss": 0.0385,
      "num_input_tokens_seen": 26503632,
      "step": 29645
    },
    {
      "epoch": 7.82539263560776,
      "grad_norm": 0.18696343898773193,
      "learning_rate": 7.81631425536257e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 26507792,
      "step": 29650
    },
    {
      "epoch": 7.8267124191632576,
      "grad_norm": 0.15291476249694824,
      "learning_rate": 7.809184864297689e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 26512304,
      "step": 29655
    },
    {
      "epoch": 7.828032202718754,
      "grad_norm": 0.032771509140729904,
      "learning_rate": 7.802058124272532e-06,
      "loss": 0.053,
      "num_input_tokens_seen": 26516816,
      "step": 29660
    },
    {
      "epoch": 7.829351986274251,
      "grad_norm": 0.24588407576084137,
      "learning_rate": 7.79493403638614e-06,
      "loss": 0.0417,
      "num_input_tokens_seen": 26521328,
      "step": 29665
    },
    {
      "epoch": 7.830671769829748,
      "grad_norm": 0.14163744449615479,
      "learning_rate": 7.787812601737132e-06,
      "loss": 0.0293,
      "num_input_tokens_seen": 26525776,
      "step": 29670
    },
    {
      "epoch": 7.831991553385245,
      "grad_norm": 0.09877733141183853,
      "learning_rate": 7.780693821423715e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 26530224,
      "step": 29675
    },
    {
      "epoch": 7.833311336940742,
      "grad_norm": 0.1099616140127182,
      "learning_rate": 7.773577696543705e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 26534768,
      "step": 29680
    },
    {
      "epoch": 7.834631120496239,
      "grad_norm": 0.16719500720500946,
      "learning_rate": 7.7664642281945e-06,
      "loss": 0.0274,
      "num_input_tokens_seen": 26539024,
      "step": 29685
    },
    {
      "epoch": 7.835950904051735,
      "grad_norm": 0.12275432795286179,
      "learning_rate": 7.759353417473072e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 26543696,
      "step": 29690
    },
    {
      "epoch": 7.837270687607233,
      "grad_norm": 0.29782289266586304,
      "learning_rate": 7.752245265476016e-06,
      "loss": 0.087,
      "num_input_tokens_seen": 26548048,
      "step": 29695
    },
    {
      "epoch": 7.838590471162729,
      "grad_norm": 0.15246348083019257,
      "learning_rate": 7.745139773299481e-06,
      "loss": 0.0337,
      "num_input_tokens_seen": 26552432,
      "step": 29700
    },
    {
      "epoch": 7.839910254718227,
      "grad_norm": 0.12156028300523758,
      "learning_rate": 7.738036942039232e-06,
      "loss": 0.0293,
      "num_input_tokens_seen": 26556752,
      "step": 29705
    },
    {
      "epoch": 7.841230038273723,
      "grad_norm": 0.7386717200279236,
      "learning_rate": 7.73093677279062e-06,
      "loss": 0.0523,
      "num_input_tokens_seen": 26561296,
      "step": 29710
    },
    {
      "epoch": 7.84254982182922,
      "grad_norm": 0.4415915310382843,
      "learning_rate": 7.72383926664857e-06,
      "loss": 0.0847,
      "num_input_tokens_seen": 26565840,
      "step": 29715
    },
    {
      "epoch": 7.843869605384717,
      "grad_norm": 0.18421350419521332,
      "learning_rate": 7.716744424707606e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 26570800,
      "step": 29720
    },
    {
      "epoch": 7.845189388940214,
      "grad_norm": 0.2864592969417572,
      "learning_rate": 7.709652248061858e-06,
      "loss": 0.0367,
      "num_input_tokens_seen": 26575408,
      "step": 29725
    },
    {
      "epoch": 7.84650917249571,
      "grad_norm": 0.04761968180537224,
      "learning_rate": 7.702562737805017e-06,
      "loss": 0.0217,
      "num_input_tokens_seen": 26579568,
      "step": 29730
    },
    {
      "epoch": 7.847828956051208,
      "grad_norm": 0.4514889717102051,
      "learning_rate": 7.695475895030365e-06,
      "loss": 0.1086,
      "num_input_tokens_seen": 26584112,
      "step": 29735
    },
    {
      "epoch": 7.849148739606704,
      "grad_norm": 0.26566755771636963,
      "learning_rate": 7.6883917208308e-06,
      "loss": 0.0692,
      "num_input_tokens_seen": 26588656,
      "step": 29740
    },
    {
      "epoch": 7.850468523162202,
      "grad_norm": 0.13077063858509064,
      "learning_rate": 7.681310216298778e-06,
      "loss": 0.0471,
      "num_input_tokens_seen": 26593072,
      "step": 29745
    },
    {
      "epoch": 7.851788306717698,
      "grad_norm": 0.15217702090740204,
      "learning_rate": 7.674231382526367e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 26597520,
      "step": 29750
    },
    {
      "epoch": 7.853108090273195,
      "grad_norm": 0.09122762829065323,
      "learning_rate": 7.667155220605198e-06,
      "loss": 0.0393,
      "num_input_tokens_seen": 26601808,
      "step": 29755
    },
    {
      "epoch": 7.854427873828692,
      "grad_norm": 0.16524642705917358,
      "learning_rate": 7.660081731626515e-06,
      "loss": 0.0278,
      "num_input_tokens_seen": 26606224,
      "step": 29760
    },
    {
      "epoch": 7.855747657384189,
      "grad_norm": 0.11907808482646942,
      "learning_rate": 7.653010916681141e-06,
      "loss": 0.0299,
      "num_input_tokens_seen": 26610576,
      "step": 29765
    },
    {
      "epoch": 7.8570674409396855,
      "grad_norm": 0.07378603518009186,
      "learning_rate": 7.645942776859472e-06,
      "loss": 0.0253,
      "num_input_tokens_seen": 26615120,
      "step": 29770
    },
    {
      "epoch": 7.858387224495183,
      "grad_norm": 0.29111722111701965,
      "learning_rate": 7.63887731325152e-06,
      "loss": 0.0726,
      "num_input_tokens_seen": 26619856,
      "step": 29775
    },
    {
      "epoch": 7.8597070080506795,
      "grad_norm": 0.4768751263618469,
      "learning_rate": 7.63181452694685e-06,
      "loss": 0.0763,
      "num_input_tokens_seen": 26624240,
      "step": 29780
    },
    {
      "epoch": 7.861026791606177,
      "grad_norm": 0.08021488040685654,
      "learning_rate": 7.624754419034644e-06,
      "loss": 0.0442,
      "num_input_tokens_seen": 26628880,
      "step": 29785
    },
    {
      "epoch": 7.8623465751616735,
      "grad_norm": 0.09418930858373642,
      "learning_rate": 7.6176969906036645e-06,
      "loss": 0.0529,
      "num_input_tokens_seen": 26633328,
      "step": 29790
    },
    {
      "epoch": 7.86366635871717,
      "grad_norm": 0.18523737788200378,
      "learning_rate": 7.610642242742242e-06,
      "loss": 0.03,
      "num_input_tokens_seen": 26638000,
      "step": 29795
    },
    {
      "epoch": 7.8649861422726675,
      "grad_norm": 0.09514988213777542,
      "learning_rate": 7.603590176538322e-06,
      "loss": 0.0428,
      "num_input_tokens_seen": 26642352,
      "step": 29800
    },
    {
      "epoch": 7.8649861422726675,
      "eval_loss": 0.06709372997283936,
      "eval_runtime": 64.8067,
      "eval_samples_per_second": 103.925,
      "eval_steps_per_second": 25.985,
      "num_input_tokens_seen": 26642352,
      "step": 29800
    },
    {
      "epoch": 7.866305925828164,
      "grad_norm": 0.7175394892692566,
      "learning_rate": 7.596540793079404e-06,
      "loss": 0.0701,
      "num_input_tokens_seen": 26646672,
      "step": 29805
    },
    {
      "epoch": 7.8676257093836615,
      "grad_norm": 0.10682795941829681,
      "learning_rate": 7.5894940934526125e-06,
      "loss": 0.0872,
      "num_input_tokens_seen": 26650992,
      "step": 29810
    },
    {
      "epoch": 7.868945492939158,
      "grad_norm": 0.24400553107261658,
      "learning_rate": 7.582450078744621e-06,
      "loss": 0.066,
      "num_input_tokens_seen": 26655280,
      "step": 29815
    },
    {
      "epoch": 7.870265276494655,
      "grad_norm": 0.15113161504268646,
      "learning_rate": 7.575408750041707e-06,
      "loss": 0.0313,
      "num_input_tokens_seen": 26659888,
      "step": 29820
    },
    {
      "epoch": 7.871585060050152,
      "grad_norm": 0.22402040660381317,
      "learning_rate": 7.568370108429732e-06,
      "loss": 0.0673,
      "num_input_tokens_seen": 26664592,
      "step": 29825
    },
    {
      "epoch": 7.872904843605649,
      "grad_norm": 0.24285975098609924,
      "learning_rate": 7.561334154994154e-06,
      "loss": 0.0553,
      "num_input_tokens_seen": 26669296,
      "step": 29830
    },
    {
      "epoch": 7.874224627161146,
      "grad_norm": 0.26852935552597046,
      "learning_rate": 7.55430089081999e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 26673584,
      "step": 29835
    },
    {
      "epoch": 7.875544410716643,
      "grad_norm": 0.18975195288658142,
      "learning_rate": 7.547270316991864e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 26677904,
      "step": 29840
    },
    {
      "epoch": 7.876864194272139,
      "grad_norm": 0.22726944088935852,
      "learning_rate": 7.5402424345939884e-06,
      "loss": 0.0543,
      "num_input_tokens_seen": 26682384,
      "step": 29845
    },
    {
      "epoch": 7.878183977827637,
      "grad_norm": 0.47393304109573364,
      "learning_rate": 7.533217244710133e-06,
      "loss": 0.0865,
      "num_input_tokens_seen": 26686672,
      "step": 29850
    },
    {
      "epoch": 7.879503761383133,
      "grad_norm": 0.11657862365245819,
      "learning_rate": 7.52619474842369e-06,
      "loss": 0.0682,
      "num_input_tokens_seen": 26690992,
      "step": 29855
    },
    {
      "epoch": 7.88082354493863,
      "grad_norm": 0.14319394528865814,
      "learning_rate": 7.519174946817597e-06,
      "loss": 0.068,
      "num_input_tokens_seen": 26695216,
      "step": 29860
    },
    {
      "epoch": 7.882143328494127,
      "grad_norm": 0.20124183595180511,
      "learning_rate": 7.512157840974407e-06,
      "loss": 0.0348,
      "num_input_tokens_seen": 26699664,
      "step": 29865
    },
    {
      "epoch": 7.883463112049624,
      "grad_norm": 0.2509338855743408,
      "learning_rate": 7.5051434319762496e-06,
      "loss": 0.0684,
      "num_input_tokens_seen": 26703920,
      "step": 29870
    },
    {
      "epoch": 7.884782895605121,
      "grad_norm": 0.10358088463544846,
      "learning_rate": 7.498131720904822e-06,
      "loss": 0.072,
      "num_input_tokens_seen": 26708464,
      "step": 29875
    },
    {
      "epoch": 7.886102679160618,
      "grad_norm": 0.4617219567298889,
      "learning_rate": 7.491122708841433e-06,
      "loss": 0.0712,
      "num_input_tokens_seen": 26712816,
      "step": 29880
    },
    {
      "epoch": 7.887422462716114,
      "grad_norm": 0.13039080798625946,
      "learning_rate": 7.4841163968669524e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 26717552,
      "step": 29885
    },
    {
      "epoch": 7.888742246271612,
      "grad_norm": 0.3037091791629791,
      "learning_rate": 7.4771127860618355e-06,
      "loss": 0.0554,
      "num_input_tokens_seen": 26722224,
      "step": 29890
    },
    {
      "epoch": 7.890062029827108,
      "grad_norm": 0.14825887978076935,
      "learning_rate": 7.470111877506139e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 26726992,
      "step": 29895
    },
    {
      "epoch": 7.891381813382605,
      "grad_norm": 0.04369297996163368,
      "learning_rate": 7.463113672279479e-06,
      "loss": 0.0543,
      "num_input_tokens_seen": 26731728,
      "step": 29900
    },
    {
      "epoch": 7.892701596938102,
      "grad_norm": 0.08232252299785614,
      "learning_rate": 7.456118171461071e-06,
      "loss": 0.0988,
      "num_input_tokens_seen": 26736304,
      "step": 29905
    },
    {
      "epoch": 7.894021380493599,
      "grad_norm": 0.276728093624115,
      "learning_rate": 7.449125376129721e-06,
      "loss": 0.0684,
      "num_input_tokens_seen": 26740720,
      "step": 29910
    },
    {
      "epoch": 7.895341164049096,
      "grad_norm": 0.3015488386154175,
      "learning_rate": 7.442135287363788e-06,
      "loss": 0.0565,
      "num_input_tokens_seen": 26745008,
      "step": 29915
    },
    {
      "epoch": 7.896660947604593,
      "grad_norm": 0.15158341825008392,
      "learning_rate": 7.435147906241247e-06,
      "loss": 0.0425,
      "num_input_tokens_seen": 26749296,
      "step": 29920
    },
    {
      "epoch": 7.897980731160089,
      "grad_norm": 0.3544863164424896,
      "learning_rate": 7.428163233839624e-06,
      "loss": 0.1373,
      "num_input_tokens_seen": 26754064,
      "step": 29925
    },
    {
      "epoch": 7.899300514715587,
      "grad_norm": 0.25933584570884705,
      "learning_rate": 7.4211812712360525e-06,
      "loss": 0.0239,
      "num_input_tokens_seen": 26758608,
      "step": 29930
    },
    {
      "epoch": 7.900620298271083,
      "grad_norm": 0.411872535943985,
      "learning_rate": 7.4142020195072464e-06,
      "loss": 0.043,
      "num_input_tokens_seen": 26763120,
      "step": 29935
    },
    {
      "epoch": 7.901940081826581,
      "grad_norm": 0.10601312667131424,
      "learning_rate": 7.407225479729479e-06,
      "loss": 0.0519,
      "num_input_tokens_seen": 26767888,
      "step": 29940
    },
    {
      "epoch": 7.903259865382077,
      "grad_norm": 0.19563254714012146,
      "learning_rate": 7.400251652978632e-06,
      "loss": 0.0444,
      "num_input_tokens_seen": 26772560,
      "step": 29945
    },
    {
      "epoch": 7.904579648937574,
      "grad_norm": 0.10898806899785995,
      "learning_rate": 7.393280540330147e-06,
      "loss": 0.0418,
      "num_input_tokens_seen": 26777424,
      "step": 29950
    },
    {
      "epoch": 7.905899432493071,
      "grad_norm": 0.056096311658620834,
      "learning_rate": 7.386312142859069e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 26781616,
      "step": 29955
    },
    {
      "epoch": 7.907219216048568,
      "grad_norm": 0.10543117672204971,
      "learning_rate": 7.379346461640008e-06,
      "loss": 0.09,
      "num_input_tokens_seen": 26786320,
      "step": 29960
    },
    {
      "epoch": 7.908538999604065,
      "grad_norm": 0.22770382463932037,
      "learning_rate": 7.372383497747149e-06,
      "loss": 0.0291,
      "num_input_tokens_seen": 26790960,
      "step": 29965
    },
    {
      "epoch": 7.909858783159562,
      "grad_norm": 0.11658193171024323,
      "learning_rate": 7.3654232522542775e-06,
      "loss": 0.0374,
      "num_input_tokens_seen": 26795248,
      "step": 29970
    },
    {
      "epoch": 7.9111785667150585,
      "grad_norm": 0.08234360814094543,
      "learning_rate": 7.358465726234756e-06,
      "loss": 0.059,
      "num_input_tokens_seen": 26799792,
      "step": 29975
    },
    {
      "epoch": 7.912498350270556,
      "grad_norm": 0.24109932780265808,
      "learning_rate": 7.351510920761512e-06,
      "loss": 0.0573,
      "num_input_tokens_seen": 26804208,
      "step": 29980
    },
    {
      "epoch": 7.9138181338260525,
      "grad_norm": 0.046731509268283844,
      "learning_rate": 7.344558836907067e-06,
      "loss": 0.0363,
      "num_input_tokens_seen": 26808752,
      "step": 29985
    },
    {
      "epoch": 7.915137917381549,
      "grad_norm": 0.06505048274993896,
      "learning_rate": 7.3376094757435285e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 26813264,
      "step": 29990
    },
    {
      "epoch": 7.9164577009370465,
      "grad_norm": 0.22991526126861572,
      "learning_rate": 7.330662838342561e-06,
      "loss": 0.0332,
      "num_input_tokens_seen": 26817712,
      "step": 29995
    },
    {
      "epoch": 7.917777484492543,
      "grad_norm": 0.06408590078353882,
      "learning_rate": 7.323718925775438e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 26822096,
      "step": 30000
    },
    {
      "epoch": 7.917777484492543,
      "eval_loss": 0.06719440966844559,
      "eval_runtime": 64.8014,
      "eval_samples_per_second": 103.933,
      "eval_steps_per_second": 25.987,
      "num_input_tokens_seen": 26822096,
      "step": 30000
    },
    {
      "epoch": 7.9190972680480405,
      "grad_norm": 0.10281378030776978,
      "learning_rate": 7.316777739112985e-06,
      "loss": 0.033,
      "num_input_tokens_seen": 26826480,
      "step": 30005
    },
    {
      "epoch": 7.920417051603537,
      "grad_norm": 0.5092872381210327,
      "learning_rate": 7.309839279425626e-06,
      "loss": 0.0853,
      "num_input_tokens_seen": 26831024,
      "step": 30010
    },
    {
      "epoch": 7.921736835159034,
      "grad_norm": 0.22064289450645447,
      "learning_rate": 7.302903547783366e-06,
      "loss": 0.0331,
      "num_input_tokens_seen": 26835312,
      "step": 30015
    },
    {
      "epoch": 7.923056618714531,
      "grad_norm": 0.3362635374069214,
      "learning_rate": 7.2959705452557644e-06,
      "loss": 0.0807,
      "num_input_tokens_seen": 26839632,
      "step": 30020
    },
    {
      "epoch": 7.924376402270028,
      "grad_norm": 0.12530982494354248,
      "learning_rate": 7.289040272911996e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 26844240,
      "step": 30025
    },
    {
      "epoch": 7.925696185825524,
      "grad_norm": 0.04797416180372238,
      "learning_rate": 7.282112731820789e-06,
      "loss": 0.099,
      "num_input_tokens_seen": 26848656,
      "step": 30030
    },
    {
      "epoch": 7.927015969381022,
      "grad_norm": 0.0678882822394371,
      "learning_rate": 7.275187923050447e-06,
      "loss": 0.0185,
      "num_input_tokens_seen": 26853168,
      "step": 30035
    },
    {
      "epoch": 7.928335752936518,
      "grad_norm": 0.029461923986673355,
      "learning_rate": 7.268265847668879e-06,
      "loss": 0.041,
      "num_input_tokens_seen": 26857808,
      "step": 30040
    },
    {
      "epoch": 7.929655536492016,
      "grad_norm": 0.03337513282895088,
      "learning_rate": 7.261346506743538e-06,
      "loss": 0.0455,
      "num_input_tokens_seen": 26861968,
      "step": 30045
    },
    {
      "epoch": 7.930975320047512,
      "grad_norm": 0.027279408648610115,
      "learning_rate": 7.254429901341486e-06,
      "loss": 0.0507,
      "num_input_tokens_seen": 26866128,
      "step": 30050
    },
    {
      "epoch": 7.932295103603009,
      "grad_norm": 0.21016177535057068,
      "learning_rate": 7.247516032529356e-06,
      "loss": 0.0854,
      "num_input_tokens_seen": 26870736,
      "step": 30055
    },
    {
      "epoch": 7.933614887158506,
      "grad_norm": 0.031663864850997925,
      "learning_rate": 7.240604901373338e-06,
      "loss": 0.0456,
      "num_input_tokens_seen": 26875408,
      "step": 30060
    },
    {
      "epoch": 7.934934670714003,
      "grad_norm": 0.040149152278900146,
      "learning_rate": 7.233696508939223e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 26879664,
      "step": 30065
    },
    {
      "epoch": 7.9362544542695,
      "grad_norm": 0.5209730267524719,
      "learning_rate": 7.226790856292376e-06,
      "loss": 0.1905,
      "num_input_tokens_seen": 26883728,
      "step": 30070
    },
    {
      "epoch": 7.937574237824997,
      "grad_norm": 0.028535690158605576,
      "learning_rate": 7.219887944497727e-06,
      "loss": 0.0278,
      "num_input_tokens_seen": 26888208,
      "step": 30075
    },
    {
      "epoch": 7.938894021380493,
      "grad_norm": 0.09450594335794449,
      "learning_rate": 7.2129877746198e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 26892560,
      "step": 30080
    },
    {
      "epoch": 7.940213804935991,
      "grad_norm": 0.03408260643482208,
      "learning_rate": 7.20609034772268e-06,
      "loss": 0.0684,
      "num_input_tokens_seen": 26897040,
      "step": 30085
    },
    {
      "epoch": 7.941533588491487,
      "grad_norm": 0.15078240633010864,
      "learning_rate": 7.19919566487004e-06,
      "loss": 0.0274,
      "num_input_tokens_seen": 26901712,
      "step": 30090
    },
    {
      "epoch": 7.942853372046985,
      "grad_norm": 1.5123096704483032,
      "learning_rate": 7.192303727125132e-06,
      "loss": 0.0534,
      "num_input_tokens_seen": 26906192,
      "step": 30095
    },
    {
      "epoch": 7.944173155602481,
      "grad_norm": 0.020771503448486328,
      "learning_rate": 7.185414535550777e-06,
      "loss": 0.0279,
      "num_input_tokens_seen": 26910832,
      "step": 30100
    },
    {
      "epoch": 7.945492939157978,
      "grad_norm": 0.37324821949005127,
      "learning_rate": 7.178528091209363e-06,
      "loss": 0.065,
      "num_input_tokens_seen": 26915184,
      "step": 30105
    },
    {
      "epoch": 7.946812722713475,
      "grad_norm": 0.08117761462926865,
      "learning_rate": 7.171644395162888e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 26919792,
      "step": 30110
    },
    {
      "epoch": 7.948132506268972,
      "grad_norm": 0.12705957889556885,
      "learning_rate": 7.164763448472881e-06,
      "loss": 0.0534,
      "num_input_tokens_seen": 26924304,
      "step": 30115
    },
    {
      "epoch": 7.949452289824469,
      "grad_norm": 0.12614706158638,
      "learning_rate": 7.157885252200491e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 26928784,
      "step": 30120
    },
    {
      "epoch": 7.950772073379966,
      "grad_norm": 0.06763449311256409,
      "learning_rate": 7.151009807406403e-06,
      "loss": 0.039,
      "num_input_tokens_seen": 26933328,
      "step": 30125
    },
    {
      "epoch": 7.952091856935462,
      "grad_norm": 0.20126107335090637,
      "learning_rate": 7.144137115150909e-06,
      "loss": 0.1024,
      "num_input_tokens_seen": 26938224,
      "step": 30130
    },
    {
      "epoch": 7.95341164049096,
      "grad_norm": 0.08997827023267746,
      "learning_rate": 7.1372671764938725e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 26942416,
      "step": 30135
    },
    {
      "epoch": 7.954731424046456,
      "grad_norm": 0.28838831186294556,
      "learning_rate": 7.130399992494705e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 26946960,
      "step": 30140
    },
    {
      "epoch": 7.956051207601953,
      "grad_norm": 0.07410097122192383,
      "learning_rate": 7.123535564212419e-06,
      "loss": 0.0408,
      "num_input_tokens_seen": 26951536,
      "step": 30145
    },
    {
      "epoch": 7.95737099115745,
      "grad_norm": 0.2847036123275757,
      "learning_rate": 7.116673892705611e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 26955888,
      "step": 30150
    },
    {
      "epoch": 7.958690774712947,
      "grad_norm": 0.08863674104213715,
      "learning_rate": 7.109814979032415e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 26960240,
      "step": 30155
    },
    {
      "epoch": 7.9600105582684435,
      "grad_norm": 0.2323164939880371,
      "learning_rate": 7.102958824250577e-06,
      "loss": 0.0421,
      "num_input_tokens_seen": 26964784,
      "step": 30160
    },
    {
      "epoch": 7.961330341823941,
      "grad_norm": 0.1313123106956482,
      "learning_rate": 7.096105429417393e-06,
      "loss": 0.0352,
      "num_input_tokens_seen": 26969072,
      "step": 30165
    },
    {
      "epoch": 7.9626501253794375,
      "grad_norm": 0.2643413841724396,
      "learning_rate": 7.0892547955897506e-06,
      "loss": 0.055,
      "num_input_tokens_seen": 26973488,
      "step": 30170
    },
    {
      "epoch": 7.963969908934935,
      "grad_norm": 0.14999638497829437,
      "learning_rate": 7.0824069238241e-06,
      "loss": 0.0498,
      "num_input_tokens_seen": 26977968,
      "step": 30175
    },
    {
      "epoch": 7.9652896924904315,
      "grad_norm": 0.04462284967303276,
      "learning_rate": 7.075561815176462e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 26982544,
      "step": 30180
    },
    {
      "epoch": 7.966609476045928,
      "grad_norm": 0.09972035884857178,
      "learning_rate": 7.068719470702445e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 26986864,
      "step": 30185
    },
    {
      "epoch": 7.9679292596014255,
      "grad_norm": 0.31688374280929565,
      "learning_rate": 7.061879891457229e-06,
      "loss": 0.0909,
      "num_input_tokens_seen": 26991504,
      "step": 30190
    },
    {
      "epoch": 7.969249043156922,
      "grad_norm": 0.28414109349250793,
      "learning_rate": 7.0550430784955515e-06,
      "loss": 0.0697,
      "num_input_tokens_seen": 26996112,
      "step": 30195
    },
    {
      "epoch": 7.9705688267124195,
      "grad_norm": 0.19878104329109192,
      "learning_rate": 7.048209032871752e-06,
      "loss": 0.0417,
      "num_input_tokens_seen": 27000688,
      "step": 30200
    },
    {
      "epoch": 7.9705688267124195,
      "eval_loss": 0.06734690070152283,
      "eval_runtime": 64.7791,
      "eval_samples_per_second": 103.969,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 27000688,
      "step": 30200
    },
    {
      "epoch": 7.971888610267916,
      "grad_norm": 0.16194979846477509,
      "learning_rate": 7.0413777556397055e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 27005488,
      "step": 30205
    },
    {
      "epoch": 7.973208393823413,
      "grad_norm": 0.12212682515382767,
      "learning_rate": 7.0345492478528925e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 27009872,
      "step": 30210
    },
    {
      "epoch": 7.97452817737891,
      "grad_norm": 0.33504918217658997,
      "learning_rate": 7.02772351056436e-06,
      "loss": 0.1048,
      "num_input_tokens_seen": 27014480,
      "step": 30215
    },
    {
      "epoch": 7.975847960934407,
      "grad_norm": 0.021263884380459785,
      "learning_rate": 7.020900544826709e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 27018640,
      "step": 30220
    },
    {
      "epoch": 7.977167744489904,
      "grad_norm": 0.09829231351613998,
      "learning_rate": 7.014080351692134e-06,
      "loss": 0.088,
      "num_input_tokens_seen": 27023088,
      "step": 30225
    },
    {
      "epoch": 7.978487528045401,
      "grad_norm": 0.23319900035858154,
      "learning_rate": 7.0072629322124024e-06,
      "loss": 0.1154,
      "num_input_tokens_seen": 27027504,
      "step": 30230
    },
    {
      "epoch": 7.979807311600897,
      "grad_norm": 0.6480019092559814,
      "learning_rate": 7.000448287438827e-06,
      "loss": 0.1265,
      "num_input_tokens_seen": 27031664,
      "step": 30235
    },
    {
      "epoch": 7.981127095156395,
      "grad_norm": 0.3714023530483246,
      "learning_rate": 6.993636418422331e-06,
      "loss": 0.0668,
      "num_input_tokens_seen": 27035792,
      "step": 30240
    },
    {
      "epoch": 7.982446878711891,
      "grad_norm": 0.2799889147281647,
      "learning_rate": 6.986827326213383e-06,
      "loss": 0.0461,
      "num_input_tokens_seen": 27040368,
      "step": 30245
    },
    {
      "epoch": 7.983766662267389,
      "grad_norm": 0.34205272793769836,
      "learning_rate": 6.9800210118620205e-06,
      "loss": 0.1031,
      "num_input_tokens_seen": 27044720,
      "step": 30250
    },
    {
      "epoch": 7.985086445822885,
      "grad_norm": 0.06961945444345474,
      "learning_rate": 6.973217476417876e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 27049232,
      "step": 30255
    },
    {
      "epoch": 7.986406229378382,
      "grad_norm": 0.18115903437137604,
      "learning_rate": 6.96641672093013e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 27053872,
      "step": 30260
    },
    {
      "epoch": 7.987726012933879,
      "grad_norm": 0.18713590502738953,
      "learning_rate": 6.95961874644755e-06,
      "loss": 0.062,
      "num_input_tokens_seen": 27058576,
      "step": 30265
    },
    {
      "epoch": 7.989045796489376,
      "grad_norm": 0.14399154484272003,
      "learning_rate": 6.952823554018476e-06,
      "loss": 0.0824,
      "num_input_tokens_seen": 27062928,
      "step": 30270
    },
    {
      "epoch": 7.990365580044872,
      "grad_norm": 0.05791733041405678,
      "learning_rate": 6.946031144690798e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 27067376,
      "step": 30275
    },
    {
      "epoch": 7.99168536360037,
      "grad_norm": 0.27518972754478455,
      "learning_rate": 6.939241519512005e-06,
      "loss": 0.0465,
      "num_input_tokens_seen": 27071696,
      "step": 30280
    },
    {
      "epoch": 7.993005147155866,
      "grad_norm": 0.06753062456846237,
      "learning_rate": 6.932454679529129e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 27076080,
      "step": 30285
    },
    {
      "epoch": 7.994324930711363,
      "grad_norm": 0.1605243682861328,
      "learning_rate": 6.925670625788791e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 27080624,
      "step": 30290
    },
    {
      "epoch": 7.99564471426686,
      "grad_norm": 0.06993427872657776,
      "learning_rate": 6.918889359337186e-06,
      "loss": 0.0926,
      "num_input_tokens_seen": 27085104,
      "step": 30295
    },
    {
      "epoch": 7.996964497822357,
      "grad_norm": 0.11666660010814667,
      "learning_rate": 6.912110881220058e-06,
      "loss": 0.0557,
      "num_input_tokens_seen": 27089456,
      "step": 30300
    },
    {
      "epoch": 7.998284281377854,
      "grad_norm": 0.25693169236183167,
      "learning_rate": 6.905335192482735e-06,
      "loss": 0.0432,
      "num_input_tokens_seen": 27094224,
      "step": 30305
    },
    {
      "epoch": 7.999604064933351,
      "grad_norm": 0.46004870533943176,
      "learning_rate": 6.8985622941701275e-06,
      "loss": 0.1014,
      "num_input_tokens_seen": 27098448,
      "step": 30310
    },
    {
      "epoch": 8.000791870133298,
      "grad_norm": 0.6272907257080078,
      "learning_rate": 6.89179218732669e-06,
      "loss": 0.1432,
      "num_input_tokens_seen": 27102112,
      "step": 30315
    },
    {
      "epoch": 8.002111653688795,
      "grad_norm": 0.051905639469623566,
      "learning_rate": 6.8850248729964595e-06,
      "loss": 0.014,
      "num_input_tokens_seen": 27106624,
      "step": 30320
    },
    {
      "epoch": 8.003431437244291,
      "grad_norm": 0.4335690438747406,
      "learning_rate": 6.8782603522230314e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 27111264,
      "step": 30325
    },
    {
      "epoch": 8.004751220799788,
      "grad_norm": 0.15811732411384583,
      "learning_rate": 6.871498626049591e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 27115872,
      "step": 30330
    },
    {
      "epoch": 8.006071004355286,
      "grad_norm": 0.21418800950050354,
      "learning_rate": 6.8647396955188875e-06,
      "loss": 0.093,
      "num_input_tokens_seen": 27120096,
      "step": 30335
    },
    {
      "epoch": 8.007390787910783,
      "grad_norm": 0.0878409668803215,
      "learning_rate": 6.857983561673218e-06,
      "loss": 0.0507,
      "num_input_tokens_seen": 27124736,
      "step": 30340
    },
    {
      "epoch": 8.00871057146628,
      "grad_norm": 0.1614827662706375,
      "learning_rate": 6.851230225554467e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 27129024,
      "step": 30345
    },
    {
      "epoch": 8.010030355021776,
      "grad_norm": 0.4831705689430237,
      "learning_rate": 6.8444796882040946e-06,
      "loss": 0.0712,
      "num_input_tokens_seen": 27133344,
      "step": 30350
    },
    {
      "epoch": 8.011350138577273,
      "grad_norm": 0.2862592041492462,
      "learning_rate": 6.837731950663106e-06,
      "loss": 0.0678,
      "num_input_tokens_seen": 27137824,
      "step": 30355
    },
    {
      "epoch": 8.01266992213277,
      "grad_norm": 0.18406331539154053,
      "learning_rate": 6.830987013972098e-06,
      "loss": 0.0216,
      "num_input_tokens_seen": 27142624,
      "step": 30360
    },
    {
      "epoch": 8.013989705688267,
      "grad_norm": 0.07311908155679703,
      "learning_rate": 6.82424487917121e-06,
      "loss": 0.0344,
      "num_input_tokens_seen": 27146848,
      "step": 30365
    },
    {
      "epoch": 8.015309489243764,
      "grad_norm": 0.1424454152584076,
      "learning_rate": 6.8175055473001735e-06,
      "loss": 0.0353,
      "num_input_tokens_seen": 27151200,
      "step": 30370
    },
    {
      "epoch": 8.01662927279926,
      "grad_norm": 0.34248077869415283,
      "learning_rate": 6.8107690193982855e-06,
      "loss": 0.0618,
      "num_input_tokens_seen": 27155712,
      "step": 30375
    },
    {
      "epoch": 8.017949056354757,
      "grad_norm": 0.26333096623420715,
      "learning_rate": 6.804035296504385e-06,
      "loss": 0.1031,
      "num_input_tokens_seen": 27160512,
      "step": 30380
    },
    {
      "epoch": 8.019268839910255,
      "grad_norm": 0.14900581538677216,
      "learning_rate": 6.797304379656916e-06,
      "loss": 0.0403,
      "num_input_tokens_seen": 27164672,
      "step": 30385
    },
    {
      "epoch": 8.020588623465752,
      "grad_norm": 0.0962800458073616,
      "learning_rate": 6.790576269893861e-06,
      "loss": 0.1023,
      "num_input_tokens_seen": 27169152,
      "step": 30390
    },
    {
      "epoch": 8.021908407021249,
      "grad_norm": 0.07552297413349152,
      "learning_rate": 6.783850968252772e-06,
      "loss": 0.1338,
      "num_input_tokens_seen": 27173696,
      "step": 30395
    },
    {
      "epoch": 8.023228190576745,
      "grad_norm": 0.48807820677757263,
      "learning_rate": 6.777128475770789e-06,
      "loss": 0.0414,
      "num_input_tokens_seen": 27178304,
      "step": 30400
    },
    {
      "epoch": 8.023228190576745,
      "eval_loss": 0.06699827313423157,
      "eval_runtime": 64.7936,
      "eval_samples_per_second": 103.945,
      "eval_steps_per_second": 25.99,
      "num_input_tokens_seen": 27178304,
      "step": 30400
    },
    {
      "epoch": 8.024547974132242,
      "grad_norm": 0.2678922712802887,
      "learning_rate": 6.77040879348459e-06,
      "loss": 0.084,
      "num_input_tokens_seen": 27182560,
      "step": 30405
    },
    {
      "epoch": 8.02586775768774,
      "grad_norm": 0.12824031710624695,
      "learning_rate": 6.763691922430443e-06,
      "loss": 0.0878,
      "num_input_tokens_seen": 27187008,
      "step": 30410
    },
    {
      "epoch": 8.027187541243237,
      "grad_norm": 0.16855241358280182,
      "learning_rate": 6.756977863644178e-06,
      "loss": 0.0682,
      "num_input_tokens_seen": 27191808,
      "step": 30415
    },
    {
      "epoch": 8.028507324798733,
      "grad_norm": 0.21065367758274078,
      "learning_rate": 6.7502666181611804e-06,
      "loss": 0.1042,
      "num_input_tokens_seen": 27196256,
      "step": 30420
    },
    {
      "epoch": 8.02982710835423,
      "grad_norm": 0.19942709803581238,
      "learning_rate": 6.743558187016405e-06,
      "loss": 0.0439,
      "num_input_tokens_seen": 27200736,
      "step": 30425
    },
    {
      "epoch": 8.031146891909726,
      "grad_norm": 0.2652882933616638,
      "learning_rate": 6.7368525712443925e-06,
      "loss": 0.0616,
      "num_input_tokens_seen": 27204928,
      "step": 30430
    },
    {
      "epoch": 8.032466675465225,
      "grad_norm": 0.16896645724773407,
      "learning_rate": 6.7301497718792155e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 27209472,
      "step": 30435
    },
    {
      "epoch": 8.033786459020721,
      "grad_norm": 0.03256889432668686,
      "learning_rate": 6.723449789954544e-06,
      "loss": 0.0326,
      "num_input_tokens_seen": 27213952,
      "step": 30440
    },
    {
      "epoch": 8.035106242576218,
      "grad_norm": 0.11549127101898193,
      "learning_rate": 6.716752626503586e-06,
      "loss": 0.0672,
      "num_input_tokens_seen": 27218528,
      "step": 30445
    },
    {
      "epoch": 8.036426026131714,
      "grad_norm": 0.04922103136777878,
      "learning_rate": 6.710058282559131e-06,
      "loss": 0.0428,
      "num_input_tokens_seen": 27223072,
      "step": 30450
    },
    {
      "epoch": 8.03774580968721,
      "grad_norm": 0.052635084837675095,
      "learning_rate": 6.703366759153545e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 27227520,
      "step": 30455
    },
    {
      "epoch": 8.039065593242707,
      "grad_norm": 0.24790987372398376,
      "learning_rate": 6.6966780573187335e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 27231712,
      "step": 30460
    },
    {
      "epoch": 8.040385376798206,
      "grad_norm": 0.10166924446821213,
      "learning_rate": 6.689992178086174e-06,
      "loss": 0.0208,
      "num_input_tokens_seen": 27236224,
      "step": 30465
    },
    {
      "epoch": 8.041705160353702,
      "grad_norm": 0.35070276260375977,
      "learning_rate": 6.683309122486925e-06,
      "loss": 0.06,
      "num_input_tokens_seen": 27240640,
      "step": 30470
    },
    {
      "epoch": 8.043024943909199,
      "grad_norm": 0.38466334342956543,
      "learning_rate": 6.676628891551584e-06,
      "loss": 0.0814,
      "num_input_tokens_seen": 27244800,
      "step": 30475
    },
    {
      "epoch": 8.044344727464695,
      "grad_norm": 0.384154349565506,
      "learning_rate": 6.6699514863103385e-06,
      "loss": 0.1041,
      "num_input_tokens_seen": 27249344,
      "step": 30480
    },
    {
      "epoch": 8.045664511020192,
      "grad_norm": 0.15531635284423828,
      "learning_rate": 6.663276907792921e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 27253920,
      "step": 30485
    },
    {
      "epoch": 8.04698429457569,
      "grad_norm": 0.2224864959716797,
      "learning_rate": 6.656605157028634e-06,
      "loss": 0.0272,
      "num_input_tokens_seen": 27258368,
      "step": 30490
    },
    {
      "epoch": 8.048304078131187,
      "grad_norm": 0.19829757511615753,
      "learning_rate": 6.649936235046358e-06,
      "loss": 0.0604,
      "num_input_tokens_seen": 27262656,
      "step": 30495
    },
    {
      "epoch": 8.049623861686683,
      "grad_norm": 0.21809379756450653,
      "learning_rate": 6.643270142874508e-06,
      "loss": 0.0455,
      "num_input_tokens_seen": 27267040,
      "step": 30500
    },
    {
      "epoch": 8.05094364524218,
      "grad_norm": 0.1967085599899292,
      "learning_rate": 6.636606881541094e-06,
      "loss": 0.055,
      "num_input_tokens_seen": 27271520,
      "step": 30505
    },
    {
      "epoch": 8.052263428797676,
      "grad_norm": 0.2552953064441681,
      "learning_rate": 6.629946452073662e-06,
      "loss": 0.0781,
      "num_input_tokens_seen": 27276384,
      "step": 30510
    },
    {
      "epoch": 8.053583212353175,
      "grad_norm": 0.3122285008430481,
      "learning_rate": 6.6232888554993375e-06,
      "loss": 0.0699,
      "num_input_tokens_seen": 27280992,
      "step": 30515
    },
    {
      "epoch": 8.054902995908671,
      "grad_norm": 0.29305362701416016,
      "learning_rate": 6.616634092844817e-06,
      "loss": 0.0948,
      "num_input_tokens_seen": 27285568,
      "step": 30520
    },
    {
      "epoch": 8.056222779464168,
      "grad_norm": 0.32683810591697693,
      "learning_rate": 6.609982165136331e-06,
      "loss": 0.0489,
      "num_input_tokens_seen": 27290272,
      "step": 30525
    },
    {
      "epoch": 8.057542563019664,
      "grad_norm": 0.16808360815048218,
      "learning_rate": 6.603333073399706e-06,
      "loss": 0.0822,
      "num_input_tokens_seen": 27294432,
      "step": 30530
    },
    {
      "epoch": 8.058862346575161,
      "grad_norm": 0.2741433084011078,
      "learning_rate": 6.596686818660308e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 27298944,
      "step": 30535
    },
    {
      "epoch": 8.06018213013066,
      "grad_norm": 0.10603894293308258,
      "learning_rate": 6.590043401943066e-06,
      "loss": 0.0666,
      "num_input_tokens_seen": 27303104,
      "step": 30540
    },
    {
      "epoch": 8.061501913686156,
      "grad_norm": 0.4956870377063751,
      "learning_rate": 6.583402824272494e-06,
      "loss": 0.0773,
      "num_input_tokens_seen": 27307392,
      "step": 30545
    },
    {
      "epoch": 8.062821697241652,
      "grad_norm": 0.12128914147615433,
      "learning_rate": 6.576765086672634e-06,
      "loss": 0.0255,
      "num_input_tokens_seen": 27312352,
      "step": 30550
    },
    {
      "epoch": 8.064141480797149,
      "grad_norm": 0.09428085386753082,
      "learning_rate": 6.57013019016712e-06,
      "loss": 0.0335,
      "num_input_tokens_seen": 27316928,
      "step": 30555
    },
    {
      "epoch": 8.065461264352646,
      "grad_norm": 0.26983922719955444,
      "learning_rate": 6.563498135779142e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 27321568,
      "step": 30560
    },
    {
      "epoch": 8.066781047908144,
      "grad_norm": 0.014601712115108967,
      "learning_rate": 6.556868924531431e-06,
      "loss": 0.0547,
      "num_input_tokens_seen": 27326112,
      "step": 30565
    },
    {
      "epoch": 8.06810083146364,
      "grad_norm": 0.049448683857917786,
      "learning_rate": 6.550242557446304e-06,
      "loss": 0.0394,
      "num_input_tokens_seen": 27330592,
      "step": 30570
    },
    {
      "epoch": 8.069420615019137,
      "grad_norm": 0.06863787025213242,
      "learning_rate": 6.543619035545634e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 27335136,
      "step": 30575
    },
    {
      "epoch": 8.070740398574634,
      "grad_norm": 0.04687635600566864,
      "learning_rate": 6.53699835985084e-06,
      "loss": 0.0201,
      "num_input_tokens_seen": 27339456,
      "step": 30580
    },
    {
      "epoch": 8.07206018213013,
      "grad_norm": 0.1384400576353073,
      "learning_rate": 6.530380531382927e-06,
      "loss": 0.039,
      "num_input_tokens_seen": 27343808,
      "step": 30585
    },
    {
      "epoch": 8.073379965685628,
      "grad_norm": 0.23200586438179016,
      "learning_rate": 6.523765551162433e-06,
      "loss": 0.0717,
      "num_input_tokens_seen": 27348320,
      "step": 30590
    },
    {
      "epoch": 8.074699749241125,
      "grad_norm": 0.1067097932100296,
      "learning_rate": 6.517153420209476e-06,
      "loss": 0.0332,
      "num_input_tokens_seen": 27352800,
      "step": 30595
    },
    {
      "epoch": 8.076019532796622,
      "grad_norm": 0.08200924098491669,
      "learning_rate": 6.510544139543739e-06,
      "loss": 0.0434,
      "num_input_tokens_seen": 27356864,
      "step": 30600
    },
    {
      "epoch": 8.076019532796622,
      "eval_loss": 0.06709827482700348,
      "eval_runtime": 64.8336,
      "eval_samples_per_second": 103.881,
      "eval_steps_per_second": 25.974,
      "num_input_tokens_seen": 27356864,
      "step": 30600
    },
    {
      "epoch": 8.077339316352118,
      "grad_norm": 0.2393959015607834,
      "learning_rate": 6.503937710184452e-06,
      "loss": 0.1075,
      "num_input_tokens_seen": 27361376,
      "step": 30605
    },
    {
      "epoch": 8.078659099907615,
      "grad_norm": 0.11751607805490494,
      "learning_rate": 6.4973341331503954e-06,
      "loss": 0.045,
      "num_input_tokens_seen": 27366112,
      "step": 30610
    },
    {
      "epoch": 8.079978883463111,
      "grad_norm": 0.1562637835741043,
      "learning_rate": 6.490733409459942e-06,
      "loss": 0.0556,
      "num_input_tokens_seen": 27370752,
      "step": 30615
    },
    {
      "epoch": 8.08129866701861,
      "grad_norm": 0.4470108151435852,
      "learning_rate": 6.484135540130995e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 27375232,
      "step": 30620
    },
    {
      "epoch": 8.082618450574106,
      "grad_norm": 0.04428371787071228,
      "learning_rate": 6.4775405261810364e-06,
      "loss": 0.061,
      "num_input_tokens_seen": 27379712,
      "step": 30625
    },
    {
      "epoch": 8.083938234129603,
      "grad_norm": 0.29352280497550964,
      "learning_rate": 6.470948368627092e-06,
      "loss": 0.1114,
      "num_input_tokens_seen": 27384000,
      "step": 30630
    },
    {
      "epoch": 8.0852580176851,
      "grad_norm": 0.36975812911987305,
      "learning_rate": 6.464359068485756e-06,
      "loss": 0.0439,
      "num_input_tokens_seen": 27388576,
      "step": 30635
    },
    {
      "epoch": 8.086577801240596,
      "grad_norm": 0.24279573559761047,
      "learning_rate": 6.457772626773195e-06,
      "loss": 0.064,
      "num_input_tokens_seen": 27393216,
      "step": 30640
    },
    {
      "epoch": 8.087897584796094,
      "grad_norm": 0.21298812329769135,
      "learning_rate": 6.451189044505104e-06,
      "loss": 0.038,
      "num_input_tokens_seen": 27397664,
      "step": 30645
    },
    {
      "epoch": 8.08921736835159,
      "grad_norm": 0.3142053186893463,
      "learning_rate": 6.44460832269676e-06,
      "loss": 0.1105,
      "num_input_tokens_seen": 27402272,
      "step": 30650
    },
    {
      "epoch": 8.090537151907087,
      "grad_norm": 0.041229307651519775,
      "learning_rate": 6.438030462363001e-06,
      "loss": 0.0737,
      "num_input_tokens_seen": 27406816,
      "step": 30655
    },
    {
      "epoch": 8.091856935462584,
      "grad_norm": 0.11774046719074249,
      "learning_rate": 6.431455464518205e-06,
      "loss": 0.0323,
      "num_input_tokens_seen": 27411552,
      "step": 30660
    },
    {
      "epoch": 8.09317671901808,
      "grad_norm": 0.2325984090566635,
      "learning_rate": 6.424883330176326e-06,
      "loss": 0.0435,
      "num_input_tokens_seen": 27416128,
      "step": 30665
    },
    {
      "epoch": 8.094496502573579,
      "grad_norm": 0.3053373396396637,
      "learning_rate": 6.418314060350864e-06,
      "loss": 0.0971,
      "num_input_tokens_seen": 27420736,
      "step": 30670
    },
    {
      "epoch": 8.095816286129075,
      "grad_norm": 0.1313740313053131,
      "learning_rate": 6.4117476560548895e-06,
      "loss": 0.0867,
      "num_input_tokens_seen": 27425376,
      "step": 30675
    },
    {
      "epoch": 8.097136069684572,
      "grad_norm": 0.05126567557454109,
      "learning_rate": 6.405184118301016e-06,
      "loss": 0.0377,
      "num_input_tokens_seen": 27429600,
      "step": 30680
    },
    {
      "epoch": 8.098455853240068,
      "grad_norm": 0.09584731608629227,
      "learning_rate": 6.398623448101434e-06,
      "loss": 0.0257,
      "num_input_tokens_seen": 27433792,
      "step": 30685
    },
    {
      "epoch": 8.099775636795565,
      "grad_norm": 0.07174982130527496,
      "learning_rate": 6.392065646467871e-06,
      "loss": 0.0257,
      "num_input_tokens_seen": 27438432,
      "step": 30690
    },
    {
      "epoch": 8.101095420351063,
      "grad_norm": 0.16257187724113464,
      "learning_rate": 6.385510714411632e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 27443136,
      "step": 30695
    },
    {
      "epoch": 8.10241520390656,
      "grad_norm": 0.038926783949136734,
      "learning_rate": 6.378958652943559e-06,
      "loss": 0.0236,
      "num_input_tokens_seen": 27447712,
      "step": 30700
    },
    {
      "epoch": 8.103734987462056,
      "grad_norm": 0.2738514840602875,
      "learning_rate": 6.3724094630740776e-06,
      "loss": 0.0624,
      "num_input_tokens_seen": 27452096,
      "step": 30705
    },
    {
      "epoch": 8.105054771017553,
      "grad_norm": 0.211082324385643,
      "learning_rate": 6.365863145813136e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 27456672,
      "step": 30710
    },
    {
      "epoch": 8.10637455457305,
      "grad_norm": 0.10475766658782959,
      "learning_rate": 6.359319702170269e-06,
      "loss": 0.0161,
      "num_input_tokens_seen": 27460832,
      "step": 30715
    },
    {
      "epoch": 8.107694338128548,
      "grad_norm": 0.1974291354417801,
      "learning_rate": 6.352779133154566e-06,
      "loss": 0.0275,
      "num_input_tokens_seen": 27465088,
      "step": 30720
    },
    {
      "epoch": 8.109014121684044,
      "grad_norm": 0.0966809093952179,
      "learning_rate": 6.346241439774648e-06,
      "loss": 0.0805,
      "num_input_tokens_seen": 27469632,
      "step": 30725
    },
    {
      "epoch": 8.110333905239541,
      "grad_norm": 0.04558365046977997,
      "learning_rate": 6.339706623038716e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 27474080,
      "step": 30730
    },
    {
      "epoch": 8.111653688795037,
      "grad_norm": 0.09342856705188751,
      "learning_rate": 6.333174683954532e-06,
      "loss": 0.0395,
      "num_input_tokens_seen": 27478496,
      "step": 30735
    },
    {
      "epoch": 8.112973472350534,
      "grad_norm": 0.08678475767374039,
      "learning_rate": 6.326645623529387e-06,
      "loss": 0.0615,
      "num_input_tokens_seen": 27482752,
      "step": 30740
    },
    {
      "epoch": 8.11429325590603,
      "grad_norm": 0.33435139060020447,
      "learning_rate": 6.320119442770156e-06,
      "loss": 0.0706,
      "num_input_tokens_seen": 27487296,
      "step": 30745
    },
    {
      "epoch": 8.115613039461529,
      "grad_norm": 0.34346190094947815,
      "learning_rate": 6.313596142683254e-06,
      "loss": 0.1007,
      "num_input_tokens_seen": 27492288,
      "step": 30750
    },
    {
      "epoch": 8.116932823017025,
      "grad_norm": 0.2475024163722992,
      "learning_rate": 6.307075724274647e-06,
      "loss": 0.0317,
      "num_input_tokens_seen": 27496576,
      "step": 30755
    },
    {
      "epoch": 8.118252606572522,
      "grad_norm": 0.18166270852088928,
      "learning_rate": 6.300558188549882e-06,
      "loss": 0.0718,
      "num_input_tokens_seen": 27500960,
      "step": 30760
    },
    {
      "epoch": 8.119572390128019,
      "grad_norm": 0.0950164943933487,
      "learning_rate": 6.29404353651403e-06,
      "loss": 0.0962,
      "num_input_tokens_seen": 27505440,
      "step": 30765
    },
    {
      "epoch": 8.120892173683515,
      "grad_norm": 0.19131559133529663,
      "learning_rate": 6.287531769171737e-06,
      "loss": 0.0557,
      "num_input_tokens_seen": 27509920,
      "step": 30770
    },
    {
      "epoch": 8.122211957239013,
      "grad_norm": 0.5464094877243042,
      "learning_rate": 6.2810228875272045e-06,
      "loss": 0.0966,
      "num_input_tokens_seen": 27514464,
      "step": 30775
    },
    {
      "epoch": 8.12353174079451,
      "grad_norm": 0.5447847843170166,
      "learning_rate": 6.274516892584179e-06,
      "loss": 0.1148,
      "num_input_tokens_seen": 27519040,
      "step": 30780
    },
    {
      "epoch": 8.124851524350007,
      "grad_norm": 0.2564816474914551,
      "learning_rate": 6.268013785345969e-06,
      "loss": 0.0743,
      "num_input_tokens_seen": 27523776,
      "step": 30785
    },
    {
      "epoch": 8.126171307905503,
      "grad_norm": 0.3408893644809723,
      "learning_rate": 6.26151356681543e-06,
      "loss": 0.0328,
      "num_input_tokens_seen": 27527936,
      "step": 30790
    },
    {
      "epoch": 8.127491091461,
      "grad_norm": 0.2511783838272095,
      "learning_rate": 6.255016237994981e-06,
      "loss": 0.0882,
      "num_input_tokens_seen": 27532128,
      "step": 30795
    },
    {
      "epoch": 8.128810875016498,
      "grad_norm": 0.26270827651023865,
      "learning_rate": 6.248521799886603e-06,
      "loss": 0.0436,
      "num_input_tokens_seen": 27536640,
      "step": 30800
    },
    {
      "epoch": 8.128810875016498,
      "eval_loss": 0.06699128448963165,
      "eval_runtime": 64.7815,
      "eval_samples_per_second": 103.965,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 27536640,
      "step": 30800
    },
    {
      "epoch": 8.130130658571995,
      "grad_norm": 0.25550466775894165,
      "learning_rate": 6.242030253491798e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 27540896,
      "step": 30805
    },
    {
      "epoch": 8.131450442127491,
      "grad_norm": 0.2523503005504608,
      "learning_rate": 6.235541599811656e-06,
      "loss": 0.0806,
      "num_input_tokens_seen": 27545152,
      "step": 30810
    },
    {
      "epoch": 8.132770225682988,
      "grad_norm": 0.08061458170413971,
      "learning_rate": 6.229055839846814e-06,
      "loss": 0.0364,
      "num_input_tokens_seen": 27549984,
      "step": 30815
    },
    {
      "epoch": 8.134090009238484,
      "grad_norm": 0.07186921685934067,
      "learning_rate": 6.222572974597455e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 27554368,
      "step": 30820
    },
    {
      "epoch": 8.135409792793983,
      "grad_norm": 0.3117089867591858,
      "learning_rate": 6.216093005063306e-06,
      "loss": 0.0568,
      "num_input_tokens_seen": 27558720,
      "step": 30825
    },
    {
      "epoch": 8.13672957634948,
      "grad_norm": 0.36000946164131165,
      "learning_rate": 6.209615932243678e-06,
      "loss": 0.0854,
      "num_input_tokens_seen": 27562976,
      "step": 30830
    },
    {
      "epoch": 8.138049359904976,
      "grad_norm": 0.0566524975001812,
      "learning_rate": 6.203141757137399e-06,
      "loss": 0.0561,
      "num_input_tokens_seen": 27567520,
      "step": 30835
    },
    {
      "epoch": 8.139369143460472,
      "grad_norm": 0.13415862619876862,
      "learning_rate": 6.196670480742886e-06,
      "loss": 0.0432,
      "num_input_tokens_seen": 27571872,
      "step": 30840
    },
    {
      "epoch": 8.140688927015969,
      "grad_norm": 0.12202560901641846,
      "learning_rate": 6.190202104058074e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 27576384,
      "step": 30845
    },
    {
      "epoch": 8.142008710571467,
      "grad_norm": 0.03247178718447685,
      "learning_rate": 6.183736628080475e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 27580800,
      "step": 30850
    },
    {
      "epoch": 8.143328494126964,
      "grad_norm": 0.36564645171165466,
      "learning_rate": 6.177274053807155e-06,
      "loss": 0.0771,
      "num_input_tokens_seen": 27585568,
      "step": 30855
    },
    {
      "epoch": 8.14464827768246,
      "grad_norm": 0.06069609895348549,
      "learning_rate": 6.170814382234713e-06,
      "loss": 0.0806,
      "num_input_tokens_seen": 27590080,
      "step": 30860
    },
    {
      "epoch": 8.145968061237957,
      "grad_norm": 0.03083733841776848,
      "learning_rate": 6.16435761435932e-06,
      "loss": 0.0215,
      "num_input_tokens_seen": 27594464,
      "step": 30865
    },
    {
      "epoch": 8.147287844793453,
      "grad_norm": 0.22972054779529572,
      "learning_rate": 6.157903751176681e-06,
      "loss": 0.0818,
      "num_input_tokens_seen": 27598816,
      "step": 30870
    },
    {
      "epoch": 8.14860762834895,
      "grad_norm": 0.28183144330978394,
      "learning_rate": 6.151452793682066e-06,
      "loss": 0.0367,
      "num_input_tokens_seen": 27603200,
      "step": 30875
    },
    {
      "epoch": 8.149927411904448,
      "grad_norm": 0.1983894556760788,
      "learning_rate": 6.145004742870305e-06,
      "loss": 0.034,
      "num_input_tokens_seen": 27607488,
      "step": 30880
    },
    {
      "epoch": 8.151247195459945,
      "grad_norm": 0.16630153357982635,
      "learning_rate": 6.138559599735752e-06,
      "loss": 0.0595,
      "num_input_tokens_seen": 27611968,
      "step": 30885
    },
    {
      "epoch": 8.152566979015441,
      "grad_norm": 0.15515772998332977,
      "learning_rate": 6.132117365272344e-06,
      "loss": 0.0859,
      "num_input_tokens_seen": 27616608,
      "step": 30890
    },
    {
      "epoch": 8.153886762570938,
      "grad_norm": 0.72199547290802,
      "learning_rate": 6.125678040473545e-06,
      "loss": 0.0723,
      "num_input_tokens_seen": 27620960,
      "step": 30895
    },
    {
      "epoch": 8.155206546126434,
      "grad_norm": 0.3908475637435913,
      "learning_rate": 6.1192416263323755e-06,
      "loss": 0.1086,
      "num_input_tokens_seen": 27625216,
      "step": 30900
    },
    {
      "epoch": 8.156526329681933,
      "grad_norm": 0.3268996775150299,
      "learning_rate": 6.112808123841424e-06,
      "loss": 0.0424,
      "num_input_tokens_seen": 27629696,
      "step": 30905
    },
    {
      "epoch": 8.15784611323743,
      "grad_norm": 0.25765612721443176,
      "learning_rate": 6.106377533992805e-06,
      "loss": 0.0705,
      "num_input_tokens_seen": 27634144,
      "step": 30910
    },
    {
      "epoch": 8.159165896792926,
      "grad_norm": 0.12074697017669678,
      "learning_rate": 6.099949857778204e-06,
      "loss": 0.0138,
      "num_input_tokens_seen": 27638656,
      "step": 30915
    },
    {
      "epoch": 8.160485680348422,
      "grad_norm": 0.2204914540052414,
      "learning_rate": 6.093525096188852e-06,
      "loss": 0.0687,
      "num_input_tokens_seen": 27642880,
      "step": 30920
    },
    {
      "epoch": 8.161805463903919,
      "grad_norm": 0.1614728420972824,
      "learning_rate": 6.087103250215518e-06,
      "loss": 0.036,
      "num_input_tokens_seen": 27647168,
      "step": 30925
    },
    {
      "epoch": 8.163125247459417,
      "grad_norm": 0.2823449373245239,
      "learning_rate": 6.080684320848537e-06,
      "loss": 0.0682,
      "num_input_tokens_seen": 27651680,
      "step": 30930
    },
    {
      "epoch": 8.164445031014914,
      "grad_norm": 0.3879050314426422,
      "learning_rate": 6.074268309077794e-06,
      "loss": 0.0497,
      "num_input_tokens_seen": 27656288,
      "step": 30935
    },
    {
      "epoch": 8.16576481457041,
      "grad_norm": 0.32613587379455566,
      "learning_rate": 6.067855215892709e-06,
      "loss": 0.0651,
      "num_input_tokens_seen": 27660640,
      "step": 30940
    },
    {
      "epoch": 8.167084598125907,
      "grad_norm": 0.06942100077867508,
      "learning_rate": 6.061445042282271e-06,
      "loss": 0.0473,
      "num_input_tokens_seen": 27665024,
      "step": 30945
    },
    {
      "epoch": 8.168404381681404,
      "grad_norm": 0.05501936748623848,
      "learning_rate": 6.055037789234999e-06,
      "loss": 0.0574,
      "num_input_tokens_seen": 27669504,
      "step": 30950
    },
    {
      "epoch": 8.169724165236902,
      "grad_norm": 0.22663424909114838,
      "learning_rate": 6.048633457738975e-06,
      "loss": 0.054,
      "num_input_tokens_seen": 27674176,
      "step": 30955
    },
    {
      "epoch": 8.171043948792398,
      "grad_norm": 0.13617445528507233,
      "learning_rate": 6.042232048781837e-06,
      "loss": 0.0641,
      "num_input_tokens_seen": 27678624,
      "step": 30960
    },
    {
      "epoch": 8.172363732347895,
      "grad_norm": 0.34475985169410706,
      "learning_rate": 6.035833563350757e-06,
      "loss": 0.0509,
      "num_input_tokens_seen": 27683168,
      "step": 30965
    },
    {
      "epoch": 8.173683515903392,
      "grad_norm": 0.029613815248012543,
      "learning_rate": 6.0294380024324525e-06,
      "loss": 0.0323,
      "num_input_tokens_seen": 27687392,
      "step": 30970
    },
    {
      "epoch": 8.175003299458888,
      "grad_norm": 0.2132827192544937,
      "learning_rate": 6.023045367013213e-06,
      "loss": 0.09,
      "num_input_tokens_seen": 27691840,
      "step": 30975
    },
    {
      "epoch": 8.176323083014386,
      "grad_norm": 0.5717481374740601,
      "learning_rate": 6.016655658078851e-06,
      "loss": 0.0719,
      "num_input_tokens_seen": 27696096,
      "step": 30980
    },
    {
      "epoch": 8.177642866569883,
      "grad_norm": 0.04379801079630852,
      "learning_rate": 6.010268876614753e-06,
      "loss": 0.0742,
      "num_input_tokens_seen": 27700608,
      "step": 30985
    },
    {
      "epoch": 8.17896265012538,
      "grad_norm": 0.45642977952957153,
      "learning_rate": 6.0038850236058266e-06,
      "loss": 0.0314,
      "num_input_tokens_seen": 27705344,
      "step": 30990
    },
    {
      "epoch": 8.180282433680876,
      "grad_norm": 0.02169051207602024,
      "learning_rate": 5.997504100036549e-06,
      "loss": 0.0434,
      "num_input_tokens_seen": 27709920,
      "step": 30995
    },
    {
      "epoch": 8.181602217236373,
      "grad_norm": 0.40946316719055176,
      "learning_rate": 5.991126106890949e-06,
      "loss": 0.0449,
      "num_input_tokens_seen": 27714496,
      "step": 31000
    },
    {
      "epoch": 8.181602217236373,
      "eval_loss": 0.06689025461673737,
      "eval_runtime": 64.8118,
      "eval_samples_per_second": 103.916,
      "eval_steps_per_second": 25.983,
      "num_input_tokens_seen": 27714496,
      "step": 31000
    },
    {
      "epoch": 8.18292200079187,
      "grad_norm": 0.2756527066230774,
      "learning_rate": 5.984751045152576e-06,
      "loss": 0.0865,
      "num_input_tokens_seen": 27718976,
      "step": 31005
    },
    {
      "epoch": 8.184241784347368,
      "grad_norm": 0.11364801973104477,
      "learning_rate": 5.978378915804553e-06,
      "loss": 0.0464,
      "num_input_tokens_seen": 27723488,
      "step": 31010
    },
    {
      "epoch": 8.185561567902864,
      "grad_norm": 0.07122594118118286,
      "learning_rate": 5.972009719829547e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 27728064,
      "step": 31015
    },
    {
      "epoch": 8.18688135145836,
      "grad_norm": 0.28265488147735596,
      "learning_rate": 5.965643458209755e-06,
      "loss": 0.08,
      "num_input_tokens_seen": 27732800,
      "step": 31020
    },
    {
      "epoch": 8.188201135013857,
      "grad_norm": 0.044520679861307144,
      "learning_rate": 5.95928013192695e-06,
      "loss": 0.0808,
      "num_input_tokens_seen": 27737536,
      "step": 31025
    },
    {
      "epoch": 8.189520918569354,
      "grad_norm": 0.21377992630004883,
      "learning_rate": 5.952919741962423e-06,
      "loss": 0.0371,
      "num_input_tokens_seen": 27741824,
      "step": 31030
    },
    {
      "epoch": 8.190840702124852,
      "grad_norm": 0.06417445093393326,
      "learning_rate": 5.946562289297042e-06,
      "loss": 0.0422,
      "num_input_tokens_seen": 27746592,
      "step": 31035
    },
    {
      "epoch": 8.192160485680349,
      "grad_norm": 0.09747159481048584,
      "learning_rate": 5.9402077749111855e-06,
      "loss": 0.0418,
      "num_input_tokens_seen": 27751104,
      "step": 31040
    },
    {
      "epoch": 8.193480269235845,
      "grad_norm": 0.19890965521335602,
      "learning_rate": 5.933856199784821e-06,
      "loss": 0.0549,
      "num_input_tokens_seen": 27755520,
      "step": 31045
    },
    {
      "epoch": 8.194800052791342,
      "grad_norm": 0.629007875919342,
      "learning_rate": 5.927507564897419e-06,
      "loss": 0.1318,
      "num_input_tokens_seen": 27760096,
      "step": 31050
    },
    {
      "epoch": 8.196119836346838,
      "grad_norm": 0.22235006093978882,
      "learning_rate": 5.9211618712280395e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 27764640,
      "step": 31055
    },
    {
      "epoch": 8.197439619902337,
      "grad_norm": 0.055925942957401276,
      "learning_rate": 5.914819119755255e-06,
      "loss": 0.0288,
      "num_input_tokens_seen": 27769024,
      "step": 31060
    },
    {
      "epoch": 8.198759403457833,
      "grad_norm": 0.3730144500732422,
      "learning_rate": 5.908479311457205e-06,
      "loss": 0.0961,
      "num_input_tokens_seen": 27773504,
      "step": 31065
    },
    {
      "epoch": 8.20007918701333,
      "grad_norm": 0.16459566354751587,
      "learning_rate": 5.902142447311559e-06,
      "loss": 0.0776,
      "num_input_tokens_seen": 27777792,
      "step": 31070
    },
    {
      "epoch": 8.201398970568826,
      "grad_norm": 0.4478456974029541,
      "learning_rate": 5.895808528295546e-06,
      "loss": 0.0845,
      "num_input_tokens_seen": 27782336,
      "step": 31075
    },
    {
      "epoch": 8.202718754124323,
      "grad_norm": 0.16906821727752686,
      "learning_rate": 5.889477555385941e-06,
      "loss": 0.0613,
      "num_input_tokens_seen": 27786752,
      "step": 31080
    },
    {
      "epoch": 8.204038537679821,
      "grad_norm": 0.3061339557170868,
      "learning_rate": 5.883149529559051e-06,
      "loss": 0.018,
      "num_input_tokens_seen": 27791008,
      "step": 31085
    },
    {
      "epoch": 8.205358321235318,
      "grad_norm": 0.09723767638206482,
      "learning_rate": 5.876824451790738e-06,
      "loss": 0.0394,
      "num_input_tokens_seen": 27795776,
      "step": 31090
    },
    {
      "epoch": 8.206678104790814,
      "grad_norm": 0.19318608939647675,
      "learning_rate": 5.87050232305642e-06,
      "loss": 0.0533,
      "num_input_tokens_seen": 27800032,
      "step": 31095
    },
    {
      "epoch": 8.207997888346311,
      "grad_norm": 0.4211634397506714,
      "learning_rate": 5.864183144331034e-06,
      "loss": 0.0673,
      "num_input_tokens_seen": 27804320,
      "step": 31100
    },
    {
      "epoch": 8.209317671901808,
      "grad_norm": 0.23600414395332336,
      "learning_rate": 5.857866916589089e-06,
      "loss": 0.0484,
      "num_input_tokens_seen": 27808544,
      "step": 31105
    },
    {
      "epoch": 8.210637455457306,
      "grad_norm": 0.03233930468559265,
      "learning_rate": 5.8515536408046216e-06,
      "loss": 0.0587,
      "num_input_tokens_seen": 27813280,
      "step": 31110
    },
    {
      "epoch": 8.211957239012802,
      "grad_norm": 0.11944656819105148,
      "learning_rate": 5.845243317951208e-06,
      "loss": 0.0519,
      "num_input_tokens_seen": 27817696,
      "step": 31115
    },
    {
      "epoch": 8.213277022568299,
      "grad_norm": 0.342307984828949,
      "learning_rate": 5.838935949001997e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 27822400,
      "step": 31120
    },
    {
      "epoch": 8.214596806123796,
      "grad_norm": 0.3261626064777374,
      "learning_rate": 5.8326315349296476e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 27826944,
      "step": 31125
    },
    {
      "epoch": 8.215916589679292,
      "grad_norm": 0.4118320345878601,
      "learning_rate": 5.826330076706396e-06,
      "loss": 0.0683,
      "num_input_tokens_seen": 27831456,
      "step": 31130
    },
    {
      "epoch": 8.21723637323479,
      "grad_norm": 0.048163801431655884,
      "learning_rate": 5.820031575303988e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 27835936,
      "step": 31135
    },
    {
      "epoch": 8.218556156790287,
      "grad_norm": 0.10341208428144455,
      "learning_rate": 5.813736031693745e-06,
      "loss": 0.0721,
      "num_input_tokens_seen": 27840320,
      "step": 31140
    },
    {
      "epoch": 8.219875940345784,
      "grad_norm": 0.08211777359247208,
      "learning_rate": 5.807443446846522e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 27844736,
      "step": 31145
    },
    {
      "epoch": 8.22119572390128,
      "grad_norm": 0.07241077721118927,
      "learning_rate": 5.801153821732699e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 27848960,
      "step": 31150
    },
    {
      "epoch": 8.222515507456777,
      "grad_norm": 0.30976414680480957,
      "learning_rate": 5.794867157322229e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 27853312,
      "step": 31155
    },
    {
      "epoch": 8.223835291012273,
      "grad_norm": 0.2636660039424896,
      "learning_rate": 5.788583454584593e-06,
      "loss": 0.0854,
      "num_input_tokens_seen": 27857920,
      "step": 31160
    },
    {
      "epoch": 8.225155074567772,
      "grad_norm": 0.13529348373413086,
      "learning_rate": 5.7823027144888075e-06,
      "loss": 0.042,
      "num_input_tokens_seen": 27862272,
      "step": 31165
    },
    {
      "epoch": 8.226474858123268,
      "grad_norm": 0.11944790929555893,
      "learning_rate": 5.776024938003455e-06,
      "loss": 0.0379,
      "num_input_tokens_seen": 27866752,
      "step": 31170
    },
    {
      "epoch": 8.227794641678765,
      "grad_norm": 0.1988510936498642,
      "learning_rate": 5.7697501260966345e-06,
      "loss": 0.0388,
      "num_input_tokens_seen": 27871552,
      "step": 31175
    },
    {
      "epoch": 8.229114425234261,
      "grad_norm": 0.23770549893379211,
      "learning_rate": 5.7634782797360145e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 27876256,
      "step": 31180
    },
    {
      "epoch": 8.230434208789758,
      "grad_norm": 0.15230992436408997,
      "learning_rate": 5.757209399888777e-06,
      "loss": 0.0194,
      "num_input_tokens_seen": 27880832,
      "step": 31185
    },
    {
      "epoch": 8.231753992345256,
      "grad_norm": 0.3588865399360657,
      "learning_rate": 5.750943487521679e-06,
      "loss": 0.0591,
      "num_input_tokens_seen": 27885120,
      "step": 31190
    },
    {
      "epoch": 8.233073775900753,
      "grad_norm": 0.2936408221721649,
      "learning_rate": 5.744680543600986e-06,
      "loss": 0.0492,
      "num_input_tokens_seen": 27889248,
      "step": 31195
    },
    {
      "epoch": 8.23439355945625,
      "grad_norm": 0.17947396636009216,
      "learning_rate": 5.738420569092537e-06,
      "loss": 0.0607,
      "num_input_tokens_seen": 27893536,
      "step": 31200
    },
    {
      "epoch": 8.23439355945625,
      "eval_loss": 0.06724120676517487,
      "eval_runtime": 64.772,
      "eval_samples_per_second": 103.98,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 27893536,
      "step": 31200
    },
    {
      "epoch": 8.235713343011746,
      "grad_norm": 0.14596471190452576,
      "learning_rate": 5.732163564961684e-06,
      "loss": 0.056,
      "num_input_tokens_seen": 27897824,
      "step": 31205
    },
    {
      "epoch": 8.237033126567242,
      "grad_norm": 0.07952891290187836,
      "learning_rate": 5.725909532173354e-06,
      "loss": 0.0471,
      "num_input_tokens_seen": 27902144,
      "step": 31210
    },
    {
      "epoch": 8.23835291012274,
      "grad_norm": 0.11572608351707458,
      "learning_rate": 5.719658471691977e-06,
      "loss": 0.0672,
      "num_input_tokens_seen": 27906400,
      "step": 31215
    },
    {
      "epoch": 8.239672693678237,
      "grad_norm": 0.05074088275432587,
      "learning_rate": 5.71341038448156e-06,
      "loss": 0.0254,
      "num_input_tokens_seen": 27910752,
      "step": 31220
    },
    {
      "epoch": 8.240992477233734,
      "grad_norm": 0.47607237100601196,
      "learning_rate": 5.707165271505635e-06,
      "loss": 0.0989,
      "num_input_tokens_seen": 27914816,
      "step": 31225
    },
    {
      "epoch": 8.24231226078923,
      "grad_norm": 0.33451172709465027,
      "learning_rate": 5.700923133727271e-06,
      "loss": 0.0586,
      "num_input_tokens_seen": 27918976,
      "step": 31230
    },
    {
      "epoch": 8.243632044344727,
      "grad_norm": 0.0521954782307148,
      "learning_rate": 5.694683972109083e-06,
      "loss": 0.0394,
      "num_input_tokens_seen": 27923584,
      "step": 31235
    },
    {
      "epoch": 8.244951827900225,
      "grad_norm": 0.2928738594055176,
      "learning_rate": 5.688447787613241e-06,
      "loss": 0.0533,
      "num_input_tokens_seen": 27928192,
      "step": 31240
    },
    {
      "epoch": 8.246271611455722,
      "grad_norm": 0.0311958696693182,
      "learning_rate": 5.6822145812014285e-06,
      "loss": 0.0363,
      "num_input_tokens_seen": 27932576,
      "step": 31245
    },
    {
      "epoch": 8.247591395011218,
      "grad_norm": 0.19424369931221008,
      "learning_rate": 5.675984353834896e-06,
      "loss": 0.0564,
      "num_input_tokens_seen": 27936960,
      "step": 31250
    },
    {
      "epoch": 8.248911178566715,
      "grad_norm": 0.26832765340805054,
      "learning_rate": 5.66975710647441e-06,
      "loss": 0.0392,
      "num_input_tokens_seen": 27941632,
      "step": 31255
    },
    {
      "epoch": 8.250230962122211,
      "grad_norm": 0.2575373649597168,
      "learning_rate": 5.663532840080304e-06,
      "loss": 0.0912,
      "num_input_tokens_seen": 27946144,
      "step": 31260
    },
    {
      "epoch": 8.251550745677708,
      "grad_norm": 0.3163471221923828,
      "learning_rate": 5.6573115556124325e-06,
      "loss": 0.0708,
      "num_input_tokens_seen": 27950880,
      "step": 31265
    },
    {
      "epoch": 8.252870529233206,
      "grad_norm": 0.12572205066680908,
      "learning_rate": 5.651093254030185e-06,
      "loss": 0.0738,
      "num_input_tokens_seen": 27955136,
      "step": 31270
    },
    {
      "epoch": 8.254190312788703,
      "grad_norm": 0.06410905718803406,
      "learning_rate": 5.644877936292514e-06,
      "loss": 0.0506,
      "num_input_tokens_seen": 27959712,
      "step": 31275
    },
    {
      "epoch": 8.2555100963442,
      "grad_norm": 0.08642936497926712,
      "learning_rate": 5.638665603357901e-06,
      "loss": 0.0346,
      "num_input_tokens_seen": 27964256,
      "step": 31280
    },
    {
      "epoch": 8.256829879899696,
      "grad_norm": 0.1099819540977478,
      "learning_rate": 5.632456256184357e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 27968800,
      "step": 31285
    },
    {
      "epoch": 8.258149663455193,
      "grad_norm": 0.2367771416902542,
      "learning_rate": 5.626249895729452e-06,
      "loss": 0.0863,
      "num_input_tokens_seen": 27973312,
      "step": 31290
    },
    {
      "epoch": 8.25946944701069,
      "grad_norm": 0.18472091853618622,
      "learning_rate": 5.620046522950273e-06,
      "loss": 0.0871,
      "num_input_tokens_seen": 27977888,
      "step": 31295
    },
    {
      "epoch": 8.260789230566187,
      "grad_norm": 0.40643802285194397,
      "learning_rate": 5.613846138803464e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 27982304,
      "step": 31300
    },
    {
      "epoch": 8.262109014121684,
      "grad_norm": 0.1791028082370758,
      "learning_rate": 5.607648744245206e-06,
      "loss": 0.051,
      "num_input_tokens_seen": 27986656,
      "step": 31305
    },
    {
      "epoch": 8.26342879767718,
      "grad_norm": 0.17201103270053864,
      "learning_rate": 5.601454340231207e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 27991072,
      "step": 31310
    },
    {
      "epoch": 8.264748581232677,
      "grad_norm": 0.18725524842739105,
      "learning_rate": 5.595262927716724e-06,
      "loss": 0.0433,
      "num_input_tokens_seen": 27995680,
      "step": 31315
    },
    {
      "epoch": 8.266068364788175,
      "grad_norm": 0.15287287533283234,
      "learning_rate": 5.589074507656561e-06,
      "loss": 0.0735,
      "num_input_tokens_seen": 28000000,
      "step": 31320
    },
    {
      "epoch": 8.267388148343672,
      "grad_norm": 0.04334359988570213,
      "learning_rate": 5.582889081005044e-06,
      "loss": 0.0606,
      "num_input_tokens_seen": 28004416,
      "step": 31325
    },
    {
      "epoch": 8.268707931899169,
      "grad_norm": 0.37684768438339233,
      "learning_rate": 5.5767066487160316e-06,
      "loss": 0.0654,
      "num_input_tokens_seen": 28008864,
      "step": 31330
    },
    {
      "epoch": 8.270027715454665,
      "grad_norm": 0.648932158946991,
      "learning_rate": 5.570527211742949e-06,
      "loss": 0.0916,
      "num_input_tokens_seen": 28013440,
      "step": 31335
    },
    {
      "epoch": 8.271347499010162,
      "grad_norm": 0.09987188875675201,
      "learning_rate": 5.564350771038731e-06,
      "loss": 0.0686,
      "num_input_tokens_seen": 28017792,
      "step": 31340
    },
    {
      "epoch": 8.27266728256566,
      "grad_norm": 0.2771715819835663,
      "learning_rate": 5.558177327555875e-06,
      "loss": 0.0973,
      "num_input_tokens_seen": 28022144,
      "step": 31345
    },
    {
      "epoch": 8.273987066121157,
      "grad_norm": 0.12438545376062393,
      "learning_rate": 5.552006882246388e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 28026848,
      "step": 31350
    },
    {
      "epoch": 8.275306849676653,
      "grad_norm": 0.4493320882320404,
      "learning_rate": 5.545839436061839e-06,
      "loss": 0.0386,
      "num_input_tokens_seen": 28031360,
      "step": 31355
    },
    {
      "epoch": 8.27662663323215,
      "grad_norm": 0.05131245404481888,
      "learning_rate": 5.539674989953331e-06,
      "loss": 0.0572,
      "num_input_tokens_seen": 28036160,
      "step": 31360
    },
    {
      "epoch": 8.277946416787646,
      "grad_norm": 0.15000054240226746,
      "learning_rate": 5.533513544871488e-06,
      "loss": 0.0484,
      "num_input_tokens_seen": 28040448,
      "step": 31365
    },
    {
      "epoch": 8.279266200343145,
      "grad_norm": 0.3825835883617401,
      "learning_rate": 5.527355101766493e-06,
      "loss": 0.0774,
      "num_input_tokens_seen": 28044768,
      "step": 31370
    },
    {
      "epoch": 8.280585983898641,
      "grad_norm": 0.0758058950304985,
      "learning_rate": 5.521199661588044e-06,
      "loss": 0.0335,
      "num_input_tokens_seen": 28049376,
      "step": 31375
    },
    {
      "epoch": 8.281905767454138,
      "grad_norm": 0.4857470989227295,
      "learning_rate": 5.5150472252853944e-06,
      "loss": 0.0931,
      "num_input_tokens_seen": 28053664,
      "step": 31380
    },
    {
      "epoch": 8.283225551009634,
      "grad_norm": 0.2806839346885681,
      "learning_rate": 5.50889779380733e-06,
      "loss": 0.0701,
      "num_input_tokens_seen": 28058112,
      "step": 31385
    },
    {
      "epoch": 8.28454533456513,
      "grad_norm": 0.09459013491868973,
      "learning_rate": 5.5027513681021605e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 28062688,
      "step": 31390
    },
    {
      "epoch": 8.285865118120629,
      "grad_norm": 0.11410725861787796,
      "learning_rate": 5.4966079491177545e-06,
      "loss": 0.0678,
      "num_input_tokens_seen": 28067392,
      "step": 31395
    },
    {
      "epoch": 8.287184901676126,
      "grad_norm": 0.08025260269641876,
      "learning_rate": 5.490467537801491e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 28071808,
      "step": 31400
    },
    {
      "epoch": 8.287184901676126,
      "eval_loss": 0.06709538400173187,
      "eval_runtime": 64.7836,
      "eval_samples_per_second": 103.961,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 28071808,
      "step": 31400
    },
    {
      "epoch": 8.288504685231622,
      "grad_norm": 0.2685471475124359,
      "learning_rate": 5.484330135100313e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 28076480,
      "step": 31405
    },
    {
      "epoch": 8.289824468787119,
      "grad_norm": 0.10487838089466095,
      "learning_rate": 5.4781957419606785e-06,
      "loss": 0.0415,
      "num_input_tokens_seen": 28081440,
      "step": 31410
    },
    {
      "epoch": 8.291144252342615,
      "grad_norm": 0.16225402057170868,
      "learning_rate": 5.472064359328577e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 28086016,
      "step": 31415
    },
    {
      "epoch": 8.292464035898112,
      "grad_norm": 0.2321801632642746,
      "learning_rate": 5.4659359881495565e-06,
      "loss": 0.0301,
      "num_input_tokens_seen": 28090304,
      "step": 31420
    },
    {
      "epoch": 8.29378381945361,
      "grad_norm": 0.05357467755675316,
      "learning_rate": 5.4598106293686916e-06,
      "loss": 0.0832,
      "num_input_tokens_seen": 28094784,
      "step": 31425
    },
    {
      "epoch": 8.295103603009107,
      "grad_norm": 0.306360125541687,
      "learning_rate": 5.45368828393058e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 28099424,
      "step": 31430
    },
    {
      "epoch": 8.296423386564603,
      "grad_norm": 0.2016575187444687,
      "learning_rate": 5.44756895277937e-06,
      "loss": 0.0626,
      "num_input_tokens_seen": 28103840,
      "step": 31435
    },
    {
      "epoch": 8.2977431701201,
      "grad_norm": 0.3366232216358185,
      "learning_rate": 5.441452636858746e-06,
      "loss": 0.078,
      "num_input_tokens_seen": 28108064,
      "step": 31440
    },
    {
      "epoch": 8.299062953675596,
      "grad_norm": 0.22584505379199982,
      "learning_rate": 5.435339337111905e-06,
      "loss": 0.0622,
      "num_input_tokens_seen": 28112768,
      "step": 31445
    },
    {
      "epoch": 8.300382737231095,
      "grad_norm": 0.3065774440765381,
      "learning_rate": 5.42922905448161e-06,
      "loss": 0.0598,
      "num_input_tokens_seen": 28117312,
      "step": 31450
    },
    {
      "epoch": 8.301702520786591,
      "grad_norm": 0.1937895119190216,
      "learning_rate": 5.423121789910129e-06,
      "loss": 0.0615,
      "num_input_tokens_seen": 28121600,
      "step": 31455
    },
    {
      "epoch": 8.303022304342088,
      "grad_norm": 0.1442263424396515,
      "learning_rate": 5.417017544339287e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 28126048,
      "step": 31460
    },
    {
      "epoch": 8.304342087897584,
      "grad_norm": 0.2206220179796219,
      "learning_rate": 5.410916318710443e-06,
      "loss": 0.0989,
      "num_input_tokens_seen": 28130656,
      "step": 31465
    },
    {
      "epoch": 8.305661871453081,
      "grad_norm": 0.22425362467765808,
      "learning_rate": 5.404818113964466e-06,
      "loss": 0.0596,
      "num_input_tokens_seen": 28135040,
      "step": 31470
    },
    {
      "epoch": 8.30698165500858,
      "grad_norm": 0.10591068863868713,
      "learning_rate": 5.398722931041792e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 28139296,
      "step": 31475
    },
    {
      "epoch": 8.308301438564076,
      "grad_norm": 0.19140157103538513,
      "learning_rate": 5.392630770882367e-06,
      "loss": 0.0508,
      "num_input_tokens_seen": 28143840,
      "step": 31480
    },
    {
      "epoch": 8.309621222119572,
      "grad_norm": 0.07222902774810791,
      "learning_rate": 5.3865416344256705e-06,
      "loss": 0.0274,
      "num_input_tokens_seen": 28148256,
      "step": 31485
    },
    {
      "epoch": 8.310941005675069,
      "grad_norm": 0.1052062138915062,
      "learning_rate": 5.380455522610742e-06,
      "loss": 0.0305,
      "num_input_tokens_seen": 28152672,
      "step": 31490
    },
    {
      "epoch": 8.312260789230566,
      "grad_norm": 0.1660299003124237,
      "learning_rate": 5.374372436376116e-06,
      "loss": 0.037,
      "num_input_tokens_seen": 28157088,
      "step": 31495
    },
    {
      "epoch": 8.313580572786064,
      "grad_norm": 0.3312130570411682,
      "learning_rate": 5.368292376659895e-06,
      "loss": 0.0784,
      "num_input_tokens_seen": 28161312,
      "step": 31500
    },
    {
      "epoch": 8.31490035634156,
      "grad_norm": 0.08026500791311264,
      "learning_rate": 5.362215344399701e-06,
      "loss": 0.0681,
      "num_input_tokens_seen": 28165536,
      "step": 31505
    },
    {
      "epoch": 8.316220139897057,
      "grad_norm": 0.4170088768005371,
      "learning_rate": 5.356141340532678e-06,
      "loss": 0.0763,
      "num_input_tokens_seen": 28170016,
      "step": 31510
    },
    {
      "epoch": 8.317539923452554,
      "grad_norm": 0.3503752648830414,
      "learning_rate": 5.350070365995522e-06,
      "loss": 0.0723,
      "num_input_tokens_seen": 28174304,
      "step": 31515
    },
    {
      "epoch": 8.31885970700805,
      "grad_norm": 0.5154256820678711,
      "learning_rate": 5.344002421724459e-06,
      "loss": 0.0453,
      "num_input_tokens_seen": 28178816,
      "step": 31520
    },
    {
      "epoch": 8.320179490563547,
      "grad_norm": 0.16560965776443481,
      "learning_rate": 5.337937508655228e-06,
      "loss": 0.0434,
      "num_input_tokens_seen": 28183456,
      "step": 31525
    },
    {
      "epoch": 8.321499274119045,
      "grad_norm": 0.41666868329048157,
      "learning_rate": 5.331875627723126e-06,
      "loss": 0.046,
      "num_input_tokens_seen": 28187840,
      "step": 31530
    },
    {
      "epoch": 8.322819057674542,
      "grad_norm": 0.1576455682516098,
      "learning_rate": 5.325816779862963e-06,
      "loss": 0.0472,
      "num_input_tokens_seen": 28192288,
      "step": 31535
    },
    {
      "epoch": 8.324138841230038,
      "grad_norm": 0.42261579632759094,
      "learning_rate": 5.319760966009102e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 28196864,
      "step": 31540
    },
    {
      "epoch": 8.325458624785535,
      "grad_norm": 0.21517950296401978,
      "learning_rate": 5.3137081870954096e-06,
      "loss": 0.0485,
      "num_input_tokens_seen": 28201568,
      "step": 31545
    },
    {
      "epoch": 8.326778408341031,
      "grad_norm": 0.25522544980049133,
      "learning_rate": 5.307658444055313e-06,
      "loss": 0.0765,
      "num_input_tokens_seen": 28205664,
      "step": 31550
    },
    {
      "epoch": 8.32809819189653,
      "grad_norm": 0.08197813481092453,
      "learning_rate": 5.301611737821749e-06,
      "loss": 0.0816,
      "num_input_tokens_seen": 28210464,
      "step": 31555
    },
    {
      "epoch": 8.329417975452026,
      "grad_norm": 0.15731588006019592,
      "learning_rate": 5.295568069327206e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 28214816,
      "step": 31560
    },
    {
      "epoch": 8.330737759007523,
      "grad_norm": 0.22606533765792847,
      "learning_rate": 5.289527439503683e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 28219296,
      "step": 31565
    },
    {
      "epoch": 8.33205754256302,
      "grad_norm": 0.2702862620353699,
      "learning_rate": 5.28348984928273e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 28223584,
      "step": 31570
    },
    {
      "epoch": 8.333377326118516,
      "grad_norm": 0.35594359040260315,
      "learning_rate": 5.27745529959541e-06,
      "loss": 0.0986,
      "num_input_tokens_seen": 28227584,
      "step": 31575
    },
    {
      "epoch": 8.334697109674014,
      "grad_norm": 0.057019129395484924,
      "learning_rate": 5.271423791372335e-06,
      "loss": 0.1018,
      "num_input_tokens_seen": 28232096,
      "step": 31580
    },
    {
      "epoch": 8.33601689322951,
      "grad_norm": 0.08245081454515457,
      "learning_rate": 5.26539532554364e-06,
      "loss": 0.0852,
      "num_input_tokens_seen": 28236640,
      "step": 31585
    },
    {
      "epoch": 8.337336676785007,
      "grad_norm": 0.3890661299228668,
      "learning_rate": 5.25936990303898e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 28241312,
      "step": 31590
    },
    {
      "epoch": 8.338656460340504,
      "grad_norm": 0.052959244698286057,
      "learning_rate": 5.253347524787555e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 28245536,
      "step": 31595
    },
    {
      "epoch": 8.339976243896,
      "grad_norm": 0.4196988046169281,
      "learning_rate": 5.2473281917181035e-06,
      "loss": 0.0963,
      "num_input_tokens_seen": 28250112,
      "step": 31600
    },
    {
      "epoch": 8.339976243896,
      "eval_loss": 0.06694076955318451,
      "eval_runtime": 64.7424,
      "eval_samples_per_second": 104.028,
      "eval_steps_per_second": 26.011,
      "num_input_tokens_seen": 28250112,
      "step": 31600
    },
    {
      "epoch": 8.341296027451499,
      "grad_norm": 0.22888000309467316,
      "learning_rate": 5.241311904758864e-06,
      "loss": 0.07,
      "num_input_tokens_seen": 28254368,
      "step": 31605
    },
    {
      "epoch": 8.342615811006995,
      "grad_norm": 0.0943085104227066,
      "learning_rate": 5.23529866483764e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 28258816,
      "step": 31610
    },
    {
      "epoch": 8.343935594562492,
      "grad_norm": 0.16361837089061737,
      "learning_rate": 5.229288472881732e-06,
      "loss": 0.0701,
      "num_input_tokens_seen": 28263392,
      "step": 31615
    },
    {
      "epoch": 8.345255378117988,
      "grad_norm": 0.2628397047519684,
      "learning_rate": 5.2232813298180025e-06,
      "loss": 0.1107,
      "num_input_tokens_seen": 28267968,
      "step": 31620
    },
    {
      "epoch": 8.346575161673485,
      "grad_norm": 0.13943825662136078,
      "learning_rate": 5.217277236572824e-06,
      "loss": 0.0731,
      "num_input_tokens_seen": 28272384,
      "step": 31625
    },
    {
      "epoch": 8.347894945228983,
      "grad_norm": 0.3473707437515259,
      "learning_rate": 5.211276194072093e-06,
      "loss": 0.098,
      "num_input_tokens_seen": 28276928,
      "step": 31630
    },
    {
      "epoch": 8.34921472878448,
      "grad_norm": 0.38137203454971313,
      "learning_rate": 5.205278203241254e-06,
      "loss": 0.0729,
      "num_input_tokens_seen": 28281280,
      "step": 31635
    },
    {
      "epoch": 8.350534512339976,
      "grad_norm": 0.33100423216819763,
      "learning_rate": 5.199283265005278e-06,
      "loss": 0.0445,
      "num_input_tokens_seen": 28285504,
      "step": 31640
    },
    {
      "epoch": 8.351854295895473,
      "grad_norm": 0.17923983931541443,
      "learning_rate": 5.193291380288648e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 28289824,
      "step": 31645
    },
    {
      "epoch": 8.35317407945097,
      "grad_norm": 0.3521307110786438,
      "learning_rate": 5.1873025500153995e-06,
      "loss": 0.0535,
      "num_input_tokens_seen": 28294528,
      "step": 31650
    },
    {
      "epoch": 8.354493863006468,
      "grad_norm": 0.2762030363082886,
      "learning_rate": 5.181316775109071e-06,
      "loss": 0.0708,
      "num_input_tokens_seen": 28299168,
      "step": 31655
    },
    {
      "epoch": 8.355813646561964,
      "grad_norm": 0.039546411484479904,
      "learning_rate": 5.1753340564927564e-06,
      "loss": 0.0341,
      "num_input_tokens_seen": 28303488,
      "step": 31660
    },
    {
      "epoch": 8.357133430117461,
      "grad_norm": 0.08518904447555542,
      "learning_rate": 5.169354395089068e-06,
      "loss": 0.0256,
      "num_input_tokens_seen": 28307968,
      "step": 31665
    },
    {
      "epoch": 8.358453213672957,
      "grad_norm": 0.056911397725343704,
      "learning_rate": 5.1633777918201346e-06,
      "loss": 0.0377,
      "num_input_tokens_seen": 28312160,
      "step": 31670
    },
    {
      "epoch": 8.359772997228454,
      "grad_norm": 0.11146537214517593,
      "learning_rate": 5.157404247607625e-06,
      "loss": 0.0656,
      "num_input_tokens_seen": 28316640,
      "step": 31675
    },
    {
      "epoch": 8.36109278078395,
      "grad_norm": 0.2967804968357086,
      "learning_rate": 5.1514337633727454e-06,
      "loss": 0.0601,
      "num_input_tokens_seen": 28321056,
      "step": 31680
    },
    {
      "epoch": 8.362412564339449,
      "grad_norm": 0.4036308825016022,
      "learning_rate": 5.145466340036206e-06,
      "loss": 0.1365,
      "num_input_tokens_seen": 28325408,
      "step": 31685
    },
    {
      "epoch": 8.363732347894945,
      "grad_norm": 0.12064474076032639,
      "learning_rate": 5.139501978518274e-06,
      "loss": 0.0488,
      "num_input_tokens_seen": 28329728,
      "step": 31690
    },
    {
      "epoch": 8.365052131450442,
      "grad_norm": 0.07563777267932892,
      "learning_rate": 5.133540679738716e-06,
      "loss": 0.0595,
      "num_input_tokens_seen": 28334208,
      "step": 31695
    },
    {
      "epoch": 8.366371915005939,
      "grad_norm": 0.10374762117862701,
      "learning_rate": 5.127582444616838e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 28338976,
      "step": 31700
    },
    {
      "epoch": 8.367691698561435,
      "grad_norm": 0.06006516143679619,
      "learning_rate": 5.121627274071486e-06,
      "loss": 0.0583,
      "num_input_tokens_seen": 28343520,
      "step": 31705
    },
    {
      "epoch": 8.369011482116933,
      "grad_norm": 0.2423616498708725,
      "learning_rate": 5.115675169021009e-06,
      "loss": 0.0782,
      "num_input_tokens_seen": 28347936,
      "step": 31710
    },
    {
      "epoch": 8.37033126567243,
      "grad_norm": 0.12826000154018402,
      "learning_rate": 5.1097261303832994e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 28352288,
      "step": 31715
    },
    {
      "epoch": 8.371651049227927,
      "grad_norm": 0.12992027401924133,
      "learning_rate": 5.103780159075788e-06,
      "loss": 0.0848,
      "num_input_tokens_seen": 28356992,
      "step": 31720
    },
    {
      "epoch": 8.372970832783423,
      "grad_norm": 0.2809655964374542,
      "learning_rate": 5.0978372560154e-06,
      "loss": 0.0514,
      "num_input_tokens_seen": 28361568,
      "step": 31725
    },
    {
      "epoch": 8.37429061633892,
      "grad_norm": 0.16390928626060486,
      "learning_rate": 5.091897422118619e-06,
      "loss": 0.0482,
      "num_input_tokens_seen": 28366272,
      "step": 31730
    },
    {
      "epoch": 8.375610399894418,
      "grad_norm": 0.060936011373996735,
      "learning_rate": 5.0859606583014305e-06,
      "loss": 0.0202,
      "num_input_tokens_seen": 28370688,
      "step": 31735
    },
    {
      "epoch": 8.376930183449915,
      "grad_norm": 0.12615734338760376,
      "learning_rate": 5.080026965479365e-06,
      "loss": 0.0375,
      "num_input_tokens_seen": 28375488,
      "step": 31740
    },
    {
      "epoch": 8.378249967005411,
      "grad_norm": 0.12842825055122375,
      "learning_rate": 5.074096344567475e-06,
      "loss": 0.0362,
      "num_input_tokens_seen": 28379904,
      "step": 31745
    },
    {
      "epoch": 8.379569750560908,
      "grad_norm": 0.6115501523017883,
      "learning_rate": 5.0681687964803294e-06,
      "loss": 0.0822,
      "num_input_tokens_seen": 28384224,
      "step": 31750
    },
    {
      "epoch": 8.380889534116404,
      "grad_norm": 0.1151709258556366,
      "learning_rate": 5.06224432213204e-06,
      "loss": 0.027,
      "num_input_tokens_seen": 28388736,
      "step": 31755
    },
    {
      "epoch": 8.382209317671903,
      "grad_norm": 0.18120890855789185,
      "learning_rate": 5.056322922436224e-06,
      "loss": 0.0552,
      "num_input_tokens_seen": 28393120,
      "step": 31760
    },
    {
      "epoch": 8.3835291012274,
      "grad_norm": 0.2290099859237671,
      "learning_rate": 5.0504045983060465e-06,
      "loss": 0.0838,
      "num_input_tokens_seen": 28397504,
      "step": 31765
    },
    {
      "epoch": 8.384848884782896,
      "grad_norm": 0.06172773241996765,
      "learning_rate": 5.044489350654183e-06,
      "loss": 0.0272,
      "num_input_tokens_seen": 28401984,
      "step": 31770
    },
    {
      "epoch": 8.386168668338392,
      "grad_norm": 0.5357003211975098,
      "learning_rate": 5.038577180392831e-06,
      "loss": 0.1154,
      "num_input_tokens_seen": 28406784,
      "step": 31775
    },
    {
      "epoch": 8.387488451893889,
      "grad_norm": 0.17018455266952515,
      "learning_rate": 5.032668088433729e-06,
      "loss": 0.0372,
      "num_input_tokens_seen": 28411168,
      "step": 31780
    },
    {
      "epoch": 8.388808235449387,
      "grad_norm": 0.09959359467029572,
      "learning_rate": 5.02676207568814e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 28415616,
      "step": 31785
    },
    {
      "epoch": 8.390128019004884,
      "grad_norm": 0.09072726219892502,
      "learning_rate": 5.02085914306683e-06,
      "loss": 0.0403,
      "num_input_tokens_seen": 28419936,
      "step": 31790
    },
    {
      "epoch": 8.39144780256038,
      "grad_norm": 0.14661361277103424,
      "learning_rate": 5.014959291480123e-06,
      "loss": 0.067,
      "num_input_tokens_seen": 28424416,
      "step": 31795
    },
    {
      "epoch": 8.392767586115877,
      "grad_norm": 0.17932093143463135,
      "learning_rate": 5.009062521837835e-06,
      "loss": 0.0301,
      "num_input_tokens_seen": 28428832,
      "step": 31800
    },
    {
      "epoch": 8.392767586115877,
      "eval_loss": 0.06704459339380264,
      "eval_runtime": 64.7313,
      "eval_samples_per_second": 104.046,
      "eval_steps_per_second": 26.015,
      "num_input_tokens_seen": 28428832,
      "step": 31800
    },
    {
      "epoch": 8.394087369671373,
      "grad_norm": 0.18261459469795227,
      "learning_rate": 5.003168835049324e-06,
      "loss": 0.0737,
      "num_input_tokens_seen": 28433344,
      "step": 31805
    },
    {
      "epoch": 8.39540715322687,
      "grad_norm": 0.07657528668642044,
      "learning_rate": 4.997278232023483e-06,
      "loss": 0.0469,
      "num_input_tokens_seen": 28437888,
      "step": 31810
    },
    {
      "epoch": 8.396726936782368,
      "grad_norm": 0.3366856575012207,
      "learning_rate": 4.9913907136687036e-06,
      "loss": 0.0579,
      "num_input_tokens_seen": 28442336,
      "step": 31815
    },
    {
      "epoch": 8.398046720337865,
      "grad_norm": 0.05351114645600319,
      "learning_rate": 4.985506280892918e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 28447168,
      "step": 31820
    },
    {
      "epoch": 8.399366503893361,
      "grad_norm": 0.19690437614917755,
      "learning_rate": 4.979624934603589e-06,
      "loss": 0.0394,
      "num_input_tokens_seen": 28451712,
      "step": 31825
    },
    {
      "epoch": 8.400686287448858,
      "grad_norm": 0.2705552577972412,
      "learning_rate": 4.97374667570768e-06,
      "loss": 0.037,
      "num_input_tokens_seen": 28456064,
      "step": 31830
    },
    {
      "epoch": 8.402006071004354,
      "grad_norm": 0.12079799175262451,
      "learning_rate": 4.967871505111704e-06,
      "loss": 0.0399,
      "num_input_tokens_seen": 28460672,
      "step": 31835
    },
    {
      "epoch": 8.403325854559853,
      "grad_norm": 0.2921290695667267,
      "learning_rate": 4.961999423721686e-06,
      "loss": 0.0587,
      "num_input_tokens_seen": 28465024,
      "step": 31840
    },
    {
      "epoch": 8.40464563811535,
      "grad_norm": 0.06648524850606918,
      "learning_rate": 4.956130432443159e-06,
      "loss": 0.0496,
      "num_input_tokens_seen": 28469472,
      "step": 31845
    },
    {
      "epoch": 8.405965421670846,
      "grad_norm": 0.19861769676208496,
      "learning_rate": 4.950264532181215e-06,
      "loss": 0.0566,
      "num_input_tokens_seen": 28473760,
      "step": 31850
    },
    {
      "epoch": 8.407285205226342,
      "grad_norm": 0.33737319707870483,
      "learning_rate": 4.944401723840433e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 28478592,
      "step": 31855
    },
    {
      "epoch": 8.408604988781839,
      "grad_norm": 0.2275693416595459,
      "learning_rate": 4.938542008324942e-06,
      "loss": 0.0418,
      "num_input_tokens_seen": 28482688,
      "step": 31860
    },
    {
      "epoch": 8.409924772337337,
      "grad_norm": 0.055784787982702255,
      "learning_rate": 4.9326853865383855e-06,
      "loss": 0.0976,
      "num_input_tokens_seen": 28487136,
      "step": 31865
    },
    {
      "epoch": 8.411244555892834,
      "grad_norm": 0.3049996495246887,
      "learning_rate": 4.926831859383918e-06,
      "loss": 0.0372,
      "num_input_tokens_seen": 28491680,
      "step": 31870
    },
    {
      "epoch": 8.41256433944833,
      "grad_norm": 0.07680006325244904,
      "learning_rate": 4.92098142776424e-06,
      "loss": 0.0623,
      "num_input_tokens_seen": 28496320,
      "step": 31875
    },
    {
      "epoch": 8.413884123003827,
      "grad_norm": 0.37665557861328125,
      "learning_rate": 4.91513409258155e-06,
      "loss": 0.0488,
      "num_input_tokens_seen": 28500928,
      "step": 31880
    },
    {
      "epoch": 8.415203906559324,
      "grad_norm": 0.3817359507083893,
      "learning_rate": 4.909289854737581e-06,
      "loss": 0.0512,
      "num_input_tokens_seen": 28505152,
      "step": 31885
    },
    {
      "epoch": 8.416523690114822,
      "grad_norm": 0.3784540295600891,
      "learning_rate": 4.903448715133602e-06,
      "loss": 0.042,
      "num_input_tokens_seen": 28509632,
      "step": 31890
    },
    {
      "epoch": 8.417843473670318,
      "grad_norm": 0.08142757415771484,
      "learning_rate": 4.897610674670372e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 28514336,
      "step": 31895
    },
    {
      "epoch": 8.419163257225815,
      "grad_norm": 0.06685979664325714,
      "learning_rate": 4.8917757342482e-06,
      "loss": 0.0236,
      "num_input_tokens_seen": 28518880,
      "step": 31900
    },
    {
      "epoch": 8.420483040781312,
      "grad_norm": 0.29032984375953674,
      "learning_rate": 4.885943894766909e-06,
      "loss": 0.0605,
      "num_input_tokens_seen": 28523232,
      "step": 31905
    },
    {
      "epoch": 8.421802824336808,
      "grad_norm": 0.377621591091156,
      "learning_rate": 4.880115157125842e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 28527776,
      "step": 31910
    },
    {
      "epoch": 8.423122607892306,
      "grad_norm": 0.20871679484844208,
      "learning_rate": 4.874289522223857e-06,
      "loss": 0.0925,
      "num_input_tokens_seen": 28532384,
      "step": 31915
    },
    {
      "epoch": 8.424442391447803,
      "grad_norm": 0.05319749936461449,
      "learning_rate": 4.868466990959339e-06,
      "loss": 0.0592,
      "num_input_tokens_seen": 28536992,
      "step": 31920
    },
    {
      "epoch": 8.4257621750033,
      "grad_norm": 0.1837724894285202,
      "learning_rate": 4.8626475642301964e-06,
      "loss": 0.0606,
      "num_input_tokens_seen": 28541600,
      "step": 31925
    },
    {
      "epoch": 8.427081958558796,
      "grad_norm": 0.3020378053188324,
      "learning_rate": 4.856831242933871e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 28545696,
      "step": 31930
    },
    {
      "epoch": 8.428401742114293,
      "grad_norm": 0.20873194932937622,
      "learning_rate": 4.851018027967294e-06,
      "loss": 0.0799,
      "num_input_tokens_seen": 28550048,
      "step": 31935
    },
    {
      "epoch": 8.429721525669791,
      "grad_norm": 0.03155635669827461,
      "learning_rate": 4.845207920226946e-06,
      "loss": 0.1012,
      "num_input_tokens_seen": 28554560,
      "step": 31940
    },
    {
      "epoch": 8.431041309225288,
      "grad_norm": 0.3218725025653839,
      "learning_rate": 4.839400920608825e-06,
      "loss": 0.0996,
      "num_input_tokens_seen": 28559008,
      "step": 31945
    },
    {
      "epoch": 8.432361092780784,
      "grad_norm": 0.28791195154190063,
      "learning_rate": 4.83359703000843e-06,
      "loss": 0.0647,
      "num_input_tokens_seen": 28563392,
      "step": 31950
    },
    {
      "epoch": 8.43368087633628,
      "grad_norm": 0.26693063974380493,
      "learning_rate": 4.827796249320804e-06,
      "loss": 0.0689,
      "num_input_tokens_seen": 28567808,
      "step": 31955
    },
    {
      "epoch": 8.435000659891777,
      "grad_norm": 0.07012023776769638,
      "learning_rate": 4.82199857944049e-06,
      "loss": 0.0545,
      "num_input_tokens_seen": 28572000,
      "step": 31960
    },
    {
      "epoch": 8.436320443447274,
      "grad_norm": 0.1225777342915535,
      "learning_rate": 4.8162040212615695e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 28576448,
      "step": 31965
    },
    {
      "epoch": 8.437640227002772,
      "grad_norm": 0.3847530484199524,
      "learning_rate": 4.810412575677639e-06,
      "loss": 0.066,
      "num_input_tokens_seen": 28581120,
      "step": 31970
    },
    {
      "epoch": 8.438960010558269,
      "grad_norm": 0.17505361139774323,
      "learning_rate": 4.804624243581801e-06,
      "loss": 0.0857,
      "num_input_tokens_seen": 28585184,
      "step": 31975
    },
    {
      "epoch": 8.440279794113765,
      "grad_norm": 0.22402746975421906,
      "learning_rate": 4.798839025866703e-06,
      "loss": 0.0603,
      "num_input_tokens_seen": 28589600,
      "step": 31980
    },
    {
      "epoch": 8.441599577669262,
      "grad_norm": 0.2544792592525482,
      "learning_rate": 4.793056923424491e-06,
      "loss": 0.0905,
      "num_input_tokens_seen": 28594240,
      "step": 31985
    },
    {
      "epoch": 8.442919361224758,
      "grad_norm": 0.2715524137020111,
      "learning_rate": 4.78727793714683e-06,
      "loss": 0.1232,
      "num_input_tokens_seen": 28598528,
      "step": 31990
    },
    {
      "epoch": 8.444239144780257,
      "grad_norm": 0.398752361536026,
      "learning_rate": 4.7815020679249285e-06,
      "loss": 0.0967,
      "num_input_tokens_seen": 28602912,
      "step": 31995
    },
    {
      "epoch": 8.445558928335753,
      "grad_norm": 0.22669674456119537,
      "learning_rate": 4.775729316649483e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 28607296,
      "step": 32000
    },
    {
      "epoch": 8.445558928335753,
      "eval_loss": 0.06695601344108582,
      "eval_runtime": 64.7583,
      "eval_samples_per_second": 104.002,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 28607296,
      "step": 32000
    },
    {
      "epoch": 8.44687871189125,
      "grad_norm": 0.19208596646785736,
      "learning_rate": 4.769959684210728e-06,
      "loss": 0.039,
      "num_input_tokens_seen": 28611648,
      "step": 32005
    },
    {
      "epoch": 8.448198495446746,
      "grad_norm": 0.39272093772888184,
      "learning_rate": 4.764193171498426e-06,
      "loss": 0.0582,
      "num_input_tokens_seen": 28616160,
      "step": 32010
    },
    {
      "epoch": 8.449518279002243,
      "grad_norm": 0.1660929173231125,
      "learning_rate": 4.75842977940183e-06,
      "loss": 0.0306,
      "num_input_tokens_seen": 28620352,
      "step": 32015
    },
    {
      "epoch": 8.450838062557741,
      "grad_norm": 0.15429633855819702,
      "learning_rate": 4.752669508809729e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 28625024,
      "step": 32020
    },
    {
      "epoch": 8.452157846113238,
      "grad_norm": 0.11691737174987793,
      "learning_rate": 4.746912360610445e-06,
      "loss": 0.059,
      "num_input_tokens_seen": 28629472,
      "step": 32025
    },
    {
      "epoch": 8.453477629668734,
      "grad_norm": 0.3263303339481354,
      "learning_rate": 4.741158335691781e-06,
      "loss": 0.0466,
      "num_input_tokens_seen": 28633728,
      "step": 32030
    },
    {
      "epoch": 8.454797413224231,
      "grad_norm": 0.17225512862205505,
      "learning_rate": 4.7354074349410994e-06,
      "loss": 0.096,
      "num_input_tokens_seen": 28638208,
      "step": 32035
    },
    {
      "epoch": 8.456117196779728,
      "grad_norm": 0.06650277227163315,
      "learning_rate": 4.729659659245245e-06,
      "loss": 0.0525,
      "num_input_tokens_seen": 28642848,
      "step": 32040
    },
    {
      "epoch": 8.457436980335226,
      "grad_norm": 0.25759896636009216,
      "learning_rate": 4.723915009490601e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 28647104,
      "step": 32045
    },
    {
      "epoch": 8.458756763890722,
      "grad_norm": 0.24824389815330505,
      "learning_rate": 4.718173486563077e-06,
      "loss": 0.0282,
      "num_input_tokens_seen": 28651488,
      "step": 32050
    },
    {
      "epoch": 8.460076547446219,
      "grad_norm": 0.08176346868276596,
      "learning_rate": 4.71243509134808e-06,
      "loss": 0.075,
      "num_input_tokens_seen": 28655872,
      "step": 32055
    },
    {
      "epoch": 8.461396331001716,
      "grad_norm": 0.2675381302833557,
      "learning_rate": 4.706699824730532e-06,
      "loss": 0.0472,
      "num_input_tokens_seen": 28660448,
      "step": 32060
    },
    {
      "epoch": 8.462716114557212,
      "grad_norm": 0.16624954342842102,
      "learning_rate": 4.700967687594901e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 28664896,
      "step": 32065
    },
    {
      "epoch": 8.464035898112709,
      "grad_norm": 0.3976919651031494,
      "learning_rate": 4.69523868082514e-06,
      "loss": 0.1168,
      "num_input_tokens_seen": 28669504,
      "step": 32070
    },
    {
      "epoch": 8.465355681668207,
      "grad_norm": 0.2948966920375824,
      "learning_rate": 4.689512805304747e-06,
      "loss": 0.0699,
      "num_input_tokens_seen": 28674016,
      "step": 32075
    },
    {
      "epoch": 8.466675465223704,
      "grad_norm": 0.16364853084087372,
      "learning_rate": 4.683790061916707e-06,
      "loss": 0.0632,
      "num_input_tokens_seen": 28678208,
      "step": 32080
    },
    {
      "epoch": 8.4679952487792,
      "grad_norm": 0.2183244228363037,
      "learning_rate": 4.678070451543551e-06,
      "loss": 0.0397,
      "num_input_tokens_seen": 28682752,
      "step": 32085
    },
    {
      "epoch": 8.469315032334697,
      "grad_norm": 0.28774094581604004,
      "learning_rate": 4.6723539750673204e-06,
      "loss": 0.1189,
      "num_input_tokens_seen": 28687200,
      "step": 32090
    },
    {
      "epoch": 8.470634815890193,
      "grad_norm": 0.08767541497945786,
      "learning_rate": 4.666640633369551e-06,
      "loss": 0.0374,
      "num_input_tokens_seen": 28691712,
      "step": 32095
    },
    {
      "epoch": 8.471954599445692,
      "grad_norm": 0.06807275116443634,
      "learning_rate": 4.660930427331323e-06,
      "loss": 0.0341,
      "num_input_tokens_seen": 28696224,
      "step": 32100
    },
    {
      "epoch": 8.473274383001188,
      "grad_norm": 0.3960869312286377,
      "learning_rate": 4.6552233578332244e-06,
      "loss": 0.069,
      "num_input_tokens_seen": 28700832,
      "step": 32105
    },
    {
      "epoch": 8.474594166556685,
      "grad_norm": 0.3786757290363312,
      "learning_rate": 4.649519425755347e-06,
      "loss": 0.075,
      "num_input_tokens_seen": 28705280,
      "step": 32110
    },
    {
      "epoch": 8.475913950112181,
      "grad_norm": 0.2847622334957123,
      "learning_rate": 4.64381863197732e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 28710048,
      "step": 32115
    },
    {
      "epoch": 8.477233733667678,
      "grad_norm": 0.2357923984527588,
      "learning_rate": 4.638120977378269e-06,
      "loss": 0.0758,
      "num_input_tokens_seen": 28714752,
      "step": 32120
    },
    {
      "epoch": 8.478553517223176,
      "grad_norm": 0.39049309492111206,
      "learning_rate": 4.632426462836848e-06,
      "loss": 0.1325,
      "num_input_tokens_seen": 28719296,
      "step": 32125
    },
    {
      "epoch": 8.479873300778673,
      "grad_norm": 0.24942511320114136,
      "learning_rate": 4.626735089231224e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 28723872,
      "step": 32130
    },
    {
      "epoch": 8.48119308433417,
      "grad_norm": 0.11647042632102966,
      "learning_rate": 4.621046857439068e-06,
      "loss": 0.0535,
      "num_input_tokens_seen": 28728672,
      "step": 32135
    },
    {
      "epoch": 8.482512867889666,
      "grad_norm": 0.05813154578208923,
      "learning_rate": 4.615361768337587e-06,
      "loss": 0.0607,
      "num_input_tokens_seen": 28733248,
      "step": 32140
    },
    {
      "epoch": 8.483832651445162,
      "grad_norm": 0.2970067262649536,
      "learning_rate": 4.6096798228034946e-06,
      "loss": 0.082,
      "num_input_tokens_seen": 28737504,
      "step": 32145
    },
    {
      "epoch": 8.48515243500066,
      "grad_norm": 0.04213309288024902,
      "learning_rate": 4.604001021713008e-06,
      "loss": 0.077,
      "num_input_tokens_seen": 28741952,
      "step": 32150
    },
    {
      "epoch": 8.486472218556157,
      "grad_norm": 0.4650309979915619,
      "learning_rate": 4.598325365941883e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 28746656,
      "step": 32155
    },
    {
      "epoch": 8.487792002111654,
      "grad_norm": 0.2871243953704834,
      "learning_rate": 4.5926528563653645e-06,
      "loss": 0.1086,
      "num_input_tokens_seen": 28750976,
      "step": 32160
    },
    {
      "epoch": 8.48911178566715,
      "grad_norm": 0.2560248374938965,
      "learning_rate": 4.5869834938582295e-06,
      "loss": 0.0926,
      "num_input_tokens_seen": 28755520,
      "step": 32165
    },
    {
      "epoch": 8.490431569222647,
      "grad_norm": 0.05088259279727936,
      "learning_rate": 4.581317279294772e-06,
      "loss": 0.0398,
      "num_input_tokens_seen": 28759936,
      "step": 32170
    },
    {
      "epoch": 8.491751352778145,
      "grad_norm": 0.0860687792301178,
      "learning_rate": 4.57565421354878e-06,
      "loss": 0.0399,
      "num_input_tokens_seen": 28764480,
      "step": 32175
    },
    {
      "epoch": 8.493071136333642,
      "grad_norm": 0.13699094951152802,
      "learning_rate": 4.569994297493579e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 28769088,
      "step": 32180
    },
    {
      "epoch": 8.494390919889138,
      "grad_norm": 0.1934330016374588,
      "learning_rate": 4.564337532002002e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 28773696,
      "step": 32185
    },
    {
      "epoch": 8.495710703444635,
      "grad_norm": 0.11536159366369247,
      "learning_rate": 4.55868391794638e-06,
      "loss": 0.0873,
      "num_input_tokens_seen": 28778272,
      "step": 32190
    },
    {
      "epoch": 8.497030487000131,
      "grad_norm": 0.2240523248910904,
      "learning_rate": 4.553033456198588e-06,
      "loss": 0.0216,
      "num_input_tokens_seen": 28782496,
      "step": 32195
    },
    {
      "epoch": 8.49835027055563,
      "grad_norm": 0.19347232580184937,
      "learning_rate": 4.54738614762999e-06,
      "loss": 0.0416,
      "num_input_tokens_seen": 28787040,
      "step": 32200
    },
    {
      "epoch": 8.49835027055563,
      "eval_loss": 0.06675978004932404,
      "eval_runtime": 64.7846,
      "eval_samples_per_second": 103.96,
      "eval_steps_per_second": 25.994,
      "num_input_tokens_seen": 28787040,
      "step": 32200
    },
    {
      "epoch": 8.499670054111126,
      "grad_norm": 0.0561528354883194,
      "learning_rate": 4.541741993111465e-06,
      "loss": 0.03,
      "num_input_tokens_seen": 28791424,
      "step": 32205
    },
    {
      "epoch": 8.500989837666623,
      "grad_norm": 0.11692935228347778,
      "learning_rate": 4.536100993513423e-06,
      "loss": 0.0189,
      "num_input_tokens_seen": 28795808,
      "step": 32210
    },
    {
      "epoch": 8.50230962122212,
      "grad_norm": 0.03888910263776779,
      "learning_rate": 4.530463149705768e-06,
      "loss": 0.0288,
      "num_input_tokens_seen": 28800288,
      "step": 32215
    },
    {
      "epoch": 8.503629404777616,
      "grad_norm": 0.2690803110599518,
      "learning_rate": 4.524828462557934e-06,
      "loss": 0.0787,
      "num_input_tokens_seen": 28804576,
      "step": 32220
    },
    {
      "epoch": 8.504949188333114,
      "grad_norm": 0.04509957507252693,
      "learning_rate": 4.5191969329388625e-06,
      "loss": 0.0657,
      "num_input_tokens_seen": 28809312,
      "step": 32225
    },
    {
      "epoch": 8.50626897188861,
      "grad_norm": 0.2366829365491867,
      "learning_rate": 4.5135685617169965e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 28813504,
      "step": 32230
    },
    {
      "epoch": 8.507588755444107,
      "grad_norm": 0.2822888195514679,
      "learning_rate": 4.507943349760313e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 28817920,
      "step": 32235
    },
    {
      "epoch": 8.508908538999604,
      "grad_norm": 0.2680598497390747,
      "learning_rate": 4.502321297936277e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 28822240,
      "step": 32240
    },
    {
      "epoch": 8.5102283225551,
      "grad_norm": 0.19376246631145477,
      "learning_rate": 4.496702407111888e-06,
      "loss": 0.0915,
      "num_input_tokens_seen": 28826848,
      "step": 32245
    },
    {
      "epoch": 8.511548106110597,
      "grad_norm": 0.08419783413410187,
      "learning_rate": 4.491086678153653e-06,
      "loss": 0.0405,
      "num_input_tokens_seen": 28831808,
      "step": 32250
    },
    {
      "epoch": 8.512867889666095,
      "grad_norm": 0.16952945291996002,
      "learning_rate": 4.485474111927579e-06,
      "loss": 0.0542,
      "num_input_tokens_seen": 28836224,
      "step": 32255
    },
    {
      "epoch": 8.514187673221592,
      "grad_norm": 0.5525137782096863,
      "learning_rate": 4.479864709299197e-06,
      "loss": 0.0905,
      "num_input_tokens_seen": 28840832,
      "step": 32260
    },
    {
      "epoch": 8.515507456777089,
      "grad_norm": 0.13719230890274048,
      "learning_rate": 4.474258471133555e-06,
      "loss": 0.0303,
      "num_input_tokens_seen": 28845504,
      "step": 32265
    },
    {
      "epoch": 8.516827240332585,
      "grad_norm": 0.04424746707081795,
      "learning_rate": 4.4686553982952014e-06,
      "loss": 0.0218,
      "num_input_tokens_seen": 28849920,
      "step": 32270
    },
    {
      "epoch": 8.518147023888082,
      "grad_norm": 0.06465354561805725,
      "learning_rate": 4.463055491648191e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 28854272,
      "step": 32275
    },
    {
      "epoch": 8.51946680744358,
      "grad_norm": 0.4480089843273163,
      "learning_rate": 4.457458752056112e-06,
      "loss": 0.1041,
      "num_input_tokens_seen": 28858784,
      "step": 32280
    },
    {
      "epoch": 8.520786590999077,
      "grad_norm": 0.05316169559955597,
      "learning_rate": 4.451865180382042e-06,
      "loss": 0.0508,
      "num_input_tokens_seen": 28863520,
      "step": 32285
    },
    {
      "epoch": 8.522106374554573,
      "grad_norm": 0.07662070542573929,
      "learning_rate": 4.4462747774885936e-06,
      "loss": 0.0408,
      "num_input_tokens_seen": 28867840,
      "step": 32290
    },
    {
      "epoch": 8.52342615811007,
      "grad_norm": 0.42241764068603516,
      "learning_rate": 4.440687544237859e-06,
      "loss": 0.0768,
      "num_input_tokens_seen": 28872192,
      "step": 32295
    },
    {
      "epoch": 8.524745941665566,
      "grad_norm": 0.2723045349121094,
      "learning_rate": 4.435103481491471e-06,
      "loss": 0.0362,
      "num_input_tokens_seen": 28876640,
      "step": 32300
    },
    {
      "epoch": 8.526065725221065,
      "grad_norm": 0.04508015513420105,
      "learning_rate": 4.429522590110569e-06,
      "loss": 0.0231,
      "num_input_tokens_seen": 28881632,
      "step": 32305
    },
    {
      "epoch": 8.527385508776561,
      "grad_norm": 0.12172966450452805,
      "learning_rate": 4.423944870955779e-06,
      "loss": 0.0481,
      "num_input_tokens_seen": 28885792,
      "step": 32310
    },
    {
      "epoch": 8.528705292332058,
      "grad_norm": 0.10600674152374268,
      "learning_rate": 4.418370324887272e-06,
      "loss": 0.0596,
      "num_input_tokens_seen": 28890464,
      "step": 32315
    },
    {
      "epoch": 8.530025075887554,
      "grad_norm": 0.06982612609863281,
      "learning_rate": 4.412798952764699e-06,
      "loss": 0.037,
      "num_input_tokens_seen": 28894816,
      "step": 32320
    },
    {
      "epoch": 8.53134485944305,
      "grad_norm": 0.4049108624458313,
      "learning_rate": 4.407230755447245e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 28899200,
      "step": 32325
    },
    {
      "epoch": 8.532664642998547,
      "grad_norm": 0.29336977005004883,
      "learning_rate": 4.401665733793598e-06,
      "loss": 0.0729,
      "num_input_tokens_seen": 28903392,
      "step": 32330
    },
    {
      "epoch": 8.533984426554046,
      "grad_norm": 0.23175837099552155,
      "learning_rate": 4.3961038886619425e-06,
      "loss": 0.0586,
      "num_input_tokens_seen": 28907872,
      "step": 32335
    },
    {
      "epoch": 8.535304210109542,
      "grad_norm": 0.15440090000629425,
      "learning_rate": 4.39054522091e-06,
      "loss": 0.0789,
      "num_input_tokens_seen": 28912480,
      "step": 32340
    },
    {
      "epoch": 8.536623993665039,
      "grad_norm": 0.24717773497104645,
      "learning_rate": 4.384989731394979e-06,
      "loss": 0.1021,
      "num_input_tokens_seen": 28916992,
      "step": 32345
    },
    {
      "epoch": 8.537943777220535,
      "grad_norm": 0.19682781398296356,
      "learning_rate": 4.379437420973598e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 28921376,
      "step": 32350
    },
    {
      "epoch": 8.539263560776032,
      "grad_norm": 0.07588683813810349,
      "learning_rate": 4.373888290502107e-06,
      "loss": 0.0332,
      "num_input_tokens_seen": 28926080,
      "step": 32355
    },
    {
      "epoch": 8.54058334433153,
      "grad_norm": 0.5153923034667969,
      "learning_rate": 4.36834234083624e-06,
      "loss": 0.1028,
      "num_input_tokens_seen": 28930624,
      "step": 32360
    },
    {
      "epoch": 8.541903127887027,
      "grad_norm": 0.2605419158935547,
      "learning_rate": 4.362799572831258e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 28935040,
      "step": 32365
    },
    {
      "epoch": 8.543222911442523,
      "grad_norm": 0.056142859160900116,
      "learning_rate": 4.35725998734193e-06,
      "loss": 0.0844,
      "num_input_tokens_seen": 28939616,
      "step": 32370
    },
    {
      "epoch": 8.54454269499802,
      "grad_norm": 0.1018192395567894,
      "learning_rate": 4.3517235852225195e-06,
      "loss": 0.0972,
      "num_input_tokens_seen": 28943968,
      "step": 32375
    },
    {
      "epoch": 8.545862478553516,
      "grad_norm": 0.17404715716838837,
      "learning_rate": 4.346190367326822e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 28948576,
      "step": 32380
    },
    {
      "epoch": 8.547182262109015,
      "grad_norm": 0.262067049741745,
      "learning_rate": 4.340660334508115e-06,
      "loss": 0.0353,
      "num_input_tokens_seen": 28952992,
      "step": 32385
    },
    {
      "epoch": 8.548502045664511,
      "grad_norm": 0.22257110476493835,
      "learning_rate": 4.335133487619206e-06,
      "loss": 0.1131,
      "num_input_tokens_seen": 28957344,
      "step": 32390
    },
    {
      "epoch": 8.549821829220008,
      "grad_norm": 0.11056710034608841,
      "learning_rate": 4.329609827512409e-06,
      "loss": 0.052,
      "num_input_tokens_seen": 28962080,
      "step": 32395
    },
    {
      "epoch": 8.551141612775504,
      "grad_norm": 0.09083539992570877,
      "learning_rate": 4.324089355039531e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 28966560,
      "step": 32400
    },
    {
      "epoch": 8.551141612775504,
      "eval_loss": 0.06670115888118744,
      "eval_runtime": 64.7671,
      "eval_samples_per_second": 103.988,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 28966560,
      "step": 32400
    },
    {
      "epoch": 8.552461396331001,
      "grad_norm": 0.20540232956409454,
      "learning_rate": 4.3185720710519075e-06,
      "loss": 0.042,
      "num_input_tokens_seen": 28970944,
      "step": 32405
    },
    {
      "epoch": 8.5537811798865,
      "grad_norm": 0.14135994017124176,
      "learning_rate": 4.3130579764003724e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 28975488,
      "step": 32410
    },
    {
      "epoch": 8.555100963441996,
      "grad_norm": 0.14959384500980377,
      "learning_rate": 4.307547071935267e-06,
      "loss": 0.1375,
      "num_input_tokens_seen": 28980000,
      "step": 32415
    },
    {
      "epoch": 8.556420746997492,
      "grad_norm": 0.22653919458389282,
      "learning_rate": 4.302039358506435e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 28984224,
      "step": 32420
    },
    {
      "epoch": 8.557740530552989,
      "grad_norm": 0.10717340558767319,
      "learning_rate": 4.296534836963245e-06,
      "loss": 0.0707,
      "num_input_tokens_seen": 28988544,
      "step": 32425
    },
    {
      "epoch": 8.559060314108486,
      "grad_norm": 0.1399032026529312,
      "learning_rate": 4.291033508154555e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 28992928,
      "step": 32430
    },
    {
      "epoch": 8.560380097663984,
      "grad_norm": 0.22474636137485504,
      "learning_rate": 4.285535372928748e-06,
      "loss": 0.0903,
      "num_input_tokens_seen": 28997408,
      "step": 32435
    },
    {
      "epoch": 8.56169988121948,
      "grad_norm": 0.522908091545105,
      "learning_rate": 4.280040432133695e-06,
      "loss": 0.0907,
      "num_input_tokens_seen": 29002112,
      "step": 32440
    },
    {
      "epoch": 8.563019664774977,
      "grad_norm": 0.13207074999809265,
      "learning_rate": 4.274548686616789e-06,
      "loss": 0.0789,
      "num_input_tokens_seen": 29006368,
      "step": 32445
    },
    {
      "epoch": 8.564339448330474,
      "grad_norm": 0.2553294003009796,
      "learning_rate": 4.2690601372249364e-06,
      "loss": 0.0981,
      "num_input_tokens_seen": 29010784,
      "step": 32450
    },
    {
      "epoch": 8.56565923188597,
      "grad_norm": 0.35335758328437805,
      "learning_rate": 4.263574784804525e-06,
      "loss": 0.0477,
      "num_input_tokens_seen": 29014848,
      "step": 32455
    },
    {
      "epoch": 8.566979015441468,
      "grad_norm": 0.026406941935420036,
      "learning_rate": 4.258092630201479e-06,
      "loss": 0.0189,
      "num_input_tokens_seen": 29019392,
      "step": 32460
    },
    {
      "epoch": 8.568298798996965,
      "grad_norm": 0.06469951570034027,
      "learning_rate": 4.252613674261202e-06,
      "loss": 0.0283,
      "num_input_tokens_seen": 29023616,
      "step": 32465
    },
    {
      "epoch": 8.569618582552462,
      "grad_norm": 0.19222770631313324,
      "learning_rate": 4.2471379178286224e-06,
      "loss": 0.0507,
      "num_input_tokens_seen": 29028128,
      "step": 32470
    },
    {
      "epoch": 8.570938366107958,
      "grad_norm": 0.4107562303543091,
      "learning_rate": 4.241665361748181e-06,
      "loss": 0.0503,
      "num_input_tokens_seen": 29032800,
      "step": 32475
    },
    {
      "epoch": 8.572258149663455,
      "grad_norm": 0.11473802477121353,
      "learning_rate": 4.2361960068637994e-06,
      "loss": 0.0319,
      "num_input_tokens_seen": 29037152,
      "step": 32480
    },
    {
      "epoch": 8.573577933218953,
      "grad_norm": 0.254913866519928,
      "learning_rate": 4.230729854018933e-06,
      "loss": 0.0509,
      "num_input_tokens_seen": 29041568,
      "step": 32485
    },
    {
      "epoch": 8.57489771677445,
      "grad_norm": 0.1560724526643753,
      "learning_rate": 4.225266904056521e-06,
      "loss": 0.0645,
      "num_input_tokens_seen": 29045824,
      "step": 32490
    },
    {
      "epoch": 8.576217500329946,
      "grad_norm": 0.18119493126869202,
      "learning_rate": 4.21980715781903e-06,
      "loss": 0.0498,
      "num_input_tokens_seen": 29050368,
      "step": 32495
    },
    {
      "epoch": 8.577537283885443,
      "grad_norm": 0.350727915763855,
      "learning_rate": 4.214350616148416e-06,
      "loss": 0.047,
      "num_input_tokens_seen": 29055008,
      "step": 32500
    },
    {
      "epoch": 8.57885706744094,
      "grad_norm": 0.11176837980747223,
      "learning_rate": 4.20889727988614e-06,
      "loss": 0.1036,
      "num_input_tokens_seen": 29059776,
      "step": 32505
    },
    {
      "epoch": 8.580176850996436,
      "grad_norm": 0.15640810132026672,
      "learning_rate": 4.20344714987318e-06,
      "loss": 0.0342,
      "num_input_tokens_seen": 29064480,
      "step": 32510
    },
    {
      "epoch": 8.581496634551934,
      "grad_norm": 0.1420358568429947,
      "learning_rate": 4.198000226950022e-06,
      "loss": 0.0794,
      "num_input_tokens_seen": 29068864,
      "step": 32515
    },
    {
      "epoch": 8.58281641810743,
      "grad_norm": 0.5217701196670532,
      "learning_rate": 4.192556511956635e-06,
      "loss": 0.1003,
      "num_input_tokens_seen": 29073344,
      "step": 32520
    },
    {
      "epoch": 8.584136201662927,
      "grad_norm": 0.31086355447769165,
      "learning_rate": 4.18711600573252e-06,
      "loss": 0.0765,
      "num_input_tokens_seen": 29077920,
      "step": 32525
    },
    {
      "epoch": 8.585455985218424,
      "grad_norm": 0.23521286249160767,
      "learning_rate": 4.181678709116671e-06,
      "loss": 0.0699,
      "num_input_tokens_seen": 29082304,
      "step": 32530
    },
    {
      "epoch": 8.58677576877392,
      "grad_norm": 0.012565506622195244,
      "learning_rate": 4.1762446229475785e-06,
      "loss": 0.0463,
      "num_input_tokens_seen": 29087008,
      "step": 32535
    },
    {
      "epoch": 8.588095552329419,
      "grad_norm": 0.38530752062797546,
      "learning_rate": 4.17081374806326e-06,
      "loss": 0.0535,
      "num_input_tokens_seen": 29091584,
      "step": 32540
    },
    {
      "epoch": 8.589415335884915,
      "grad_norm": 0.3136151134967804,
      "learning_rate": 4.165386085301212e-06,
      "loss": 0.0753,
      "num_input_tokens_seen": 29096128,
      "step": 32545
    },
    {
      "epoch": 8.590735119440412,
      "grad_norm": 0.16378270089626312,
      "learning_rate": 4.1599616354984525e-06,
      "loss": 0.0734,
      "num_input_tokens_seen": 29100448,
      "step": 32550
    },
    {
      "epoch": 8.592054902995908,
      "grad_norm": 0.09566398710012436,
      "learning_rate": 4.154540399491508e-06,
      "loss": 0.086,
      "num_input_tokens_seen": 29104800,
      "step": 32555
    },
    {
      "epoch": 8.593374686551405,
      "grad_norm": 0.32277849316596985,
      "learning_rate": 4.149122378116394e-06,
      "loss": 0.0318,
      "num_input_tokens_seen": 29109088,
      "step": 32560
    },
    {
      "epoch": 8.594694470106903,
      "grad_norm": 0.1684873402118683,
      "learning_rate": 4.14370757220863e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 29113696,
      "step": 32565
    },
    {
      "epoch": 8.5960142536624,
      "grad_norm": 0.2711392343044281,
      "learning_rate": 4.138295982603263e-06,
      "loss": 0.0478,
      "num_input_tokens_seen": 29118272,
      "step": 32570
    },
    {
      "epoch": 8.597334037217896,
      "grad_norm": 0.051284972578287125,
      "learning_rate": 4.132887610134814e-06,
      "loss": 0.0337,
      "num_input_tokens_seen": 29122976,
      "step": 32575
    },
    {
      "epoch": 8.598653820773393,
      "grad_norm": 0.06220335513353348,
      "learning_rate": 4.127482455637335e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 29127104,
      "step": 32580
    },
    {
      "epoch": 8.59997360432889,
      "grad_norm": 0.13155677914619446,
      "learning_rate": 4.1220805199443545e-06,
      "loss": 0.0238,
      "num_input_tokens_seen": 29131424,
      "step": 32585
    },
    {
      "epoch": 8.601293387884386,
      "grad_norm": 0.1772831380367279,
      "learning_rate": 4.116681803888925e-06,
      "loss": 0.0982,
      "num_input_tokens_seen": 29136064,
      "step": 32590
    },
    {
      "epoch": 8.602613171439884,
      "grad_norm": 0.03344983607530594,
      "learning_rate": 4.111286308303605e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 29140224,
      "step": 32595
    },
    {
      "epoch": 8.603932954995381,
      "grad_norm": 0.24147240817546844,
      "learning_rate": 4.105894034020433e-06,
      "loss": 0.0619,
      "num_input_tokens_seen": 29144544,
      "step": 32600
    },
    {
      "epoch": 8.603932954995381,
      "eval_loss": 0.06694675236940384,
      "eval_runtime": 64.7223,
      "eval_samples_per_second": 104.06,
      "eval_steps_per_second": 26.019,
      "num_input_tokens_seen": 29144544,
      "step": 32600
    },
    {
      "epoch": 8.605252738550877,
      "grad_norm": 0.3302669823169708,
      "learning_rate": 4.100504981870975e-06,
      "loss": 0.0307,
      "num_input_tokens_seen": 29149056,
      "step": 32605
    },
    {
      "epoch": 8.606572522106374,
      "grad_norm": 0.31716346740722656,
      "learning_rate": 4.0951191526862915e-06,
      "loss": 0.066,
      "num_input_tokens_seen": 29153504,
      "step": 32610
    },
    {
      "epoch": 8.60789230566187,
      "grad_norm": 0.18275701999664307,
      "learning_rate": 4.089736547296938e-06,
      "loss": 0.026,
      "num_input_tokens_seen": 29157728,
      "step": 32615
    },
    {
      "epoch": 8.609212089217369,
      "grad_norm": 0.0660560131072998,
      "learning_rate": 4.08435716653299e-06,
      "loss": 0.0543,
      "num_input_tokens_seen": 29161984,
      "step": 32620
    },
    {
      "epoch": 8.610531872772865,
      "grad_norm": 0.2171325385570526,
      "learning_rate": 4.0789810112240005e-06,
      "loss": 0.0534,
      "num_input_tokens_seen": 29166144,
      "step": 32625
    },
    {
      "epoch": 8.611851656328362,
      "grad_norm": 0.20636442303657532,
      "learning_rate": 4.073608082199057e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 29170656,
      "step": 32630
    },
    {
      "epoch": 8.613171439883859,
      "grad_norm": 0.30510881543159485,
      "learning_rate": 4.068238380286718e-06,
      "loss": 0.0713,
      "num_input_tokens_seen": 29175424,
      "step": 32635
    },
    {
      "epoch": 8.614491223439355,
      "grad_norm": 0.17648358643054962,
      "learning_rate": 4.062871906315072e-06,
      "loss": 0.0574,
      "num_input_tokens_seen": 29179840,
      "step": 32640
    },
    {
      "epoch": 8.615811006994853,
      "grad_norm": 0.2471001297235489,
      "learning_rate": 4.057508661111686e-06,
      "loss": 0.0786,
      "num_input_tokens_seen": 29184320,
      "step": 32645
    },
    {
      "epoch": 8.61713079055035,
      "grad_norm": 0.07106463611125946,
      "learning_rate": 4.052148645503648e-06,
      "loss": 0.0405,
      "num_input_tokens_seen": 29188704,
      "step": 32650
    },
    {
      "epoch": 8.618450574105847,
      "grad_norm": 0.06340433657169342,
      "learning_rate": 4.046791860317531e-06,
      "loss": 0.0282,
      "num_input_tokens_seen": 29193472,
      "step": 32655
    },
    {
      "epoch": 8.619770357661343,
      "grad_norm": 0.19910138845443726,
      "learning_rate": 4.041438306379431e-06,
      "loss": 0.0515,
      "num_input_tokens_seen": 29197760,
      "step": 32660
    },
    {
      "epoch": 8.62109014121684,
      "grad_norm": 0.10229451954364777,
      "learning_rate": 4.036087984514916e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 29202272,
      "step": 32665
    },
    {
      "epoch": 8.622409924772338,
      "grad_norm": 0.07130244374275208,
      "learning_rate": 4.030740895549084e-06,
      "loss": 0.0917,
      "num_input_tokens_seen": 29206496,
      "step": 32670
    },
    {
      "epoch": 8.623729708327835,
      "grad_norm": 0.3523802161216736,
      "learning_rate": 4.025397040306531e-06,
      "loss": 0.0774,
      "num_input_tokens_seen": 29211200,
      "step": 32675
    },
    {
      "epoch": 8.625049491883331,
      "grad_norm": 0.1125662624835968,
      "learning_rate": 4.0200564196113285e-06,
      "loss": 0.1065,
      "num_input_tokens_seen": 29215584,
      "step": 32680
    },
    {
      "epoch": 8.626369275438828,
      "grad_norm": 0.14605417847633362,
      "learning_rate": 4.014719034287079e-06,
      "loss": 0.0402,
      "num_input_tokens_seen": 29219808,
      "step": 32685
    },
    {
      "epoch": 8.627689058994324,
      "grad_norm": 0.3915920555591583,
      "learning_rate": 4.0093848851568775e-06,
      "loss": 0.1012,
      "num_input_tokens_seen": 29224224,
      "step": 32690
    },
    {
      "epoch": 8.629008842549823,
      "grad_norm": 0.17859698832035065,
      "learning_rate": 4.004053973043304e-06,
      "loss": 0.1052,
      "num_input_tokens_seen": 29228576,
      "step": 32695
    },
    {
      "epoch": 8.63032862610532,
      "grad_norm": 0.19770224392414093,
      "learning_rate": 3.998726298768465e-06,
      "loss": 0.0946,
      "num_input_tokens_seen": 29233216,
      "step": 32700
    },
    {
      "epoch": 8.631648409660816,
      "grad_norm": 0.16487587988376617,
      "learning_rate": 3.99340186315395e-06,
      "loss": 0.095,
      "num_input_tokens_seen": 29237760,
      "step": 32705
    },
    {
      "epoch": 8.632968193216312,
      "grad_norm": 0.3935829997062683,
      "learning_rate": 3.988080667020849e-06,
      "loss": 0.0706,
      "num_input_tokens_seen": 29242144,
      "step": 32710
    },
    {
      "epoch": 8.634287976771809,
      "grad_norm": 0.3497520089149475,
      "learning_rate": 3.982762711189766e-06,
      "loss": 0.0677,
      "num_input_tokens_seen": 29246944,
      "step": 32715
    },
    {
      "epoch": 8.635607760327307,
      "grad_norm": 0.3890731930732727,
      "learning_rate": 3.977447996480785e-06,
      "loss": 0.067,
      "num_input_tokens_seen": 29251200,
      "step": 32720
    },
    {
      "epoch": 8.636927543882804,
      "grad_norm": 0.034144505858421326,
      "learning_rate": 3.97213652371351e-06,
      "loss": 0.0623,
      "num_input_tokens_seen": 29256000,
      "step": 32725
    },
    {
      "epoch": 8.6382473274383,
      "grad_norm": 0.17078763246536255,
      "learning_rate": 3.966828293707042e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 29260032,
      "step": 32730
    },
    {
      "epoch": 8.639567110993797,
      "grad_norm": 0.3603339195251465,
      "learning_rate": 3.961523307279963e-06,
      "loss": 0.1263,
      "num_input_tokens_seen": 29264480,
      "step": 32735
    },
    {
      "epoch": 8.640886894549293,
      "grad_norm": 0.034257672727108,
      "learning_rate": 3.956221565250382e-06,
      "loss": 0.0732,
      "num_input_tokens_seen": 29268864,
      "step": 32740
    },
    {
      "epoch": 8.642206678104792,
      "grad_norm": 0.09919828176498413,
      "learning_rate": 3.950923068435883e-06,
      "loss": 0.0503,
      "num_input_tokens_seen": 29273216,
      "step": 32745
    },
    {
      "epoch": 8.643526461660288,
      "grad_norm": 0.3247840106487274,
      "learning_rate": 3.945627817653566e-06,
      "loss": 0.0707,
      "num_input_tokens_seen": 29277760,
      "step": 32750
    },
    {
      "epoch": 8.644846245215785,
      "grad_norm": 0.46490582823753357,
      "learning_rate": 3.9403358137200335e-06,
      "loss": 0.058,
      "num_input_tokens_seen": 29281984,
      "step": 32755
    },
    {
      "epoch": 8.646166028771281,
      "grad_norm": 0.14247538149356842,
      "learning_rate": 3.9350470574513605e-06,
      "loss": 0.086,
      "num_input_tokens_seen": 29286496,
      "step": 32760
    },
    {
      "epoch": 8.647485812326778,
      "grad_norm": 0.34124991297721863,
      "learning_rate": 3.9297615496631525e-06,
      "loss": 0.0583,
      "num_input_tokens_seen": 29290944,
      "step": 32765
    },
    {
      "epoch": 8.648805595882274,
      "grad_norm": 0.09631524235010147,
      "learning_rate": 3.924479291170505e-06,
      "loss": 0.052,
      "num_input_tokens_seen": 29295328,
      "step": 32770
    },
    {
      "epoch": 8.650125379437773,
      "grad_norm": 0.15627343952655792,
      "learning_rate": 3.919200282788002e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 29299840,
      "step": 32775
    },
    {
      "epoch": 8.65144516299327,
      "grad_norm": 0.23588554561138153,
      "learning_rate": 3.913924525329726e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 29304896,
      "step": 32780
    },
    {
      "epoch": 8.652764946548766,
      "grad_norm": 0.25933998823165894,
      "learning_rate": 3.908652019609279e-06,
      "loss": 0.1102,
      "num_input_tokens_seen": 29309728,
      "step": 32785
    },
    {
      "epoch": 8.654084730104262,
      "grad_norm": 0.05708885192871094,
      "learning_rate": 3.9033827664397364e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 29314112,
      "step": 32790
    },
    {
      "epoch": 8.655404513659759,
      "grad_norm": 0.21943096816539764,
      "learning_rate": 3.898116766633694e-06,
      "loss": 0.052,
      "num_input_tokens_seen": 29318752,
      "step": 32795
    },
    {
      "epoch": 8.656724297215257,
      "grad_norm": 0.5523438453674316,
      "learning_rate": 3.8928540210032225e-06,
      "loss": 0.1309,
      "num_input_tokens_seen": 29323040,
      "step": 32800
    },
    {
      "epoch": 8.656724297215257,
      "eval_loss": 0.06685389578342438,
      "eval_runtime": 64.7446,
      "eval_samples_per_second": 104.024,
      "eval_steps_per_second": 26.01,
      "num_input_tokens_seen": 29323040,
      "step": 32800
    },
    {
      "epoch": 8.658044080770754,
      "grad_norm": 0.27125513553619385,
      "learning_rate": 3.887594530359909e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 29327680,
      "step": 32805
    },
    {
      "epoch": 8.65936386432625,
      "grad_norm": 0.13196101784706116,
      "learning_rate": 3.88233829551484e-06,
      "loss": 0.0372,
      "num_input_tokens_seen": 29332192,
      "step": 32810
    },
    {
      "epoch": 8.660683647881747,
      "grad_norm": 0.052428774535655975,
      "learning_rate": 3.877085317278581e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 29336608,
      "step": 32815
    },
    {
      "epoch": 8.662003431437244,
      "grad_norm": 0.16101226210594177,
      "learning_rate": 3.87183559646122e-06,
      "loss": 0.082,
      "num_input_tokens_seen": 29341248,
      "step": 32820
    },
    {
      "epoch": 8.663323214992742,
      "grad_norm": 0.19853097200393677,
      "learning_rate": 3.866589133872317e-06,
      "loss": 0.0594,
      "num_input_tokens_seen": 29345632,
      "step": 32825
    },
    {
      "epoch": 8.664642998548238,
      "grad_norm": 0.27099448442459106,
      "learning_rate": 3.861345930320948e-06,
      "loss": 0.0854,
      "num_input_tokens_seen": 29350304,
      "step": 32830
    },
    {
      "epoch": 8.665962782103735,
      "grad_norm": 0.23485668003559113,
      "learning_rate": 3.856105986615688e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 29354976,
      "step": 32835
    },
    {
      "epoch": 8.667282565659232,
      "grad_norm": 0.02862987481057644,
      "learning_rate": 3.850869303564589e-06,
      "loss": 0.0183,
      "num_input_tokens_seen": 29359456,
      "step": 32840
    },
    {
      "epoch": 8.668602349214728,
      "grad_norm": 0.058950502425432205,
      "learning_rate": 3.845635881975226e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 29363840,
      "step": 32845
    },
    {
      "epoch": 8.669922132770226,
      "grad_norm": 0.1413072943687439,
      "learning_rate": 3.840405722654647e-06,
      "loss": 0.0311,
      "num_input_tokens_seen": 29368256,
      "step": 32850
    },
    {
      "epoch": 8.671241916325723,
      "grad_norm": 0.042121268808841705,
      "learning_rate": 3.835178826409419e-06,
      "loss": 0.0672,
      "num_input_tokens_seen": 29372512,
      "step": 32855
    },
    {
      "epoch": 8.67256169988122,
      "grad_norm": 0.15638349950313568,
      "learning_rate": 3.8299551940455895e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 29377088,
      "step": 32860
    },
    {
      "epoch": 8.673881483436716,
      "grad_norm": 0.35553884506225586,
      "learning_rate": 3.824734826368703e-06,
      "loss": 0.0501,
      "num_input_tokens_seen": 29381536,
      "step": 32865
    },
    {
      "epoch": 8.675201266992213,
      "grad_norm": 0.5077926516532898,
      "learning_rate": 3.819517724183813e-06,
      "loss": 0.0772,
      "num_input_tokens_seen": 29385888,
      "step": 32870
    },
    {
      "epoch": 8.67652105054771,
      "grad_norm": 0.09993810951709747,
      "learning_rate": 3.8143038882954648e-06,
      "loss": 0.0515,
      "num_input_tokens_seen": 29390400,
      "step": 32875
    },
    {
      "epoch": 8.677840834103208,
      "grad_norm": 0.23732170462608337,
      "learning_rate": 3.8090933195076867e-06,
      "loss": 0.0806,
      "num_input_tokens_seen": 29395136,
      "step": 32880
    },
    {
      "epoch": 8.679160617658704,
      "grad_norm": 0.24257096648216248,
      "learning_rate": 3.8038860186240198e-06,
      "loss": 0.0435,
      "num_input_tokens_seen": 29399648,
      "step": 32885
    },
    {
      "epoch": 8.6804804012142,
      "grad_norm": 0.17076610028743744,
      "learning_rate": 3.7986819864475026e-06,
      "loss": 0.0477,
      "num_input_tokens_seen": 29404064,
      "step": 32890
    },
    {
      "epoch": 8.681800184769697,
      "grad_norm": 0.3292805552482605,
      "learning_rate": 3.793481223780651e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 29408416,
      "step": 32895
    },
    {
      "epoch": 8.683119968325194,
      "grad_norm": 0.28284069895744324,
      "learning_rate": 3.788283731425496e-06,
      "loss": 0.0927,
      "num_input_tokens_seen": 29412768,
      "step": 32900
    },
    {
      "epoch": 8.684439751880692,
      "grad_norm": 0.1924552172422409,
      "learning_rate": 3.7830895101835488e-06,
      "loss": 0.0582,
      "num_input_tokens_seen": 29417376,
      "step": 32905
    },
    {
      "epoch": 8.685759535436189,
      "grad_norm": 0.15752220153808594,
      "learning_rate": 3.7778985608558274e-06,
      "loss": 0.1213,
      "num_input_tokens_seen": 29421952,
      "step": 32910
    },
    {
      "epoch": 8.687079318991685,
      "grad_norm": 0.4603945314884186,
      "learning_rate": 3.7727108842428443e-06,
      "loss": 0.0703,
      "num_input_tokens_seen": 29426208,
      "step": 32915
    },
    {
      "epoch": 8.688399102547182,
      "grad_norm": 0.2732532024383545,
      "learning_rate": 3.7675264811446065e-06,
      "loss": 0.0487,
      "num_input_tokens_seen": 29430752,
      "step": 32920
    },
    {
      "epoch": 8.689718886102678,
      "grad_norm": 0.27809938788414,
      "learning_rate": 3.7623453523605994e-06,
      "loss": 0.0839,
      "num_input_tokens_seen": 29435168,
      "step": 32925
    },
    {
      "epoch": 8.691038669658177,
      "grad_norm": 0.17244605720043182,
      "learning_rate": 3.757167498689834e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 29439776,
      "step": 32930
    },
    {
      "epoch": 8.692358453213673,
      "grad_norm": 0.2921839654445648,
      "learning_rate": 3.7519929209307914e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 29444384,
      "step": 32935
    },
    {
      "epoch": 8.69367823676917,
      "grad_norm": 0.2505919933319092,
      "learning_rate": 3.746821619881463e-06,
      "loss": 0.0341,
      "num_input_tokens_seen": 29449024,
      "step": 32940
    },
    {
      "epoch": 8.694998020324666,
      "grad_norm": 0.3557567000389099,
      "learning_rate": 3.74165359633932e-06,
      "loss": 0.0422,
      "num_input_tokens_seen": 29453728,
      "step": 32945
    },
    {
      "epoch": 8.696317803880163,
      "grad_norm": 0.16278469562530518,
      "learning_rate": 3.736488851101341e-06,
      "loss": 0.0814,
      "num_input_tokens_seen": 29458208,
      "step": 32950
    },
    {
      "epoch": 8.697637587435661,
      "grad_norm": 0.07389740645885468,
      "learning_rate": 3.7313273849640035e-06,
      "loss": 0.046,
      "num_input_tokens_seen": 29462208,
      "step": 32955
    },
    {
      "epoch": 8.698957370991158,
      "grad_norm": 0.15894150733947754,
      "learning_rate": 3.7261691987232533e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 29466752,
      "step": 32960
    },
    {
      "epoch": 8.700277154546654,
      "grad_norm": 0.23662126064300537,
      "learning_rate": 3.7210142931745575e-06,
      "loss": 0.0491,
      "num_input_tokens_seen": 29471360,
      "step": 32965
    },
    {
      "epoch": 8.701596938102151,
      "grad_norm": 0.030453141778707504,
      "learning_rate": 3.7158626691128712e-06,
      "loss": 0.0784,
      "num_input_tokens_seen": 29475584,
      "step": 32970
    },
    {
      "epoch": 8.702916721657648,
      "grad_norm": 0.024369509890675545,
      "learning_rate": 3.710714327332629e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 29479936,
      "step": 32975
    },
    {
      "epoch": 8.704236505213146,
      "grad_norm": 0.21548347175121307,
      "learning_rate": 3.7055692686277815e-06,
      "loss": 0.0989,
      "num_input_tokens_seen": 29484512,
      "step": 32980
    },
    {
      "epoch": 8.705556288768642,
      "grad_norm": 0.3104400038719177,
      "learning_rate": 3.70042749379175e-06,
      "loss": 0.0625,
      "num_input_tokens_seen": 29489056,
      "step": 32985
    },
    {
      "epoch": 8.706876072324139,
      "grad_norm": 0.042929720133543015,
      "learning_rate": 3.6952890036174693e-06,
      "loss": 0.0556,
      "num_input_tokens_seen": 29493472,
      "step": 32990
    },
    {
      "epoch": 8.708195855879636,
      "grad_norm": 0.08773769438266754,
      "learning_rate": 3.690153798897353e-06,
      "loss": 0.1146,
      "num_input_tokens_seen": 29497792,
      "step": 32995
    },
    {
      "epoch": 8.709515639435132,
      "grad_norm": 0.2040826380252838,
      "learning_rate": 3.6850218804233225e-06,
      "loss": 0.0829,
      "num_input_tokens_seen": 29502176,
      "step": 33000
    },
    {
      "epoch": 8.709515639435132,
      "eval_loss": 0.06704409420490265,
      "eval_runtime": 64.7625,
      "eval_samples_per_second": 103.995,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 29502176,
      "step": 33000
    },
    {
      "epoch": 8.71083542299063,
      "grad_norm": 0.2500505745410919,
      "learning_rate": 3.679893248986779e-06,
      "loss": 0.0497,
      "num_input_tokens_seen": 29506560,
      "step": 33005
    },
    {
      "epoch": 8.712155206546127,
      "grad_norm": 0.28358033299446106,
      "learning_rate": 3.6747679053786147e-06,
      "loss": 0.0626,
      "num_input_tokens_seen": 29511008,
      "step": 33010
    },
    {
      "epoch": 8.713474990101624,
      "grad_norm": 0.0624716617166996,
      "learning_rate": 3.669645850389228e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 29515456,
      "step": 33015
    },
    {
      "epoch": 8.71479477365712,
      "grad_norm": 0.20820358395576477,
      "learning_rate": 3.664527084808514e-06,
      "loss": 0.0793,
      "num_input_tokens_seen": 29520416,
      "step": 33020
    },
    {
      "epoch": 8.716114557212617,
      "grad_norm": 0.3063121736049652,
      "learning_rate": 3.6594116094258337e-06,
      "loss": 0.1164,
      "num_input_tokens_seen": 29525024,
      "step": 33025
    },
    {
      "epoch": 8.717434340768115,
      "grad_norm": 0.11276159435510635,
      "learning_rate": 3.6542994250300665e-06,
      "loss": 0.0824,
      "num_input_tokens_seen": 29529600,
      "step": 33030
    },
    {
      "epoch": 8.718754124323612,
      "grad_norm": 0.4005969762802124,
      "learning_rate": 3.6491905324095825e-06,
      "loss": 0.0337,
      "num_input_tokens_seen": 29534304,
      "step": 33035
    },
    {
      "epoch": 8.720073907879108,
      "grad_norm": 0.5101680755615234,
      "learning_rate": 3.644084932352221e-06,
      "loss": 0.0623,
      "num_input_tokens_seen": 29538848,
      "step": 33040
    },
    {
      "epoch": 8.721393691434605,
      "grad_norm": 0.49138447642326355,
      "learning_rate": 3.6389826256453457e-06,
      "loss": 0.097,
      "num_input_tokens_seen": 29543296,
      "step": 33045
    },
    {
      "epoch": 8.722713474990101,
      "grad_norm": 0.1468307077884674,
      "learning_rate": 3.633883613075781e-06,
      "loss": 0.032,
      "num_input_tokens_seen": 29547744,
      "step": 33050
    },
    {
      "epoch": 8.724033258545598,
      "grad_norm": 0.17765173316001892,
      "learning_rate": 3.6287878954298693e-06,
      "loss": 0.035,
      "num_input_tokens_seen": 29552384,
      "step": 33055
    },
    {
      "epoch": 8.725353042101096,
      "grad_norm": 0.11306607723236084,
      "learning_rate": 3.6236954734934354e-06,
      "loss": 0.1001,
      "num_input_tokens_seen": 29557248,
      "step": 33060
    },
    {
      "epoch": 8.726672825656593,
      "grad_norm": 0.21788683533668518,
      "learning_rate": 3.618606348051784e-06,
      "loss": 0.0359,
      "num_input_tokens_seen": 29561472,
      "step": 33065
    },
    {
      "epoch": 8.72799260921209,
      "grad_norm": 0.038152098655700684,
      "learning_rate": 3.6135205198897376e-06,
      "loss": 0.0187,
      "num_input_tokens_seen": 29565632,
      "step": 33070
    },
    {
      "epoch": 8.729312392767586,
      "grad_norm": 0.14211511611938477,
      "learning_rate": 3.6084379897915854e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 29569984,
      "step": 33075
    },
    {
      "epoch": 8.730632176323082,
      "grad_norm": 0.2154967039823532,
      "learning_rate": 3.6033587585411115e-06,
      "loss": 0.0874,
      "num_input_tokens_seen": 29574592,
      "step": 33080
    },
    {
      "epoch": 8.73195195987858,
      "grad_norm": 0.08838847279548645,
      "learning_rate": 3.5982828269216117e-06,
      "loss": 0.0272,
      "num_input_tokens_seen": 29578944,
      "step": 33085
    },
    {
      "epoch": 8.733271743434077,
      "grad_norm": 0.19743730127811432,
      "learning_rate": 3.593210195715843e-06,
      "loss": 0.0542,
      "num_input_tokens_seen": 29583744,
      "step": 33090
    },
    {
      "epoch": 8.734591526989574,
      "grad_norm": 0.5086268186569214,
      "learning_rate": 3.5881408657060773e-06,
      "loss": 0.0941,
      "num_input_tokens_seen": 29587968,
      "step": 33095
    },
    {
      "epoch": 8.73591131054507,
      "grad_norm": 0.048844076693058014,
      "learning_rate": 3.583074837674075e-06,
      "loss": 0.0334,
      "num_input_tokens_seen": 29592736,
      "step": 33100
    },
    {
      "epoch": 8.737231094100567,
      "grad_norm": 0.43333449959754944,
      "learning_rate": 3.578012112401069e-06,
      "loss": 0.087,
      "num_input_tokens_seen": 29597024,
      "step": 33105
    },
    {
      "epoch": 8.738550877656065,
      "grad_norm": 0.19507119059562683,
      "learning_rate": 3.5729526906677996e-06,
      "loss": 0.0663,
      "num_input_tokens_seen": 29601600,
      "step": 33110
    },
    {
      "epoch": 8.739870661211562,
      "grad_norm": 0.4819417893886566,
      "learning_rate": 3.5678965732545007e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 29606496,
      "step": 33115
    },
    {
      "epoch": 8.741190444767058,
      "grad_norm": 0.15235202014446259,
      "learning_rate": 3.562843760940876e-06,
      "loss": 0.0741,
      "num_input_tokens_seen": 29611040,
      "step": 33120
    },
    {
      "epoch": 8.742510228322555,
      "grad_norm": 0.26340052485466003,
      "learning_rate": 3.5577942545061473e-06,
      "loss": 0.047,
      "num_input_tokens_seen": 29615616,
      "step": 33125
    },
    {
      "epoch": 8.743830011878051,
      "grad_norm": 0.14818225800991058,
      "learning_rate": 3.5527480547289967e-06,
      "loss": 0.1005,
      "num_input_tokens_seen": 29620128,
      "step": 33130
    },
    {
      "epoch": 8.745149795433548,
      "grad_norm": 0.47261500358581543,
      "learning_rate": 3.547705162387624e-06,
      "loss": 0.0642,
      "num_input_tokens_seen": 29624608,
      "step": 33135
    },
    {
      "epoch": 8.746469578989046,
      "grad_norm": 0.11401692777872086,
      "learning_rate": 3.542665578259699e-06,
      "loss": 0.0455,
      "num_input_tokens_seen": 29629376,
      "step": 33140
    },
    {
      "epoch": 8.747789362544543,
      "grad_norm": 0.11041387170553207,
      "learning_rate": 3.5376293031223945e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 29633984,
      "step": 33145
    },
    {
      "epoch": 8.74910914610004,
      "grad_norm": 0.08082600682973862,
      "learning_rate": 3.5325963377523614e-06,
      "loss": 0.05,
      "num_input_tokens_seen": 29638592,
      "step": 33150
    },
    {
      "epoch": 8.750428929655536,
      "grad_norm": 0.07581301778554916,
      "learning_rate": 3.5275666829257536e-06,
      "loss": 0.0359,
      "num_input_tokens_seen": 29643232,
      "step": 33155
    },
    {
      "epoch": 8.751748713211033,
      "grad_norm": 0.103957399725914,
      "learning_rate": 3.5225403394181955e-06,
      "loss": 0.0357,
      "num_input_tokens_seen": 29647808,
      "step": 33160
    },
    {
      "epoch": 8.75306849676653,
      "grad_norm": 0.2724948227405548,
      "learning_rate": 3.517517308004828e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 29652192,
      "step": 33165
    },
    {
      "epoch": 8.754388280322027,
      "grad_norm": 0.1904197484254837,
      "learning_rate": 3.512497589460251e-06,
      "loss": 0.0794,
      "num_input_tokens_seen": 29656384,
      "step": 33170
    },
    {
      "epoch": 8.755708063877524,
      "grad_norm": 0.42544230818748474,
      "learning_rate": 3.5074811845585727e-06,
      "loss": 0.0814,
      "num_input_tokens_seen": 29660768,
      "step": 33175
    },
    {
      "epoch": 8.75702784743302,
      "grad_norm": 0.38817551732063293,
      "learning_rate": 3.5024680940733937e-06,
      "loss": 0.0683,
      "num_input_tokens_seen": 29664960,
      "step": 33180
    },
    {
      "epoch": 8.758347630988517,
      "grad_norm": 0.1555711179971695,
      "learning_rate": 3.4974583187777852e-06,
      "loss": 0.0518,
      "num_input_tokens_seen": 29669600,
      "step": 33185
    },
    {
      "epoch": 8.759667414544015,
      "grad_norm": 0.05048147961497307,
      "learning_rate": 3.4924518594443204e-06,
      "loss": 0.1023,
      "num_input_tokens_seen": 29674048,
      "step": 33190
    },
    {
      "epoch": 8.760987198099512,
      "grad_norm": 0.2714904248714447,
      "learning_rate": 3.4874487168450682e-06,
      "loss": 0.0746,
      "num_input_tokens_seen": 29678240,
      "step": 33195
    },
    {
      "epoch": 8.762306981655009,
      "grad_norm": 0.0266673993319273,
      "learning_rate": 3.482448891751558e-06,
      "loss": 0.0213,
      "num_input_tokens_seen": 29682560,
      "step": 33200
    },
    {
      "epoch": 8.762306981655009,
      "eval_loss": 0.06676588952541351,
      "eval_runtime": 64.819,
      "eval_samples_per_second": 103.905,
      "eval_steps_per_second": 25.98,
      "num_input_tokens_seen": 29682560,
      "step": 33200
    },
    {
      "epoch": 8.763626765210505,
      "grad_norm": 0.28558051586151123,
      "learning_rate": 3.477452384934843e-06,
      "loss": 0.0659,
      "num_input_tokens_seen": 29686880,
      "step": 33205
    },
    {
      "epoch": 8.764946548766002,
      "grad_norm": 0.2627415657043457,
      "learning_rate": 3.472459197165434e-06,
      "loss": 0.0602,
      "num_input_tokens_seen": 29691296,
      "step": 33210
    },
    {
      "epoch": 8.7662663323215,
      "grad_norm": 0.42825135588645935,
      "learning_rate": 3.4674693292133518e-06,
      "loss": 0.0771,
      "num_input_tokens_seen": 29695616,
      "step": 33215
    },
    {
      "epoch": 8.767586115876997,
      "grad_norm": 0.056689005345106125,
      "learning_rate": 3.4624827818480977e-06,
      "loss": 0.0451,
      "num_input_tokens_seen": 29700064,
      "step": 33220
    },
    {
      "epoch": 8.768905899432493,
      "grad_norm": 0.11603164672851562,
      "learning_rate": 3.4574995558386474e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 29704928,
      "step": 33225
    },
    {
      "epoch": 8.77022568298799,
      "grad_norm": 0.18743230402469635,
      "learning_rate": 3.452519651953487e-06,
      "loss": 0.0307,
      "num_input_tokens_seen": 29709280,
      "step": 33230
    },
    {
      "epoch": 8.771545466543486,
      "grad_norm": 0.21699371933937073,
      "learning_rate": 3.447543070960585e-06,
      "loss": 0.0649,
      "num_input_tokens_seen": 29713760,
      "step": 33235
    },
    {
      "epoch": 8.772865250098985,
      "grad_norm": 0.23706123232841492,
      "learning_rate": 3.4425698136273778e-06,
      "loss": 0.0986,
      "num_input_tokens_seen": 29718176,
      "step": 33240
    },
    {
      "epoch": 8.774185033654481,
      "grad_norm": 0.08144243061542511,
      "learning_rate": 3.437599880720821e-06,
      "loss": 0.0241,
      "num_input_tokens_seen": 29722720,
      "step": 33245
    },
    {
      "epoch": 8.775504817209978,
      "grad_norm": 0.05081043764948845,
      "learning_rate": 3.4326332730073267e-06,
      "loss": 0.0474,
      "num_input_tokens_seen": 29727360,
      "step": 33250
    },
    {
      "epoch": 8.776824600765474,
      "grad_norm": 0.08954569697380066,
      "learning_rate": 3.427669991252813e-06,
      "loss": 0.097,
      "num_input_tokens_seen": 29731744,
      "step": 33255
    },
    {
      "epoch": 8.77814438432097,
      "grad_norm": 0.04469580948352814,
      "learning_rate": 3.42271003622269e-06,
      "loss": 0.0417,
      "num_input_tokens_seen": 29736160,
      "step": 33260
    },
    {
      "epoch": 8.779464167876469,
      "grad_norm": 0.519282341003418,
      "learning_rate": 3.4177534086818286e-06,
      "loss": 0.0715,
      "num_input_tokens_seen": 29740512,
      "step": 33265
    },
    {
      "epoch": 8.780783951431966,
      "grad_norm": 0.2236897051334381,
      "learning_rate": 3.412800109394612e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 29744928,
      "step": 33270
    },
    {
      "epoch": 8.782103734987462,
      "grad_norm": 0.31672611832618713,
      "learning_rate": 3.4078501391249044e-06,
      "loss": 0.1019,
      "num_input_tokens_seen": 29749312,
      "step": 33275
    },
    {
      "epoch": 8.783423518542959,
      "grad_norm": 0.25734931230545044,
      "learning_rate": 3.4029034986360453e-06,
      "loss": 0.0892,
      "num_input_tokens_seen": 29753728,
      "step": 33280
    },
    {
      "epoch": 8.784743302098455,
      "grad_norm": 0.17762035131454468,
      "learning_rate": 3.397960188690877e-06,
      "loss": 0.08,
      "num_input_tokens_seen": 29758400,
      "step": 33285
    },
    {
      "epoch": 8.786063085653954,
      "grad_norm": 0.29525870084762573,
      "learning_rate": 3.393020210051717e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 29762848,
      "step": 33290
    },
    {
      "epoch": 8.78738286920945,
      "grad_norm": 0.34065327048301697,
      "learning_rate": 3.3880835634803655e-06,
      "loss": 0.0674,
      "num_input_tokens_seen": 29767200,
      "step": 33295
    },
    {
      "epoch": 8.788702652764947,
      "grad_norm": 0.05425555631518364,
      "learning_rate": 3.383150249738126e-06,
      "loss": 0.0356,
      "num_input_tokens_seen": 29771424,
      "step": 33300
    },
    {
      "epoch": 8.790022436320443,
      "grad_norm": 0.2203356772661209,
      "learning_rate": 3.3782202695857663e-06,
      "loss": 0.0385,
      "num_input_tokens_seen": 29775584,
      "step": 33305
    },
    {
      "epoch": 8.79134221987594,
      "grad_norm": 0.455761194229126,
      "learning_rate": 3.373293623783558e-06,
      "loss": 0.0725,
      "num_input_tokens_seen": 29780096,
      "step": 33310
    },
    {
      "epoch": 8.792662003431436,
      "grad_norm": 0.061299826949834824,
      "learning_rate": 3.368370313091257e-06,
      "loss": 0.023,
      "num_input_tokens_seen": 29784480,
      "step": 33315
    },
    {
      "epoch": 8.793981786986935,
      "grad_norm": 0.0508180633187294,
      "learning_rate": 3.363450338268087e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 29788928,
      "step": 33320
    },
    {
      "epoch": 8.795301570542431,
      "grad_norm": 0.0725937932729721,
      "learning_rate": 3.358533700072783e-06,
      "loss": 0.0492,
      "num_input_tokens_seen": 29793408,
      "step": 33325
    },
    {
      "epoch": 8.796621354097928,
      "grad_norm": 0.18679532408714294,
      "learning_rate": 3.3536203992635377e-06,
      "loss": 0.0697,
      "num_input_tokens_seen": 29797920,
      "step": 33330
    },
    {
      "epoch": 8.797941137653424,
      "grad_norm": 0.2164284884929657,
      "learning_rate": 3.348710436598057e-06,
      "loss": 0.0549,
      "num_input_tokens_seen": 29802144,
      "step": 33335
    },
    {
      "epoch": 8.799260921208921,
      "grad_norm": 0.1605917066335678,
      "learning_rate": 3.3438038128335155e-06,
      "loss": 0.0296,
      "num_input_tokens_seen": 29806496,
      "step": 33340
    },
    {
      "epoch": 8.80058070476442,
      "grad_norm": 0.14203490316867828,
      "learning_rate": 3.338900528726571e-06,
      "loss": 0.0581,
      "num_input_tokens_seen": 29811328,
      "step": 33345
    },
    {
      "epoch": 8.801900488319916,
      "grad_norm": 0.12017769366502762,
      "learning_rate": 3.3340005850333812e-06,
      "loss": 0.0701,
      "num_input_tokens_seen": 29815840,
      "step": 33350
    },
    {
      "epoch": 8.803220271875412,
      "grad_norm": 0.24080802500247955,
      "learning_rate": 3.329103982509568e-06,
      "loss": 0.0656,
      "num_input_tokens_seen": 29820416,
      "step": 33355
    },
    {
      "epoch": 8.804540055430909,
      "grad_norm": 0.209104523062706,
      "learning_rate": 3.324210721910259e-06,
      "loss": 0.0318,
      "num_input_tokens_seen": 29824736,
      "step": 33360
    },
    {
      "epoch": 8.805859838986406,
      "grad_norm": 0.2165743112564087,
      "learning_rate": 3.319320803990053e-06,
      "loss": 0.0301,
      "num_input_tokens_seen": 29829216,
      "step": 33365
    },
    {
      "epoch": 8.807179622541904,
      "grad_norm": 0.13622023165225983,
      "learning_rate": 3.3144342295030274e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 29833824,
      "step": 33370
    },
    {
      "epoch": 8.8084994060974,
      "grad_norm": 0.16241422295570374,
      "learning_rate": 3.309550999202765e-06,
      "loss": 0.1075,
      "num_input_tokens_seen": 29838016,
      "step": 33375
    },
    {
      "epoch": 8.809819189652897,
      "grad_norm": 0.4399169981479645,
      "learning_rate": 3.3046711138423197e-06,
      "loss": 0.0499,
      "num_input_tokens_seen": 29842400,
      "step": 33380
    },
    {
      "epoch": 8.811138973208394,
      "grad_norm": 0.408474326133728,
      "learning_rate": 3.2997945741742255e-06,
      "loss": 0.0399,
      "num_input_tokens_seen": 29847136,
      "step": 33385
    },
    {
      "epoch": 8.81245875676389,
      "grad_norm": 0.34309136867523193,
      "learning_rate": 3.2949213809505082e-06,
      "loss": 0.0719,
      "num_input_tokens_seen": 29851776,
      "step": 33390
    },
    {
      "epoch": 8.813778540319387,
      "grad_norm": 0.18612755835056305,
      "learning_rate": 3.2900515349226834e-06,
      "loss": 0.0319,
      "num_input_tokens_seen": 29856320,
      "step": 33395
    },
    {
      "epoch": 8.815098323874885,
      "grad_norm": 0.05157690867781639,
      "learning_rate": 3.285185036841731e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 29860768,
      "step": 33400
    },
    {
      "epoch": 8.815098323874885,
      "eval_loss": 0.06660357117652893,
      "eval_runtime": 64.7488,
      "eval_samples_per_second": 104.017,
      "eval_steps_per_second": 26.008,
      "num_input_tokens_seen": 29860768,
      "step": 33400
    },
    {
      "epoch": 8.816418107430382,
      "grad_norm": 0.5171669721603394,
      "learning_rate": 3.2803218874581377e-06,
      "loss": 0.0538,
      "num_input_tokens_seen": 29865056,
      "step": 33405
    },
    {
      "epoch": 8.817737890985878,
      "grad_norm": 0.05674148350954056,
      "learning_rate": 3.2754620875218494e-06,
      "loss": 0.0328,
      "num_input_tokens_seen": 29869344,
      "step": 33410
    },
    {
      "epoch": 8.819057674541375,
      "grad_norm": 0.03309844434261322,
      "learning_rate": 3.2706056377823146e-06,
      "loss": 0.0272,
      "num_input_tokens_seen": 29873824,
      "step": 33415
    },
    {
      "epoch": 8.820377458096871,
      "grad_norm": 0.618338406085968,
      "learning_rate": 3.2657525389884647e-06,
      "loss": 0.089,
      "num_input_tokens_seen": 29878496,
      "step": 33420
    },
    {
      "epoch": 8.82169724165237,
      "grad_norm": 0.1796351820230484,
      "learning_rate": 3.260902791888698e-06,
      "loss": 0.0346,
      "num_input_tokens_seen": 29882752,
      "step": 33425
    },
    {
      "epoch": 8.823017025207866,
      "grad_norm": 0.4590378701686859,
      "learning_rate": 3.2560563972309166e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 29887072,
      "step": 33430
    },
    {
      "epoch": 8.824336808763363,
      "grad_norm": 0.2503621578216553,
      "learning_rate": 3.251213355762489e-06,
      "loss": 0.1501,
      "num_input_tokens_seen": 29891680,
      "step": 33435
    },
    {
      "epoch": 8.82565659231886,
      "grad_norm": 0.24146400392055511,
      "learning_rate": 3.2463736682302707e-06,
      "loss": 0.0496,
      "num_input_tokens_seen": 29896320,
      "step": 33440
    },
    {
      "epoch": 8.826976375874356,
      "grad_norm": 0.18247416615486145,
      "learning_rate": 3.2415373353806124e-06,
      "loss": 0.0846,
      "num_input_tokens_seen": 29901152,
      "step": 33445
    },
    {
      "epoch": 8.828296159429854,
      "grad_norm": 0.1882723718881607,
      "learning_rate": 3.236704357959322e-06,
      "loss": 0.0269,
      "num_input_tokens_seen": 29905888,
      "step": 33450
    },
    {
      "epoch": 8.82961594298535,
      "grad_norm": 0.1875351071357727,
      "learning_rate": 3.2318747367117154e-06,
      "loss": 0.0498,
      "num_input_tokens_seen": 29910496,
      "step": 33455
    },
    {
      "epoch": 8.830935726540847,
      "grad_norm": 0.15903311967849731,
      "learning_rate": 3.227048472382585e-06,
      "loss": 0.0433,
      "num_input_tokens_seen": 29914816,
      "step": 33460
    },
    {
      "epoch": 8.832255510096344,
      "grad_norm": 0.41671276092529297,
      "learning_rate": 3.2222255657161915e-06,
      "loss": 0.09,
      "num_input_tokens_seen": 29919200,
      "step": 33465
    },
    {
      "epoch": 8.83357529365184,
      "grad_norm": 0.23736391961574554,
      "learning_rate": 3.2174060174562924e-06,
      "loss": 0.0314,
      "num_input_tokens_seen": 29923424,
      "step": 33470
    },
    {
      "epoch": 8.834895077207339,
      "grad_norm": 0.31572583317756653,
      "learning_rate": 3.2125898283461298e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 29927648,
      "step": 33475
    },
    {
      "epoch": 8.836214860762835,
      "grad_norm": 0.0662103220820427,
      "learning_rate": 3.207776999128406e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 29931712,
      "step": 33480
    },
    {
      "epoch": 8.837534644318332,
      "grad_norm": 0.12377462536096573,
      "learning_rate": 3.202967530545331e-06,
      "loss": 0.0469,
      "num_input_tokens_seen": 29935744,
      "step": 33485
    },
    {
      "epoch": 8.838854427873828,
      "grad_norm": 0.3124590218067169,
      "learning_rate": 3.1981614233385778e-06,
      "loss": 0.0434,
      "num_input_tokens_seen": 29939840,
      "step": 33490
    },
    {
      "epoch": 8.840174211429325,
      "grad_norm": 0.33301085233688354,
      "learning_rate": 3.1933586782493115e-06,
      "loss": 0.1164,
      "num_input_tokens_seen": 29944320,
      "step": 33495
    },
    {
      "epoch": 8.841493994984823,
      "grad_norm": 0.10111606866121292,
      "learning_rate": 3.188559296018184e-06,
      "loss": 0.0673,
      "num_input_tokens_seen": 29949184,
      "step": 33500
    },
    {
      "epoch": 8.84281377854032,
      "grad_norm": 0.1912315934896469,
      "learning_rate": 3.1837632773853098e-06,
      "loss": 0.0554,
      "num_input_tokens_seen": 29953504,
      "step": 33505
    },
    {
      "epoch": 8.844133562095816,
      "grad_norm": 0.1832091361284256,
      "learning_rate": 3.178970623090294e-06,
      "loss": 0.0216,
      "num_input_tokens_seen": 29958016,
      "step": 33510
    },
    {
      "epoch": 8.845453345651313,
      "grad_norm": 0.04191828519105911,
      "learning_rate": 3.174181333872234e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 29962432,
      "step": 33515
    },
    {
      "epoch": 8.84677312920681,
      "grad_norm": 0.10754439234733582,
      "learning_rate": 3.169395410469686e-06,
      "loss": 0.0599,
      "num_input_tokens_seen": 29966784,
      "step": 33520
    },
    {
      "epoch": 8.848092912762308,
      "grad_norm": 0.057430148124694824,
      "learning_rate": 3.164612853620713e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 29971744,
      "step": 33525
    },
    {
      "epoch": 8.849412696317804,
      "grad_norm": 0.2920302748680115,
      "learning_rate": 3.1598336640628333e-06,
      "loss": 0.0307,
      "num_input_tokens_seen": 29976224,
      "step": 33530
    },
    {
      "epoch": 8.850732479873301,
      "grad_norm": 0.15682421624660492,
      "learning_rate": 3.155057842533063e-06,
      "loss": 0.047,
      "num_input_tokens_seen": 29980384,
      "step": 33535
    },
    {
      "epoch": 8.852052263428797,
      "grad_norm": 0.08103614300489426,
      "learning_rate": 3.1502853897678984e-06,
      "loss": 0.0395,
      "num_input_tokens_seen": 29984768,
      "step": 33540
    },
    {
      "epoch": 8.853372046984294,
      "grad_norm": 0.20594191551208496,
      "learning_rate": 3.1455163065033017e-06,
      "loss": 0.0397,
      "num_input_tokens_seen": 29989472,
      "step": 33545
    },
    {
      "epoch": 8.854691830539792,
      "grad_norm": 0.050733353942632675,
      "learning_rate": 3.140750593474734e-06,
      "loss": 0.0696,
      "num_input_tokens_seen": 29993920,
      "step": 33550
    },
    {
      "epoch": 8.856011614095289,
      "grad_norm": 0.18081626296043396,
      "learning_rate": 3.1359882514171294e-06,
      "loss": 0.0779,
      "num_input_tokens_seen": 29998048,
      "step": 33555
    },
    {
      "epoch": 8.857331397650785,
      "grad_norm": 0.029113732278347015,
      "learning_rate": 3.1312292810648903e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 30002496,
      "step": 33560
    },
    {
      "epoch": 8.858651181206282,
      "grad_norm": 0.10736019164323807,
      "learning_rate": 3.1264736831519204e-06,
      "loss": 0.0279,
      "num_input_tokens_seen": 30006816,
      "step": 33565
    },
    {
      "epoch": 8.859970964761779,
      "grad_norm": 0.03093256987631321,
      "learning_rate": 3.1217214584115863e-06,
      "loss": 0.0413,
      "num_input_tokens_seen": 30011552,
      "step": 33570
    },
    {
      "epoch": 8.861290748317275,
      "grad_norm": 0.17004123330116272,
      "learning_rate": 3.116972607576746e-06,
      "loss": 0.0732,
      "num_input_tokens_seen": 30016224,
      "step": 33575
    },
    {
      "epoch": 8.862610531872773,
      "grad_norm": 0.3231600821018219,
      "learning_rate": 3.1122271313797303e-06,
      "loss": 0.0746,
      "num_input_tokens_seen": 30020544,
      "step": 33580
    },
    {
      "epoch": 8.86393031542827,
      "grad_norm": 0.1653924435377121,
      "learning_rate": 3.107485030552343e-06,
      "loss": 0.0851,
      "num_input_tokens_seen": 30025120,
      "step": 33585
    },
    {
      "epoch": 8.865250098983767,
      "grad_norm": 0.1316758692264557,
      "learning_rate": 3.1027463058258848e-06,
      "loss": 0.0542,
      "num_input_tokens_seen": 30029696,
      "step": 33590
    },
    {
      "epoch": 8.866569882539263,
      "grad_norm": 0.04609846696257591,
      "learning_rate": 3.0980109579311273e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 30034080,
      "step": 33595
    },
    {
      "epoch": 8.86788966609476,
      "grad_norm": 0.1789936125278473,
      "learning_rate": 3.093278987598314e-06,
      "loss": 0.0547,
      "num_input_tokens_seen": 30038624,
      "step": 33600
    },
    {
      "epoch": 8.86788966609476,
      "eval_loss": 0.06689382344484329,
      "eval_runtime": 64.7544,
      "eval_samples_per_second": 104.008,
      "eval_steps_per_second": 26.006,
      "num_input_tokens_seen": 30038624,
      "step": 33600
    },
    {
      "epoch": 8.869209449650258,
      "grad_norm": 0.2320082187652588,
      "learning_rate": 3.0885503955571826e-06,
      "loss": 0.0414,
      "num_input_tokens_seen": 30042944,
      "step": 33605
    },
    {
      "epoch": 8.870529233205755,
      "grad_norm": 0.35803115367889404,
      "learning_rate": 3.0838251825369313e-06,
      "loss": 0.0669,
      "num_input_tokens_seen": 30047328,
      "step": 33610
    },
    {
      "epoch": 8.871849016761251,
      "grad_norm": 0.19429190456867218,
      "learning_rate": 3.0791033492662517e-06,
      "loss": 0.0526,
      "num_input_tokens_seen": 30051680,
      "step": 33615
    },
    {
      "epoch": 8.873168800316748,
      "grad_norm": 0.24337060749530792,
      "learning_rate": 3.0743848964733203e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 30056064,
      "step": 33620
    },
    {
      "epoch": 8.874488583872244,
      "grad_norm": 0.2758674919605255,
      "learning_rate": 3.0696698248857625e-06,
      "loss": 0.044,
      "num_input_tokens_seen": 30060448,
      "step": 33625
    },
    {
      "epoch": 8.875808367427743,
      "grad_norm": 0.28566503524780273,
      "learning_rate": 3.0649581352307192e-06,
      "loss": 0.0767,
      "num_input_tokens_seen": 30064928,
      "step": 33630
    },
    {
      "epoch": 8.87712815098324,
      "grad_norm": 0.11672279238700867,
      "learning_rate": 3.060249828234776e-06,
      "loss": 0.0532,
      "num_input_tokens_seen": 30069024,
      "step": 33635
    },
    {
      "epoch": 8.878447934538736,
      "grad_norm": 0.19393907487392426,
      "learning_rate": 3.055544904624025e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 30073504,
      "step": 33640
    },
    {
      "epoch": 8.879767718094232,
      "grad_norm": 0.04693886265158653,
      "learning_rate": 3.050843365124026e-06,
      "loss": 0.0386,
      "num_input_tokens_seen": 30077664,
      "step": 33645
    },
    {
      "epoch": 8.881087501649729,
      "grad_norm": 0.14179402589797974,
      "learning_rate": 3.0461452104598083e-06,
      "loss": 0.0416,
      "num_input_tokens_seen": 30081888,
      "step": 33650
    },
    {
      "epoch": 8.882407285205227,
      "grad_norm": 0.32260459661483765,
      "learning_rate": 3.0414504413558836e-06,
      "loss": 0.0563,
      "num_input_tokens_seen": 30086144,
      "step": 33655
    },
    {
      "epoch": 8.883727068760724,
      "grad_norm": 0.2694810628890991,
      "learning_rate": 3.0367590585362564e-06,
      "loss": 0.0428,
      "num_input_tokens_seen": 30090496,
      "step": 33660
    },
    {
      "epoch": 8.88504685231622,
      "grad_norm": 0.2997215688228607,
      "learning_rate": 3.0320710627243813e-06,
      "loss": 0.0787,
      "num_input_tokens_seen": 30094976,
      "step": 33665
    },
    {
      "epoch": 8.886366635871717,
      "grad_norm": 0.16100257635116577,
      "learning_rate": 3.027386454643222e-06,
      "loss": 0.0614,
      "num_input_tokens_seen": 30099712,
      "step": 33670
    },
    {
      "epoch": 8.887686419427213,
      "grad_norm": 0.3621823787689209,
      "learning_rate": 3.0227052350151914e-06,
      "loss": 0.0534,
      "num_input_tokens_seen": 30104128,
      "step": 33675
    },
    {
      "epoch": 8.88900620298271,
      "grad_norm": 0.14674434065818787,
      "learning_rate": 3.0180274045621957e-06,
      "loss": 0.1146,
      "num_input_tokens_seen": 30108384,
      "step": 33680
    },
    {
      "epoch": 8.890325986538208,
      "grad_norm": 0.2379944622516632,
      "learning_rate": 3.013352964005625e-06,
      "loss": 0.0676,
      "num_input_tokens_seen": 30112640,
      "step": 33685
    },
    {
      "epoch": 8.891645770093705,
      "grad_norm": 0.12532909214496613,
      "learning_rate": 3.0086819140663218e-06,
      "loss": 0.0348,
      "num_input_tokens_seen": 30117376,
      "step": 33690
    },
    {
      "epoch": 8.892965553649201,
      "grad_norm": 0.05897604301571846,
      "learning_rate": 3.0040142554646265e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 30121696,
      "step": 33695
    },
    {
      "epoch": 8.894285337204698,
      "grad_norm": 0.08731305599212646,
      "learning_rate": 2.999349988920361e-06,
      "loss": 0.0693,
      "num_input_tokens_seen": 30125920,
      "step": 33700
    },
    {
      "epoch": 8.895605120760194,
      "grad_norm": 0.07779120653867722,
      "learning_rate": 2.994689115152796e-06,
      "loss": 0.071,
      "num_input_tokens_seen": 30130112,
      "step": 33705
    },
    {
      "epoch": 8.896924904315693,
      "grad_norm": 0.20681162178516388,
      "learning_rate": 2.9900316348807105e-06,
      "loss": 0.0491,
      "num_input_tokens_seen": 30134432,
      "step": 33710
    },
    {
      "epoch": 8.89824468787119,
      "grad_norm": 0.22345609962940216,
      "learning_rate": 2.985377548822338e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 30138976,
      "step": 33715
    },
    {
      "epoch": 8.899564471426686,
      "grad_norm": 0.16095314919948578,
      "learning_rate": 2.980726857695404e-06,
      "loss": 0.0403,
      "num_input_tokens_seen": 30143552,
      "step": 33720
    },
    {
      "epoch": 8.900884254982182,
      "grad_norm": 0.07145502418279648,
      "learning_rate": 2.9760795622171017e-06,
      "loss": 0.0493,
      "num_input_tokens_seen": 30148192,
      "step": 33725
    },
    {
      "epoch": 8.902204038537679,
      "grad_norm": 0.05089440196752548,
      "learning_rate": 2.971435663104094e-06,
      "loss": 0.0635,
      "num_input_tokens_seen": 30152960,
      "step": 33730
    },
    {
      "epoch": 8.903523822093177,
      "grad_norm": 0.11006798595190048,
      "learning_rate": 2.9667951610725385e-06,
      "loss": 0.0839,
      "num_input_tokens_seen": 30157568,
      "step": 33735
    },
    {
      "epoch": 8.904843605648674,
      "grad_norm": 0.39502623677253723,
      "learning_rate": 2.9621580568380575e-06,
      "loss": 0.0841,
      "num_input_tokens_seen": 30162144,
      "step": 33740
    },
    {
      "epoch": 8.90616338920417,
      "grad_norm": 0.40727654099464417,
      "learning_rate": 2.9575243511157453e-06,
      "loss": 0.0803,
      "num_input_tokens_seen": 30166720,
      "step": 33745
    },
    {
      "epoch": 8.907483172759667,
      "grad_norm": 0.3528825044631958,
      "learning_rate": 2.952894044620186e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 30171296,
      "step": 33750
    },
    {
      "epoch": 8.908802956315164,
      "grad_norm": 0.26231250166893005,
      "learning_rate": 2.948267138065419e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 30175968,
      "step": 33755
    },
    {
      "epoch": 8.910122739870662,
      "grad_norm": 0.4106236398220062,
      "learning_rate": 2.943643632164983e-06,
      "loss": 0.0422,
      "num_input_tokens_seen": 30180448,
      "step": 33760
    },
    {
      "epoch": 8.911442523426159,
      "grad_norm": 0.43454307317733765,
      "learning_rate": 2.939023527631879e-06,
      "loss": 0.0582,
      "num_input_tokens_seen": 30184832,
      "step": 33765
    },
    {
      "epoch": 8.912762306981655,
      "grad_norm": 0.20565663278102875,
      "learning_rate": 2.934406825178576e-06,
      "loss": 0.0396,
      "num_input_tokens_seen": 30189440,
      "step": 33770
    },
    {
      "epoch": 8.914082090537152,
      "grad_norm": 0.3171520233154297,
      "learning_rate": 2.9297935255170357e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 30193728,
      "step": 33775
    },
    {
      "epoch": 8.915401874092648,
      "grad_norm": 0.3234643340110779,
      "learning_rate": 2.925183629358691e-06,
      "loss": 0.0836,
      "num_input_tokens_seen": 30198368,
      "step": 33780
    },
    {
      "epoch": 8.916721657648147,
      "grad_norm": 0.4608983099460602,
      "learning_rate": 2.9205771374144346e-06,
      "loss": 0.1413,
      "num_input_tokens_seen": 30202944,
      "step": 33785
    },
    {
      "epoch": 8.918041441203643,
      "grad_norm": 0.05254804715514183,
      "learning_rate": 2.915974050394657e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 30207680,
      "step": 33790
    },
    {
      "epoch": 8.91936122475914,
      "grad_norm": 0.05397457256913185,
      "learning_rate": 2.9113743690092067e-06,
      "loss": 0.0533,
      "num_input_tokens_seen": 30211968,
      "step": 33795
    },
    {
      "epoch": 8.920681008314636,
      "grad_norm": 0.04846327379345894,
      "learning_rate": 2.906778093967402e-06,
      "loss": 0.0384,
      "num_input_tokens_seen": 30216384,
      "step": 33800
    },
    {
      "epoch": 8.920681008314636,
      "eval_loss": 0.06673185527324677,
      "eval_runtime": 64.7602,
      "eval_samples_per_second": 103.999,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 30216384,
      "step": 33800
    },
    {
      "epoch": 8.922000791870133,
      "grad_norm": 0.03824539855122566,
      "learning_rate": 2.9021852259780656e-06,
      "loss": 0.0225,
      "num_input_tokens_seen": 30220800,
      "step": 33805
    },
    {
      "epoch": 8.923320575425631,
      "grad_norm": 0.13985612988471985,
      "learning_rate": 2.8975957657494583e-06,
      "loss": 0.0376,
      "num_input_tokens_seen": 30225472,
      "step": 33810
    },
    {
      "epoch": 8.924640358981128,
      "grad_norm": 0.22610527276992798,
      "learning_rate": 2.8930097139893417e-06,
      "loss": 0.0274,
      "num_input_tokens_seen": 30229984,
      "step": 33815
    },
    {
      "epoch": 8.925960142536624,
      "grad_norm": 0.12828852236270905,
      "learning_rate": 2.888427071404945e-06,
      "loss": 0.0425,
      "num_input_tokens_seen": 30234528,
      "step": 33820
    },
    {
      "epoch": 8.92727992609212,
      "grad_norm": 0.17715567350387573,
      "learning_rate": 2.8838478387029606e-06,
      "loss": 0.0709,
      "num_input_tokens_seen": 30238976,
      "step": 33825
    },
    {
      "epoch": 8.928599709647617,
      "grad_norm": 0.24659322202205658,
      "learning_rate": 2.8792720165895737e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 30243520,
      "step": 33830
    },
    {
      "epoch": 8.929919493203116,
      "grad_norm": 0.21981750428676605,
      "learning_rate": 2.874699605770423e-06,
      "loss": 0.064,
      "num_input_tokens_seen": 30248128,
      "step": 33835
    },
    {
      "epoch": 8.931239276758612,
      "grad_norm": 0.26891186833381653,
      "learning_rate": 2.8701306069506383e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 30252672,
      "step": 33840
    },
    {
      "epoch": 8.932559060314109,
      "grad_norm": 0.21992447972297668,
      "learning_rate": 2.8655650208348178e-06,
      "loss": 0.0446,
      "num_input_tokens_seen": 30257216,
      "step": 33845
    },
    {
      "epoch": 8.933878843869605,
      "grad_norm": 0.10089381784200668,
      "learning_rate": 2.8610028481270257e-06,
      "loss": 0.0392,
      "num_input_tokens_seen": 30261600,
      "step": 33850
    },
    {
      "epoch": 8.935198627425102,
      "grad_norm": 0.10193891078233719,
      "learning_rate": 2.856444089530813e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 30266112,
      "step": 33855
    },
    {
      "epoch": 8.936518410980598,
      "grad_norm": 0.22850273549556732,
      "learning_rate": 2.8518887457491955e-06,
      "loss": 0.068,
      "num_input_tokens_seen": 30270688,
      "step": 33860
    },
    {
      "epoch": 8.937838194536097,
      "grad_norm": 0.4412859380245209,
      "learning_rate": 2.8473368174846666e-06,
      "loss": 0.0604,
      "num_input_tokens_seen": 30275360,
      "step": 33865
    },
    {
      "epoch": 8.939157978091593,
      "grad_norm": 0.19657257199287415,
      "learning_rate": 2.842788305439184e-06,
      "loss": 0.0578,
      "num_input_tokens_seen": 30279968,
      "step": 33870
    },
    {
      "epoch": 8.94047776164709,
      "grad_norm": 0.23637180030345917,
      "learning_rate": 2.8382432103141925e-06,
      "loss": 0.0669,
      "num_input_tokens_seen": 30284512,
      "step": 33875
    },
    {
      "epoch": 8.941797545202586,
      "grad_norm": 0.09823627024888992,
      "learning_rate": 2.833701532810598e-06,
      "loss": 0.1378,
      "num_input_tokens_seen": 30288992,
      "step": 33880
    },
    {
      "epoch": 8.943117328758083,
      "grad_norm": 0.05842132866382599,
      "learning_rate": 2.8291632736287877e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 30293024,
      "step": 33885
    },
    {
      "epoch": 8.944437112313581,
      "grad_norm": 0.28283172845840454,
      "learning_rate": 2.824628433468615e-06,
      "loss": 0.08,
      "num_input_tokens_seen": 30297504,
      "step": 33890
    },
    {
      "epoch": 8.945756895869078,
      "grad_norm": 0.23988932371139526,
      "learning_rate": 2.8200970130294073e-06,
      "loss": 0.0469,
      "num_input_tokens_seen": 30301696,
      "step": 33895
    },
    {
      "epoch": 8.947076679424574,
      "grad_norm": 0.2807973027229309,
      "learning_rate": 2.8155690130099775e-06,
      "loss": 0.029,
      "num_input_tokens_seen": 30306176,
      "step": 33900
    },
    {
      "epoch": 8.948396462980071,
      "grad_norm": 0.1333424299955368,
      "learning_rate": 2.8110444341085895e-06,
      "loss": 0.0477,
      "num_input_tokens_seen": 30310752,
      "step": 33905
    },
    {
      "epoch": 8.949716246535568,
      "grad_norm": 0.044917792081832886,
      "learning_rate": 2.806523277022996e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 30315392,
      "step": 33910
    },
    {
      "epoch": 8.951036030091066,
      "grad_norm": 0.0968228131532669,
      "learning_rate": 2.802005542450409e-06,
      "loss": 0.0233,
      "num_input_tokens_seen": 30319680,
      "step": 33915
    },
    {
      "epoch": 8.952355813646562,
      "grad_norm": 0.09316549450159073,
      "learning_rate": 2.797491231087526e-06,
      "loss": 0.0302,
      "num_input_tokens_seen": 30324448,
      "step": 33920
    },
    {
      "epoch": 8.953675597202059,
      "grad_norm": 0.5087403655052185,
      "learning_rate": 2.7929803436305137e-06,
      "loss": 0.0482,
      "num_input_tokens_seen": 30328672,
      "step": 33925
    },
    {
      "epoch": 8.954995380757556,
      "grad_norm": 0.22161389887332916,
      "learning_rate": 2.788472880774998e-06,
      "loss": 0.0807,
      "num_input_tokens_seen": 30333056,
      "step": 33930
    },
    {
      "epoch": 8.956315164313052,
      "grad_norm": 0.1899174004793167,
      "learning_rate": 2.7839688432160977e-06,
      "loss": 0.0741,
      "num_input_tokens_seen": 30337632,
      "step": 33935
    },
    {
      "epoch": 8.957634947868549,
      "grad_norm": 0.31969016790390015,
      "learning_rate": 2.779468231648383e-06,
      "loss": 0.0551,
      "num_input_tokens_seen": 30342176,
      "step": 33940
    },
    {
      "epoch": 8.958954731424047,
      "grad_norm": 0.22740787267684937,
      "learning_rate": 2.774971046765906e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 30346464,
      "step": 33945
    },
    {
      "epoch": 8.960274514979544,
      "grad_norm": 0.2248426377773285,
      "learning_rate": 2.770477289262194e-06,
      "loss": 0.0731,
      "num_input_tokens_seen": 30351168,
      "step": 33950
    },
    {
      "epoch": 8.96159429853504,
      "grad_norm": 0.0767684280872345,
      "learning_rate": 2.765986959830233e-06,
      "loss": 0.0369,
      "num_input_tokens_seen": 30355648,
      "step": 33955
    },
    {
      "epoch": 8.962914082090537,
      "grad_norm": 0.08161349594593048,
      "learning_rate": 2.761500059162492e-06,
      "loss": 0.0844,
      "num_input_tokens_seen": 30360512,
      "step": 33960
    },
    {
      "epoch": 8.964233865646033,
      "grad_norm": 0.11311199516057968,
      "learning_rate": 2.757016587950914e-06,
      "loss": 0.0217,
      "num_input_tokens_seen": 30365024,
      "step": 33965
    },
    {
      "epoch": 8.965553649201532,
      "grad_norm": 0.06902402639389038,
      "learning_rate": 2.752536546886897e-06,
      "loss": 0.0747,
      "num_input_tokens_seen": 30369536,
      "step": 33970
    },
    {
      "epoch": 8.966873432757028,
      "grad_norm": 0.0739562138915062,
      "learning_rate": 2.7480599366613234e-06,
      "loss": 0.048,
      "num_input_tokens_seen": 30374048,
      "step": 33975
    },
    {
      "epoch": 8.968193216312525,
      "grad_norm": 0.06103876233100891,
      "learning_rate": 2.7435867579645473e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 30378496,
      "step": 33980
    },
    {
      "epoch": 8.969512999868021,
      "grad_norm": 0.17259518802165985,
      "learning_rate": 2.739117011486378e-06,
      "loss": 0.0707,
      "num_input_tokens_seen": 30382976,
      "step": 33985
    },
    {
      "epoch": 8.970832783423518,
      "grad_norm": 0.023606494069099426,
      "learning_rate": 2.7346506979161216e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 30387424,
      "step": 33990
    },
    {
      "epoch": 8.972152566979016,
      "grad_norm": 0.08904796093702316,
      "learning_rate": 2.7301878179425227e-06,
      "loss": 0.0537,
      "num_input_tokens_seen": 30391904,
      "step": 33995
    },
    {
      "epoch": 8.973472350534513,
      "grad_norm": 0.1865626722574234,
      "learning_rate": 2.7257283722538244e-06,
      "loss": 0.0605,
      "num_input_tokens_seen": 30395872,
      "step": 34000
    },
    {
      "epoch": 8.973472350534513,
      "eval_loss": 0.06677081435918808,
      "eval_runtime": 64.7564,
      "eval_samples_per_second": 104.005,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 30395872,
      "step": 34000
    },
    {
      "epoch": 8.97479213409001,
      "grad_norm": 0.02121318131685257,
      "learning_rate": 2.7212723615377326e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 30400480,
      "step": 34005
    },
    {
      "epoch": 8.976111917645506,
      "grad_norm": 0.2629278004169464,
      "learning_rate": 2.7168197864814145e-06,
      "loss": 0.095,
      "num_input_tokens_seen": 30405088,
      "step": 34010
    },
    {
      "epoch": 8.977431701201002,
      "grad_norm": 0.27620548009872437,
      "learning_rate": 2.712370647771509e-06,
      "loss": 0.0809,
      "num_input_tokens_seen": 30409600,
      "step": 34015
    },
    {
      "epoch": 8.9787514847565,
      "grad_norm": 0.4787909686565399,
      "learning_rate": 2.707924946094137e-06,
      "loss": 0.1249,
      "num_input_tokens_seen": 30413728,
      "step": 34020
    },
    {
      "epoch": 8.980071268311997,
      "grad_norm": 0.3036344051361084,
      "learning_rate": 2.7034826821348723e-06,
      "loss": 0.0345,
      "num_input_tokens_seen": 30417952,
      "step": 34025
    },
    {
      "epoch": 8.981391051867494,
      "grad_norm": 0.03586472198367119,
      "learning_rate": 2.6990438565787786e-06,
      "loss": 0.0712,
      "num_input_tokens_seen": 30422016,
      "step": 34030
    },
    {
      "epoch": 8.98271083542299,
      "grad_norm": 0.10260725021362305,
      "learning_rate": 2.6946084701103714e-06,
      "loss": 0.0407,
      "num_input_tokens_seen": 30426336,
      "step": 34035
    },
    {
      "epoch": 8.984030618978487,
      "grad_norm": 0.1300048679113388,
      "learning_rate": 2.6901765234136428e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 30430592,
      "step": 34040
    },
    {
      "epoch": 8.985350402533985,
      "grad_norm": 0.04834938049316406,
      "learning_rate": 2.685748017172063e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 30435232,
      "step": 34045
    },
    {
      "epoch": 8.986670186089482,
      "grad_norm": 0.22620095312595367,
      "learning_rate": 2.681322952068549e-06,
      "loss": 0.0661,
      "num_input_tokens_seen": 30439840,
      "step": 34050
    },
    {
      "epoch": 8.987989969644978,
      "grad_norm": 0.2332710325717926,
      "learning_rate": 2.6769013287855137e-06,
      "loss": 0.0471,
      "num_input_tokens_seen": 30443840,
      "step": 34055
    },
    {
      "epoch": 8.989309753200475,
      "grad_norm": 0.0665879175066948,
      "learning_rate": 2.6724831480048286e-06,
      "loss": 0.0481,
      "num_input_tokens_seen": 30448128,
      "step": 34060
    },
    {
      "epoch": 8.990629536755971,
      "grad_norm": 0.07129260897636414,
      "learning_rate": 2.66806841040782e-06,
      "loss": 0.0586,
      "num_input_tokens_seen": 30452544,
      "step": 34065
    },
    {
      "epoch": 8.99194932031147,
      "grad_norm": 0.019878221675753593,
      "learning_rate": 2.6636571166753083e-06,
      "loss": 0.0299,
      "num_input_tokens_seen": 30457088,
      "step": 34070
    },
    {
      "epoch": 8.993269103866966,
      "grad_norm": 0.03723651170730591,
      "learning_rate": 2.6592492674875598e-06,
      "loss": 0.035,
      "num_input_tokens_seen": 30461472,
      "step": 34075
    },
    {
      "epoch": 8.994588887422463,
      "grad_norm": 0.026776984333992004,
      "learning_rate": 2.6548448635243305e-06,
      "loss": 0.0263,
      "num_input_tokens_seen": 30466080,
      "step": 34080
    },
    {
      "epoch": 8.99590867097796,
      "grad_norm": 0.34618017077445984,
      "learning_rate": 2.650443905464828e-06,
      "loss": 0.059,
      "num_input_tokens_seen": 30470624,
      "step": 34085
    },
    {
      "epoch": 8.997228454533456,
      "grad_norm": 0.15927749872207642,
      "learning_rate": 2.646046393987739e-06,
      "loss": 0.0805,
      "num_input_tokens_seen": 30475488,
      "step": 34090
    },
    {
      "epoch": 8.998548238088954,
      "grad_norm": 0.1500425934791565,
      "learning_rate": 2.64165232977121e-06,
      "loss": 0.1179,
      "num_input_tokens_seen": 30480064,
      "step": 34095
    },
    {
      "epoch": 8.99986802164445,
      "grad_norm": 0.2765823304653168,
      "learning_rate": 2.6372617134928695e-06,
      "loss": 0.0756,
      "num_input_tokens_seen": 30484480,
      "step": 34100
    },
    {
      "epoch": 9.001055826844398,
      "grad_norm": 0.14811013638973236,
      "learning_rate": 2.6328745458297943e-06,
      "loss": 0.0748,
      "num_input_tokens_seen": 30488672,
      "step": 34105
    },
    {
      "epoch": 9.002375610399895,
      "grad_norm": 0.3726392388343811,
      "learning_rate": 2.6284908274585546e-06,
      "loss": 0.0803,
      "num_input_tokens_seen": 30492960,
      "step": 34110
    },
    {
      "epoch": 9.003695393955391,
      "grad_norm": 0.24268601834774017,
      "learning_rate": 2.6241105590551595e-06,
      "loss": 0.091,
      "num_input_tokens_seen": 30497120,
      "step": 34115
    },
    {
      "epoch": 9.005015177510888,
      "grad_norm": 0.16123487055301666,
      "learning_rate": 2.6197337412951105e-06,
      "loss": 0.0302,
      "num_input_tokens_seen": 30501664,
      "step": 34120
    },
    {
      "epoch": 9.006334961066385,
      "grad_norm": 0.24793678522109985,
      "learning_rate": 2.6153603748533705e-06,
      "loss": 0.0551,
      "num_input_tokens_seen": 30506080,
      "step": 34125
    },
    {
      "epoch": 9.007654744621883,
      "grad_norm": 0.11255044490098953,
      "learning_rate": 2.6109904604043585e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 30510912,
      "step": 34130
    },
    {
      "epoch": 9.00897452817738,
      "grad_norm": 0.20901846885681152,
      "learning_rate": 2.6066239986219765e-06,
      "loss": 0.0383,
      "num_input_tokens_seen": 30515616,
      "step": 34135
    },
    {
      "epoch": 9.010294311732876,
      "grad_norm": 0.37452232837677,
      "learning_rate": 2.602260990179592e-06,
      "loss": 0.0772,
      "num_input_tokens_seen": 30519744,
      "step": 34140
    },
    {
      "epoch": 9.011614095288373,
      "grad_norm": 0.7276430130004883,
      "learning_rate": 2.5979014357500248e-06,
      "loss": 0.0209,
      "num_input_tokens_seen": 30524160,
      "step": 34145
    },
    {
      "epoch": 9.012933878843869,
      "grad_norm": 0.11489791423082352,
      "learning_rate": 2.5935453360055844e-06,
      "loss": 0.0511,
      "num_input_tokens_seen": 30528768,
      "step": 34150
    },
    {
      "epoch": 9.014253662399366,
      "grad_norm": 0.3230082392692566,
      "learning_rate": 2.5891926916180283e-06,
      "loss": 0.0453,
      "num_input_tokens_seen": 30533408,
      "step": 34155
    },
    {
      "epoch": 9.015573445954864,
      "grad_norm": 0.04839889332652092,
      "learning_rate": 2.5848435032585883e-06,
      "loss": 0.0713,
      "num_input_tokens_seen": 30538336,
      "step": 34160
    },
    {
      "epoch": 9.01689322951036,
      "grad_norm": 0.16654208302497864,
      "learning_rate": 2.58049777159797e-06,
      "loss": 0.0597,
      "num_input_tokens_seen": 30542816,
      "step": 34165
    },
    {
      "epoch": 9.018213013065857,
      "grad_norm": 0.06050413101911545,
      "learning_rate": 2.576155497306332e-06,
      "loss": 0.0854,
      "num_input_tokens_seen": 30547392,
      "step": 34170
    },
    {
      "epoch": 9.019532796621354,
      "grad_norm": 0.071187324821949,
      "learning_rate": 2.57181668105331e-06,
      "loss": 0.0589,
      "num_input_tokens_seen": 30551712,
      "step": 34175
    },
    {
      "epoch": 9.02085258017685,
      "grad_norm": 0.16261057555675507,
      "learning_rate": 2.567481323508014e-06,
      "loss": 0.0663,
      "num_input_tokens_seen": 30556224,
      "step": 34180
    },
    {
      "epoch": 9.022172363732349,
      "grad_norm": 0.14340335130691528,
      "learning_rate": 2.5631494253389954e-06,
      "loss": 0.0505,
      "num_input_tokens_seen": 30560608,
      "step": 34185
    },
    {
      "epoch": 9.023492147287845,
      "grad_norm": 0.2412203997373581,
      "learning_rate": 2.5588209872142997e-06,
      "loss": 0.0391,
      "num_input_tokens_seen": 30565120,
      "step": 34190
    },
    {
      "epoch": 9.024811930843342,
      "grad_norm": 0.05711883306503296,
      "learning_rate": 2.5544960098014186e-06,
      "loss": 0.0646,
      "num_input_tokens_seen": 30569344,
      "step": 34195
    },
    {
      "epoch": 9.026131714398838,
      "grad_norm": 0.06810269504785538,
      "learning_rate": 2.550174493767318e-06,
      "loss": 0.029,
      "num_input_tokens_seen": 30573920,
      "step": 34200
    },
    {
      "epoch": 9.026131714398838,
      "eval_loss": 0.06678295135498047,
      "eval_runtime": 64.7567,
      "eval_samples_per_second": 104.005,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 30573920,
      "step": 34200
    },
    {
      "epoch": 9.027451497954335,
      "grad_norm": 0.3600969612598419,
      "learning_rate": 2.545856439778438e-06,
      "loss": 0.0885,
      "num_input_tokens_seen": 30578464,
      "step": 34205
    },
    {
      "epoch": 9.028771281509833,
      "grad_norm": 0.38368111848831177,
      "learning_rate": 2.541541848500667e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 30583072,
      "step": 34210
    },
    {
      "epoch": 9.03009106506533,
      "grad_norm": 0.21490925550460815,
      "learning_rate": 2.5372307205993733e-06,
      "loss": 0.0293,
      "num_input_tokens_seen": 30587520,
      "step": 34215
    },
    {
      "epoch": 9.031410848620826,
      "grad_norm": 0.063166543841362,
      "learning_rate": 2.5329230567393917e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 30592192,
      "step": 34220
    },
    {
      "epoch": 9.032730632176323,
      "grad_norm": 0.16628064215183258,
      "learning_rate": 2.5286188575850164e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 30596736,
      "step": 34225
    },
    {
      "epoch": 9.03405041573182,
      "grad_norm": 0.06292189657688141,
      "learning_rate": 2.5243181237999984e-06,
      "loss": 0.0576,
      "num_input_tokens_seen": 30601312,
      "step": 34230
    },
    {
      "epoch": 9.035370199287318,
      "grad_norm": 0.054911937564611435,
      "learning_rate": 2.520020856047578e-06,
      "loss": 0.0715,
      "num_input_tokens_seen": 30605728,
      "step": 34235
    },
    {
      "epoch": 9.036689982842814,
      "grad_norm": 0.25926274061203003,
      "learning_rate": 2.515727054990438e-06,
      "loss": 0.0753,
      "num_input_tokens_seen": 30610208,
      "step": 34240
    },
    {
      "epoch": 9.03800976639831,
      "grad_norm": 0.21004317700862885,
      "learning_rate": 2.511436721290747e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 30614560,
      "step": 34245
    },
    {
      "epoch": 9.039329549953807,
      "grad_norm": 0.31882160902023315,
      "learning_rate": 2.5071498556101164e-06,
      "loss": 0.0776,
      "num_input_tokens_seen": 30618944,
      "step": 34250
    },
    {
      "epoch": 9.040649333509304,
      "grad_norm": 0.16440081596374512,
      "learning_rate": 2.5028664586096485e-06,
      "loss": 0.0728,
      "num_input_tokens_seen": 30623456,
      "step": 34255
    },
    {
      "epoch": 9.041969117064802,
      "grad_norm": 0.06106419861316681,
      "learning_rate": 2.498586530949881e-06,
      "loss": 0.0366,
      "num_input_tokens_seen": 30627648,
      "step": 34260
    },
    {
      "epoch": 9.043288900620299,
      "grad_norm": 0.1860082894563675,
      "learning_rate": 2.4943100732908427e-06,
      "loss": 0.0584,
      "num_input_tokens_seen": 30632096,
      "step": 34265
    },
    {
      "epoch": 9.044608684175795,
      "grad_norm": 0.19812174141407013,
      "learning_rate": 2.4900370862920188e-06,
      "loss": 0.0562,
      "num_input_tokens_seen": 30636224,
      "step": 34270
    },
    {
      "epoch": 9.045928467731292,
      "grad_norm": 0.02721893973648548,
      "learning_rate": 2.4857675706123518e-06,
      "loss": 0.0297,
      "num_input_tokens_seen": 30640896,
      "step": 34275
    },
    {
      "epoch": 9.047248251286788,
      "grad_norm": 0.18956708908081055,
      "learning_rate": 2.4815015269102543e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 30645216,
      "step": 34280
    },
    {
      "epoch": 9.048568034842285,
      "grad_norm": 0.05388544872403145,
      "learning_rate": 2.477238955843611e-06,
      "loss": 0.066,
      "num_input_tokens_seen": 30649376,
      "step": 34285
    },
    {
      "epoch": 9.049887818397783,
      "grad_norm": 0.2598036825656891,
      "learning_rate": 2.4729798580697573e-06,
      "loss": 0.05,
      "num_input_tokens_seen": 30653888,
      "step": 34290
    },
    {
      "epoch": 9.05120760195328,
      "grad_norm": 0.19086545705795288,
      "learning_rate": 2.4687242342455034e-06,
      "loss": 0.0329,
      "num_input_tokens_seen": 30658592,
      "step": 34295
    },
    {
      "epoch": 9.052527385508776,
      "grad_norm": 0.12978072464466095,
      "learning_rate": 2.4644720850271196e-06,
      "loss": 0.0717,
      "num_input_tokens_seen": 30663008,
      "step": 34300
    },
    {
      "epoch": 9.053847169064273,
      "grad_norm": 0.3398091793060303,
      "learning_rate": 2.4602234110703364e-06,
      "loss": 0.0458,
      "num_input_tokens_seen": 30667456,
      "step": 34305
    },
    {
      "epoch": 9.05516695261977,
      "grad_norm": 0.09530085325241089,
      "learning_rate": 2.4559782130303576e-06,
      "loss": 0.0495,
      "num_input_tokens_seen": 30671808,
      "step": 34310
    },
    {
      "epoch": 9.056486736175268,
      "grad_norm": 0.08488544821739197,
      "learning_rate": 2.451736491561843e-06,
      "loss": 0.0185,
      "num_input_tokens_seen": 30676544,
      "step": 34315
    },
    {
      "epoch": 9.057806519730764,
      "grad_norm": 0.4272072911262512,
      "learning_rate": 2.4474982473189163e-06,
      "loss": 0.1205,
      "num_input_tokens_seen": 30681280,
      "step": 34320
    },
    {
      "epoch": 9.059126303286261,
      "grad_norm": 0.11024122685194016,
      "learning_rate": 2.4432634809551796e-06,
      "loss": 0.0546,
      "num_input_tokens_seen": 30685984,
      "step": 34325
    },
    {
      "epoch": 9.060446086841758,
      "grad_norm": 0.37305113673210144,
      "learning_rate": 2.439032193123675e-06,
      "loss": 0.0769,
      "num_input_tokens_seen": 30690304,
      "step": 34330
    },
    {
      "epoch": 9.061765870397254,
      "grad_norm": 0.3001219630241394,
      "learning_rate": 2.4348043844769297e-06,
      "loss": 0.0472,
      "num_input_tokens_seen": 30694752,
      "step": 34335
    },
    {
      "epoch": 9.063085653952752,
      "grad_norm": 0.2659684121608734,
      "learning_rate": 2.4305800556669146e-06,
      "loss": 0.0742,
      "num_input_tokens_seen": 30699200,
      "step": 34340
    },
    {
      "epoch": 9.064405437508249,
      "grad_norm": 0.18000783026218414,
      "learning_rate": 2.426359207345083e-06,
      "loss": 0.0558,
      "num_input_tokens_seen": 30703904,
      "step": 34345
    },
    {
      "epoch": 9.065725221063746,
      "grad_norm": 0.27730628848075867,
      "learning_rate": 2.4221418401623396e-06,
      "loss": 0.0556,
      "num_input_tokens_seen": 30708128,
      "step": 34350
    },
    {
      "epoch": 9.067045004619242,
      "grad_norm": 0.09412769228219986,
      "learning_rate": 2.4179279547690557e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 30712480,
      "step": 34355
    },
    {
      "epoch": 9.068364788174739,
      "grad_norm": 0.08771993219852448,
      "learning_rate": 2.413717551815062e-06,
      "loss": 0.0807,
      "num_input_tokens_seen": 30716928,
      "step": 34360
    },
    {
      "epoch": 9.069684571730237,
      "grad_norm": 0.16841833293437958,
      "learning_rate": 2.409510631949666e-06,
      "loss": 0.0283,
      "num_input_tokens_seen": 30721920,
      "step": 34365
    },
    {
      "epoch": 9.071004355285734,
      "grad_norm": 0.1209721565246582,
      "learning_rate": 2.405307195821618e-06,
      "loss": 0.0847,
      "num_input_tokens_seen": 30726496,
      "step": 34370
    },
    {
      "epoch": 9.07232413884123,
      "grad_norm": 0.07847556471824646,
      "learning_rate": 2.4011072440791372e-06,
      "loss": 0.0775,
      "num_input_tokens_seen": 30731072,
      "step": 34375
    },
    {
      "epoch": 9.073643922396727,
      "grad_norm": 0.22748929262161255,
      "learning_rate": 2.3969107773699233e-06,
      "loss": 0.1007,
      "num_input_tokens_seen": 30735488,
      "step": 34380
    },
    {
      "epoch": 9.074963705952223,
      "grad_norm": 0.2644062042236328,
      "learning_rate": 2.3927177963411096e-06,
      "loss": 0.0482,
      "num_input_tokens_seen": 30740256,
      "step": 34385
    },
    {
      "epoch": 9.076283489507722,
      "grad_norm": 0.34842315316200256,
      "learning_rate": 2.3885283016393144e-06,
      "loss": 0.1058,
      "num_input_tokens_seen": 30744704,
      "step": 34390
    },
    {
      "epoch": 9.077603273063218,
      "grad_norm": 0.2846302390098572,
      "learning_rate": 2.3843422939106076e-06,
      "loss": 0.0838,
      "num_input_tokens_seen": 30749120,
      "step": 34395
    },
    {
      "epoch": 9.078923056618715,
      "grad_norm": 0.0994100570678711,
      "learning_rate": 2.380159773800525e-06,
      "loss": 0.0564,
      "num_input_tokens_seen": 30753536,
      "step": 34400
    },
    {
      "epoch": 9.078923056618715,
      "eval_loss": 0.0668676346540451,
      "eval_runtime": 64.733,
      "eval_samples_per_second": 104.043,
      "eval_steps_per_second": 26.015,
      "num_input_tokens_seen": 30753536,
      "step": 34400
    },
    {
      "epoch": 9.080242840174211,
      "grad_norm": 0.2420606166124344,
      "learning_rate": 2.3759807419540675e-06,
      "loss": 0.0872,
      "num_input_tokens_seen": 30758016,
      "step": 34405
    },
    {
      "epoch": 9.081562623729708,
      "grad_norm": 0.045871444046497345,
      "learning_rate": 2.3718051990156835e-06,
      "loss": 0.0547,
      "num_input_tokens_seen": 30762432,
      "step": 34410
    },
    {
      "epoch": 9.082882407285204,
      "grad_norm": 0.3601616621017456,
      "learning_rate": 2.367633145629311e-06,
      "loss": 0.0954,
      "num_input_tokens_seen": 30766944,
      "step": 34415
    },
    {
      "epoch": 9.084202190840703,
      "grad_norm": 0.1162792444229126,
      "learning_rate": 2.363464582438316e-06,
      "loss": 0.0408,
      "num_input_tokens_seen": 30771392,
      "step": 34420
    },
    {
      "epoch": 9.0855219743962,
      "grad_norm": 0.2849809229373932,
      "learning_rate": 2.3592995100855526e-06,
      "loss": 0.0962,
      "num_input_tokens_seen": 30775744,
      "step": 34425
    },
    {
      "epoch": 9.086841757951696,
      "grad_norm": 0.021222257986664772,
      "learning_rate": 2.3551379292133273e-06,
      "loss": 0.089,
      "num_input_tokens_seen": 30780224,
      "step": 34430
    },
    {
      "epoch": 9.088161541507192,
      "grad_norm": 0.19661718606948853,
      "learning_rate": 2.3509798404634047e-06,
      "loss": 0.0506,
      "num_input_tokens_seen": 30785152,
      "step": 34435
    },
    {
      "epoch": 9.089481325062689,
      "grad_norm": 0.08007462322711945,
      "learning_rate": 2.346825244477019e-06,
      "loss": 0.0318,
      "num_input_tokens_seen": 30789760,
      "step": 34440
    },
    {
      "epoch": 9.090801108618187,
      "grad_norm": 0.12793056666851044,
      "learning_rate": 2.3426741418948545e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 30794336,
      "step": 34445
    },
    {
      "epoch": 9.092120892173684,
      "grad_norm": 0.2705821394920349,
      "learning_rate": 2.3385265333570715e-06,
      "loss": 0.0845,
      "num_input_tokens_seen": 30798944,
      "step": 34450
    },
    {
      "epoch": 9.09344067572918,
      "grad_norm": 0.240452378988266,
      "learning_rate": 2.334382419503278e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 30803200,
      "step": 34455
    },
    {
      "epoch": 9.094760459284677,
      "grad_norm": 0.2875104546546936,
      "learning_rate": 2.3302418009725465e-06,
      "loss": 0.0802,
      "num_input_tokens_seen": 30807776,
      "step": 34460
    },
    {
      "epoch": 9.096080242840173,
      "grad_norm": 0.2622779607772827,
      "learning_rate": 2.326104678403415e-06,
      "loss": 0.0503,
      "num_input_tokens_seen": 30812064,
      "step": 34465
    },
    {
      "epoch": 9.097400026395672,
      "grad_norm": 0.37460052967071533,
      "learning_rate": 2.321971052433883e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 30816384,
      "step": 34470
    },
    {
      "epoch": 9.098719809951168,
      "grad_norm": 0.4294571578502655,
      "learning_rate": 2.3178409237014004e-06,
      "loss": 0.0912,
      "num_input_tokens_seen": 30820800,
      "step": 34475
    },
    {
      "epoch": 9.100039593506665,
      "grad_norm": 0.09951585531234741,
      "learning_rate": 2.313714292842889e-06,
      "loss": 0.0259,
      "num_input_tokens_seen": 30825312,
      "step": 34480
    },
    {
      "epoch": 9.101359377062161,
      "grad_norm": 0.051830872893333435,
      "learning_rate": 2.309591160494734e-06,
      "loss": 0.0729,
      "num_input_tokens_seen": 30829536,
      "step": 34485
    },
    {
      "epoch": 9.102679160617658,
      "grad_norm": 0.21221770346164703,
      "learning_rate": 2.305471527292763e-06,
      "loss": 0.0387,
      "num_input_tokens_seen": 30833856,
      "step": 34490
    },
    {
      "epoch": 9.103998944173156,
      "grad_norm": 0.07946129888296127,
      "learning_rate": 2.3013553938722817e-06,
      "loss": 0.0703,
      "num_input_tokens_seen": 30838560,
      "step": 34495
    },
    {
      "epoch": 9.105318727728653,
      "grad_norm": 0.19412478804588318,
      "learning_rate": 2.297242760868043e-06,
      "loss": 0.0911,
      "num_input_tokens_seen": 30842848,
      "step": 34500
    },
    {
      "epoch": 9.10663851128415,
      "grad_norm": 0.2251584231853485,
      "learning_rate": 2.2931336289142735e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 30847328,
      "step": 34505
    },
    {
      "epoch": 9.107958294839646,
      "grad_norm": 0.0960526242852211,
      "learning_rate": 2.289027998644655e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 30852000,
      "step": 34510
    },
    {
      "epoch": 9.109278078395143,
      "grad_norm": 0.08653322607278824,
      "learning_rate": 2.2849258706923228e-06,
      "loss": 0.06,
      "num_input_tokens_seen": 30856512,
      "step": 34515
    },
    {
      "epoch": 9.110597861950641,
      "grad_norm": 0.04940854012966156,
      "learning_rate": 2.2808272456898705e-06,
      "loss": 0.0515,
      "num_input_tokens_seen": 30860896,
      "step": 34520
    },
    {
      "epoch": 9.111917645506137,
      "grad_norm": 0.2512252926826477,
      "learning_rate": 2.2767321242693707e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 30865504,
      "step": 34525
    },
    {
      "epoch": 9.113237429061634,
      "grad_norm": 0.26358452439308167,
      "learning_rate": 2.272640507062329e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 30870080,
      "step": 34530
    },
    {
      "epoch": 9.11455721261713,
      "grad_norm": 0.10302529484033585,
      "learning_rate": 2.2685523946997382e-06,
      "loss": 0.0461,
      "num_input_tokens_seen": 30874048,
      "step": 34535
    },
    {
      "epoch": 9.115876996172627,
      "grad_norm": 0.024606136605143547,
      "learning_rate": 2.2644677878120245e-06,
      "loss": 0.0802,
      "num_input_tokens_seen": 30878624,
      "step": 34540
    },
    {
      "epoch": 9.117196779728125,
      "grad_norm": 0.11763212829828262,
      "learning_rate": 2.2603866870290897e-06,
      "loss": 0.0399,
      "num_input_tokens_seen": 30883264,
      "step": 34545
    },
    {
      "epoch": 9.118516563283622,
      "grad_norm": 0.06879474967718124,
      "learning_rate": 2.256309092980294e-06,
      "loss": 0.0326,
      "num_input_tokens_seen": 30887712,
      "step": 34550
    },
    {
      "epoch": 9.119836346839119,
      "grad_norm": 0.26831358671188354,
      "learning_rate": 2.252235006294448e-06,
      "loss": 0.0735,
      "num_input_tokens_seen": 30891872,
      "step": 34555
    },
    {
      "epoch": 9.121156130394615,
      "grad_norm": 0.21659240126609802,
      "learning_rate": 2.2481644275998333e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 30896096,
      "step": 34560
    },
    {
      "epoch": 9.122475913950112,
      "grad_norm": 0.6196189522743225,
      "learning_rate": 2.2440973575241832e-06,
      "loss": 0.0711,
      "num_input_tokens_seen": 30900480,
      "step": 34565
    },
    {
      "epoch": 9.123795697505608,
      "grad_norm": 0.05162690952420235,
      "learning_rate": 2.240033796694685e-06,
      "loss": 0.0261,
      "num_input_tokens_seen": 30905248,
      "step": 34570
    },
    {
      "epoch": 9.125115481061107,
      "grad_norm": 0.18388321995735168,
      "learning_rate": 2.235973745737999e-06,
      "loss": 0.0894,
      "num_input_tokens_seen": 30909824,
      "step": 34575
    },
    {
      "epoch": 9.126435264616603,
      "grad_norm": 0.17426635324954987,
      "learning_rate": 2.2319172052802263e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 30914240,
      "step": 34580
    },
    {
      "epoch": 9.1277550481721,
      "grad_norm": 0.08125588297843933,
      "learning_rate": 2.2278641759469477e-06,
      "loss": 0.0804,
      "num_input_tokens_seen": 30918624,
      "step": 34585
    },
    {
      "epoch": 9.129074831727596,
      "grad_norm": 0.2528926134109497,
      "learning_rate": 2.2238146583631825e-06,
      "loss": 0.0466,
      "num_input_tokens_seen": 30922816,
      "step": 34590
    },
    {
      "epoch": 9.130394615283093,
      "grad_norm": 0.16479706764221191,
      "learning_rate": 2.2197686531534256e-06,
      "loss": 0.0484,
      "num_input_tokens_seen": 30927296,
      "step": 34595
    },
    {
      "epoch": 9.131714398838591,
      "grad_norm": 0.5828876495361328,
      "learning_rate": 2.2157261609416087e-06,
      "loss": 0.1248,
      "num_input_tokens_seen": 30931776,
      "step": 34600
    },
    {
      "epoch": 9.131714398838591,
      "eval_loss": 0.0668899342417717,
      "eval_runtime": 64.7287,
      "eval_samples_per_second": 104.05,
      "eval_steps_per_second": 26.016,
      "num_input_tokens_seen": 30931776,
      "step": 34600
    },
    {
      "epoch": 9.133034182394088,
      "grad_norm": 0.3236760199069977,
      "learning_rate": 2.211687182351149e-06,
      "loss": 0.0595,
      "num_input_tokens_seen": 30936160,
      "step": 34605
    },
    {
      "epoch": 9.134353965949584,
      "grad_norm": 0.027450822293758392,
      "learning_rate": 2.2076517180048993e-06,
      "loss": 0.0496,
      "num_input_tokens_seen": 30940416,
      "step": 34610
    },
    {
      "epoch": 9.13567374950508,
      "grad_norm": 0.22930891811847687,
      "learning_rate": 2.2036197685251834e-06,
      "loss": 0.0593,
      "num_input_tokens_seen": 30944448,
      "step": 34615
    },
    {
      "epoch": 9.136993533060577,
      "grad_norm": 0.22151413559913635,
      "learning_rate": 2.199591334533771e-06,
      "loss": 0.064,
      "num_input_tokens_seen": 30948672,
      "step": 34620
    },
    {
      "epoch": 9.138313316616076,
      "grad_norm": 0.06415620446205139,
      "learning_rate": 2.1955664166519036e-06,
      "loss": 0.0455,
      "num_input_tokens_seen": 30953248,
      "step": 34625
    },
    {
      "epoch": 9.139633100171572,
      "grad_norm": 0.11579170823097229,
      "learning_rate": 2.1915450155002793e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 30957760,
      "step": 34630
    },
    {
      "epoch": 9.140952883727069,
      "grad_norm": 0.09477139264345169,
      "learning_rate": 2.187527131699038e-06,
      "loss": 0.0963,
      "num_input_tokens_seen": 30962432,
      "step": 34635
    },
    {
      "epoch": 9.142272667282565,
      "grad_norm": 0.09049534797668457,
      "learning_rate": 2.18351276586779e-06,
      "loss": 0.0566,
      "num_input_tokens_seen": 30966624,
      "step": 34640
    },
    {
      "epoch": 9.143592450838062,
      "grad_norm": 0.08964726328849792,
      "learning_rate": 2.1795019186256092e-06,
      "loss": 0.0379,
      "num_input_tokens_seen": 30970816,
      "step": 34645
    },
    {
      "epoch": 9.14491223439356,
      "grad_norm": 0.048677053302526474,
      "learning_rate": 2.1754945905910094e-06,
      "loss": 0.088,
      "num_input_tokens_seen": 30975200,
      "step": 34650
    },
    {
      "epoch": 9.146232017949057,
      "grad_norm": 0.09771867841482162,
      "learning_rate": 2.171490782381977e-06,
      "loss": 0.0601,
      "num_input_tokens_seen": 30979680,
      "step": 34655
    },
    {
      "epoch": 9.147551801504553,
      "grad_norm": 0.05459333583712578,
      "learning_rate": 2.1674904946159425e-06,
      "loss": 0.0318,
      "num_input_tokens_seen": 30984160,
      "step": 34660
    },
    {
      "epoch": 9.14887158506005,
      "grad_norm": 0.17346391081809998,
      "learning_rate": 2.16349372790981e-06,
      "loss": 0.0767,
      "num_input_tokens_seen": 30988768,
      "step": 34665
    },
    {
      "epoch": 9.150191368615546,
      "grad_norm": 0.23871064186096191,
      "learning_rate": 2.159500482879928e-06,
      "loss": 0.0828,
      "num_input_tokens_seen": 30993536,
      "step": 34670
    },
    {
      "epoch": 9.151511152171043,
      "grad_norm": 0.07678995281457901,
      "learning_rate": 2.155510760142096e-06,
      "loss": 0.0707,
      "num_input_tokens_seen": 30998080,
      "step": 34675
    },
    {
      "epoch": 9.152830935726541,
      "grad_norm": 0.2861528694629669,
      "learning_rate": 2.151524560311588e-06,
      "loss": 0.041,
      "num_input_tokens_seen": 31002816,
      "step": 34680
    },
    {
      "epoch": 9.154150719282038,
      "grad_norm": 0.15417854487895966,
      "learning_rate": 2.147541884003129e-06,
      "loss": 0.0551,
      "num_input_tokens_seen": 31007232,
      "step": 34685
    },
    {
      "epoch": 9.155470502837534,
      "grad_norm": 0.155519500374794,
      "learning_rate": 2.1435627318308895e-06,
      "loss": 0.0512,
      "num_input_tokens_seen": 31011904,
      "step": 34690
    },
    {
      "epoch": 9.156790286393031,
      "grad_norm": 0.3594532907009125,
      "learning_rate": 2.139587104408511e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 31016640,
      "step": 34695
    },
    {
      "epoch": 9.158110069948528,
      "grad_norm": 0.20393440127372742,
      "learning_rate": 2.1356150023490783e-06,
      "loss": 0.0544,
      "num_input_tokens_seen": 31021440,
      "step": 34700
    },
    {
      "epoch": 9.159429853504026,
      "grad_norm": 0.09567544609308243,
      "learning_rate": 2.1316464262651464e-06,
      "loss": 0.0576,
      "num_input_tokens_seen": 31026016,
      "step": 34705
    },
    {
      "epoch": 9.160749637059522,
      "grad_norm": 0.16381627321243286,
      "learning_rate": 2.1276813767687224e-06,
      "loss": 0.0876,
      "num_input_tokens_seen": 31030688,
      "step": 34710
    },
    {
      "epoch": 9.162069420615019,
      "grad_norm": 0.19003786146640778,
      "learning_rate": 2.123719854471254e-06,
      "loss": 0.0295,
      "num_input_tokens_seen": 31035200,
      "step": 34715
    },
    {
      "epoch": 9.163389204170516,
      "grad_norm": 0.33006224036216736,
      "learning_rate": 2.119761859983668e-06,
      "loss": 0.0813,
      "num_input_tokens_seen": 31039616,
      "step": 34720
    },
    {
      "epoch": 9.164708987726012,
      "grad_norm": 0.10440487414598465,
      "learning_rate": 2.1158073939163386e-06,
      "loss": 0.0316,
      "num_input_tokens_seen": 31044000,
      "step": 34725
    },
    {
      "epoch": 9.16602877128151,
      "grad_norm": 0.19474081695079803,
      "learning_rate": 2.111856456879088e-06,
      "loss": 0.0558,
      "num_input_tokens_seen": 31048480,
      "step": 34730
    },
    {
      "epoch": 9.167348554837007,
      "grad_norm": 0.023104684427380562,
      "learning_rate": 2.1079090494811993e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 31052928,
      "step": 34735
    },
    {
      "epoch": 9.168668338392504,
      "grad_norm": 0.027050459757447243,
      "learning_rate": 2.103965172331418e-06,
      "loss": 0.0954,
      "num_input_tokens_seen": 31057408,
      "step": 34740
    },
    {
      "epoch": 9.169988121948,
      "grad_norm": 0.2050352841615677,
      "learning_rate": 2.100024826037933e-06,
      "loss": 0.0452,
      "num_input_tokens_seen": 31061952,
      "step": 34745
    },
    {
      "epoch": 9.171307905503497,
      "grad_norm": 0.09616685658693314,
      "learning_rate": 2.0960880112084027e-06,
      "loss": 0.0757,
      "num_input_tokens_seen": 31066496,
      "step": 34750
    },
    {
      "epoch": 9.172627689058995,
      "grad_norm": 0.30344218015670776,
      "learning_rate": 2.092154728449927e-06,
      "loss": 0.0536,
      "num_input_tokens_seen": 31071104,
      "step": 34755
    },
    {
      "epoch": 9.173947472614492,
      "grad_norm": 0.0866456851363182,
      "learning_rate": 2.0882249783690687e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 31075424,
      "step": 34760
    },
    {
      "epoch": 9.175267256169988,
      "grad_norm": 0.19021546840667725,
      "learning_rate": 2.084298761571851e-06,
      "loss": 0.0932,
      "num_input_tokens_seen": 31079584,
      "step": 34765
    },
    {
      "epoch": 9.176587039725485,
      "grad_norm": 0.1955973207950592,
      "learning_rate": 2.080376078663737e-06,
      "loss": 0.0566,
      "num_input_tokens_seen": 31084256,
      "step": 34770
    },
    {
      "epoch": 9.177906823280981,
      "grad_norm": 0.11902330070734024,
      "learning_rate": 2.0764569302496593e-06,
      "loss": 0.1028,
      "num_input_tokens_seen": 31088512,
      "step": 34775
    },
    {
      "epoch": 9.17922660683648,
      "grad_norm": 0.1522032767534256,
      "learning_rate": 2.0725413169339957e-06,
      "loss": 0.0561,
      "num_input_tokens_seen": 31092960,
      "step": 34780
    },
    {
      "epoch": 9.180546390391976,
      "grad_norm": 0.22749671339988708,
      "learning_rate": 2.068629239320588e-06,
      "loss": 0.065,
      "num_input_tokens_seen": 31097120,
      "step": 34785
    },
    {
      "epoch": 9.181866173947473,
      "grad_norm": 0.11153033375740051,
      "learning_rate": 2.064720698012726e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 31101632,
      "step": 34790
    },
    {
      "epoch": 9.18318595750297,
      "grad_norm": 0.15709128975868225,
      "learning_rate": 2.0608156936131522e-06,
      "loss": 0.0393,
      "num_input_tokens_seen": 31106240,
      "step": 34795
    },
    {
      "epoch": 9.184505741058466,
      "grad_norm": 0.2918480634689331,
      "learning_rate": 2.056914226724074e-06,
      "loss": 0.0469,
      "num_input_tokens_seen": 31110592,
      "step": 34800
    },
    {
      "epoch": 9.184505741058466,
      "eval_loss": 0.06680532544851303,
      "eval_runtime": 64.7679,
      "eval_samples_per_second": 103.987,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 31110592,
      "step": 34800
    },
    {
      "epoch": 9.185825524613964,
      "grad_norm": 0.4884013533592224,
      "learning_rate": 2.0530162979471385e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 31115008,
      "step": 34805
    },
    {
      "epoch": 9.18714530816946,
      "grad_norm": 0.04927368834614754,
      "learning_rate": 2.0491219078834667e-06,
      "loss": 0.1058,
      "num_input_tokens_seen": 31119264,
      "step": 34810
    },
    {
      "epoch": 9.188465091724957,
      "grad_norm": 0.17808640003204346,
      "learning_rate": 2.045231057133612e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 31123872,
      "step": 34815
    },
    {
      "epoch": 9.189784875280454,
      "grad_norm": 0.22612524032592773,
      "learning_rate": 2.0413437462975944e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 31128512,
      "step": 34820
    },
    {
      "epoch": 9.19110465883595,
      "grad_norm": 0.23320885002613068,
      "learning_rate": 2.0374599759748843e-06,
      "loss": 0.0906,
      "num_input_tokens_seen": 31133152,
      "step": 34825
    },
    {
      "epoch": 9.192424442391447,
      "grad_norm": 0.5458745956420898,
      "learning_rate": 2.033579746764419e-06,
      "loss": 0.046,
      "num_input_tokens_seen": 31137568,
      "step": 34830
    },
    {
      "epoch": 9.193744225946945,
      "grad_norm": 0.24943146109580994,
      "learning_rate": 2.029703059264565e-06,
      "loss": 0.0368,
      "num_input_tokens_seen": 31142112,
      "step": 34835
    },
    {
      "epoch": 9.195064009502442,
      "grad_norm": 0.2626420855522156,
      "learning_rate": 2.02582991407316e-06,
      "loss": 0.0624,
      "num_input_tokens_seen": 31146400,
      "step": 34840
    },
    {
      "epoch": 9.196383793057938,
      "grad_norm": 0.1136518269777298,
      "learning_rate": 2.0219603117874992e-06,
      "loss": 0.0567,
      "num_input_tokens_seen": 31150816,
      "step": 34845
    },
    {
      "epoch": 9.197703576613435,
      "grad_norm": 0.08701056241989136,
      "learning_rate": 2.0180942530043156e-06,
      "loss": 0.0286,
      "num_input_tokens_seen": 31155360,
      "step": 34850
    },
    {
      "epoch": 9.199023360168932,
      "grad_norm": 0.208920419216156,
      "learning_rate": 2.0142317383198107e-06,
      "loss": 0.0741,
      "num_input_tokens_seen": 31159968,
      "step": 34855
    },
    {
      "epoch": 9.20034314372443,
      "grad_norm": 0.32918015122413635,
      "learning_rate": 2.0103727683296243e-06,
      "loss": 0.1289,
      "num_input_tokens_seen": 31164256,
      "step": 34860
    },
    {
      "epoch": 9.201662927279926,
      "grad_norm": 0.09645234793424606,
      "learning_rate": 2.0065173436288636e-06,
      "loss": 0.0305,
      "num_input_tokens_seen": 31168704,
      "step": 34865
    },
    {
      "epoch": 9.202982710835423,
      "grad_norm": 0.06839535385370255,
      "learning_rate": 2.002665464812087e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 31173344,
      "step": 34870
    },
    {
      "epoch": 9.20430249439092,
      "grad_norm": 0.13438846170902252,
      "learning_rate": 1.998817132473291e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 31177408,
      "step": 34875
    },
    {
      "epoch": 9.205622277946416,
      "grad_norm": 0.11136575043201447,
      "learning_rate": 1.9949723472059507e-06,
      "loss": 0.0247,
      "num_input_tokens_seen": 31181888,
      "step": 34880
    },
    {
      "epoch": 9.206942061501914,
      "grad_norm": 0.19053339958190918,
      "learning_rate": 1.9911311096029726e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 31186560,
      "step": 34885
    },
    {
      "epoch": 9.208261845057411,
      "grad_norm": 0.36149442195892334,
      "learning_rate": 1.9872934202567224e-06,
      "loss": 0.0257,
      "num_input_tokens_seen": 31190816,
      "step": 34890
    },
    {
      "epoch": 9.209581628612908,
      "grad_norm": 0.24522201716899872,
      "learning_rate": 1.9834592797590257e-06,
      "loss": 0.0622,
      "num_input_tokens_seen": 31195040,
      "step": 34895
    },
    {
      "epoch": 9.210901412168404,
      "grad_norm": 0.11872867494821548,
      "learning_rate": 1.979628688701149e-06,
      "loss": 0.0285,
      "num_input_tokens_seen": 31199328,
      "step": 34900
    },
    {
      "epoch": 9.2122211957239,
      "grad_norm": 0.07938193529844284,
      "learning_rate": 1.9758016476738193e-06,
      "loss": 0.0569,
      "num_input_tokens_seen": 31204160,
      "step": 34905
    },
    {
      "epoch": 9.213540979279399,
      "grad_norm": 0.05569290369749069,
      "learning_rate": 1.971978157267221e-06,
      "loss": 0.0553,
      "num_input_tokens_seen": 31208416,
      "step": 34910
    },
    {
      "epoch": 9.214860762834896,
      "grad_norm": 0.25793763995170593,
      "learning_rate": 1.968158218070973e-06,
      "loss": 0.0869,
      "num_input_tokens_seen": 31212928,
      "step": 34915
    },
    {
      "epoch": 9.216180546390392,
      "grad_norm": 0.10364700108766556,
      "learning_rate": 1.9643418306741682e-06,
      "loss": 0.0424,
      "num_input_tokens_seen": 31217568,
      "step": 34920
    },
    {
      "epoch": 9.217500329945889,
      "grad_norm": 0.06759428977966309,
      "learning_rate": 1.9605289956653337e-06,
      "loss": 0.0613,
      "num_input_tokens_seen": 31221920,
      "step": 34925
    },
    {
      "epoch": 9.218820113501385,
      "grad_norm": 0.06491217017173767,
      "learning_rate": 1.9567197136324626e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 31226272,
      "step": 34930
    },
    {
      "epoch": 9.220139897056884,
      "grad_norm": 0.027667446061968803,
      "learning_rate": 1.9529139851629935e-06,
      "loss": 0.0696,
      "num_input_tokens_seen": 31230752,
      "step": 34935
    },
    {
      "epoch": 9.22145968061238,
      "grad_norm": 0.5168484449386597,
      "learning_rate": 1.949111810843812e-06,
      "loss": 0.0821,
      "num_input_tokens_seen": 31235040,
      "step": 34940
    },
    {
      "epoch": 9.222779464167877,
      "grad_norm": 0.19354355335235596,
      "learning_rate": 1.9453131912612694e-06,
      "loss": 0.1061,
      "num_input_tokens_seen": 31239488,
      "step": 34945
    },
    {
      "epoch": 9.224099247723373,
      "grad_norm": 0.14623889327049255,
      "learning_rate": 1.941518127001149e-06,
      "loss": 0.0806,
      "num_input_tokens_seen": 31243904,
      "step": 34950
    },
    {
      "epoch": 9.22541903127887,
      "grad_norm": 0.23593467473983765,
      "learning_rate": 1.9377266186487107e-06,
      "loss": 0.0476,
      "num_input_tokens_seen": 31248352,
      "step": 34955
    },
    {
      "epoch": 9.226738814834366,
      "grad_norm": 0.15503300726413727,
      "learning_rate": 1.9339386667886483e-06,
      "loss": 0.0247,
      "num_input_tokens_seen": 31252928,
      "step": 34960
    },
    {
      "epoch": 9.228058598389865,
      "grad_norm": 0.2180701494216919,
      "learning_rate": 1.9301542720051024e-06,
      "loss": 0.0405,
      "num_input_tokens_seen": 31257376,
      "step": 34965
    },
    {
      "epoch": 9.229378381945361,
      "grad_norm": 0.1182580441236496,
      "learning_rate": 1.926373434881684e-06,
      "loss": 0.0982,
      "num_input_tokens_seen": 31261984,
      "step": 34970
    },
    {
      "epoch": 9.230698165500858,
      "grad_norm": 0.2371140569448471,
      "learning_rate": 1.9225961560014468e-06,
      "loss": 0.026,
      "num_input_tokens_seen": 31266400,
      "step": 34975
    },
    {
      "epoch": 9.232017949056354,
      "grad_norm": 0.08317003399133682,
      "learning_rate": 1.918822435946885e-06,
      "loss": 0.0313,
      "num_input_tokens_seen": 31270880,
      "step": 34980
    },
    {
      "epoch": 9.23333773261185,
      "grad_norm": 0.08405255526304245,
      "learning_rate": 1.915052275299961e-06,
      "loss": 0.0966,
      "num_input_tokens_seen": 31275392,
      "step": 34985
    },
    {
      "epoch": 9.23465751616735,
      "grad_norm": 0.04278353601694107,
      "learning_rate": 1.9112856746420854e-06,
      "loss": 0.0719,
      "num_input_tokens_seen": 31279488,
      "step": 34990
    },
    {
      "epoch": 9.235977299722846,
      "grad_norm": 0.14153872430324554,
      "learning_rate": 1.907522634554104e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 31283584,
      "step": 34995
    },
    {
      "epoch": 9.237297083278342,
      "grad_norm": 0.5269731283187866,
      "learning_rate": 1.9037631556163337e-06,
      "loss": 0.1252,
      "num_input_tokens_seen": 31288160,
      "step": 35000
    },
    {
      "epoch": 9.237297083278342,
      "eval_loss": 0.06683162599802017,
      "eval_runtime": 64.7802,
      "eval_samples_per_second": 103.967,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 31288160,
      "step": 35000
    },
    {
      "epoch": 9.238616866833839,
      "grad_norm": 0.2745518982410431,
      "learning_rate": 1.9000072384085272e-06,
      "loss": 0.071,
      "num_input_tokens_seen": 31292672,
      "step": 35005
    },
    {
      "epoch": 9.239936650389335,
      "grad_norm": 0.16345472633838654,
      "learning_rate": 1.8962548835098987e-06,
      "loss": 0.0736,
      "num_input_tokens_seen": 31296992,
      "step": 35010
    },
    {
      "epoch": 9.241256433944834,
      "grad_norm": 0.19953536987304688,
      "learning_rate": 1.8925060914991077e-06,
      "loss": 0.0796,
      "num_input_tokens_seen": 31301472,
      "step": 35015
    },
    {
      "epoch": 9.24257621750033,
      "grad_norm": 0.3499477505683899,
      "learning_rate": 1.888760862954264e-06,
      "loss": 0.0761,
      "num_input_tokens_seen": 31306240,
      "step": 35020
    },
    {
      "epoch": 9.243896001055827,
      "grad_norm": 0.1604841947555542,
      "learning_rate": 1.8850191984529309e-06,
      "loss": 0.0579,
      "num_input_tokens_seen": 31310336,
      "step": 35025
    },
    {
      "epoch": 9.245215784611323,
      "grad_norm": 0.2717927098274231,
      "learning_rate": 1.8812810985721186e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 31314752,
      "step": 35030
    },
    {
      "epoch": 9.24653556816682,
      "grad_norm": 0.20730164647102356,
      "learning_rate": 1.8775465638882856e-06,
      "loss": 0.0707,
      "num_input_tokens_seen": 31319520,
      "step": 35035
    },
    {
      "epoch": 9.247855351722318,
      "grad_norm": 0.04383067414164543,
      "learning_rate": 1.8738155949773517e-06,
      "loss": 0.0927,
      "num_input_tokens_seen": 31323968,
      "step": 35040
    },
    {
      "epoch": 9.249175135277815,
      "grad_norm": 0.21296025812625885,
      "learning_rate": 1.8700881924146707e-06,
      "loss": 0.0663,
      "num_input_tokens_seen": 31328768,
      "step": 35045
    },
    {
      "epoch": 9.250494918833311,
      "grad_norm": 0.07832084596157074,
      "learning_rate": 1.8663643567750577e-06,
      "loss": 0.0651,
      "num_input_tokens_seen": 31332992,
      "step": 35050
    },
    {
      "epoch": 9.251814702388808,
      "grad_norm": 0.3525379002094269,
      "learning_rate": 1.8626440886327813e-06,
      "loss": 0.089,
      "num_input_tokens_seen": 31337696,
      "step": 35055
    },
    {
      "epoch": 9.253134485944305,
      "grad_norm": 0.11978326737880707,
      "learning_rate": 1.8589273885615432e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 31342304,
      "step": 35060
    },
    {
      "epoch": 9.254454269499803,
      "grad_norm": 0.6307327151298523,
      "learning_rate": 1.8552142571345133e-06,
      "loss": 0.1236,
      "num_input_tokens_seen": 31346752,
      "step": 35065
    },
    {
      "epoch": 9.2557740530553,
      "grad_norm": 0.26586130261421204,
      "learning_rate": 1.8515046949243025e-06,
      "loss": 0.0599,
      "num_input_tokens_seen": 31351328,
      "step": 35070
    },
    {
      "epoch": 9.257093836610796,
      "grad_norm": 0.19320237636566162,
      "learning_rate": 1.8477987025029674e-06,
      "loss": 0.0419,
      "num_input_tokens_seen": 31355776,
      "step": 35075
    },
    {
      "epoch": 9.258413620166293,
      "grad_norm": 0.1549493819475174,
      "learning_rate": 1.8440962804420232e-06,
      "loss": 0.097,
      "num_input_tokens_seen": 31360320,
      "step": 35080
    },
    {
      "epoch": 9.259733403721789,
      "grad_norm": 0.1618567258119583,
      "learning_rate": 1.8403974293124265e-06,
      "loss": 0.0763,
      "num_input_tokens_seen": 31364768,
      "step": 35085
    },
    {
      "epoch": 9.261053187277287,
      "grad_norm": 0.1198483258485794,
      "learning_rate": 1.8367021496845854e-06,
      "loss": 0.0538,
      "num_input_tokens_seen": 31368928,
      "step": 35090
    },
    {
      "epoch": 9.262372970832784,
      "grad_norm": 0.12135984748601913,
      "learning_rate": 1.8330104421283662e-06,
      "loss": 0.0811,
      "num_input_tokens_seen": 31373440,
      "step": 35095
    },
    {
      "epoch": 9.26369275438828,
      "grad_norm": 0.30855053663253784,
      "learning_rate": 1.8293223072130717e-06,
      "loss": 0.029,
      "num_input_tokens_seen": 31377728,
      "step": 35100
    },
    {
      "epoch": 9.265012537943777,
      "grad_norm": 0.19386233389377594,
      "learning_rate": 1.8256377455074525e-06,
      "loss": 0.0748,
      "num_input_tokens_seen": 31381888,
      "step": 35105
    },
    {
      "epoch": 9.266332321499274,
      "grad_norm": 0.32154449820518494,
      "learning_rate": 1.8219567575797263e-06,
      "loss": 0.0617,
      "num_input_tokens_seen": 31386208,
      "step": 35110
    },
    {
      "epoch": 9.26765210505477,
      "grad_norm": 0.27865394949913025,
      "learning_rate": 1.8182793439975365e-06,
      "loss": 0.0502,
      "num_input_tokens_seen": 31390464,
      "step": 35115
    },
    {
      "epoch": 9.268971888610269,
      "grad_norm": 0.189579576253891,
      "learning_rate": 1.8146055053279958e-06,
      "loss": 0.0775,
      "num_input_tokens_seen": 31394848,
      "step": 35120
    },
    {
      "epoch": 9.270291672165765,
      "grad_norm": 0.10871469974517822,
      "learning_rate": 1.8109352421376486e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 31398944,
      "step": 35125
    },
    {
      "epoch": 9.271611455721262,
      "grad_norm": 0.3476543426513672,
      "learning_rate": 1.8072685549924972e-06,
      "loss": 0.0775,
      "num_input_tokens_seen": 31403264,
      "step": 35130
    },
    {
      "epoch": 9.272931239276758,
      "grad_norm": 0.21213936805725098,
      "learning_rate": 1.8036054444579982e-06,
      "loss": 0.0631,
      "num_input_tokens_seen": 31407552,
      "step": 35135
    },
    {
      "epoch": 9.274251022832255,
      "grad_norm": 0.47305503487586975,
      "learning_rate": 1.7999459110990407e-06,
      "loss": 0.0842,
      "num_input_tokens_seen": 31412384,
      "step": 35140
    },
    {
      "epoch": 9.275570806387753,
      "grad_norm": 0.06577346473932266,
      "learning_rate": 1.7962899554799712e-06,
      "loss": 0.0436,
      "num_input_tokens_seen": 31416640,
      "step": 35145
    },
    {
      "epoch": 9.27689058994325,
      "grad_norm": 0.22955182194709778,
      "learning_rate": 1.7926375781645937e-06,
      "loss": 0.0703,
      "num_input_tokens_seen": 31421408,
      "step": 35150
    },
    {
      "epoch": 9.278210373498746,
      "grad_norm": 0.036228280514478683,
      "learning_rate": 1.7889887797161359e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 31425952,
      "step": 35155
    },
    {
      "epoch": 9.279530157054243,
      "grad_norm": 0.1186600998044014,
      "learning_rate": 1.7853435606973028e-06,
      "loss": 0.0463,
      "num_input_tokens_seen": 31430368,
      "step": 35160
    },
    {
      "epoch": 9.28084994060974,
      "grad_norm": 0.054083045572042465,
      "learning_rate": 1.781701921670223e-06,
      "loss": 0.047,
      "num_input_tokens_seen": 31434848,
      "step": 35165
    },
    {
      "epoch": 9.282169724165238,
      "grad_norm": 0.09619026631116867,
      "learning_rate": 1.7780638631964886e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 31438880,
      "step": 35170
    },
    {
      "epoch": 9.283489507720734,
      "grad_norm": 0.32648539543151855,
      "learning_rate": 1.7744293858371314e-06,
      "loss": 0.0578,
      "num_input_tokens_seen": 31443200,
      "step": 35175
    },
    {
      "epoch": 9.28480929127623,
      "grad_norm": 0.11920803040266037,
      "learning_rate": 1.770798490152631e-06,
      "loss": 0.0496,
      "num_input_tokens_seen": 31447776,
      "step": 35180
    },
    {
      "epoch": 9.286129074831727,
      "grad_norm": 0.23472970724105835,
      "learning_rate": 1.767171176702917e-06,
      "loss": 0.0519,
      "num_input_tokens_seen": 31452000,
      "step": 35185
    },
    {
      "epoch": 9.287448858387224,
      "grad_norm": 0.1074891984462738,
      "learning_rate": 1.7635474460473755e-06,
      "loss": 0.0241,
      "num_input_tokens_seen": 31456544,
      "step": 35190
    },
    {
      "epoch": 9.288768641942722,
      "grad_norm": 0.130157932639122,
      "learning_rate": 1.7599272987448206e-06,
      "loss": 0.044,
      "num_input_tokens_seen": 31461088,
      "step": 35195
    },
    {
      "epoch": 9.290088425498219,
      "grad_norm": 0.14141687750816345,
      "learning_rate": 1.7563107353535362e-06,
      "loss": 0.0335,
      "num_input_tokens_seen": 31465760,
      "step": 35200
    },
    {
      "epoch": 9.290088425498219,
      "eval_loss": 0.066804438829422,
      "eval_runtime": 64.7256,
      "eval_samples_per_second": 104.055,
      "eval_steps_per_second": 26.018,
      "num_input_tokens_seen": 31465760,
      "step": 35200
    },
    {
      "epoch": 9.291408209053715,
      "grad_norm": 0.3006802797317505,
      "learning_rate": 1.7526977564312263e-06,
      "loss": 0.1077,
      "num_input_tokens_seen": 31470112,
      "step": 35205
    },
    {
      "epoch": 9.292727992609212,
      "grad_norm": 0.3751697540283203,
      "learning_rate": 1.7490883625350701e-06,
      "loss": 0.0585,
      "num_input_tokens_seen": 31474592,
      "step": 35210
    },
    {
      "epoch": 9.294047776164708,
      "grad_norm": 0.052346792072057724,
      "learning_rate": 1.7454825542216807e-06,
      "loss": 0.0653,
      "num_input_tokens_seen": 31478912,
      "step": 35215
    },
    {
      "epoch": 9.295367559720205,
      "grad_norm": 0.3811892867088318,
      "learning_rate": 1.7418803320471105e-06,
      "loss": 0.0471,
      "num_input_tokens_seen": 31483232,
      "step": 35220
    },
    {
      "epoch": 9.296687343275703,
      "grad_norm": 0.3555549085140228,
      "learning_rate": 1.7382816965668737e-06,
      "loss": 0.0602,
      "num_input_tokens_seen": 31488160,
      "step": 35225
    },
    {
      "epoch": 9.2980071268312,
      "grad_norm": 0.13075213134288788,
      "learning_rate": 1.7346866483359285e-06,
      "loss": 0.0828,
      "num_input_tokens_seen": 31492448,
      "step": 35230
    },
    {
      "epoch": 9.299326910386696,
      "grad_norm": 0.08017893880605698,
      "learning_rate": 1.7310951879086657e-06,
      "loss": 0.0503,
      "num_input_tokens_seen": 31497056,
      "step": 35235
    },
    {
      "epoch": 9.300646693942193,
      "grad_norm": 0.35733428597450256,
      "learning_rate": 1.7275073158389471e-06,
      "loss": 0.058,
      "num_input_tokens_seen": 31501408,
      "step": 35240
    },
    {
      "epoch": 9.30196647749769,
      "grad_norm": 0.04155261069536209,
      "learning_rate": 1.723923032680061e-06,
      "loss": 0.0369,
      "num_input_tokens_seen": 31505920,
      "step": 35245
    },
    {
      "epoch": 9.303286261053188,
      "grad_norm": 0.20637236535549164,
      "learning_rate": 1.7203423389847428e-06,
      "loss": 0.0494,
      "num_input_tokens_seen": 31510496,
      "step": 35250
    },
    {
      "epoch": 9.304606044608684,
      "grad_norm": 0.45904818177223206,
      "learning_rate": 1.7167652353051928e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 31515040,
      "step": 35255
    },
    {
      "epoch": 9.305925828164181,
      "grad_norm": 0.40912407636642456,
      "learning_rate": 1.7131917221930333e-06,
      "loss": 0.058,
      "num_input_tokens_seen": 31519424,
      "step": 35260
    },
    {
      "epoch": 9.307245611719678,
      "grad_norm": 0.535129964351654,
      "learning_rate": 1.7096218001993513e-06,
      "loss": 0.082,
      "num_input_tokens_seen": 31523808,
      "step": 35265
    },
    {
      "epoch": 9.308565395275174,
      "grad_norm": 0.39120668172836304,
      "learning_rate": 1.706055469874676e-06,
      "loss": 0.0864,
      "num_input_tokens_seen": 31528288,
      "step": 35270
    },
    {
      "epoch": 9.309885178830672,
      "grad_norm": 0.20317073166370392,
      "learning_rate": 1.702492731768976e-06,
      "loss": 0.0587,
      "num_input_tokens_seen": 31532768,
      "step": 35275
    },
    {
      "epoch": 9.311204962386169,
      "grad_norm": 0.4517119228839874,
      "learning_rate": 1.6989335864316724e-06,
      "loss": 0.0817,
      "num_input_tokens_seen": 31537120,
      "step": 35280
    },
    {
      "epoch": 9.312524745941666,
      "grad_norm": 0.42405879497528076,
      "learning_rate": 1.6953780344116265e-06,
      "loss": 0.0734,
      "num_input_tokens_seen": 31541888,
      "step": 35285
    },
    {
      "epoch": 9.313844529497162,
      "grad_norm": 0.3776664435863495,
      "learning_rate": 1.6918260762571497e-06,
      "loss": 0.0968,
      "num_input_tokens_seen": 31546624,
      "step": 35290
    },
    {
      "epoch": 9.315164313052659,
      "grad_norm": 0.27456656098365784,
      "learning_rate": 1.6882777125160093e-06,
      "loss": 0.0439,
      "num_input_tokens_seen": 31550688,
      "step": 35295
    },
    {
      "epoch": 9.316484096608157,
      "grad_norm": 0.08221334964036942,
      "learning_rate": 1.6847329437353899e-06,
      "loss": 0.0437,
      "num_input_tokens_seen": 31554848,
      "step": 35300
    },
    {
      "epoch": 9.317803880163654,
      "grad_norm": 0.13294507563114166,
      "learning_rate": 1.6811917704619511e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 31559488,
      "step": 35305
    },
    {
      "epoch": 9.31912366371915,
      "grad_norm": 0.5683545470237732,
      "learning_rate": 1.67765419324179e-06,
      "loss": 0.0929,
      "num_input_tokens_seen": 31564000,
      "step": 35310
    },
    {
      "epoch": 9.320443447274647,
      "grad_norm": 0.19954471290111542,
      "learning_rate": 1.6741202126204364e-06,
      "loss": 0.0251,
      "num_input_tokens_seen": 31568544,
      "step": 35315
    },
    {
      "epoch": 9.321763230830143,
      "grad_norm": 0.04624074697494507,
      "learning_rate": 1.6705898291428767e-06,
      "loss": 0.017,
      "num_input_tokens_seen": 31572992,
      "step": 35320
    },
    {
      "epoch": 9.323083014385642,
      "grad_norm": 0.10103760659694672,
      "learning_rate": 1.6670630433535395e-06,
      "loss": 0.0432,
      "num_input_tokens_seen": 31577248,
      "step": 35325
    },
    {
      "epoch": 9.324402797941138,
      "grad_norm": 0.2893691062927246,
      "learning_rate": 1.6635398557962979e-06,
      "loss": 0.1044,
      "num_input_tokens_seen": 31581664,
      "step": 35330
    },
    {
      "epoch": 9.325722581496635,
      "grad_norm": 0.1032041609287262,
      "learning_rate": 1.660020267014481e-06,
      "loss": 0.0688,
      "num_input_tokens_seen": 31586048,
      "step": 35335
    },
    {
      "epoch": 9.327042365052131,
      "grad_norm": 0.31559255719184875,
      "learning_rate": 1.6565042775508438e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 31590336,
      "step": 35340
    },
    {
      "epoch": 9.328362148607628,
      "grad_norm": 0.40643203258514404,
      "learning_rate": 1.6529918879475997e-06,
      "loss": 0.0914,
      "num_input_tokens_seen": 31594688,
      "step": 35345
    },
    {
      "epoch": 9.329681932163126,
      "grad_norm": 0.2483128160238266,
      "learning_rate": 1.6494830987464043e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 31599104,
      "step": 35350
    },
    {
      "epoch": 9.331001715718623,
      "grad_norm": 0.4615936279296875,
      "learning_rate": 1.6459779104883555e-06,
      "loss": 0.0796,
      "num_input_tokens_seen": 31603360,
      "step": 35355
    },
    {
      "epoch": 9.33232149927412,
      "grad_norm": 0.12768855690956116,
      "learning_rate": 1.6424763237140013e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 31607648,
      "step": 35360
    },
    {
      "epoch": 9.333641282829616,
      "grad_norm": 0.10510478168725967,
      "learning_rate": 1.6389783389633207e-06,
      "loss": 0.0681,
      "num_input_tokens_seen": 31611904,
      "step": 35365
    },
    {
      "epoch": 9.334961066385112,
      "grad_norm": 0.04251127690076828,
      "learning_rate": 1.6354839567757546e-06,
      "loss": 0.0202,
      "num_input_tokens_seen": 31616544,
      "step": 35370
    },
    {
      "epoch": 9.336280849940609,
      "grad_norm": 0.08850254118442535,
      "learning_rate": 1.6319931776901831e-06,
      "loss": 0.0648,
      "num_input_tokens_seen": 31621216,
      "step": 35375
    },
    {
      "epoch": 9.337600633496107,
      "grad_norm": 0.16239258646965027,
      "learning_rate": 1.6285060022449229e-06,
      "loss": 0.0557,
      "num_input_tokens_seen": 31625728,
      "step": 35380
    },
    {
      "epoch": 9.338920417051604,
      "grad_norm": 0.04191487282514572,
      "learning_rate": 1.6250224309777434e-06,
      "loss": 0.032,
      "num_input_tokens_seen": 31630304,
      "step": 35385
    },
    {
      "epoch": 9.3402402006071,
      "grad_norm": 0.40563708543777466,
      "learning_rate": 1.6215424644258515e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 31634688,
      "step": 35390
    },
    {
      "epoch": 9.341559984162597,
      "grad_norm": 0.28846850991249084,
      "learning_rate": 1.6180661031259036e-06,
      "loss": 0.0687,
      "num_input_tokens_seen": 31638816,
      "step": 35395
    },
    {
      "epoch": 9.342879767718093,
      "grad_norm": 0.19480672478675842,
      "learning_rate": 1.614593347613999e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 31643168,
      "step": 35400
    },
    {
      "epoch": 9.342879767718093,
      "eval_loss": 0.06670345366001129,
      "eval_runtime": 64.7249,
      "eval_samples_per_second": 104.056,
      "eval_steps_per_second": 26.018,
      "num_input_tokens_seen": 31643168,
      "step": 35400
    },
    {
      "epoch": 9.344199551273592,
      "grad_norm": 0.08062545955181122,
      "learning_rate": 1.6111241984256758e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 31647680,
      "step": 35405
    },
    {
      "epoch": 9.345519334829088,
      "grad_norm": 0.15169300138950348,
      "learning_rate": 1.6076586560959257e-06,
      "loss": 0.0848,
      "num_input_tokens_seen": 31652032,
      "step": 35410
    },
    {
      "epoch": 9.346839118384585,
      "grad_norm": 0.11037168651819229,
      "learning_rate": 1.604196721159182e-06,
      "loss": 0.0422,
      "num_input_tokens_seen": 31656640,
      "step": 35415
    },
    {
      "epoch": 9.348158901940081,
      "grad_norm": 0.438476026058197,
      "learning_rate": 1.6007383941493092e-06,
      "loss": 0.0759,
      "num_input_tokens_seen": 31661088,
      "step": 35420
    },
    {
      "epoch": 9.349478685495578,
      "grad_norm": 0.18475303053855896,
      "learning_rate": 1.5972836755996285e-06,
      "loss": 0.0365,
      "num_input_tokens_seen": 31665536,
      "step": 35425
    },
    {
      "epoch": 9.350798469051076,
      "grad_norm": 0.21896690130233765,
      "learning_rate": 1.5938325660429076e-06,
      "loss": 0.067,
      "num_input_tokens_seen": 31670016,
      "step": 35430
    },
    {
      "epoch": 9.352118252606573,
      "grad_norm": 0.2817040979862213,
      "learning_rate": 1.5903850660113378e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 31674432,
      "step": 35435
    },
    {
      "epoch": 9.35343803616207,
      "grad_norm": 0.1353224515914917,
      "learning_rate": 1.5869411760365826e-06,
      "loss": 0.0566,
      "num_input_tokens_seen": 31678944,
      "step": 35440
    },
    {
      "epoch": 9.354757819717566,
      "grad_norm": 0.3669642210006714,
      "learning_rate": 1.58350089664972e-06,
      "loss": 0.0716,
      "num_input_tokens_seen": 31683424,
      "step": 35445
    },
    {
      "epoch": 9.356077603273063,
      "grad_norm": 0.23217715322971344,
      "learning_rate": 1.5800642283812865e-06,
      "loss": 0.073,
      "num_input_tokens_seen": 31687552,
      "step": 35450
    },
    {
      "epoch": 9.357397386828561,
      "grad_norm": 0.12688122689723969,
      "learning_rate": 1.5766311717612698e-06,
      "loss": 0.0452,
      "num_input_tokens_seen": 31691712,
      "step": 35455
    },
    {
      "epoch": 9.358717170384057,
      "grad_norm": 0.1995408833026886,
      "learning_rate": 1.5732017273190818e-06,
      "loss": 0.0276,
      "num_input_tokens_seen": 31696032,
      "step": 35460
    },
    {
      "epoch": 9.360036953939554,
      "grad_norm": 0.27183258533477783,
      "learning_rate": 1.5697758955835806e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 31700736,
      "step": 35465
    },
    {
      "epoch": 9.36135673749505,
      "grad_norm": 0.202050119638443,
      "learning_rate": 1.566353677083085e-06,
      "loss": 0.0288,
      "num_input_tokens_seen": 31705536,
      "step": 35470
    },
    {
      "epoch": 9.362676521050547,
      "grad_norm": 0.5705959796905518,
      "learning_rate": 1.562935072345334e-06,
      "loss": 0.1186,
      "num_input_tokens_seen": 31710016,
      "step": 35475
    },
    {
      "epoch": 9.363996304606044,
      "grad_norm": 0.40491166710853577,
      "learning_rate": 1.5595200818975281e-06,
      "loss": 0.0879,
      "num_input_tokens_seen": 31714592,
      "step": 35480
    },
    {
      "epoch": 9.365316088161542,
      "grad_norm": 0.1162884309887886,
      "learning_rate": 1.5561087062662905e-06,
      "loss": 0.0647,
      "num_input_tokens_seen": 31719040,
      "step": 35485
    },
    {
      "epoch": 9.366635871717039,
      "grad_norm": 0.04190465062856674,
      "learning_rate": 1.5527009459777087e-06,
      "loss": 0.0537,
      "num_input_tokens_seen": 31723616,
      "step": 35490
    },
    {
      "epoch": 9.367955655272535,
      "grad_norm": 0.43932339549064636,
      "learning_rate": 1.5492968015572984e-06,
      "loss": 0.1108,
      "num_input_tokens_seen": 31727840,
      "step": 35495
    },
    {
      "epoch": 9.369275438828032,
      "grad_norm": 0.23918014764785767,
      "learning_rate": 1.5458962735300203e-06,
      "loss": 0.0906,
      "num_input_tokens_seen": 31732096,
      "step": 35500
    },
    {
      "epoch": 9.370595222383528,
      "grad_norm": 0.3836345374584198,
      "learning_rate": 1.54249936242028e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 31736512,
      "step": 35505
    },
    {
      "epoch": 9.371915005939027,
      "grad_norm": 0.16582264006137848,
      "learning_rate": 1.5391060687519222e-06,
      "loss": 0.0753,
      "num_input_tokens_seen": 31741152,
      "step": 35510
    },
    {
      "epoch": 9.373234789494523,
      "grad_norm": 0.21266137063503265,
      "learning_rate": 1.5357163930482367e-06,
      "loss": 0.0501,
      "num_input_tokens_seen": 31745600,
      "step": 35515
    },
    {
      "epoch": 9.37455457305002,
      "grad_norm": 0.02760969288647175,
      "learning_rate": 1.532330335831955e-06,
      "loss": 0.0315,
      "num_input_tokens_seen": 31750016,
      "step": 35520
    },
    {
      "epoch": 9.375874356605516,
      "grad_norm": 0.6111297607421875,
      "learning_rate": 1.5289478976252491e-06,
      "loss": 0.0773,
      "num_input_tokens_seen": 31754976,
      "step": 35525
    },
    {
      "epoch": 9.377194140161013,
      "grad_norm": 0.10423564165830612,
      "learning_rate": 1.5255690789497345e-06,
      "loss": 0.0312,
      "num_input_tokens_seen": 31759232,
      "step": 35530
    },
    {
      "epoch": 9.378513923716511,
      "grad_norm": 0.10101795196533203,
      "learning_rate": 1.5221938803264641e-06,
      "loss": 0.0454,
      "num_input_tokens_seen": 31763904,
      "step": 35535
    },
    {
      "epoch": 9.379833707272008,
      "grad_norm": 0.0902305319905281,
      "learning_rate": 1.518822302275938e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 31768320,
      "step": 35540
    },
    {
      "epoch": 9.381153490827504,
      "grad_norm": 0.47667238116264343,
      "learning_rate": 1.5154543453180958e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 31773056,
      "step": 35545
    },
    {
      "epoch": 9.382473274383,
      "grad_norm": 0.13630710542201996,
      "learning_rate": 1.5120900099723167e-06,
      "loss": 0.0187,
      "num_input_tokens_seen": 31777600,
      "step": 35550
    },
    {
      "epoch": 9.383793057938497,
      "grad_norm": 0.3094789385795593,
      "learning_rate": 1.5087292967574273e-06,
      "loss": 0.0429,
      "num_input_tokens_seen": 31782144,
      "step": 35555
    },
    {
      "epoch": 9.385112841493996,
      "grad_norm": 0.03494316712021828,
      "learning_rate": 1.5053722061916908e-06,
      "loss": 0.0375,
      "num_input_tokens_seen": 31786912,
      "step": 35560
    },
    {
      "epoch": 9.386432625049492,
      "grad_norm": 0.05522536113858223,
      "learning_rate": 1.5020187387928124e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 31791296,
      "step": 35565
    },
    {
      "epoch": 9.387752408604989,
      "grad_norm": 0.07607729732990265,
      "learning_rate": 1.4986688950779343e-06,
      "loss": 0.0392,
      "num_input_tokens_seen": 31795808,
      "step": 35570
    },
    {
      "epoch": 9.389072192160485,
      "grad_norm": 0.09176502376794815,
      "learning_rate": 1.495322675563654e-06,
      "loss": 0.0596,
      "num_input_tokens_seen": 31800288,
      "step": 35575
    },
    {
      "epoch": 9.390391975715982,
      "grad_norm": 0.22456508874893188,
      "learning_rate": 1.4919800807659922e-06,
      "loss": 0.0412,
      "num_input_tokens_seen": 31804544,
      "step": 35580
    },
    {
      "epoch": 9.39171175927148,
      "grad_norm": 0.20803193747997284,
      "learning_rate": 1.4886411112004255e-06,
      "loss": 0.0258,
      "num_input_tokens_seen": 31808864,
      "step": 35585
    },
    {
      "epoch": 9.393031542826977,
      "grad_norm": 0.0972791388630867,
      "learning_rate": 1.4853057673818588e-06,
      "loss": 0.0172,
      "num_input_tokens_seen": 31813408,
      "step": 35590
    },
    {
      "epoch": 9.394351326382473,
      "grad_norm": 0.34281662106513977,
      "learning_rate": 1.481974049824647e-06,
      "loss": 0.0867,
      "num_input_tokens_seen": 31817632,
      "step": 35595
    },
    {
      "epoch": 9.39567110993797,
      "grad_norm": 0.3948444426059723,
      "learning_rate": 1.4786459590425849e-06,
      "loss": 0.0852,
      "num_input_tokens_seen": 31821856,
      "step": 35600
    },
    {
      "epoch": 9.39567110993797,
      "eval_loss": 0.0668850988149643,
      "eval_runtime": 64.717,
      "eval_samples_per_second": 104.068,
      "eval_steps_per_second": 26.021,
      "num_input_tokens_seen": 31821856,
      "step": 35600
    },
    {
      "epoch": 9.396990893493467,
      "grad_norm": 0.18731388449668884,
      "learning_rate": 1.4753214955489036e-06,
      "loss": 0.0533,
      "num_input_tokens_seen": 31826400,
      "step": 35605
    },
    {
      "epoch": 9.398310677048965,
      "grad_norm": 0.1431075781583786,
      "learning_rate": 1.4720006598562737e-06,
      "loss": 0.0953,
      "num_input_tokens_seen": 31830592,
      "step": 35610
    },
    {
      "epoch": 9.399630460604461,
      "grad_norm": 0.0659983903169632,
      "learning_rate": 1.4686834524768185e-06,
      "loss": 0.021,
      "num_input_tokens_seen": 31834880,
      "step": 35615
    },
    {
      "epoch": 9.400950244159958,
      "grad_norm": 0.3530988097190857,
      "learning_rate": 1.4653698739220844e-06,
      "loss": 0.0296,
      "num_input_tokens_seen": 31839296,
      "step": 35620
    },
    {
      "epoch": 9.402270027715455,
      "grad_norm": 0.21844987571239471,
      "learning_rate": 1.4620599247030715e-06,
      "loss": 0.0614,
      "num_input_tokens_seen": 31843840,
      "step": 35625
    },
    {
      "epoch": 9.403589811270951,
      "grad_norm": 0.2784656882286072,
      "learning_rate": 1.4587536053302125e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 31848128,
      "step": 35630
    },
    {
      "epoch": 9.404909594826448,
      "grad_norm": 0.5678418874740601,
      "learning_rate": 1.4554509163133862e-06,
      "loss": 0.0842,
      "num_input_tokens_seen": 31852544,
      "step": 35635
    },
    {
      "epoch": 9.406229378381946,
      "grad_norm": 0.25789692997932434,
      "learning_rate": 1.4521518581619098e-06,
      "loss": 0.0603,
      "num_input_tokens_seen": 31856768,
      "step": 35640
    },
    {
      "epoch": 9.407549161937443,
      "grad_norm": 0.07239355146884918,
      "learning_rate": 1.4488564313845348e-06,
      "loss": 0.0204,
      "num_input_tokens_seen": 31861376,
      "step": 35645
    },
    {
      "epoch": 9.408868945492939,
      "grad_norm": 0.25704944133758545,
      "learning_rate": 1.4455646364894603e-06,
      "loss": 0.0676,
      "num_input_tokens_seen": 31865728,
      "step": 35650
    },
    {
      "epoch": 9.410188729048436,
      "grad_norm": 0.14750495553016663,
      "learning_rate": 1.4422764739843247e-06,
      "loss": 0.1043,
      "num_input_tokens_seen": 31870048,
      "step": 35655
    },
    {
      "epoch": 9.411508512603932,
      "grad_norm": 0.3360860347747803,
      "learning_rate": 1.4389919443762e-06,
      "loss": 0.1965,
      "num_input_tokens_seen": 31874432,
      "step": 35660
    },
    {
      "epoch": 9.41282829615943,
      "grad_norm": 0.04006439819931984,
      "learning_rate": 1.4357110481716063e-06,
      "loss": 0.0496,
      "num_input_tokens_seen": 31878752,
      "step": 35665
    },
    {
      "epoch": 9.414148079714927,
      "grad_norm": 0.394654244184494,
      "learning_rate": 1.4324337858764941e-06,
      "loss": 0.0691,
      "num_input_tokens_seen": 31883328,
      "step": 35670
    },
    {
      "epoch": 9.415467863270424,
      "grad_norm": 0.16913703083992004,
      "learning_rate": 1.4291601579962622e-06,
      "loss": 0.0949,
      "num_input_tokens_seen": 31887776,
      "step": 35675
    },
    {
      "epoch": 9.41678764682592,
      "grad_norm": 0.322200745344162,
      "learning_rate": 1.42589016503574e-06,
      "loss": 0.0432,
      "num_input_tokens_seen": 31891904,
      "step": 35680
    },
    {
      "epoch": 9.418107430381417,
      "grad_norm": 0.17461909353733063,
      "learning_rate": 1.4226238074992099e-06,
      "loss": 0.0622,
      "num_input_tokens_seen": 31896128,
      "step": 35685
    },
    {
      "epoch": 9.419427213936915,
      "grad_norm": 0.16690261662006378,
      "learning_rate": 1.4193610858903778e-06,
      "loss": 0.061,
      "num_input_tokens_seen": 31900672,
      "step": 35690
    },
    {
      "epoch": 9.420746997492412,
      "grad_norm": 0.11034580320119858,
      "learning_rate": 1.416102000712402e-06,
      "loss": 0.0597,
      "num_input_tokens_seen": 31905184,
      "step": 35695
    },
    {
      "epoch": 9.422066781047908,
      "grad_norm": 0.21503998339176178,
      "learning_rate": 1.4128465524678668e-06,
      "loss": 0.0322,
      "num_input_tokens_seen": 31909504,
      "step": 35700
    },
    {
      "epoch": 9.423386564603405,
      "grad_norm": 0.1960384100675583,
      "learning_rate": 1.4095947416588124e-06,
      "loss": 0.078,
      "num_input_tokens_seen": 31914112,
      "step": 35705
    },
    {
      "epoch": 9.424706348158901,
      "grad_norm": 0.13333863019943237,
      "learning_rate": 1.4063465687866983e-06,
      "loss": 0.1114,
      "num_input_tokens_seen": 31918560,
      "step": 35710
    },
    {
      "epoch": 9.4260261317144,
      "grad_norm": 0.5157410502433777,
      "learning_rate": 1.4031020343524438e-06,
      "loss": 0.05,
      "num_input_tokens_seen": 31923520,
      "step": 35715
    },
    {
      "epoch": 9.427345915269896,
      "grad_norm": 0.2986643314361572,
      "learning_rate": 1.3998611388563926e-06,
      "loss": 0.0849,
      "num_input_tokens_seen": 31927840,
      "step": 35720
    },
    {
      "epoch": 9.428665698825393,
      "grad_norm": 0.11249393224716187,
      "learning_rate": 1.3966238827983314e-06,
      "loss": 0.0687,
      "num_input_tokens_seen": 31932416,
      "step": 35725
    },
    {
      "epoch": 9.42998548238089,
      "grad_norm": 0.2712875306606293,
      "learning_rate": 1.393390266677483e-06,
      "loss": 0.0749,
      "num_input_tokens_seen": 31936704,
      "step": 35730
    },
    {
      "epoch": 9.431305265936386,
      "grad_norm": 0.055935170501470566,
      "learning_rate": 1.3901602909925204e-06,
      "loss": 0.0861,
      "num_input_tokens_seen": 31941120,
      "step": 35735
    },
    {
      "epoch": 9.432625049491884,
      "grad_norm": 0.14575591683387756,
      "learning_rate": 1.3869339562415373e-06,
      "loss": 0.0886,
      "num_input_tokens_seen": 31945760,
      "step": 35740
    },
    {
      "epoch": 9.43394483304738,
      "grad_norm": 0.046151384711265564,
      "learning_rate": 1.38371126292208e-06,
      "loss": 0.0401,
      "num_input_tokens_seen": 31950336,
      "step": 35745
    },
    {
      "epoch": 9.435264616602877,
      "grad_norm": 0.20803791284561157,
      "learning_rate": 1.3804922115311286e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 31954560,
      "step": 35750
    },
    {
      "epoch": 9.436584400158374,
      "grad_norm": 0.14611060917377472,
      "learning_rate": 1.3772768025650945e-06,
      "loss": 0.0424,
      "num_input_tokens_seen": 31958944,
      "step": 35755
    },
    {
      "epoch": 9.43790418371387,
      "grad_norm": 0.3255484104156494,
      "learning_rate": 1.3740650365198448e-06,
      "loss": 0.0568,
      "num_input_tokens_seen": 31963456,
      "step": 35760
    },
    {
      "epoch": 9.439223967269367,
      "grad_norm": 0.03696254640817642,
      "learning_rate": 1.3708569138906612e-06,
      "loss": 0.0741,
      "num_input_tokens_seen": 31967776,
      "step": 35765
    },
    {
      "epoch": 9.440543750824865,
      "grad_norm": 0.09446676075458527,
      "learning_rate": 1.367652435172287e-06,
      "loss": 0.0453,
      "num_input_tokens_seen": 31972256,
      "step": 35770
    },
    {
      "epoch": 9.441863534380362,
      "grad_norm": 0.3016657829284668,
      "learning_rate": 1.364451600858893e-06,
      "loss": 0.0484,
      "num_input_tokens_seen": 31976544,
      "step": 35775
    },
    {
      "epoch": 9.443183317935858,
      "grad_norm": 0.3186556100845337,
      "learning_rate": 1.3612544114440823e-06,
      "loss": 0.053,
      "num_input_tokens_seen": 31981408,
      "step": 35780
    },
    {
      "epoch": 9.444503101491355,
      "grad_norm": 0.27121850848197937,
      "learning_rate": 1.3580608674209072e-06,
      "loss": 0.0511,
      "num_input_tokens_seen": 31985600,
      "step": 35785
    },
    {
      "epoch": 9.445822885046852,
      "grad_norm": 0.05777192860841751,
      "learning_rate": 1.3548709692818434e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 31990016,
      "step": 35790
    },
    {
      "epoch": 9.44714266860235,
      "grad_norm": 0.17754770815372467,
      "learning_rate": 1.3516847175188223e-06,
      "loss": 0.0303,
      "num_input_tokens_seen": 31994240,
      "step": 35795
    },
    {
      "epoch": 9.448462452157846,
      "grad_norm": 0.1456814855337143,
      "learning_rate": 1.348502112623204e-06,
      "loss": 0.0511,
      "num_input_tokens_seen": 31998368,
      "step": 35800
    },
    {
      "epoch": 9.448462452157846,
      "eval_loss": 0.06698223948478699,
      "eval_runtime": 64.7626,
      "eval_samples_per_second": 103.995,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 31998368,
      "step": 35800
    },
    {
      "epoch": 9.449782235713343,
      "grad_norm": 0.3033802807331085,
      "learning_rate": 1.3453231550857787e-06,
      "loss": 0.0647,
      "num_input_tokens_seen": 32002880,
      "step": 35805
    },
    {
      "epoch": 9.45110201926884,
      "grad_norm": 0.11059413850307465,
      "learning_rate": 1.3421478453967878e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 32007552,
      "step": 35810
    },
    {
      "epoch": 9.452421802824336,
      "grad_norm": 0.12480597198009491,
      "learning_rate": 1.3389761840459065e-06,
      "loss": 0.0317,
      "num_input_tokens_seen": 32012000,
      "step": 35815
    },
    {
      "epoch": 9.453741586379834,
      "grad_norm": 0.33015137910842896,
      "learning_rate": 1.3358081715222376e-06,
      "loss": 0.0873,
      "num_input_tokens_seen": 32016544,
      "step": 35820
    },
    {
      "epoch": 9.455061369935331,
      "grad_norm": 0.34156304597854614,
      "learning_rate": 1.3326438083143295e-06,
      "loss": 0.057,
      "num_input_tokens_seen": 32021376,
      "step": 35825
    },
    {
      "epoch": 9.456381153490828,
      "grad_norm": 0.16523858904838562,
      "learning_rate": 1.3294830949101723e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 32025824,
      "step": 35830
    },
    {
      "epoch": 9.457700937046324,
      "grad_norm": 0.13245216012001038,
      "learning_rate": 1.3263260317971815e-06,
      "loss": 0.0603,
      "num_input_tokens_seen": 32030368,
      "step": 35835
    },
    {
      "epoch": 9.45902072060182,
      "grad_norm": 0.09716157615184784,
      "learning_rate": 1.3231726194622208e-06,
      "loss": 0.0405,
      "num_input_tokens_seen": 32034848,
      "step": 35840
    },
    {
      "epoch": 9.460340504157319,
      "grad_norm": 0.2539297342300415,
      "learning_rate": 1.3200228583915814e-06,
      "loss": 0.0466,
      "num_input_tokens_seen": 32039296,
      "step": 35845
    },
    {
      "epoch": 9.461660287712816,
      "grad_norm": 0.07760278135538101,
      "learning_rate": 1.3168767490709971e-06,
      "loss": 0.0486,
      "num_input_tokens_seen": 32043648,
      "step": 35850
    },
    {
      "epoch": 9.462980071268312,
      "grad_norm": 0.21851150691509247,
      "learning_rate": 1.3137342919856437e-06,
      "loss": 0.0218,
      "num_input_tokens_seen": 32048352,
      "step": 35855
    },
    {
      "epoch": 9.464299854823809,
      "grad_norm": 0.2527139484882355,
      "learning_rate": 1.310595487620117e-06,
      "loss": 0.0531,
      "num_input_tokens_seen": 32052736,
      "step": 35860
    },
    {
      "epoch": 9.465619638379305,
      "grad_norm": 0.3916867673397064,
      "learning_rate": 1.3074603364584715e-06,
      "loss": 0.07,
      "num_input_tokens_seen": 32056896,
      "step": 35865
    },
    {
      "epoch": 9.466939421934804,
      "grad_norm": 0.2598065137863159,
      "learning_rate": 1.3043288389841758e-06,
      "loss": 0.0478,
      "num_input_tokens_seen": 32061440,
      "step": 35870
    },
    {
      "epoch": 9.4682592054903,
      "grad_norm": 0.1841055303812027,
      "learning_rate": 1.3012009956801546e-06,
      "loss": 0.0388,
      "num_input_tokens_seen": 32065536,
      "step": 35875
    },
    {
      "epoch": 9.469578989045797,
      "grad_norm": 0.22664779424667358,
      "learning_rate": 1.2980768070287586e-06,
      "loss": 0.065,
      "num_input_tokens_seen": 32070112,
      "step": 35880
    },
    {
      "epoch": 9.470898772601293,
      "grad_norm": 0.16580571234226227,
      "learning_rate": 1.2949562735117716e-06,
      "loss": 0.07,
      "num_input_tokens_seen": 32075040,
      "step": 35885
    },
    {
      "epoch": 9.47221855615679,
      "grad_norm": 0.060632552951574326,
      "learning_rate": 1.291839395610428e-06,
      "loss": 0.0479,
      "num_input_tokens_seen": 32079520,
      "step": 35890
    },
    {
      "epoch": 9.473538339712288,
      "grad_norm": 0.06873712688684464,
      "learning_rate": 1.2887261738053852e-06,
      "loss": 0.0416,
      "num_input_tokens_seen": 32083808,
      "step": 35895
    },
    {
      "epoch": 9.474858123267785,
      "grad_norm": 0.43802589178085327,
      "learning_rate": 1.2856166085767396e-06,
      "loss": 0.07,
      "num_input_tokens_seen": 32088000,
      "step": 35900
    },
    {
      "epoch": 9.476177906823281,
      "grad_norm": 0.12196921557188034,
      "learning_rate": 1.2825107004040272e-06,
      "loss": 0.027,
      "num_input_tokens_seen": 32092224,
      "step": 35905
    },
    {
      "epoch": 9.477497690378778,
      "grad_norm": 0.1386118084192276,
      "learning_rate": 1.2794084497662146e-06,
      "loss": 0.0565,
      "num_input_tokens_seen": 32096928,
      "step": 35910
    },
    {
      "epoch": 9.478817473934274,
      "grad_norm": 0.48510780930519104,
      "learning_rate": 1.276309857141711e-06,
      "loss": 0.0555,
      "num_input_tokens_seen": 32101440,
      "step": 35915
    },
    {
      "epoch": 9.48013725748977,
      "grad_norm": 0.060436997562646866,
      "learning_rate": 1.273214923008359e-06,
      "loss": 0.0756,
      "num_input_tokens_seen": 32105760,
      "step": 35920
    },
    {
      "epoch": 9.48145704104527,
      "grad_norm": 0.6558877229690552,
      "learning_rate": 1.2701236478434352e-06,
      "loss": 0.0598,
      "num_input_tokens_seen": 32110048,
      "step": 35925
    },
    {
      "epoch": 9.482776824600766,
      "grad_norm": 0.039301253855228424,
      "learning_rate": 1.2670360321236502e-06,
      "loss": 0.048,
      "num_input_tokens_seen": 32114368,
      "step": 35930
    },
    {
      "epoch": 9.484096608156262,
      "grad_norm": 0.22750680148601532,
      "learning_rate": 1.2639520763251617e-06,
      "loss": 0.0913,
      "num_input_tokens_seen": 32118912,
      "step": 35935
    },
    {
      "epoch": 9.485416391711759,
      "grad_norm": 0.08800280094146729,
      "learning_rate": 1.2608717809235448e-06,
      "loss": 0.0607,
      "num_input_tokens_seen": 32123296,
      "step": 35940
    },
    {
      "epoch": 9.486736175267255,
      "grad_norm": 0.31629496812820435,
      "learning_rate": 1.2577951463938282e-06,
      "loss": 0.0744,
      "num_input_tokens_seen": 32127584,
      "step": 35945
    },
    {
      "epoch": 9.488055958822754,
      "grad_norm": 0.03660682961344719,
      "learning_rate": 1.2547221732104569e-06,
      "loss": 0.0404,
      "num_input_tokens_seen": 32132000,
      "step": 35950
    },
    {
      "epoch": 9.48937574237825,
      "grad_norm": 0.11558277904987335,
      "learning_rate": 1.25165286184733e-06,
      "loss": 0.0521,
      "num_input_tokens_seen": 32136672,
      "step": 35955
    },
    {
      "epoch": 9.490695525933747,
      "grad_norm": 0.2368285059928894,
      "learning_rate": 1.248587212777777e-06,
      "loss": 0.067,
      "num_input_tokens_seen": 32141024,
      "step": 35960
    },
    {
      "epoch": 9.492015309489243,
      "grad_norm": 0.0584837831556797,
      "learning_rate": 1.2455252264745532e-06,
      "loss": 0.0237,
      "num_input_tokens_seen": 32145568,
      "step": 35965
    },
    {
      "epoch": 9.49333509304474,
      "grad_norm": 0.2167876958847046,
      "learning_rate": 1.2424669034098528e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 32150176,
      "step": 35970
    },
    {
      "epoch": 9.494654876600238,
      "grad_norm": 0.2450532168149948,
      "learning_rate": 1.2394122440553185e-06,
      "loss": 0.041,
      "num_input_tokens_seen": 32154752,
      "step": 35975
    },
    {
      "epoch": 9.495974660155735,
      "grad_norm": 0.47548699378967285,
      "learning_rate": 1.2363612488820037e-06,
      "loss": 0.0955,
      "num_input_tokens_seen": 32159456,
      "step": 35980
    },
    {
      "epoch": 9.497294443711231,
      "grad_norm": 0.21347351372241974,
      "learning_rate": 1.2333139183604208e-06,
      "loss": 0.085,
      "num_input_tokens_seen": 32164224,
      "step": 35985
    },
    {
      "epoch": 9.498614227266728,
      "grad_norm": 0.14385399222373962,
      "learning_rate": 1.2302702529604998e-06,
      "loss": 0.0311,
      "num_input_tokens_seen": 32168864,
      "step": 35990
    },
    {
      "epoch": 9.499934010822225,
      "grad_norm": 0.12284520268440247,
      "learning_rate": 1.227230253151615e-06,
      "loss": 0.0398,
      "num_input_tokens_seen": 32173504,
      "step": 35995
    },
    {
      "epoch": 9.501253794377723,
      "grad_norm": 0.08230874687433243,
      "learning_rate": 1.2241939194025748e-06,
      "loss": 0.0541,
      "num_input_tokens_seen": 32178176,
      "step": 36000
    },
    {
      "epoch": 9.501253794377723,
      "eval_loss": 0.06689131259918213,
      "eval_runtime": 64.7572,
      "eval_samples_per_second": 104.004,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 32178176,
      "step": 36000
    },
    {
      "epoch": 9.50257357793322,
      "grad_norm": 0.05430793762207031,
      "learning_rate": 1.2211612521816156e-06,
      "loss": 0.0998,
      "num_input_tokens_seen": 32182592,
      "step": 36005
    },
    {
      "epoch": 9.503893361488716,
      "grad_norm": 0.06356596946716309,
      "learning_rate": 1.2181322519564137e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 32187360,
      "step": 36010
    },
    {
      "epoch": 9.505213145044213,
      "grad_norm": 0.16719351708889008,
      "learning_rate": 1.2151069191940839e-06,
      "loss": 0.0973,
      "num_input_tokens_seen": 32191808,
      "step": 36015
    },
    {
      "epoch": 9.506532928599709,
      "grad_norm": 0.21490101516246796,
      "learning_rate": 1.2120852543611644e-06,
      "loss": 0.0855,
      "num_input_tokens_seen": 32196352,
      "step": 36020
    },
    {
      "epoch": 9.507852712155206,
      "grad_norm": 0.19186356663703918,
      "learning_rate": 1.2090672579236379e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 32200864,
      "step": 36025
    },
    {
      "epoch": 9.509172495710704,
      "grad_norm": 0.3806540369987488,
      "learning_rate": 1.2060529303469126e-06,
      "loss": 0.0498,
      "num_input_tokens_seen": 32205472,
      "step": 36030
    },
    {
      "epoch": 9.5104922792662,
      "grad_norm": 0.09993086755275726,
      "learning_rate": 1.2030422720958445e-06,
      "loss": 0.0284,
      "num_input_tokens_seen": 32209888,
      "step": 36035
    },
    {
      "epoch": 9.511812062821697,
      "grad_norm": 0.18758989870548248,
      "learning_rate": 1.200035283634704e-06,
      "loss": 0.0331,
      "num_input_tokens_seen": 32214336,
      "step": 36040
    },
    {
      "epoch": 9.513131846377194,
      "grad_norm": 0.032845307141542435,
      "learning_rate": 1.1970319654272144e-06,
      "loss": 0.0421,
      "num_input_tokens_seen": 32218816,
      "step": 36045
    },
    {
      "epoch": 9.51445162993269,
      "grad_norm": 0.1916522979736328,
      "learning_rate": 1.1940323179365192e-06,
      "loss": 0.0951,
      "num_input_tokens_seen": 32223200,
      "step": 36050
    },
    {
      "epoch": 9.515771413488189,
      "grad_norm": 0.026084665209054947,
      "learning_rate": 1.1910363416252095e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 32227488,
      "step": 36055
    },
    {
      "epoch": 9.517091197043685,
      "grad_norm": 0.22550491988658905,
      "learning_rate": 1.1880440369552964e-06,
      "loss": 0.0449,
      "num_input_tokens_seen": 32231872,
      "step": 36060
    },
    {
      "epoch": 9.518410980599182,
      "grad_norm": 0.2937047779560089,
      "learning_rate": 1.1850554043882328e-06,
      "loss": 0.0637,
      "num_input_tokens_seen": 32236160,
      "step": 36065
    },
    {
      "epoch": 9.519730764154678,
      "grad_norm": 0.23751212656497955,
      "learning_rate": 1.1820704443849028e-06,
      "loss": 0.0426,
      "num_input_tokens_seen": 32240704,
      "step": 36070
    },
    {
      "epoch": 9.521050547710175,
      "grad_norm": 0.08320045471191406,
      "learning_rate": 1.1790891574056219e-06,
      "loss": 0.0364,
      "num_input_tokens_seen": 32245088,
      "step": 36075
    },
    {
      "epoch": 9.522370331265673,
      "grad_norm": 0.33854562044143677,
      "learning_rate": 1.1761115439101523e-06,
      "loss": 0.091,
      "num_input_tokens_seen": 32249632,
      "step": 36080
    },
    {
      "epoch": 9.52369011482117,
      "grad_norm": 0.1391405314207077,
      "learning_rate": 1.1731376043576659e-06,
      "loss": 0.0339,
      "num_input_tokens_seen": 32254176,
      "step": 36085
    },
    {
      "epoch": 9.525009898376666,
      "grad_norm": 0.5929086804389954,
      "learning_rate": 1.1701673392067875e-06,
      "loss": 0.1383,
      "num_input_tokens_seen": 32258528,
      "step": 36090
    },
    {
      "epoch": 9.526329681932163,
      "grad_norm": 0.21276335418224335,
      "learning_rate": 1.1672007489155757e-06,
      "loss": 0.0709,
      "num_input_tokens_seen": 32262688,
      "step": 36095
    },
    {
      "epoch": 9.52764946548766,
      "grad_norm": 0.08716940879821777,
      "learning_rate": 1.164237833941506e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 32267520,
      "step": 36100
    },
    {
      "epoch": 9.528969249043158,
      "grad_norm": 0.05854455381631851,
      "learning_rate": 1.1612785947415022e-06,
      "loss": 0.0611,
      "num_input_tokens_seen": 32271808,
      "step": 36105
    },
    {
      "epoch": 9.530289032598654,
      "grad_norm": 0.461655855178833,
      "learning_rate": 1.1583230317719185e-06,
      "loss": 0.0717,
      "num_input_tokens_seen": 32276288,
      "step": 36110
    },
    {
      "epoch": 9.53160881615415,
      "grad_norm": 0.15409792959690094,
      "learning_rate": 1.1553711454885318e-06,
      "loss": 0.0509,
      "num_input_tokens_seen": 32280512,
      "step": 36115
    },
    {
      "epoch": 9.532928599709647,
      "grad_norm": 0.15361681580543518,
      "learning_rate": 1.152422936346567e-06,
      "loss": 0.0405,
      "num_input_tokens_seen": 32284832,
      "step": 36120
    },
    {
      "epoch": 9.534248383265144,
      "grad_norm": 0.1338394731283188,
      "learning_rate": 1.1494784048006718e-06,
      "loss": 0.0568,
      "num_input_tokens_seen": 32289440,
      "step": 36125
    },
    {
      "epoch": 9.535568166820642,
      "grad_norm": 0.1780836284160614,
      "learning_rate": 1.1465375513049326e-06,
      "loss": 0.0495,
      "num_input_tokens_seen": 32294112,
      "step": 36130
    },
    {
      "epoch": 9.536887950376139,
      "grad_norm": 0.3479895293712616,
      "learning_rate": 1.1436003763128616e-06,
      "loss": 0.0618,
      "num_input_tokens_seen": 32298528,
      "step": 36135
    },
    {
      "epoch": 9.538207733931635,
      "grad_norm": 0.24477459490299225,
      "learning_rate": 1.1406668802774106e-06,
      "loss": 0.0537,
      "num_input_tokens_seen": 32303072,
      "step": 36140
    },
    {
      "epoch": 9.539527517487132,
      "grad_norm": 0.19615603983402252,
      "learning_rate": 1.137737063650965e-06,
      "loss": 0.0965,
      "num_input_tokens_seen": 32307712,
      "step": 36145
    },
    {
      "epoch": 9.540847301042628,
      "grad_norm": 0.22162428498268127,
      "learning_rate": 1.1348109268853323e-06,
      "loss": 0.0616,
      "num_input_tokens_seen": 32312256,
      "step": 36150
    },
    {
      "epoch": 9.542167084598127,
      "grad_norm": 0.15427495539188385,
      "learning_rate": 1.1318884704317634e-06,
      "loss": 0.0343,
      "num_input_tokens_seen": 32316672,
      "step": 36155
    },
    {
      "epoch": 9.543486868153623,
      "grad_norm": 0.27769675850868225,
      "learning_rate": 1.1289696947409417e-06,
      "loss": 0.0506,
      "num_input_tokens_seen": 32321024,
      "step": 36160
    },
    {
      "epoch": 9.54480665170912,
      "grad_norm": 0.11984102427959442,
      "learning_rate": 1.126054600262974e-06,
      "loss": 0.0781,
      "num_input_tokens_seen": 32325728,
      "step": 36165
    },
    {
      "epoch": 9.546126435264616,
      "grad_norm": 0.25897517800331116,
      "learning_rate": 1.1231431874474064e-06,
      "loss": 0.0445,
      "num_input_tokens_seen": 32330112,
      "step": 36170
    },
    {
      "epoch": 9.547446218820113,
      "grad_norm": 0.12067486345767975,
      "learning_rate": 1.12023545674321e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 32334496,
      "step": 36175
    },
    {
      "epoch": 9.54876600237561,
      "grad_norm": 0.38571301102638245,
      "learning_rate": 1.117331408598804e-06,
      "loss": 0.1041,
      "num_input_tokens_seen": 32338848,
      "step": 36180
    },
    {
      "epoch": 9.550085785931108,
      "grad_norm": 0.23605024814605713,
      "learning_rate": 1.1144310434620191e-06,
      "loss": 0.0316,
      "num_input_tokens_seen": 32343296,
      "step": 36185
    },
    {
      "epoch": 9.551405569486604,
      "grad_norm": 0.20349544286727905,
      "learning_rate": 1.1115343617801365e-06,
      "loss": 0.0992,
      "num_input_tokens_seen": 32347904,
      "step": 36190
    },
    {
      "epoch": 9.552725353042101,
      "grad_norm": 0.4745585322380066,
      "learning_rate": 1.1086413639998515e-06,
      "loss": 0.0865,
      "num_input_tokens_seen": 32352320,
      "step": 36195
    },
    {
      "epoch": 9.554045136597598,
      "grad_norm": 0.1492098867893219,
      "learning_rate": 1.1057520505673103e-06,
      "loss": 0.0533,
      "num_input_tokens_seen": 32356768,
      "step": 36200
    },
    {
      "epoch": 9.554045136597598,
      "eval_loss": 0.06703390926122665,
      "eval_runtime": 64.7595,
      "eval_samples_per_second": 104.0,
      "eval_steps_per_second": 26.004,
      "num_input_tokens_seen": 32356768,
      "step": 36200
    },
    {
      "epoch": 9.555364920153094,
      "grad_norm": 0.1634102314710617,
      "learning_rate": 1.1028664219280727e-06,
      "loss": 0.046,
      "num_input_tokens_seen": 32361344,
      "step": 36205
    },
    {
      "epoch": 9.556684703708592,
      "grad_norm": 0.1981695592403412,
      "learning_rate": 1.0999844785271468e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 32366048,
      "step": 36210
    },
    {
      "epoch": 9.558004487264089,
      "grad_norm": 0.05717066675424576,
      "learning_rate": 1.097106220808955e-06,
      "loss": 0.0367,
      "num_input_tokens_seen": 32370304,
      "step": 36215
    },
    {
      "epoch": 9.559324270819586,
      "grad_norm": 0.200364887714386,
      "learning_rate": 1.0942316492173698e-06,
      "loss": 0.0428,
      "num_input_tokens_seen": 32374720,
      "step": 36220
    },
    {
      "epoch": 9.560644054375082,
      "grad_norm": 0.11499189585447311,
      "learning_rate": 1.0913607641956841e-06,
      "loss": 0.0232,
      "num_input_tokens_seen": 32378944,
      "step": 36225
    },
    {
      "epoch": 9.561963837930579,
      "grad_norm": 0.08159168809652328,
      "learning_rate": 1.0884935661866213e-06,
      "loss": 0.0511,
      "num_input_tokens_seen": 32383392,
      "step": 36230
    },
    {
      "epoch": 9.563283621486077,
      "grad_norm": 0.1983921229839325,
      "learning_rate": 1.0856300556323418e-06,
      "loss": 0.0822,
      "num_input_tokens_seen": 32387840,
      "step": 36235
    },
    {
      "epoch": 9.564603405041574,
      "grad_norm": 0.0377018116414547,
      "learning_rate": 1.0827702329744365e-06,
      "loss": 0.0575,
      "num_input_tokens_seen": 32392224,
      "step": 36240
    },
    {
      "epoch": 9.56592318859707,
      "grad_norm": 0.08685025572776794,
      "learning_rate": 1.0799140986539197e-06,
      "loss": 0.0431,
      "num_input_tokens_seen": 32396800,
      "step": 36245
    },
    {
      "epoch": 9.567242972152567,
      "grad_norm": 0.20333470404148102,
      "learning_rate": 1.0770616531112526e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 32401216,
      "step": 36250
    },
    {
      "epoch": 9.568562755708063,
      "grad_norm": 0.44525718688964844,
      "learning_rate": 1.0742128967863085e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 32405888,
      "step": 36255
    },
    {
      "epoch": 9.569882539263562,
      "grad_norm": 0.4472443163394928,
      "learning_rate": 1.071367830118411e-06,
      "loss": 0.1023,
      "num_input_tokens_seen": 32410816,
      "step": 36260
    },
    {
      "epoch": 9.571202322819058,
      "grad_norm": 0.2957662045955658,
      "learning_rate": 1.068526453546298e-06,
      "loss": 0.0906,
      "num_input_tokens_seen": 32415328,
      "step": 36265
    },
    {
      "epoch": 9.572522106374555,
      "grad_norm": 0.13187162578105927,
      "learning_rate": 1.0656887675081467e-06,
      "loss": 0.0482,
      "num_input_tokens_seen": 32419616,
      "step": 36270
    },
    {
      "epoch": 9.573841889930051,
      "grad_norm": 0.3729113042354584,
      "learning_rate": 1.0628547724415628e-06,
      "loss": 0.0511,
      "num_input_tokens_seen": 32424160,
      "step": 36275
    },
    {
      "epoch": 9.575161673485548,
      "grad_norm": 0.30321475863456726,
      "learning_rate": 1.0600244687835881e-06,
      "loss": 0.0874,
      "num_input_tokens_seen": 32428608,
      "step": 36280
    },
    {
      "epoch": 9.576481457041044,
      "grad_norm": 0.23957787454128265,
      "learning_rate": 1.0571978569706876e-06,
      "loss": 0.0547,
      "num_input_tokens_seen": 32433536,
      "step": 36285
    },
    {
      "epoch": 9.577801240596543,
      "grad_norm": 0.346936970949173,
      "learning_rate": 1.0543749374387652e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 32438144,
      "step": 36290
    },
    {
      "epoch": 9.57912102415204,
      "grad_norm": 0.041708122938871384,
      "learning_rate": 1.051555710623142e-06,
      "loss": 0.0743,
      "num_input_tokens_seen": 32442720,
      "step": 36295
    },
    {
      "epoch": 9.580440807707536,
      "grad_norm": 0.25884315371513367,
      "learning_rate": 1.0487401769585847e-06,
      "loss": 0.0285,
      "num_input_tokens_seen": 32447072,
      "step": 36300
    },
    {
      "epoch": 9.581760591263032,
      "grad_norm": 0.21954306960105896,
      "learning_rate": 1.0459283368792845e-06,
      "loss": 0.1061,
      "num_input_tokens_seen": 32451680,
      "step": 36305
    },
    {
      "epoch": 9.583080374818529,
      "grad_norm": 0.35791873931884766,
      "learning_rate": 1.043120190818858e-06,
      "loss": 0.0734,
      "num_input_tokens_seen": 32456256,
      "step": 36310
    },
    {
      "epoch": 9.584400158374027,
      "grad_norm": 0.6011964678764343,
      "learning_rate": 1.0403157392103596e-06,
      "loss": 0.0976,
      "num_input_tokens_seen": 32461120,
      "step": 36315
    },
    {
      "epoch": 9.585719941929524,
      "grad_norm": 0.3216552138328552,
      "learning_rate": 1.0375149824862735e-06,
      "loss": 0.0704,
      "num_input_tokens_seen": 32465728,
      "step": 36320
    },
    {
      "epoch": 9.58703972548502,
      "grad_norm": 0.10040421783924103,
      "learning_rate": 1.034717921078507e-06,
      "loss": 0.1019,
      "num_input_tokens_seen": 32470048,
      "step": 36325
    },
    {
      "epoch": 9.588359509040517,
      "grad_norm": 0.34894701838493347,
      "learning_rate": 1.0319245554184009e-06,
      "loss": 0.0328,
      "num_input_tokens_seen": 32474656,
      "step": 36330
    },
    {
      "epoch": 9.589679292596013,
      "grad_norm": 0.0894196480512619,
      "learning_rate": 1.0291348859367361e-06,
      "loss": 0.1103,
      "num_input_tokens_seen": 32478976,
      "step": 36335
    },
    {
      "epoch": 9.590999076151512,
      "grad_norm": 0.42518916726112366,
      "learning_rate": 1.0263489130637016e-06,
      "loss": 0.0977,
      "num_input_tokens_seen": 32483680,
      "step": 36340
    },
    {
      "epoch": 9.592318859707008,
      "grad_norm": 0.05858302861452103,
      "learning_rate": 1.0235666372289427e-06,
      "loss": 0.086,
      "num_input_tokens_seen": 32488128,
      "step": 36345
    },
    {
      "epoch": 9.593638643262505,
      "grad_norm": 0.14783096313476562,
      "learning_rate": 1.0207880588615076e-06,
      "loss": 0.046,
      "num_input_tokens_seen": 32492704,
      "step": 36350
    },
    {
      "epoch": 9.594958426818001,
      "grad_norm": 0.2992841303348541,
      "learning_rate": 1.0180131783898984e-06,
      "loss": 0.0853,
      "num_input_tokens_seen": 32497152,
      "step": 36355
    },
    {
      "epoch": 9.596278210373498,
      "grad_norm": 0.33625030517578125,
      "learning_rate": 1.0152419962420362e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 32501536,
      "step": 36360
    },
    {
      "epoch": 9.597597993928996,
      "grad_norm": 0.10060246288776398,
      "learning_rate": 1.0124745128452685e-06,
      "loss": 0.0573,
      "num_input_tokens_seen": 32505856,
      "step": 36365
    },
    {
      "epoch": 9.598917777484493,
      "grad_norm": 0.1423308402299881,
      "learning_rate": 1.0097107286263758e-06,
      "loss": 0.0287,
      "num_input_tokens_seen": 32510592,
      "step": 36370
    },
    {
      "epoch": 9.60023756103999,
      "grad_norm": 0.08338498324155807,
      "learning_rate": 1.00695064401157e-06,
      "loss": 0.053,
      "num_input_tokens_seen": 32515008,
      "step": 36375
    },
    {
      "epoch": 9.601557344595486,
      "grad_norm": 0.24650920927524567,
      "learning_rate": 1.0041942594264886e-06,
      "loss": 0.0654,
      "num_input_tokens_seen": 32519680,
      "step": 36380
    },
    {
      "epoch": 9.602877128150983,
      "grad_norm": 0.291219025850296,
      "learning_rate": 1.001441575296208e-06,
      "loss": 0.0513,
      "num_input_tokens_seen": 32524064,
      "step": 36385
    },
    {
      "epoch": 9.604196911706481,
      "grad_norm": 0.09708461165428162,
      "learning_rate": 9.986925920452139e-07,
      "loss": 0.0209,
      "num_input_tokens_seen": 32528864,
      "step": 36390
    },
    {
      "epoch": 9.605516695261977,
      "grad_norm": 0.40332335233688354,
      "learning_rate": 9.959473100974475e-07,
      "loss": 0.1139,
      "num_input_tokens_seen": 32533440,
      "step": 36395
    },
    {
      "epoch": 9.606836478817474,
      "grad_norm": 0.43742066621780396,
      "learning_rate": 9.932057298762564e-07,
      "loss": 0.1163,
      "num_input_tokens_seen": 32537792,
      "step": 36400
    },
    {
      "epoch": 9.606836478817474,
      "eval_loss": 0.06680143624544144,
      "eval_runtime": 64.7362,
      "eval_samples_per_second": 104.038,
      "eval_steps_per_second": 26.013,
      "num_input_tokens_seen": 32537792,
      "step": 36400
    },
    {
      "epoch": 9.60815626237297,
      "grad_norm": 0.46489986777305603,
      "learning_rate": 9.90467851804433e-07,
      "loss": 0.1172,
      "num_input_tokens_seen": 32542528,
      "step": 36405
    },
    {
      "epoch": 9.609476045928467,
      "grad_norm": 0.29599010944366455,
      "learning_rate": 9.877336763041895e-07,
      "loss": 0.0525,
      "num_input_tokens_seen": 32546784,
      "step": 36410
    },
    {
      "epoch": 9.610795829483965,
      "grad_norm": 0.05403849855065346,
      "learning_rate": 9.850032037971662e-07,
      "loss": 0.0482,
      "num_input_tokens_seen": 32551040,
      "step": 36415
    },
    {
      "epoch": 9.612115613039462,
      "grad_norm": 0.15851284563541412,
      "learning_rate": 9.822764347044406e-07,
      "loss": 0.0247,
      "num_input_tokens_seen": 32555456,
      "step": 36420
    },
    {
      "epoch": 9.613435396594959,
      "grad_norm": 0.08824492990970612,
      "learning_rate": 9.795533694465175e-07,
      "loss": 0.043,
      "num_input_tokens_seen": 32559840,
      "step": 36425
    },
    {
      "epoch": 9.614755180150455,
      "grad_norm": 0.24742934107780457,
      "learning_rate": 9.768340084433197e-07,
      "loss": 0.0498,
      "num_input_tokens_seen": 32564064,
      "step": 36430
    },
    {
      "epoch": 9.616074963705952,
      "grad_norm": 0.2141820639371872,
      "learning_rate": 9.741183521142143e-07,
      "loss": 0.1092,
      "num_input_tokens_seen": 32568736,
      "step": 36435
    },
    {
      "epoch": 9.61739474726145,
      "grad_norm": 0.19324392080307007,
      "learning_rate": 9.714064008779889e-07,
      "loss": 0.0383,
      "num_input_tokens_seen": 32572864,
      "step": 36440
    },
    {
      "epoch": 9.618714530816947,
      "grad_norm": 0.5264033675193787,
      "learning_rate": 9.686981551528584e-07,
      "loss": 0.0754,
      "num_input_tokens_seen": 32577216,
      "step": 36445
    },
    {
      "epoch": 9.620034314372443,
      "grad_norm": 0.09812907874584198,
      "learning_rate": 9.65993615356467e-07,
      "loss": 0.0214,
      "num_input_tokens_seen": 32581504,
      "step": 36450
    },
    {
      "epoch": 9.62135409792794,
      "grad_norm": 0.11764270812273026,
      "learning_rate": 9.632927819058917e-07,
      "loss": 0.0573,
      "num_input_tokens_seen": 32586144,
      "step": 36455
    },
    {
      "epoch": 9.622673881483436,
      "grad_norm": 0.13817362487316132,
      "learning_rate": 9.605956552176305e-07,
      "loss": 0.042,
      "num_input_tokens_seen": 32590688,
      "step": 36460
    },
    {
      "epoch": 9.623993665038933,
      "grad_norm": 0.5800774693489075,
      "learning_rate": 9.579022357076223e-07,
      "loss": 0.0741,
      "num_input_tokens_seen": 32595296,
      "step": 36465
    },
    {
      "epoch": 9.625313448594431,
      "grad_norm": 0.058949973434209824,
      "learning_rate": 9.552125237912158e-07,
      "loss": 0.0316,
      "num_input_tokens_seen": 32599648,
      "step": 36470
    },
    {
      "epoch": 9.626633232149928,
      "grad_norm": 0.1207602247595787,
      "learning_rate": 9.525265198832096e-07,
      "loss": 0.0809,
      "num_input_tokens_seen": 32604000,
      "step": 36475
    },
    {
      "epoch": 9.627953015705424,
      "grad_norm": 0.7709706425666809,
      "learning_rate": 9.498442243978112e-07,
      "loss": 0.0914,
      "num_input_tokens_seen": 32608416,
      "step": 36480
    },
    {
      "epoch": 9.62927279926092,
      "grad_norm": 0.13213668763637543,
      "learning_rate": 9.471656377486649e-07,
      "loss": 0.0349,
      "num_input_tokens_seen": 32612832,
      "step": 36485
    },
    {
      "epoch": 9.630592582816417,
      "grad_norm": 0.08719389885663986,
      "learning_rate": 9.444907603488456e-07,
      "loss": 0.0639,
      "num_input_tokens_seen": 32617408,
      "step": 36490
    },
    {
      "epoch": 9.631912366371916,
      "grad_norm": 0.08487800508737564,
      "learning_rate": 9.418195926108514e-07,
      "loss": 0.0546,
      "num_input_tokens_seen": 32621856,
      "step": 36495
    },
    {
      "epoch": 9.633232149927412,
      "grad_norm": 0.2714018225669861,
      "learning_rate": 9.391521349466053e-07,
      "loss": 0.0473,
      "num_input_tokens_seen": 32626176,
      "step": 36500
    },
    {
      "epoch": 9.634551933482909,
      "grad_norm": 0.30420243740081787,
      "learning_rate": 9.364883877674758e-07,
      "loss": 0.0773,
      "num_input_tokens_seen": 32630688,
      "step": 36505
    },
    {
      "epoch": 9.635871717038405,
      "grad_norm": 0.6089054942131042,
      "learning_rate": 9.33828351484231e-07,
      "loss": 0.0818,
      "num_input_tokens_seen": 32634976,
      "step": 36510
    },
    {
      "epoch": 9.637191500593902,
      "grad_norm": 0.05980704724788666,
      "learning_rate": 9.311720265070906e-07,
      "loss": 0.0575,
      "num_input_tokens_seen": 32639584,
      "step": 36515
    },
    {
      "epoch": 9.6385112841494,
      "grad_norm": 0.11449901759624481,
      "learning_rate": 9.285194132456931e-07,
      "loss": 0.0412,
      "num_input_tokens_seen": 32644128,
      "step": 36520
    },
    {
      "epoch": 9.639831067704897,
      "grad_norm": 0.36789384484291077,
      "learning_rate": 9.258705121091032e-07,
      "loss": 0.0732,
      "num_input_tokens_seen": 32648480,
      "step": 36525
    },
    {
      "epoch": 9.641150851260393,
      "grad_norm": 0.26520252227783203,
      "learning_rate": 9.232253235058136e-07,
      "loss": 0.0421,
      "num_input_tokens_seen": 32653216,
      "step": 36530
    },
    {
      "epoch": 9.64247063481589,
      "grad_norm": 0.04465131461620331,
      "learning_rate": 9.205838478437478e-07,
      "loss": 0.0296,
      "num_input_tokens_seen": 32657920,
      "step": 36535
    },
    {
      "epoch": 9.643790418371387,
      "grad_norm": 0.058188553899526596,
      "learning_rate": 9.179460855302524e-07,
      "loss": 0.0551,
      "num_input_tokens_seen": 32662432,
      "step": 36540
    },
    {
      "epoch": 9.645110201926883,
      "grad_norm": 0.12666666507720947,
      "learning_rate": 9.153120369721046e-07,
      "loss": 0.0831,
      "num_input_tokens_seen": 32666944,
      "step": 36545
    },
    {
      "epoch": 9.646429985482381,
      "grad_norm": 0.03768955171108246,
      "learning_rate": 9.126817025755103e-07,
      "loss": 0.0564,
      "num_input_tokens_seen": 32671680,
      "step": 36550
    },
    {
      "epoch": 9.647749769037878,
      "grad_norm": 0.5245255827903748,
      "learning_rate": 9.100550827460947e-07,
      "loss": 0.0873,
      "num_input_tokens_seen": 32676128,
      "step": 36555
    },
    {
      "epoch": 9.649069552593375,
      "grad_norm": 0.37384510040283203,
      "learning_rate": 9.0743217788892e-07,
      "loss": 0.1058,
      "num_input_tokens_seen": 32680544,
      "step": 36560
    },
    {
      "epoch": 9.650389336148871,
      "grad_norm": 0.1574496477842331,
      "learning_rate": 9.048129884084683e-07,
      "loss": 0.074,
      "num_input_tokens_seen": 32684800,
      "step": 36565
    },
    {
      "epoch": 9.651709119704368,
      "grad_norm": 0.024178965017199516,
      "learning_rate": 9.021975147086553e-07,
      "loss": 0.0335,
      "num_input_tokens_seen": 32689152,
      "step": 36570
    },
    {
      "epoch": 9.653028903259866,
      "grad_norm": 0.17039723694324493,
      "learning_rate": 8.995857571928141e-07,
      "loss": 0.0637,
      "num_input_tokens_seen": 32693440,
      "step": 36575
    },
    {
      "epoch": 9.654348686815363,
      "grad_norm": 0.23729397356510162,
      "learning_rate": 8.969777162637139e-07,
      "loss": 0.0434,
      "num_input_tokens_seen": 32698048,
      "step": 36580
    },
    {
      "epoch": 9.655668470370859,
      "grad_norm": 0.21668079495429993,
      "learning_rate": 8.943733923235525e-07,
      "loss": 0.0783,
      "num_input_tokens_seen": 32702304,
      "step": 36585
    },
    {
      "epoch": 9.656988253926356,
      "grad_norm": 0.14011089503765106,
      "learning_rate": 8.917727857739394e-07,
      "loss": 0.0562,
      "num_input_tokens_seen": 32706368,
      "step": 36590
    },
    {
      "epoch": 9.658308037481852,
      "grad_norm": 0.19592523574829102,
      "learning_rate": 8.891758970159258e-07,
      "loss": 0.0522,
      "num_input_tokens_seen": 32710528,
      "step": 36595
    },
    {
      "epoch": 9.65962782103735,
      "grad_norm": 0.15823659300804138,
      "learning_rate": 8.86582726449986e-07,
      "loss": 0.0261,
      "num_input_tokens_seen": 32714880,
      "step": 36600
    },
    {
      "epoch": 9.65962782103735,
      "eval_loss": 0.06666872650384903,
      "eval_runtime": 64.7204,
      "eval_samples_per_second": 104.063,
      "eval_steps_per_second": 26.02,
      "num_input_tokens_seen": 32714880,
      "step": 36600
    },
    {
      "epoch": 9.660947604592847,
      "grad_norm": 0.05643278360366821,
      "learning_rate": 8.839932744760165e-07,
      "loss": 0.0479,
      "num_input_tokens_seen": 32719584,
      "step": 36605
    },
    {
      "epoch": 9.662267388148344,
      "grad_norm": 0.14952188730239868,
      "learning_rate": 8.814075414933482e-07,
      "loss": 0.0433,
      "num_input_tokens_seen": 32723776,
      "step": 36610
    },
    {
      "epoch": 9.66358717170384,
      "grad_norm": 1.0100500583648682,
      "learning_rate": 8.788255279007257e-07,
      "loss": 0.1414,
      "num_input_tokens_seen": 32728544,
      "step": 36615
    },
    {
      "epoch": 9.664906955259337,
      "grad_norm": 0.26396429538726807,
      "learning_rate": 8.762472340963362e-07,
      "loss": 0.0365,
      "num_input_tokens_seen": 32733056,
      "step": 36620
    },
    {
      "epoch": 9.666226738814835,
      "grad_norm": 0.1437600702047348,
      "learning_rate": 8.736726604777811e-07,
      "loss": 0.0321,
      "num_input_tokens_seen": 32737408,
      "step": 36625
    },
    {
      "epoch": 9.667546522370332,
      "grad_norm": 0.2120397537946701,
      "learning_rate": 8.711018074420901e-07,
      "loss": 0.0727,
      "num_input_tokens_seen": 32741888,
      "step": 36630
    },
    {
      "epoch": 9.668866305925828,
      "grad_norm": 0.24308793246746063,
      "learning_rate": 8.685346753857209e-07,
      "loss": 0.0552,
      "num_input_tokens_seen": 32746432,
      "step": 36635
    },
    {
      "epoch": 9.670186089481325,
      "grad_norm": 0.37679094076156616,
      "learning_rate": 8.659712647045654e-07,
      "loss": 0.0588,
      "num_input_tokens_seen": 32750880,
      "step": 36640
    },
    {
      "epoch": 9.671505873036821,
      "grad_norm": 0.32582390308380127,
      "learning_rate": 8.634115757939209e-07,
      "loss": 0.0536,
      "num_input_tokens_seen": 32755040,
      "step": 36645
    },
    {
      "epoch": 9.67282565659232,
      "grad_norm": 0.045773304998874664,
      "learning_rate": 8.608556090485387e-07,
      "loss": 0.0664,
      "num_input_tokens_seen": 32759264,
      "step": 36650
    },
    {
      "epoch": 9.674145440147816,
      "grad_norm": 0.24038860201835632,
      "learning_rate": 8.583033648625671e-07,
      "loss": 0.0875,
      "num_input_tokens_seen": 32763840,
      "step": 36655
    },
    {
      "epoch": 9.675465223703313,
      "grad_norm": 0.10279779136180878,
      "learning_rate": 8.557548436295998e-07,
      "loss": 0.0255,
      "num_input_tokens_seen": 32768192,
      "step": 36660
    },
    {
      "epoch": 9.67678500725881,
      "grad_norm": 0.02744360640645027,
      "learning_rate": 8.532100457426556e-07,
      "loss": 0.0589,
      "num_input_tokens_seen": 32772384,
      "step": 36665
    },
    {
      "epoch": 9.678104790814306,
      "grad_norm": 0.18834277987480164,
      "learning_rate": 8.506689715941679e-07,
      "loss": 0.0456,
      "num_input_tokens_seen": 32776832,
      "step": 36670
    },
    {
      "epoch": 9.679424574369804,
      "grad_norm": 0.09027508646249771,
      "learning_rate": 8.481316215760011e-07,
      "loss": 0.016,
      "num_input_tokens_seen": 32781120,
      "step": 36675
    },
    {
      "epoch": 9.6807443579253,
      "grad_norm": 0.4615442752838135,
      "learning_rate": 8.455979960794558e-07,
      "loss": 0.0676,
      "num_input_tokens_seen": 32785760,
      "step": 36680
    },
    {
      "epoch": 9.682064141480797,
      "grad_norm": 0.41708117723464966,
      "learning_rate": 8.430680954952364e-07,
      "loss": 0.0846,
      "num_input_tokens_seen": 32790624,
      "step": 36685
    },
    {
      "epoch": 9.683383925036294,
      "grad_norm": 0.0963064432144165,
      "learning_rate": 8.405419202134974e-07,
      "loss": 0.0589,
      "num_input_tokens_seen": 32795296,
      "step": 36690
    },
    {
      "epoch": 9.68470370859179,
      "grad_norm": 0.031728584319353104,
      "learning_rate": 8.380194706237993e-07,
      "loss": 0.0786,
      "num_input_tokens_seen": 32800096,
      "step": 36695
    },
    {
      "epoch": 9.686023492147289,
      "grad_norm": 0.1846175640821457,
      "learning_rate": 8.355007471151366e-07,
      "loss": 0.0335,
      "num_input_tokens_seen": 32804736,
      "step": 36700
    },
    {
      "epoch": 9.687343275702785,
      "grad_norm": 0.1352980136871338,
      "learning_rate": 8.329857500759292e-07,
      "loss": 0.0557,
      "num_input_tokens_seen": 32809152,
      "step": 36705
    },
    {
      "epoch": 9.688663059258282,
      "grad_norm": 0.2586682438850403,
      "learning_rate": 8.304744798940194e-07,
      "loss": 0.0825,
      "num_input_tokens_seen": 32813152,
      "step": 36710
    },
    {
      "epoch": 9.689982842813778,
      "grad_norm": 0.3628913462162018,
      "learning_rate": 8.279669369566756e-07,
      "loss": 0.0418,
      "num_input_tokens_seen": 32817536,
      "step": 36715
    },
    {
      "epoch": 9.691302626369275,
      "grad_norm": 0.027791297063231468,
      "learning_rate": 8.254631216505993e-07,
      "loss": 0.058,
      "num_input_tokens_seen": 32822080,
      "step": 36720
    },
    {
      "epoch": 9.692622409924772,
      "grad_norm": 0.3089745044708252,
      "learning_rate": 8.229630343619038e-07,
      "loss": 0.0797,
      "num_input_tokens_seen": 32826496,
      "step": 36725
    },
    {
      "epoch": 9.69394219348027,
      "grad_norm": 0.11913476139307022,
      "learning_rate": 8.204666754761392e-07,
      "loss": 0.0654,
      "num_input_tokens_seen": 32831072,
      "step": 36730
    },
    {
      "epoch": 9.695261977035766,
      "grad_norm": 0.08050766587257385,
      "learning_rate": 8.179740453782669e-07,
      "loss": 0.0341,
      "num_input_tokens_seen": 32835424,
      "step": 36735
    },
    {
      "epoch": 9.696581760591263,
      "grad_norm": 0.06045312434434891,
      "learning_rate": 8.154851444526907e-07,
      "loss": 0.0319,
      "num_input_tokens_seen": 32839648,
      "step": 36740
    },
    {
      "epoch": 9.69790154414676,
      "grad_norm": 0.32110732793807983,
      "learning_rate": 8.129999730832283e-07,
      "loss": 0.0537,
      "num_input_tokens_seen": 32844544,
      "step": 36745
    },
    {
      "epoch": 9.699221327702256,
      "grad_norm": 0.23077958822250366,
      "learning_rate": 8.105185316531178e-07,
      "loss": 0.0399,
      "num_input_tokens_seen": 32849152,
      "step": 36750
    },
    {
      "epoch": 9.700541111257754,
      "grad_norm": 0.16372065246105194,
      "learning_rate": 8.08040820545039e-07,
      "loss": 0.046,
      "num_input_tokens_seen": 32853632,
      "step": 36755
    },
    {
      "epoch": 9.701860894813251,
      "grad_norm": 0.22356246411800385,
      "learning_rate": 8.055668401410782e-07,
      "loss": 0.0751,
      "num_input_tokens_seen": 32857952,
      "step": 36760
    },
    {
      "epoch": 9.703180678368748,
      "grad_norm": 0.4643203616142273,
      "learning_rate": 8.030965908227578e-07,
      "loss": 0.0483,
      "num_input_tokens_seen": 32862528,
      "step": 36765
    },
    {
      "epoch": 9.704500461924244,
      "grad_norm": 0.15463142096996307,
      "learning_rate": 8.006300729710203e-07,
      "loss": 0.034,
      "num_input_tokens_seen": 32866784,
      "step": 36770
    },
    {
      "epoch": 9.70582024547974,
      "grad_norm": 0.1039866954088211,
      "learning_rate": 7.981672869662337e-07,
      "loss": 0.0279,
      "num_input_tokens_seen": 32871072,
      "step": 36775
    },
    {
      "epoch": 9.707140029035239,
      "grad_norm": 0.41785889863967896,
      "learning_rate": 7.957082331881888e-07,
      "loss": 0.1147,
      "num_input_tokens_seen": 32875680,
      "step": 36780
    },
    {
      "epoch": 9.708459812590736,
      "grad_norm": 0.33697253465652466,
      "learning_rate": 7.932529120161069e-07,
      "loss": 0.0674,
      "num_input_tokens_seen": 32880032,
      "step": 36785
    },
    {
      "epoch": 9.709779596146232,
      "grad_norm": 0.34266966581344604,
      "learning_rate": 7.908013238286243e-07,
      "loss": 0.062,
      "num_input_tokens_seen": 32884608,
      "step": 36790
    },
    {
      "epoch": 9.711099379701729,
      "grad_norm": 0.10386625677347183,
      "learning_rate": 7.883534690038136e-07,
      "loss": 0.0427,
      "num_input_tokens_seen": 32889152,
      "step": 36795
    },
    {
      "epoch": 9.712419163257225,
      "grad_norm": 0.23419667780399323,
      "learning_rate": 7.859093479191559e-07,
      "loss": 0.0462,
      "num_input_tokens_seen": 32893312,
      "step": 36800
    },
    {
      "epoch": 9.712419163257225,
      "eval_loss": 0.06686647236347198,
      "eval_runtime": 64.7556,
      "eval_samples_per_second": 104.006,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 32893312,
      "step": 36800
    },
    {
      "epoch": 9.713738946812724,
      "grad_norm": 0.1796492040157318,
      "learning_rate": 7.834689609515722e-07,
      "loss": 0.0275,
      "num_input_tokens_seen": 32898016,
      "step": 36805
    },
    {
      "epoch": 9.71505873036822,
      "grad_norm": 0.25671592354774475,
      "learning_rate": 7.810323084774002e-07,
      "loss": 0.0467,
      "num_input_tokens_seen": 32902432,
      "step": 36810
    },
    {
      "epoch": 9.716378513923717,
      "grad_norm": 0.3190504014492035,
      "learning_rate": 7.785993908723976e-07,
      "loss": 0.0625,
      "num_input_tokens_seen": 32906816,
      "step": 36815
    },
    {
      "epoch": 9.717698297479213,
      "grad_norm": 0.1363082379102707,
      "learning_rate": 7.761702085117534e-07,
      "loss": 0.0466,
      "num_input_tokens_seen": 32911520,
      "step": 36820
    },
    {
      "epoch": 9.71901808103471,
      "grad_norm": 0.08445966988801956,
      "learning_rate": 7.737447617700844e-07,
      "loss": 0.0401,
      "num_input_tokens_seen": 32915968,
      "step": 36825
    },
    {
      "epoch": 9.720337864590206,
      "grad_norm": 0.049805037677288055,
      "learning_rate": 7.713230510214136e-07,
      "loss": 0.0825,
      "num_input_tokens_seen": 32920640,
      "step": 36830
    },
    {
      "epoch": 9.721657648145705,
      "grad_norm": 0.1078580841422081,
      "learning_rate": 7.689050766392092e-07,
      "loss": 0.0553,
      "num_input_tokens_seen": 32925088,
      "step": 36835
    },
    {
      "epoch": 9.722977431701201,
      "grad_norm": 0.01730385422706604,
      "learning_rate": 7.664908389963477e-07,
      "loss": 0.0209,
      "num_input_tokens_seen": 32929376,
      "step": 36840
    },
    {
      "epoch": 9.724297215256698,
      "grad_norm": 0.19974097609519958,
      "learning_rate": 7.64080338465134e-07,
      "loss": 0.0972,
      "num_input_tokens_seen": 32933952,
      "step": 36845
    },
    {
      "epoch": 9.725616998812194,
      "grad_norm": 0.22803504765033722,
      "learning_rate": 7.616735754173043e-07,
      "loss": 0.0371,
      "num_input_tokens_seen": 32938304,
      "step": 36850
    },
    {
      "epoch": 9.72693678236769,
      "grad_norm": 0.24974143505096436,
      "learning_rate": 7.592705502240005e-07,
      "loss": 0.075,
      "num_input_tokens_seen": 32942720,
      "step": 36855
    },
    {
      "epoch": 9.72825656592319,
      "grad_norm": 0.08736464381217957,
      "learning_rate": 7.568712632558095e-07,
      "loss": 0.042,
      "num_input_tokens_seen": 32947296,
      "step": 36860
    },
    {
      "epoch": 9.729576349478686,
      "grad_norm": 0.1285615861415863,
      "learning_rate": 7.544757148827297e-07,
      "loss": 0.0465,
      "num_input_tokens_seen": 32951648,
      "step": 36865
    },
    {
      "epoch": 9.730896133034182,
      "grad_norm": 0.35199403762817383,
      "learning_rate": 7.520839054741797e-07,
      "loss": 0.1016,
      "num_input_tokens_seen": 32955872,
      "step": 36870
    },
    {
      "epoch": 9.732215916589679,
      "grad_norm": 0.34824544191360474,
      "learning_rate": 7.496958353990113e-07,
      "loss": 0.0519,
      "num_input_tokens_seen": 32960224,
      "step": 36875
    },
    {
      "epoch": 9.733535700145175,
      "grad_norm": 0.6092756390571594,
      "learning_rate": 7.473115050254941e-07,
      "loss": 0.0563,
      "num_input_tokens_seen": 32964832,
      "step": 36880
    },
    {
      "epoch": 9.734855483700674,
      "grad_norm": 0.13450933992862701,
      "learning_rate": 7.449309147213173e-07,
      "loss": 0.03,
      "num_input_tokens_seen": 32969344,
      "step": 36885
    },
    {
      "epoch": 9.73617526725617,
      "grad_norm": 0.14798036217689514,
      "learning_rate": 7.425540648536067e-07,
      "loss": 0.0451,
      "num_input_tokens_seen": 32973984,
      "step": 36890
    },
    {
      "epoch": 9.737495050811667,
      "grad_norm": 0.23962444067001343,
      "learning_rate": 7.40180955788894e-07,
      "loss": 0.0529,
      "num_input_tokens_seen": 32978272,
      "step": 36895
    },
    {
      "epoch": 9.738814834367163,
      "grad_norm": 0.24147900938987732,
      "learning_rate": 7.378115878931474e-07,
      "loss": 0.0714,
      "num_input_tokens_seen": 32982976,
      "step": 36900
    },
    {
      "epoch": 9.74013461792266,
      "grad_norm": 0.5479511618614197,
      "learning_rate": 7.354459615317527e-07,
      "loss": 0.0873,
      "num_input_tokens_seen": 32987520,
      "step": 36905
    },
    {
      "epoch": 9.741454401478158,
      "grad_norm": 0.06193297728896141,
      "learning_rate": 7.33084077069518e-07,
      "loss": 0.0636,
      "num_input_tokens_seen": 32991840,
      "step": 36910
    },
    {
      "epoch": 9.742774185033655,
      "grad_norm": 0.07315385341644287,
      "learning_rate": 7.307259348706768e-07,
      "loss": 0.0553,
      "num_input_tokens_seen": 32996192,
      "step": 36915
    },
    {
      "epoch": 9.744093968589151,
      "grad_norm": 0.4768737554550171,
      "learning_rate": 7.283715352988801e-07,
      "loss": 0.0252,
      "num_input_tokens_seen": 33000576,
      "step": 36920
    },
    {
      "epoch": 9.745413752144648,
      "grad_norm": 0.36762920022010803,
      "learning_rate": 7.260208787172068e-07,
      "loss": 0.1161,
      "num_input_tokens_seen": 33005056,
      "step": 36925
    },
    {
      "epoch": 9.746733535700145,
      "grad_norm": 0.4105690121650696,
      "learning_rate": 7.23673965488167e-07,
      "loss": 0.0841,
      "num_input_tokens_seen": 33009408,
      "step": 36930
    },
    {
      "epoch": 9.748053319255643,
      "grad_norm": 0.1448710858821869,
      "learning_rate": 7.213307959736709e-07,
      "loss": 0.0401,
      "num_input_tokens_seen": 33013952,
      "step": 36935
    },
    {
      "epoch": 9.74937310281114,
      "grad_norm": 0.07691355794668198,
      "learning_rate": 7.189913705350715e-07,
      "loss": 0.0322,
      "num_input_tokens_seen": 33018272,
      "step": 36940
    },
    {
      "epoch": 9.750692886366636,
      "grad_norm": 0.1460392326116562,
      "learning_rate": 7.166556895331411e-07,
      "loss": 0.0356,
      "num_input_tokens_seen": 33022528,
      "step": 36945
    },
    {
      "epoch": 9.752012669922133,
      "grad_norm": 0.05647468939423561,
      "learning_rate": 7.143237533280639e-07,
      "loss": 0.0344,
      "num_input_tokens_seen": 33027520,
      "step": 36950
    },
    {
      "epoch": 9.75333245347763,
      "grad_norm": 0.22515201568603516,
      "learning_rate": 7.119955622794578e-07,
      "loss": 0.0754,
      "num_input_tokens_seen": 33032032,
      "step": 36955
    },
    {
      "epoch": 9.754652237033127,
      "grad_norm": 0.5260065197944641,
      "learning_rate": 7.096711167463577e-07,
      "loss": 0.0703,
      "num_input_tokens_seen": 33036128,
      "step": 36960
    },
    {
      "epoch": 9.755972020588624,
      "grad_norm": 0.2403654307126999,
      "learning_rate": 7.073504170872213e-07,
      "loss": 0.0985,
      "num_input_tokens_seen": 33040448,
      "step": 36965
    },
    {
      "epoch": 9.75729180414412,
      "grad_norm": 0.18120531737804413,
      "learning_rate": 7.05033463659932e-07,
      "loss": 0.0462,
      "num_input_tokens_seen": 33044960,
      "step": 36970
    },
    {
      "epoch": 9.758611587699617,
      "grad_norm": 0.24936379492282867,
      "learning_rate": 7.027202568217928e-07,
      "loss": 0.0378,
      "num_input_tokens_seen": 33049312,
      "step": 36975
    },
    {
      "epoch": 9.759931371255114,
      "grad_norm": 0.08598634600639343,
      "learning_rate": 7.004107969295293e-07,
      "loss": 0.02,
      "num_input_tokens_seen": 33053632,
      "step": 36980
    },
    {
      "epoch": 9.76125115481061,
      "grad_norm": 0.13366326689720154,
      "learning_rate": 6.9810508433929e-07,
      "loss": 0.0389,
      "num_input_tokens_seen": 33057856,
      "step": 36985
    },
    {
      "epoch": 9.762570938366109,
      "grad_norm": 0.04107854142785072,
      "learning_rate": 6.958031194066406e-07,
      "loss": 0.0464,
      "num_input_tokens_seen": 33062592,
      "step": 36990
    },
    {
      "epoch": 9.763890721921605,
      "grad_norm": 0.11019769310951233,
      "learning_rate": 6.935049024865776e-07,
      "loss": 0.0237,
      "num_input_tokens_seen": 33067104,
      "step": 36995
    },
    {
      "epoch": 9.765210505477102,
      "grad_norm": 0.07969275861978531,
      "learning_rate": 6.912104339335118e-07,
      "loss": 0.0218,
      "num_input_tokens_seen": 33071840,
      "step": 37000
    },
    {
      "epoch": 9.765210505477102,
      "eval_loss": 0.06681240350008011,
      "eval_runtime": 64.7068,
      "eval_samples_per_second": 104.085,
      "eval_steps_per_second": 26.025,
      "num_input_tokens_seen": 33071840,
      "step": 37000
    },
    {
      "epoch": 9.766530289032598,
      "grad_norm": 0.15063218772411346,
      "learning_rate": 6.889197141012799e-07,
      "loss": 0.0416,
      "num_input_tokens_seen": 33076352,
      "step": 37005
    },
    {
      "epoch": 9.767850072588095,
      "grad_norm": 0.14502611756324768,
      "learning_rate": 6.866327433431435e-07,
      "loss": 0.074,
      "num_input_tokens_seen": 33080992,
      "step": 37010
    },
    {
      "epoch": 9.769169856143593,
      "grad_norm": 0.03428981080651283,
      "learning_rate": 6.843495220117735e-07,
      "loss": 0.1014,
      "num_input_tokens_seen": 33085216,
      "step": 37015
    },
    {
      "epoch": 9.77048963969909,
      "grad_norm": 0.05172092467546463,
      "learning_rate": 6.820700504592798e-07,
      "loss": 0.0452,
      "num_input_tokens_seen": 33089440,
      "step": 37020
    },
    {
      "epoch": 9.771809423254586,
      "grad_norm": 0.2527451515197754,
      "learning_rate": 6.797943290371839e-07,
      "loss": 0.0225,
      "num_input_tokens_seen": 33093824,
      "step": 37025
    },
    {
      "epoch": 9.773129206810083,
      "grad_norm": 0.16107957065105438,
      "learning_rate": 6.775223580964274e-07,
      "loss": 0.0357,
      "num_input_tokens_seen": 33098272,
      "step": 37030
    },
    {
      "epoch": 9.77444899036558,
      "grad_norm": 0.2943838834762573,
      "learning_rate": 6.7525413798738e-07,
      "loss": 0.118,
      "num_input_tokens_seen": 33102784,
      "step": 37035
    },
    {
      "epoch": 9.775768773921078,
      "grad_norm": 0.0790591612458229,
      "learning_rate": 6.729896690598259e-07,
      "loss": 0.0633,
      "num_input_tokens_seen": 33106976,
      "step": 37040
    },
    {
      "epoch": 9.777088557476574,
      "grad_norm": 0.20587407052516937,
      "learning_rate": 6.707289516629772e-07,
      "loss": 0.0348,
      "num_input_tokens_seen": 33111264,
      "step": 37045
    },
    {
      "epoch": 9.77840834103207,
      "grad_norm": 0.16879156231880188,
      "learning_rate": 6.684719861454692e-07,
      "loss": 0.0531,
      "num_input_tokens_seen": 33115904,
      "step": 37050
    },
    {
      "epoch": 9.779728124587567,
      "grad_norm": 0.04438408091664314,
      "learning_rate": 6.662187728553481e-07,
      "loss": 0.0592,
      "num_input_tokens_seen": 33120576,
      "step": 37055
    },
    {
      "epoch": 9.781047908143064,
      "grad_norm": 0.12252353131771088,
      "learning_rate": 6.639693121400892e-07,
      "loss": 0.0638,
      "num_input_tokens_seen": 33124992,
      "step": 37060
    },
    {
      "epoch": 9.782367691698562,
      "grad_norm": 0.3266213536262512,
      "learning_rate": 6.617236043465868e-07,
      "loss": 0.0559,
      "num_input_tokens_seen": 33129408,
      "step": 37065
    },
    {
      "epoch": 9.783687475254059,
      "grad_norm": 0.1434156447649002,
      "learning_rate": 6.594816498211587e-07,
      "loss": 0.0385,
      "num_input_tokens_seen": 33134016,
      "step": 37070
    },
    {
      "epoch": 9.785007258809555,
      "grad_norm": 0.27777019143104553,
      "learning_rate": 6.572434489095447e-07,
      "loss": 0.0677,
      "num_input_tokens_seen": 33138720,
      "step": 37075
    },
    {
      "epoch": 9.786327042365052,
      "grad_norm": 0.34602275490760803,
      "learning_rate": 6.550090019568994e-07,
      "loss": 0.0871,
      "num_input_tokens_seen": 33143072,
      "step": 37080
    },
    {
      "epoch": 9.787646825920548,
      "grad_norm": 0.24594658613204956,
      "learning_rate": 6.527783093078027e-07,
      "loss": 0.0518,
      "num_input_tokens_seen": 33147616,
      "step": 37085
    },
    {
      "epoch": 9.788966609476045,
      "grad_norm": 0.07750890403985977,
      "learning_rate": 6.5055137130626e-07,
      "loss": 0.0744,
      "num_input_tokens_seen": 33152256,
      "step": 37090
    },
    {
      "epoch": 9.790286393031543,
      "grad_norm": 0.09458242356777191,
      "learning_rate": 6.483281882956854e-07,
      "loss": 0.0657,
      "num_input_tokens_seen": 33156640,
      "step": 37095
    },
    {
      "epoch": 9.79160617658704,
      "grad_norm": 0.12512488663196564,
      "learning_rate": 6.461087606189298e-07,
      "loss": 0.0524,
      "num_input_tokens_seen": 33161376,
      "step": 37100
    },
    {
      "epoch": 9.792925960142536,
      "grad_norm": 0.10112882405519485,
      "learning_rate": 6.438930886182554e-07,
      "loss": 0.0389,
      "num_input_tokens_seen": 33165568,
      "step": 37105
    },
    {
      "epoch": 9.794245743698033,
      "grad_norm": 0.05718434974551201,
      "learning_rate": 6.416811726353417e-07,
      "loss": 0.0601,
      "num_input_tokens_seen": 33169760,
      "step": 37110
    },
    {
      "epoch": 9.79556552725353,
      "grad_norm": 0.2776090204715729,
      "learning_rate": 6.394730130112991e-07,
      "loss": 0.0375,
      "num_input_tokens_seen": 33174464,
      "step": 37115
    },
    {
      "epoch": 9.796885310809028,
      "grad_norm": 0.1050463393330574,
      "learning_rate": 6.372686100866471e-07,
      "loss": 0.0588,
      "num_input_tokens_seen": 33179008,
      "step": 37120
    },
    {
      "epoch": 9.798205094364524,
      "grad_norm": 0.025791127234697342,
      "learning_rate": 6.350679642013413e-07,
      "loss": 0.1004,
      "num_input_tokens_seen": 33183360,
      "step": 37125
    },
    {
      "epoch": 9.799524877920021,
      "grad_norm": 0.42565640807151794,
      "learning_rate": 6.328710756947437e-07,
      "loss": 0.035,
      "num_input_tokens_seen": 33187648,
      "step": 37130
    },
    {
      "epoch": 9.800844661475518,
      "grad_norm": 0.3045409023761749,
      "learning_rate": 6.306779449056416e-07,
      "loss": 0.0555,
      "num_input_tokens_seen": 33192352,
      "step": 37135
    },
    {
      "epoch": 9.802164445031014,
      "grad_norm": 0.2856617271900177,
      "learning_rate": 6.284885721722422e-07,
      "loss": 0.0589,
      "num_input_tokens_seen": 33196864,
      "step": 37140
    },
    {
      "epoch": 9.803484228586512,
      "grad_norm": 0.0676911398768425,
      "learning_rate": 6.26302957832181e-07,
      "loss": 0.0575,
      "num_input_tokens_seen": 33201568,
      "step": 37145
    },
    {
      "epoch": 9.804804012142009,
      "grad_norm": 0.045551009476184845,
      "learning_rate": 6.241211022224997e-07,
      "loss": 0.0453,
      "num_input_tokens_seen": 33206112,
      "step": 37150
    },
    {
      "epoch": 9.806123795697506,
      "grad_norm": 0.0717528834939003,
      "learning_rate": 6.219430056796732e-07,
      "loss": 0.036,
      "num_input_tokens_seen": 33210496,
      "step": 37155
    },
    {
      "epoch": 9.807443579253002,
      "grad_norm": 0.173980712890625,
      "learning_rate": 6.19768668539586e-07,
      "loss": 0.0346,
      "num_input_tokens_seen": 33215264,
      "step": 37160
    },
    {
      "epoch": 9.808763362808499,
      "grad_norm": 0.09196390956640244,
      "learning_rate": 6.175980911375528e-07,
      "loss": 0.0225,
      "num_input_tokens_seen": 33219680,
      "step": 37165
    },
    {
      "epoch": 9.810083146363997,
      "grad_norm": 0.08689911663532257,
      "learning_rate": 6.154312738083034e-07,
      "loss": 0.0395,
      "num_input_tokens_seen": 33224352,
      "step": 37170
    },
    {
      "epoch": 9.811402929919494,
      "grad_norm": 0.09039895981550217,
      "learning_rate": 6.132682168859843e-07,
      "loss": 0.0324,
      "num_input_tokens_seen": 33228736,
      "step": 37175
    },
    {
      "epoch": 9.81272271347499,
      "grad_norm": 0.10126292705535889,
      "learning_rate": 6.111089207041704e-07,
      "loss": 0.0776,
      "num_input_tokens_seen": 33233600,
      "step": 37180
    },
    {
      "epoch": 9.814042497030487,
      "grad_norm": 0.10259318351745605,
      "learning_rate": 6.089533855958507e-07,
      "loss": 0.0372,
      "num_input_tokens_seen": 33238048,
      "step": 37185
    },
    {
      "epoch": 9.815362280585983,
      "grad_norm": 0.229819655418396,
      "learning_rate": 6.068016118934372e-07,
      "loss": 0.0296,
      "num_input_tokens_seen": 33242400,
      "step": 37190
    },
    {
      "epoch": 9.816682064141482,
      "grad_norm": 0.09456592053174973,
      "learning_rate": 6.04653599928759e-07,
      "loss": 0.0216,
      "num_input_tokens_seen": 33246848,
      "step": 37195
    },
    {
      "epoch": 9.818001847696978,
      "grad_norm": 0.2497648000717163,
      "learning_rate": 6.025093500330675e-07,
      "loss": 0.0752,
      "num_input_tokens_seen": 33251936,
      "step": 37200
    },
    {
      "epoch": 9.818001847696978,
      "eval_loss": 0.06687842309474945,
      "eval_runtime": 64.7714,
      "eval_samples_per_second": 103.981,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 33251936,
      "step": 37200
    },
    {
      "epoch": 9.819321631252475,
      "grad_norm": 0.14527736604213715,
      "learning_rate": 6.003688625370291e-07,
      "loss": 0.0424,
      "num_input_tokens_seen": 33256288,
      "step": 37205
    },
    {
      "epoch": 9.820641414807971,
      "grad_norm": 0.02420186623930931,
      "learning_rate": 5.982321377707406e-07,
      "loss": 0.0373,
      "num_input_tokens_seen": 33260928,
      "step": 37210
    },
    {
      "epoch": 9.821961198363468,
      "grad_norm": 0.11463820934295654,
      "learning_rate": 5.96099176063708e-07,
      "loss": 0.105,
      "num_input_tokens_seen": 33265344,
      "step": 37215
    },
    {
      "epoch": 9.823280981918966,
      "grad_norm": 0.3090822398662567,
      "learning_rate": 5.93969977744857e-07,
      "loss": 0.075,
      "num_input_tokens_seen": 33269728,
      "step": 37220
    },
    {
      "epoch": 9.824600765474463,
      "grad_norm": 0.32196441292762756,
      "learning_rate": 5.918445431425445e-07,
      "loss": 0.0914,
      "num_input_tokens_seen": 33274176,
      "step": 37225
    },
    {
      "epoch": 9.82592054902996,
      "grad_norm": 0.05316760018467903,
      "learning_rate": 5.897228725845333e-07,
      "loss": 0.04,
      "num_input_tokens_seen": 33278592,
      "step": 37230
    },
    {
      "epoch": 9.827240332585456,
      "grad_norm": 0.18035797774791718,
      "learning_rate": 5.876049663980171e-07,
      "loss": 0.0393,
      "num_input_tokens_seen": 33282912,
      "step": 37235
    },
    {
      "epoch": 9.828560116140952,
      "grad_norm": 0.5084364414215088,
      "learning_rate": 5.854908249095959e-07,
      "loss": 0.0717,
      "num_input_tokens_seen": 33287360,
      "step": 37240
    },
    {
      "epoch": 9.82987989969645,
      "grad_norm": 0.07268796116113663,
      "learning_rate": 5.833804484453031e-07,
      "loss": 0.0609,
      "num_input_tokens_seen": 33291904,
      "step": 37245
    },
    {
      "epoch": 9.831199683251947,
      "grad_norm": 0.24577440321445465,
      "learning_rate": 5.81273837330587e-07,
      "loss": 0.0623,
      "num_input_tokens_seen": 33296256,
      "step": 37250
    },
    {
      "epoch": 9.832519466807444,
      "grad_norm": 0.08234730362892151,
      "learning_rate": 5.791709918903071e-07,
      "loss": 0.0855,
      "num_input_tokens_seen": 33300736,
      "step": 37255
    },
    {
      "epoch": 9.83383925036294,
      "grad_norm": 0.168877974152565,
      "learning_rate": 5.770719124487483e-07,
      "loss": 0.0362,
      "num_input_tokens_seen": 33305248,
      "step": 37260
    },
    {
      "epoch": 9.835159033918437,
      "grad_norm": 0.4873526096343994,
      "learning_rate": 5.749765993296241e-07,
      "loss": 0.0736,
      "num_input_tokens_seen": 33309760,
      "step": 37265
    },
    {
      "epoch": 9.836478817473933,
      "grad_norm": 0.21717356145381927,
      "learning_rate": 5.728850528560509e-07,
      "loss": 0.0194,
      "num_input_tokens_seen": 33314048,
      "step": 37270
    },
    {
      "epoch": 9.837798601029432,
      "grad_norm": 0.22532346844673157,
      "learning_rate": 5.707972733505707e-07,
      "loss": 0.1178,
      "num_input_tokens_seen": 33318624,
      "step": 37275
    },
    {
      "epoch": 9.839118384584928,
      "grad_norm": 0.20967541635036469,
      "learning_rate": 5.687132611351509e-07,
      "loss": 0.0622,
      "num_input_tokens_seen": 33322912,
      "step": 37280
    },
    {
      "epoch": 9.840438168140425,
      "grad_norm": 0.08733822405338287,
      "learning_rate": 5.666330165311651e-07,
      "loss": 0.0306,
      "num_input_tokens_seen": 33327648,
      "step": 37285
    },
    {
      "epoch": 9.841757951695921,
      "grad_norm": 0.27452874183654785,
      "learning_rate": 5.645565398594204e-07,
      "loss": 0.0643,
      "num_input_tokens_seen": 33331872,
      "step": 37290
    },
    {
      "epoch": 9.843077735251418,
      "grad_norm": 0.22627541422843933,
      "learning_rate": 5.624838314401304e-07,
      "loss": 0.048,
      "num_input_tokens_seen": 33336352,
      "step": 37295
    },
    {
      "epoch": 9.844397518806916,
      "grad_norm": 0.03566306084394455,
      "learning_rate": 5.604148915929336e-07,
      "loss": 0.1266,
      "num_input_tokens_seen": 33340576,
      "step": 37300
    },
    {
      "epoch": 9.845717302362413,
      "grad_norm": 0.08761882781982422,
      "learning_rate": 5.583497206368887e-07,
      "loss": 0.0732,
      "num_input_tokens_seen": 33345088,
      "step": 37305
    },
    {
      "epoch": 9.84703708591791,
      "grad_norm": 0.48690420389175415,
      "learning_rate": 5.562883188904688e-07,
      "loss": 0.1253,
      "num_input_tokens_seen": 33349472,
      "step": 37310
    },
    {
      "epoch": 9.848356869473406,
      "grad_norm": 0.11547759175300598,
      "learning_rate": 5.542306866715724e-07,
      "loss": 0.0714,
      "num_input_tokens_seen": 33354048,
      "step": 37315
    },
    {
      "epoch": 9.849676653028903,
      "grad_norm": 0.06528135389089584,
      "learning_rate": 5.52176824297504e-07,
      "loss": 0.0574,
      "num_input_tokens_seen": 33358848,
      "step": 37320
    },
    {
      "epoch": 9.850996436584401,
      "grad_norm": 0.18517811596393585,
      "learning_rate": 5.501267320850018e-07,
      "loss": 0.0499,
      "num_input_tokens_seen": 33363296,
      "step": 37325
    },
    {
      "epoch": 9.852316220139897,
      "grad_norm": 0.48187902569770813,
      "learning_rate": 5.480804103502157e-07,
      "loss": 0.0585,
      "num_input_tokens_seen": 33367648,
      "step": 37330
    },
    {
      "epoch": 9.853636003695394,
      "grad_norm": 0.1787024289369583,
      "learning_rate": 5.460378594087101e-07,
      "loss": 0.0695,
      "num_input_tokens_seen": 33372224,
      "step": 37335
    },
    {
      "epoch": 9.85495578725089,
      "grad_norm": 0.23202332854270935,
      "learning_rate": 5.439990795754773e-07,
      "loss": 0.0426,
      "num_input_tokens_seen": 33376416,
      "step": 37340
    },
    {
      "epoch": 9.856275570806387,
      "grad_norm": 0.25234633684158325,
      "learning_rate": 5.419640711649188e-07,
      "loss": 0.0827,
      "num_input_tokens_seen": 33380800,
      "step": 37345
    },
    {
      "epoch": 9.857595354361884,
      "grad_norm": 0.2695445120334625,
      "learning_rate": 5.399328344908583e-07,
      "loss": 0.0567,
      "num_input_tokens_seen": 33385472,
      "step": 37350
    },
    {
      "epoch": 9.858915137917382,
      "grad_norm": 0.2858934700489044,
      "learning_rate": 5.379053698665399e-07,
      "loss": 0.0308,
      "num_input_tokens_seen": 33389856,
      "step": 37355
    },
    {
      "epoch": 9.860234921472879,
      "grad_norm": 0.33976802229881287,
      "learning_rate": 5.358816776046216e-07,
      "loss": 0.0952,
      "num_input_tokens_seen": 33394528,
      "step": 37360
    },
    {
      "epoch": 9.861554705028375,
      "grad_norm": 0.03156353160738945,
      "learning_rate": 5.338617580171817e-07,
      "loss": 0.0422,
      "num_input_tokens_seen": 33398816,
      "step": 37365
    },
    {
      "epoch": 9.862874488583872,
      "grad_norm": 0.19284354150295258,
      "learning_rate": 5.318456114157239e-07,
      "loss": 0.0682,
      "num_input_tokens_seen": 33403488,
      "step": 37370
    },
    {
      "epoch": 9.864194272139368,
      "grad_norm": 0.373263418674469,
      "learning_rate": 5.298332381111576e-07,
      "loss": 0.041,
      "num_input_tokens_seen": 33408000,
      "step": 37375
    },
    {
      "epoch": 9.865514055694867,
      "grad_norm": 0.168899267911911,
      "learning_rate": 5.27824638413818e-07,
      "loss": 0.0353,
      "num_input_tokens_seen": 33412864,
      "step": 37380
    },
    {
      "epoch": 9.866833839250363,
      "grad_norm": 0.11835721135139465,
      "learning_rate": 5.258198126334546e-07,
      "loss": 0.055,
      "num_input_tokens_seen": 33417120,
      "step": 37385
    },
    {
      "epoch": 9.86815362280586,
      "grad_norm": 0.3403611481189728,
      "learning_rate": 5.238187610792367e-07,
      "loss": 0.0791,
      "num_input_tokens_seen": 33421760,
      "step": 37390
    },
    {
      "epoch": 9.869473406361356,
      "grad_norm": 0.04644545167684555,
      "learning_rate": 5.218214840597563e-07,
      "loss": 0.0263,
      "num_input_tokens_seen": 33426272,
      "step": 37395
    },
    {
      "epoch": 9.870793189916853,
      "grad_norm": 0.10087738931179047,
      "learning_rate": 5.198279818830115e-07,
      "loss": 0.019,
      "num_input_tokens_seen": 33431008,
      "step": 37400
    },
    {
      "epoch": 9.870793189916853,
      "eval_loss": 0.06666062027215958,
      "eval_runtime": 64.7631,
      "eval_samples_per_second": 103.994,
      "eval_steps_per_second": 26.002,
      "num_input_tokens_seen": 33431008,
      "step": 37400
    },
    {
      "epoch": 9.872112973472351,
      "grad_norm": 0.2766406834125519,
      "learning_rate": 5.178382548564287e-07,
      "loss": 0.0729,
      "num_input_tokens_seen": 33435520,
      "step": 37405
    },
    {
      "epoch": 9.873432757027848,
      "grad_norm": 0.4182945191860199,
      "learning_rate": 5.15852303286854e-07,
      "loss": 0.1098,
      "num_input_tokens_seen": 33439808,
      "step": 37410
    },
    {
      "epoch": 9.874752540583344,
      "grad_norm": 0.10201513022184372,
      "learning_rate": 5.138701274805396e-07,
      "loss": 0.0644,
      "num_input_tokens_seen": 33444128,
      "step": 37415
    },
    {
      "epoch": 9.87607232413884,
      "grad_norm": 0.2746305465698242,
      "learning_rate": 5.118917277431606e-07,
      "loss": 0.0358,
      "num_input_tokens_seen": 33448800,
      "step": 37420
    },
    {
      "epoch": 9.877392107694337,
      "grad_norm": 0.5421135425567627,
      "learning_rate": 5.099171043798145e-07,
      "loss": 0.0769,
      "num_input_tokens_seen": 33452960,
      "step": 37425
    },
    {
      "epoch": 9.878711891249836,
      "grad_norm": 0.17085377871990204,
      "learning_rate": 5.079462576950133e-07,
      "loss": 0.0386,
      "num_input_tokens_seen": 33457728,
      "step": 37430
    },
    {
      "epoch": 9.880031674805332,
      "grad_norm": 0.19559040665626526,
      "learning_rate": 5.059791879926862e-07,
      "loss": 0.0328,
      "num_input_tokens_seen": 33461984,
      "step": 37435
    },
    {
      "epoch": 9.881351458360829,
      "grad_norm": 0.07823667675256729,
      "learning_rate": 5.040158955761793e-07,
      "loss": 0.0812,
      "num_input_tokens_seen": 33466272,
      "step": 37440
    },
    {
      "epoch": 9.882671241916325,
      "grad_norm": 0.15807582437992096,
      "learning_rate": 5.020563807482559e-07,
      "loss": 0.0417,
      "num_input_tokens_seen": 33470912,
      "step": 37445
    },
    {
      "epoch": 9.883991025471822,
      "grad_norm": 0.1242532879114151,
      "learning_rate": 5.001006438110995e-07,
      "loss": 0.0218,
      "num_input_tokens_seen": 33475072,
      "step": 37450
    },
    {
      "epoch": 9.88531080902732,
      "grad_norm": 0.06011911854147911,
      "learning_rate": 4.981486850663075e-07,
      "loss": 0.0574,
      "num_input_tokens_seen": 33480064,
      "step": 37455
    },
    {
      "epoch": 9.886630592582817,
      "grad_norm": 0.21927376091480255,
      "learning_rate": 4.962005048149005e-07,
      "loss": 0.0295,
      "num_input_tokens_seen": 33484608,
      "step": 37460
    },
    {
      "epoch": 9.887950376138313,
      "grad_norm": 0.08316655457019806,
      "learning_rate": 4.942561033573073e-07,
      "loss": 0.0764,
      "num_input_tokens_seen": 33489408,
      "step": 37465
    },
    {
      "epoch": 9.88927015969381,
      "grad_norm": 0.21824263036251068,
      "learning_rate": 4.923154809933827e-07,
      "loss": 0.0311,
      "num_input_tokens_seen": 33494272,
      "step": 37470
    },
    {
      "epoch": 9.890589943249307,
      "grad_norm": 0.13372664153575897,
      "learning_rate": 4.903786380223957e-07,
      "loss": 0.0595,
      "num_input_tokens_seen": 33498720,
      "step": 37475
    },
    {
      "epoch": 9.891909726804805,
      "grad_norm": 0.12491300702095032,
      "learning_rate": 4.884455747430266e-07,
      "loss": 0.0409,
      "num_input_tokens_seen": 33503552,
      "step": 37480
    },
    {
      "epoch": 9.893229510360301,
      "grad_norm": 0.03278215602040291,
      "learning_rate": 4.865162914533816e-07,
      "loss": 0.0318,
      "num_input_tokens_seen": 33508032,
      "step": 37485
    },
    {
      "epoch": 9.894549293915798,
      "grad_norm": 0.1143200695514679,
      "learning_rate": 4.845907884509809e-07,
      "loss": 0.0284,
      "num_input_tokens_seen": 33512224,
      "step": 37490
    },
    {
      "epoch": 9.895869077471295,
      "grad_norm": 0.048936448991298676,
      "learning_rate": 4.82669066032762e-07,
      "loss": 0.0637,
      "num_input_tokens_seen": 33516896,
      "step": 37495
    },
    {
      "epoch": 9.897188861026791,
      "grad_norm": 0.17584718763828278,
      "learning_rate": 4.807511244950768e-07,
      "loss": 0.0489,
      "num_input_tokens_seen": 33521088,
      "step": 37500
    },
    {
      "epoch": 9.89850864458229,
      "grad_norm": 0.04286818578839302,
      "learning_rate": 4.788369641336943e-07,
      "loss": 0.0499,
      "num_input_tokens_seen": 33525664,
      "step": 37505
    },
    {
      "epoch": 9.899828428137786,
      "grad_norm": 0.17666275799274445,
      "learning_rate": 4.769265852438032e-07,
      "loss": 0.0461,
      "num_input_tokens_seen": 33530112,
      "step": 37510
    },
    {
      "epoch": 9.901148211693283,
      "grad_norm": 0.22586558759212494,
      "learning_rate": 4.750199881200124e-07,
      "loss": 0.0759,
      "num_input_tokens_seen": 33534496,
      "step": 37515
    },
    {
      "epoch": 9.902467995248779,
      "grad_norm": 0.340863436460495,
      "learning_rate": 4.7311717305633664e-07,
      "loss": 0.0652,
      "num_input_tokens_seen": 33539072,
      "step": 37520
    },
    {
      "epoch": 9.903787778804276,
      "grad_norm": 0.14083009958267212,
      "learning_rate": 4.7121814034621623e-07,
      "loss": 0.0815,
      "num_input_tokens_seen": 33543616,
      "step": 37525
    },
    {
      "epoch": 9.905107562359772,
      "grad_norm": 0.2481112778186798,
      "learning_rate": 4.693228902825114e-07,
      "loss": 0.0774,
      "num_input_tokens_seen": 33547936,
      "step": 37530
    },
    {
      "epoch": 9.90642734591527,
      "grad_norm": 0.22487303614616394,
      "learning_rate": 4.6743142315748277e-07,
      "loss": 0.0308,
      "num_input_tokens_seen": 33552224,
      "step": 37535
    },
    {
      "epoch": 9.907747129470767,
      "grad_norm": 0.05817550793290138,
      "learning_rate": 4.655437392628276e-07,
      "loss": 0.0632,
      "num_input_tokens_seen": 33556704,
      "step": 37540
    },
    {
      "epoch": 9.909066913026264,
      "grad_norm": 0.12761804461479187,
      "learning_rate": 4.636598388896463e-07,
      "loss": 0.0501,
      "num_input_tokens_seen": 33561632,
      "step": 37545
    },
    {
      "epoch": 9.91038669658176,
      "grad_norm": 0.09301698952913284,
      "learning_rate": 4.6177972232845925e-07,
      "loss": 0.0534,
      "num_input_tokens_seen": 33566208,
      "step": 37550
    },
    {
      "epoch": 9.911706480137257,
      "grad_norm": 0.28809693455696106,
      "learning_rate": 4.5990338986920953e-07,
      "loss": 0.0973,
      "num_input_tokens_seen": 33570880,
      "step": 37555
    },
    {
      "epoch": 9.913026263692755,
      "grad_norm": 0.2248685657978058,
      "learning_rate": 4.5803084180124633e-07,
      "loss": 0.0605,
      "num_input_tokens_seen": 33575936,
      "step": 37560
    },
    {
      "epoch": 9.914346047248252,
      "grad_norm": 0.15298062562942505,
      "learning_rate": 4.561620784133386e-07,
      "loss": 0.0503,
      "num_input_tokens_seen": 33580064,
      "step": 37565
    },
    {
      "epoch": 9.915665830803748,
      "grad_norm": 0.4139944016933441,
      "learning_rate": 4.5429709999367796e-07,
      "loss": 0.0711,
      "num_input_tokens_seen": 33584480,
      "step": 37570
    },
    {
      "epoch": 9.916985614359245,
      "grad_norm": 0.13759741187095642,
      "learning_rate": 4.5243590682986223e-07,
      "loss": 0.0558,
      "num_input_tokens_seen": 33588640,
      "step": 37575
    },
    {
      "epoch": 9.918305397914741,
      "grad_norm": 0.09520643204450607,
      "learning_rate": 4.5057849920891735e-07,
      "loss": 0.0644,
      "num_input_tokens_seen": 33593344,
      "step": 37580
    },
    {
      "epoch": 9.91962518147024,
      "grad_norm": 0.14969949424266815,
      "learning_rate": 4.487248774172698e-07,
      "loss": 0.0456,
      "num_input_tokens_seen": 33597600,
      "step": 37585
    },
    {
      "epoch": 9.920944965025736,
      "grad_norm": 0.2798534631729126,
      "learning_rate": 4.4687504174077965e-07,
      "loss": 0.0523,
      "num_input_tokens_seen": 33601920,
      "step": 37590
    },
    {
      "epoch": 9.922264748581233,
      "grad_norm": 0.12155798822641373,
      "learning_rate": 4.450289924647133e-07,
      "loss": 0.0552,
      "num_input_tokens_seen": 33606368,
      "step": 37595
    },
    {
      "epoch": 9.92358453213673,
      "grad_norm": 0.33616411685943604,
      "learning_rate": 4.431867298737513e-07,
      "loss": 0.0513,
      "num_input_tokens_seen": 33610816,
      "step": 37600
    },
    {
      "epoch": 9.92358453213673,
      "eval_loss": 0.06668636202812195,
      "eval_runtime": 64.7816,
      "eval_samples_per_second": 103.965,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 33610816,
      "step": 37600
    },
    {
      "epoch": 9.924904315692226,
      "grad_norm": 0.17602252960205078,
      "learning_rate": 4.41348254251997e-07,
      "loss": 0.0626,
      "num_input_tokens_seen": 33615104,
      "step": 37605
    },
    {
      "epoch": 9.926224099247724,
      "grad_norm": 0.03280019015073776,
      "learning_rate": 4.395135658829652e-07,
      "loss": 0.0649,
      "num_input_tokens_seen": 33619520,
      "step": 37610
    },
    {
      "epoch": 9.92754388280322,
      "grad_norm": 0.306650847196579,
      "learning_rate": 4.376826650495852e-07,
      "loss": 0.0497,
      "num_input_tokens_seen": 33624224,
      "step": 37615
    },
    {
      "epoch": 9.928863666358717,
      "grad_norm": 0.18307961523532867,
      "learning_rate": 4.358555520342117e-07,
      "loss": 0.0397,
      "num_input_tokens_seen": 33628512,
      "step": 37620
    },
    {
      "epoch": 9.930183449914214,
      "grad_norm": 0.026564642786979675,
      "learning_rate": 4.3403222711860257e-07,
      "loss": 0.0429,
      "num_input_tokens_seen": 33632896,
      "step": 37625
    },
    {
      "epoch": 9.93150323346971,
      "grad_norm": 0.159596249461174,
      "learning_rate": 4.3221269058394133e-07,
      "loss": 0.094,
      "num_input_tokens_seen": 33637312,
      "step": 37630
    },
    {
      "epoch": 9.932823017025207,
      "grad_norm": 0.17609955370426178,
      "learning_rate": 4.303969427108173e-07,
      "loss": 0.013,
      "num_input_tokens_seen": 33641824,
      "step": 37635
    },
    {
      "epoch": 9.934142800580705,
      "grad_norm": 0.3578086495399475,
      "learning_rate": 4.2858498377924825e-07,
      "loss": 0.1042,
      "num_input_tokens_seen": 33646528,
      "step": 37640
    },
    {
      "epoch": 9.935462584136202,
      "grad_norm": 0.2368783801794052,
      "learning_rate": 4.267768140686579e-07,
      "loss": 0.0533,
      "num_input_tokens_seen": 33650880,
      "step": 37645
    },
    {
      "epoch": 9.936782367691698,
      "grad_norm": 0.3872343897819519,
      "learning_rate": 4.2497243385788975e-07,
      "loss": 0.1125,
      "num_input_tokens_seen": 33655360,
      "step": 37650
    },
    {
      "epoch": 9.938102151247195,
      "grad_norm": 0.2644798159599304,
      "learning_rate": 4.231718434251991e-07,
      "loss": 0.0535,
      "num_input_tokens_seen": 33659840,
      "step": 37655
    },
    {
      "epoch": 9.939421934802692,
      "grad_norm": 0.07587499916553497,
      "learning_rate": 4.213750430482666e-07,
      "loss": 0.0241,
      "num_input_tokens_seen": 33664352,
      "step": 37660
    },
    {
      "epoch": 9.94074171835819,
      "grad_norm": 0.053557366132736206,
      "learning_rate": 4.1958203300417054e-07,
      "loss": 0.051,
      "num_input_tokens_seen": 33668864,
      "step": 37665
    },
    {
      "epoch": 9.942061501913686,
      "grad_norm": 0.14698892831802368,
      "learning_rate": 4.177928135694259e-07,
      "loss": 0.0404,
      "num_input_tokens_seen": 33673312,
      "step": 37670
    },
    {
      "epoch": 9.943381285469183,
      "grad_norm": 0.05708412453532219,
      "learning_rate": 4.1600738501994807e-07,
      "loss": 0.0533,
      "num_input_tokens_seen": 33677728,
      "step": 37675
    },
    {
      "epoch": 9.94470106902468,
      "grad_norm": 0.36210569739341736,
      "learning_rate": 4.1422574763107237e-07,
      "loss": 0.0999,
      "num_input_tokens_seen": 33682528,
      "step": 37680
    },
    {
      "epoch": 9.946020852580176,
      "grad_norm": 0.14235402643680573,
      "learning_rate": 4.124479016775512e-07,
      "loss": 0.0557,
      "num_input_tokens_seen": 33687296,
      "step": 37685
    },
    {
      "epoch": 9.947340636135674,
      "grad_norm": 0.19021719694137573,
      "learning_rate": 4.106738474335514e-07,
      "loss": 0.0711,
      "num_input_tokens_seen": 33691584,
      "step": 37690
    },
    {
      "epoch": 9.948660419691171,
      "grad_norm": 0.15559323132038116,
      "learning_rate": 4.089035851726486e-07,
      "loss": 0.0451,
      "num_input_tokens_seen": 33695808,
      "step": 37695
    },
    {
      "epoch": 9.949980203246668,
      "grad_norm": 0.11610660701990128,
      "learning_rate": 4.0713711516784937e-07,
      "loss": 0.0555,
      "num_input_tokens_seen": 33700352,
      "step": 37700
    },
    {
      "epoch": 9.951299986802164,
      "grad_norm": 0.19101186096668243,
      "learning_rate": 4.05374437691558e-07,
      "loss": 0.0493,
      "num_input_tokens_seen": 33704704,
      "step": 37705
    },
    {
      "epoch": 9.95261977035766,
      "grad_norm": 0.10566519945859909,
      "learning_rate": 4.036155530156044e-07,
      "loss": 0.0219,
      "num_input_tokens_seen": 33709376,
      "step": 37710
    },
    {
      "epoch": 9.953939553913159,
      "grad_norm": 0.029912447556853294,
      "learning_rate": 4.018604614112298e-07,
      "loss": 0.1215,
      "num_input_tokens_seen": 33713888,
      "step": 37715
    },
    {
      "epoch": 9.955259337468656,
      "grad_norm": 0.24787446856498718,
      "learning_rate": 4.0010916314908996e-07,
      "loss": 0.0858,
      "num_input_tokens_seen": 33718304,
      "step": 37720
    },
    {
      "epoch": 9.956579121024152,
      "grad_norm": 0.14572301506996155,
      "learning_rate": 3.983616584992578e-07,
      "loss": 0.0427,
      "num_input_tokens_seen": 33722912,
      "step": 37725
    },
    {
      "epoch": 9.957898904579649,
      "grad_norm": 0.20772379636764526,
      "learning_rate": 3.9661794773122595e-07,
      "loss": 0.051,
      "num_input_tokens_seen": 33727520,
      "step": 37730
    },
    {
      "epoch": 9.959218688135145,
      "grad_norm": 0.3043055534362793,
      "learning_rate": 3.9487803111388777e-07,
      "loss": 0.0748,
      "num_input_tokens_seen": 33731904,
      "step": 37735
    },
    {
      "epoch": 9.960538471690644,
      "grad_norm": 0.1559484601020813,
      "learning_rate": 3.9314190891556747e-07,
      "loss": 0.035,
      "num_input_tokens_seen": 33736928,
      "step": 37740
    },
    {
      "epoch": 9.96185825524614,
      "grad_norm": 0.06001834571361542,
      "learning_rate": 3.914095814039925e-07,
      "loss": 0.0274,
      "num_input_tokens_seen": 33741408,
      "step": 37745
    },
    {
      "epoch": 9.963178038801637,
      "grad_norm": 0.34496524930000305,
      "learning_rate": 3.896810488463104e-07,
      "loss": 0.0329,
      "num_input_tokens_seen": 33746144,
      "step": 37750
    },
    {
      "epoch": 9.964497822357133,
      "grad_norm": 0.05874822288751602,
      "learning_rate": 3.8795631150908565e-07,
      "loss": 0.0372,
      "num_input_tokens_seen": 33750688,
      "step": 37755
    },
    {
      "epoch": 9.96581760591263,
      "grad_norm": 0.34706056118011475,
      "learning_rate": 3.862353696582888e-07,
      "loss": 0.0399,
      "num_input_tokens_seen": 33754592,
      "step": 37760
    },
    {
      "epoch": 9.967137389468128,
      "grad_norm": 0.2014765590429306,
      "learning_rate": 3.8451822355931313e-07,
      "loss": 0.1041,
      "num_input_tokens_seen": 33759104,
      "step": 37765
    },
    {
      "epoch": 9.968457173023625,
      "grad_norm": 0.4670551121234894,
      "learning_rate": 3.82804873476969e-07,
      "loss": 0.0533,
      "num_input_tokens_seen": 33763680,
      "step": 37770
    },
    {
      "epoch": 9.969776956579121,
      "grad_norm": 0.11517295241355896,
      "learning_rate": 3.810953196754702e-07,
      "loss": 0.0653,
      "num_input_tokens_seen": 33768352,
      "step": 37775
    },
    {
      "epoch": 9.971096740134618,
      "grad_norm": 0.07209106534719467,
      "learning_rate": 3.793895624184529e-07,
      "loss": 0.0386,
      "num_input_tokens_seen": 33772576,
      "step": 37780
    },
    {
      "epoch": 9.972416523690114,
      "grad_norm": 0.11723534017801285,
      "learning_rate": 3.776876019689679e-07,
      "loss": 0.1142,
      "num_input_tokens_seen": 33777184,
      "step": 37785
    },
    {
      "epoch": 9.97373630724561,
      "grad_norm": 0.06859634071588516,
      "learning_rate": 3.7598943858947743e-07,
      "loss": 0.0363,
      "num_input_tokens_seen": 33781696,
      "step": 37790
    },
    {
      "epoch": 9.97505609080111,
      "grad_norm": 0.20572331547737122,
      "learning_rate": 3.742950725418637e-07,
      "loss": 0.0534,
      "num_input_tokens_seen": 33786176,
      "step": 37795
    },
    {
      "epoch": 9.976375874356606,
      "grad_norm": 0.5612384080886841,
      "learning_rate": 3.726045040874093e-07,
      "loss": 0.0667,
      "num_input_tokens_seen": 33790528,
      "step": 37800
    },
    {
      "epoch": 9.976375874356606,
      "eval_loss": 0.06681814044713974,
      "eval_runtime": 64.7803,
      "eval_samples_per_second": 103.967,
      "eval_steps_per_second": 25.996,
      "num_input_tokens_seen": 33790528,
      "step": 37800
    },
    {
      "epoch": 9.977695657912102,
      "grad_norm": 0.33468860387802124,
      "learning_rate": 3.709177334868308e-07,
      "loss": 0.0872,
      "num_input_tokens_seen": 33794752,
      "step": 37805
    },
    {
      "epoch": 9.979015441467599,
      "grad_norm": 0.22791466116905212,
      "learning_rate": 3.692347610002478e-07,
      "loss": 0.0705,
      "num_input_tokens_seen": 33799264,
      "step": 37810
    },
    {
      "epoch": 9.980335225023095,
      "grad_norm": 0.0702073723077774,
      "learning_rate": 3.675555868871916e-07,
      "loss": 0.03,
      "num_input_tokens_seen": 33803776,
      "step": 37815
    },
    {
      "epoch": 9.981655008578594,
      "grad_norm": 0.0674089640378952,
      "learning_rate": 3.658802114066162e-07,
      "loss": 0.039,
      "num_input_tokens_seen": 33808288,
      "step": 37820
    },
    {
      "epoch": 9.98297479213409,
      "grad_norm": 0.2751138210296631,
      "learning_rate": 3.6420863481688437e-07,
      "loss": 0.0402,
      "num_input_tokens_seen": 33812832,
      "step": 37825
    },
    {
      "epoch": 9.984294575689587,
      "grad_norm": 0.11854992806911469,
      "learning_rate": 3.625408573757705e-07,
      "loss": 0.061,
      "num_input_tokens_seen": 33817120,
      "step": 37830
    },
    {
      "epoch": 9.985614359245083,
      "grad_norm": 0.07495156675577164,
      "learning_rate": 3.608768793404743e-07,
      "loss": 0.0474,
      "num_input_tokens_seen": 33821504,
      "step": 37835
    },
    {
      "epoch": 9.98693414280058,
      "grad_norm": 0.2974012494087219,
      "learning_rate": 3.592167009675934e-07,
      "loss": 0.0942,
      "num_input_tokens_seen": 33825664,
      "step": 37840
    },
    {
      "epoch": 9.988253926356078,
      "grad_norm": 0.3808560073375702,
      "learning_rate": 3.575603225131563e-07,
      "loss": 0.0662,
      "num_input_tokens_seen": 33830048,
      "step": 37845
    },
    {
      "epoch": 9.989573709911575,
      "grad_norm": 0.1966271549463272,
      "learning_rate": 3.55907744232592e-07,
      "loss": 0.0554,
      "num_input_tokens_seen": 33834624,
      "step": 37850
    },
    {
      "epoch": 9.990893493467071,
      "grad_norm": 0.09760642051696777,
      "learning_rate": 3.5425896638075217e-07,
      "loss": 0.0304,
      "num_input_tokens_seen": 33838816,
      "step": 37855
    },
    {
      "epoch": 9.992213277022568,
      "grad_norm": 0.15175718069076538,
      "learning_rate": 3.5261398921189736e-07,
      "loss": 0.0711,
      "num_input_tokens_seen": 33843328,
      "step": 37860
    },
    {
      "epoch": 9.993533060578065,
      "grad_norm": 0.3343449831008911,
      "learning_rate": 3.509728129797024e-07,
      "loss": 0.0624,
      "num_input_tokens_seen": 33847680,
      "step": 37865
    },
    {
      "epoch": 9.994852844133563,
      "grad_norm": 0.0773860514163971,
      "learning_rate": 3.4933543793725656e-07,
      "loss": 0.0792,
      "num_input_tokens_seen": 33851968,
      "step": 37870
    },
    {
      "epoch": 9.99617262768906,
      "grad_norm": 0.2404259592294693,
      "learning_rate": 3.4770186433707163e-07,
      "loss": 0.0504,
      "num_input_tokens_seen": 33856576,
      "step": 37875
    },
    {
      "epoch": 9.997492411244556,
      "grad_norm": 0.194298654794693,
      "learning_rate": 3.4607209243105453e-07,
      "loss": 0.0362,
      "num_input_tokens_seen": 33861120,
      "step": 37880
    },
    {
      "epoch": 9.998812194800053,
      "grad_norm": 0.10631931573152542,
      "learning_rate": 3.444461224705431e-07,
      "loss": 0.0388,
      "num_input_tokens_seen": 33865824,
      "step": 37885
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.12102352827787399,
      "learning_rate": 3.4282395470628116e-07,
      "loss": 0.0448,
      "num_input_tokens_seen": 33869824,
      "step": 37890
    },
    {
      "epoch": 10.001319783555497,
      "grad_norm": 0.3837428390979767,
      "learning_rate": 3.4120558938842417e-07,
      "loss": 0.0978,
      "num_input_tokens_seen": 33873856,
      "step": 37895
    },
    {
      "epoch": 10.002639567110993,
      "grad_norm": 0.12204327434301376,
      "learning_rate": 3.395910267665503e-07,
      "loss": 0.044,
      "num_input_tokens_seen": 33878528,
      "step": 37900
    },
    {
      "epoch": 10.003959350666491,
      "grad_norm": 0.19019757211208344,
      "learning_rate": 3.3798026708964094e-07,
      "loss": 0.0866,
      "num_input_tokens_seen": 33882944,
      "step": 37905
    },
    {
      "epoch": 10.005279134221988,
      "grad_norm": 0.15790686011314392,
      "learning_rate": 3.3637331060609456e-07,
      "loss": 0.0405,
      "num_input_tokens_seen": 33887360,
      "step": 37910
    },
    {
      "epoch": 10.006598917777485,
      "grad_norm": 0.03515373170375824,
      "learning_rate": 3.3477015756372966e-07,
      "loss": 0.0534,
      "num_input_tokens_seen": 33891840,
      "step": 37915
    },
    {
      "epoch": 10.007918701332981,
      "grad_norm": 0.08442790806293488,
      "learning_rate": 3.3317080820976785e-07,
      "loss": 0.0231,
      "num_input_tokens_seen": 33896384,
      "step": 37920
    },
    {
      "epoch": 10.009238484888478,
      "grad_norm": 0.1246090680360794,
      "learning_rate": 3.315752627908508e-07,
      "loss": 0.0657,
      "num_input_tokens_seen": 33900992,
      "step": 37925
    },
    {
      "epoch": 10.010558268443976,
      "grad_norm": 0.13627082109451294,
      "learning_rate": 3.299835215530317e-07,
      "loss": 0.0228,
      "num_input_tokens_seen": 33905152,
      "step": 37930
    },
    {
      "epoch": 10.011878051999473,
      "grad_norm": 0.4496801197528839,
      "learning_rate": 3.2839558474177245e-07,
      "loss": 0.0903,
      "num_input_tokens_seen": 33909600,
      "step": 37935
    },
    {
      "epoch": 10.01319783555497,
      "grad_norm": 0.08077823370695114,
      "learning_rate": 3.2681145260196056e-07,
      "loss": 0.0537,
      "num_input_tokens_seen": 33913792,
      "step": 37940
    },
    {
      "epoch": 10.014517619110466,
      "grad_norm": 0.3165336847305298,
      "learning_rate": 3.252311253778839e-07,
      "loss": 0.0262,
      "num_input_tokens_seen": 33918336,
      "step": 37945
    },
    {
      "epoch": 10.015837402665962,
      "grad_norm": 0.28487715125083923,
      "learning_rate": 3.2365460331325034e-07,
      "loss": 0.0646,
      "num_input_tokens_seen": 33922528,
      "step": 37950
    },
    {
      "epoch": 10.01715718622146,
      "grad_norm": 0.2868373990058899,
      "learning_rate": 3.2208188665117934e-07,
      "loss": 0.0934,
      "num_input_tokens_seen": 33927168,
      "step": 37955
    },
    {
      "epoch": 10.018476969776957,
      "grad_norm": 0.21680396795272827,
      "learning_rate": 3.205129756342018e-07,
      "loss": 0.0518,
      "num_input_tokens_seen": 33931744,
      "step": 37960
    },
    {
      "epoch": 10.019796753332454,
      "grad_norm": 0.11099917441606522,
      "learning_rate": 3.189478705042659e-07,
      "loss": 0.0818,
      "num_input_tokens_seen": 33936256,
      "step": 37965
    },
    {
      "epoch": 10.02111653688795,
      "grad_norm": 0.222688689827919,
      "learning_rate": 3.173865715027341e-07,
      "loss": 0.0909,
      "num_input_tokens_seen": 33940576,
      "step": 37970
    },
    {
      "epoch": 10.022436320443447,
      "grad_norm": 0.053455181419849396,
      "learning_rate": 3.158290788703694e-07,
      "loss": 0.0476,
      "num_input_tokens_seen": 33944896,
      "step": 37975
    },
    {
      "epoch": 10.023756103998943,
      "grad_norm": 0.0686720609664917,
      "learning_rate": 3.1427539284736297e-07,
      "loss": 0.042,
      "num_input_tokens_seen": 33949472,
      "step": 37980
    },
    {
      "epoch": 10.025075887554442,
      "grad_norm": 0.376705139875412,
      "learning_rate": 3.127255136733093e-07,
      "loss": 0.0444,
      "num_input_tokens_seen": 33954080,
      "step": 37985
    },
    {
      "epoch": 10.026395671109938,
      "grad_norm": 0.05477975681424141,
      "learning_rate": 3.1117944158722544e-07,
      "loss": 0.1289,
      "num_input_tokens_seen": 33958336,
      "step": 37990
    },
    {
      "epoch": 10.027715454665435,
      "grad_norm": 0.052923668175935745,
      "learning_rate": 3.0963717682752635e-07,
      "loss": 0.0183,
      "num_input_tokens_seen": 33962976,
      "step": 37995
    },
    {
      "epoch": 10.029035238220931,
      "grad_norm": 0.26281264424324036,
      "learning_rate": 3.080987196320578e-07,
      "loss": 0.0793,
      "num_input_tokens_seen": 33967520,
      "step": 38000
    },
    {
      "epoch": 10.029035238220931,
      "eval_loss": 0.06681524962186813,
      "eval_runtime": 64.731,
      "eval_samples_per_second": 104.046,
      "eval_steps_per_second": 26.015,
      "num_input_tokens_seen": 33967520,
      "step": 38000
    },
    {
      "epoch": 10.030355021776428,
      "grad_norm": 0.08204266428947449,
      "learning_rate": 3.065640702380607e-07,
      "loss": 0.089,
      "num_input_tokens_seen": 33971872,
      "step": 38005
    },
    {
      "epoch": 10.031674805331926,
      "grad_norm": 0.30541324615478516,
      "learning_rate": 3.050332288822011e-07,
      "loss": 0.0634,
      "num_input_tokens_seen": 33976320,
      "step": 38010
    },
    {
      "epoch": 10.032994588887423,
      "grad_norm": 0.36135199666023254,
      "learning_rate": 3.035061958005542e-07,
      "loss": 0.0542,
      "num_input_tokens_seen": 33980672,
      "step": 38015
    },
    {
      "epoch": 10.03431437244292,
      "grad_norm": 0.22293494641780853,
      "learning_rate": 3.019829712286093e-07,
      "loss": 0.1233,
      "num_input_tokens_seen": 33985152,
      "step": 38020
    },
    {
      "epoch": 10.035634155998416,
      "grad_norm": 0.11124400049448013,
      "learning_rate": 3.004635554012647e-07,
      "loss": 0.0371,
      "num_input_tokens_seen": 33989536,
      "step": 38025
    },
    {
      "epoch": 10.036953939553912,
      "grad_norm": 0.11182738095521927,
      "learning_rate": 2.9894794855283017e-07,
      "loss": 0.0667,
      "num_input_tokens_seen": 33993856,
      "step": 38030
    },
    {
      "epoch": 10.03827372310941,
      "grad_norm": 0.02259814366698265,
      "learning_rate": 2.9743615091703816e-07,
      "loss": 0.0361,
      "num_input_tokens_seen": 33998336,
      "step": 38035
    },
    {
      "epoch": 10.039593506664907,
      "grad_norm": 0.05230380967259407,
      "learning_rate": 2.959281627270216e-07,
      "loss": 0.0983,
      "num_input_tokens_seen": 34003136,
      "step": 38040
    },
    {
      "epoch": 10.040913290220404,
      "grad_norm": 0.14785458147525787,
      "learning_rate": 2.944239842153362e-07,
      "loss": 0.0775,
      "num_input_tokens_seen": 34007616,
      "step": 38045
    },
    {
      "epoch": 10.0422330737759,
      "grad_norm": 0.06608415395021439,
      "learning_rate": 2.929236156139381e-07,
      "loss": 0.0348,
      "num_input_tokens_seen": 34012320,
      "step": 38050
    },
    {
      "epoch": 10.043552857331397,
      "grad_norm": 0.25637203454971313,
      "learning_rate": 2.9142705715420883e-07,
      "loss": 0.0391,
      "num_input_tokens_seen": 34016736,
      "step": 38055
    },
    {
      "epoch": 10.044872640886895,
      "grad_norm": 0.12935493886470795,
      "learning_rate": 2.8993430906693595e-07,
      "loss": 0.0995,
      "num_input_tokens_seen": 34021312,
      "step": 38060
    },
    {
      "epoch": 10.046192424442392,
      "grad_norm": 0.3180655837059021,
      "learning_rate": 2.88445371582316e-07,
      "loss": 0.0555,
      "num_input_tokens_seen": 34025792,
      "step": 38065
    },
    {
      "epoch": 10.047512207997888,
      "grad_norm": 0.17793750762939453,
      "learning_rate": 2.8696024492996796e-07,
      "loss": 0.0722,
      "num_input_tokens_seen": 34029920,
      "step": 38070
    },
    {
      "epoch": 10.048831991553385,
      "grad_norm": 0.14906224608421326,
      "learning_rate": 2.854789293389115e-07,
      "loss": 0.0334,
      "num_input_tokens_seen": 34034368,
      "step": 38075
    },
    {
      "epoch": 10.050151775108882,
      "grad_norm": 0.2949332594871521,
      "learning_rate": 2.8400142503758606e-07,
      "loss": 0.0561,
      "num_input_tokens_seen": 34038784,
      "step": 38080
    },
    {
      "epoch": 10.05147155866438,
      "grad_norm": 0.5170860886573792,
      "learning_rate": 2.8252773225384276e-07,
      "loss": 0.0939,
      "num_input_tokens_seen": 34043040,
      "step": 38085
    },
    {
      "epoch": 10.052791342219876,
      "grad_norm": 0.20082075893878937,
      "learning_rate": 2.8105785121494143e-07,
      "loss": 0.0579,
      "num_input_tokens_seen": 34047840,
      "step": 38090
    },
    {
      "epoch": 10.054111125775373,
      "grad_norm": 0.3481748104095459,
      "learning_rate": 2.795917821475563e-07,
      "loss": 0.0858,
      "num_input_tokens_seen": 34052480,
      "step": 38095
    },
    {
      "epoch": 10.05543090933087,
      "grad_norm": 0.2764076292514801,
      "learning_rate": 2.78129525277776e-07,
      "loss": 0.1093,
      "num_input_tokens_seen": 34056800,
      "step": 38100
    },
    {
      "epoch": 10.056750692886366,
      "grad_norm": 0.23656082153320312,
      "learning_rate": 2.766710808310952e-07,
      "loss": 0.0506,
      "num_input_tokens_seen": 34060896,
      "step": 38105
    },
    {
      "epoch": 10.058070476441863,
      "grad_norm": 0.28911080956459045,
      "learning_rate": 2.7521644903242827e-07,
      "loss": 0.0785,
      "num_input_tokens_seen": 34065184,
      "step": 38110
    },
    {
      "epoch": 10.059390259997361,
      "grad_norm": 0.03436349332332611,
      "learning_rate": 2.7376563010609593e-07,
      "loss": 0.1204,
      "num_input_tokens_seen": 34069536,
      "step": 38115
    },
    {
      "epoch": 10.060710043552858,
      "grad_norm": 0.2678040564060211,
      "learning_rate": 2.72318624275833e-07,
      "loss": 0.0605,
      "num_input_tokens_seen": 34074016,
      "step": 38120
    },
    {
      "epoch": 10.062029827108354,
      "grad_norm": 0.08466766774654388,
      "learning_rate": 2.7087543176478324e-07,
      "loss": 0.0767,
      "num_input_tokens_seen": 34078560,
      "step": 38125
    },
    {
      "epoch": 10.06334961066385,
      "grad_norm": 0.1350928246974945,
      "learning_rate": 2.694360527955103e-07,
      "loss": 0.0401,
      "num_input_tokens_seen": 34083168,
      "step": 38130
    },
    {
      "epoch": 10.064669394219347,
      "grad_norm": 0.22209849953651428,
      "learning_rate": 2.680004875899811e-07,
      "loss": 0.057,
      "num_input_tokens_seen": 34087296,
      "step": 38135
    },
    {
      "epoch": 10.065989177774846,
      "grad_norm": 0.3528316020965576,
      "learning_rate": 2.665687363695768e-07,
      "loss": 0.06,
      "num_input_tokens_seen": 34092032,
      "step": 38140
    },
    {
      "epoch": 10.067308961330342,
      "grad_norm": 0.23281915485858917,
      "learning_rate": 2.6514079935509584e-07,
      "loss": 0.0737,
      "num_input_tokens_seen": 34096608,
      "step": 38145
    },
    {
      "epoch": 10.068628744885839,
      "grad_norm": 0.5810974836349487,
      "learning_rate": 2.6371667676673983e-07,
      "loss": 0.0994,
      "num_input_tokens_seen": 34101440,
      "step": 38150
    },
    {
      "epoch": 10.069948528441335,
      "grad_norm": 0.02259673923254013,
      "learning_rate": 2.6229636882412755e-07,
      "loss": 0.0402,
      "num_input_tokens_seen": 34105792,
      "step": 38155
    },
    {
      "epoch": 10.071268311996832,
      "grad_norm": 0.24426713585853577,
      "learning_rate": 2.6087987574628935e-07,
      "loss": 0.0732,
      "num_input_tokens_seen": 34110112,
      "step": 38160
    },
    {
      "epoch": 10.07258809555233,
      "grad_norm": 0.14291857182979584,
      "learning_rate": 2.5946719775166437e-07,
      "loss": 0.0718,
      "num_input_tokens_seen": 34114560,
      "step": 38165
    },
    {
      "epoch": 10.073907879107827,
      "grad_norm": 0.31020388007164,
      "learning_rate": 2.5805833505810616e-07,
      "loss": 0.0842,
      "num_input_tokens_seen": 34119072,
      "step": 38170
    },
    {
      "epoch": 10.075227662663323,
      "grad_norm": 0.1958678811788559,
      "learning_rate": 2.566532878828798e-07,
      "loss": 0.0877,
      "num_input_tokens_seen": 34123552,
      "step": 38175
    },
    {
      "epoch": 10.07654744621882,
      "grad_norm": 0.19787365198135376,
      "learning_rate": 2.552520564426619e-07,
      "loss": 0.0434,
      "num_input_tokens_seen": 34127840,
      "step": 38180
    },
    {
      "epoch": 10.077867229774316,
      "grad_norm": 0.09934801608324051,
      "learning_rate": 2.5385464095353803e-07,
      "loss": 0.0294,
      "num_input_tokens_seen": 34132128,
      "step": 38185
    },
    {
      "epoch": 10.079187013329815,
      "grad_norm": 0.06508375704288483,
      "learning_rate": 2.5246104163100804e-07,
      "loss": 0.0673,
      "num_input_tokens_seen": 34136576,
      "step": 38190
    },
    {
      "epoch": 10.080506796885311,
      "grad_norm": 0.16275034844875336,
      "learning_rate": 2.510712586899833e-07,
      "loss": 0.0479,
      "num_input_tokens_seen": 34140960,
      "step": 38195
    },
    {
      "epoch": 10.081826580440808,
      "grad_norm": 0.15083876252174377,
      "learning_rate": 2.4968529234478124e-07,
      "loss": 0.0775,
      "num_input_tokens_seen": 34145280,
      "step": 38200
    },
    {
      "epoch": 10.081826580440808,
      "eval_loss": 0.06662769615650177,
      "eval_runtime": 64.7578,
      "eval_samples_per_second": 104.003,
      "eval_steps_per_second": 26.005,
      "num_input_tokens_seen": 34145280,
      "step": 38200
    },
    {
      "epoch": 10.083146363996304,
      "grad_norm": 0.35546180605888367,
      "learning_rate": 2.483031428091448e-07,
      "loss": 0.0511,
      "num_input_tokens_seen": 34149632,
      "step": 38205
    },
    {
      "epoch": 10.084466147551801,
      "grad_norm": 0.2029927521944046,
      "learning_rate": 2.469248102962091e-07,
      "loss": 0.0465,
      "num_input_tokens_seen": 34154528,
      "step": 38210
    },
    {
      "epoch": 10.0857859311073,
      "grad_norm": 0.16105994582176208,
      "learning_rate": 2.4555029501853455e-07,
      "loss": 0.0907,
      "num_input_tokens_seen": 34158752,
      "step": 38215
    },
    {
      "epoch": 10.087105714662796,
      "grad_norm": 0.04617764428257942,
      "learning_rate": 2.441795971880906e-07,
      "loss": 0.0571,
      "num_input_tokens_seen": 34163584,
      "step": 38220
    },
    {
      "epoch": 10.088425498218292,
      "grad_norm": 0.17850059270858765,
      "learning_rate": 2.4281271701625255e-07,
      "loss": 0.0541,
      "num_input_tokens_seen": 34168000,
      "step": 38225
    },
    {
      "epoch": 10.089745281773789,
      "grad_norm": 0.37063658237457275,
      "learning_rate": 2.4144965471381007e-07,
      "loss": 0.0735,
      "num_input_tokens_seen": 34172512,
      "step": 38230
    },
    {
      "epoch": 10.091065065329285,
      "grad_norm": 0.477384090423584,
      "learning_rate": 2.400904104909674e-07,
      "loss": 0.0717,
      "num_input_tokens_seen": 34177280,
      "step": 38235
    },
    {
      "epoch": 10.092384848884782,
      "grad_norm": 0.09601424634456635,
      "learning_rate": 2.3873498455733725e-07,
      "loss": 0.0263,
      "num_input_tokens_seen": 34181888,
      "step": 38240
    },
    {
      "epoch": 10.09370463244028,
      "grad_norm": 0.09946175664663315,
      "learning_rate": 2.3738337712194137e-07,
      "loss": 0.0233,
      "num_input_tokens_seen": 34186144,
      "step": 38245
    },
    {
      "epoch": 10.095024415995777,
      "grad_norm": 0.18595488369464874,
      "learning_rate": 2.3603558839321305e-07,
      "loss": 0.046,
      "num_input_tokens_seen": 34190688,
      "step": 38250
    },
    {
      "epoch": 10.096344199551273,
      "grad_norm": 0.0821559727191925,
      "learning_rate": 2.3469161857900267e-07,
      "loss": 0.0178,
      "num_input_tokens_seen": 34195200,
      "step": 38255
    },
    {
      "epoch": 10.09766398310677,
      "grad_norm": 0.1861674189567566,
      "learning_rate": 2.3335146788656393e-07,
      "loss": 0.0685,
      "num_input_tokens_seen": 34199840,
      "step": 38260
    },
    {
      "epoch": 10.098983766662267,
      "grad_norm": 0.0501311793923378,
      "learning_rate": 2.3201513652256757e-07,
      "loss": 0.0383,
      "num_input_tokens_seen": 34204064,
      "step": 38265
    },
    {
      "epoch": 10.100303550217765,
      "grad_norm": 0.12444113194942474,
      "learning_rate": 2.3068262469308766e-07,
      "loss": 0.0694,
      "num_input_tokens_seen": 34208256,
      "step": 38270
    },
    {
      "epoch": 10.101623333773261,
      "grad_norm": 0.09051330387592316,
      "learning_rate": 2.2935393260362093e-07,
      "loss": 0.0419,
      "num_input_tokens_seen": 34212448,
      "step": 38275
    },
    {
      "epoch": 10.102943117328758,
      "grad_norm": 0.17154602706432343,
      "learning_rate": 2.2802906045906458e-07,
      "loss": 0.0908,
      "num_input_tokens_seen": 34216800,
      "step": 38280
    },
    {
      "epoch": 10.104262900884255,
      "grad_norm": 0.03760075941681862,
      "learning_rate": 2.2670800846373018e-07,
      "loss": 0.0664,
      "num_input_tokens_seen": 34221216,
      "step": 38285
    },
    {
      "epoch": 10.105582684439751,
      "grad_norm": 0.23714080452919006,
      "learning_rate": 2.2539077682134367e-07,
      "loss": 0.0582,
      "num_input_tokens_seen": 34225888,
      "step": 38290
    },
    {
      "epoch": 10.10690246799525,
      "grad_norm": 0.4629157483577728,
      "learning_rate": 2.2407736573503423e-07,
      "loss": 0.0697,
      "num_input_tokens_seen": 34230496,
      "step": 38295
    },
    {
      "epoch": 10.108222251550746,
      "grad_norm": 0.048339519649744034,
      "learning_rate": 2.2276777540735093e-07,
      "loss": 0.0518,
      "num_input_tokens_seen": 34235008,
      "step": 38300
    },
    {
      "epoch": 10.109542035106243,
      "grad_norm": 0.08449076116085052,
      "learning_rate": 2.2146200604024613e-07,
      "loss": 0.0222,
      "num_input_tokens_seen": 34239680,
      "step": 38305
    },
    {
      "epoch": 10.11086181866174,
      "grad_norm": 0.3369671404361725,
      "learning_rate": 2.2016005783508375e-07,
      "loss": 0.1095,
      "num_input_tokens_seen": 34244096,
      "step": 38310
    },
    {
      "epoch": 10.112181602217236,
      "grad_norm": 0.03911479189991951,
      "learning_rate": 2.1886193099264763e-07,
      "loss": 0.0919,
      "num_input_tokens_seen": 34248384,
      "step": 38315
    },
    {
      "epoch": 10.113501385772734,
      "grad_norm": 0.239566832780838,
      "learning_rate": 2.175676257131165e-07,
      "loss": 0.0785,
      "num_input_tokens_seen": 34252512,
      "step": 38320
    },
    {
      "epoch": 10.11482116932823,
      "grad_norm": 0.10681136697530746,
      "learning_rate": 2.162771421960974e-07,
      "loss": 0.1113,
      "num_input_tokens_seen": 34257088,
      "step": 38325
    },
    {
      "epoch": 10.116140952883727,
      "grad_norm": 0.5920694470405579,
      "learning_rate": 2.1499048064059224e-07,
      "loss": 0.0837,
      "num_input_tokens_seen": 34261248,
      "step": 38330
    },
    {
      "epoch": 10.117460736439224,
      "grad_norm": 0.37838441133499146,
      "learning_rate": 2.1370764124502285e-07,
      "loss": 0.0439,
      "num_input_tokens_seen": 34265568,
      "step": 38335
    },
    {
      "epoch": 10.11878051999472,
      "grad_norm": 0.11948937922716141,
      "learning_rate": 2.1242862420721988e-07,
      "loss": 0.0379,
      "num_input_tokens_seen": 34269920,
      "step": 38340
    },
    {
      "epoch": 10.120100303550219,
      "grad_norm": 0.218428835272789,
      "learning_rate": 2.1115342972442276e-07,
      "loss": 0.0619,
      "num_input_tokens_seen": 34274432,
      "step": 38345
    },
    {
      "epoch": 10.121420087105715,
      "grad_norm": 0.17791852355003357,
      "learning_rate": 2.0988205799328252e-07,
      "loss": 0.0434,
      "num_input_tokens_seen": 34278688,
      "step": 38350
    },
    {
      "epoch": 10.122739870661212,
      "grad_norm": 0.16509293019771576,
      "learning_rate": 2.0861450920986182e-07,
      "loss": 0.0375,
      "num_input_tokens_seen": 34283200,
      "step": 38355
    },
    {
      "epoch": 10.124059654216708,
      "grad_norm": 0.1823548674583435,
      "learning_rate": 2.07350783569632e-07,
      "loss": 0.1131,
      "num_input_tokens_seen": 34288160,
      "step": 38360
    },
    {
      "epoch": 10.125379437772205,
      "grad_norm": 0.07168129831552505,
      "learning_rate": 2.060908812674761e-07,
      "loss": 0.0366,
      "num_input_tokens_seen": 34292640,
      "step": 38365
    },
    {
      "epoch": 10.126699221327701,
      "grad_norm": 0.26080265641212463,
      "learning_rate": 2.0483480249768317e-07,
      "loss": 0.0714,
      "num_input_tokens_seen": 34297088,
      "step": 38370
    },
    {
      "epoch": 10.1280190048832,
      "grad_norm": 0.4193274676799774,
      "learning_rate": 2.035825474539621e-07,
      "loss": 0.0781,
      "num_input_tokens_seen": 34301600,
      "step": 38375
    },
    {
      "epoch": 10.129338788438696,
      "grad_norm": 0.27161845564842224,
      "learning_rate": 2.0233411632942235e-07,
      "loss": 0.0241,
      "num_input_tokens_seen": 34306240,
      "step": 38380
    },
    {
      "epoch": 10.130658571994193,
      "grad_norm": 0.1561221033334732,
      "learning_rate": 2.0108950931658764e-07,
      "loss": 0.0325,
      "num_input_tokens_seen": 34310880,
      "step": 38385
    },
    {
      "epoch": 10.13197835554969,
      "grad_norm": 0.2053440660238266,
      "learning_rate": 1.998487266073934e-07,
      "loss": 0.0852,
      "num_input_tokens_seen": 34315392,
      "step": 38390
    },
    {
      "epoch": 10.133298139105186,
      "grad_norm": 0.17807281017303467,
      "learning_rate": 1.986117683931865e-07,
      "loss": 0.0448,
      "num_input_tokens_seen": 34319968,
      "step": 38395
    },
    {
      "epoch": 10.134617922660684,
      "grad_norm": 0.2973754405975342,
      "learning_rate": 1.9737863486471442e-07,
      "loss": 0.0978,
      "num_input_tokens_seen": 34324448,
      "step": 38400
    },
    {
      "epoch": 10.134617922660684,
      "eval_loss": 0.06676899641752243,
      "eval_runtime": 64.7742,
      "eval_samples_per_second": 103.977,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 34324448,
      "step": 38400
    },
    {
      "epoch": 10.13593770621618,
      "grad_norm": 0.17091618478298187,
      "learning_rate": 1.9614932621215e-07,
      "loss": 0.0356,
      "num_input_tokens_seen": 34329152,
      "step": 38405
    },
    {
      "epoch": 10.137257489771677,
      "grad_norm": 0.16999801993370056,
      "learning_rate": 1.9492384262506102e-07,
      "loss": 0.0732,
      "num_input_tokens_seen": 34333408,
      "step": 38410
    },
    {
      "epoch": 10.138577273327174,
      "grad_norm": 0.1775590479373932,
      "learning_rate": 1.9370218429243524e-07,
      "loss": 0.0502,
      "num_input_tokens_seen": 34337760,
      "step": 38415
    },
    {
      "epoch": 10.13989705688267,
      "grad_norm": 0.4389783442020416,
      "learning_rate": 1.9248435140267197e-07,
      "loss": 0.0614,
      "num_input_tokens_seen": 34342272,
      "step": 38420
    },
    {
      "epoch": 10.141216840438169,
      "grad_norm": 0.14308783411979675,
      "learning_rate": 1.9127034414356814e-07,
      "loss": 0.043,
      "num_input_tokens_seen": 34346464,
      "step": 38425
    },
    {
      "epoch": 10.142536623993665,
      "grad_norm": 0.3823065161705017,
      "learning_rate": 1.9006016270234627e-07,
      "loss": 0.0734,
      "num_input_tokens_seen": 34350880,
      "step": 38430
    },
    {
      "epoch": 10.143856407549162,
      "grad_norm": 0.18779703974723816,
      "learning_rate": 1.888538072656293e-07,
      "loss": 0.0375,
      "num_input_tokens_seen": 34355456,
      "step": 38435
    },
    {
      "epoch": 10.145176191104659,
      "grad_norm": 0.10997649282217026,
      "learning_rate": 1.8765127801944893e-07,
      "loss": 0.0297,
      "num_input_tokens_seen": 34359968,
      "step": 38440
    },
    {
      "epoch": 10.146495974660155,
      "grad_norm": 0.07946807891130447,
      "learning_rate": 1.8645257514925406e-07,
      "loss": 0.1066,
      "num_input_tokens_seen": 34364352,
      "step": 38445
    },
    {
      "epoch": 10.147815758215653,
      "grad_norm": 0.2856985628604889,
      "learning_rate": 1.8525769883989685e-07,
      "loss": 0.0673,
      "num_input_tokens_seen": 34368608,
      "step": 38450
    },
    {
      "epoch": 10.14913554177115,
      "grad_norm": 0.3269865810871124,
      "learning_rate": 1.8406664927564654e-07,
      "loss": 0.0866,
      "num_input_tokens_seen": 34373504,
      "step": 38455
    },
    {
      "epoch": 10.150455325326647,
      "grad_norm": 0.24784094095230103,
      "learning_rate": 1.8287942664017566e-07,
      "loss": 0.0598,
      "num_input_tokens_seen": 34378080,
      "step": 38460
    },
    {
      "epoch": 10.151775108882143,
      "grad_norm": 0.13810604810714722,
      "learning_rate": 1.8169603111656552e-07,
      "loss": 0.0553,
      "num_input_tokens_seen": 34382784,
      "step": 38465
    },
    {
      "epoch": 10.15309489243764,
      "grad_norm": 0.2855387032032013,
      "learning_rate": 1.805164628873146e-07,
      "loss": 0.059,
      "num_input_tokens_seen": 34387424,
      "step": 38470
    },
    {
      "epoch": 10.154414675993138,
      "grad_norm": 0.26491403579711914,
      "learning_rate": 1.793407221343274e-07,
      "loss": 0.0338,
      "num_input_tokens_seen": 34391840,
      "step": 38475
    },
    {
      "epoch": 10.155734459548635,
      "grad_norm": 0.4147125482559204,
      "learning_rate": 1.781688090389172e-07,
      "loss": 0.036,
      "num_input_tokens_seen": 34396320,
      "step": 38480
    },
    {
      "epoch": 10.157054243104131,
      "grad_norm": 0.5289273262023926,
      "learning_rate": 1.770007237818061e-07,
      "loss": 0.1146,
      "num_input_tokens_seen": 34400640,
      "step": 38485
    },
    {
      "epoch": 10.158374026659628,
      "grad_norm": 0.3145584166049957,
      "learning_rate": 1.7583646654313059e-07,
      "loss": 0.0827,
      "num_input_tokens_seen": 34405280,
      "step": 38490
    },
    {
      "epoch": 10.159693810215124,
      "grad_norm": 0.111534982919693,
      "learning_rate": 1.7467603750242757e-07,
      "loss": 0.0547,
      "num_input_tokens_seen": 34409696,
      "step": 38495
    },
    {
      "epoch": 10.16101359377062,
      "grad_norm": 0.3558240532875061,
      "learning_rate": 1.7351943683865944e-07,
      "loss": 0.0423,
      "num_input_tokens_seen": 34414496,
      "step": 38500
    },
    {
      "epoch": 10.162333377326119,
      "grad_norm": 0.45740658044815063,
      "learning_rate": 1.723666647301808e-07,
      "loss": 0.0615,
      "num_input_tokens_seen": 34419008,
      "step": 38505
    },
    {
      "epoch": 10.163653160881616,
      "grad_norm": 0.0971350222826004,
      "learning_rate": 1.712177213547661e-07,
      "loss": 0.0435,
      "num_input_tokens_seen": 34423264,
      "step": 38510
    },
    {
      "epoch": 10.164972944437112,
      "grad_norm": 0.23266413807868958,
      "learning_rate": 1.7007260688959581e-07,
      "loss": 0.054,
      "num_input_tokens_seen": 34427872,
      "step": 38515
    },
    {
      "epoch": 10.166292727992609,
      "grad_norm": 0.12821246683597565,
      "learning_rate": 1.68931321511262e-07,
      "loss": 0.0337,
      "num_input_tokens_seen": 34432416,
      "step": 38520
    },
    {
      "epoch": 10.167612511548105,
      "grad_norm": 0.058961283415555954,
      "learning_rate": 1.6779386539576835e-07,
      "loss": 0.0582,
      "num_input_tokens_seen": 34436736,
      "step": 38525
    },
    {
      "epoch": 10.168932295103604,
      "grad_norm": 0.09676984697580338,
      "learning_rate": 1.666602387185162e-07,
      "loss": 0.043,
      "num_input_tokens_seen": 34441152,
      "step": 38530
    },
    {
      "epoch": 10.1702520786591,
      "grad_norm": 0.24321171641349792,
      "learning_rate": 1.655304416543352e-07,
      "loss": 0.0809,
      "num_input_tokens_seen": 34445472,
      "step": 38535
    },
    {
      "epoch": 10.171571862214597,
      "grad_norm": 0.3261115550994873,
      "learning_rate": 1.6440447437744698e-07,
      "loss": 0.0604,
      "num_input_tokens_seen": 34450016,
      "step": 38540
    },
    {
      "epoch": 10.172891645770093,
      "grad_norm": 0.3651960790157318,
      "learning_rate": 1.6328233706149332e-07,
      "loss": 0.0634,
      "num_input_tokens_seen": 34454208,
      "step": 38545
    },
    {
      "epoch": 10.17421142932559,
      "grad_norm": 0.48833152651786804,
      "learning_rate": 1.6216402987951906e-07,
      "loss": 0.0677,
      "num_input_tokens_seen": 34458784,
      "step": 38550
    },
    {
      "epoch": 10.175531212881088,
      "grad_norm": 0.2033245712518692,
      "learning_rate": 1.6104955300398627e-07,
      "loss": 0.0611,
      "num_input_tokens_seen": 34463360,
      "step": 38555
    },
    {
      "epoch": 10.176850996436585,
      "grad_norm": 0.22157514095306396,
      "learning_rate": 1.5993890660675748e-07,
      "loss": 0.0627,
      "num_input_tokens_seen": 34467872,
      "step": 38560
    },
    {
      "epoch": 10.178170779992081,
      "grad_norm": 0.4708595275878906,
      "learning_rate": 1.5883209085910678e-07,
      "loss": 0.0811,
      "num_input_tokens_seen": 34472672,
      "step": 38565
    },
    {
      "epoch": 10.179490563547578,
      "grad_norm": 0.20167233049869537,
      "learning_rate": 1.5772910593172264e-07,
      "loss": 0.0262,
      "num_input_tokens_seen": 34476896,
      "step": 38570
    },
    {
      "epoch": 10.180810347103074,
      "grad_norm": 0.10514634102582932,
      "learning_rate": 1.5662995199469954e-07,
      "loss": 0.0439,
      "num_input_tokens_seen": 34481472,
      "step": 38575
    },
    {
      "epoch": 10.182130130658573,
      "grad_norm": 0.3564639091491699,
      "learning_rate": 1.5553462921753802e-07,
      "loss": 0.0708,
      "num_input_tokens_seen": 34485920,
      "step": 38580
    },
    {
      "epoch": 10.18344991421407,
      "grad_norm": 0.1412670612335205,
      "learning_rate": 1.544431377691502e-07,
      "loss": 0.0848,
      "num_input_tokens_seen": 34490144,
      "step": 38585
    },
    {
      "epoch": 10.184769697769566,
      "grad_norm": 0.1433810144662857,
      "learning_rate": 1.5335547781785975e-07,
      "loss": 0.0309,
      "num_input_tokens_seen": 34494752,
      "step": 38590
    },
    {
      "epoch": 10.186089481325062,
      "grad_norm": 0.0487697534263134,
      "learning_rate": 1.5227164953139917e-07,
      "loss": 0.0231,
      "num_input_tokens_seen": 34498976,
      "step": 38595
    },
    {
      "epoch": 10.187409264880559,
      "grad_norm": 0.05024844408035278,
      "learning_rate": 1.511916530769042e-07,
      "loss": 0.0125,
      "num_input_tokens_seen": 34503808,
      "step": 38600
    },
    {
      "epoch": 10.187409264880559,
      "eval_loss": 0.06674957275390625,
      "eval_runtime": 64.7698,
      "eval_samples_per_second": 103.984,
      "eval_steps_per_second": 26.0,
      "num_input_tokens_seen": 34503808,
      "step": 38600
    },
    {
      "epoch": 10.188729048436057,
      "grad_norm": 0.5583072304725647,
      "learning_rate": 1.5011548862092773e-07,
      "loss": 0.0985,
      "num_input_tokens_seen": 34508224,
      "step": 38605
    },
    {
      "epoch": 10.190048831991554,
      "grad_norm": 0.17431506514549255,
      "learning_rate": 1.490431563294231e-07,
      "loss": 0.0278,
      "num_input_tokens_seen": 34512736,
      "step": 38610
    },
    {
      "epoch": 10.19136861554705,
      "grad_norm": 0.04053434729576111,
      "learning_rate": 1.4797465636776365e-07,
      "loss": 0.0419,
      "num_input_tokens_seen": 34517152,
      "step": 38615
    },
    {
      "epoch": 10.192688399102547,
      "grad_norm": 0.34291166067123413,
      "learning_rate": 1.4690998890072027e-07,
      "loss": 0.0899,
      "num_input_tokens_seen": 34521632,
      "step": 38620
    },
    {
      "epoch": 10.194008182658044,
      "grad_norm": 0.0928865298628807,
      "learning_rate": 1.4584915409248112e-07,
      "loss": 0.0161,
      "num_input_tokens_seen": 34525984,
      "step": 38625
    },
    {
      "epoch": 10.19532796621354,
      "grad_norm": 0.06816121935844421,
      "learning_rate": 1.4479215210663754e-07,
      "loss": 0.023,
      "num_input_tokens_seen": 34530624,
      "step": 38630
    },
    {
      "epoch": 10.196647749769038,
      "grad_norm": 0.5474709272384644,
      "learning_rate": 1.4373898310619528e-07,
      "loss": 0.0737,
      "num_input_tokens_seen": 34535232,
      "step": 38635
    },
    {
      "epoch": 10.197967533324535,
      "grad_norm": 0.2091810554265976,
      "learning_rate": 1.4268964725356604e-07,
      "loss": 0.1289,
      "num_input_tokens_seen": 34539712,
      "step": 38640
    },
    {
      "epoch": 10.199287316880032,
      "grad_norm": 0.024008432403206825,
      "learning_rate": 1.4164414471056764e-07,
      "loss": 0.0339,
      "num_input_tokens_seen": 34544384,
      "step": 38645
    },
    {
      "epoch": 10.200607100435528,
      "grad_norm": 0.18757879734039307,
      "learning_rate": 1.4060247563843497e-07,
      "loss": 0.0562,
      "num_input_tokens_seen": 34548928,
      "step": 38650
    },
    {
      "epoch": 10.201926883991025,
      "grad_norm": 0.2560552954673767,
      "learning_rate": 1.3956464019780068e-07,
      "loss": 0.0755,
      "num_input_tokens_seen": 34553344,
      "step": 38655
    },
    {
      "epoch": 10.203246667546523,
      "grad_norm": 0.21417447924613953,
      "learning_rate": 1.385306385487145e-07,
      "loss": 0.0756,
      "num_input_tokens_seen": 34557728,
      "step": 38660
    },
    {
      "epoch": 10.20456645110202,
      "grad_norm": 0.1422772854566574,
      "learning_rate": 1.3750047085063222e-07,
      "loss": 0.0658,
      "num_input_tokens_seen": 34562208,
      "step": 38665
    },
    {
      "epoch": 10.205886234657516,
      "grad_norm": 0.15437009930610657,
      "learning_rate": 1.3647413726242119e-07,
      "loss": 0.0434,
      "num_input_tokens_seen": 34566592,
      "step": 38670
    },
    {
      "epoch": 10.207206018213013,
      "grad_norm": 0.03055115044116974,
      "learning_rate": 1.3545163794235205e-07,
      "loss": 0.0489,
      "num_input_tokens_seen": 34571552,
      "step": 38675
    },
    {
      "epoch": 10.20852580176851,
      "grad_norm": 0.1291097104549408,
      "learning_rate": 1.3443297304810698e-07,
      "loss": 0.0555,
      "num_input_tokens_seen": 34575840,
      "step": 38680
    },
    {
      "epoch": 10.209845585324008,
      "grad_norm": 0.2938077747821808,
      "learning_rate": 1.3341814273677977e-07,
      "loss": 0.0557,
      "num_input_tokens_seen": 34580512,
      "step": 38685
    },
    {
      "epoch": 10.211165368879504,
      "grad_norm": 0.542168378829956,
      "learning_rate": 1.324071471648647e-07,
      "loss": 0.1073,
      "num_input_tokens_seen": 34585024,
      "step": 38690
    },
    {
      "epoch": 10.212485152435,
      "grad_norm": 0.561034619808197,
      "learning_rate": 1.3139998648827312e-07,
      "loss": 0.0827,
      "num_input_tokens_seen": 34589664,
      "step": 38695
    },
    {
      "epoch": 10.213804935990497,
      "grad_norm": 0.12570880353450775,
      "learning_rate": 1.3039666086232526e-07,
      "loss": 0.0225,
      "num_input_tokens_seen": 34594144,
      "step": 38700
    },
    {
      "epoch": 10.215124719545994,
      "grad_norm": 0.15776684880256653,
      "learning_rate": 1.2939717044174183e-07,
      "loss": 0.0319,
      "num_input_tokens_seen": 34598592,
      "step": 38705
    },
    {
      "epoch": 10.216444503101492,
      "grad_norm": 0.04353850707411766,
      "learning_rate": 1.284015153806578e-07,
      "loss": 0.0489,
      "num_input_tokens_seen": 34602752,
      "step": 38710
    },
    {
      "epoch": 10.217764286656989,
      "grad_norm": 0.5946187376976013,
      "learning_rate": 1.274096958326171e-07,
      "loss": 0.0749,
      "num_input_tokens_seen": 34607584,
      "step": 38715
    },
    {
      "epoch": 10.219084070212485,
      "grad_norm": 0.04369615763425827,
      "learning_rate": 1.2642171195056952e-07,
      "loss": 0.0368,
      "num_input_tokens_seen": 34612416,
      "step": 38720
    },
    {
      "epoch": 10.220403853767982,
      "grad_norm": 0.21111132204532623,
      "learning_rate": 1.2543756388687377e-07,
      "loss": 0.0795,
      "num_input_tokens_seen": 34616672,
      "step": 38725
    },
    {
      "epoch": 10.221723637323478,
      "grad_norm": 0.08242810517549515,
      "learning_rate": 1.2445725179330014e-07,
      "loss": 0.0335,
      "num_input_tokens_seen": 34620736,
      "step": 38730
    },
    {
      "epoch": 10.223043420878977,
      "grad_norm": 0.10774124413728714,
      "learning_rate": 1.2348077582102212e-07,
      "loss": 0.0522,
      "num_input_tokens_seen": 34625120,
      "step": 38735
    },
    {
      "epoch": 10.224363204434473,
      "grad_norm": 0.2842857241630554,
      "learning_rate": 1.2250813612062762e-07,
      "loss": 0.0874,
      "num_input_tokens_seen": 34629824,
      "step": 38740
    },
    {
      "epoch": 10.22568298798997,
      "grad_norm": 0.14024296402931213,
      "learning_rate": 1.215393328421105e-07,
      "loss": 0.051,
      "num_input_tokens_seen": 34634400,
      "step": 38745
    },
    {
      "epoch": 10.227002771545466,
      "grad_norm": 0.09930836409330368,
      "learning_rate": 1.2057436613486796e-07,
      "loss": 0.031,
      "num_input_tokens_seen": 34638656,
      "step": 38750
    },
    {
      "epoch": 10.228322555100963,
      "grad_norm": 0.4378281235694885,
      "learning_rate": 1.1961323614771424e-07,
      "loss": 0.0953,
      "num_input_tokens_seen": 34643328,
      "step": 38755
    },
    {
      "epoch": 10.229642338656461,
      "grad_norm": 0.16187579929828644,
      "learning_rate": 1.1865594302886418e-07,
      "loss": 0.0417,
      "num_input_tokens_seen": 34648000,
      "step": 38760
    },
    {
      "epoch": 10.230962122211958,
      "grad_norm": 0.3063742518424988,
      "learning_rate": 1.1770248692594687e-07,
      "loss": 0.0704,
      "num_input_tokens_seen": 34652224,
      "step": 38765
    },
    {
      "epoch": 10.232281905767454,
      "grad_norm": 0.3048715591430664,
      "learning_rate": 1.167528679859975e-07,
      "loss": 0.0729,
      "num_input_tokens_seen": 34656736,
      "step": 38770
    },
    {
      "epoch": 10.233601689322951,
      "grad_norm": 0.38002267479896545,
      "learning_rate": 1.1580708635545446e-07,
      "loss": 0.0546,
      "num_input_tokens_seen": 34661216,
      "step": 38775
    },
    {
      "epoch": 10.234921472878447,
      "grad_norm": 0.06072096899151802,
      "learning_rate": 1.1486514218017885e-07,
      "loss": 0.025,
      "num_input_tokens_seen": 34665888,
      "step": 38780
    },
    {
      "epoch": 10.236241256433944,
      "grad_norm": 0.42840471863746643,
      "learning_rate": 1.1392703560542117e-07,
      "loss": 0.0362,
      "num_input_tokens_seen": 34670400,
      "step": 38785
    },
    {
      "epoch": 10.237561039989442,
      "grad_norm": 0.06651969254016876,
      "learning_rate": 1.129927667758518e-07,
      "loss": 0.0381,
      "num_input_tokens_seen": 34674784,
      "step": 38790
    },
    {
      "epoch": 10.238880823544939,
      "grad_norm": 0.082082100212574,
      "learning_rate": 1.1206233583554992e-07,
      "loss": 0.0551,
      "num_input_tokens_seen": 34679328,
      "step": 38795
    },
    {
      "epoch": 10.240200607100435,
      "grad_norm": 0.03843453899025917,
      "learning_rate": 1.1113574292799523e-07,
      "loss": 0.0467,
      "num_input_tokens_seen": 34683552,
      "step": 38800
    },
    {
      "epoch": 10.240200607100435,
      "eval_loss": 0.06684853881597519,
      "eval_runtime": 64.7989,
      "eval_samples_per_second": 103.937,
      "eval_steps_per_second": 25.988,
      "num_input_tokens_seen": 34683552,
      "step": 38800
    },
    {
      "epoch": 10.241520390655932,
      "grad_norm": 0.10521683096885681,
      "learning_rate": 1.1021298819608449e-07,
      "loss": 0.0527,
      "num_input_tokens_seen": 34687840,
      "step": 38805
    },
    {
      "epoch": 10.242840174211429,
      "grad_norm": 0.12652160227298737,
      "learning_rate": 1.0929407178211226e-07,
      "loss": 0.077,
      "num_input_tokens_seen": 34692352,
      "step": 38810
    },
    {
      "epoch": 10.244159957766927,
      "grad_norm": 0.02760947123169899,
      "learning_rate": 1.0837899382779293e-07,
      "loss": 0.021,
      "num_input_tokens_seen": 34697152,
      "step": 38815
    },
    {
      "epoch": 10.245479741322423,
      "grad_norm": 0.7245965600013733,
      "learning_rate": 1.0746775447423862e-07,
      "loss": 0.0916,
      "num_input_tokens_seen": 34701472,
      "step": 38820
    },
    {
      "epoch": 10.24679952487792,
      "grad_norm": 0.08462116867303848,
      "learning_rate": 1.0656035386197583e-07,
      "loss": 0.0212,
      "num_input_tokens_seen": 34705888,
      "step": 38825
    },
    {
      "epoch": 10.248119308433417,
      "grad_norm": 0.12168659269809723,
      "learning_rate": 1.0565679213093982e-07,
      "loss": 0.0444,
      "num_input_tokens_seen": 34710208,
      "step": 38830
    },
    {
      "epoch": 10.249439091988913,
      "grad_norm": 0.05977329611778259,
      "learning_rate": 1.0475706942046638e-07,
      "loss": 0.0383,
      "num_input_tokens_seen": 34714336,
      "step": 38835
    },
    {
      "epoch": 10.250758875544411,
      "grad_norm": 0.40691497921943665,
      "learning_rate": 1.0386118586930282e-07,
      "loss": 0.0631,
      "num_input_tokens_seen": 34718816,
      "step": 38840
    },
    {
      "epoch": 10.252078659099908,
      "grad_norm": 0.4241376519203186,
      "learning_rate": 1.0296914161561367e-07,
      "loss": 0.0838,
      "num_input_tokens_seen": 34723360,
      "step": 38845
    },
    {
      "epoch": 10.253398442655405,
      "grad_norm": 0.20379526913166046,
      "learning_rate": 1.0208093679695552e-07,
      "loss": 0.0441,
      "num_input_tokens_seen": 34727616,
      "step": 38850
    },
    {
      "epoch": 10.254718226210901,
      "grad_norm": 0.06051342934370041,
      "learning_rate": 1.0119657155030493e-07,
      "loss": 0.0408,
      "num_input_tokens_seen": 34732032,
      "step": 38855
    },
    {
      "epoch": 10.256038009766398,
      "grad_norm": 0.13274140655994415,
      "learning_rate": 1.003160460120417e-07,
      "loss": 0.0287,
      "num_input_tokens_seen": 34736512,
      "step": 38860
    },
    {
      "epoch": 10.257357793321896,
      "grad_norm": 0.13771241903305054,
      "learning_rate": 9.943936031795165e-08,
      "loss": 0.0457,
      "num_input_tokens_seen": 34740960,
      "step": 38865
    },
    {
      "epoch": 10.258677576877393,
      "grad_norm": 0.3219413757324219,
      "learning_rate": 9.856651460323219e-08,
      "loss": 0.0745,
      "num_input_tokens_seen": 34745376,
      "step": 38870
    },
    {
      "epoch": 10.25999736043289,
      "grad_norm": 0.02981734462082386,
      "learning_rate": 9.769750900248953e-08,
      "loss": 0.0394,
      "num_input_tokens_seen": 34749600,
      "step": 38875
    },
    {
      "epoch": 10.261317143988386,
      "grad_norm": 0.18825046718120575,
      "learning_rate": 9.683234364973038e-08,
      "loss": 0.0265,
      "num_input_tokens_seen": 34754176,
      "step": 38880
    },
    {
      "epoch": 10.262636927543882,
      "grad_norm": 0.18589605391025543,
      "learning_rate": 9.597101867837854e-08,
      "loss": 0.0306,
      "num_input_tokens_seen": 34758528,
      "step": 38885
    },
    {
      "epoch": 10.263956711099379,
      "grad_norm": 0.48391279578208923,
      "learning_rate": 9.511353422125835e-08,
      "loss": 0.1284,
      "num_input_tokens_seen": 34762944,
      "step": 38890
    },
    {
      "epoch": 10.265276494654877,
      "grad_norm": 0.1390886753797531,
      "learning_rate": 9.42598904106029e-08,
      "loss": 0.0599,
      "num_input_tokens_seen": 34767232,
      "step": 38895
    },
    {
      "epoch": 10.266596278210374,
      "grad_norm": 0.282479852437973,
      "learning_rate": 9.341008737806245e-08,
      "loss": 0.0337,
      "num_input_tokens_seen": 34771584,
      "step": 38900
    },
    {
      "epoch": 10.26791606176587,
      "grad_norm": 0.07049477845430374,
      "learning_rate": 9.256412525467661e-08,
      "loss": 0.0635,
      "num_input_tokens_seen": 34776000,
      "step": 38905
    },
    {
      "epoch": 10.269235845321367,
      "grad_norm": 0.32042673230171204,
      "learning_rate": 9.172200417091326e-08,
      "loss": 0.0411,
      "num_input_tokens_seen": 34780384,
      "step": 38910
    },
    {
      "epoch": 10.270555628876863,
      "grad_norm": 0.051038529723882675,
      "learning_rate": 9.088372425663239e-08,
      "loss": 0.0569,
      "num_input_tokens_seen": 34784896,
      "step": 38915
    },
    {
      "epoch": 10.271875412432362,
      "grad_norm": 0.1753125786781311,
      "learning_rate": 9.004928564110837e-08,
      "loss": 0.0439,
      "num_input_tokens_seen": 34789696,
      "step": 38920
    },
    {
      "epoch": 10.273195195987858,
      "grad_norm": 0.5092338919639587,
      "learning_rate": 8.92186884530244e-08,
      "loss": 0.0841,
      "num_input_tokens_seen": 34794112,
      "step": 38925
    },
    {
      "epoch": 10.274514979543355,
      "grad_norm": 0.022948477417230606,
      "learning_rate": 8.83919328204641e-08,
      "loss": 0.0281,
      "num_input_tokens_seen": 34798432,
      "step": 38930
    },
    {
      "epoch": 10.275834763098851,
      "grad_norm": 0.22265799343585968,
      "learning_rate": 8.756901887093105e-08,
      "loss": 0.0816,
      "num_input_tokens_seen": 34802656,
      "step": 38935
    },
    {
      "epoch": 10.277154546654348,
      "grad_norm": 0.2962473928928375,
      "learning_rate": 8.674994673132098e-08,
      "loss": 0.0667,
      "num_input_tokens_seen": 34807264,
      "step": 38940
    },
    {
      "epoch": 10.278474330209846,
      "grad_norm": 0.09947988390922546,
      "learning_rate": 8.593471652794949e-08,
      "loss": 0.0867,
      "num_input_tokens_seen": 34811872,
      "step": 38945
    },
    {
      "epoch": 10.279794113765343,
      "grad_norm": 0.49483397603034973,
      "learning_rate": 8.512332838653548e-08,
      "loss": 0.0949,
      "num_input_tokens_seen": 34816608,
      "step": 38950
    },
    {
      "epoch": 10.28111389732084,
      "grad_norm": 0.19445523619651794,
      "learning_rate": 8.431578243220106e-08,
      "loss": 0.0491,
      "num_input_tokens_seen": 34821024,
      "step": 38955
    },
    {
      "epoch": 10.282433680876336,
      "grad_norm": 0.12636226415634155,
      "learning_rate": 8.351207878948552e-08,
      "loss": 0.0232,
      "num_input_tokens_seen": 34825280,
      "step": 38960
    },
    {
      "epoch": 10.283753464431832,
      "grad_norm": 0.3059556484222412,
      "learning_rate": 8.271221758232583e-08,
      "loss": 0.0819,
      "num_input_tokens_seen": 34830048,
      "step": 38965
    },
    {
      "epoch": 10.28507324798733,
      "grad_norm": 0.05188632383942604,
      "learning_rate": 8.191619893407332e-08,
      "loss": 0.0406,
      "num_input_tokens_seen": 34834464,
      "step": 38970
    },
    {
      "epoch": 10.286393031542827,
      "grad_norm": 0.032690707594156265,
      "learning_rate": 8.112402296748534e-08,
      "loss": 0.0224,
      "num_input_tokens_seen": 34839136,
      "step": 38975
    },
    {
      "epoch": 10.287712815098324,
      "grad_norm": 0.25292280316352844,
      "learning_rate": 8.033568980471973e-08,
      "loss": 0.1075,
      "num_input_tokens_seen": 34843104,
      "step": 38980
    },
    {
      "epoch": 10.28903259865382,
      "grad_norm": 0.3211628496646881,
      "learning_rate": 7.955119956735146e-08,
      "loss": 0.0731,
      "num_input_tokens_seen": 34847296,
      "step": 38985
    },
    {
      "epoch": 10.290352382209317,
      "grad_norm": 0.08910756558179855,
      "learning_rate": 7.877055237636155e-08,
      "loss": 0.0591,
      "num_input_tokens_seen": 34851744,
      "step": 38990
    },
    {
      "epoch": 10.291672165764815,
      "grad_norm": 0.07351075857877731,
      "learning_rate": 7.79937483521287e-08,
      "loss": 0.0296,
      "num_input_tokens_seen": 34856416,
      "step": 38995
    },
    {
      "epoch": 10.292991949320312,
      "grad_norm": 0.24101482331752777,
      "learning_rate": 7.722078761444873e-08,
      "loss": 0.0458,
      "num_input_tokens_seen": 34860896,
      "step": 39000
    },
    {
      "epoch": 10.292991949320312,
      "eval_loss": 0.0668007880449295,
      "eval_runtime": 64.7625,
      "eval_samples_per_second": 103.995,
      "eval_steps_per_second": 26.003,
      "num_input_tokens_seen": 34860896,
      "step": 39000
    },
    {
      "epoch": 10.294311732875808,
      "grad_norm": 0.2994737923145294,
      "learning_rate": 7.645167028252631e-08,
      "loss": 0.1298,
      "num_input_tokens_seen": 34865312,
      "step": 39005
    },
    {
      "epoch": 10.295631516431305,
      "grad_norm": 0.1882815957069397,
      "learning_rate": 7.568639647496379e-08,
      "loss": 0.0703,
      "num_input_tokens_seen": 34869856,
      "step": 39010
    },
    {
      "epoch": 10.296951299986802,
      "grad_norm": 0.07314976304769516,
      "learning_rate": 7.492496630977508e-08,
      "loss": 0.0919,
      "num_input_tokens_seen": 34874592,
      "step": 39015
    },
    {
      "epoch": 10.2982710835423,
      "grad_norm": 0.2699569761753082,
      "learning_rate": 7.416737990438571e-08,
      "loss": 0.0446,
      "num_input_tokens_seen": 34879072,
      "step": 39020
    },
    {
      "epoch": 10.299590867097796,
      "grad_norm": 0.12427210062742233,
      "learning_rate": 7.341363737562445e-08,
      "loss": 0.0769,
      "num_input_tokens_seen": 34883552,
      "step": 39025
    },
    {
      "epoch": 10.300910650653293,
      "grad_norm": 0.08230319619178772,
      "learning_rate": 7.266373883972887e-08,
      "loss": 0.121,
      "num_input_tokens_seen": 34888032,
      "step": 39030
    },
    {
      "epoch": 10.30223043420879,
      "grad_norm": 0.40962591767311096,
      "learning_rate": 7.191768441233981e-08,
      "loss": 0.0952,
      "num_input_tokens_seen": 34892576,
      "step": 39035
    },
    {
      "epoch": 10.303550217764286,
      "grad_norm": 0.1081714779138565,
      "learning_rate": 7.11754742085069e-08,
      "loss": 0.076,
      "num_input_tokens_seen": 34897056,
      "step": 39040
    },
    {
      "epoch": 10.304870001319783,
      "grad_norm": 0.14868226647377014,
      "learning_rate": 7.043710834269413e-08,
      "loss": 0.0335,
      "num_input_tokens_seen": 34901568,
      "step": 39045
    },
    {
      "epoch": 10.306189784875281,
      "grad_norm": 0.34856677055358887,
      "learning_rate": 6.970258692876319e-08,
      "loss": 0.0931,
      "num_input_tokens_seen": 34906016,
      "step": 39050
    },
    {
      "epoch": 10.307509568430778,
      "grad_norm": 0.1221441924571991,
      "learning_rate": 6.897191007998738e-08,
      "loss": 0.0569,
      "num_input_tokens_seen": 34910688,
      "step": 39055
    },
    {
      "epoch": 10.308829351986274,
      "grad_norm": 0.20173712074756622,
      "learning_rate": 6.824507790904599e-08,
      "loss": 0.0799,
      "num_input_tokens_seen": 34915424,
      "step": 39060
    },
    {
      "epoch": 10.31014913554177,
      "grad_norm": 0.22462844848632812,
      "learning_rate": 6.752209052802439e-08,
      "loss": 0.0891,
      "num_input_tokens_seen": 34919488,
      "step": 39065
    },
    {
      "epoch": 10.311468919097267,
      "grad_norm": 0.09147199243307114,
      "learning_rate": 6.680294804841946e-08,
      "loss": 0.0543,
      "num_input_tokens_seen": 34924288,
      "step": 39070
    },
    {
      "epoch": 10.312788702652766,
      "grad_norm": 0.06847822666168213,
      "learning_rate": 6.608765058112865e-08,
      "loss": 0.0138,
      "num_input_tokens_seen": 34928928,
      "step": 39075
    },
    {
      "epoch": 10.314108486208262,
      "grad_norm": 0.31936952471733093,
      "learning_rate": 6.537619823646368e-08,
      "loss": 0.0521,
      "num_input_tokens_seen": 34933216,
      "step": 39080
    },
    {
      "epoch": 10.315428269763759,
      "grad_norm": 0.06189334765076637,
      "learning_rate": 6.466859112413404e-08,
      "loss": 0.1031,
      "num_input_tokens_seen": 34937440,
      "step": 39085
    },
    {
      "epoch": 10.316748053319255,
      "grad_norm": 0.06286962330341339,
      "learning_rate": 6.39648293532663e-08,
      "loss": 0.0366,
      "num_input_tokens_seen": 34942208,
      "step": 39090
    },
    {
      "epoch": 10.318067836874752,
      "grad_norm": 0.13972745835781097,
      "learning_rate": 6.32649130323848e-08,
      "loss": 0.0666,
      "num_input_tokens_seen": 34946592,
      "step": 39095
    },
    {
      "epoch": 10.31938762043025,
      "grad_norm": 0.22997868061065674,
      "learning_rate": 6.256884226943094e-08,
      "loss": 0.0647,
      "num_input_tokens_seen": 34950912,
      "step": 39100
    },
    {
      "epoch": 10.320707403985747,
      "grad_norm": 0.13109838962554932,
      "learning_rate": 6.187661717174386e-08,
      "loss": 0.0696,
      "num_input_tokens_seen": 34955296,
      "step": 39105
    },
    {
      "epoch": 10.322027187541243,
      "grad_norm": 0.19327491521835327,
      "learning_rate": 6.118823784607708e-08,
      "loss": 0.0688,
      "num_input_tokens_seen": 34959712,
      "step": 39110
    },
    {
      "epoch": 10.32334697109674,
      "grad_norm": 0.140794038772583,
      "learning_rate": 6.050370439858178e-08,
      "loss": 0.047,
      "num_input_tokens_seen": 34964000,
      "step": 39115
    },
    {
      "epoch": 10.324666754652236,
      "grad_norm": 0.32161054015159607,
      "learning_rate": 5.98230169348235e-08,
      "loss": 0.0584,
      "num_input_tokens_seen": 34968352,
      "step": 39120
    },
    {
      "epoch": 10.325986538207735,
      "grad_norm": 0.14388413727283478,
      "learning_rate": 5.914617555977664e-08,
      "loss": 0.0315,
      "num_input_tokens_seen": 34972640,
      "step": 39125
    },
    {
      "epoch": 10.327306321763231,
      "grad_norm": 0.026323627680540085,
      "learning_rate": 5.8473180377816017e-08,
      "loss": 0.0579,
      "num_input_tokens_seen": 34977248,
      "step": 39130
    },
    {
      "epoch": 10.328626105318728,
      "grad_norm": 0.12550987303256989,
      "learning_rate": 5.780403149272251e-08,
      "loss": 0.0765,
      "num_input_tokens_seen": 34981760,
      "step": 39135
    },
    {
      "epoch": 10.329945888874224,
      "grad_norm": 0.20249582827091217,
      "learning_rate": 5.7138729007694126e-08,
      "loss": 0.0367,
      "num_input_tokens_seen": 34986048,
      "step": 39140
    },
    {
      "epoch": 10.331265672429721,
      "grad_norm": 0.411763995885849,
      "learning_rate": 5.64772730253238e-08,
      "loss": 0.0677,
      "num_input_tokens_seen": 34990304,
      "step": 39145
    },
    {
      "epoch": 10.33258545598522,
      "grad_norm": 0.2721242606639862,
      "learning_rate": 5.5819663647618814e-08,
      "loss": 0.047,
      "num_input_tokens_seen": 34994624,
      "step": 39150
    },
    {
      "epoch": 10.333905239540716,
      "grad_norm": 0.1430835723876953,
      "learning_rate": 5.5165900975989723e-08,
      "loss": 0.0421,
      "num_input_tokens_seen": 34999008,
      "step": 39155
    },
    {
      "epoch": 10.335225023096212,
      "grad_norm": 0.3239065408706665,
      "learning_rate": 5.451598511125311e-08,
      "loss": 0.0896,
      "num_input_tokens_seen": 35003552,
      "step": 39160
    },
    {
      "epoch": 10.336544806651709,
      "grad_norm": 0.42054077982902527,
      "learning_rate": 5.3869916153637124e-08,
      "loss": 0.0981,
      "num_input_tokens_seen": 35008064,
      "step": 39165
    },
    {
      "epoch": 10.337864590207205,
      "grad_norm": 0.43364962935447693,
      "learning_rate": 5.322769420277318e-08,
      "loss": 0.1049,
      "num_input_tokens_seen": 35012672,
      "step": 39170
    },
    {
      "epoch": 10.339184373762702,
      "grad_norm": 0.3862692713737488,
      "learning_rate": 5.258931935769873e-08,
      "loss": 0.1081,
      "num_input_tokens_seen": 35017088,
      "step": 39175
    },
    {
      "epoch": 10.3405041573182,
      "grad_norm": 0.42801719903945923,
      "learning_rate": 5.19547917168628e-08,
      "loss": 0.092,
      "num_input_tokens_seen": 35021568,
      "step": 39180
    },
    {
      "epoch": 10.341823940873697,
      "grad_norm": 0.05797917768359184,
      "learning_rate": 5.13241113781121e-08,
      "loss": 0.0503,
      "num_input_tokens_seen": 35025920,
      "step": 39185
    },
    {
      "epoch": 10.343143724429193,
      "grad_norm": 0.058744050562381744,
      "learning_rate": 5.0697278438707755e-08,
      "loss": 0.0513,
      "num_input_tokens_seen": 35030240,
      "step": 39190
    },
    {
      "epoch": 10.34446350798469,
      "grad_norm": 0.037497036159038544,
      "learning_rate": 5.0074292995316854e-08,
      "loss": 0.0316,
      "num_input_tokens_seen": 35034848,
      "step": 39195
    },
    {
      "epoch": 10.345783291540187,
      "grad_norm": 0.25086164474487305,
      "learning_rate": 4.945515514400978e-08,
      "loss": 0.0469,
      "num_input_tokens_seen": 35039424,
      "step": 39200
    },
    {
      "epoch": 10.345783291540187,
      "eval_loss": 0.06669439375400543,
      "eval_runtime": 64.7685,
      "eval_samples_per_second": 103.986,
      "eval_steps_per_second": 26.0,
      "num_input_tokens_seen": 35039424,
      "step": 39200
    },
    {
      "epoch": 10.347103075095685,
      "grad_norm": 0.356332391500473,
      "learning_rate": 4.883986498026571e-08,
      "loss": 0.1522,
      "num_input_tokens_seen": 35043936,
      "step": 39205
    },
    {
      "epoch": 10.348422858651181,
      "grad_norm": 0.6044077277183533,
      "learning_rate": 4.822842259896987e-08,
      "loss": 0.0977,
      "num_input_tokens_seen": 35048096,
      "step": 39210
    },
    {
      "epoch": 10.349742642206678,
      "grad_norm": 0.06820203363895416,
      "learning_rate": 4.762082809441626e-08,
      "loss": 0.0104,
      "num_input_tokens_seen": 35052416,
      "step": 39215
    },
    {
      "epoch": 10.351062425762175,
      "grad_norm": 0.42420727014541626,
      "learning_rate": 4.7017081560302156e-08,
      "loss": 0.0581,
      "num_input_tokens_seen": 35057056,
      "step": 39220
    },
    {
      "epoch": 10.352382209317671,
      "grad_norm": 0.19865663349628448,
      "learning_rate": 4.6417183089730866e-08,
      "loss": 0.0763,
      "num_input_tokens_seen": 35061312,
      "step": 39225
    },
    {
      "epoch": 10.35370199287317,
      "grad_norm": 0.06255370378494263,
      "learning_rate": 4.5821132775217265e-08,
      "loss": 0.0201,
      "num_input_tokens_seen": 35065888,
      "step": 39230
    },
    {
      "epoch": 10.355021776428666,
      "grad_norm": 0.10420253127813339,
      "learning_rate": 4.5228930708679504e-08,
      "loss": 0.0518,
      "num_input_tokens_seen": 35070624,
      "step": 39235
    },
    {
      "epoch": 10.356341559984163,
      "grad_norm": 0.18582554161548615,
      "learning_rate": 4.464057698144175e-08,
      "loss": 0.044,
      "num_input_tokens_seen": 35074784,
      "step": 39240
    },
    {
      "epoch": 10.35766134353966,
      "grad_norm": 0.3424522578716278,
      "learning_rate": 4.4056071684236974e-08,
      "loss": 0.0561,
      "num_input_tokens_seen": 35079296,
      "step": 39245
    },
    {
      "epoch": 10.358981127095156,
      "grad_norm": 0.2922587990760803,
      "learning_rate": 4.347541490719864e-08,
      "loss": 0.0807,
      "num_input_tokens_seen": 35084032,
      "step": 39250
    },
    {
      "epoch": 10.360300910650654,
      "grad_norm": 0.03479823097586632,
      "learning_rate": 4.2898606739877336e-08,
      "loss": 0.0426,
      "num_input_tokens_seen": 35088128,
      "step": 39255
    },
    {
      "epoch": 10.36162069420615,
      "grad_norm": 0.2777692377567291,
      "learning_rate": 4.232564727122135e-08,
      "loss": 0.0544,
      "num_input_tokens_seen": 35092928,
      "step": 39260
    },
    {
      "epoch": 10.362940477761647,
      "grad_norm": 0.35067906975746155,
      "learning_rate": 4.1756536589585004e-08,
      "loss": 0.063,
      "num_input_tokens_seen": 35097632,
      "step": 39265
    },
    {
      "epoch": 10.364260261317144,
      "grad_norm": 0.09923063963651657,
      "learning_rate": 4.119127478273976e-08,
      "loss": 0.0276,
      "num_input_tokens_seen": 35102080,
      "step": 39270
    },
    {
      "epoch": 10.36558004487264,
      "grad_norm": 0.031164998188614845,
      "learning_rate": 4.062986193784923e-08,
      "loss": 0.0227,
      "num_input_tokens_seen": 35106784,
      "step": 39275
    },
    {
      "epoch": 10.366899828428139,
      "grad_norm": 0.09088560938835144,
      "learning_rate": 4.007229814149416e-08,
      "loss": 0.0377,
      "num_input_tokens_seen": 35110976,
      "step": 39280
    },
    {
      "epoch": 10.368219611983635,
      "grad_norm": 0.5445138812065125,
      "learning_rate": 3.951858347965576e-08,
      "loss": 0.0727,
      "num_input_tokens_seen": 35115456,
      "step": 39285
    },
    {
      "epoch": 10.369539395539132,
      "grad_norm": 0.21864992380142212,
      "learning_rate": 3.896871803772684e-08,
      "loss": 0.066,
      "num_input_tokens_seen": 35119520,
      "step": 39290
    },
    {
      "epoch": 10.370859179094628,
      "grad_norm": 0.06065914407372475,
      "learning_rate": 3.842270190050068e-08,
      "loss": 0.0841,
      "num_input_tokens_seen": 35124224,
      "step": 39295
    },
    {
      "epoch": 10.372178962650125,
      "grad_norm": 0.01906857080757618,
      "learning_rate": 3.7880535152179376e-08,
      "loss": 0.0901,
      "num_input_tokens_seen": 35128224,
      "step": 39300
    },
    {
      "epoch": 10.373498746205623,
      "grad_norm": 0.3996250331401825,
      "learning_rate": 3.734221787637382e-08,
      "loss": 0.0873,
      "num_input_tokens_seen": 35132832,
      "step": 39305
    },
    {
      "epoch": 10.37481852976112,
      "grad_norm": 0.11153167486190796,
      "learning_rate": 3.680775015609817e-08,
      "loss": 0.0928,
      "num_input_tokens_seen": 35137120,
      "step": 39310
    },
    {
      "epoch": 10.376138313316616,
      "grad_norm": 0.2349633276462555,
      "learning_rate": 3.627713207377537e-08,
      "loss": 0.0469,
      "num_input_tokens_seen": 35141600,
      "step": 39315
    },
    {
      "epoch": 10.377458096872113,
      "grad_norm": 0.08981253206729889,
      "learning_rate": 3.575036371123164e-08,
      "loss": 0.0244,
      "num_input_tokens_seen": 35146272,
      "step": 39320
    },
    {
      "epoch": 10.37877788042761,
      "grad_norm": 0.1727363020181656,
      "learning_rate": 3.5227445149704776e-08,
      "loss": 0.0836,
      "num_input_tokens_seen": 35150912,
      "step": 39325
    },
    {
      "epoch": 10.380097663983106,
      "grad_norm": 0.25248101353645325,
      "learning_rate": 3.470837646983027e-08,
      "loss": 0.0477,
      "num_input_tokens_seen": 35155360,
      "step": 39330
    },
    {
      "epoch": 10.381417447538604,
      "grad_norm": 0.13595643639564514,
      "learning_rate": 3.419315775165799e-08,
      "loss": 0.0436,
      "num_input_tokens_seen": 35160032,
      "step": 39335
    },
    {
      "epoch": 10.3827372310941,
      "grad_norm": 0.24405348300933838,
      "learning_rate": 3.368178907464103e-08,
      "loss": 0.0472,
      "num_input_tokens_seen": 35164544,
      "step": 39340
    },
    {
      "epoch": 10.384057014649597,
      "grad_norm": 0.12438478320837021,
      "learning_rate": 3.317427051763855e-08,
      "loss": 0.0488,
      "num_input_tokens_seen": 35168832,
      "step": 39345
    },
    {
      "epoch": 10.385376798205094,
      "grad_norm": 0.3066204786300659,
      "learning_rate": 3.267060215891571e-08,
      "loss": 0.0826,
      "num_input_tokens_seen": 35173440,
      "step": 39350
    },
    {
      "epoch": 10.38669658176059,
      "grad_norm": 0.27836695313453674,
      "learning_rate": 3.217078407614649e-08,
      "loss": 0.1218,
      "num_input_tokens_seen": 35178080,
      "step": 39355
    },
    {
      "epoch": 10.388016365316089,
      "grad_norm": 0.21126212179660797,
      "learning_rate": 3.1674816346405345e-08,
      "loss": 0.0501,
      "num_input_tokens_seen": 35182496,
      "step": 39360
    },
    {
      "epoch": 10.389336148871585,
      "grad_norm": 0.2908795177936554,
      "learning_rate": 3.11826990461811e-08,
      "loss": 0.0346,
      "num_input_tokens_seen": 35186720,
      "step": 39365
    },
    {
      "epoch": 10.390655932427082,
      "grad_norm": 0.08373452723026276,
      "learning_rate": 3.069443225136304e-08,
      "loss": 0.0728,
      "num_input_tokens_seen": 35190976,
      "step": 39370
    },
    {
      "epoch": 10.391975715982579,
      "grad_norm": 0.20170536637306213,
      "learning_rate": 3.021001603724372e-08,
      "loss": 0.0595,
      "num_input_tokens_seen": 35195360,
      "step": 39375
    },
    {
      "epoch": 10.393295499538075,
      "grad_norm": 0.40421366691589355,
      "learning_rate": 2.9729450478532818e-08,
      "loss": 0.0944,
      "num_input_tokens_seen": 35199744,
      "step": 39380
    },
    {
      "epoch": 10.394615283093573,
      "grad_norm": 0.17769184708595276,
      "learning_rate": 2.9252735649337726e-08,
      "loss": 0.0431,
      "num_input_tokens_seen": 35204544,
      "step": 39385
    },
    {
      "epoch": 10.39593506664907,
      "grad_norm": 0.1792316734790802,
      "learning_rate": 2.8779871623171863e-08,
      "loss": 0.0263,
      "num_input_tokens_seen": 35209152,
      "step": 39390
    },
    {
      "epoch": 10.397254850204567,
      "grad_norm": 0.6637542843818665,
      "learning_rate": 2.8310858472957448e-08,
      "loss": 0.1141,
      "num_input_tokens_seen": 35213440,
      "step": 39395
    },
    {
      "epoch": 10.398574633760063,
      "grad_norm": 0.21643762290477753,
      "learning_rate": 2.784569627101996e-08,
      "loss": 0.0359,
      "num_input_tokens_seen": 35217792,
      "step": 39400
    },
    {
      "epoch": 10.398574633760063,
      "eval_loss": 0.06686842441558838,
      "eval_runtime": 64.7817,
      "eval_samples_per_second": 103.964,
      "eval_steps_per_second": 25.995,
      "num_input_tokens_seen": 35217792,
      "step": 39400
    },
    {
      "epoch": 10.39989441731556,
      "grad_norm": 0.060691509395837784,
      "learning_rate": 2.738438508909924e-08,
      "loss": 0.0526,
      "num_input_tokens_seen": 35222272,
      "step": 39405
    },
    {
      "epoch": 10.401214200871058,
      "grad_norm": 0.08169547468423843,
      "learning_rate": 2.692692499833005e-08,
      "loss": 0.03,
      "num_input_tokens_seen": 35226752,
      "step": 39410
    },
    {
      "epoch": 10.402533984426555,
      "grad_norm": 0.31232625246047974,
      "learning_rate": 2.647331606926151e-08,
      "loss": 0.0442,
      "num_input_tokens_seen": 35231168,
      "step": 39415
    },
    {
      "epoch": 10.403853767982051,
      "grad_norm": 0.017157819122076035,
      "learning_rate": 2.6023558371843225e-08,
      "loss": 0.0394,
      "num_input_tokens_seen": 35235712,
      "step": 39420
    },
    {
      "epoch": 10.405173551537548,
      "grad_norm": 0.06694570183753967,
      "learning_rate": 2.557765197543638e-08,
      "loss": 0.0437,
      "num_input_tokens_seen": 35240416,
      "step": 39425
    },
    {
      "epoch": 10.406493335093044,
      "grad_norm": 0.43080952763557434,
      "learning_rate": 2.513559694880263e-08,
      "loss": 0.0674,
      "num_input_tokens_seen": 35245184,
      "step": 39430
    },
    {
      "epoch": 10.40781311864854,
      "grad_norm": 0.09999874979257584,
      "learning_rate": 2.469739336011523e-08,
      "loss": 0.0813,
      "num_input_tokens_seen": 35249280,
      "step": 39435
    },
    {
      "epoch": 10.409132902204039,
      "grad_norm": 0.02377139776945114,
      "learning_rate": 2.4263041276947894e-08,
      "loss": 0.02,
      "num_input_tokens_seen": 35253728,
      "step": 39440
    },
    {
      "epoch": 10.410452685759536,
      "grad_norm": 0.4324994385242462,
      "learning_rate": 2.3832540766283164e-08,
      "loss": 0.0716,
      "num_input_tokens_seen": 35258368,
      "step": 39445
    },
    {
      "epoch": 10.411772469315032,
      "grad_norm": 0.17599305510520935,
      "learning_rate": 2.3405891894512366e-08,
      "loss": 0.0479,
      "num_input_tokens_seen": 35262560,
      "step": 39450
    },
    {
      "epoch": 10.413092252870529,
      "grad_norm": 0.5903554558753967,
      "learning_rate": 2.29830947274301e-08,
      "loss": 0.0473,
      "num_input_tokens_seen": 35266912,
      "step": 39455
    },
    {
      "epoch": 10.414412036426025,
      "grad_norm": 0.23585903644561768,
      "learning_rate": 2.2564149330231432e-08,
      "loss": 0.0427,
      "num_input_tokens_seen": 35271424,
      "step": 39460
    },
    {
      "epoch": 10.415731819981524,
      "grad_norm": 0.03456173837184906,
      "learning_rate": 2.2149055767528572e-08,
      "loss": 0.0544,
      "num_input_tokens_seen": 35275808,
      "step": 39465
    },
    {
      "epoch": 10.41705160353702,
      "grad_norm": 0.057969797402620316,
      "learning_rate": 2.1737814103334197e-08,
      "loss": 0.055,
      "num_input_tokens_seen": 35280448,
      "step": 39470
    },
    {
      "epoch": 10.418371387092517,
      "grad_norm": 0.17995183169841766,
      "learning_rate": 2.1330424401064253e-08,
      "loss": 0.04,
      "num_input_tokens_seen": 35284800,
      "step": 39475
    },
    {
      "epoch": 10.419691170648013,
      "grad_norm": 0.10369167476892471,
      "learning_rate": 2.092688672354348e-08,
      "loss": 0.0624,
      "num_input_tokens_seen": 35289248,
      "step": 39480
    },
    {
      "epoch": 10.42101095420351,
      "grad_norm": 0.14830124378204346,
      "learning_rate": 2.0527201133005435e-08,
      "loss": 0.0387,
      "num_input_tokens_seen": 35293760,
      "step": 39485
    },
    {
      "epoch": 10.422330737759008,
      "grad_norm": 0.021682003512978554,
      "learning_rate": 2.0131367691084148e-08,
      "loss": 0.0574,
      "num_input_tokens_seen": 35298176,
      "step": 39490
    },
    {
      "epoch": 10.423650521314505,
      "grad_norm": 0.19973953068256378,
      "learning_rate": 1.9739386458819675e-08,
      "loss": 0.0898,
      "num_input_tokens_seen": 35302656,
      "step": 39495
    },
    {
      "epoch": 10.424970304870001,
      "grad_norm": 0.19754983484745026,
      "learning_rate": 1.9351257496666442e-08,
      "loss": 0.1166,
      "num_input_tokens_seen": 35307168,
      "step": 39500
    },
    {
      "epoch": 10.426290088425498,
      "grad_norm": 0.5130003690719604,
      "learning_rate": 1.896698086447657e-08,
      "loss": 0.1357,
      "num_input_tokens_seen": 35311360,
      "step": 39505
    },
    {
      "epoch": 10.427609871980994,
      "grad_norm": 0.14373548328876495,
      "learning_rate": 1.8586556621505436e-08,
      "loss": 0.0347,
      "num_input_tokens_seen": 35316000,
      "step": 39510
    },
    {
      "epoch": 10.428929655536493,
      "grad_norm": 0.08314194530248642,
      "learning_rate": 1.820998482642833e-08,
      "loss": 0.0791,
      "num_input_tokens_seen": 35320096,
      "step": 39515
    },
    {
      "epoch": 10.43024943909199,
      "grad_norm": 0.0866573378443718,
      "learning_rate": 1.7837265537309912e-08,
      "loss": 0.0739,
      "num_input_tokens_seen": 35324416,
      "step": 39520
    },
    {
      "epoch": 10.431569222647486,
      "grad_norm": 0.554697573184967,
      "learning_rate": 1.7468398811629206e-08,
      "loss": 0.0651,
      "num_input_tokens_seen": 35329216,
      "step": 39525
    },
    {
      "epoch": 10.432889006202982,
      "grad_norm": 0.13776478171348572,
      "learning_rate": 1.710338470627404e-08,
      "loss": 0.0362,
      "num_input_tokens_seen": 35333856,
      "step": 39530
    },
    {
      "epoch": 10.434208789758479,
      "grad_norm": 0.09810629487037659,
      "learning_rate": 1.6742223277529945e-08,
      "loss": 0.0826,
      "num_input_tokens_seen": 35338272,
      "step": 39535
    },
    {
      "epoch": 10.435528573313977,
      "grad_norm": 0.053270068019628525,
      "learning_rate": 1.6384914581094036e-08,
      "loss": 0.0225,
      "num_input_tokens_seen": 35342848,
      "step": 39540
    },
    {
      "epoch": 10.436848356869474,
      "grad_norm": 0.5734820365905762,
      "learning_rate": 1.6031458672069455e-08,
      "loss": 0.0738,
      "num_input_tokens_seen": 35347392,
      "step": 39545
    },
    {
      "epoch": 10.43816814042497,
      "grad_norm": 0.04700048640370369,
      "learning_rate": 1.5681855604962602e-08,
      "loss": 0.0815,
      "num_input_tokens_seen": 35351744,
      "step": 39550
    },
    {
      "epoch": 10.439487923980467,
      "grad_norm": 0.192489892244339,
      "learning_rate": 1.5336105433683135e-08,
      "loss": 0.0196,
      "num_input_tokens_seen": 35355968,
      "step": 39555
    },
    {
      "epoch": 10.440807707535964,
      "grad_norm": 0.1723727285861969,
      "learning_rate": 1.499420821155506e-08,
      "loss": 0.0553,
      "num_input_tokens_seen": 35360448,
      "step": 39560
    },
    {
      "epoch": 10.442127491091462,
      "grad_norm": 0.0815168246626854,
      "learning_rate": 1.4656163991302874e-08,
      "loss": 0.0162,
      "num_input_tokens_seen": 35364640,
      "step": 39565
    },
    {
      "epoch": 10.443447274646958,
      "grad_norm": 0.17705528438091278,
      "learning_rate": 1.4321972825051544e-08,
      "loss": 0.0775,
      "num_input_tokens_seen": 35368864,
      "step": 39570
    },
    {
      "epoch": 10.444767058202455,
      "grad_norm": 0.19994628429412842,
      "learning_rate": 1.3991634764345951e-08,
      "loss": 0.0445,
      "num_input_tokens_seen": 35373408,
      "step": 39575
    },
    {
      "epoch": 10.446086841757952,
      "grad_norm": 0.09063997864723206,
      "learning_rate": 1.3665149860120352e-08,
      "loss": 0.0294,
      "num_input_tokens_seen": 35377952,
      "step": 39580
    },
    {
      "epoch": 10.447406625313448,
      "grad_norm": 0.11073332279920578,
      "learning_rate": 1.3342518162728912e-08,
      "loss": 0.0561,
      "num_input_tokens_seen": 35382560,
      "step": 39585
    },
    {
      "epoch": 10.448726408868945,
      "grad_norm": 0.057739559561014175,
      "learning_rate": 1.30237397219235e-08,
      "loss": 0.1048,
      "num_input_tokens_seen": 35387168,
      "step": 39590
    },
    {
      "epoch": 10.450046192424443,
      "grad_norm": 0.2522336542606354,
      "learning_rate": 1.2708814586862016e-08,
      "loss": 0.0405,
      "num_input_tokens_seen": 35391744,
      "step": 39595
    },
    {
      "epoch": 10.45136597597994,
      "grad_norm": 0.12819427251815796,
      "learning_rate": 1.2397742806111168e-08,
      "loss": 0.0202,
      "num_input_tokens_seen": 35396000,
      "step": 39600
    },
    {
      "epoch": 10.45136597597994,
      "eval_loss": 0.06686899065971375,
      "eval_runtime": 64.7708,
      "eval_samples_per_second": 103.982,
      "eval_steps_per_second": 25.999,
      "num_input_tokens_seen": 35396000,
      "step": 39600
    },
    {
      "epoch": 10.452685759535436,
      "grad_norm": 0.15743660926818848,
      "learning_rate": 1.209052442764369e-08,
      "loss": 0.0504,
      "num_input_tokens_seen": 35400480,
      "step": 39605
    },
    {
      "epoch": 10.454005543090933,
      "grad_norm": 0.18735115230083466,
      "learning_rate": 1.17871594988328e-08,
      "loss": 0.0294,
      "num_input_tokens_seen": 35404672,
      "step": 39610
    },
    {
      "epoch": 10.45532532664643,
      "grad_norm": 0.029971200972795486,
      "learning_rate": 1.1487648066466072e-08,
      "loss": 0.0442,
      "num_input_tokens_seen": 35409216,
      "step": 39615
    },
    {
      "epoch": 10.456645110201928,
      "grad_norm": 0.23768869042396545,
      "learning_rate": 1.1191990176728784e-08,
      "loss": 0.0648,
      "num_input_tokens_seen": 35414112,
      "step": 39620
    },
    {
      "epoch": 10.457964893757424,
      "grad_norm": 0.31516918540000916,
      "learning_rate": 1.0900185875215018e-08,
      "loss": 0.0471,
      "num_input_tokens_seen": 35418656,
      "step": 39625
    },
    {
      "epoch": 10.45928467731292,
      "grad_norm": 0.2805478870868683,
      "learning_rate": 1.0612235206924891e-08,
      "loss": 0.0766,
      "num_input_tokens_seen": 35423360,
      "step": 39630
    },
    {
      "epoch": 10.460604460868417,
      "grad_norm": 0.12993866205215454,
      "learning_rate": 1.0328138216264549e-08,
      "loss": 0.083,
      "num_input_tokens_seen": 35428032,
      "step": 39635
    },
    {
      "epoch": 10.461924244423914,
      "grad_norm": 0.18811824917793274,
      "learning_rate": 1.004789494704339e-08,
      "loss": 0.0578,
      "num_input_tokens_seen": 35432448,
      "step": 39640
    },
    {
      "epoch": 10.463244027979412,
      "grad_norm": 0.05997803062200546,
      "learning_rate": 9.771505442482397e-09,
      "loss": 0.0379,
      "num_input_tokens_seen": 35436736,
      "step": 39645
    },
    {
      "epoch": 10.464563811534909,
      "grad_norm": 0.1549091637134552,
      "learning_rate": 9.498969745200259e-09,
      "loss": 0.0481,
      "num_input_tokens_seen": 35441248,
      "step": 39650
    },
    {
      "epoch": 10.465883595090405,
      "grad_norm": 0.11711782962083817,
      "learning_rate": 9.230287897230017e-09,
      "loss": 0.0695,
      "num_input_tokens_seen": 35445888,
      "step": 39655
    },
    {
      "epoch": 10.467203378645902,
      "grad_norm": 0.14941294491291046,
      "learning_rate": 8.965459940002419e-09,
      "loss": 0.0865,
      "num_input_tokens_seen": 35450336,
      "step": 39660
    },
    {
      "epoch": 10.468523162201398,
      "grad_norm": 0.12813811004161835,
      "learning_rate": 8.704485914357019e-09,
      "loss": 0.0786,
      "num_input_tokens_seen": 35454592,
      "step": 39665
    },
    {
      "epoch": 10.469842945756897,
      "grad_norm": 0.19397613406181335,
      "learning_rate": 8.447365860539402e-09,
      "loss": 0.0603,
      "num_input_tokens_seen": 35459200,
      "step": 39670
    },
    {
      "epoch": 10.471162729312393,
      "grad_norm": 0.24467119574546814,
      "learning_rate": 8.194099818201184e-09,
      "loss": 0.0614,
      "num_input_tokens_seen": 35463776,
      "step": 39675
    },
    {
      "epoch": 10.47248251286789,
      "grad_norm": 0.21178002655506134,
      "learning_rate": 7.944687826400011e-09,
      "loss": 0.0624,
      "num_input_tokens_seen": 35468224,
      "step": 39680
    },
    {
      "epoch": 10.473802296423386,
      "grad_norm": 0.19043120741844177,
      "learning_rate": 7.699129923599557e-09,
      "loss": 0.0345,
      "num_input_tokens_seen": 35472832,
      "step": 39685
    },
    {
      "epoch": 10.475122079978883,
      "grad_norm": 0.08776219189167023,
      "learning_rate": 7.457426147663982e-09,
      "loss": 0.0324,
      "num_input_tokens_seen": 35477312,
      "step": 39690
    },
    {
      "epoch": 10.47644186353438,
      "grad_norm": 0.09339522570371628,
      "learning_rate": 7.219576535871797e-09,
      "loss": 0.0454,
      "num_input_tokens_seen": 35481696,
      "step": 39695
    },
    {
      "epoch": 10.477761647089878,
      "grad_norm": 0.09532131254673004,
      "learning_rate": 6.985581124896445e-09,
      "loss": 0.038,
      "num_input_tokens_seen": 35486112,
      "step": 39700
    },
    {
      "epoch": 10.479081430645374,
      "grad_norm": 0.11481470614671707,
      "learning_rate": 6.755439950828501e-09,
      "loss": 0.0566,
      "num_input_tokens_seen": 35490368,
      "step": 39705
    },
    {
      "epoch": 10.480401214200871,
      "grad_norm": 0.16774976253509521,
      "learning_rate": 6.5291530491562444e-09,
      "loss": 0.0506,
      "num_input_tokens_seen": 35494656,
      "step": 39710
    },
    {
      "epoch": 10.481720997756367,
      "grad_norm": 0.14974111318588257,
      "learning_rate": 6.3067204547739845e-09,
      "loss": 0.056,
      "num_input_tokens_seen": 35499360,
      "step": 39715
    },
    {
      "epoch": 10.483040781311864,
      "grad_norm": 0.4242684543132782,
      "learning_rate": 6.088142201987612e-09,
      "loss": 0.0302,
      "num_input_tokens_seen": 35503520,
      "step": 39720
    },
    {
      "epoch": 10.484360564867362,
      "grad_norm": 0.2316826730966568,
      "learning_rate": 5.873418324503499e-09,
      "loss": 0.0638,
      "num_input_tokens_seen": 35508032,
      "step": 39725
    },
    {
      "epoch": 10.485680348422859,
      "grad_norm": 0.3321036994457245,
      "learning_rate": 5.6625488554340465e-09,
      "loss": 0.0383,
      "num_input_tokens_seen": 35512608,
      "step": 39730
    },
    {
      "epoch": 10.487000131978355,
      "grad_norm": 0.23089765012264252,
      "learning_rate": 5.455533827297688e-09,
      "loss": 0.0468,
      "num_input_tokens_seen": 35517312,
      "step": 39735
    },
    {
      "epoch": 10.488319915533852,
      "grad_norm": 0.07289786636829376,
      "learning_rate": 5.252373272018885e-09,
      "loss": 0.0503,
      "num_input_tokens_seen": 35522112,
      "step": 39740
    },
    {
      "epoch": 10.489639699089349,
      "grad_norm": 0.17986682057380676,
      "learning_rate": 5.053067220925356e-09,
      "loss": 0.0714,
      "num_input_tokens_seen": 35526560,
      "step": 39745
    },
    {
      "epoch": 10.490959482644847,
      "grad_norm": 0.4868500232696533,
      "learning_rate": 4.857615704759177e-09,
      "loss": 0.0916,
      "num_input_tokens_seen": 35530752,
      "step": 39750
    },
    {
      "epoch": 10.492279266200343,
      "grad_norm": 0.464214563369751,
      "learning_rate": 4.666018753654577e-09,
      "loss": 0.1001,
      "num_input_tokens_seen": 35535168,
      "step": 39755
    },
    {
      "epoch": 10.49359904975584,
      "grad_norm": 0.20315559208393097,
      "learning_rate": 4.478276397162917e-09,
      "loss": 0.07,
      "num_input_tokens_seen": 35539712,
      "step": 39760
    },
    {
      "epoch": 10.494918833311337,
      "grad_norm": 0.15449482202529907,
      "learning_rate": 4.294388664233262e-09,
      "loss": 0.0869,
      "num_input_tokens_seen": 35544096,
      "step": 39765
    },
    {
      "epoch": 10.496238616866833,
      "grad_norm": 0.31987449526786804,
      "learning_rate": 4.114355583223484e-09,
      "loss": 0.0401,
      "num_input_tokens_seen": 35548672,
      "step": 39770
    },
    {
      "epoch": 10.497558400422331,
      "grad_norm": 0.064491406083107,
      "learning_rate": 3.9381771818974845e-09,
      "loss": 0.0222,
      "num_input_tokens_seen": 35553216,
      "step": 39775
    },
    {
      "epoch": 10.498878183977828,
      "grad_norm": 0.3632766604423523,
      "learning_rate": 3.765853487427973e-09,
      "loss": 0.0982,
      "num_input_tokens_seen": 35557696,
      "step": 39780
    },
    {
      "epoch": 10.500197967533325,
      "grad_norm": 0.23432210087776184,
      "learning_rate": 3.5973845263825857e-09,
      "loss": 0.047,
      "num_input_tokens_seen": 35562208,
      "step": 39785
    },
    {
      "epoch": 10.501517751088821,
      "grad_norm": 0.10265903174877167,
      "learning_rate": 3.4327703247488684e-09,
      "loss": 0.0635,
      "num_input_tokens_seen": 35566720,
      "step": 39790
    },
    {
      "epoch": 10.502837534644318,
      "grad_norm": 0.08026956766843796,
      "learning_rate": 3.2720109079037443e-09,
      "loss": 0.0679,
      "num_input_tokens_seen": 35570912,
      "step": 39795
    },
    {
      "epoch": 10.504157318199816,
      "grad_norm": 0.24375790357589722,
      "learning_rate": 3.1151063006468193e-09,
      "loss": 0.0659,
      "num_input_tokens_seen": 35575872,
      "step": 39800
    },
    {
      "epoch": 10.504157318199816,
      "eval_loss": 0.06686899065971375,
      "eval_runtime": 64.7668,
      "eval_samples_per_second": 103.988,
      "eval_steps_per_second": 26.001,
      "num_input_tokens_seen": 35575872,
      "step": 39800
    },
    {
      "epoch": 10.505477101755313,
      "grad_norm": 0.24508266150951385,
      "learning_rate": 2.962056527169854e-09,
      "loss": 0.0278,
      "num_input_tokens_seen": 35580256,
      "step": 39805
    },
    {
      "epoch": 10.50679688531081,
      "grad_norm": 0.12731514871120453,
      "learning_rate": 2.8128616110761898e-09,
      "loss": 0.0373,
      "num_input_tokens_seen": 35584512,
      "step": 39810
    },
    {
      "epoch": 10.508116668866306,
      "grad_norm": 0.4986494481563568,
      "learning_rate": 2.6675215753724223e-09,
      "loss": 0.1155,
      "num_input_tokens_seen": 35589088,
      "step": 39815
    },
    {
      "epoch": 10.509436452421802,
      "grad_norm": 0.09775055199861526,
      "learning_rate": 2.5260364424739557e-09,
      "loss": 0.0526,
      "num_input_tokens_seen": 35593472,
      "step": 39820
    },
    {
      "epoch": 10.5107562359773,
      "grad_norm": 0.397988885641098,
      "learning_rate": 2.3884062341994475e-09,
      "loss": 0.0641,
      "num_input_tokens_seen": 35598144,
      "step": 39825
    },
    {
      "epoch": 10.512076019532797,
      "grad_norm": 0.39882686734199524,
      "learning_rate": 2.25463097177081e-09,
      "loss": 0.0782,
      "num_input_tokens_seen": 35602784,
      "step": 39830
    },
    {
      "epoch": 10.513395803088294,
      "grad_norm": 0.07363387942314148,
      "learning_rate": 2.1247106758215397e-09,
      "loss": 0.0439,
      "num_input_tokens_seen": 35607296,
      "step": 39835
    },
    {
      "epoch": 10.51471558664379,
      "grad_norm": 0.07311610132455826,
      "learning_rate": 1.998645366382834e-09,
      "loss": 0.0286,
      "num_input_tokens_seen": 35611520,
      "step": 39840
    },
    {
      "epoch": 10.516035370199287,
      "grad_norm": 0.09407701343297958,
      "learning_rate": 1.876435062897475e-09,
      "loss": 0.0276,
      "num_input_tokens_seen": 35616000,
      "step": 39845
    },
    {
      "epoch": 10.517355153754785,
      "grad_norm": 0.04539132118225098,
      "learning_rate": 1.758079784211497e-09,
      "loss": 0.0653,
      "num_input_tokens_seen": 35620320,
      "step": 39850
    },
    {
      "epoch": 10.518674937310282,
      "grad_norm": 0.6947534680366516,
      "learning_rate": 1.6435795485797434e-09,
      "loss": 0.0629,
      "num_input_tokens_seen": 35624736,
      "step": 39855
    },
    {
      "epoch": 10.519994720865778,
      "grad_norm": 0.15846362709999084,
      "learning_rate": 1.5329343736547596e-09,
      "loss": 0.0618,
      "num_input_tokens_seen": 35629088,
      "step": 39860
    },
    {
      "epoch": 10.521314504421275,
      "grad_norm": 0.2503301501274109,
      "learning_rate": 1.4261442765006739e-09,
      "loss": 0.0372,
      "num_input_tokens_seen": 35633472,
      "step": 39865
    },
    {
      "epoch": 10.522634287976771,
      "grad_norm": 0.18638908863067627,
      "learning_rate": 1.3232092735876445e-09,
      "loss": 0.0949,
      "num_input_tokens_seen": 35638048,
      "step": 39870
    },
    {
      "epoch": 10.523954071532268,
      "grad_norm": 0.35070744156837463,
      "learning_rate": 1.2241293807918607e-09,
      "loss": 0.0731,
      "num_input_tokens_seen": 35642592,
      "step": 39875
    },
    {
      "epoch": 10.525273855087766,
      "grad_norm": 0.08850501477718353,
      "learning_rate": 1.128904613387216e-09,
      "loss": 0.022,
      "num_input_tokens_seen": 35647104,
      "step": 39880
    },
    {
      "epoch": 10.526593638643263,
      "grad_norm": 0.2380620390176773,
      "learning_rate": 1.0375349860591853e-09,
      "loss": 0.085,
      "num_input_tokens_seen": 35651392,
      "step": 39885
    },
    {
      "epoch": 10.52791342219876,
      "grad_norm": 0.26449471712112427,
      "learning_rate": 9.5002051290205e-10,
      "loss": 0.0659,
      "num_input_tokens_seen": 35656000,
      "step": 39890
    },
    {
      "epoch": 10.529233205754256,
      "grad_norm": 0.19770123064517975,
      "learning_rate": 8.663612074077954e-10,
      "loss": 0.0394,
      "num_input_tokens_seen": 35660480,
      "step": 39895
    },
    {
      "epoch": 10.530552989309752,
      "grad_norm": 0.04451938718557358,
      "learning_rate": 7.865570824799884e-10,
      "loss": 0.0685,
      "num_input_tokens_seen": 35664704,
      "step": 39900
    },
    {
      "epoch": 10.53187277286525,
      "grad_norm": 0.30129677057266235,
      "learning_rate": 7.106081504254514e-10,
      "loss": 0.091,
      "num_input_tokens_seen": 35669056,
      "step": 39905
    },
    {
      "epoch": 10.533192556420747,
      "grad_norm": 0.08751536905765533,
      "learning_rate": 6.385144229570372e-10,
      "loss": 0.0417,
      "num_input_tokens_seen": 35673632,
      "step": 39910
    },
    {
      "epoch": 10.534512339976244,
      "grad_norm": 0.286141961812973,
      "learning_rate": 5.70275911190854e-10,
      "loss": 0.0407,
      "num_input_tokens_seen": 35678112,
      "step": 39915
    },
    {
      "epoch": 10.53583212353174,
      "grad_norm": 0.2553674876689911,
      "learning_rate": 5.058926256490403e-10,
      "loss": 0.0595,
      "num_input_tokens_seen": 35682592,
      "step": 39920
    },
    {
      "epoch": 10.537151907087237,
      "grad_norm": 0.08961944282054901,
      "learning_rate": 4.4536457626254134e-10,
      "loss": 0.0481,
      "num_input_tokens_seen": 35687328,
      "step": 39925
    },
    {
      "epoch": 10.538471690642735,
      "grad_norm": 0.048031508922576904,
      "learning_rate": 3.88691772365557e-10,
      "loss": 0.0262,
      "num_input_tokens_seen": 35691904,
      "step": 39930
    },
    {
      "epoch": 10.539791474198232,
      "grad_norm": 0.10910078883171082,
      "learning_rate": 3.358742226955425e-10,
      "loss": 0.0495,
      "num_input_tokens_seen": 35696288,
      "step": 39935
    },
    {
      "epoch": 10.541111257753728,
      "grad_norm": 0.17059040069580078,
      "learning_rate": 2.8691193539875925e-10,
      "loss": 0.1079,
      "num_input_tokens_seen": 35700640,
      "step": 39940
    },
    {
      "epoch": 10.542431041309225,
      "grad_norm": 0.08324994891881943,
      "learning_rate": 2.418049180274995e-10,
      "loss": 0.0444,
      "num_input_tokens_seen": 35704960,
      "step": 39945
    },
    {
      "epoch": 10.543750824864722,
      "grad_norm": 0.48782598972320557,
      "learning_rate": 2.005531775373104e-10,
      "loss": 0.0892,
      "num_input_tokens_seen": 35709760,
      "step": 39950
    },
    {
      "epoch": 10.545070608420218,
      "grad_norm": 0.32518601417541504,
      "learning_rate": 1.6315672028699435e-10,
      "loss": 0.0411,
      "num_input_tokens_seen": 35714048,
      "step": 39955
    },
    {
      "epoch": 10.546390391975716,
      "grad_norm": 0.06300792098045349,
      "learning_rate": 1.2961555204693555e-10,
      "loss": 0.0377,
      "num_input_tokens_seen": 35719104,
      "step": 39960
    },
    {
      "epoch": 10.547710175531213,
      "grad_norm": 0.2515268325805664,
      "learning_rate": 9.992967798799768e-11,
      "loss": 0.0774,
      "num_input_tokens_seen": 35723744,
      "step": 39965
    },
    {
      "epoch": 10.54902995908671,
      "grad_norm": 0.0424208790063858,
      "learning_rate": 7.409910268707521e-11,
      "loss": 0.0456,
      "num_input_tokens_seen": 35728448,
      "step": 39970
    },
    {
      "epoch": 10.550349742642206,
      "grad_norm": 0.06364504247903824,
      "learning_rate": 5.212383012986877e-11,
      "loss": 0.0483,
      "num_input_tokens_seen": 35732736,
      "step": 39975
    },
    {
      "epoch": 10.551669526197703,
      "grad_norm": 0.07357128709554672,
      "learning_rate": 3.400386370533415e-11,
      "loss": 0.0287,
      "num_input_tokens_seen": 35737184,
      "step": 39980
    },
    {
      "epoch": 10.552989309753201,
      "grad_norm": 0.11814670264720917,
      "learning_rate": 1.9739206205682258e-11,
      "loss": 0.016,
      "num_input_tokens_seen": 35741472,
      "step": 39985
    },
    {
      "epoch": 10.554309093308698,
      "grad_norm": 0.23330411314964294,
      "learning_rate": 9.329859829154685e-12,
      "loss": 0.0831,
      "num_input_tokens_seen": 35746048,
      "step": 39990
    },
    {
      "epoch": 10.555628876864194,
      "grad_norm": 0.11445839703083038,
      "learning_rate": 2.7758261855748148e-12,
      "loss": 0.0448,
      "num_input_tokens_seen": 35750592,
      "step": 39995
    },
    {
      "epoch": 10.55694866041969,
      "grad_norm": 0.04129907861351967,
      "learning_rate": 7.710628524559838e-14,
      "loss": 0.0661,
      "num_input_tokens_seen": 35754976,
      "step": 40000
    },
    {
      "epoch": 10.55694866041969,
      "eval_loss": 0.06686899065971375,
      "eval_runtime": 64.7753,
      "eval_samples_per_second": 103.975,
      "eval_steps_per_second": 25.998,
      "num_input_tokens_seen": 35754976,
      "step": 40000
    },
    {
      "epoch": 10.55694866041969,
      "num_input_tokens_seen": 35754976,
      "step": 40000,
      "total_flos": 1.6100725874609357e+18,
      "train_loss": 0.06669788992283866,
      "train_runtime": 35858.3327,
      "train_samples_per_second": 17.848,
      "train_steps_per_second": 1.116
    }
  ],
  "logging_steps": 5,
  "max_steps": 40000,
  "num_input_tokens_seen": 35754976,
  "num_train_epochs": 11,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6100725874609357e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}